{ "best_metric": null, "best_model_checkpoint": null, "epoch": 4.996455696202531, "eval_steps": 500, "global_step": 1230, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.004050632911392405, "grad_norm": 6.624785885078098, "learning_rate": 3.2520325203252037e-07, "loss": 1.0869, "step": 1 }, { "epoch": 0.00810126582278481, "grad_norm": 6.200452677459058, "learning_rate": 6.504065040650407e-07, "loss": 1.0282, "step": 2 }, { "epoch": 0.012151898734177215, "grad_norm": 7.405411786468841, "learning_rate": 9.75609756097561e-07, "loss": 1.1469, "step": 3 }, { "epoch": 0.01620253164556962, "grad_norm": 6.328549097366735, "learning_rate": 1.3008130081300815e-06, "loss": 1.0559, "step": 4 }, { "epoch": 0.020253164556962026, "grad_norm": 6.231281278939459, "learning_rate": 1.6260162601626018e-06, "loss": 1.0945, "step": 5 }, { "epoch": 0.02430379746835443, "grad_norm": 6.095238093207984, "learning_rate": 1.951219512195122e-06, "loss": 1.053, "step": 6 }, { "epoch": 0.028354430379746835, "grad_norm": 5.175768961619925, "learning_rate": 2.2764227642276426e-06, "loss": 1.0058, "step": 7 }, { "epoch": 0.03240506329113924, "grad_norm": 4.616809675474139, "learning_rate": 2.601626016260163e-06, "loss": 0.99, "step": 8 }, { "epoch": 0.03645569620253165, "grad_norm": 2.675312329569691, "learning_rate": 2.926829268292683e-06, "loss": 0.9928, "step": 9 }, { "epoch": 0.04050632911392405, "grad_norm": 2.4409804211191255, "learning_rate": 3.2520325203252037e-06, "loss": 0.9427, "step": 10 }, { "epoch": 0.044556962025316456, "grad_norm": 2.110406507653588, "learning_rate": 3.577235772357724e-06, "loss": 0.9139, "step": 11 }, { "epoch": 0.04860759493670886, "grad_norm": 4.226471475722353, "learning_rate": 3.902439024390244e-06, "loss": 0.9528, "step": 12 }, { "epoch": 0.052658227848101265, "grad_norm": 4.516331952533838, "learning_rate": 4.227642276422765e-06, "loss": 0.9402, "step": 13 }, { "epoch": 0.05670886075949367, "grad_norm": 4.519934045159026, "learning_rate": 4.552845528455285e-06, "loss": 0.9597, "step": 14 }, { "epoch": 0.060759493670886074, "grad_norm": 3.9489355564608046, "learning_rate": 4.8780487804878055e-06, "loss": 0.8695, "step": 15 }, { "epoch": 0.06481012658227848, "grad_norm": 3.374626058913794, "learning_rate": 5.203252032520326e-06, "loss": 0.8705, "step": 16 }, { "epoch": 0.06886075949367089, "grad_norm": 3.284268432039881, "learning_rate": 5.528455284552846e-06, "loss": 0.8929, "step": 17 }, { "epoch": 0.0729113924050633, "grad_norm": 2.716435739753283, "learning_rate": 5.853658536585366e-06, "loss": 0.8456, "step": 18 }, { "epoch": 0.0769620253164557, "grad_norm": 2.4557399305182503, "learning_rate": 6.178861788617887e-06, "loss": 0.8616, "step": 19 }, { "epoch": 0.0810126582278481, "grad_norm": 2.0201083791022896, "learning_rate": 6.504065040650407e-06, "loss": 0.824, "step": 20 }, { "epoch": 0.08506329113924051, "grad_norm": 1.824878989378493, "learning_rate": 6.829268292682928e-06, "loss": 0.8234, "step": 21 }, { "epoch": 0.08911392405063291, "grad_norm": 2.123832940014624, "learning_rate": 7.154471544715448e-06, "loss": 0.8446, "step": 22 }, { "epoch": 0.09316455696202532, "grad_norm": 1.8986036825425516, "learning_rate": 7.4796747967479676e-06, "loss": 0.793, "step": 23 }, { "epoch": 0.09721518987341772, "grad_norm": 2.0037810776821043, "learning_rate": 7.804878048780489e-06, "loss": 0.8136, "step": 24 }, { "epoch": 0.10126582278481013, "grad_norm": 1.6279589787837645, "learning_rate": 8.130081300813009e-06, "loss": 0.8058, "step": 25 }, { "epoch": 0.10531645569620253, "grad_norm": 1.4900945540818282, "learning_rate": 8.45528455284553e-06, "loss": 0.8014, "step": 26 }, { "epoch": 0.10936708860759493, "grad_norm": 1.5635653737418427, "learning_rate": 8.78048780487805e-06, "loss": 0.8118, "step": 27 }, { "epoch": 0.11341772151898734, "grad_norm": 1.7383451343217446, "learning_rate": 9.10569105691057e-06, "loss": 0.7759, "step": 28 }, { "epoch": 0.11746835443037974, "grad_norm": 1.403221362687963, "learning_rate": 9.43089430894309e-06, "loss": 0.7941, "step": 29 }, { "epoch": 0.12151898734177215, "grad_norm": 1.1031288393644516, "learning_rate": 9.756097560975611e-06, "loss": 0.7646, "step": 30 }, { "epoch": 0.12556962025316457, "grad_norm": 1.0112033648071796, "learning_rate": 1.008130081300813e-05, "loss": 0.7441, "step": 31 }, { "epoch": 0.12962025316455697, "grad_norm": 1.0885747633753935, "learning_rate": 1.0406504065040652e-05, "loss": 0.758, "step": 32 }, { "epoch": 0.13367088607594937, "grad_norm": 1.1349612637255222, "learning_rate": 1.0731707317073172e-05, "loss": 0.7628, "step": 33 }, { "epoch": 0.13772151898734178, "grad_norm": 0.8622911905599927, "learning_rate": 1.1056910569105692e-05, "loss": 0.7703, "step": 34 }, { "epoch": 0.14177215189873418, "grad_norm": 0.8789183352497403, "learning_rate": 1.1382113821138213e-05, "loss": 0.7488, "step": 35 }, { "epoch": 0.1458227848101266, "grad_norm": 0.880450925584051, "learning_rate": 1.1707317073170731e-05, "loss": 0.7613, "step": 36 }, { "epoch": 0.149873417721519, "grad_norm": 0.8399237999265943, "learning_rate": 1.2032520325203254e-05, "loss": 0.7232, "step": 37 }, { "epoch": 0.1539240506329114, "grad_norm": 0.7740386019857722, "learning_rate": 1.2357723577235774e-05, "loss": 0.7418, "step": 38 }, { "epoch": 0.1579746835443038, "grad_norm": 0.7786625641197614, "learning_rate": 1.2682926829268294e-05, "loss": 0.7246, "step": 39 }, { "epoch": 0.1620253164556962, "grad_norm": 0.7959006479460693, "learning_rate": 1.3008130081300815e-05, "loss": 0.7127, "step": 40 }, { "epoch": 0.1660759493670886, "grad_norm": 0.7152450156278007, "learning_rate": 1.3333333333333333e-05, "loss": 0.7408, "step": 41 }, { "epoch": 0.17012658227848101, "grad_norm": 0.671506162628021, "learning_rate": 1.3658536585365855e-05, "loss": 0.7157, "step": 42 }, { "epoch": 0.17417721518987342, "grad_norm": 0.6491791849786066, "learning_rate": 1.3983739837398376e-05, "loss": 0.7269, "step": 43 }, { "epoch": 0.17822784810126582, "grad_norm": 0.6697041546646614, "learning_rate": 1.4308943089430896e-05, "loss": 0.7079, "step": 44 }, { "epoch": 0.18227848101265823, "grad_norm": 0.7448720067717243, "learning_rate": 1.4634146341463415e-05, "loss": 0.7357, "step": 45 }, { "epoch": 0.18632911392405063, "grad_norm": 0.647414907047492, "learning_rate": 1.4959349593495935e-05, "loss": 0.7137, "step": 46 }, { "epoch": 0.19037974683544304, "grad_norm": 0.7186345067175797, "learning_rate": 1.528455284552846e-05, "loss": 0.7126, "step": 47 }, { "epoch": 0.19443037974683544, "grad_norm": 0.729956700815, "learning_rate": 1.5609756097560978e-05, "loss": 0.7222, "step": 48 }, { "epoch": 0.19848101265822785, "grad_norm": 0.6839914113101827, "learning_rate": 1.5934959349593496e-05, "loss": 0.7355, "step": 49 }, { "epoch": 0.20253164556962025, "grad_norm": 0.6733294559938702, "learning_rate": 1.6260162601626018e-05, "loss": 0.7186, "step": 50 }, { "epoch": 0.20658227848101265, "grad_norm": 0.6411736614792108, "learning_rate": 1.6585365853658537e-05, "loss": 0.714, "step": 51 }, { "epoch": 0.21063291139240506, "grad_norm": 0.7015278901933064, "learning_rate": 1.691056910569106e-05, "loss": 0.6998, "step": 52 }, { "epoch": 0.21468354430379746, "grad_norm": 0.7203520569361187, "learning_rate": 1.7235772357723578e-05, "loss": 0.712, "step": 53 }, { "epoch": 0.21873417721518987, "grad_norm": 0.633915555869983, "learning_rate": 1.75609756097561e-05, "loss": 0.7054, "step": 54 }, { "epoch": 0.22278481012658227, "grad_norm": 0.7440771374898305, "learning_rate": 1.788617886178862e-05, "loss": 0.7503, "step": 55 }, { "epoch": 0.22683544303797468, "grad_norm": 0.7218494654014811, "learning_rate": 1.821138211382114e-05, "loss": 0.704, "step": 56 }, { "epoch": 0.23088607594936708, "grad_norm": 0.6812418529670542, "learning_rate": 1.8536585365853663e-05, "loss": 0.724, "step": 57 }, { "epoch": 0.23493670886075949, "grad_norm": 0.6918847439334124, "learning_rate": 1.886178861788618e-05, "loss": 0.6937, "step": 58 }, { "epoch": 0.2389873417721519, "grad_norm": 0.6944158703133283, "learning_rate": 1.91869918699187e-05, "loss": 0.7121, "step": 59 }, { "epoch": 0.2430379746835443, "grad_norm": 0.6062682553077197, "learning_rate": 1.9512195121951222e-05, "loss": 0.7088, "step": 60 }, { "epoch": 0.2470886075949367, "grad_norm": 0.7974875318358634, "learning_rate": 1.983739837398374e-05, "loss": 0.6911, "step": 61 }, { "epoch": 0.25113924050632913, "grad_norm": 0.831396563034242, "learning_rate": 2.016260162601626e-05, "loss": 0.6881, "step": 62 }, { "epoch": 0.25518987341772154, "grad_norm": 0.7359032150740006, "learning_rate": 2.048780487804878e-05, "loss": 0.7042, "step": 63 }, { "epoch": 0.25924050632911394, "grad_norm": 0.8217399249777904, "learning_rate": 2.0813008130081303e-05, "loss": 0.7065, "step": 64 }, { "epoch": 0.26329113924050634, "grad_norm": 0.7298081961103227, "learning_rate": 2.1138211382113822e-05, "loss": 0.7042, "step": 65 }, { "epoch": 0.26734177215189875, "grad_norm": 0.6629499500603125, "learning_rate": 2.1463414634146344e-05, "loss": 0.6966, "step": 66 }, { "epoch": 0.27139240506329115, "grad_norm": 0.7234152409452838, "learning_rate": 2.1788617886178863e-05, "loss": 0.6849, "step": 67 }, { "epoch": 0.27544303797468356, "grad_norm": 0.5925226747864875, "learning_rate": 2.2113821138211385e-05, "loss": 0.6924, "step": 68 }, { "epoch": 0.27949367088607596, "grad_norm": 0.8011325863302172, "learning_rate": 2.2439024390243907e-05, "loss": 0.6846, "step": 69 }, { "epoch": 0.28354430379746837, "grad_norm": 0.8418776539773919, "learning_rate": 2.2764227642276426e-05, "loss": 0.6931, "step": 70 }, { "epoch": 0.28759493670886077, "grad_norm": 0.6996102135201194, "learning_rate": 2.3089430894308948e-05, "loss": 0.6971, "step": 71 }, { "epoch": 0.2916455696202532, "grad_norm": 0.7375707837026545, "learning_rate": 2.3414634146341463e-05, "loss": 0.7029, "step": 72 }, { "epoch": 0.2956962025316456, "grad_norm": 0.665954924717082, "learning_rate": 2.3739837398373985e-05, "loss": 0.6843, "step": 73 }, { "epoch": 0.299746835443038, "grad_norm": 0.8406257055299563, "learning_rate": 2.4065040650406507e-05, "loss": 0.6727, "step": 74 }, { "epoch": 0.3037974683544304, "grad_norm": 0.852588155865969, "learning_rate": 2.4390243902439026e-05, "loss": 0.6905, "step": 75 }, { "epoch": 0.3078481012658228, "grad_norm": 0.630855051206748, "learning_rate": 2.4715447154471548e-05, "loss": 0.686, "step": 76 }, { "epoch": 0.3118987341772152, "grad_norm": 0.7607449127156211, "learning_rate": 2.5040650406504066e-05, "loss": 0.7105, "step": 77 }, { "epoch": 0.3159493670886076, "grad_norm": 0.7031118012805995, "learning_rate": 2.536585365853659e-05, "loss": 0.6599, "step": 78 }, { "epoch": 0.32, "grad_norm": 0.799233483869746, "learning_rate": 2.569105691056911e-05, "loss": 0.6907, "step": 79 }, { "epoch": 0.3240506329113924, "grad_norm": 0.7155749883653362, "learning_rate": 2.601626016260163e-05, "loss": 0.7055, "step": 80 }, { "epoch": 0.3281012658227848, "grad_norm": 0.7633448774814235, "learning_rate": 2.634146341463415e-05, "loss": 0.6948, "step": 81 }, { "epoch": 0.3321518987341772, "grad_norm": 0.7907660987059516, "learning_rate": 2.6666666666666667e-05, "loss": 0.6724, "step": 82 }, { "epoch": 0.3362025316455696, "grad_norm": 0.6480627847681337, "learning_rate": 2.699186991869919e-05, "loss": 0.6799, "step": 83 }, { "epoch": 0.34025316455696203, "grad_norm": 0.7015005230267416, "learning_rate": 2.731707317073171e-05, "loss": 0.6904, "step": 84 }, { "epoch": 0.34430379746835443, "grad_norm": 0.6913337979812311, "learning_rate": 2.764227642276423e-05, "loss": 0.6862, "step": 85 }, { "epoch": 0.34835443037974684, "grad_norm": 0.622059078181869, "learning_rate": 2.796747967479675e-05, "loss": 0.6766, "step": 86 }, { "epoch": 0.35240506329113924, "grad_norm": 0.8091424177849006, "learning_rate": 2.829268292682927e-05, "loss": 0.6867, "step": 87 }, { "epoch": 0.35645569620253165, "grad_norm": 0.8551133793705099, "learning_rate": 2.8617886178861792e-05, "loss": 0.6758, "step": 88 }, { "epoch": 0.36050632911392405, "grad_norm": 1.0628875020904676, "learning_rate": 2.8943089430894314e-05, "loss": 0.6568, "step": 89 }, { "epoch": 0.36455696202531646, "grad_norm": 0.7607206030352549, "learning_rate": 2.926829268292683e-05, "loss": 0.6744, "step": 90 }, { "epoch": 0.36860759493670886, "grad_norm": 0.8502157969637446, "learning_rate": 2.959349593495935e-05, "loss": 0.6516, "step": 91 }, { "epoch": 0.37265822784810126, "grad_norm": 0.8691673980682065, "learning_rate": 2.991869918699187e-05, "loss": 0.6724, "step": 92 }, { "epoch": 0.37670886075949367, "grad_norm": 0.7664510359203496, "learning_rate": 3.0243902439024392e-05, "loss": 0.6883, "step": 93 }, { "epoch": 0.3807594936708861, "grad_norm": 0.9731106625784032, "learning_rate": 3.056910569105692e-05, "loss": 0.7074, "step": 94 }, { "epoch": 0.3848101265822785, "grad_norm": 0.7321488493617986, "learning_rate": 3.089430894308943e-05, "loss": 0.6796, "step": 95 }, { "epoch": 0.3888607594936709, "grad_norm": 1.003465245773365, "learning_rate": 3.1219512195121955e-05, "loss": 0.6849, "step": 96 }, { "epoch": 0.3929113924050633, "grad_norm": 0.8705847295598682, "learning_rate": 3.154471544715447e-05, "loss": 0.6745, "step": 97 }, { "epoch": 0.3969620253164557, "grad_norm": 0.7859005246603915, "learning_rate": 3.186991869918699e-05, "loss": 0.6836, "step": 98 }, { "epoch": 0.4010126582278481, "grad_norm": 0.8454684069603275, "learning_rate": 3.2195121951219514e-05, "loss": 0.7027, "step": 99 }, { "epoch": 0.4050632911392405, "grad_norm": 0.9305549521725919, "learning_rate": 3.2520325203252037e-05, "loss": 0.6842, "step": 100 }, { "epoch": 0.4091139240506329, "grad_norm": 0.7729576505621848, "learning_rate": 3.284552845528456e-05, "loss": 0.6711, "step": 101 }, { "epoch": 0.4131645569620253, "grad_norm": 0.9659525581149131, "learning_rate": 3.3170731707317074e-05, "loss": 0.6856, "step": 102 }, { "epoch": 0.4172151898734177, "grad_norm": 0.7850679905734488, "learning_rate": 3.3495934959349596e-05, "loss": 0.6709, "step": 103 }, { "epoch": 0.4212658227848101, "grad_norm": 0.9742160045176127, "learning_rate": 3.382113821138212e-05, "loss": 0.7099, "step": 104 }, { "epoch": 0.4253164556962025, "grad_norm": 0.8554150455378559, "learning_rate": 3.414634146341463e-05, "loss": 0.6425, "step": 105 }, { "epoch": 0.4293670886075949, "grad_norm": 1.0505667681945667, "learning_rate": 3.4471544715447155e-05, "loss": 0.6955, "step": 106 }, { "epoch": 0.43341772151898733, "grad_norm": 0.6989944364608391, "learning_rate": 3.479674796747968e-05, "loss": 0.694, "step": 107 }, { "epoch": 0.43746835443037974, "grad_norm": 1.0109537624395293, "learning_rate": 3.51219512195122e-05, "loss": 0.6655, "step": 108 }, { "epoch": 0.44151898734177214, "grad_norm": 0.7744851994727527, "learning_rate": 3.544715447154472e-05, "loss": 0.6722, "step": 109 }, { "epoch": 0.44556962025316454, "grad_norm": 0.8610898608570061, "learning_rate": 3.577235772357724e-05, "loss": 0.658, "step": 110 }, { "epoch": 0.44962025316455695, "grad_norm": 0.7259094239057211, "learning_rate": 3.609756097560976e-05, "loss": 0.6538, "step": 111 }, { "epoch": 0.45367088607594935, "grad_norm": 0.8234062726420098, "learning_rate": 3.642276422764228e-05, "loss": 0.6563, "step": 112 }, { "epoch": 0.45772151898734176, "grad_norm": 0.822784090845906, "learning_rate": 3.67479674796748e-05, "loss": 0.6947, "step": 113 }, { "epoch": 0.46177215189873416, "grad_norm": 0.752119517808012, "learning_rate": 3.7073170731707325e-05, "loss": 0.7006, "step": 114 }, { "epoch": 0.46582278481012657, "grad_norm": 0.8689152458242883, "learning_rate": 3.739837398373984e-05, "loss": 0.6626, "step": 115 }, { "epoch": 0.46987341772151897, "grad_norm": 0.7638050122473946, "learning_rate": 3.772357723577236e-05, "loss": 0.6765, "step": 116 }, { "epoch": 0.4739240506329114, "grad_norm": 0.9353449560290353, "learning_rate": 3.804878048780488e-05, "loss": 0.694, "step": 117 }, { "epoch": 0.4779746835443038, "grad_norm": 0.6623778558218107, "learning_rate": 3.83739837398374e-05, "loss": 0.678, "step": 118 }, { "epoch": 0.4820253164556962, "grad_norm": 0.9448599028830834, "learning_rate": 3.869918699186992e-05, "loss": 0.6771, "step": 119 }, { "epoch": 0.4860759493670886, "grad_norm": 1.0415989638053555, "learning_rate": 3.9024390243902444e-05, "loss": 0.6749, "step": 120 }, { "epoch": 0.490126582278481, "grad_norm": 0.8266997987342358, "learning_rate": 3.9349593495934966e-05, "loss": 0.6584, "step": 121 }, { "epoch": 0.4941772151898734, "grad_norm": 0.762671322273082, "learning_rate": 3.967479674796748e-05, "loss": 0.6718, "step": 122 }, { "epoch": 0.4982278481012658, "grad_norm": 0.6400965150860211, "learning_rate": 4e-05, "loss": 0.6366, "step": 123 }, { "epoch": 0.5022784810126583, "grad_norm": 0.6182844416049017, "learning_rate": 3.999991946137476e-05, "loss": 0.6567, "step": 124 }, { "epoch": 0.5063291139240507, "grad_norm": 0.8232333372187765, "learning_rate": 3.999967784614766e-05, "loss": 0.6754, "step": 125 }, { "epoch": 0.5103797468354431, "grad_norm": 0.6666408525413208, "learning_rate": 3.9999275156264656e-05, "loss": 0.6727, "step": 126 }, { "epoch": 0.5144303797468355, "grad_norm": 0.740786754901356, "learning_rate": 3.999871139496895e-05, "loss": 0.6711, "step": 127 }, { "epoch": 0.5184810126582279, "grad_norm": 1.0687720053771328, "learning_rate": 3.9997986566800995e-05, "loss": 0.7002, "step": 128 }, { "epoch": 0.5225316455696203, "grad_norm": 0.9197037189220846, "learning_rate": 3.999710067759846e-05, "loss": 0.6903, "step": 129 }, { "epoch": 0.5265822784810127, "grad_norm": 0.8007301016162489, "learning_rate": 3.999605373449617e-05, "loss": 0.6755, "step": 130 }, { "epoch": 0.5306329113924051, "grad_norm": 0.7317146224040869, "learning_rate": 3.9994845745926075e-05, "loss": 0.6524, "step": 131 }, { "epoch": 0.5346835443037975, "grad_norm": 0.7339871789719086, "learning_rate": 3.999347672161713e-05, "loss": 0.6692, "step": 132 }, { "epoch": 0.5387341772151899, "grad_norm": 0.726640771207049, "learning_rate": 3.999194667259528e-05, "loss": 0.6743, "step": 133 }, { "epoch": 0.5427848101265823, "grad_norm": 0.809203860626538, "learning_rate": 3.999025561118334e-05, "loss": 0.6667, "step": 134 }, { "epoch": 0.5468354430379747, "grad_norm": 0.7053249389272829, "learning_rate": 3.998840355100086e-05, "loss": 0.6662, "step": 135 }, { "epoch": 0.5508860759493671, "grad_norm": 1.0357333014884562, "learning_rate": 3.998639050696409e-05, "loss": 0.6816, "step": 136 }, { "epoch": 0.5549367088607595, "grad_norm": 0.9344963028615187, "learning_rate": 3.998421649528582e-05, "loss": 0.6738, "step": 137 }, { "epoch": 0.5589873417721519, "grad_norm": 0.6586341481670294, "learning_rate": 3.9981881533475234e-05, "loss": 0.6688, "step": 138 }, { "epoch": 0.5630379746835443, "grad_norm": 0.7681994173895635, "learning_rate": 3.997938564033779e-05, "loss": 0.6736, "step": 139 }, { "epoch": 0.5670886075949367, "grad_norm": 0.9180288651172342, "learning_rate": 3.9976728835975064e-05, "loss": 0.6596, "step": 140 }, { "epoch": 0.5711392405063291, "grad_norm": 0.7415584723395408, "learning_rate": 3.9973911141784605e-05, "loss": 0.6754, "step": 141 }, { "epoch": 0.5751898734177215, "grad_norm": 0.6147208060871178, "learning_rate": 3.997093258045973e-05, "loss": 0.6694, "step": 142 }, { "epoch": 0.579240506329114, "grad_norm": 0.8172138140270159, "learning_rate": 3.996779317598936e-05, "loss": 0.6607, "step": 143 }, { "epoch": 0.5832911392405064, "grad_norm": 0.7974067463406005, "learning_rate": 3.996449295365782e-05, "loss": 0.6331, "step": 144 }, { "epoch": 0.5873417721518988, "grad_norm": 0.7459564938974032, "learning_rate": 3.996103194004467e-05, "loss": 0.6896, "step": 145 }, { "epoch": 0.5913924050632912, "grad_norm": 0.8699450857437874, "learning_rate": 3.995741016302441e-05, "loss": 0.6521, "step": 146 }, { "epoch": 0.5954430379746836, "grad_norm": 0.6213680988375035, "learning_rate": 3.9953627651766364e-05, "loss": 0.6627, "step": 147 }, { "epoch": 0.599493670886076, "grad_norm": 0.6512838029710066, "learning_rate": 3.9949684436734325e-05, "loss": 0.6634, "step": 148 }, { "epoch": 0.6035443037974684, "grad_norm": 0.6194244066819361, "learning_rate": 3.994558054968643e-05, "loss": 0.6657, "step": 149 }, { "epoch": 0.6075949367088608, "grad_norm": 0.6740091410866387, "learning_rate": 3.994131602367481e-05, "loss": 0.6723, "step": 150 }, { "epoch": 0.6116455696202532, "grad_norm": 0.7733842642886334, "learning_rate": 3.9936890893045376e-05, "loss": 0.6696, "step": 151 }, { "epoch": 0.6156962025316456, "grad_norm": 0.6715445954457137, "learning_rate": 3.993230519343752e-05, "loss": 0.6244, "step": 152 }, { "epoch": 0.619746835443038, "grad_norm": 0.737229129318353, "learning_rate": 3.992755896178383e-05, "loss": 0.7008, "step": 153 }, { "epoch": 0.6237974683544304, "grad_norm": 0.9471534262324407, "learning_rate": 3.992265223630981e-05, "loss": 0.6596, "step": 154 }, { "epoch": 0.6278481012658228, "grad_norm": 0.6876701441374604, "learning_rate": 3.991758505653355e-05, "loss": 0.6606, "step": 155 }, { "epoch": 0.6318987341772152, "grad_norm": 0.6470795079853546, "learning_rate": 3.991235746326543e-05, "loss": 0.661, "step": 156 }, { "epoch": 0.6359493670886076, "grad_norm": 0.6106865370715696, "learning_rate": 3.9906969498607745e-05, "loss": 0.6531, "step": 157 }, { "epoch": 0.64, "grad_norm": 0.6833625723707898, "learning_rate": 3.990142120595444e-05, "loss": 0.6563, "step": 158 }, { "epoch": 0.6440506329113924, "grad_norm": 0.7199132195325162, "learning_rate": 3.98957126299907e-05, "loss": 0.6752, "step": 159 }, { "epoch": 0.6481012658227848, "grad_norm": 0.5533803154741237, "learning_rate": 3.9889843816692596e-05, "loss": 0.6881, "step": 160 }, { "epoch": 0.6521518987341772, "grad_norm": 0.5595195072702748, "learning_rate": 3.9883814813326766e-05, "loss": 0.65, "step": 161 }, { "epoch": 0.6562025316455696, "grad_norm": 0.5961373172964936, "learning_rate": 3.9877625668449956e-05, "loss": 0.6675, "step": 162 }, { "epoch": 0.660253164556962, "grad_norm": 0.5342452514670545, "learning_rate": 3.98712764319087e-05, "loss": 0.6547, "step": 163 }, { "epoch": 0.6643037974683544, "grad_norm": 0.8161616535949966, "learning_rate": 3.9864767154838864e-05, "loss": 0.6525, "step": 164 }, { "epoch": 0.6683544303797468, "grad_norm": 0.6051658113613877, "learning_rate": 3.9858097889665277e-05, "loss": 0.6643, "step": 165 }, { "epoch": 0.6724050632911392, "grad_norm": 0.5723099583443331, "learning_rate": 3.985126869010129e-05, "loss": 0.6667, "step": 166 }, { "epoch": 0.6764556962025317, "grad_norm": 0.6002949576413271, "learning_rate": 3.984427961114833e-05, "loss": 0.6433, "step": 167 }, { "epoch": 0.6805063291139241, "grad_norm": 0.5857606594617282, "learning_rate": 3.9837130709095475e-05, "loss": 0.6608, "step": 168 }, { "epoch": 0.6845569620253165, "grad_norm": 0.5551757418354418, "learning_rate": 3.982982204151901e-05, "loss": 0.6771, "step": 169 }, { "epoch": 0.6886075949367089, "grad_norm": 0.5785128990018641, "learning_rate": 3.982235366728193e-05, "loss": 0.6215, "step": 170 }, { "epoch": 0.6926582278481013, "grad_norm": 0.6621902416530179, "learning_rate": 3.9814725646533505e-05, "loss": 0.659, "step": 171 }, { "epoch": 0.6967088607594937, "grad_norm": 0.5446411800961785, "learning_rate": 3.9806938040708746e-05, "loss": 0.6511, "step": 172 }, { "epoch": 0.7007594936708861, "grad_norm": 0.7994153229660889, "learning_rate": 3.9798990912527976e-05, "loss": 0.6605, "step": 173 }, { "epoch": 0.7048101265822785, "grad_norm": 0.6030689381949175, "learning_rate": 3.979088432599627e-05, "loss": 0.6445, "step": 174 }, { "epoch": 0.7088607594936709, "grad_norm": 0.6770720098832523, "learning_rate": 3.9782618346402964e-05, "loss": 0.672, "step": 175 }, { "epoch": 0.7129113924050633, "grad_norm": 0.6444968125233007, "learning_rate": 3.977419304032111e-05, "loss": 0.6859, "step": 176 }, { "epoch": 0.7169620253164557, "grad_norm": 0.6766127329175744, "learning_rate": 3.976560847560697e-05, "loss": 0.698, "step": 177 }, { "epoch": 0.7210126582278481, "grad_norm": 0.6061304102190777, "learning_rate": 3.9756864721399456e-05, "loss": 0.6396, "step": 178 }, { "epoch": 0.7250632911392405, "grad_norm": 0.7629116439736471, "learning_rate": 3.974796184811956e-05, "loss": 0.6371, "step": 179 }, { "epoch": 0.7291139240506329, "grad_norm": 0.6341834278376568, "learning_rate": 3.973889992746979e-05, "loss": 0.659, "step": 180 }, { "epoch": 0.7331645569620253, "grad_norm": 0.7111849767219653, "learning_rate": 3.972967903243361e-05, "loss": 0.67, "step": 181 }, { "epoch": 0.7372151898734177, "grad_norm": 0.6177591627692965, "learning_rate": 3.972029923727486e-05, "loss": 0.6188, "step": 182 }, { "epoch": 0.7412658227848101, "grad_norm": 0.6276690393362324, "learning_rate": 3.971076061753709e-05, "loss": 0.6415, "step": 183 }, { "epoch": 0.7453164556962025, "grad_norm": 0.6979336768707899, "learning_rate": 3.9701063250043066e-05, "loss": 0.6676, "step": 184 }, { "epoch": 0.7493670886075949, "grad_norm": 0.6904234620569265, "learning_rate": 3.969120721289402e-05, "loss": 0.6618, "step": 185 }, { "epoch": 0.7534177215189873, "grad_norm": 0.556659247738762, "learning_rate": 3.9681192585469146e-05, "loss": 0.6486, "step": 186 }, { "epoch": 0.7574683544303797, "grad_norm": 0.622384070406256, "learning_rate": 3.9671019448424865e-05, "loss": 0.6275, "step": 187 }, { "epoch": 0.7615189873417721, "grad_norm": 0.6612943016823112, "learning_rate": 3.966068788369422e-05, "loss": 0.6709, "step": 188 }, { "epoch": 0.7655696202531646, "grad_norm": 0.4955618017275831, "learning_rate": 3.965019797448622e-05, "loss": 0.6198, "step": 189 }, { "epoch": 0.769620253164557, "grad_norm": 0.5363572418793552, "learning_rate": 3.963954980528515e-05, "loss": 0.669, "step": 190 }, { "epoch": 0.7736708860759494, "grad_norm": 0.4847235239090232, "learning_rate": 3.9628743461849905e-05, "loss": 0.609, "step": 191 }, { "epoch": 0.7777215189873418, "grad_norm": 0.5327651812963341, "learning_rate": 3.961777903121329e-05, "loss": 0.65, "step": 192 }, { "epoch": 0.7817721518987342, "grad_norm": 0.5939665642855341, "learning_rate": 3.960665660168131e-05, "loss": 0.6636, "step": 193 }, { "epoch": 0.7858227848101266, "grad_norm": 0.5231210882110316, "learning_rate": 3.9595376262832485e-05, "loss": 0.6345, "step": 194 }, { "epoch": 0.789873417721519, "grad_norm": 0.6067922608022028, "learning_rate": 3.9583938105517127e-05, "loss": 0.6292, "step": 195 }, { "epoch": 0.7939240506329114, "grad_norm": 0.5353454387573278, "learning_rate": 3.957234222185657e-05, "loss": 0.6233, "step": 196 }, { "epoch": 0.7979746835443038, "grad_norm": 0.5656756765399391, "learning_rate": 3.9560588705242474e-05, "loss": 0.6443, "step": 197 }, { "epoch": 0.8020253164556962, "grad_norm": 0.5256348929205377, "learning_rate": 3.954867765033605e-05, "loss": 0.6543, "step": 198 }, { "epoch": 0.8060759493670886, "grad_norm": 0.5535689291188907, "learning_rate": 3.953660915306728e-05, "loss": 0.683, "step": 199 }, { "epoch": 0.810126582278481, "grad_norm": 0.4917113007903338, "learning_rate": 3.952438331063419e-05, "loss": 0.6325, "step": 200 }, { "epoch": 0.8141772151898734, "grad_norm": 0.5925868093726666, "learning_rate": 3.951200022150205e-05, "loss": 0.62, "step": 201 }, { "epoch": 0.8182278481012658, "grad_norm": 0.49057439506147305, "learning_rate": 3.949945998540253e-05, "loss": 0.6352, "step": 202 }, { "epoch": 0.8222784810126582, "grad_norm": 0.5638200390734699, "learning_rate": 3.9486762703332993e-05, "loss": 0.6443, "step": 203 }, { "epoch": 0.8263291139240506, "grad_norm": 0.5289828498529209, "learning_rate": 3.947390847755559e-05, "loss": 0.6374, "step": 204 }, { "epoch": 0.830379746835443, "grad_norm": 0.603119379400923, "learning_rate": 3.946089741159648e-05, "loss": 0.6456, "step": 205 }, { "epoch": 0.8344303797468354, "grad_norm": 0.5326759760845043, "learning_rate": 3.944772961024501e-05, "loss": 0.6614, "step": 206 }, { "epoch": 0.8384810126582278, "grad_norm": 0.5335135685694539, "learning_rate": 3.943440517955285e-05, "loss": 0.6089, "step": 207 }, { "epoch": 0.8425316455696202, "grad_norm": 0.6004183706977569, "learning_rate": 3.9420924226833126e-05, "loss": 0.6669, "step": 208 }, { "epoch": 0.8465822784810126, "grad_norm": 0.48026436839432624, "learning_rate": 3.9407286860659566e-05, "loss": 0.6561, "step": 209 }, { "epoch": 0.850632911392405, "grad_norm": 0.5384566944595727, "learning_rate": 3.9393493190865657e-05, "loss": 0.6467, "step": 210 }, { "epoch": 0.8546835443037974, "grad_norm": 0.4716707035336216, "learning_rate": 3.937954332854371e-05, "loss": 0.6154, "step": 211 }, { "epoch": 0.8587341772151899, "grad_norm": 0.49988250612839363, "learning_rate": 3.9365437386044016e-05, "loss": 0.6233, "step": 212 }, { "epoch": 0.8627848101265823, "grad_norm": 0.5341648878124812, "learning_rate": 3.935117547697387e-05, "loss": 0.648, "step": 213 }, { "epoch": 0.8668354430379747, "grad_norm": 1.286054280326782, "learning_rate": 3.933675771619675e-05, "loss": 0.6213, "step": 214 }, { "epoch": 0.8708860759493671, "grad_norm": 0.5090635817337975, "learning_rate": 3.932218421983131e-05, "loss": 0.644, "step": 215 }, { "epoch": 0.8749367088607595, "grad_norm": 0.5373609147621803, "learning_rate": 3.9307455105250484e-05, "loss": 0.6444, "step": 216 }, { "epoch": 0.8789873417721519, "grad_norm": 0.5448570773910852, "learning_rate": 3.929257049108054e-05, "loss": 0.6298, "step": 217 }, { "epoch": 0.8830379746835443, "grad_norm": 0.6127066764852339, "learning_rate": 3.927753049720011e-05, "loss": 0.6471, "step": 218 }, { "epoch": 0.8870886075949367, "grad_norm": 0.6077264689352353, "learning_rate": 3.9262335244739234e-05, "loss": 0.6256, "step": 219 }, { "epoch": 0.8911392405063291, "grad_norm": 0.605431710735135, "learning_rate": 3.92469848560784e-05, "loss": 0.6328, "step": 220 }, { "epoch": 0.8951898734177215, "grad_norm": 0.5619802320021483, "learning_rate": 3.923147945484751e-05, "loss": 0.6522, "step": 221 }, { "epoch": 0.8992405063291139, "grad_norm": 0.5543828051796416, "learning_rate": 3.9215819165924956e-05, "loss": 0.606, "step": 222 }, { "epoch": 0.9032911392405063, "grad_norm": 2.4905998765771034, "learning_rate": 3.920000411543654e-05, "loss": 0.6546, "step": 223 }, { "epoch": 0.9073417721518987, "grad_norm": 0.8281219756063728, "learning_rate": 3.9184034430754495e-05, "loss": 0.6609, "step": 224 }, { "epoch": 0.9113924050632911, "grad_norm": 0.7779629470191791, "learning_rate": 3.916791024049648e-05, "loss": 0.6598, "step": 225 }, { "epoch": 0.9154430379746835, "grad_norm": 0.6242858226210395, "learning_rate": 3.91516316745245e-05, "loss": 0.6365, "step": 226 }, { "epoch": 0.9194936708860759, "grad_norm": 0.7378987738617464, "learning_rate": 3.913519886394389e-05, "loss": 0.6425, "step": 227 }, { "epoch": 0.9235443037974683, "grad_norm": 0.7456626691714603, "learning_rate": 3.911861194110225e-05, "loss": 0.6256, "step": 228 }, { "epoch": 0.9275949367088607, "grad_norm": 2.213517553675689, "learning_rate": 3.910187103958837e-05, "loss": 0.6094, "step": 229 }, { "epoch": 0.9316455696202531, "grad_norm": 0.9604025874506216, "learning_rate": 3.908497629423117e-05, "loss": 0.6559, "step": 230 }, { "epoch": 0.9356962025316455, "grad_norm": 0.5854555011412973, "learning_rate": 3.9067927841098614e-05, "loss": 0.6452, "step": 231 }, { "epoch": 0.9397468354430379, "grad_norm": 0.711168392588542, "learning_rate": 3.9050725817496594e-05, "loss": 0.6103, "step": 232 }, { "epoch": 0.9437974683544303, "grad_norm": 0.6536043333530047, "learning_rate": 3.9033370361967844e-05, "loss": 0.6404, "step": 233 }, { "epoch": 0.9478481012658228, "grad_norm": 0.7538435891493747, "learning_rate": 3.901586161429081e-05, "loss": 0.6574, "step": 234 }, { "epoch": 0.9518987341772152, "grad_norm": 0.643547895945557, "learning_rate": 3.8998199715478545e-05, "loss": 0.6511, "step": 235 }, { "epoch": 0.9559493670886076, "grad_norm": 0.5498716703152023, "learning_rate": 3.8980384807777564e-05, "loss": 0.6231, "step": 236 }, { "epoch": 0.96, "grad_norm": 0.6026552860896944, "learning_rate": 3.896241703466667e-05, "loss": 0.6442, "step": 237 }, { "epoch": 0.9640506329113924, "grad_norm": 0.6449581737337179, "learning_rate": 3.894429654085585e-05, "loss": 0.633, "step": 238 }, { "epoch": 0.9681012658227848, "grad_norm": 0.7176566783737499, "learning_rate": 3.892602347228505e-05, "loss": 0.6348, "step": 239 }, { "epoch": 0.9721518987341772, "grad_norm": 0.5635331324052674, "learning_rate": 3.890759797612307e-05, "loss": 0.6348, "step": 240 }, { "epoch": 0.9762025316455696, "grad_norm": 0.6875837601307951, "learning_rate": 3.888902020076632e-05, "loss": 0.627, "step": 241 }, { "epoch": 0.980253164556962, "grad_norm": 0.5861012542559801, "learning_rate": 3.887029029583764e-05, "loss": 0.6648, "step": 242 }, { "epoch": 0.9843037974683544, "grad_norm": 0.6702179875322679, "learning_rate": 3.8851408412185125e-05, "loss": 0.6337, "step": 243 }, { "epoch": 0.9883544303797468, "grad_norm": 0.5825908235997014, "learning_rate": 3.8832374701880855e-05, "loss": 0.6234, "step": 244 }, { "epoch": 0.9924050632911392, "grad_norm": 0.818475474507664, "learning_rate": 3.881318931821972e-05, "loss": 0.633, "step": 245 }, { "epoch": 0.9964556962025316, "grad_norm": 0.5572117462996143, "learning_rate": 3.879385241571817e-05, "loss": 0.6507, "step": 246 }, { "epoch": 1.000506329113924, "grad_norm": 0.7251564007876375, "learning_rate": 3.8774364150112955e-05, "loss": 0.6159, "step": 247 }, { "epoch": 1.0045569620253165, "grad_norm": 0.6791374979763907, "learning_rate": 3.8754724678359884e-05, "loss": 0.524, "step": 248 }, { "epoch": 1.0086075949367088, "grad_norm": 0.6371850629053237, "learning_rate": 3.873493415863256e-05, "loss": 0.5415, "step": 249 }, { "epoch": 1.0126582278481013, "grad_norm": 0.652173125276497, "learning_rate": 3.871499275032111e-05, "loss": 0.5066, "step": 250 }, { "epoch": 1.0167088607594936, "grad_norm": 0.6090924223954526, "learning_rate": 3.869490061403091e-05, "loss": 0.5085, "step": 251 }, { "epoch": 1.0207594936708861, "grad_norm": 0.666824387976362, "learning_rate": 3.867465791158124e-05, "loss": 0.4872, "step": 252 }, { "epoch": 1.0248101265822784, "grad_norm": 0.5456798347309395, "learning_rate": 3.865426480600407e-05, "loss": 0.5352, "step": 253 }, { "epoch": 1.028860759493671, "grad_norm": 0.7020936294447288, "learning_rate": 3.863372146154264e-05, "loss": 0.5043, "step": 254 }, { "epoch": 1.0329113924050632, "grad_norm": 0.6310022115024625, "learning_rate": 3.861302804365024e-05, "loss": 0.4877, "step": 255 }, { "epoch": 1.0369620253164558, "grad_norm": 0.6717198745053748, "learning_rate": 3.85921847189888e-05, "loss": 0.5177, "step": 256 }, { "epoch": 1.041012658227848, "grad_norm": 0.576748162167274, "learning_rate": 3.85711916554276e-05, "loss": 0.5142, "step": 257 }, { "epoch": 1.0450632911392406, "grad_norm": 0.6767184744885569, "learning_rate": 3.85500490220419e-05, "loss": 0.5003, "step": 258 }, { "epoch": 1.0491139240506329, "grad_norm": 0.5693859931694916, "learning_rate": 3.852875698911154e-05, "loss": 0.5264, "step": 259 }, { "epoch": 1.0531645569620254, "grad_norm": 0.6749070811258204, "learning_rate": 3.850731572811963e-05, "loss": 0.5135, "step": 260 }, { "epoch": 1.0572151898734177, "grad_norm": 0.6771749956300589, "learning_rate": 3.848572541175116e-05, "loss": 0.5179, "step": 261 }, { "epoch": 1.0612658227848102, "grad_norm": 0.8033857424721887, "learning_rate": 3.846398621389154e-05, "loss": 0.5176, "step": 262 }, { "epoch": 1.0653164556962025, "grad_norm": 0.8300768170581386, "learning_rate": 3.84420983096253e-05, "loss": 0.5193, "step": 263 }, { "epoch": 1.069367088607595, "grad_norm": 0.6622884840104873, "learning_rate": 3.8420061875234606e-05, "loss": 0.4984, "step": 264 }, { "epoch": 1.0734177215189873, "grad_norm": 0.6717057925807328, "learning_rate": 3.839787708819787e-05, "loss": 0.5055, "step": 265 }, { "epoch": 1.0774683544303798, "grad_norm": 0.715456350614785, "learning_rate": 3.8375544127188325e-05, "loss": 0.5321, "step": 266 }, { "epoch": 1.081518987341772, "grad_norm": 0.619959038777474, "learning_rate": 3.8353063172072564e-05, "loss": 0.5185, "step": 267 }, { "epoch": 1.0855696202531646, "grad_norm": 0.5721729501205592, "learning_rate": 3.8330434403909105e-05, "loss": 0.5011, "step": 268 }, { "epoch": 1.089620253164557, "grad_norm": 0.5746153515792959, "learning_rate": 3.8307658004946934e-05, "loss": 0.5202, "step": 269 }, { "epoch": 1.0936708860759494, "grad_norm": 0.6295865773050041, "learning_rate": 3.8284734158624046e-05, "loss": 0.5495, "step": 270 }, { "epoch": 1.0977215189873417, "grad_norm": 0.6107979930610131, "learning_rate": 3.826166304956594e-05, "loss": 0.5387, "step": 271 }, { "epoch": 1.1017721518987342, "grad_norm": 0.6318488310071322, "learning_rate": 3.8238444863584164e-05, "loss": 0.4998, "step": 272 }, { "epoch": 1.1058227848101265, "grad_norm": 0.489714691423169, "learning_rate": 3.821507978767479e-05, "loss": 0.5069, "step": 273 }, { "epoch": 1.109873417721519, "grad_norm": 0.574481329436951, "learning_rate": 3.819156801001693e-05, "loss": 0.5218, "step": 274 }, { "epoch": 1.1139240506329113, "grad_norm": 0.594005639153082, "learning_rate": 3.816790971997121e-05, "loss": 0.5126, "step": 275 }, { "epoch": 1.1179746835443038, "grad_norm": 0.5235360262043505, "learning_rate": 3.8144105108078246e-05, "loss": 0.5116, "step": 276 }, { "epoch": 1.1220253164556961, "grad_norm": 0.5253060012309337, "learning_rate": 3.81201543660571e-05, "loss": 0.5172, "step": 277 }, { "epoch": 1.1260759493670887, "grad_norm": 0.6680439808600942, "learning_rate": 3.809605768680377e-05, "loss": 0.5254, "step": 278 }, { "epoch": 1.130126582278481, "grad_norm": 0.5161502480319025, "learning_rate": 3.807181526438958e-05, "loss": 0.5492, "step": 279 }, { "epoch": 1.1341772151898735, "grad_norm": 0.6948514837862666, "learning_rate": 3.8047427294059697e-05, "loss": 0.5076, "step": 280 }, { "epoch": 1.1382278481012658, "grad_norm": 0.5265523425536754, "learning_rate": 3.802289397223145e-05, "loss": 0.5184, "step": 281 }, { "epoch": 1.1422784810126583, "grad_norm": 0.6390305012817301, "learning_rate": 3.7998215496492854e-05, "loss": 0.5187, "step": 282 }, { "epoch": 1.1463291139240506, "grad_norm": 0.5637590441315751, "learning_rate": 3.797339206560096e-05, "loss": 0.5148, "step": 283 }, { "epoch": 1.150379746835443, "grad_norm": 0.6207285324885199, "learning_rate": 3.794842387948027e-05, "loss": 0.5262, "step": 284 }, { "epoch": 1.1544303797468354, "grad_norm": 0.5639800552702728, "learning_rate": 3.7923311139221114e-05, "loss": 0.5258, "step": 285 }, { "epoch": 1.158481012658228, "grad_norm": 0.5864188906931026, "learning_rate": 3.7898054047078054e-05, "loss": 0.4925, "step": 286 }, { "epoch": 1.1625316455696202, "grad_norm": 0.5440400764378178, "learning_rate": 3.787265280646825e-05, "loss": 0.4869, "step": 287 }, { "epoch": 1.1665822784810127, "grad_norm": 0.6737701941757426, "learning_rate": 3.7847107621969786e-05, "loss": 0.515, "step": 288 }, { "epoch": 1.170632911392405, "grad_norm": 0.7740443569900187, "learning_rate": 3.7821418699320064e-05, "loss": 0.5214, "step": 289 }, { "epoch": 1.1746835443037975, "grad_norm": 0.49510579621613565, "learning_rate": 3.7795586245414145e-05, "loss": 0.5148, "step": 290 }, { "epoch": 1.1787341772151898, "grad_norm": 0.7058614506978594, "learning_rate": 3.776961046830306e-05, "loss": 0.5059, "step": 291 }, { "epoch": 1.1827848101265823, "grad_norm": 0.5405623837521485, "learning_rate": 3.774349157719215e-05, "loss": 0.5191, "step": 292 }, { "epoch": 1.1868354430379746, "grad_norm": 0.5745719868544726, "learning_rate": 3.7717229782439365e-05, "loss": 0.5028, "step": 293 }, { "epoch": 1.1908860759493671, "grad_norm": 0.6124661661526899, "learning_rate": 3.769082529555359e-05, "loss": 0.53, "step": 294 }, { "epoch": 1.1949367088607594, "grad_norm": 0.5516172625900615, "learning_rate": 3.766427832919294e-05, "loss": 0.5147, "step": 295 }, { "epoch": 1.198987341772152, "grad_norm": 0.49741751333716705, "learning_rate": 3.7637589097163024e-05, "loss": 0.4972, "step": 296 }, { "epoch": 1.2030379746835442, "grad_norm": 0.6137748961732178, "learning_rate": 3.761075781441526e-05, "loss": 0.4964, "step": 297 }, { "epoch": 1.2070886075949367, "grad_norm": 0.5263692659572872, "learning_rate": 3.75837846970451e-05, "loss": 0.5116, "step": 298 }, { "epoch": 1.211139240506329, "grad_norm": 0.5742493128471297, "learning_rate": 3.755666996229032e-05, "loss": 0.5167, "step": 299 }, { "epoch": 1.2151898734177216, "grad_norm": 0.45916454744842167, "learning_rate": 3.752941382852927e-05, "loss": 0.5033, "step": 300 }, { "epoch": 1.2192405063291138, "grad_norm": 0.48007831463224854, "learning_rate": 3.7502016515279115e-05, "loss": 0.5167, "step": 301 }, { "epoch": 1.2232911392405064, "grad_norm": 0.4555753864072745, "learning_rate": 3.7474478243194043e-05, "loss": 0.5083, "step": 302 }, { "epoch": 1.2273417721518987, "grad_norm": 0.48774480177991236, "learning_rate": 3.744679923406351e-05, "loss": 0.519, "step": 303 }, { "epoch": 1.2313924050632912, "grad_norm": 0.48719356731213154, "learning_rate": 3.741897971081043e-05, "loss": 0.5267, "step": 304 }, { "epoch": 1.2354430379746835, "grad_norm": 0.5302603705098167, "learning_rate": 3.739101989748946e-05, "loss": 0.5221, "step": 305 }, { "epoch": 1.239493670886076, "grad_norm": 0.5772441997759046, "learning_rate": 3.7362920019285066e-05, "loss": 0.5118, "step": 306 }, { "epoch": 1.2435443037974683, "grad_norm": 0.5229996868421704, "learning_rate": 3.73346803025098e-05, "loss": 0.4969, "step": 307 }, { "epoch": 1.2475949367088608, "grad_norm": 0.512436282126902, "learning_rate": 3.730630097460247e-05, "loss": 0.5125, "step": 308 }, { "epoch": 1.251645569620253, "grad_norm": 0.6070861286111274, "learning_rate": 3.727778226412628e-05, "loss": 0.523, "step": 309 }, { "epoch": 1.2556962025316456, "grad_norm": 0.478703744591041, "learning_rate": 3.7249124400767006e-05, "loss": 0.5161, "step": 310 }, { "epoch": 1.259746835443038, "grad_norm": 0.5858786231108589, "learning_rate": 3.722032761533114e-05, "loss": 0.5254, "step": 311 }, { "epoch": 1.2637974683544304, "grad_norm": 0.4737677265525278, "learning_rate": 3.719139213974403e-05, "loss": 0.4963, "step": 312 }, { "epoch": 1.2678481012658227, "grad_norm": 0.5487332963245154, "learning_rate": 3.7162318207048006e-05, "loss": 0.5183, "step": 313 }, { "epoch": 1.2718987341772152, "grad_norm": 0.49022720281635207, "learning_rate": 3.713310605140055e-05, "loss": 0.5287, "step": 314 }, { "epoch": 1.2759493670886077, "grad_norm": 0.5300306196288976, "learning_rate": 3.710375590807233e-05, "loss": 0.5244, "step": 315 }, { "epoch": 1.28, "grad_norm": 0.5041026240549328, "learning_rate": 3.7074268013445365e-05, "loss": 0.5044, "step": 316 }, { "epoch": 1.2840506329113923, "grad_norm": 0.504600856731068, "learning_rate": 3.7044642605011114e-05, "loss": 0.5246, "step": 317 }, { "epoch": 1.2881012658227848, "grad_norm": 0.49238726100858277, "learning_rate": 3.701487992136854e-05, "loss": 0.4994, "step": 318 }, { "epoch": 1.2921518987341774, "grad_norm": 0.4961604946702843, "learning_rate": 3.69849802022222e-05, "loss": 0.4961, "step": 319 }, { "epoch": 1.2962025316455696, "grad_norm": 0.5518361924441585, "learning_rate": 3.6954943688380334e-05, "loss": 0.5194, "step": 320 }, { "epoch": 1.300253164556962, "grad_norm": 0.52146889519393, "learning_rate": 3.692477062175289e-05, "loss": 0.5194, "step": 321 }, { "epoch": 1.3043037974683545, "grad_norm": 0.6172520057045061, "learning_rate": 3.689446124534958e-05, "loss": 0.5283, "step": 322 }, { "epoch": 1.308354430379747, "grad_norm": 0.5105091630311582, "learning_rate": 3.686401580327799e-05, "loss": 0.5207, "step": 323 }, { "epoch": 1.3124050632911393, "grad_norm": 0.6300811159273395, "learning_rate": 3.683343454074149e-05, "loss": 0.4975, "step": 324 }, { "epoch": 1.3164556962025316, "grad_norm": 0.5777228791513471, "learning_rate": 3.6802717704037386e-05, "loss": 0.5135, "step": 325 }, { "epoch": 1.320506329113924, "grad_norm": 0.6818089189015873, "learning_rate": 3.6771865540554855e-05, "loss": 0.52, "step": 326 }, { "epoch": 1.3245569620253166, "grad_norm": 0.6346795397275034, "learning_rate": 3.674087829877297e-05, "loss": 0.5307, "step": 327 }, { "epoch": 1.3286075949367089, "grad_norm": 0.6208828815103379, "learning_rate": 3.6709756228258735e-05, "loss": 0.4998, "step": 328 }, { "epoch": 1.3326582278481012, "grad_norm": 0.6065338579190919, "learning_rate": 3.667849957966501e-05, "loss": 0.5206, "step": 329 }, { "epoch": 1.3367088607594937, "grad_norm": 0.6642899479726576, "learning_rate": 3.6647108604728546e-05, "loss": 0.502, "step": 330 }, { "epoch": 1.3407594936708862, "grad_norm": 0.527427958728121, "learning_rate": 3.661558355626795e-05, "loss": 0.4888, "step": 331 }, { "epoch": 1.3448101265822785, "grad_norm": 0.5486908192136926, "learning_rate": 3.658392468818163e-05, "loss": 0.5124, "step": 332 }, { "epoch": 1.3488607594936708, "grad_norm": 0.6228024717445497, "learning_rate": 3.655213225544574e-05, "loss": 0.5144, "step": 333 }, { "epoch": 1.3529113924050633, "grad_norm": 0.49480680627973356, "learning_rate": 3.652020651411218e-05, "loss": 0.5128, "step": 334 }, { "epoch": 1.3569620253164558, "grad_norm": 0.7229220335087813, "learning_rate": 3.6488147721306474e-05, "loss": 0.4936, "step": 335 }, { "epoch": 1.3610126582278481, "grad_norm": 0.5447758383005893, "learning_rate": 3.645595613522574e-05, "loss": 0.5142, "step": 336 }, { "epoch": 1.3650632911392404, "grad_norm": 0.6087713312397548, "learning_rate": 3.642363201513657e-05, "loss": 0.5172, "step": 337 }, { "epoch": 1.369113924050633, "grad_norm": 0.5511030656116345, "learning_rate": 3.6391175621373006e-05, "loss": 0.5486, "step": 338 }, { "epoch": 1.3731645569620254, "grad_norm": 0.5469300829226, "learning_rate": 3.6358587215334355e-05, "loss": 0.5255, "step": 339 }, { "epoch": 1.3772151898734177, "grad_norm": 0.5087308413840432, "learning_rate": 3.632586705948318e-05, "loss": 0.4897, "step": 340 }, { "epoch": 1.38126582278481, "grad_norm": 0.5303509719998464, "learning_rate": 3.629301541734311e-05, "loss": 0.5229, "step": 341 }, { "epoch": 1.3853164556962025, "grad_norm": 0.5706099496031939, "learning_rate": 3.626003255349676e-05, "loss": 0.5458, "step": 342 }, { "epoch": 1.389367088607595, "grad_norm": 0.5973683650286797, "learning_rate": 3.622691873358357e-05, "loss": 0.4956, "step": 343 }, { "epoch": 1.3934177215189874, "grad_norm": 0.4722411635520934, "learning_rate": 3.61936742242977e-05, "loss": 0.522, "step": 344 }, { "epoch": 1.3974683544303796, "grad_norm": 0.6589983411015343, "learning_rate": 3.6160299293385864e-05, "loss": 0.5008, "step": 345 }, { "epoch": 1.4015189873417722, "grad_norm": 0.49678683251744515, "learning_rate": 3.612679420964516e-05, "loss": 0.5148, "step": 346 }, { "epoch": 1.4055696202531647, "grad_norm": 0.6286411504585694, "learning_rate": 3.609315924292092e-05, "loss": 0.5166, "step": 347 }, { "epoch": 1.409620253164557, "grad_norm": 0.5521005596997488, "learning_rate": 3.6059394664104554e-05, "loss": 0.5197, "step": 348 }, { "epoch": 1.4136708860759493, "grad_norm": 0.6486258369820274, "learning_rate": 3.602550074513133e-05, "loss": 0.5124, "step": 349 }, { "epoch": 1.4177215189873418, "grad_norm": 0.5400332185823478, "learning_rate": 3.599147775897822e-05, "loss": 0.5149, "step": 350 }, { "epoch": 1.4217721518987343, "grad_norm": 0.46045837089806574, "learning_rate": 3.595732597966167e-05, "loss": 0.5018, "step": 351 }, { "epoch": 1.4258227848101266, "grad_norm": 0.5215152511823903, "learning_rate": 3.592304568223542e-05, "loss": 0.4917, "step": 352 }, { "epoch": 1.4298734177215189, "grad_norm": 0.5100807754614216, "learning_rate": 3.588863714278826e-05, "loss": 0.509, "step": 353 }, { "epoch": 1.4339240506329114, "grad_norm": 0.5701826426610781, "learning_rate": 3.585410063844186e-05, "loss": 0.5228, "step": 354 }, { "epoch": 1.437974683544304, "grad_norm": 0.5060428233904736, "learning_rate": 3.581943644734846e-05, "loss": 0.5216, "step": 355 }, { "epoch": 1.4420253164556962, "grad_norm": 0.5466837639241389, "learning_rate": 3.578464484868869e-05, "loss": 0.5014, "step": 356 }, { "epoch": 1.4460759493670885, "grad_norm": 0.4674490197139673, "learning_rate": 3.5749726122669316e-05, "loss": 0.5385, "step": 357 }, { "epoch": 1.450126582278481, "grad_norm": 0.5189607876103682, "learning_rate": 3.5714680550520943e-05, "loss": 0.4981, "step": 358 }, { "epoch": 1.4541772151898735, "grad_norm": 0.4940997686233705, "learning_rate": 3.5679508414495794e-05, "loss": 0.5132, "step": 359 }, { "epoch": 1.4582278481012658, "grad_norm": 0.5483385094097736, "learning_rate": 3.564420999786543e-05, "loss": 0.5248, "step": 360 }, { "epoch": 1.4622784810126581, "grad_norm": 0.5099038477816391, "learning_rate": 3.560878558491842e-05, "loss": 0.5383, "step": 361 }, { "epoch": 1.4663291139240506, "grad_norm": 0.5418249215189241, "learning_rate": 3.5573235460958145e-05, "loss": 0.494, "step": 362 }, { "epoch": 1.4703797468354431, "grad_norm": 0.5247391940596513, "learning_rate": 3.553755991230039e-05, "loss": 0.499, "step": 363 }, { "epoch": 1.4744303797468354, "grad_norm": 0.4864289818800211, "learning_rate": 3.5501759226271144e-05, "loss": 0.5239, "step": 364 }, { "epoch": 1.4784810126582277, "grad_norm": 0.5691865287791761, "learning_rate": 3.546583369120419e-05, "loss": 0.5158, "step": 365 }, { "epoch": 1.4825316455696202, "grad_norm": 0.4614653923346937, "learning_rate": 3.5429783596438864e-05, "loss": 0.5303, "step": 366 }, { "epoch": 1.4865822784810128, "grad_norm": 0.511804481955538, "learning_rate": 3.539360923231766e-05, "loss": 0.5147, "step": 367 }, { "epoch": 1.490632911392405, "grad_norm": 0.5266638269631707, "learning_rate": 3.535731089018394e-05, "loss": 0.5125, "step": 368 }, { "epoch": 1.4946835443037974, "grad_norm": 0.4473345034352312, "learning_rate": 3.532088886237956e-05, "loss": 0.5122, "step": 369 }, { "epoch": 1.4987341772151899, "grad_norm": 0.5318281547813267, "learning_rate": 3.528434344224253e-05, "loss": 0.5276, "step": 370 }, { "epoch": 1.5027848101265824, "grad_norm": 0.47935497874823674, "learning_rate": 3.524767492410464e-05, "loss": 0.5074, "step": 371 }, { "epoch": 1.5068354430379747, "grad_norm": 0.4688150943292483, "learning_rate": 3.521088360328908e-05, "loss": 0.5199, "step": 372 }, { "epoch": 1.510886075949367, "grad_norm": 0.5033024056533049, "learning_rate": 3.517396977610811e-05, "loss": 0.5286, "step": 373 }, { "epoch": 1.5149367088607595, "grad_norm": 0.5334860025902037, "learning_rate": 3.5136933739860595e-05, "loss": 0.4892, "step": 374 }, { "epoch": 1.518987341772152, "grad_norm": 0.49611088038305934, "learning_rate": 3.509977579282971e-05, "loss": 0.5307, "step": 375 }, { "epoch": 1.5230379746835443, "grad_norm": 0.5856919554339662, "learning_rate": 3.5062496234280424e-05, "loss": 0.5034, "step": 376 }, { "epoch": 1.5270886075949366, "grad_norm": 0.46115206670625825, "learning_rate": 3.502509536445719e-05, "loss": 0.4874, "step": 377 }, { "epoch": 1.531139240506329, "grad_norm": 0.5038803771607964, "learning_rate": 3.498757348458147e-05, "loss": 0.5268, "step": 378 }, { "epoch": 1.5351898734177216, "grad_norm": 0.4800559102302341, "learning_rate": 3.4949930896849324e-05, "loss": 0.496, "step": 379 }, { "epoch": 1.539240506329114, "grad_norm": 0.4775925235565738, "learning_rate": 3.491216790442899e-05, "loss": 0.5128, "step": 380 }, { "epoch": 1.5432911392405062, "grad_norm": 0.4942010053430929, "learning_rate": 3.487428481145839e-05, "loss": 0.5101, "step": 381 }, { "epoch": 1.5473417721518987, "grad_norm": 0.4758724550066122, "learning_rate": 3.483628192304278e-05, "loss": 0.5018, "step": 382 }, { "epoch": 1.5513924050632912, "grad_norm": 0.5962374485600993, "learning_rate": 3.479815954525219e-05, "loss": 0.5179, "step": 383 }, { "epoch": 1.5554430379746835, "grad_norm": 0.46186173676664, "learning_rate": 3.475991798511899e-05, "loss": 0.4886, "step": 384 }, { "epoch": 1.5594936708860758, "grad_norm": 0.4902003864815707, "learning_rate": 3.4721557550635464e-05, "loss": 0.537, "step": 385 }, { "epoch": 1.5635443037974683, "grad_norm": 0.5294424539520522, "learning_rate": 3.468307855075128e-05, "loss": 0.5211, "step": 386 }, { "epoch": 1.5675949367088609, "grad_norm": 0.47140842440765923, "learning_rate": 3.4644481295371005e-05, "loss": 0.4978, "step": 387 }, { "epoch": 1.5716455696202531, "grad_norm": 0.5221373920968913, "learning_rate": 3.460576609535163e-05, "loss": 0.5373, "step": 388 }, { "epoch": 1.5756962025316454, "grad_norm": 0.5318652474487179, "learning_rate": 3.456693326250006e-05, "loss": 0.5157, "step": 389 }, { "epoch": 1.579746835443038, "grad_norm": 0.4920055975247876, "learning_rate": 3.452798310957058e-05, "loss": 0.5284, "step": 390 }, { "epoch": 1.5837974683544305, "grad_norm": 0.5945671712356142, "learning_rate": 3.4488915950262386e-05, "loss": 0.538, "step": 391 }, { "epoch": 1.5878481012658228, "grad_norm": 0.46360899871238315, "learning_rate": 3.4449732099216985e-05, "loss": 0.4901, "step": 392 }, { "epoch": 1.591898734177215, "grad_norm": 0.5780340331091144, "learning_rate": 3.441043187201574e-05, "loss": 0.5335, "step": 393 }, { "epoch": 1.5959493670886076, "grad_norm": 0.5788711031049214, "learning_rate": 3.437101558517728e-05, "loss": 0.5151, "step": 394 }, { "epoch": 1.6, "grad_norm": 0.46735027786229294, "learning_rate": 3.433148355615496e-05, "loss": 0.5234, "step": 395 }, { "epoch": 1.6040506329113924, "grad_norm": 0.5627501905819234, "learning_rate": 3.4291836103334294e-05, "loss": 0.5397, "step": 396 }, { "epoch": 1.6081012658227847, "grad_norm": 0.5056468874301293, "learning_rate": 3.425207354603043e-05, "loss": 0.5055, "step": 397 }, { "epoch": 1.6121518987341772, "grad_norm": 0.5634726983118219, "learning_rate": 3.421219620448553e-05, "loss": 0.5241, "step": 398 }, { "epoch": 1.6162025316455697, "grad_norm": 0.546854032679703, "learning_rate": 3.417220439986623e-05, "loss": 0.5155, "step": 399 }, { "epoch": 1.620253164556962, "grad_norm": 0.5968033690560174, "learning_rate": 3.4132098454261024e-05, "loss": 0.5078, "step": 400 }, { "epoch": 1.6243037974683543, "grad_norm": 0.6131556649642994, "learning_rate": 3.4091878690677676e-05, "loss": 0.5203, "step": 401 }, { "epoch": 1.6283544303797468, "grad_norm": 0.5548547269819257, "learning_rate": 3.405154543304065e-05, "loss": 0.5267, "step": 402 }, { "epoch": 1.6324050632911393, "grad_norm": 0.7199541955876111, "learning_rate": 3.401109900618843e-05, "loss": 0.5205, "step": 403 }, { "epoch": 1.6364556962025316, "grad_norm": 0.5265261115637422, "learning_rate": 3.3970539735870996e-05, "loss": 0.5327, "step": 404 }, { "epoch": 1.640506329113924, "grad_norm": 0.677544345698355, "learning_rate": 3.392986794874714e-05, "loss": 0.5251, "step": 405 }, { "epoch": 1.6445569620253164, "grad_norm": 0.4968802880952616, "learning_rate": 3.388908397238184e-05, "loss": 0.4876, "step": 406 }, { "epoch": 1.648607594936709, "grad_norm": 0.6487825991579651, "learning_rate": 3.384818813524362e-05, "loss": 0.5329, "step": 407 }, { "epoch": 1.6526582278481012, "grad_norm": 0.4868978083486541, "learning_rate": 3.380718076670195e-05, "loss": 0.5104, "step": 408 }, { "epoch": 1.6567088607594935, "grad_norm": 0.5067290106950973, "learning_rate": 3.376606219702454e-05, "loss": 0.4958, "step": 409 }, { "epoch": 1.660759493670886, "grad_norm": 0.46992209009965685, "learning_rate": 3.372483275737468e-05, "loss": 0.5069, "step": 410 }, { "epoch": 1.6648101265822786, "grad_norm": 0.49231326662490643, "learning_rate": 3.368349277980861e-05, "loss": 0.5114, "step": 411 }, { "epoch": 1.6688607594936709, "grad_norm": 0.490224422821406, "learning_rate": 3.3642042597272844e-05, "loss": 0.5121, "step": 412 }, { "epoch": 1.6729113924050631, "grad_norm": 0.5034707125345705, "learning_rate": 3.360048254360144e-05, "loss": 0.511, "step": 413 }, { "epoch": 1.6769620253164557, "grad_norm": 0.4736863070403787, "learning_rate": 3.355881295351336e-05, "loss": 0.5076, "step": 414 }, { "epoch": 1.6810126582278482, "grad_norm": 0.4987038948668379, "learning_rate": 3.351703416260975e-05, "loss": 0.5128, "step": 415 }, { "epoch": 1.6850632911392405, "grad_norm": 0.49483477222102845, "learning_rate": 3.347514650737126e-05, "loss": 0.5072, "step": 416 }, { "epoch": 1.6891139240506328, "grad_norm": 0.4930685669643292, "learning_rate": 3.3433150325155295e-05, "loss": 0.5192, "step": 417 }, { "epoch": 1.6931645569620253, "grad_norm": 0.5667965860370128, "learning_rate": 3.339104595419334e-05, "loss": 0.5183, "step": 418 }, { "epoch": 1.6972151898734178, "grad_norm": 0.5835608429856615, "learning_rate": 3.3348833733588204e-05, "loss": 0.5006, "step": 419 }, { "epoch": 1.70126582278481, "grad_norm": 0.5234336252545885, "learning_rate": 3.3306514003311305e-05, "loss": 0.5422, "step": 420 }, { "epoch": 1.7053164556962024, "grad_norm": 0.6736520073202333, "learning_rate": 3.326408710419996e-05, "loss": 0.5096, "step": 421 }, { "epoch": 1.709367088607595, "grad_norm": 0.5877047348076696, "learning_rate": 3.322155337795454e-05, "loss": 0.5076, "step": 422 }, { "epoch": 1.7134177215189874, "grad_norm": 0.5160215031535551, "learning_rate": 3.317891316713587e-05, "loss": 0.5337, "step": 423 }, { "epoch": 1.7174683544303797, "grad_norm": 0.5665042273809753, "learning_rate": 3.313616681516231e-05, "loss": 0.5178, "step": 424 }, { "epoch": 1.721518987341772, "grad_norm": 0.5791896550859417, "learning_rate": 3.309331466630713e-05, "loss": 0.5121, "step": 425 }, { "epoch": 1.7255696202531645, "grad_norm": 0.5409898591870768, "learning_rate": 3.305035706569563e-05, "loss": 0.5061, "step": 426 }, { "epoch": 1.729620253164557, "grad_norm": 0.5574960025907992, "learning_rate": 3.3007294359302433e-05, "loss": 0.5052, "step": 427 }, { "epoch": 1.7336708860759493, "grad_norm": 0.48296271581109007, "learning_rate": 3.296412689394864e-05, "loss": 0.5006, "step": 428 }, { "epoch": 1.7377215189873416, "grad_norm": 0.4471404242745036, "learning_rate": 3.292085501729909e-05, "loss": 0.5339, "step": 429 }, { "epoch": 1.7417721518987341, "grad_norm": 0.4937576022453591, "learning_rate": 3.2877479077859534e-05, "loss": 0.5227, "step": 430 }, { "epoch": 1.7458227848101266, "grad_norm": 0.464424900505568, "learning_rate": 3.283399942497381e-05, "loss": 0.4975, "step": 431 }, { "epoch": 1.749873417721519, "grad_norm": 0.5093568215746925, "learning_rate": 3.279041640882108e-05, "loss": 0.5171, "step": 432 }, { "epoch": 1.7539240506329112, "grad_norm": 0.5957300552875353, "learning_rate": 3.2746730380412964e-05, "loss": 0.502, "step": 433 }, { "epoch": 1.7579746835443038, "grad_norm": 0.46056137671547087, "learning_rate": 3.2702941691590726e-05, "loss": 0.5167, "step": 434 }, { "epoch": 1.7620253164556963, "grad_norm": 0.5375542714017787, "learning_rate": 3.265905069502244e-05, "loss": 0.5254, "step": 435 }, { "epoch": 1.7660759493670886, "grad_norm": 0.4768478166912632, "learning_rate": 3.261505774420016e-05, "loss": 0.4868, "step": 436 }, { "epoch": 1.7701265822784809, "grad_norm": 0.5875160553493708, "learning_rate": 3.257096319343707e-05, "loss": 0.5225, "step": 437 }, { "epoch": 1.7741772151898734, "grad_norm": 0.5101253537984762, "learning_rate": 3.2526767397864614e-05, "loss": 0.5216, "step": 438 }, { "epoch": 1.7782278481012659, "grad_norm": 0.5254803995612166, "learning_rate": 3.248247071342966e-05, "loss": 0.51, "step": 439 }, { "epoch": 1.7822784810126582, "grad_norm": 0.4581699535437198, "learning_rate": 3.243807349689161e-05, "loss": 0.4841, "step": 440 }, { "epoch": 1.7863291139240505, "grad_norm": 0.48493333446131365, "learning_rate": 3.2393576105819544e-05, "loss": 0.4912, "step": 441 }, { "epoch": 1.790379746835443, "grad_norm": 0.5091411373672954, "learning_rate": 3.2348978898589333e-05, "loss": 0.5002, "step": 442 }, { "epoch": 1.7944303797468355, "grad_norm": 0.4983202676968306, "learning_rate": 3.230428223438075e-05, "loss": 0.5029, "step": 443 }, { "epoch": 1.7984810126582278, "grad_norm": 0.49615157047368424, "learning_rate": 3.225948647317459e-05, "loss": 0.5103, "step": 444 }, { "epoch": 1.80253164556962, "grad_norm": 0.48067261139585143, "learning_rate": 3.2214591975749745e-05, "loss": 0.5281, "step": 445 }, { "epoch": 1.8065822784810126, "grad_norm": 0.4874135374319231, "learning_rate": 3.216959910368034e-05, "loss": 0.4957, "step": 446 }, { "epoch": 1.8106329113924051, "grad_norm": 0.4977507424348179, "learning_rate": 3.212450821933277e-05, "loss": 0.5188, "step": 447 }, { "epoch": 1.8146835443037974, "grad_norm": 0.4410900667045802, "learning_rate": 3.207931968586281e-05, "loss": 0.5056, "step": 448 }, { "epoch": 1.81873417721519, "grad_norm": 0.5109984585995753, "learning_rate": 3.203403386721272e-05, "loss": 0.5091, "step": 449 }, { "epoch": 1.8227848101265822, "grad_norm": 0.47467805157500004, "learning_rate": 3.1988651128108245e-05, "loss": 0.5135, "step": 450 }, { "epoch": 1.8268354430379747, "grad_norm": 0.45632146202450985, "learning_rate": 3.194317183405573e-05, "loss": 0.5065, "step": 451 }, { "epoch": 1.8308860759493673, "grad_norm": 0.4654786068912121, "learning_rate": 3.189759635133914e-05, "loss": 0.5215, "step": 452 }, { "epoch": 1.8349367088607595, "grad_norm": 0.45432133686029297, "learning_rate": 3.185192504701718e-05, "loss": 0.5059, "step": 453 }, { "epoch": 1.8389873417721518, "grad_norm": 0.45382160454659676, "learning_rate": 3.1806158288920234e-05, "loss": 0.5136, "step": 454 }, { "epoch": 1.8430379746835444, "grad_norm": 0.4222207935384708, "learning_rate": 3.1760296445647477e-05, "loss": 0.4915, "step": 455 }, { "epoch": 1.8470886075949369, "grad_norm": 0.5121813049579058, "learning_rate": 3.1714339886563896e-05, "loss": 0.5056, "step": 456 }, { "epoch": 1.8511392405063292, "grad_norm": 0.4540701326273945, "learning_rate": 3.166828898179731e-05, "loss": 0.5397, "step": 457 }, { "epoch": 1.8551898734177215, "grad_norm": 0.5933196880357904, "learning_rate": 3.162214410223536e-05, "loss": 0.5042, "step": 458 }, { "epoch": 1.859240506329114, "grad_norm": 0.4393684088516174, "learning_rate": 3.157590561952257e-05, "loss": 0.4998, "step": 459 }, { "epoch": 1.8632911392405065, "grad_norm": 0.5328626724421073, "learning_rate": 3.152957390605732e-05, "loss": 0.5094, "step": 460 }, { "epoch": 1.8673417721518988, "grad_norm": 0.4757303215865902, "learning_rate": 3.148314933498886e-05, "loss": 0.5098, "step": 461 }, { "epoch": 1.871392405063291, "grad_norm": 0.496016094238223, "learning_rate": 3.143663228021431e-05, "loss": 0.5059, "step": 462 }, { "epoch": 1.8754430379746836, "grad_norm": 0.4741167574325065, "learning_rate": 3.1390023116375624e-05, "loss": 0.4969, "step": 463 }, { "epoch": 1.879493670886076, "grad_norm": 0.5086934776189659, "learning_rate": 3.134332221885661e-05, "loss": 0.54, "step": 464 }, { "epoch": 1.8835443037974684, "grad_norm": 0.4883325856656935, "learning_rate": 3.129652996377987e-05, "loss": 0.5185, "step": 465 }, { "epoch": 1.8875949367088607, "grad_norm": 0.4820413776199168, "learning_rate": 3.12496467280038e-05, "loss": 0.5078, "step": 466 }, { "epoch": 1.8916455696202532, "grad_norm": 0.44364880986494715, "learning_rate": 3.120267288911952e-05, "loss": 0.4946, "step": 467 }, { "epoch": 1.8956962025316457, "grad_norm": 0.44895681437726687, "learning_rate": 3.11556088254479e-05, "loss": 0.53, "step": 468 }, { "epoch": 1.899746835443038, "grad_norm": 0.42952223562957287, "learning_rate": 3.11084549160364e-05, "loss": 0.5041, "step": 469 }, { "epoch": 1.9037974683544303, "grad_norm": 0.43562050219940446, "learning_rate": 3.106121154065615e-05, "loss": 0.4933, "step": 470 }, { "epoch": 1.9078481012658228, "grad_norm": 0.5192748515708394, "learning_rate": 3.1013879079798805e-05, "loss": 0.5018, "step": 471 }, { "epoch": 1.9118987341772153, "grad_norm": 0.47045259653996796, "learning_rate": 3.096645791467348e-05, "loss": 0.5149, "step": 472 }, { "epoch": 1.9159493670886076, "grad_norm": 0.527608264953023, "learning_rate": 3.091894842720373e-05, "loss": 0.494, "step": 473 }, { "epoch": 1.92, "grad_norm": 0.4908344909761733, "learning_rate": 3.0871351000024425e-05, "loss": 0.5127, "step": 474 }, { "epoch": 1.9240506329113924, "grad_norm": 0.605642234735486, "learning_rate": 3.0823666016478716e-05, "loss": 0.5166, "step": 475 }, { "epoch": 1.928101265822785, "grad_norm": 0.49507490419400624, "learning_rate": 3.0775893860614896e-05, "loss": 0.4925, "step": 476 }, { "epoch": 1.9321518987341773, "grad_norm": 0.5840654276739927, "learning_rate": 3.0728034917183336e-05, "loss": 0.4952, "step": 477 }, { "epoch": 1.9362025316455695, "grad_norm": 0.44823086604951656, "learning_rate": 3.06800895716334e-05, "loss": 0.4964, "step": 478 }, { "epoch": 1.940253164556962, "grad_norm": 0.5673805208463173, "learning_rate": 3.063205821011029e-05, "loss": 0.5121, "step": 479 }, { "epoch": 1.9443037974683546, "grad_norm": 0.502933063176208, "learning_rate": 3.0583941219452016e-05, "loss": 0.5031, "step": 480 }, { "epoch": 1.9483544303797469, "grad_norm": 0.5458164066557897, "learning_rate": 3.053573898718618e-05, "loss": 0.5353, "step": 481 }, { "epoch": 1.9524050632911392, "grad_norm": 0.48004494091231725, "learning_rate": 3.0487451901526956e-05, "loss": 0.4906, "step": 482 }, { "epoch": 1.9564556962025317, "grad_norm": 0.47446267637321776, "learning_rate": 3.0439080351371875e-05, "loss": 0.527, "step": 483 }, { "epoch": 1.9605063291139242, "grad_norm": 0.5015305248571487, "learning_rate": 3.0390624726298764e-05, "loss": 0.501, "step": 484 }, { "epoch": 1.9645569620253165, "grad_norm": 0.4669555681315626, "learning_rate": 3.034208541656255e-05, "loss": 0.5104, "step": 485 }, { "epoch": 1.9686075949367088, "grad_norm": 0.5065764933354725, "learning_rate": 3.029346281309218e-05, "loss": 0.495, "step": 486 }, { "epoch": 1.9726582278481013, "grad_norm": 0.4677314070569522, "learning_rate": 3.0244757307487415e-05, "loss": 0.5184, "step": 487 }, { "epoch": 1.9767088607594938, "grad_norm": 0.44531769150968215, "learning_rate": 3.019596929201569e-05, "loss": 0.5238, "step": 488 }, { "epoch": 1.980759493670886, "grad_norm": 0.4731455131291045, "learning_rate": 3.0147099159608985e-05, "loss": 0.4994, "step": 489 }, { "epoch": 1.9848101265822784, "grad_norm": 0.44994000890861724, "learning_rate": 3.0098147303860616e-05, "loss": 0.5118, "step": 490 }, { "epoch": 1.988860759493671, "grad_norm": 0.5379334987342066, "learning_rate": 3.0049114119022117e-05, "loss": 0.5013, "step": 491 }, { "epoch": 1.9929113924050634, "grad_norm": 0.468410973755851, "learning_rate": 3.0000000000000004e-05, "loss": 0.4972, "step": 492 }, { "epoch": 1.9969620253164557, "grad_norm": 0.4704015328065598, "learning_rate": 2.995080534235264e-05, "loss": 0.5144, "step": 493 }, { "epoch": 2.001012658227848, "grad_norm": 0.5392127747822606, "learning_rate": 2.9901530542287044e-05, "loss": 0.4859, "step": 494 }, { "epoch": 2.0050632911392405, "grad_norm": 0.695831441326847, "learning_rate": 2.9852175996655676e-05, "loss": 0.3744, "step": 495 }, { "epoch": 2.009113924050633, "grad_norm": 0.7151065772937333, "learning_rate": 2.980274210295326e-05, "loss": 0.3851, "step": 496 }, { "epoch": 2.013164556962025, "grad_norm": 0.7649931656911287, "learning_rate": 2.9753229259313578e-05, "loss": 0.3738, "step": 497 }, { "epoch": 2.0172151898734176, "grad_norm": 0.5507821643378089, "learning_rate": 2.9703637864506274e-05, "loss": 0.3742, "step": 498 }, { "epoch": 2.02126582278481, "grad_norm": 0.635496795474154, "learning_rate": 2.965396831793362e-05, "loss": 0.3514, "step": 499 }, { "epoch": 2.0253164556962027, "grad_norm": 0.49877445528997133, "learning_rate": 2.9604221019627316e-05, "loss": 0.3452, "step": 500 }, { "epoch": 2.0293670886075947, "grad_norm": 0.5302674054287261, "learning_rate": 2.955439637024526e-05, "loss": 0.3543, "step": 501 }, { "epoch": 2.0334177215189873, "grad_norm": 0.5730374741171816, "learning_rate": 2.9504494771068334e-05, "loss": 0.3696, "step": 502 }, { "epoch": 2.0374683544303798, "grad_norm": 0.8250632633577367, "learning_rate": 2.9454516623997156e-05, "loss": 0.3683, "step": 503 }, { "epoch": 2.0415189873417723, "grad_norm": 0.447503282492855, "learning_rate": 2.9404462331548847e-05, "loss": 0.3687, "step": 504 }, { "epoch": 2.0455696202531644, "grad_norm": 0.6001111827572677, "learning_rate": 2.93543322968538e-05, "loss": 0.3521, "step": 505 }, { "epoch": 2.049620253164557, "grad_norm": 0.5017860514218542, "learning_rate": 2.9304126923652428e-05, "loss": 0.3744, "step": 506 }, { "epoch": 2.0536708860759494, "grad_norm": 0.5302715628657263, "learning_rate": 2.9253846616291896e-05, "loss": 0.3597, "step": 507 }, { "epoch": 2.057721518987342, "grad_norm": 0.49834459708029594, "learning_rate": 2.9203491779722896e-05, "loss": 0.3664, "step": 508 }, { "epoch": 2.061772151898734, "grad_norm": 0.5508619229414398, "learning_rate": 2.9153062819496357e-05, "loss": 0.3653, "step": 509 }, { "epoch": 2.0658227848101265, "grad_norm": 0.4518794483223388, "learning_rate": 2.9102560141760178e-05, "loss": 0.371, "step": 510 }, { "epoch": 2.069873417721519, "grad_norm": 0.548619396235738, "learning_rate": 2.9051984153256004e-05, "loss": 0.3723, "step": 511 }, { "epoch": 2.0739240506329115, "grad_norm": 0.4959729492348829, "learning_rate": 2.900133526131588e-05, "loss": 0.3583, "step": 512 }, { "epoch": 2.0779746835443036, "grad_norm": 0.48204729368023747, "learning_rate": 2.8950613873859025e-05, "loss": 0.3623, "step": 513 }, { "epoch": 2.082025316455696, "grad_norm": 0.41676636896869457, "learning_rate": 2.8899820399388515e-05, "loss": 0.3553, "step": 514 }, { "epoch": 2.0860759493670886, "grad_norm": 0.5339065353493071, "learning_rate": 2.8848955246988012e-05, "loss": 0.3608, "step": 515 }, { "epoch": 2.090126582278481, "grad_norm": 0.477090993130748, "learning_rate": 2.879801882631847e-05, "loss": 0.3575, "step": 516 }, { "epoch": 2.094177215189873, "grad_norm": 0.6161262278829153, "learning_rate": 2.8747011547614808e-05, "loss": 0.3653, "step": 517 }, { "epoch": 2.0982278481012657, "grad_norm": 0.45815658741847126, "learning_rate": 2.8695933821682635e-05, "loss": 0.3728, "step": 518 }, { "epoch": 2.1022784810126582, "grad_norm": 0.5946643396009783, "learning_rate": 2.864478605989494e-05, "loss": 0.3573, "step": 519 }, { "epoch": 2.1063291139240508, "grad_norm": 0.42934261830778025, "learning_rate": 2.8593568674188765e-05, "loss": 0.3689, "step": 520 }, { "epoch": 2.110379746835443, "grad_norm": 0.5400163912777278, "learning_rate": 2.8542282077061892e-05, "loss": 0.3735, "step": 521 }, { "epoch": 2.1144303797468353, "grad_norm": 0.42436718729865663, "learning_rate": 2.8490926681569523e-05, "loss": 0.3707, "step": 522 }, { "epoch": 2.118481012658228, "grad_norm": 0.5302102711222236, "learning_rate": 2.8439502901320956e-05, "loss": 0.3517, "step": 523 }, { "epoch": 2.1225316455696204, "grad_norm": 0.4309130261095578, "learning_rate": 2.8388011150476237e-05, "loss": 0.355, "step": 524 }, { "epoch": 2.1265822784810124, "grad_norm": 0.5025397480081368, "learning_rate": 2.8336451843742866e-05, "loss": 0.385, "step": 525 }, { "epoch": 2.130632911392405, "grad_norm": 0.42213653467162143, "learning_rate": 2.8284825396372387e-05, "loss": 0.3478, "step": 526 }, { "epoch": 2.1346835443037975, "grad_norm": 0.4781066435866388, "learning_rate": 2.8233132224157132e-05, "loss": 0.3469, "step": 527 }, { "epoch": 2.13873417721519, "grad_norm": 0.4160770498553429, "learning_rate": 2.8181372743426805e-05, "loss": 0.3466, "step": 528 }, { "epoch": 2.1427848101265825, "grad_norm": 0.4468796121696071, "learning_rate": 2.8129547371045128e-05, "loss": 0.3521, "step": 529 }, { "epoch": 2.1468354430379746, "grad_norm": 0.43890689002896555, "learning_rate": 2.8077656524406534e-05, "loss": 0.3485, "step": 530 }, { "epoch": 2.150886075949367, "grad_norm": 0.44102452788636826, "learning_rate": 2.802570062143278e-05, "loss": 0.3504, "step": 531 }, { "epoch": 2.1549367088607596, "grad_norm": 0.4820267235291308, "learning_rate": 2.7973680080569555e-05, "loss": 0.3655, "step": 532 }, { "epoch": 2.1589873417721517, "grad_norm": 0.41869321203548565, "learning_rate": 2.792159532078314e-05, "loss": 0.3592, "step": 533 }, { "epoch": 2.163037974683544, "grad_norm": 0.4481187191206551, "learning_rate": 2.7869446761557033e-05, "loss": 0.388, "step": 534 }, { "epoch": 2.1670886075949367, "grad_norm": 0.46918472817475343, "learning_rate": 2.781723482288857e-05, "loss": 0.3708, "step": 535 }, { "epoch": 2.1711392405063292, "grad_norm": 0.4321744428443935, "learning_rate": 2.7764959925285517e-05, "loss": 0.3736, "step": 536 }, { "epoch": 2.1751898734177217, "grad_norm": 0.4728294986189712, "learning_rate": 2.771262248976272e-05, "loss": 0.3636, "step": 537 }, { "epoch": 2.179240506329114, "grad_norm": 0.4150013205562907, "learning_rate": 2.7660222937838677e-05, "loss": 0.3655, "step": 538 }, { "epoch": 2.1832911392405063, "grad_norm": 0.44436807480120943, "learning_rate": 2.7607761691532186e-05, "loss": 0.3398, "step": 539 }, { "epoch": 2.187341772151899, "grad_norm": 0.42804042979588103, "learning_rate": 2.7555239173358916e-05, "loss": 0.3342, "step": 540 }, { "epoch": 2.191392405063291, "grad_norm": 0.459105534456635, "learning_rate": 2.7502655806328e-05, "loss": 0.3422, "step": 541 }, { "epoch": 2.1954430379746834, "grad_norm": 0.421946066259788, "learning_rate": 2.7450012013938648e-05, "loss": 0.356, "step": 542 }, { "epoch": 2.199493670886076, "grad_norm": 0.4339993067959308, "learning_rate": 2.739730822017673e-05, "loss": 0.3626, "step": 543 }, { "epoch": 2.2035443037974685, "grad_norm": 0.46721231605114905, "learning_rate": 2.7344544849511355e-05, "loss": 0.3411, "step": 544 }, { "epoch": 2.207594936708861, "grad_norm": 0.40841009406746465, "learning_rate": 2.7291722326891456e-05, "loss": 0.3537, "step": 545 }, { "epoch": 2.211645569620253, "grad_norm": 0.4387701246799477, "learning_rate": 2.723884107774236e-05, "loss": 0.3506, "step": 546 }, { "epoch": 2.2156962025316456, "grad_norm": 0.42322959218327016, "learning_rate": 2.718590152796239e-05, "loss": 0.3514, "step": 547 }, { "epoch": 2.219746835443038, "grad_norm": 0.44789519854522963, "learning_rate": 2.71329041039194e-05, "loss": 0.3487, "step": 548 }, { "epoch": 2.22379746835443, "grad_norm": 0.4559295375248663, "learning_rate": 2.7079849232447357e-05, "loss": 0.3623, "step": 549 }, { "epoch": 2.2278481012658227, "grad_norm": 0.4317717953076896, "learning_rate": 2.7026737340842895e-05, "loss": 0.3702, "step": 550 }, { "epoch": 2.231898734177215, "grad_norm": 0.4201410630665602, "learning_rate": 2.697356885686189e-05, "loss": 0.3604, "step": 551 }, { "epoch": 2.2359493670886077, "grad_norm": 0.46070440182361144, "learning_rate": 2.6920344208716014e-05, "loss": 0.361, "step": 552 }, { "epoch": 2.24, "grad_norm": 0.4096320946466004, "learning_rate": 2.6867063825069252e-05, "loss": 0.3378, "step": 553 }, { "epoch": 2.2440506329113923, "grad_norm": 0.4497795220269693, "learning_rate": 2.6813728135034494e-05, "loss": 0.3638, "step": 554 }, { "epoch": 2.248101265822785, "grad_norm": 0.4458354870939812, "learning_rate": 2.6760337568170056e-05, "loss": 0.3225, "step": 555 }, { "epoch": 2.2521518987341773, "grad_norm": 0.424291400048105, "learning_rate": 2.6706892554476226e-05, "loss": 0.3694, "step": 556 }, { "epoch": 2.2562025316455694, "grad_norm": 0.4487618701884435, "learning_rate": 2.6653393524391795e-05, "loss": 0.3557, "step": 557 }, { "epoch": 2.260253164556962, "grad_norm": 0.46304637147607175, "learning_rate": 2.6599840908790592e-05, "loss": 0.3684, "step": 558 }, { "epoch": 2.2643037974683544, "grad_norm": 0.40826554743242455, "learning_rate": 2.6546235138978028e-05, "loss": 0.3467, "step": 559 }, { "epoch": 2.268354430379747, "grad_norm": 0.4127920001452329, "learning_rate": 2.6492576646687597e-05, "loss": 0.3484, "step": 560 }, { "epoch": 2.2724050632911394, "grad_norm": 0.40418565738524603, "learning_rate": 2.6438865864077425e-05, "loss": 0.3601, "step": 561 }, { "epoch": 2.2764556962025315, "grad_norm": 0.4493020320857425, "learning_rate": 2.6385103223726766e-05, "loss": 0.3578, "step": 562 }, { "epoch": 2.280506329113924, "grad_norm": 0.44524818213507594, "learning_rate": 2.6331289158632537e-05, "loss": 0.3651, "step": 563 }, { "epoch": 2.2845569620253166, "grad_norm": 0.3941483997701693, "learning_rate": 2.6277424102205817e-05, "loss": 0.3474, "step": 564 }, { "epoch": 2.2886075949367086, "grad_norm": 0.4300354716090139, "learning_rate": 2.6223508488268374e-05, "loss": 0.3687, "step": 565 }, { "epoch": 2.292658227848101, "grad_norm": 0.43860366652783234, "learning_rate": 2.6169542751049148e-05, "loss": 0.3684, "step": 566 }, { "epoch": 2.2967088607594937, "grad_norm": 0.41777243442160766, "learning_rate": 2.6115527325180754e-05, "loss": 0.3722, "step": 567 }, { "epoch": 2.300759493670886, "grad_norm": 0.42381082427200023, "learning_rate": 2.606146264569603e-05, "loss": 0.3597, "step": 568 }, { "epoch": 2.3048101265822787, "grad_norm": 0.42723955438706035, "learning_rate": 2.6007349148024447e-05, "loss": 0.3464, "step": 569 }, { "epoch": 2.3088607594936708, "grad_norm": 0.39151647068857737, "learning_rate": 2.5953187267988694e-05, "loss": 0.382, "step": 570 }, { "epoch": 2.3129113924050633, "grad_norm": 0.4392992909373022, "learning_rate": 2.5898977441801097e-05, "loss": 0.3712, "step": 571 }, { "epoch": 2.316962025316456, "grad_norm": 0.42952082103649686, "learning_rate": 2.584472010606015e-05, "loss": 0.3814, "step": 572 }, { "epoch": 2.321012658227848, "grad_norm": 0.4108917407423964, "learning_rate": 2.5790415697746976e-05, "loss": 0.366, "step": 573 }, { "epoch": 2.3250632911392404, "grad_norm": 0.40623559035806933, "learning_rate": 2.5736064654221808e-05, "loss": 0.3434, "step": 574 }, { "epoch": 2.329113924050633, "grad_norm": 0.3988960679507731, "learning_rate": 2.568166741322048e-05, "loss": 0.3545, "step": 575 }, { "epoch": 2.3331645569620254, "grad_norm": 0.4183928088684155, "learning_rate": 2.56272244128509e-05, "loss": 0.3475, "step": 576 }, { "epoch": 2.337215189873418, "grad_norm": 0.39118287350598396, "learning_rate": 2.55727360915895e-05, "loss": 0.3859, "step": 577 }, { "epoch": 2.34126582278481, "grad_norm": 0.43625634113583966, "learning_rate": 2.5518202888277734e-05, "loss": 0.3587, "step": 578 }, { "epoch": 2.3453164556962025, "grad_norm": 0.394586444767902, "learning_rate": 2.5463625242118523e-05, "loss": 0.349, "step": 579 }, { "epoch": 2.349367088607595, "grad_norm": 0.46254387542092346, "learning_rate": 2.5409003592672723e-05, "loss": 0.3635, "step": 580 }, { "epoch": 2.353417721518987, "grad_norm": 0.45245303135411824, "learning_rate": 2.535433837985559e-05, "loss": 0.3616, "step": 581 }, { "epoch": 2.3574683544303796, "grad_norm": 0.4588127419666294, "learning_rate": 2.529963004393324e-05, "loss": 0.3682, "step": 582 }, { "epoch": 2.361518987341772, "grad_norm": 0.43212709498921925, "learning_rate": 2.524487902551908e-05, "loss": 0.3662, "step": 583 }, { "epoch": 2.3655696202531646, "grad_norm": 0.44749974269856424, "learning_rate": 2.519008576557029e-05, "loss": 0.3627, "step": 584 }, { "epoch": 2.369620253164557, "grad_norm": 0.4129086578445037, "learning_rate": 2.5135250705384254e-05, "loss": 0.377, "step": 585 }, { "epoch": 2.3736708860759492, "grad_norm": 0.4174842893909308, "learning_rate": 2.5080374286595007e-05, "loss": 0.3835, "step": 586 }, { "epoch": 2.3777215189873417, "grad_norm": 0.44098350239013845, "learning_rate": 2.5025456951169677e-05, "loss": 0.3917, "step": 587 }, { "epoch": 2.3817721518987343, "grad_norm": 0.45235824910014083, "learning_rate": 2.4970499141404942e-05, "loss": 0.3698, "step": 588 }, { "epoch": 2.3858227848101268, "grad_norm": 0.42582901427923087, "learning_rate": 2.491550129992345e-05, "loss": 0.3497, "step": 589 }, { "epoch": 2.389873417721519, "grad_norm": 0.45213281050287046, "learning_rate": 2.486046386967024e-05, "loss": 0.3527, "step": 590 }, { "epoch": 2.3939240506329114, "grad_norm": 0.38808965957397995, "learning_rate": 2.4805387293909214e-05, "loss": 0.3442, "step": 591 }, { "epoch": 2.397974683544304, "grad_norm": 0.44916351797559684, "learning_rate": 2.4750272016219552e-05, "loss": 0.3622, "step": 592 }, { "epoch": 2.4020253164556964, "grad_norm": 0.4314932941976515, "learning_rate": 2.4695118480492114e-05, "loss": 0.3441, "step": 593 }, { "epoch": 2.4060759493670885, "grad_norm": 0.4059212627402688, "learning_rate": 2.4639927130925898e-05, "loss": 0.3759, "step": 594 }, { "epoch": 2.410126582278481, "grad_norm": 0.42171911760061603, "learning_rate": 2.458469841202444e-05, "loss": 0.3883, "step": 595 }, { "epoch": 2.4141772151898735, "grad_norm": 0.42785959842337395, "learning_rate": 2.452943276859226e-05, "loss": 0.3796, "step": 596 }, { "epoch": 2.418227848101266, "grad_norm": 0.42021209017441347, "learning_rate": 2.447413064573125e-05, "loss": 0.3574, "step": 597 }, { "epoch": 2.422278481012658, "grad_norm": 0.43426343404989143, "learning_rate": 2.4418792488837095e-05, "loss": 0.3667, "step": 598 }, { "epoch": 2.4263291139240506, "grad_norm": 0.42527366270927647, "learning_rate": 2.4363418743595713e-05, "loss": 0.3705, "step": 599 }, { "epoch": 2.430379746835443, "grad_norm": 0.537809347326925, "learning_rate": 2.430800985597963e-05, "loss": 0.3746, "step": 600 }, { "epoch": 2.4344303797468356, "grad_norm": 0.42491402519602317, "learning_rate": 2.4252566272244415e-05, "loss": 0.3715, "step": 601 }, { "epoch": 2.4384810126582277, "grad_norm": 0.4419601879640057, "learning_rate": 2.4197088438925063e-05, "loss": 0.3347, "step": 602 }, { "epoch": 2.44253164556962, "grad_norm": 0.5208890697361118, "learning_rate": 2.4141576802832417e-05, "loss": 0.3945, "step": 603 }, { "epoch": 2.4465822784810127, "grad_norm": 0.4162484811464081, "learning_rate": 2.408603181104957e-05, "loss": 0.3756, "step": 604 }, { "epoch": 2.4506329113924052, "grad_norm": 0.4660957306285697, "learning_rate": 2.4030453910928245e-05, "loss": 0.3817, "step": 605 }, { "epoch": 2.4546835443037973, "grad_norm": 0.4372089109520261, "learning_rate": 2.397484355008521e-05, "loss": 0.3655, "step": 606 }, { "epoch": 2.45873417721519, "grad_norm": 0.44614542686581043, "learning_rate": 2.3919201176398662e-05, "loss": 0.342, "step": 607 }, { "epoch": 2.4627848101265823, "grad_norm": 0.4330615711895265, "learning_rate": 2.3863527238004633e-05, "loss": 0.3601, "step": 608 }, { "epoch": 2.466835443037975, "grad_norm": 0.46834732357217707, "learning_rate": 2.380782218329337e-05, "loss": 0.37, "step": 609 }, { "epoch": 2.470886075949367, "grad_norm": 0.390007771856772, "learning_rate": 2.3752086460905725e-05, "loss": 0.3523, "step": 610 }, { "epoch": 2.4749367088607594, "grad_norm": 0.4657167672239417, "learning_rate": 2.3696320519729544e-05, "loss": 0.3727, "step": 611 }, { "epoch": 2.478987341772152, "grad_norm": 0.4333592456257204, "learning_rate": 2.3640524808896045e-05, "loss": 0.3633, "step": 612 }, { "epoch": 2.4830379746835445, "grad_norm": 0.399258968831537, "learning_rate": 2.3584699777776222e-05, "loss": 0.3515, "step": 613 }, { "epoch": 2.4870886075949366, "grad_norm": 0.4490484559654212, "learning_rate": 2.3528845875977195e-05, "loss": 0.36, "step": 614 }, { "epoch": 2.491139240506329, "grad_norm": 0.41339383161933335, "learning_rate": 2.3472963553338614e-05, "loss": 0.3429, "step": 615 }, { "epoch": 2.4951898734177216, "grad_norm": 0.41266110811696854, "learning_rate": 2.341705325992901e-05, "loss": 0.3629, "step": 616 }, { "epoch": 2.499240506329114, "grad_norm": 0.44238796227059746, "learning_rate": 2.336111544604222e-05, "loss": 0.3561, "step": 617 }, { "epoch": 2.503291139240506, "grad_norm": 0.39447919130473247, "learning_rate": 2.33051505621937e-05, "loss": 0.3647, "step": 618 }, { "epoch": 2.5073417721518987, "grad_norm": 0.4202083830411146, "learning_rate": 2.324915905911693e-05, "loss": 0.3537, "step": 619 }, { "epoch": 2.511392405063291, "grad_norm": 0.411402248118739, "learning_rate": 2.319314138775977e-05, "loss": 0.3802, "step": 620 }, { "epoch": 2.5154430379746833, "grad_norm": 0.42779655909148945, "learning_rate": 2.3137097999280856e-05, "loss": 0.3739, "step": 621 }, { "epoch": 2.519493670886076, "grad_norm": 0.4292897780760611, "learning_rate": 2.308102934504593e-05, "loss": 0.3616, "step": 622 }, { "epoch": 2.5235443037974683, "grad_norm": 0.406666253539991, "learning_rate": 2.3024935876624222e-05, "loss": 0.3755, "step": 623 }, { "epoch": 2.527594936708861, "grad_norm": 0.3984972700442974, "learning_rate": 2.2968818045784813e-05, "loss": 0.3543, "step": 624 }, { "epoch": 2.5316455696202533, "grad_norm": 0.4109109547361494, "learning_rate": 2.2912676304493006e-05, "loss": 0.3677, "step": 625 }, { "epoch": 2.5356962025316454, "grad_norm": 0.41741345276506364, "learning_rate": 2.2856511104906668e-05, "loss": 0.3764, "step": 626 }, { "epoch": 2.539746835443038, "grad_norm": 0.37498463713518154, "learning_rate": 2.2800322899372586e-05, "loss": 0.3483, "step": 627 }, { "epoch": 2.5437974683544304, "grad_norm": 0.42046996477452747, "learning_rate": 2.2744112140422844e-05, "loss": 0.3846, "step": 628 }, { "epoch": 2.547848101265823, "grad_norm": 0.3976314125639926, "learning_rate": 2.2687879280771177e-05, "loss": 0.3796, "step": 629 }, { "epoch": 2.5518987341772155, "grad_norm": 0.4119655836914932, "learning_rate": 2.26316247733093e-05, "loss": 0.3621, "step": 630 }, { "epoch": 2.5559493670886075, "grad_norm": 0.4245630033468918, "learning_rate": 2.257534907110328e-05, "loss": 0.3638, "step": 631 }, { "epoch": 2.56, "grad_norm": 0.3958197296595268, "learning_rate": 2.2519052627389882e-05, "loss": 0.3844, "step": 632 }, { "epoch": 2.5640506329113926, "grad_norm": 0.4001252447593726, "learning_rate": 2.246273589557294e-05, "loss": 0.3593, "step": 633 }, { "epoch": 2.5681012658227846, "grad_norm": 0.4094429347958576, "learning_rate": 2.240639932921966e-05, "loss": 0.353, "step": 634 }, { "epoch": 2.572151898734177, "grad_norm": 0.40437456126375293, "learning_rate": 2.2350043382056995e-05, "loss": 0.3605, "step": 635 }, { "epoch": 2.5762025316455697, "grad_norm": 0.379282586692726, "learning_rate": 2.2293668507968015e-05, "loss": 0.3746, "step": 636 }, { "epoch": 2.580253164556962, "grad_norm": 0.40586313323622447, "learning_rate": 2.2237275160988186e-05, "loss": 0.3629, "step": 637 }, { "epoch": 2.5843037974683547, "grad_norm": 0.41173202990130003, "learning_rate": 2.2180863795301787e-05, "loss": 0.3478, "step": 638 }, { "epoch": 2.5883544303797468, "grad_norm": 0.3913630054594112, "learning_rate": 2.212443486523819e-05, "loss": 0.3489, "step": 639 }, { "epoch": 2.5924050632911393, "grad_norm": 0.3939337078838234, "learning_rate": 2.2067988825268243e-05, "loss": 0.3755, "step": 640 }, { "epoch": 2.596455696202532, "grad_norm": 0.4156468843896675, "learning_rate": 2.2011526130000596e-05, "loss": 0.3519, "step": 641 }, { "epoch": 2.600506329113924, "grad_norm": 0.4185952943382881, "learning_rate": 2.1955047234178038e-05, "loss": 0.3659, "step": 642 }, { "epoch": 2.6045569620253164, "grad_norm": 0.4095435551469605, "learning_rate": 2.1898552592673825e-05, "loss": 0.3793, "step": 643 }, { "epoch": 2.608607594936709, "grad_norm": 0.42265294107836815, "learning_rate": 2.184204266048803e-05, "loss": 0.3672, "step": 644 }, { "epoch": 2.6126582278481014, "grad_norm": 0.43609316837209283, "learning_rate": 2.1785517892743887e-05, "loss": 0.3754, "step": 645 }, { "epoch": 2.616708860759494, "grad_norm": 0.41639780030491114, "learning_rate": 2.17289787446841e-05, "loss": 0.3587, "step": 646 }, { "epoch": 2.620759493670886, "grad_norm": 0.43039197406928303, "learning_rate": 2.1672425671667198e-05, "loss": 0.3665, "step": 647 }, { "epoch": 2.6248101265822785, "grad_norm": 0.41687515535881947, "learning_rate": 2.161585912916385e-05, "loss": 0.3729, "step": 648 }, { "epoch": 2.628860759493671, "grad_norm": 0.42582354428960456, "learning_rate": 2.1559279572753214e-05, "loss": 0.3633, "step": 649 }, { "epoch": 2.632911392405063, "grad_norm": 0.42130224274837264, "learning_rate": 2.1502687458119268e-05, "loss": 0.3738, "step": 650 }, { "epoch": 2.6369620253164556, "grad_norm": 0.3901457170746934, "learning_rate": 2.1446083241047116e-05, "loss": 0.3494, "step": 651 }, { "epoch": 2.641012658227848, "grad_norm": 0.38626703919799216, "learning_rate": 2.1389467377419333e-05, "loss": 0.3776, "step": 652 }, { "epoch": 2.6450632911392407, "grad_norm": 0.38603210700534973, "learning_rate": 2.133284032321232e-05, "loss": 0.3707, "step": 653 }, { "epoch": 2.649113924050633, "grad_norm": 0.3902553002718617, "learning_rate": 2.1276202534492566e-05, "loss": 0.3576, "step": 654 }, { "epoch": 2.6531645569620252, "grad_norm": 0.41149707582106493, "learning_rate": 2.121955446741306e-05, "loss": 0.3515, "step": 655 }, { "epoch": 2.6572151898734178, "grad_norm": 0.4122760427813406, "learning_rate": 2.1162896578209517e-05, "loss": 0.3819, "step": 656 }, { "epoch": 2.6612658227848103, "grad_norm": 0.3586750289656699, "learning_rate": 2.1106229323196813e-05, "loss": 0.3585, "step": 657 }, { "epoch": 2.6653164556962023, "grad_norm": 0.39078771638964505, "learning_rate": 2.1049553158765214e-05, "loss": 0.364, "step": 658 }, { "epoch": 2.669367088607595, "grad_norm": 0.4108415504619746, "learning_rate": 2.0992868541376764e-05, "loss": 0.3554, "step": 659 }, { "epoch": 2.6734177215189874, "grad_norm": 0.38816937915667366, "learning_rate": 2.093617592756158e-05, "loss": 0.3468, "step": 660 }, { "epoch": 2.67746835443038, "grad_norm": 0.44037509719142226, "learning_rate": 2.0879475773914167e-05, "loss": 0.3641, "step": 661 }, { "epoch": 2.6815189873417724, "grad_norm": 0.40187462576853417, "learning_rate": 2.082276853708978e-05, "loss": 0.3674, "step": 662 }, { "epoch": 2.6855696202531645, "grad_norm": 0.39124853696121803, "learning_rate": 2.076605467380071e-05, "loss": 0.3568, "step": 663 }, { "epoch": 2.689620253164557, "grad_norm": 0.39587944264820185, "learning_rate": 2.0709334640812613e-05, "loss": 0.3783, "step": 664 }, { "epoch": 2.6936708860759495, "grad_norm": 0.38980814544995757, "learning_rate": 2.0652608894940824e-05, "loss": 0.3753, "step": 665 }, { "epoch": 2.6977215189873416, "grad_norm": 0.3907568051488874, "learning_rate": 2.0595877893046722e-05, "loss": 0.3832, "step": 666 }, { "epoch": 2.701772151898734, "grad_norm": 0.40389267435131954, "learning_rate": 2.0539142092033985e-05, "loss": 0.3572, "step": 667 }, { "epoch": 2.7058227848101266, "grad_norm": 0.3830225006126153, "learning_rate": 2.048240194884496e-05, "loss": 0.3474, "step": 668 }, { "epoch": 2.709873417721519, "grad_norm": 0.39785070093480995, "learning_rate": 2.042565792045695e-05, "loss": 0.369, "step": 669 }, { "epoch": 2.7139240506329116, "grad_norm": 0.36808209183129964, "learning_rate": 2.036891046387857e-05, "loss": 0.3669, "step": 670 }, { "epoch": 2.7179746835443037, "grad_norm": 0.4020153422768235, "learning_rate": 2.0312160036146036e-05, "loss": 0.3711, "step": 671 }, { "epoch": 2.7220253164556962, "grad_norm": 0.4038935034661924, "learning_rate": 2.025540709431948e-05, "loss": 0.3573, "step": 672 }, { "epoch": 2.7260759493670887, "grad_norm": 0.375762752717514, "learning_rate": 2.0198652095479298e-05, "loss": 0.3712, "step": 673 }, { "epoch": 2.730126582278481, "grad_norm": 0.404146341356224, "learning_rate": 2.014189549672245e-05, "loss": 0.3612, "step": 674 }, { "epoch": 2.7341772151898733, "grad_norm": 0.4191745855828847, "learning_rate": 2.0085137755158776e-05, "loss": 0.3851, "step": 675 }, { "epoch": 2.738227848101266, "grad_norm": 0.4232211868528661, "learning_rate": 2.0028379327907327e-05, "loss": 0.3504, "step": 676 }, { "epoch": 2.7422784810126584, "grad_norm": 0.39115210398931993, "learning_rate": 1.9971620672092676e-05, "loss": 0.3388, "step": 677 }, { "epoch": 2.746329113924051, "grad_norm": 0.41830028057627483, "learning_rate": 1.991486224484123e-05, "loss": 0.3403, "step": 678 }, { "epoch": 2.750379746835443, "grad_norm": 0.4074961950776157, "learning_rate": 1.985810450327756e-05, "loss": 0.3627, "step": 679 }, { "epoch": 2.7544303797468355, "grad_norm": 0.40152451125026484, "learning_rate": 1.9801347904520706e-05, "loss": 0.384, "step": 680 }, { "epoch": 2.758481012658228, "grad_norm": 0.4068148248104805, "learning_rate": 1.974459290568053e-05, "loss": 0.3434, "step": 681 }, { "epoch": 2.76253164556962, "grad_norm": 0.372855985089429, "learning_rate": 1.968783996385397e-05, "loss": 0.3488, "step": 682 }, { "epoch": 2.7665822784810126, "grad_norm": 0.3859316478976, "learning_rate": 1.963108953612143e-05, "loss": 0.3664, "step": 683 }, { "epoch": 2.770632911392405, "grad_norm": 0.3866334873713681, "learning_rate": 1.9574342079543056e-05, "loss": 0.354, "step": 684 }, { "epoch": 2.7746835443037976, "grad_norm": 0.3989903851827184, "learning_rate": 1.9517598051155046e-05, "loss": 0.3469, "step": 685 }, { "epoch": 2.77873417721519, "grad_norm": 0.3932069757104323, "learning_rate": 1.9460857907966025e-05, "loss": 0.3736, "step": 686 }, { "epoch": 2.782784810126582, "grad_norm": 0.4075203344509612, "learning_rate": 1.9404122106953285e-05, "loss": 0.3363, "step": 687 }, { "epoch": 2.7868354430379747, "grad_norm": 0.3824202711678968, "learning_rate": 1.9347391105059176e-05, "loss": 0.3592, "step": 688 }, { "epoch": 2.790886075949367, "grad_norm": 0.3838072641216809, "learning_rate": 1.92906653591874e-05, "loss": 0.3633, "step": 689 }, { "epoch": 2.7949367088607593, "grad_norm": 0.38939607411612165, "learning_rate": 1.9233945326199295e-05, "loss": 0.3646, "step": 690 }, { "epoch": 2.798987341772152, "grad_norm": 0.41021722695065643, "learning_rate": 1.917723146291022e-05, "loss": 0.3532, "step": 691 }, { "epoch": 2.8030379746835443, "grad_norm": 0.38188097248629643, "learning_rate": 1.912052422608584e-05, "loss": 0.3549, "step": 692 }, { "epoch": 2.807088607594937, "grad_norm": 0.3858276102131695, "learning_rate": 1.9063824072438428e-05, "loss": 0.3487, "step": 693 }, { "epoch": 2.8111392405063294, "grad_norm": 0.3785916021441111, "learning_rate": 1.9007131458623246e-05, "loss": 0.3501, "step": 694 }, { "epoch": 2.8151898734177214, "grad_norm": 0.39449823505093207, "learning_rate": 1.895044684123479e-05, "loss": 0.373, "step": 695 }, { "epoch": 2.819240506329114, "grad_norm": 0.38023174441781477, "learning_rate": 1.8893770676803194e-05, "loss": 0.3488, "step": 696 }, { "epoch": 2.8232911392405065, "grad_norm": 0.4235084819020359, "learning_rate": 1.8837103421790486e-05, "loss": 0.3652, "step": 697 }, { "epoch": 2.8273417721518985, "grad_norm": 0.37539140303649166, "learning_rate": 1.8780445532586952e-05, "loss": 0.3757, "step": 698 }, { "epoch": 2.831392405063291, "grad_norm": 0.37407702170136553, "learning_rate": 1.872379746550743e-05, "loss": 0.3742, "step": 699 }, { "epoch": 2.8354430379746836, "grad_norm": 0.45500301571733787, "learning_rate": 1.866715967678769e-05, "loss": 0.3766, "step": 700 }, { "epoch": 2.839493670886076, "grad_norm": 0.3848531725389483, "learning_rate": 1.861053262258067e-05, "loss": 0.3639, "step": 701 }, { "epoch": 2.8435443037974686, "grad_norm": 0.37860213206360066, "learning_rate": 1.8553916758952897e-05, "loss": 0.3507, "step": 702 }, { "epoch": 2.8475949367088607, "grad_norm": 0.36158201774167414, "learning_rate": 1.8497312541880735e-05, "loss": 0.3615, "step": 703 }, { "epoch": 2.851645569620253, "grad_norm": 0.3795905730431033, "learning_rate": 1.8440720427246786e-05, "loss": 0.3568, "step": 704 }, { "epoch": 2.8556962025316457, "grad_norm": 0.4258803314215164, "learning_rate": 1.8384140870836157e-05, "loss": 0.3596, "step": 705 }, { "epoch": 2.8597468354430378, "grad_norm": 0.3741177481362283, "learning_rate": 1.8327574328332806e-05, "loss": 0.3527, "step": 706 }, { "epoch": 2.8637974683544303, "grad_norm": 0.42143120630588304, "learning_rate": 1.8271021255315906e-05, "loss": 0.3575, "step": 707 }, { "epoch": 2.867848101265823, "grad_norm": 0.40366931855135835, "learning_rate": 1.8214482107256117e-05, "loss": 0.3765, "step": 708 }, { "epoch": 2.8718987341772153, "grad_norm": 0.3839365764123498, "learning_rate": 1.8157957339511968e-05, "loss": 0.3471, "step": 709 }, { "epoch": 2.875949367088608, "grad_norm": 0.412063998284568, "learning_rate": 1.8101447407326182e-05, "loss": 0.3321, "step": 710 }, { "epoch": 2.88, "grad_norm": 0.38270803452224483, "learning_rate": 1.8044952765821966e-05, "loss": 0.3661, "step": 711 }, { "epoch": 2.8840506329113924, "grad_norm": 0.44264785646752625, "learning_rate": 1.7988473869999407e-05, "loss": 0.35, "step": 712 }, { "epoch": 2.888101265822785, "grad_norm": 0.3937704118885698, "learning_rate": 1.7932011174731764e-05, "loss": 0.347, "step": 713 }, { "epoch": 2.892151898734177, "grad_norm": 0.4511477525404023, "learning_rate": 1.7875565134761817e-05, "loss": 0.3671, "step": 714 }, { "epoch": 2.8962025316455695, "grad_norm": 0.40949054324515993, "learning_rate": 1.7819136204698226e-05, "loss": 0.3665, "step": 715 }, { "epoch": 2.900253164556962, "grad_norm": 0.41257251951258794, "learning_rate": 1.776272483901182e-05, "loss": 0.3982, "step": 716 }, { "epoch": 2.9043037974683545, "grad_norm": 0.4119014216180005, "learning_rate": 1.7706331492031995e-05, "loss": 0.3624, "step": 717 }, { "epoch": 2.908354430379747, "grad_norm": 0.3897715836303693, "learning_rate": 1.764995661794301e-05, "loss": 0.3609, "step": 718 }, { "epoch": 2.912405063291139, "grad_norm": 0.38518855070926666, "learning_rate": 1.759360067078035e-05, "loss": 0.3501, "step": 719 }, { "epoch": 2.9164556962025316, "grad_norm": 0.4336213073351749, "learning_rate": 1.7537264104427064e-05, "loss": 0.4073, "step": 720 }, { "epoch": 2.920506329113924, "grad_norm": 0.3578639579108877, "learning_rate": 1.748094737261012e-05, "loss": 0.355, "step": 721 }, { "epoch": 2.9245569620253162, "grad_norm": 0.4306053301261114, "learning_rate": 1.7424650928896726e-05, "loss": 0.3633, "step": 722 }, { "epoch": 2.9286075949367087, "grad_norm": 0.4229027599327127, "learning_rate": 1.7368375226690712e-05, "loss": 0.3697, "step": 723 }, { "epoch": 2.9326582278481013, "grad_norm": 0.4061351666253514, "learning_rate": 1.731212071922883e-05, "loss": 0.3502, "step": 724 }, { "epoch": 2.9367088607594938, "grad_norm": 0.40773016513547056, "learning_rate": 1.7255887859577156e-05, "loss": 0.3582, "step": 725 }, { "epoch": 2.9407594936708863, "grad_norm": 0.41958417337025833, "learning_rate": 1.7199677100627427e-05, "loss": 0.3682, "step": 726 }, { "epoch": 2.9448101265822784, "grad_norm": 0.3908890260730899, "learning_rate": 1.7143488895093343e-05, "loss": 0.3479, "step": 727 }, { "epoch": 2.948860759493671, "grad_norm": 0.3744774439461002, "learning_rate": 1.7087323695506994e-05, "loss": 0.3645, "step": 728 }, { "epoch": 2.9529113924050634, "grad_norm": 0.42765392019347565, "learning_rate": 1.7031181954215194e-05, "loss": 0.3365, "step": 729 }, { "epoch": 2.9569620253164555, "grad_norm": 0.38554435414039245, "learning_rate": 1.6975064123375788e-05, "loss": 0.3834, "step": 730 }, { "epoch": 2.961012658227848, "grad_norm": 0.36914332838694985, "learning_rate": 1.6918970654954084e-05, "loss": 0.3577, "step": 731 }, { "epoch": 2.9650632911392405, "grad_norm": 0.3763300225419106, "learning_rate": 1.686290200071915e-05, "loss": 0.3513, "step": 732 }, { "epoch": 2.969113924050633, "grad_norm": 0.3563745162703487, "learning_rate": 1.6806858612240234e-05, "loss": 0.3697, "step": 733 }, { "epoch": 2.9731645569620255, "grad_norm": 0.383054146302113, "learning_rate": 1.6750840940883078e-05, "loss": 0.3354, "step": 734 }, { "epoch": 2.9772151898734176, "grad_norm": 0.3683483008022521, "learning_rate": 1.6694849437806305e-05, "loss": 0.345, "step": 735 }, { "epoch": 2.98126582278481, "grad_norm": 0.3880294163970687, "learning_rate": 1.663888455395778e-05, "loss": 0.3568, "step": 736 }, { "epoch": 2.9853164556962026, "grad_norm": 0.3790401856310894, "learning_rate": 1.6582946740070995e-05, "loss": 0.3494, "step": 737 }, { "epoch": 2.9893670886075947, "grad_norm": 0.3811047547311447, "learning_rate": 1.6527036446661396e-05, "loss": 0.3745, "step": 738 }, { "epoch": 2.993417721518987, "grad_norm": 0.384737903852753, "learning_rate": 1.6471154124022818e-05, "loss": 0.3603, "step": 739 }, { "epoch": 2.9974683544303797, "grad_norm": 0.41825059165726863, "learning_rate": 1.6415300222223788e-05, "loss": 0.3481, "step": 740 }, { "epoch": 3.0015189873417722, "grad_norm": 0.4187669392837981, "learning_rate": 1.6359475191103958e-05, "loss": 0.3312, "step": 741 }, { "epoch": 3.0055696202531648, "grad_norm": 0.5340132009077057, "learning_rate": 1.6303679480270466e-05, "loss": 0.2582, "step": 742 }, { "epoch": 3.009620253164557, "grad_norm": 0.3813761058746791, "learning_rate": 1.624791353909428e-05, "loss": 0.2494, "step": 743 }, { "epoch": 3.0136708860759494, "grad_norm": 0.6806903349900925, "learning_rate": 1.619217781670663e-05, "loss": 0.2675, "step": 744 }, { "epoch": 3.017721518987342, "grad_norm": 0.525894597083129, "learning_rate": 1.6136472761995373e-05, "loss": 0.262, "step": 745 }, { "epoch": 3.0217721518987344, "grad_norm": 0.5099318343499517, "learning_rate": 1.608079882360134e-05, "loss": 0.2415, "step": 746 }, { "epoch": 3.0258227848101265, "grad_norm": 0.5189718473502425, "learning_rate": 1.60251564499148e-05, "loss": 0.2477, "step": 747 }, { "epoch": 3.029873417721519, "grad_norm": 0.415924314945101, "learning_rate": 1.596954608907176e-05, "loss": 0.2323, "step": 748 }, { "epoch": 3.0339240506329115, "grad_norm": 0.42143734820355283, "learning_rate": 1.591396818895043e-05, "loss": 0.2449, "step": 749 }, { "epoch": 3.037974683544304, "grad_norm": 0.4340132311296633, "learning_rate": 1.585842319716759e-05, "loss": 0.2505, "step": 750 }, { "epoch": 3.042025316455696, "grad_norm": 0.4314895347330589, "learning_rate": 1.5802911561074944e-05, "loss": 0.2458, "step": 751 }, { "epoch": 3.0460759493670886, "grad_norm": 0.37088412140651555, "learning_rate": 1.5747433727755595e-05, "loss": 0.2648, "step": 752 }, { "epoch": 3.050126582278481, "grad_norm": 0.4092790462545322, "learning_rate": 1.5691990144020376e-05, "loss": 0.2705, "step": 753 }, { "epoch": 3.0541772151898736, "grad_norm": 0.41009393392726623, "learning_rate": 1.5636581256404297e-05, "loss": 0.2472, "step": 754 }, { "epoch": 3.0582278481012657, "grad_norm": 0.3833361244399568, "learning_rate": 1.558120751116291e-05, "loss": 0.2691, "step": 755 }, { "epoch": 3.062278481012658, "grad_norm": 0.3404507306897185, "learning_rate": 1.552586935426876e-05, "loss": 0.2581, "step": 756 }, { "epoch": 3.0663291139240507, "grad_norm": 0.3552771289596434, "learning_rate": 1.547056723140774e-05, "loss": 0.24, "step": 757 }, { "epoch": 3.0703797468354432, "grad_norm": 0.35858738321784706, "learning_rate": 1.5415301587975565e-05, "loss": 0.2406, "step": 758 }, { "epoch": 3.0744303797468353, "grad_norm": 0.33666487030180614, "learning_rate": 1.536007286907411e-05, "loss": 0.2545, "step": 759 }, { "epoch": 3.078481012658228, "grad_norm": 0.355852600297072, "learning_rate": 1.5304881519507896e-05, "loss": 0.2764, "step": 760 }, { "epoch": 3.0825316455696203, "grad_norm": 0.33286998917832283, "learning_rate": 1.5249727983780453e-05, "loss": 0.2637, "step": 761 }, { "epoch": 3.086582278481013, "grad_norm": 0.36741335272547737, "learning_rate": 1.5194612706090786e-05, "loss": 0.2401, "step": 762 }, { "epoch": 3.090632911392405, "grad_norm": 0.35653789430129285, "learning_rate": 1.5139536130329771e-05, "loss": 0.2656, "step": 763 }, { "epoch": 3.0946835443037974, "grad_norm": 0.34547831124651707, "learning_rate": 1.508449870007656e-05, "loss": 0.2718, "step": 764 }, { "epoch": 3.09873417721519, "grad_norm": 0.3426845915437223, "learning_rate": 1.5029500858595056e-05, "loss": 0.2616, "step": 765 }, { "epoch": 3.1027848101265825, "grad_norm": 0.33577529654897814, "learning_rate": 1.4974543048830328e-05, "loss": 0.2447, "step": 766 }, { "epoch": 3.1068354430379745, "grad_norm": 0.34576775036611856, "learning_rate": 1.4919625713405e-05, "loss": 0.2247, "step": 767 }, { "epoch": 3.110886075949367, "grad_norm": 0.3373349012455855, "learning_rate": 1.4864749294615756e-05, "loss": 0.2453, "step": 768 }, { "epoch": 3.1149367088607596, "grad_norm": 0.35208957483416214, "learning_rate": 1.4809914234429716e-05, "loss": 0.2145, "step": 769 }, { "epoch": 3.118987341772152, "grad_norm": 0.3720549324821614, "learning_rate": 1.4755120974480923e-05, "loss": 0.2672, "step": 770 }, { "epoch": 3.123037974683544, "grad_norm": 0.34281804954099415, "learning_rate": 1.4700369956066771e-05, "loss": 0.2559, "step": 771 }, { "epoch": 3.1270886075949367, "grad_norm": 0.3673550262154981, "learning_rate": 1.4645661620144413e-05, "loss": 0.2377, "step": 772 }, { "epoch": 3.131139240506329, "grad_norm": 0.3396708215875063, "learning_rate": 1.4590996407327284e-05, "loss": 0.2474, "step": 773 }, { "epoch": 3.1351898734177217, "grad_norm": 0.34420313954906107, "learning_rate": 1.4536374757881487e-05, "loss": 0.2493, "step": 774 }, { "epoch": 3.1392405063291138, "grad_norm": 0.3397696924399664, "learning_rate": 1.4481797111722271e-05, "loss": 0.2382, "step": 775 }, { "epoch": 3.1432911392405063, "grad_norm": 0.365159789739419, "learning_rate": 1.4427263908410507e-05, "loss": 0.2467, "step": 776 }, { "epoch": 3.147341772151899, "grad_norm": 0.33432501985384155, "learning_rate": 1.4372775587149108e-05, "loss": 0.2324, "step": 777 }, { "epoch": 3.1513924050632913, "grad_norm": 0.3411978973364612, "learning_rate": 1.4318332586779522e-05, "loss": 0.2555, "step": 778 }, { "epoch": 3.1554430379746834, "grad_norm": 0.3335311146099223, "learning_rate": 1.4263935345778202e-05, "loss": 0.2452, "step": 779 }, { "epoch": 3.159493670886076, "grad_norm": 0.34180843916521214, "learning_rate": 1.420958430225303e-05, "loss": 0.2382, "step": 780 }, { "epoch": 3.1635443037974684, "grad_norm": 0.34054541282607065, "learning_rate": 1.415527989393985e-05, "loss": 0.236, "step": 781 }, { "epoch": 3.167594936708861, "grad_norm": 0.338478120193187, "learning_rate": 1.410102255819891e-05, "loss": 0.2376, "step": 782 }, { "epoch": 3.171645569620253, "grad_norm": 0.3410136620917761, "learning_rate": 1.404681273201131e-05, "loss": 0.2409, "step": 783 }, { "epoch": 3.1756962025316455, "grad_norm": 0.33055973009128514, "learning_rate": 1.399265085197556e-05, "loss": 0.2279, "step": 784 }, { "epoch": 3.179746835443038, "grad_norm": 0.3368147024091749, "learning_rate": 1.393853735430398e-05, "loss": 0.2606, "step": 785 }, { "epoch": 3.1837974683544306, "grad_norm": 0.32889678853296306, "learning_rate": 1.3884472674819246e-05, "loss": 0.2691, "step": 786 }, { "epoch": 3.1878481012658226, "grad_norm": 0.3421466668509608, "learning_rate": 1.3830457248950864e-05, "loss": 0.2225, "step": 787 }, { "epoch": 3.191898734177215, "grad_norm": 0.32007229446495716, "learning_rate": 1.377649151173163e-05, "loss": 0.2689, "step": 788 }, { "epoch": 3.1959493670886077, "grad_norm": 0.3370701534854479, "learning_rate": 1.3722575897794181e-05, "loss": 0.2124, "step": 789 }, { "epoch": 3.2, "grad_norm": 0.3326791185783512, "learning_rate": 1.3668710841367472e-05, "loss": 0.2312, "step": 790 }, { "epoch": 3.2040506329113922, "grad_norm": 0.3140591478380161, "learning_rate": 1.361489677627324e-05, "loss": 0.2454, "step": 791 }, { "epoch": 3.2081012658227848, "grad_norm": 0.32788743630011824, "learning_rate": 1.3561134135922585e-05, "loss": 0.2554, "step": 792 }, { "epoch": 3.2121518987341773, "grad_norm": 0.3303877587838723, "learning_rate": 1.350742335331241e-05, "loss": 0.2522, "step": 793 }, { "epoch": 3.21620253164557, "grad_norm": 0.34349591458954015, "learning_rate": 1.345376486102198e-05, "loss": 0.2443, "step": 794 }, { "epoch": 3.220253164556962, "grad_norm": 0.3475433855403918, "learning_rate": 1.3400159091209414e-05, "loss": 0.2533, "step": 795 }, { "epoch": 3.2243037974683544, "grad_norm": 0.32978025213351375, "learning_rate": 1.3346606475608216e-05, "loss": 0.2549, "step": 796 }, { "epoch": 3.228354430379747, "grad_norm": 0.33108983606116793, "learning_rate": 1.3293107445523781e-05, "loss": 0.2525, "step": 797 }, { "epoch": 3.2324050632911394, "grad_norm": 0.33095538449600553, "learning_rate": 1.3239662431829949e-05, "loss": 0.2629, "step": 798 }, { "epoch": 3.2364556962025315, "grad_norm": 0.33035531718608313, "learning_rate": 1.3186271864965509e-05, "loss": 0.2584, "step": 799 }, { "epoch": 3.240506329113924, "grad_norm": 0.34792023800785943, "learning_rate": 1.3132936174930756e-05, "loss": 0.2397, "step": 800 }, { "epoch": 3.2445569620253165, "grad_norm": 0.33285929671796327, "learning_rate": 1.3079655791283995e-05, "loss": 0.2485, "step": 801 }, { "epoch": 3.248607594936709, "grad_norm": 0.3313167754837024, "learning_rate": 1.3026431143138108e-05, "loss": 0.2483, "step": 802 }, { "epoch": 3.252658227848101, "grad_norm": 0.3142251863248109, "learning_rate": 1.2973262659157114e-05, "loss": 0.2545, "step": 803 }, { "epoch": 3.2567088607594936, "grad_norm": 0.32779015196626443, "learning_rate": 1.2920150767552651e-05, "loss": 0.2385, "step": 804 }, { "epoch": 3.260759493670886, "grad_norm": 0.32506851647544355, "learning_rate": 1.2867095896080607e-05, "loss": 0.2437, "step": 805 }, { "epoch": 3.2648101265822786, "grad_norm": 0.34641260270181123, "learning_rate": 1.2814098472037612e-05, "loss": 0.2289, "step": 806 }, { "epoch": 3.2688607594936707, "grad_norm": 0.32282556727220124, "learning_rate": 1.276115892225764e-05, "loss": 0.2568, "step": 807 }, { "epoch": 3.2729113924050632, "grad_norm": 0.31769789295026635, "learning_rate": 1.2708277673108555e-05, "loss": 0.2593, "step": 808 }, { "epoch": 3.2769620253164558, "grad_norm": 0.3139359335418637, "learning_rate": 1.2655455150488649e-05, "loss": 0.2396, "step": 809 }, { "epoch": 3.2810126582278483, "grad_norm": 0.3219527303846401, "learning_rate": 1.2602691779823272e-05, "loss": 0.2402, "step": 810 }, { "epoch": 3.2850632911392403, "grad_norm": 0.3173293669200885, "learning_rate": 1.2549987986061355e-05, "loss": 0.2591, "step": 811 }, { "epoch": 3.289113924050633, "grad_norm": 0.3165762718416237, "learning_rate": 1.2497344193672005e-05, "loss": 0.2362, "step": 812 }, { "epoch": 3.2931645569620254, "grad_norm": 0.3367866363055433, "learning_rate": 1.2444760826641092e-05, "loss": 0.2247, "step": 813 }, { "epoch": 3.297215189873418, "grad_norm": 0.32474847962458453, "learning_rate": 1.2392238308467817e-05, "loss": 0.2473, "step": 814 }, { "epoch": 3.30126582278481, "grad_norm": 0.34021416159528145, "learning_rate": 1.2339777062161326e-05, "loss": 0.2429, "step": 815 }, { "epoch": 3.3053164556962025, "grad_norm": 0.3392648492372397, "learning_rate": 1.2287377510237293e-05, "loss": 0.2418, "step": 816 }, { "epoch": 3.309367088607595, "grad_norm": 0.34236435928530345, "learning_rate": 1.2235040074714488e-05, "loss": 0.236, "step": 817 }, { "epoch": 3.3134177215189875, "grad_norm": 0.3359411366730203, "learning_rate": 1.2182765177111434e-05, "loss": 0.2533, "step": 818 }, { "epoch": 3.3174683544303796, "grad_norm": 0.316660917903016, "learning_rate": 1.213055323844297e-05, "loss": 0.2246, "step": 819 }, { "epoch": 3.321518987341772, "grad_norm": 0.3447633994601158, "learning_rate": 1.2078404679216864e-05, "loss": 0.2448, "step": 820 }, { "epoch": 3.3255696202531646, "grad_norm": 0.3334131375997866, "learning_rate": 1.2026319919430458e-05, "loss": 0.2671, "step": 821 }, { "epoch": 3.329620253164557, "grad_norm": 0.3043408910379511, "learning_rate": 1.1974299378567227e-05, "loss": 0.2469, "step": 822 }, { "epoch": 3.333670886075949, "grad_norm": 0.3485297347924261, "learning_rate": 1.1922343475593462e-05, "loss": 0.2471, "step": 823 }, { "epoch": 3.3377215189873417, "grad_norm": 0.3496515679278481, "learning_rate": 1.187045262895488e-05, "loss": 0.2584, "step": 824 }, { "epoch": 3.3417721518987342, "grad_norm": 0.33777841299069644, "learning_rate": 1.1818627256573203e-05, "loss": 0.2458, "step": 825 }, { "epoch": 3.3458227848101267, "grad_norm": 0.3388430030570932, "learning_rate": 1.1766867775842864e-05, "loss": 0.2841, "step": 826 }, { "epoch": 3.349873417721519, "grad_norm": 0.3343265582211163, "learning_rate": 1.1715174603627615e-05, "loss": 0.2631, "step": 827 }, { "epoch": 3.3539240506329113, "grad_norm": 0.33963144370162196, "learning_rate": 1.1663548156257147e-05, "loss": 0.2394, "step": 828 }, { "epoch": 3.357974683544304, "grad_norm": 0.32939994339766104, "learning_rate": 1.161198884952377e-05, "loss": 0.256, "step": 829 }, { "epoch": 3.3620253164556964, "grad_norm": 0.3263111084705086, "learning_rate": 1.1560497098679056e-05, "loss": 0.2612, "step": 830 }, { "epoch": 3.3660759493670884, "grad_norm": 0.33382840078866466, "learning_rate": 1.1509073318430479e-05, "loss": 0.2465, "step": 831 }, { "epoch": 3.370126582278481, "grad_norm": 0.3012907224731824, "learning_rate": 1.1457717922938116e-05, "loss": 0.2785, "step": 832 }, { "epoch": 3.3741772151898735, "grad_norm": 0.3362159037222815, "learning_rate": 1.1406431325811233e-05, "loss": 0.2529, "step": 833 }, { "epoch": 3.378227848101266, "grad_norm": 0.33304279432150524, "learning_rate": 1.135521394010506e-05, "loss": 0.2401, "step": 834 }, { "epoch": 3.382278481012658, "grad_norm": 0.30820503805029403, "learning_rate": 1.1304066178317367e-05, "loss": 0.2475, "step": 835 }, { "epoch": 3.3863291139240506, "grad_norm": 0.31765611243091474, "learning_rate": 1.1252988452385199e-05, "loss": 0.2454, "step": 836 }, { "epoch": 3.390379746835443, "grad_norm": 0.34301183514320877, "learning_rate": 1.1201981173681536e-05, "loss": 0.2573, "step": 837 }, { "epoch": 3.3944303797468356, "grad_norm": 0.3322364541472241, "learning_rate": 1.1151044753011991e-05, "loss": 0.2567, "step": 838 }, { "epoch": 3.3984810126582277, "grad_norm": 0.3243299970725727, "learning_rate": 1.1100179600611491e-05, "loss": 0.2657, "step": 839 }, { "epoch": 3.40253164556962, "grad_norm": 0.32617231391982937, "learning_rate": 1.1049386126140985e-05, "loss": 0.2351, "step": 840 }, { "epoch": 3.4065822784810127, "grad_norm": 0.32365279405179853, "learning_rate": 1.0998664738684128e-05, "loss": 0.2553, "step": 841 }, { "epoch": 3.410632911392405, "grad_norm": 0.3315572236689557, "learning_rate": 1.0948015846744e-05, "loss": 0.2607, "step": 842 }, { "epoch": 3.4146835443037973, "grad_norm": 0.3165449649007485, "learning_rate": 1.0897439858239832e-05, "loss": 0.2295, "step": 843 }, { "epoch": 3.41873417721519, "grad_norm": 0.3261655099962764, "learning_rate": 1.0846937180503652e-05, "loss": 0.2383, "step": 844 }, { "epoch": 3.4227848101265823, "grad_norm": 0.37025280486743617, "learning_rate": 1.0796508220277117e-05, "loss": 0.245, "step": 845 }, { "epoch": 3.426835443037975, "grad_norm": 0.32429664923708024, "learning_rate": 1.0746153383708107e-05, "loss": 0.2525, "step": 846 }, { "epoch": 3.430886075949367, "grad_norm": 0.362914973205453, "learning_rate": 1.0695873076347579e-05, "loss": 0.2512, "step": 847 }, { "epoch": 3.4349367088607594, "grad_norm": 0.3289311355977165, "learning_rate": 1.0645667703146205e-05, "loss": 0.2605, "step": 848 }, { "epoch": 3.438987341772152, "grad_norm": 0.3152264391263699, "learning_rate": 1.0595537668451161e-05, "loss": 0.2554, "step": 849 }, { "epoch": 3.4430379746835444, "grad_norm": 0.3290099132645217, "learning_rate": 1.0545483376002854e-05, "loss": 0.2528, "step": 850 }, { "epoch": 3.4470886075949365, "grad_norm": 0.33416714948885995, "learning_rate": 1.0495505228931676e-05, "loss": 0.2525, "step": 851 }, { "epoch": 3.451139240506329, "grad_norm": 0.316489327189384, "learning_rate": 1.044560362975474e-05, "loss": 0.2326, "step": 852 }, { "epoch": 3.4551898734177215, "grad_norm": 0.3413285823634335, "learning_rate": 1.0395778980372695e-05, "loss": 0.2516, "step": 853 }, { "epoch": 3.459240506329114, "grad_norm": 0.3250802468562503, "learning_rate": 1.0346031682066381e-05, "loss": 0.2464, "step": 854 }, { "epoch": 3.463291139240506, "grad_norm": 0.32934045534705053, "learning_rate": 1.0296362135493724e-05, "loss": 0.2368, "step": 855 }, { "epoch": 3.4673417721518986, "grad_norm": 0.34086586825959375, "learning_rate": 1.0246770740686422e-05, "loss": 0.2506, "step": 856 }, { "epoch": 3.471392405063291, "grad_norm": 0.3272954298953066, "learning_rate": 1.0197257897046743e-05, "loss": 0.2645, "step": 857 }, { "epoch": 3.4754430379746837, "grad_norm": 0.34269394940560416, "learning_rate": 1.014782400334433e-05, "loss": 0.2511, "step": 858 }, { "epoch": 3.479493670886076, "grad_norm": 0.33606736882799987, "learning_rate": 1.009846945771296e-05, "loss": 0.2371, "step": 859 }, { "epoch": 3.4835443037974683, "grad_norm": 0.31424441870710573, "learning_rate": 1.0049194657647363e-05, "loss": 0.2394, "step": 860 }, { "epoch": 3.487594936708861, "grad_norm": 0.3360587977759128, "learning_rate": 1.0000000000000006e-05, "loss": 0.2325, "step": 861 }, { "epoch": 3.4916455696202533, "grad_norm": 0.3275406987113783, "learning_rate": 9.950885880977891e-06, "loss": 0.2177, "step": 862 }, { "epoch": 3.4956962025316454, "grad_norm": 0.33051935774994695, "learning_rate": 9.901852696139382e-06, "loss": 0.2337, "step": 863 }, { "epoch": 3.499746835443038, "grad_norm": 0.31879717879278685, "learning_rate": 9.852900840391027e-06, "loss": 0.2313, "step": 864 }, { "epoch": 3.5037974683544304, "grad_norm": 0.325337924750545, "learning_rate": 9.804030707984313e-06, "loss": 0.245, "step": 865 }, { "epoch": 3.507848101265823, "grad_norm": 0.3407643531196284, "learning_rate": 9.755242692512599e-06, "loss": 0.2198, "step": 866 }, { "epoch": 3.5118987341772154, "grad_norm": 0.32084530721980775, "learning_rate": 9.70653718690782e-06, "loss": 0.2462, "step": 867 }, { "epoch": 3.5159493670886075, "grad_norm": 0.3303261558593866, "learning_rate": 9.657914583437454e-06, "loss": 0.2577, "step": 868 }, { "epoch": 3.52, "grad_norm": 0.33701143669528816, "learning_rate": 9.609375273701246e-06, "loss": 0.2633, "step": 869 }, { "epoch": 3.5240506329113925, "grad_norm": 0.31508039243874353, "learning_rate": 9.560919648628133e-06, "loss": 0.2568, "step": 870 }, { "epoch": 3.5281012658227846, "grad_norm": 0.31716714182818834, "learning_rate": 9.512548098473047e-06, "loss": 0.2465, "step": 871 }, { "epoch": 3.532151898734177, "grad_norm": 0.3301984307765389, "learning_rate": 9.464261012813825e-06, "loss": 0.2447, "step": 872 }, { "epoch": 3.5362025316455696, "grad_norm": 0.3131062547645162, "learning_rate": 9.416058780547987e-06, "loss": 0.2381, "step": 873 }, { "epoch": 3.540253164556962, "grad_norm": 0.3199631885972641, "learning_rate": 9.367941789889714e-06, "loss": 0.2327, "step": 874 }, { "epoch": 3.5443037974683547, "grad_norm": 0.31889179258811057, "learning_rate": 9.319910428366607e-06, "loss": 0.2329, "step": 875 }, { "epoch": 3.5483544303797467, "grad_norm": 0.33290517232262956, "learning_rate": 9.271965082816667e-06, "loss": 0.2292, "step": 876 }, { "epoch": 3.5524050632911393, "grad_norm": 0.3236875722257051, "learning_rate": 9.224106139385111e-06, "loss": 0.2411, "step": 877 }, { "epoch": 3.5564556962025318, "grad_norm": 0.32445954815743033, "learning_rate": 9.176333983521291e-06, "loss": 0.259, "step": 878 }, { "epoch": 3.560506329113924, "grad_norm": 0.3259235147861101, "learning_rate": 9.12864899997558e-06, "loss": 0.2255, "step": 879 }, { "epoch": 3.5645569620253164, "grad_norm": 0.32479672091624906, "learning_rate": 9.08105157279628e-06, "loss": 0.2573, "step": 880 }, { "epoch": 3.568607594936709, "grad_norm": 0.32225244634753175, "learning_rate": 9.03354208532653e-06, "loss": 0.2208, "step": 881 }, { "epoch": 3.5726582278481014, "grad_norm": 0.34542529513528525, "learning_rate": 8.986120920201205e-06, "loss": 0.2555, "step": 882 }, { "epoch": 3.576708860759494, "grad_norm": 0.3171855104412481, "learning_rate": 8.938788459343852e-06, "loss": 0.228, "step": 883 }, { "epoch": 3.580759493670886, "grad_norm": 0.3237536083676936, "learning_rate": 8.8915450839636e-06, "loss": 0.2349, "step": 884 }, { "epoch": 3.5848101265822785, "grad_norm": 0.32810385684129223, "learning_rate": 8.844391174552116e-06, "loss": 0.2558, "step": 885 }, { "epoch": 3.588860759493671, "grad_norm": 0.3259850465396525, "learning_rate": 8.797327110880479e-06, "loss": 0.2377, "step": 886 }, { "epoch": 3.592911392405063, "grad_norm": 0.3144510027122409, "learning_rate": 8.750353271996206e-06, "loss": 0.2378, "step": 887 }, { "epoch": 3.5969620253164556, "grad_norm": 0.31419279754527085, "learning_rate": 8.703470036220132e-06, "loss": 0.244, "step": 888 }, { "epoch": 3.601012658227848, "grad_norm": 0.32745895012685106, "learning_rate": 8.656677781143394e-06, "loss": 0.2598, "step": 889 }, { "epoch": 3.6050632911392406, "grad_norm": 0.32268068349040896, "learning_rate": 8.609976883624377e-06, "loss": 0.2437, "step": 890 }, { "epoch": 3.609113924050633, "grad_norm": 0.3192539229656004, "learning_rate": 8.563367719785698e-06, "loss": 0.25, "step": 891 }, { "epoch": 3.613164556962025, "grad_norm": 0.30083106222362105, "learning_rate": 8.516850665011138e-06, "loss": 0.2447, "step": 892 }, { "epoch": 3.6172151898734177, "grad_norm": 0.31160273463640004, "learning_rate": 8.47042609394269e-06, "loss": 0.2613, "step": 893 }, { "epoch": 3.6212658227848102, "grad_norm": 0.3142882069519253, "learning_rate": 8.424094380477432e-06, "loss": 0.2257, "step": 894 }, { "epoch": 3.6253164556962023, "grad_norm": 0.3473725274691121, "learning_rate": 8.37785589776465e-06, "loss": 0.2484, "step": 895 }, { "epoch": 3.629367088607595, "grad_norm": 0.324687543842604, "learning_rate": 8.331711018202694e-06, "loss": 0.24, "step": 896 }, { "epoch": 3.6334177215189873, "grad_norm": 0.32794330091169954, "learning_rate": 8.285660113436104e-06, "loss": 0.2247, "step": 897 }, { "epoch": 3.63746835443038, "grad_norm": 0.31627474783655, "learning_rate": 8.239703554352527e-06, "loss": 0.2459, "step": 898 }, { "epoch": 3.6415189873417724, "grad_norm": 0.35581783952918783, "learning_rate": 8.193841711079775e-06, "loss": 0.266, "step": 899 }, { "epoch": 3.6455696202531644, "grad_norm": 0.3073150633114822, "learning_rate": 8.148074952982828e-06, "loss": 0.2548, "step": 900 }, { "epoch": 3.649620253164557, "grad_norm": 0.3223743834115442, "learning_rate": 8.102403648660859e-06, "loss": 0.2627, "step": 901 }, { "epoch": 3.6536708860759495, "grad_norm": 0.35006657704903715, "learning_rate": 8.056828165944282e-06, "loss": 0.2586, "step": 902 }, { "epoch": 3.6577215189873415, "grad_norm": 0.32725328586267505, "learning_rate": 8.011348871891762e-06, "loss": 0.2399, "step": 903 }, { "epoch": 3.661772151898734, "grad_norm": 0.3245376799702508, "learning_rate": 7.965966132787287e-06, "loss": 0.2619, "step": 904 }, { "epoch": 3.6658227848101266, "grad_norm": 0.31184741206268446, "learning_rate": 7.920680314137189e-06, "loss": 0.2429, "step": 905 }, { "epoch": 3.669873417721519, "grad_norm": 0.3297366040045418, "learning_rate": 7.875491780667246e-06, "loss": 0.2466, "step": 906 }, { "epoch": 3.6739240506329116, "grad_norm": 0.3180505408841936, "learning_rate": 7.830400896319667e-06, "loss": 0.2442, "step": 907 }, { "epoch": 3.6779746835443037, "grad_norm": 0.3113282922547884, "learning_rate": 7.785408024250259e-06, "loss": 0.2396, "step": 908 }, { "epoch": 3.682025316455696, "grad_norm": 0.3158575915934227, "learning_rate": 7.74051352682542e-06, "loss": 0.2296, "step": 909 }, { "epoch": 3.6860759493670887, "grad_norm": 0.3192848722960496, "learning_rate": 7.695717765619257e-06, "loss": 0.2176, "step": 910 }, { "epoch": 3.690126582278481, "grad_norm": 0.31732597125368706, "learning_rate": 7.651021101410673e-06, "loss": 0.2221, "step": 911 }, { "epoch": 3.6941772151898733, "grad_norm": 0.31939131680463134, "learning_rate": 7.606423894180464e-06, "loss": 0.2302, "step": 912 }, { "epoch": 3.698227848101266, "grad_norm": 0.3157407509932589, "learning_rate": 7.56192650310839e-06, "loss": 0.255, "step": 913 }, { "epoch": 3.7022784810126583, "grad_norm": 0.3164429044871905, "learning_rate": 7.517529286570349e-06, "loss": 0.2314, "step": 914 }, { "epoch": 3.706329113924051, "grad_norm": 0.31509366290094887, "learning_rate": 7.473232602135387e-06, "loss": 0.2541, "step": 915 }, { "epoch": 3.710379746835443, "grad_norm": 0.30849230612455386, "learning_rate": 7.429036806562935e-06, "loss": 0.2464, "step": 916 }, { "epoch": 3.7144303797468354, "grad_norm": 0.33691557754692547, "learning_rate": 7.3849422557998455e-06, "loss": 0.2422, "step": 917 }, { "epoch": 3.718481012658228, "grad_norm": 0.32972727698344684, "learning_rate": 7.340949304977567e-06, "loss": 0.2371, "step": 918 }, { "epoch": 3.72253164556962, "grad_norm": 0.3170110750156512, "learning_rate": 7.297058308409282e-06, "loss": 0.2194, "step": 919 }, { "epoch": 3.7265822784810125, "grad_norm": 0.3026975111502901, "learning_rate": 7.25326961958704e-06, "loss": 0.2564, "step": 920 }, { "epoch": 3.730632911392405, "grad_norm": 0.33061879118185444, "learning_rate": 7.209583591178921e-06, "loss": 0.2566, "step": 921 }, { "epoch": 3.7346835443037976, "grad_norm": 0.3412130900113667, "learning_rate": 7.1660005750261925e-06, "loss": 0.2374, "step": 922 }, { "epoch": 3.73873417721519, "grad_norm": 0.3317091002284077, "learning_rate": 7.1225209221404765e-06, "loss": 0.2476, "step": 923 }, { "epoch": 3.742784810126582, "grad_norm": 0.30896745044242235, "learning_rate": 7.079144982700909e-06, "loss": 0.251, "step": 924 }, { "epoch": 3.7468354430379747, "grad_norm": 0.3163678968479515, "learning_rate": 7.0358731060513695e-06, "loss": 0.2416, "step": 925 }, { "epoch": 3.750886075949367, "grad_norm": 0.301355034065856, "learning_rate": 6.99270564069757e-06, "loss": 0.2484, "step": 926 }, { "epoch": 3.7549367088607593, "grad_norm": 0.30301528728281407, "learning_rate": 6.949642934304375e-06, "loss": 0.244, "step": 927 }, { "epoch": 3.7589873417721518, "grad_norm": 0.3202267184866359, "learning_rate": 6.906685333692871e-06, "loss": 0.2363, "step": 928 }, { "epoch": 3.7630379746835443, "grad_norm": 0.3345567724023094, "learning_rate": 6.86383318483769e-06, "loss": 0.253, "step": 929 }, { "epoch": 3.767088607594937, "grad_norm": 0.3242475807722324, "learning_rate": 6.821086832864139e-06, "loss": 0.2395, "step": 930 }, { "epoch": 3.7711392405063293, "grad_norm": 0.3064818934593835, "learning_rate": 6.77844662204546e-06, "loss": 0.2416, "step": 931 }, { "epoch": 3.7751898734177214, "grad_norm": 0.3167362195199549, "learning_rate": 6.7359128958000455e-06, "loss": 0.2295, "step": 932 }, { "epoch": 3.779240506329114, "grad_norm": 0.31675169841473133, "learning_rate": 6.693485996688695e-06, "loss": 0.2343, "step": 933 }, { "epoch": 3.7832911392405064, "grad_norm": 0.31434067908622354, "learning_rate": 6.651166266411801e-06, "loss": 0.2334, "step": 934 }, { "epoch": 3.7873417721518985, "grad_norm": 0.31752479487632806, "learning_rate": 6.6089540458066725e-06, "loss": 0.2474, "step": 935 }, { "epoch": 3.791392405063291, "grad_norm": 0.3166658092354889, "learning_rate": 6.566849674844711e-06, "loss": 0.2339, "step": 936 }, { "epoch": 3.7954430379746835, "grad_norm": 0.31048307232323685, "learning_rate": 6.524853492628747e-06, "loss": 0.2529, "step": 937 }, { "epoch": 3.799493670886076, "grad_norm": 0.30901140828572604, "learning_rate": 6.4829658373902536e-06, "loss": 0.2735, "step": 938 }, { "epoch": 3.8035443037974686, "grad_norm": 0.31004760922194097, "learning_rate": 6.441187046486648e-06, "loss": 0.2421, "step": 939 }, { "epoch": 3.8075949367088606, "grad_norm": 0.3029769136604379, "learning_rate": 6.399517456398567e-06, "loss": 0.235, "step": 940 }, { "epoch": 3.811645569620253, "grad_norm": 0.3122734146149477, "learning_rate": 6.357957402727164e-06, "loss": 0.2361, "step": 941 }, { "epoch": 3.8156962025316457, "grad_norm": 0.32418138639612093, "learning_rate": 6.316507220191395e-06, "loss": 0.2263, "step": 942 }, { "epoch": 3.8197468354430377, "grad_norm": 0.31246206309617663, "learning_rate": 6.275167242625331e-06, "loss": 0.2695, "step": 943 }, { "epoch": 3.8237974683544302, "grad_norm": 0.29286133661831876, "learning_rate": 6.233937802975471e-06, "loss": 0.2504, "step": 944 }, { "epoch": 3.8278481012658228, "grad_norm": 0.3135541662543286, "learning_rate": 6.192819233298046e-06, "loss": 0.234, "step": 945 }, { "epoch": 3.8318987341772153, "grad_norm": 0.31921120235605915, "learning_rate": 6.151811864756383e-06, "loss": 0.2524, "step": 946 }, { "epoch": 3.835949367088608, "grad_norm": 0.3112927864136072, "learning_rate": 6.1109160276181655e-06, "loss": 0.2068, "step": 947 }, { "epoch": 3.84, "grad_norm": 0.324706442369522, "learning_rate": 6.070132051252868e-06, "loss": 0.25, "step": 948 }, { "epoch": 3.8440506329113924, "grad_norm": 0.3360753389839713, "learning_rate": 6.0294602641290034e-06, "loss": 0.245, "step": 949 }, { "epoch": 3.848101265822785, "grad_norm": 0.3119670308394329, "learning_rate": 5.988900993811575e-06, "loss": 0.2596, "step": 950 }, { "epoch": 3.852151898734177, "grad_norm": 0.3110924712820628, "learning_rate": 5.948454566959363e-06, "loss": 0.2357, "step": 951 }, { "epoch": 3.8562025316455695, "grad_norm": 0.30487620843917185, "learning_rate": 5.908121309322328e-06, "loss": 0.2306, "step": 952 }, { "epoch": 3.860253164556962, "grad_norm": 0.3140647442546888, "learning_rate": 5.867901545738976e-06, "loss": 0.2383, "step": 953 }, { "epoch": 3.8643037974683545, "grad_norm": 0.3228168955763839, "learning_rate": 5.827795600133774e-06, "loss": 0.243, "step": 954 }, { "epoch": 3.868354430379747, "grad_norm": 0.2976157721758022, "learning_rate": 5.787803795514466e-06, "loss": 0.2568, "step": 955 }, { "epoch": 3.872405063291139, "grad_norm": 0.3274947433964772, "learning_rate": 5.747926453969576e-06, "loss": 0.2286, "step": 956 }, { "epoch": 3.8764556962025316, "grad_norm": 0.31308287134226026, "learning_rate": 5.708163896665708e-06, "loss": 0.2574, "step": 957 }, { "epoch": 3.880506329113924, "grad_norm": 0.3091402652448197, "learning_rate": 5.668516443845047e-06, "loss": 0.2423, "step": 958 }, { "epoch": 3.884556962025316, "grad_norm": 0.3065004039478424, "learning_rate": 5.6289844148227225e-06, "loss": 0.2464, "step": 959 }, { "epoch": 3.8886075949367087, "grad_norm": 0.30704648021091846, "learning_rate": 5.5895681279842615e-06, "loss": 0.2503, "step": 960 }, { "epoch": 3.8926582278481012, "grad_norm": 0.315939429835143, "learning_rate": 5.550267900783019e-06, "loss": 0.256, "step": 961 }, { "epoch": 3.8967088607594937, "grad_norm": 0.3174391961277774, "learning_rate": 5.511084049737623e-06, "loss": 0.2561, "step": 962 }, { "epoch": 3.9007594936708863, "grad_norm": 0.30081920013364144, "learning_rate": 5.4720168904294215e-06, "loss": 0.2565, "step": 963 }, { "epoch": 3.9048101265822783, "grad_norm": 0.30287155306179453, "learning_rate": 5.433066737499948e-06, "loss": 0.2357, "step": 964 }, { "epoch": 3.908860759493671, "grad_norm": 0.3164382027573777, "learning_rate": 5.394233904648376e-06, "loss": 0.2529, "step": 965 }, { "epoch": 3.9129113924050634, "grad_norm": 0.3081647669568605, "learning_rate": 5.355518704628997e-06, "loss": 0.2343, "step": 966 }, { "epoch": 3.9169620253164554, "grad_norm": 0.3058497107937708, "learning_rate": 5.316921449248731e-06, "loss": 0.2676, "step": 967 }, { "epoch": 3.921012658227848, "grad_norm": 0.3011588763203766, "learning_rate": 5.278442449364538e-06, "loss": 0.2478, "step": 968 }, { "epoch": 3.9250632911392405, "grad_norm": 0.30099733381541477, "learning_rate": 5.240082014881016e-06, "loss": 0.2508, "step": 969 }, { "epoch": 3.929113924050633, "grad_norm": 0.3101341838605249, "learning_rate": 5.201840454747822e-06, "loss": 0.2372, "step": 970 }, { "epoch": 3.9331645569620255, "grad_norm": 0.29545637240704836, "learning_rate": 5.163718076957223e-06, "loss": 0.2248, "step": 971 }, { "epoch": 3.9372151898734176, "grad_norm": 0.3104786522761715, "learning_rate": 5.125715188541609e-06, "loss": 0.2143, "step": 972 }, { "epoch": 3.94126582278481, "grad_norm": 0.2906271491437359, "learning_rate": 5.087832095571021e-06, "loss": 0.2552, "step": 973 }, { "epoch": 3.9453164556962026, "grad_norm": 0.32364495715811137, "learning_rate": 5.0500691031506766e-06, "loss": 0.2565, "step": 974 }, { "epoch": 3.9493670886075947, "grad_norm": 0.30384068689987337, "learning_rate": 5.01242651541854e-06, "loss": 0.2309, "step": 975 }, { "epoch": 3.953417721518987, "grad_norm": 0.30754833748773924, "learning_rate": 4.974904635542815e-06, "loss": 0.234, "step": 976 }, { "epoch": 3.9574683544303797, "grad_norm": 0.3166074561107965, "learning_rate": 4.937503765719582e-06, "loss": 0.2341, "step": 977 }, { "epoch": 3.961518987341772, "grad_norm": 0.3137037030968802, "learning_rate": 4.900224207170299e-06, "loss": 0.2136, "step": 978 }, { "epoch": 3.9655696202531647, "grad_norm": 0.30181633124693624, "learning_rate": 4.8630662601394065e-06, "loss": 0.2355, "step": 979 }, { "epoch": 3.969620253164557, "grad_norm": 0.3019481857019933, "learning_rate": 4.8260302238918995e-06, "loss": 0.2749, "step": 980 }, { "epoch": 3.9736708860759493, "grad_norm": 0.3101448607231526, "learning_rate": 4.789116396710924e-06, "loss": 0.2398, "step": 981 }, { "epoch": 3.977721518987342, "grad_norm": 0.2914659477534417, "learning_rate": 4.752325075895368e-06, "loss": 0.2599, "step": 982 }, { "epoch": 3.981772151898734, "grad_norm": 0.3138191515770842, "learning_rate": 4.715656557757473e-06, "loss": 0.2337, "step": 983 }, { "epoch": 3.9858227848101264, "grad_norm": 0.3072310812454874, "learning_rate": 4.679111137620442e-06, "loss": 0.2612, "step": 984 }, { "epoch": 3.989873417721519, "grad_norm": 0.31291324593474173, "learning_rate": 4.6426891098160585e-06, "loss": 0.2388, "step": 985 }, { "epoch": 3.9939240506329114, "grad_norm": 0.3043405388766245, "learning_rate": 4.6063907676823474e-06, "loss": 0.2405, "step": 986 }, { "epoch": 3.997974683544304, "grad_norm": 0.305208388824953, "learning_rate": 4.570216403561141e-06, "loss": 0.2218, "step": 987 }, { "epoch": 4.01620253164557, "grad_norm": 0.4544430305418597, "learning_rate": 4.534166308795815e-06, "loss": 0.2026, "step": 988 }, { "epoch": 4.020253164556962, "grad_norm": 0.3946813418425237, "learning_rate": 4.498240773728859e-06, "loss": 0.2021, "step": 989 }, { "epoch": 4.024303797468354, "grad_norm": 0.32160552796417163, "learning_rate": 4.462440087699609e-06, "loss": 0.1915, "step": 990 }, { "epoch": 4.028354430379747, "grad_norm": 0.289764526754448, "learning_rate": 4.426764539041861e-06, "loss": 0.1772, "step": 991 }, { "epoch": 4.032405063291139, "grad_norm": 0.4026623824644641, "learning_rate": 4.391214415081582e-06, "loss": 0.1973, "step": 992 }, { "epoch": 4.036455696202531, "grad_norm": 0.43831303279141715, "learning_rate": 4.355790002134579e-06, "loss": 0.1794, "step": 993 }, { "epoch": 4.040506329113924, "grad_norm": 0.3996350759575781, "learning_rate": 4.320491585504207e-06, "loss": 0.1889, "step": 994 }, { "epoch": 4.044556962025316, "grad_norm": 0.34425357300964066, "learning_rate": 4.2853194494790615e-06, "loss": 0.1946, "step": 995 }, { "epoch": 4.048607594936709, "grad_norm": 0.3156143124946476, "learning_rate": 4.250273877330691e-06, "loss": 0.1898, "step": 996 }, { "epoch": 4.052658227848101, "grad_norm": 0.3536760467500462, "learning_rate": 4.215355151311313e-06, "loss": 0.1792, "step": 997 }, { "epoch": 4.056708860759493, "grad_norm": 0.3420137927068648, "learning_rate": 4.180563552651542e-06, "loss": 0.1652, "step": 998 }, { "epoch": 4.060759493670886, "grad_norm": 0.346601422640507, "learning_rate": 4.145899361558147e-06, "loss": 0.1779, "step": 999 }, { "epoch": 4.0648101265822785, "grad_norm": 0.26082159858408727, "learning_rate": 4.111362857211738e-06, "loss": 0.1399, "step": 1000 }, { "epoch": 4.0688607594936705, "grad_norm": 0.28030775812670333, "learning_rate": 4.076954317764592e-06, "loss": 0.1849, "step": 1001 }, { "epoch": 4.0729113924050635, "grad_norm": 0.2915960647839899, "learning_rate": 4.042674020338335e-06, "loss": 0.1986, "step": 1002 }, { "epoch": 4.076962025316456, "grad_norm": 0.2942639233441812, "learning_rate": 4.0085222410217835e-06, "loss": 0.1637, "step": 1003 }, { "epoch": 4.0810126582278485, "grad_norm": 0.34101127952367577, "learning_rate": 3.974499254868674e-06, "loss": 0.1909, "step": 1004 }, { "epoch": 4.085063291139241, "grad_norm": 0.2645895113108063, "learning_rate": 3.940605335895451e-06, "loss": 0.1743, "step": 1005 }, { "epoch": 4.089113924050633, "grad_norm": 0.2651679890277567, "learning_rate": 3.90684075707908e-06, "loss": 0.166, "step": 1006 }, { "epoch": 4.093164556962026, "grad_norm": 0.2846465590203964, "learning_rate": 3.8732057903548505e-06, "loss": 0.191, "step": 1007 }, { "epoch": 4.097215189873418, "grad_norm": 0.28489924793137017, "learning_rate": 3.8397007066141375e-06, "loss": 0.1821, "step": 1008 }, { "epoch": 4.10126582278481, "grad_norm": 0.2791339836956419, "learning_rate": 3.806325775702304e-06, "loss": 0.1773, "step": 1009 }, { "epoch": 4.105316455696203, "grad_norm": 0.274470262243458, "learning_rate": 3.773081266416434e-06, "loss": 0.1817, "step": 1010 }, { "epoch": 4.109367088607595, "grad_norm": 0.2673142685705358, "learning_rate": 3.739967446503245e-06, "loss": 0.1881, "step": 1011 }, { "epoch": 4.113417721518988, "grad_norm": 0.2758907460336557, "learning_rate": 3.706984582656894e-06, "loss": 0.198, "step": 1012 }, { "epoch": 4.11746835443038, "grad_norm": 0.3123527055068546, "learning_rate": 3.6741329405168237e-06, "loss": 0.1874, "step": 1013 }, { "epoch": 4.121518987341772, "grad_norm": 0.27809572903637314, "learning_rate": 3.641412784665648e-06, "loss": 0.2054, "step": 1014 }, { "epoch": 4.125569620253165, "grad_norm": 0.2611647205671966, "learning_rate": 3.608824378627005e-06, "loss": 0.1751, "step": 1015 }, { "epoch": 4.129620253164557, "grad_norm": 0.27743307017106494, "learning_rate": 3.5763679848634337e-06, "loss": 0.1985, "step": 1016 }, { "epoch": 4.133670886075949, "grad_norm": 0.26432508910600966, "learning_rate": 3.544043864774269e-06, "loss": 0.192, "step": 1017 }, { "epoch": 4.137721518987342, "grad_norm": 0.28856287842699146, "learning_rate": 3.5118522786935282e-06, "loss": 0.1894, "step": 1018 }, { "epoch": 4.141772151898734, "grad_norm": 0.26942132824906156, "learning_rate": 3.479793485887819e-06, "loss": 0.1834, "step": 1019 }, { "epoch": 4.145822784810127, "grad_norm": 0.25074614540667206, "learning_rate": 3.4478677445542653e-06, "loss": 0.1902, "step": 1020 }, { "epoch": 4.149873417721519, "grad_norm": 0.2529476202994876, "learning_rate": 3.4160753118183767e-06, "loss": 0.212, "step": 1021 }, { "epoch": 4.153924050632911, "grad_norm": 0.2531141699244196, "learning_rate": 3.3844164437320527e-06, "loss": 0.1944, "step": 1022 }, { "epoch": 4.157974683544304, "grad_norm": 0.2505244293906678, "learning_rate": 3.3528913952714558e-06, "loss": 0.1778, "step": 1023 }, { "epoch": 4.162025316455696, "grad_norm": 0.2635044074795311, "learning_rate": 3.321500420335e-06, "loss": 0.1932, "step": 1024 }, { "epoch": 4.166075949367088, "grad_norm": 0.26006340027449526, "learning_rate": 3.290243771741275e-06, "loss": 0.1894, "step": 1025 }, { "epoch": 4.170126582278481, "grad_norm": 0.2508891880070159, "learning_rate": 3.2591217012270325e-06, "loss": 0.1601, "step": 1026 }, { "epoch": 4.174177215189873, "grad_norm": 0.26186277083184717, "learning_rate": 3.228134459445149e-06, "loss": 0.2, "step": 1027 }, { "epoch": 4.178227848101266, "grad_norm": 0.26319518447112034, "learning_rate": 3.1972822959626205e-06, "loss": 0.1946, "step": 1028 }, { "epoch": 4.182278481012658, "grad_norm": 0.2561836068674533, "learning_rate": 3.166565459258513e-06, "loss": 0.1991, "step": 1029 }, { "epoch": 4.18632911392405, "grad_norm": 0.25365286271173654, "learning_rate": 3.1359841967220193e-06, "loss": 0.1748, "step": 1030 }, { "epoch": 4.190379746835443, "grad_norm": 0.2574993743649099, "learning_rate": 3.105538754650419e-06, "loss": 0.2036, "step": 1031 }, { "epoch": 4.194430379746835, "grad_norm": 0.32053424908768163, "learning_rate": 3.07522937824712e-06, "loss": 0.1938, "step": 1032 }, { "epoch": 4.1984810126582275, "grad_norm": 0.2541809864993721, "learning_rate": 3.0450563116196697e-06, "loss": 0.1828, "step": 1033 }, { "epoch": 4.2025316455696204, "grad_norm": 0.2632077322451832, "learning_rate": 3.0150197977778008e-06, "loss": 0.1934, "step": 1034 }, { "epoch": 4.2065822784810125, "grad_norm": 0.26101281778353175, "learning_rate": 2.985120078631465e-06, "loss": 0.1941, "step": 1035 }, { "epoch": 4.2106329113924055, "grad_norm": 0.252119864229832, "learning_rate": 2.9553573949888893e-06, "loss": 0.1653, "step": 1036 }, { "epoch": 4.2146835443037975, "grad_norm": 0.24800660155790208, "learning_rate": 2.9257319865546384e-06, "loss": 0.1925, "step": 1037 }, { "epoch": 4.21873417721519, "grad_norm": 0.2553613513898262, "learning_rate": 2.896244091927678e-06, "loss": 0.1702, "step": 1038 }, { "epoch": 4.222784810126583, "grad_norm": 0.2569970126888527, "learning_rate": 2.8668939485994584e-06, "loss": 0.2091, "step": 1039 }, { "epoch": 4.226835443037975, "grad_norm": 0.2502075325354623, "learning_rate": 2.837681792951994e-06, "loss": 0.1985, "step": 1040 }, { "epoch": 4.230886075949367, "grad_norm": 0.25593655927135783, "learning_rate": 2.808607860255981e-06, "loss": 0.1556, "step": 1041 }, { "epoch": 4.23493670886076, "grad_norm": 0.2537355982407617, "learning_rate": 2.7796723846688634e-06, "loss": 0.1844, "step": 1042 }, { "epoch": 4.238987341772152, "grad_norm": 0.25562795318297593, "learning_rate": 2.7508755992329937e-06, "loss": 0.1919, "step": 1043 }, { "epoch": 4.243037974683545, "grad_norm": 0.25168797416608457, "learning_rate": 2.722217735873718e-06, "loss": 0.205, "step": 1044 }, { "epoch": 4.247088607594937, "grad_norm": 0.26092755182349586, "learning_rate": 2.6936990253975315e-06, "loss": 0.1717, "step": 1045 }, { "epoch": 4.251139240506329, "grad_norm": 0.25573369165192844, "learning_rate": 2.665319697490205e-06, "loss": 0.1638, "step": 1046 }, { "epoch": 4.255189873417722, "grad_norm": 0.2465267079357583, "learning_rate": 2.637079980714945e-06, "loss": 0.1736, "step": 1047 }, { "epoch": 4.259240506329114, "grad_norm": 0.24614844128021915, "learning_rate": 2.6089801025105453e-06, "loss": 0.1805, "step": 1048 }, { "epoch": 4.263291139240506, "grad_norm": 0.25470311850987126, "learning_rate": 2.581020289189571e-06, "loss": 0.1834, "step": 1049 }, { "epoch": 4.267341772151899, "grad_norm": 0.25481445028873284, "learning_rate": 2.553200765936501e-06, "loss": 0.2028, "step": 1050 }, { "epoch": 4.271392405063291, "grad_norm": 0.25786560596126634, "learning_rate": 2.525521756805962e-06, "loss": 0.1835, "step": 1051 }, { "epoch": 4.275443037974684, "grad_norm": 0.26342423150701644, "learning_rate": 2.497983484720885e-06, "loss": 0.1557, "step": 1052 }, { "epoch": 4.279493670886076, "grad_norm": 0.24321420032481653, "learning_rate": 2.470586171470728e-06, "loss": 0.1505, "step": 1053 }, { "epoch": 4.283544303797468, "grad_norm": 0.24481266145739186, "learning_rate": 2.4433300377096836e-06, "loss": 0.1879, "step": 1054 }, { "epoch": 4.287594936708861, "grad_norm": 0.2529326550666124, "learning_rate": 2.4162153029549073e-06, "loss": 0.1772, "step": 1055 }, { "epoch": 4.291645569620253, "grad_norm": 0.24421865207225718, "learning_rate": 2.3892421855847458e-06, "loss": 0.1889, "step": 1056 }, { "epoch": 4.295696202531645, "grad_norm": 0.26154752105218115, "learning_rate": 2.362410902836978e-06, "loss": 0.2053, "step": 1057 }, { "epoch": 4.299746835443038, "grad_norm": 0.2648334992575525, "learning_rate": 2.3357216708070653e-06, "loss": 0.1877, "step": 1058 }, { "epoch": 4.30379746835443, "grad_norm": 0.23916309002500608, "learning_rate": 2.309174704446411e-06, "loss": 0.1994, "step": 1059 }, { "epoch": 4.307848101265823, "grad_norm": 0.25771608294396603, "learning_rate": 2.2827702175606437e-06, "loss": 0.1796, "step": 1060 }, { "epoch": 4.311898734177215, "grad_norm": 0.2552465226932246, "learning_rate": 2.256508422807855e-06, "loss": 0.1896, "step": 1061 }, { "epoch": 4.315949367088607, "grad_norm": 0.26046289506544945, "learning_rate": 2.230389531696946e-06, "loss": 0.1923, "step": 1062 }, { "epoch": 4.32, "grad_norm": 0.2544150766940966, "learning_rate": 2.204413754585857e-06, "loss": 0.1777, "step": 1063 }, { "epoch": 4.324050632911392, "grad_norm": 0.2560990593008856, "learning_rate": 2.1785813006799406e-06, "loss": 0.1649, "step": 1064 }, { "epoch": 4.328101265822784, "grad_norm": 0.25970014031149447, "learning_rate": 2.1528923780302224e-06, "loss": 0.1568, "step": 1065 }, { "epoch": 4.332151898734177, "grad_norm": 0.2567082018630218, "learning_rate": 2.127347193531757e-06, "loss": 0.1769, "step": 1066 }, { "epoch": 4.3362025316455695, "grad_norm": 0.26452589828138373, "learning_rate": 2.101945952921942e-06, "loss": 0.1992, "step": 1067 }, { "epoch": 4.340253164556962, "grad_norm": 0.24857793796593425, "learning_rate": 2.0766888607788906e-06, "loss": 0.1819, "step": 1068 }, { "epoch": 4.3443037974683545, "grad_norm": 0.24675498688694508, "learning_rate": 2.0515761205197337e-06, "loss": 0.1691, "step": 1069 }, { "epoch": 4.348354430379747, "grad_norm": 0.256463671424538, "learning_rate": 2.0266079343990453e-06, "loss": 0.2088, "step": 1070 }, { "epoch": 4.3524050632911395, "grad_norm": 0.25922623917952464, "learning_rate": 2.0017845035071494e-06, "loss": 0.2014, "step": 1071 }, { "epoch": 4.356455696202532, "grad_norm": 0.24838065849298144, "learning_rate": 1.9771060277685537e-06, "loss": 0.1844, "step": 1072 }, { "epoch": 4.360506329113924, "grad_norm": 0.24939445386292858, "learning_rate": 1.95257270594031e-06, "loss": 0.1832, "step": 1073 }, { "epoch": 4.364556962025317, "grad_norm": 0.2524068438283658, "learning_rate": 1.9281847356104188e-06, "loss": 0.2063, "step": 1074 }, { "epoch": 4.368607594936709, "grad_norm": 0.25458117561497323, "learning_rate": 1.9039423131962365e-06, "loss": 0.1775, "step": 1075 }, { "epoch": 4.372658227848102, "grad_norm": 0.25067976644291745, "learning_rate": 1.8798456339429027e-06, "loss": 0.1639, "step": 1076 }, { "epoch": 4.376708860759494, "grad_norm": 0.2589811808827528, "learning_rate": 1.8558948919217612e-06, "loss": 0.1773, "step": 1077 }, { "epoch": 4.380759493670886, "grad_norm": 0.25872969926051365, "learning_rate": 1.8320902800287954e-06, "loss": 0.1962, "step": 1078 }, { "epoch": 4.384810126582279, "grad_norm": 0.24598775436169146, "learning_rate": 1.8084319899830726e-06, "loss": 0.1634, "step": 1079 }, { "epoch": 4.388860759493671, "grad_norm": 0.2539112854368541, "learning_rate": 1.7849202123252097e-06, "loss": 0.1805, "step": 1080 }, { "epoch": 4.392911392405063, "grad_norm": 0.25545107969719233, "learning_rate": 1.7615551364158401e-06, "loss": 0.1825, "step": 1081 }, { "epoch": 4.396962025316456, "grad_norm": 0.2946980186461365, "learning_rate": 1.738336950434061e-06, "loss": 0.1707, "step": 1082 }, { "epoch": 4.401012658227848, "grad_norm": 0.2626123797411181, "learning_rate": 1.715265841375957e-06, "loss": 0.1913, "step": 1083 }, { "epoch": 4.405063291139241, "grad_norm": 0.2631556879674905, "learning_rate": 1.6923419950530684e-06, "loss": 0.1893, "step": 1084 }, { "epoch": 4.409113924050633, "grad_norm": 0.24915189531960352, "learning_rate": 1.6695655960909008e-06, "loss": 0.1746, "step": 1085 }, { "epoch": 4.413164556962025, "grad_norm": 0.26029962370078746, "learning_rate": 1.646936827927441e-06, "loss": 0.1944, "step": 1086 }, { "epoch": 4.417215189873418, "grad_norm": 0.25242678258709134, "learning_rate": 1.6244558728116766e-06, "loss": 0.1665, "step": 1087 }, { "epoch": 4.42126582278481, "grad_norm": 0.243408007335565, "learning_rate": 1.6021229118021265e-06, "loss": 0.1803, "step": 1088 }, { "epoch": 4.425316455696202, "grad_norm": 0.25531696179789826, "learning_rate": 1.5799381247653967e-06, "loss": 0.198, "step": 1089 }, { "epoch": 4.429367088607595, "grad_norm": 0.2580625342293257, "learning_rate": 1.5579016903747013e-06, "loss": 0.1866, "step": 1090 }, { "epoch": 4.433417721518987, "grad_norm": 0.250425243567716, "learning_rate": 1.5360137861084656e-06, "loss": 0.1899, "step": 1091 }, { "epoch": 4.43746835443038, "grad_norm": 0.25290477283246743, "learning_rate": 1.5142745882488475e-06, "loss": 0.2123, "step": 1092 }, { "epoch": 4.441518987341772, "grad_norm": 0.2691447584866871, "learning_rate": 1.4926842718803691e-06, "loss": 0.1496, "step": 1093 }, { "epoch": 4.445569620253164, "grad_norm": 0.25027186072320673, "learning_rate": 1.4712430108884657e-06, "loss": 0.1885, "step": 1094 }, { "epoch": 4.449620253164557, "grad_norm": 0.2531269655780145, "learning_rate": 1.4499509779581078e-06, "loss": 0.1875, "step": 1095 }, { "epoch": 4.453670886075949, "grad_norm": 0.25253309014302383, "learning_rate": 1.4288083445723988e-06, "loss": 0.1977, "step": 1096 }, { "epoch": 4.457721518987341, "grad_norm": 0.2606055005187019, "learning_rate": 1.4078152810112045e-06, "loss": 0.1681, "step": 1097 }, { "epoch": 4.461772151898734, "grad_norm": 0.2537916660342403, "learning_rate": 1.3869719563497697e-06, "loss": 0.1964, "step": 1098 }, { "epoch": 4.465822784810126, "grad_norm": 0.25157831935083874, "learning_rate": 1.3662785384573663e-06, "loss": 0.1667, "step": 1099 }, { "epoch": 4.469873417721519, "grad_norm": 0.2563554782011798, "learning_rate": 1.3457351939959383e-06, "loss": 0.2253, "step": 1100 }, { "epoch": 4.473924050632911, "grad_norm": 0.2540516789276785, "learning_rate": 1.3253420884187551e-06, "loss": 0.2134, "step": 1101 }, { "epoch": 4.4779746835443035, "grad_norm": 0.2610359280104465, "learning_rate": 1.3050993859690953e-06, "loss": 0.1648, "step": 1102 }, { "epoch": 4.4820253164556965, "grad_norm": 0.2595006599110775, "learning_rate": 1.2850072496788869e-06, "loss": 0.1856, "step": 1103 }, { "epoch": 4.4860759493670885, "grad_norm": 0.2633668160155252, "learning_rate": 1.2650658413674434e-06, "loss": 0.1963, "step": 1104 }, { "epoch": 4.490126582278481, "grad_norm": 0.2522505327681858, "learning_rate": 1.2452753216401226e-06, "loss": 0.1681, "step": 1105 }, { "epoch": 4.494177215189874, "grad_norm": 0.2558502259985488, "learning_rate": 1.2256358498870503e-06, "loss": 0.1784, "step": 1106 }, { "epoch": 4.498227848101266, "grad_norm": 0.2591872026805913, "learning_rate": 1.2061475842818337e-06, "loss": 0.1895, "step": 1107 }, { "epoch": 4.502278481012659, "grad_norm": 0.2478854790531783, "learning_rate": 1.1868106817802816e-06, "loss": 0.185, "step": 1108 }, { "epoch": 4.506329113924051, "grad_norm": 0.24381072603834908, "learning_rate": 1.1676252981191482e-06, "loss": 0.1863, "step": 1109 }, { "epoch": 4.510379746835443, "grad_norm": 0.2559458979130056, "learning_rate": 1.1485915878148823e-06, "loss": 0.1898, "step": 1110 }, { "epoch": 4.514430379746836, "grad_norm": 0.24395240031556958, "learning_rate": 1.1297097041623584e-06, "loss": 0.1961, "step": 1111 }, { "epoch": 4.518481012658228, "grad_norm": 0.2641512824431329, "learning_rate": 1.1109797992336847e-06, "loss": 0.1667, "step": 1112 }, { "epoch": 4.52253164556962, "grad_norm": 0.2597602342293569, "learning_rate": 1.092402023876933e-06, "loss": 0.1578, "step": 1113 }, { "epoch": 4.526582278481013, "grad_norm": 0.27078544141116534, "learning_rate": 1.0739765277149527e-06, "loss": 0.1943, "step": 1114 }, { "epoch": 4.530632911392405, "grad_norm": 0.2593050343149852, "learning_rate": 1.0557034591441596e-06, "loss": 0.1916, "step": 1115 }, { "epoch": 4.534683544303798, "grad_norm": 0.25169106240031747, "learning_rate": 1.0375829653333324e-06, "loss": 0.21, "step": 1116 }, { "epoch": 4.53873417721519, "grad_norm": 0.23942414892766467, "learning_rate": 1.0196151922224385e-06, "loss": 0.1965, "step": 1117 }, { "epoch": 4.542784810126582, "grad_norm": 0.24787070799155525, "learning_rate": 1.0018002845214526e-06, "loss": 0.2052, "step": 1118 }, { "epoch": 4.546835443037975, "grad_norm": 0.25229547602551256, "learning_rate": 9.841383857091947e-07, "loss": 0.1706, "step": 1119 }, { "epoch": 4.550886075949367, "grad_norm": 0.25324981190671353, "learning_rate": 9.666296380321616e-07, "loss": 0.1837, "step": 1120 }, { "epoch": 4.55493670886076, "grad_norm": 0.2737674628910203, "learning_rate": 9.492741825034124e-07, "loss": 0.1848, "step": 1121 }, { "epoch": 4.558987341772152, "grad_norm": 0.25184350560510377, "learning_rate": 9.320721589013892e-07, "loss": 0.1855, "step": 1122 }, { "epoch": 4.563037974683544, "grad_norm": 0.24859363536918, "learning_rate": 9.150237057688339e-07, "loss": 0.1953, "step": 1123 }, { "epoch": 4.567088607594937, "grad_norm": 0.25342356624464996, "learning_rate": 8.981289604116328e-07, "loss": 0.176, "step": 1124 }, { "epoch": 4.571139240506329, "grad_norm": 0.2569979881577742, "learning_rate": 8.813880588977542e-07, "loss": 0.1817, "step": 1125 }, { "epoch": 4.575189873417721, "grad_norm": 0.24649384380113296, "learning_rate": 8.648011360561126e-07, "loss": 0.1852, "step": 1126 }, { "epoch": 4.579240506329114, "grad_norm": 0.24843245310870876, "learning_rate": 8.483683254755037e-07, "loss": 0.178, "step": 1127 }, { "epoch": 4.583291139240506, "grad_norm": 0.2531428470150896, "learning_rate": 8.320897595035227e-07, "loss": 0.1615, "step": 1128 }, { "epoch": 4.587341772151898, "grad_norm": 0.2627098870122786, "learning_rate": 8.159655692455093e-07, "loss": 0.1723, "step": 1129 }, { "epoch": 4.591392405063291, "grad_norm": 0.24879547643526537, "learning_rate": 7.999958845634648e-07, "loss": 0.1538, "step": 1130 }, { "epoch": 4.595443037974683, "grad_norm": 0.24603586651150372, "learning_rate": 7.841808340750478e-07, "loss": 0.1909, "step": 1131 }, { "epoch": 4.599493670886076, "grad_norm": 0.26016029188387546, "learning_rate": 7.685205451524869e-07, "loss": 0.1602, "step": 1132 }, { "epoch": 4.603544303797468, "grad_norm": 0.253635017481058, "learning_rate": 7.530151439216027e-07, "loss": 0.1846, "step": 1133 }, { "epoch": 4.6075949367088604, "grad_norm": 0.25939775102804735, "learning_rate": 7.376647552607675e-07, "loss": 0.1628, "step": 1134 }, { "epoch": 4.611645569620253, "grad_norm": 0.2667684180237091, "learning_rate": 7.224695027998963e-07, "loss": 0.1689, "step": 1135 }, { "epoch": 4.6156962025316455, "grad_norm": 0.2518433612860597, "learning_rate": 7.07429508919466e-07, "loss": 0.1833, "step": 1136 }, { "epoch": 4.619746835443038, "grad_norm": 0.2584396603985693, "learning_rate": 6.925448947495206e-07, "loss": 0.1804, "step": 1137 }, { "epoch": 4.6237974683544305, "grad_norm": 0.2597885748667056, "learning_rate": 6.778157801686936e-07, "loss": 0.1968, "step": 1138 }, { "epoch": 4.627848101265823, "grad_norm": 0.25139364997113467, "learning_rate": 6.632422838032515e-07, "loss": 0.1808, "step": 1139 }, { "epoch": 4.6318987341772155, "grad_norm": 0.24645559490797012, "learning_rate": 6.488245230261281e-07, "loss": 0.1712, "step": 1140 }, { "epoch": 4.635949367088608, "grad_norm": 0.2476305329944155, "learning_rate": 6.345626139559868e-07, "loss": 0.1764, "step": 1141 }, { "epoch": 4.64, "grad_norm": 0.2422692728540795, "learning_rate": 6.204566714562866e-07, "loss": 0.1668, "step": 1142 }, { "epoch": 4.644050632911393, "grad_norm": 0.25453337534225423, "learning_rate": 6.06506809134344e-07, "loss": 0.1787, "step": 1143 }, { "epoch": 4.648101265822785, "grad_norm": 0.2572606439239162, "learning_rate": 5.927131393404373e-07, "loss": 0.1929, "step": 1144 }, { "epoch": 4.652151898734177, "grad_norm": 0.24770792818730872, "learning_rate": 5.790757731668817e-07, "loss": 0.1797, "step": 1145 }, { "epoch": 4.65620253164557, "grad_norm": 0.24483497175066982, "learning_rate": 5.655948204471507e-07, "loss": 0.1991, "step": 1146 }, { "epoch": 4.660253164556962, "grad_norm": 0.2501910019084601, "learning_rate": 5.522703897549875e-07, "loss": 0.2111, "step": 1147 }, { "epoch": 4.664303797468355, "grad_norm": 0.24590789931978552, "learning_rate": 5.391025884035239e-07, "loss": 0.1722, "step": 1148 }, { "epoch": 4.668354430379747, "grad_norm": 0.24842769809819695, "learning_rate": 5.260915224444207e-07, "loss": 0.1778, "step": 1149 }, { "epoch": 4.672405063291139, "grad_norm": 0.24682657160827814, "learning_rate": 5.132372966670129e-07, "loss": 0.1842, "step": 1150 }, { "epoch": 4.676455696202532, "grad_norm": 0.2520379201823887, "learning_rate": 5.005400145974704e-07, "loss": 0.2043, "step": 1151 }, { "epoch": 4.680506329113924, "grad_norm": 0.251024301267163, "learning_rate": 4.879997784979562e-07, "loss": 0.1709, "step": 1152 }, { "epoch": 4.684556962025317, "grad_norm": 0.25043932666538693, "learning_rate": 4.7561668936580984e-07, "loss": 0.2059, "step": 1153 }, { "epoch": 4.688607594936709, "grad_norm": 0.25203107895275306, "learning_rate": 4.6339084693272306e-07, "loss": 0.1989, "step": 1154 }, { "epoch": 4.692658227848101, "grad_norm": 0.24417957780876035, "learning_rate": 4.5132234966395847e-07, "loss": 0.1816, "step": 1155 }, { "epoch": 4.696708860759494, "grad_norm": 0.25741601206336934, "learning_rate": 4.3941129475752795e-07, "loss": 0.1758, "step": 1156 }, { "epoch": 4.700759493670886, "grad_norm": 0.24722312662023863, "learning_rate": 4.27657778143431e-07, "loss": 0.1934, "step": 1157 }, { "epoch": 4.704810126582278, "grad_norm": 0.2626144396945982, "learning_rate": 4.1606189448287757e-07, "loss": 0.1955, "step": 1158 }, { "epoch": 4.708860759493671, "grad_norm": 0.2510612142972368, "learning_rate": 4.046237371675177e-07, "loss": 0.2042, "step": 1159 }, { "epoch": 4.712911392405063, "grad_norm": 0.2584532396115231, "learning_rate": 3.9334339831869963e-07, "loss": 0.1866, "step": 1160 }, { "epoch": 4.716962025316455, "grad_norm": 0.25936596324114375, "learning_rate": 3.8222096878671955e-07, "loss": 0.1793, "step": 1161 }, { "epoch": 4.721012658227848, "grad_norm": 0.2547310919023514, "learning_rate": 3.7125653815009545e-07, "loss": 0.1813, "step": 1162 }, { "epoch": 4.72506329113924, "grad_norm": 0.24657907627773162, "learning_rate": 3.6045019471484974e-07, "loss": 0.1821, "step": 1163 }, { "epoch": 4.729113924050633, "grad_norm": 0.24387121965321734, "learning_rate": 3.498020255137813e-07, "loss": 0.186, "step": 1164 }, { "epoch": 4.733164556962025, "grad_norm": 0.2543241990232599, "learning_rate": 3.393121163057811e-07, "loss": 0.163, "step": 1165 }, { "epoch": 4.737215189873417, "grad_norm": 0.2521649604325065, "learning_rate": 3.289805515751399e-07, "loss": 0.1683, "step": 1166 }, { "epoch": 4.74126582278481, "grad_norm": 0.25134933569848844, "learning_rate": 3.188074145308573e-07, "loss": 0.1811, "step": 1167 }, { "epoch": 4.745316455696202, "grad_norm": 0.246500193770118, "learning_rate": 3.087927871059804e-07, "loss": 0.1836, "step": 1168 }, { "epoch": 4.749367088607595, "grad_norm": 0.2491306515148887, "learning_rate": 2.989367499569418e-07, "loss": 0.1826, "step": 1169 }, { "epoch": 4.7534177215189874, "grad_norm": 0.2468674040278226, "learning_rate": 2.8923938246290917e-07, "loss": 0.2089, "step": 1170 }, { "epoch": 4.7574683544303795, "grad_norm": 0.25995282053858326, "learning_rate": 2.7970076272514804e-07, "loss": 0.1669, "step": 1171 }, { "epoch": 4.7615189873417725, "grad_norm": 0.24694880673651368, "learning_rate": 2.703209675663887e-07, "loss": 0.1933, "step": 1172 }, { "epoch": 4.7655696202531646, "grad_norm": 0.24865632506109372, "learning_rate": 2.6110007253021374e-07, "loss": 0.1796, "step": 1173 }, { "epoch": 4.769620253164557, "grad_norm": 0.25020557218222106, "learning_rate": 2.520381518804471e-07, "loss": 0.204, "step": 1174 }, { "epoch": 4.77367088607595, "grad_norm": 0.2442764471445744, "learning_rate": 2.4313527860054585e-07, "loss": 0.1822, "step": 1175 }, { "epoch": 4.777721518987342, "grad_norm": 0.24415860547428622, "learning_rate": 2.343915243930317e-07, "loss": 0.1942, "step": 1176 }, { "epoch": 4.781772151898734, "grad_norm": 0.252310203497333, "learning_rate": 2.2580695967889367e-07, "loss": 0.1809, "step": 1177 }, { "epoch": 4.785822784810127, "grad_norm": 0.24865312053784286, "learning_rate": 2.1738165359704189e-07, "loss": 0.1646, "step": 1178 }, { "epoch": 4.789873417721519, "grad_norm": 0.2585472096747889, "learning_rate": 2.0911567400373257e-07, "loss": 0.1897, "step": 1179 }, { "epoch": 4.793924050632912, "grad_norm": 0.24794413504625462, "learning_rate": 2.0100908747202607e-07, "loss": 0.1804, "step": 1180 }, { "epoch": 4.797974683544304, "grad_norm": 0.2559261130811481, "learning_rate": 1.9306195929125638e-07, "loss": 0.1862, "step": 1181 }, { "epoch": 4.802025316455696, "grad_norm": 0.237937585332866, "learning_rate": 1.8527435346650247e-07, "loss": 0.1724, "step": 1182 }, { "epoch": 4.806075949367089, "grad_norm": 0.244077100950966, "learning_rate": 1.7764633271807108e-07, "loss": 0.1986, "step": 1183 }, { "epoch": 4.810126582278481, "grad_norm": 0.26516142285800254, "learning_rate": 1.7017795848099262e-07, "loss": 0.1857, "step": 1184 }, { "epoch": 4.814177215189874, "grad_norm": 0.250567419121132, "learning_rate": 1.6286929090452596e-07, "loss": 0.1745, "step": 1185 }, { "epoch": 4.818227848101266, "grad_norm": 0.24757469090696477, "learning_rate": 1.557203888516745e-07, "loss": 0.2018, "step": 1186 }, { "epoch": 4.822278481012658, "grad_norm": 0.2545326647713915, "learning_rate": 1.487313098987131e-07, "loss": 0.1791, "step": 1187 }, { "epoch": 4.826329113924051, "grad_norm": 0.24943308999651193, "learning_rate": 1.4190211033472402e-07, "loss": 0.1728, "step": 1188 }, { "epoch": 4.830379746835443, "grad_norm": 0.25001559956305575, "learning_rate": 1.3523284516113955e-07, "loss": 0.1789, "step": 1189 }, { "epoch": 4.834430379746835, "grad_norm": 0.26323164457536896, "learning_rate": 1.2872356809130682e-07, "loss": 0.171, "step": 1190 }, { "epoch": 4.838481012658228, "grad_norm": 0.26396679656888944, "learning_rate": 1.2237433155004807e-07, "loss": 0.1925, "step": 1191 }, { "epoch": 4.84253164556962, "grad_norm": 0.25192878604296975, "learning_rate": 1.1618518667323886e-07, "loss": 0.1855, "step": 1192 }, { "epoch": 4.846582278481012, "grad_norm": 0.2420196417606236, "learning_rate": 1.1015618330740385e-07, "loss": 0.1907, "step": 1193 }, { "epoch": 4.850632911392405, "grad_norm": 0.24955788011406163, "learning_rate": 1.042873700093061e-07, "loss": 0.1678, "step": 1194 }, { "epoch": 4.854683544303797, "grad_norm": 0.2608037056265271, "learning_rate": 9.857879404556291e-08, "loss": 0.1894, "step": 1195 }, { "epoch": 4.85873417721519, "grad_norm": 0.24676259396376363, "learning_rate": 9.303050139225722e-08, "loss": 0.1957, "step": 1196 }, { "epoch": 4.862784810126582, "grad_norm": 0.24480281357536524, "learning_rate": 8.76425367345779e-08, "loss": 0.1709, "step": 1197 }, { "epoch": 4.866835443037974, "grad_norm": 0.24929410669368673, "learning_rate": 8.241494346644897e-08, "loss": 0.1924, "step": 1198 }, { "epoch": 4.870886075949367, "grad_norm": 0.294429795211666, "learning_rate": 7.734776369019204e-08, "loss": 0.1719, "step": 1199 }, { "epoch": 4.874936708860759, "grad_norm": 0.25402729434435933, "learning_rate": 7.244103821617332e-08, "loss": 0.1726, "step": 1200 }, { "epoch": 4.878987341772152, "grad_norm": 0.24764148201081101, "learning_rate": 6.769480656248606e-08, "loss": 0.1965, "step": 1201 }, { "epoch": 4.883037974683544, "grad_norm": 0.25281846261051283, "learning_rate": 6.310910695462635e-08, "loss": 0.1681, "step": 1202 }, { "epoch": 4.8870886075949365, "grad_norm": 0.24171271396034813, "learning_rate": 5.8683976325191185e-08, "loss": 0.1682, "step": 1203 }, { "epoch": 4.891139240506329, "grad_norm": 0.249834087286601, "learning_rate": 5.4419450313571984e-08, "loss": 0.2096, "step": 1204 }, { "epoch": 4.8951898734177215, "grad_norm": 0.2454138951563455, "learning_rate": 5.031556326567488e-08, "loss": 0.1944, "step": 1205 }, { "epoch": 4.899240506329114, "grad_norm": 0.24753875486973362, "learning_rate": 4.637234823364312e-08, "loss": 0.2026, "step": 1206 }, { "epoch": 4.9032911392405065, "grad_norm": 0.25577664469846784, "learning_rate": 4.258983697558838e-08, "loss": 0.1851, "step": 1207 }, { "epoch": 4.907341772151899, "grad_norm": 0.26742267219413896, "learning_rate": 3.896805995533548e-08, "loss": 0.178, "step": 1208 }, { "epoch": 4.911392405063291, "grad_norm": 0.24284269185672722, "learning_rate": 3.550704634218028e-08, "loss": 0.2029, "step": 1209 }, { "epoch": 4.915443037974684, "grad_norm": 0.25047584310799537, "learning_rate": 3.2206824010647676e-08, "loss": 0.1822, "step": 1210 }, { "epoch": 4.919493670886076, "grad_norm": 0.23727596838951365, "learning_rate": 2.9067419540278476e-08, "loss": 0.1821, "step": 1211 }, { "epoch": 4.923544303797469, "grad_norm": 0.2452879786042457, "learning_rate": 2.6088858215400638e-08, "loss": 0.1926, "step": 1212 }, { "epoch": 4.927594936708861, "grad_norm": 0.2516624880677422, "learning_rate": 2.3271164024940564e-08, "loss": 0.1909, "step": 1213 }, { "epoch": 4.931645569620253, "grad_norm": 0.26483130167250796, "learning_rate": 2.061435966221881e-08, "loss": 0.1872, "step": 1214 }, { "epoch": 4.935696202531646, "grad_norm": 0.2512784059603234, "learning_rate": 1.811846652477245e-08, "loss": 0.1959, "step": 1215 }, { "epoch": 4.939746835443038, "grad_norm": 0.24160335669731411, "learning_rate": 1.5783504714184106e-08, "loss": 0.1821, "step": 1216 }, { "epoch": 4.943797468354431, "grad_norm": 0.24189665983034153, "learning_rate": 1.360949303591097e-08, "loss": 0.1785, "step": 1217 }, { "epoch": 4.947848101265823, "grad_norm": 0.2473303956903516, "learning_rate": 1.1596448999144916e-08, "loss": 0.2115, "step": 1218 }, { "epoch": 4.951898734177215, "grad_norm": 0.2512014798619976, "learning_rate": 9.744388816668172e-09, "loss": 0.1868, "step": 1219 }, { "epoch": 4.955949367088608, "grad_norm": 0.24680806960261, "learning_rate": 8.05332740472009e-09, "loss": 0.182, "step": 1220 }, { "epoch": 4.96, "grad_norm": 0.239232513972375, "learning_rate": 6.523278382872811e-09, "loss": 0.1949, "step": 1221 }, { "epoch": 4.964050632911392, "grad_norm": 0.24205325896387045, "learning_rate": 5.15425407393133e-09, "loss": 0.1806, "step": 1222 }, { "epoch": 4.968101265822785, "grad_norm": 0.24869617215050532, "learning_rate": 3.94626550383137e-09, "loss": 0.1842, "step": 1223 }, { "epoch": 4.972151898734177, "grad_norm": 0.2532437393475568, "learning_rate": 2.899322401546112e-09, "loss": 0.1792, "step": 1224 }, { "epoch": 4.976202531645569, "grad_norm": 0.2448465920018649, "learning_rate": 2.013433199010706e-09, "loss": 0.1809, "step": 1225 }, { "epoch": 4.980253164556962, "grad_norm": 0.25346801640282185, "learning_rate": 1.2886050310556563e-09, "loss": 0.1723, "step": 1226 }, { "epoch": 4.984303797468354, "grad_norm": 0.24482451223918622, "learning_rate": 7.248437353468695e-10, "loss": 0.2027, "step": 1227 }, { "epoch": 4.988354430379747, "grad_norm": 0.24103018302100987, "learning_rate": 3.221538523412449e-10, "loss": 0.1795, "step": 1228 }, { "epoch": 4.992405063291139, "grad_norm": 0.24592806814860735, "learning_rate": 8.053862524670663e-11, "loss": 0.1748, "step": 1229 }, { "epoch": 4.996455696202531, "grad_norm": 0.23401249039930813, "learning_rate": 0.0, "loss": 0.1637, "step": 1230 }, { "epoch": 4.996455696202531, "step": 1230, "total_flos": 2.5035122508038144e+18, "train_loss": 0.036447300247060575, "train_runtime": 7869.5507, "train_samples_per_second": 20.077, "train_steps_per_second": 0.156 } ], "logging_steps": 1, "max_steps": 1230, "num_input_tokens_seen": 0, "num_train_epochs": 5, "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": 2.5035122508038144e+18, "train_batch_size": 1, "trial_name": null, "trial_params": null }