a1_math_open2math / trainer_state.json
sedrickkeh's picture
End of training
de50131 verified
raw
history blame
214 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.982278481012658,
"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": 6.321949967627026,
"learning_rate": 3.2520325203252037e-07,
"loss": 0.8141,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 6.8199983661094405,
"learning_rate": 6.504065040650407e-07,
"loss": 0.8713,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 6.488320756546987,
"learning_rate": 9.75609756097561e-07,
"loss": 0.8253,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 6.14756306041393,
"learning_rate": 1.3008130081300815e-06,
"loss": 0.8193,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 5.744390117853706,
"learning_rate": 1.6260162601626018e-06,
"loss": 0.7855,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 5.837831874637069,
"learning_rate": 1.951219512195122e-06,
"loss": 0.7919,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 4.760133383421948,
"learning_rate": 2.2764227642276426e-06,
"loss": 0.7711,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 4.526236582456803,
"learning_rate": 2.601626016260163e-06,
"loss": 0.7808,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.503762766767712,
"learning_rate": 2.926829268292683e-06,
"loss": 0.729,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.201334010378042,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.7223,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.1499273953386484,
"learning_rate": 3.577235772357724e-06,
"loss": 0.7202,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.639093383403756,
"learning_rate": 3.902439024390244e-06,
"loss": 0.6952,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 3.9852149217528727,
"learning_rate": 4.227642276422765e-06,
"loss": 0.698,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.15534916687844,
"learning_rate": 4.552845528455285e-06,
"loss": 0.6919,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 3.754246201173943,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.655,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 3.161136109475919,
"learning_rate": 5.203252032520326e-06,
"loss": 0.6466,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 3.1286336894242113,
"learning_rate": 5.528455284552846e-06,
"loss": 0.6278,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.577346617949884,
"learning_rate": 5.853658536585366e-06,
"loss": 0.6306,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.8025143025619421,
"learning_rate": 6.178861788617887e-06,
"loss": 0.6218,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.2577613059384753,
"learning_rate": 6.504065040650407e-06,
"loss": 0.5791,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.6181155189034264,
"learning_rate": 6.829268292682928e-06,
"loss": 0.6001,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.793784411815894,
"learning_rate": 7.154471544715448e-06,
"loss": 0.567,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.6907198762228053,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.5983,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.1032930466738173,
"learning_rate": 7.804878048780489e-06,
"loss": 0.5616,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.8523637386284916,
"learning_rate": 8.130081300813009e-06,
"loss": 0.5588,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.0833824522356301,
"learning_rate": 8.45528455284553e-06,
"loss": 0.5509,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 0.9677882990698344,
"learning_rate": 8.78048780487805e-06,
"loss": 0.5417,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.7657210985969838,
"learning_rate": 9.10569105691057e-06,
"loss": 0.5686,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.7555845734088903,
"learning_rate": 9.43089430894309e-06,
"loss": 0.531,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.8580210052381343,
"learning_rate": 9.756097560975611e-06,
"loss": 0.5224,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.7472448625043672,
"learning_rate": 1.008130081300813e-05,
"loss": 0.5535,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.670670324834762,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.5337,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.7915524694680136,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.5295,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.7420731305225214,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.5458,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.5569955812943824,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.5211,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.8909023535136663,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.5303,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.6756952861926832,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.5057,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.6534758768977116,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.5234,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.6540161113459456,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.52,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.6080189331237126,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.4928,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.5695849638215544,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.5006,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.5598796639427857,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.5256,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.5260667616215371,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.4807,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.5277664304653565,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.4887,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.7715052576716896,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.4925,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.5327793458049991,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.5137,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.8455866790570215,
"learning_rate": 1.528455284552846e-05,
"loss": 0.4967,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.5246032628806733,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.4827,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.7017859091084974,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.485,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.5419608444288071,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.4728,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.7747289344509529,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.4805,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.5766170401965557,
"learning_rate": 1.691056910569106e-05,
"loss": 0.4886,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.8567303949607963,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.4944,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.7924905871444238,
"learning_rate": 1.75609756097561e-05,
"loss": 0.5082,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.6239719432421978,
"learning_rate": 1.788617886178862e-05,
"loss": 0.4789,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.7421717186887025,
"learning_rate": 1.821138211382114e-05,
"loss": 0.5003,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.5281281080449155,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.4789,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.6729188603217315,
"learning_rate": 1.886178861788618e-05,
"loss": 0.4749,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.6523756438492239,
"learning_rate": 1.91869918699187e-05,
"loss": 0.4581,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.6897005546219833,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.4909,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.631717057092728,
"learning_rate": 1.983739837398374e-05,
"loss": 0.4586,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.8011112836439717,
"learning_rate": 2.016260162601626e-05,
"loss": 0.4645,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.5869506410782408,
"learning_rate": 2.048780487804878e-05,
"loss": 0.4478,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.6945130585336375,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.4766,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.7645702933354891,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.4843,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.5971978894332436,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.4756,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.8121918029670555,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.4737,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.6353249105066672,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.4556,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.7443156760151339,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.4515,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.6905504581294317,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.4816,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.7079802065086289,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.4681,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.9216652823365556,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.4833,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.8691664428250214,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.4836,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.9202761372432186,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.4622,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 1.0776468919217712,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.4545,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.952394253984643,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.4639,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.9983502506500074,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.4729,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 1.0077484068793887,
"learning_rate": 2.536585365853659e-05,
"loss": 0.4608,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.8174019841834125,
"learning_rate": 2.569105691056911e-05,
"loss": 0.4875,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.9950055496459871,
"learning_rate": 2.601626016260163e-05,
"loss": 0.4627,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 1.2612628579013692,
"learning_rate": 2.634146341463415e-05,
"loss": 0.4561,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.9165206986847768,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.4622,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.941538407664318,
"learning_rate": 2.699186991869919e-05,
"loss": 0.4574,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.2694698822800556,
"learning_rate": 2.731707317073171e-05,
"loss": 0.4579,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.6569643892669981,
"learning_rate": 2.764227642276423e-05,
"loss": 0.4618,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 1.2796113052821287,
"learning_rate": 2.796747967479675e-05,
"loss": 0.4756,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.7672013060339831,
"learning_rate": 2.829268292682927e-05,
"loss": 0.4633,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.9450614178585994,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.4732,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 1.1054754152296005,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.4655,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 1.055952282424391,
"learning_rate": 2.926829268292683e-05,
"loss": 0.4446,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.9338167690491658,
"learning_rate": 2.959349593495935e-05,
"loss": 0.434,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 1.0322866760209968,
"learning_rate": 2.991869918699187e-05,
"loss": 0.4576,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 1.003970762602946,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.4687,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 1.065734570184499,
"learning_rate": 3.056910569105692e-05,
"loss": 0.4713,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.777249002943878,
"learning_rate": 3.089430894308943e-05,
"loss": 0.425,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 1.0062688351790485,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.4663,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.8269409863453074,
"learning_rate": 3.154471544715447e-05,
"loss": 0.4639,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.9625596710107783,
"learning_rate": 3.186991869918699e-05,
"loss": 0.4345,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.9444181913846895,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.4536,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.1264072732182475,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.4467,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.9369043755410301,
"learning_rate": 3.284552845528456e-05,
"loss": 0.4553,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 1.1523424036107508,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.4696,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.8777163014271671,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.481,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 1.0967852812965075,
"learning_rate": 3.382113821138212e-05,
"loss": 0.4803,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.8147757607633831,
"learning_rate": 3.414634146341463e-05,
"loss": 0.4678,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.8853309357938409,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.4368,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.9564210431734049,
"learning_rate": 3.479674796747968e-05,
"loss": 0.4701,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.9690662289828335,
"learning_rate": 3.51219512195122e-05,
"loss": 0.4808,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.7858176985420097,
"learning_rate": 3.544715447154472e-05,
"loss": 0.457,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.083866081000804,
"learning_rate": 3.577235772357724e-05,
"loss": 0.4483,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 0.8656989772716427,
"learning_rate": 3.609756097560976e-05,
"loss": 0.455,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.9734460499136482,
"learning_rate": 3.642276422764228e-05,
"loss": 0.4536,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.0629044896780213,
"learning_rate": 3.67479674796748e-05,
"loss": 0.4564,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.9472612738313319,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.4399,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.7700681900785087,
"learning_rate": 3.739837398373984e-05,
"loss": 0.4481,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.8600872899913012,
"learning_rate": 3.772357723577236e-05,
"loss": 0.4588,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.7869178801024028,
"learning_rate": 3.804878048780488e-05,
"loss": 0.4614,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.9409557851649036,
"learning_rate": 3.83739837398374e-05,
"loss": 0.4608,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.9508703852174584,
"learning_rate": 3.869918699186992e-05,
"loss": 0.4334,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.2869892158810599,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.4602,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.861135831418999,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.4673,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 1.2362980881789531,
"learning_rate": 3.967479674796748e-05,
"loss": 0.4645,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 1.1135472161270028,
"learning_rate": 4e-05,
"loss": 0.45,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.7333454071757762,
"learning_rate": 3.999991946137476e-05,
"loss": 0.4565,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.019242570249519,
"learning_rate": 3.999967784614766e-05,
"loss": 0.453,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.7322353273688116,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.456,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.7946693623120755,
"learning_rate": 3.999871139496895e-05,
"loss": 0.4498,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 1.0388449239066677,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.4498,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 1.135110124648114,
"learning_rate": 3.999710067759846e-05,
"loss": 0.4404,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 1.1372737967625657,
"learning_rate": 3.999605373449617e-05,
"loss": 0.4592,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.8630816426425227,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.4644,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.8500736727727053,
"learning_rate": 3.999347672161713e-05,
"loss": 0.4879,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.8947651963339159,
"learning_rate": 3.999194667259528e-05,
"loss": 0.4588,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.9039882713956191,
"learning_rate": 3.999025561118334e-05,
"loss": 0.4505,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.235880416312397,
"learning_rate": 3.998840355100086e-05,
"loss": 0.4733,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 1.10516076615186,
"learning_rate": 3.998639050696409e-05,
"loss": 0.4522,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.7969459482130627,
"learning_rate": 3.998421649528582e-05,
"loss": 0.4657,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 1.340455735884462,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.4816,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.6699872282344217,
"learning_rate": 3.997938564033779e-05,
"loss": 0.438,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.9923431056316986,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.4561,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.7495611181405863,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.4332,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.7674138712027692,
"learning_rate": 3.997093258045973e-05,
"loss": 0.4558,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.8579357755761065,
"learning_rate": 3.996779317598936e-05,
"loss": 0.4642,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.7466935715508728,
"learning_rate": 3.996449295365782e-05,
"loss": 0.452,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.8146596329769796,
"learning_rate": 3.996103194004467e-05,
"loss": 0.4573,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.6654182512833235,
"learning_rate": 3.995741016302441e-05,
"loss": 0.4515,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 1.0291498511074015,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.4525,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.9119972665400535,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.4662,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 1.1943249324314062,
"learning_rate": 3.994558054968643e-05,
"loss": 0.4591,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.9196475326700626,
"learning_rate": 3.994131602367481e-05,
"loss": 0.4501,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 1.0217956144731273,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.4504,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.8151062564999557,
"learning_rate": 3.993230519343752e-05,
"loss": 0.4746,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.9561310208920438,
"learning_rate": 3.992755896178383e-05,
"loss": 0.4477,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.8005929855598839,
"learning_rate": 3.992265223630981e-05,
"loss": 0.4539,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.9343914591308471,
"learning_rate": 3.991758505653355e-05,
"loss": 0.4494,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.7559765076377314,
"learning_rate": 3.991235746326543e-05,
"loss": 0.4448,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.9686746165212624,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.4576,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.8624291865765256,
"learning_rate": 3.990142120595444e-05,
"loss": 0.4587,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.7522461516967455,
"learning_rate": 3.98957126299907e-05,
"loss": 0.4387,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.9943549183174756,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.4641,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.8648214233500979,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.4333,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.9144199833628481,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.4311,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.6160331606022642,
"learning_rate": 3.98712764319087e-05,
"loss": 0.4195,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.9840925294923062,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.438,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.6896596247092802,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.4408,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.8786166641505603,
"learning_rate": 3.985126869010129e-05,
"loss": 0.433,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.9294817890537319,
"learning_rate": 3.984427961114833e-05,
"loss": 0.4464,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.7712211401634979,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.4453,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.8776465716094822,
"learning_rate": 3.982982204151901e-05,
"loss": 0.445,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.8703531138467249,
"learning_rate": 3.982235366728193e-05,
"loss": 0.4899,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.9213802713321433,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.4411,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 1.1166837882667735,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.4539,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.9519590857159009,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.4351,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.6923171154268185,
"learning_rate": 3.979088432599627e-05,
"loss": 0.4587,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 1.0816732564236065,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.4458,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.9351491146863047,
"learning_rate": 3.977419304032111e-05,
"loss": 0.4548,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.786837290901519,
"learning_rate": 3.976560847560697e-05,
"loss": 0.4664,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.6461030383201106,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.456,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.8028222141406482,
"learning_rate": 3.974796184811956e-05,
"loss": 0.435,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.7100026473934186,
"learning_rate": 3.973889992746979e-05,
"loss": 0.4622,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.7375145537972061,
"learning_rate": 3.972967903243361e-05,
"loss": 0.4268,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.5820506433646312,
"learning_rate": 3.972029923727486e-05,
"loss": 0.441,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.7625667547273004,
"learning_rate": 3.971076061753709e-05,
"loss": 0.4433,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.7502233311018757,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.4525,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.894604683458091,
"learning_rate": 3.969120721289402e-05,
"loss": 0.4526,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.8287719382727782,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.4472,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.711131139808416,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.4482,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.9777462845344229,
"learning_rate": 3.966068788369422e-05,
"loss": 0.4621,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 1.250352414503313,
"learning_rate": 3.965019797448622e-05,
"loss": 0.4463,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.6005486076518204,
"learning_rate": 3.963954980528515e-05,
"loss": 0.4674,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.8057912255057355,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.4453,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.713221240764434,
"learning_rate": 3.961777903121329e-05,
"loss": 0.4542,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.6831748835092057,
"learning_rate": 3.960665660168131e-05,
"loss": 0.4595,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.6829569879923053,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.4538,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.7565038644117508,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.4364,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.681551621984661,
"learning_rate": 3.957234222185657e-05,
"loss": 0.4448,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.7116603526580798,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.4413,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.6586362577180774,
"learning_rate": 3.954867765033605e-05,
"loss": 0.4606,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.7193161060190371,
"learning_rate": 3.953660915306728e-05,
"loss": 0.4318,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.6164839371157329,
"learning_rate": 3.952438331063419e-05,
"loss": 0.4586,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.849329431146727,
"learning_rate": 3.951200022150205e-05,
"loss": 0.4476,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.8337537239344638,
"learning_rate": 3.949945998540253e-05,
"loss": 0.446,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.8891521719348002,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.4456,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.7275452200919326,
"learning_rate": 3.947390847755559e-05,
"loss": 0.4577,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.7677006463582865,
"learning_rate": 3.946089741159648e-05,
"loss": 0.4296,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.726219044247643,
"learning_rate": 3.944772961024501e-05,
"loss": 0.4598,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.8712519666180191,
"learning_rate": 3.943440517955285e-05,
"loss": 0.4269,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.7736034660576943,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.455,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.5639469627986099,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.4289,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.8109880420039349,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.4226,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.5989567522977639,
"learning_rate": 3.937954332854371e-05,
"loss": 0.427,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.5812948529796553,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.4183,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.7890820798188667,
"learning_rate": 3.935117547697387e-05,
"loss": 0.4171,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.7594574392233743,
"learning_rate": 3.933675771619675e-05,
"loss": 0.441,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.5768984502401866,
"learning_rate": 3.932218421983131e-05,
"loss": 0.44,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.8051037948139778,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.4357,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.9534768918141695,
"learning_rate": 3.929257049108054e-05,
"loss": 0.4684,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.5672340060184335,
"learning_rate": 3.927753049720011e-05,
"loss": 0.4476,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.7729358434145178,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.4617,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.73363675365063,
"learning_rate": 3.92469848560784e-05,
"loss": 0.4325,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.6816762357696874,
"learning_rate": 3.923147945484751e-05,
"loss": 0.4702,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.7804679701912035,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.4426,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.8071790122937856,
"learning_rate": 3.920000411543654e-05,
"loss": 0.4406,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.6148119450322652,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.4369,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.6845736818913868,
"learning_rate": 3.916791024049648e-05,
"loss": 0.4154,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.6977925119330053,
"learning_rate": 3.91516316745245e-05,
"loss": 0.4457,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.5733329379222073,
"learning_rate": 3.913519886394389e-05,
"loss": 0.439,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.9146722529543119,
"learning_rate": 3.911861194110225e-05,
"loss": 0.4467,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 1.1862523912499803,
"learning_rate": 3.910187103958837e-05,
"loss": 0.4399,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.7031886799401439,
"learning_rate": 3.908497629423117e-05,
"loss": 0.4365,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 1.0276939467382085,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.4577,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.7552633433962823,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.4374,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.8137481666394772,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.4444,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.7716880563723987,
"learning_rate": 3.901586161429081e-05,
"loss": 0.4405,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.789436580699524,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.4231,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.5431259468129148,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.4347,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.6307559995378647,
"learning_rate": 3.896241703466667e-05,
"loss": 0.4119,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.5319816816300302,
"learning_rate": 3.894429654085585e-05,
"loss": 0.4574,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.5244026649382629,
"learning_rate": 3.892602347228505e-05,
"loss": 0.4592,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.585928885455211,
"learning_rate": 3.890759797612307e-05,
"loss": 0.4369,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.5723493604296689,
"learning_rate": 3.888902020076632e-05,
"loss": 0.4287,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.5975318664501996,
"learning_rate": 3.887029029583764e-05,
"loss": 0.4312,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.6768588704055118,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.441,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.6010752967507942,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.4136,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.6197608643777256,
"learning_rate": 3.881318931821972e-05,
"loss": 0.4218,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.5902047119679896,
"learning_rate": 3.879385241571817e-05,
"loss": 0.4406,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 0.6612460734303174,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.4935,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.5932922847514694,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.3662,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.5301115451685868,
"learning_rate": 3.873493415863256e-05,
"loss": 0.3684,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.6070925202795017,
"learning_rate": 3.871499275032111e-05,
"loss": 0.3749,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.6223103378368309,
"learning_rate": 3.869490061403091e-05,
"loss": 0.3693,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 0.5926131873276659,
"learning_rate": 3.867465791158124e-05,
"loss": 0.3822,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.5822583703652573,
"learning_rate": 3.865426480600407e-05,
"loss": 0.3525,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 0.6642197900714935,
"learning_rate": 3.863372146154264e-05,
"loss": 0.3877,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.5393015463691317,
"learning_rate": 3.861302804365024e-05,
"loss": 0.3652,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.7091442404104473,
"learning_rate": 3.85921847189888e-05,
"loss": 0.3943,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.6795099681960469,
"learning_rate": 3.85711916554276e-05,
"loss": 0.3727,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.7071539459434137,
"learning_rate": 3.85500490220419e-05,
"loss": 0.3957,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.6693365089263168,
"learning_rate": 3.852875698911154e-05,
"loss": 0.3893,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.7122192116507396,
"learning_rate": 3.850731572811963e-05,
"loss": 0.4033,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.639231562089442,
"learning_rate": 3.848572541175116e-05,
"loss": 0.3741,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.7027477942731284,
"learning_rate": 3.846398621389154e-05,
"loss": 0.3954,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.5444536598691878,
"learning_rate": 3.84420983096253e-05,
"loss": 0.3475,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.6523494542191236,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.3655,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.6626519871848646,
"learning_rate": 3.839787708819787e-05,
"loss": 0.367,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.7644467399750277,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.3656,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.6973611727954419,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.3752,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.5858740716211058,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.3667,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.6101456636712379,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.3827,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.5292638190420572,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.3679,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.6337058668543881,
"learning_rate": 3.826166304956594e-05,
"loss": 0.3847,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.6122962027854479,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.3794,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.5789757692251489,
"learning_rate": 3.821507978767479e-05,
"loss": 0.3739,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.5827618687766665,
"learning_rate": 3.819156801001693e-05,
"loss": 0.3493,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.5241311744105478,
"learning_rate": 3.816790971997121e-05,
"loss": 0.3631,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.6650795720874478,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.3902,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.47903471214970306,
"learning_rate": 3.81201543660571e-05,
"loss": 0.3627,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.6415214755319258,
"learning_rate": 3.809605768680377e-05,
"loss": 0.353,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.6745891003534598,
"learning_rate": 3.807181526438958e-05,
"loss": 0.3861,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.4712793149747154,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.3423,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.6371351586212599,
"learning_rate": 3.802289397223145e-05,
"loss": 0.3938,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.5027575051790194,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.3608,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.62036324908681,
"learning_rate": 3.797339206560096e-05,
"loss": 0.3712,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.4696951764734086,
"learning_rate": 3.794842387948027e-05,
"loss": 0.364,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.6847070791153437,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.4222,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.5028974924625986,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.3457,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.6357976911009015,
"learning_rate": 3.787265280646825e-05,
"loss": 0.3808,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.6710094894371826,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.3614,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.6979652247242085,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.3539,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.6412796422091839,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.3616,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.6288252839057961,
"learning_rate": 3.776961046830306e-05,
"loss": 0.3548,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.6303812700893479,
"learning_rate": 3.774349157719215e-05,
"loss": 0.3851,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.6443938531880744,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.3537,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.6647476441531147,
"learning_rate": 3.769082529555359e-05,
"loss": 0.3654,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.6651477281386762,
"learning_rate": 3.766427832919294e-05,
"loss": 0.3838,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.6822201752044151,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.3828,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.7452462739950095,
"learning_rate": 3.761075781441526e-05,
"loss": 0.3819,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 0.6734446509912946,
"learning_rate": 3.75837846970451e-05,
"loss": 0.3628,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.7842013804470751,
"learning_rate": 3.755666996229032e-05,
"loss": 0.3488,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.6564415347716268,
"learning_rate": 3.752941382852927e-05,
"loss": 0.4075,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.702215320335481,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.3397,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.5996239900881898,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.3644,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.7985092822728791,
"learning_rate": 3.744679923406351e-05,
"loss": 0.3533,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.7108749109970185,
"learning_rate": 3.741897971081043e-05,
"loss": 0.4022,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.7894282890259561,
"learning_rate": 3.739101989748946e-05,
"loss": 0.3889,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.5722671265684276,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.3852,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.7260180956667716,
"learning_rate": 3.73346803025098e-05,
"loss": 0.3486,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 0.6148201704838376,
"learning_rate": 3.730630097460247e-05,
"loss": 0.3872,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.6973012911229917,
"learning_rate": 3.727778226412628e-05,
"loss": 0.3286,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 0.6927303295301698,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.4217,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.685915499159247,
"learning_rate": 3.722032761533114e-05,
"loss": 0.3805,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.6528232638481228,
"learning_rate": 3.719139213974403e-05,
"loss": 0.3799,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 0.5294444785209557,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.3702,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.6356293012245876,
"learning_rate": 3.713310605140055e-05,
"loss": 0.4098,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.6016229467266194,
"learning_rate": 3.710375590807233e-05,
"loss": 0.3505,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.7315418258500226,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.3742,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.6410776599431915,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.3452,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.5691344411543786,
"learning_rate": 3.701487992136854e-05,
"loss": 0.3641,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 0.754905396260898,
"learning_rate": 3.69849802022222e-05,
"loss": 0.3931,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.6579048292188042,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.3689,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.5701287157677051,
"learning_rate": 3.692477062175289e-05,
"loss": 0.3685,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.6370159192308493,
"learning_rate": 3.689446124534958e-05,
"loss": 0.3848,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.5027845963458968,
"learning_rate": 3.686401580327799e-05,
"loss": 0.3619,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.5194413979343931,
"learning_rate": 3.683343454074149e-05,
"loss": 0.3744,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.5077262308475304,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.3615,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.5672673530659689,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.3958,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.6070969870187868,
"learning_rate": 3.674087829877297e-05,
"loss": 0.3406,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.6333247815022708,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.3978,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.5493541847755329,
"learning_rate": 3.667849957966501e-05,
"loss": 0.3642,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.6470167702031546,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.3742,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.5723736586321807,
"learning_rate": 3.661558355626795e-05,
"loss": 0.4043,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.670128455838125,
"learning_rate": 3.658392468818163e-05,
"loss": 0.3941,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.6389286930803958,
"learning_rate": 3.655213225544574e-05,
"loss": 0.3427,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.813759126336383,
"learning_rate": 3.652020651411218e-05,
"loss": 0.3913,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.6648340562731355,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.3861,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.6279827554457417,
"learning_rate": 3.645595613522574e-05,
"loss": 0.3686,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.5609390115470597,
"learning_rate": 3.642363201513657e-05,
"loss": 0.3885,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.7131205506748883,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.3677,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.658020647764234,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.3691,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.5619366320421245,
"learning_rate": 3.632586705948318e-05,
"loss": 0.3714,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.7581286623523374,
"learning_rate": 3.629301541734311e-05,
"loss": 0.3874,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.6338884085407034,
"learning_rate": 3.626003255349676e-05,
"loss": 0.3566,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.7708176378333006,
"learning_rate": 3.622691873358357e-05,
"loss": 0.3872,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.7423045083463986,
"learning_rate": 3.61936742242977e-05,
"loss": 0.3579,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.7162037125133675,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.3637,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.7602099116406451,
"learning_rate": 3.612679420964516e-05,
"loss": 0.3684,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.7184879230750214,
"learning_rate": 3.609315924292092e-05,
"loss": 0.373,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.848128338272079,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.3734,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.7263424805763969,
"learning_rate": 3.602550074513133e-05,
"loss": 0.3705,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.7028777869669544,
"learning_rate": 3.599147775897822e-05,
"loss": 0.4049,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.5045229808849552,
"learning_rate": 3.595732597966167e-05,
"loss": 0.3271,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.8549355187969817,
"learning_rate": 3.592304568223542e-05,
"loss": 0.4126,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.7727190655873146,
"learning_rate": 3.588863714278826e-05,
"loss": 0.392,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.8237553535731215,
"learning_rate": 3.585410063844186e-05,
"loss": 0.3603,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.8019415066924191,
"learning_rate": 3.581943644734846e-05,
"loss": 0.3989,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.9438078801936826,
"learning_rate": 3.578464484868869e-05,
"loss": 0.3572,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.7897761430190083,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.4,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.6925314962969537,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.3516,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.9489998015024487,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.3822,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.5624048306355566,
"learning_rate": 3.564420999786543e-05,
"loss": 0.3565,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.8280906844224754,
"learning_rate": 3.560878558491842e-05,
"loss": 0.3874,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.6828403594096736,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.3693,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.7578497815337718,
"learning_rate": 3.553755991230039e-05,
"loss": 0.3844,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.7470592093279689,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.353,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.6944291210616202,
"learning_rate": 3.546583369120419e-05,
"loss": 0.3525,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.6942545685247483,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.3836,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.8267553019390819,
"learning_rate": 3.539360923231766e-05,
"loss": 0.3968,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.6506326819719855,
"learning_rate": 3.535731089018394e-05,
"loss": 0.3929,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.639686441506425,
"learning_rate": 3.532088886237956e-05,
"loss": 0.3706,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.5311097733277504,
"learning_rate": 3.528434344224253e-05,
"loss": 0.3706,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.7085498526182703,
"learning_rate": 3.524767492410464e-05,
"loss": 0.3912,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.4762416495602507,
"learning_rate": 3.521088360328908e-05,
"loss": 0.3287,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.7116776216671409,
"learning_rate": 3.517396977610811e-05,
"loss": 0.379,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.5518904428559207,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.3965,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.6921137434829213,
"learning_rate": 3.509977579282971e-05,
"loss": 0.3822,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.5122919199591476,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.3906,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.6211698991267575,
"learning_rate": 3.502509536445719e-05,
"loss": 0.3651,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.5455202404064395,
"learning_rate": 3.498757348458147e-05,
"loss": 0.3936,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.48537962197966916,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.3277,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.6064638020338912,
"learning_rate": 3.491216790442899e-05,
"loss": 0.379,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.5513304792733938,
"learning_rate": 3.487428481145839e-05,
"loss": 0.3638,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.5431331472577596,
"learning_rate": 3.483628192304278e-05,
"loss": 0.3599,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.6230431545056061,
"learning_rate": 3.479815954525219e-05,
"loss": 0.3694,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.5052767800249827,
"learning_rate": 3.475991798511899e-05,
"loss": 0.3637,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.6046185739426602,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.4046,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.6198513569733549,
"learning_rate": 3.468307855075128e-05,
"loss": 0.3566,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.5410715231452782,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.3563,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.6723255398222208,
"learning_rate": 3.460576609535163e-05,
"loss": 0.3915,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.5618668941118388,
"learning_rate": 3.456693326250006e-05,
"loss": 0.3815,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.604309103743189,
"learning_rate": 3.452798310957058e-05,
"loss": 0.3909,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.5545189079850223,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.3598,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.6295926777135528,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.4072,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.5846643270127403,
"learning_rate": 3.441043187201574e-05,
"loss": 0.3795,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.6059541198640588,
"learning_rate": 3.437101558517728e-05,
"loss": 0.3884,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.6089980662317442,
"learning_rate": 3.433148355615496e-05,
"loss": 0.383,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.5699557429027935,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.3541,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.5845858988338071,
"learning_rate": 3.425207354603043e-05,
"loss": 0.3957,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.5775730383175001,
"learning_rate": 3.421219620448553e-05,
"loss": 0.368,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.4680672894451932,
"learning_rate": 3.417220439986623e-05,
"loss": 0.3495,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.6379508805745,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.4089,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.5100733786994194,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.3719,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.5765699750578449,
"learning_rate": 3.405154543304065e-05,
"loss": 0.3633,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.47710084467917785,
"learning_rate": 3.401109900618843e-05,
"loss": 0.3641,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.6663263071231945,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.4089,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.5397536860894295,
"learning_rate": 3.392986794874714e-05,
"loss": 0.3775,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.5295626494837684,
"learning_rate": 3.388908397238184e-05,
"loss": 0.3676,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.5731544053759818,
"learning_rate": 3.384818813524362e-05,
"loss": 0.382,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.578332213328555,
"learning_rate": 3.380718076670195e-05,
"loss": 0.3726,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.4950916022540452,
"learning_rate": 3.376606219702454e-05,
"loss": 0.3955,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.5097314345822723,
"learning_rate": 3.372483275737468e-05,
"loss": 0.3673,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.5936361324057824,
"learning_rate": 3.368349277980861e-05,
"loss": 0.4053,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.4857911033248024,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.3531,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.5260042077488933,
"learning_rate": 3.360048254360144e-05,
"loss": 0.3856,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.5417361001125799,
"learning_rate": 3.355881295351336e-05,
"loss": 0.3729,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.5814360675306575,
"learning_rate": 3.351703416260975e-05,
"loss": 0.3739,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.47000010429312095,
"learning_rate": 3.347514650737126e-05,
"loss": 0.3682,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.6468705930036914,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.4003,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.5067219971347707,
"learning_rate": 3.339104595419334e-05,
"loss": 0.366,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.47629712467535473,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.351,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.5523034974776373,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.3948,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.489811047204724,
"learning_rate": 3.326408710419996e-05,
"loss": 0.3901,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.5342996834213659,
"learning_rate": 3.322155337795454e-05,
"loss": 0.3459,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.5431926630102648,
"learning_rate": 3.317891316713587e-05,
"loss": 0.3796,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.4559862541955908,
"learning_rate": 3.313616681516231e-05,
"loss": 0.3526,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.5797114411110708,
"learning_rate": 3.309331466630713e-05,
"loss": 0.3888,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.5043779907174283,
"learning_rate": 3.305035706569563e-05,
"loss": 0.3837,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.6845579563626579,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.4108,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.49762999420579973,
"learning_rate": 3.296412689394864e-05,
"loss": 0.3326,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.5891402247247638,
"learning_rate": 3.292085501729909e-05,
"loss": 0.3746,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.5140421265435756,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.3657,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.5334256252935604,
"learning_rate": 3.283399942497381e-05,
"loss": 0.3651,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.531139981740974,
"learning_rate": 3.279041640882108e-05,
"loss": 0.3677,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.4405552493678426,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.3791,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.5483242106349978,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.3916,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.4397977629042849,
"learning_rate": 3.265905069502244e-05,
"loss": 0.3651,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.5198827273321849,
"learning_rate": 3.261505774420016e-05,
"loss": 0.3758,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.5188965785556219,
"learning_rate": 3.257096319343707e-05,
"loss": 0.373,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.573604526321178,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.3758,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.4904091276181341,
"learning_rate": 3.248247071342966e-05,
"loss": 0.3817,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.6418551864783973,
"learning_rate": 3.243807349689161e-05,
"loss": 0.4068,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.5716766604932132,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.3759,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.6256480959774622,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.3779,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.7870799910202776,
"learning_rate": 3.230428223438075e-05,
"loss": 0.3708,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.5591004426216337,
"learning_rate": 3.225948647317459e-05,
"loss": 0.3547,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.7442181870529674,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.4077,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.5071374600495145,
"learning_rate": 3.216959910368034e-05,
"loss": 0.4163,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.577473946625859,
"learning_rate": 3.212450821933277e-05,
"loss": 0.3805,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.5213570473780911,
"learning_rate": 3.207931968586281e-05,
"loss": 0.3534,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.4911233633145584,
"learning_rate": 3.203403386721272e-05,
"loss": 0.3666,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.5714084857494942,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.3778,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.4711721972964105,
"learning_rate": 3.194317183405573e-05,
"loss": 0.3676,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.5972918685466335,
"learning_rate": 3.189759635133914e-05,
"loss": 0.3839,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.6231327543317907,
"learning_rate": 3.185192504701718e-05,
"loss": 0.3515,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.5365055918467982,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.3946,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.5183045687230502,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.3508,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.5709262199011997,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.3624,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.5456976038511028,
"learning_rate": 3.166828898179731e-05,
"loss": 0.3574,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.5554342270306488,
"learning_rate": 3.162214410223536e-05,
"loss": 0.3796,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.6423204438691494,
"learning_rate": 3.157590561952257e-05,
"loss": 0.3994,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.46149449426401207,
"learning_rate": 3.152957390605732e-05,
"loss": 0.3513,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.5725559475340252,
"learning_rate": 3.148314933498886e-05,
"loss": 0.3485,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.47013469438249095,
"learning_rate": 3.143663228021431e-05,
"loss": 0.3594,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.4516902661278115,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.3806,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.49313863292068383,
"learning_rate": 3.134332221885661e-05,
"loss": 0.3671,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.48658193203203526,
"learning_rate": 3.129652996377987e-05,
"loss": 0.3667,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.4952997422151929,
"learning_rate": 3.12496467280038e-05,
"loss": 0.3733,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.5096746521040342,
"learning_rate": 3.120267288911952e-05,
"loss": 0.3595,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.5574979513032707,
"learning_rate": 3.11556088254479e-05,
"loss": 0.3915,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.49870863628648027,
"learning_rate": 3.11084549160364e-05,
"loss": 0.3648,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.6025653444308415,
"learning_rate": 3.106121154065615e-05,
"loss": 0.3731,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.46770279907822626,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.3678,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.62141502625386,
"learning_rate": 3.096645791467348e-05,
"loss": 0.3743,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.5077986967531527,
"learning_rate": 3.091894842720373e-05,
"loss": 0.3638,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.5627789947841831,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.3786,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.6050096071798715,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.3992,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.6721391729061993,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.3538,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.5999797243621217,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.3788,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.6986835565894822,
"learning_rate": 3.06800895716334e-05,
"loss": 0.3887,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.5378988215362247,
"learning_rate": 3.063205821011029e-05,
"loss": 0.3505,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.6035634994875964,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.3797,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.5328491648119086,
"learning_rate": 3.053573898718618e-05,
"loss": 0.3885,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.5901491018290067,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.3882,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.4547757655530949,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.3899,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.43725575077592166,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.368,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.5297611626390727,
"learning_rate": 3.034208541656255e-05,
"loss": 0.3779,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.477166338795374,
"learning_rate": 3.029346281309218e-05,
"loss": 0.3949,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.5106969175175945,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.3454,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.5237675224403795,
"learning_rate": 3.019596929201569e-05,
"loss": 0.3932,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.5018250454888903,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.353,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.4563181603833332,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.3651,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.5811080663473511,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.402,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.4382085751836705,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.3596,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.4922053397427758,
"learning_rate": 2.995080534235264e-05,
"loss": 0.3779,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.593093739525878,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.3756,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.5538959854870248,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.3027,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 0.673865705032606,
"learning_rate": 2.980274210295326e-05,
"loss": 0.2929,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.5038399335144141,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.2659,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.48885097579864534,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.2897,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.6080462027271569,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3093,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.5384879099143883,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.272,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.44079217583128627,
"learning_rate": 2.955439637024526e-05,
"loss": 0.272,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.5497734893916537,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.3098,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.39209424482394517,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.2455,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.5898620467370751,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.2962,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.46905699987094907,
"learning_rate": 2.93543322968538e-05,
"loss": 0.2973,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.5249500860974198,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.2565,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.6439663370389226,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.3001,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.587555970688215,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3044,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.5712393204054836,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.2834,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.48078530418199455,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.2635,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.6305415181396017,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.321,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.5735906834945632,
"learning_rate": 2.900133526131588e-05,
"loss": 0.2573,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.5048125863712263,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.283,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.5771311715981097,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.2629,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.505156258462438,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.2853,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.6013734391608286,
"learning_rate": 2.879801882631847e-05,
"loss": 0.2907,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.3757545464005975,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.2777,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.5780568316112885,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.2742,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.46405442848213085,
"learning_rate": 2.864478605989494e-05,
"loss": 0.311,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.48236471009341964,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.2474,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.4894258456721074,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.294,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.3804278155783774,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.2661,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.5161510673870723,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.2931,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.47294926374256985,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.2644,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.5301634773070928,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.2892,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.43494814671782417,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.2714,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.4768542495789591,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.2889,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.44917709767549946,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.2895,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.4681611162223507,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3083,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.4326487812211854,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.2748,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.40461783465747436,
"learning_rate": 2.802570062143278e-05,
"loss": 0.2679,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.4602497626590127,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.2982,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.4158026794906077,
"learning_rate": 2.792159532078314e-05,
"loss": 0.2643,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.39922433133197477,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.257,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.4463871631574887,
"learning_rate": 2.781723482288857e-05,
"loss": 0.296,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.39923365542690586,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.2635,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.49091645018374835,
"learning_rate": 2.771262248976272e-05,
"loss": 0.2945,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.40577509231911546,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.2569,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.49742619896883133,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3117,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.44373660187789804,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.2668,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.4268279200532454,
"learning_rate": 2.7502655806328e-05,
"loss": 0.2796,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.48588023788458623,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.2691,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.5363467661660773,
"learning_rate": 2.739730822017673e-05,
"loss": 0.3421,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.4037882144645396,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.2749,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.5326243592531117,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.2767,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.45171916819682023,
"learning_rate": 2.723884107774236e-05,
"loss": 0.2964,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.522107836213339,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3181,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.4146238073474943,
"learning_rate": 2.71329041039194e-05,
"loss": 0.2648,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.48507414204137594,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.2676,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.4548578287811572,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.2747,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.46061773501856995,
"learning_rate": 2.697356885686189e-05,
"loss": 0.2731,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.520861760288944,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3054,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.4179535650037209,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.2664,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.46896180356185097,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.2952,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.4735363051256582,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.2948,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.4232937039750707,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.2624,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.5196940894807031,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3036,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.5071138086359843,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.2909,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.5146427989986204,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.2762,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.5354532326077166,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.3105,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.443082597353416,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.2457,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.49647525542453036,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3052,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.4060061839556909,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.2596,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.473254927025622,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.2812,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.4290575398830281,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.2914,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.512355115735953,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.2795,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.4849182590053955,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3046,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.4652580975889665,
"learning_rate": 2.606146264569603e-05,
"loss": 0.2776,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.5054995590668088,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.2841,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.40720863977550853,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.2997,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.5564492883739593,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.2953,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.39609946528207157,
"learning_rate": 2.584472010606015e-05,
"loss": 0.2949,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.6572882994104452,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.304,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.38970923154733966,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.2529,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.5868013564536844,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3134,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.4633872753042094,
"learning_rate": 2.56272244128509e-05,
"loss": 0.2651,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.4908438542203886,
"learning_rate": 2.55727360915895e-05,
"loss": 0.2532,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.47208486504666963,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.3105,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.564565537034114,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.2812,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.4106061508874553,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3084,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.4786843298858375,
"learning_rate": 2.535433837985559e-05,
"loss": 0.2772,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.4194564879815133,
"learning_rate": 2.529963004393324e-05,
"loss": 0.2965,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.4130393227546154,
"learning_rate": 2.524487902551908e-05,
"loss": 0.2819,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.4678320355660151,
"learning_rate": 2.519008576557029e-05,
"loss": 0.2951,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.4198026120276588,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.2776,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.4324821539705962,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.2952,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.39135633655407837,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.2956,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.47148090243237756,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.2733,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.4298285179750515,
"learning_rate": 2.491550129992345e-05,
"loss": 0.3112,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.3835993090114841,
"learning_rate": 2.486046386967024e-05,
"loss": 0.2429,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.4320143826366396,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.2803,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.4094315005131263,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.292,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.440375048742197,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.2867,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.4928371367658382,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.289,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.4340500056097791,
"learning_rate": 2.458469841202444e-05,
"loss": 0.2772,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.3980291957675104,
"learning_rate": 2.452943276859226e-05,
"loss": 0.2668,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.41543339083449476,
"learning_rate": 2.447413064573125e-05,
"loss": 0.2904,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.44856637869572913,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.2793,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.4562052204229528,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.3071,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.5959302866213294,
"learning_rate": 2.430800985597963e-05,
"loss": 0.3283,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.3741770683748045,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.2645,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.45667394024324337,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.2674,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.46063229625144747,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.2716,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.43970384507369575,
"learning_rate": 2.408603181104957e-05,
"loss": 0.3052,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.38661011435995857,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.2564,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.3952789468905448,
"learning_rate": 2.397484355008521e-05,
"loss": 0.2719,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.3866003509597451,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.2818,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.4249803576555808,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.3025,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.4220466976945481,
"learning_rate": 2.380782218329337e-05,
"loss": 0.2831,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.35497046932730186,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.2596,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.42127218531199895,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.3195,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.38805030738043556,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.2906,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.3315743982735547,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.254,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.38013961574533955,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.2811,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.40030740877649773,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.3056,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.378837141033372,
"learning_rate": 2.341705325992901e-05,
"loss": 0.3081,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.3776478865324526,
"learning_rate": 2.336111544604222e-05,
"loss": 0.2763,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.3927727439578414,
"learning_rate": 2.33051505621937e-05,
"loss": 0.286,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.3544326210682019,
"learning_rate": 2.324915905911693e-05,
"loss": 0.2676,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.36617029268471696,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3015,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.32282547047554155,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.2582,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.38001447152372375,
"learning_rate": 2.308102934504593e-05,
"loss": 0.2889,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.3995208038526333,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.301,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.3683266320349205,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.2562,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.4079616069291649,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3164,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.32981703454268513,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.2421,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.4255864047759635,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3163,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.4061085652124247,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.2826,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.4047488432742045,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.2801,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.3714180599209369,
"learning_rate": 2.26316247733093e-05,
"loss": 0.2833,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.48434481198829926,
"learning_rate": 2.257534907110328e-05,
"loss": 0.2853,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.4020387923903253,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.2849,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.4362417585172134,
"learning_rate": 2.246273589557294e-05,
"loss": 0.2973,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.5489057726408276,
"learning_rate": 2.240639932921966e-05,
"loss": 0.3164,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.36985399471488123,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.2943,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.35835533356045157,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.2463,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.3802873562364821,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.2992,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.37507945035050516,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.2854,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.40480106137928634,
"learning_rate": 2.212443486523819e-05,
"loss": 0.294,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.3466099727691219,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.2829,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.3891405147954072,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.2989,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.4132357320826912,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.2618,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.3900917568124276,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3144,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.358101745683651,
"learning_rate": 2.184204266048803e-05,
"loss": 0.2878,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.3477300438804496,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.2699,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.3866787472286775,
"learning_rate": 2.17289787446841e-05,
"loss": 0.2714,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.37343111841630233,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.2936,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.37381450559996626,
"learning_rate": 2.161585912916385e-05,
"loss": 0.2899,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.39377728116098054,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.2902,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.3671876985649111,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.274,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.41051229226746877,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.3057,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.45051344454766024,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.3014,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.4109785157331815,
"learning_rate": 2.133284032321232e-05,
"loss": 0.2579,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.44695211051783246,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.2685,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.4557215230925412,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3026,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.4539460445825398,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.2762,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.5264259693317037,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.308,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.44543036408656983,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.2633,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.42670515306005696,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3149,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.4817430336149499,
"learning_rate": 2.093617592756158e-05,
"loss": 0.2883,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.40983849369515485,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.2987,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.36970532474235995,
"learning_rate": 2.082276853708978e-05,
"loss": 0.2554,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.3752489456838842,
"learning_rate": 2.076605467380071e-05,
"loss": 0.2739,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.341224852156229,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.2822,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.34113297343219817,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.2653,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.3570666289525047,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.2878,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.3529330112371475,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.2721,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.36615525653480974,
"learning_rate": 2.048240194884496e-05,
"loss": 0.2929,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.3308640258881196,
"learning_rate": 2.042565792045695e-05,
"loss": 0.2504,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.402899373034655,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3059,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.3754585891881446,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.2845,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.4219300354913725,
"learning_rate": 2.025540709431948e-05,
"loss": 0.2981,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.3747487242247928,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3071,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.40297142278616027,
"learning_rate": 2.014189549672245e-05,
"loss": 0.2849,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.42195909584567604,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.3006,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.3795628554658524,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.2757,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.40412625879847486,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.2966,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.42034283889260926,
"learning_rate": 1.991486224484123e-05,
"loss": 0.2615,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.36164754395761955,
"learning_rate": 1.985810450327756e-05,
"loss": 0.2873,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.38435321002785144,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.2777,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.3840939395096417,
"learning_rate": 1.974459290568053e-05,
"loss": 0.2775,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.39238754686185523,
"learning_rate": 1.968783996385397e-05,
"loss": 0.2746,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.4008545714323698,
"learning_rate": 1.963108953612143e-05,
"loss": 0.2754,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.3433750776144352,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.2735,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.3999610758353946,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.2887,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.36306407650088524,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.2914,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.4727077178333078,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.306,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.3557658623865322,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.27,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.4692384610547744,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3049,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.3656205333114426,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.2682,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.509054237210849,
"learning_rate": 1.917723146291022e-05,
"loss": 0.2804,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.3949548113001569,
"learning_rate": 1.912052422608584e-05,
"loss": 0.2818,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.4086513643714326,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.2871,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.47858137092813735,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.2558,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.39793590965016695,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3024,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.5651284751751702,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.2827,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.3696402717881142,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.2637,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.4490188601890236,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3013,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.4749337414873632,
"learning_rate": 1.872379746550743e-05,
"loss": 0.308,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.3551323674605517,
"learning_rate": 1.866715967678769e-05,
"loss": 0.2747,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.43394071460729716,
"learning_rate": 1.861053262258067e-05,
"loss": 0.2638,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.3943672226551438,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.302,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.38193776260582096,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.287,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.4027832200171129,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.2905,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.34913817935652525,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.2792,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.35962937512287657,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.2841,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.3860240365752726,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.267,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.4219953491673746,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3138,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.3654631320972551,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.2816,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.3888057825575905,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.2843,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.3453588178308771,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.2855,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.40339483991087344,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.2735,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.37606984624863327,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.3168,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.30331086810556135,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.2474,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.4182463503798477,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.3167,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.32121121262837876,
"learning_rate": 1.776272483901182e-05,
"loss": 0.2699,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.388934107624758,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.2828,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.3343410609177138,
"learning_rate": 1.764995661794301e-05,
"loss": 0.2783,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.35478551369844147,
"learning_rate": 1.759360067078035e-05,
"loss": 0.2717,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.34965251001999614,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.2628,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.3659394279282846,
"learning_rate": 1.748094737261012e-05,
"loss": 0.2721,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.387394154035108,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.2937,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.34966977342594396,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.2673,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.42172612721604574,
"learning_rate": 1.731212071922883e-05,
"loss": 0.3018,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.3545010712860689,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.2815,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.3528236287028636,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.2638,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.4020107282387961,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3075,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.347893536407178,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.2816,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.4018438494290215,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.2956,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.3595472096136101,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.298,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.34243842282928755,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.2495,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.3464411698525787,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3105,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.34714613505077985,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.2874,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.3740363357313951,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.2933,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.37356597425900867,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.2801,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.3361851992803513,
"learning_rate": 1.663888455395778e-05,
"loss": 0.2993,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.42028589259894766,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3065,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.3071700416488511,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.2552,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.35611592311540813,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.2841,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.3647689086158393,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.304,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.401309643919786,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.2726,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.4494879446247661,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2086,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.35381986287484263,
"learning_rate": 1.624791353909428e-05,
"loss": 0.1998,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.6571210713631405,
"learning_rate": 1.619217781670663e-05,
"loss": 0.2126,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.38493755101395094,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.1762,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.43153963795665495,
"learning_rate": 1.608079882360134e-05,
"loss": 0.1932,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.4556834033422859,
"learning_rate": 1.60251564499148e-05,
"loss": 0.2118,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.38457269229223884,
"learning_rate": 1.596954608907176e-05,
"loss": 0.1792,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.3827861040724339,
"learning_rate": 1.591396818895043e-05,
"loss": 0.192,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.40422200285693727,
"learning_rate": 1.585842319716759e-05,
"loss": 0.1863,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.39608316826023104,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.2117,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.37506560895077706,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.1959,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.3239979617907035,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.187,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.3526233318512141,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.1885,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.3643454883397354,
"learning_rate": 1.558120751116291e-05,
"loss": 0.2096,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.3340850194985118,
"learning_rate": 1.552586935426876e-05,
"loss": 0.19,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.39603528545617583,
"learning_rate": 1.547056723140774e-05,
"loss": 0.1922,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.32947913485758346,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.1834,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.3447655936199811,
"learning_rate": 1.536007286907411e-05,
"loss": 0.1906,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.3490837474095369,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.1837,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.3252215490168291,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.1794,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.3506484502188549,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.2067,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.313253311171457,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.1985,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.3195519625607016,
"learning_rate": 1.508449870007656e-05,
"loss": 0.181,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.31497916208230764,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.1875,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.3214370030018471,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.1937,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.3280108037126933,
"learning_rate": 1.4919625713405e-05,
"loss": 0.1819,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.3281325064357066,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.195,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.3633544380453866,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.2028,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.30386641627069044,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.1679,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.33964352912661855,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.1955,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.3334583956202426,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.1589,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.3464035542502072,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.2142,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.3543989297743814,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2043,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.3280051263906032,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.1863,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.3421642474344214,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2002,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.3271810313746488,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.1882,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.3273951186875465,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.1916,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.3279943208981525,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.2165,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.31562835087780816,
"learning_rate": 1.420958430225303e-05,
"loss": 0.1927,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.303071191813408,
"learning_rate": 1.415527989393985e-05,
"loss": 0.1843,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.32843643189392013,
"learning_rate": 1.410102255819891e-05,
"loss": 0.2031,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.3602073809373371,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2052,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.31321416662455503,
"learning_rate": 1.399265085197556e-05,
"loss": 0.1771,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.3503656228243467,
"learning_rate": 1.393853735430398e-05,
"loss": 0.2161,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.32715290923669305,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.1925,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.29752128349034096,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.1752,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.31972021759865543,
"learning_rate": 1.377649151173163e-05,
"loss": 0.1819,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.35726438164714275,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.1963,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.32657403835389953,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.1875,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.36686751256191136,
"learning_rate": 1.361489677627324e-05,
"loss": 0.1827,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.3061089433884727,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.1795,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.30810380778079943,
"learning_rate": 1.350742335331241e-05,
"loss": 0.187,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.3572034008758245,
"learning_rate": 1.345376486102198e-05,
"loss": 0.2043,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.3117954977640782,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.1932,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.3466191479430294,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.2103,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.3082684496422822,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.1861,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.29947712149768213,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.1713,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.406977376750653,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2089,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.3173709159412448,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.1884,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.3036748862583886,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.1685,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.3378481692454428,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.1951,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.3423786435754868,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.1947,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.310644923713298,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.1793,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.3375710704663777,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.1871,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.3241117553714247,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.1985,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.31781233005326803,
"learning_rate": 1.276115892225764e-05,
"loss": 0.1954,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.3001392963396394,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.1647,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.33180034602268044,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.2065,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.3138339001813041,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.1906,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.33717246587815036,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.2073,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.28423208679217404,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.1953,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.30851047673086524,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.1931,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.3145171436431576,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.2309,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.3223081171004941,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.1727,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.3228554485027876,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.1954,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.33149853624937964,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2085,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.33239332776375285,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.2072,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.3151556169006731,
"learning_rate": 1.213055323844297e-05,
"loss": 0.184,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.36122777980596416,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.2048,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.3391039648143441,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.204,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.29154332300476044,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.17,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.37247782957764125,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2017,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.3128434680774391,
"learning_rate": 1.187045262895488e-05,
"loss": 0.1996,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.31596160079454044,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.1838,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.3514880734477051,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.1926,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.29629312060832774,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.1742,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.3725673122948238,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.2032,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.3179295018010224,
"learning_rate": 1.161198884952377e-05,
"loss": 0.1869,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.35456916784028575,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2074,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.2816457152404054,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.1675,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.3415124872732163,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2182,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.3052600764423698,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.1836,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.34908442255408284,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2043,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.31787067969809163,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.2075,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.2971433762374366,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.1949,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.33136975313195804,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.185,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.3584672145064533,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.1873,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.3348274763473414,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.205,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.31005174635379884,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.1706,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.3568890761984914,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.1936,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.28907935627872994,
"learning_rate": 1.0948015846744e-05,
"loss": 0.1762,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.32458059949344953,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.2063,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.34332649514479757,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.1967,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.2786500854730323,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.1746,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.3311157392292299,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.2072,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.30390316626074027,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.1874,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.30774439446802043,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.203,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.3160892865510697,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.1824,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.3308126971387281,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.1806,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.3195998826434039,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.1963,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.2917501342813611,
"learning_rate": 1.044560362975474e-05,
"loss": 0.1704,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.36026427134491057,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.2285,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.29912877103625146,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.1856,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.3169965387444753,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.208,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.310433406732971,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.181,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.3004522245276902,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.1946,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.3097588823988763,
"learning_rate": 1.014782400334433e-05,
"loss": 0.1941,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.31820842468511207,
"learning_rate": 1.009846945771296e-05,
"loss": 0.196,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.33356981423137827,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.2165,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.28800031979569146,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.171,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.33065718033134617,
"learning_rate": 9.950885880977891e-06,
"loss": 0.2055,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.31291653519652063,
"learning_rate": 9.901852696139382e-06,
"loss": 0.1744,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.31192393862117695,
"learning_rate": 9.852900840391027e-06,
"loss": 0.197,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.30411125682895074,
"learning_rate": 9.804030707984313e-06,
"loss": 0.1864,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.30448389622791167,
"learning_rate": 9.755242692512599e-06,
"loss": 0.2062,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.27657187249138393,
"learning_rate": 9.70653718690782e-06,
"loss": 0.1897,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.3016953465130358,
"learning_rate": 9.657914583437454e-06,
"loss": 0.1772,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.3247755524259392,
"learning_rate": 9.609375273701246e-06,
"loss": 0.1905,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.2983230668484799,
"learning_rate": 9.560919648628133e-06,
"loss": 0.1952,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.2791566581949605,
"learning_rate": 9.512548098473047e-06,
"loss": 0.1664,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.35017218642140213,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2249,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.28345241423673895,
"learning_rate": 9.416058780547987e-06,
"loss": 0.1797,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.31510783958162175,
"learning_rate": 9.367941789889714e-06,
"loss": 0.196,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.2929298018229853,
"learning_rate": 9.319910428366607e-06,
"loss": 0.1709,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.32183200298485964,
"learning_rate": 9.271965082816667e-06,
"loss": 0.2259,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.29037135712232304,
"learning_rate": 9.224106139385111e-06,
"loss": 0.1781,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.2954694359343009,
"learning_rate": 9.176333983521291e-06,
"loss": 0.2023,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.28392218674290415,
"learning_rate": 9.12864899997558e-06,
"loss": 0.1817,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.3166746150962137,
"learning_rate": 9.08105157279628e-06,
"loss": 0.208,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.26366872226419663,
"learning_rate": 9.03354208532653e-06,
"loss": 0.1584,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.30064909486334657,
"learning_rate": 8.986120920201205e-06,
"loss": 0.1866,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.29000392149457344,
"learning_rate": 8.938788459343852e-06,
"loss": 0.1887,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.31619291346988887,
"learning_rate": 8.8915450839636e-06,
"loss": 0.2016,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.3152321705642546,
"learning_rate": 8.844391174552116e-06,
"loss": 0.1902,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.29210465260853147,
"learning_rate": 8.797327110880479e-06,
"loss": 0.1811,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.2821742617092896,
"learning_rate": 8.750353271996206e-06,
"loss": 0.1795,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.30863737057768914,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2107,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.2967985671369567,
"learning_rate": 8.656677781143394e-06,
"loss": 0.1866,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.31083807726294765,
"learning_rate": 8.609976883624377e-06,
"loss": 0.2044,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.29925262315020507,
"learning_rate": 8.563367719785698e-06,
"loss": 0.1839,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.3058087682714933,
"learning_rate": 8.516850665011138e-06,
"loss": 0.1937,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.2959978586081765,
"learning_rate": 8.47042609394269e-06,
"loss": 0.1915,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.2994009310830393,
"learning_rate": 8.424094380477432e-06,
"loss": 0.1918,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.30614594951124896,
"learning_rate": 8.37785589776465e-06,
"loss": 0.1991,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.2912263282848892,
"learning_rate": 8.331711018202694e-06,
"loss": 0.1766,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.28702940481303857,
"learning_rate": 8.285660113436104e-06,
"loss": 0.1911,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.30292694042839297,
"learning_rate": 8.239703554352527e-06,
"loss": 0.1929,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.3363721348732952,
"learning_rate": 8.193841711079775e-06,
"loss": 0.2116,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.30769324600182746,
"learning_rate": 8.148074952982828e-06,
"loss": 0.1998,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.2665452495739093,
"learning_rate": 8.102403648660859e-06,
"loss": 0.175,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.3309048251440096,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2076,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.2982087067531819,
"learning_rate": 8.011348871891762e-06,
"loss": 0.1961,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.2836320030639386,
"learning_rate": 7.965966132787287e-06,
"loss": 0.1706,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.35586942415067974,
"learning_rate": 7.920680314137189e-06,
"loss": 0.208,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.2815691121569045,
"learning_rate": 7.875491780667246e-06,
"loss": 0.187,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.31244157099153974,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2129,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.3287067053160449,
"learning_rate": 7.785408024250259e-06,
"loss": 0.1917,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.29527485301754425,
"learning_rate": 7.74051352682542e-06,
"loss": 0.1727,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.2949233211556494,
"learning_rate": 7.695717765619257e-06,
"loss": 0.2023,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.2892121692711779,
"learning_rate": 7.651021101410673e-06,
"loss": 0.1931,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.31921025675722897,
"learning_rate": 7.606423894180464e-06,
"loss": 0.2027,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.2968896195733594,
"learning_rate": 7.56192650310839e-06,
"loss": 0.1986,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.2875040707705201,
"learning_rate": 7.517529286570349e-06,
"loss": 0.1758,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.3190329112177124,
"learning_rate": 7.473232602135387e-06,
"loss": 0.1965,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.30296831487578185,
"learning_rate": 7.429036806562935e-06,
"loss": 0.1806,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.2989677867701075,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.184,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.30858261563547773,
"learning_rate": 7.340949304977567e-06,
"loss": 0.2095,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.2984757369262141,
"learning_rate": 7.297058308409282e-06,
"loss": 0.1708,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.3001343895517145,
"learning_rate": 7.25326961958704e-06,
"loss": 0.1806,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.29363586955334803,
"learning_rate": 7.209583591178921e-06,
"loss": 0.1779,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.30968578964945676,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.1784,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.29082653174118994,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.1965,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.3161166327625153,
"learning_rate": 7.079144982700909e-06,
"loss": 0.1938,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.2778980112666072,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.1871,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.2966650163803844,
"learning_rate": 6.99270564069757e-06,
"loss": 0.2079,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.2727125086958448,
"learning_rate": 6.949642934304375e-06,
"loss": 0.1813,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.2955855841232463,
"learning_rate": 6.906685333692871e-06,
"loss": 0.1997,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.2545364928566164,
"learning_rate": 6.86383318483769e-06,
"loss": 0.1591,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.3062250182200471,
"learning_rate": 6.821086832864139e-06,
"loss": 0.2144,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.2871265807752657,
"learning_rate": 6.77844662204546e-06,
"loss": 0.188,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.31970451210835016,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2112,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.29026557012652804,
"learning_rate": 6.693485996688695e-06,
"loss": 0.1981,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.2596021997538026,
"learning_rate": 6.651166266411801e-06,
"loss": 0.1599,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.3226744236070473,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.2245,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.2847793064786697,
"learning_rate": 6.566849674844711e-06,
"loss": 0.1681,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.3048641054079504,
"learning_rate": 6.524853492628747e-06,
"loss": 0.21,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.2823660471380863,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.1761,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.29849587350943424,
"learning_rate": 6.441187046486648e-06,
"loss": 0.1915,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.2875693758074018,
"learning_rate": 6.399517456398567e-06,
"loss": 0.182,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.2979569039845659,
"learning_rate": 6.357957402727164e-06,
"loss": 0.1968,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.284565109183977,
"learning_rate": 6.316507220191395e-06,
"loss": 0.1674,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.3156466114115207,
"learning_rate": 6.275167242625331e-06,
"loss": 0.2032,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.3243927142657716,
"learning_rate": 6.233937802975471e-06,
"loss": 0.2077,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.33545119928304246,
"learning_rate": 6.192819233298046e-06,
"loss": 0.1969,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.2809887206645705,
"learning_rate": 6.151811864756383e-06,
"loss": 0.1753,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.2656486392682715,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.1778,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.30729710993141873,
"learning_rate": 6.070132051252868e-06,
"loss": 0.2058,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.28115403310953035,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.1731,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.2882922365627405,
"learning_rate": 5.988900993811575e-06,
"loss": 0.1853,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.31048135342854727,
"learning_rate": 5.948454566959363e-06,
"loss": 0.2045,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.29347847074939143,
"learning_rate": 5.908121309322328e-06,
"loss": 0.2006,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.2739576318077019,
"learning_rate": 5.867901545738976e-06,
"loss": 0.1737,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.30444142480886927,
"learning_rate": 5.827795600133774e-06,
"loss": 0.1944,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.29893231263164494,
"learning_rate": 5.787803795514466e-06,
"loss": 0.1871,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.2854602562825505,
"learning_rate": 5.747926453969576e-06,
"loss": 0.196,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.3008248412136313,
"learning_rate": 5.708163896665708e-06,
"loss": 0.1882,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.3042234115651594,
"learning_rate": 5.668516443845047e-06,
"loss": 0.1864,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.34602939807509464,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.1951,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.29492181548124774,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.1808,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.28110583359541996,
"learning_rate": 5.550267900783019e-06,
"loss": 0.1846,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.28112620545717126,
"learning_rate": 5.511084049737623e-06,
"loss": 0.1852,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.2780573155105353,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.1827,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.31989118827293017,
"learning_rate": 5.433066737499948e-06,
"loss": 0.1973,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.30173947072324303,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2137,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.25447586746076944,
"learning_rate": 5.355518704628997e-06,
"loss": 0.1646,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.28809770911547616,
"learning_rate": 5.316921449248731e-06,
"loss": 0.2023,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.25447155570297353,
"learning_rate": 5.278442449364538e-06,
"loss": 0.1814,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.3039890117016232,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2015,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.27167974508025344,
"learning_rate": 5.201840454747822e-06,
"loss": 0.1706,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.2919896568042387,
"learning_rate": 5.163718076957223e-06,
"loss": 0.1887,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.28979190438938723,
"learning_rate": 5.125715188541609e-06,
"loss": 0.1883,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.2929541662782835,
"learning_rate": 5.087832095571021e-06,
"loss": 0.1886,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.26653511993806295,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.1648,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.2889585686341497,
"learning_rate": 5.01242651541854e-06,
"loss": 0.2008,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.2730512461799011,
"learning_rate": 4.974904635542815e-06,
"loss": 0.1882,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.2781127224884192,
"learning_rate": 4.937503765719582e-06,
"loss": 0.1829,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.29218309416945726,
"learning_rate": 4.900224207170299e-06,
"loss": 0.1654,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.3194147318321097,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2067,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.28451262204031147,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.1881,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.2534625168364004,
"learning_rate": 4.789116396710924e-06,
"loss": 0.1752,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.2951388665034386,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2096,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.27567794757903297,
"learning_rate": 4.715656557757473e-06,
"loss": 0.174,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.3218754739920183,
"learning_rate": 4.679111137620442e-06,
"loss": 0.2058,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.2625141028415917,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.1806,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.27518626502017246,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2026,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.2841734726200722,
"learning_rate": 4.570216403561141e-06,
"loss": 0.2049,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.3589457043110656,
"learning_rate": 4.534166308795815e-06,
"loss": 0.1791,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.39596384513274746,
"learning_rate": 4.498240773728859e-06,
"loss": 0.1362,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.3287650535415528,
"learning_rate": 4.462440087699609e-06,
"loss": 0.1221,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.29967411439591723,
"learning_rate": 4.426764539041861e-06,
"loss": 0.1424,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 0.27778102592244514,
"learning_rate": 4.391214415081582e-06,
"loss": 0.1248,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.40607190928621323,
"learning_rate": 4.355790002134579e-06,
"loss": 0.1465,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.41268438093106974,
"learning_rate": 4.320491585504207e-06,
"loss": 0.1235,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.4100640193291354,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.133,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.3405599764955851,
"learning_rate": 4.250273877330691e-06,
"loss": 0.1201,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.32813856782170825,
"learning_rate": 4.215355151311313e-06,
"loss": 0.1333,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.3102510351504104,
"learning_rate": 4.180563552651542e-06,
"loss": 0.1327,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.31918981670876106,
"learning_rate": 4.145899361558147e-06,
"loss": 0.1298,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.3644829031172187,
"learning_rate": 4.111362857211738e-06,
"loss": 0.1499,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.28258614903805207,
"learning_rate": 4.076954317764592e-06,
"loss": 0.1155,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.29425430217510745,
"learning_rate": 4.042674020338335e-06,
"loss": 0.1314,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.29428146851287923,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.1334,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.28186963741661236,
"learning_rate": 3.974499254868674e-06,
"loss": 0.1226,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.30782725810299055,
"learning_rate": 3.940605335895451e-06,
"loss": 0.1428,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.3017183401102668,
"learning_rate": 3.90684075707908e-06,
"loss": 0.1494,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.2581850240152474,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.1126,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.2897310763971961,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.1339,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.28455620452264013,
"learning_rate": 3.806325775702304e-06,
"loss": 0.1421,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.2509460056179919,
"learning_rate": 3.773081266416434e-06,
"loss": 0.1232,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.270819084575095,
"learning_rate": 3.739967446503245e-06,
"loss": 0.1339,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.2804316539217467,
"learning_rate": 3.706984582656894e-06,
"loss": 0.1312,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.24737886444298046,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.1252,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.24447892542107744,
"learning_rate": 3.641412784665648e-06,
"loss": 0.114,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.29569668334185906,
"learning_rate": 3.608824378627005e-06,
"loss": 0.1492,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.26039382291253754,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.1162,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.26796659886696267,
"learning_rate": 3.544043864774269e-06,
"loss": 0.136,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.27443189515482236,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.1226,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.2699612074955308,
"learning_rate": 3.479793485887819e-06,
"loss": 0.1371,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.24662335848294042,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.111,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.24586145781885288,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.116,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.26895715314902124,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.1399,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.29233486328267794,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.1381,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.26218520577316584,
"learning_rate": 3.321500420335e-06,
"loss": 0.1275,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.24266021768997115,
"learning_rate": 3.290243771741275e-06,
"loss": 0.1296,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.2745605116212587,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.139,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.24658396712838926,
"learning_rate": 3.228134459445149e-06,
"loss": 0.1108,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.2813535383495792,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.1373,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.25561439310277395,
"learning_rate": 3.166565459258513e-06,
"loss": 0.1193,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.28204992292062014,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.1489,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.25011757042485444,
"learning_rate": 3.105538754650419e-06,
"loss": 0.1319,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.26253966588369254,
"learning_rate": 3.07522937824712e-06,
"loss": 0.133,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.24221371863159286,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.1171,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.2603851173282285,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.1401,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.2603837687393077,
"learning_rate": 2.985120078631465e-06,
"loss": 0.13,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.26709468144933496,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.1456,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.2538274862775357,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.1193,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.27773396437019,
"learning_rate": 2.896244091927678e-06,
"loss": 0.1363,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.2541197383392023,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.1161,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.250961272561924,
"learning_rate": 2.837681792951994e-06,
"loss": 0.1254,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.2609031546948184,
"learning_rate": 2.808607860255981e-06,
"loss": 0.136,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.2515513535966848,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.126,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.2807888742068889,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.1398,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.2385032064523665,
"learning_rate": 2.722217735873718e-06,
"loss": 0.1092,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.2579769642907314,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.1289,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.26882355436731425,
"learning_rate": 2.665319697490205e-06,
"loss": 0.1297,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.25295257589528325,
"learning_rate": 2.637079980714945e-06,
"loss": 0.1217,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.2739150895106317,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.1322,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.25912489380194964,
"learning_rate": 2.581020289189571e-06,
"loss": 0.1195,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.2741673395973818,
"learning_rate": 2.553200765936501e-06,
"loss": 0.1362,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.2413401370696669,
"learning_rate": 2.525521756805962e-06,
"loss": 0.1147,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.26630958783866227,
"learning_rate": 2.497983484720885e-06,
"loss": 0.1297,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.26291969125366876,
"learning_rate": 2.470586171470728e-06,
"loss": 0.1324,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.26378494295362404,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.1271,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.23919728783132116,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.1144,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.26684723426885637,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.139,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.2743194037357143,
"learning_rate": 2.362410902836978e-06,
"loss": 0.128,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.23999051767575408,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.11,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.24679988354720991,
"learning_rate": 2.309174704446411e-06,
"loss": 0.1156,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.27155453663236306,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.1455,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.26491792073489767,
"learning_rate": 2.256508422807855e-06,
"loss": 0.1382,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.26570743942714775,
"learning_rate": 2.230389531696946e-06,
"loss": 0.1166,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.2554176181484278,
"learning_rate": 2.204413754585857e-06,
"loss": 0.1265,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.24741025710909176,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.1139,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.26284591325539014,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.1236,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.27335147818648414,
"learning_rate": 2.127347193531757e-06,
"loss": 0.1401,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.26883064529528083,
"learning_rate": 2.101945952921942e-06,
"loss": 0.1368,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.23853421212821707,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.1109,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.2727370318547846,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.1329,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.2497770728835698,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.1093,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.2677943872854694,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.1285,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.2606505193123236,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.1173,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.269928462235847,
"learning_rate": 1.95257270594031e-06,
"loss": 0.1265,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.259277247109757,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.1252,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.25858115317568253,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.1289,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.2840322774550965,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.1254,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.24905006887573852,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.1149,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.2618379024740677,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.1328,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.2518341518233997,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.1174,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.27310513254048646,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.134,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.27032373546251937,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.1346,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.24288458789157585,
"learning_rate": 1.738336950434061e-06,
"loss": 0.1136,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.2570702251034397,
"learning_rate": 1.715265841375957e-06,
"loss": 0.1257,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.2662443781196174,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.1273,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.26797290578434907,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.1233,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.23098102614957047,
"learning_rate": 1.646936827927441e-06,
"loss": 0.1108,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.2603599350853924,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.1274,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.26011085568654807,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.1304,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.25995164909524976,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.1279,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.26909597068065433,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.1307,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.2678303682341107,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.121,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.2562235279652154,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.1246,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.2694203877220934,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.1204,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.26386715051798615,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.1458,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.26604644946907513,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.1444,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.23369168784943897,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.1073,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.2563614311000625,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.1205,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.2882340031834417,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.1406,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.2623106241099142,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.126,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.2538155222692844,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.1268,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.2649170780803851,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.1313,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.24424216278693126,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.1156,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.2560416290299474,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.1424,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.251740630384699,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.1233,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.2578490397897112,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.1323,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.25862543489960477,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.1266,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.2647117959267297,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.1293,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.2661612952781214,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.131,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.2544822550218989,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.118,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.2598701226488315,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.1177,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.2613652093317228,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.1295,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.2451563010347704,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.1177,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.25742755452508254,
"learning_rate": 1.092402023876933e-06,
"loss": 0.1276,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.26951924368393837,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.1247,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.28049822960859466,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.1243,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.2600987563614626,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.1219,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.25295215640469537,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.1231,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.26519388319812437,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.1303,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.242708643877295,
"learning_rate": 9.841383857091947e-07,
"loss": 0.1078,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.2711553579079075,
"learning_rate": 9.666296380321616e-07,
"loss": 0.1323,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.2578196384637373,
"learning_rate": 9.492741825034124e-07,
"loss": 0.1289,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.26507357592670955,
"learning_rate": 9.320721589013892e-07,
"loss": 0.1285,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.27129941042350764,
"learning_rate": 9.150237057688339e-07,
"loss": 0.1346,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.26709735293387654,
"learning_rate": 8.981289604116328e-07,
"loss": 0.1268,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.256236663541204,
"learning_rate": 8.813880588977542e-07,
"loss": 0.1202,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.2337779204936602,
"learning_rate": 8.648011360561126e-07,
"loss": 0.1177,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.2914939532619924,
"learning_rate": 8.483683254755037e-07,
"loss": 0.1469,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.2708281410322519,
"learning_rate": 8.320897595035227e-07,
"loss": 0.1274,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.25997021858246977,
"learning_rate": 8.159655692455093e-07,
"loss": 0.1322,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.26019801818181765,
"learning_rate": 7.999958845634648e-07,
"loss": 0.143,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.25855491162601285,
"learning_rate": 7.841808340750478e-07,
"loss": 0.1334,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.26190956258871806,
"learning_rate": 7.685205451524869e-07,
"loss": 0.1133,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.276632072381936,
"learning_rate": 7.530151439216027e-07,
"loss": 0.1221,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.2570694119646122,
"learning_rate": 7.376647552607675e-07,
"loss": 0.1312,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.24432997026382733,
"learning_rate": 7.224695027998963e-07,
"loss": 0.1151,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.24353983269347979,
"learning_rate": 7.07429508919466e-07,
"loss": 0.1222,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.26948207927554296,
"learning_rate": 6.925448947495206e-07,
"loss": 0.1417,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.2551507976402425,
"learning_rate": 6.778157801686936e-07,
"loss": 0.1244,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.2541107685976849,
"learning_rate": 6.632422838032515e-07,
"loss": 0.1158,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.2754875790963283,
"learning_rate": 6.488245230261281e-07,
"loss": 0.1448,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.24853107236887742,
"learning_rate": 6.345626139559868e-07,
"loss": 0.123,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.24697525133813494,
"learning_rate": 6.204566714562866e-07,
"loss": 0.1128,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.2584977510487406,
"learning_rate": 6.06506809134344e-07,
"loss": 0.1303,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.25439249139182657,
"learning_rate": 5.927131393404373e-07,
"loss": 0.1234,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.2592165951680047,
"learning_rate": 5.790757731668817e-07,
"loss": 0.1296,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.2518189800002725,
"learning_rate": 5.655948204471507e-07,
"loss": 0.1304,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.27001250698983703,
"learning_rate": 5.522703897549875e-07,
"loss": 0.1468,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.25190557916659384,
"learning_rate": 5.391025884035239e-07,
"loss": 0.1184,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.2762292291942114,
"learning_rate": 5.260915224444207e-07,
"loss": 0.1499,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.23833381724957228,
"learning_rate": 5.132372966670129e-07,
"loss": 0.1056,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.2704245662003936,
"learning_rate": 5.005400145974704e-07,
"loss": 0.1245,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.2547098951516238,
"learning_rate": 4.879997784979562e-07,
"loss": 0.12,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.25162096422139796,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.1339,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.25531657048902096,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.1252,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.24718135243319847,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.1194,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.2863013918942534,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.1517,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.2426083749016937,
"learning_rate": 4.27657778143431e-07,
"loss": 0.119,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.22895838557814863,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.1118,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.2533033359841471,
"learning_rate": 4.046237371675177e-07,
"loss": 0.1225,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.26240079707084923,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.1254,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.26026805719850005,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.1292,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.2448139891501939,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.1145,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.24784074074466406,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.1176,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.2609601779646533,
"learning_rate": 3.498020255137813e-07,
"loss": 0.1351,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.24870646850927244,
"learning_rate": 3.393121163057811e-07,
"loss": 0.1265,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.2542850157850641,
"learning_rate": 3.289805515751399e-07,
"loss": 0.1229,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.24171322241447227,
"learning_rate": 3.188074145308573e-07,
"loss": 0.118,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.26475052873619753,
"learning_rate": 3.087927871059804e-07,
"loss": 0.1428,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.25318325334786984,
"learning_rate": 2.989367499569418e-07,
"loss": 0.126,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.24857028300028816,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.1176,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.25176464720805947,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.1189,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.26245734947622534,
"learning_rate": 2.703209675663887e-07,
"loss": 0.1367,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.26545586487287304,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.14,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.23109472537362258,
"learning_rate": 2.520381518804471e-07,
"loss": 0.1075,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.2658663303052665,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.1438,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.2526916532900411,
"learning_rate": 2.343915243930317e-07,
"loss": 0.119,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.25428749306683057,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.1185,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.23948945234154093,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.1179,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.2641467753564119,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.1404,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.24467646747499316,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.1155,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.2517328299161815,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.1218,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.24992450681796907,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.1426,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.25802895683516724,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.1393,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.25373746104333555,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.1241,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.2648923215321204,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.1248,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.25140740088004593,
"learning_rate": 1.557203888516745e-07,
"loss": 0.1226,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.25849524396062906,
"learning_rate": 1.487313098987131e-07,
"loss": 0.1377,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.23825372537211095,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.1102,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.27636552194474695,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.1356,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.2538271768343137,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.1155,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.2513505134624802,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.1182,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.2539610676637153,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.1381,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.2595875178765861,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.123,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.2585712344947893,
"learning_rate": 1.042873700093061e-07,
"loss": 0.1277,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.23610808571399003,
"learning_rate": 9.857879404556291e-08,
"loss": 0.1057,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.2735641427233854,
"learning_rate": 9.303050139225722e-08,
"loss": 0.1396,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.26211710742361943,
"learning_rate": 8.76425367345779e-08,
"loss": 0.1361,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.24241183202555797,
"learning_rate": 8.241494346644897e-08,
"loss": 0.1187,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.2425117743837619,
"learning_rate": 7.734776369019204e-08,
"loss": 0.1113,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.25953359183027047,
"learning_rate": 7.244103821617332e-08,
"loss": 0.131,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.25705101342158604,
"learning_rate": 6.769480656248606e-08,
"loss": 0.1342,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.2783677188970859,
"learning_rate": 6.310910695462635e-08,
"loss": 0.1434,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.23536894615794246,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.1067,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.24321681870388476,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.1189,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.28846256976901835,
"learning_rate": 5.031556326567488e-08,
"loss": 0.1547,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.23757926987780167,
"learning_rate": 4.637234823364312e-08,
"loss": 0.1111,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.26173623534339346,
"learning_rate": 4.258983697558838e-08,
"loss": 0.1231,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.25474628390137055,
"learning_rate": 3.896805995533548e-08,
"loss": 0.1191,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.24283676751545663,
"learning_rate": 3.550704634218028e-08,
"loss": 0.1184,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.25368452672205216,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.126,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.27917065783505746,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.1458,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.23812029899313444,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.1163,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.2596356338526681,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.145,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.24093858237607285,
"learning_rate": 2.061435966221881e-08,
"loss": 0.1254,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.25733136003635987,
"learning_rate": 1.811846652477245e-08,
"loss": 0.1323,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.2638747442953262,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.1417,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.2404858304458219,
"learning_rate": 1.360949303591097e-08,
"loss": 0.1083,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.24949097712968396,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.121,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.2624937512774275,
"learning_rate": 9.744388816668172e-09,
"loss": 0.1338,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.2434212504771701,
"learning_rate": 8.05332740472009e-09,
"loss": 0.1151,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.2423322969906,
"learning_rate": 6.523278382872811e-09,
"loss": 0.1265,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.2727610283250997,
"learning_rate": 5.15425407393133e-09,
"loss": 0.1321,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.2537617996233366,
"learning_rate": 3.94626550383137e-09,
"loss": 0.1267,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.24306738166922487,
"learning_rate": 2.899322401546112e-09,
"loss": 0.1231,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.2603517242295158,
"learning_rate": 2.013433199010706e-09,
"loss": 0.1303,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.24096720721756443,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.1116,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.2660291728757945,
"learning_rate": 7.248437353468695e-10,
"loss": 0.1263,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.24622362431011433,
"learning_rate": 3.221538523412449e-10,
"loss": 0.1177,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.249220573122714,
"learning_rate": 8.053862524670663e-11,
"loss": 0.1251,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.2500548152601822,
"learning_rate": 0.0,
"loss": 0.1359,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 2.312324413665575e+18,
"train_loss": 0.2930771143637537,
"train_runtime": 126133.8526,
"train_samples_per_second": 1.253,
"train_steps_per_second": 0.01
}
],
"logging_steps": 1.0,
"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": 2.312324413665575e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}