a1_math_tiger_lab_math / trainer_state.json
ryanmarten's picture
End of training
1642f8f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.988354430379747,
"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": 8.60050859242951,
"learning_rate": 3.2520325203252037e-07,
"loss": 0.969,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 8.423505186968669,
"learning_rate": 6.504065040650407e-07,
"loss": 0.9424,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 8.423789103784614,
"learning_rate": 9.75609756097561e-07,
"loss": 0.9377,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 8.333376293174801,
"learning_rate": 1.3008130081300815e-06,
"loss": 0.9433,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 8.125366772622959,
"learning_rate": 1.6260162601626018e-06,
"loss": 0.9395,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 7.886774993170616,
"learning_rate": 1.951219512195122e-06,
"loss": 0.9527,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 6.516927537329475,
"learning_rate": 2.2764227642276426e-06,
"loss": 0.8405,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 5.953831187827548,
"learning_rate": 2.601626016260163e-06,
"loss": 0.8524,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 3.331974530222765,
"learning_rate": 2.926829268292683e-06,
"loss": 0.8034,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 3.231715510228559,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.8076,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.705301196260624,
"learning_rate": 3.577235772357724e-06,
"loss": 0.7613,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.9941787231305206,
"learning_rate": 3.902439024390244e-06,
"loss": 0.748,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 4.1024513997250285,
"learning_rate": 4.227642276422765e-06,
"loss": 0.7473,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 3.714190245207122,
"learning_rate": 4.552845528455285e-06,
"loss": 0.7006,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 3.469401553292364,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.7168,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 2.8270382188524943,
"learning_rate": 5.203252032520326e-06,
"loss": 0.6694,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.568606682968395,
"learning_rate": 5.528455284552846e-06,
"loss": 0.6312,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.243834593714799,
"learning_rate": 5.853658536585366e-06,
"loss": 0.6176,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.8221410934709408,
"learning_rate": 6.178861788617887e-06,
"loss": 0.5926,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.554483898050658,
"learning_rate": 6.504065040650407e-06,
"loss": 0.5915,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.5757722049418954,
"learning_rate": 6.829268292682928e-06,
"loss": 0.5827,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.907104204955401,
"learning_rate": 7.154471544715448e-06,
"loss": 0.5768,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.8572036397431497,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.5598,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.2940879435831047,
"learning_rate": 7.804878048780489e-06,
"loss": 0.5749,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.09135766893973,
"learning_rate": 8.130081300813009e-06,
"loss": 0.5697,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.4667313806179185,
"learning_rate": 8.45528455284553e-06,
"loss": 0.5435,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.2885910928038973,
"learning_rate": 8.78048780487805e-06,
"loss": 0.5336,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.8948396112091365,
"learning_rate": 9.10569105691057e-06,
"loss": 0.5268,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.8989877896568506,
"learning_rate": 9.43089430894309e-06,
"loss": 0.5303,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.8511966306015549,
"learning_rate": 9.756097560975611e-06,
"loss": 0.5073,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.8546644644236359,
"learning_rate": 1.008130081300813e-05,
"loss": 0.5257,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.711941423415603,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.5341,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.7963428650806776,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.4947,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.904146115275363,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.5055,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.7707682373539683,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.4977,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.7774330705205216,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.4834,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.811633008196302,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.4931,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.6711097120544246,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.4982,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.6883380963905577,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.4874,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.7666065139588681,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.4904,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.6692395280393769,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.4928,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.8128868207086424,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.4696,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.7007013593792586,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.4819,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.680617258765567,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.4776,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.6320671250674865,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.4784,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.7121457067636096,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.4763,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.6232298164806577,
"learning_rate": 1.528455284552846e-05,
"loss": 0.4748,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.748963344455475,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.4608,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.6866920380175642,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.4801,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.7931523189459185,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.4785,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.5768178838331772,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.462,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.6204644622396734,
"learning_rate": 1.691056910569106e-05,
"loss": 0.4541,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.7242874471853671,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.4808,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.7804989685126362,
"learning_rate": 1.75609756097561e-05,
"loss": 0.4816,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.7115641778333726,
"learning_rate": 1.788617886178862e-05,
"loss": 0.4546,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.7673017818414998,
"learning_rate": 1.821138211382114e-05,
"loss": 0.4692,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.7267670016656781,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.4772,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.7653857619039075,
"learning_rate": 1.886178861788618e-05,
"loss": 0.4498,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.6277902700178508,
"learning_rate": 1.91869918699187e-05,
"loss": 0.452,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.690654910146877,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.4663,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.8074890632737652,
"learning_rate": 1.983739837398374e-05,
"loss": 0.4483,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.8298615347680025,
"learning_rate": 2.016260162601626e-05,
"loss": 0.4662,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.8214980972627007,
"learning_rate": 2.048780487804878e-05,
"loss": 0.4607,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.7950759077184323,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.4572,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.8177179985927894,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.4429,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.6334670247750424,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.4757,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.9630381435609074,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.4578,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.7246715308628239,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.4634,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.9222964607530834,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.4433,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.8249784632890622,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.4461,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.8170481875367926,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.4588,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.83028131101939,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.47,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.9133576066145385,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.4362,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.9310292593475188,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.4643,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.7932298247517076,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.4421,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.9487272623960453,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.4402,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.6634823896126352,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.427,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.9301078679219085,
"learning_rate": 2.536585365853659e-05,
"loss": 0.4407,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.8588209652019435,
"learning_rate": 2.569105691056911e-05,
"loss": 0.4607,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.8775096916300468,
"learning_rate": 2.601626016260163e-05,
"loss": 0.4444,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.8104963729650422,
"learning_rate": 2.634146341463415e-05,
"loss": 0.4506,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 0.6499259690641334,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.4467,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.8205791493640167,
"learning_rate": 2.699186991869919e-05,
"loss": 0.4397,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 0.794944475341976,
"learning_rate": 2.731707317073171e-05,
"loss": 0.4609,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.9234059173453114,
"learning_rate": 2.764227642276423e-05,
"loss": 0.4293,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.7385948937062723,
"learning_rate": 2.796747967479675e-05,
"loss": 0.4557,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 0.7960003521234281,
"learning_rate": 2.829268292682927e-05,
"loss": 0.449,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.961901898644528,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.4491,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.801670423473668,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.45,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.7772098037903413,
"learning_rate": 2.926829268292683e-05,
"loss": 0.4539,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.8177826520488204,
"learning_rate": 2.959349593495935e-05,
"loss": 0.4459,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.644520824916339,
"learning_rate": 2.991869918699187e-05,
"loss": 0.4584,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.823014897749393,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.4582,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.7651327945058568,
"learning_rate": 3.056910569105692e-05,
"loss": 0.4637,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.790116532121274,
"learning_rate": 3.089430894308943e-05,
"loss": 0.4427,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.7660439085601293,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.4477,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.6785882879685713,
"learning_rate": 3.154471544715447e-05,
"loss": 0.437,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.7976586901492347,
"learning_rate": 3.186991869918699e-05,
"loss": 0.4539,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.8459179053419076,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.4335,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.7178628375822179,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.4288,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.6866977001741795,
"learning_rate": 3.284552845528456e-05,
"loss": 0.4335,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.699674851008956,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.4195,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.6973720882505813,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.444,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 0.9291479560784863,
"learning_rate": 3.382113821138212e-05,
"loss": 0.4576,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.7535393463583298,
"learning_rate": 3.414634146341463e-05,
"loss": 0.4707,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.6828084610082468,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.4477,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.7890423760139403,
"learning_rate": 3.479674796747968e-05,
"loss": 0.4367,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.6680376784941415,
"learning_rate": 3.51219512195122e-05,
"loss": 0.4617,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.6668920016709673,
"learning_rate": 3.544715447154472e-05,
"loss": 0.4566,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 0.6528682844262823,
"learning_rate": 3.577235772357724e-05,
"loss": 0.4636,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 0.7064596166579137,
"learning_rate": 3.609756097560976e-05,
"loss": 0.4294,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 0.9252746704460473,
"learning_rate": 3.642276422764228e-05,
"loss": 0.4457,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 0.8768446022773372,
"learning_rate": 3.67479674796748e-05,
"loss": 0.4397,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.8719065994426215,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.4555,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.7569306867938328,
"learning_rate": 3.739837398373984e-05,
"loss": 0.4488,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.9406318320197966,
"learning_rate": 3.772357723577236e-05,
"loss": 0.4438,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.7697797254148875,
"learning_rate": 3.804878048780488e-05,
"loss": 0.4503,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 0.9412175612177326,
"learning_rate": 3.83739837398374e-05,
"loss": 0.4358,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.9674665121182286,
"learning_rate": 3.869918699186992e-05,
"loss": 0.436,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.1342527079030693,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.452,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.892716066070595,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.4478,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 0.7350098937595472,
"learning_rate": 3.967479674796748e-05,
"loss": 0.4681,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 0.6614283203261005,
"learning_rate": 4e-05,
"loss": 0.4553,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.7278287758761491,
"learning_rate": 3.999991946137476e-05,
"loss": 0.4331,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.7402479133210104,
"learning_rate": 3.999967784614766e-05,
"loss": 0.4415,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.6759241100889606,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.4342,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.741164636260101,
"learning_rate": 3.999871139496895e-05,
"loss": 0.4504,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.7724147522649781,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.4485,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.7208451095082749,
"learning_rate": 3.999710067759846e-05,
"loss": 0.4492,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.6849305109206492,
"learning_rate": 3.999605373449617e-05,
"loss": 0.4373,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.7055891631915774,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.4361,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.640564090936282,
"learning_rate": 3.999347672161713e-05,
"loss": 0.4402,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.7103584094453766,
"learning_rate": 3.999194667259528e-05,
"loss": 0.4397,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.6518735841049043,
"learning_rate": 3.999025561118334e-05,
"loss": 0.4328,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.691355558522432,
"learning_rate": 3.998840355100086e-05,
"loss": 0.4256,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.702793594166305,
"learning_rate": 3.998639050696409e-05,
"loss": 0.4586,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.5039209227683813,
"learning_rate": 3.998421649528582e-05,
"loss": 0.4277,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.6318483703659126,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.4696,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.6823212582791374,
"learning_rate": 3.997938564033779e-05,
"loss": 0.4256,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.7709504624080329,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.4364,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.5951094274214065,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.4389,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.7024974129264948,
"learning_rate": 3.997093258045973e-05,
"loss": 0.4288,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.611229649001157,
"learning_rate": 3.996779317598936e-05,
"loss": 0.433,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.6762904949732498,
"learning_rate": 3.996449295365782e-05,
"loss": 0.4356,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.7611846640228084,
"learning_rate": 3.996103194004467e-05,
"loss": 0.4239,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.664344732069834,
"learning_rate": 3.995741016302441e-05,
"loss": 0.434,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.691147112765389,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.4565,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.6794060678126981,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.4325,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.6640483191341551,
"learning_rate": 3.994558054968643e-05,
"loss": 0.4597,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.5590047728247275,
"learning_rate": 3.994131602367481e-05,
"loss": 0.4355,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.646910742916938,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.4432,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.7809043963374647,
"learning_rate": 3.993230519343752e-05,
"loss": 0.4457,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.6450749677161336,
"learning_rate": 3.992755896178383e-05,
"loss": 0.4448,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.677354990125155,
"learning_rate": 3.992265223630981e-05,
"loss": 0.4457,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.6804924504346401,
"learning_rate": 3.991758505653355e-05,
"loss": 0.435,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.6215974989971642,
"learning_rate": 3.991235746326543e-05,
"loss": 0.4257,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.945339907961734,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.4557,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.6444869100031965,
"learning_rate": 3.990142120595444e-05,
"loss": 0.4339,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.758626459038366,
"learning_rate": 3.98957126299907e-05,
"loss": 0.4259,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.6606889747923732,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.4343,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.6311457652722658,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.4507,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.5342029001738626,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.4485,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.5847037769799401,
"learning_rate": 3.98712764319087e-05,
"loss": 0.4453,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.5730237675728661,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.4472,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.48543432152634547,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.442,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.6504087772521976,
"learning_rate": 3.985126869010129e-05,
"loss": 0.4585,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.5289749882281461,
"learning_rate": 3.984427961114833e-05,
"loss": 0.4422,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.5822806814014805,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.4235,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.535357059394722,
"learning_rate": 3.982982204151901e-05,
"loss": 0.4387,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.6214884967663846,
"learning_rate": 3.982235366728193e-05,
"loss": 0.4486,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.5561622404383417,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.4394,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.594128542751291,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.4427,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.5503349716727172,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.4378,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.5324598524029474,
"learning_rate": 3.979088432599627e-05,
"loss": 0.4103,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.510743388729473,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.445,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.6099795499158108,
"learning_rate": 3.977419304032111e-05,
"loss": 0.4313,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.5071388729944235,
"learning_rate": 3.976560847560697e-05,
"loss": 0.4247,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.47299670483098205,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.4587,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.5567478790742811,
"learning_rate": 3.974796184811956e-05,
"loss": 0.428,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.6239042806653022,
"learning_rate": 3.973889992746979e-05,
"loss": 0.4307,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.6739528323670255,
"learning_rate": 3.972967903243361e-05,
"loss": 0.4234,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.6616565356251114,
"learning_rate": 3.972029923727486e-05,
"loss": 0.4388,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.6409931042591177,
"learning_rate": 3.971076061753709e-05,
"loss": 0.4371,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.7673737338767278,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.4301,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.7834551402192652,
"learning_rate": 3.969120721289402e-05,
"loss": 0.427,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.5772781128118456,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.4346,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.7648828136648163,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.4016,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.5861005461139777,
"learning_rate": 3.966068788369422e-05,
"loss": 0.4222,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.5985355700541802,
"learning_rate": 3.965019797448622e-05,
"loss": 0.4354,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.5425722570915043,
"learning_rate": 3.963954980528515e-05,
"loss": 0.4196,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.537295139906245,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.437,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.5901195585204905,
"learning_rate": 3.961777903121329e-05,
"loss": 0.4138,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.5223833127735348,
"learning_rate": 3.960665660168131e-05,
"loss": 0.4305,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.6861763328430944,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.4274,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.5225360120647239,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.4293,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.659000587625896,
"learning_rate": 3.957234222185657e-05,
"loss": 0.4174,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.5189082110820011,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.4433,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.6592746943777263,
"learning_rate": 3.954867765033605e-05,
"loss": 0.4341,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.4739728989314021,
"learning_rate": 3.953660915306728e-05,
"loss": 0.4434,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.6410455756977287,
"learning_rate": 3.952438331063419e-05,
"loss": 0.4144,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.5080499331603797,
"learning_rate": 3.951200022150205e-05,
"loss": 0.4251,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.6533284797663386,
"learning_rate": 3.949945998540253e-05,
"loss": 0.4088,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.4503736486689305,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.418,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.6279282798767103,
"learning_rate": 3.947390847755559e-05,
"loss": 0.4573,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.6007575638024936,
"learning_rate": 3.946089741159648e-05,
"loss": 0.4284,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.579586983029723,
"learning_rate": 3.944772961024501e-05,
"loss": 0.4394,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.5728909129444062,
"learning_rate": 3.943440517955285e-05,
"loss": 0.4121,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.5194913300009117,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.4291,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.5939874032073263,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.4202,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.5027845317639019,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.4509,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.660136263442682,
"learning_rate": 3.937954332854371e-05,
"loss": 0.4377,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.4463457255862533,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.4334,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.6031784396505738,
"learning_rate": 3.935117547697387e-05,
"loss": 0.4255,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.49338195916197075,
"learning_rate": 3.933675771619675e-05,
"loss": 0.4359,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.5001516665691574,
"learning_rate": 3.932218421983131e-05,
"loss": 0.446,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.5211872620880744,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.4466,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.47557692265437085,
"learning_rate": 3.929257049108054e-05,
"loss": 0.4358,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.4765040527936316,
"learning_rate": 3.927753049720011e-05,
"loss": 0.4354,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.46310732849159575,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.4253,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.4698157460869811,
"learning_rate": 3.92469848560784e-05,
"loss": 0.4201,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.5176393360791873,
"learning_rate": 3.923147945484751e-05,
"loss": 0.436,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.47746528233027774,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.4146,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.5689490737426188,
"learning_rate": 3.920000411543654e-05,
"loss": 0.4187,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.5846281123728453,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.4392,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.5215120471010646,
"learning_rate": 3.916791024049648e-05,
"loss": 0.4389,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.571214545514271,
"learning_rate": 3.91516316745245e-05,
"loss": 0.4393,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.6160924402410662,
"learning_rate": 3.913519886394389e-05,
"loss": 0.4221,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.4953573724208773,
"learning_rate": 3.911861194110225e-05,
"loss": 0.4542,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.5109388724856437,
"learning_rate": 3.910187103958837e-05,
"loss": 0.3994,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.47788300461093164,
"learning_rate": 3.908497629423117e-05,
"loss": 0.4255,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.5708326207610074,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.421,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.5098685085659405,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.4172,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.6412551678390915,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.3965,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.48215424858758393,
"learning_rate": 3.901586161429081e-05,
"loss": 0.4288,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.555927209843123,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.4216,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.53321208662847,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.4284,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.5049577658830666,
"learning_rate": 3.896241703466667e-05,
"loss": 0.4241,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.563287295763608,
"learning_rate": 3.894429654085585e-05,
"loss": 0.413,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.522917359045758,
"learning_rate": 3.892602347228505e-05,
"loss": 0.4336,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.5321845194266547,
"learning_rate": 3.890759797612307e-05,
"loss": 0.4506,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.6007658444571387,
"learning_rate": 3.888902020076632e-05,
"loss": 0.424,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.4473399931086473,
"learning_rate": 3.887029029583764e-05,
"loss": 0.4338,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.6297777106221042,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.4271,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.4416011135138063,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.4179,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.5419732873274982,
"learning_rate": 3.881318931821972e-05,
"loss": 0.4178,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.49720532499952114,
"learning_rate": 3.879385241571817e-05,
"loss": 0.4103,
"step": 246
},
{
"epoch": 1.0020253164556963,
"grad_norm": 0.5419038447795362,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.3997,
"step": 247
},
{
"epoch": 1.0060759493670886,
"grad_norm": 0.5222351795042967,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.3466,
"step": 248
},
{
"epoch": 1.010126582278481,
"grad_norm": 0.49612353640755147,
"learning_rate": 3.873493415863256e-05,
"loss": 0.3485,
"step": 249
},
{
"epoch": 1.0141772151898734,
"grad_norm": 0.5426089288404474,
"learning_rate": 3.871499275032111e-05,
"loss": 0.3559,
"step": 250
},
{
"epoch": 1.0182278481012659,
"grad_norm": 0.5676204870581877,
"learning_rate": 3.869490061403091e-05,
"loss": 0.3341,
"step": 251
},
{
"epoch": 1.0222784810126582,
"grad_norm": 0.5572433155250933,
"learning_rate": 3.867465791158124e-05,
"loss": 0.3315,
"step": 252
},
{
"epoch": 1.0263291139240507,
"grad_norm": 0.5100950863399575,
"learning_rate": 3.865426480600407e-05,
"loss": 0.3262,
"step": 253
},
{
"epoch": 1.030379746835443,
"grad_norm": 0.5122908559202103,
"learning_rate": 3.863372146154264e-05,
"loss": 0.3405,
"step": 254
},
{
"epoch": 1.0344303797468355,
"grad_norm": 0.5943294829189683,
"learning_rate": 3.861302804365024e-05,
"loss": 0.3529,
"step": 255
},
{
"epoch": 1.0384810126582278,
"grad_norm": 0.5133871727301151,
"learning_rate": 3.85921847189888e-05,
"loss": 0.3324,
"step": 256
},
{
"epoch": 1.0425316455696203,
"grad_norm": 0.5637331251652807,
"learning_rate": 3.85711916554276e-05,
"loss": 0.3424,
"step": 257
},
{
"epoch": 1.0465822784810126,
"grad_norm": 0.5433735456037284,
"learning_rate": 3.85500490220419e-05,
"loss": 0.3359,
"step": 258
},
{
"epoch": 1.0506329113924051,
"grad_norm": 0.465175321123351,
"learning_rate": 3.852875698911154e-05,
"loss": 0.3432,
"step": 259
},
{
"epoch": 1.0546835443037974,
"grad_norm": 0.5278732982030209,
"learning_rate": 3.850731572811963e-05,
"loss": 0.3442,
"step": 260
},
{
"epoch": 1.05873417721519,
"grad_norm": 0.5044337950359139,
"learning_rate": 3.848572541175116e-05,
"loss": 0.3332,
"step": 261
},
{
"epoch": 1.0627848101265822,
"grad_norm": 0.5027260170610465,
"learning_rate": 3.846398621389154e-05,
"loss": 0.346,
"step": 262
},
{
"epoch": 1.0668354430379747,
"grad_norm": 0.4429004498299103,
"learning_rate": 3.84420983096253e-05,
"loss": 0.3295,
"step": 263
},
{
"epoch": 1.070886075949367,
"grad_norm": 0.5349128221708259,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.3542,
"step": 264
},
{
"epoch": 1.0749367088607595,
"grad_norm": 0.4709963359885837,
"learning_rate": 3.839787708819787e-05,
"loss": 0.3573,
"step": 265
},
{
"epoch": 1.0789873417721518,
"grad_norm": 0.49016415577732697,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.3463,
"step": 266
},
{
"epoch": 1.0830379746835443,
"grad_norm": 0.6063859771858053,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.3459,
"step": 267
},
{
"epoch": 1.0870886075949366,
"grad_norm": 0.43423722743270077,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.3268,
"step": 268
},
{
"epoch": 1.0911392405063292,
"grad_norm": 0.5835279900568977,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.3405,
"step": 269
},
{
"epoch": 1.0951898734177214,
"grad_norm": 0.5060047891808482,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.3444,
"step": 270
},
{
"epoch": 1.099240506329114,
"grad_norm": 0.4848328852049688,
"learning_rate": 3.826166304956594e-05,
"loss": 0.3238,
"step": 271
},
{
"epoch": 1.1032911392405063,
"grad_norm": 0.5367525224643638,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.3651,
"step": 272
},
{
"epoch": 1.1073417721518988,
"grad_norm": 0.4545537312314633,
"learning_rate": 3.821507978767479e-05,
"loss": 0.315,
"step": 273
},
{
"epoch": 1.111392405063291,
"grad_norm": 0.5068980490215698,
"learning_rate": 3.819156801001693e-05,
"loss": 0.358,
"step": 274
},
{
"epoch": 1.1154430379746836,
"grad_norm": 0.4562876401401247,
"learning_rate": 3.816790971997121e-05,
"loss": 0.3344,
"step": 275
},
{
"epoch": 1.1194936708860759,
"grad_norm": 0.5171467723474229,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.3338,
"step": 276
},
{
"epoch": 1.1235443037974684,
"grad_norm": 0.4645827364499015,
"learning_rate": 3.81201543660571e-05,
"loss": 0.3338,
"step": 277
},
{
"epoch": 1.1275949367088607,
"grad_norm": 0.5183143146782586,
"learning_rate": 3.809605768680377e-05,
"loss": 0.3319,
"step": 278
},
{
"epoch": 1.1316455696202532,
"grad_norm": 0.4711704893932822,
"learning_rate": 3.807181526438958e-05,
"loss": 0.341,
"step": 279
},
{
"epoch": 1.1356962025316455,
"grad_norm": 0.6647820422700838,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.3495,
"step": 280
},
{
"epoch": 1.139746835443038,
"grad_norm": 0.5191788015791382,
"learning_rate": 3.802289397223145e-05,
"loss": 0.3312,
"step": 281
},
{
"epoch": 1.1437974683544303,
"grad_norm": 0.72284115863236,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.3539,
"step": 282
},
{
"epoch": 1.1478481012658228,
"grad_norm": 0.4985203971118566,
"learning_rate": 3.797339206560096e-05,
"loss": 0.3631,
"step": 283
},
{
"epoch": 1.1518987341772151,
"grad_norm": 0.5887596449929491,
"learning_rate": 3.794842387948027e-05,
"loss": 0.3312,
"step": 284
},
{
"epoch": 1.1559493670886076,
"grad_norm": 0.5477637016674187,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.3446,
"step": 285
},
{
"epoch": 1.16,
"grad_norm": 0.5076129800993857,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.3552,
"step": 286
},
{
"epoch": 1.1640506329113924,
"grad_norm": 0.5209228572705362,
"learning_rate": 3.787265280646825e-05,
"loss": 0.3231,
"step": 287
},
{
"epoch": 1.1681012658227847,
"grad_norm": 0.455329499799213,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.3405,
"step": 288
},
{
"epoch": 1.1721518987341772,
"grad_norm": 0.489984130434488,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.3461,
"step": 289
},
{
"epoch": 1.1762025316455695,
"grad_norm": 0.47956358562579,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.3316,
"step": 290
},
{
"epoch": 1.180253164556962,
"grad_norm": 0.431452462371144,
"learning_rate": 3.776961046830306e-05,
"loss": 0.3349,
"step": 291
},
{
"epoch": 1.1843037974683543,
"grad_norm": 0.5035595138297086,
"learning_rate": 3.774349157719215e-05,
"loss": 0.3486,
"step": 292
},
{
"epoch": 1.1883544303797469,
"grad_norm": 0.4274156595571704,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.3455,
"step": 293
},
{
"epoch": 1.1924050632911392,
"grad_norm": 0.5591066767501923,
"learning_rate": 3.769082529555359e-05,
"loss": 0.3472,
"step": 294
},
{
"epoch": 1.1964556962025317,
"grad_norm": 0.4815606521888572,
"learning_rate": 3.766427832919294e-05,
"loss": 0.3417,
"step": 295
},
{
"epoch": 1.200506329113924,
"grad_norm": 0.5494903096938466,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.3368,
"step": 296
},
{
"epoch": 1.2045569620253165,
"grad_norm": 0.4864755590483048,
"learning_rate": 3.761075781441526e-05,
"loss": 0.3265,
"step": 297
},
{
"epoch": 1.2086075949367088,
"grad_norm": 0.6113576020110734,
"learning_rate": 3.75837846970451e-05,
"loss": 0.3421,
"step": 298
},
{
"epoch": 1.2126582278481013,
"grad_norm": 0.5439540632619371,
"learning_rate": 3.755666996229032e-05,
"loss": 0.3383,
"step": 299
},
{
"epoch": 1.2167088607594936,
"grad_norm": 0.585798211608132,
"learning_rate": 3.752941382852927e-05,
"loss": 0.3443,
"step": 300
},
{
"epoch": 1.220759493670886,
"grad_norm": 0.6202561820731924,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.3367,
"step": 301
},
{
"epoch": 1.2248101265822784,
"grad_norm": 0.4380522754817675,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.3517,
"step": 302
},
{
"epoch": 1.228860759493671,
"grad_norm": 0.5277787426983394,
"learning_rate": 3.744679923406351e-05,
"loss": 0.3493,
"step": 303
},
{
"epoch": 1.2329113924050632,
"grad_norm": 0.4964771513470393,
"learning_rate": 3.741897971081043e-05,
"loss": 0.333,
"step": 304
},
{
"epoch": 1.2369620253164557,
"grad_norm": 0.43718244013980334,
"learning_rate": 3.739101989748946e-05,
"loss": 0.3607,
"step": 305
},
{
"epoch": 1.241012658227848,
"grad_norm": 0.5269067237553792,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.345,
"step": 306
},
{
"epoch": 1.2450632911392405,
"grad_norm": 0.455581482231718,
"learning_rate": 3.73346803025098e-05,
"loss": 0.3543,
"step": 307
},
{
"epoch": 1.2491139240506328,
"grad_norm": 0.4859358773912525,
"learning_rate": 3.730630097460247e-05,
"loss": 0.345,
"step": 308
},
{
"epoch": 1.2531645569620253,
"grad_norm": 0.42313510456710657,
"learning_rate": 3.727778226412628e-05,
"loss": 0.3291,
"step": 309
},
{
"epoch": 1.2572151898734178,
"grad_norm": 0.4936030557682094,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.335,
"step": 310
},
{
"epoch": 1.2612658227848101,
"grad_norm": 0.4931861553197091,
"learning_rate": 3.722032761533114e-05,
"loss": 0.3243,
"step": 311
},
{
"epoch": 1.2653164556962024,
"grad_norm": 0.5820988598890653,
"learning_rate": 3.719139213974403e-05,
"loss": 0.3544,
"step": 312
},
{
"epoch": 1.269367088607595,
"grad_norm": 0.4755799346084926,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.3394,
"step": 313
},
{
"epoch": 1.2734177215189875,
"grad_norm": 0.5721641021554791,
"learning_rate": 3.713310605140055e-05,
"loss": 0.3418,
"step": 314
},
{
"epoch": 1.2774683544303798,
"grad_norm": 0.5043061478968807,
"learning_rate": 3.710375590807233e-05,
"loss": 0.3359,
"step": 315
},
{
"epoch": 1.281518987341772,
"grad_norm": 0.5323224896907358,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.3488,
"step": 316
},
{
"epoch": 1.2855696202531646,
"grad_norm": 0.5201920341548326,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.3304,
"step": 317
},
{
"epoch": 1.289620253164557,
"grad_norm": 0.488254277834751,
"learning_rate": 3.701487992136854e-05,
"loss": 0.3252,
"step": 318
},
{
"epoch": 1.2936708860759494,
"grad_norm": 0.5500821805421587,
"learning_rate": 3.69849802022222e-05,
"loss": 0.3433,
"step": 319
},
{
"epoch": 1.2977215189873417,
"grad_norm": 0.45156010274915204,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.3395,
"step": 320
},
{
"epoch": 1.3017721518987342,
"grad_norm": 0.48526794249779276,
"learning_rate": 3.692477062175289e-05,
"loss": 0.3392,
"step": 321
},
{
"epoch": 1.3058227848101267,
"grad_norm": 0.4682754280669763,
"learning_rate": 3.689446124534958e-05,
"loss": 0.3488,
"step": 322
},
{
"epoch": 1.309873417721519,
"grad_norm": 0.4593607484163409,
"learning_rate": 3.686401580327799e-05,
"loss": 0.3238,
"step": 323
},
{
"epoch": 1.3139240506329113,
"grad_norm": 0.5108078826320113,
"learning_rate": 3.683343454074149e-05,
"loss": 0.3494,
"step": 324
},
{
"epoch": 1.3179746835443038,
"grad_norm": 0.4435586660299157,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.3388,
"step": 325
},
{
"epoch": 1.3220253164556963,
"grad_norm": 0.4747487492540269,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.3357,
"step": 326
},
{
"epoch": 1.3260759493670886,
"grad_norm": 0.5102607207092316,
"learning_rate": 3.674087829877297e-05,
"loss": 0.346,
"step": 327
},
{
"epoch": 1.330126582278481,
"grad_norm": 0.4705442622402744,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.3278,
"step": 328
},
{
"epoch": 1.3341772151898734,
"grad_norm": 0.4730288051912772,
"learning_rate": 3.667849957966501e-05,
"loss": 0.338,
"step": 329
},
{
"epoch": 1.338227848101266,
"grad_norm": 0.4981545418490215,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.3343,
"step": 330
},
{
"epoch": 1.3422784810126582,
"grad_norm": 0.46869682229018955,
"learning_rate": 3.661558355626795e-05,
"loss": 0.3634,
"step": 331
},
{
"epoch": 1.3463291139240505,
"grad_norm": 0.4627119743476647,
"learning_rate": 3.658392468818163e-05,
"loss": 0.3573,
"step": 332
},
{
"epoch": 1.350379746835443,
"grad_norm": 0.4487706126102673,
"learning_rate": 3.655213225544574e-05,
"loss": 0.3399,
"step": 333
},
{
"epoch": 1.3544303797468356,
"grad_norm": 0.42863977541270265,
"learning_rate": 3.652020651411218e-05,
"loss": 0.3344,
"step": 334
},
{
"epoch": 1.3584810126582278,
"grad_norm": 0.4979304005053725,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.3359,
"step": 335
},
{
"epoch": 1.3625316455696201,
"grad_norm": 0.41289220874040117,
"learning_rate": 3.645595613522574e-05,
"loss": 0.3506,
"step": 336
},
{
"epoch": 1.3665822784810127,
"grad_norm": 0.4545155289876307,
"learning_rate": 3.642363201513657e-05,
"loss": 0.3172,
"step": 337
},
{
"epoch": 1.3706329113924052,
"grad_norm": 0.48420474245242145,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.3344,
"step": 338
},
{
"epoch": 1.3746835443037975,
"grad_norm": 0.3940819761194091,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.3133,
"step": 339
},
{
"epoch": 1.3787341772151898,
"grad_norm": 0.43172833694720353,
"learning_rate": 3.632586705948318e-05,
"loss": 0.3164,
"step": 340
},
{
"epoch": 1.3827848101265823,
"grad_norm": 0.4740384661927722,
"learning_rate": 3.629301541734311e-05,
"loss": 0.3347,
"step": 341
},
{
"epoch": 1.3868354430379748,
"grad_norm": 0.4607314637285381,
"learning_rate": 3.626003255349676e-05,
"loss": 0.3366,
"step": 342
},
{
"epoch": 1.390886075949367,
"grad_norm": 0.47776775898791735,
"learning_rate": 3.622691873358357e-05,
"loss": 0.3617,
"step": 343
},
{
"epoch": 1.3949367088607594,
"grad_norm": 0.4650571143372117,
"learning_rate": 3.61936742242977e-05,
"loss": 0.3265,
"step": 344
},
{
"epoch": 1.398987341772152,
"grad_norm": 0.43776704116241977,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.3314,
"step": 345
},
{
"epoch": 1.4030379746835444,
"grad_norm": 0.53722899055503,
"learning_rate": 3.612679420964516e-05,
"loss": 0.3868,
"step": 346
},
{
"epoch": 1.4070886075949367,
"grad_norm": 0.4574233069664573,
"learning_rate": 3.609315924292092e-05,
"loss": 0.3707,
"step": 347
},
{
"epoch": 1.411139240506329,
"grad_norm": 0.47559684796327534,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.3249,
"step": 348
},
{
"epoch": 1.4151898734177215,
"grad_norm": 0.5776632508710674,
"learning_rate": 3.602550074513133e-05,
"loss": 0.3264,
"step": 349
},
{
"epoch": 1.419240506329114,
"grad_norm": 0.4650152370111365,
"learning_rate": 3.599147775897822e-05,
"loss": 0.3502,
"step": 350
},
{
"epoch": 1.4232911392405063,
"grad_norm": 0.4590811874722296,
"learning_rate": 3.595732597966167e-05,
"loss": 0.3624,
"step": 351
},
{
"epoch": 1.4273417721518986,
"grad_norm": 0.5243081152507931,
"learning_rate": 3.592304568223542e-05,
"loss": 0.348,
"step": 352
},
{
"epoch": 1.4313924050632911,
"grad_norm": 0.42990496727758737,
"learning_rate": 3.588863714278826e-05,
"loss": 0.3285,
"step": 353
},
{
"epoch": 1.4354430379746836,
"grad_norm": 0.4268688887869658,
"learning_rate": 3.585410063844186e-05,
"loss": 0.345,
"step": 354
},
{
"epoch": 1.439493670886076,
"grad_norm": 0.5008507829040156,
"learning_rate": 3.581943644734846e-05,
"loss": 0.3336,
"step": 355
},
{
"epoch": 1.4435443037974682,
"grad_norm": 0.49019135145659387,
"learning_rate": 3.578464484868869e-05,
"loss": 0.3274,
"step": 356
},
{
"epoch": 1.4475949367088607,
"grad_norm": 0.5690164842964275,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.376,
"step": 357
},
{
"epoch": 1.4516455696202533,
"grad_norm": 0.4715098711903623,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.3471,
"step": 358
},
{
"epoch": 1.4556962025316456,
"grad_norm": 0.49318796194138703,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.3399,
"step": 359
},
{
"epoch": 1.4597468354430378,
"grad_norm": 0.45474847136632557,
"learning_rate": 3.564420999786543e-05,
"loss": 0.33,
"step": 360
},
{
"epoch": 1.4637974683544304,
"grad_norm": 0.4349976173012787,
"learning_rate": 3.560878558491842e-05,
"loss": 0.3511,
"step": 361
},
{
"epoch": 1.4678481012658229,
"grad_norm": 0.552632691535068,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.3457,
"step": 362
},
{
"epoch": 1.4718987341772152,
"grad_norm": 0.45962032651318907,
"learning_rate": 3.553755991230039e-05,
"loss": 0.3517,
"step": 363
},
{
"epoch": 1.4759493670886075,
"grad_norm": 0.47566647117710764,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.3564,
"step": 364
},
{
"epoch": 1.48,
"grad_norm": 0.5180887552634797,
"learning_rate": 3.546583369120419e-05,
"loss": 0.3269,
"step": 365
},
{
"epoch": 1.4840506329113925,
"grad_norm": 0.39690034720757134,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.3406,
"step": 366
},
{
"epoch": 1.4881012658227848,
"grad_norm": 0.5808319265217144,
"learning_rate": 3.539360923231766e-05,
"loss": 0.3369,
"step": 367
},
{
"epoch": 1.492151898734177,
"grad_norm": 0.5113970047872217,
"learning_rate": 3.535731089018394e-05,
"loss": 0.3412,
"step": 368
},
{
"epoch": 1.4962025316455696,
"grad_norm": 0.5113335497046099,
"learning_rate": 3.532088886237956e-05,
"loss": 0.3184,
"step": 369
},
{
"epoch": 1.5002531645569621,
"grad_norm": 0.5350306048680086,
"learning_rate": 3.528434344224253e-05,
"loss": 0.3516,
"step": 370
},
{
"epoch": 1.5043037974683544,
"grad_norm": 0.4404685479544306,
"learning_rate": 3.524767492410464e-05,
"loss": 0.3491,
"step": 371
},
{
"epoch": 1.5083544303797467,
"grad_norm": 0.48299279596982975,
"learning_rate": 3.521088360328908e-05,
"loss": 0.3401,
"step": 372
},
{
"epoch": 1.5124050632911392,
"grad_norm": 0.43324798152233696,
"learning_rate": 3.517396977610811e-05,
"loss": 0.3481,
"step": 373
},
{
"epoch": 1.5164556962025317,
"grad_norm": 0.5165352867706462,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.3327,
"step": 374
},
{
"epoch": 1.520506329113924,
"grad_norm": 0.44779867129891665,
"learning_rate": 3.509977579282971e-05,
"loss": 0.3223,
"step": 375
},
{
"epoch": 1.5245569620253163,
"grad_norm": 0.6277205232823583,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.3474,
"step": 376
},
{
"epoch": 1.5286075949367088,
"grad_norm": 0.4703537464154026,
"learning_rate": 3.502509536445719e-05,
"loss": 0.3363,
"step": 377
},
{
"epoch": 1.5326582278481014,
"grad_norm": 0.6685602802689383,
"learning_rate": 3.498757348458147e-05,
"loss": 0.3482,
"step": 378
},
{
"epoch": 1.5367088607594936,
"grad_norm": 0.5295036258962079,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.3507,
"step": 379
},
{
"epoch": 1.540759493670886,
"grad_norm": 0.5948599593958653,
"learning_rate": 3.491216790442899e-05,
"loss": 0.3628,
"step": 380
},
{
"epoch": 1.5448101265822785,
"grad_norm": 0.5150117992625173,
"learning_rate": 3.487428481145839e-05,
"loss": 0.3442,
"step": 381
},
{
"epoch": 1.548860759493671,
"grad_norm": 0.5344983126974742,
"learning_rate": 3.483628192304278e-05,
"loss": 0.3566,
"step": 382
},
{
"epoch": 1.5529113924050633,
"grad_norm": 0.49029696236035836,
"learning_rate": 3.479815954525219e-05,
"loss": 0.3321,
"step": 383
},
{
"epoch": 1.5569620253164556,
"grad_norm": 0.5659607274383949,
"learning_rate": 3.475991798511899e-05,
"loss": 0.3448,
"step": 384
},
{
"epoch": 1.561012658227848,
"grad_norm": 0.5555834776087583,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.3696,
"step": 385
},
{
"epoch": 1.5650632911392406,
"grad_norm": 0.4336914179646585,
"learning_rate": 3.468307855075128e-05,
"loss": 0.3224,
"step": 386
},
{
"epoch": 1.5691139240506329,
"grad_norm": 0.5567806418532919,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.3375,
"step": 387
},
{
"epoch": 1.5731645569620252,
"grad_norm": 0.5182802446210885,
"learning_rate": 3.460576609535163e-05,
"loss": 0.3634,
"step": 388
},
{
"epoch": 1.5772151898734177,
"grad_norm": 0.4546638420134034,
"learning_rate": 3.456693326250006e-05,
"loss": 0.3492,
"step": 389
},
{
"epoch": 1.5812658227848102,
"grad_norm": 0.496837511593122,
"learning_rate": 3.452798310957058e-05,
"loss": 0.332,
"step": 390
},
{
"epoch": 1.5853164556962025,
"grad_norm": 0.4386734791881964,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.3423,
"step": 391
},
{
"epoch": 1.5893670886075948,
"grad_norm": 0.40361896196628283,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.3526,
"step": 392
},
{
"epoch": 1.5934177215189873,
"grad_norm": 0.4643379831158051,
"learning_rate": 3.441043187201574e-05,
"loss": 0.3586,
"step": 393
},
{
"epoch": 1.5974683544303798,
"grad_norm": 0.4497316403866711,
"learning_rate": 3.437101558517728e-05,
"loss": 0.3534,
"step": 394
},
{
"epoch": 1.6015189873417721,
"grad_norm": 0.43260131736848206,
"learning_rate": 3.433148355615496e-05,
"loss": 0.3667,
"step": 395
},
{
"epoch": 1.6055696202531644,
"grad_norm": 0.4441369516065414,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.3424,
"step": 396
},
{
"epoch": 1.609620253164557,
"grad_norm": 0.4518236476050238,
"learning_rate": 3.425207354603043e-05,
"loss": 0.3512,
"step": 397
},
{
"epoch": 1.6136708860759494,
"grad_norm": 0.49862616312042674,
"learning_rate": 3.421219620448553e-05,
"loss": 0.3498,
"step": 398
},
{
"epoch": 1.6177215189873417,
"grad_norm": 0.45755848064562843,
"learning_rate": 3.417220439986623e-05,
"loss": 0.358,
"step": 399
},
{
"epoch": 1.621772151898734,
"grad_norm": 0.4461400677070297,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.346,
"step": 400
},
{
"epoch": 1.6258227848101265,
"grad_norm": 0.47025219457129375,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.3545,
"step": 401
},
{
"epoch": 1.629873417721519,
"grad_norm": 0.422345351887796,
"learning_rate": 3.405154543304065e-05,
"loss": 0.3244,
"step": 402
},
{
"epoch": 1.6339240506329114,
"grad_norm": 0.4876898489787585,
"learning_rate": 3.401109900618843e-05,
"loss": 0.35,
"step": 403
},
{
"epoch": 1.6379746835443036,
"grad_norm": 0.4666322768545651,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.3504,
"step": 404
},
{
"epoch": 1.6420253164556962,
"grad_norm": 0.48523942898437433,
"learning_rate": 3.392986794874714e-05,
"loss": 0.3163,
"step": 405
},
{
"epoch": 1.6460759493670887,
"grad_norm": 0.4134871460694911,
"learning_rate": 3.388908397238184e-05,
"loss": 0.3398,
"step": 406
},
{
"epoch": 1.650126582278481,
"grad_norm": 0.5651484399657075,
"learning_rate": 3.384818813524362e-05,
"loss": 0.3558,
"step": 407
},
{
"epoch": 1.6541772151898733,
"grad_norm": 0.42747303190523184,
"learning_rate": 3.380718076670195e-05,
"loss": 0.3257,
"step": 408
},
{
"epoch": 1.6582278481012658,
"grad_norm": 0.5353303205347961,
"learning_rate": 3.376606219702454e-05,
"loss": 0.3494,
"step": 409
},
{
"epoch": 1.6622784810126583,
"grad_norm": 0.41523248262664014,
"learning_rate": 3.372483275737468e-05,
"loss": 0.3485,
"step": 410
},
{
"epoch": 1.6663291139240506,
"grad_norm": 0.4094140345762293,
"learning_rate": 3.368349277980861e-05,
"loss": 0.341,
"step": 411
},
{
"epoch": 1.6703797468354429,
"grad_norm": 0.38868509872695123,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.3482,
"step": 412
},
{
"epoch": 1.6744303797468354,
"grad_norm": 0.39443287320036574,
"learning_rate": 3.360048254360144e-05,
"loss": 0.3397,
"step": 413
},
{
"epoch": 1.678481012658228,
"grad_norm": 0.4250026126972739,
"learning_rate": 3.355881295351336e-05,
"loss": 0.3527,
"step": 414
},
{
"epoch": 1.6825316455696202,
"grad_norm": 0.4137545299750722,
"learning_rate": 3.351703416260975e-05,
"loss": 0.3544,
"step": 415
},
{
"epoch": 1.6865822784810125,
"grad_norm": 0.4219491426384593,
"learning_rate": 3.347514650737126e-05,
"loss": 0.3439,
"step": 416
},
{
"epoch": 1.690632911392405,
"grad_norm": 0.4623000638970958,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.3352,
"step": 417
},
{
"epoch": 1.6946835443037975,
"grad_norm": 0.43643006804174744,
"learning_rate": 3.339104595419334e-05,
"loss": 0.3391,
"step": 418
},
{
"epoch": 1.6987341772151898,
"grad_norm": 0.4332093639640789,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.3509,
"step": 419
},
{
"epoch": 1.7027848101265821,
"grad_norm": 0.42678609222525493,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.3262,
"step": 420
},
{
"epoch": 1.7068354430379746,
"grad_norm": 0.4342190057825798,
"learning_rate": 3.326408710419996e-05,
"loss": 0.3457,
"step": 421
},
{
"epoch": 1.7108860759493671,
"grad_norm": 0.4327013552926118,
"learning_rate": 3.322155337795454e-05,
"loss": 0.3192,
"step": 422
},
{
"epoch": 1.7149367088607594,
"grad_norm": 0.43498519605253916,
"learning_rate": 3.317891316713587e-05,
"loss": 0.3338,
"step": 423
},
{
"epoch": 1.7189873417721517,
"grad_norm": 0.4367957872060851,
"learning_rate": 3.313616681516231e-05,
"loss": 0.3339,
"step": 424
},
{
"epoch": 1.7230379746835442,
"grad_norm": 0.3921957559339082,
"learning_rate": 3.309331466630713e-05,
"loss": 0.3375,
"step": 425
},
{
"epoch": 1.7270886075949368,
"grad_norm": 0.42332525181710096,
"learning_rate": 3.305035706569563e-05,
"loss": 0.3344,
"step": 426
},
{
"epoch": 1.7311392405063293,
"grad_norm": 0.3917721720842058,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.3473,
"step": 427
},
{
"epoch": 1.7351898734177216,
"grad_norm": 0.45845075503248084,
"learning_rate": 3.296412689394864e-05,
"loss": 0.3429,
"step": 428
},
{
"epoch": 1.7392405063291139,
"grad_norm": 0.4586873015408533,
"learning_rate": 3.292085501729909e-05,
"loss": 0.35,
"step": 429
},
{
"epoch": 1.7432911392405064,
"grad_norm": 0.4094250569477021,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.3358,
"step": 430
},
{
"epoch": 1.747341772151899,
"grad_norm": 0.39992672372932947,
"learning_rate": 3.283399942497381e-05,
"loss": 0.3322,
"step": 431
},
{
"epoch": 1.7513924050632912,
"grad_norm": 0.4215758274792743,
"learning_rate": 3.279041640882108e-05,
"loss": 0.3412,
"step": 432
},
{
"epoch": 1.7554430379746835,
"grad_norm": 0.38802729722073165,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.3331,
"step": 433
},
{
"epoch": 1.759493670886076,
"grad_norm": 0.48326146087648203,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.3507,
"step": 434
},
{
"epoch": 1.7635443037974685,
"grad_norm": 0.4048376326658082,
"learning_rate": 3.265905069502244e-05,
"loss": 0.3532,
"step": 435
},
{
"epoch": 1.7675949367088608,
"grad_norm": 0.4438475670600104,
"learning_rate": 3.261505774420016e-05,
"loss": 0.3388,
"step": 436
},
{
"epoch": 1.771645569620253,
"grad_norm": 0.4531961705347585,
"learning_rate": 3.257096319343707e-05,
"loss": 0.3344,
"step": 437
},
{
"epoch": 1.7756962025316456,
"grad_norm": 0.46674650389107364,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.3456,
"step": 438
},
{
"epoch": 1.7797468354430381,
"grad_norm": 0.43602942777078435,
"learning_rate": 3.248247071342966e-05,
"loss": 0.3535,
"step": 439
},
{
"epoch": 1.7837974683544304,
"grad_norm": 0.4758598061496191,
"learning_rate": 3.243807349689161e-05,
"loss": 0.3428,
"step": 440
},
{
"epoch": 1.7878481012658227,
"grad_norm": 0.4807280010424283,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.3381,
"step": 441
},
{
"epoch": 1.7918987341772152,
"grad_norm": 0.41254412607272317,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.3411,
"step": 442
},
{
"epoch": 1.7959493670886078,
"grad_norm": 0.49097290746336436,
"learning_rate": 3.230428223438075e-05,
"loss": 0.324,
"step": 443
},
{
"epoch": 1.8,
"grad_norm": 0.44016168003364103,
"learning_rate": 3.225948647317459e-05,
"loss": 0.3521,
"step": 444
},
{
"epoch": 1.8040506329113923,
"grad_norm": 0.4303152931782457,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.3433,
"step": 445
},
{
"epoch": 1.8081012658227849,
"grad_norm": 0.49628684568320935,
"learning_rate": 3.216959910368034e-05,
"loss": 0.3355,
"step": 446
},
{
"epoch": 1.8121518987341774,
"grad_norm": 0.5125348697434309,
"learning_rate": 3.212450821933277e-05,
"loss": 0.3452,
"step": 447
},
{
"epoch": 1.8162025316455697,
"grad_norm": 0.44841380754723664,
"learning_rate": 3.207931968586281e-05,
"loss": 0.3339,
"step": 448
},
{
"epoch": 1.820253164556962,
"grad_norm": 0.4714659069013098,
"learning_rate": 3.203403386721272e-05,
"loss": 0.3414,
"step": 449
},
{
"epoch": 1.8243037974683545,
"grad_norm": 0.422067046243334,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.3347,
"step": 450
},
{
"epoch": 1.828354430379747,
"grad_norm": 0.4507371550831096,
"learning_rate": 3.194317183405573e-05,
"loss": 0.3342,
"step": 451
},
{
"epoch": 1.8324050632911393,
"grad_norm": 0.43959501542668594,
"learning_rate": 3.189759635133914e-05,
"loss": 0.3521,
"step": 452
},
{
"epoch": 1.8364556962025316,
"grad_norm": 0.4055635892395125,
"learning_rate": 3.185192504701718e-05,
"loss": 0.3447,
"step": 453
},
{
"epoch": 1.840506329113924,
"grad_norm": 0.4683142518285758,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.3457,
"step": 454
},
{
"epoch": 1.8445569620253166,
"grad_norm": 0.40358513543100455,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.3276,
"step": 455
},
{
"epoch": 1.848607594936709,
"grad_norm": 0.44605234425918083,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.3496,
"step": 456
},
{
"epoch": 1.8526582278481012,
"grad_norm": 0.4289458165614119,
"learning_rate": 3.166828898179731e-05,
"loss": 0.3533,
"step": 457
},
{
"epoch": 1.8567088607594937,
"grad_norm": 0.4259369481221902,
"learning_rate": 3.162214410223536e-05,
"loss": 0.3519,
"step": 458
},
{
"epoch": 1.8607594936708862,
"grad_norm": 0.4448018894256399,
"learning_rate": 3.157590561952257e-05,
"loss": 0.3311,
"step": 459
},
{
"epoch": 1.8648101265822785,
"grad_norm": 0.4448926440284337,
"learning_rate": 3.152957390605732e-05,
"loss": 0.3287,
"step": 460
},
{
"epoch": 1.8688607594936708,
"grad_norm": 0.4650203577999011,
"learning_rate": 3.148314933498886e-05,
"loss": 0.3333,
"step": 461
},
{
"epoch": 1.8729113924050633,
"grad_norm": 0.3772123114234287,
"learning_rate": 3.143663228021431e-05,
"loss": 0.3344,
"step": 462
},
{
"epoch": 1.8769620253164558,
"grad_norm": 0.45144865020969566,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.3282,
"step": 463
},
{
"epoch": 1.8810126582278481,
"grad_norm": 0.4233116643799818,
"learning_rate": 3.134332221885661e-05,
"loss": 0.3457,
"step": 464
},
{
"epoch": 1.8850632911392404,
"grad_norm": 0.4497241502433581,
"learning_rate": 3.129652996377987e-05,
"loss": 0.3332,
"step": 465
},
{
"epoch": 1.889113924050633,
"grad_norm": 0.4643660083704921,
"learning_rate": 3.12496467280038e-05,
"loss": 0.3468,
"step": 466
},
{
"epoch": 1.8931645569620255,
"grad_norm": 0.3887037394125834,
"learning_rate": 3.120267288911952e-05,
"loss": 0.3373,
"step": 467
},
{
"epoch": 1.8972151898734178,
"grad_norm": 0.4886383570720335,
"learning_rate": 3.11556088254479e-05,
"loss": 0.3317,
"step": 468
},
{
"epoch": 1.90126582278481,
"grad_norm": 0.43845791934937917,
"learning_rate": 3.11084549160364e-05,
"loss": 0.3285,
"step": 469
},
{
"epoch": 1.9053164556962026,
"grad_norm": 0.5123568210469412,
"learning_rate": 3.106121154065615e-05,
"loss": 0.3312,
"step": 470
},
{
"epoch": 1.909367088607595,
"grad_norm": 0.41107450451781563,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.3573,
"step": 471
},
{
"epoch": 1.9134177215189874,
"grad_norm": 0.39184376605735205,
"learning_rate": 3.096645791467348e-05,
"loss": 0.3221,
"step": 472
},
{
"epoch": 1.9174683544303797,
"grad_norm": 0.4937703779303941,
"learning_rate": 3.091894842720373e-05,
"loss": 0.3376,
"step": 473
},
{
"epoch": 1.9215189873417722,
"grad_norm": 0.4226113403013667,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.3506,
"step": 474
},
{
"epoch": 1.9255696202531647,
"grad_norm": 0.4071096817626689,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.3637,
"step": 475
},
{
"epoch": 1.929620253164557,
"grad_norm": 0.49468807399365394,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.3522,
"step": 476
},
{
"epoch": 1.9336708860759493,
"grad_norm": 0.434363265384187,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.3402,
"step": 477
},
{
"epoch": 1.9377215189873418,
"grad_norm": 0.5014683122540653,
"learning_rate": 3.06800895716334e-05,
"loss": 0.3336,
"step": 478
},
{
"epoch": 1.9417721518987343,
"grad_norm": 0.4040768682247225,
"learning_rate": 3.063205821011029e-05,
"loss": 0.3327,
"step": 479
},
{
"epoch": 1.9458227848101266,
"grad_norm": 0.49643069920792376,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.3274,
"step": 480
},
{
"epoch": 1.949873417721519,
"grad_norm": 0.43117285432124963,
"learning_rate": 3.053573898718618e-05,
"loss": 0.3285,
"step": 481
},
{
"epoch": 1.9539240506329114,
"grad_norm": 0.48179878097540396,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.3595,
"step": 482
},
{
"epoch": 1.957974683544304,
"grad_norm": 0.41951117559667356,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.3311,
"step": 483
},
{
"epoch": 1.9620253164556962,
"grad_norm": 0.4822408512190162,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.3344,
"step": 484
},
{
"epoch": 1.9660759493670885,
"grad_norm": 0.4104198615518152,
"learning_rate": 3.034208541656255e-05,
"loss": 0.3457,
"step": 485
},
{
"epoch": 1.970126582278481,
"grad_norm": 0.40702393354554134,
"learning_rate": 3.029346281309218e-05,
"loss": 0.3567,
"step": 486
},
{
"epoch": 1.9741772151898735,
"grad_norm": 0.40491634947012106,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.3563,
"step": 487
},
{
"epoch": 1.9782278481012658,
"grad_norm": 0.43900637194161607,
"learning_rate": 3.019596929201569e-05,
"loss": 0.3448,
"step": 488
},
{
"epoch": 1.9822784810126581,
"grad_norm": 0.3970494071100939,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.3454,
"step": 489
},
{
"epoch": 1.9863291139240506,
"grad_norm": 0.47304311748105615,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.3408,
"step": 490
},
{
"epoch": 1.9903797468354432,
"grad_norm": 0.3958444694886608,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.3355,
"step": 491
},
{
"epoch": 1.9944303797468355,
"grad_norm": 0.423596733955689,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.3506,
"step": 492
},
{
"epoch": 1.9984810126582278,
"grad_norm": 0.40149059412649674,
"learning_rate": 2.995080534235264e-05,
"loss": 0.3358,
"step": 493
},
{
"epoch": 2.0040506329113925,
"grad_norm": 0.5618395396322275,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.2677,
"step": 494
},
{
"epoch": 2.008101265822785,
"grad_norm": 0.43332389821765266,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.2599,
"step": 495
},
{
"epoch": 2.012151898734177,
"grad_norm": 0.8898708356620648,
"learning_rate": 2.980274210295326e-05,
"loss": 0.2408,
"step": 496
},
{
"epoch": 2.0162025316455696,
"grad_norm": 0.5326469324916266,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.2407,
"step": 497
},
{
"epoch": 2.020253164556962,
"grad_norm": 0.4480468452908142,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.2439,
"step": 498
},
{
"epoch": 2.0243037974683546,
"grad_norm": 0.5163564436662306,
"learning_rate": 2.965396831793362e-05,
"loss": 0.2498,
"step": 499
},
{
"epoch": 2.0283544303797467,
"grad_norm": 0.4355899196981215,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.2266,
"step": 500
},
{
"epoch": 2.0324050632911392,
"grad_norm": 0.4866785258074407,
"learning_rate": 2.955439637024526e-05,
"loss": 0.2338,
"step": 501
},
{
"epoch": 2.0364556962025318,
"grad_norm": 0.4459895160464478,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.23,
"step": 502
},
{
"epoch": 2.0405063291139243,
"grad_norm": 0.4710127778153047,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.2417,
"step": 503
},
{
"epoch": 2.0445569620253163,
"grad_norm": 0.4796994704356595,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.2251,
"step": 504
},
{
"epoch": 2.048607594936709,
"grad_norm": 0.4448677179747505,
"learning_rate": 2.93543322968538e-05,
"loss": 0.255,
"step": 505
},
{
"epoch": 2.0526582278481014,
"grad_norm": 0.4867032470547771,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.2435,
"step": 506
},
{
"epoch": 2.056708860759494,
"grad_norm": 0.38036003994373163,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.2382,
"step": 507
},
{
"epoch": 2.060759493670886,
"grad_norm": 0.4615325550113931,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.2528,
"step": 508
},
{
"epoch": 2.0648101265822785,
"grad_norm": 0.3697695958561701,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.2033,
"step": 509
},
{
"epoch": 2.068860759493671,
"grad_norm": 0.44900575516259744,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.2232,
"step": 510
},
{
"epoch": 2.0729113924050635,
"grad_norm": 0.39540837089314645,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.2355,
"step": 511
},
{
"epoch": 2.0769620253164556,
"grad_norm": 0.5125442084462707,
"learning_rate": 2.900133526131588e-05,
"loss": 0.236,
"step": 512
},
{
"epoch": 2.081012658227848,
"grad_norm": 0.3946511856446207,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.2282,
"step": 513
},
{
"epoch": 2.0850632911392406,
"grad_norm": 0.44624477212672314,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.2316,
"step": 514
},
{
"epoch": 2.089113924050633,
"grad_norm": 0.4066301898996315,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.2437,
"step": 515
},
{
"epoch": 2.093164556962025,
"grad_norm": 0.4038351943710213,
"learning_rate": 2.879801882631847e-05,
"loss": 0.2364,
"step": 516
},
{
"epoch": 2.0972151898734177,
"grad_norm": 0.39744101939085463,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.2203,
"step": 517
},
{
"epoch": 2.1012658227848102,
"grad_norm": 0.4019894557200102,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.2243,
"step": 518
},
{
"epoch": 2.1053164556962027,
"grad_norm": 0.3772037399759273,
"learning_rate": 2.864478605989494e-05,
"loss": 0.2447,
"step": 519
},
{
"epoch": 2.109367088607595,
"grad_norm": 0.37849548296786273,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.2342,
"step": 520
},
{
"epoch": 2.1134177215189873,
"grad_norm": 0.3740054978541879,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.221,
"step": 521
},
{
"epoch": 2.11746835443038,
"grad_norm": 0.3840835351343469,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.2528,
"step": 522
},
{
"epoch": 2.1215189873417724,
"grad_norm": 0.39834513777476704,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.2225,
"step": 523
},
{
"epoch": 2.1255696202531644,
"grad_norm": 0.4295113947861478,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.2315,
"step": 524
},
{
"epoch": 2.129620253164557,
"grad_norm": 0.39674885133667204,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.2331,
"step": 525
},
{
"epoch": 2.1336708860759495,
"grad_norm": 0.46099058333411763,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.2306,
"step": 526
},
{
"epoch": 2.137721518987342,
"grad_norm": 0.4309971070548834,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.2328,
"step": 527
},
{
"epoch": 2.141772151898734,
"grad_norm": 0.4492383217435781,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.2358,
"step": 528
},
{
"epoch": 2.1458227848101266,
"grad_norm": 0.4039958022135034,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.2247,
"step": 529
},
{
"epoch": 2.149873417721519,
"grad_norm": 0.4423000415127838,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.2572,
"step": 530
},
{
"epoch": 2.1539240506329116,
"grad_norm": 0.40085773329839475,
"learning_rate": 2.802570062143278e-05,
"loss": 0.2335,
"step": 531
},
{
"epoch": 2.1579746835443037,
"grad_norm": 0.4131339203383251,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.241,
"step": 532
},
{
"epoch": 2.162025316455696,
"grad_norm": 0.41062623575147883,
"learning_rate": 2.792159532078314e-05,
"loss": 0.2369,
"step": 533
},
{
"epoch": 2.1660759493670887,
"grad_norm": 0.3831237298841066,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.2426,
"step": 534
},
{
"epoch": 2.170126582278481,
"grad_norm": 0.41997970973393056,
"learning_rate": 2.781723482288857e-05,
"loss": 0.2588,
"step": 535
},
{
"epoch": 2.1741772151898733,
"grad_norm": 0.36744392734133985,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.2486,
"step": 536
},
{
"epoch": 2.178227848101266,
"grad_norm": 0.4473651252600172,
"learning_rate": 2.771262248976272e-05,
"loss": 0.2399,
"step": 537
},
{
"epoch": 2.1822784810126583,
"grad_norm": 0.38162747533780605,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.2353,
"step": 538
},
{
"epoch": 2.186329113924051,
"grad_norm": 0.4295843143512743,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.2314,
"step": 539
},
{
"epoch": 2.190379746835443,
"grad_norm": 0.3904140247931754,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.2177,
"step": 540
},
{
"epoch": 2.1944303797468354,
"grad_norm": 0.4029921961374907,
"learning_rate": 2.7502655806328e-05,
"loss": 0.2271,
"step": 541
},
{
"epoch": 2.198481012658228,
"grad_norm": 0.37840697259365813,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.2245,
"step": 542
},
{
"epoch": 2.2025316455696204,
"grad_norm": 0.40425243135524036,
"learning_rate": 2.739730822017673e-05,
"loss": 0.2477,
"step": 543
},
{
"epoch": 2.2065822784810125,
"grad_norm": 0.4064986221703155,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.239,
"step": 544
},
{
"epoch": 2.210632911392405,
"grad_norm": 0.3976327307640877,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.2396,
"step": 545
},
{
"epoch": 2.2146835443037975,
"grad_norm": 0.42200507237736906,
"learning_rate": 2.723884107774236e-05,
"loss": 0.2366,
"step": 546
},
{
"epoch": 2.21873417721519,
"grad_norm": 0.3572953266870805,
"learning_rate": 2.718590152796239e-05,
"loss": 0.2425,
"step": 547
},
{
"epoch": 2.222784810126582,
"grad_norm": 0.37287670811516976,
"learning_rate": 2.71329041039194e-05,
"loss": 0.2136,
"step": 548
},
{
"epoch": 2.2268354430379746,
"grad_norm": 0.40520472248178296,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.2351,
"step": 549
},
{
"epoch": 2.230886075949367,
"grad_norm": 0.43583701290424326,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.2272,
"step": 550
},
{
"epoch": 2.2349367088607597,
"grad_norm": 0.4205873440552366,
"learning_rate": 2.697356885686189e-05,
"loss": 0.2494,
"step": 551
},
{
"epoch": 2.2389873417721518,
"grad_norm": 0.36867114510321086,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.2366,
"step": 552
},
{
"epoch": 2.2430379746835443,
"grad_norm": 0.4234207404044553,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.231,
"step": 553
},
{
"epoch": 2.247088607594937,
"grad_norm": 0.38767496069041685,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.2319,
"step": 554
},
{
"epoch": 2.2511392405063293,
"grad_norm": 0.457679134334366,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.2255,
"step": 555
},
{
"epoch": 2.2551898734177214,
"grad_norm": 0.38605488222731854,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.2397,
"step": 556
},
{
"epoch": 2.259240506329114,
"grad_norm": 0.420789112774985,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.2251,
"step": 557
},
{
"epoch": 2.2632911392405064,
"grad_norm": 0.3940403545330923,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.2528,
"step": 558
},
{
"epoch": 2.267341772151899,
"grad_norm": 0.38661777482856624,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.2255,
"step": 559
},
{
"epoch": 2.271392405063291,
"grad_norm": 0.41145365386503907,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.2234,
"step": 560
},
{
"epoch": 2.2754430379746835,
"grad_norm": 0.4228133056004388,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.2455,
"step": 561
},
{
"epoch": 2.279493670886076,
"grad_norm": 0.4211471300689591,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.2417,
"step": 562
},
{
"epoch": 2.2835443037974685,
"grad_norm": 0.3997006582022732,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.2374,
"step": 563
},
{
"epoch": 2.2875949367088606,
"grad_norm": 0.4292346135291654,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.2393,
"step": 564
},
{
"epoch": 2.291645569620253,
"grad_norm": 0.3817682779370123,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.2299,
"step": 565
},
{
"epoch": 2.2956962025316456,
"grad_norm": 0.39285773666710005,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.2324,
"step": 566
},
{
"epoch": 2.299746835443038,
"grad_norm": 0.3930657617514451,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.2491,
"step": 567
},
{
"epoch": 2.3037974683544302,
"grad_norm": 0.3757254248017907,
"learning_rate": 2.606146264569603e-05,
"loss": 0.2214,
"step": 568
},
{
"epoch": 2.3078481012658227,
"grad_norm": 0.43039762510946655,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.2243,
"step": 569
},
{
"epoch": 2.3118987341772153,
"grad_norm": 0.3678147083602104,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.2259,
"step": 570
},
{
"epoch": 2.3159493670886078,
"grad_norm": 0.4344436514927903,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.2449,
"step": 571
},
{
"epoch": 2.32,
"grad_norm": 0.42021366436590923,
"learning_rate": 2.584472010606015e-05,
"loss": 0.2582,
"step": 572
},
{
"epoch": 2.3240506329113924,
"grad_norm": 0.39771892157128197,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.2424,
"step": 573
},
{
"epoch": 2.328101265822785,
"grad_norm": 0.4465485960338361,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.2436,
"step": 574
},
{
"epoch": 2.3321518987341774,
"grad_norm": 0.4190857243984731,
"learning_rate": 2.568166741322048e-05,
"loss": 0.2355,
"step": 575
},
{
"epoch": 2.3362025316455695,
"grad_norm": 0.4329855588980875,
"learning_rate": 2.56272244128509e-05,
"loss": 0.2287,
"step": 576
},
{
"epoch": 2.340253164556962,
"grad_norm": 0.46592746001911406,
"learning_rate": 2.55727360915895e-05,
"loss": 0.2343,
"step": 577
},
{
"epoch": 2.3443037974683545,
"grad_norm": 0.41935839680318565,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.219,
"step": 578
},
{
"epoch": 2.348354430379747,
"grad_norm": 0.4560876140527052,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.2428,
"step": 579
},
{
"epoch": 2.352405063291139,
"grad_norm": 0.4310363715768755,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.2554,
"step": 580
},
{
"epoch": 2.3564556962025316,
"grad_norm": 0.42192908006763724,
"learning_rate": 2.535433837985559e-05,
"loss": 0.2399,
"step": 581
},
{
"epoch": 2.360506329113924,
"grad_norm": 0.4181435626849493,
"learning_rate": 2.529963004393324e-05,
"loss": 0.2544,
"step": 582
},
{
"epoch": 2.3645569620253166,
"grad_norm": 0.4512644007070888,
"learning_rate": 2.524487902551908e-05,
"loss": 0.2207,
"step": 583
},
{
"epoch": 2.3686075949367087,
"grad_norm": 0.3910855802325268,
"learning_rate": 2.519008576557029e-05,
"loss": 0.2364,
"step": 584
},
{
"epoch": 2.372658227848101,
"grad_norm": 0.41851684593162275,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.2207,
"step": 585
},
{
"epoch": 2.3767088607594937,
"grad_norm": 0.39737462734599194,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.249,
"step": 586
},
{
"epoch": 2.3807594936708862,
"grad_norm": 0.4203827972687883,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.2439,
"step": 587
},
{
"epoch": 2.3848101265822783,
"grad_norm": 0.43853518785710766,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.2284,
"step": 588
},
{
"epoch": 2.388860759493671,
"grad_norm": 0.43186745860197445,
"learning_rate": 2.491550129992345e-05,
"loss": 0.2378,
"step": 589
},
{
"epoch": 2.3929113924050633,
"grad_norm": 0.44202264255798535,
"learning_rate": 2.486046386967024e-05,
"loss": 0.2621,
"step": 590
},
{
"epoch": 2.396962025316456,
"grad_norm": 0.41845562824930554,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.2481,
"step": 591
},
{
"epoch": 2.401012658227848,
"grad_norm": 0.40088506105057425,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.23,
"step": 592
},
{
"epoch": 2.4050632911392404,
"grad_norm": 0.4244950102771436,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.246,
"step": 593
},
{
"epoch": 2.409113924050633,
"grad_norm": 0.39684088298349546,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.2439,
"step": 594
},
{
"epoch": 2.4131645569620255,
"grad_norm": 0.4221208098247973,
"learning_rate": 2.458469841202444e-05,
"loss": 0.2604,
"step": 595
},
{
"epoch": 2.4172151898734175,
"grad_norm": 0.3998545685662563,
"learning_rate": 2.452943276859226e-05,
"loss": 0.2414,
"step": 596
},
{
"epoch": 2.42126582278481,
"grad_norm": 0.420069725679426,
"learning_rate": 2.447413064573125e-05,
"loss": 0.2241,
"step": 597
},
{
"epoch": 2.4253164556962026,
"grad_norm": 0.36544772639522166,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.2483,
"step": 598
},
{
"epoch": 2.429367088607595,
"grad_norm": 0.39663682270541706,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.233,
"step": 599
},
{
"epoch": 2.433417721518987,
"grad_norm": 0.38757378540584453,
"learning_rate": 2.430800985597963e-05,
"loss": 0.2402,
"step": 600
},
{
"epoch": 2.4374683544303797,
"grad_norm": 0.42534994290944345,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.2368,
"step": 601
},
{
"epoch": 2.441518987341772,
"grad_norm": 0.38777890280952854,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.2472,
"step": 602
},
{
"epoch": 2.4455696202531647,
"grad_norm": 0.38461619334046515,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.2374,
"step": 603
},
{
"epoch": 2.449620253164557,
"grad_norm": 0.3575331884288786,
"learning_rate": 2.408603181104957e-05,
"loss": 0.2364,
"step": 604
},
{
"epoch": 2.4536708860759493,
"grad_norm": 0.3809260154830202,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.2434,
"step": 605
},
{
"epoch": 2.457721518987342,
"grad_norm": 0.3674682488873355,
"learning_rate": 2.397484355008521e-05,
"loss": 0.2388,
"step": 606
},
{
"epoch": 2.4617721518987343,
"grad_norm": 0.39264609162869163,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.2505,
"step": 607
},
{
"epoch": 2.4658227848101264,
"grad_norm": 0.38934083584223306,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.2566,
"step": 608
},
{
"epoch": 2.469873417721519,
"grad_norm": 0.38223984149305384,
"learning_rate": 2.380782218329337e-05,
"loss": 0.2316,
"step": 609
},
{
"epoch": 2.4739240506329114,
"grad_norm": 0.40525892824168447,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.2336,
"step": 610
},
{
"epoch": 2.477974683544304,
"grad_norm": 0.4045613950648491,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.2422,
"step": 611
},
{
"epoch": 2.482025316455696,
"grad_norm": 0.4511792868284245,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.2266,
"step": 612
},
{
"epoch": 2.4860759493670885,
"grad_norm": 0.4009787983037842,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.2428,
"step": 613
},
{
"epoch": 2.490126582278481,
"grad_norm": 0.42450156703272257,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.237,
"step": 614
},
{
"epoch": 2.4941772151898736,
"grad_norm": 0.39774338372984,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.2366,
"step": 615
},
{
"epoch": 2.4982278481012656,
"grad_norm": 0.470069226269624,
"learning_rate": 2.341705325992901e-05,
"loss": 0.2433,
"step": 616
},
{
"epoch": 2.502278481012658,
"grad_norm": 0.4351053918202678,
"learning_rate": 2.336111544604222e-05,
"loss": 0.2297,
"step": 617
},
{
"epoch": 2.5063291139240507,
"grad_norm": 0.4408716045706729,
"learning_rate": 2.33051505621937e-05,
"loss": 0.2327,
"step": 618
},
{
"epoch": 2.510379746835443,
"grad_norm": 0.4109059069583198,
"learning_rate": 2.324915905911693e-05,
"loss": 0.2442,
"step": 619
},
{
"epoch": 2.5144303797468357,
"grad_norm": 0.4312895485399082,
"learning_rate": 2.319314138775977e-05,
"loss": 0.215,
"step": 620
},
{
"epoch": 2.5184810126582278,
"grad_norm": 0.4109399930753428,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.2462,
"step": 621
},
{
"epoch": 2.5225316455696203,
"grad_norm": 0.39640058064286876,
"learning_rate": 2.308102934504593e-05,
"loss": 0.2359,
"step": 622
},
{
"epoch": 2.526582278481013,
"grad_norm": 0.40852775032864325,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.239,
"step": 623
},
{
"epoch": 2.530632911392405,
"grad_norm": 0.38191606948924234,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.2398,
"step": 624
},
{
"epoch": 2.5346835443037974,
"grad_norm": 0.38233337359412944,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.2459,
"step": 625
},
{
"epoch": 2.53873417721519,
"grad_norm": 0.40495506229913353,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.2431,
"step": 626
},
{
"epoch": 2.5427848101265824,
"grad_norm": 0.39885188481152023,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.2279,
"step": 627
},
{
"epoch": 2.546835443037975,
"grad_norm": 0.4033926560537853,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.2454,
"step": 628
},
{
"epoch": 2.550886075949367,
"grad_norm": 0.4046861759717986,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.232,
"step": 629
},
{
"epoch": 2.5549367088607595,
"grad_norm": 0.40064571005257354,
"learning_rate": 2.26316247733093e-05,
"loss": 0.2253,
"step": 630
},
{
"epoch": 2.558987341772152,
"grad_norm": 0.40585718502439483,
"learning_rate": 2.257534907110328e-05,
"loss": 0.2468,
"step": 631
},
{
"epoch": 2.563037974683544,
"grad_norm": 0.3749249871574529,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.2445,
"step": 632
},
{
"epoch": 2.5670886075949366,
"grad_norm": 0.39450012078657726,
"learning_rate": 2.246273589557294e-05,
"loss": 0.2223,
"step": 633
},
{
"epoch": 2.571139240506329,
"grad_norm": 0.41159350647444576,
"learning_rate": 2.240639932921966e-05,
"loss": 0.2519,
"step": 634
},
{
"epoch": 2.5751898734177217,
"grad_norm": 0.37339520740011106,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.2446,
"step": 635
},
{
"epoch": 2.579240506329114,
"grad_norm": 0.4451016925255843,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.2403,
"step": 636
},
{
"epoch": 2.5832911392405062,
"grad_norm": 0.3933793997091228,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.2378,
"step": 637
},
{
"epoch": 2.5873417721518988,
"grad_norm": 0.38558785026972425,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.2409,
"step": 638
},
{
"epoch": 2.5913924050632913,
"grad_norm": 0.431814564579042,
"learning_rate": 2.212443486523819e-05,
"loss": 0.2248,
"step": 639
},
{
"epoch": 2.5954430379746833,
"grad_norm": 0.3575451603726466,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.236,
"step": 640
},
{
"epoch": 2.599493670886076,
"grad_norm": 0.3789258187901637,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.2288,
"step": 641
},
{
"epoch": 2.6035443037974684,
"grad_norm": 0.4123602283365587,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.2343,
"step": 642
},
{
"epoch": 2.607594936708861,
"grad_norm": 0.3936037747343463,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.2236,
"step": 643
},
{
"epoch": 2.6116455696202534,
"grad_norm": 0.3738910173528121,
"learning_rate": 2.184204266048803e-05,
"loss": 0.2332,
"step": 644
},
{
"epoch": 2.6156962025316455,
"grad_norm": 0.41213419669726276,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.2312,
"step": 645
},
{
"epoch": 2.619746835443038,
"grad_norm": 0.38195201926011796,
"learning_rate": 2.17289787446841e-05,
"loss": 0.2479,
"step": 646
},
{
"epoch": 2.6237974683544305,
"grad_norm": 0.39167428290524103,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.226,
"step": 647
},
{
"epoch": 2.6278481012658226,
"grad_norm": 0.4101363147918156,
"learning_rate": 2.161585912916385e-05,
"loss": 0.2447,
"step": 648
},
{
"epoch": 2.631898734177215,
"grad_norm": 0.4228280861088826,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.2309,
"step": 649
},
{
"epoch": 2.6359493670886076,
"grad_norm": 0.3683613437482635,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.2325,
"step": 650
},
{
"epoch": 2.64,
"grad_norm": 0.41673071854036436,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.2311,
"step": 651
},
{
"epoch": 2.6440506329113926,
"grad_norm": 0.3998500191675193,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.231,
"step": 652
},
{
"epoch": 2.6481012658227847,
"grad_norm": 0.38188751036544943,
"learning_rate": 2.133284032321232e-05,
"loss": 0.2289,
"step": 653
},
{
"epoch": 2.6521518987341772,
"grad_norm": 0.4268820058400474,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.2508,
"step": 654
},
{
"epoch": 2.6562025316455697,
"grad_norm": 0.35328899376965733,
"learning_rate": 2.121955446741306e-05,
"loss": 0.239,
"step": 655
},
{
"epoch": 2.660253164556962,
"grad_norm": 0.4064467649561817,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.2458,
"step": 656
},
{
"epoch": 2.6643037974683543,
"grad_norm": 0.37863588831528056,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.2087,
"step": 657
},
{
"epoch": 2.668354430379747,
"grad_norm": 0.3618838994727545,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.2261,
"step": 658
},
{
"epoch": 2.6724050632911394,
"grad_norm": 0.42935303912812894,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.2413,
"step": 659
},
{
"epoch": 2.676455696202532,
"grad_norm": 0.37955875527596716,
"learning_rate": 2.093617592756158e-05,
"loss": 0.2415,
"step": 660
},
{
"epoch": 2.680506329113924,
"grad_norm": 0.38224879403564355,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.2365,
"step": 661
},
{
"epoch": 2.6845569620253165,
"grad_norm": 0.42050380546223004,
"learning_rate": 2.082276853708978e-05,
"loss": 0.2281,
"step": 662
},
{
"epoch": 2.688607594936709,
"grad_norm": 0.35681884021846644,
"learning_rate": 2.076605467380071e-05,
"loss": 0.23,
"step": 663
},
{
"epoch": 2.692658227848101,
"grad_norm": 0.40904500896021107,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.2478,
"step": 664
},
{
"epoch": 2.6967088607594936,
"grad_norm": 0.41030736517436145,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.2307,
"step": 665
},
{
"epoch": 2.700759493670886,
"grad_norm": 0.3724714516647774,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.254,
"step": 666
},
{
"epoch": 2.7048101265822786,
"grad_norm": 0.397666061823016,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.2482,
"step": 667
},
{
"epoch": 2.708860759493671,
"grad_norm": 0.3557003932445275,
"learning_rate": 2.048240194884496e-05,
"loss": 0.2333,
"step": 668
},
{
"epoch": 2.712911392405063,
"grad_norm": 0.36491429120875646,
"learning_rate": 2.042565792045695e-05,
"loss": 0.2404,
"step": 669
},
{
"epoch": 2.7169620253164557,
"grad_norm": 0.3722584181059093,
"learning_rate": 2.036891046387857e-05,
"loss": 0.2291,
"step": 670
},
{
"epoch": 2.721012658227848,
"grad_norm": 0.3645979227009828,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.2168,
"step": 671
},
{
"epoch": 2.7250632911392403,
"grad_norm": 0.37139278094623096,
"learning_rate": 2.025540709431948e-05,
"loss": 0.2262,
"step": 672
},
{
"epoch": 2.729113924050633,
"grad_norm": 0.3517294436799045,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.2335,
"step": 673
},
{
"epoch": 2.7331645569620253,
"grad_norm": 0.34678356193489135,
"learning_rate": 2.014189549672245e-05,
"loss": 0.2465,
"step": 674
},
{
"epoch": 2.737215189873418,
"grad_norm": 0.3841880269015426,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.2231,
"step": 675
},
{
"epoch": 2.7412658227848103,
"grad_norm": 0.37191772608715434,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.2595,
"step": 676
},
{
"epoch": 2.7453164556962024,
"grad_norm": 0.36961027223969195,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.2409,
"step": 677
},
{
"epoch": 2.749367088607595,
"grad_norm": 0.3946954386302438,
"learning_rate": 1.991486224484123e-05,
"loss": 0.2461,
"step": 678
},
{
"epoch": 2.7534177215189874,
"grad_norm": 0.3505511786173827,
"learning_rate": 1.985810450327756e-05,
"loss": 0.2454,
"step": 679
},
{
"epoch": 2.7574683544303795,
"grad_norm": 0.38546077857392547,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.2466,
"step": 680
},
{
"epoch": 2.761518987341772,
"grad_norm": 0.3518760312724095,
"learning_rate": 1.974459290568053e-05,
"loss": 0.2275,
"step": 681
},
{
"epoch": 2.7655696202531646,
"grad_norm": 0.3661859644814656,
"learning_rate": 1.968783996385397e-05,
"loss": 0.2385,
"step": 682
},
{
"epoch": 2.769620253164557,
"grad_norm": 0.4116662522777588,
"learning_rate": 1.963108953612143e-05,
"loss": 0.222,
"step": 683
},
{
"epoch": 2.7736708860759496,
"grad_norm": 0.36958715171088097,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.2264,
"step": 684
},
{
"epoch": 2.7777215189873417,
"grad_norm": 0.3871376760624998,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.2354,
"step": 685
},
{
"epoch": 2.781772151898734,
"grad_norm": 0.39949090563007705,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.2357,
"step": 686
},
{
"epoch": 2.7858227848101267,
"grad_norm": 0.35599855940173336,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.2477,
"step": 687
},
{
"epoch": 2.7898734177215188,
"grad_norm": 0.40710198983201223,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.2375,
"step": 688
},
{
"epoch": 2.7939240506329113,
"grad_norm": 0.3748625136319348,
"learning_rate": 1.92906653591874e-05,
"loss": 0.2231,
"step": 689
},
{
"epoch": 2.797974683544304,
"grad_norm": 0.36587564580339216,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.2327,
"step": 690
},
{
"epoch": 2.8020253164556963,
"grad_norm": 0.3942246890598021,
"learning_rate": 1.917723146291022e-05,
"loss": 0.2441,
"step": 691
},
{
"epoch": 2.806075949367089,
"grad_norm": 0.37475350075595404,
"learning_rate": 1.912052422608584e-05,
"loss": 0.2112,
"step": 692
},
{
"epoch": 2.810126582278481,
"grad_norm": 0.38751150324553396,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.2364,
"step": 693
},
{
"epoch": 2.8141772151898734,
"grad_norm": 0.3815367409479136,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.2484,
"step": 694
},
{
"epoch": 2.818227848101266,
"grad_norm": 0.42166819798688926,
"learning_rate": 1.895044684123479e-05,
"loss": 0.2462,
"step": 695
},
{
"epoch": 2.822278481012658,
"grad_norm": 0.4123883198781062,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.2391,
"step": 696
},
{
"epoch": 2.8263291139240505,
"grad_norm": 0.41624241391733957,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.2563,
"step": 697
},
{
"epoch": 2.830379746835443,
"grad_norm": 0.35284113045155685,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.2218,
"step": 698
},
{
"epoch": 2.8344303797468355,
"grad_norm": 0.43512671998164176,
"learning_rate": 1.872379746550743e-05,
"loss": 0.2305,
"step": 699
},
{
"epoch": 2.838481012658228,
"grad_norm": 0.4163436209372256,
"learning_rate": 1.866715967678769e-05,
"loss": 0.2365,
"step": 700
},
{
"epoch": 2.84253164556962,
"grad_norm": 0.3758087266792836,
"learning_rate": 1.861053262258067e-05,
"loss": 0.2312,
"step": 701
},
{
"epoch": 2.8465822784810126,
"grad_norm": 0.44742611293399265,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.2517,
"step": 702
},
{
"epoch": 2.850632911392405,
"grad_norm": 0.3636246580573936,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.2356,
"step": 703
},
{
"epoch": 2.8546835443037972,
"grad_norm": 0.3860441927158775,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.2421,
"step": 704
},
{
"epoch": 2.8587341772151897,
"grad_norm": 0.4262273938933396,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.2208,
"step": 705
},
{
"epoch": 2.8627848101265823,
"grad_norm": 0.3430157346254683,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.234,
"step": 706
},
{
"epoch": 2.8668354430379748,
"grad_norm": 0.3787743301392553,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.2351,
"step": 707
},
{
"epoch": 2.8708860759493673,
"grad_norm": 0.3569176962972927,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.2234,
"step": 708
},
{
"epoch": 2.8749367088607594,
"grad_norm": 0.41249453656723134,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.2315,
"step": 709
},
{
"epoch": 2.878987341772152,
"grad_norm": 0.406992726547366,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.2285,
"step": 710
},
{
"epoch": 2.8830379746835444,
"grad_norm": 0.3788897405826842,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.2433,
"step": 711
},
{
"epoch": 2.8870886075949365,
"grad_norm": 0.3857383970794428,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.2458,
"step": 712
},
{
"epoch": 2.891139240506329,
"grad_norm": 0.35333306176399426,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.2384,
"step": 713
},
{
"epoch": 2.8951898734177215,
"grad_norm": 0.355563818724427,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.2349,
"step": 714
},
{
"epoch": 2.899240506329114,
"grad_norm": 0.37604324484364293,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.2491,
"step": 715
},
{
"epoch": 2.9032911392405065,
"grad_norm": 0.39647108051149355,
"learning_rate": 1.776272483901182e-05,
"loss": 0.2454,
"step": 716
},
{
"epoch": 2.9073417721518986,
"grad_norm": 0.3342208792388238,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.229,
"step": 717
},
{
"epoch": 2.911392405063291,
"grad_norm": 0.3725911952825882,
"learning_rate": 1.764995661794301e-05,
"loss": 0.218,
"step": 718
},
{
"epoch": 2.9154430379746836,
"grad_norm": 0.36340125759200664,
"learning_rate": 1.759360067078035e-05,
"loss": 0.2287,
"step": 719
},
{
"epoch": 2.9194936708860757,
"grad_norm": 0.33375238936719676,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.2432,
"step": 720
},
{
"epoch": 2.923544303797468,
"grad_norm": 0.3788950337275911,
"learning_rate": 1.748094737261012e-05,
"loss": 0.2343,
"step": 721
},
{
"epoch": 2.9275949367088607,
"grad_norm": 0.3538110993119652,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.2428,
"step": 722
},
{
"epoch": 2.9316455696202532,
"grad_norm": 0.3729954386825288,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.2386,
"step": 723
},
{
"epoch": 2.9356962025316458,
"grad_norm": 0.38551988945392307,
"learning_rate": 1.731212071922883e-05,
"loss": 0.2369,
"step": 724
},
{
"epoch": 2.939746835443038,
"grad_norm": 0.36840481153599924,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.2384,
"step": 725
},
{
"epoch": 2.9437974683544303,
"grad_norm": 0.34747676897380914,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.23,
"step": 726
},
{
"epoch": 2.947848101265823,
"grad_norm": 0.3351699701907833,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.2238,
"step": 727
},
{
"epoch": 2.951898734177215,
"grad_norm": 0.3775018486935368,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.228,
"step": 728
},
{
"epoch": 2.9559493670886074,
"grad_norm": 0.32364765979959403,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.2327,
"step": 729
},
{
"epoch": 2.96,
"grad_norm": 0.33733007188599073,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.2461,
"step": 730
},
{
"epoch": 2.9640506329113925,
"grad_norm": 0.35148360201503265,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.2573,
"step": 731
},
{
"epoch": 2.968101265822785,
"grad_norm": 0.3314040120806177,
"learning_rate": 1.686290200071915e-05,
"loss": 0.2326,
"step": 732
},
{
"epoch": 2.972151898734177,
"grad_norm": 0.3343485105579084,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.2157,
"step": 733
},
{
"epoch": 2.9762025316455696,
"grad_norm": 0.36331680384820336,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.2443,
"step": 734
},
{
"epoch": 2.980253164556962,
"grad_norm": 0.34556931760620235,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.2312,
"step": 735
},
{
"epoch": 2.984303797468354,
"grad_norm": 0.3507676711418869,
"learning_rate": 1.663888455395778e-05,
"loss": 0.2283,
"step": 736
},
{
"epoch": 2.9883544303797467,
"grad_norm": 0.3513847214307472,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.2298,
"step": 737
},
{
"epoch": 2.992405063291139,
"grad_norm": 0.3781032493546456,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.236,
"step": 738
},
{
"epoch": 2.9964556962025317,
"grad_norm": 0.3744270869608403,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.2372,
"step": 739
},
{
"epoch": 3.002025316455696,
"grad_norm": 0.3600906601907665,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.1793,
"step": 740
},
{
"epoch": 3.0060759493670886,
"grad_norm": 0.40112459476458795,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.1535,
"step": 741
},
{
"epoch": 3.010126582278481,
"grad_norm": 0.31946734086205714,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.1361,
"step": 742
},
{
"epoch": 3.0141772151898736,
"grad_norm": 0.45398609669928935,
"learning_rate": 1.624791353909428e-05,
"loss": 0.1587,
"step": 743
},
{
"epoch": 3.0182278481012657,
"grad_norm": 0.5118730334195539,
"learning_rate": 1.619217781670663e-05,
"loss": 0.1507,
"step": 744
},
{
"epoch": 3.022278481012658,
"grad_norm": 0.40123654886944016,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.1275,
"step": 745
},
{
"epoch": 3.0263291139240507,
"grad_norm": 0.384680351695138,
"learning_rate": 1.608079882360134e-05,
"loss": 0.1674,
"step": 746
},
{
"epoch": 3.030379746835443,
"grad_norm": 0.43882673815165796,
"learning_rate": 1.60251564499148e-05,
"loss": 0.1493,
"step": 747
},
{
"epoch": 3.0344303797468353,
"grad_norm": 0.43264385504958114,
"learning_rate": 1.596954608907176e-05,
"loss": 0.1487,
"step": 748
},
{
"epoch": 3.038481012658228,
"grad_norm": 0.36609589157662253,
"learning_rate": 1.591396818895043e-05,
"loss": 0.1475,
"step": 749
},
{
"epoch": 3.0425316455696203,
"grad_norm": 0.36518691908192064,
"learning_rate": 1.585842319716759e-05,
"loss": 0.156,
"step": 750
},
{
"epoch": 3.046582278481013,
"grad_norm": 0.38339614602998345,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.1526,
"step": 751
},
{
"epoch": 3.050632911392405,
"grad_norm": 0.3783817956537679,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.1364,
"step": 752
},
{
"epoch": 3.0546835443037974,
"grad_norm": 0.3853354718823504,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.1295,
"step": 753
},
{
"epoch": 3.05873417721519,
"grad_norm": 0.3683473649618834,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.1514,
"step": 754
},
{
"epoch": 3.0627848101265824,
"grad_norm": 0.3465534602257483,
"learning_rate": 1.558120751116291e-05,
"loss": 0.1567,
"step": 755
},
{
"epoch": 3.0668354430379745,
"grad_norm": 0.3714963265762931,
"learning_rate": 1.552586935426876e-05,
"loss": 0.1728,
"step": 756
},
{
"epoch": 3.070886075949367,
"grad_norm": 0.33554881871550446,
"learning_rate": 1.547056723140774e-05,
"loss": 0.1434,
"step": 757
},
{
"epoch": 3.0749367088607595,
"grad_norm": 0.3389842353601179,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.1583,
"step": 758
},
{
"epoch": 3.078987341772152,
"grad_norm": 0.36692130539979984,
"learning_rate": 1.536007286907411e-05,
"loss": 0.1392,
"step": 759
},
{
"epoch": 3.083037974683544,
"grad_norm": 0.335896722920729,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.1531,
"step": 760
},
{
"epoch": 3.0870886075949366,
"grad_norm": 0.3436782248390633,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.1561,
"step": 761
},
{
"epoch": 3.091139240506329,
"grad_norm": 0.33955286157688236,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.1493,
"step": 762
},
{
"epoch": 3.0951898734177217,
"grad_norm": 0.3494545971498695,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.1431,
"step": 763
},
{
"epoch": 3.0992405063291137,
"grad_norm": 0.3306178856674918,
"learning_rate": 1.508449870007656e-05,
"loss": 0.1331,
"step": 764
},
{
"epoch": 3.1032911392405063,
"grad_norm": 0.3330672805483444,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.1475,
"step": 765
},
{
"epoch": 3.1073417721518988,
"grad_norm": 0.3324128464305078,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.1477,
"step": 766
},
{
"epoch": 3.1113924050632913,
"grad_norm": 0.34436929599385435,
"learning_rate": 1.4919625713405e-05,
"loss": 0.1611,
"step": 767
},
{
"epoch": 3.1154430379746834,
"grad_norm": 0.3368151330938911,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.1653,
"step": 768
},
{
"epoch": 3.119493670886076,
"grad_norm": 0.34720567495849886,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.1417,
"step": 769
},
{
"epoch": 3.1235443037974684,
"grad_norm": 0.3530142848197358,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.1507,
"step": 770
},
{
"epoch": 3.127594936708861,
"grad_norm": 0.3361376412881378,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.1397,
"step": 771
},
{
"epoch": 3.131645569620253,
"grad_norm": 0.3317399724813452,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.1557,
"step": 772
},
{
"epoch": 3.1356962025316455,
"grad_norm": 0.3289389068959089,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.1526,
"step": 773
},
{
"epoch": 3.139746835443038,
"grad_norm": 0.34039827534002765,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.1317,
"step": 774
},
{
"epoch": 3.1437974683544305,
"grad_norm": 0.328819129245934,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.1594,
"step": 775
},
{
"epoch": 3.1478481012658226,
"grad_norm": 0.31820287555053806,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.1391,
"step": 776
},
{
"epoch": 3.151898734177215,
"grad_norm": 0.35484334897879966,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.1651,
"step": 777
},
{
"epoch": 3.1559493670886076,
"grad_norm": 0.34311164598281474,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.1777,
"step": 778
},
{
"epoch": 3.16,
"grad_norm": 0.3444479815627733,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.1417,
"step": 779
},
{
"epoch": 3.164050632911392,
"grad_norm": 0.32181567420062174,
"learning_rate": 1.420958430225303e-05,
"loss": 0.1623,
"step": 780
},
{
"epoch": 3.1681012658227847,
"grad_norm": 0.33586457295958505,
"learning_rate": 1.415527989393985e-05,
"loss": 0.1303,
"step": 781
},
{
"epoch": 3.1721518987341772,
"grad_norm": 0.3271133845198112,
"learning_rate": 1.410102255819891e-05,
"loss": 0.1467,
"step": 782
},
{
"epoch": 3.1762025316455698,
"grad_norm": 0.33559647858172137,
"learning_rate": 1.404681273201131e-05,
"loss": 0.1443,
"step": 783
},
{
"epoch": 3.180253164556962,
"grad_norm": 0.34043174550253524,
"learning_rate": 1.399265085197556e-05,
"loss": 0.1427,
"step": 784
},
{
"epoch": 3.1843037974683543,
"grad_norm": 0.32740460988815545,
"learning_rate": 1.393853735430398e-05,
"loss": 0.1529,
"step": 785
},
{
"epoch": 3.188354430379747,
"grad_norm": 0.3442761869384193,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.1489,
"step": 786
},
{
"epoch": 3.1924050632911394,
"grad_norm": 0.3363383235200708,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.1385,
"step": 787
},
{
"epoch": 3.1964556962025314,
"grad_norm": 0.3414985699103966,
"learning_rate": 1.377649151173163e-05,
"loss": 0.1442,
"step": 788
},
{
"epoch": 3.200506329113924,
"grad_norm": 0.3405344647297418,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.1691,
"step": 789
},
{
"epoch": 3.2045569620253165,
"grad_norm": 0.3426422538891704,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.1528,
"step": 790
},
{
"epoch": 3.208607594936709,
"grad_norm": 0.3302376021153577,
"learning_rate": 1.361489677627324e-05,
"loss": 0.1371,
"step": 791
},
{
"epoch": 3.212658227848101,
"grad_norm": 0.33450688020849223,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.1483,
"step": 792
},
{
"epoch": 3.2167088607594936,
"grad_norm": 0.3250605796983243,
"learning_rate": 1.350742335331241e-05,
"loss": 0.1612,
"step": 793
},
{
"epoch": 3.220759493670886,
"grad_norm": 0.3572227713607356,
"learning_rate": 1.345376486102198e-05,
"loss": 0.1417,
"step": 794
},
{
"epoch": 3.2248101265822786,
"grad_norm": 0.3492236899367037,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.1513,
"step": 795
},
{
"epoch": 3.2288607594936707,
"grad_norm": 0.34966109970897974,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.1542,
"step": 796
},
{
"epoch": 3.232911392405063,
"grad_norm": 0.3456159387792808,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.1581,
"step": 797
},
{
"epoch": 3.2369620253164557,
"grad_norm": 0.3684470770014698,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.1487,
"step": 798
},
{
"epoch": 3.2410126582278482,
"grad_norm": 0.335170380212365,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.1494,
"step": 799
},
{
"epoch": 3.2450632911392403,
"grad_norm": 0.36440020050453553,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.1434,
"step": 800
},
{
"epoch": 3.249113924050633,
"grad_norm": 0.34430774761951694,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.1463,
"step": 801
},
{
"epoch": 3.2531645569620253,
"grad_norm": 0.32462122014595196,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.1567,
"step": 802
},
{
"epoch": 3.257215189873418,
"grad_norm": 0.3257397408244913,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.1367,
"step": 803
},
{
"epoch": 3.26126582278481,
"grad_norm": 0.32946993206642855,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.1545,
"step": 804
},
{
"epoch": 3.2653164556962024,
"grad_norm": 0.32915263683300783,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.1814,
"step": 805
},
{
"epoch": 3.269367088607595,
"grad_norm": 0.3115846798768745,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.1401,
"step": 806
},
{
"epoch": 3.2734177215189875,
"grad_norm": 0.32342652750048984,
"learning_rate": 1.276115892225764e-05,
"loss": 0.1604,
"step": 807
},
{
"epoch": 3.27746835443038,
"grad_norm": 0.3224655427735403,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.1607,
"step": 808
},
{
"epoch": 3.281518987341772,
"grad_norm": 0.3156663544018741,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.1559,
"step": 809
},
{
"epoch": 3.2855696202531646,
"grad_norm": 0.3338094396117485,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.157,
"step": 810
},
{
"epoch": 3.289620253164557,
"grad_norm": 0.33265037181577406,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.1518,
"step": 811
},
{
"epoch": 3.293670886075949,
"grad_norm": 0.32824153396715305,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.1569,
"step": 812
},
{
"epoch": 3.2977215189873417,
"grad_norm": 0.33014068084729403,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.1486,
"step": 813
},
{
"epoch": 3.301772151898734,
"grad_norm": 0.32705533306830464,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.145,
"step": 814
},
{
"epoch": 3.3058227848101267,
"grad_norm": 0.34127080593717846,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.136,
"step": 815
},
{
"epoch": 3.309873417721519,
"grad_norm": 0.33531413258140885,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.1476,
"step": 816
},
{
"epoch": 3.3139240506329113,
"grad_norm": 0.32686612342082694,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.1528,
"step": 817
},
{
"epoch": 3.317974683544304,
"grad_norm": 0.33387249382370926,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.1435,
"step": 818
},
{
"epoch": 3.3220253164556963,
"grad_norm": 0.3274956893334011,
"learning_rate": 1.213055323844297e-05,
"loss": 0.1382,
"step": 819
},
{
"epoch": 3.3260759493670884,
"grad_norm": 0.30841593713237336,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.1442,
"step": 820
},
{
"epoch": 3.330126582278481,
"grad_norm": 0.33058804249374335,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.1474,
"step": 821
},
{
"epoch": 3.3341772151898734,
"grad_norm": 0.30706302501836696,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.1575,
"step": 822
},
{
"epoch": 3.338227848101266,
"grad_norm": 0.32471745180822503,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.1567,
"step": 823
},
{
"epoch": 3.3422784810126585,
"grad_norm": 0.344128818496442,
"learning_rate": 1.187045262895488e-05,
"loss": 0.1503,
"step": 824
},
{
"epoch": 3.3463291139240505,
"grad_norm": 0.314633594776157,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.1452,
"step": 825
},
{
"epoch": 3.350379746835443,
"grad_norm": 0.33844388747545395,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.1429,
"step": 826
},
{
"epoch": 3.3544303797468356,
"grad_norm": 0.3126608941319487,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.1731,
"step": 827
},
{
"epoch": 3.3584810126582276,
"grad_norm": 0.3238521248506667,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.1388,
"step": 828
},
{
"epoch": 3.36253164556962,
"grad_norm": 0.32567642967647065,
"learning_rate": 1.161198884952377e-05,
"loss": 0.1495,
"step": 829
},
{
"epoch": 3.3665822784810127,
"grad_norm": 0.32285358999256974,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.144,
"step": 830
},
{
"epoch": 3.370632911392405,
"grad_norm": 0.33121836345646644,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.1712,
"step": 831
},
{
"epoch": 3.3746835443037977,
"grad_norm": 0.3387250195501694,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.1377,
"step": 832
},
{
"epoch": 3.3787341772151898,
"grad_norm": 0.32023655279061575,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.1415,
"step": 833
},
{
"epoch": 3.3827848101265823,
"grad_norm": 0.3400757069876044,
"learning_rate": 1.135521394010506e-05,
"loss": 0.1473,
"step": 834
},
{
"epoch": 3.386835443037975,
"grad_norm": 0.32627331366207873,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.1396,
"step": 835
},
{
"epoch": 3.390886075949367,
"grad_norm": 0.3347890725689411,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.1379,
"step": 836
},
{
"epoch": 3.3949367088607594,
"grad_norm": 0.31585503769414186,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.1516,
"step": 837
},
{
"epoch": 3.398987341772152,
"grad_norm": 0.3147811005888908,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.1598,
"step": 838
},
{
"epoch": 3.4030379746835444,
"grad_norm": 0.31270479096018927,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.1491,
"step": 839
},
{
"epoch": 3.407088607594937,
"grad_norm": 0.3190028539193045,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.1586,
"step": 840
},
{
"epoch": 3.411139240506329,
"grad_norm": 0.3306343643475561,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.1392,
"step": 841
},
{
"epoch": 3.4151898734177215,
"grad_norm": 0.31932460633092097,
"learning_rate": 1.0948015846744e-05,
"loss": 0.1493,
"step": 842
},
{
"epoch": 3.419240506329114,
"grad_norm": 0.3150570329753684,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.1677,
"step": 843
},
{
"epoch": 3.423291139240506,
"grad_norm": 0.32855232549241686,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.1378,
"step": 844
},
{
"epoch": 3.4273417721518986,
"grad_norm": 0.31304344419391555,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.1528,
"step": 845
},
{
"epoch": 3.431392405063291,
"grad_norm": 0.3148903832346968,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.1483,
"step": 846
},
{
"epoch": 3.4354430379746836,
"grad_norm": 0.3238844126726166,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.1491,
"step": 847
},
{
"epoch": 3.439493670886076,
"grad_norm": 0.3083870149813768,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.1563,
"step": 848
},
{
"epoch": 3.4435443037974682,
"grad_norm": 0.31175023425988124,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.1521,
"step": 849
},
{
"epoch": 3.4475949367088607,
"grad_norm": 0.30834303223330456,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.1546,
"step": 850
},
{
"epoch": 3.4516455696202533,
"grad_norm": 0.30822292493536,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.1646,
"step": 851
},
{
"epoch": 3.4556962025316453,
"grad_norm": 0.31579636834798597,
"learning_rate": 1.044560362975474e-05,
"loss": 0.1453,
"step": 852
},
{
"epoch": 3.459746835443038,
"grad_norm": 0.31255732324874436,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.1481,
"step": 853
},
{
"epoch": 3.4637974683544304,
"grad_norm": 0.31760054095060636,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.1482,
"step": 854
},
{
"epoch": 3.467848101265823,
"grad_norm": 0.29418983216578304,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.1472,
"step": 855
},
{
"epoch": 3.4718987341772154,
"grad_norm": 0.29728994932873315,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.1706,
"step": 856
},
{
"epoch": 3.4759493670886075,
"grad_norm": 0.31186257329326367,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.1518,
"step": 857
},
{
"epoch": 3.48,
"grad_norm": 0.32024056756043545,
"learning_rate": 1.014782400334433e-05,
"loss": 0.1757,
"step": 858
},
{
"epoch": 3.4840506329113925,
"grad_norm": 0.3154311416490225,
"learning_rate": 1.009846945771296e-05,
"loss": 0.1656,
"step": 859
},
{
"epoch": 3.4881012658227846,
"grad_norm": 0.3163522764413754,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.1326,
"step": 860
},
{
"epoch": 3.492151898734177,
"grad_norm": 0.3300040988297701,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.1541,
"step": 861
},
{
"epoch": 3.4962025316455696,
"grad_norm": 0.3145858611881271,
"learning_rate": 9.950885880977891e-06,
"loss": 0.1563,
"step": 862
},
{
"epoch": 3.500253164556962,
"grad_norm": 0.3277730339885276,
"learning_rate": 9.901852696139382e-06,
"loss": 0.139,
"step": 863
},
{
"epoch": 3.5043037974683546,
"grad_norm": 0.3106799936651795,
"learning_rate": 9.852900840391027e-06,
"loss": 0.1584,
"step": 864
},
{
"epoch": 3.5083544303797467,
"grad_norm": 0.33276093272438817,
"learning_rate": 9.804030707984313e-06,
"loss": 0.1586,
"step": 865
},
{
"epoch": 3.512405063291139,
"grad_norm": 0.33134914451699193,
"learning_rate": 9.755242692512599e-06,
"loss": 0.1565,
"step": 866
},
{
"epoch": 3.5164556962025317,
"grad_norm": 0.32534875837135097,
"learning_rate": 9.70653718690782e-06,
"loss": 0.1372,
"step": 867
},
{
"epoch": 3.520506329113924,
"grad_norm": 0.31219333409599265,
"learning_rate": 9.657914583437454e-06,
"loss": 0.141,
"step": 868
},
{
"epoch": 3.5245569620253163,
"grad_norm": 0.3193893921954364,
"learning_rate": 9.609375273701246e-06,
"loss": 0.1426,
"step": 869
},
{
"epoch": 3.528607594936709,
"grad_norm": 0.3040882704862924,
"learning_rate": 9.560919648628133e-06,
"loss": 0.1528,
"step": 870
},
{
"epoch": 3.5326582278481014,
"grad_norm": 0.3121526259321463,
"learning_rate": 9.512548098473047e-06,
"loss": 0.1325,
"step": 871
},
{
"epoch": 3.536708860759494,
"grad_norm": 0.31048995965626336,
"learning_rate": 9.464261012813825e-06,
"loss": 0.1633,
"step": 872
},
{
"epoch": 3.540759493670886,
"grad_norm": 0.33150633815349817,
"learning_rate": 9.416058780547987e-06,
"loss": 0.1286,
"step": 873
},
{
"epoch": 3.5448101265822785,
"grad_norm": 0.33175210296824387,
"learning_rate": 9.367941789889714e-06,
"loss": 0.1589,
"step": 874
},
{
"epoch": 3.548860759493671,
"grad_norm": 0.3177411263280725,
"learning_rate": 9.319910428366607e-06,
"loss": 0.1462,
"step": 875
},
{
"epoch": 3.552911392405063,
"grad_norm": 0.3171146555994605,
"learning_rate": 9.271965082816667e-06,
"loss": 0.1558,
"step": 876
},
{
"epoch": 3.5569620253164556,
"grad_norm": 0.3284774879688557,
"learning_rate": 9.224106139385111e-06,
"loss": 0.1408,
"step": 877
},
{
"epoch": 3.561012658227848,
"grad_norm": 0.31756244072777584,
"learning_rate": 9.176333983521291e-06,
"loss": 0.1761,
"step": 878
},
{
"epoch": 3.5650632911392406,
"grad_norm": 0.3070134274588859,
"learning_rate": 9.12864899997558e-06,
"loss": 0.1598,
"step": 879
},
{
"epoch": 3.569113924050633,
"grad_norm": 0.30888753006616976,
"learning_rate": 9.08105157279628e-06,
"loss": 0.1504,
"step": 880
},
{
"epoch": 3.573164556962025,
"grad_norm": 0.30413896195203427,
"learning_rate": 9.03354208532653e-06,
"loss": 0.143,
"step": 881
},
{
"epoch": 3.5772151898734177,
"grad_norm": 0.3217296223275184,
"learning_rate": 8.986120920201205e-06,
"loss": 0.1581,
"step": 882
},
{
"epoch": 3.58126582278481,
"grad_norm": 0.30487131954365876,
"learning_rate": 8.938788459343852e-06,
"loss": 0.1425,
"step": 883
},
{
"epoch": 3.5853164556962023,
"grad_norm": 0.3162572950354029,
"learning_rate": 8.8915450839636e-06,
"loss": 0.1513,
"step": 884
},
{
"epoch": 3.589367088607595,
"grad_norm": 0.32322611417886105,
"learning_rate": 8.844391174552116e-06,
"loss": 0.1568,
"step": 885
},
{
"epoch": 3.5934177215189873,
"grad_norm": 0.3196035238805944,
"learning_rate": 8.797327110880479e-06,
"loss": 0.1516,
"step": 886
},
{
"epoch": 3.59746835443038,
"grad_norm": 0.3132503914933546,
"learning_rate": 8.750353271996206e-06,
"loss": 0.1515,
"step": 887
},
{
"epoch": 3.6015189873417723,
"grad_norm": 0.31108275487126247,
"learning_rate": 8.703470036220132e-06,
"loss": 0.1414,
"step": 888
},
{
"epoch": 3.6055696202531644,
"grad_norm": 0.31707219828712396,
"learning_rate": 8.656677781143394e-06,
"loss": 0.1445,
"step": 889
},
{
"epoch": 3.609620253164557,
"grad_norm": 0.32720525894486463,
"learning_rate": 8.609976883624377e-06,
"loss": 0.1284,
"step": 890
},
{
"epoch": 3.6136708860759494,
"grad_norm": 0.3132731690282026,
"learning_rate": 8.563367719785698e-06,
"loss": 0.1538,
"step": 891
},
{
"epoch": 3.6177215189873415,
"grad_norm": 0.302103591737446,
"learning_rate": 8.516850665011138e-06,
"loss": 0.1427,
"step": 892
},
{
"epoch": 3.621772151898734,
"grad_norm": 0.3168504230367718,
"learning_rate": 8.47042609394269e-06,
"loss": 0.1606,
"step": 893
},
{
"epoch": 3.6258227848101265,
"grad_norm": 0.30861920079408717,
"learning_rate": 8.424094380477432e-06,
"loss": 0.1728,
"step": 894
},
{
"epoch": 3.629873417721519,
"grad_norm": 0.31416113908914206,
"learning_rate": 8.37785589776465e-06,
"loss": 0.1464,
"step": 895
},
{
"epoch": 3.6339240506329116,
"grad_norm": 0.30941284378202744,
"learning_rate": 8.331711018202694e-06,
"loss": 0.1548,
"step": 896
},
{
"epoch": 3.6379746835443036,
"grad_norm": 0.31786720994450446,
"learning_rate": 8.285660113436104e-06,
"loss": 0.1413,
"step": 897
},
{
"epoch": 3.642025316455696,
"grad_norm": 0.32734758147087367,
"learning_rate": 8.239703554352527e-06,
"loss": 0.144,
"step": 898
},
{
"epoch": 3.6460759493670887,
"grad_norm": 0.3204817422016448,
"learning_rate": 8.193841711079775e-06,
"loss": 0.1555,
"step": 899
},
{
"epoch": 3.6501265822784807,
"grad_norm": 0.31831437680567265,
"learning_rate": 8.148074952982828e-06,
"loss": 0.1715,
"step": 900
},
{
"epoch": 3.6541772151898733,
"grad_norm": 0.3098456134320011,
"learning_rate": 8.102403648660859e-06,
"loss": 0.1524,
"step": 901
},
{
"epoch": 3.6582278481012658,
"grad_norm": 0.30847820863010345,
"learning_rate": 8.056828165944282e-06,
"loss": 0.1352,
"step": 902
},
{
"epoch": 3.6622784810126583,
"grad_norm": 0.31534247043106844,
"learning_rate": 8.011348871891762e-06,
"loss": 0.1284,
"step": 903
},
{
"epoch": 3.666329113924051,
"grad_norm": 0.31396316743925956,
"learning_rate": 7.965966132787287e-06,
"loss": 0.1428,
"step": 904
},
{
"epoch": 3.670379746835443,
"grad_norm": 0.32403598023577934,
"learning_rate": 7.920680314137189e-06,
"loss": 0.1474,
"step": 905
},
{
"epoch": 3.6744303797468354,
"grad_norm": 0.3084643318774696,
"learning_rate": 7.875491780667246e-06,
"loss": 0.1234,
"step": 906
},
{
"epoch": 3.678481012658228,
"grad_norm": 0.31863264665300517,
"learning_rate": 7.830400896319667e-06,
"loss": 0.1404,
"step": 907
},
{
"epoch": 3.68253164556962,
"grad_norm": 0.31273029576787814,
"learning_rate": 7.785408024250259e-06,
"loss": 0.1428,
"step": 908
},
{
"epoch": 3.6865822784810125,
"grad_norm": 0.3117999018019873,
"learning_rate": 7.74051352682542e-06,
"loss": 0.1641,
"step": 909
},
{
"epoch": 3.690632911392405,
"grad_norm": 0.30210445222602483,
"learning_rate": 7.695717765619257e-06,
"loss": 0.1366,
"step": 910
},
{
"epoch": 3.6946835443037975,
"grad_norm": 0.30831759392492664,
"learning_rate": 7.651021101410673e-06,
"loss": 0.1384,
"step": 911
},
{
"epoch": 3.69873417721519,
"grad_norm": 0.30889782049820874,
"learning_rate": 7.606423894180464e-06,
"loss": 0.1407,
"step": 912
},
{
"epoch": 3.702784810126582,
"grad_norm": 0.33121870875369763,
"learning_rate": 7.56192650310839e-06,
"loss": 0.1422,
"step": 913
},
{
"epoch": 3.7068354430379746,
"grad_norm": 0.3272473481772408,
"learning_rate": 7.517529286570349e-06,
"loss": 0.1484,
"step": 914
},
{
"epoch": 3.710886075949367,
"grad_norm": 0.3108862426063153,
"learning_rate": 7.473232602135387e-06,
"loss": 0.1458,
"step": 915
},
{
"epoch": 3.714936708860759,
"grad_norm": 0.3157909196973785,
"learning_rate": 7.429036806562935e-06,
"loss": 0.1392,
"step": 916
},
{
"epoch": 3.7189873417721517,
"grad_norm": 0.3154785899599137,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.1532,
"step": 917
},
{
"epoch": 3.7230379746835442,
"grad_norm": 0.3074460755367421,
"learning_rate": 7.340949304977567e-06,
"loss": 0.1529,
"step": 918
},
{
"epoch": 3.7270886075949368,
"grad_norm": 0.3217617215195575,
"learning_rate": 7.297058308409282e-06,
"loss": 0.1528,
"step": 919
},
{
"epoch": 3.7311392405063293,
"grad_norm": 0.3176713599694772,
"learning_rate": 7.25326961958704e-06,
"loss": 0.1473,
"step": 920
},
{
"epoch": 3.7351898734177214,
"grad_norm": 0.3093560862145486,
"learning_rate": 7.209583591178921e-06,
"loss": 0.1621,
"step": 921
},
{
"epoch": 3.739240506329114,
"grad_norm": 0.31863951939020296,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.1456,
"step": 922
},
{
"epoch": 3.7432911392405064,
"grad_norm": 0.31650626138914595,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.1545,
"step": 923
},
{
"epoch": 3.747341772151899,
"grad_norm": 0.3055529961755167,
"learning_rate": 7.079144982700909e-06,
"loss": 0.13,
"step": 924
},
{
"epoch": 3.7513924050632914,
"grad_norm": 0.3184781864001659,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.1427,
"step": 925
},
{
"epoch": 3.7554430379746835,
"grad_norm": 0.28523111174985777,
"learning_rate": 6.99270564069757e-06,
"loss": 0.1474,
"step": 926
},
{
"epoch": 3.759493670886076,
"grad_norm": 0.31071209017070234,
"learning_rate": 6.949642934304375e-06,
"loss": 0.1626,
"step": 927
},
{
"epoch": 3.7635443037974685,
"grad_norm": 0.302262300381582,
"learning_rate": 6.906685333692871e-06,
"loss": 0.1324,
"step": 928
},
{
"epoch": 3.7675949367088606,
"grad_norm": 0.31551430535862385,
"learning_rate": 6.86383318483769e-06,
"loss": 0.1359,
"step": 929
},
{
"epoch": 3.771645569620253,
"grad_norm": 0.33360830285085646,
"learning_rate": 6.821086832864139e-06,
"loss": 0.1405,
"step": 930
},
{
"epoch": 3.7756962025316456,
"grad_norm": 0.30261792597241466,
"learning_rate": 6.77844662204546e-06,
"loss": 0.1478,
"step": 931
},
{
"epoch": 3.779746835443038,
"grad_norm": 0.3201711639219672,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.1359,
"step": 932
},
{
"epoch": 3.7837974683544306,
"grad_norm": 0.32201159173473287,
"learning_rate": 6.693485996688695e-06,
"loss": 0.1415,
"step": 933
},
{
"epoch": 3.7878481012658227,
"grad_norm": 0.3100263075890855,
"learning_rate": 6.651166266411801e-06,
"loss": 0.139,
"step": 934
},
{
"epoch": 3.7918987341772152,
"grad_norm": 0.30122838992572076,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.1497,
"step": 935
},
{
"epoch": 3.7959493670886078,
"grad_norm": 0.31307995981699527,
"learning_rate": 6.566849674844711e-06,
"loss": 0.1447,
"step": 936
},
{
"epoch": 3.8,
"grad_norm": 0.30556143432159344,
"learning_rate": 6.524853492628747e-06,
"loss": 0.1487,
"step": 937
},
{
"epoch": 3.8040506329113923,
"grad_norm": 0.29966683678769784,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.1481,
"step": 938
},
{
"epoch": 3.808101265822785,
"grad_norm": 0.3095579671811141,
"learning_rate": 6.441187046486648e-06,
"loss": 0.1475,
"step": 939
},
{
"epoch": 3.8121518987341774,
"grad_norm": 0.3029826638381046,
"learning_rate": 6.399517456398567e-06,
"loss": 0.1354,
"step": 940
},
{
"epoch": 3.81620253164557,
"grad_norm": 0.30414812618992987,
"learning_rate": 6.357957402727164e-06,
"loss": 0.1597,
"step": 941
},
{
"epoch": 3.820253164556962,
"grad_norm": 0.3189904752571902,
"learning_rate": 6.316507220191395e-06,
"loss": 0.1395,
"step": 942
},
{
"epoch": 3.8243037974683545,
"grad_norm": 0.2951418088734427,
"learning_rate": 6.275167242625331e-06,
"loss": 0.1523,
"step": 943
},
{
"epoch": 3.828354430379747,
"grad_norm": 0.3220483048493563,
"learning_rate": 6.233937802975471e-06,
"loss": 0.1514,
"step": 944
},
{
"epoch": 3.832405063291139,
"grad_norm": 0.3182432026663171,
"learning_rate": 6.192819233298046e-06,
"loss": 0.1418,
"step": 945
},
{
"epoch": 3.8364556962025316,
"grad_norm": 0.29791919976472975,
"learning_rate": 6.151811864756383e-06,
"loss": 0.1443,
"step": 946
},
{
"epoch": 3.840506329113924,
"grad_norm": 0.31965093087417046,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.1576,
"step": 947
},
{
"epoch": 3.8445569620253166,
"grad_norm": 0.2976599357168331,
"learning_rate": 6.070132051252868e-06,
"loss": 0.1427,
"step": 948
},
{
"epoch": 3.848607594936709,
"grad_norm": 0.30229914755862275,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.1476,
"step": 949
},
{
"epoch": 3.852658227848101,
"grad_norm": 0.31699148290929,
"learning_rate": 5.988900993811575e-06,
"loss": 0.1425,
"step": 950
},
{
"epoch": 3.8567088607594937,
"grad_norm": 0.30845292350407805,
"learning_rate": 5.948454566959363e-06,
"loss": 0.1435,
"step": 951
},
{
"epoch": 3.8607594936708862,
"grad_norm": 0.3135108770184291,
"learning_rate": 5.908121309322328e-06,
"loss": 0.1367,
"step": 952
},
{
"epoch": 3.8648101265822783,
"grad_norm": 0.31144864728121,
"learning_rate": 5.867901545738976e-06,
"loss": 0.1331,
"step": 953
},
{
"epoch": 3.868860759493671,
"grad_norm": 0.3070375017165889,
"learning_rate": 5.827795600133774e-06,
"loss": 0.1538,
"step": 954
},
{
"epoch": 3.8729113924050633,
"grad_norm": 0.3007615017937698,
"learning_rate": 5.787803795514466e-06,
"loss": 0.1528,
"step": 955
},
{
"epoch": 3.876962025316456,
"grad_norm": 0.3112366053157739,
"learning_rate": 5.747926453969576e-06,
"loss": 0.1336,
"step": 956
},
{
"epoch": 3.8810126582278484,
"grad_norm": 0.30237308707031535,
"learning_rate": 5.708163896665708e-06,
"loss": 0.1466,
"step": 957
},
{
"epoch": 3.8850632911392404,
"grad_norm": 0.29278690844447913,
"learning_rate": 5.668516443845047e-06,
"loss": 0.1454,
"step": 958
},
{
"epoch": 3.889113924050633,
"grad_norm": 0.3076842685938107,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.1365,
"step": 959
},
{
"epoch": 3.8931645569620255,
"grad_norm": 0.3065568690319972,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.1447,
"step": 960
},
{
"epoch": 3.8972151898734175,
"grad_norm": 0.3087839285877794,
"learning_rate": 5.550267900783019e-06,
"loss": 0.1324,
"step": 961
},
{
"epoch": 3.90126582278481,
"grad_norm": 0.3085139972067695,
"learning_rate": 5.511084049737623e-06,
"loss": 0.1485,
"step": 962
},
{
"epoch": 3.9053164556962026,
"grad_norm": 0.3006026532929946,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.1359,
"step": 963
},
{
"epoch": 3.909367088607595,
"grad_norm": 0.31662604188775245,
"learning_rate": 5.433066737499948e-06,
"loss": 0.1512,
"step": 964
},
{
"epoch": 3.9134177215189876,
"grad_norm": 0.30729412825139274,
"learning_rate": 5.394233904648376e-06,
"loss": 0.1514,
"step": 965
},
{
"epoch": 3.9174683544303797,
"grad_norm": 0.3127328084086466,
"learning_rate": 5.355518704628997e-06,
"loss": 0.1335,
"step": 966
},
{
"epoch": 3.921518987341772,
"grad_norm": 0.3145239072516153,
"learning_rate": 5.316921449248731e-06,
"loss": 0.1634,
"step": 967
},
{
"epoch": 3.9255696202531647,
"grad_norm": 0.31047336370576323,
"learning_rate": 5.278442449364538e-06,
"loss": 0.1402,
"step": 968
},
{
"epoch": 3.9296202531645568,
"grad_norm": 0.30961625461142755,
"learning_rate": 5.240082014881016e-06,
"loss": 0.1369,
"step": 969
},
{
"epoch": 3.9336708860759493,
"grad_norm": 0.3202714232730156,
"learning_rate": 5.201840454747822e-06,
"loss": 0.1577,
"step": 970
},
{
"epoch": 3.937721518987342,
"grad_norm": 0.318132080725971,
"learning_rate": 5.163718076957223e-06,
"loss": 0.1523,
"step": 971
},
{
"epoch": 3.9417721518987343,
"grad_norm": 0.30223478779442664,
"learning_rate": 5.125715188541609e-06,
"loss": 0.1358,
"step": 972
},
{
"epoch": 3.945822784810127,
"grad_norm": 0.30677035114001355,
"learning_rate": 5.087832095571021e-06,
"loss": 0.139,
"step": 973
},
{
"epoch": 3.949873417721519,
"grad_norm": 0.2922897352411125,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.1477,
"step": 974
},
{
"epoch": 3.9539240506329114,
"grad_norm": 0.3146431831771742,
"learning_rate": 5.01242651541854e-06,
"loss": 0.1501,
"step": 975
},
{
"epoch": 3.957974683544304,
"grad_norm": 0.3105171249261358,
"learning_rate": 4.974904635542815e-06,
"loss": 0.1415,
"step": 976
},
{
"epoch": 3.962025316455696,
"grad_norm": 0.30649426641977023,
"learning_rate": 4.937503765719582e-06,
"loss": 0.1428,
"step": 977
},
{
"epoch": 3.9660759493670885,
"grad_norm": 0.31008762004185564,
"learning_rate": 4.900224207170299e-06,
"loss": 0.1383,
"step": 978
},
{
"epoch": 3.970126582278481,
"grad_norm": 0.3056115389547901,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.1489,
"step": 979
},
{
"epoch": 3.9741772151898735,
"grad_norm": 0.3016434598099687,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.16,
"step": 980
},
{
"epoch": 3.978227848101266,
"grad_norm": 0.30802520630849095,
"learning_rate": 4.789116396710924e-06,
"loss": 0.1361,
"step": 981
},
{
"epoch": 3.982278481012658,
"grad_norm": 0.2980134900139868,
"learning_rate": 4.752325075895368e-06,
"loss": 0.1629,
"step": 982
},
{
"epoch": 3.9863291139240506,
"grad_norm": 0.3020071680285737,
"learning_rate": 4.715656557757473e-06,
"loss": 0.1543,
"step": 983
},
{
"epoch": 3.990379746835443,
"grad_norm": 0.29976506406843656,
"learning_rate": 4.679111137620442e-06,
"loss": 0.1473,
"step": 984
},
{
"epoch": 3.9944303797468352,
"grad_norm": 0.3031642699528218,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.1444,
"step": 985
},
{
"epoch": 3.9984810126582278,
"grad_norm": 0.2981308616751398,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.1323,
"step": 986
},
{
"epoch": 4.004050632911392,
"grad_norm": 0.3558340206881297,
"learning_rate": 4.570216403561141e-06,
"loss": 0.0924,
"step": 987
},
{
"epoch": 4.008101265822785,
"grad_norm": 0.34502200102323105,
"learning_rate": 4.534166308795815e-06,
"loss": 0.1093,
"step": 988
},
{
"epoch": 4.012151898734177,
"grad_norm": 0.27968856718252394,
"learning_rate": 4.498240773728859e-06,
"loss": 0.1046,
"step": 989
},
{
"epoch": 4.01620253164557,
"grad_norm": 0.23225492857126287,
"learning_rate": 4.462440087699609e-06,
"loss": 0.0813,
"step": 990
},
{
"epoch": 4.020253164556962,
"grad_norm": 0.28989877221589594,
"learning_rate": 4.426764539041861e-06,
"loss": 0.1065,
"step": 991
},
{
"epoch": 4.024303797468354,
"grad_norm": 0.3181456528890928,
"learning_rate": 4.391214415081582e-06,
"loss": 0.1031,
"step": 992
},
{
"epoch": 4.028354430379747,
"grad_norm": 0.3681158388476263,
"learning_rate": 4.355790002134579e-06,
"loss": 0.1076,
"step": 993
},
{
"epoch": 4.032405063291139,
"grad_norm": 0.37347183490679275,
"learning_rate": 4.320491585504207e-06,
"loss": 0.1081,
"step": 994
},
{
"epoch": 4.036455696202531,
"grad_norm": 0.2956715309826616,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.0826,
"step": 995
},
{
"epoch": 4.040506329113924,
"grad_norm": 0.29571620690783396,
"learning_rate": 4.250273877330691e-06,
"loss": 0.1029,
"step": 996
},
{
"epoch": 4.044556962025316,
"grad_norm": 0.27287568832424863,
"learning_rate": 4.215355151311313e-06,
"loss": 0.0878,
"step": 997
},
{
"epoch": 4.048607594936709,
"grad_norm": 0.2809500239608938,
"learning_rate": 4.180563552651542e-06,
"loss": 0.1122,
"step": 998
},
{
"epoch": 4.052658227848101,
"grad_norm": 0.30895645891773005,
"learning_rate": 4.145899361558147e-06,
"loss": 0.0923,
"step": 999
},
{
"epoch": 4.056708860759493,
"grad_norm": 0.29045054051528174,
"learning_rate": 4.111362857211738e-06,
"loss": 0.1091,
"step": 1000
},
{
"epoch": 4.060759493670886,
"grad_norm": 0.2585141072444677,
"learning_rate": 4.076954317764592e-06,
"loss": 0.1202,
"step": 1001
},
{
"epoch": 4.0648101265822785,
"grad_norm": 0.26043167920330096,
"learning_rate": 4.042674020338335e-06,
"loss": 0.0993,
"step": 1002
},
{
"epoch": 4.0688607594936705,
"grad_norm": 0.24378583614081153,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.0956,
"step": 1003
},
{
"epoch": 4.0729113924050635,
"grad_norm": 0.2641291440181553,
"learning_rate": 3.974499254868674e-06,
"loss": 0.0924,
"step": 1004
},
{
"epoch": 4.076962025316456,
"grad_norm": 0.2620807499207312,
"learning_rate": 3.940605335895451e-06,
"loss": 0.1046,
"step": 1005
},
{
"epoch": 4.0810126582278485,
"grad_norm": 0.26451206478014955,
"learning_rate": 3.90684075707908e-06,
"loss": 0.0826,
"step": 1006
},
{
"epoch": 4.085063291139241,
"grad_norm": 0.288765866196326,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.1115,
"step": 1007
},
{
"epoch": 4.089113924050633,
"grad_norm": 0.2721583540496314,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.1028,
"step": 1008
},
{
"epoch": 4.093164556962026,
"grad_norm": 0.25598987590399264,
"learning_rate": 3.806325775702304e-06,
"loss": 0.0858,
"step": 1009
},
{
"epoch": 4.097215189873418,
"grad_norm": 0.2656508082042958,
"learning_rate": 3.773081266416434e-06,
"loss": 0.1109,
"step": 1010
},
{
"epoch": 4.10126582278481,
"grad_norm": 0.24286308777968169,
"learning_rate": 3.739967446503245e-06,
"loss": 0.0932,
"step": 1011
},
{
"epoch": 4.105316455696203,
"grad_norm": 0.24994705125108346,
"learning_rate": 3.706984582656894e-06,
"loss": 0.0968,
"step": 1012
},
{
"epoch": 4.109367088607595,
"grad_norm": 0.2503446461499507,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.1002,
"step": 1013
},
{
"epoch": 4.113417721518988,
"grad_norm": 0.24391898423195854,
"learning_rate": 3.641412784665648e-06,
"loss": 0.1134,
"step": 1014
},
{
"epoch": 4.11746835443038,
"grad_norm": 0.24663616713133288,
"learning_rate": 3.608824378627005e-06,
"loss": 0.1179,
"step": 1015
},
{
"epoch": 4.121518987341772,
"grad_norm": 0.26043251975664705,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.079,
"step": 1016
},
{
"epoch": 4.125569620253165,
"grad_norm": 0.2721184323686397,
"learning_rate": 3.544043864774269e-06,
"loss": 0.1135,
"step": 1017
},
{
"epoch": 4.129620253164557,
"grad_norm": 0.25164066299594573,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.1076,
"step": 1018
},
{
"epoch": 4.133670886075949,
"grad_norm": 0.2432517118473237,
"learning_rate": 3.479793485887819e-06,
"loss": 0.1133,
"step": 1019
},
{
"epoch": 4.137721518987342,
"grad_norm": 0.2459812444209647,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.0994,
"step": 1020
},
{
"epoch": 4.141772151898734,
"grad_norm": 0.24816460558784398,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.1113,
"step": 1021
},
{
"epoch": 4.145822784810127,
"grad_norm": 0.2590212213001976,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.0849,
"step": 1022
},
{
"epoch": 4.149873417721519,
"grad_norm": 0.24518439306306553,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.092,
"step": 1023
},
{
"epoch": 4.153924050632911,
"grad_norm": 0.25628725601098773,
"learning_rate": 3.321500420335e-06,
"loss": 0.1116,
"step": 1024
},
{
"epoch": 4.157974683544304,
"grad_norm": 0.2714742091648926,
"learning_rate": 3.290243771741275e-06,
"loss": 0.0988,
"step": 1025
},
{
"epoch": 4.162025316455696,
"grad_norm": 0.2605441940720298,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.0864,
"step": 1026
},
{
"epoch": 4.166075949367088,
"grad_norm": 0.24949131837446628,
"learning_rate": 3.228134459445149e-06,
"loss": 0.0963,
"step": 1027
},
{
"epoch": 4.170126582278481,
"grad_norm": 0.250173751137564,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.1097,
"step": 1028
},
{
"epoch": 4.174177215189873,
"grad_norm": 0.24985337512944367,
"learning_rate": 3.166565459258513e-06,
"loss": 0.0964,
"step": 1029
},
{
"epoch": 4.178227848101266,
"grad_norm": 0.23178076967916716,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.093,
"step": 1030
},
{
"epoch": 4.182278481012658,
"grad_norm": 0.2628378505603558,
"learning_rate": 3.105538754650419e-06,
"loss": 0.0822,
"step": 1031
},
{
"epoch": 4.18632911392405,
"grad_norm": 0.2406927544062428,
"learning_rate": 3.07522937824712e-06,
"loss": 0.0799,
"step": 1032
},
{
"epoch": 4.190379746835443,
"grad_norm": 0.24189835843142532,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.0892,
"step": 1033
},
{
"epoch": 4.194430379746835,
"grad_norm": 0.24990880722433306,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.1008,
"step": 1034
},
{
"epoch": 4.1984810126582275,
"grad_norm": 0.25305889555752237,
"learning_rate": 2.985120078631465e-06,
"loss": 0.1016,
"step": 1035
},
{
"epoch": 4.2025316455696204,
"grad_norm": 0.24306626701983633,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.1074,
"step": 1036
},
{
"epoch": 4.2065822784810125,
"grad_norm": 0.24904544287056438,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.0895,
"step": 1037
},
{
"epoch": 4.2106329113924055,
"grad_norm": 0.2541399778644939,
"learning_rate": 2.896244091927678e-06,
"loss": 0.1172,
"step": 1038
},
{
"epoch": 4.2146835443037975,
"grad_norm": 0.23645425381997176,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.1022,
"step": 1039
},
{
"epoch": 4.21873417721519,
"grad_norm": 0.24244324738195547,
"learning_rate": 2.837681792951994e-06,
"loss": 0.117,
"step": 1040
},
{
"epoch": 4.222784810126583,
"grad_norm": 0.25448228968235304,
"learning_rate": 2.808607860255981e-06,
"loss": 0.1084,
"step": 1041
},
{
"epoch": 4.226835443037975,
"grad_norm": 0.24308797054960865,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.0974,
"step": 1042
},
{
"epoch": 4.230886075949367,
"grad_norm": 0.22606936677942935,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.0876,
"step": 1043
},
{
"epoch": 4.23493670886076,
"grad_norm": 0.2388742841176979,
"learning_rate": 2.722217735873718e-06,
"loss": 0.1056,
"step": 1044
},
{
"epoch": 4.238987341772152,
"grad_norm": 0.24480376753047847,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.0952,
"step": 1045
},
{
"epoch": 4.243037974683545,
"grad_norm": 0.23088187815727887,
"learning_rate": 2.665319697490205e-06,
"loss": 0.0885,
"step": 1046
},
{
"epoch": 4.247088607594937,
"grad_norm": 0.24346170446122484,
"learning_rate": 2.637079980714945e-06,
"loss": 0.1084,
"step": 1047
},
{
"epoch": 4.251139240506329,
"grad_norm": 0.23457773266789336,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.0894,
"step": 1048
},
{
"epoch": 4.255189873417722,
"grad_norm": 0.2439345969754364,
"learning_rate": 2.581020289189571e-06,
"loss": 0.0952,
"step": 1049
},
{
"epoch": 4.259240506329114,
"grad_norm": 0.2634376203307505,
"learning_rate": 2.553200765936501e-06,
"loss": 0.0915,
"step": 1050
},
{
"epoch": 4.263291139240506,
"grad_norm": 0.2730053226305486,
"learning_rate": 2.525521756805962e-06,
"loss": 0.1004,
"step": 1051
},
{
"epoch": 4.267341772151899,
"grad_norm": 0.2541307549799196,
"learning_rate": 2.497983484720885e-06,
"loss": 0.1099,
"step": 1052
},
{
"epoch": 4.271392405063291,
"grad_norm": 0.2432864258744948,
"learning_rate": 2.470586171470728e-06,
"loss": 0.0935,
"step": 1053
},
{
"epoch": 4.275443037974684,
"grad_norm": 0.2539523951074236,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.0933,
"step": 1054
},
{
"epoch": 4.279493670886076,
"grad_norm": 0.24133102578897433,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.092,
"step": 1055
},
{
"epoch": 4.283544303797468,
"grad_norm": 0.2434020007764359,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.1001,
"step": 1056
},
{
"epoch": 4.287594936708861,
"grad_norm": 0.2603179024595173,
"learning_rate": 2.362410902836978e-06,
"loss": 0.1146,
"step": 1057
},
{
"epoch": 4.291645569620253,
"grad_norm": 0.2450296666045376,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.1131,
"step": 1058
},
{
"epoch": 4.295696202531645,
"grad_norm": 0.24078097790109815,
"learning_rate": 2.309174704446411e-06,
"loss": 0.0872,
"step": 1059
},
{
"epoch": 4.299746835443038,
"grad_norm": 0.24925869682679436,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.0961,
"step": 1060
},
{
"epoch": 4.30379746835443,
"grad_norm": 0.2529656209388516,
"learning_rate": 2.256508422807855e-06,
"loss": 0.1033,
"step": 1061
},
{
"epoch": 4.307848101265823,
"grad_norm": 0.24281278596429273,
"learning_rate": 2.230389531696946e-06,
"loss": 0.0939,
"step": 1062
},
{
"epoch": 4.311898734177215,
"grad_norm": 0.2510786654189413,
"learning_rate": 2.204413754585857e-06,
"loss": 0.0913,
"step": 1063
},
{
"epoch": 4.315949367088607,
"grad_norm": 0.25854597344422414,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.0987,
"step": 1064
},
{
"epoch": 4.32,
"grad_norm": 0.2524253297260038,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.1085,
"step": 1065
},
{
"epoch": 4.324050632911392,
"grad_norm": 0.24843532734817617,
"learning_rate": 2.127347193531757e-06,
"loss": 0.0781,
"step": 1066
},
{
"epoch": 4.328101265822784,
"grad_norm": 0.24459625006523203,
"learning_rate": 2.101945952921942e-06,
"loss": 0.0916,
"step": 1067
},
{
"epoch": 4.332151898734177,
"grad_norm": 0.2469781446294768,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.0894,
"step": 1068
},
{
"epoch": 4.3362025316455695,
"grad_norm": 0.23572355242382517,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.097,
"step": 1069
},
{
"epoch": 4.340253164556962,
"grad_norm": 0.24782256408066786,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.0889,
"step": 1070
},
{
"epoch": 4.3443037974683545,
"grad_norm": 0.23220469202174557,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.0905,
"step": 1071
},
{
"epoch": 4.348354430379747,
"grad_norm": 0.23264327044508826,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.0874,
"step": 1072
},
{
"epoch": 4.3524050632911395,
"grad_norm": 0.23918168586813177,
"learning_rate": 1.95257270594031e-06,
"loss": 0.0945,
"step": 1073
},
{
"epoch": 4.356455696202532,
"grad_norm": 0.24240743083751845,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.1052,
"step": 1074
},
{
"epoch": 4.360506329113924,
"grad_norm": 0.2413307394993314,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.1124,
"step": 1075
},
{
"epoch": 4.364556962025317,
"grad_norm": 0.2404355661314276,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.0954,
"step": 1076
},
{
"epoch": 4.368607594936709,
"grad_norm": 0.2535468987055318,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.117,
"step": 1077
},
{
"epoch": 4.372658227848102,
"grad_norm": 0.2332213068929193,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.1109,
"step": 1078
},
{
"epoch": 4.376708860759494,
"grad_norm": 0.25513693608968524,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.0768,
"step": 1079
},
{
"epoch": 4.380759493670886,
"grad_norm": 0.2367725058743154,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.0948,
"step": 1080
},
{
"epoch": 4.384810126582279,
"grad_norm": 0.24396953812212605,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.1041,
"step": 1081
},
{
"epoch": 4.388860759493671,
"grad_norm": 0.2421347596085938,
"learning_rate": 1.738336950434061e-06,
"loss": 0.1105,
"step": 1082
},
{
"epoch": 4.392911392405063,
"grad_norm": 0.25828614608841544,
"learning_rate": 1.715265841375957e-06,
"loss": 0.12,
"step": 1083
},
{
"epoch": 4.396962025316456,
"grad_norm": 0.24264569613262377,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.0935,
"step": 1084
},
{
"epoch": 4.401012658227848,
"grad_norm": 0.2505364348049405,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.0962,
"step": 1085
},
{
"epoch": 4.405063291139241,
"grad_norm": 0.241735504355208,
"learning_rate": 1.646936827927441e-06,
"loss": 0.1218,
"step": 1086
},
{
"epoch": 4.409113924050633,
"grad_norm": 0.23566557403360938,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.0907,
"step": 1087
},
{
"epoch": 4.413164556962025,
"grad_norm": 0.24141473110324943,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.115,
"step": 1088
},
{
"epoch": 4.417215189873418,
"grad_norm": 0.24990499426421084,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.1054,
"step": 1089
},
{
"epoch": 4.42126582278481,
"grad_norm": 0.23610736181021164,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.0894,
"step": 1090
},
{
"epoch": 4.425316455696202,
"grad_norm": 0.2322412685806906,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.0953,
"step": 1091
},
{
"epoch": 4.429367088607595,
"grad_norm": 0.24387401661254499,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.1049,
"step": 1092
},
{
"epoch": 4.433417721518987,
"grad_norm": 0.24479750232114988,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.0974,
"step": 1093
},
{
"epoch": 4.43746835443038,
"grad_norm": 0.23681098463082925,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.0954,
"step": 1094
},
{
"epoch": 4.441518987341772,
"grad_norm": 0.2559607310623418,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.0789,
"step": 1095
},
{
"epoch": 4.445569620253164,
"grad_norm": 0.25171997255457323,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.0881,
"step": 1096
},
{
"epoch": 4.449620253164557,
"grad_norm": 0.2550882053028517,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.0909,
"step": 1097
},
{
"epoch": 4.453670886075949,
"grad_norm": 0.2419666511654711,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.1046,
"step": 1098
},
{
"epoch": 4.457721518987341,
"grad_norm": 0.2420262730750023,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.0996,
"step": 1099
},
{
"epoch": 4.461772151898734,
"grad_norm": 0.24634519411073572,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.0901,
"step": 1100
},
{
"epoch": 4.465822784810126,
"grad_norm": 0.2558315608985439,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.1019,
"step": 1101
},
{
"epoch": 4.469873417721519,
"grad_norm": 0.2488172879642797,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.0973,
"step": 1102
},
{
"epoch": 4.473924050632911,
"grad_norm": 0.2391676198388007,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.0856,
"step": 1103
},
{
"epoch": 4.4779746835443035,
"grad_norm": 0.2374739554850736,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.1011,
"step": 1104
},
{
"epoch": 4.4820253164556965,
"grad_norm": 0.24396909124782282,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.0938,
"step": 1105
},
{
"epoch": 4.4860759493670885,
"grad_norm": 0.24359795000088286,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.0814,
"step": 1106
},
{
"epoch": 4.490126582278481,
"grad_norm": 0.25084104617503883,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.0941,
"step": 1107
},
{
"epoch": 4.494177215189874,
"grad_norm": 0.2463426717307214,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.088,
"step": 1108
},
{
"epoch": 4.498227848101266,
"grad_norm": 0.22730972545528128,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.1111,
"step": 1109
},
{
"epoch": 4.502278481012659,
"grad_norm": 0.23943911545446325,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.0914,
"step": 1110
},
{
"epoch": 4.506329113924051,
"grad_norm": 0.23792429859521874,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.0975,
"step": 1111
},
{
"epoch": 4.510379746835443,
"grad_norm": 0.2228990079573159,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.0975,
"step": 1112
},
{
"epoch": 4.514430379746836,
"grad_norm": 0.2525812653312759,
"learning_rate": 1.092402023876933e-06,
"loss": 0.1229,
"step": 1113
},
{
"epoch": 4.518481012658228,
"grad_norm": 0.23828397666703893,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.0834,
"step": 1114
},
{
"epoch": 4.52253164556962,
"grad_norm": 0.23802015620719824,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.1005,
"step": 1115
},
{
"epoch": 4.526582278481013,
"grad_norm": 0.25184217838320233,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.113,
"step": 1116
},
{
"epoch": 4.530632911392405,
"grad_norm": 0.23029124991723612,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.0948,
"step": 1117
},
{
"epoch": 4.534683544303798,
"grad_norm": 0.24740843160042195,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.1091,
"step": 1118
},
{
"epoch": 4.53873417721519,
"grad_norm": 0.2510285524365786,
"learning_rate": 9.841383857091947e-07,
"loss": 0.0902,
"step": 1119
},
{
"epoch": 4.542784810126582,
"grad_norm": 0.23080794325116022,
"learning_rate": 9.666296380321616e-07,
"loss": 0.096,
"step": 1120
},
{
"epoch": 4.546835443037975,
"grad_norm": 0.24569708827469888,
"learning_rate": 9.492741825034124e-07,
"loss": 0.0868,
"step": 1121
},
{
"epoch": 4.550886075949367,
"grad_norm": 0.24028877738357918,
"learning_rate": 9.320721589013892e-07,
"loss": 0.1021,
"step": 1122
},
{
"epoch": 4.55493670886076,
"grad_norm": 0.2450270024690331,
"learning_rate": 9.150237057688339e-07,
"loss": 0.0887,
"step": 1123
},
{
"epoch": 4.558987341772152,
"grad_norm": 0.24005946921808805,
"learning_rate": 8.981289604116328e-07,
"loss": 0.0935,
"step": 1124
},
{
"epoch": 4.563037974683544,
"grad_norm": 0.24150753434320682,
"learning_rate": 8.813880588977542e-07,
"loss": 0.1066,
"step": 1125
},
{
"epoch": 4.567088607594937,
"grad_norm": 0.2538260480500532,
"learning_rate": 8.648011360561126e-07,
"loss": 0.1093,
"step": 1126
},
{
"epoch": 4.571139240506329,
"grad_norm": 0.23720379866353322,
"learning_rate": 8.483683254755037e-07,
"loss": 0.0871,
"step": 1127
},
{
"epoch": 4.575189873417721,
"grad_norm": 0.24003751012408142,
"learning_rate": 8.320897595035227e-07,
"loss": 0.1051,
"step": 1128
},
{
"epoch": 4.579240506329114,
"grad_norm": 0.2389937646784995,
"learning_rate": 8.159655692455093e-07,
"loss": 0.0905,
"step": 1129
},
{
"epoch": 4.583291139240506,
"grad_norm": 0.24940976854248162,
"learning_rate": 7.999958845634648e-07,
"loss": 0.1068,
"step": 1130
},
{
"epoch": 4.587341772151898,
"grad_norm": 0.24680502836750384,
"learning_rate": 7.841808340750478e-07,
"loss": 0.0873,
"step": 1131
},
{
"epoch": 4.591392405063291,
"grad_norm": 0.24864159863980986,
"learning_rate": 7.685205451524869e-07,
"loss": 0.0964,
"step": 1132
},
{
"epoch": 4.595443037974683,
"grad_norm": 0.229623243983979,
"learning_rate": 7.530151439216027e-07,
"loss": 0.1152,
"step": 1133
},
{
"epoch": 4.599493670886076,
"grad_norm": 0.24862102604000794,
"learning_rate": 7.376647552607675e-07,
"loss": 0.086,
"step": 1134
},
{
"epoch": 4.603544303797468,
"grad_norm": 0.23424671428210067,
"learning_rate": 7.224695027998963e-07,
"loss": 0.1096,
"step": 1135
},
{
"epoch": 4.6075949367088604,
"grad_norm": 0.255905845138719,
"learning_rate": 7.07429508919466e-07,
"loss": 0.1007,
"step": 1136
},
{
"epoch": 4.611645569620253,
"grad_norm": 0.2355994849617788,
"learning_rate": 6.925448947495206e-07,
"loss": 0.0926,
"step": 1137
},
{
"epoch": 4.6156962025316455,
"grad_norm": 0.28090779257535353,
"learning_rate": 6.778157801686936e-07,
"loss": 0.0919,
"step": 1138
},
{
"epoch": 4.619746835443038,
"grad_norm": 0.2418843234150777,
"learning_rate": 6.632422838032515e-07,
"loss": 0.0947,
"step": 1139
},
{
"epoch": 4.6237974683544305,
"grad_norm": 0.24623892575568362,
"learning_rate": 6.488245230261281e-07,
"loss": 0.1099,
"step": 1140
},
{
"epoch": 4.627848101265823,
"grad_norm": 0.2449815983940914,
"learning_rate": 6.345626139559868e-07,
"loss": 0.0876,
"step": 1141
},
{
"epoch": 4.6318987341772155,
"grad_norm": 0.2528744286931328,
"learning_rate": 6.204566714562866e-07,
"loss": 0.0895,
"step": 1142
},
{
"epoch": 4.635949367088608,
"grad_norm": 0.24676327741856846,
"learning_rate": 6.06506809134344e-07,
"loss": 0.1065,
"step": 1143
},
{
"epoch": 4.64,
"grad_norm": 0.24720130401995472,
"learning_rate": 5.927131393404373e-07,
"loss": 0.083,
"step": 1144
},
{
"epoch": 4.644050632911393,
"grad_norm": 0.23763894896425716,
"learning_rate": 5.790757731668817e-07,
"loss": 0.1076,
"step": 1145
},
{
"epoch": 4.648101265822785,
"grad_norm": 0.24872849976616943,
"learning_rate": 5.655948204471507e-07,
"loss": 0.0894,
"step": 1146
},
{
"epoch": 4.652151898734177,
"grad_norm": 0.25924259798660626,
"learning_rate": 5.522703897549875e-07,
"loss": 0.0964,
"step": 1147
},
{
"epoch": 4.65620253164557,
"grad_norm": 0.24033796140258157,
"learning_rate": 5.391025884035239e-07,
"loss": 0.1,
"step": 1148
},
{
"epoch": 4.660253164556962,
"grad_norm": 0.2453653937574591,
"learning_rate": 5.260915224444207e-07,
"loss": 0.1122,
"step": 1149
},
{
"epoch": 4.664303797468355,
"grad_norm": 0.2503879316898561,
"learning_rate": 5.132372966670129e-07,
"loss": 0.1103,
"step": 1150
},
{
"epoch": 4.668354430379747,
"grad_norm": 0.24192654207904535,
"learning_rate": 5.005400145974704e-07,
"loss": 0.0831,
"step": 1151
},
{
"epoch": 4.672405063291139,
"grad_norm": 0.22948005110593975,
"learning_rate": 4.879997784979562e-07,
"loss": 0.114,
"step": 1152
},
{
"epoch": 4.676455696202532,
"grad_norm": 0.23727160783460854,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.0835,
"step": 1153
},
{
"epoch": 4.680506329113924,
"grad_norm": 0.23860061454826773,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.1277,
"step": 1154
},
{
"epoch": 4.684556962025317,
"grad_norm": 0.2343074390542557,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.0882,
"step": 1155
},
{
"epoch": 4.688607594936709,
"grad_norm": 0.24240326599668569,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.0781,
"step": 1156
},
{
"epoch": 4.692658227848101,
"grad_norm": 0.24404648160277057,
"learning_rate": 4.27657778143431e-07,
"loss": 0.1087,
"step": 1157
},
{
"epoch": 4.696708860759494,
"grad_norm": 0.24084834675136038,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.0854,
"step": 1158
},
{
"epoch": 4.700759493670886,
"grad_norm": 0.23754137731458308,
"learning_rate": 4.046237371675177e-07,
"loss": 0.1098,
"step": 1159
},
{
"epoch": 4.704810126582278,
"grad_norm": 0.24040759902052825,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.1053,
"step": 1160
},
{
"epoch": 4.708860759493671,
"grad_norm": 0.22857889900841014,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.1085,
"step": 1161
},
{
"epoch": 4.712911392405063,
"grad_norm": 0.24650758630669575,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.0852,
"step": 1162
},
{
"epoch": 4.716962025316455,
"grad_norm": 0.24085697346761076,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.0871,
"step": 1163
},
{
"epoch": 4.721012658227848,
"grad_norm": 0.23605634680708412,
"learning_rate": 3.498020255137813e-07,
"loss": 0.0914,
"step": 1164
},
{
"epoch": 4.72506329113924,
"grad_norm": 0.22708890094136716,
"learning_rate": 3.393121163057811e-07,
"loss": 0.1006,
"step": 1165
},
{
"epoch": 4.729113924050633,
"grad_norm": 0.2325615975591414,
"learning_rate": 3.289805515751399e-07,
"loss": 0.0915,
"step": 1166
},
{
"epoch": 4.733164556962025,
"grad_norm": 0.24938130986724005,
"learning_rate": 3.188074145308573e-07,
"loss": 0.0918,
"step": 1167
},
{
"epoch": 4.737215189873417,
"grad_norm": 0.24389080798474863,
"learning_rate": 3.087927871059804e-07,
"loss": 0.0866,
"step": 1168
},
{
"epoch": 4.74126582278481,
"grad_norm": 0.23312269444421146,
"learning_rate": 2.989367499569418e-07,
"loss": 0.1078,
"step": 1169
},
{
"epoch": 4.745316455696202,
"grad_norm": 0.23536897416882208,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.0881,
"step": 1170
},
{
"epoch": 4.749367088607595,
"grad_norm": 0.23388675203708603,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.0724,
"step": 1171
},
{
"epoch": 4.7534177215189874,
"grad_norm": 0.23587498318832023,
"learning_rate": 2.703209675663887e-07,
"loss": 0.1075,
"step": 1172
},
{
"epoch": 4.7574683544303795,
"grad_norm": 0.23189251928255353,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.0974,
"step": 1173
},
{
"epoch": 4.7615189873417725,
"grad_norm": 0.24603080192411414,
"learning_rate": 2.520381518804471e-07,
"loss": 0.0952,
"step": 1174
},
{
"epoch": 4.7655696202531646,
"grad_norm": 0.23630857290581672,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.1026,
"step": 1175
},
{
"epoch": 4.769620253164557,
"grad_norm": 0.2412884340447727,
"learning_rate": 2.343915243930317e-07,
"loss": 0.0897,
"step": 1176
},
{
"epoch": 4.77367088607595,
"grad_norm": 0.23883098258089952,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.0837,
"step": 1177
},
{
"epoch": 4.777721518987342,
"grad_norm": 0.23943491058590177,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.0955,
"step": 1178
},
{
"epoch": 4.781772151898734,
"grad_norm": 0.2609199600820985,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.1036,
"step": 1179
},
{
"epoch": 4.785822784810127,
"grad_norm": 0.24455726606312378,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.0987,
"step": 1180
},
{
"epoch": 4.789873417721519,
"grad_norm": 0.2471131270061707,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.0839,
"step": 1181
},
{
"epoch": 4.793924050632912,
"grad_norm": 0.23507637479437243,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.1016,
"step": 1182
},
{
"epoch": 4.797974683544304,
"grad_norm": 0.24310307764426214,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.0856,
"step": 1183
},
{
"epoch": 4.802025316455696,
"grad_norm": 0.2440816092257352,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.097,
"step": 1184
},
{
"epoch": 4.806075949367089,
"grad_norm": 0.24799959574504893,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.1082,
"step": 1185
},
{
"epoch": 4.810126582278481,
"grad_norm": 0.2280227650383461,
"learning_rate": 1.557203888516745e-07,
"loss": 0.1058,
"step": 1186
},
{
"epoch": 4.814177215189874,
"grad_norm": 0.22542133526692434,
"learning_rate": 1.487313098987131e-07,
"loss": 0.0776,
"step": 1187
},
{
"epoch": 4.818227848101266,
"grad_norm": 0.23867692091512954,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.0881,
"step": 1188
},
{
"epoch": 4.822278481012658,
"grad_norm": 0.24137712943345913,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.0883,
"step": 1189
},
{
"epoch": 4.826329113924051,
"grad_norm": 0.23256938686055353,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.1026,
"step": 1190
},
{
"epoch": 4.830379746835443,
"grad_norm": 0.2528468011084582,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.1092,
"step": 1191
},
{
"epoch": 4.834430379746835,
"grad_norm": 0.2461740594948306,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.0951,
"step": 1192
},
{
"epoch": 4.838481012658228,
"grad_norm": 0.2352626616439655,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.1012,
"step": 1193
},
{
"epoch": 4.84253164556962,
"grad_norm": 0.23804539894697463,
"learning_rate": 1.042873700093061e-07,
"loss": 0.1034,
"step": 1194
},
{
"epoch": 4.846582278481012,
"grad_norm": 0.2365995323730331,
"learning_rate": 9.857879404556291e-08,
"loss": 0.0976,
"step": 1195
},
{
"epoch": 4.850632911392405,
"grad_norm": 0.23642202812552812,
"learning_rate": 9.303050139225722e-08,
"loss": 0.0765,
"step": 1196
},
{
"epoch": 4.854683544303797,
"grad_norm": 0.2411165251866619,
"learning_rate": 8.76425367345779e-08,
"loss": 0.0958,
"step": 1197
},
{
"epoch": 4.85873417721519,
"grad_norm": 0.2410019851203311,
"learning_rate": 8.241494346644897e-08,
"loss": 0.0866,
"step": 1198
},
{
"epoch": 4.862784810126582,
"grad_norm": 0.24131494190819616,
"learning_rate": 7.734776369019204e-08,
"loss": 0.0893,
"step": 1199
},
{
"epoch": 4.866835443037974,
"grad_norm": 0.24815461206418943,
"learning_rate": 7.244103821617332e-08,
"loss": 0.0989,
"step": 1200
},
{
"epoch": 4.870886075949367,
"grad_norm": 0.24150984132692127,
"learning_rate": 6.769480656248606e-08,
"loss": 0.0822,
"step": 1201
},
{
"epoch": 4.874936708860759,
"grad_norm": 0.25719306807324976,
"learning_rate": 6.310910695462635e-08,
"loss": 0.107,
"step": 1202
},
{
"epoch": 4.878987341772152,
"grad_norm": 0.2317457850974086,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.0866,
"step": 1203
},
{
"epoch": 4.883037974683544,
"grad_norm": 0.2571145810155104,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.0942,
"step": 1204
},
{
"epoch": 4.8870886075949365,
"grad_norm": 0.2254803964874467,
"learning_rate": 5.031556326567488e-08,
"loss": 0.1028,
"step": 1205
},
{
"epoch": 4.891139240506329,
"grad_norm": 0.23570863432112923,
"learning_rate": 4.637234823364312e-08,
"loss": 0.098,
"step": 1206
},
{
"epoch": 4.8951898734177215,
"grad_norm": 0.23830344222510758,
"learning_rate": 4.258983697558838e-08,
"loss": 0.0936,
"step": 1207
},
{
"epoch": 4.899240506329114,
"grad_norm": 0.23112610298137415,
"learning_rate": 3.896805995533548e-08,
"loss": 0.0803,
"step": 1208
},
{
"epoch": 4.9032911392405065,
"grad_norm": 0.2334027490200919,
"learning_rate": 3.550704634218028e-08,
"loss": 0.088,
"step": 1209
},
{
"epoch": 4.907341772151899,
"grad_norm": 0.2402425812713402,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.0859,
"step": 1210
},
{
"epoch": 4.911392405063291,
"grad_norm": 0.25871776069249913,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.1015,
"step": 1211
},
{
"epoch": 4.915443037974684,
"grad_norm": 0.249514520184064,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.0898,
"step": 1212
},
{
"epoch": 4.919493670886076,
"grad_norm": 0.2511029745233125,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.0926,
"step": 1213
},
{
"epoch": 4.923544303797469,
"grad_norm": 0.2287507603237362,
"learning_rate": 2.061435966221881e-08,
"loss": 0.096,
"step": 1214
},
{
"epoch": 4.927594936708861,
"grad_norm": 0.23918941881733063,
"learning_rate": 1.811846652477245e-08,
"loss": 0.0893,
"step": 1215
},
{
"epoch": 4.931645569620253,
"grad_norm": 0.24876239478744344,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.0983,
"step": 1216
},
{
"epoch": 4.935696202531646,
"grad_norm": 0.252333115620244,
"learning_rate": 1.360949303591097e-08,
"loss": 0.1073,
"step": 1217
},
{
"epoch": 4.939746835443038,
"grad_norm": 0.23499459548929333,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.1031,
"step": 1218
},
{
"epoch": 4.943797468354431,
"grad_norm": 0.2373320826001164,
"learning_rate": 9.744388816668172e-09,
"loss": 0.1115,
"step": 1219
},
{
"epoch": 4.947848101265823,
"grad_norm": 0.24526182189416965,
"learning_rate": 8.05332740472009e-09,
"loss": 0.1076,
"step": 1220
},
{
"epoch": 4.951898734177215,
"grad_norm": 0.2468980157184546,
"learning_rate": 6.523278382872811e-09,
"loss": 0.0996,
"step": 1221
},
{
"epoch": 4.955949367088608,
"grad_norm": 0.2386286533902945,
"learning_rate": 5.15425407393133e-09,
"loss": 0.0986,
"step": 1222
},
{
"epoch": 4.96,
"grad_norm": 0.2457689427776535,
"learning_rate": 3.94626550383137e-09,
"loss": 0.1111,
"step": 1223
},
{
"epoch": 4.964050632911392,
"grad_norm": 0.23414243464992812,
"learning_rate": 2.899322401546112e-09,
"loss": 0.113,
"step": 1224
},
{
"epoch": 4.968101265822785,
"grad_norm": 0.2524232717601357,
"learning_rate": 2.013433199010706e-09,
"loss": 0.0945,
"step": 1225
},
{
"epoch": 4.972151898734177,
"grad_norm": 0.24706219245078723,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.1177,
"step": 1226
},
{
"epoch": 4.976202531645569,
"grad_norm": 0.24119557097762662,
"learning_rate": 7.248437353468695e-10,
"loss": 0.0853,
"step": 1227
},
{
"epoch": 4.980253164556962,
"grad_norm": 0.22556143577600035,
"learning_rate": 3.221538523412449e-10,
"loss": 0.1064,
"step": 1228
},
{
"epoch": 4.984303797468354,
"grad_norm": 0.23711275845499924,
"learning_rate": 8.053862524670663e-11,
"loss": 0.0864,
"step": 1229
},
{
"epoch": 4.988354430379747,
"grad_norm": 0.23032004014765722,
"learning_rate": 0.0,
"loss": 0.1078,
"step": 1230
},
{
"epoch": 4.988354430379747,
"step": 1230,
"total_flos": 1.4699010678136504e+18,
"train_loss": 0.2606043583372744,
"train_runtime": 56685.2385,
"train_samples_per_second": 2.787,
"train_steps_per_second": 0.022
}
],
"logging_steps": 1,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.4699010678136504e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}