c1_math_0d_1s / trainer_state.json
neginr's picture
End of training
fe2e39f verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.982278481012658,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004050632911392405,
"grad_norm": 6.289286440867451,
"learning_rate": 3.2520325203252037e-07,
"loss": 0.8117,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 6.411352922223598,
"learning_rate": 6.504065040650407e-07,
"loss": 0.8223,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 6.644571145569796,
"learning_rate": 9.75609756097561e-07,
"loss": 0.8423,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 6.00584624655917,
"learning_rate": 1.3008130081300815e-06,
"loss": 0.7969,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 5.644933112731298,
"learning_rate": 1.6260162601626018e-06,
"loss": 0.7933,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 5.689696581638928,
"learning_rate": 1.951219512195122e-06,
"loss": 0.8007,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 4.289026967630556,
"learning_rate": 2.2764227642276426e-06,
"loss": 0.7618,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 3.9862935855124286,
"learning_rate": 2.601626016260163e-06,
"loss": 0.7481,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 2.3023467152215087,
"learning_rate": 2.926829268292683e-06,
"loss": 0.7314,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.054086235079138,
"learning_rate": 3.2520325203252037e-06,
"loss": 0.7299,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 1.9948337131465352,
"learning_rate": 3.577235772357724e-06,
"loss": 0.7284,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 3.786905371050804,
"learning_rate": 3.902439024390244e-06,
"loss": 0.7158,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 4.311401999723257,
"learning_rate": 4.227642276422765e-06,
"loss": 0.7346,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.213448649900101,
"learning_rate": 4.552845528455285e-06,
"loss": 0.7261,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 4.043013627110665,
"learning_rate": 4.8780487804878055e-06,
"loss": 0.7011,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 3.026604295475713,
"learning_rate": 5.203252032520326e-06,
"loss": 0.6568,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.839592336858255,
"learning_rate": 5.528455284552846e-06,
"loss": 0.6707,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.0929690496439153,
"learning_rate": 5.853658536585366e-06,
"loss": 0.627,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.4364071650633696,
"learning_rate": 6.178861788617887e-06,
"loss": 0.6252,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.1952585952649277,
"learning_rate": 6.504065040650407e-06,
"loss": 0.6254,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.2310789101034125,
"learning_rate": 6.829268292682928e-06,
"loss": 0.6158,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.3729283604620675,
"learning_rate": 7.154471544715448e-06,
"loss": 0.5994,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.0657395025675407,
"learning_rate": 7.4796747967479676e-06,
"loss": 0.6112,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 0.847993639540014,
"learning_rate": 7.804878048780489e-06,
"loss": 0.5976,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 0.830866491104629,
"learning_rate": 8.130081300813009e-06,
"loss": 0.5796,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 0.8380282302179068,
"learning_rate": 8.45528455284553e-06,
"loss": 0.5741,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 0.6587415284463399,
"learning_rate": 8.78048780487805e-06,
"loss": 0.5789,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 0.7299245146646142,
"learning_rate": 9.10569105691057e-06,
"loss": 0.5469,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 0.7718127662693516,
"learning_rate": 9.43089430894309e-06,
"loss": 0.5565,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 0.5690570514199657,
"learning_rate": 9.756097560975611e-06,
"loss": 0.5473,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.5633463156881316,
"learning_rate": 1.008130081300813e-05,
"loss": 0.5416,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 0.7879157324734113,
"learning_rate": 1.0406504065040652e-05,
"loss": 0.5474,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.6650492617136702,
"learning_rate": 1.0731707317073172e-05,
"loss": 0.5451,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.4981939622483124,
"learning_rate": 1.1056910569105692e-05,
"loss": 0.5477,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.6563321107758356,
"learning_rate": 1.1382113821138213e-05,
"loss": 0.5514,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.607120964091458,
"learning_rate": 1.1707317073170731e-05,
"loss": 0.5543,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.5428151685860222,
"learning_rate": 1.2032520325203254e-05,
"loss": 0.5292,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.5963012585667554,
"learning_rate": 1.2357723577235774e-05,
"loss": 0.5248,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.42129376730018997,
"learning_rate": 1.2682926829268294e-05,
"loss": 0.5335,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 0.5074282166425899,
"learning_rate": 1.3008130081300815e-05,
"loss": 0.5223,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.5420075624882711,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.517,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.4330982760599787,
"learning_rate": 1.3658536585365855e-05,
"loss": 0.5218,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.5062879399439018,
"learning_rate": 1.3983739837398376e-05,
"loss": 0.5228,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.4792893960150176,
"learning_rate": 1.4308943089430896e-05,
"loss": 0.5261,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 0.472479138035049,
"learning_rate": 1.4634146341463415e-05,
"loss": 0.4972,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 0.490767684497388,
"learning_rate": 1.4959349593495935e-05,
"loss": 0.523,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.44970228479293617,
"learning_rate": 1.528455284552846e-05,
"loss": 0.5233,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.4107099209819031,
"learning_rate": 1.5609756097560978e-05,
"loss": 0.4909,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.3994795637755471,
"learning_rate": 1.5934959349593496e-05,
"loss": 0.5057,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.4195671453492299,
"learning_rate": 1.6260162601626018e-05,
"loss": 0.498,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 0.4617998656669926,
"learning_rate": 1.6585365853658537e-05,
"loss": 0.5079,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.39604560052809173,
"learning_rate": 1.691056910569106e-05,
"loss": 0.4904,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.4608554689869722,
"learning_rate": 1.7235772357723578e-05,
"loss": 0.4996,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.44627902789826684,
"learning_rate": 1.75609756097561e-05,
"loss": 0.507,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.4789017227841,
"learning_rate": 1.788617886178862e-05,
"loss": 0.5104,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.5428298812276173,
"learning_rate": 1.821138211382114e-05,
"loss": 0.4926,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.4425535810215728,
"learning_rate": 1.8536585365853663e-05,
"loss": 0.4972,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.4935473918906659,
"learning_rate": 1.886178861788618e-05,
"loss": 0.5059,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.40606764353211605,
"learning_rate": 1.91869918699187e-05,
"loss": 0.505,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.48257555263525026,
"learning_rate": 1.9512195121951222e-05,
"loss": 0.491,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.4726804098213924,
"learning_rate": 1.983739837398374e-05,
"loss": 0.4754,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.5540234375400078,
"learning_rate": 2.016260162601626e-05,
"loss": 0.4864,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.5686469504438126,
"learning_rate": 2.048780487804878e-05,
"loss": 0.4998,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.6394882057288443,
"learning_rate": 2.0813008130081303e-05,
"loss": 0.4854,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.7124333711470291,
"learning_rate": 2.1138211382113822e-05,
"loss": 0.4999,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.7071054827643117,
"learning_rate": 2.1463414634146344e-05,
"loss": 0.481,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.8727842511352433,
"learning_rate": 2.1788617886178863e-05,
"loss": 0.4936,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 1.0570991299090151,
"learning_rate": 2.2113821138211385e-05,
"loss": 0.4885,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.5127641027514478,
"learning_rate": 2.2439024390243907e-05,
"loss": 0.4741,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.7889586398884805,
"learning_rate": 2.2764227642276426e-05,
"loss": 0.4884,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.9294275944845564,
"learning_rate": 2.3089430894308948e-05,
"loss": 0.4828,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.5836953443566243,
"learning_rate": 2.3414634146341463e-05,
"loss": 0.4697,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.958189958326084,
"learning_rate": 2.3739837398373985e-05,
"loss": 0.4793,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.9071302982806327,
"learning_rate": 2.4065040650406507e-05,
"loss": 0.4752,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.6480761391033341,
"learning_rate": 2.4390243902439026e-05,
"loss": 0.489,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 1.243923957890688,
"learning_rate": 2.4715447154471548e-05,
"loss": 0.4847,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.6979778982676609,
"learning_rate": 2.5040650406504066e-05,
"loss": 0.4845,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.9478902101121978,
"learning_rate": 2.536585365853659e-05,
"loss": 0.4791,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.598992634641955,
"learning_rate": 2.569105691056911e-05,
"loss": 0.467,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 0.9920266117225077,
"learning_rate": 2.601626016260163e-05,
"loss": 0.4798,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.9099389246819358,
"learning_rate": 2.634146341463415e-05,
"loss": 0.485,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 1.0047156774095378,
"learning_rate": 2.6666666666666667e-05,
"loss": 0.4668,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.9567318651482277,
"learning_rate": 2.699186991869919e-05,
"loss": 0.4881,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 0.7609421395872493,
"learning_rate": 2.731707317073171e-05,
"loss": 0.4795,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 0.8100771537625544,
"learning_rate": 2.764227642276423e-05,
"loss": 0.4734,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 1.1478272623261274,
"learning_rate": 2.796747967479675e-05,
"loss": 0.4694,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 1.2152558471391968,
"learning_rate": 2.829268292682927e-05,
"loss": 0.4758,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.7133289256889849,
"learning_rate": 2.8617886178861792e-05,
"loss": 0.4778,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 0.7177980044342595,
"learning_rate": 2.8943089430894314e-05,
"loss": 0.4658,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 0.7225464803728915,
"learning_rate": 2.926829268292683e-05,
"loss": 0.4703,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 0.7363209857924066,
"learning_rate": 2.959349593495935e-05,
"loss": 0.4842,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.884650387700513,
"learning_rate": 2.991869918699187e-05,
"loss": 0.4822,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.5766899821561234,
"learning_rate": 3.0243902439024392e-05,
"loss": 0.4626,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.7409529887529124,
"learning_rate": 3.056910569105692e-05,
"loss": 0.4788,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.726602059663951,
"learning_rate": 3.089430894308943e-05,
"loss": 0.4703,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 0.807595875662809,
"learning_rate": 3.1219512195121955e-05,
"loss": 0.4724,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 0.7597881617330193,
"learning_rate": 3.154471544715447e-05,
"loss": 0.4697,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 0.7861608334463179,
"learning_rate": 3.186991869918699e-05,
"loss": 0.4714,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 0.9324308945122123,
"learning_rate": 3.2195121951219514e-05,
"loss": 0.4808,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 0.8666098622966577,
"learning_rate": 3.2520325203252037e-05,
"loss": 0.4726,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 0.7511963565925498,
"learning_rate": 3.284552845528456e-05,
"loss": 0.4646,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 0.69972216469334,
"learning_rate": 3.3170731707317074e-05,
"loss": 0.4713,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 0.6975462120398396,
"learning_rate": 3.3495934959349596e-05,
"loss": 0.4621,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 0.596348640868867,
"learning_rate": 3.382113821138212e-05,
"loss": 0.465,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 0.803863463179983,
"learning_rate": 3.414634146341463e-05,
"loss": 0.4713,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 0.6306117365192425,
"learning_rate": 3.4471544715447155e-05,
"loss": 0.4804,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 0.9658263906643764,
"learning_rate": 3.479674796747968e-05,
"loss": 0.454,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 0.8655682645682903,
"learning_rate": 3.51219512195122e-05,
"loss": 0.4622,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 0.775275668170993,
"learning_rate": 3.544715447154472e-05,
"loss": 0.4718,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.1182495950955502,
"learning_rate": 3.577235772357724e-05,
"loss": 0.4673,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 0.8806067466575437,
"learning_rate": 3.609756097560976e-05,
"loss": 0.4803,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 1.0982265257858705,
"learning_rate": 3.642276422764228e-05,
"loss": 0.4763,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 0.5351968649632319,
"learning_rate": 3.67479674796748e-05,
"loss": 0.4616,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 0.5998862213661691,
"learning_rate": 3.7073170731707325e-05,
"loss": 0.4532,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 0.6290080471411021,
"learning_rate": 3.739837398373984e-05,
"loss": 0.4394,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 0.5408412079353866,
"learning_rate": 3.772357723577236e-05,
"loss": 0.4557,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 0.8538446844872392,
"learning_rate": 3.804878048780488e-05,
"loss": 0.4725,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 1.157386611094185,
"learning_rate": 3.83739837398374e-05,
"loss": 0.4733,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 0.767270736976893,
"learning_rate": 3.869918699186992e-05,
"loss": 0.4696,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 0.4826603491703123,
"learning_rate": 3.9024390243902444e-05,
"loss": 0.4657,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 0.5270734570057883,
"learning_rate": 3.9349593495934966e-05,
"loss": 0.4627,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 0.5111144588315328,
"learning_rate": 3.967479674796748e-05,
"loss": 0.4686,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 0.5940903317280773,
"learning_rate": 4e-05,
"loss": 0.4474,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 0.6440528872648922,
"learning_rate": 3.999991946137476e-05,
"loss": 0.4697,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 0.7295528155183251,
"learning_rate": 3.999967784614766e-05,
"loss": 0.4742,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.5797937875483542,
"learning_rate": 3.9999275156264656e-05,
"loss": 0.4663,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 0.665678450613836,
"learning_rate": 3.999871139496895e-05,
"loss": 0.4627,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 0.65016711132788,
"learning_rate": 3.9997986566800995e-05,
"loss": 0.453,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 0.6779809075004015,
"learning_rate": 3.999710067759846e-05,
"loss": 0.4663,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 0.6518433737962406,
"learning_rate": 3.999605373449617e-05,
"loss": 0.467,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.5892623597426443,
"learning_rate": 3.9994845745926075e-05,
"loss": 0.4532,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 0.7799208448578804,
"learning_rate": 3.999347672161713e-05,
"loss": 0.4601,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 0.6954783140691935,
"learning_rate": 3.999194667259528e-05,
"loss": 0.4579,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 0.7534666937743358,
"learning_rate": 3.999025561118334e-05,
"loss": 0.469,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 0.9078723076658732,
"learning_rate": 3.998840355100086e-05,
"loss": 0.473,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 0.8369818898401147,
"learning_rate": 3.998639050696409e-05,
"loss": 0.4524,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 0.6743515486835778,
"learning_rate": 3.998421649528582e-05,
"loss": 0.4607,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 0.6999867847877168,
"learning_rate": 3.9981881533475234e-05,
"loss": 0.4616,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 0.836681264713539,
"learning_rate": 3.997938564033779e-05,
"loss": 0.4573,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.49962433155554775,
"learning_rate": 3.9976728835975064e-05,
"loss": 0.4537,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.7700003071987674,
"learning_rate": 3.9973911141784605e-05,
"loss": 0.4667,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 0.8706638665465914,
"learning_rate": 3.997093258045973e-05,
"loss": 0.4686,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.6804575696530022,
"learning_rate": 3.996779317598936e-05,
"loss": 0.4539,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 0.5653994335533975,
"learning_rate": 3.996449295365782e-05,
"loss": 0.4749,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 0.6543970680610559,
"learning_rate": 3.996103194004467e-05,
"loss": 0.4586,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.5647572754450547,
"learning_rate": 3.995741016302441e-05,
"loss": 0.4503,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 0.4658126103025371,
"learning_rate": 3.9953627651766364e-05,
"loss": 0.4629,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 0.5045936091074238,
"learning_rate": 3.9949684436734325e-05,
"loss": 0.4461,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 0.5449744481495109,
"learning_rate": 3.994558054968643e-05,
"loss": 0.4523,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 0.5813976715734154,
"learning_rate": 3.994131602367481e-05,
"loss": 0.4506,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 0.5862428321010292,
"learning_rate": 3.9936890893045376e-05,
"loss": 0.4507,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 0.5794111916069671,
"learning_rate": 3.993230519343752e-05,
"loss": 0.4781,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 0.5047498462017435,
"learning_rate": 3.992755896178383e-05,
"loss": 0.4525,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 0.6241611099712497,
"learning_rate": 3.992265223630981e-05,
"loss": 0.4603,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 0.6930547912848078,
"learning_rate": 3.991758505653355e-05,
"loss": 0.4587,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 0.5848949350304259,
"learning_rate": 3.991235746326543e-05,
"loss": 0.4619,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 0.5744991899106858,
"learning_rate": 3.9906969498607745e-05,
"loss": 0.4481,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 0.5860499006093769,
"learning_rate": 3.990142120595444e-05,
"loss": 0.4543,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 0.6334059691522708,
"learning_rate": 3.98957126299907e-05,
"loss": 0.4608,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 0.8185368765644289,
"learning_rate": 3.9889843816692596e-05,
"loss": 0.463,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 0.8396827873112954,
"learning_rate": 3.9883814813326766e-05,
"loss": 0.4703,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 0.5989637235373214,
"learning_rate": 3.9877625668449956e-05,
"loss": 0.4645,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 0.698731543056972,
"learning_rate": 3.98712764319087e-05,
"loss": 0.4598,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 0.720348170276798,
"learning_rate": 3.9864767154838864e-05,
"loss": 0.4628,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 0.5079192315751321,
"learning_rate": 3.9858097889665277e-05,
"loss": 0.4512,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 0.7031773164786286,
"learning_rate": 3.985126869010129e-05,
"loss": 0.4544,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 0.5855610109838101,
"learning_rate": 3.984427961114833e-05,
"loss": 0.4491,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 0.40014115475137835,
"learning_rate": 3.9837130709095475e-05,
"loss": 0.4567,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 0.5311323814763691,
"learning_rate": 3.982982204151901e-05,
"loss": 0.4567,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 0.5325345666210406,
"learning_rate": 3.982235366728193e-05,
"loss": 0.4518,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 0.629058588343926,
"learning_rate": 3.9814725646533505e-05,
"loss": 0.4626,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 0.5502515585661516,
"learning_rate": 3.9806938040708746e-05,
"loss": 0.4546,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 0.38201029063787334,
"learning_rate": 3.9798990912527976e-05,
"loss": 0.4512,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 0.44095572178739817,
"learning_rate": 3.979088432599627e-05,
"loss": 0.4454,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 0.38888683821899017,
"learning_rate": 3.9782618346402964e-05,
"loss": 0.4316,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 0.5013098384530591,
"learning_rate": 3.977419304032111e-05,
"loss": 0.4515,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 0.4459392310555741,
"learning_rate": 3.976560847560697e-05,
"loss": 0.4533,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 0.5714976794735979,
"learning_rate": 3.9756864721399456e-05,
"loss": 0.4455,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 0.5951831397819918,
"learning_rate": 3.974796184811956e-05,
"loss": 0.4542,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 0.43105179678098815,
"learning_rate": 3.973889992746979e-05,
"loss": 0.4384,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 0.46431044033135804,
"learning_rate": 3.972967903243361e-05,
"loss": 0.441,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 0.5371062575375862,
"learning_rate": 3.972029923727486e-05,
"loss": 0.4578,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.44863672621201073,
"learning_rate": 3.971076061753709e-05,
"loss": 0.4364,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 0.42336173158989476,
"learning_rate": 3.9701063250043066e-05,
"loss": 0.4506,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.48882736772803514,
"learning_rate": 3.969120721289402e-05,
"loss": 0.4653,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 0.441248904779168,
"learning_rate": 3.9681192585469146e-05,
"loss": 0.4503,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 0.5424187023168797,
"learning_rate": 3.9671019448424865e-05,
"loss": 0.4352,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 0.6389688824901799,
"learning_rate": 3.966068788369422e-05,
"loss": 0.4436,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 0.5547140656204106,
"learning_rate": 3.965019797448622e-05,
"loss": 0.4641,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 0.5435867557098174,
"learning_rate": 3.963954980528515e-05,
"loss": 0.444,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 0.5570416582237089,
"learning_rate": 3.9628743461849905e-05,
"loss": 0.4487,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 0.5448033334291909,
"learning_rate": 3.961777903121329e-05,
"loss": 0.4453,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 0.6599635892949658,
"learning_rate": 3.960665660168131e-05,
"loss": 0.4422,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 0.44333504691274517,
"learning_rate": 3.9595376262832485e-05,
"loss": 0.4437,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 0.5559964773452943,
"learning_rate": 3.9583938105517127e-05,
"loss": 0.4483,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 0.5429531628144375,
"learning_rate": 3.957234222185657e-05,
"loss": 0.4378,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 0.43813959518527673,
"learning_rate": 3.9560588705242474e-05,
"loss": 0.4472,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 0.4569547024115376,
"learning_rate": 3.954867765033605e-05,
"loss": 0.4492,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 0.4975126891635034,
"learning_rate": 3.953660915306728e-05,
"loss": 0.4404,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 0.46596598935405503,
"learning_rate": 3.952438331063419e-05,
"loss": 0.4287,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 0.6123734472023292,
"learning_rate": 3.951200022150205e-05,
"loss": 0.44,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 0.48127551594901163,
"learning_rate": 3.949945998540253e-05,
"loss": 0.4267,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 0.5284122905875202,
"learning_rate": 3.9486762703332993e-05,
"loss": 0.441,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 0.5220414712193395,
"learning_rate": 3.947390847755559e-05,
"loss": 0.4477,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 0.6538724898783044,
"learning_rate": 3.946089741159648e-05,
"loss": 0.4396,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 0.5129162038564671,
"learning_rate": 3.944772961024501e-05,
"loss": 0.4419,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 0.6015110917476078,
"learning_rate": 3.943440517955285e-05,
"loss": 0.4426,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.6910979459156473,
"learning_rate": 3.9420924226833126e-05,
"loss": 0.4429,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.5754181211878954,
"learning_rate": 3.9407286860659566e-05,
"loss": 0.4476,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.5227278105477676,
"learning_rate": 3.9393493190865657e-05,
"loss": 0.4511,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.46117622212961085,
"learning_rate": 3.937954332854371e-05,
"loss": 0.4415,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.4995377517973534,
"learning_rate": 3.9365437386044016e-05,
"loss": 0.4491,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 0.556263094546522,
"learning_rate": 3.935117547697387e-05,
"loss": 0.4447,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.5682978972159433,
"learning_rate": 3.933675771619675e-05,
"loss": 0.4253,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 0.6144422068014421,
"learning_rate": 3.932218421983131e-05,
"loss": 0.4515,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 0.46132294618176495,
"learning_rate": 3.9307455105250484e-05,
"loss": 0.4458,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 0.40869572471716076,
"learning_rate": 3.929257049108054e-05,
"loss": 0.4289,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 0.5416941280431147,
"learning_rate": 3.927753049720011e-05,
"loss": 0.4452,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 0.4233327054044143,
"learning_rate": 3.9262335244739234e-05,
"loss": 0.4452,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 0.4284817764691901,
"learning_rate": 3.92469848560784e-05,
"loss": 0.4331,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 0.3361811505692687,
"learning_rate": 3.923147945484751e-05,
"loss": 0.4392,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.43385613624492697,
"learning_rate": 3.9215819165924956e-05,
"loss": 0.4266,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 0.43851188186727963,
"learning_rate": 3.920000411543654e-05,
"loss": 0.4441,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.43447910773779613,
"learning_rate": 3.9184034430754495e-05,
"loss": 0.4516,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 0.599617878612872,
"learning_rate": 3.916791024049648e-05,
"loss": 0.446,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.4546007781119821,
"learning_rate": 3.91516316745245e-05,
"loss": 0.4484,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 0.5123426024916937,
"learning_rate": 3.913519886394389e-05,
"loss": 0.4353,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.5904221761830782,
"learning_rate": 3.911861194110225e-05,
"loss": 0.4439,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.5618201070858226,
"learning_rate": 3.910187103958837e-05,
"loss": 0.4398,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.44893349136287336,
"learning_rate": 3.908497629423117e-05,
"loss": 0.4497,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.5048509695105279,
"learning_rate": 3.9067927841098614e-05,
"loss": 0.4515,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.5953427075906875,
"learning_rate": 3.9050725817496594e-05,
"loss": 0.4329,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.541980126215068,
"learning_rate": 3.9033370361967844e-05,
"loss": 0.4396,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.7363116422843797,
"learning_rate": 3.901586161429081e-05,
"loss": 0.4418,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.648192752747525,
"learning_rate": 3.8998199715478545e-05,
"loss": 0.4393,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.5280181692955821,
"learning_rate": 3.8980384807777564e-05,
"loss": 0.4379,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.4723170963514602,
"learning_rate": 3.896241703466667e-05,
"loss": 0.4345,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.49049354509477183,
"learning_rate": 3.894429654085585e-05,
"loss": 0.4436,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.6466886948107943,
"learning_rate": 3.892602347228505e-05,
"loss": 0.4565,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.6299267769694589,
"learning_rate": 3.890759797612307e-05,
"loss": 0.4354,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.45009409730161787,
"learning_rate": 3.888902020076632e-05,
"loss": 0.4414,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.40939123822934986,
"learning_rate": 3.887029029583764e-05,
"loss": 0.4358,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 0.4612786172532384,
"learning_rate": 3.8851408412185125e-05,
"loss": 0.4454,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.46218774434298215,
"learning_rate": 3.8832374701880855e-05,
"loss": 0.4366,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.47567161588704576,
"learning_rate": 3.881318931821972e-05,
"loss": 0.4444,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.43180696224741477,
"learning_rate": 3.879385241571817e-05,
"loss": 0.4389,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 0.5630603369599148,
"learning_rate": 3.8774364150112955e-05,
"loss": 0.4248,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.542597266274369,
"learning_rate": 3.8754724678359884e-05,
"loss": 0.3969,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.453386828945273,
"learning_rate": 3.873493415863256e-05,
"loss": 0.3991,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.504842582038129,
"learning_rate": 3.871499275032111e-05,
"loss": 0.4077,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.3770858755088825,
"learning_rate": 3.869490061403091e-05,
"loss": 0.4067,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 0.46049701077815336,
"learning_rate": 3.867465791158124e-05,
"loss": 0.4074,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.4969346197219919,
"learning_rate": 3.865426480600407e-05,
"loss": 0.3985,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 0.44273528750009683,
"learning_rate": 3.863372146154264e-05,
"loss": 0.4003,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.49043245312013806,
"learning_rate": 3.861302804365024e-05,
"loss": 0.3953,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.48707961631597874,
"learning_rate": 3.85921847189888e-05,
"loss": 0.3984,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.4669555432446554,
"learning_rate": 3.85711916554276e-05,
"loss": 0.4004,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.5084030266444322,
"learning_rate": 3.85500490220419e-05,
"loss": 0.3982,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.4863962732196203,
"learning_rate": 3.852875698911154e-05,
"loss": 0.4009,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.40530993702963364,
"learning_rate": 3.850731572811963e-05,
"loss": 0.3944,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.45039473492403514,
"learning_rate": 3.848572541175116e-05,
"loss": 0.4049,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.4635221903321783,
"learning_rate": 3.846398621389154e-05,
"loss": 0.3973,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.6612205279783843,
"learning_rate": 3.84420983096253e-05,
"loss": 0.3997,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.7269274754547639,
"learning_rate": 3.8420061875234606e-05,
"loss": 0.4039,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.5919997009881557,
"learning_rate": 3.839787708819787e-05,
"loss": 0.4086,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.4044837302914207,
"learning_rate": 3.8375544127188325e-05,
"loss": 0.3957,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.6162228067154731,
"learning_rate": 3.8353063172072564e-05,
"loss": 0.395,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.5946707632340674,
"learning_rate": 3.8330434403909105e-05,
"loss": 0.4003,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.5269924347281333,
"learning_rate": 3.8307658004946934e-05,
"loss": 0.4015,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.44652286402443364,
"learning_rate": 3.8284734158624046e-05,
"loss": 0.3944,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.4410235647471469,
"learning_rate": 3.826166304956594e-05,
"loss": 0.3949,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.5322676256452101,
"learning_rate": 3.8238444863584164e-05,
"loss": 0.4135,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.4042624234263607,
"learning_rate": 3.821507978767479e-05,
"loss": 0.3904,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.38186342689129166,
"learning_rate": 3.819156801001693e-05,
"loss": 0.3991,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.40873038052789123,
"learning_rate": 3.816790971997121e-05,
"loss": 0.3826,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.5249982118352632,
"learning_rate": 3.8144105108078246e-05,
"loss": 0.3993,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.47042488389773346,
"learning_rate": 3.81201543660571e-05,
"loss": 0.3966,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.3993243236445046,
"learning_rate": 3.809605768680377e-05,
"loss": 0.4133,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.513144753930664,
"learning_rate": 3.807181526438958e-05,
"loss": 0.4081,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.49329547014684716,
"learning_rate": 3.8047427294059697e-05,
"loss": 0.4044,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.39861655771255705,
"learning_rate": 3.802289397223145e-05,
"loss": 0.404,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.36804544022379176,
"learning_rate": 3.7998215496492854e-05,
"loss": 0.4067,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.36601643128827627,
"learning_rate": 3.797339206560096e-05,
"loss": 0.4152,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.5895362890875693,
"learning_rate": 3.794842387948027e-05,
"loss": 0.3882,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.5606492429352079,
"learning_rate": 3.7923311139221114e-05,
"loss": 0.3946,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.38521094388495397,
"learning_rate": 3.7898054047078054e-05,
"loss": 0.3975,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.5164984759881424,
"learning_rate": 3.787265280646825e-05,
"loss": 0.406,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.48293319556082587,
"learning_rate": 3.7847107621969786e-05,
"loss": 0.4143,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.5291608691303267,
"learning_rate": 3.7821418699320064e-05,
"loss": 0.3939,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.34123008839052865,
"learning_rate": 3.7795586245414145e-05,
"loss": 0.406,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.4774517201883595,
"learning_rate": 3.776961046830306e-05,
"loss": 0.3953,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.4902061645285676,
"learning_rate": 3.774349157719215e-05,
"loss": 0.4014,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.6236665287459114,
"learning_rate": 3.7717229782439365e-05,
"loss": 0.4045,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.5747353385793417,
"learning_rate": 3.769082529555359e-05,
"loss": 0.3932,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.3480386299543018,
"learning_rate": 3.766427832919294e-05,
"loss": 0.4013,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.5486841406903501,
"learning_rate": 3.7637589097163024e-05,
"loss": 0.4017,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.5895768942496205,
"learning_rate": 3.761075781441526e-05,
"loss": 0.4144,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 0.4204118912255965,
"learning_rate": 3.75837846970451e-05,
"loss": 0.4035,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.44030646505638565,
"learning_rate": 3.755666996229032e-05,
"loss": 0.3925,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.45504428200482067,
"learning_rate": 3.752941382852927e-05,
"loss": 0.3954,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.4944398611273607,
"learning_rate": 3.7502016515279115e-05,
"loss": 0.3992,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.4164148764281623,
"learning_rate": 3.7474478243194043e-05,
"loss": 0.3904,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.35864593141396667,
"learning_rate": 3.744679923406351e-05,
"loss": 0.3906,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.4925704070854084,
"learning_rate": 3.741897971081043e-05,
"loss": 0.3959,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.40997505630222425,
"learning_rate": 3.739101989748946e-05,
"loss": 0.4117,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.3524366453164234,
"learning_rate": 3.7362920019285066e-05,
"loss": 0.408,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.42610922299428183,
"learning_rate": 3.73346803025098e-05,
"loss": 0.3964,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 0.5555272513034818,
"learning_rate": 3.730630097460247e-05,
"loss": 0.3963,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.5353738627967675,
"learning_rate": 3.727778226412628e-05,
"loss": 0.4108,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 0.3949024751750218,
"learning_rate": 3.7249124400767006e-05,
"loss": 0.3891,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.41464396218368893,
"learning_rate": 3.722032761533114e-05,
"loss": 0.3861,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.4681022898592037,
"learning_rate": 3.719139213974403e-05,
"loss": 0.4119,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 0.4634337378725406,
"learning_rate": 3.7162318207048006e-05,
"loss": 0.3945,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.3720937694569255,
"learning_rate": 3.713310605140055e-05,
"loss": 0.4136,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.5015614926610431,
"learning_rate": 3.710375590807233e-05,
"loss": 0.403,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.48784671979821564,
"learning_rate": 3.7074268013445365e-05,
"loss": 0.402,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.3612502147159279,
"learning_rate": 3.7044642605011114e-05,
"loss": 0.4048,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.4074735658627379,
"learning_rate": 3.701487992136854e-05,
"loss": 0.3911,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 0.46401796819077695,
"learning_rate": 3.69849802022222e-05,
"loss": 0.3918,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.46339161464643086,
"learning_rate": 3.6954943688380334e-05,
"loss": 0.4083,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.35733666099195366,
"learning_rate": 3.692477062175289e-05,
"loss": 0.3995,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.6266903389890947,
"learning_rate": 3.689446124534958e-05,
"loss": 0.4123,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.6324641517461221,
"learning_rate": 3.686401580327799e-05,
"loss": 0.3847,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.4244595544772277,
"learning_rate": 3.683343454074149e-05,
"loss": 0.4091,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.4289455737910442,
"learning_rate": 3.6802717704037386e-05,
"loss": 0.4065,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.48570192066213214,
"learning_rate": 3.6771865540554855e-05,
"loss": 0.3978,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.4467857331865248,
"learning_rate": 3.674087829877297e-05,
"loss": 0.4086,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.3477475897434,
"learning_rate": 3.6709756228258735e-05,
"loss": 0.3994,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.3676296230998324,
"learning_rate": 3.667849957966501e-05,
"loss": 0.3887,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.4199310439513471,
"learning_rate": 3.6647108604728546e-05,
"loss": 0.4025,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.3202410450614145,
"learning_rate": 3.661558355626795e-05,
"loss": 0.3966,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.38468234564036785,
"learning_rate": 3.658392468818163e-05,
"loss": 0.4111,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.40202127791555364,
"learning_rate": 3.655213225544574e-05,
"loss": 0.3957,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.3296995214744877,
"learning_rate": 3.652020651411218e-05,
"loss": 0.3948,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.35892742411809325,
"learning_rate": 3.6488147721306474e-05,
"loss": 0.4187,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.32760060247923606,
"learning_rate": 3.645595613522574e-05,
"loss": 0.3953,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.32966841601353614,
"learning_rate": 3.642363201513657e-05,
"loss": 0.4037,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.37644695813019746,
"learning_rate": 3.6391175621373006e-05,
"loss": 0.405,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.35666171907069205,
"learning_rate": 3.6358587215334355e-05,
"loss": 0.392,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.3513255032874715,
"learning_rate": 3.632586705948318e-05,
"loss": 0.3844,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.35121130917648047,
"learning_rate": 3.629301541734311e-05,
"loss": 0.4106,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.41600525563286134,
"learning_rate": 3.626003255349676e-05,
"loss": 0.4006,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.4594848339940611,
"learning_rate": 3.622691873358357e-05,
"loss": 0.3822,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.3984000602429221,
"learning_rate": 3.61936742242977e-05,
"loss": 0.4159,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.4647724131446653,
"learning_rate": 3.6160299293385864e-05,
"loss": 0.3965,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.3387320370162092,
"learning_rate": 3.612679420964516e-05,
"loss": 0.407,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.40518850139691875,
"learning_rate": 3.609315924292092e-05,
"loss": 0.4076,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.2825779173779976,
"learning_rate": 3.6059394664104554e-05,
"loss": 0.4068,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.3251268825393156,
"learning_rate": 3.602550074513133e-05,
"loss": 0.3967,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.30999791647339003,
"learning_rate": 3.599147775897822e-05,
"loss": 0.38,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.33899032125719414,
"learning_rate": 3.595732597966167e-05,
"loss": 0.4052,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.38581386277109225,
"learning_rate": 3.592304568223542e-05,
"loss": 0.3883,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.4862689315556685,
"learning_rate": 3.588863714278826e-05,
"loss": 0.3989,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.4645010391640131,
"learning_rate": 3.585410063844186e-05,
"loss": 0.4021,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.3659558314265265,
"learning_rate": 3.581943644734846e-05,
"loss": 0.3995,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.30510267520443013,
"learning_rate": 3.578464484868869e-05,
"loss": 0.3879,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.37361149207393457,
"learning_rate": 3.5749726122669316e-05,
"loss": 0.3904,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.40731810418857856,
"learning_rate": 3.5714680550520943e-05,
"loss": 0.4028,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.46785736876329653,
"learning_rate": 3.5679508414495794e-05,
"loss": 0.3998,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.48699388593872606,
"learning_rate": 3.564420999786543e-05,
"loss": 0.4119,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.3363009378539564,
"learning_rate": 3.560878558491842e-05,
"loss": 0.3974,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.48045189520328946,
"learning_rate": 3.5573235460958145e-05,
"loss": 0.4005,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.4415747586648743,
"learning_rate": 3.553755991230039e-05,
"loss": 0.3926,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.3172347606553149,
"learning_rate": 3.5501759226271144e-05,
"loss": 0.3826,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.37704312926109246,
"learning_rate": 3.546583369120419e-05,
"loss": 0.3982,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.43502617558701606,
"learning_rate": 3.5429783596438864e-05,
"loss": 0.3967,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.4480215217600431,
"learning_rate": 3.539360923231766e-05,
"loss": 0.3953,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.30911243182397774,
"learning_rate": 3.535731089018394e-05,
"loss": 0.4016,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.4703090961285979,
"learning_rate": 3.532088886237956e-05,
"loss": 0.3834,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.445172378647541,
"learning_rate": 3.528434344224253e-05,
"loss": 0.4018,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.38518582771225723,
"learning_rate": 3.524767492410464e-05,
"loss": 0.4006,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.2918172509851109,
"learning_rate": 3.521088360328908e-05,
"loss": 0.3967,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.3597485166244472,
"learning_rate": 3.517396977610811e-05,
"loss": 0.3976,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.4005310260177406,
"learning_rate": 3.5136933739860595e-05,
"loss": 0.3949,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.38301326815612147,
"learning_rate": 3.509977579282971e-05,
"loss": 0.3947,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.3677176245848521,
"learning_rate": 3.5062496234280424e-05,
"loss": 0.4082,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.4581953490647303,
"learning_rate": 3.502509536445719e-05,
"loss": 0.3935,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.3485156654738773,
"learning_rate": 3.498757348458147e-05,
"loss": 0.4011,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.34358178579807835,
"learning_rate": 3.4949930896849324e-05,
"loss": 0.406,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.4250402195470649,
"learning_rate": 3.491216790442899e-05,
"loss": 0.3974,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.3688011039985426,
"learning_rate": 3.487428481145839e-05,
"loss": 0.3983,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.30623145467211005,
"learning_rate": 3.483628192304278e-05,
"loss": 0.3939,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.35071532897192453,
"learning_rate": 3.479815954525219e-05,
"loss": 0.3954,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.3558677043302736,
"learning_rate": 3.475991798511899e-05,
"loss": 0.3988,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.32743443161129737,
"learning_rate": 3.4721557550635464e-05,
"loss": 0.3933,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.3803713288272411,
"learning_rate": 3.468307855075128e-05,
"loss": 0.3981,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.37903458289371206,
"learning_rate": 3.4644481295371005e-05,
"loss": 0.3959,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.37566162983616397,
"learning_rate": 3.460576609535163e-05,
"loss": 0.4184,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.4322400271024682,
"learning_rate": 3.456693326250006e-05,
"loss": 0.4052,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.5243530237790323,
"learning_rate": 3.452798310957058e-05,
"loss": 0.4061,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.40877643090075516,
"learning_rate": 3.4488915950262386e-05,
"loss": 0.3981,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.42895110162794303,
"learning_rate": 3.4449732099216985e-05,
"loss": 0.4064,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.49086593661199257,
"learning_rate": 3.441043187201574e-05,
"loss": 0.3915,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.385908134905131,
"learning_rate": 3.437101558517728e-05,
"loss": 0.4101,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.46339864541728526,
"learning_rate": 3.433148355615496e-05,
"loss": 0.3855,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.4331539705442504,
"learning_rate": 3.4291836103334294e-05,
"loss": 0.4011,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.36681750800235013,
"learning_rate": 3.425207354603043e-05,
"loss": 0.4054,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.3150816174338742,
"learning_rate": 3.421219620448553e-05,
"loss": 0.3879,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.4441806827141326,
"learning_rate": 3.417220439986623e-05,
"loss": 0.3835,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.37195007640561,
"learning_rate": 3.4132098454261024e-05,
"loss": 0.3895,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.4224165982781685,
"learning_rate": 3.4091878690677676e-05,
"loss": 0.4073,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.40400620290445277,
"learning_rate": 3.405154543304065e-05,
"loss": 0.3892,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.3823164058587053,
"learning_rate": 3.401109900618843e-05,
"loss": 0.4075,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.37479334808778586,
"learning_rate": 3.3970539735870996e-05,
"loss": 0.3837,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.417101598199313,
"learning_rate": 3.392986794874714e-05,
"loss": 0.4092,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.4497583230514636,
"learning_rate": 3.388908397238184e-05,
"loss": 0.4076,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.36912346186165795,
"learning_rate": 3.384818813524362e-05,
"loss": 0.3894,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.36952211661266465,
"learning_rate": 3.380718076670195e-05,
"loss": 0.4005,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.4029010250241343,
"learning_rate": 3.376606219702454e-05,
"loss": 0.4112,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.3767804832862671,
"learning_rate": 3.372483275737468e-05,
"loss": 0.4099,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.3804506364606566,
"learning_rate": 3.368349277980861e-05,
"loss": 0.4039,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.3815191733108533,
"learning_rate": 3.3642042597272844e-05,
"loss": 0.4016,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.553435101295078,
"learning_rate": 3.360048254360144e-05,
"loss": 0.4014,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.3843702434283537,
"learning_rate": 3.355881295351336e-05,
"loss": 0.4032,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.4384223903351665,
"learning_rate": 3.351703416260975e-05,
"loss": 0.4025,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.4412023925436807,
"learning_rate": 3.347514650737126e-05,
"loss": 0.3914,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.40555156097801,
"learning_rate": 3.3433150325155295e-05,
"loss": 0.3852,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.3968528151047142,
"learning_rate": 3.339104595419334e-05,
"loss": 0.3919,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.41714277121143756,
"learning_rate": 3.3348833733588204e-05,
"loss": 0.4023,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.4077257128917511,
"learning_rate": 3.3306514003311305e-05,
"loss": 0.4041,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.38943547236706616,
"learning_rate": 3.326408710419996e-05,
"loss": 0.3995,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.35313543339869247,
"learning_rate": 3.322155337795454e-05,
"loss": 0.3999,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.3903823489046648,
"learning_rate": 3.317891316713587e-05,
"loss": 0.4039,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.39249421468008494,
"learning_rate": 3.313616681516231e-05,
"loss": 0.4099,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.32769643574881757,
"learning_rate": 3.309331466630713e-05,
"loss": 0.3869,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.3727948772088613,
"learning_rate": 3.305035706569563e-05,
"loss": 0.4004,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.3184699156018918,
"learning_rate": 3.3007294359302433e-05,
"loss": 0.3947,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.3456683722712126,
"learning_rate": 3.296412689394864e-05,
"loss": 0.4027,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.3572134584709407,
"learning_rate": 3.292085501729909e-05,
"loss": 0.3948,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.3363363438308483,
"learning_rate": 3.2877479077859534e-05,
"loss": 0.4,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.4348808192287335,
"learning_rate": 3.283399942497381e-05,
"loss": 0.4016,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.34203282471274027,
"learning_rate": 3.279041640882108e-05,
"loss": 0.3834,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.39814461503610665,
"learning_rate": 3.2746730380412964e-05,
"loss": 0.3858,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.3108541890067744,
"learning_rate": 3.2702941691590726e-05,
"loss": 0.4042,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.4051565773444433,
"learning_rate": 3.265905069502244e-05,
"loss": 0.394,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.3520993300770663,
"learning_rate": 3.261505774420016e-05,
"loss": 0.3986,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.2875105063202822,
"learning_rate": 3.257096319343707e-05,
"loss": 0.3818,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.33516882407530546,
"learning_rate": 3.2526767397864614e-05,
"loss": 0.4011,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.3396041076298476,
"learning_rate": 3.248247071342966e-05,
"loss": 0.3916,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.3304240763571902,
"learning_rate": 3.243807349689161e-05,
"loss": 0.3937,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.3074003923985243,
"learning_rate": 3.2393576105819544e-05,
"loss": 0.3838,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.33804453681119867,
"learning_rate": 3.2348978898589333e-05,
"loss": 0.4025,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.30604802839373896,
"learning_rate": 3.230428223438075e-05,
"loss": 0.3942,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.3100391529473118,
"learning_rate": 3.225948647317459e-05,
"loss": 0.3956,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.31368798473450227,
"learning_rate": 3.2214591975749745e-05,
"loss": 0.3997,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.33758671902544274,
"learning_rate": 3.216959910368034e-05,
"loss": 0.3973,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.31283602187439497,
"learning_rate": 3.212450821933277e-05,
"loss": 0.3995,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.30817764811412074,
"learning_rate": 3.207931968586281e-05,
"loss": 0.395,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.28480311829205945,
"learning_rate": 3.203403386721272e-05,
"loss": 0.3911,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.29733846031960015,
"learning_rate": 3.1988651128108245e-05,
"loss": 0.3915,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.34262054838404155,
"learning_rate": 3.194317183405573e-05,
"loss": 0.3966,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.3137356864508664,
"learning_rate": 3.189759635133914e-05,
"loss": 0.3966,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.3089394510651776,
"learning_rate": 3.185192504701718e-05,
"loss": 0.3882,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.30008334910102913,
"learning_rate": 3.1806158288920234e-05,
"loss": 0.4028,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.2963297990143214,
"learning_rate": 3.1760296445647477e-05,
"loss": 0.4014,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.3222360448259214,
"learning_rate": 3.1714339886563896e-05,
"loss": 0.3928,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.34552892757784825,
"learning_rate": 3.166828898179731e-05,
"loss": 0.394,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.2929734867536495,
"learning_rate": 3.162214410223536e-05,
"loss": 0.4008,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.3609568833376742,
"learning_rate": 3.157590561952257e-05,
"loss": 0.3938,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.30909312160634284,
"learning_rate": 3.152957390605732e-05,
"loss": 0.3969,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.3641003114460583,
"learning_rate": 3.148314933498886e-05,
"loss": 0.3875,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.29341386396636554,
"learning_rate": 3.143663228021431e-05,
"loss": 0.39,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.3251473549704783,
"learning_rate": 3.1390023116375624e-05,
"loss": 0.3782,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.36907917693310177,
"learning_rate": 3.134332221885661e-05,
"loss": 0.3961,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.35970816787975196,
"learning_rate": 3.129652996377987e-05,
"loss": 0.393,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.3751996352039081,
"learning_rate": 3.12496467280038e-05,
"loss": 0.4065,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.3860819959033839,
"learning_rate": 3.120267288911952e-05,
"loss": 0.389,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.33543625307909375,
"learning_rate": 3.11556088254479e-05,
"loss": 0.3754,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.4107455471734404,
"learning_rate": 3.11084549160364e-05,
"loss": 0.3959,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.3211466321475857,
"learning_rate": 3.106121154065615e-05,
"loss": 0.3991,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.4411307999971255,
"learning_rate": 3.1013879079798805e-05,
"loss": 0.4036,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.36478962071366283,
"learning_rate": 3.096645791467348e-05,
"loss": 0.3924,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.5011110307969944,
"learning_rate": 3.091894842720373e-05,
"loss": 0.399,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.4688382262194142,
"learning_rate": 3.0871351000024425e-05,
"loss": 0.3969,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.3886038997726484,
"learning_rate": 3.0823666016478716e-05,
"loss": 0.3893,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.49765534103108916,
"learning_rate": 3.0775893860614896e-05,
"loss": 0.3977,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.3476475176529487,
"learning_rate": 3.0728034917183336e-05,
"loss": 0.4042,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.4634997096530916,
"learning_rate": 3.06800895716334e-05,
"loss": 0.3955,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.319476062834144,
"learning_rate": 3.063205821011029e-05,
"loss": 0.4084,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.4795488219712325,
"learning_rate": 3.0583941219452016e-05,
"loss": 0.4076,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.300558885406406,
"learning_rate": 3.053573898718618e-05,
"loss": 0.3902,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.40658565160517957,
"learning_rate": 3.0487451901526956e-05,
"loss": 0.3932,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.4455393957063756,
"learning_rate": 3.0439080351371875e-05,
"loss": 0.395,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.4721914364913131,
"learning_rate": 3.0390624726298764e-05,
"loss": 0.3992,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.39895621985425217,
"learning_rate": 3.034208541656255e-05,
"loss": 0.3912,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.5272326901250234,
"learning_rate": 3.029346281309218e-05,
"loss": 0.3995,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.32916196690489496,
"learning_rate": 3.0244757307487415e-05,
"loss": 0.4,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.46433777724675757,
"learning_rate": 3.019596929201569e-05,
"loss": 0.3903,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.35763271922320455,
"learning_rate": 3.0147099159608985e-05,
"loss": 0.3873,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.3636612211414771,
"learning_rate": 3.0098147303860616e-05,
"loss": 0.3967,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.40961520771872423,
"learning_rate": 3.0049114119022117e-05,
"loss": 0.3903,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.31559692503340986,
"learning_rate": 3.0000000000000004e-05,
"loss": 0.3882,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.37739648644562823,
"learning_rate": 2.995080534235264e-05,
"loss": 0.3861,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.3263327039266923,
"learning_rate": 2.9901530542287044e-05,
"loss": 0.3834,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.4452796619879132,
"learning_rate": 2.9852175996655676e-05,
"loss": 0.3469,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 0.8028971843518358,
"learning_rate": 2.980274210295326e-05,
"loss": 0.3483,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.6264175484520534,
"learning_rate": 2.9753229259313578e-05,
"loss": 0.3435,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.5643944824598596,
"learning_rate": 2.9703637864506274e-05,
"loss": 0.3406,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.41640396335712615,
"learning_rate": 2.965396831793362e-05,
"loss": 0.3424,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.6347002427567898,
"learning_rate": 2.9604221019627316e-05,
"loss": 0.337,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.6677114391319677,
"learning_rate": 2.955439637024526e-05,
"loss": 0.3468,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.4073561309865602,
"learning_rate": 2.9504494771068334e-05,
"loss": 0.3472,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.5972166131206924,
"learning_rate": 2.9454516623997156e-05,
"loss": 0.354,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.5727006341610796,
"learning_rate": 2.9404462331548847e-05,
"loss": 0.3548,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.4067560386772968,
"learning_rate": 2.93543322968538e-05,
"loss": 0.3452,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.5616013544997751,
"learning_rate": 2.9304126923652428e-05,
"loss": 0.337,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.44100074212037643,
"learning_rate": 2.9253846616291896e-05,
"loss": 0.3467,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.4742378283420198,
"learning_rate": 2.9203491779722896e-05,
"loss": 0.3291,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.49561023315964275,
"learning_rate": 2.9153062819496357e-05,
"loss": 0.3464,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.3938745663195535,
"learning_rate": 2.9102560141760178e-05,
"loss": 0.3322,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.43775301322833815,
"learning_rate": 2.9051984153256004e-05,
"loss": 0.3376,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.4455229449290395,
"learning_rate": 2.900133526131588e-05,
"loss": 0.3269,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.41138958995685565,
"learning_rate": 2.8950613873859025e-05,
"loss": 0.3544,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.3689178366321214,
"learning_rate": 2.8899820399388515e-05,
"loss": 0.3482,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.46180476855294955,
"learning_rate": 2.8848955246988012e-05,
"loss": 0.3489,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.3300636258967015,
"learning_rate": 2.879801882631847e-05,
"loss": 0.3389,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.38553278955846865,
"learning_rate": 2.8747011547614808e-05,
"loss": 0.3389,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.37854369263953663,
"learning_rate": 2.8695933821682635e-05,
"loss": 0.3502,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.3440848273499826,
"learning_rate": 2.864478605989494e-05,
"loss": 0.331,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.3161683110987583,
"learning_rate": 2.8593568674188765e-05,
"loss": 0.3399,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.31845568366465205,
"learning_rate": 2.8542282077061892e-05,
"loss": 0.3421,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.3231083549305791,
"learning_rate": 2.8490926681569523e-05,
"loss": 0.3533,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.34051933690548764,
"learning_rate": 2.8439502901320956e-05,
"loss": 0.3491,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.30406153829802873,
"learning_rate": 2.8388011150476237e-05,
"loss": 0.3389,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.348462744561755,
"learning_rate": 2.8336451843742866e-05,
"loss": 0.3337,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.3206691304592402,
"learning_rate": 2.8284825396372387e-05,
"loss": 0.344,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.30444214655942686,
"learning_rate": 2.8233132224157132e-05,
"loss": 0.3406,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.35866785575979204,
"learning_rate": 2.8181372743426805e-05,
"loss": 0.3379,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.29182072097802647,
"learning_rate": 2.8129547371045128e-05,
"loss": 0.3404,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.31619672606172383,
"learning_rate": 2.8077656524406534e-05,
"loss": 0.3541,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.3163527673176312,
"learning_rate": 2.802570062143278e-05,
"loss": 0.3396,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.31067591956210633,
"learning_rate": 2.7973680080569555e-05,
"loss": 0.3322,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.3098737386924152,
"learning_rate": 2.792159532078314e-05,
"loss": 0.3455,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.38643670920609946,
"learning_rate": 2.7869446761557033e-05,
"loss": 0.341,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.3323781770929014,
"learning_rate": 2.781723482288857e-05,
"loss": 0.3386,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.30363928029340137,
"learning_rate": 2.7764959925285517e-05,
"loss": 0.3369,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.4633818290521634,
"learning_rate": 2.771262248976272e-05,
"loss": 0.3465,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.3343997025155383,
"learning_rate": 2.7660222937838677e-05,
"loss": 0.3355,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.4094018710407227,
"learning_rate": 2.7607761691532186e-05,
"loss": 0.3404,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.37220527554096383,
"learning_rate": 2.7555239173358916e-05,
"loss": 0.3323,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.3528127658581048,
"learning_rate": 2.7502655806328e-05,
"loss": 0.3399,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.4153830115002887,
"learning_rate": 2.7450012013938648e-05,
"loss": 0.3389,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.33314476316302893,
"learning_rate": 2.739730822017673e-05,
"loss": 0.3388,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.4328018718748249,
"learning_rate": 2.7344544849511355e-05,
"loss": 0.348,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.31571291157595194,
"learning_rate": 2.7291722326891456e-05,
"loss": 0.357,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.472995985365362,
"learning_rate": 2.723884107774236e-05,
"loss": 0.3555,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.3125899511385642,
"learning_rate": 2.718590152796239e-05,
"loss": 0.3516,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.4384550736830633,
"learning_rate": 2.71329041039194e-05,
"loss": 0.335,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.3246502016614766,
"learning_rate": 2.7079849232447357e-05,
"loss": 0.3396,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.33847751777373575,
"learning_rate": 2.7026737340842895e-05,
"loss": 0.3363,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.35425912624820227,
"learning_rate": 2.697356885686189e-05,
"loss": 0.3508,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.3005216006556085,
"learning_rate": 2.6920344208716014e-05,
"loss": 0.3493,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.3453118659517926,
"learning_rate": 2.6867063825069252e-05,
"loss": 0.3337,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.29328368710669106,
"learning_rate": 2.6813728135034494e-05,
"loss": 0.343,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.3149124938390566,
"learning_rate": 2.6760337568170056e-05,
"loss": 0.3441,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.32883069805920623,
"learning_rate": 2.6706892554476226e-05,
"loss": 0.345,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.33477156327514984,
"learning_rate": 2.6653393524391795e-05,
"loss": 0.3344,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.2732994841130123,
"learning_rate": 2.6599840908790592e-05,
"loss": 0.34,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.30945778363167287,
"learning_rate": 2.6546235138978028e-05,
"loss": 0.3446,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.29094922986733196,
"learning_rate": 2.6492576646687597e-05,
"loss": 0.3431,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.33125233210606964,
"learning_rate": 2.6438865864077425e-05,
"loss": 0.3545,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.3097091959558416,
"learning_rate": 2.6385103223726766e-05,
"loss": 0.3478,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.31067285657329574,
"learning_rate": 2.6331289158632537e-05,
"loss": 0.3482,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.2713293165293498,
"learning_rate": 2.6277424102205817e-05,
"loss": 0.3319,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.24913151266731806,
"learning_rate": 2.6223508488268374e-05,
"loss": 0.3531,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.29079158897256224,
"learning_rate": 2.6169542751049148e-05,
"loss": 0.3357,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.25757504558340244,
"learning_rate": 2.6115527325180754e-05,
"loss": 0.3403,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.30543713505921266,
"learning_rate": 2.606146264569603e-05,
"loss": 0.3597,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.27705347229983185,
"learning_rate": 2.6007349148024447e-05,
"loss": 0.3393,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.2788597198788762,
"learning_rate": 2.5953187267988694e-05,
"loss": 0.3433,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.2413969348890641,
"learning_rate": 2.5898977441801097e-05,
"loss": 0.3319,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.2658279065400016,
"learning_rate": 2.584472010606015e-05,
"loss": 0.3483,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.23575836232122874,
"learning_rate": 2.5790415697746976e-05,
"loss": 0.3426,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.23665844553863064,
"learning_rate": 2.5736064654221808e-05,
"loss": 0.3333,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.24009244825176532,
"learning_rate": 2.568166741322048e-05,
"loss": 0.3374,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.24657580563150652,
"learning_rate": 2.56272244128509e-05,
"loss": 0.3364,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.25471854470983096,
"learning_rate": 2.55727360915895e-05,
"loss": 0.3347,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.2699963246188958,
"learning_rate": 2.5518202888277734e-05,
"loss": 0.3467,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.3152759355348449,
"learning_rate": 2.5463625242118523e-05,
"loss": 0.3494,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.2952276581764004,
"learning_rate": 2.5409003592672723e-05,
"loss": 0.3429,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.27733721490915486,
"learning_rate": 2.535433837985559e-05,
"loss": 0.3294,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.2804560607134888,
"learning_rate": 2.529963004393324e-05,
"loss": 0.3376,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.37006916741120854,
"learning_rate": 2.524487902551908e-05,
"loss": 0.3543,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.2989070154345703,
"learning_rate": 2.519008576557029e-05,
"loss": 0.3404,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.2982490575616926,
"learning_rate": 2.5135250705384254e-05,
"loss": 0.3364,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.2630842427210206,
"learning_rate": 2.5080374286595007e-05,
"loss": 0.3446,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.32117941779063747,
"learning_rate": 2.5025456951169677e-05,
"loss": 0.3511,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.2521433374488664,
"learning_rate": 2.4970499141404942e-05,
"loss": 0.3347,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.2904537870964669,
"learning_rate": 2.491550129992345e-05,
"loss": 0.3417,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.28947611732842793,
"learning_rate": 2.486046386967024e-05,
"loss": 0.3399,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.3142796001350565,
"learning_rate": 2.4805387293909214e-05,
"loss": 0.3354,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.2984267704246262,
"learning_rate": 2.4750272016219552e-05,
"loss": 0.3394,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.3040445173693886,
"learning_rate": 2.4695118480492114e-05,
"loss": 0.3519,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.3186839718285975,
"learning_rate": 2.4639927130925898e-05,
"loss": 0.3418,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.2925558077639244,
"learning_rate": 2.458469841202444e-05,
"loss": 0.3492,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.2919096994367376,
"learning_rate": 2.452943276859226e-05,
"loss": 0.3437,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.3057436484875994,
"learning_rate": 2.447413064573125e-05,
"loss": 0.3364,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.3009016581403,
"learning_rate": 2.4418792488837095e-05,
"loss": 0.3545,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.3214757264551615,
"learning_rate": 2.4363418743595713e-05,
"loss": 0.3453,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.30248170266973096,
"learning_rate": 2.430800985597963e-05,
"loss": 0.3449,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.27226966832775207,
"learning_rate": 2.4252566272244415e-05,
"loss": 0.3494,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.350105453019002,
"learning_rate": 2.4197088438925063e-05,
"loss": 0.3457,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.28819155544489444,
"learning_rate": 2.4141576802832417e-05,
"loss": 0.3452,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.3084961176141965,
"learning_rate": 2.408603181104957e-05,
"loss": 0.3451,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.293580255742804,
"learning_rate": 2.4030453910928245e-05,
"loss": 0.3477,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.29182576151138706,
"learning_rate": 2.397484355008521e-05,
"loss": 0.3538,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.294267337585354,
"learning_rate": 2.3919201176398662e-05,
"loss": 0.3454,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.23383871706354864,
"learning_rate": 2.3863527238004633e-05,
"loss": 0.3399,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.2985314685866447,
"learning_rate": 2.380782218329337e-05,
"loss": 0.3396,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.2614286458146216,
"learning_rate": 2.3752086460905725e-05,
"loss": 0.3491,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.26615317831365887,
"learning_rate": 2.3696320519729544e-05,
"loss": 0.3398,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.29050410214068184,
"learning_rate": 2.3640524808896045e-05,
"loss": 0.342,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.2747518654549072,
"learning_rate": 2.3584699777776222e-05,
"loss": 0.3498,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.2968868266930472,
"learning_rate": 2.3528845875977195e-05,
"loss": 0.3551,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.2727825336388592,
"learning_rate": 2.3472963553338614e-05,
"loss": 0.3363,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.28885936953467806,
"learning_rate": 2.341705325992901e-05,
"loss": 0.3508,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.2669292288939344,
"learning_rate": 2.336111544604222e-05,
"loss": 0.3378,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.2734063638755421,
"learning_rate": 2.33051505621937e-05,
"loss": 0.3464,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.28462093842983577,
"learning_rate": 2.324915905911693e-05,
"loss": 0.3482,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.251490962278227,
"learning_rate": 2.319314138775977e-05,
"loss": 0.3399,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.29907777121067447,
"learning_rate": 2.3137097999280856e-05,
"loss": 0.3382,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.2492864701030473,
"learning_rate": 2.308102934504593e-05,
"loss": 0.3441,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.27442087808570453,
"learning_rate": 2.3024935876624222e-05,
"loss": 0.3406,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.24373496800520394,
"learning_rate": 2.2968818045784813e-05,
"loss": 0.3402,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.25792759297013385,
"learning_rate": 2.2912676304493006e-05,
"loss": 0.3398,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.24849263146486716,
"learning_rate": 2.2856511104906668e-05,
"loss": 0.3377,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.27611688436305837,
"learning_rate": 2.2800322899372586e-05,
"loss": 0.3455,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.26578180165577875,
"learning_rate": 2.2744112140422844e-05,
"loss": 0.3417,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.2882524097792701,
"learning_rate": 2.2687879280771177e-05,
"loss": 0.3416,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.2913068589227096,
"learning_rate": 2.26316247733093e-05,
"loss": 0.3446,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.22610315824823993,
"learning_rate": 2.257534907110328e-05,
"loss": 0.328,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.2857859193162673,
"learning_rate": 2.2519052627389882e-05,
"loss": 0.3429,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.2311453129436023,
"learning_rate": 2.246273589557294e-05,
"loss": 0.3334,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.26685314888543143,
"learning_rate": 2.240639932921966e-05,
"loss": 0.3515,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.2947413613130862,
"learning_rate": 2.2350043382056995e-05,
"loss": 0.3392,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.2509411653995594,
"learning_rate": 2.2293668507968015e-05,
"loss": 0.3282,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.3182881434752697,
"learning_rate": 2.2237275160988186e-05,
"loss": 0.3382,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.2698222732105087,
"learning_rate": 2.2180863795301787e-05,
"loss": 0.3444,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.2648894692362686,
"learning_rate": 2.212443486523819e-05,
"loss": 0.3388,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.2801673036695503,
"learning_rate": 2.2067988825268243e-05,
"loss": 0.3577,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.2538877305303437,
"learning_rate": 2.2011526130000596e-05,
"loss": 0.3344,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.3048365401280818,
"learning_rate": 2.1955047234178038e-05,
"loss": 0.3495,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.24162846998779197,
"learning_rate": 2.1898552592673825e-05,
"loss": 0.3421,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.2989691064626192,
"learning_rate": 2.184204266048803e-05,
"loss": 0.3483,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.27349897265306655,
"learning_rate": 2.1785517892743887e-05,
"loss": 0.3433,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.28817094107949187,
"learning_rate": 2.17289787446841e-05,
"loss": 0.3439,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.29302268049770963,
"learning_rate": 2.1672425671667198e-05,
"loss": 0.3466,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 0.2589490476168462,
"learning_rate": 2.161585912916385e-05,
"loss": 0.3319,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.3113297299173182,
"learning_rate": 2.1559279572753214e-05,
"loss": 0.344,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.27772518260112466,
"learning_rate": 2.1502687458119268e-05,
"loss": 0.3481,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.2693132474492103,
"learning_rate": 2.1446083241047116e-05,
"loss": 0.3339,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.2931615995036157,
"learning_rate": 2.1389467377419333e-05,
"loss": 0.3407,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.2361642266582641,
"learning_rate": 2.133284032321232e-05,
"loss": 0.3483,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.33508652068660527,
"learning_rate": 2.1276202534492566e-05,
"loss": 0.3327,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.2231517560470486,
"learning_rate": 2.121955446741306e-05,
"loss": 0.3472,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.3037276806211225,
"learning_rate": 2.1162896578209517e-05,
"loss": 0.3395,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.25862858064974625,
"learning_rate": 2.1106229323196813e-05,
"loss": 0.347,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.2422351452573029,
"learning_rate": 2.1049553158765214e-05,
"loss": 0.3313,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.3116277672983508,
"learning_rate": 2.0992868541376764e-05,
"loss": 0.3299,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.2701980230143509,
"learning_rate": 2.093617592756158e-05,
"loss": 0.3367,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.2901751053300537,
"learning_rate": 2.0879475773914167e-05,
"loss": 0.3369,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.31050393943682025,
"learning_rate": 2.082276853708978e-05,
"loss": 0.3387,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.25796037759748397,
"learning_rate": 2.076605467380071e-05,
"loss": 0.3551,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.3196788251304496,
"learning_rate": 2.0709334640812613e-05,
"loss": 0.3425,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.25283390905698283,
"learning_rate": 2.0652608894940824e-05,
"loss": 0.3419,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.3085952239278846,
"learning_rate": 2.0595877893046722e-05,
"loss": 0.3398,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.2610904887981524,
"learning_rate": 2.0539142092033985e-05,
"loss": 0.3368,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.28349751049550226,
"learning_rate": 2.048240194884496e-05,
"loss": 0.3332,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.31070113287192974,
"learning_rate": 2.042565792045695e-05,
"loss": 0.3474,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.29405184460560635,
"learning_rate": 2.036891046387857e-05,
"loss": 0.3516,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.3203409771676577,
"learning_rate": 2.0312160036146036e-05,
"loss": 0.3469,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.3312967107623391,
"learning_rate": 2.025540709431948e-05,
"loss": 0.3419,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.3141854988500302,
"learning_rate": 2.0198652095479298e-05,
"loss": 0.3302,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.31660065679624555,
"learning_rate": 2.014189549672245e-05,
"loss": 0.338,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.28583913076459294,
"learning_rate": 2.0085137755158776e-05,
"loss": 0.3403,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.32392242835653706,
"learning_rate": 2.0028379327907327e-05,
"loss": 0.3368,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.25767523958710553,
"learning_rate": 1.9971620672092676e-05,
"loss": 0.3491,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.2846034103203795,
"learning_rate": 1.991486224484123e-05,
"loss": 0.3482,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.28733611247025054,
"learning_rate": 1.985810450327756e-05,
"loss": 0.3411,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.2988403424707842,
"learning_rate": 1.9801347904520706e-05,
"loss": 0.3464,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.2674784020826904,
"learning_rate": 1.974459290568053e-05,
"loss": 0.3428,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.2682242571716748,
"learning_rate": 1.968783996385397e-05,
"loss": 0.337,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.27673860895617775,
"learning_rate": 1.963108953612143e-05,
"loss": 0.3519,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.2729473542558062,
"learning_rate": 1.9574342079543056e-05,
"loss": 0.3404,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.29013990546938373,
"learning_rate": 1.9517598051155046e-05,
"loss": 0.3428,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.2895723873466397,
"learning_rate": 1.9460857907966025e-05,
"loss": 0.3378,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.2314424114677329,
"learning_rate": 1.9404122106953285e-05,
"loss": 0.3316,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.3047244186392826,
"learning_rate": 1.9347391105059176e-05,
"loss": 0.3431,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.2531643096248829,
"learning_rate": 1.92906653591874e-05,
"loss": 0.3388,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.26350411012957775,
"learning_rate": 1.9233945326199295e-05,
"loss": 0.3422,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.26142344587841004,
"learning_rate": 1.917723146291022e-05,
"loss": 0.3269,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.2564077248987844,
"learning_rate": 1.912052422608584e-05,
"loss": 0.3442,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.24276857520164982,
"learning_rate": 1.9063824072438428e-05,
"loss": 0.3352,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.25170567535802923,
"learning_rate": 1.9007131458623246e-05,
"loss": 0.3343,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.2718055243926052,
"learning_rate": 1.895044684123479e-05,
"loss": 0.3471,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.24006190904944114,
"learning_rate": 1.8893770676803194e-05,
"loss": 0.3419,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.32451553097198343,
"learning_rate": 1.8837103421790486e-05,
"loss": 0.3354,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.219335584606544,
"learning_rate": 1.8780445532586952e-05,
"loss": 0.3355,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.2787423195880007,
"learning_rate": 1.872379746550743e-05,
"loss": 0.3358,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.2594118384403886,
"learning_rate": 1.866715967678769e-05,
"loss": 0.3453,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.27675766294944687,
"learning_rate": 1.861053262258067e-05,
"loss": 0.3407,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.2526874632453657,
"learning_rate": 1.8553916758952897e-05,
"loss": 0.3518,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.25400302652225804,
"learning_rate": 1.8497312541880735e-05,
"loss": 0.3395,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.2500463875292118,
"learning_rate": 1.8440720427246786e-05,
"loss": 0.3512,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.25262569467408647,
"learning_rate": 1.8384140870836157e-05,
"loss": 0.3458,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.2637766719711524,
"learning_rate": 1.8327574328332806e-05,
"loss": 0.3459,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.24392409798267242,
"learning_rate": 1.8271021255315906e-05,
"loss": 0.3504,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.23583964577140543,
"learning_rate": 1.8214482107256117e-05,
"loss": 0.3416,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.25114224323903916,
"learning_rate": 1.8157957339511968e-05,
"loss": 0.3389,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.24578741015821304,
"learning_rate": 1.8101447407326182e-05,
"loss": 0.3419,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.24288171471441683,
"learning_rate": 1.8044952765821966e-05,
"loss": 0.3526,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.24190959855781663,
"learning_rate": 1.7988473869999407e-05,
"loss": 0.3429,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.2707158117071247,
"learning_rate": 1.7932011174731764e-05,
"loss": 0.3468,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.22832900905011716,
"learning_rate": 1.7875565134761817e-05,
"loss": 0.3388,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.2494138541889771,
"learning_rate": 1.7819136204698226e-05,
"loss": 0.3488,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.27696358030492535,
"learning_rate": 1.776272483901182e-05,
"loss": 0.3386,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.23818728567546324,
"learning_rate": 1.7706331492031995e-05,
"loss": 0.3429,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.2696041224255792,
"learning_rate": 1.764995661794301e-05,
"loss": 0.3381,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.23727228512048343,
"learning_rate": 1.759360067078035e-05,
"loss": 0.3303,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.2633137369284493,
"learning_rate": 1.7537264104427064e-05,
"loss": 0.3364,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.25054589526362825,
"learning_rate": 1.748094737261012e-05,
"loss": 0.3455,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.2506959350864187,
"learning_rate": 1.7424650928896726e-05,
"loss": 0.3359,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.25380281083272876,
"learning_rate": 1.7368375226690712e-05,
"loss": 0.3378,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.26093363640738276,
"learning_rate": 1.731212071922883e-05,
"loss": 0.3442,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.25128125709168087,
"learning_rate": 1.7255887859577156e-05,
"loss": 0.3367,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.2684290419582953,
"learning_rate": 1.7199677100627427e-05,
"loss": 0.3531,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.2911525152128766,
"learning_rate": 1.7143488895093343e-05,
"loss": 0.3353,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.23940486798909638,
"learning_rate": 1.7087323695506994e-05,
"loss": 0.3398,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.29167891994872785,
"learning_rate": 1.7031181954215194e-05,
"loss": 0.3548,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.2541222622223528,
"learning_rate": 1.6975064123375788e-05,
"loss": 0.3474,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.25956402020717384,
"learning_rate": 1.6918970654954084e-05,
"loss": 0.3357,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.2599382675747497,
"learning_rate": 1.686290200071915e-05,
"loss": 0.3461,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.24212768771235257,
"learning_rate": 1.6806858612240234e-05,
"loss": 0.3408,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.25663512714723946,
"learning_rate": 1.6750840940883078e-05,
"loss": 0.3475,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.2367487866168618,
"learning_rate": 1.6694849437806305e-05,
"loss": 0.3348,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.2674224692787709,
"learning_rate": 1.663888455395778e-05,
"loss": 0.3456,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.2708017390579473,
"learning_rate": 1.6582946740070995e-05,
"loss": 0.3478,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.24752619723866218,
"learning_rate": 1.6527036446661396e-05,
"loss": 0.3326,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.27484296113900814,
"learning_rate": 1.6471154124022818e-05,
"loss": 0.3488,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.2371100610767601,
"learning_rate": 1.6415300222223788e-05,
"loss": 0.3386,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.2692602746630848,
"learning_rate": 1.6359475191103958e-05,
"loss": 0.3237,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.3156111617022998,
"learning_rate": 1.6303679480270466e-05,
"loss": 0.2827,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.30127366371366354,
"learning_rate": 1.624791353909428e-05,
"loss": 0.2878,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 0.39709771079648887,
"learning_rate": 1.619217781670663e-05,
"loss": 0.2965,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.3151484279975859,
"learning_rate": 1.6136472761995373e-05,
"loss": 0.2809,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.35842955545454147,
"learning_rate": 1.608079882360134e-05,
"loss": 0.2878,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.30831372607628604,
"learning_rate": 1.60251564499148e-05,
"loss": 0.2954,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.3135829071538165,
"learning_rate": 1.596954608907176e-05,
"loss": 0.2897,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.3407544765199251,
"learning_rate": 1.591396818895043e-05,
"loss": 0.297,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.29361395589020123,
"learning_rate": 1.585842319716759e-05,
"loss": 0.2934,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.33768021833094336,
"learning_rate": 1.5802911561074944e-05,
"loss": 0.2846,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.28017680559675795,
"learning_rate": 1.5747433727755595e-05,
"loss": 0.2798,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.3327482705898758,
"learning_rate": 1.5691990144020376e-05,
"loss": 0.2951,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.29642586163007373,
"learning_rate": 1.5636581256404297e-05,
"loss": 0.2839,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.31578943706920176,
"learning_rate": 1.558120751116291e-05,
"loss": 0.287,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.2799816075064646,
"learning_rate": 1.552586935426876e-05,
"loss": 0.2797,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.2556363696735114,
"learning_rate": 1.547056723140774e-05,
"loss": 0.2852,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.2857722790807482,
"learning_rate": 1.5415301587975565e-05,
"loss": 0.2903,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.27053500462268903,
"learning_rate": 1.536007286907411e-05,
"loss": 0.2836,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.2500570978345988,
"learning_rate": 1.5304881519507896e-05,
"loss": 0.2803,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.30149503340113026,
"learning_rate": 1.5249727983780453e-05,
"loss": 0.2863,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.23172594047028766,
"learning_rate": 1.5194612706090786e-05,
"loss": 0.2943,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.289766147771541,
"learning_rate": 1.5139536130329771e-05,
"loss": 0.2915,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.2576367699842231,
"learning_rate": 1.508449870007656e-05,
"loss": 0.2948,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.25434380082780517,
"learning_rate": 1.5029500858595056e-05,
"loss": 0.2908,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.26857994728962137,
"learning_rate": 1.4974543048830328e-05,
"loss": 0.3,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.2459492122331002,
"learning_rate": 1.4919625713405e-05,
"loss": 0.2881,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.25224788294263195,
"learning_rate": 1.4864749294615756e-05,
"loss": 0.2909,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.25848740486417204,
"learning_rate": 1.4809914234429716e-05,
"loss": 0.2987,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.25090100498089385,
"learning_rate": 1.4755120974480923e-05,
"loss": 0.2843,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.24897037451879647,
"learning_rate": 1.4700369956066771e-05,
"loss": 0.2836,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.2322523794899346,
"learning_rate": 1.4645661620144413e-05,
"loss": 0.2795,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.2554379008079068,
"learning_rate": 1.4590996407327284e-05,
"loss": 0.2918,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.23420851578819923,
"learning_rate": 1.4536374757881487e-05,
"loss": 0.2892,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.2356128095034264,
"learning_rate": 1.4481797111722271e-05,
"loss": 0.2843,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.2178291639763976,
"learning_rate": 1.4427263908410507e-05,
"loss": 0.2806,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.252165425768089,
"learning_rate": 1.4372775587149108e-05,
"loss": 0.2847,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.22831246322484872,
"learning_rate": 1.4318332586779522e-05,
"loss": 0.2886,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.24588205498834698,
"learning_rate": 1.4263935345778202e-05,
"loss": 0.286,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.21683955827715043,
"learning_rate": 1.420958430225303e-05,
"loss": 0.2849,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.23346846570650973,
"learning_rate": 1.415527989393985e-05,
"loss": 0.2836,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.24488767733508154,
"learning_rate": 1.410102255819891e-05,
"loss": 0.2836,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.2365554352876381,
"learning_rate": 1.404681273201131e-05,
"loss": 0.2951,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.22662246822749615,
"learning_rate": 1.399265085197556e-05,
"loss": 0.2908,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.26408091981878606,
"learning_rate": 1.393853735430398e-05,
"loss": 0.3002,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.2074389108319708,
"learning_rate": 1.3884472674819246e-05,
"loss": 0.2865,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.2644960654806502,
"learning_rate": 1.3830457248950864e-05,
"loss": 0.2851,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.2222341963986168,
"learning_rate": 1.377649151173163e-05,
"loss": 0.2914,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.2368687024004962,
"learning_rate": 1.3722575897794181e-05,
"loss": 0.2899,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.2111317621941381,
"learning_rate": 1.3668710841367472e-05,
"loss": 0.2671,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.2298800111178283,
"learning_rate": 1.361489677627324e-05,
"loss": 0.2892,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.22805858640009558,
"learning_rate": 1.3561134135922585e-05,
"loss": 0.2777,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.22936534511216974,
"learning_rate": 1.350742335331241e-05,
"loss": 0.2978,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.21166978249455493,
"learning_rate": 1.345376486102198e-05,
"loss": 0.2799,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.24569537978375075,
"learning_rate": 1.3400159091209414e-05,
"loss": 0.2795,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.22086410872412113,
"learning_rate": 1.3346606475608216e-05,
"loss": 0.2923,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.24985654336681337,
"learning_rate": 1.3293107445523781e-05,
"loss": 0.2838,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.21621385193626544,
"learning_rate": 1.3239662431829949e-05,
"loss": 0.2872,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.2214228625089786,
"learning_rate": 1.3186271864965509e-05,
"loss": 0.2886,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.2699827844859046,
"learning_rate": 1.3132936174930756e-05,
"loss": 0.2828,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.22135467688908,
"learning_rate": 1.3079655791283995e-05,
"loss": 0.2923,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.2634931999796635,
"learning_rate": 1.3026431143138108e-05,
"loss": 0.2932,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.23234406415476377,
"learning_rate": 1.2973262659157114e-05,
"loss": 0.293,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.21387162519415606,
"learning_rate": 1.2920150767552651e-05,
"loss": 0.2952,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.22511272987263006,
"learning_rate": 1.2867095896080607e-05,
"loss": 0.2915,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.23859410677174248,
"learning_rate": 1.2814098472037612e-05,
"loss": 0.2835,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.21829364265847162,
"learning_rate": 1.276115892225764e-05,
"loss": 0.2914,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.21864774786166938,
"learning_rate": 1.2708277673108555e-05,
"loss": 0.2809,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.22302957722419284,
"learning_rate": 1.2655455150488649e-05,
"loss": 0.2875,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.24094022355582217,
"learning_rate": 1.2602691779823272e-05,
"loss": 0.2854,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.23803199364579494,
"learning_rate": 1.2549987986061355e-05,
"loss": 0.2886,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.24541856208498114,
"learning_rate": 1.2497344193672005e-05,
"loss": 0.2865,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.2590780483474395,
"learning_rate": 1.2444760826641092e-05,
"loss": 0.2886,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.23048433540018162,
"learning_rate": 1.2392238308467817e-05,
"loss": 0.2788,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.2957610695678421,
"learning_rate": 1.2339777062161326e-05,
"loss": 0.2966,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.24418563707108268,
"learning_rate": 1.2287377510237293e-05,
"loss": 0.2994,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.23336860175947938,
"learning_rate": 1.2235040074714488e-05,
"loss": 0.2899,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.2476399037114659,
"learning_rate": 1.2182765177111434e-05,
"loss": 0.2892,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.21888115928533464,
"learning_rate": 1.213055323844297e-05,
"loss": 0.2864,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.23118375096734753,
"learning_rate": 1.2078404679216864e-05,
"loss": 0.293,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.2194811286081674,
"learning_rate": 1.2026319919430458e-05,
"loss": 0.2806,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.22995377339499487,
"learning_rate": 1.1974299378567227e-05,
"loss": 0.2928,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.2582062406995477,
"learning_rate": 1.1922343475593462e-05,
"loss": 0.2871,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.22808392761672233,
"learning_rate": 1.187045262895488e-05,
"loss": 0.2783,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.21914223127298438,
"learning_rate": 1.1818627256573203e-05,
"loss": 0.2941,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.22698944764243292,
"learning_rate": 1.1766867775842864e-05,
"loss": 0.3023,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.24881309782456482,
"learning_rate": 1.1715174603627615e-05,
"loss": 0.2813,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.22459712615873217,
"learning_rate": 1.1663548156257147e-05,
"loss": 0.29,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.24374552861810447,
"learning_rate": 1.161198884952377e-05,
"loss": 0.2957,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.23719997286010455,
"learning_rate": 1.1560497098679056e-05,
"loss": 0.2902,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.2076039276524381,
"learning_rate": 1.1509073318430479e-05,
"loss": 0.2796,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.2403595675276985,
"learning_rate": 1.1457717922938116e-05,
"loss": 0.2974,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.21919759722851223,
"learning_rate": 1.1406431325811233e-05,
"loss": 0.2758,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.24277947100145036,
"learning_rate": 1.135521394010506e-05,
"loss": 0.2819,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.231690175837689,
"learning_rate": 1.1304066178317367e-05,
"loss": 0.2911,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.2117615777377464,
"learning_rate": 1.1252988452385199e-05,
"loss": 0.2876,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.21460751990692337,
"learning_rate": 1.1201981173681536e-05,
"loss": 0.2925,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.22447004836038562,
"learning_rate": 1.1151044753011991e-05,
"loss": 0.2829,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.22088196652281472,
"learning_rate": 1.1100179600611491e-05,
"loss": 0.284,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.21429351878089375,
"learning_rate": 1.1049386126140985e-05,
"loss": 0.2863,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.23788725971155097,
"learning_rate": 1.0998664738684128e-05,
"loss": 0.2962,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.22570340168228623,
"learning_rate": 1.0948015846744e-05,
"loss": 0.2812,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.23137730145347762,
"learning_rate": 1.0897439858239832e-05,
"loss": 0.2851,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.25575990384944103,
"learning_rate": 1.0846937180503652e-05,
"loss": 0.2916,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.22295115359388806,
"learning_rate": 1.0796508220277117e-05,
"loss": 0.2926,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.23875012677686833,
"learning_rate": 1.0746153383708107e-05,
"loss": 0.293,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.24476072051305994,
"learning_rate": 1.0695873076347579e-05,
"loss": 0.2983,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.22113594879011647,
"learning_rate": 1.0645667703146205e-05,
"loss": 0.2995,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.23276823233591484,
"learning_rate": 1.0595537668451161e-05,
"loss": 0.2872,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.24233116300782792,
"learning_rate": 1.0545483376002854e-05,
"loss": 0.2773,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.23118867189954087,
"learning_rate": 1.0495505228931676e-05,
"loss": 0.2829,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.21394019267555825,
"learning_rate": 1.044560362975474e-05,
"loss": 0.2887,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.20862319221126666,
"learning_rate": 1.0395778980372695e-05,
"loss": 0.2859,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.24719155100138368,
"learning_rate": 1.0346031682066381e-05,
"loss": 0.2881,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.23561230822092516,
"learning_rate": 1.0296362135493724e-05,
"loss": 0.2901,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.22946508851977754,
"learning_rate": 1.0246770740686422e-05,
"loss": 0.2897,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.2517837520044373,
"learning_rate": 1.0197257897046743e-05,
"loss": 0.29,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.2327906411501294,
"learning_rate": 1.014782400334433e-05,
"loss": 0.2912,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.20982966785975699,
"learning_rate": 1.009846945771296e-05,
"loss": 0.2831,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.24564768077924384,
"learning_rate": 1.0049194657647363e-05,
"loss": 0.2915,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.20977323088060756,
"learning_rate": 1.0000000000000006e-05,
"loss": 0.2896,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.23101403588796296,
"learning_rate": 9.950885880977891e-06,
"loss": 0.284,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.2237214320337388,
"learning_rate": 9.901852696139382e-06,
"loss": 0.2951,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.21000550603019255,
"learning_rate": 9.852900840391027e-06,
"loss": 0.2916,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.21766420921098908,
"learning_rate": 9.804030707984313e-06,
"loss": 0.2968,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.19973189732440422,
"learning_rate": 9.755242692512599e-06,
"loss": 0.2799,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.20981235468681173,
"learning_rate": 9.70653718690782e-06,
"loss": 0.2942,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.21318505046853867,
"learning_rate": 9.657914583437454e-06,
"loss": 0.2824,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.21398111939255257,
"learning_rate": 9.609375273701246e-06,
"loss": 0.2856,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.19374378884196078,
"learning_rate": 9.560919648628133e-06,
"loss": 0.2858,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.20856460468708332,
"learning_rate": 9.512548098473047e-06,
"loss": 0.281,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.19695897500258908,
"learning_rate": 9.464261012813825e-06,
"loss": 0.2884,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.1959705482968179,
"learning_rate": 9.416058780547987e-06,
"loss": 0.2863,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.19160361821608274,
"learning_rate": 9.367941789889714e-06,
"loss": 0.298,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.2224063361546436,
"learning_rate": 9.319910428366607e-06,
"loss": 0.2968,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.20448257183821464,
"learning_rate": 9.271965082816667e-06,
"loss": 0.2947,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.1884339921582658,
"learning_rate": 9.224106139385111e-06,
"loss": 0.2936,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.22659761058592914,
"learning_rate": 9.176333983521291e-06,
"loss": 0.2958,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.2040722003418143,
"learning_rate": 9.12864899997558e-06,
"loss": 0.2913,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.2118898151058537,
"learning_rate": 9.08105157279628e-06,
"loss": 0.2845,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.22077857047453583,
"learning_rate": 9.03354208532653e-06,
"loss": 0.2819,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.21203332399808764,
"learning_rate": 8.986120920201205e-06,
"loss": 0.2968,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.21543615533183721,
"learning_rate": 8.938788459343852e-06,
"loss": 0.2833,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.1932930435958138,
"learning_rate": 8.8915450839636e-06,
"loss": 0.2877,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.21524460745741267,
"learning_rate": 8.844391174552116e-06,
"loss": 0.2926,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.2130080184600181,
"learning_rate": 8.797327110880479e-06,
"loss": 0.2862,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.21175719170898077,
"learning_rate": 8.750353271996206e-06,
"loss": 0.2831,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.21134333849439502,
"learning_rate": 8.703470036220132e-06,
"loss": 0.2885,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.1982470815557418,
"learning_rate": 8.656677781143394e-06,
"loss": 0.2866,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.20955219847681394,
"learning_rate": 8.609976883624377e-06,
"loss": 0.2944,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.21574578508285683,
"learning_rate": 8.563367719785698e-06,
"loss": 0.2921,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.2012385194952385,
"learning_rate": 8.516850665011138e-06,
"loss": 0.2868,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.20966975522927905,
"learning_rate": 8.47042609394269e-06,
"loss": 0.2868,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.19073181056640115,
"learning_rate": 8.424094380477432e-06,
"loss": 0.2914,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.20345886233302707,
"learning_rate": 8.37785589776465e-06,
"loss": 0.2903,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.20192351669088612,
"learning_rate": 8.331711018202694e-06,
"loss": 0.2908,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.19390335153182522,
"learning_rate": 8.285660113436104e-06,
"loss": 0.2869,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.1986795613001007,
"learning_rate": 8.239703554352527e-06,
"loss": 0.2784,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.20924408698077596,
"learning_rate": 8.193841711079775e-06,
"loss": 0.2964,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.20393986812144824,
"learning_rate": 8.148074952982828e-06,
"loss": 0.2846,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.19926080815144537,
"learning_rate": 8.102403648660859e-06,
"loss": 0.2803,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.1915616756426903,
"learning_rate": 8.056828165944282e-06,
"loss": 0.2807,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.1960584764431038,
"learning_rate": 8.011348871891762e-06,
"loss": 0.2863,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.1885435930482296,
"learning_rate": 7.965966132787287e-06,
"loss": 0.2936,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.21280659194653984,
"learning_rate": 7.920680314137189e-06,
"loss": 0.2853,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.1959952315181185,
"learning_rate": 7.875491780667246e-06,
"loss": 0.2927,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.21244878590119928,
"learning_rate": 7.830400896319667e-06,
"loss": 0.2887,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.20689216583239142,
"learning_rate": 7.785408024250259e-06,
"loss": 0.2967,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.20210372852546088,
"learning_rate": 7.74051352682542e-06,
"loss": 0.2871,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.1908513878936331,
"learning_rate": 7.695717765619257e-06,
"loss": 0.2937,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.20995184567207462,
"learning_rate": 7.651021101410673e-06,
"loss": 0.2894,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.20285618942718117,
"learning_rate": 7.606423894180464e-06,
"loss": 0.2859,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.20188817556717023,
"learning_rate": 7.56192650310839e-06,
"loss": 0.2919,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.2124422141065988,
"learning_rate": 7.517529286570349e-06,
"loss": 0.2943,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.19875770213857916,
"learning_rate": 7.473232602135387e-06,
"loss": 0.2921,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.190009214159585,
"learning_rate": 7.429036806562935e-06,
"loss": 0.2877,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.21801869287591186,
"learning_rate": 7.3849422557998455e-06,
"loss": 0.2835,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.1894529033709107,
"learning_rate": 7.340949304977567e-06,
"loss": 0.2801,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.20208802058691822,
"learning_rate": 7.297058308409282e-06,
"loss": 0.291,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.21583624345087335,
"learning_rate": 7.25326961958704e-06,
"loss": 0.2929,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.20416162166182467,
"learning_rate": 7.209583591178921e-06,
"loss": 0.2928,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.2101250452017512,
"learning_rate": 7.1660005750261925e-06,
"loss": 0.2968,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.20765810154610526,
"learning_rate": 7.1225209221404765e-06,
"loss": 0.2853,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.18920835636352015,
"learning_rate": 7.079144982700909e-06,
"loss": 0.2999,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.1968041407836695,
"learning_rate": 7.0358731060513695e-06,
"loss": 0.2881,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.20911022378066524,
"learning_rate": 6.99270564069757e-06,
"loss": 0.2878,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.19268314796727823,
"learning_rate": 6.949642934304375e-06,
"loss": 0.2992,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.19543690963516147,
"learning_rate": 6.906685333692871e-06,
"loss": 0.2837,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.20542554414093706,
"learning_rate": 6.86383318483769e-06,
"loss": 0.2921,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.20099188254729913,
"learning_rate": 6.821086832864139e-06,
"loss": 0.2866,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.2055782596963447,
"learning_rate": 6.77844662204546e-06,
"loss": 0.2851,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.20243169893643642,
"learning_rate": 6.7359128958000455e-06,
"loss": 0.2862,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.20049989179390335,
"learning_rate": 6.693485996688695e-06,
"loss": 0.2888,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.19442744425875272,
"learning_rate": 6.651166266411801e-06,
"loss": 0.2815,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.19383474100460565,
"learning_rate": 6.6089540458066725e-06,
"loss": 0.2917,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.2301586190857945,
"learning_rate": 6.566849674844711e-06,
"loss": 0.2884,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.20195302646430932,
"learning_rate": 6.524853492628747e-06,
"loss": 0.2943,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.20204172009107085,
"learning_rate": 6.4829658373902536e-06,
"loss": 0.2778,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.18812700101949584,
"learning_rate": 6.441187046486648e-06,
"loss": 0.2807,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.19883358712550522,
"learning_rate": 6.399517456398567e-06,
"loss": 0.2945,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.18838184845753486,
"learning_rate": 6.357957402727164e-06,
"loss": 0.2887,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.19037908885991836,
"learning_rate": 6.316507220191395e-06,
"loss": 0.2847,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.2017858502673898,
"learning_rate": 6.275167242625331e-06,
"loss": 0.3006,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.20426482703446378,
"learning_rate": 6.233937802975471e-06,
"loss": 0.2915,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.19338269662366372,
"learning_rate": 6.192819233298046e-06,
"loss": 0.2849,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.20051099532877387,
"learning_rate": 6.151811864756383e-06,
"loss": 0.2985,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.18797652239273668,
"learning_rate": 6.1109160276181655e-06,
"loss": 0.2886,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.20423579584123683,
"learning_rate": 6.070132051252868e-06,
"loss": 0.2961,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.20035032382907575,
"learning_rate": 6.0294602641290034e-06,
"loss": 0.2903,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.19826720893384456,
"learning_rate": 5.988900993811575e-06,
"loss": 0.2853,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.20025790313974998,
"learning_rate": 5.948454566959363e-06,
"loss": 0.2931,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.18517063364536782,
"learning_rate": 5.908121309322328e-06,
"loss": 0.2847,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.20704799528109608,
"learning_rate": 5.867901545738976e-06,
"loss": 0.2842,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.19694546864120693,
"learning_rate": 5.827795600133774e-06,
"loss": 0.2973,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.19951294545912585,
"learning_rate": 5.787803795514466e-06,
"loss": 0.2881,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.19486156635944588,
"learning_rate": 5.747926453969576e-06,
"loss": 0.2892,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.1937183497624411,
"learning_rate": 5.708163896665708e-06,
"loss": 0.2923,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.19064527936971076,
"learning_rate": 5.668516443845047e-06,
"loss": 0.2883,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.19415987083346353,
"learning_rate": 5.6289844148227225e-06,
"loss": 0.2921,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.2032751232030011,
"learning_rate": 5.5895681279842615e-06,
"loss": 0.293,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.1935688344061829,
"learning_rate": 5.550267900783019e-06,
"loss": 0.2825,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.18890423864407663,
"learning_rate": 5.511084049737623e-06,
"loss": 0.2829,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.1893513765809406,
"learning_rate": 5.4720168904294215e-06,
"loss": 0.2886,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.19867131999531198,
"learning_rate": 5.433066737499948e-06,
"loss": 0.2842,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.19124326076683942,
"learning_rate": 5.394233904648376e-06,
"loss": 0.2822,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.19062535201404687,
"learning_rate": 5.355518704628997e-06,
"loss": 0.2877,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.19955437421620836,
"learning_rate": 5.316921449248731e-06,
"loss": 0.286,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.18230726669447045,
"learning_rate": 5.278442449364538e-06,
"loss": 0.2845,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.18531256562093587,
"learning_rate": 5.240082014881016e-06,
"loss": 0.2917,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.19916037832912253,
"learning_rate": 5.201840454747822e-06,
"loss": 0.2814,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.19407343521842813,
"learning_rate": 5.163718076957223e-06,
"loss": 0.2752,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.18008156862889893,
"learning_rate": 5.125715188541609e-06,
"loss": 0.298,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.18882998241016782,
"learning_rate": 5.087832095571021e-06,
"loss": 0.2828,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.18568275618158667,
"learning_rate": 5.0500691031506766e-06,
"loss": 0.2762,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.20577112324183755,
"learning_rate": 5.01242651541854e-06,
"loss": 0.282,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.1810205297443181,
"learning_rate": 4.974904635542815e-06,
"loss": 0.2918,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.1840046239909962,
"learning_rate": 4.937503765719582e-06,
"loss": 0.2744,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.18951558793468745,
"learning_rate": 4.900224207170299e-06,
"loss": 0.29,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.18591438947243705,
"learning_rate": 4.8630662601394065e-06,
"loss": 0.2943,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.17769454962473497,
"learning_rate": 4.8260302238918995e-06,
"loss": 0.2867,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.18911107564559626,
"learning_rate": 4.789116396710924e-06,
"loss": 0.2919,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.18089315754001642,
"learning_rate": 4.752325075895368e-06,
"loss": 0.2864,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.18404084636078338,
"learning_rate": 4.715656557757473e-06,
"loss": 0.277,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.18920952364471869,
"learning_rate": 4.679111137620442e-06,
"loss": 0.2956,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.19963513040181288,
"learning_rate": 4.6426891098160585e-06,
"loss": 0.2826,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.18516435281698163,
"learning_rate": 4.6063907676823474e-06,
"loss": 0.2968,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.1840123042677989,
"learning_rate": 4.570216403561141e-06,
"loss": 0.2864,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.22072408961623452,
"learning_rate": 4.534166308795815e-06,
"loss": 0.2641,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.3109826011715752,
"learning_rate": 4.498240773728859e-06,
"loss": 0.2597,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.21236319969064427,
"learning_rate": 4.462440087699609e-06,
"loss": 0.2384,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.29206449345959545,
"learning_rate": 4.426764539041861e-06,
"loss": 0.2517,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 0.3593353647970361,
"learning_rate": 4.391214415081582e-06,
"loss": 0.2487,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 0.23566310126261486,
"learning_rate": 4.355790002134579e-06,
"loss": 0.2476,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.2570501090990083,
"learning_rate": 4.320491585504207e-06,
"loss": 0.2472,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.30135765737185005,
"learning_rate": 4.2853194494790615e-06,
"loss": 0.2407,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.2557921233963976,
"learning_rate": 4.250273877330691e-06,
"loss": 0.267,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.19909136494839366,
"learning_rate": 4.215355151311313e-06,
"loss": 0.2476,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.24394013730263225,
"learning_rate": 4.180563552651542e-06,
"loss": 0.2557,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.26616730104188213,
"learning_rate": 4.145899361558147e-06,
"loss": 0.245,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.2487998597257695,
"learning_rate": 4.111362857211738e-06,
"loss": 0.2609,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.20619886799698964,
"learning_rate": 4.076954317764592e-06,
"loss": 0.2455,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.22175153348900697,
"learning_rate": 4.042674020338335e-06,
"loss": 0.252,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.24177161518297863,
"learning_rate": 4.0085222410217835e-06,
"loss": 0.2502,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.22673198406378195,
"learning_rate": 3.974499254868674e-06,
"loss": 0.2518,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.20733063196106952,
"learning_rate": 3.940605335895451e-06,
"loss": 0.2541,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.2263650584704158,
"learning_rate": 3.90684075707908e-06,
"loss": 0.2377,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.23151955879078134,
"learning_rate": 3.8732057903548505e-06,
"loss": 0.2507,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.20862842872727,
"learning_rate": 3.8397007066141375e-06,
"loss": 0.2541,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.19805497897835905,
"learning_rate": 3.806325775702304e-06,
"loss": 0.2584,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.20685686705211082,
"learning_rate": 3.773081266416434e-06,
"loss": 0.2548,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.19835993215243292,
"learning_rate": 3.739967446503245e-06,
"loss": 0.2412,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.20177591163670675,
"learning_rate": 3.706984582656894e-06,
"loss": 0.2476,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.19243166166464343,
"learning_rate": 3.6741329405168237e-06,
"loss": 0.2527,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.20346556270601246,
"learning_rate": 3.641412784665648e-06,
"loss": 0.259,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.20377354489841681,
"learning_rate": 3.608824378627005e-06,
"loss": 0.2526,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.19924779656644798,
"learning_rate": 3.5763679848634337e-06,
"loss": 0.2483,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.186132410789942,
"learning_rate": 3.544043864774269e-06,
"loss": 0.2497,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.18151252708417862,
"learning_rate": 3.5118522786935282e-06,
"loss": 0.2416,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.19510392877213434,
"learning_rate": 3.479793485887819e-06,
"loss": 0.2448,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.18549318213975774,
"learning_rate": 3.4478677445542653e-06,
"loss": 0.2373,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.19469833813574267,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.2458,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.20256798101676946,
"learning_rate": 3.3844164437320527e-06,
"loss": 0.2563,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.1957450473581279,
"learning_rate": 3.3528913952714558e-06,
"loss": 0.2526,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.18747220149731358,
"learning_rate": 3.321500420335e-06,
"loss": 0.2494,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.19281296753431634,
"learning_rate": 3.290243771741275e-06,
"loss": 0.2493,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.19354228820810698,
"learning_rate": 3.2591217012270325e-06,
"loss": 0.2553,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.18101425711464936,
"learning_rate": 3.228134459445149e-06,
"loss": 0.2556,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.1867844401128252,
"learning_rate": 3.1972822959626205e-06,
"loss": 0.2563,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.1833600017743403,
"learning_rate": 3.166565459258513e-06,
"loss": 0.2524,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.19501531487463217,
"learning_rate": 3.1359841967220193e-06,
"loss": 0.2592,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.18374759708564484,
"learning_rate": 3.105538754650419e-06,
"loss": 0.2624,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.18997673952401567,
"learning_rate": 3.07522937824712e-06,
"loss": 0.2467,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.18187177427479284,
"learning_rate": 3.0450563116196697e-06,
"loss": 0.2504,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.17798725192627368,
"learning_rate": 3.0150197977778008e-06,
"loss": 0.2535,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.18033236237833208,
"learning_rate": 2.985120078631465e-06,
"loss": 0.2547,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.26071317120735477,
"learning_rate": 2.9553573949888893e-06,
"loss": 0.2459,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.17339670866347376,
"learning_rate": 2.9257319865546384e-06,
"loss": 0.2576,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.1820739553045229,
"learning_rate": 2.896244091927678e-06,
"loss": 0.2529,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.1712223792125586,
"learning_rate": 2.8668939485994584e-06,
"loss": 0.2578,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.1818944866179739,
"learning_rate": 2.837681792951994e-06,
"loss": 0.2526,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.17688755335491166,
"learning_rate": 2.808607860255981e-06,
"loss": 0.2505,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.18689822302150585,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.2459,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.18376357219967507,
"learning_rate": 2.7508755992329937e-06,
"loss": 0.2652,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.18806676683925647,
"learning_rate": 2.722217735873718e-06,
"loss": 0.2567,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.17481314126508632,
"learning_rate": 2.6936990253975315e-06,
"loss": 0.2544,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.18903307352782164,
"learning_rate": 2.665319697490205e-06,
"loss": 0.2439,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.19042810357857942,
"learning_rate": 2.637079980714945e-06,
"loss": 0.2467,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.17829902698368333,
"learning_rate": 2.6089801025105453e-06,
"loss": 0.2393,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.17677027043656585,
"learning_rate": 2.581020289189571e-06,
"loss": 0.2525,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.17170524359681746,
"learning_rate": 2.553200765936501e-06,
"loss": 0.2418,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.18523487923139656,
"learning_rate": 2.525521756805962e-06,
"loss": 0.2455,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.17912600847095658,
"learning_rate": 2.497983484720885e-06,
"loss": 0.2513,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.18595695654680286,
"learning_rate": 2.470586171470728e-06,
"loss": 0.2473,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.1858014091266137,
"learning_rate": 2.4433300377096836e-06,
"loss": 0.2554,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.17417293685339805,
"learning_rate": 2.4162153029549073e-06,
"loss": 0.2374,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.1763826973194314,
"learning_rate": 2.3892421855847458e-06,
"loss": 0.2421,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.18205535896709768,
"learning_rate": 2.362410902836978e-06,
"loss": 0.2496,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.1801290446797605,
"learning_rate": 2.3357216708070653e-06,
"loss": 0.2465,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.18230648174987035,
"learning_rate": 2.309174704446411e-06,
"loss": 0.2481,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.17715522047587867,
"learning_rate": 2.2827702175606437e-06,
"loss": 0.2472,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.176881947470453,
"learning_rate": 2.256508422807855e-06,
"loss": 0.2511,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.17911363179562015,
"learning_rate": 2.230389531696946e-06,
"loss": 0.2493,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.18823396258855427,
"learning_rate": 2.204413754585857e-06,
"loss": 0.252,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.1735444113203639,
"learning_rate": 2.1785813006799406e-06,
"loss": 0.2458,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.186967173493876,
"learning_rate": 2.1528923780302224e-06,
"loss": 0.2407,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.18652307842236215,
"learning_rate": 2.127347193531757e-06,
"loss": 0.2625,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.17308668585646309,
"learning_rate": 2.101945952921942e-06,
"loss": 0.2549,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.1799473077516383,
"learning_rate": 2.0766888607788906e-06,
"loss": 0.2547,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.18091756609667253,
"learning_rate": 2.0515761205197337e-06,
"loss": 0.2497,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.1820584589359309,
"learning_rate": 2.0266079343990453e-06,
"loss": 0.2469,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.1804786151972986,
"learning_rate": 2.0017845035071494e-06,
"loss": 0.2448,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.1837393959244212,
"learning_rate": 1.9771060277685537e-06,
"loss": 0.2495,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.17452456750219528,
"learning_rate": 1.95257270594031e-06,
"loss": 0.2518,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.1817947077951402,
"learning_rate": 1.9281847356104188e-06,
"loss": 0.2563,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.18516194875677014,
"learning_rate": 1.9039423131962365e-06,
"loss": 0.2463,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.17191157518559985,
"learning_rate": 1.8798456339429027e-06,
"loss": 0.2485,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.18466685317832518,
"learning_rate": 1.8558948919217612e-06,
"loss": 0.2512,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.1698517872303968,
"learning_rate": 1.8320902800287954e-06,
"loss": 0.2366,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.17353926129245326,
"learning_rate": 1.8084319899830726e-06,
"loss": 0.2603,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.18289742566754488,
"learning_rate": 1.7849202123252097e-06,
"loss": 0.2366,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.17158402493628286,
"learning_rate": 1.7615551364158401e-06,
"loss": 0.2477,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.17591116120574607,
"learning_rate": 1.738336950434061e-06,
"loss": 0.2529,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.18207272730814367,
"learning_rate": 1.715265841375957e-06,
"loss": 0.2584,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.18185597817734117,
"learning_rate": 1.6923419950530684e-06,
"loss": 0.2471,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.17132893814373068,
"learning_rate": 1.6695655960909008e-06,
"loss": 0.2411,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.1783927692021874,
"learning_rate": 1.646936827927441e-06,
"loss": 0.2497,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.17421005525839167,
"learning_rate": 1.6244558728116766e-06,
"loss": 0.2406,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.18044210926761958,
"learning_rate": 1.6021229118021265e-06,
"loss": 0.2569,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.17457531742122362,
"learning_rate": 1.5799381247653967e-06,
"loss": 0.2469,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.1758778142704245,
"learning_rate": 1.5579016903747013e-06,
"loss": 0.2391,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.17994068847159217,
"learning_rate": 1.5360137861084656e-06,
"loss": 0.2537,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.1769282651187894,
"learning_rate": 1.5142745882488475e-06,
"loss": 0.2488,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.17169382234744698,
"learning_rate": 1.4926842718803691e-06,
"loss": 0.2592,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.18297004245559711,
"learning_rate": 1.4712430108884657e-06,
"loss": 0.2451,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.17718295977486975,
"learning_rate": 1.4499509779581078e-06,
"loss": 0.2549,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.17675602416953473,
"learning_rate": 1.4288083445723988e-06,
"loss": 0.2562,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.20049345925956275,
"learning_rate": 1.4078152810112045e-06,
"loss": 0.2576,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.17321190722043345,
"learning_rate": 1.3869719563497697e-06,
"loss": 0.2414,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.17066424690024606,
"learning_rate": 1.3662785384573663e-06,
"loss": 0.2665,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.17543248967190675,
"learning_rate": 1.3457351939959383e-06,
"loss": 0.2507,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.17649411478293311,
"learning_rate": 1.3253420884187551e-06,
"loss": 0.2402,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.1823517589199631,
"learning_rate": 1.3050993859690953e-06,
"loss": 0.2478,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.18018587568581015,
"learning_rate": 1.2850072496788869e-06,
"loss": 0.2436,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.1826993931646273,
"learning_rate": 1.2650658413674434e-06,
"loss": 0.2544,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.1740995992677903,
"learning_rate": 1.2452753216401226e-06,
"loss": 0.2552,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.19008930800907914,
"learning_rate": 1.2256358498870503e-06,
"loss": 0.2455,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.17125534175650167,
"learning_rate": 1.2061475842818337e-06,
"loss": 0.2509,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.17256380143481248,
"learning_rate": 1.1868106817802816e-06,
"loss": 0.2519,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.1689997682687718,
"learning_rate": 1.1676252981191482e-06,
"loss": 0.2454,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.17316927560779485,
"learning_rate": 1.1485915878148823e-06,
"loss": 0.2636,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.18262334057376045,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.2387,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.1687388390428301,
"learning_rate": 1.1109797992336847e-06,
"loss": 0.2536,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.16617481992721225,
"learning_rate": 1.092402023876933e-06,
"loss": 0.2468,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.1697433087924109,
"learning_rate": 1.0739765277149527e-06,
"loss": 0.2509,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.1687524980417368,
"learning_rate": 1.0557034591441596e-06,
"loss": 0.2589,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.1744555788062188,
"learning_rate": 1.0375829653333324e-06,
"loss": 0.2602,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.17347611549426256,
"learning_rate": 1.0196151922224385e-06,
"loss": 0.2451,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.17463170591569654,
"learning_rate": 1.0018002845214526e-06,
"loss": 0.2506,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.17780713737116136,
"learning_rate": 9.841383857091947e-07,
"loss": 0.2543,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.17087633118757817,
"learning_rate": 9.666296380321616e-07,
"loss": 0.2481,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.16912464786273665,
"learning_rate": 9.492741825034124e-07,
"loss": 0.2378,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.17622355565061887,
"learning_rate": 9.320721589013892e-07,
"loss": 0.2593,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.17306833817082815,
"learning_rate": 9.150237057688339e-07,
"loss": 0.2639,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.17388570296282102,
"learning_rate": 8.981289604116328e-07,
"loss": 0.2542,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.16994037508550494,
"learning_rate": 8.813880588977542e-07,
"loss": 0.2443,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.18359979752980382,
"learning_rate": 8.648011360561126e-07,
"loss": 0.2511,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.1712949363159629,
"learning_rate": 8.483683254755037e-07,
"loss": 0.2486,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.16940283794778085,
"learning_rate": 8.320897595035227e-07,
"loss": 0.2468,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.17443376131870597,
"learning_rate": 8.159655692455093e-07,
"loss": 0.2518,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.17557039046882578,
"learning_rate": 7.999958845634648e-07,
"loss": 0.2523,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.17331034965403483,
"learning_rate": 7.841808340750478e-07,
"loss": 0.2528,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.17697800764814217,
"learning_rate": 7.685205451524869e-07,
"loss": 0.2564,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.17344983968860042,
"learning_rate": 7.530151439216027e-07,
"loss": 0.2443,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.17936426356667903,
"learning_rate": 7.376647552607675e-07,
"loss": 0.2538,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.16813762739519647,
"learning_rate": 7.224695027998963e-07,
"loss": 0.2433,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.19680820423275083,
"learning_rate": 7.07429508919466e-07,
"loss": 0.252,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.16908942180138517,
"learning_rate": 6.925448947495206e-07,
"loss": 0.2573,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.1730566974009618,
"learning_rate": 6.778157801686936e-07,
"loss": 0.2581,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.1678836063760352,
"learning_rate": 6.632422838032515e-07,
"loss": 0.2483,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.16612436332077993,
"learning_rate": 6.488245230261281e-07,
"loss": 0.2464,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.1654186902907426,
"learning_rate": 6.345626139559868e-07,
"loss": 0.2578,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.17034006616024444,
"learning_rate": 6.204566714562866e-07,
"loss": 0.2449,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.17374395025532308,
"learning_rate": 6.06506809134344e-07,
"loss": 0.2465,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.16323109200491942,
"learning_rate": 5.927131393404373e-07,
"loss": 0.2491,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.16684897938931856,
"learning_rate": 5.790757731668817e-07,
"loss": 0.2459,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.17394360665479516,
"learning_rate": 5.655948204471507e-07,
"loss": 0.2479,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.16567849979497767,
"learning_rate": 5.522703897549875e-07,
"loss": 0.2466,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.1690957503984618,
"learning_rate": 5.391025884035239e-07,
"loss": 0.2463,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.1665479196222732,
"learning_rate": 5.260915224444207e-07,
"loss": 0.252,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.1750284581320013,
"learning_rate": 5.132372966670129e-07,
"loss": 0.2474,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.17689413274012905,
"learning_rate": 5.005400145974704e-07,
"loss": 0.2484,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.17769738286796868,
"learning_rate": 4.879997784979562e-07,
"loss": 0.2489,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.17585029833620466,
"learning_rate": 4.7561668936580984e-07,
"loss": 0.2544,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.16554313876423332,
"learning_rate": 4.6339084693272306e-07,
"loss": 0.2578,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.16976940013863526,
"learning_rate": 4.5132234966395847e-07,
"loss": 0.2552,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.1668551331959679,
"learning_rate": 4.3941129475752795e-07,
"loss": 0.2556,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.16885707195382957,
"learning_rate": 4.27657778143431e-07,
"loss": 0.2452,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.1728477388385542,
"learning_rate": 4.1606189448287757e-07,
"loss": 0.245,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.17170877232096673,
"learning_rate": 4.046237371675177e-07,
"loss": 0.2598,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.16780833423882327,
"learning_rate": 3.9334339831869963e-07,
"loss": 0.2503,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.17082630575857077,
"learning_rate": 3.8222096878671955e-07,
"loss": 0.2499,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.17068578796531464,
"learning_rate": 3.7125653815009545e-07,
"loss": 0.243,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.15960703039403926,
"learning_rate": 3.6045019471484974e-07,
"loss": 0.258,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.16807642704203282,
"learning_rate": 3.498020255137813e-07,
"loss": 0.2488,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.16983811302675064,
"learning_rate": 3.393121163057811e-07,
"loss": 0.2339,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.17255103748140893,
"learning_rate": 3.289805515751399e-07,
"loss": 0.2522,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.16843604873751813,
"learning_rate": 3.188074145308573e-07,
"loss": 0.2565,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.1699341863542953,
"learning_rate": 3.087927871059804e-07,
"loss": 0.2576,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.17802972178594334,
"learning_rate": 2.989367499569418e-07,
"loss": 0.2502,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.1687873492242771,
"learning_rate": 2.8923938246290917e-07,
"loss": 0.2467,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.1661452994021217,
"learning_rate": 2.7970076272514804e-07,
"loss": 0.2536,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.17022765378454666,
"learning_rate": 2.703209675663887e-07,
"loss": 0.2545,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.16790197972025744,
"learning_rate": 2.6110007253021374e-07,
"loss": 0.2559,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.1603074846812276,
"learning_rate": 2.520381518804471e-07,
"loss": 0.2484,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.163264552947334,
"learning_rate": 2.4313527860054585e-07,
"loss": 0.253,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.16466433046030246,
"learning_rate": 2.343915243930317e-07,
"loss": 0.2529,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.17093653941089904,
"learning_rate": 2.2580695967889367e-07,
"loss": 0.2399,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.16299879541314322,
"learning_rate": 2.1738165359704189e-07,
"loss": 0.2457,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.16793799224262962,
"learning_rate": 2.0911567400373257e-07,
"loss": 0.251,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.16964946690083713,
"learning_rate": 2.0100908747202607e-07,
"loss": 0.2358,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.16865386853884656,
"learning_rate": 1.9306195929125638e-07,
"loss": 0.2544,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.17702025401265184,
"learning_rate": 1.8527435346650247e-07,
"loss": 0.2449,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.1709617956513804,
"learning_rate": 1.7764633271807108e-07,
"loss": 0.2456,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.16795100167272917,
"learning_rate": 1.7017795848099262e-07,
"loss": 0.2555,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.17146292584311662,
"learning_rate": 1.6286929090452596e-07,
"loss": 0.2543,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.1721752166751522,
"learning_rate": 1.557203888516745e-07,
"loss": 0.2532,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.17364894258594546,
"learning_rate": 1.487313098987131e-07,
"loss": 0.2556,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.16934434326740458,
"learning_rate": 1.4190211033472402e-07,
"loss": 0.2572,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.17149888635143243,
"learning_rate": 1.3523284516113955e-07,
"loss": 0.2472,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.17105633916975987,
"learning_rate": 1.2872356809130682e-07,
"loss": 0.2485,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.1687374126020292,
"learning_rate": 1.2237433155004807e-07,
"loss": 0.2553,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.16885237641367337,
"learning_rate": 1.1618518667323886e-07,
"loss": 0.2396,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.167241762804619,
"learning_rate": 1.1015618330740385e-07,
"loss": 0.2447,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.16371504335932077,
"learning_rate": 1.042873700093061e-07,
"loss": 0.2529,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.1716075575854276,
"learning_rate": 9.857879404556291e-08,
"loss": 0.2488,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.16938576759049462,
"learning_rate": 9.303050139225722e-08,
"loss": 0.2568,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.18056411515659196,
"learning_rate": 8.76425367345779e-08,
"loss": 0.2392,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.16526525533324823,
"learning_rate": 8.241494346644897e-08,
"loss": 0.2639,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.16568720097721706,
"learning_rate": 7.734776369019204e-08,
"loss": 0.2577,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.16956324701549735,
"learning_rate": 7.244103821617332e-08,
"loss": 0.2475,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.16608451241946118,
"learning_rate": 6.769480656248606e-08,
"loss": 0.2524,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.16930635463363944,
"learning_rate": 6.310910695462635e-08,
"loss": 0.2473,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.17253719868375242,
"learning_rate": 5.8683976325191185e-08,
"loss": 0.2485,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.1615145677671123,
"learning_rate": 5.4419450313571984e-08,
"loss": 0.2458,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.16918133195162982,
"learning_rate": 5.031556326567488e-08,
"loss": 0.2452,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.17784415188143848,
"learning_rate": 4.637234823364312e-08,
"loss": 0.2425,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.16221630783489605,
"learning_rate": 4.258983697558838e-08,
"loss": 0.2493,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.17007940690315088,
"learning_rate": 3.896805995533548e-08,
"loss": 0.2475,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.16314474822593517,
"learning_rate": 3.550704634218028e-08,
"loss": 0.2432,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.1623631720199163,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.2457,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.1680909670955153,
"learning_rate": 2.9067419540278476e-08,
"loss": 0.251,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.1703022795967642,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.2505,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.16380909204407015,
"learning_rate": 2.3271164024940564e-08,
"loss": 0.2541,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.16711446403938132,
"learning_rate": 2.061435966221881e-08,
"loss": 0.2535,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.1688660267035651,
"learning_rate": 1.811846652477245e-08,
"loss": 0.2494,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.17077291061974495,
"learning_rate": 1.5783504714184106e-08,
"loss": 0.2445,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.17710019945064726,
"learning_rate": 1.360949303591097e-08,
"loss": 0.2577,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.16439073169671775,
"learning_rate": 1.1596448999144916e-08,
"loss": 0.2568,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.17228232654851597,
"learning_rate": 9.744388816668172e-09,
"loss": 0.2573,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.1610969116768146,
"learning_rate": 8.05332740472009e-09,
"loss": 0.2587,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.16945283902147948,
"learning_rate": 6.523278382872811e-09,
"loss": 0.2424,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.1746964402496059,
"learning_rate": 5.15425407393133e-09,
"loss": 0.246,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.17382363741663856,
"learning_rate": 3.94626550383137e-09,
"loss": 0.246,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.15992952950115583,
"learning_rate": 2.899322401546112e-09,
"loss": 0.2454,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.1612347770984825,
"learning_rate": 2.013433199010706e-09,
"loss": 0.2466,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.16957435139067184,
"learning_rate": 1.2886050310556563e-09,
"loss": 0.2534,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.16770099846830772,
"learning_rate": 7.248437353468695e-10,
"loss": 0.2447,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.16617845220730612,
"learning_rate": 3.221538523412449e-10,
"loss": 0.258,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.17095643973651026,
"learning_rate": 8.053862524670663e-11,
"loss": 0.2548,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.16380214195179532,
"learning_rate": 0.0,
"loss": 0.2509,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 5.92233089038372e+18,
"train_loss": 0.35468531735544284,
"train_runtime": 54703.1114,
"train_samples_per_second": 2.888,
"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": 5.92233089038372e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}