e1_science_longest_phi / trainer_state.json
sedrickkeh's picture
End of training
5b9268c verified
raw
history blame
214 kB
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 4.982278481012658,
"eval_steps": 500,
"global_step": 1230,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004050632911392405,
"grad_norm": 7.847215205064711,
"learning_rate": 3.2520325203252037e-07,
"loss": 1.9982,
"step": 1
},
{
"epoch": 0.00810126582278481,
"grad_norm": 8.118371364481783,
"learning_rate": 6.504065040650407e-07,
"loss": 2.0381,
"step": 2
},
{
"epoch": 0.012151898734177215,
"grad_norm": 8.325391506899951,
"learning_rate": 9.75609756097561e-07,
"loss": 2.0453,
"step": 3
},
{
"epoch": 0.01620253164556962,
"grad_norm": 7.976897345811807,
"learning_rate": 1.3008130081300815e-06,
"loss": 2.028,
"step": 4
},
{
"epoch": 0.020253164556962026,
"grad_norm": 7.414882093650217,
"learning_rate": 1.6260162601626018e-06,
"loss": 2.0125,
"step": 5
},
{
"epoch": 0.02430379746835443,
"grad_norm": 7.297901662018341,
"learning_rate": 1.951219512195122e-06,
"loss": 1.9921,
"step": 6
},
{
"epoch": 0.028354430379746835,
"grad_norm": 5.904528386502779,
"learning_rate": 2.2764227642276426e-06,
"loss": 1.949,
"step": 7
},
{
"epoch": 0.03240506329113924,
"grad_norm": 5.521581028173608,
"learning_rate": 2.601626016260163e-06,
"loss": 1.9634,
"step": 8
},
{
"epoch": 0.03645569620253165,
"grad_norm": 3.1009627264687176,
"learning_rate": 2.926829268292683e-06,
"loss": 1.9135,
"step": 9
},
{
"epoch": 0.04050632911392405,
"grad_norm": 2.8188521429613784,
"learning_rate": 3.2520325203252037e-06,
"loss": 1.8837,
"step": 10
},
{
"epoch": 0.044556962025316456,
"grad_norm": 2.5938632551385385,
"learning_rate": 3.577235772357724e-06,
"loss": 1.8482,
"step": 11
},
{
"epoch": 0.04860759493670886,
"grad_norm": 4.461885160537773,
"learning_rate": 3.902439024390244e-06,
"loss": 1.8531,
"step": 12
},
{
"epoch": 0.052658227848101265,
"grad_norm": 4.307874551204792,
"learning_rate": 4.227642276422765e-06,
"loss": 1.867,
"step": 13
},
{
"epoch": 0.05670886075949367,
"grad_norm": 4.251403352719513,
"learning_rate": 4.552845528455285e-06,
"loss": 1.8472,
"step": 14
},
{
"epoch": 0.060759493670886074,
"grad_norm": 3.981630275189901,
"learning_rate": 4.8780487804878055e-06,
"loss": 1.862,
"step": 15
},
{
"epoch": 0.06481012658227848,
"grad_norm": 2.6913054641048504,
"learning_rate": 5.203252032520326e-06,
"loss": 1.8048,
"step": 16
},
{
"epoch": 0.06886075949367089,
"grad_norm": 2.389851642755812,
"learning_rate": 5.528455284552846e-06,
"loss": 1.7984,
"step": 17
},
{
"epoch": 0.0729113924050633,
"grad_norm": 2.062460554966612,
"learning_rate": 5.853658536585366e-06,
"loss": 1.7918,
"step": 18
},
{
"epoch": 0.0769620253164557,
"grad_norm": 1.7133677678107693,
"learning_rate": 6.178861788617887e-06,
"loss": 1.7315,
"step": 19
},
{
"epoch": 0.0810126582278481,
"grad_norm": 1.2552880305904963,
"learning_rate": 6.504065040650407e-06,
"loss": 1.7161,
"step": 20
},
{
"epoch": 0.08506329113924051,
"grad_norm": 1.31577278981272,
"learning_rate": 6.829268292682928e-06,
"loss": 1.7287,
"step": 21
},
{
"epoch": 0.08911392405063291,
"grad_norm": 1.54471514198393,
"learning_rate": 7.154471544715448e-06,
"loss": 1.7381,
"step": 22
},
{
"epoch": 0.09316455696202532,
"grad_norm": 1.5118056250819454,
"learning_rate": 7.4796747967479676e-06,
"loss": 1.7433,
"step": 23
},
{
"epoch": 0.09721518987341772,
"grad_norm": 1.298704775338139,
"learning_rate": 7.804878048780489e-06,
"loss": 1.6982,
"step": 24
},
{
"epoch": 0.10126582278481013,
"grad_norm": 1.1762640749687219,
"learning_rate": 8.130081300813009e-06,
"loss": 1.729,
"step": 25
},
{
"epoch": 0.10531645569620253,
"grad_norm": 1.1702375647654428,
"learning_rate": 8.45528455284553e-06,
"loss": 1.7218,
"step": 26
},
{
"epoch": 0.10936708860759493,
"grad_norm": 1.3089658140660163,
"learning_rate": 8.78048780487805e-06,
"loss": 1.7053,
"step": 27
},
{
"epoch": 0.11341772151898734,
"grad_norm": 1.3649993238953593,
"learning_rate": 9.10569105691057e-06,
"loss": 1.6359,
"step": 28
},
{
"epoch": 0.11746835443037974,
"grad_norm": 1.168939657303398,
"learning_rate": 9.43089430894309e-06,
"loss": 1.6659,
"step": 29
},
{
"epoch": 0.12151898734177215,
"grad_norm": 1.0166253206261076,
"learning_rate": 9.756097560975611e-06,
"loss": 1.7124,
"step": 30
},
{
"epoch": 0.12556962025316457,
"grad_norm": 0.9928008393474003,
"learning_rate": 1.008130081300813e-05,
"loss": 1.6721,
"step": 31
},
{
"epoch": 0.12962025316455697,
"grad_norm": 1.0763632880402774,
"learning_rate": 1.0406504065040652e-05,
"loss": 1.6238,
"step": 32
},
{
"epoch": 0.13367088607594937,
"grad_norm": 0.9507407797334282,
"learning_rate": 1.0731707317073172e-05,
"loss": 1.6648,
"step": 33
},
{
"epoch": 0.13772151898734178,
"grad_norm": 0.8454844482173011,
"learning_rate": 1.1056910569105692e-05,
"loss": 1.6548,
"step": 34
},
{
"epoch": 0.14177215189873418,
"grad_norm": 0.8875109172801641,
"learning_rate": 1.1382113821138213e-05,
"loss": 1.5952,
"step": 35
},
{
"epoch": 0.1458227848101266,
"grad_norm": 0.9170884660128061,
"learning_rate": 1.1707317073170731e-05,
"loss": 1.6896,
"step": 36
},
{
"epoch": 0.149873417721519,
"grad_norm": 0.9080996638249322,
"learning_rate": 1.2032520325203254e-05,
"loss": 1.6404,
"step": 37
},
{
"epoch": 0.1539240506329114,
"grad_norm": 0.7876089230137557,
"learning_rate": 1.2357723577235774e-05,
"loss": 1.6752,
"step": 38
},
{
"epoch": 0.1579746835443038,
"grad_norm": 0.902274585590205,
"learning_rate": 1.2682926829268294e-05,
"loss": 1.7078,
"step": 39
},
{
"epoch": 0.1620253164556962,
"grad_norm": 1.0305812589647725,
"learning_rate": 1.3008130081300815e-05,
"loss": 1.6619,
"step": 40
},
{
"epoch": 0.1660759493670886,
"grad_norm": 0.7808685670625216,
"learning_rate": 1.3333333333333333e-05,
"loss": 1.6484,
"step": 41
},
{
"epoch": 0.17012658227848101,
"grad_norm": 0.9779080728629154,
"learning_rate": 1.3658536585365855e-05,
"loss": 1.6185,
"step": 42
},
{
"epoch": 0.17417721518987342,
"grad_norm": 0.9985506394085919,
"learning_rate": 1.3983739837398376e-05,
"loss": 1.6635,
"step": 43
},
{
"epoch": 0.17822784810126582,
"grad_norm": 0.8531800069091916,
"learning_rate": 1.4308943089430896e-05,
"loss": 1.5966,
"step": 44
},
{
"epoch": 0.18227848101265823,
"grad_norm": 1.0052999686723205,
"learning_rate": 1.4634146341463415e-05,
"loss": 1.6052,
"step": 45
},
{
"epoch": 0.18632911392405063,
"grad_norm": 1.161717126123582,
"learning_rate": 1.4959349593495935e-05,
"loss": 1.6122,
"step": 46
},
{
"epoch": 0.19037974683544304,
"grad_norm": 0.8200420184710876,
"learning_rate": 1.528455284552846e-05,
"loss": 1.646,
"step": 47
},
{
"epoch": 0.19443037974683544,
"grad_norm": 0.8849959719190451,
"learning_rate": 1.5609756097560978e-05,
"loss": 1.6191,
"step": 48
},
{
"epoch": 0.19848101265822785,
"grad_norm": 0.8720684555816829,
"learning_rate": 1.5934959349593496e-05,
"loss": 1.6421,
"step": 49
},
{
"epoch": 0.20253164556962025,
"grad_norm": 0.8203732280508076,
"learning_rate": 1.6260162601626018e-05,
"loss": 1.5899,
"step": 50
},
{
"epoch": 0.20658227848101265,
"grad_norm": 1.0605813419983763,
"learning_rate": 1.6585365853658537e-05,
"loss": 1.6163,
"step": 51
},
{
"epoch": 0.21063291139240506,
"grad_norm": 0.7576271631450012,
"learning_rate": 1.691056910569106e-05,
"loss": 1.6319,
"step": 52
},
{
"epoch": 0.21468354430379746,
"grad_norm": 0.9388841105062257,
"learning_rate": 1.7235772357723578e-05,
"loss": 1.5778,
"step": 53
},
{
"epoch": 0.21873417721518987,
"grad_norm": 0.7683629896800243,
"learning_rate": 1.75609756097561e-05,
"loss": 1.6393,
"step": 54
},
{
"epoch": 0.22278481012658227,
"grad_norm": 0.7871519087076295,
"learning_rate": 1.788617886178862e-05,
"loss": 1.6554,
"step": 55
},
{
"epoch": 0.22683544303797468,
"grad_norm": 0.7314534427246892,
"learning_rate": 1.821138211382114e-05,
"loss": 1.612,
"step": 56
},
{
"epoch": 0.23088607594936708,
"grad_norm": 0.8014108375014888,
"learning_rate": 1.8536585365853663e-05,
"loss": 1.6456,
"step": 57
},
{
"epoch": 0.23493670886075949,
"grad_norm": 0.7431732897991109,
"learning_rate": 1.886178861788618e-05,
"loss": 1.6012,
"step": 58
},
{
"epoch": 0.2389873417721519,
"grad_norm": 0.8130493401887201,
"learning_rate": 1.91869918699187e-05,
"loss": 1.5887,
"step": 59
},
{
"epoch": 0.2430379746835443,
"grad_norm": 0.9158315508300362,
"learning_rate": 1.9512195121951222e-05,
"loss": 1.6151,
"step": 60
},
{
"epoch": 0.2470886075949367,
"grad_norm": 0.7930105549765682,
"learning_rate": 1.983739837398374e-05,
"loss": 1.6052,
"step": 61
},
{
"epoch": 0.25113924050632913,
"grad_norm": 0.7199672585716881,
"learning_rate": 2.016260162601626e-05,
"loss": 1.553,
"step": 62
},
{
"epoch": 0.25518987341772154,
"grad_norm": 0.7866737187751454,
"learning_rate": 2.048780487804878e-05,
"loss": 1.6353,
"step": 63
},
{
"epoch": 0.25924050632911394,
"grad_norm": 0.6327380873432442,
"learning_rate": 2.0813008130081303e-05,
"loss": 1.5779,
"step": 64
},
{
"epoch": 0.26329113924050634,
"grad_norm": 0.6943783535313671,
"learning_rate": 2.1138211382113822e-05,
"loss": 1.6249,
"step": 65
},
{
"epoch": 0.26734177215189875,
"grad_norm": 0.8824211610808923,
"learning_rate": 2.1463414634146344e-05,
"loss": 1.5978,
"step": 66
},
{
"epoch": 0.27139240506329115,
"grad_norm": 0.8298400737835392,
"learning_rate": 2.1788617886178863e-05,
"loss": 1.631,
"step": 67
},
{
"epoch": 0.27544303797468356,
"grad_norm": 0.8977931300693257,
"learning_rate": 2.2113821138211385e-05,
"loss": 1.5866,
"step": 68
},
{
"epoch": 0.27949367088607596,
"grad_norm": 0.8278143907690333,
"learning_rate": 2.2439024390243907e-05,
"loss": 1.6047,
"step": 69
},
{
"epoch": 0.28354430379746837,
"grad_norm": 0.7367869172327068,
"learning_rate": 2.2764227642276426e-05,
"loss": 1.6547,
"step": 70
},
{
"epoch": 0.28759493670886077,
"grad_norm": 0.8241383153406088,
"learning_rate": 2.3089430894308948e-05,
"loss": 1.6349,
"step": 71
},
{
"epoch": 0.2916455696202532,
"grad_norm": 0.7088875053423881,
"learning_rate": 2.3414634146341463e-05,
"loss": 1.6258,
"step": 72
},
{
"epoch": 0.2956962025316456,
"grad_norm": 0.6276887757574976,
"learning_rate": 2.3739837398373985e-05,
"loss": 1.5816,
"step": 73
},
{
"epoch": 0.299746835443038,
"grad_norm": 0.8056067152733327,
"learning_rate": 2.4065040650406507e-05,
"loss": 1.538,
"step": 74
},
{
"epoch": 0.3037974683544304,
"grad_norm": 0.7203557277218589,
"learning_rate": 2.4390243902439026e-05,
"loss": 1.6259,
"step": 75
},
{
"epoch": 0.3078481012658228,
"grad_norm": 0.7789694000596235,
"learning_rate": 2.4715447154471548e-05,
"loss": 1.6563,
"step": 76
},
{
"epoch": 0.3118987341772152,
"grad_norm": 0.92186502761971,
"learning_rate": 2.5040650406504066e-05,
"loss": 1.5768,
"step": 77
},
{
"epoch": 0.3159493670886076,
"grad_norm": 0.8260442936753352,
"learning_rate": 2.536585365853659e-05,
"loss": 1.6252,
"step": 78
},
{
"epoch": 0.32,
"grad_norm": 0.9315838831041934,
"learning_rate": 2.569105691056911e-05,
"loss": 1.6378,
"step": 79
},
{
"epoch": 0.3240506329113924,
"grad_norm": 1.002787649570172,
"learning_rate": 2.601626016260163e-05,
"loss": 1.6094,
"step": 80
},
{
"epoch": 0.3281012658227848,
"grad_norm": 0.8765188165180666,
"learning_rate": 2.634146341463415e-05,
"loss": 1.6234,
"step": 81
},
{
"epoch": 0.3321518987341772,
"grad_norm": 1.0330609769982448,
"learning_rate": 2.6666666666666667e-05,
"loss": 1.5921,
"step": 82
},
{
"epoch": 0.3362025316455696,
"grad_norm": 0.9218441307054872,
"learning_rate": 2.699186991869919e-05,
"loss": 1.5863,
"step": 83
},
{
"epoch": 0.34025316455696203,
"grad_norm": 1.5237435928219774,
"learning_rate": 2.731707317073171e-05,
"loss": 1.6111,
"step": 84
},
{
"epoch": 0.34430379746835443,
"grad_norm": 1.1819183706830643,
"learning_rate": 2.764227642276423e-05,
"loss": 1.6162,
"step": 85
},
{
"epoch": 0.34835443037974684,
"grad_norm": 0.9066750451577599,
"learning_rate": 2.796747967479675e-05,
"loss": 1.5831,
"step": 86
},
{
"epoch": 0.35240506329113924,
"grad_norm": 1.8507934085983357,
"learning_rate": 2.829268292682927e-05,
"loss": 1.5824,
"step": 87
},
{
"epoch": 0.35645569620253165,
"grad_norm": 0.942460091869486,
"learning_rate": 2.8617886178861792e-05,
"loss": 1.5984,
"step": 88
},
{
"epoch": 0.36050632911392405,
"grad_norm": 1.482038628492743,
"learning_rate": 2.8943089430894314e-05,
"loss": 1.6142,
"step": 89
},
{
"epoch": 0.36455696202531646,
"grad_norm": 1.1904998380228067,
"learning_rate": 2.926829268292683e-05,
"loss": 1.6198,
"step": 90
},
{
"epoch": 0.36860759493670886,
"grad_norm": 1.2293327085459305,
"learning_rate": 2.959349593495935e-05,
"loss": 1.6268,
"step": 91
},
{
"epoch": 0.37265822784810126,
"grad_norm": 0.8448684793933755,
"learning_rate": 2.991869918699187e-05,
"loss": 1.5651,
"step": 92
},
{
"epoch": 0.37670886075949367,
"grad_norm": 0.7806371393454177,
"learning_rate": 3.0243902439024392e-05,
"loss": 1.6143,
"step": 93
},
{
"epoch": 0.3807594936708861,
"grad_norm": 0.7952571866396014,
"learning_rate": 3.056910569105692e-05,
"loss": 1.6172,
"step": 94
},
{
"epoch": 0.3848101265822785,
"grad_norm": 0.8174236379684373,
"learning_rate": 3.089430894308943e-05,
"loss": 1.5563,
"step": 95
},
{
"epoch": 0.3888607594936709,
"grad_norm": 1.2847356687407787,
"learning_rate": 3.1219512195121955e-05,
"loss": 1.5974,
"step": 96
},
{
"epoch": 0.3929113924050633,
"grad_norm": 1.5043708857784812,
"learning_rate": 3.154471544715447e-05,
"loss": 1.6028,
"step": 97
},
{
"epoch": 0.3969620253164557,
"grad_norm": 1.231527885416442,
"learning_rate": 3.186991869918699e-05,
"loss": 1.5829,
"step": 98
},
{
"epoch": 0.4010126582278481,
"grad_norm": 1.2510484356360152,
"learning_rate": 3.2195121951219514e-05,
"loss": 1.6125,
"step": 99
},
{
"epoch": 0.4050632911392405,
"grad_norm": 1.831315583002477,
"learning_rate": 3.2520325203252037e-05,
"loss": 1.5524,
"step": 100
},
{
"epoch": 0.4091139240506329,
"grad_norm": 1.0220238484728084,
"learning_rate": 3.284552845528456e-05,
"loss": 1.5993,
"step": 101
},
{
"epoch": 0.4131645569620253,
"grad_norm": 1.261718682807198,
"learning_rate": 3.3170731707317074e-05,
"loss": 1.595,
"step": 102
},
{
"epoch": 0.4172151898734177,
"grad_norm": 1.5474639851346301,
"learning_rate": 3.3495934959349596e-05,
"loss": 1.5947,
"step": 103
},
{
"epoch": 0.4212658227848101,
"grad_norm": 0.8494176478462546,
"learning_rate": 3.382113821138212e-05,
"loss": 1.6148,
"step": 104
},
{
"epoch": 0.4253164556962025,
"grad_norm": 1.506890457873893,
"learning_rate": 3.414634146341463e-05,
"loss": 1.6119,
"step": 105
},
{
"epoch": 0.4293670886075949,
"grad_norm": 1.4888101972324823,
"learning_rate": 3.4471544715447155e-05,
"loss": 1.5778,
"step": 106
},
{
"epoch": 0.43341772151898733,
"grad_norm": 1.1982258075896894,
"learning_rate": 3.479674796747968e-05,
"loss": 1.6129,
"step": 107
},
{
"epoch": 0.43746835443037974,
"grad_norm": 1.9885907749884228,
"learning_rate": 3.51219512195122e-05,
"loss": 1.5872,
"step": 108
},
{
"epoch": 0.44151898734177214,
"grad_norm": 1.2663317071872546,
"learning_rate": 3.544715447154472e-05,
"loss": 1.5838,
"step": 109
},
{
"epoch": 0.44556962025316454,
"grad_norm": 1.8814028532254368,
"learning_rate": 3.577235772357724e-05,
"loss": 1.599,
"step": 110
},
{
"epoch": 0.44962025316455695,
"grad_norm": 1.256754306930896,
"learning_rate": 3.609756097560976e-05,
"loss": 1.6471,
"step": 111
},
{
"epoch": 0.45367088607594935,
"grad_norm": 2.072122235697746,
"learning_rate": 3.642276422764228e-05,
"loss": 1.5838,
"step": 112
},
{
"epoch": 0.45772151898734176,
"grad_norm": 1.4265741800910663,
"learning_rate": 3.67479674796748e-05,
"loss": 1.591,
"step": 113
},
{
"epoch": 0.46177215189873416,
"grad_norm": 1.722896990240273,
"learning_rate": 3.7073170731707325e-05,
"loss": 1.5874,
"step": 114
},
{
"epoch": 0.46582278481012657,
"grad_norm": 1.5147875655719207,
"learning_rate": 3.739837398373984e-05,
"loss": 1.6329,
"step": 115
},
{
"epoch": 0.46987341772151897,
"grad_norm": 2.0448494387378418,
"learning_rate": 3.772357723577236e-05,
"loss": 1.6307,
"step": 116
},
{
"epoch": 0.4739240506329114,
"grad_norm": 1.9003995375239326,
"learning_rate": 3.804878048780488e-05,
"loss": 1.6209,
"step": 117
},
{
"epoch": 0.4779746835443038,
"grad_norm": 1.7224453409265608,
"learning_rate": 3.83739837398374e-05,
"loss": 1.5795,
"step": 118
},
{
"epoch": 0.4820253164556962,
"grad_norm": 1.507905685449079,
"learning_rate": 3.869918699186992e-05,
"loss": 1.6152,
"step": 119
},
{
"epoch": 0.4860759493670886,
"grad_norm": 1.928881181984129,
"learning_rate": 3.9024390243902444e-05,
"loss": 1.5406,
"step": 120
},
{
"epoch": 0.490126582278481,
"grad_norm": 1.6006291980220415,
"learning_rate": 3.9349593495934966e-05,
"loss": 1.5796,
"step": 121
},
{
"epoch": 0.4941772151898734,
"grad_norm": 2.0741034734031847,
"learning_rate": 3.967479674796748e-05,
"loss": 1.6065,
"step": 122
},
{
"epoch": 0.4982278481012658,
"grad_norm": 2.043822426867333,
"learning_rate": 4e-05,
"loss": 1.5538,
"step": 123
},
{
"epoch": 0.5022784810126583,
"grad_norm": 1.3713830846769242,
"learning_rate": 3.999991946137476e-05,
"loss": 1.5839,
"step": 124
},
{
"epoch": 0.5063291139240507,
"grad_norm": 1.7589331838759688,
"learning_rate": 3.999967784614766e-05,
"loss": 1.57,
"step": 125
},
{
"epoch": 0.5103797468354431,
"grad_norm": 0.852556816647701,
"learning_rate": 3.9999275156264656e-05,
"loss": 1.591,
"step": 126
},
{
"epoch": 0.5144303797468355,
"grad_norm": 1.7720016439988622,
"learning_rate": 3.999871139496895e-05,
"loss": 1.6176,
"step": 127
},
{
"epoch": 0.5184810126582279,
"grad_norm": 1.4248896903131925,
"learning_rate": 3.9997986566800995e-05,
"loss": 1.5998,
"step": 128
},
{
"epoch": 0.5225316455696203,
"grad_norm": 1.4494065974829493,
"learning_rate": 3.999710067759846e-05,
"loss": 1.5879,
"step": 129
},
{
"epoch": 0.5265822784810127,
"grad_norm": 1.6384658380827737,
"learning_rate": 3.999605373449617e-05,
"loss": 1.568,
"step": 130
},
{
"epoch": 0.5306329113924051,
"grad_norm": 0.9475749576549228,
"learning_rate": 3.9994845745926075e-05,
"loss": 1.5807,
"step": 131
},
{
"epoch": 0.5346835443037975,
"grad_norm": 1.371590248071306,
"learning_rate": 3.999347672161713e-05,
"loss": 1.5761,
"step": 132
},
{
"epoch": 0.5387341772151899,
"grad_norm": 1.2740051492809579,
"learning_rate": 3.999194667259528e-05,
"loss": 1.5922,
"step": 133
},
{
"epoch": 0.5427848101265823,
"grad_norm": 1.4767463059710102,
"learning_rate": 3.999025561118334e-05,
"loss": 1.6104,
"step": 134
},
{
"epoch": 0.5468354430379747,
"grad_norm": 1.1587675393966255,
"learning_rate": 3.998840355100086e-05,
"loss": 1.5368,
"step": 135
},
{
"epoch": 0.5508860759493671,
"grad_norm": 1.3034471000576926,
"learning_rate": 3.998639050696409e-05,
"loss": 1.5669,
"step": 136
},
{
"epoch": 0.5549367088607595,
"grad_norm": 1.144549909316193,
"learning_rate": 3.998421649528582e-05,
"loss": 1.59,
"step": 137
},
{
"epoch": 0.5589873417721519,
"grad_norm": 1.2140765344592768,
"learning_rate": 3.9981881533475234e-05,
"loss": 1.6011,
"step": 138
},
{
"epoch": 0.5630379746835443,
"grad_norm": 1.4641177194965138,
"learning_rate": 3.997938564033779e-05,
"loss": 1.5912,
"step": 139
},
{
"epoch": 0.5670886075949367,
"grad_norm": 0.9463843307113535,
"learning_rate": 3.9976728835975064e-05,
"loss": 1.5991,
"step": 140
},
{
"epoch": 0.5711392405063291,
"grad_norm": 0.9150445295623888,
"learning_rate": 3.9973911141784605e-05,
"loss": 1.5624,
"step": 141
},
{
"epoch": 0.5751898734177215,
"grad_norm": 1.1602450829890467,
"learning_rate": 3.997093258045973e-05,
"loss": 1.6069,
"step": 142
},
{
"epoch": 0.579240506329114,
"grad_norm": 0.930611412613755,
"learning_rate": 3.996779317598936e-05,
"loss": 1.6192,
"step": 143
},
{
"epoch": 0.5832911392405064,
"grad_norm": 1.2102529607004326,
"learning_rate": 3.996449295365782e-05,
"loss": 1.6006,
"step": 144
},
{
"epoch": 0.5873417721518988,
"grad_norm": 1.291265402745687,
"learning_rate": 3.996103194004467e-05,
"loss": 1.5373,
"step": 145
},
{
"epoch": 0.5913924050632912,
"grad_norm": 0.9863710302216822,
"learning_rate": 3.995741016302441e-05,
"loss": 1.6103,
"step": 146
},
{
"epoch": 0.5954430379746836,
"grad_norm": 1.5473006369019264,
"learning_rate": 3.9953627651766364e-05,
"loss": 1.5693,
"step": 147
},
{
"epoch": 0.599493670886076,
"grad_norm": 1.0361281770658093,
"learning_rate": 3.9949684436734325e-05,
"loss": 1.6195,
"step": 148
},
{
"epoch": 0.6035443037974684,
"grad_norm": 1.606757235635811,
"learning_rate": 3.994558054968643e-05,
"loss": 1.5899,
"step": 149
},
{
"epoch": 0.6075949367088608,
"grad_norm": 1.0357566989565048,
"learning_rate": 3.994131602367481e-05,
"loss": 1.6078,
"step": 150
},
{
"epoch": 0.6116455696202532,
"grad_norm": 1.7719873906197807,
"learning_rate": 3.9936890893045376e-05,
"loss": 1.5389,
"step": 151
},
{
"epoch": 0.6156962025316456,
"grad_norm": 1.4561226045515816,
"learning_rate": 3.993230519343752e-05,
"loss": 1.568,
"step": 152
},
{
"epoch": 0.619746835443038,
"grad_norm": 1.3236955653162061,
"learning_rate": 3.992755896178383e-05,
"loss": 1.5881,
"step": 153
},
{
"epoch": 0.6237974683544304,
"grad_norm": 1.1950674923016316,
"learning_rate": 3.992265223630981e-05,
"loss": 1.6284,
"step": 154
},
{
"epoch": 0.6278481012658228,
"grad_norm": 1.4555070838895614,
"learning_rate": 3.991758505653355e-05,
"loss": 1.6078,
"step": 155
},
{
"epoch": 0.6318987341772152,
"grad_norm": 1.1437448662575191,
"learning_rate": 3.991235746326543e-05,
"loss": 1.602,
"step": 156
},
{
"epoch": 0.6359493670886076,
"grad_norm": 1.1460308283228857,
"learning_rate": 3.9906969498607745e-05,
"loss": 1.5544,
"step": 157
},
{
"epoch": 0.64,
"grad_norm": 1.1944970685440386,
"learning_rate": 3.990142120595444e-05,
"loss": 1.5552,
"step": 158
},
{
"epoch": 0.6440506329113924,
"grad_norm": 1.5102473707630062,
"learning_rate": 3.98957126299907e-05,
"loss": 1.5653,
"step": 159
},
{
"epoch": 0.6481012658227848,
"grad_norm": 1.149891104159494,
"learning_rate": 3.9889843816692596e-05,
"loss": 1.6216,
"step": 160
},
{
"epoch": 0.6521518987341772,
"grad_norm": 1.4922729128533523,
"learning_rate": 3.9883814813326766e-05,
"loss": 1.5875,
"step": 161
},
{
"epoch": 0.6562025316455696,
"grad_norm": 1.604546433339899,
"learning_rate": 3.9877625668449956e-05,
"loss": 1.6111,
"step": 162
},
{
"epoch": 0.660253164556962,
"grad_norm": 1.1351739270464305,
"learning_rate": 3.98712764319087e-05,
"loss": 1.6198,
"step": 163
},
{
"epoch": 0.6643037974683544,
"grad_norm": 2.159223841819358,
"learning_rate": 3.9864767154838864e-05,
"loss": 1.5918,
"step": 164
},
{
"epoch": 0.6683544303797468,
"grad_norm": 1.7491675906876265,
"learning_rate": 3.9858097889665277e-05,
"loss": 1.5849,
"step": 165
},
{
"epoch": 0.6724050632911392,
"grad_norm": 1.9418161428063012,
"learning_rate": 3.985126869010129e-05,
"loss": 1.6148,
"step": 166
},
{
"epoch": 0.6764556962025317,
"grad_norm": 1.3450862232226632,
"learning_rate": 3.984427961114833e-05,
"loss": 1.5824,
"step": 167
},
{
"epoch": 0.6805063291139241,
"grad_norm": 2.059158241855422,
"learning_rate": 3.9837130709095475e-05,
"loss": 1.5851,
"step": 168
},
{
"epoch": 0.6845569620253165,
"grad_norm": 1.3409360424111665,
"learning_rate": 3.982982204151901e-05,
"loss": 1.5661,
"step": 169
},
{
"epoch": 0.6886075949367089,
"grad_norm": 2.537298008754443,
"learning_rate": 3.982235366728193e-05,
"loss": 1.5818,
"step": 170
},
{
"epoch": 0.6926582278481013,
"grad_norm": 2.05813011561159,
"learning_rate": 3.9814725646533505e-05,
"loss": 1.5978,
"step": 171
},
{
"epoch": 0.6967088607594937,
"grad_norm": 1.7093823809919668,
"learning_rate": 3.9806938040708746e-05,
"loss": 1.606,
"step": 172
},
{
"epoch": 0.7007594936708861,
"grad_norm": 1.4014891835401502,
"learning_rate": 3.9798990912527976e-05,
"loss": 1.5901,
"step": 173
},
{
"epoch": 0.7048101265822785,
"grad_norm": 2.3299372277814205,
"learning_rate": 3.979088432599627e-05,
"loss": 1.5788,
"step": 174
},
{
"epoch": 0.7088607594936709,
"grad_norm": 1.7759156796366395,
"learning_rate": 3.9782618346402964e-05,
"loss": 1.5161,
"step": 175
},
{
"epoch": 0.7129113924050633,
"grad_norm": 2.2463318403068095,
"learning_rate": 3.977419304032111e-05,
"loss": 1.5736,
"step": 176
},
{
"epoch": 0.7169620253164557,
"grad_norm": 2.3758263984499415,
"learning_rate": 3.976560847560697e-05,
"loss": 1.5394,
"step": 177
},
{
"epoch": 0.7210126582278481,
"grad_norm": 1.3562792032993587,
"learning_rate": 3.9756864721399456e-05,
"loss": 1.6076,
"step": 178
},
{
"epoch": 0.7250632911392405,
"grad_norm": 1.8302645577150987,
"learning_rate": 3.974796184811956e-05,
"loss": 1.581,
"step": 179
},
{
"epoch": 0.7291139240506329,
"grad_norm": 1.2092387612036046,
"learning_rate": 3.973889992746979e-05,
"loss": 1.5802,
"step": 180
},
{
"epoch": 0.7331645569620253,
"grad_norm": 1.0758481694654696,
"learning_rate": 3.972967903243361e-05,
"loss": 1.5728,
"step": 181
},
{
"epoch": 0.7372151898734177,
"grad_norm": 1.497653985590649,
"learning_rate": 3.972029923727486e-05,
"loss": 1.5407,
"step": 182
},
{
"epoch": 0.7412658227848101,
"grad_norm": 0.808715800043573,
"learning_rate": 3.971076061753709e-05,
"loss": 1.5946,
"step": 183
},
{
"epoch": 0.7453164556962025,
"grad_norm": 1.6047938818401482,
"learning_rate": 3.9701063250043066e-05,
"loss": 1.5752,
"step": 184
},
{
"epoch": 0.7493670886075949,
"grad_norm": 0.9956070747173383,
"learning_rate": 3.969120721289402e-05,
"loss": 1.5782,
"step": 185
},
{
"epoch": 0.7534177215189873,
"grad_norm": 1.7343952261747875,
"learning_rate": 3.9681192585469146e-05,
"loss": 1.5753,
"step": 186
},
{
"epoch": 0.7574683544303797,
"grad_norm": 1.1797985020740651,
"learning_rate": 3.9671019448424865e-05,
"loss": 1.5803,
"step": 187
},
{
"epoch": 0.7615189873417721,
"grad_norm": 2.007230297474155,
"learning_rate": 3.966068788369422e-05,
"loss": 1.5974,
"step": 188
},
{
"epoch": 0.7655696202531646,
"grad_norm": 1.2270846225406717,
"learning_rate": 3.965019797448622e-05,
"loss": 1.5789,
"step": 189
},
{
"epoch": 0.769620253164557,
"grad_norm": 2.188132759874859,
"learning_rate": 3.963954980528515e-05,
"loss": 1.6077,
"step": 190
},
{
"epoch": 0.7736708860759494,
"grad_norm": 1.7323230901482374,
"learning_rate": 3.9628743461849905e-05,
"loss": 1.611,
"step": 191
},
{
"epoch": 0.7777215189873418,
"grad_norm": 1.7477178734391814,
"learning_rate": 3.961777903121329e-05,
"loss": 1.5692,
"step": 192
},
{
"epoch": 0.7817721518987342,
"grad_norm": 1.6302264686653782,
"learning_rate": 3.960665660168131e-05,
"loss": 1.5712,
"step": 193
},
{
"epoch": 0.7858227848101266,
"grad_norm": 1.6154039152444681,
"learning_rate": 3.9595376262832485e-05,
"loss": 1.5899,
"step": 194
},
{
"epoch": 0.789873417721519,
"grad_norm": 1.359222884064358,
"learning_rate": 3.9583938105517127e-05,
"loss": 1.5108,
"step": 195
},
{
"epoch": 0.7939240506329114,
"grad_norm": 1.479698748648423,
"learning_rate": 3.957234222185657e-05,
"loss": 1.561,
"step": 196
},
{
"epoch": 0.7979746835443038,
"grad_norm": 1.1884673335323912,
"learning_rate": 3.9560588705242474e-05,
"loss": 1.5902,
"step": 197
},
{
"epoch": 0.8020253164556962,
"grad_norm": 1.4925166936435508,
"learning_rate": 3.954867765033605e-05,
"loss": 1.573,
"step": 198
},
{
"epoch": 0.8060759493670886,
"grad_norm": 1.1716671908121525,
"learning_rate": 3.953660915306728e-05,
"loss": 1.6285,
"step": 199
},
{
"epoch": 0.810126582278481,
"grad_norm": 1.668078348813483,
"learning_rate": 3.952438331063419e-05,
"loss": 1.5884,
"step": 200
},
{
"epoch": 0.8141772151898734,
"grad_norm": 1.2212038713550868,
"learning_rate": 3.951200022150205e-05,
"loss": 1.5983,
"step": 201
},
{
"epoch": 0.8182278481012658,
"grad_norm": 1.747826343668385,
"learning_rate": 3.949945998540253e-05,
"loss": 1.5999,
"step": 202
},
{
"epoch": 0.8222784810126582,
"grad_norm": 1.4466206351462139,
"learning_rate": 3.9486762703332993e-05,
"loss": 1.5653,
"step": 203
},
{
"epoch": 0.8263291139240506,
"grad_norm": 1.6694343999848376,
"learning_rate": 3.947390847755559e-05,
"loss": 1.5855,
"step": 204
},
{
"epoch": 0.830379746835443,
"grad_norm": 1.6538519174272073,
"learning_rate": 3.946089741159648e-05,
"loss": 1.5656,
"step": 205
},
{
"epoch": 0.8344303797468354,
"grad_norm": 1.021458496880297,
"learning_rate": 3.944772961024501e-05,
"loss": 1.5732,
"step": 206
},
{
"epoch": 0.8384810126582278,
"grad_norm": 1.0221397258971954,
"learning_rate": 3.943440517955285e-05,
"loss": 1.5552,
"step": 207
},
{
"epoch": 0.8425316455696202,
"grad_norm": 0.8841437897029805,
"learning_rate": 3.9420924226833126e-05,
"loss": 1.5845,
"step": 208
},
{
"epoch": 0.8465822784810126,
"grad_norm": 0.8349089254097803,
"learning_rate": 3.9407286860659566e-05,
"loss": 1.5508,
"step": 209
},
{
"epoch": 0.850632911392405,
"grad_norm": 0.8888150952749206,
"learning_rate": 3.9393493190865657e-05,
"loss": 1.5659,
"step": 210
},
{
"epoch": 0.8546835443037974,
"grad_norm": 0.8258668401791126,
"learning_rate": 3.937954332854371e-05,
"loss": 1.6155,
"step": 211
},
{
"epoch": 0.8587341772151899,
"grad_norm": 0.9287592566995861,
"learning_rate": 3.9365437386044016e-05,
"loss": 1.5361,
"step": 212
},
{
"epoch": 0.8627848101265823,
"grad_norm": 1.3421707560865352,
"learning_rate": 3.935117547697387e-05,
"loss": 1.612,
"step": 213
},
{
"epoch": 0.8668354430379747,
"grad_norm": 0.7638889961734745,
"learning_rate": 3.933675771619675e-05,
"loss": 1.6064,
"step": 214
},
{
"epoch": 0.8708860759493671,
"grad_norm": 1.4055734439634364,
"learning_rate": 3.932218421983131e-05,
"loss": 1.5818,
"step": 215
},
{
"epoch": 0.8749367088607595,
"grad_norm": 1.0170290849425898,
"learning_rate": 3.9307455105250484e-05,
"loss": 1.6152,
"step": 216
},
{
"epoch": 0.8789873417721519,
"grad_norm": 1.4298208022901506,
"learning_rate": 3.929257049108054e-05,
"loss": 1.6049,
"step": 217
},
{
"epoch": 0.8830379746835443,
"grad_norm": 1.2905088137253589,
"learning_rate": 3.927753049720011e-05,
"loss": 1.566,
"step": 218
},
{
"epoch": 0.8870886075949367,
"grad_norm": 1.1230522496040476,
"learning_rate": 3.9262335244739234e-05,
"loss": 1.6444,
"step": 219
},
{
"epoch": 0.8911392405063291,
"grad_norm": 1.0633429558020715,
"learning_rate": 3.92469848560784e-05,
"loss": 1.6075,
"step": 220
},
{
"epoch": 0.8951898734177215,
"grad_norm": 1.0526273603327927,
"learning_rate": 3.923147945484751e-05,
"loss": 1.5928,
"step": 221
},
{
"epoch": 0.8992405063291139,
"grad_norm": 0.7681797451181619,
"learning_rate": 3.9215819165924956e-05,
"loss": 1.5524,
"step": 222
},
{
"epoch": 0.9032911392405063,
"grad_norm": 1.13394709242423,
"learning_rate": 3.920000411543654e-05,
"loss": 1.5608,
"step": 223
},
{
"epoch": 0.9073417721518987,
"grad_norm": 0.6555315956109813,
"learning_rate": 3.9184034430754495e-05,
"loss": 1.5775,
"step": 224
},
{
"epoch": 0.9113924050632911,
"grad_norm": 1.272177420172778,
"learning_rate": 3.916791024049648e-05,
"loss": 1.5881,
"step": 225
},
{
"epoch": 0.9154430379746835,
"grad_norm": 0.71317039595053,
"learning_rate": 3.91516316745245e-05,
"loss": 1.5478,
"step": 226
},
{
"epoch": 0.9194936708860759,
"grad_norm": 1.2817667936148507,
"learning_rate": 3.913519886394389e-05,
"loss": 1.5641,
"step": 227
},
{
"epoch": 0.9235443037974683,
"grad_norm": 0.8324043548497349,
"learning_rate": 3.911861194110225e-05,
"loss": 1.6218,
"step": 228
},
{
"epoch": 0.9275949367088607,
"grad_norm": 0.8872789575504182,
"learning_rate": 3.910187103958837e-05,
"loss": 1.6018,
"step": 229
},
{
"epoch": 0.9316455696202531,
"grad_norm": 0.8096960357864582,
"learning_rate": 3.908497629423117e-05,
"loss": 1.5837,
"step": 230
},
{
"epoch": 0.9356962025316455,
"grad_norm": 0.8933864789968333,
"learning_rate": 3.9067927841098614e-05,
"loss": 1.5612,
"step": 231
},
{
"epoch": 0.9397468354430379,
"grad_norm": 0.8365690640706857,
"learning_rate": 3.9050725817496594e-05,
"loss": 1.6191,
"step": 232
},
{
"epoch": 0.9437974683544303,
"grad_norm": 0.8052206445071723,
"learning_rate": 3.9033370361967844e-05,
"loss": 1.5998,
"step": 233
},
{
"epoch": 0.9478481012658228,
"grad_norm": 0.8843198814122848,
"learning_rate": 3.901586161429081e-05,
"loss": 1.5865,
"step": 234
},
{
"epoch": 0.9518987341772152,
"grad_norm": 0.8176413321171877,
"learning_rate": 3.8998199715478545e-05,
"loss": 1.5678,
"step": 235
},
{
"epoch": 0.9559493670886076,
"grad_norm": 0.8389449201271935,
"learning_rate": 3.8980384807777564e-05,
"loss": 1.5673,
"step": 236
},
{
"epoch": 0.96,
"grad_norm": 0.7726546826806036,
"learning_rate": 3.896241703466667e-05,
"loss": 1.5869,
"step": 237
},
{
"epoch": 0.9640506329113924,
"grad_norm": 0.7473764148492749,
"learning_rate": 3.894429654085585e-05,
"loss": 1.5586,
"step": 238
},
{
"epoch": 0.9681012658227848,
"grad_norm": 0.831830969821002,
"learning_rate": 3.892602347228505e-05,
"loss": 1.5985,
"step": 239
},
{
"epoch": 0.9721518987341772,
"grad_norm": 0.7701004465922224,
"learning_rate": 3.890759797612307e-05,
"loss": 1.5944,
"step": 240
},
{
"epoch": 0.9762025316455696,
"grad_norm": 0.9153333354667467,
"learning_rate": 3.888902020076632e-05,
"loss": 1.5654,
"step": 241
},
{
"epoch": 0.980253164556962,
"grad_norm": 0.8698897068049768,
"learning_rate": 3.887029029583764e-05,
"loss": 1.546,
"step": 242
},
{
"epoch": 0.9843037974683544,
"grad_norm": 1.006804543475018,
"learning_rate": 3.8851408412185125e-05,
"loss": 1.5805,
"step": 243
},
{
"epoch": 0.9883544303797468,
"grad_norm": 0.7338620565329317,
"learning_rate": 3.8832374701880855e-05,
"loss": 1.6143,
"step": 244
},
{
"epoch": 0.9924050632911392,
"grad_norm": 0.8738249191028905,
"learning_rate": 3.881318931821972e-05,
"loss": 1.6058,
"step": 245
},
{
"epoch": 0.9964556962025316,
"grad_norm": 0.7713496513493373,
"learning_rate": 3.879385241571817e-05,
"loss": 1.5994,
"step": 246
},
{
"epoch": 1.000506329113924,
"grad_norm": 0.7971214887109749,
"learning_rate": 3.8774364150112955e-05,
"loss": 1.7212,
"step": 247
},
{
"epoch": 1.0045569620253165,
"grad_norm": 0.7575614419880381,
"learning_rate": 3.8754724678359884e-05,
"loss": 1.5468,
"step": 248
},
{
"epoch": 1.0086075949367088,
"grad_norm": 0.7625221031591854,
"learning_rate": 3.873493415863256e-05,
"loss": 1.4965,
"step": 249
},
{
"epoch": 1.0126582278481013,
"grad_norm": 0.740357589310458,
"learning_rate": 3.871499275032111e-05,
"loss": 1.5369,
"step": 250
},
{
"epoch": 1.0167088607594936,
"grad_norm": 0.8981497389265131,
"learning_rate": 3.869490061403091e-05,
"loss": 1.4559,
"step": 251
},
{
"epoch": 1.0207594936708861,
"grad_norm": 0.7545975444494873,
"learning_rate": 3.867465791158124e-05,
"loss": 1.5567,
"step": 252
},
{
"epoch": 1.0248101265822784,
"grad_norm": 0.9625328770271231,
"learning_rate": 3.865426480600407e-05,
"loss": 1.5231,
"step": 253
},
{
"epoch": 1.028860759493671,
"grad_norm": 0.919486346916154,
"learning_rate": 3.863372146154264e-05,
"loss": 1.5317,
"step": 254
},
{
"epoch": 1.0329113924050632,
"grad_norm": 0.7725619350118018,
"learning_rate": 3.861302804365024e-05,
"loss": 1.4929,
"step": 255
},
{
"epoch": 1.0369620253164558,
"grad_norm": 0.9495447240178623,
"learning_rate": 3.85921847189888e-05,
"loss": 1.4926,
"step": 256
},
{
"epoch": 1.041012658227848,
"grad_norm": 0.6880334630308214,
"learning_rate": 3.85711916554276e-05,
"loss": 1.4992,
"step": 257
},
{
"epoch": 1.0450632911392406,
"grad_norm": 0.8491174325270926,
"learning_rate": 3.85500490220419e-05,
"loss": 1.4804,
"step": 258
},
{
"epoch": 1.0491139240506329,
"grad_norm": 0.9002544799676502,
"learning_rate": 3.852875698911154e-05,
"loss": 1.5223,
"step": 259
},
{
"epoch": 1.0531645569620254,
"grad_norm": 0.8746116095426576,
"learning_rate": 3.850731572811963e-05,
"loss": 1.4964,
"step": 260
},
{
"epoch": 1.0572151898734177,
"grad_norm": 0.7631369211843546,
"learning_rate": 3.848572541175116e-05,
"loss": 1.503,
"step": 261
},
{
"epoch": 1.0612658227848102,
"grad_norm": 0.8338055531900532,
"learning_rate": 3.846398621389154e-05,
"loss": 1.5025,
"step": 262
},
{
"epoch": 1.0653164556962025,
"grad_norm": 0.834562999861234,
"learning_rate": 3.84420983096253e-05,
"loss": 1.5227,
"step": 263
},
{
"epoch": 1.069367088607595,
"grad_norm": 0.7886264325194775,
"learning_rate": 3.8420061875234606e-05,
"loss": 1.4999,
"step": 264
},
{
"epoch": 1.0734177215189873,
"grad_norm": 0.904504028141716,
"learning_rate": 3.839787708819787e-05,
"loss": 1.4741,
"step": 265
},
{
"epoch": 1.0774683544303798,
"grad_norm": 0.7719612733288665,
"learning_rate": 3.8375544127188325e-05,
"loss": 1.4958,
"step": 266
},
{
"epoch": 1.081518987341772,
"grad_norm": 0.7691921458884292,
"learning_rate": 3.8353063172072564e-05,
"loss": 1.5215,
"step": 267
},
{
"epoch": 1.0855696202531646,
"grad_norm": 0.7018232201542657,
"learning_rate": 3.8330434403909105e-05,
"loss": 1.5023,
"step": 268
},
{
"epoch": 1.089620253164557,
"grad_norm": 0.8643383376005677,
"learning_rate": 3.8307658004946934e-05,
"loss": 1.5479,
"step": 269
},
{
"epoch": 1.0936708860759494,
"grad_norm": 0.6612417005994365,
"learning_rate": 3.8284734158624046e-05,
"loss": 1.4954,
"step": 270
},
{
"epoch": 1.0977215189873417,
"grad_norm": 0.6639305159139212,
"learning_rate": 3.826166304956594e-05,
"loss": 1.4614,
"step": 271
},
{
"epoch": 1.1017721518987342,
"grad_norm": 0.7696414652946932,
"learning_rate": 3.8238444863584164e-05,
"loss": 1.5256,
"step": 272
},
{
"epoch": 1.1058227848101265,
"grad_norm": 0.7818101103162195,
"learning_rate": 3.821507978767479e-05,
"loss": 1.4511,
"step": 273
},
{
"epoch": 1.109873417721519,
"grad_norm": 0.7805433282889446,
"learning_rate": 3.819156801001693e-05,
"loss": 1.5408,
"step": 274
},
{
"epoch": 1.1139240506329113,
"grad_norm": 0.6513921685435425,
"learning_rate": 3.816790971997121e-05,
"loss": 1.501,
"step": 275
},
{
"epoch": 1.1179746835443038,
"grad_norm": 0.8278883869216707,
"learning_rate": 3.8144105108078246e-05,
"loss": 1.4705,
"step": 276
},
{
"epoch": 1.1220253164556961,
"grad_norm": 0.6399105172529602,
"learning_rate": 3.81201543660571e-05,
"loss": 1.4786,
"step": 277
},
{
"epoch": 1.1260759493670887,
"grad_norm": 0.7883852619667905,
"learning_rate": 3.809605768680377e-05,
"loss": 1.5281,
"step": 278
},
{
"epoch": 1.130126582278481,
"grad_norm": 0.623031067087336,
"learning_rate": 3.807181526438958e-05,
"loss": 1.509,
"step": 279
},
{
"epoch": 1.1341772151898735,
"grad_norm": 0.7473280758164969,
"learning_rate": 3.8047427294059697e-05,
"loss": 1.5015,
"step": 280
},
{
"epoch": 1.1382278481012658,
"grad_norm": 0.5776605767857038,
"learning_rate": 3.802289397223145e-05,
"loss": 1.5389,
"step": 281
},
{
"epoch": 1.1422784810126583,
"grad_norm": 0.7567360955900058,
"learning_rate": 3.7998215496492854e-05,
"loss": 1.4893,
"step": 282
},
{
"epoch": 1.1463291139240506,
"grad_norm": 0.7497591435454941,
"learning_rate": 3.797339206560096e-05,
"loss": 1.4611,
"step": 283
},
{
"epoch": 1.150379746835443,
"grad_norm": 0.6223051909518278,
"learning_rate": 3.794842387948027e-05,
"loss": 1.5232,
"step": 284
},
{
"epoch": 1.1544303797468354,
"grad_norm": 0.8065231251198678,
"learning_rate": 3.7923311139221114e-05,
"loss": 1.4779,
"step": 285
},
{
"epoch": 1.158481012658228,
"grad_norm": 0.6694403426289538,
"learning_rate": 3.7898054047078054e-05,
"loss": 1.4466,
"step": 286
},
{
"epoch": 1.1625316455696202,
"grad_norm": 0.6919450541447565,
"learning_rate": 3.787265280646825e-05,
"loss": 1.5343,
"step": 287
},
{
"epoch": 1.1665822784810127,
"grad_norm": 0.6203161355284065,
"learning_rate": 3.7847107621969786e-05,
"loss": 1.4997,
"step": 288
},
{
"epoch": 1.170632911392405,
"grad_norm": 0.6336127452514888,
"learning_rate": 3.7821418699320064e-05,
"loss": 1.5338,
"step": 289
},
{
"epoch": 1.1746835443037975,
"grad_norm": 0.729058122847704,
"learning_rate": 3.7795586245414145e-05,
"loss": 1.5478,
"step": 290
},
{
"epoch": 1.1787341772151898,
"grad_norm": 0.6108334325568491,
"learning_rate": 3.776961046830306e-05,
"loss": 1.4685,
"step": 291
},
{
"epoch": 1.1827848101265823,
"grad_norm": 0.6600283359992037,
"learning_rate": 3.774349157719215e-05,
"loss": 1.481,
"step": 292
},
{
"epoch": 1.1868354430379746,
"grad_norm": 0.7085181595556652,
"learning_rate": 3.7717229782439365e-05,
"loss": 1.5057,
"step": 293
},
{
"epoch": 1.1908860759493671,
"grad_norm": 0.7674105573921582,
"learning_rate": 3.769082529555359e-05,
"loss": 1.5171,
"step": 294
},
{
"epoch": 1.1949367088607594,
"grad_norm": 0.7109479824388855,
"learning_rate": 3.766427832919294e-05,
"loss": 1.4944,
"step": 295
},
{
"epoch": 1.198987341772152,
"grad_norm": 0.8171631469731844,
"learning_rate": 3.7637589097163024e-05,
"loss": 1.4961,
"step": 296
},
{
"epoch": 1.2030379746835442,
"grad_norm": 0.7623954129140406,
"learning_rate": 3.761075781441526e-05,
"loss": 1.5355,
"step": 297
},
{
"epoch": 1.2070886075949367,
"grad_norm": 1.024015887717862,
"learning_rate": 3.75837846970451e-05,
"loss": 1.4807,
"step": 298
},
{
"epoch": 1.211139240506329,
"grad_norm": 0.7998275714747115,
"learning_rate": 3.755666996229032e-05,
"loss": 1.4925,
"step": 299
},
{
"epoch": 1.2151898734177216,
"grad_norm": 0.7200576137865807,
"learning_rate": 3.752941382852927e-05,
"loss": 1.4688,
"step": 300
},
{
"epoch": 1.2192405063291138,
"grad_norm": 0.779361410637896,
"learning_rate": 3.7502016515279115e-05,
"loss": 1.4575,
"step": 301
},
{
"epoch": 1.2232911392405064,
"grad_norm": 0.8133990331020858,
"learning_rate": 3.7474478243194043e-05,
"loss": 1.5074,
"step": 302
},
{
"epoch": 1.2273417721518987,
"grad_norm": 0.6440184984751446,
"learning_rate": 3.744679923406351e-05,
"loss": 1.5373,
"step": 303
},
{
"epoch": 1.2313924050632912,
"grad_norm": 0.9188588820549559,
"learning_rate": 3.741897971081043e-05,
"loss": 1.4778,
"step": 304
},
{
"epoch": 1.2354430379746835,
"grad_norm": 0.9203654791553199,
"learning_rate": 3.739101989748946e-05,
"loss": 1.5134,
"step": 305
},
{
"epoch": 1.239493670886076,
"grad_norm": 0.685303504646411,
"learning_rate": 3.7362920019285066e-05,
"loss": 1.4946,
"step": 306
},
{
"epoch": 1.2435443037974683,
"grad_norm": 0.7967551479920514,
"learning_rate": 3.73346803025098e-05,
"loss": 1.4688,
"step": 307
},
{
"epoch": 1.2475949367088608,
"grad_norm": 0.8320493721328717,
"learning_rate": 3.730630097460247e-05,
"loss": 1.4856,
"step": 308
},
{
"epoch": 1.251645569620253,
"grad_norm": 0.6007039585975955,
"learning_rate": 3.727778226412628e-05,
"loss": 1.4907,
"step": 309
},
{
"epoch": 1.2556962025316456,
"grad_norm": 0.8805301251182075,
"learning_rate": 3.7249124400767006e-05,
"loss": 1.5149,
"step": 310
},
{
"epoch": 1.259746835443038,
"grad_norm": 0.7451409167627974,
"learning_rate": 3.722032761533114e-05,
"loss": 1.4487,
"step": 311
},
{
"epoch": 1.2637974683544304,
"grad_norm": 0.6424030921279024,
"learning_rate": 3.719139213974403e-05,
"loss": 1.5298,
"step": 312
},
{
"epoch": 1.2678481012658227,
"grad_norm": 0.73400373037577,
"learning_rate": 3.7162318207048006e-05,
"loss": 1.4996,
"step": 313
},
{
"epoch": 1.2718987341772152,
"grad_norm": 0.6141139744218045,
"learning_rate": 3.713310605140055e-05,
"loss": 1.4841,
"step": 314
},
{
"epoch": 1.2759493670886077,
"grad_norm": 0.6259410761238472,
"learning_rate": 3.710375590807233e-05,
"loss": 1.5284,
"step": 315
},
{
"epoch": 1.28,
"grad_norm": 0.7269666605348997,
"learning_rate": 3.7074268013445365e-05,
"loss": 1.4859,
"step": 316
},
{
"epoch": 1.2840506329113923,
"grad_norm": 0.5615204896696946,
"learning_rate": 3.7044642605011114e-05,
"loss": 1.5129,
"step": 317
},
{
"epoch": 1.2881012658227848,
"grad_norm": 0.6770639931154463,
"learning_rate": 3.701487992136854e-05,
"loss": 1.4756,
"step": 318
},
{
"epoch": 1.2921518987341774,
"grad_norm": 0.6541552102786158,
"learning_rate": 3.69849802022222e-05,
"loss": 1.5214,
"step": 319
},
{
"epoch": 1.2962025316455696,
"grad_norm": 0.6088213814877959,
"learning_rate": 3.6954943688380334e-05,
"loss": 1.4812,
"step": 320
},
{
"epoch": 1.300253164556962,
"grad_norm": 0.7114158268211753,
"learning_rate": 3.692477062175289e-05,
"loss": 1.4893,
"step": 321
},
{
"epoch": 1.3043037974683545,
"grad_norm": 0.7171695678615494,
"learning_rate": 3.689446124534958e-05,
"loss": 1.5551,
"step": 322
},
{
"epoch": 1.308354430379747,
"grad_norm": 0.7092584355413049,
"learning_rate": 3.686401580327799e-05,
"loss": 1.4999,
"step": 323
},
{
"epoch": 1.3124050632911393,
"grad_norm": 0.7326665388714326,
"learning_rate": 3.683343454074149e-05,
"loss": 1.536,
"step": 324
},
{
"epoch": 1.3164556962025316,
"grad_norm": 0.7948278377722184,
"learning_rate": 3.6802717704037386e-05,
"loss": 1.4967,
"step": 325
},
{
"epoch": 1.320506329113924,
"grad_norm": 0.557613972849223,
"learning_rate": 3.6771865540554855e-05,
"loss": 1.4712,
"step": 326
},
{
"epoch": 1.3245569620253166,
"grad_norm": 0.6520587717594549,
"learning_rate": 3.674087829877297e-05,
"loss": 1.5364,
"step": 327
},
{
"epoch": 1.3286075949367089,
"grad_norm": 0.6216151179266305,
"learning_rate": 3.6709756228258735e-05,
"loss": 1.3998,
"step": 328
},
{
"epoch": 1.3326582278481012,
"grad_norm": 0.5756683958767114,
"learning_rate": 3.667849957966501e-05,
"loss": 1.4893,
"step": 329
},
{
"epoch": 1.3367088607594937,
"grad_norm": 0.7425008598614916,
"learning_rate": 3.6647108604728546e-05,
"loss": 1.4872,
"step": 330
},
{
"epoch": 1.3407594936708862,
"grad_norm": 0.6595361672990061,
"learning_rate": 3.661558355626795e-05,
"loss": 1.4859,
"step": 331
},
{
"epoch": 1.3448101265822785,
"grad_norm": 0.5419726130190505,
"learning_rate": 3.658392468818163e-05,
"loss": 1.4827,
"step": 332
},
{
"epoch": 1.3488607594936708,
"grad_norm": 0.6917351988790177,
"learning_rate": 3.655213225544574e-05,
"loss": 1.4889,
"step": 333
},
{
"epoch": 1.3529113924050633,
"grad_norm": 0.6330532161387704,
"learning_rate": 3.652020651411218e-05,
"loss": 1.5102,
"step": 334
},
{
"epoch": 1.3569620253164558,
"grad_norm": 0.5590336432810067,
"learning_rate": 3.6488147721306474e-05,
"loss": 1.5268,
"step": 335
},
{
"epoch": 1.3610126582278481,
"grad_norm": 0.5417133345618953,
"learning_rate": 3.645595613522574e-05,
"loss": 1.5219,
"step": 336
},
{
"epoch": 1.3650632911392404,
"grad_norm": 0.6653025897105774,
"learning_rate": 3.642363201513657e-05,
"loss": 1.4973,
"step": 337
},
{
"epoch": 1.369113924050633,
"grad_norm": 0.6586434656461185,
"learning_rate": 3.6391175621373006e-05,
"loss": 1.4828,
"step": 338
},
{
"epoch": 1.3731645569620254,
"grad_norm": 0.5609701409755484,
"learning_rate": 3.6358587215334355e-05,
"loss": 1.5192,
"step": 339
},
{
"epoch": 1.3772151898734177,
"grad_norm": 0.6711154924759631,
"learning_rate": 3.632586705948318e-05,
"loss": 1.4642,
"step": 340
},
{
"epoch": 1.38126582278481,
"grad_norm": 0.5849958805664981,
"learning_rate": 3.629301541734311e-05,
"loss": 1.5174,
"step": 341
},
{
"epoch": 1.3853164556962025,
"grad_norm": 0.5436108011891131,
"learning_rate": 3.626003255349676e-05,
"loss": 1.542,
"step": 342
},
{
"epoch": 1.389367088607595,
"grad_norm": 0.6469728564138539,
"learning_rate": 3.622691873358357e-05,
"loss": 1.4833,
"step": 343
},
{
"epoch": 1.3934177215189874,
"grad_norm": 0.5354431391598756,
"learning_rate": 3.61936742242977e-05,
"loss": 1.4933,
"step": 344
},
{
"epoch": 1.3974683544303796,
"grad_norm": 0.5739267124473401,
"learning_rate": 3.6160299293385864e-05,
"loss": 1.5341,
"step": 345
},
{
"epoch": 1.4015189873417722,
"grad_norm": 0.6454280713774461,
"learning_rate": 3.612679420964516e-05,
"loss": 1.4805,
"step": 346
},
{
"epoch": 1.4055696202531647,
"grad_norm": 0.6018490625414357,
"learning_rate": 3.609315924292092e-05,
"loss": 1.4913,
"step": 347
},
{
"epoch": 1.409620253164557,
"grad_norm": 0.582859622033331,
"learning_rate": 3.6059394664104554e-05,
"loss": 1.5215,
"step": 348
},
{
"epoch": 1.4136708860759493,
"grad_norm": 0.6234658245757503,
"learning_rate": 3.602550074513133e-05,
"loss": 1.5038,
"step": 349
},
{
"epoch": 1.4177215189873418,
"grad_norm": 0.5928213672519045,
"learning_rate": 3.599147775897822e-05,
"loss": 1.4802,
"step": 350
},
{
"epoch": 1.4217721518987343,
"grad_norm": 0.5035521779745382,
"learning_rate": 3.595732597966167e-05,
"loss": 1.4951,
"step": 351
},
{
"epoch": 1.4258227848101266,
"grad_norm": 0.6095478423725893,
"learning_rate": 3.592304568223542e-05,
"loss": 1.5147,
"step": 352
},
{
"epoch": 1.4298734177215189,
"grad_norm": 0.7126416698754735,
"learning_rate": 3.588863714278826e-05,
"loss": 1.51,
"step": 353
},
{
"epoch": 1.4339240506329114,
"grad_norm": 0.5261712072728097,
"learning_rate": 3.585410063844186e-05,
"loss": 1.4927,
"step": 354
},
{
"epoch": 1.437974683544304,
"grad_norm": 0.5289417694427142,
"learning_rate": 3.581943644734846e-05,
"loss": 1.4783,
"step": 355
},
{
"epoch": 1.4420253164556962,
"grad_norm": 0.5547280338838183,
"learning_rate": 3.578464484868869e-05,
"loss": 1.5049,
"step": 356
},
{
"epoch": 1.4460759493670885,
"grad_norm": 0.6048909361307292,
"learning_rate": 3.5749726122669316e-05,
"loss": 1.4819,
"step": 357
},
{
"epoch": 1.450126582278481,
"grad_norm": 0.5702887927923851,
"learning_rate": 3.5714680550520943e-05,
"loss": 1.484,
"step": 358
},
{
"epoch": 1.4541772151898735,
"grad_norm": 0.585776473245831,
"learning_rate": 3.5679508414495794e-05,
"loss": 1.4989,
"step": 359
},
{
"epoch": 1.4582278481012658,
"grad_norm": 0.5910529108885565,
"learning_rate": 3.564420999786543e-05,
"loss": 1.5405,
"step": 360
},
{
"epoch": 1.4622784810126581,
"grad_norm": 0.5580970606287686,
"learning_rate": 3.560878558491842e-05,
"loss": 1.4991,
"step": 361
},
{
"epoch": 1.4663291139240506,
"grad_norm": 0.5513711896458328,
"learning_rate": 3.5573235460958145e-05,
"loss": 1.5177,
"step": 362
},
{
"epoch": 1.4703797468354431,
"grad_norm": 0.5890963711777422,
"learning_rate": 3.553755991230039e-05,
"loss": 1.4917,
"step": 363
},
{
"epoch": 1.4744303797468354,
"grad_norm": 0.531495484096474,
"learning_rate": 3.5501759226271144e-05,
"loss": 1.4697,
"step": 364
},
{
"epoch": 1.4784810126582277,
"grad_norm": 0.5984175886719494,
"learning_rate": 3.546583369120419e-05,
"loss": 1.5065,
"step": 365
},
{
"epoch": 1.4825316455696202,
"grad_norm": 0.5441753080212223,
"learning_rate": 3.5429783596438864e-05,
"loss": 1.4734,
"step": 366
},
{
"epoch": 1.4865822784810128,
"grad_norm": 0.5926562081076293,
"learning_rate": 3.539360923231766e-05,
"loss": 1.4945,
"step": 367
},
{
"epoch": 1.490632911392405,
"grad_norm": 0.6392477553501779,
"learning_rate": 3.535731089018394e-05,
"loss": 1.4815,
"step": 368
},
{
"epoch": 1.4946835443037974,
"grad_norm": 0.575886512992848,
"learning_rate": 3.532088886237956e-05,
"loss": 1.5024,
"step": 369
},
{
"epoch": 1.4987341772151899,
"grad_norm": 0.612520576238627,
"learning_rate": 3.528434344224253e-05,
"loss": 1.4614,
"step": 370
},
{
"epoch": 1.5027848101265824,
"grad_norm": 0.5563329223452396,
"learning_rate": 3.524767492410464e-05,
"loss": 1.4731,
"step": 371
},
{
"epoch": 1.5068354430379747,
"grad_norm": 0.5901452191766194,
"learning_rate": 3.521088360328908e-05,
"loss": 1.5217,
"step": 372
},
{
"epoch": 1.510886075949367,
"grad_norm": 0.5754742594560367,
"learning_rate": 3.517396977610811e-05,
"loss": 1.4698,
"step": 373
},
{
"epoch": 1.5149367088607595,
"grad_norm": 0.634521664736648,
"learning_rate": 3.5136933739860595e-05,
"loss": 1.4577,
"step": 374
},
{
"epoch": 1.518987341772152,
"grad_norm": 0.5842596441296014,
"learning_rate": 3.509977579282971e-05,
"loss": 1.5055,
"step": 375
},
{
"epoch": 1.5230379746835443,
"grad_norm": 0.5790601452705227,
"learning_rate": 3.5062496234280424e-05,
"loss": 1.505,
"step": 376
},
{
"epoch": 1.5270886075949366,
"grad_norm": 0.7055095024577823,
"learning_rate": 3.502509536445719e-05,
"loss": 1.495,
"step": 377
},
{
"epoch": 1.531139240506329,
"grad_norm": 0.6011186222335092,
"learning_rate": 3.498757348458147e-05,
"loss": 1.507,
"step": 378
},
{
"epoch": 1.5351898734177216,
"grad_norm": 0.6241043605770006,
"learning_rate": 3.4949930896849324e-05,
"loss": 1.499,
"step": 379
},
{
"epoch": 1.539240506329114,
"grad_norm": 0.631061818576299,
"learning_rate": 3.491216790442899e-05,
"loss": 1.5207,
"step": 380
},
{
"epoch": 1.5432911392405062,
"grad_norm": 0.5564243492401068,
"learning_rate": 3.487428481145839e-05,
"loss": 1.4916,
"step": 381
},
{
"epoch": 1.5473417721518987,
"grad_norm": 0.6133124981648787,
"learning_rate": 3.483628192304278e-05,
"loss": 1.4894,
"step": 382
},
{
"epoch": 1.5513924050632912,
"grad_norm": 0.5599684952631202,
"learning_rate": 3.479815954525219e-05,
"loss": 1.4736,
"step": 383
},
{
"epoch": 1.5554430379746835,
"grad_norm": 0.7198083998953423,
"learning_rate": 3.475991798511899e-05,
"loss": 1.5119,
"step": 384
},
{
"epoch": 1.5594936708860758,
"grad_norm": 0.6749256906633588,
"learning_rate": 3.4721557550635464e-05,
"loss": 1.5429,
"step": 385
},
{
"epoch": 1.5635443037974683,
"grad_norm": 0.6342598299793724,
"learning_rate": 3.468307855075128e-05,
"loss": 1.4603,
"step": 386
},
{
"epoch": 1.5675949367088609,
"grad_norm": 0.6938637336609185,
"learning_rate": 3.4644481295371005e-05,
"loss": 1.5081,
"step": 387
},
{
"epoch": 1.5716455696202531,
"grad_norm": 0.663023205386894,
"learning_rate": 3.460576609535163e-05,
"loss": 1.4922,
"step": 388
},
{
"epoch": 1.5756962025316454,
"grad_norm": 0.6085658968712089,
"learning_rate": 3.456693326250006e-05,
"loss": 1.4933,
"step": 389
},
{
"epoch": 1.579746835443038,
"grad_norm": 0.5843184464679858,
"learning_rate": 3.452798310957058e-05,
"loss": 1.4549,
"step": 390
},
{
"epoch": 1.5837974683544305,
"grad_norm": 0.6687391090633804,
"learning_rate": 3.4488915950262386e-05,
"loss": 1.5285,
"step": 391
},
{
"epoch": 1.5878481012658228,
"grad_norm": 0.6813419860742185,
"learning_rate": 3.4449732099216985e-05,
"loss": 1.5596,
"step": 392
},
{
"epoch": 1.591898734177215,
"grad_norm": 0.6321299607313734,
"learning_rate": 3.441043187201574e-05,
"loss": 1.5184,
"step": 393
},
{
"epoch": 1.5959493670886076,
"grad_norm": 0.6771449459044923,
"learning_rate": 3.437101558517728e-05,
"loss": 1.4652,
"step": 394
},
{
"epoch": 1.6,
"grad_norm": 0.670943318467807,
"learning_rate": 3.433148355615496e-05,
"loss": 1.534,
"step": 395
},
{
"epoch": 1.6040506329113924,
"grad_norm": 0.5791458591273599,
"learning_rate": 3.4291836103334294e-05,
"loss": 1.5218,
"step": 396
},
{
"epoch": 1.6081012658227847,
"grad_norm": 0.5389984012187755,
"learning_rate": 3.425207354603043e-05,
"loss": 1.4884,
"step": 397
},
{
"epoch": 1.6121518987341772,
"grad_norm": 0.5324198721992685,
"learning_rate": 3.421219620448553e-05,
"loss": 1.4759,
"step": 398
},
{
"epoch": 1.6162025316455697,
"grad_norm": 0.6078623053527729,
"learning_rate": 3.417220439986623e-05,
"loss": 1.4883,
"step": 399
},
{
"epoch": 1.620253164556962,
"grad_norm": 0.49341628096924994,
"learning_rate": 3.4132098454261024e-05,
"loss": 1.5128,
"step": 400
},
{
"epoch": 1.6243037974683543,
"grad_norm": 0.6116979272342264,
"learning_rate": 3.4091878690677676e-05,
"loss": 1.4221,
"step": 401
},
{
"epoch": 1.6283544303797468,
"grad_norm": 0.6731345844330359,
"learning_rate": 3.405154543304065e-05,
"loss": 1.5549,
"step": 402
},
{
"epoch": 1.6324050632911393,
"grad_norm": 0.5277943611060174,
"learning_rate": 3.401109900618843e-05,
"loss": 1.4729,
"step": 403
},
{
"epoch": 1.6364556962025316,
"grad_norm": 0.7206049056569799,
"learning_rate": 3.3970539735870996e-05,
"loss": 1.4874,
"step": 404
},
{
"epoch": 1.640506329113924,
"grad_norm": 0.5392930415043001,
"learning_rate": 3.392986794874714e-05,
"loss": 1.5133,
"step": 405
},
{
"epoch": 1.6445569620253164,
"grad_norm": 0.5801433064568511,
"learning_rate": 3.388908397238184e-05,
"loss": 1.5622,
"step": 406
},
{
"epoch": 1.648607594936709,
"grad_norm": 0.559312081920737,
"learning_rate": 3.384818813524362e-05,
"loss": 1.5292,
"step": 407
},
{
"epoch": 1.6526582278481012,
"grad_norm": 0.5745694534709397,
"learning_rate": 3.380718076670195e-05,
"loss": 1.5115,
"step": 408
},
{
"epoch": 1.6567088607594935,
"grad_norm": 0.5640971947233311,
"learning_rate": 3.376606219702454e-05,
"loss": 1.5033,
"step": 409
},
{
"epoch": 1.660759493670886,
"grad_norm": 0.6056302218739594,
"learning_rate": 3.372483275737468e-05,
"loss": 1.5252,
"step": 410
},
{
"epoch": 1.6648101265822786,
"grad_norm": 0.6447044237129445,
"learning_rate": 3.368349277980861e-05,
"loss": 1.5096,
"step": 411
},
{
"epoch": 1.6688607594936709,
"grad_norm": 0.6636789639167576,
"learning_rate": 3.3642042597272844e-05,
"loss": 1.5061,
"step": 412
},
{
"epoch": 1.6729113924050631,
"grad_norm": 0.7471252404365849,
"learning_rate": 3.360048254360144e-05,
"loss": 1.5252,
"step": 413
},
{
"epoch": 1.6769620253164557,
"grad_norm": 0.6691307572658453,
"learning_rate": 3.355881295351336e-05,
"loss": 1.5393,
"step": 414
},
{
"epoch": 1.6810126582278482,
"grad_norm": 0.6050295457287403,
"learning_rate": 3.351703416260975e-05,
"loss": 1.4831,
"step": 415
},
{
"epoch": 1.6850632911392405,
"grad_norm": 0.5975157621446497,
"learning_rate": 3.347514650737126e-05,
"loss": 1.508,
"step": 416
},
{
"epoch": 1.6891139240506328,
"grad_norm": 0.5738352973641911,
"learning_rate": 3.3433150325155295e-05,
"loss": 1.5152,
"step": 417
},
{
"epoch": 1.6931645569620253,
"grad_norm": 0.5427612194734308,
"learning_rate": 3.339104595419334e-05,
"loss": 1.5127,
"step": 418
},
{
"epoch": 1.6972151898734178,
"grad_norm": 0.5531938499817652,
"learning_rate": 3.3348833733588204e-05,
"loss": 1.5103,
"step": 419
},
{
"epoch": 1.70126582278481,
"grad_norm": 0.5957039724618606,
"learning_rate": 3.3306514003311305e-05,
"loss": 1.5125,
"step": 420
},
{
"epoch": 1.7053164556962024,
"grad_norm": 0.6189730030916409,
"learning_rate": 3.326408710419996e-05,
"loss": 1.5217,
"step": 421
},
{
"epoch": 1.709367088607595,
"grad_norm": 0.5549514605327449,
"learning_rate": 3.322155337795454e-05,
"loss": 1.4787,
"step": 422
},
{
"epoch": 1.7134177215189874,
"grad_norm": 0.6127855241654301,
"learning_rate": 3.317891316713587e-05,
"loss": 1.4546,
"step": 423
},
{
"epoch": 1.7174683544303797,
"grad_norm": 0.6406415166265726,
"learning_rate": 3.313616681516231e-05,
"loss": 1.5002,
"step": 424
},
{
"epoch": 1.721518987341772,
"grad_norm": 0.6484623694181115,
"learning_rate": 3.309331466630713e-05,
"loss": 1.4777,
"step": 425
},
{
"epoch": 1.7255696202531645,
"grad_norm": 0.552030172200037,
"learning_rate": 3.305035706569563e-05,
"loss": 1.5124,
"step": 426
},
{
"epoch": 1.729620253164557,
"grad_norm": 0.6051034868261812,
"learning_rate": 3.3007294359302433e-05,
"loss": 1.5263,
"step": 427
},
{
"epoch": 1.7336708860759493,
"grad_norm": 0.5631528312354831,
"learning_rate": 3.296412689394864e-05,
"loss": 1.5018,
"step": 428
},
{
"epoch": 1.7377215189873416,
"grad_norm": 0.6269437147196972,
"learning_rate": 3.292085501729909e-05,
"loss": 1.496,
"step": 429
},
{
"epoch": 1.7417721518987341,
"grad_norm": 0.6052651126146383,
"learning_rate": 3.2877479077859534e-05,
"loss": 1.5097,
"step": 430
},
{
"epoch": 1.7458227848101266,
"grad_norm": 0.6489784627555383,
"learning_rate": 3.283399942497381e-05,
"loss": 1.513,
"step": 431
},
{
"epoch": 1.749873417721519,
"grad_norm": 0.5261700461188061,
"learning_rate": 3.279041640882108e-05,
"loss": 1.5102,
"step": 432
},
{
"epoch": 1.7539240506329112,
"grad_norm": 0.6682169269425776,
"learning_rate": 3.2746730380412964e-05,
"loss": 1.509,
"step": 433
},
{
"epoch": 1.7579746835443038,
"grad_norm": 0.5549809897755573,
"learning_rate": 3.2702941691590726e-05,
"loss": 1.5035,
"step": 434
},
{
"epoch": 1.7620253164556963,
"grad_norm": 0.5872702715563124,
"learning_rate": 3.265905069502244e-05,
"loss": 1.5014,
"step": 435
},
{
"epoch": 1.7660759493670886,
"grad_norm": 0.6203031657988995,
"learning_rate": 3.261505774420016e-05,
"loss": 1.4801,
"step": 436
},
{
"epoch": 1.7701265822784809,
"grad_norm": 0.5500803243943837,
"learning_rate": 3.257096319343707e-05,
"loss": 1.5102,
"step": 437
},
{
"epoch": 1.7741772151898734,
"grad_norm": 0.5693719595325595,
"learning_rate": 3.2526767397864614e-05,
"loss": 1.5091,
"step": 438
},
{
"epoch": 1.7782278481012659,
"grad_norm": 0.4880281770789791,
"learning_rate": 3.248247071342966e-05,
"loss": 1.5249,
"step": 439
},
{
"epoch": 1.7822784810126582,
"grad_norm": 0.6313663711773843,
"learning_rate": 3.243807349689161e-05,
"loss": 1.4863,
"step": 440
},
{
"epoch": 1.7863291139240505,
"grad_norm": 0.473075262584268,
"learning_rate": 3.2393576105819544e-05,
"loss": 1.5055,
"step": 441
},
{
"epoch": 1.790379746835443,
"grad_norm": 0.5540268066217658,
"learning_rate": 3.2348978898589333e-05,
"loss": 1.4913,
"step": 442
},
{
"epoch": 1.7944303797468355,
"grad_norm": 0.5883344943997437,
"learning_rate": 3.230428223438075e-05,
"loss": 1.5341,
"step": 443
},
{
"epoch": 1.7984810126582278,
"grad_norm": 0.5533422270178031,
"learning_rate": 3.225948647317459e-05,
"loss": 1.513,
"step": 444
},
{
"epoch": 1.80253164556962,
"grad_norm": 0.5926761417611788,
"learning_rate": 3.2214591975749745e-05,
"loss": 1.5038,
"step": 445
},
{
"epoch": 1.8065822784810126,
"grad_norm": 0.5552971931256268,
"learning_rate": 3.216959910368034e-05,
"loss": 1.532,
"step": 446
},
{
"epoch": 1.8106329113924051,
"grad_norm": 0.5568267864173251,
"learning_rate": 3.212450821933277e-05,
"loss": 1.4885,
"step": 447
},
{
"epoch": 1.8146835443037974,
"grad_norm": 0.5066504639529399,
"learning_rate": 3.207931968586281e-05,
"loss": 1.4739,
"step": 448
},
{
"epoch": 1.81873417721519,
"grad_norm": 0.6382395667149189,
"learning_rate": 3.203403386721272e-05,
"loss": 1.477,
"step": 449
},
{
"epoch": 1.8227848101265822,
"grad_norm": 0.6097180468273785,
"learning_rate": 3.1988651128108245e-05,
"loss": 1.4645,
"step": 450
},
{
"epoch": 1.8268354430379747,
"grad_norm": 0.47615061502524475,
"learning_rate": 3.194317183405573e-05,
"loss": 1.5203,
"step": 451
},
{
"epoch": 1.8308860759493673,
"grad_norm": 0.6629344403838621,
"learning_rate": 3.189759635133914e-05,
"loss": 1.4913,
"step": 452
},
{
"epoch": 1.8349367088607595,
"grad_norm": 0.5184096192612507,
"learning_rate": 3.185192504701718e-05,
"loss": 1.4645,
"step": 453
},
{
"epoch": 1.8389873417721518,
"grad_norm": 0.5434259635149664,
"learning_rate": 3.1806158288920234e-05,
"loss": 1.5049,
"step": 454
},
{
"epoch": 1.8430379746835444,
"grad_norm": 0.5405827473111477,
"learning_rate": 3.1760296445647477e-05,
"loss": 1.4935,
"step": 455
},
{
"epoch": 1.8470886075949369,
"grad_norm": 0.5920962789408867,
"learning_rate": 3.1714339886563896e-05,
"loss": 1.4086,
"step": 456
},
{
"epoch": 1.8511392405063292,
"grad_norm": 0.6241449926822576,
"learning_rate": 3.166828898179731e-05,
"loss": 1.4934,
"step": 457
},
{
"epoch": 1.8551898734177215,
"grad_norm": 0.47720946189721014,
"learning_rate": 3.162214410223536e-05,
"loss": 1.5014,
"step": 458
},
{
"epoch": 1.859240506329114,
"grad_norm": 0.5251553414296696,
"learning_rate": 3.157590561952257e-05,
"loss": 1.4733,
"step": 459
},
{
"epoch": 1.8632911392405065,
"grad_norm": 0.5357689990037319,
"learning_rate": 3.152957390605732e-05,
"loss": 1.527,
"step": 460
},
{
"epoch": 1.8673417721518988,
"grad_norm": 0.5187281023997085,
"learning_rate": 3.148314933498886e-05,
"loss": 1.5015,
"step": 461
},
{
"epoch": 1.871392405063291,
"grad_norm": 0.5776652045167943,
"learning_rate": 3.143663228021431e-05,
"loss": 1.513,
"step": 462
},
{
"epoch": 1.8754430379746836,
"grad_norm": 0.5038762978532217,
"learning_rate": 3.1390023116375624e-05,
"loss": 1.4386,
"step": 463
},
{
"epoch": 1.879493670886076,
"grad_norm": 0.6027624101026343,
"learning_rate": 3.134332221885661e-05,
"loss": 1.4705,
"step": 464
},
{
"epoch": 1.8835443037974684,
"grad_norm": 0.554106946326009,
"learning_rate": 3.129652996377987e-05,
"loss": 1.5154,
"step": 465
},
{
"epoch": 1.8875949367088607,
"grad_norm": 0.525625471929272,
"learning_rate": 3.12496467280038e-05,
"loss": 1.4831,
"step": 466
},
{
"epoch": 1.8916455696202532,
"grad_norm": 0.550242897127108,
"learning_rate": 3.120267288911952e-05,
"loss": 1.4678,
"step": 467
},
{
"epoch": 1.8956962025316457,
"grad_norm": 0.48638010866342946,
"learning_rate": 3.11556088254479e-05,
"loss": 1.5194,
"step": 468
},
{
"epoch": 1.899746835443038,
"grad_norm": 0.563782329064369,
"learning_rate": 3.11084549160364e-05,
"loss": 1.5055,
"step": 469
},
{
"epoch": 1.9037974683544303,
"grad_norm": 0.537922396416219,
"learning_rate": 3.106121154065615e-05,
"loss": 1.4896,
"step": 470
},
{
"epoch": 1.9078481012658228,
"grad_norm": 0.5137423175524479,
"learning_rate": 3.1013879079798805e-05,
"loss": 1.4749,
"step": 471
},
{
"epoch": 1.9118987341772153,
"grad_norm": 0.5930965130673588,
"learning_rate": 3.096645791467348e-05,
"loss": 1.4947,
"step": 472
},
{
"epoch": 1.9159493670886076,
"grad_norm": 0.5242969337593649,
"learning_rate": 3.091894842720373e-05,
"loss": 1.4867,
"step": 473
},
{
"epoch": 1.92,
"grad_norm": 0.5552538475364676,
"learning_rate": 3.0871351000024425e-05,
"loss": 1.5014,
"step": 474
},
{
"epoch": 1.9240506329113924,
"grad_norm": 0.5792885810294173,
"learning_rate": 3.0823666016478716e-05,
"loss": 1.4601,
"step": 475
},
{
"epoch": 1.928101265822785,
"grad_norm": 0.6609970476810498,
"learning_rate": 3.0775893860614896e-05,
"loss": 1.4963,
"step": 476
},
{
"epoch": 1.9321518987341773,
"grad_norm": 0.5327851105022522,
"learning_rate": 3.0728034917183336e-05,
"loss": 1.5076,
"step": 477
},
{
"epoch": 1.9362025316455695,
"grad_norm": 0.5505211824117172,
"learning_rate": 3.06800895716334e-05,
"loss": 1.5318,
"step": 478
},
{
"epoch": 1.940253164556962,
"grad_norm": 0.49430857392574856,
"learning_rate": 3.063205821011029e-05,
"loss": 1.4933,
"step": 479
},
{
"epoch": 1.9443037974683546,
"grad_norm": 0.5413707638418142,
"learning_rate": 3.0583941219452016e-05,
"loss": 1.5031,
"step": 480
},
{
"epoch": 1.9483544303797469,
"grad_norm": 0.545358661749885,
"learning_rate": 3.053573898718618e-05,
"loss": 1.4707,
"step": 481
},
{
"epoch": 1.9524050632911392,
"grad_norm": 0.502749490024255,
"learning_rate": 3.0487451901526956e-05,
"loss": 1.497,
"step": 482
},
{
"epoch": 1.9564556962025317,
"grad_norm": 0.5291790058456792,
"learning_rate": 3.0439080351371875e-05,
"loss": 1.495,
"step": 483
},
{
"epoch": 1.9605063291139242,
"grad_norm": 0.6451798822453867,
"learning_rate": 3.0390624726298764e-05,
"loss": 1.4965,
"step": 484
},
{
"epoch": 1.9645569620253165,
"grad_norm": 0.6110816450151265,
"learning_rate": 3.034208541656255e-05,
"loss": 1.4595,
"step": 485
},
{
"epoch": 1.9686075949367088,
"grad_norm": 0.7081348318600158,
"learning_rate": 3.029346281309218e-05,
"loss": 1.5326,
"step": 486
},
{
"epoch": 1.9726582278481013,
"grad_norm": 0.6316148728533176,
"learning_rate": 3.0244757307487415e-05,
"loss": 1.4628,
"step": 487
},
{
"epoch": 1.9767088607594938,
"grad_norm": 0.5106208228250058,
"learning_rate": 3.019596929201569e-05,
"loss": 1.5317,
"step": 488
},
{
"epoch": 1.980759493670886,
"grad_norm": 0.5619113292836765,
"learning_rate": 3.0147099159608985e-05,
"loss": 1.5259,
"step": 489
},
{
"epoch": 1.9848101265822784,
"grad_norm": 0.5243214851088815,
"learning_rate": 3.0098147303860616e-05,
"loss": 1.4726,
"step": 490
},
{
"epoch": 1.988860759493671,
"grad_norm": 0.4788003971926545,
"learning_rate": 3.0049114119022117e-05,
"loss": 1.4612,
"step": 491
},
{
"epoch": 1.9929113924050634,
"grad_norm": 0.5199778414412846,
"learning_rate": 3.0000000000000004e-05,
"loss": 1.5243,
"step": 492
},
{
"epoch": 1.9969620253164557,
"grad_norm": 0.4829884400025195,
"learning_rate": 2.995080534235264e-05,
"loss": 1.5375,
"step": 493
},
{
"epoch": 2.001012658227848,
"grad_norm": 0.6270269762496853,
"learning_rate": 2.9901530542287044e-05,
"loss": 1.6159,
"step": 494
},
{
"epoch": 2.0050632911392405,
"grad_norm": 0.6038549059549236,
"learning_rate": 2.9852175996655676e-05,
"loss": 1.3199,
"step": 495
},
{
"epoch": 2.009113924050633,
"grad_norm": 0.6830642752849071,
"learning_rate": 2.980274210295326e-05,
"loss": 1.3264,
"step": 496
},
{
"epoch": 2.013164556962025,
"grad_norm": 0.8173245872623136,
"learning_rate": 2.9753229259313578e-05,
"loss": 1.3827,
"step": 497
},
{
"epoch": 2.0172151898734176,
"grad_norm": 0.8713400808099322,
"learning_rate": 2.9703637864506274e-05,
"loss": 1.3627,
"step": 498
},
{
"epoch": 2.02126582278481,
"grad_norm": 0.8927931504581594,
"learning_rate": 2.965396831793362e-05,
"loss": 1.3602,
"step": 499
},
{
"epoch": 2.0253164556962027,
"grad_norm": 0.7192181138426044,
"learning_rate": 2.9604221019627316e-05,
"loss": 1.365,
"step": 500
},
{
"epoch": 2.0293670886075947,
"grad_norm": 0.9639961443281839,
"learning_rate": 2.955439637024526e-05,
"loss": 1.3617,
"step": 501
},
{
"epoch": 2.0334177215189873,
"grad_norm": 0.8335735175838374,
"learning_rate": 2.9504494771068334e-05,
"loss": 1.3974,
"step": 502
},
{
"epoch": 2.0374683544303798,
"grad_norm": 0.756539272995457,
"learning_rate": 2.9454516623997156e-05,
"loss": 1.3752,
"step": 503
},
{
"epoch": 2.0415189873417723,
"grad_norm": 0.9260807963433714,
"learning_rate": 2.9404462331548847e-05,
"loss": 1.3596,
"step": 504
},
{
"epoch": 2.0455696202531644,
"grad_norm": 0.7264827751010255,
"learning_rate": 2.93543322968538e-05,
"loss": 1.3676,
"step": 505
},
{
"epoch": 2.049620253164557,
"grad_norm": 0.8883638779160667,
"learning_rate": 2.9304126923652428e-05,
"loss": 1.345,
"step": 506
},
{
"epoch": 2.0536708860759494,
"grad_norm": 0.7769710364399213,
"learning_rate": 2.9253846616291896e-05,
"loss": 1.3657,
"step": 507
},
{
"epoch": 2.057721518987342,
"grad_norm": 0.7802710136816351,
"learning_rate": 2.9203491779722896e-05,
"loss": 1.3743,
"step": 508
},
{
"epoch": 2.061772151898734,
"grad_norm": 0.7183941778191146,
"learning_rate": 2.9153062819496357e-05,
"loss": 1.3378,
"step": 509
},
{
"epoch": 2.0658227848101265,
"grad_norm": 0.8357409793745743,
"learning_rate": 2.9102560141760178e-05,
"loss": 1.3683,
"step": 510
},
{
"epoch": 2.069873417721519,
"grad_norm": 0.7422409987856647,
"learning_rate": 2.9051984153256004e-05,
"loss": 1.3683,
"step": 511
},
{
"epoch": 2.0739240506329115,
"grad_norm": 0.9473275418257886,
"learning_rate": 2.900133526131588e-05,
"loss": 1.3903,
"step": 512
},
{
"epoch": 2.0779746835443036,
"grad_norm": 0.6896977038694101,
"learning_rate": 2.8950613873859025e-05,
"loss": 1.3909,
"step": 513
},
{
"epoch": 2.082025316455696,
"grad_norm": 0.7082824291137971,
"learning_rate": 2.8899820399388515e-05,
"loss": 1.3669,
"step": 514
},
{
"epoch": 2.0860759493670886,
"grad_norm": 0.6178864837819786,
"learning_rate": 2.8848955246988012e-05,
"loss": 1.3464,
"step": 515
},
{
"epoch": 2.090126582278481,
"grad_norm": 0.6164241337429134,
"learning_rate": 2.879801882631847e-05,
"loss": 1.3852,
"step": 516
},
{
"epoch": 2.094177215189873,
"grad_norm": 0.6609208140739328,
"learning_rate": 2.8747011547614808e-05,
"loss": 1.3621,
"step": 517
},
{
"epoch": 2.0982278481012657,
"grad_norm": 0.6722630348786474,
"learning_rate": 2.8695933821682635e-05,
"loss": 1.3589,
"step": 518
},
{
"epoch": 2.1022784810126582,
"grad_norm": 0.652845284013913,
"learning_rate": 2.864478605989494e-05,
"loss": 1.3789,
"step": 519
},
{
"epoch": 2.1063291139240508,
"grad_norm": 0.5401668408479317,
"learning_rate": 2.8593568674188765e-05,
"loss": 1.3679,
"step": 520
},
{
"epoch": 2.110379746835443,
"grad_norm": 0.6368266414017458,
"learning_rate": 2.8542282077061892e-05,
"loss": 1.3553,
"step": 521
},
{
"epoch": 2.1144303797468353,
"grad_norm": 0.5119023403529904,
"learning_rate": 2.8490926681569523e-05,
"loss": 1.3722,
"step": 522
},
{
"epoch": 2.118481012658228,
"grad_norm": 0.6361624372829713,
"learning_rate": 2.8439502901320956e-05,
"loss": 1.3638,
"step": 523
},
{
"epoch": 2.1225316455696204,
"grad_norm": 0.5698564117540146,
"learning_rate": 2.8388011150476237e-05,
"loss": 1.3264,
"step": 524
},
{
"epoch": 2.1265822784810124,
"grad_norm": 0.5824294446617527,
"learning_rate": 2.8336451843742866e-05,
"loss": 1.3621,
"step": 525
},
{
"epoch": 2.130632911392405,
"grad_norm": 0.5911361424068198,
"learning_rate": 2.8284825396372387e-05,
"loss": 1.3171,
"step": 526
},
{
"epoch": 2.1346835443037975,
"grad_norm": 0.530417179122293,
"learning_rate": 2.8233132224157132e-05,
"loss": 1.3423,
"step": 527
},
{
"epoch": 2.13873417721519,
"grad_norm": 0.5192158027431856,
"learning_rate": 2.8181372743426805e-05,
"loss": 1.3579,
"step": 528
},
{
"epoch": 2.1427848101265825,
"grad_norm": 0.5464533677215277,
"learning_rate": 2.8129547371045128e-05,
"loss": 1.2951,
"step": 529
},
{
"epoch": 2.1468354430379746,
"grad_norm": 0.5798268240831366,
"learning_rate": 2.8077656524406534e-05,
"loss": 1.3737,
"step": 530
},
{
"epoch": 2.150886075949367,
"grad_norm": 0.5881406672020669,
"learning_rate": 2.802570062143278e-05,
"loss": 1.328,
"step": 531
},
{
"epoch": 2.1549367088607596,
"grad_norm": 0.5583032189172379,
"learning_rate": 2.7973680080569555e-05,
"loss": 1.3601,
"step": 532
},
{
"epoch": 2.1589873417721517,
"grad_norm": 0.5725150734314741,
"learning_rate": 2.792159532078314e-05,
"loss": 1.3159,
"step": 533
},
{
"epoch": 2.163037974683544,
"grad_norm": 0.5364321499435175,
"learning_rate": 2.7869446761557033e-05,
"loss": 1.3856,
"step": 534
},
{
"epoch": 2.1670886075949367,
"grad_norm": 0.5639395027525452,
"learning_rate": 2.781723482288857e-05,
"loss": 1.3448,
"step": 535
},
{
"epoch": 2.1711392405063292,
"grad_norm": 0.5460099379579837,
"learning_rate": 2.7764959925285517e-05,
"loss": 1.3619,
"step": 536
},
{
"epoch": 2.1751898734177217,
"grad_norm": 0.5797198541695741,
"learning_rate": 2.771262248976272e-05,
"loss": 1.32,
"step": 537
},
{
"epoch": 2.179240506329114,
"grad_norm": 0.5738176935226658,
"learning_rate": 2.7660222937838677e-05,
"loss": 1.395,
"step": 538
},
{
"epoch": 2.1832911392405063,
"grad_norm": 0.5699013381714293,
"learning_rate": 2.7607761691532186e-05,
"loss": 1.4144,
"step": 539
},
{
"epoch": 2.187341772151899,
"grad_norm": 0.5299299090993017,
"learning_rate": 2.7555239173358916e-05,
"loss": 1.3245,
"step": 540
},
{
"epoch": 2.191392405063291,
"grad_norm": 0.5822358146728666,
"learning_rate": 2.7502655806328e-05,
"loss": 1.3944,
"step": 541
},
{
"epoch": 2.1954430379746834,
"grad_norm": 0.5427419208597998,
"learning_rate": 2.7450012013938648e-05,
"loss": 1.36,
"step": 542
},
{
"epoch": 2.199493670886076,
"grad_norm": 0.529970238074713,
"learning_rate": 2.739730822017673e-05,
"loss": 1.3257,
"step": 543
},
{
"epoch": 2.2035443037974685,
"grad_norm": 0.5972475097109792,
"learning_rate": 2.7344544849511355e-05,
"loss": 1.3959,
"step": 544
},
{
"epoch": 2.207594936708861,
"grad_norm": 0.5089159025065362,
"learning_rate": 2.7291722326891456e-05,
"loss": 1.3755,
"step": 545
},
{
"epoch": 2.211645569620253,
"grad_norm": 0.6521040802633509,
"learning_rate": 2.723884107774236e-05,
"loss": 1.3825,
"step": 546
},
{
"epoch": 2.2156962025316456,
"grad_norm": 0.5224779248357908,
"learning_rate": 2.718590152796239e-05,
"loss": 1.3381,
"step": 547
},
{
"epoch": 2.219746835443038,
"grad_norm": 0.5876916111309168,
"learning_rate": 2.71329041039194e-05,
"loss": 1.3339,
"step": 548
},
{
"epoch": 2.22379746835443,
"grad_norm": 0.4895130522120739,
"learning_rate": 2.7079849232447357e-05,
"loss": 1.366,
"step": 549
},
{
"epoch": 2.2278481012658227,
"grad_norm": 0.578668722305487,
"learning_rate": 2.7026737340842895e-05,
"loss": 1.3853,
"step": 550
},
{
"epoch": 2.231898734177215,
"grad_norm": 0.6037806572989179,
"learning_rate": 2.697356885686189e-05,
"loss": 1.3514,
"step": 551
},
{
"epoch": 2.2359493670886077,
"grad_norm": 0.4884453358967491,
"learning_rate": 2.6920344208716014e-05,
"loss": 1.3669,
"step": 552
},
{
"epoch": 2.24,
"grad_norm": 0.5434144424801743,
"learning_rate": 2.6867063825069252e-05,
"loss": 1.3292,
"step": 553
},
{
"epoch": 2.2440506329113923,
"grad_norm": 0.5684299261005658,
"learning_rate": 2.6813728135034494e-05,
"loss": 1.3959,
"step": 554
},
{
"epoch": 2.248101265822785,
"grad_norm": 0.5012974590965793,
"learning_rate": 2.6760337568170056e-05,
"loss": 1.3948,
"step": 555
},
{
"epoch": 2.2521518987341773,
"grad_norm": 0.5022633411926101,
"learning_rate": 2.6706892554476226e-05,
"loss": 1.347,
"step": 556
},
{
"epoch": 2.2562025316455694,
"grad_norm": 0.6067156569917771,
"learning_rate": 2.6653393524391795e-05,
"loss": 1.3439,
"step": 557
},
{
"epoch": 2.260253164556962,
"grad_norm": 0.4751611038042102,
"learning_rate": 2.6599840908790592e-05,
"loss": 1.3454,
"step": 558
},
{
"epoch": 2.2643037974683544,
"grad_norm": 0.50760722285793,
"learning_rate": 2.6546235138978028e-05,
"loss": 1.3125,
"step": 559
},
{
"epoch": 2.268354430379747,
"grad_norm": 0.5017694833404508,
"learning_rate": 2.6492576646687597e-05,
"loss": 1.3279,
"step": 560
},
{
"epoch": 2.2724050632911394,
"grad_norm": 0.4877761253662506,
"learning_rate": 2.6438865864077425e-05,
"loss": 1.3502,
"step": 561
},
{
"epoch": 2.2764556962025315,
"grad_norm": 0.5074158797545743,
"learning_rate": 2.6385103223726766e-05,
"loss": 1.3275,
"step": 562
},
{
"epoch": 2.280506329113924,
"grad_norm": 0.5343561617178727,
"learning_rate": 2.6331289158632537e-05,
"loss": 1.3621,
"step": 563
},
{
"epoch": 2.2845569620253166,
"grad_norm": 0.5051125525631159,
"learning_rate": 2.6277424102205817e-05,
"loss": 1.3591,
"step": 564
},
{
"epoch": 2.2886075949367086,
"grad_norm": 0.5534118023507625,
"learning_rate": 2.6223508488268374e-05,
"loss": 1.3605,
"step": 565
},
{
"epoch": 2.292658227848101,
"grad_norm": 0.49932718671796916,
"learning_rate": 2.6169542751049148e-05,
"loss": 1.3831,
"step": 566
},
{
"epoch": 2.2967088607594937,
"grad_norm": 0.534977294085128,
"learning_rate": 2.6115527325180754e-05,
"loss": 1.3555,
"step": 567
},
{
"epoch": 2.300759493670886,
"grad_norm": 0.49278062943010353,
"learning_rate": 2.606146264569603e-05,
"loss": 1.3589,
"step": 568
},
{
"epoch": 2.3048101265822787,
"grad_norm": 0.51589107542057,
"learning_rate": 2.6007349148024447e-05,
"loss": 1.3654,
"step": 569
},
{
"epoch": 2.3088607594936708,
"grad_norm": 0.49748338146557586,
"learning_rate": 2.5953187267988694e-05,
"loss": 1.3663,
"step": 570
},
{
"epoch": 2.3129113924050633,
"grad_norm": 0.5828347680054642,
"learning_rate": 2.5898977441801097e-05,
"loss": 1.3692,
"step": 571
},
{
"epoch": 2.316962025316456,
"grad_norm": 0.5270596039492975,
"learning_rate": 2.584472010606015e-05,
"loss": 1.3657,
"step": 572
},
{
"epoch": 2.321012658227848,
"grad_norm": 0.5068767214203126,
"learning_rate": 2.5790415697746976e-05,
"loss": 1.3643,
"step": 573
},
{
"epoch": 2.3250632911392404,
"grad_norm": 0.563914839318865,
"learning_rate": 2.5736064654221808e-05,
"loss": 1.3826,
"step": 574
},
{
"epoch": 2.329113924050633,
"grad_norm": 0.4890074216904571,
"learning_rate": 2.568166741322048e-05,
"loss": 1.3826,
"step": 575
},
{
"epoch": 2.3331645569620254,
"grad_norm": 0.6206541863603688,
"learning_rate": 2.56272244128509e-05,
"loss": 1.3406,
"step": 576
},
{
"epoch": 2.337215189873418,
"grad_norm": 0.48630665227618886,
"learning_rate": 2.55727360915895e-05,
"loss": 1.3944,
"step": 577
},
{
"epoch": 2.34126582278481,
"grad_norm": 0.5602308808042233,
"learning_rate": 2.5518202888277734e-05,
"loss": 1.3679,
"step": 578
},
{
"epoch": 2.3453164556962025,
"grad_norm": 0.5220901481938744,
"learning_rate": 2.5463625242118523e-05,
"loss": 1.3236,
"step": 579
},
{
"epoch": 2.349367088607595,
"grad_norm": 0.5515051884484689,
"learning_rate": 2.5409003592672723e-05,
"loss": 1.3378,
"step": 580
},
{
"epoch": 2.353417721518987,
"grad_norm": 0.69032010851292,
"learning_rate": 2.535433837985559e-05,
"loss": 1.3386,
"step": 581
},
{
"epoch": 2.3574683544303796,
"grad_norm": 0.5151019642222949,
"learning_rate": 2.529963004393324e-05,
"loss": 1.3717,
"step": 582
},
{
"epoch": 2.361518987341772,
"grad_norm": 0.7006026160385288,
"learning_rate": 2.524487902551908e-05,
"loss": 1.3383,
"step": 583
},
{
"epoch": 2.3655696202531646,
"grad_norm": 0.5050102629136717,
"learning_rate": 2.519008576557029e-05,
"loss": 1.3635,
"step": 584
},
{
"epoch": 2.369620253164557,
"grad_norm": 0.5915750521460854,
"learning_rate": 2.5135250705384254e-05,
"loss": 1.3676,
"step": 585
},
{
"epoch": 2.3736708860759492,
"grad_norm": 0.46568532040269967,
"learning_rate": 2.5080374286595007e-05,
"loss": 1.3105,
"step": 586
},
{
"epoch": 2.3777215189873417,
"grad_norm": 0.4477751526039323,
"learning_rate": 2.5025456951169677e-05,
"loss": 1.3884,
"step": 587
},
{
"epoch": 2.3817721518987343,
"grad_norm": 0.5308165272225487,
"learning_rate": 2.4970499141404942e-05,
"loss": 1.3629,
"step": 588
},
{
"epoch": 2.3858227848101268,
"grad_norm": 0.4960568705144957,
"learning_rate": 2.491550129992345e-05,
"loss": 1.3857,
"step": 589
},
{
"epoch": 2.389873417721519,
"grad_norm": 0.577813504119723,
"learning_rate": 2.486046386967024e-05,
"loss": 1.3967,
"step": 590
},
{
"epoch": 2.3939240506329114,
"grad_norm": 0.5013382714916023,
"learning_rate": 2.4805387293909214e-05,
"loss": 1.4013,
"step": 591
},
{
"epoch": 2.397974683544304,
"grad_norm": 0.5061178533603924,
"learning_rate": 2.4750272016219552e-05,
"loss": 1.3716,
"step": 592
},
{
"epoch": 2.4020253164556964,
"grad_norm": 0.5140639524634738,
"learning_rate": 2.4695118480492114e-05,
"loss": 1.3928,
"step": 593
},
{
"epoch": 2.4060759493670885,
"grad_norm": 0.47457349214124095,
"learning_rate": 2.4639927130925898e-05,
"loss": 1.4012,
"step": 594
},
{
"epoch": 2.410126582278481,
"grad_norm": 0.5132773680139003,
"learning_rate": 2.458469841202444e-05,
"loss": 1.3397,
"step": 595
},
{
"epoch": 2.4141772151898735,
"grad_norm": 0.48609597307976943,
"learning_rate": 2.452943276859226e-05,
"loss": 1.3583,
"step": 596
},
{
"epoch": 2.418227848101266,
"grad_norm": 0.5063791207242632,
"learning_rate": 2.447413064573125e-05,
"loss": 1.3842,
"step": 597
},
{
"epoch": 2.422278481012658,
"grad_norm": 0.4437406042134229,
"learning_rate": 2.4418792488837095e-05,
"loss": 1.3333,
"step": 598
},
{
"epoch": 2.4263291139240506,
"grad_norm": 0.6080060352478218,
"learning_rate": 2.4363418743595713e-05,
"loss": 1.3888,
"step": 599
},
{
"epoch": 2.430379746835443,
"grad_norm": 0.4970067953656869,
"learning_rate": 2.430800985597963e-05,
"loss": 1.3055,
"step": 600
},
{
"epoch": 2.4344303797468356,
"grad_norm": 0.4649625278624322,
"learning_rate": 2.4252566272244415e-05,
"loss": 1.4005,
"step": 601
},
{
"epoch": 2.4384810126582277,
"grad_norm": 0.5430208821796478,
"learning_rate": 2.4197088438925063e-05,
"loss": 1.3834,
"step": 602
},
{
"epoch": 2.44253164556962,
"grad_norm": 0.5438548845752229,
"learning_rate": 2.4141576802832417e-05,
"loss": 1.3831,
"step": 603
},
{
"epoch": 2.4465822784810127,
"grad_norm": 0.5046491785548559,
"learning_rate": 2.408603181104957e-05,
"loss": 1.3573,
"step": 604
},
{
"epoch": 2.4506329113924052,
"grad_norm": 0.6244288454605619,
"learning_rate": 2.4030453910928245e-05,
"loss": 1.3855,
"step": 605
},
{
"epoch": 2.4546835443037973,
"grad_norm": 0.48201833128748,
"learning_rate": 2.397484355008521e-05,
"loss": 1.3442,
"step": 606
},
{
"epoch": 2.45873417721519,
"grad_norm": 0.5521719590369817,
"learning_rate": 2.3919201176398662e-05,
"loss": 1.3805,
"step": 607
},
{
"epoch": 2.4627848101265823,
"grad_norm": 0.5670509901984843,
"learning_rate": 2.3863527238004633e-05,
"loss": 1.3759,
"step": 608
},
{
"epoch": 2.466835443037975,
"grad_norm": 0.48000361713751394,
"learning_rate": 2.380782218329337e-05,
"loss": 1.3616,
"step": 609
},
{
"epoch": 2.470886075949367,
"grad_norm": 0.5196774373756327,
"learning_rate": 2.3752086460905725e-05,
"loss": 1.3519,
"step": 610
},
{
"epoch": 2.4749367088607594,
"grad_norm": 0.5377927335939715,
"learning_rate": 2.3696320519729544e-05,
"loss": 1.3627,
"step": 611
},
{
"epoch": 2.478987341772152,
"grad_norm": 0.4445929230953333,
"learning_rate": 2.3640524808896045e-05,
"loss": 1.3693,
"step": 612
},
{
"epoch": 2.4830379746835445,
"grad_norm": 0.4821249793310173,
"learning_rate": 2.3584699777776222e-05,
"loss": 1.333,
"step": 613
},
{
"epoch": 2.4870886075949366,
"grad_norm": 0.5062933560086222,
"learning_rate": 2.3528845875977195e-05,
"loss": 1.3587,
"step": 614
},
{
"epoch": 2.491139240506329,
"grad_norm": 0.5675878908561979,
"learning_rate": 2.3472963553338614e-05,
"loss": 1.3728,
"step": 615
},
{
"epoch": 2.4951898734177216,
"grad_norm": 0.519474022215299,
"learning_rate": 2.341705325992901e-05,
"loss": 1.3964,
"step": 616
},
{
"epoch": 2.499240506329114,
"grad_norm": 0.640843359001659,
"learning_rate": 2.336111544604222e-05,
"loss": 1.381,
"step": 617
},
{
"epoch": 2.503291139240506,
"grad_norm": 0.43215741948326625,
"learning_rate": 2.33051505621937e-05,
"loss": 1.3593,
"step": 618
},
{
"epoch": 2.5073417721518987,
"grad_norm": 0.649123773548095,
"learning_rate": 2.324915905911693e-05,
"loss": 1.3346,
"step": 619
},
{
"epoch": 2.511392405063291,
"grad_norm": 0.5126124834001767,
"learning_rate": 2.319314138775977e-05,
"loss": 1.3519,
"step": 620
},
{
"epoch": 2.5154430379746833,
"grad_norm": 0.5883071615163765,
"learning_rate": 2.3137097999280856e-05,
"loss": 1.3971,
"step": 621
},
{
"epoch": 2.519493670886076,
"grad_norm": 0.569421856014863,
"learning_rate": 2.308102934504593e-05,
"loss": 1.3478,
"step": 622
},
{
"epoch": 2.5235443037974683,
"grad_norm": 0.4506569817976574,
"learning_rate": 2.3024935876624222e-05,
"loss": 1.3606,
"step": 623
},
{
"epoch": 2.527594936708861,
"grad_norm": 0.5606377778407629,
"learning_rate": 2.2968818045784813e-05,
"loss": 1.3419,
"step": 624
},
{
"epoch": 2.5316455696202533,
"grad_norm": 0.5417549002524105,
"learning_rate": 2.2912676304493006e-05,
"loss": 1.3187,
"step": 625
},
{
"epoch": 2.5356962025316454,
"grad_norm": 0.7217930610319852,
"learning_rate": 2.2856511104906668e-05,
"loss": 1.4344,
"step": 626
},
{
"epoch": 2.539746835443038,
"grad_norm": 0.4894005593125535,
"learning_rate": 2.2800322899372586e-05,
"loss": 1.3438,
"step": 627
},
{
"epoch": 2.5437974683544304,
"grad_norm": 0.5287762361939871,
"learning_rate": 2.2744112140422844e-05,
"loss": 1.3551,
"step": 628
},
{
"epoch": 2.547848101265823,
"grad_norm": 0.6447063293725471,
"learning_rate": 2.2687879280771177e-05,
"loss": 1.3767,
"step": 629
},
{
"epoch": 2.5518987341772155,
"grad_norm": 0.5694309458116713,
"learning_rate": 2.26316247733093e-05,
"loss": 1.3649,
"step": 630
},
{
"epoch": 2.5559493670886075,
"grad_norm": 0.5461660026022614,
"learning_rate": 2.257534907110328e-05,
"loss": 1.3975,
"step": 631
},
{
"epoch": 2.56,
"grad_norm": 0.49311372881290866,
"learning_rate": 2.2519052627389882e-05,
"loss": 1.3231,
"step": 632
},
{
"epoch": 2.5640506329113926,
"grad_norm": 0.6029249291091654,
"learning_rate": 2.246273589557294e-05,
"loss": 1.3868,
"step": 633
},
{
"epoch": 2.5681012658227846,
"grad_norm": 0.539568472620005,
"learning_rate": 2.240639932921966e-05,
"loss": 1.3288,
"step": 634
},
{
"epoch": 2.572151898734177,
"grad_norm": 0.5782714566057904,
"learning_rate": 2.2350043382056995e-05,
"loss": 1.3888,
"step": 635
},
{
"epoch": 2.5762025316455697,
"grad_norm": 0.5007187656812244,
"learning_rate": 2.2293668507968015e-05,
"loss": 1.3465,
"step": 636
},
{
"epoch": 2.580253164556962,
"grad_norm": 0.5873556118598268,
"learning_rate": 2.2237275160988186e-05,
"loss": 1.3637,
"step": 637
},
{
"epoch": 2.5843037974683547,
"grad_norm": 0.49790216823250455,
"learning_rate": 2.2180863795301787e-05,
"loss": 1.364,
"step": 638
},
{
"epoch": 2.5883544303797468,
"grad_norm": 0.5203878773632066,
"learning_rate": 2.212443486523819e-05,
"loss": 1.3543,
"step": 639
},
{
"epoch": 2.5924050632911393,
"grad_norm": 0.5272820645350005,
"learning_rate": 2.2067988825268243e-05,
"loss": 1.3977,
"step": 640
},
{
"epoch": 2.596455696202532,
"grad_norm": 0.509875089392387,
"learning_rate": 2.2011526130000596e-05,
"loss": 1.3894,
"step": 641
},
{
"epoch": 2.600506329113924,
"grad_norm": 0.4893029196976772,
"learning_rate": 2.1955047234178038e-05,
"loss": 1.353,
"step": 642
},
{
"epoch": 2.6045569620253164,
"grad_norm": 0.6294279919135927,
"learning_rate": 2.1898552592673825e-05,
"loss": 1.3466,
"step": 643
},
{
"epoch": 2.608607594936709,
"grad_norm": 0.46474033470537707,
"learning_rate": 2.184204266048803e-05,
"loss": 1.3663,
"step": 644
},
{
"epoch": 2.6126582278481014,
"grad_norm": 0.4293416091791631,
"learning_rate": 2.1785517892743887e-05,
"loss": 1.3204,
"step": 645
},
{
"epoch": 2.616708860759494,
"grad_norm": 0.5326955040626106,
"learning_rate": 2.17289787446841e-05,
"loss": 1.3847,
"step": 646
},
{
"epoch": 2.620759493670886,
"grad_norm": 0.45759570055148363,
"learning_rate": 2.1672425671667198e-05,
"loss": 1.4166,
"step": 647
},
{
"epoch": 2.6248101265822785,
"grad_norm": 1.315876474513302,
"learning_rate": 2.161585912916385e-05,
"loss": 1.3613,
"step": 648
},
{
"epoch": 2.628860759493671,
"grad_norm": 0.5078909914347068,
"learning_rate": 2.1559279572753214e-05,
"loss": 1.3508,
"step": 649
},
{
"epoch": 2.632911392405063,
"grad_norm": 0.44390668767523317,
"learning_rate": 2.1502687458119268e-05,
"loss": 1.3623,
"step": 650
},
{
"epoch": 2.6369620253164556,
"grad_norm": 0.5197350076344358,
"learning_rate": 2.1446083241047116e-05,
"loss": 1.399,
"step": 651
},
{
"epoch": 2.641012658227848,
"grad_norm": 0.5147661259879136,
"learning_rate": 2.1389467377419333e-05,
"loss": 1.3347,
"step": 652
},
{
"epoch": 2.6450632911392407,
"grad_norm": 0.4592590449870372,
"learning_rate": 2.133284032321232e-05,
"loss": 1.3724,
"step": 653
},
{
"epoch": 2.649113924050633,
"grad_norm": 0.4873517721177637,
"learning_rate": 2.1276202534492566e-05,
"loss": 1.394,
"step": 654
},
{
"epoch": 2.6531645569620252,
"grad_norm": 0.5065560939497591,
"learning_rate": 2.121955446741306e-05,
"loss": 1.3723,
"step": 655
},
{
"epoch": 2.6572151898734178,
"grad_norm": 0.45469258749675684,
"learning_rate": 2.1162896578209517e-05,
"loss": 1.4001,
"step": 656
},
{
"epoch": 2.6612658227848103,
"grad_norm": 0.4070370251642974,
"learning_rate": 2.1106229323196813e-05,
"loss": 1.3017,
"step": 657
},
{
"epoch": 2.6653164556962023,
"grad_norm": 0.5100097179611837,
"learning_rate": 2.1049553158765214e-05,
"loss": 1.3983,
"step": 658
},
{
"epoch": 2.669367088607595,
"grad_norm": 0.46516764171183955,
"learning_rate": 2.0992868541376764e-05,
"loss": 1.3807,
"step": 659
},
{
"epoch": 2.6734177215189874,
"grad_norm": 0.4658796518899716,
"learning_rate": 2.093617592756158e-05,
"loss": 1.3626,
"step": 660
},
{
"epoch": 2.67746835443038,
"grad_norm": 0.5017011936095186,
"learning_rate": 2.0879475773914167e-05,
"loss": 1.388,
"step": 661
},
{
"epoch": 2.6815189873417724,
"grad_norm": 0.4848442183124176,
"learning_rate": 2.082276853708978e-05,
"loss": 1.3827,
"step": 662
},
{
"epoch": 2.6855696202531645,
"grad_norm": 0.4579711799559738,
"learning_rate": 2.076605467380071e-05,
"loss": 1.3836,
"step": 663
},
{
"epoch": 2.689620253164557,
"grad_norm": 0.48882820795597665,
"learning_rate": 2.0709334640812613e-05,
"loss": 1.3742,
"step": 664
},
{
"epoch": 2.6936708860759495,
"grad_norm": 0.45788224294385377,
"learning_rate": 2.0652608894940824e-05,
"loss": 1.3528,
"step": 665
},
{
"epoch": 2.6977215189873416,
"grad_norm": 0.4575872827159122,
"learning_rate": 2.0595877893046722e-05,
"loss": 1.3258,
"step": 666
},
{
"epoch": 2.701772151898734,
"grad_norm": 0.4668309329139266,
"learning_rate": 2.0539142092033985e-05,
"loss": 1.3504,
"step": 667
},
{
"epoch": 2.7058227848101266,
"grad_norm": 0.48956394769910566,
"learning_rate": 2.048240194884496e-05,
"loss": 1.3416,
"step": 668
},
{
"epoch": 2.709873417721519,
"grad_norm": 0.45464060516929933,
"learning_rate": 2.042565792045695e-05,
"loss": 1.3372,
"step": 669
},
{
"epoch": 2.7139240506329116,
"grad_norm": 0.582719640878663,
"learning_rate": 2.036891046387857e-05,
"loss": 1.3425,
"step": 670
},
{
"epoch": 2.7179746835443037,
"grad_norm": 0.49775567071406507,
"learning_rate": 2.0312160036146036e-05,
"loss": 1.3794,
"step": 671
},
{
"epoch": 2.7220253164556962,
"grad_norm": 0.5346025532810886,
"learning_rate": 2.025540709431948e-05,
"loss": 1.3811,
"step": 672
},
{
"epoch": 2.7260759493670887,
"grad_norm": 0.47780944890492116,
"learning_rate": 2.0198652095479298e-05,
"loss": 1.3865,
"step": 673
},
{
"epoch": 2.730126582278481,
"grad_norm": 0.4920528650774528,
"learning_rate": 2.014189549672245e-05,
"loss": 1.366,
"step": 674
},
{
"epoch": 2.7341772151898733,
"grad_norm": 0.42732735977895653,
"learning_rate": 2.0085137755158776e-05,
"loss": 1.3575,
"step": 675
},
{
"epoch": 2.738227848101266,
"grad_norm": 0.466694780282947,
"learning_rate": 2.0028379327907327e-05,
"loss": 1.352,
"step": 676
},
{
"epoch": 2.7422784810126584,
"grad_norm": 0.5324414916619556,
"learning_rate": 1.9971620672092676e-05,
"loss": 1.3894,
"step": 677
},
{
"epoch": 2.746329113924051,
"grad_norm": 0.46013071376924036,
"learning_rate": 1.991486224484123e-05,
"loss": 1.3609,
"step": 678
},
{
"epoch": 2.750379746835443,
"grad_norm": 0.46770631020616743,
"learning_rate": 1.985810450327756e-05,
"loss": 1.3594,
"step": 679
},
{
"epoch": 2.7544303797468355,
"grad_norm": 0.5788720640719298,
"learning_rate": 1.9801347904520706e-05,
"loss": 1.3712,
"step": 680
},
{
"epoch": 2.758481012658228,
"grad_norm": 0.48789100450376954,
"learning_rate": 1.974459290568053e-05,
"loss": 1.3811,
"step": 681
},
{
"epoch": 2.76253164556962,
"grad_norm": 0.5046154952266603,
"learning_rate": 1.968783996385397e-05,
"loss": 1.3928,
"step": 682
},
{
"epoch": 2.7665822784810126,
"grad_norm": 0.5918145177138942,
"learning_rate": 1.963108953612143e-05,
"loss": 1.3847,
"step": 683
},
{
"epoch": 2.770632911392405,
"grad_norm": 0.5002293628038581,
"learning_rate": 1.9574342079543056e-05,
"loss": 1.3643,
"step": 684
},
{
"epoch": 2.7746835443037976,
"grad_norm": 0.6643693678038608,
"learning_rate": 1.9517598051155046e-05,
"loss": 1.3615,
"step": 685
},
{
"epoch": 2.77873417721519,
"grad_norm": 0.4699047969368661,
"learning_rate": 1.9460857907966025e-05,
"loss": 1.3556,
"step": 686
},
{
"epoch": 2.782784810126582,
"grad_norm": 0.5874804779899586,
"learning_rate": 1.9404122106953285e-05,
"loss": 1.338,
"step": 687
},
{
"epoch": 2.7868354430379747,
"grad_norm": 0.5528417380352973,
"learning_rate": 1.9347391105059176e-05,
"loss": 1.3563,
"step": 688
},
{
"epoch": 2.790886075949367,
"grad_norm": 0.5228764655192333,
"learning_rate": 1.92906653591874e-05,
"loss": 1.3714,
"step": 689
},
{
"epoch": 2.7949367088607593,
"grad_norm": 0.5200925564749559,
"learning_rate": 1.9233945326199295e-05,
"loss": 1.3148,
"step": 690
},
{
"epoch": 2.798987341772152,
"grad_norm": 0.5894413957895569,
"learning_rate": 1.917723146291022e-05,
"loss": 1.348,
"step": 691
},
{
"epoch": 2.8030379746835443,
"grad_norm": 0.5146990432672198,
"learning_rate": 1.912052422608584e-05,
"loss": 1.3691,
"step": 692
},
{
"epoch": 2.807088607594937,
"grad_norm": 0.5020219524310903,
"learning_rate": 1.9063824072438428e-05,
"loss": 1.3517,
"step": 693
},
{
"epoch": 2.8111392405063294,
"grad_norm": 0.512147833913391,
"learning_rate": 1.9007131458623246e-05,
"loss": 1.3716,
"step": 694
},
{
"epoch": 2.8151898734177214,
"grad_norm": 0.4926451125070377,
"learning_rate": 1.895044684123479e-05,
"loss": 1.4015,
"step": 695
},
{
"epoch": 2.819240506329114,
"grad_norm": 0.457156388582585,
"learning_rate": 1.8893770676803194e-05,
"loss": 1.3909,
"step": 696
},
{
"epoch": 2.8232911392405065,
"grad_norm": 0.5309143582708816,
"learning_rate": 1.8837103421790486e-05,
"loss": 1.3795,
"step": 697
},
{
"epoch": 2.8273417721518985,
"grad_norm": 0.4849641246827914,
"learning_rate": 1.8780445532586952e-05,
"loss": 1.3579,
"step": 698
},
{
"epoch": 2.831392405063291,
"grad_norm": 0.5274459665605216,
"learning_rate": 1.872379746550743e-05,
"loss": 1.3322,
"step": 699
},
{
"epoch": 2.8354430379746836,
"grad_norm": 0.5473942979229001,
"learning_rate": 1.866715967678769e-05,
"loss": 1.3754,
"step": 700
},
{
"epoch": 2.839493670886076,
"grad_norm": 0.4375893498045572,
"learning_rate": 1.861053262258067e-05,
"loss": 1.3472,
"step": 701
},
{
"epoch": 2.8435443037974686,
"grad_norm": 0.5355251467783154,
"learning_rate": 1.8553916758952897e-05,
"loss": 1.4121,
"step": 702
},
{
"epoch": 2.8475949367088607,
"grad_norm": 0.46385166409103157,
"learning_rate": 1.8497312541880735e-05,
"loss": 1.3683,
"step": 703
},
{
"epoch": 2.851645569620253,
"grad_norm": 0.48895102591265643,
"learning_rate": 1.8440720427246786e-05,
"loss": 1.3628,
"step": 704
},
{
"epoch": 2.8556962025316457,
"grad_norm": 0.522310241855455,
"learning_rate": 1.8384140870836157e-05,
"loss": 1.3738,
"step": 705
},
{
"epoch": 2.8597468354430378,
"grad_norm": 0.4322128501874774,
"learning_rate": 1.8327574328332806e-05,
"loss": 1.3593,
"step": 706
},
{
"epoch": 2.8637974683544303,
"grad_norm": 0.4895822550252396,
"learning_rate": 1.8271021255315906e-05,
"loss": 1.3464,
"step": 707
},
{
"epoch": 2.867848101265823,
"grad_norm": 0.5006144800664172,
"learning_rate": 1.8214482107256117e-05,
"loss": 1.4042,
"step": 708
},
{
"epoch": 2.8718987341772153,
"grad_norm": 0.46795035265840706,
"learning_rate": 1.8157957339511968e-05,
"loss": 1.3605,
"step": 709
},
{
"epoch": 2.875949367088608,
"grad_norm": 0.5908811064528944,
"learning_rate": 1.8101447407326182e-05,
"loss": 1.3762,
"step": 710
},
{
"epoch": 2.88,
"grad_norm": 0.5298504786252352,
"learning_rate": 1.8044952765821966e-05,
"loss": 1.338,
"step": 711
},
{
"epoch": 2.8840506329113924,
"grad_norm": 0.4403847715033649,
"learning_rate": 1.7988473869999407e-05,
"loss": 1.3426,
"step": 712
},
{
"epoch": 2.888101265822785,
"grad_norm": 0.6520515446276053,
"learning_rate": 1.7932011174731764e-05,
"loss": 1.4322,
"step": 713
},
{
"epoch": 2.892151898734177,
"grad_norm": 0.47931855495666337,
"learning_rate": 1.7875565134761817e-05,
"loss": 1.3803,
"step": 714
},
{
"epoch": 2.8962025316455695,
"grad_norm": 0.5494826091875872,
"learning_rate": 1.7819136204698226e-05,
"loss": 1.3575,
"step": 715
},
{
"epoch": 2.900253164556962,
"grad_norm": 0.4679418757798025,
"learning_rate": 1.776272483901182e-05,
"loss": 1.376,
"step": 716
},
{
"epoch": 2.9043037974683545,
"grad_norm": 0.52453789103102,
"learning_rate": 1.7706331492031995e-05,
"loss": 1.389,
"step": 717
},
{
"epoch": 2.908354430379747,
"grad_norm": 0.5147392134548955,
"learning_rate": 1.764995661794301e-05,
"loss": 1.3333,
"step": 718
},
{
"epoch": 2.912405063291139,
"grad_norm": 0.5311938690227374,
"learning_rate": 1.759360067078035e-05,
"loss": 1.3652,
"step": 719
},
{
"epoch": 2.9164556962025316,
"grad_norm": 0.5280707181481894,
"learning_rate": 1.7537264104427064e-05,
"loss": 1.3554,
"step": 720
},
{
"epoch": 2.920506329113924,
"grad_norm": 0.589726995501214,
"learning_rate": 1.748094737261012e-05,
"loss": 1.3596,
"step": 721
},
{
"epoch": 2.9245569620253162,
"grad_norm": 0.533176235802329,
"learning_rate": 1.7424650928896726e-05,
"loss": 1.361,
"step": 722
},
{
"epoch": 2.9286075949367087,
"grad_norm": 0.5146061671554711,
"learning_rate": 1.7368375226690712e-05,
"loss": 1.3841,
"step": 723
},
{
"epoch": 2.9326582278481013,
"grad_norm": 0.5527457612199279,
"learning_rate": 1.731212071922883e-05,
"loss": 1.3688,
"step": 724
},
{
"epoch": 2.9367088607594938,
"grad_norm": 0.44482398427382,
"learning_rate": 1.7255887859577156e-05,
"loss": 1.3612,
"step": 725
},
{
"epoch": 2.9407594936708863,
"grad_norm": 0.6527024430737688,
"learning_rate": 1.7199677100627427e-05,
"loss": 1.4448,
"step": 726
},
{
"epoch": 2.9448101265822784,
"grad_norm": 0.43714728256303365,
"learning_rate": 1.7143488895093343e-05,
"loss": 1.3492,
"step": 727
},
{
"epoch": 2.948860759493671,
"grad_norm": 0.4498447995189133,
"learning_rate": 1.7087323695506994e-05,
"loss": 1.373,
"step": 728
},
{
"epoch": 2.9529113924050634,
"grad_norm": 0.5402338253230159,
"learning_rate": 1.7031181954215194e-05,
"loss": 1.3319,
"step": 729
},
{
"epoch": 2.9569620253164555,
"grad_norm": 0.450972957151404,
"learning_rate": 1.6975064123375788e-05,
"loss": 1.3603,
"step": 730
},
{
"epoch": 2.961012658227848,
"grad_norm": 0.5087099591224562,
"learning_rate": 1.6918970654954084e-05,
"loss": 1.3686,
"step": 731
},
{
"epoch": 2.9650632911392405,
"grad_norm": 0.47460835285388125,
"learning_rate": 1.686290200071915e-05,
"loss": 1.3475,
"step": 732
},
{
"epoch": 2.969113924050633,
"grad_norm": 0.46719231906682135,
"learning_rate": 1.6806858612240234e-05,
"loss": 1.3435,
"step": 733
},
{
"epoch": 2.9731645569620255,
"grad_norm": 0.5031819921853149,
"learning_rate": 1.6750840940883078e-05,
"loss": 1.3688,
"step": 734
},
{
"epoch": 2.9772151898734176,
"grad_norm": 0.4938814615920837,
"learning_rate": 1.6694849437806305e-05,
"loss": 1.3529,
"step": 735
},
{
"epoch": 2.98126582278481,
"grad_norm": 0.4310710256321563,
"learning_rate": 1.663888455395778e-05,
"loss": 1.326,
"step": 736
},
{
"epoch": 2.9853164556962026,
"grad_norm": 0.47846231936156836,
"learning_rate": 1.6582946740070995e-05,
"loss": 1.3841,
"step": 737
},
{
"epoch": 2.9893670886075947,
"grad_norm": 0.5030138677852755,
"learning_rate": 1.6527036446661396e-05,
"loss": 1.3897,
"step": 738
},
{
"epoch": 2.993417721518987,
"grad_norm": 0.5101424378682257,
"learning_rate": 1.6471154124022818e-05,
"loss": 1.3545,
"step": 739
},
{
"epoch": 2.9974683544303797,
"grad_norm": 0.4588170453313183,
"learning_rate": 1.6415300222223788e-05,
"loss": 1.373,
"step": 740
},
{
"epoch": 3.0015189873417722,
"grad_norm": 0.5498728164802428,
"learning_rate": 1.6359475191103958e-05,
"loss": 1.3975,
"step": 741
},
{
"epoch": 3.0055696202531648,
"grad_norm": 0.7318135053062929,
"learning_rate": 1.6303679480270466e-05,
"loss": 1.1844,
"step": 742
},
{
"epoch": 3.009620253164557,
"grad_norm": 0.7401310439754623,
"learning_rate": 1.624791353909428e-05,
"loss": 1.181,
"step": 743
},
{
"epoch": 3.0136708860759494,
"grad_norm": 1.148283440746444,
"learning_rate": 1.619217781670663e-05,
"loss": 1.2048,
"step": 744
},
{
"epoch": 3.017721518987342,
"grad_norm": 0.8447275755194898,
"learning_rate": 1.6136472761995373e-05,
"loss": 1.2069,
"step": 745
},
{
"epoch": 3.0217721518987344,
"grad_norm": 0.9663187393176085,
"learning_rate": 1.608079882360134e-05,
"loss": 1.2022,
"step": 746
},
{
"epoch": 3.0258227848101265,
"grad_norm": 0.7899567345200027,
"learning_rate": 1.60251564499148e-05,
"loss": 1.1848,
"step": 747
},
{
"epoch": 3.029873417721519,
"grad_norm": 0.7733332081280031,
"learning_rate": 1.596954608907176e-05,
"loss": 1.2127,
"step": 748
},
{
"epoch": 3.0339240506329115,
"grad_norm": 0.9030333005385875,
"learning_rate": 1.591396818895043e-05,
"loss": 1.1644,
"step": 749
},
{
"epoch": 3.037974683544304,
"grad_norm": 0.880698555783657,
"learning_rate": 1.585842319716759e-05,
"loss": 1.1767,
"step": 750
},
{
"epoch": 3.042025316455696,
"grad_norm": 0.86834484012468,
"learning_rate": 1.5802911561074944e-05,
"loss": 1.1844,
"step": 751
},
{
"epoch": 3.0460759493670886,
"grad_norm": 0.974206098062169,
"learning_rate": 1.5747433727755595e-05,
"loss": 1.1594,
"step": 752
},
{
"epoch": 3.050126582278481,
"grad_norm": 0.8260134857420244,
"learning_rate": 1.5691990144020376e-05,
"loss": 1.2156,
"step": 753
},
{
"epoch": 3.0541772151898736,
"grad_norm": 0.9139601365597755,
"learning_rate": 1.5636581256404297e-05,
"loss": 1.1761,
"step": 754
},
{
"epoch": 3.0582278481012657,
"grad_norm": 0.9358534197869692,
"learning_rate": 1.558120751116291e-05,
"loss": 1.1884,
"step": 755
},
{
"epoch": 3.062278481012658,
"grad_norm": 0.8277825293271527,
"learning_rate": 1.552586935426876e-05,
"loss": 1.2192,
"step": 756
},
{
"epoch": 3.0663291139240507,
"grad_norm": 0.7325931179601817,
"learning_rate": 1.547056723140774e-05,
"loss": 1.1699,
"step": 757
},
{
"epoch": 3.0703797468354432,
"grad_norm": 0.8705770750209227,
"learning_rate": 1.5415301587975565e-05,
"loss": 1.1831,
"step": 758
},
{
"epoch": 3.0744303797468353,
"grad_norm": 0.8129247635277906,
"learning_rate": 1.536007286907411e-05,
"loss": 1.1951,
"step": 759
},
{
"epoch": 3.078481012658228,
"grad_norm": 0.6563406996157911,
"learning_rate": 1.5304881519507896e-05,
"loss": 1.1645,
"step": 760
},
{
"epoch": 3.0825316455696203,
"grad_norm": 0.8519456175220197,
"learning_rate": 1.5249727983780453e-05,
"loss": 1.2052,
"step": 761
},
{
"epoch": 3.086582278481013,
"grad_norm": 0.79137795972044,
"learning_rate": 1.5194612706090786e-05,
"loss": 1.1686,
"step": 762
},
{
"epoch": 3.090632911392405,
"grad_norm": 0.7562421788859691,
"learning_rate": 1.5139536130329771e-05,
"loss": 1.2185,
"step": 763
},
{
"epoch": 3.0946835443037974,
"grad_norm": 0.7568394231325494,
"learning_rate": 1.508449870007656e-05,
"loss": 1.1706,
"step": 764
},
{
"epoch": 3.09873417721519,
"grad_norm": 0.5768007505548911,
"learning_rate": 1.5029500858595056e-05,
"loss": 1.1679,
"step": 765
},
{
"epoch": 3.1027848101265825,
"grad_norm": 0.6902669619491462,
"learning_rate": 1.4974543048830328e-05,
"loss": 1.1662,
"step": 766
},
{
"epoch": 3.1068354430379745,
"grad_norm": 0.6642773113931104,
"learning_rate": 1.4919625713405e-05,
"loss": 1.1907,
"step": 767
},
{
"epoch": 3.110886075949367,
"grad_norm": 0.5983737349045666,
"learning_rate": 1.4864749294615756e-05,
"loss": 1.1773,
"step": 768
},
{
"epoch": 3.1149367088607596,
"grad_norm": 0.6269859786536192,
"learning_rate": 1.4809914234429716e-05,
"loss": 1.2145,
"step": 769
},
{
"epoch": 3.118987341772152,
"grad_norm": 0.588759246761377,
"learning_rate": 1.4755120974480923e-05,
"loss": 1.1871,
"step": 770
},
{
"epoch": 3.123037974683544,
"grad_norm": 0.48229682367078225,
"learning_rate": 1.4700369956066771e-05,
"loss": 1.1754,
"step": 771
},
{
"epoch": 3.1270886075949367,
"grad_norm": 0.5020499560681886,
"learning_rate": 1.4645661620144413e-05,
"loss": 1.1671,
"step": 772
},
{
"epoch": 3.131139240506329,
"grad_norm": 0.5347137380011857,
"learning_rate": 1.4590996407327284e-05,
"loss": 1.1687,
"step": 773
},
{
"epoch": 3.1351898734177217,
"grad_norm": 0.5029692117875568,
"learning_rate": 1.4536374757881487e-05,
"loss": 1.1804,
"step": 774
},
{
"epoch": 3.1392405063291138,
"grad_norm": 0.5219964598885257,
"learning_rate": 1.4481797111722271e-05,
"loss": 1.1626,
"step": 775
},
{
"epoch": 3.1432911392405063,
"grad_norm": 0.5530206481948411,
"learning_rate": 1.4427263908410507e-05,
"loss": 1.1958,
"step": 776
},
{
"epoch": 3.147341772151899,
"grad_norm": 0.5102655724117371,
"learning_rate": 1.4372775587149108e-05,
"loss": 1.2114,
"step": 777
},
{
"epoch": 3.1513924050632913,
"grad_norm": 0.5272264853814373,
"learning_rate": 1.4318332586779522e-05,
"loss": 1.186,
"step": 778
},
{
"epoch": 3.1554430379746834,
"grad_norm": 0.5072251631322804,
"learning_rate": 1.4263935345778202e-05,
"loss": 1.1867,
"step": 779
},
{
"epoch": 3.159493670886076,
"grad_norm": 0.5135113383531033,
"learning_rate": 1.420958430225303e-05,
"loss": 1.2153,
"step": 780
},
{
"epoch": 3.1635443037974684,
"grad_norm": 0.4938335840571743,
"learning_rate": 1.415527989393985e-05,
"loss": 1.1736,
"step": 781
},
{
"epoch": 3.167594936708861,
"grad_norm": 0.4741225210393059,
"learning_rate": 1.410102255819891e-05,
"loss": 1.1489,
"step": 782
},
{
"epoch": 3.171645569620253,
"grad_norm": 0.4587838207766825,
"learning_rate": 1.404681273201131e-05,
"loss": 1.1734,
"step": 783
},
{
"epoch": 3.1756962025316455,
"grad_norm": 0.47461010764688993,
"learning_rate": 1.399265085197556e-05,
"loss": 1.1967,
"step": 784
},
{
"epoch": 3.179746835443038,
"grad_norm": 0.4765428501359228,
"learning_rate": 1.393853735430398e-05,
"loss": 1.2209,
"step": 785
},
{
"epoch": 3.1837974683544306,
"grad_norm": 0.4737580168734725,
"learning_rate": 1.3884472674819246e-05,
"loss": 1.2388,
"step": 786
},
{
"epoch": 3.1878481012658226,
"grad_norm": 0.5063631696485895,
"learning_rate": 1.3830457248950864e-05,
"loss": 1.1768,
"step": 787
},
{
"epoch": 3.191898734177215,
"grad_norm": 0.457701638263518,
"learning_rate": 1.377649151173163e-05,
"loss": 1.1927,
"step": 788
},
{
"epoch": 3.1959493670886077,
"grad_norm": 0.4744286498388194,
"learning_rate": 1.3722575897794181e-05,
"loss": 1.2437,
"step": 789
},
{
"epoch": 3.2,
"grad_norm": 0.49122416369860755,
"learning_rate": 1.3668710841367472e-05,
"loss": 1.1773,
"step": 790
},
{
"epoch": 3.2040506329113922,
"grad_norm": 0.5253880378579754,
"learning_rate": 1.361489677627324e-05,
"loss": 1.1961,
"step": 791
},
{
"epoch": 3.2081012658227848,
"grad_norm": 0.5021774094036692,
"learning_rate": 1.3561134135922585e-05,
"loss": 1.1605,
"step": 792
},
{
"epoch": 3.2121518987341773,
"grad_norm": 0.5683235214419485,
"learning_rate": 1.350742335331241e-05,
"loss": 1.191,
"step": 793
},
{
"epoch": 3.21620253164557,
"grad_norm": 0.4936045939364357,
"learning_rate": 1.345376486102198e-05,
"loss": 1.1784,
"step": 794
},
{
"epoch": 3.220253164556962,
"grad_norm": 0.5339803853015289,
"learning_rate": 1.3400159091209414e-05,
"loss": 1.196,
"step": 795
},
{
"epoch": 3.2243037974683544,
"grad_norm": 0.5487650825874216,
"learning_rate": 1.3346606475608216e-05,
"loss": 1.1929,
"step": 796
},
{
"epoch": 3.228354430379747,
"grad_norm": 0.49858946881945393,
"learning_rate": 1.3293107445523781e-05,
"loss": 1.1573,
"step": 797
},
{
"epoch": 3.2324050632911394,
"grad_norm": 0.5389358103848386,
"learning_rate": 1.3239662431829949e-05,
"loss": 1.169,
"step": 798
},
{
"epoch": 3.2364556962025315,
"grad_norm": 0.6001499241703325,
"learning_rate": 1.3186271864965509e-05,
"loss": 1.1986,
"step": 799
},
{
"epoch": 3.240506329113924,
"grad_norm": 0.5246546250389919,
"learning_rate": 1.3132936174930756e-05,
"loss": 1.1825,
"step": 800
},
{
"epoch": 3.2445569620253165,
"grad_norm": 0.5746473005956125,
"learning_rate": 1.3079655791283995e-05,
"loss": 1.2213,
"step": 801
},
{
"epoch": 3.248607594936709,
"grad_norm": 0.5166822255724837,
"learning_rate": 1.3026431143138108e-05,
"loss": 1.1929,
"step": 802
},
{
"epoch": 3.252658227848101,
"grad_norm": 0.4964133998313581,
"learning_rate": 1.2973262659157114e-05,
"loss": 1.1876,
"step": 803
},
{
"epoch": 3.2567088607594936,
"grad_norm": 0.5606526011019413,
"learning_rate": 1.2920150767552651e-05,
"loss": 1.1768,
"step": 804
},
{
"epoch": 3.260759493670886,
"grad_norm": 0.5516669832357598,
"learning_rate": 1.2867095896080607e-05,
"loss": 1.208,
"step": 805
},
{
"epoch": 3.2648101265822786,
"grad_norm": 0.4825063711423262,
"learning_rate": 1.2814098472037612e-05,
"loss": 1.1701,
"step": 806
},
{
"epoch": 3.2688607594936707,
"grad_norm": 0.5073750300405783,
"learning_rate": 1.276115892225764e-05,
"loss": 1.1861,
"step": 807
},
{
"epoch": 3.2729113924050632,
"grad_norm": 0.49004428463748356,
"learning_rate": 1.2708277673108555e-05,
"loss": 1.2183,
"step": 808
},
{
"epoch": 3.2769620253164558,
"grad_norm": 0.48012476826545997,
"learning_rate": 1.2655455150488649e-05,
"loss": 1.2214,
"step": 809
},
{
"epoch": 3.2810126582278483,
"grad_norm": 0.46556224609445296,
"learning_rate": 1.2602691779823272e-05,
"loss": 1.1585,
"step": 810
},
{
"epoch": 3.2850632911392403,
"grad_norm": 0.48035351965931655,
"learning_rate": 1.2549987986061355e-05,
"loss": 1.2585,
"step": 811
},
{
"epoch": 3.289113924050633,
"grad_norm": 0.4842695212758225,
"learning_rate": 1.2497344193672005e-05,
"loss": 1.1892,
"step": 812
},
{
"epoch": 3.2931645569620254,
"grad_norm": 0.45585987844914,
"learning_rate": 1.2444760826641092e-05,
"loss": 1.2025,
"step": 813
},
{
"epoch": 3.297215189873418,
"grad_norm": 0.45587361027255563,
"learning_rate": 1.2392238308467817e-05,
"loss": 1.1589,
"step": 814
},
{
"epoch": 3.30126582278481,
"grad_norm": 0.508189814801608,
"learning_rate": 1.2339777062161326e-05,
"loss": 1.182,
"step": 815
},
{
"epoch": 3.3053164556962025,
"grad_norm": 0.46094790094170046,
"learning_rate": 1.2287377510237293e-05,
"loss": 1.1742,
"step": 816
},
{
"epoch": 3.309367088607595,
"grad_norm": 0.47673705951617174,
"learning_rate": 1.2235040074714488e-05,
"loss": 1.2013,
"step": 817
},
{
"epoch": 3.3134177215189875,
"grad_norm": 0.49893199092234597,
"learning_rate": 1.2182765177111434e-05,
"loss": 1.2011,
"step": 818
},
{
"epoch": 3.3174683544303796,
"grad_norm": 0.4705121091668679,
"learning_rate": 1.213055323844297e-05,
"loss": 1.2187,
"step": 819
},
{
"epoch": 3.321518987341772,
"grad_norm": 0.44096451216694016,
"learning_rate": 1.2078404679216864e-05,
"loss": 1.1733,
"step": 820
},
{
"epoch": 3.3255696202531646,
"grad_norm": 0.5107184422366595,
"learning_rate": 1.2026319919430458e-05,
"loss": 1.21,
"step": 821
},
{
"epoch": 3.329620253164557,
"grad_norm": 0.4709351010525272,
"learning_rate": 1.1974299378567227e-05,
"loss": 1.1932,
"step": 822
},
{
"epoch": 3.333670886075949,
"grad_norm": 0.46340799583857495,
"learning_rate": 1.1922343475593462e-05,
"loss": 1.2164,
"step": 823
},
{
"epoch": 3.3377215189873417,
"grad_norm": 0.46579288603011765,
"learning_rate": 1.187045262895488e-05,
"loss": 1.1952,
"step": 824
},
{
"epoch": 3.3417721518987342,
"grad_norm": 0.48599420112659575,
"learning_rate": 1.1818627256573203e-05,
"loss": 1.187,
"step": 825
},
{
"epoch": 3.3458227848101267,
"grad_norm": 0.47935233409639044,
"learning_rate": 1.1766867775842864e-05,
"loss": 1.2113,
"step": 826
},
{
"epoch": 3.349873417721519,
"grad_norm": 0.48620713197435733,
"learning_rate": 1.1715174603627615e-05,
"loss": 1.2,
"step": 827
},
{
"epoch": 3.3539240506329113,
"grad_norm": 0.488189452416017,
"learning_rate": 1.1663548156257147e-05,
"loss": 1.1646,
"step": 828
},
{
"epoch": 3.357974683544304,
"grad_norm": 0.4594298851918262,
"learning_rate": 1.161198884952377e-05,
"loss": 1.1528,
"step": 829
},
{
"epoch": 3.3620253164556964,
"grad_norm": 0.5123726121545096,
"learning_rate": 1.1560497098679056e-05,
"loss": 1.1969,
"step": 830
},
{
"epoch": 3.3660759493670884,
"grad_norm": 0.4782521990621781,
"learning_rate": 1.1509073318430479e-05,
"loss": 1.1944,
"step": 831
},
{
"epoch": 3.370126582278481,
"grad_norm": 0.47825731213937417,
"learning_rate": 1.1457717922938116e-05,
"loss": 1.189,
"step": 832
},
{
"epoch": 3.3741772151898735,
"grad_norm": 0.4714563871820282,
"learning_rate": 1.1406431325811233e-05,
"loss": 1.1606,
"step": 833
},
{
"epoch": 3.378227848101266,
"grad_norm": 0.54060818615823,
"learning_rate": 1.135521394010506e-05,
"loss": 1.2306,
"step": 834
},
{
"epoch": 3.382278481012658,
"grad_norm": 0.49303832342856624,
"learning_rate": 1.1304066178317367e-05,
"loss": 1.1801,
"step": 835
},
{
"epoch": 3.3863291139240506,
"grad_norm": 0.46974402407904536,
"learning_rate": 1.1252988452385199e-05,
"loss": 1.174,
"step": 836
},
{
"epoch": 3.390379746835443,
"grad_norm": 0.4793450631061359,
"learning_rate": 1.1201981173681536e-05,
"loss": 1.1878,
"step": 837
},
{
"epoch": 3.3944303797468356,
"grad_norm": 0.481924383779943,
"learning_rate": 1.1151044753011991e-05,
"loss": 1.114,
"step": 838
},
{
"epoch": 3.3984810126582277,
"grad_norm": 0.4685668313632894,
"learning_rate": 1.1100179600611491e-05,
"loss": 1.1856,
"step": 839
},
{
"epoch": 3.40253164556962,
"grad_norm": 0.44618662890824595,
"learning_rate": 1.1049386126140985e-05,
"loss": 1.2039,
"step": 840
},
{
"epoch": 3.4065822784810127,
"grad_norm": 0.5070835542681827,
"learning_rate": 1.0998664738684128e-05,
"loss": 1.206,
"step": 841
},
{
"epoch": 3.410632911392405,
"grad_norm": 0.5894659479522416,
"learning_rate": 1.0948015846744e-05,
"loss": 1.2185,
"step": 842
},
{
"epoch": 3.4146835443037973,
"grad_norm": 0.439635247014645,
"learning_rate": 1.0897439858239832e-05,
"loss": 1.1524,
"step": 843
},
{
"epoch": 3.41873417721519,
"grad_norm": 0.4676108961535113,
"learning_rate": 1.0846937180503652e-05,
"loss": 1.232,
"step": 844
},
{
"epoch": 3.4227848101265823,
"grad_norm": 0.5792101461228593,
"learning_rate": 1.0796508220277117e-05,
"loss": 1.2059,
"step": 845
},
{
"epoch": 3.426835443037975,
"grad_norm": 0.46136475173119734,
"learning_rate": 1.0746153383708107e-05,
"loss": 1.1889,
"step": 846
},
{
"epoch": 3.430886075949367,
"grad_norm": 0.4702521974194089,
"learning_rate": 1.0695873076347579e-05,
"loss": 1.1788,
"step": 847
},
{
"epoch": 3.4349367088607594,
"grad_norm": 0.5350633887364539,
"learning_rate": 1.0645667703146205e-05,
"loss": 1.2268,
"step": 848
},
{
"epoch": 3.438987341772152,
"grad_norm": 0.46852767286925967,
"learning_rate": 1.0595537668451161e-05,
"loss": 1.142,
"step": 849
},
{
"epoch": 3.4430379746835444,
"grad_norm": 0.4345168222315947,
"learning_rate": 1.0545483376002854e-05,
"loss": 1.1471,
"step": 850
},
{
"epoch": 3.4470886075949365,
"grad_norm": 0.5141905359017521,
"learning_rate": 1.0495505228931676e-05,
"loss": 1.1949,
"step": 851
},
{
"epoch": 3.451139240506329,
"grad_norm": 0.5143393659820328,
"learning_rate": 1.044560362975474e-05,
"loss": 1.1809,
"step": 852
},
{
"epoch": 3.4551898734177215,
"grad_norm": 0.4437882872703027,
"learning_rate": 1.0395778980372695e-05,
"loss": 1.1833,
"step": 853
},
{
"epoch": 3.459240506329114,
"grad_norm": 0.4547613145646534,
"learning_rate": 1.0346031682066381e-05,
"loss": 1.2131,
"step": 854
},
{
"epoch": 3.463291139240506,
"grad_norm": 0.449818963867369,
"learning_rate": 1.0296362135493724e-05,
"loss": 1.1944,
"step": 855
},
{
"epoch": 3.4673417721518986,
"grad_norm": 0.4707760524445298,
"learning_rate": 1.0246770740686422e-05,
"loss": 1.1852,
"step": 856
},
{
"epoch": 3.471392405063291,
"grad_norm": 0.4479226647418359,
"learning_rate": 1.0197257897046743e-05,
"loss": 1.188,
"step": 857
},
{
"epoch": 3.4754430379746837,
"grad_norm": 0.44421296616235334,
"learning_rate": 1.014782400334433e-05,
"loss": 1.1639,
"step": 858
},
{
"epoch": 3.479493670886076,
"grad_norm": 0.4685332380552682,
"learning_rate": 1.009846945771296e-05,
"loss": 1.2063,
"step": 859
},
{
"epoch": 3.4835443037974683,
"grad_norm": 0.4622531239826464,
"learning_rate": 1.0049194657647363e-05,
"loss": 1.2388,
"step": 860
},
{
"epoch": 3.487594936708861,
"grad_norm": 0.4884993275501679,
"learning_rate": 1.0000000000000006e-05,
"loss": 1.2087,
"step": 861
},
{
"epoch": 3.4916455696202533,
"grad_norm": 0.46531534403228825,
"learning_rate": 9.950885880977891e-06,
"loss": 1.1504,
"step": 862
},
{
"epoch": 3.4956962025316454,
"grad_norm": 0.44311483783248734,
"learning_rate": 9.901852696139382e-06,
"loss": 1.2113,
"step": 863
},
{
"epoch": 3.499746835443038,
"grad_norm": 0.4535812525978502,
"learning_rate": 9.852900840391027e-06,
"loss": 1.1606,
"step": 864
},
{
"epoch": 3.5037974683544304,
"grad_norm": 0.4734674237289631,
"learning_rate": 9.804030707984313e-06,
"loss": 1.1721,
"step": 865
},
{
"epoch": 3.507848101265823,
"grad_norm": 0.4462774659926807,
"learning_rate": 9.755242692512599e-06,
"loss": 1.1739,
"step": 866
},
{
"epoch": 3.5118987341772154,
"grad_norm": 0.44692425955377624,
"learning_rate": 9.70653718690782e-06,
"loss": 1.1948,
"step": 867
},
{
"epoch": 3.5159493670886075,
"grad_norm": 0.45971817968348894,
"learning_rate": 9.657914583437454e-06,
"loss": 1.2079,
"step": 868
},
{
"epoch": 3.52,
"grad_norm": 0.5275413297641675,
"learning_rate": 9.609375273701246e-06,
"loss": 1.2194,
"step": 869
},
{
"epoch": 3.5240506329113925,
"grad_norm": 0.43784953141237026,
"learning_rate": 9.560919648628133e-06,
"loss": 1.1794,
"step": 870
},
{
"epoch": 3.5281012658227846,
"grad_norm": 0.47383425965244025,
"learning_rate": 9.512548098473047e-06,
"loss": 1.1798,
"step": 871
},
{
"epoch": 3.532151898734177,
"grad_norm": 0.46227193525141175,
"learning_rate": 9.464261012813825e-06,
"loss": 1.1895,
"step": 872
},
{
"epoch": 3.5362025316455696,
"grad_norm": 0.47259936457284035,
"learning_rate": 9.416058780547987e-06,
"loss": 1.1798,
"step": 873
},
{
"epoch": 3.540253164556962,
"grad_norm": 0.4418738131614512,
"learning_rate": 9.367941789889714e-06,
"loss": 1.1769,
"step": 874
},
{
"epoch": 3.5443037974683547,
"grad_norm": 0.46017709071091956,
"learning_rate": 9.319910428366607e-06,
"loss": 1.223,
"step": 875
},
{
"epoch": 3.5483544303797467,
"grad_norm": 0.46306910970972753,
"learning_rate": 9.271965082816667e-06,
"loss": 1.1755,
"step": 876
},
{
"epoch": 3.5524050632911393,
"grad_norm": 0.4926279436488095,
"learning_rate": 9.224106139385111e-06,
"loss": 1.1833,
"step": 877
},
{
"epoch": 3.5564556962025318,
"grad_norm": 0.4730655594558266,
"learning_rate": 9.176333983521291e-06,
"loss": 1.1825,
"step": 878
},
{
"epoch": 3.560506329113924,
"grad_norm": 0.44736366541760053,
"learning_rate": 9.12864899997558e-06,
"loss": 1.1862,
"step": 879
},
{
"epoch": 3.5645569620253164,
"grad_norm": 0.4462938617941005,
"learning_rate": 9.08105157279628e-06,
"loss": 1.1918,
"step": 880
},
{
"epoch": 3.568607594936709,
"grad_norm": 0.4514315835008869,
"learning_rate": 9.03354208532653e-06,
"loss": 1.2047,
"step": 881
},
{
"epoch": 3.5726582278481014,
"grad_norm": 0.49164364998372273,
"learning_rate": 8.986120920201205e-06,
"loss": 1.2255,
"step": 882
},
{
"epoch": 3.576708860759494,
"grad_norm": 0.4715100053040883,
"learning_rate": 8.938788459343852e-06,
"loss": 1.1803,
"step": 883
},
{
"epoch": 3.580759493670886,
"grad_norm": 0.4672091262988023,
"learning_rate": 8.8915450839636e-06,
"loss": 1.1905,
"step": 884
},
{
"epoch": 3.5848101265822785,
"grad_norm": 0.4915315187325219,
"learning_rate": 8.844391174552116e-06,
"loss": 1.2341,
"step": 885
},
{
"epoch": 3.588860759493671,
"grad_norm": 0.47862012880895344,
"learning_rate": 8.797327110880479e-06,
"loss": 1.1899,
"step": 886
},
{
"epoch": 3.592911392405063,
"grad_norm": 0.4608213219311107,
"learning_rate": 8.750353271996206e-06,
"loss": 1.1852,
"step": 887
},
{
"epoch": 3.5969620253164556,
"grad_norm": 0.4531479265754058,
"learning_rate": 8.703470036220132e-06,
"loss": 1.1537,
"step": 888
},
{
"epoch": 3.601012658227848,
"grad_norm": 0.4653013691526645,
"learning_rate": 8.656677781143394e-06,
"loss": 1.2008,
"step": 889
},
{
"epoch": 3.6050632911392406,
"grad_norm": 0.48810952672996405,
"learning_rate": 8.609976883624377e-06,
"loss": 1.1871,
"step": 890
},
{
"epoch": 3.609113924050633,
"grad_norm": 0.47207989882285856,
"learning_rate": 8.563367719785698e-06,
"loss": 1.2123,
"step": 891
},
{
"epoch": 3.613164556962025,
"grad_norm": 0.473056944497011,
"learning_rate": 8.516850665011138e-06,
"loss": 1.2205,
"step": 892
},
{
"epoch": 3.6172151898734177,
"grad_norm": 0.4668591101989504,
"learning_rate": 8.47042609394269e-06,
"loss": 1.2189,
"step": 893
},
{
"epoch": 3.6212658227848102,
"grad_norm": 0.4721470255239711,
"learning_rate": 8.424094380477432e-06,
"loss": 1.1764,
"step": 894
},
{
"epoch": 3.6253164556962023,
"grad_norm": 0.4681266412395909,
"learning_rate": 8.37785589776465e-06,
"loss": 1.1697,
"step": 895
},
{
"epoch": 3.629367088607595,
"grad_norm": 0.43580087554756336,
"learning_rate": 8.331711018202694e-06,
"loss": 1.1484,
"step": 896
},
{
"epoch": 3.6334177215189873,
"grad_norm": 0.4704859317414002,
"learning_rate": 8.285660113436104e-06,
"loss": 1.206,
"step": 897
},
{
"epoch": 3.63746835443038,
"grad_norm": 0.4728675612951488,
"learning_rate": 8.239703554352527e-06,
"loss": 1.1764,
"step": 898
},
{
"epoch": 3.6415189873417724,
"grad_norm": 0.4310100623893338,
"learning_rate": 8.193841711079775e-06,
"loss": 1.2127,
"step": 899
},
{
"epoch": 3.6455696202531644,
"grad_norm": 0.4593650109525007,
"learning_rate": 8.148074952982828e-06,
"loss": 1.1801,
"step": 900
},
{
"epoch": 3.649620253164557,
"grad_norm": 0.4543625641632022,
"learning_rate": 8.102403648660859e-06,
"loss": 1.1989,
"step": 901
},
{
"epoch": 3.6536708860759495,
"grad_norm": 0.4342838441956303,
"learning_rate": 8.056828165944282e-06,
"loss": 1.1951,
"step": 902
},
{
"epoch": 3.6577215189873415,
"grad_norm": 0.4572131339012011,
"learning_rate": 8.011348871891762e-06,
"loss": 1.2158,
"step": 903
},
{
"epoch": 3.661772151898734,
"grad_norm": 0.4542056016988432,
"learning_rate": 7.965966132787287e-06,
"loss": 1.173,
"step": 904
},
{
"epoch": 3.6658227848101266,
"grad_norm": 0.4710732010355873,
"learning_rate": 7.920680314137189e-06,
"loss": 1.2126,
"step": 905
},
{
"epoch": 3.669873417721519,
"grad_norm": 0.4261185485019166,
"learning_rate": 7.875491780667246e-06,
"loss": 1.2012,
"step": 906
},
{
"epoch": 3.6739240506329116,
"grad_norm": 0.44455493579396504,
"learning_rate": 7.830400896319667e-06,
"loss": 1.219,
"step": 907
},
{
"epoch": 3.6779746835443037,
"grad_norm": 0.4813522058749255,
"learning_rate": 7.785408024250259e-06,
"loss": 1.1552,
"step": 908
},
{
"epoch": 3.682025316455696,
"grad_norm": 0.46465774385138625,
"learning_rate": 7.74051352682542e-06,
"loss": 1.2066,
"step": 909
},
{
"epoch": 3.6860759493670887,
"grad_norm": 0.42186898899121494,
"learning_rate": 7.695717765619257e-06,
"loss": 1.1532,
"step": 910
},
{
"epoch": 3.690126582278481,
"grad_norm": 0.44984796848572456,
"learning_rate": 7.651021101410673e-06,
"loss": 1.2133,
"step": 911
},
{
"epoch": 3.6941772151898733,
"grad_norm": 0.5706532982625352,
"learning_rate": 7.606423894180464e-06,
"loss": 1.2484,
"step": 912
},
{
"epoch": 3.698227848101266,
"grad_norm": 0.44132397960863395,
"learning_rate": 7.56192650310839e-06,
"loss": 1.1744,
"step": 913
},
{
"epoch": 3.7022784810126583,
"grad_norm": 0.46392863363008124,
"learning_rate": 7.517529286570349e-06,
"loss": 1.2114,
"step": 914
},
{
"epoch": 3.706329113924051,
"grad_norm": 0.5002936499736561,
"learning_rate": 7.473232602135387e-06,
"loss": 1.1931,
"step": 915
},
{
"epoch": 3.710379746835443,
"grad_norm": 0.47115728777523164,
"learning_rate": 7.429036806562935e-06,
"loss": 1.2103,
"step": 916
},
{
"epoch": 3.7144303797468354,
"grad_norm": 0.4547633417349095,
"learning_rate": 7.3849422557998455e-06,
"loss": 1.1816,
"step": 917
},
{
"epoch": 3.718481012658228,
"grad_norm": 0.4742317414039737,
"learning_rate": 7.340949304977567e-06,
"loss": 1.1596,
"step": 918
},
{
"epoch": 3.72253164556962,
"grad_norm": 0.4294432763784577,
"learning_rate": 7.297058308409282e-06,
"loss": 1.1833,
"step": 919
},
{
"epoch": 3.7265822784810125,
"grad_norm": 0.45505063479738284,
"learning_rate": 7.25326961958704e-06,
"loss": 1.2139,
"step": 920
},
{
"epoch": 3.730632911392405,
"grad_norm": 0.4402403813785672,
"learning_rate": 7.209583591178921e-06,
"loss": 1.2269,
"step": 921
},
{
"epoch": 3.7346835443037976,
"grad_norm": 0.4632785319500067,
"learning_rate": 7.1660005750261925e-06,
"loss": 1.1566,
"step": 922
},
{
"epoch": 3.73873417721519,
"grad_norm": 0.4520442537051577,
"learning_rate": 7.1225209221404765e-06,
"loss": 1.1991,
"step": 923
},
{
"epoch": 3.742784810126582,
"grad_norm": 0.4589950550774543,
"learning_rate": 7.079144982700909e-06,
"loss": 1.205,
"step": 924
},
{
"epoch": 3.7468354430379747,
"grad_norm": 0.4368486602947587,
"learning_rate": 7.0358731060513695e-06,
"loss": 1.2086,
"step": 925
},
{
"epoch": 3.750886075949367,
"grad_norm": 0.45463407383062476,
"learning_rate": 6.99270564069757e-06,
"loss": 1.2008,
"step": 926
},
{
"epoch": 3.7549367088607593,
"grad_norm": 0.4509774303376656,
"learning_rate": 6.949642934304375e-06,
"loss": 1.2102,
"step": 927
},
{
"epoch": 3.7589873417721518,
"grad_norm": 0.43729035733011673,
"learning_rate": 6.906685333692871e-06,
"loss": 1.216,
"step": 928
},
{
"epoch": 3.7630379746835443,
"grad_norm": 0.42968497927013155,
"learning_rate": 6.86383318483769e-06,
"loss": 1.2156,
"step": 929
},
{
"epoch": 3.767088607594937,
"grad_norm": 0.45175021956683625,
"learning_rate": 6.821086832864139e-06,
"loss": 1.214,
"step": 930
},
{
"epoch": 3.7711392405063293,
"grad_norm": 0.45770015816763676,
"learning_rate": 6.77844662204546e-06,
"loss": 1.1851,
"step": 931
},
{
"epoch": 3.7751898734177214,
"grad_norm": 0.45914185920821743,
"learning_rate": 6.7359128958000455e-06,
"loss": 1.2194,
"step": 932
},
{
"epoch": 3.779240506329114,
"grad_norm": 0.42777001768389034,
"learning_rate": 6.693485996688695e-06,
"loss": 1.1657,
"step": 933
},
{
"epoch": 3.7832911392405064,
"grad_norm": 0.43575550842681043,
"learning_rate": 6.651166266411801e-06,
"loss": 1.1958,
"step": 934
},
{
"epoch": 3.7873417721518985,
"grad_norm": 0.4478477319423756,
"learning_rate": 6.6089540458066725e-06,
"loss": 1.2226,
"step": 935
},
{
"epoch": 3.791392405063291,
"grad_norm": 0.43216419382189575,
"learning_rate": 6.566849674844711e-06,
"loss": 1.1959,
"step": 936
},
{
"epoch": 3.7954430379746835,
"grad_norm": 0.42830882336178555,
"learning_rate": 6.524853492628747e-06,
"loss": 1.1834,
"step": 937
},
{
"epoch": 3.799493670886076,
"grad_norm": 0.44457714761746375,
"learning_rate": 6.4829658373902536e-06,
"loss": 1.1641,
"step": 938
},
{
"epoch": 3.8035443037974686,
"grad_norm": 0.44085622421536563,
"learning_rate": 6.441187046486648e-06,
"loss": 1.2076,
"step": 939
},
{
"epoch": 3.8075949367088606,
"grad_norm": 0.4504502564520498,
"learning_rate": 6.399517456398567e-06,
"loss": 1.1817,
"step": 940
},
{
"epoch": 3.811645569620253,
"grad_norm": 0.4650088218613446,
"learning_rate": 6.357957402727164e-06,
"loss": 1.1957,
"step": 941
},
{
"epoch": 3.8156962025316457,
"grad_norm": 0.43626871831001507,
"learning_rate": 6.316507220191395e-06,
"loss": 1.1481,
"step": 942
},
{
"epoch": 3.8197468354430377,
"grad_norm": 0.4348886118447035,
"learning_rate": 6.275167242625331e-06,
"loss": 1.174,
"step": 943
},
{
"epoch": 3.8237974683544302,
"grad_norm": 0.4348858443556247,
"learning_rate": 6.233937802975471e-06,
"loss": 1.2098,
"step": 944
},
{
"epoch": 3.8278481012658228,
"grad_norm": 0.43300893543191227,
"learning_rate": 6.192819233298046e-06,
"loss": 1.1989,
"step": 945
},
{
"epoch": 3.8318987341772153,
"grad_norm": 0.4200899094588128,
"learning_rate": 6.151811864756383e-06,
"loss": 1.1859,
"step": 946
},
{
"epoch": 3.835949367088608,
"grad_norm": 0.4600679013018507,
"learning_rate": 6.1109160276181655e-06,
"loss": 1.2197,
"step": 947
},
{
"epoch": 3.84,
"grad_norm": 0.44026344762521014,
"learning_rate": 6.070132051252868e-06,
"loss": 1.1665,
"step": 948
},
{
"epoch": 3.8440506329113924,
"grad_norm": 0.48940350809940986,
"learning_rate": 6.0294602641290034e-06,
"loss": 1.1762,
"step": 949
},
{
"epoch": 3.848101265822785,
"grad_norm": 0.41606132056683315,
"learning_rate": 5.988900993811575e-06,
"loss": 1.1792,
"step": 950
},
{
"epoch": 3.852151898734177,
"grad_norm": 0.46593597737790243,
"learning_rate": 5.948454566959363e-06,
"loss": 1.1975,
"step": 951
},
{
"epoch": 3.8562025316455695,
"grad_norm": 0.43881524130833627,
"learning_rate": 5.908121309322328e-06,
"loss": 1.203,
"step": 952
},
{
"epoch": 3.860253164556962,
"grad_norm": 0.5070659398795176,
"learning_rate": 5.867901545738976e-06,
"loss": 1.1959,
"step": 953
},
{
"epoch": 3.8643037974683545,
"grad_norm": 0.4364066285046679,
"learning_rate": 5.827795600133774e-06,
"loss": 1.1717,
"step": 954
},
{
"epoch": 3.868354430379747,
"grad_norm": 0.43867768812607355,
"learning_rate": 5.787803795514466e-06,
"loss": 1.2056,
"step": 955
},
{
"epoch": 3.872405063291139,
"grad_norm": 0.4374270198735127,
"learning_rate": 5.747926453969576e-06,
"loss": 1.2057,
"step": 956
},
{
"epoch": 3.8764556962025316,
"grad_norm": 0.4709887667582948,
"learning_rate": 5.708163896665708e-06,
"loss": 1.1683,
"step": 957
},
{
"epoch": 3.880506329113924,
"grad_norm": 0.4346217905043107,
"learning_rate": 5.668516443845047e-06,
"loss": 1.227,
"step": 958
},
{
"epoch": 3.884556962025316,
"grad_norm": 0.4369141086716946,
"learning_rate": 5.6289844148227225e-06,
"loss": 1.1787,
"step": 959
},
{
"epoch": 3.8886075949367087,
"grad_norm": 0.4532679627290179,
"learning_rate": 5.5895681279842615e-06,
"loss": 1.2008,
"step": 960
},
{
"epoch": 3.8926582278481012,
"grad_norm": 0.45289754751397,
"learning_rate": 5.550267900783019e-06,
"loss": 1.1677,
"step": 961
},
{
"epoch": 3.8967088607594937,
"grad_norm": 0.45202644930721614,
"learning_rate": 5.511084049737623e-06,
"loss": 1.188,
"step": 962
},
{
"epoch": 3.9007594936708863,
"grad_norm": 0.44437497962180034,
"learning_rate": 5.4720168904294215e-06,
"loss": 1.1805,
"step": 963
},
{
"epoch": 3.9048101265822783,
"grad_norm": 0.4352514721804859,
"learning_rate": 5.433066737499948e-06,
"loss": 1.1965,
"step": 964
},
{
"epoch": 3.908860759493671,
"grad_norm": 0.47205924536701205,
"learning_rate": 5.394233904648376e-06,
"loss": 1.208,
"step": 965
},
{
"epoch": 3.9129113924050634,
"grad_norm": 0.4410340168369851,
"learning_rate": 5.355518704628997e-06,
"loss": 1.2068,
"step": 966
},
{
"epoch": 3.9169620253164554,
"grad_norm": 0.4265251666169479,
"learning_rate": 5.316921449248731e-06,
"loss": 1.1679,
"step": 967
},
{
"epoch": 3.921012658227848,
"grad_norm": 0.47374942883952564,
"learning_rate": 5.278442449364538e-06,
"loss": 1.2246,
"step": 968
},
{
"epoch": 3.9250632911392405,
"grad_norm": 0.43229777764230665,
"learning_rate": 5.240082014881016e-06,
"loss": 1.1867,
"step": 969
},
{
"epoch": 3.929113924050633,
"grad_norm": 0.43568087230040575,
"learning_rate": 5.201840454747822e-06,
"loss": 1.2001,
"step": 970
},
{
"epoch": 3.9331645569620255,
"grad_norm": 0.4338546571174067,
"learning_rate": 5.163718076957223e-06,
"loss": 1.1949,
"step": 971
},
{
"epoch": 3.9372151898734176,
"grad_norm": 0.4147302796244111,
"learning_rate": 5.125715188541609e-06,
"loss": 1.1938,
"step": 972
},
{
"epoch": 3.94126582278481,
"grad_norm": 0.43101249019710225,
"learning_rate": 5.087832095571021e-06,
"loss": 1.2196,
"step": 973
},
{
"epoch": 3.9453164556962026,
"grad_norm": 0.43965970550780703,
"learning_rate": 5.0500691031506766e-06,
"loss": 1.1741,
"step": 974
},
{
"epoch": 3.9493670886075947,
"grad_norm": 0.421174589443329,
"learning_rate": 5.01242651541854e-06,
"loss": 1.2319,
"step": 975
},
{
"epoch": 3.953417721518987,
"grad_norm": 0.41790012870754506,
"learning_rate": 4.974904635542815e-06,
"loss": 1.1787,
"step": 976
},
{
"epoch": 3.9574683544303797,
"grad_norm": 0.4207838171371075,
"learning_rate": 4.937503765719582e-06,
"loss": 1.1769,
"step": 977
},
{
"epoch": 3.961518987341772,
"grad_norm": 0.4303961445395076,
"learning_rate": 4.900224207170299e-06,
"loss": 1.1967,
"step": 978
},
{
"epoch": 3.9655696202531647,
"grad_norm": 0.43288352619944287,
"learning_rate": 4.8630662601394065e-06,
"loss": 1.2287,
"step": 979
},
{
"epoch": 3.969620253164557,
"grad_norm": 0.41920154174822266,
"learning_rate": 4.8260302238918995e-06,
"loss": 1.1987,
"step": 980
},
{
"epoch": 3.9736708860759493,
"grad_norm": 0.41554721134256767,
"learning_rate": 4.789116396710924e-06,
"loss": 1.1755,
"step": 981
},
{
"epoch": 3.977721518987342,
"grad_norm": 0.4217927184280148,
"learning_rate": 4.752325075895368e-06,
"loss": 1.1682,
"step": 982
},
{
"epoch": 3.981772151898734,
"grad_norm": 0.43630664873048447,
"learning_rate": 4.715656557757473e-06,
"loss": 1.2061,
"step": 983
},
{
"epoch": 3.9858227848101264,
"grad_norm": 0.44145295297760756,
"learning_rate": 4.679111137620442e-06,
"loss": 1.2136,
"step": 984
},
{
"epoch": 3.989873417721519,
"grad_norm": 0.41582502976690094,
"learning_rate": 4.6426891098160585e-06,
"loss": 1.1832,
"step": 985
},
{
"epoch": 3.9939240506329114,
"grad_norm": 0.4294946257422308,
"learning_rate": 4.6063907676823474e-06,
"loss": 1.2244,
"step": 986
},
{
"epoch": 3.997974683544304,
"grad_norm": 0.4563056993415526,
"learning_rate": 4.570216403561141e-06,
"loss": 1.2564,
"step": 987
},
{
"epoch": 4.002025316455696,
"grad_norm": 0.6648494628298313,
"learning_rate": 4.534166308795815e-06,
"loss": 1.2164,
"step": 988
},
{
"epoch": 4.006075949367089,
"grad_norm": 0.8930332304680204,
"learning_rate": 4.498240773728859e-06,
"loss": 1.0295,
"step": 989
},
{
"epoch": 4.010126582278481,
"grad_norm": 0.7414785774518974,
"learning_rate": 4.462440087699609e-06,
"loss": 1.0571,
"step": 990
},
{
"epoch": 4.014177215189873,
"grad_norm": 0.5981219180444515,
"learning_rate": 4.426764539041861e-06,
"loss": 1.0431,
"step": 991
},
{
"epoch": 4.018227848101266,
"grad_norm": 1.054563245431186,
"learning_rate": 4.391214415081582e-06,
"loss": 1.056,
"step": 992
},
{
"epoch": 4.022278481012658,
"grad_norm": 1.2009427533644643,
"learning_rate": 4.355790002134579e-06,
"loss": 1.0044,
"step": 993
},
{
"epoch": 4.02632911392405,
"grad_norm": 0.9491591416813697,
"learning_rate": 4.320491585504207e-06,
"loss": 1.0221,
"step": 994
},
{
"epoch": 4.030379746835443,
"grad_norm": 0.8082220993975204,
"learning_rate": 4.2853194494790615e-06,
"loss": 1.0191,
"step": 995
},
{
"epoch": 4.034430379746835,
"grad_norm": 0.8265196663638914,
"learning_rate": 4.250273877330691e-06,
"loss": 1.0417,
"step": 996
},
{
"epoch": 4.038481012658228,
"grad_norm": 0.8178147613815071,
"learning_rate": 4.215355151311313e-06,
"loss": 1.0493,
"step": 997
},
{
"epoch": 4.04253164556962,
"grad_norm": 0.8318326111776609,
"learning_rate": 4.180563552651542e-06,
"loss": 1.0423,
"step": 998
},
{
"epoch": 4.046582278481012,
"grad_norm": 0.7455501159095413,
"learning_rate": 4.145899361558147e-06,
"loss": 1.0179,
"step": 999
},
{
"epoch": 4.050632911392405,
"grad_norm": 0.6473417180867634,
"learning_rate": 4.111362857211738e-06,
"loss": 1.0123,
"step": 1000
},
{
"epoch": 4.054683544303797,
"grad_norm": 0.711875735736804,
"learning_rate": 4.076954317764592e-06,
"loss": 1.0624,
"step": 1001
},
{
"epoch": 4.0587341772151895,
"grad_norm": 0.7378468549051612,
"learning_rate": 4.042674020338335e-06,
"loss": 1.0258,
"step": 1002
},
{
"epoch": 4.062784810126582,
"grad_norm": 0.8467148228531438,
"learning_rate": 4.0085222410217835e-06,
"loss": 1.0637,
"step": 1003
},
{
"epoch": 4.0668354430379745,
"grad_norm": 0.6391825850223135,
"learning_rate": 3.974499254868674e-06,
"loss": 0.9961,
"step": 1004
},
{
"epoch": 4.0708860759493675,
"grad_norm": 0.5795422601312817,
"learning_rate": 3.940605335895451e-06,
"loss": 1.063,
"step": 1005
},
{
"epoch": 4.0749367088607595,
"grad_norm": 0.6313178323535786,
"learning_rate": 3.90684075707908e-06,
"loss": 1.0832,
"step": 1006
},
{
"epoch": 4.078987341772152,
"grad_norm": 0.6743175735477777,
"learning_rate": 3.8732057903548505e-06,
"loss": 1.0001,
"step": 1007
},
{
"epoch": 4.083037974683545,
"grad_norm": 0.6119184310381399,
"learning_rate": 3.8397007066141375e-06,
"loss": 1.0339,
"step": 1008
},
{
"epoch": 4.087088607594937,
"grad_norm": 0.5776325125572207,
"learning_rate": 3.806325775702304e-06,
"loss": 1.0148,
"step": 1009
},
{
"epoch": 4.091139240506329,
"grad_norm": 0.5574075491677055,
"learning_rate": 3.773081266416434e-06,
"loss": 1.0525,
"step": 1010
},
{
"epoch": 4.095189873417722,
"grad_norm": 0.5860144628295528,
"learning_rate": 3.739967446503245e-06,
"loss": 1.0414,
"step": 1011
},
{
"epoch": 4.099240506329114,
"grad_norm": 0.6310531738894284,
"learning_rate": 3.706984582656894e-06,
"loss": 1.0666,
"step": 1012
},
{
"epoch": 4.103291139240507,
"grad_norm": 0.5932343275916686,
"learning_rate": 3.6741329405168237e-06,
"loss": 1.0467,
"step": 1013
},
{
"epoch": 4.107341772151899,
"grad_norm": 0.5184830664650083,
"learning_rate": 3.641412784665648e-06,
"loss": 1.0276,
"step": 1014
},
{
"epoch": 4.111392405063291,
"grad_norm": 0.55861134641298,
"learning_rate": 3.608824378627005e-06,
"loss": 1.0416,
"step": 1015
},
{
"epoch": 4.115443037974684,
"grad_norm": 0.591684063957714,
"learning_rate": 3.5763679848634337e-06,
"loss": 1.0355,
"step": 1016
},
{
"epoch": 4.119493670886076,
"grad_norm": 0.5770463563150893,
"learning_rate": 3.544043864774269e-06,
"loss": 1.0385,
"step": 1017
},
{
"epoch": 4.123544303797468,
"grad_norm": 0.5192014195797624,
"learning_rate": 3.5118522786935282e-06,
"loss": 1.0153,
"step": 1018
},
{
"epoch": 4.127594936708861,
"grad_norm": 0.5435757932136609,
"learning_rate": 3.479793485887819e-06,
"loss": 1.1015,
"step": 1019
},
{
"epoch": 4.131645569620253,
"grad_norm": 0.5027685034978651,
"learning_rate": 3.4478677445542653e-06,
"loss": 1.0311,
"step": 1020
},
{
"epoch": 4.135696202531646,
"grad_norm": 0.5277614796471473,
"learning_rate": 3.4160753118183767e-06,
"loss": 0.9711,
"step": 1021
},
{
"epoch": 4.139746835443038,
"grad_norm": 0.5735832412584578,
"learning_rate": 3.3844164437320527e-06,
"loss": 1.0318,
"step": 1022
},
{
"epoch": 4.14379746835443,
"grad_norm": 0.5394626986300108,
"learning_rate": 3.3528913952714558e-06,
"loss": 1.0508,
"step": 1023
},
{
"epoch": 4.147848101265823,
"grad_norm": 0.50310525867152,
"learning_rate": 3.321500420335e-06,
"loss": 1.0293,
"step": 1024
},
{
"epoch": 4.151898734177215,
"grad_norm": 0.494832783315943,
"learning_rate": 3.290243771741275e-06,
"loss": 1.0105,
"step": 1025
},
{
"epoch": 4.155949367088607,
"grad_norm": 0.5027696662808417,
"learning_rate": 3.2591217012270325e-06,
"loss": 1.0311,
"step": 1026
},
{
"epoch": 4.16,
"grad_norm": 0.5153811278363972,
"learning_rate": 3.228134459445149e-06,
"loss": 1.0092,
"step": 1027
},
{
"epoch": 4.164050632911392,
"grad_norm": 0.5364259426765821,
"learning_rate": 3.1972822959626205e-06,
"loss": 1.0219,
"step": 1028
},
{
"epoch": 4.168101265822785,
"grad_norm": 0.47939297207668596,
"learning_rate": 3.166565459258513e-06,
"loss": 0.9861,
"step": 1029
},
{
"epoch": 4.172151898734177,
"grad_norm": 0.49522577673967944,
"learning_rate": 3.1359841967220193e-06,
"loss": 1.0395,
"step": 1030
},
{
"epoch": 4.176202531645569,
"grad_norm": 0.5070580663610541,
"learning_rate": 3.105538754650419e-06,
"loss": 1.0197,
"step": 1031
},
{
"epoch": 4.180253164556962,
"grad_norm": 0.506147517370805,
"learning_rate": 3.07522937824712e-06,
"loss": 1.0017,
"step": 1032
},
{
"epoch": 4.184303797468354,
"grad_norm": 0.5212788333643146,
"learning_rate": 3.0450563116196697e-06,
"loss": 1.0295,
"step": 1033
},
{
"epoch": 4.188354430379746,
"grad_norm": 0.4895817314382802,
"learning_rate": 3.0150197977778008e-06,
"loss": 1.0229,
"step": 1034
},
{
"epoch": 4.192405063291139,
"grad_norm": 0.5031245335169913,
"learning_rate": 2.985120078631465e-06,
"loss": 1.0075,
"step": 1035
},
{
"epoch": 4.1964556962025314,
"grad_norm": 0.4929480459004963,
"learning_rate": 2.9553573949888893e-06,
"loss": 1.0267,
"step": 1036
},
{
"epoch": 4.200506329113924,
"grad_norm": 0.5020815822166971,
"learning_rate": 2.9257319865546384e-06,
"loss": 1.0137,
"step": 1037
},
{
"epoch": 4.2045569620253165,
"grad_norm": 0.529311889660767,
"learning_rate": 2.896244091927678e-06,
"loss": 1.0582,
"step": 1038
},
{
"epoch": 4.2086075949367086,
"grad_norm": 0.5301022336328415,
"learning_rate": 2.8668939485994584e-06,
"loss": 1.0166,
"step": 1039
},
{
"epoch": 4.2126582278481015,
"grad_norm": 0.49086604629993147,
"learning_rate": 2.837681792951994e-06,
"loss": 1.0203,
"step": 1040
},
{
"epoch": 4.216708860759494,
"grad_norm": 0.49503618715887493,
"learning_rate": 2.808607860255981e-06,
"loss": 1.0207,
"step": 1041
},
{
"epoch": 4.220759493670886,
"grad_norm": 0.49017513498190507,
"learning_rate": 2.7796723846688634e-06,
"loss": 0.999,
"step": 1042
},
{
"epoch": 4.224810126582279,
"grad_norm": 0.5233685368853027,
"learning_rate": 2.7508755992329937e-06,
"loss": 1.0242,
"step": 1043
},
{
"epoch": 4.228860759493671,
"grad_norm": 0.5222763758063168,
"learning_rate": 2.722217735873718e-06,
"loss": 1.0154,
"step": 1044
},
{
"epoch": 4.232911392405064,
"grad_norm": 0.5107652124387514,
"learning_rate": 2.6936990253975315e-06,
"loss": 1.031,
"step": 1045
},
{
"epoch": 4.236962025316456,
"grad_norm": 0.48514666042768434,
"learning_rate": 2.665319697490205e-06,
"loss": 1.03,
"step": 1046
},
{
"epoch": 4.241012658227848,
"grad_norm": 0.495741114757844,
"learning_rate": 2.637079980714945e-06,
"loss": 1.0372,
"step": 1047
},
{
"epoch": 4.245063291139241,
"grad_norm": 0.4949146568589931,
"learning_rate": 2.6089801025105453e-06,
"loss": 1.0589,
"step": 1048
},
{
"epoch": 4.249113924050633,
"grad_norm": 0.4781215145697032,
"learning_rate": 2.581020289189571e-06,
"loss": 1.0224,
"step": 1049
},
{
"epoch": 4.253164556962025,
"grad_norm": 0.5189377689522933,
"learning_rate": 2.553200765936501e-06,
"loss": 1.019,
"step": 1050
},
{
"epoch": 4.257215189873418,
"grad_norm": 0.4928514386916176,
"learning_rate": 2.525521756805962e-06,
"loss": 1.0438,
"step": 1051
},
{
"epoch": 4.26126582278481,
"grad_norm": 0.488831421936607,
"learning_rate": 2.497983484720885e-06,
"loss": 1.0237,
"step": 1052
},
{
"epoch": 4.265316455696203,
"grad_norm": 0.489616357382862,
"learning_rate": 2.470586171470728e-06,
"loss": 1.0442,
"step": 1053
},
{
"epoch": 4.269367088607595,
"grad_norm": 0.49799471161919906,
"learning_rate": 2.4433300377096836e-06,
"loss": 1.0041,
"step": 1054
},
{
"epoch": 4.273417721518987,
"grad_norm": 0.48818358433401354,
"learning_rate": 2.4162153029549073e-06,
"loss": 1.0051,
"step": 1055
},
{
"epoch": 4.27746835443038,
"grad_norm": 0.5056047796627267,
"learning_rate": 2.3892421855847458e-06,
"loss": 1.0446,
"step": 1056
},
{
"epoch": 4.281518987341772,
"grad_norm": 0.5021345690114758,
"learning_rate": 2.362410902836978e-06,
"loss": 1.0067,
"step": 1057
},
{
"epoch": 4.285569620253165,
"grad_norm": 0.4844342255078532,
"learning_rate": 2.3357216708070653e-06,
"loss": 1.005,
"step": 1058
},
{
"epoch": 4.289620253164557,
"grad_norm": 0.4791109601211824,
"learning_rate": 2.309174704446411e-06,
"loss": 1.0347,
"step": 1059
},
{
"epoch": 4.293670886075949,
"grad_norm": 0.4959380894275428,
"learning_rate": 2.2827702175606437e-06,
"loss": 1.0413,
"step": 1060
},
{
"epoch": 4.297721518987342,
"grad_norm": 0.5018394064529311,
"learning_rate": 2.256508422807855e-06,
"loss": 1.0113,
"step": 1061
},
{
"epoch": 4.301772151898734,
"grad_norm": 0.49452041770516303,
"learning_rate": 2.230389531696946e-06,
"loss": 1.0193,
"step": 1062
},
{
"epoch": 4.305822784810126,
"grad_norm": 0.4984330983837433,
"learning_rate": 2.204413754585857e-06,
"loss": 1.0399,
"step": 1063
},
{
"epoch": 4.309873417721519,
"grad_norm": 0.48724856199613237,
"learning_rate": 2.1785813006799406e-06,
"loss": 1.0067,
"step": 1064
},
{
"epoch": 4.313924050632911,
"grad_norm": 0.48455407978175363,
"learning_rate": 2.1528923780302224e-06,
"loss": 1.027,
"step": 1065
},
{
"epoch": 4.317974683544303,
"grad_norm": 0.4975607365156767,
"learning_rate": 2.127347193531757e-06,
"loss": 1.0553,
"step": 1066
},
{
"epoch": 4.322025316455696,
"grad_norm": 0.5111965617758655,
"learning_rate": 2.101945952921942e-06,
"loss": 1.0408,
"step": 1067
},
{
"epoch": 4.326075949367088,
"grad_norm": 0.5490026053969088,
"learning_rate": 2.0766888607788906e-06,
"loss": 1.0247,
"step": 1068
},
{
"epoch": 4.330126582278481,
"grad_norm": 0.47593780889137427,
"learning_rate": 2.0515761205197337e-06,
"loss": 1.013,
"step": 1069
},
{
"epoch": 4.334177215189873,
"grad_norm": 0.48561015384877726,
"learning_rate": 2.0266079343990453e-06,
"loss": 1.0467,
"step": 1070
},
{
"epoch": 4.3382278481012655,
"grad_norm": 0.4859209009151709,
"learning_rate": 2.0017845035071494e-06,
"loss": 1.0296,
"step": 1071
},
{
"epoch": 4.3422784810126585,
"grad_norm": 0.4864620196645959,
"learning_rate": 1.9771060277685537e-06,
"loss": 1.0349,
"step": 1072
},
{
"epoch": 4.3463291139240505,
"grad_norm": 0.4818807521322039,
"learning_rate": 1.95257270594031e-06,
"loss": 1.0161,
"step": 1073
},
{
"epoch": 4.3503797468354435,
"grad_norm": 0.4899629434313933,
"learning_rate": 1.9281847356104188e-06,
"loss": 1.0549,
"step": 1074
},
{
"epoch": 4.3544303797468356,
"grad_norm": 0.48345202310583923,
"learning_rate": 1.9039423131962365e-06,
"loss": 1.0584,
"step": 1075
},
{
"epoch": 4.358481012658228,
"grad_norm": 0.4808398649560067,
"learning_rate": 1.8798456339429027e-06,
"loss": 1.0495,
"step": 1076
},
{
"epoch": 4.362531645569621,
"grad_norm": 0.481782094320306,
"learning_rate": 1.8558948919217612e-06,
"loss": 1.0342,
"step": 1077
},
{
"epoch": 4.366582278481013,
"grad_norm": 0.4842090461817237,
"learning_rate": 1.8320902800287954e-06,
"loss": 1.0462,
"step": 1078
},
{
"epoch": 4.370632911392405,
"grad_norm": 0.47867328875254544,
"learning_rate": 1.8084319899830726e-06,
"loss": 1.0321,
"step": 1079
},
{
"epoch": 4.374683544303798,
"grad_norm": 0.477225584187067,
"learning_rate": 1.7849202123252097e-06,
"loss": 1.0489,
"step": 1080
},
{
"epoch": 4.37873417721519,
"grad_norm": 0.47575499456553855,
"learning_rate": 1.7615551364158401e-06,
"loss": 1.0221,
"step": 1081
},
{
"epoch": 4.382784810126582,
"grad_norm": 0.48711378872419137,
"learning_rate": 1.738336950434061e-06,
"loss": 1.0366,
"step": 1082
},
{
"epoch": 4.386835443037975,
"grad_norm": 0.49166278801974295,
"learning_rate": 1.715265841375957e-06,
"loss": 1.0019,
"step": 1083
},
{
"epoch": 4.390886075949367,
"grad_norm": 0.5007401472925583,
"learning_rate": 1.6923419950530684e-06,
"loss": 1.0303,
"step": 1084
},
{
"epoch": 4.39493670886076,
"grad_norm": 0.4776527466957639,
"learning_rate": 1.6695655960909008e-06,
"loss": 1.0208,
"step": 1085
},
{
"epoch": 4.398987341772152,
"grad_norm": 0.47948482379272545,
"learning_rate": 1.646936827927441e-06,
"loss": 1.0179,
"step": 1086
},
{
"epoch": 4.403037974683544,
"grad_norm": 0.47419413471802496,
"learning_rate": 1.6244558728116766e-06,
"loss": 1.011,
"step": 1087
},
{
"epoch": 4.407088607594937,
"grad_norm": 0.5070140511323763,
"learning_rate": 1.6021229118021265e-06,
"loss": 1.0355,
"step": 1088
},
{
"epoch": 4.411139240506329,
"grad_norm": 0.4836476916226125,
"learning_rate": 1.5799381247653967e-06,
"loss": 1.0356,
"step": 1089
},
{
"epoch": 4.415189873417722,
"grad_norm": 0.4871698040608055,
"learning_rate": 1.5579016903747013e-06,
"loss": 1.0304,
"step": 1090
},
{
"epoch": 4.419240506329114,
"grad_norm": 0.4762111259621133,
"learning_rate": 1.5360137861084656e-06,
"loss": 1.0199,
"step": 1091
},
{
"epoch": 4.423291139240506,
"grad_norm": 0.4792087869115727,
"learning_rate": 1.5142745882488475e-06,
"loss": 1.0395,
"step": 1092
},
{
"epoch": 4.427341772151899,
"grad_norm": 0.4684431480421043,
"learning_rate": 1.4926842718803691e-06,
"loss": 1.0017,
"step": 1093
},
{
"epoch": 4.431392405063291,
"grad_norm": 0.4836130884463581,
"learning_rate": 1.4712430108884657e-06,
"loss": 1.016,
"step": 1094
},
{
"epoch": 4.435443037974683,
"grad_norm": 0.4990085289311973,
"learning_rate": 1.4499509779581078e-06,
"loss": 1.0408,
"step": 1095
},
{
"epoch": 4.439493670886076,
"grad_norm": 0.48728033665512704,
"learning_rate": 1.4288083445723988e-06,
"loss": 1.0278,
"step": 1096
},
{
"epoch": 4.443544303797468,
"grad_norm": 0.48291649257580044,
"learning_rate": 1.4078152810112045e-06,
"loss": 1.0237,
"step": 1097
},
{
"epoch": 4.44759493670886,
"grad_norm": 0.4814426736745352,
"learning_rate": 1.3869719563497697e-06,
"loss": 1.013,
"step": 1098
},
{
"epoch": 4.451645569620253,
"grad_norm": 0.4903071027050568,
"learning_rate": 1.3662785384573663e-06,
"loss": 1.0401,
"step": 1099
},
{
"epoch": 4.455696202531645,
"grad_norm": 0.4818543817261408,
"learning_rate": 1.3457351939959383e-06,
"loss": 1.0456,
"step": 1100
},
{
"epoch": 4.459746835443038,
"grad_norm": 0.47156894773906993,
"learning_rate": 1.3253420884187551e-06,
"loss": 1.0485,
"step": 1101
},
{
"epoch": 4.46379746835443,
"grad_norm": 0.4741992641955142,
"learning_rate": 1.3050993859690953e-06,
"loss": 1.0346,
"step": 1102
},
{
"epoch": 4.467848101265822,
"grad_norm": 0.4984170319803511,
"learning_rate": 1.2850072496788869e-06,
"loss": 1.0112,
"step": 1103
},
{
"epoch": 4.471898734177215,
"grad_norm": 0.4811587743083635,
"learning_rate": 1.2650658413674434e-06,
"loss": 1.0488,
"step": 1104
},
{
"epoch": 4.4759493670886075,
"grad_norm": 0.4718129781069484,
"learning_rate": 1.2452753216401226e-06,
"loss": 1.0337,
"step": 1105
},
{
"epoch": 4.48,
"grad_norm": 0.484280096819518,
"learning_rate": 1.2256358498870503e-06,
"loss": 1.0139,
"step": 1106
},
{
"epoch": 4.4840506329113925,
"grad_norm": 0.4945922353851416,
"learning_rate": 1.2061475842818337e-06,
"loss": 1.0654,
"step": 1107
},
{
"epoch": 4.488101265822785,
"grad_norm": 0.48768552062820975,
"learning_rate": 1.1868106817802816e-06,
"loss": 1.0006,
"step": 1108
},
{
"epoch": 4.4921518987341775,
"grad_norm": 0.482817100660586,
"learning_rate": 1.1676252981191482e-06,
"loss": 1.0321,
"step": 1109
},
{
"epoch": 4.49620253164557,
"grad_norm": 0.48703185662799664,
"learning_rate": 1.1485915878148823e-06,
"loss": 1.0179,
"step": 1110
},
{
"epoch": 4.500253164556962,
"grad_norm": 0.47458434510469427,
"learning_rate": 1.1297097041623584e-06,
"loss": 0.9761,
"step": 1111
},
{
"epoch": 4.504303797468355,
"grad_norm": 0.48107375477514064,
"learning_rate": 1.1109797992336847e-06,
"loss": 1.0456,
"step": 1112
},
{
"epoch": 4.508354430379747,
"grad_norm": 0.489385621176898,
"learning_rate": 1.092402023876933e-06,
"loss": 1.0024,
"step": 1113
},
{
"epoch": 4.512405063291139,
"grad_norm": 0.5041699301711228,
"learning_rate": 1.0739765277149527e-06,
"loss": 1.0321,
"step": 1114
},
{
"epoch": 4.516455696202532,
"grad_norm": 0.4760607088761475,
"learning_rate": 1.0557034591441596e-06,
"loss": 1.0237,
"step": 1115
},
{
"epoch": 4.520506329113924,
"grad_norm": 0.4824308912990539,
"learning_rate": 1.0375829653333324e-06,
"loss": 1.0419,
"step": 1116
},
{
"epoch": 4.524556962025317,
"grad_norm": 0.4790941776388887,
"learning_rate": 1.0196151922224385e-06,
"loss": 1.0191,
"step": 1117
},
{
"epoch": 4.528607594936709,
"grad_norm": 0.47193873682609344,
"learning_rate": 1.0018002845214526e-06,
"loss": 1.023,
"step": 1118
},
{
"epoch": 4.532658227848101,
"grad_norm": 0.484849281435286,
"learning_rate": 9.841383857091947e-07,
"loss": 1.0292,
"step": 1119
},
{
"epoch": 4.536708860759494,
"grad_norm": 0.4788534698344297,
"learning_rate": 9.666296380321616e-07,
"loss": 1.0274,
"step": 1120
},
{
"epoch": 4.540759493670886,
"grad_norm": 0.48849101696832803,
"learning_rate": 9.492741825034124e-07,
"loss": 1.0135,
"step": 1121
},
{
"epoch": 4.544810126582279,
"grad_norm": 0.4918848543821711,
"learning_rate": 9.320721589013892e-07,
"loss": 1.043,
"step": 1122
},
{
"epoch": 4.548860759493671,
"grad_norm": 0.47487392739310375,
"learning_rate": 9.150237057688339e-07,
"loss": 1.0523,
"step": 1123
},
{
"epoch": 4.552911392405063,
"grad_norm": 0.47710443572170036,
"learning_rate": 8.981289604116328e-07,
"loss": 1.0403,
"step": 1124
},
{
"epoch": 4.556962025316456,
"grad_norm": 0.46476093092952403,
"learning_rate": 8.813880588977542e-07,
"loss": 1.0039,
"step": 1125
},
{
"epoch": 4.561012658227848,
"grad_norm": 0.48487982450160316,
"learning_rate": 8.648011360561126e-07,
"loss": 1.0473,
"step": 1126
},
{
"epoch": 4.56506329113924,
"grad_norm": 0.4785365158329638,
"learning_rate": 8.483683254755037e-07,
"loss": 1.0185,
"step": 1127
},
{
"epoch": 4.569113924050633,
"grad_norm": 0.4795042885435069,
"learning_rate": 8.320897595035227e-07,
"loss": 1.0521,
"step": 1128
},
{
"epoch": 4.573164556962025,
"grad_norm": 0.47523010335282034,
"learning_rate": 8.159655692455093e-07,
"loss": 0.9765,
"step": 1129
},
{
"epoch": 4.577215189873417,
"grad_norm": 0.4806816508847355,
"learning_rate": 7.999958845634648e-07,
"loss": 1.0449,
"step": 1130
},
{
"epoch": 4.58126582278481,
"grad_norm": 0.46892660207611964,
"learning_rate": 7.841808340750478e-07,
"loss": 1.0009,
"step": 1131
},
{
"epoch": 4.585316455696202,
"grad_norm": 0.4756054360431227,
"learning_rate": 7.685205451524869e-07,
"loss": 1.0286,
"step": 1132
},
{
"epoch": 4.589367088607595,
"grad_norm": 0.49171331136202295,
"learning_rate": 7.530151439216027e-07,
"loss": 1.0427,
"step": 1133
},
{
"epoch": 4.593417721518987,
"grad_norm": 0.5038809286061726,
"learning_rate": 7.376647552607675e-07,
"loss": 1.0561,
"step": 1134
},
{
"epoch": 4.597468354430379,
"grad_norm": 0.48538159780754087,
"learning_rate": 7.224695027998963e-07,
"loss": 1.0141,
"step": 1135
},
{
"epoch": 4.601518987341772,
"grad_norm": 0.471694239248044,
"learning_rate": 7.07429508919466e-07,
"loss": 1.0397,
"step": 1136
},
{
"epoch": 4.605569620253164,
"grad_norm": 0.48453217713430835,
"learning_rate": 6.925448947495206e-07,
"loss": 1.0371,
"step": 1137
},
{
"epoch": 4.609620253164557,
"grad_norm": 0.48376360133986257,
"learning_rate": 6.778157801686936e-07,
"loss": 1.0312,
"step": 1138
},
{
"epoch": 4.613670886075949,
"grad_norm": 0.47225993140716027,
"learning_rate": 6.632422838032515e-07,
"loss": 0.992,
"step": 1139
},
{
"epoch": 4.6177215189873415,
"grad_norm": 0.4914180964666429,
"learning_rate": 6.488245230261281e-07,
"loss": 1.0437,
"step": 1140
},
{
"epoch": 4.6217721518987345,
"grad_norm": 0.46848019391996826,
"learning_rate": 6.345626139559868e-07,
"loss": 1.0391,
"step": 1141
},
{
"epoch": 4.6258227848101265,
"grad_norm": 0.46488262378232276,
"learning_rate": 6.204566714562866e-07,
"loss": 1.0267,
"step": 1142
},
{
"epoch": 4.629873417721519,
"grad_norm": 0.48116196057980615,
"learning_rate": 6.06506809134344e-07,
"loss": 1.078,
"step": 1143
},
{
"epoch": 4.633924050632912,
"grad_norm": 0.4678757744397705,
"learning_rate": 5.927131393404373e-07,
"loss": 1.021,
"step": 1144
},
{
"epoch": 4.637974683544304,
"grad_norm": 0.46378512922378357,
"learning_rate": 5.790757731668817e-07,
"loss": 0.9848,
"step": 1145
},
{
"epoch": 4.642025316455696,
"grad_norm": 0.47146273729087723,
"learning_rate": 5.655948204471507e-07,
"loss": 1.0526,
"step": 1146
},
{
"epoch": 4.646075949367089,
"grad_norm": 0.47010142212387407,
"learning_rate": 5.522703897549875e-07,
"loss": 1.0185,
"step": 1147
},
{
"epoch": 4.650126582278481,
"grad_norm": 0.47874344870527885,
"learning_rate": 5.391025884035239e-07,
"loss": 1.0342,
"step": 1148
},
{
"epoch": 4.654177215189874,
"grad_norm": 0.47209827135252297,
"learning_rate": 5.260915224444207e-07,
"loss": 1.0003,
"step": 1149
},
{
"epoch": 4.658227848101266,
"grad_norm": 0.47383748536160947,
"learning_rate": 5.132372966670129e-07,
"loss": 1.0674,
"step": 1150
},
{
"epoch": 4.662278481012658,
"grad_norm": 0.46748107894738844,
"learning_rate": 5.005400145974704e-07,
"loss": 1.0622,
"step": 1151
},
{
"epoch": 4.666329113924051,
"grad_norm": 0.48569562484679113,
"learning_rate": 4.879997784979562e-07,
"loss": 1.0083,
"step": 1152
},
{
"epoch": 4.670379746835443,
"grad_norm": 0.4719304414396807,
"learning_rate": 4.7561668936580984e-07,
"loss": 1.0283,
"step": 1153
},
{
"epoch": 4.674430379746836,
"grad_norm": 0.4700956625856296,
"learning_rate": 4.6339084693272306e-07,
"loss": 1.0672,
"step": 1154
},
{
"epoch": 4.678481012658228,
"grad_norm": 0.4725724476212699,
"learning_rate": 4.5132234966395847e-07,
"loss": 1.0142,
"step": 1155
},
{
"epoch": 4.68253164556962,
"grad_norm": 0.4937842543469435,
"learning_rate": 4.3941129475752795e-07,
"loss": 1.0212,
"step": 1156
},
{
"epoch": 4.686582278481013,
"grad_norm": 0.47017098765681525,
"learning_rate": 4.27657778143431e-07,
"loss": 1.0061,
"step": 1157
},
{
"epoch": 4.690632911392405,
"grad_norm": 0.4887028686113434,
"learning_rate": 4.1606189448287757e-07,
"loss": 1.053,
"step": 1158
},
{
"epoch": 4.694683544303797,
"grad_norm": 0.4770372063164144,
"learning_rate": 4.046237371675177e-07,
"loss": 1.0284,
"step": 1159
},
{
"epoch": 4.69873417721519,
"grad_norm": 0.47460652100412226,
"learning_rate": 3.9334339831869963e-07,
"loss": 1.0361,
"step": 1160
},
{
"epoch": 4.702784810126582,
"grad_norm": 0.4780593775841714,
"learning_rate": 3.8222096878671955e-07,
"loss": 1.0073,
"step": 1161
},
{
"epoch": 4.706835443037974,
"grad_norm": 0.4790006302885801,
"learning_rate": 3.7125653815009545e-07,
"loss": 1.0336,
"step": 1162
},
{
"epoch": 4.710886075949367,
"grad_norm": 0.4722476492756577,
"learning_rate": 3.6045019471484974e-07,
"loss": 1.0377,
"step": 1163
},
{
"epoch": 4.714936708860759,
"grad_norm": 0.4724052708299559,
"learning_rate": 3.498020255137813e-07,
"loss": 1.0375,
"step": 1164
},
{
"epoch": 4.718987341772152,
"grad_norm": 0.46954039826245286,
"learning_rate": 3.393121163057811e-07,
"loss": 1.0516,
"step": 1165
},
{
"epoch": 4.723037974683544,
"grad_norm": 0.48810890784968497,
"learning_rate": 3.289805515751399e-07,
"loss": 1.0439,
"step": 1166
},
{
"epoch": 4.727088607594936,
"grad_norm": 0.4898239532362736,
"learning_rate": 3.188074145308573e-07,
"loss": 1.0234,
"step": 1167
},
{
"epoch": 4.731139240506329,
"grad_norm": 0.48293379386930085,
"learning_rate": 3.087927871059804e-07,
"loss": 1.0347,
"step": 1168
},
{
"epoch": 4.735189873417721,
"grad_norm": 0.48131476805301,
"learning_rate": 2.989367499569418e-07,
"loss": 1.0607,
"step": 1169
},
{
"epoch": 4.739240506329114,
"grad_norm": 0.47813435031135887,
"learning_rate": 2.8923938246290917e-07,
"loss": 1.002,
"step": 1170
},
{
"epoch": 4.743291139240506,
"grad_norm": 0.47344172168122706,
"learning_rate": 2.7970076272514804e-07,
"loss": 1.0568,
"step": 1171
},
{
"epoch": 4.7473417721518985,
"grad_norm": 0.46369421502976976,
"learning_rate": 2.703209675663887e-07,
"loss": 1.0213,
"step": 1172
},
{
"epoch": 4.751392405063291,
"grad_norm": 0.5004924633877847,
"learning_rate": 2.6110007253021374e-07,
"loss": 1.0364,
"step": 1173
},
{
"epoch": 4.7554430379746835,
"grad_norm": 0.47082162981783754,
"learning_rate": 2.520381518804471e-07,
"loss": 1.016,
"step": 1174
},
{
"epoch": 4.759493670886076,
"grad_norm": 0.4628987416864406,
"learning_rate": 2.4313527860054585e-07,
"loss": 1.0291,
"step": 1175
},
{
"epoch": 4.7635443037974685,
"grad_norm": 0.47377583848524435,
"learning_rate": 2.343915243930317e-07,
"loss": 1.0302,
"step": 1176
},
{
"epoch": 4.767594936708861,
"grad_norm": 0.47125321903468215,
"learning_rate": 2.2580695967889367e-07,
"loss": 1.033,
"step": 1177
},
{
"epoch": 4.7716455696202535,
"grad_norm": 0.47355465157768784,
"learning_rate": 2.1738165359704189e-07,
"loss": 1.023,
"step": 1178
},
{
"epoch": 4.775696202531646,
"grad_norm": 0.4825777145892887,
"learning_rate": 2.0911567400373257e-07,
"loss": 1.0542,
"step": 1179
},
{
"epoch": 4.779746835443038,
"grad_norm": 0.49526669442762405,
"learning_rate": 2.0100908747202607e-07,
"loss": 1.0826,
"step": 1180
},
{
"epoch": 4.783797468354431,
"grad_norm": 0.4875982709589614,
"learning_rate": 1.9306195929125638e-07,
"loss": 1.0343,
"step": 1181
},
{
"epoch": 4.787848101265823,
"grad_norm": 0.47784792353204963,
"learning_rate": 1.8527435346650247e-07,
"loss": 1.0299,
"step": 1182
},
{
"epoch": 4.791898734177215,
"grad_norm": 0.4749887842722522,
"learning_rate": 1.7764633271807108e-07,
"loss": 1.0448,
"step": 1183
},
{
"epoch": 4.795949367088608,
"grad_norm": 0.47219808386523454,
"learning_rate": 1.7017795848099262e-07,
"loss": 1.0004,
"step": 1184
},
{
"epoch": 4.8,
"grad_norm": 0.4676601402797075,
"learning_rate": 1.6286929090452596e-07,
"loss": 1.0283,
"step": 1185
},
{
"epoch": 4.804050632911393,
"grad_norm": 0.4637623806104882,
"learning_rate": 1.557203888516745e-07,
"loss": 1.0186,
"step": 1186
},
{
"epoch": 4.808101265822785,
"grad_norm": 0.460733791342502,
"learning_rate": 1.487313098987131e-07,
"loss": 1.0161,
"step": 1187
},
{
"epoch": 4.812151898734177,
"grad_norm": 0.47967421708572616,
"learning_rate": 1.4190211033472402e-07,
"loss": 1.0546,
"step": 1188
},
{
"epoch": 4.81620253164557,
"grad_norm": 0.479533402810503,
"learning_rate": 1.3523284516113955e-07,
"loss": 1.0131,
"step": 1189
},
{
"epoch": 4.820253164556962,
"grad_norm": 0.47403607416857346,
"learning_rate": 1.2872356809130682e-07,
"loss": 1.0398,
"step": 1190
},
{
"epoch": 4.824303797468355,
"grad_norm": 0.4724209766625126,
"learning_rate": 1.2237433155004807e-07,
"loss": 1.015,
"step": 1191
},
{
"epoch": 4.828354430379747,
"grad_norm": 0.48350666418897736,
"learning_rate": 1.1618518667323886e-07,
"loss": 1.0631,
"step": 1192
},
{
"epoch": 4.832405063291139,
"grad_norm": 0.4796119650452368,
"learning_rate": 1.1015618330740385e-07,
"loss": 1.0529,
"step": 1193
},
{
"epoch": 4.836455696202532,
"grad_norm": 0.4811473431862424,
"learning_rate": 1.042873700093061e-07,
"loss": 1.0194,
"step": 1194
},
{
"epoch": 4.840506329113924,
"grad_norm": 0.4755229156346968,
"learning_rate": 9.857879404556291e-08,
"loss": 1.0329,
"step": 1195
},
{
"epoch": 4.844556962025316,
"grad_norm": 0.4704293190680227,
"learning_rate": 9.303050139225722e-08,
"loss": 1.0448,
"step": 1196
},
{
"epoch": 4.848607594936709,
"grad_norm": 0.48136097335461175,
"learning_rate": 8.76425367345779e-08,
"loss": 1.0522,
"step": 1197
},
{
"epoch": 4.852658227848101,
"grad_norm": 0.4748540823441084,
"learning_rate": 8.241494346644897e-08,
"loss": 1.0548,
"step": 1198
},
{
"epoch": 4.856708860759493,
"grad_norm": 0.4792177907735774,
"learning_rate": 7.734776369019204e-08,
"loss": 1.0333,
"step": 1199
},
{
"epoch": 4.860759493670886,
"grad_norm": 0.4618070110379755,
"learning_rate": 7.244103821617332e-08,
"loss": 1.0236,
"step": 1200
},
{
"epoch": 4.864810126582278,
"grad_norm": 0.47306676777869594,
"learning_rate": 6.769480656248606e-08,
"loss": 1.083,
"step": 1201
},
{
"epoch": 4.868860759493671,
"grad_norm": 0.48177195894999014,
"learning_rate": 6.310910695462635e-08,
"loss": 1.0439,
"step": 1202
},
{
"epoch": 4.872911392405063,
"grad_norm": 0.4660468942821506,
"learning_rate": 5.8683976325191185e-08,
"loss": 1.0116,
"step": 1203
},
{
"epoch": 4.876962025316455,
"grad_norm": 0.46498160286141055,
"learning_rate": 5.4419450313571984e-08,
"loss": 1.0596,
"step": 1204
},
{
"epoch": 4.881012658227848,
"grad_norm": 0.47268724331555906,
"learning_rate": 5.031556326567488e-08,
"loss": 1.0589,
"step": 1205
},
{
"epoch": 4.88506329113924,
"grad_norm": 0.46719341052099717,
"learning_rate": 4.637234823364312e-08,
"loss": 1.0171,
"step": 1206
},
{
"epoch": 4.889113924050633,
"grad_norm": 0.5073553638914423,
"learning_rate": 4.258983697558838e-08,
"loss": 1.0387,
"step": 1207
},
{
"epoch": 4.8931645569620255,
"grad_norm": 0.47412027941257645,
"learning_rate": 3.896805995533548e-08,
"loss": 1.0073,
"step": 1208
},
{
"epoch": 4.8972151898734175,
"grad_norm": 0.48702211978962884,
"learning_rate": 3.550704634218028e-08,
"loss": 1.0171,
"step": 1209
},
{
"epoch": 4.9012658227848105,
"grad_norm": 0.45798990589577987,
"learning_rate": 3.2206824010647676e-08,
"loss": 0.9958,
"step": 1210
},
{
"epoch": 4.905316455696203,
"grad_norm": 0.4993787692797108,
"learning_rate": 2.9067419540278476e-08,
"loss": 1.0492,
"step": 1211
},
{
"epoch": 4.909367088607595,
"grad_norm": 0.4872385075544387,
"learning_rate": 2.6088858215400638e-08,
"loss": 0.9802,
"step": 1212
},
{
"epoch": 4.913417721518988,
"grad_norm": 0.49451137864228306,
"learning_rate": 2.3271164024940564e-08,
"loss": 1.0641,
"step": 1213
},
{
"epoch": 4.91746835443038,
"grad_norm": 0.4680117671074222,
"learning_rate": 2.061435966221881e-08,
"loss": 1.0542,
"step": 1214
},
{
"epoch": 4.921518987341772,
"grad_norm": 0.47931195308429947,
"learning_rate": 1.811846652477245e-08,
"loss": 0.9968,
"step": 1215
},
{
"epoch": 4.925569620253165,
"grad_norm": 0.4795412696612002,
"learning_rate": 1.5783504714184106e-08,
"loss": 1.0361,
"step": 1216
},
{
"epoch": 4.929620253164557,
"grad_norm": 0.4759766727935285,
"learning_rate": 1.360949303591097e-08,
"loss": 1.0665,
"step": 1217
},
{
"epoch": 4.93367088607595,
"grad_norm": 0.47629798680377317,
"learning_rate": 1.1596448999144916e-08,
"loss": 1.0348,
"step": 1218
},
{
"epoch": 4.937721518987342,
"grad_norm": 0.5133190947930638,
"learning_rate": 9.744388816668172e-09,
"loss": 1.0415,
"step": 1219
},
{
"epoch": 4.941772151898734,
"grad_norm": 0.4591473909159565,
"learning_rate": 8.05332740472009e-09,
"loss": 0.9989,
"step": 1220
},
{
"epoch": 4.945822784810127,
"grad_norm": 0.46855779428349553,
"learning_rate": 6.523278382872811e-09,
"loss": 1.0254,
"step": 1221
},
{
"epoch": 4.949873417721519,
"grad_norm": 0.4802234890648042,
"learning_rate": 5.15425407393133e-09,
"loss": 0.9957,
"step": 1222
},
{
"epoch": 4.953924050632912,
"grad_norm": 0.4792301942511197,
"learning_rate": 3.94626550383137e-09,
"loss": 1.0177,
"step": 1223
},
{
"epoch": 4.957974683544304,
"grad_norm": 0.4806546973271331,
"learning_rate": 2.899322401546112e-09,
"loss": 1.0076,
"step": 1224
},
{
"epoch": 4.962025316455696,
"grad_norm": 0.47147047821360555,
"learning_rate": 2.013433199010706e-09,
"loss": 0.9959,
"step": 1225
},
{
"epoch": 4.966075949367089,
"grad_norm": 0.48448776994213283,
"learning_rate": 1.2886050310556563e-09,
"loss": 1.0182,
"step": 1226
},
{
"epoch": 4.970126582278481,
"grad_norm": 0.46414942163203693,
"learning_rate": 7.248437353468695e-10,
"loss": 0.9741,
"step": 1227
},
{
"epoch": 4.974177215189873,
"grad_norm": 0.47008297704441965,
"learning_rate": 3.221538523412449e-10,
"loss": 1.035,
"step": 1228
},
{
"epoch": 4.978227848101266,
"grad_norm": 0.4671723112858263,
"learning_rate": 8.053862524670663e-11,
"loss": 0.9961,
"step": 1229
},
{
"epoch": 4.982278481012658,
"grad_norm": 0.4674606149102945,
"learning_rate": 0.0,
"loss": 1.0393,
"step": 1230
},
{
"epoch": 4.982278481012658,
"step": 1230,
"total_flos": 1.949760664301609e+18,
"train_loss": 1.3435062479197495,
"train_runtime": 12265.1955,
"train_samples_per_second": 12.882,
"train_steps_per_second": 0.1
}
],
"logging_steps": 1.0,
"max_steps": 1230,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.949760664301609e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}