tbench_traces_sharegptv1 / trainer_state.json
EtashGuha's picture
End of training
905ae03 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 7.0,
"eval_steps": 500,
"global_step": 567,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.012345679012345678,
"grad_norm": 9.008057481638138,
"learning_rate": 3.5087719298245616e-07,
"loss": 0.6743,
"step": 1
},
{
"epoch": 0.024691358024691357,
"grad_norm": 7.6701365166507625,
"learning_rate": 7.017543859649123e-07,
"loss": 0.576,
"step": 2
},
{
"epoch": 0.037037037037037035,
"grad_norm": 8.97726725259136,
"learning_rate": 1.0526315789473685e-06,
"loss": 0.6663,
"step": 3
},
{
"epoch": 0.04938271604938271,
"grad_norm": 7.225115208278711,
"learning_rate": 1.4035087719298246e-06,
"loss": 0.5696,
"step": 4
},
{
"epoch": 0.06172839506172839,
"grad_norm": 7.33497956558816,
"learning_rate": 1.7543859649122807e-06,
"loss": 0.5746,
"step": 5
},
{
"epoch": 0.07407407407407407,
"grad_norm": 6.771581012013294,
"learning_rate": 2.105263157894737e-06,
"loss": 0.5319,
"step": 6
},
{
"epoch": 0.08641975308641975,
"grad_norm": 5.723788798376614,
"learning_rate": 2.456140350877193e-06,
"loss": 0.4894,
"step": 7
},
{
"epoch": 0.09876543209876543,
"grad_norm": 5.4088640282382565,
"learning_rate": 2.8070175438596493e-06,
"loss": 0.517,
"step": 8
},
{
"epoch": 0.1111111111111111,
"grad_norm": 2.153808698110969,
"learning_rate": 3.157894736842105e-06,
"loss": 0.4509,
"step": 9
},
{
"epoch": 0.12345679012345678,
"grad_norm": 2.0408212548859406,
"learning_rate": 3.5087719298245615e-06,
"loss": 0.4467,
"step": 10
},
{
"epoch": 0.13580246913580246,
"grad_norm": 1.5459963386370055,
"learning_rate": 3.859649122807018e-06,
"loss": 0.4118,
"step": 11
},
{
"epoch": 0.14814814814814814,
"grad_norm": 1.5814209627310287,
"learning_rate": 4.210526315789474e-06,
"loss": 0.3888,
"step": 12
},
{
"epoch": 0.16049382716049382,
"grad_norm": 1.87145877734503,
"learning_rate": 4.56140350877193e-06,
"loss": 0.4058,
"step": 13
},
{
"epoch": 0.1728395061728395,
"grad_norm": 1.3756801350478975,
"learning_rate": 4.912280701754386e-06,
"loss": 0.3741,
"step": 14
},
{
"epoch": 0.18518518518518517,
"grad_norm": 1.1752142353172148,
"learning_rate": 5.263157894736842e-06,
"loss": 0.3995,
"step": 15
},
{
"epoch": 0.19753086419753085,
"grad_norm": 1.418292428476638,
"learning_rate": 5.6140350877192985e-06,
"loss": 0.385,
"step": 16
},
{
"epoch": 0.20987654320987653,
"grad_norm": 1.16124552621937,
"learning_rate": 5.964912280701755e-06,
"loss": 0.36,
"step": 17
},
{
"epoch": 0.2222222222222222,
"grad_norm": 0.9870066082850208,
"learning_rate": 6.31578947368421e-06,
"loss": 0.3499,
"step": 18
},
{
"epoch": 0.2345679012345679,
"grad_norm": 0.9366608437505954,
"learning_rate": 6.666666666666667e-06,
"loss": 0.3513,
"step": 19
},
{
"epoch": 0.24691358024691357,
"grad_norm": 1.1198168697388675,
"learning_rate": 7.017543859649123e-06,
"loss": 0.3412,
"step": 20
},
{
"epoch": 0.25925925925925924,
"grad_norm": 0.877398481087362,
"learning_rate": 7.368421052631579e-06,
"loss": 0.3298,
"step": 21
},
{
"epoch": 0.2716049382716049,
"grad_norm": 0.9702377137387861,
"learning_rate": 7.719298245614036e-06,
"loss": 0.3365,
"step": 22
},
{
"epoch": 0.2839506172839506,
"grad_norm": 0.7984859557261315,
"learning_rate": 8.070175438596492e-06,
"loss": 0.3025,
"step": 23
},
{
"epoch": 0.2962962962962963,
"grad_norm": 0.7050324311912214,
"learning_rate": 8.421052631578948e-06,
"loss": 0.2703,
"step": 24
},
{
"epoch": 0.30864197530864196,
"grad_norm": 0.9126479851972958,
"learning_rate": 8.771929824561405e-06,
"loss": 0.2939,
"step": 25
},
{
"epoch": 0.32098765432098764,
"grad_norm": 0.8796740071209702,
"learning_rate": 9.12280701754386e-06,
"loss": 0.304,
"step": 26
},
{
"epoch": 0.3333333333333333,
"grad_norm": 0.9446428136109001,
"learning_rate": 9.473684210526315e-06,
"loss": 0.3092,
"step": 27
},
{
"epoch": 0.345679012345679,
"grad_norm": 0.8426809268341795,
"learning_rate": 9.824561403508772e-06,
"loss": 0.2732,
"step": 28
},
{
"epoch": 0.35802469135802467,
"grad_norm": 0.8732085201361185,
"learning_rate": 1.017543859649123e-05,
"loss": 0.2745,
"step": 29
},
{
"epoch": 0.37037037037037035,
"grad_norm": 0.827942537306075,
"learning_rate": 1.0526315789473684e-05,
"loss": 0.2824,
"step": 30
},
{
"epoch": 0.38271604938271603,
"grad_norm": 0.9093812676850115,
"learning_rate": 1.0877192982456142e-05,
"loss": 0.2796,
"step": 31
},
{
"epoch": 0.3950617283950617,
"grad_norm": 0.7240207676930677,
"learning_rate": 1.1228070175438597e-05,
"loss": 0.2413,
"step": 32
},
{
"epoch": 0.4074074074074074,
"grad_norm": 0.8837553684406467,
"learning_rate": 1.1578947368421053e-05,
"loss": 0.2654,
"step": 33
},
{
"epoch": 0.41975308641975306,
"grad_norm": 0.9620696328942127,
"learning_rate": 1.192982456140351e-05,
"loss": 0.2604,
"step": 34
},
{
"epoch": 0.43209876543209874,
"grad_norm": 0.8315152131518835,
"learning_rate": 1.2280701754385966e-05,
"loss": 0.2374,
"step": 35
},
{
"epoch": 0.4444444444444444,
"grad_norm": 0.7954427643282206,
"learning_rate": 1.263157894736842e-05,
"loss": 0.2599,
"step": 36
},
{
"epoch": 0.4567901234567901,
"grad_norm": 0.8097701562489309,
"learning_rate": 1.2982456140350879e-05,
"loss": 0.2336,
"step": 37
},
{
"epoch": 0.4691358024691358,
"grad_norm": 0.7679487776985499,
"learning_rate": 1.3333333333333333e-05,
"loss": 0.1984,
"step": 38
},
{
"epoch": 0.48148148148148145,
"grad_norm": 0.8186214855532091,
"learning_rate": 1.3684210526315791e-05,
"loss": 0.2257,
"step": 39
},
{
"epoch": 0.49382716049382713,
"grad_norm": 0.8486399745760126,
"learning_rate": 1.4035087719298246e-05,
"loss": 0.2203,
"step": 40
},
{
"epoch": 0.5061728395061729,
"grad_norm": 0.7515522994828953,
"learning_rate": 1.4385964912280704e-05,
"loss": 0.2033,
"step": 41
},
{
"epoch": 0.5185185185185185,
"grad_norm": 0.8186675607439609,
"learning_rate": 1.4736842105263159e-05,
"loss": 0.2035,
"step": 42
},
{
"epoch": 0.5308641975308642,
"grad_norm": 0.8105881928417837,
"learning_rate": 1.5087719298245615e-05,
"loss": 0.2067,
"step": 43
},
{
"epoch": 0.5432098765432098,
"grad_norm": 0.7099905049117751,
"learning_rate": 1.543859649122807e-05,
"loss": 0.1619,
"step": 44
},
{
"epoch": 0.5555555555555556,
"grad_norm": 0.7178850952198459,
"learning_rate": 1.578947368421053e-05,
"loss": 0.1763,
"step": 45
},
{
"epoch": 0.5679012345679012,
"grad_norm": 0.9051580828689527,
"learning_rate": 1.6140350877192984e-05,
"loss": 0.1777,
"step": 46
},
{
"epoch": 0.5802469135802469,
"grad_norm": 0.776120251351539,
"learning_rate": 1.649122807017544e-05,
"loss": 0.1533,
"step": 47
},
{
"epoch": 0.5925925925925926,
"grad_norm": 0.6923150710928373,
"learning_rate": 1.6842105263157896e-05,
"loss": 0.1471,
"step": 48
},
{
"epoch": 0.6049382716049383,
"grad_norm": 0.8308137534219192,
"learning_rate": 1.719298245614035e-05,
"loss": 0.1661,
"step": 49
},
{
"epoch": 0.6172839506172839,
"grad_norm": 0.6920144332725509,
"learning_rate": 1.754385964912281e-05,
"loss": 0.1544,
"step": 50
},
{
"epoch": 0.6296296296296297,
"grad_norm": 0.8936796942488102,
"learning_rate": 1.7894736842105264e-05,
"loss": 0.1517,
"step": 51
},
{
"epoch": 0.6419753086419753,
"grad_norm": 0.7606903567238646,
"learning_rate": 1.824561403508772e-05,
"loss": 0.1436,
"step": 52
},
{
"epoch": 0.654320987654321,
"grad_norm": 0.7983567717631593,
"learning_rate": 1.8596491228070176e-05,
"loss": 0.1504,
"step": 53
},
{
"epoch": 0.6666666666666666,
"grad_norm": 0.6932938639060522,
"learning_rate": 1.894736842105263e-05,
"loss": 0.1323,
"step": 54
},
{
"epoch": 0.6790123456790124,
"grad_norm": 0.765978207523316,
"learning_rate": 1.929824561403509e-05,
"loss": 0.1584,
"step": 55
},
{
"epoch": 0.691358024691358,
"grad_norm": 0.6706267081010999,
"learning_rate": 1.9649122807017544e-05,
"loss": 0.1226,
"step": 56
},
{
"epoch": 0.7037037037037037,
"grad_norm": 0.7576546029397313,
"learning_rate": 2e-05,
"loss": 0.1107,
"step": 57
},
{
"epoch": 0.7160493827160493,
"grad_norm": 0.631172942746403,
"learning_rate": 1.999981027348727e-05,
"loss": 0.1269,
"step": 58
},
{
"epoch": 0.7283950617283951,
"grad_norm": 0.7763194682579061,
"learning_rate": 1.999924110114831e-05,
"loss": 0.1106,
"step": 59
},
{
"epoch": 0.7407407407407407,
"grad_norm": 0.6620307624287203,
"learning_rate": 1.9998292504580528e-05,
"loss": 0.1264,
"step": 60
},
{
"epoch": 0.7530864197530864,
"grad_norm": 0.6018372333318557,
"learning_rate": 1.999696451977872e-05,
"loss": 0.1064,
"step": 61
},
{
"epoch": 0.7654320987654321,
"grad_norm": 0.5505324516461024,
"learning_rate": 1.999525719713366e-05,
"loss": 0.0924,
"step": 62
},
{
"epoch": 0.7777777777777778,
"grad_norm": 0.6290628543373665,
"learning_rate": 1.9993170601430233e-05,
"loss": 0.1015,
"step": 63
},
{
"epoch": 0.7901234567901234,
"grad_norm": 0.5654581976010135,
"learning_rate": 1.9990704811844934e-05,
"loss": 0.0955,
"step": 64
},
{
"epoch": 0.8024691358024691,
"grad_norm": 0.6143207836622308,
"learning_rate": 1.99878599219429e-05,
"loss": 0.1045,
"step": 65
},
{
"epoch": 0.8148148148148148,
"grad_norm": 0.6398660896422187,
"learning_rate": 1.9984636039674342e-05,
"loss": 0.0873,
"step": 66
},
{
"epoch": 0.8271604938271605,
"grad_norm": 0.6398310694406929,
"learning_rate": 1.9981033287370443e-05,
"loss": 0.1089,
"step": 67
},
{
"epoch": 0.8395061728395061,
"grad_norm": 0.6948653198177049,
"learning_rate": 1.9977051801738733e-05,
"loss": 0.1364,
"step": 68
},
{
"epoch": 0.8518518518518519,
"grad_norm": 0.6005769842034276,
"learning_rate": 1.997269173385788e-05,
"loss": 0.092,
"step": 69
},
{
"epoch": 0.8641975308641975,
"grad_norm": 0.6977886617269095,
"learning_rate": 1.996795324917199e-05,
"loss": 0.1083,
"step": 70
},
{
"epoch": 0.8765432098765432,
"grad_norm": 0.6065814882574566,
"learning_rate": 1.9962836527484296e-05,
"loss": 0.0818,
"step": 71
},
{
"epoch": 0.8888888888888888,
"grad_norm": 0.6615848160240534,
"learning_rate": 1.9957341762950346e-05,
"loss": 0.0981,
"step": 72
},
{
"epoch": 0.9012345679012346,
"grad_norm": 0.5931653814016081,
"learning_rate": 1.9951469164070647e-05,
"loss": 0.083,
"step": 73
},
{
"epoch": 0.9135802469135802,
"grad_norm": 0.5186114507696477,
"learning_rate": 1.9945218953682736e-05,
"loss": 0.0809,
"step": 74
},
{
"epoch": 0.9259259259259259,
"grad_norm": 0.573114746431398,
"learning_rate": 1.993859136895274e-05,
"loss": 0.0848,
"step": 75
},
{
"epoch": 0.9382716049382716,
"grad_norm": 0.5344071067213606,
"learning_rate": 1.9931586661366363e-05,
"loss": 0.0815,
"step": 76
},
{
"epoch": 0.9506172839506173,
"grad_norm": 0.6377105774122757,
"learning_rate": 1.992420509671936e-05,
"loss": 0.0879,
"step": 77
},
{
"epoch": 0.9629629629629629,
"grad_norm": 0.5209583162843623,
"learning_rate": 1.991644695510743e-05,
"loss": 0.0805,
"step": 78
},
{
"epoch": 0.9753086419753086,
"grad_norm": 0.417423387838958,
"learning_rate": 1.9908312530915603e-05,
"loss": 0.0634,
"step": 79
},
{
"epoch": 0.9876543209876543,
"grad_norm": 0.5230529318454548,
"learning_rate": 1.9899802132807073e-05,
"loss": 0.074,
"step": 80
},
{
"epoch": 1.0,
"grad_norm": 0.571800306158981,
"learning_rate": 1.9890916083711463e-05,
"loss": 0.0896,
"step": 81
},
{
"epoch": 1.0123456790123457,
"grad_norm": 0.48654258298002473,
"learning_rate": 1.9881654720812594e-05,
"loss": 0.0616,
"step": 82
},
{
"epoch": 1.0246913580246915,
"grad_norm": 0.49061163663393764,
"learning_rate": 1.9872018395535694e-05,
"loss": 0.0546,
"step": 83
},
{
"epoch": 1.037037037037037,
"grad_norm": 0.5362863528158126,
"learning_rate": 1.9862007473534026e-05,
"loss": 0.0633,
"step": 84
},
{
"epoch": 1.0493827160493827,
"grad_norm": 0.48289786827410613,
"learning_rate": 1.9851622334675065e-05,
"loss": 0.0651,
"step": 85
},
{
"epoch": 1.0617283950617284,
"grad_norm": 0.4191994966523693,
"learning_rate": 1.9840863373026046e-05,
"loss": 0.0378,
"step": 86
},
{
"epoch": 1.074074074074074,
"grad_norm": 0.48414987204753185,
"learning_rate": 1.982973099683902e-05,
"loss": 0.0479,
"step": 87
},
{
"epoch": 1.0864197530864197,
"grad_norm": 0.47673601747945576,
"learning_rate": 1.9818225628535372e-05,
"loss": 0.0482,
"step": 88
},
{
"epoch": 1.0987654320987654,
"grad_norm": 0.4668133839134151,
"learning_rate": 1.9806347704689778e-05,
"loss": 0.0691,
"step": 89
},
{
"epoch": 1.1111111111111112,
"grad_norm": 0.40912483085288526,
"learning_rate": 1.979409767601366e-05,
"loss": 0.0473,
"step": 90
},
{
"epoch": 1.123456790123457,
"grad_norm": 0.46287321439131024,
"learning_rate": 1.9781476007338058e-05,
"loss": 0.0487,
"step": 91
},
{
"epoch": 1.1358024691358024,
"grad_norm": 0.46265781731853756,
"learning_rate": 1.9768483177596008e-05,
"loss": 0.0548,
"step": 92
},
{
"epoch": 1.1481481481481481,
"grad_norm": 0.43307375489466543,
"learning_rate": 1.975511967980437e-05,
"loss": 0.045,
"step": 93
},
{
"epoch": 1.1604938271604939,
"grad_norm": 0.4715524387710138,
"learning_rate": 1.9741386021045105e-05,
"loss": 0.0417,
"step": 94
},
{
"epoch": 1.1728395061728394,
"grad_norm": 0.42072406617196306,
"learning_rate": 1.972728272244605e-05,
"loss": 0.0472,
"step": 95
},
{
"epoch": 1.1851851851851851,
"grad_norm": 0.43172521185288276,
"learning_rate": 1.971281031916114e-05,
"loss": 0.0377,
"step": 96
},
{
"epoch": 1.1975308641975309,
"grad_norm": 0.4497872448243483,
"learning_rate": 1.9697969360350098e-05,
"loss": 0.042,
"step": 97
},
{
"epoch": 1.2098765432098766,
"grad_norm": 0.4015427782061893,
"learning_rate": 1.968276040915759e-05,
"loss": 0.0467,
"step": 98
},
{
"epoch": 1.2222222222222223,
"grad_norm": 0.46433315248898455,
"learning_rate": 1.9667184042691877e-05,
"loss": 0.0781,
"step": 99
},
{
"epoch": 1.2345679012345678,
"grad_norm": 0.4607108700832256,
"learning_rate": 1.965124085200289e-05,
"loss": 0.0484,
"step": 100
},
{
"epoch": 1.2469135802469136,
"grad_norm": 0.4580474315592068,
"learning_rate": 1.9634931442059833e-05,
"loss": 0.0497,
"step": 101
},
{
"epoch": 1.2592592592592593,
"grad_norm": 0.4376923035257714,
"learning_rate": 1.961825643172819e-05,
"loss": 0.0372,
"step": 102
},
{
"epoch": 1.2716049382716048,
"grad_norm": 0.47470979724718027,
"learning_rate": 1.9601216453746285e-05,
"loss": 0.0412,
"step": 103
},
{
"epoch": 1.2839506172839505,
"grad_norm": 0.48607721607358306,
"learning_rate": 1.9583812154701227e-05,
"loss": 0.063,
"step": 104
},
{
"epoch": 1.2962962962962963,
"grad_norm": 0.3698617332173753,
"learning_rate": 1.956604419500441e-05,
"loss": 0.0531,
"step": 105
},
{
"epoch": 1.308641975308642,
"grad_norm": 0.4394488729697042,
"learning_rate": 1.9547913248866447e-05,
"loss": 0.0629,
"step": 106
},
{
"epoch": 1.3209876543209877,
"grad_norm": 0.36774414517262893,
"learning_rate": 1.9529420004271568e-05,
"loss": 0.0465,
"step": 107
},
{
"epoch": 1.3333333333333333,
"grad_norm": 0.4084826218592715,
"learning_rate": 1.9510565162951538e-05,
"loss": 0.0375,
"step": 108
},
{
"epoch": 1.345679012345679,
"grad_norm": 0.3974187863812691,
"learning_rate": 1.9491349440359014e-05,
"loss": 0.0372,
"step": 109
},
{
"epoch": 1.3580246913580247,
"grad_norm": 0.43146975975134916,
"learning_rate": 1.9471773565640405e-05,
"loss": 0.0366,
"step": 110
},
{
"epoch": 1.3703703703703702,
"grad_norm": 0.3812925184125817,
"learning_rate": 1.94518382816082e-05,
"loss": 0.0432,
"step": 111
},
{
"epoch": 1.382716049382716,
"grad_norm": 0.5295352324623606,
"learning_rate": 1.9431544344712776e-05,
"loss": 0.0643,
"step": 112
},
{
"epoch": 1.3950617283950617,
"grad_norm": 0.46291492570157333,
"learning_rate": 1.9410892525013717e-05,
"loss": 0.0419,
"step": 113
},
{
"epoch": 1.4074074074074074,
"grad_norm": 0.3646553984739498,
"learning_rate": 1.938988360615057e-05,
"loss": 0.0376,
"step": 114
},
{
"epoch": 1.4197530864197532,
"grad_norm": 0.4120840867522972,
"learning_rate": 1.9368518385313108e-05,
"loss": 0.0277,
"step": 115
},
{
"epoch": 1.4320987654320987,
"grad_norm": 0.3678320151928319,
"learning_rate": 1.9346797673211108e-05,
"loss": 0.0472,
"step": 116
},
{
"epoch": 1.4444444444444444,
"grad_norm": 0.3509880150630426,
"learning_rate": 1.932472229404356e-05,
"loss": 0.0532,
"step": 117
},
{
"epoch": 1.4567901234567902,
"grad_norm": 0.5013873051103777,
"learning_rate": 1.9302293085467404e-05,
"loss": 0.0468,
"step": 118
},
{
"epoch": 1.4691358024691357,
"grad_norm": 0.37461635613832134,
"learning_rate": 1.927951089856575e-05,
"loss": 0.0393,
"step": 119
},
{
"epoch": 1.4814814814814814,
"grad_norm": 0.6569993579112918,
"learning_rate": 1.9256376597815565e-05,
"loss": 0.0506,
"step": 120
},
{
"epoch": 1.4938271604938271,
"grad_norm": 0.3606538758277819,
"learning_rate": 1.9232891061054896e-05,
"loss": 0.0309,
"step": 121
},
{
"epoch": 1.5061728395061729,
"grad_norm": 0.4489552042946805,
"learning_rate": 1.920905517944954e-05,
"loss": 0.0495,
"step": 122
},
{
"epoch": 1.5185185185185186,
"grad_norm": 0.3640183154186618,
"learning_rate": 1.9184869857459233e-05,
"loss": 0.0294,
"step": 123
},
{
"epoch": 1.5308641975308643,
"grad_norm": 0.4987631065093897,
"learning_rate": 1.916033601280334e-05,
"loss": 0.0533,
"step": 124
},
{
"epoch": 1.5432098765432098,
"grad_norm": 0.40319894875583845,
"learning_rate": 1.913545457642601e-05,
"loss": 0.0386,
"step": 125
},
{
"epoch": 1.5555555555555556,
"grad_norm": 0.3044060536431661,
"learning_rate": 1.9110226492460886e-05,
"loss": 0.0342,
"step": 126
},
{
"epoch": 1.567901234567901,
"grad_norm": 0.36919281340072985,
"learning_rate": 1.9084652718195237e-05,
"loss": 0.034,
"step": 127
},
{
"epoch": 1.5802469135802468,
"grad_norm": 0.4554316645468543,
"learning_rate": 1.9058734224033673e-05,
"loss": 0.0442,
"step": 128
},
{
"epoch": 1.5925925925925926,
"grad_norm": 0.3860637450416904,
"learning_rate": 1.903247199346129e-05,
"loss": 0.0351,
"step": 129
},
{
"epoch": 1.6049382716049383,
"grad_norm": 0.4145904111255904,
"learning_rate": 1.9005867023006374e-05,
"loss": 0.0538,
"step": 130
},
{
"epoch": 1.617283950617284,
"grad_norm": 0.37152108124728694,
"learning_rate": 1.8978920322202582e-05,
"loss": 0.0372,
"step": 131
},
{
"epoch": 1.6296296296296298,
"grad_norm": 0.3533748489120987,
"learning_rate": 1.8951632913550625e-05,
"loss": 0.0347,
"step": 132
},
{
"epoch": 1.6419753086419753,
"grad_norm": 0.41307921206850035,
"learning_rate": 1.892400583247948e-05,
"loss": 0.0467,
"step": 133
},
{
"epoch": 1.654320987654321,
"grad_norm": 0.3623309750124559,
"learning_rate": 1.88960401273071e-05,
"loss": 0.0339,
"step": 134
},
{
"epoch": 1.6666666666666665,
"grad_norm": 0.32342684047692133,
"learning_rate": 1.886773685920062e-05,
"loss": 0.0475,
"step": 135
},
{
"epoch": 1.6790123456790123,
"grad_norm": 0.2751221763038461,
"learning_rate": 1.8839097102136123e-05,
"loss": 0.0411,
"step": 136
},
{
"epoch": 1.691358024691358,
"grad_norm": 0.4309033939961943,
"learning_rate": 1.8810121942857848e-05,
"loss": 0.0499,
"step": 137
},
{
"epoch": 1.7037037037037037,
"grad_norm": 0.40460125045082745,
"learning_rate": 1.878081248083698e-05,
"loss": 0.0704,
"step": 138
},
{
"epoch": 1.7160493827160495,
"grad_norm": 0.33237280139195097,
"learning_rate": 1.8751169828229927e-05,
"loss": 0.0288,
"step": 139
},
{
"epoch": 1.7283950617283952,
"grad_norm": 0.33276164181630363,
"learning_rate": 1.872119510983611e-05,
"loss": 0.0283,
"step": 140
},
{
"epoch": 1.7407407407407407,
"grad_norm": 0.33846042235910045,
"learning_rate": 1.8690889463055285e-05,
"loss": 0.0346,
"step": 141
},
{
"epoch": 1.7530864197530864,
"grad_norm": 0.2983835469497029,
"learning_rate": 1.866025403784439e-05,
"loss": 0.0204,
"step": 142
},
{
"epoch": 1.765432098765432,
"grad_norm": 0.32286554382633115,
"learning_rate": 1.86292899966739e-05,
"loss": 0.0564,
"step": 143
},
{
"epoch": 1.7777777777777777,
"grad_norm": 0.29494797125215033,
"learning_rate": 1.8597998514483724e-05,
"loss": 0.0199,
"step": 144
},
{
"epoch": 1.7901234567901234,
"grad_norm": 0.31485674246853385,
"learning_rate": 1.856638077863863e-05,
"loss": 0.0246,
"step": 145
},
{
"epoch": 1.8024691358024691,
"grad_norm": 0.3553330613625899,
"learning_rate": 1.853443798888316e-05,
"loss": 0.0428,
"step": 146
},
{
"epoch": 1.8148148148148149,
"grad_norm": 0.4094124323752043,
"learning_rate": 1.8502171357296144e-05,
"loss": 0.0325,
"step": 147
},
{
"epoch": 1.8271604938271606,
"grad_norm": 0.4512274400089503,
"learning_rate": 1.8469582108244672e-05,
"loss": 0.0316,
"step": 148
},
{
"epoch": 1.8395061728395061,
"grad_norm": 0.3386207896347789,
"learning_rate": 1.8436671478337666e-05,
"loss": 0.0264,
"step": 149
},
{
"epoch": 1.8518518518518519,
"grad_norm": 0.30377308765711836,
"learning_rate": 1.840344071637893e-05,
"loss": 0.0256,
"step": 150
},
{
"epoch": 1.8641975308641974,
"grad_norm": 0.3047593078910109,
"learning_rate": 1.836989108331978e-05,
"loss": 0.0406,
"step": 151
},
{
"epoch": 1.876543209876543,
"grad_norm": 0.3453918783190626,
"learning_rate": 1.8336023852211197e-05,
"loss": 0.0293,
"step": 152
},
{
"epoch": 1.8888888888888888,
"grad_norm": 0.34020059500334965,
"learning_rate": 1.8301840308155507e-05,
"loss": 0.0376,
"step": 153
},
{
"epoch": 1.9012345679012346,
"grad_norm": 0.3898075678508035,
"learning_rate": 1.8267341748257636e-05,
"loss": 0.0339,
"step": 154
},
{
"epoch": 1.9135802469135803,
"grad_norm": 0.5039729606941286,
"learning_rate": 1.8232529481575874e-05,
"loss": 0.0237,
"step": 155
},
{
"epoch": 1.925925925925926,
"grad_norm": 0.39939423124671763,
"learning_rate": 1.8197404829072214e-05,
"loss": 0.0567,
"step": 156
},
{
"epoch": 1.9382716049382716,
"grad_norm": 0.3110373822433647,
"learning_rate": 1.816196912356222e-05,
"loss": 0.0263,
"step": 157
},
{
"epoch": 1.9506172839506173,
"grad_norm": 0.2730898361055386,
"learning_rate": 1.812622370966446e-05,
"loss": 0.0167,
"step": 158
},
{
"epoch": 1.9629629629629628,
"grad_norm": 0.3740775592122447,
"learning_rate": 1.8090169943749477e-05,
"loss": 0.0366,
"step": 159
},
{
"epoch": 1.9753086419753085,
"grad_norm": 0.3517482972136562,
"learning_rate": 1.8053809193888326e-05,
"loss": 0.0343,
"step": 160
},
{
"epoch": 1.9876543209876543,
"grad_norm": 0.3448201470407539,
"learning_rate": 1.8017142839800667e-05,
"loss": 0.0407,
"step": 161
},
{
"epoch": 2.0,
"grad_norm": 0.31463659127167554,
"learning_rate": 1.7980172272802398e-05,
"loss": 0.0326,
"step": 162
},
{
"epoch": 2.0123456790123457,
"grad_norm": 0.3119319557778762,
"learning_rate": 1.794289889575286e-05,
"loss": 0.0162,
"step": 163
},
{
"epoch": 2.0246913580246915,
"grad_norm": 0.32152041344430515,
"learning_rate": 1.7905324123001634e-05,
"loss": 0.0253,
"step": 164
},
{
"epoch": 2.037037037037037,
"grad_norm": 0.2773957765251134,
"learning_rate": 1.7867449380334834e-05,
"loss": 0.0203,
"step": 165
},
{
"epoch": 2.049382716049383,
"grad_norm": 0.22908870780721655,
"learning_rate": 1.782927610492103e-05,
"loss": 0.0186,
"step": 166
},
{
"epoch": 2.0617283950617282,
"grad_norm": 0.2955956970727338,
"learning_rate": 1.7790805745256703e-05,
"loss": 0.02,
"step": 167
},
{
"epoch": 2.074074074074074,
"grad_norm": 0.34426895222254217,
"learning_rate": 1.77520397611113e-05,
"loss": 0.0206,
"step": 168
},
{
"epoch": 2.0864197530864197,
"grad_norm": 0.3394383274911477,
"learning_rate": 1.771297962347181e-05,
"loss": 0.021,
"step": 169
},
{
"epoch": 2.0987654320987654,
"grad_norm": 0.3174782568269176,
"learning_rate": 1.767362681448697e-05,
"loss": 0.0223,
"step": 170
},
{
"epoch": 2.111111111111111,
"grad_norm": 0.28793132810552485,
"learning_rate": 1.763398282741103e-05,
"loss": 0.015,
"step": 171
},
{
"epoch": 2.123456790123457,
"grad_norm": 0.25904756194914924,
"learning_rate": 1.7594049166547073e-05,
"loss": 0.0152,
"step": 172
},
{
"epoch": 2.1358024691358026,
"grad_norm": 0.28872886550519916,
"learning_rate": 1.7553827347189937e-05,
"loss": 0.0218,
"step": 173
},
{
"epoch": 2.148148148148148,
"grad_norm": 0.3383745922979738,
"learning_rate": 1.7513318895568734e-05,
"loss": 0.0186,
"step": 174
},
{
"epoch": 2.1604938271604937,
"grad_norm": 0.3433669022459947,
"learning_rate": 1.747252534878891e-05,
"loss": 0.0244,
"step": 175
},
{
"epoch": 2.1728395061728394,
"grad_norm": 0.3096900118277167,
"learning_rate": 1.7431448254773943e-05,
"loss": 0.0247,
"step": 176
},
{
"epoch": 2.185185185185185,
"grad_norm": 0.3160661155338934,
"learning_rate": 1.7390089172206594e-05,
"loss": 0.0204,
"step": 177
},
{
"epoch": 2.197530864197531,
"grad_norm": 0.3031506421614271,
"learning_rate": 1.7348449670469758e-05,
"loss": 0.0247,
"step": 178
},
{
"epoch": 2.2098765432098766,
"grad_norm": 0.29699710884206193,
"learning_rate": 1.7306531329586933e-05,
"loss": 0.0212,
"step": 179
},
{
"epoch": 2.2222222222222223,
"grad_norm": 0.27055995416522066,
"learning_rate": 1.7264335740162244e-05,
"loss": 0.0141,
"step": 180
},
{
"epoch": 2.234567901234568,
"grad_norm": 0.2513720939615437,
"learning_rate": 1.7221864503320093e-05,
"loss": 0.0176,
"step": 181
},
{
"epoch": 2.246913580246914,
"grad_norm": 0.21463355334635534,
"learning_rate": 1.717911923064442e-05,
"loss": 0.0195,
"step": 182
},
{
"epoch": 2.259259259259259,
"grad_norm": 0.3057377106103357,
"learning_rate": 1.7136101544117526e-05,
"loss": 0.0216,
"step": 183
},
{
"epoch": 2.271604938271605,
"grad_norm": 0.29775222163099,
"learning_rate": 1.7092813076058536e-05,
"loss": 0.0275,
"step": 184
},
{
"epoch": 2.2839506172839505,
"grad_norm": 0.3055276236869669,
"learning_rate": 1.7049255469061476e-05,
"loss": 0.0344,
"step": 185
},
{
"epoch": 2.2962962962962963,
"grad_norm": 0.21549260314464128,
"learning_rate": 1.700543037593291e-05,
"loss": 0.0168,
"step": 186
},
{
"epoch": 2.308641975308642,
"grad_norm": 0.2617363419016656,
"learning_rate": 1.696133945962927e-05,
"loss": 0.0223,
"step": 187
},
{
"epoch": 2.3209876543209877,
"grad_norm": 0.24161925121125105,
"learning_rate": 1.6916984393193704e-05,
"loss": 0.0149,
"step": 188
},
{
"epoch": 2.3333333333333335,
"grad_norm": 0.3097073679033491,
"learning_rate": 1.687236685969263e-05,
"loss": 0.0278,
"step": 189
},
{
"epoch": 2.3456790123456788,
"grad_norm": 0.25261420518988137,
"learning_rate": 1.6827488552151855e-05,
"loss": 0.0209,
"step": 190
},
{
"epoch": 2.3580246913580245,
"grad_norm": 0.29919385437925194,
"learning_rate": 1.678235117349234e-05,
"loss": 0.021,
"step": 191
},
{
"epoch": 2.3703703703703702,
"grad_norm": 0.26627559608374113,
"learning_rate": 1.6736956436465573e-05,
"loss": 0.0156,
"step": 192
},
{
"epoch": 2.382716049382716,
"grad_norm": 0.315756273819458,
"learning_rate": 1.6691306063588583e-05,
"loss": 0.0119,
"step": 193
},
{
"epoch": 2.3950617283950617,
"grad_norm": 0.3521313813287778,
"learning_rate": 1.664540178707858e-05,
"loss": 0.0197,
"step": 194
},
{
"epoch": 2.4074074074074074,
"grad_norm": 0.23890807245570367,
"learning_rate": 1.659924534878723e-05,
"loss": 0.0161,
"step": 195
},
{
"epoch": 2.419753086419753,
"grad_norm": 0.23229593095773818,
"learning_rate": 1.655283850013454e-05,
"loss": 0.0082,
"step": 196
},
{
"epoch": 2.432098765432099,
"grad_norm": 0.2488497617155846,
"learning_rate": 1.650618300204242e-05,
"loss": 0.0154,
"step": 197
},
{
"epoch": 2.4444444444444446,
"grad_norm": 0.2846644237202055,
"learning_rate": 1.6459280624867876e-05,
"loss": 0.017,
"step": 198
},
{
"epoch": 2.45679012345679,
"grad_norm": 0.3307306653758879,
"learning_rate": 1.6412133148335786e-05,
"loss": 0.0253,
"step": 199
},
{
"epoch": 2.4691358024691357,
"grad_norm": 0.401524089447455,
"learning_rate": 1.6364742361471416e-05,
"loss": 0.0248,
"step": 200
},
{
"epoch": 2.4814814814814814,
"grad_norm": 0.24557760407685852,
"learning_rate": 1.631711006253251e-05,
"loss": 0.0219,
"step": 201
},
{
"epoch": 2.493827160493827,
"grad_norm": 0.2231254084874177,
"learning_rate": 1.626923805894107e-05,
"loss": 0.0185,
"step": 202
},
{
"epoch": 2.506172839506173,
"grad_norm": 0.2746670149754389,
"learning_rate": 1.6221128167214742e-05,
"loss": 0.0198,
"step": 203
},
{
"epoch": 2.5185185185185186,
"grad_norm": 0.22454413205332946,
"learning_rate": 1.617278221289793e-05,
"loss": 0.0131,
"step": 204
},
{
"epoch": 2.5308641975308643,
"grad_norm": 0.23942689316296767,
"learning_rate": 1.61242020304925e-05,
"loss": 0.0216,
"step": 205
},
{
"epoch": 2.5432098765432096,
"grad_norm": 0.3167065348394166,
"learning_rate": 1.607538946338817e-05,
"loss": 0.015,
"step": 206
},
{
"epoch": 2.5555555555555554,
"grad_norm": 0.3522580744129786,
"learning_rate": 1.6026346363792565e-05,
"loss": 0.0284,
"step": 207
},
{
"epoch": 2.567901234567901,
"grad_norm": 0.2988578750438895,
"learning_rate": 1.5977074592660936e-05,
"loss": 0.0166,
"step": 208
},
{
"epoch": 2.580246913580247,
"grad_norm": 0.2119054892726317,
"learning_rate": 1.592757601962555e-05,
"loss": 0.0096,
"step": 209
},
{
"epoch": 2.5925925925925926,
"grad_norm": 0.20741064911092552,
"learning_rate": 1.5877852522924733e-05,
"loss": 0.0191,
"step": 210
},
{
"epoch": 2.6049382716049383,
"grad_norm": 0.26294749753775415,
"learning_rate": 1.582790598933161e-05,
"loss": 0.0265,
"step": 211
},
{
"epoch": 2.617283950617284,
"grad_norm": 0.2521344397345761,
"learning_rate": 1.5777738314082514e-05,
"loss": 0.0174,
"step": 212
},
{
"epoch": 2.6296296296296298,
"grad_norm": 0.2090224535341981,
"learning_rate": 1.5727351400805054e-05,
"loss": 0.0159,
"step": 213
},
{
"epoch": 2.6419753086419755,
"grad_norm": 0.21033486854745503,
"learning_rate": 1.5676747161445903e-05,
"loss": 0.0236,
"step": 214
},
{
"epoch": 2.6543209876543212,
"grad_norm": 0.19657648924979762,
"learning_rate": 1.5625927516198235e-05,
"loss": 0.0178,
"step": 215
},
{
"epoch": 2.6666666666666665,
"grad_norm": 0.2411052605645195,
"learning_rate": 1.5574894393428856e-05,
"loss": 0.0111,
"step": 216
},
{
"epoch": 2.6790123456790123,
"grad_norm": 0.31954942273313647,
"learning_rate": 1.552364972960506e-05,
"loss": 0.026,
"step": 217
},
{
"epoch": 2.691358024691358,
"grad_norm": 0.2838958762793406,
"learning_rate": 1.5472195469221115e-05,
"loss": 0.0309,
"step": 218
},
{
"epoch": 2.7037037037037037,
"grad_norm": 0.20041777320173626,
"learning_rate": 1.5420533564724495e-05,
"loss": 0.0183,
"step": 219
},
{
"epoch": 2.7160493827160495,
"grad_norm": 0.2450116676808879,
"learning_rate": 1.5368665976441802e-05,
"loss": 0.0209,
"step": 220
},
{
"epoch": 2.728395061728395,
"grad_norm": 0.23961047907329863,
"learning_rate": 1.5316594672504362e-05,
"loss": 0.0177,
"step": 221
},
{
"epoch": 2.7407407407407405,
"grad_norm": 0.22107765498822843,
"learning_rate": 1.526432162877356e-05,
"loss": 0.0115,
"step": 222
},
{
"epoch": 2.753086419753086,
"grad_norm": 0.283474952878893,
"learning_rate": 1.5211848828765852e-05,
"loss": 0.034,
"step": 223
},
{
"epoch": 2.765432098765432,
"grad_norm": 0.21699559982290464,
"learning_rate": 1.5159178263577512e-05,
"loss": 0.0101,
"step": 224
},
{
"epoch": 2.7777777777777777,
"grad_norm": 0.22698992910116217,
"learning_rate": 1.510631193180907e-05,
"loss": 0.0251,
"step": 225
},
{
"epoch": 2.7901234567901234,
"grad_norm": 0.21582178770974225,
"learning_rate": 1.5053251839489482e-05,
"loss": 0.0108,
"step": 226
},
{
"epoch": 2.802469135802469,
"grad_norm": 0.26829257899099024,
"learning_rate": 1.5000000000000002e-05,
"loss": 0.0209,
"step": 227
},
{
"epoch": 2.814814814814815,
"grad_norm": 0.3507275921603246,
"learning_rate": 1.4946558433997792e-05,
"loss": 0.0264,
"step": 228
},
{
"epoch": 2.8271604938271606,
"grad_norm": 0.17972895444262435,
"learning_rate": 1.4892929169339237e-05,
"loss": 0.0064,
"step": 229
},
{
"epoch": 2.8395061728395063,
"grad_norm": 0.2617950316333363,
"learning_rate": 1.4839114241003017e-05,
"loss": 0.0205,
"step": 230
},
{
"epoch": 2.851851851851852,
"grad_norm": 0.21854432114775668,
"learning_rate": 1.4785115691012866e-05,
"loss": 0.0133,
"step": 231
},
{
"epoch": 2.8641975308641974,
"grad_norm": 0.3086185413115422,
"learning_rate": 1.4730935568360103e-05,
"loss": 0.0258,
"step": 232
},
{
"epoch": 2.876543209876543,
"grad_norm": 0.26360881813895337,
"learning_rate": 1.4676575928925869e-05,
"loss": 0.0122,
"step": 233
},
{
"epoch": 2.888888888888889,
"grad_norm": 0.19397354302951686,
"learning_rate": 1.4622038835403135e-05,
"loss": 0.0189,
"step": 234
},
{
"epoch": 2.9012345679012346,
"grad_norm": 0.27942768599957285,
"learning_rate": 1.4567326357218408e-05,
"loss": 0.025,
"step": 235
},
{
"epoch": 2.9135802469135803,
"grad_norm": 0.2950242324226194,
"learning_rate": 1.451244057045323e-05,
"loss": 0.0365,
"step": 236
},
{
"epoch": 2.925925925925926,
"grad_norm": 0.17975187594689376,
"learning_rate": 1.4457383557765385e-05,
"loss": 0.0188,
"step": 237
},
{
"epoch": 2.9382716049382713,
"grad_norm": 0.23376665230749175,
"learning_rate": 1.4402157408309876e-05,
"loss": 0.0099,
"step": 238
},
{
"epoch": 2.950617283950617,
"grad_norm": 0.19019995851535298,
"learning_rate": 1.4346764217659652e-05,
"loss": 0.0162,
"step": 239
},
{
"epoch": 2.962962962962963,
"grad_norm": 0.2214940835243357,
"learning_rate": 1.429120608772609e-05,
"loss": 0.0237,
"step": 240
},
{
"epoch": 2.9753086419753085,
"grad_norm": 0.2627103929594842,
"learning_rate": 1.4235485126679244e-05,
"loss": 0.0157,
"step": 241
},
{
"epoch": 2.9876543209876543,
"grad_norm": 0.2682992652202373,
"learning_rate": 1.4179603448867836e-05,
"loss": 0.0298,
"step": 242
},
{
"epoch": 3.0,
"grad_norm": 0.298232489464775,
"learning_rate": 1.4123563174739036e-05,
"loss": 0.0152,
"step": 243
},
{
"epoch": 3.0123456790123457,
"grad_norm": 0.12740632304155414,
"learning_rate": 1.4067366430758004e-05,
"loss": 0.0059,
"step": 244
},
{
"epoch": 3.0246913580246915,
"grad_norm": 0.19794558417137054,
"learning_rate": 1.4011015349327188e-05,
"loss": 0.0088,
"step": 245
},
{
"epoch": 3.037037037037037,
"grad_norm": 0.1922683196961256,
"learning_rate": 1.3954512068705425e-05,
"loss": 0.0081,
"step": 246
},
{
"epoch": 3.049382716049383,
"grad_norm": 0.17829186059186022,
"learning_rate": 1.3897858732926794e-05,
"loss": 0.0071,
"step": 247
},
{
"epoch": 3.0617283950617282,
"grad_norm": 0.14264428973058538,
"learning_rate": 1.3841057491719261e-05,
"loss": 0.0096,
"step": 248
},
{
"epoch": 3.074074074074074,
"grad_norm": 0.2266151450801307,
"learning_rate": 1.3784110500423104e-05,
"loss": 0.0154,
"step": 249
},
{
"epoch": 3.0864197530864197,
"grad_norm": 0.2197194981703044,
"learning_rate": 1.372701991990914e-05,
"loss": 0.0231,
"step": 250
},
{
"epoch": 3.0987654320987654,
"grad_norm": 0.21131738021284832,
"learning_rate": 1.3669787916496722e-05,
"loss": 0.0094,
"step": 251
},
{
"epoch": 3.111111111111111,
"grad_norm": 0.20464969805446995,
"learning_rate": 1.3612416661871532e-05,
"loss": 0.0085,
"step": 252
},
{
"epoch": 3.123456790123457,
"grad_norm": 0.21532674194312212,
"learning_rate": 1.355490833300318e-05,
"loss": 0.0087,
"step": 253
},
{
"epoch": 3.1358024691358026,
"grad_norm": 0.16104981288328868,
"learning_rate": 1.3497265112062613e-05,
"loss": 0.0055,
"step": 254
},
{
"epoch": 3.148148148148148,
"grad_norm": 0.24076058570830933,
"learning_rate": 1.3439489186339283e-05,
"loss": 0.008,
"step": 255
},
{
"epoch": 3.1604938271604937,
"grad_norm": 0.2604821262949657,
"learning_rate": 1.3381582748158173e-05,
"loss": 0.0231,
"step": 256
},
{
"epoch": 3.1728395061728394,
"grad_norm": 0.24580486733045098,
"learning_rate": 1.3323547994796597e-05,
"loss": 0.0174,
"step": 257
},
{
"epoch": 3.185185185185185,
"grad_norm": 0.3154617969549976,
"learning_rate": 1.3265387128400833e-05,
"loss": 0.0092,
"step": 258
},
{
"epoch": 3.197530864197531,
"grad_norm": 0.2224133620360006,
"learning_rate": 1.3207102355902553e-05,
"loss": 0.0129,
"step": 259
},
{
"epoch": 3.2098765432098766,
"grad_norm": 0.281666620283078,
"learning_rate": 1.314869588893508e-05,
"loss": 0.0114,
"step": 260
},
{
"epoch": 3.2222222222222223,
"grad_norm": 0.22582566658451753,
"learning_rate": 1.3090169943749475e-05,
"loss": 0.0125,
"step": 261
},
{
"epoch": 3.234567901234568,
"grad_norm": 0.20743608754259807,
"learning_rate": 1.3031526741130435e-05,
"loss": 0.0133,
"step": 262
},
{
"epoch": 3.246913580246914,
"grad_norm": 0.1823778790146933,
"learning_rate": 1.2972768506312028e-05,
"loss": 0.0097,
"step": 263
},
{
"epoch": 3.259259259259259,
"grad_norm": 0.24794430856710525,
"learning_rate": 1.2913897468893249e-05,
"loss": 0.0081,
"step": 264
},
{
"epoch": 3.271604938271605,
"grad_norm": 0.22571751396689432,
"learning_rate": 1.2854915862753424e-05,
"loss": 0.0114,
"step": 265
},
{
"epoch": 3.2839506172839505,
"grad_norm": 0.19127790036678968,
"learning_rate": 1.279582592596744e-05,
"loss": 0.009,
"step": 266
},
{
"epoch": 3.2962962962962963,
"grad_norm": 0.21071669945822885,
"learning_rate": 1.2736629900720832e-05,
"loss": 0.0155,
"step": 267
},
{
"epoch": 3.308641975308642,
"grad_norm": 0.15143256661237275,
"learning_rate": 1.2677330033224681e-05,
"loss": 0.0065,
"step": 268
},
{
"epoch": 3.3209876543209877,
"grad_norm": 0.23256148298533083,
"learning_rate": 1.2617928573630405e-05,
"loss": 0.0124,
"step": 269
},
{
"epoch": 3.3333333333333335,
"grad_norm": 0.22582573521296015,
"learning_rate": 1.2558427775944357e-05,
"loss": 0.0216,
"step": 270
},
{
"epoch": 3.3456790123456788,
"grad_norm": 0.13193162285788634,
"learning_rate": 1.2498829897942308e-05,
"loss": 0.011,
"step": 271
},
{
"epoch": 3.3580246913580245,
"grad_norm": 0.312976420046679,
"learning_rate": 1.2439137201083772e-05,
"loss": 0.0106,
"step": 272
},
{
"epoch": 3.3703703703703702,
"grad_norm": 0.19625961746503687,
"learning_rate": 1.2379351950426188e-05,
"loss": 0.0089,
"step": 273
},
{
"epoch": 3.382716049382716,
"grad_norm": 0.16447584005359223,
"learning_rate": 1.2319476414538982e-05,
"loss": 0.0063,
"step": 274
},
{
"epoch": 3.3950617283950617,
"grad_norm": 0.1965358769188583,
"learning_rate": 1.2259512865417478e-05,
"loss": 0.0079,
"step": 275
},
{
"epoch": 3.4074074074074074,
"grad_norm": 0.2160518293938132,
"learning_rate": 1.2199463578396688e-05,
"loss": 0.0115,
"step": 276
},
{
"epoch": 3.419753086419753,
"grad_norm": 0.15127756963407038,
"learning_rate": 1.2139330832064975e-05,
"loss": 0.0094,
"step": 277
},
{
"epoch": 3.432098765432099,
"grad_norm": 0.15670354823656912,
"learning_rate": 1.2079116908177592e-05,
"loss": 0.0067,
"step": 278
},
{
"epoch": 3.4444444444444446,
"grad_norm": 0.1832029547236401,
"learning_rate": 1.2018824091570103e-05,
"loss": 0.0076,
"step": 279
},
{
"epoch": 3.45679012345679,
"grad_norm": 0.1847955538396133,
"learning_rate": 1.1958454670071671e-05,
"loss": 0.0068,
"step": 280
},
{
"epoch": 3.4691358024691357,
"grad_norm": 0.16708923166356573,
"learning_rate": 1.1898010934418261e-05,
"loss": 0.0077,
"step": 281
},
{
"epoch": 3.4814814814814814,
"grad_norm": 0.219254384910188,
"learning_rate": 1.1837495178165706e-05,
"loss": 0.0136,
"step": 282
},
{
"epoch": 3.493827160493827,
"grad_norm": 0.3031509215534536,
"learning_rate": 1.177690969760269e-05,
"loss": 0.0078,
"step": 283
},
{
"epoch": 3.506172839506173,
"grad_norm": 0.20587394181029134,
"learning_rate": 1.1716256791663599e-05,
"loss": 0.0107,
"step": 284
},
{
"epoch": 3.5185185185185186,
"grad_norm": 0.18349458503479488,
"learning_rate": 1.16555387618413e-05,
"loss": 0.0114,
"step": 285
},
{
"epoch": 3.5308641975308643,
"grad_norm": 0.21953256031427928,
"learning_rate": 1.159475791209981e-05,
"loss": 0.021,
"step": 286
},
{
"epoch": 3.5432098765432096,
"grad_norm": 0.16802927221598177,
"learning_rate": 1.1533916548786856e-05,
"loss": 0.017,
"step": 287
},
{
"epoch": 3.5555555555555554,
"grad_norm": 0.17437530961491682,
"learning_rate": 1.1473016980546377e-05,
"loss": 0.0082,
"step": 288
},
{
"epoch": 3.567901234567901,
"grad_norm": 0.18753660249190723,
"learning_rate": 1.1412061518230916e-05,
"loss": 0.0104,
"step": 289
},
{
"epoch": 3.580246913580247,
"grad_norm": 0.1961816817103867,
"learning_rate": 1.135105247481393e-05,
"loss": 0.0151,
"step": 290
},
{
"epoch": 3.5925925925925926,
"grad_norm": 0.21880871777689972,
"learning_rate": 1.1289992165302036e-05,
"loss": 0.0114,
"step": 291
},
{
"epoch": 3.6049382716049383,
"grad_norm": 0.1704995800988171,
"learning_rate": 1.1228882906647142e-05,
"loss": 0.0082,
"step": 292
},
{
"epoch": 3.617283950617284,
"grad_norm": 0.2617854020277945,
"learning_rate": 1.1167727017658562e-05,
"loss": 0.0116,
"step": 293
},
{
"epoch": 3.6296296296296298,
"grad_norm": 0.17168277914622568,
"learning_rate": 1.1106526818915008e-05,
"loss": 0.0083,
"step": 294
},
{
"epoch": 3.6419753086419755,
"grad_norm": 0.23889824190420522,
"learning_rate": 1.1045284632676535e-05,
"loss": 0.0147,
"step": 295
},
{
"epoch": 3.6543209876543212,
"grad_norm": 0.1539777031660912,
"learning_rate": 1.0984002782796427e-05,
"loss": 0.0164,
"step": 296
},
{
"epoch": 3.6666666666666665,
"grad_norm": 0.15189002638648533,
"learning_rate": 1.092268359463302e-05,
"loss": 0.0093,
"step": 297
},
{
"epoch": 3.6790123456790123,
"grad_norm": 0.1931624710226531,
"learning_rate": 1.0861329394961461e-05,
"loss": 0.0159,
"step": 298
},
{
"epoch": 3.691358024691358,
"grad_norm": 0.19079183444633152,
"learning_rate": 1.0799942511885417e-05,
"loss": 0.0123,
"step": 299
},
{
"epoch": 3.7037037037037037,
"grad_norm": 0.19983669205081916,
"learning_rate": 1.073852527474874e-05,
"loss": 0.0107,
"step": 300
},
{
"epoch": 3.7160493827160495,
"grad_norm": 0.20550325855050947,
"learning_rate": 1.0677080014047076e-05,
"loss": 0.004,
"step": 301
},
{
"epoch": 3.728395061728395,
"grad_norm": 0.2164226790088601,
"learning_rate": 1.0615609061339431e-05,
"loss": 0.011,
"step": 302
},
{
"epoch": 3.7407407407407405,
"grad_norm": 0.14412057991078386,
"learning_rate": 1.05541147491597e-05,
"loss": 0.0051,
"step": 303
},
{
"epoch": 3.753086419753086,
"grad_norm": 0.2650461598250637,
"learning_rate": 1.049259941092817e-05,
"loss": 0.0101,
"step": 304
},
{
"epoch": 3.765432098765432,
"grad_norm": 0.30085979106051064,
"learning_rate": 1.0431065380862959e-05,
"loss": 0.0132,
"step": 305
},
{
"epoch": 3.7777777777777777,
"grad_norm": 0.24498181798018345,
"learning_rate": 1.0369514993891451e-05,
"loss": 0.0106,
"step": 306
},
{
"epoch": 3.7901234567901234,
"grad_norm": 0.2082184557258639,
"learning_rate": 1.0307950585561705e-05,
"loss": 0.0109,
"step": 307
},
{
"epoch": 3.802469135802469,
"grad_norm": 0.1840429660303778,
"learning_rate": 1.0246374491953823e-05,
"loss": 0.0057,
"step": 308
},
{
"epoch": 3.814814814814815,
"grad_norm": 0.1543192699947808,
"learning_rate": 1.01847890495913e-05,
"loss": 0.0057,
"step": 309
},
{
"epoch": 3.8271604938271606,
"grad_norm": 0.31085221751147407,
"learning_rate": 1.0123196595352385e-05,
"loss": 0.0182,
"step": 310
},
{
"epoch": 3.8395061728395063,
"grad_norm": 0.1871258490697948,
"learning_rate": 1.0061599466381388e-05,
"loss": 0.0145,
"step": 311
},
{
"epoch": 3.851851851851852,
"grad_norm": 0.1551317940376091,
"learning_rate": 1e-05,
"loss": 0.0064,
"step": 312
},
{
"epoch": 3.8641975308641974,
"grad_norm": 0.12067080386439995,
"learning_rate": 9.938400533618615e-06,
"loss": 0.0086,
"step": 313
},
{
"epoch": 3.876543209876543,
"grad_norm": 0.14504074540723555,
"learning_rate": 9.876803404647617e-06,
"loss": 0.0047,
"step": 314
},
{
"epoch": 3.888888888888889,
"grad_norm": 0.2391873275575988,
"learning_rate": 9.815210950408703e-06,
"loss": 0.0075,
"step": 315
},
{
"epoch": 3.9012345679012346,
"grad_norm": 0.20353348393758977,
"learning_rate": 9.753625508046183e-06,
"loss": 0.0227,
"step": 316
},
{
"epoch": 3.9135802469135803,
"grad_norm": 0.20785860276808238,
"learning_rate": 9.692049414438298e-06,
"loss": 0.0191,
"step": 317
},
{
"epoch": 3.925925925925926,
"grad_norm": 0.1771402757741428,
"learning_rate": 9.630485006108554e-06,
"loss": 0.0074,
"step": 318
},
{
"epoch": 3.9382716049382713,
"grad_norm": 0.23082418089890744,
"learning_rate": 9.568934619137048e-06,
"loss": 0.0181,
"step": 319
},
{
"epoch": 3.950617283950617,
"grad_norm": 0.1985690884699008,
"learning_rate": 9.507400589071833e-06,
"loss": 0.0116,
"step": 320
},
{
"epoch": 3.962962962962963,
"grad_norm": 0.18052282236312095,
"learning_rate": 9.445885250840301e-06,
"loss": 0.008,
"step": 321
},
{
"epoch": 3.9753086419753085,
"grad_norm": 0.18304117085309365,
"learning_rate": 9.384390938660572e-06,
"loss": 0.0075,
"step": 322
},
{
"epoch": 3.9876543209876543,
"grad_norm": 0.1352378832885016,
"learning_rate": 9.322919985952926e-06,
"loss": 0.0043,
"step": 323
},
{
"epoch": 4.0,
"grad_norm": 0.1888441520514094,
"learning_rate": 9.261474725251261e-06,
"loss": 0.0066,
"step": 324
},
{
"epoch": 4.012345679012346,
"grad_norm": 0.17576556583410355,
"learning_rate": 9.200057488114585e-06,
"loss": 0.0111,
"step": 325
},
{
"epoch": 4.0246913580246915,
"grad_norm": 0.1786262879981403,
"learning_rate": 9.138670605038542e-06,
"loss": 0.0051,
"step": 326
},
{
"epoch": 4.037037037037037,
"grad_norm": 0.18990489010264094,
"learning_rate": 9.07731640536698e-06,
"loss": 0.0071,
"step": 327
},
{
"epoch": 4.049382716049383,
"grad_norm": 0.12318159558833343,
"learning_rate": 9.015997217203574e-06,
"loss": 0.0034,
"step": 328
},
{
"epoch": 4.061728395061729,
"grad_norm": 0.13751334569727902,
"learning_rate": 8.954715367323468e-06,
"loss": 0.0063,
"step": 329
},
{
"epoch": 4.074074074074074,
"grad_norm": 0.13146054154726655,
"learning_rate": 8.893473181084993e-06,
"loss": 0.0047,
"step": 330
},
{
"epoch": 4.08641975308642,
"grad_norm": 0.15881793905029615,
"learning_rate": 8.83227298234144e-06,
"loss": 0.0086,
"step": 331
},
{
"epoch": 4.098765432098766,
"grad_norm": 0.12142724241545133,
"learning_rate": 8.771117093352861e-06,
"loss": 0.0051,
"step": 332
},
{
"epoch": 4.111111111111111,
"grad_norm": 0.2286294288273942,
"learning_rate": 8.71000783469797e-06,
"loss": 0.0121,
"step": 333
},
{
"epoch": 4.1234567901234565,
"grad_norm": 0.16035361370524753,
"learning_rate": 8.648947525186073e-06,
"loss": 0.005,
"step": 334
},
{
"epoch": 4.135802469135802,
"grad_norm": 0.1828509697608793,
"learning_rate": 8.58793848176909e-06,
"loss": 0.0058,
"step": 335
},
{
"epoch": 4.148148148148148,
"grad_norm": 0.18114371230252813,
"learning_rate": 8.526983019453624e-06,
"loss": 0.0041,
"step": 336
},
{
"epoch": 4.160493827160494,
"grad_norm": 0.13123399087407325,
"learning_rate": 8.466083451213145e-06,
"loss": 0.0055,
"step": 337
},
{
"epoch": 4.172839506172839,
"grad_norm": 0.15545804752592987,
"learning_rate": 8.405242087900192e-06,
"loss": 0.0031,
"step": 338
},
{
"epoch": 4.185185185185185,
"grad_norm": 0.14836867939239134,
"learning_rate": 8.3444612381587e-06,
"loss": 0.0105,
"step": 339
},
{
"epoch": 4.197530864197531,
"grad_norm": 0.205738715941424,
"learning_rate": 8.283743208336403e-06,
"loss": 0.0055,
"step": 340
},
{
"epoch": 4.209876543209877,
"grad_norm": 0.21183835509861776,
"learning_rate": 8.223090302397313e-06,
"loss": 0.006,
"step": 341
},
{
"epoch": 4.222222222222222,
"grad_norm": 0.15257239381928547,
"learning_rate": 8.162504821834296e-06,
"loss": 0.0047,
"step": 342
},
{
"epoch": 4.234567901234568,
"grad_norm": 0.11658733556890352,
"learning_rate": 8.101989065581742e-06,
"loss": 0.0068,
"step": 343
},
{
"epoch": 4.246913580246914,
"grad_norm": 0.13573163174587216,
"learning_rate": 8.041545329928332e-06,
"loss": 0.0123,
"step": 344
},
{
"epoch": 4.2592592592592595,
"grad_norm": 0.13777699068418175,
"learning_rate": 7.9811759084299e-06,
"loss": 0.0097,
"step": 345
},
{
"epoch": 4.271604938271605,
"grad_norm": 0.15763381172079863,
"learning_rate": 7.92088309182241e-06,
"loss": 0.0087,
"step": 346
},
{
"epoch": 4.283950617283951,
"grad_norm": 0.15479605806346516,
"learning_rate": 7.860669167935028e-06,
"loss": 0.0032,
"step": 347
},
{
"epoch": 4.296296296296296,
"grad_norm": 0.1586828874171454,
"learning_rate": 7.800536421603317e-06,
"loss": 0.008,
"step": 348
},
{
"epoch": 4.308641975308642,
"grad_norm": 0.15327141154299848,
"learning_rate": 7.740487134582527e-06,
"loss": 0.0059,
"step": 349
},
{
"epoch": 4.320987654320987,
"grad_norm": 0.140301040005369,
"learning_rate": 7.680523585461021e-06,
"loss": 0.0062,
"step": 350
},
{
"epoch": 4.333333333333333,
"grad_norm": 0.12438760539558087,
"learning_rate": 7.620648049573815e-06,
"loss": 0.0033,
"step": 351
},
{
"epoch": 4.345679012345679,
"grad_norm": 0.15251811244541832,
"learning_rate": 7.560862798916229e-06,
"loss": 0.0066,
"step": 352
},
{
"epoch": 4.3580246913580245,
"grad_norm": 0.12017480297474846,
"learning_rate": 7.501170102057691e-06,
"loss": 0.0033,
"step": 353
},
{
"epoch": 4.37037037037037,
"grad_norm": 0.07857457205022518,
"learning_rate": 7.441572224055644e-06,
"loss": 0.0021,
"step": 354
},
{
"epoch": 4.382716049382716,
"grad_norm": 0.1278711203666051,
"learning_rate": 7.382071426369597e-06,
"loss": 0.0064,
"step": 355
},
{
"epoch": 4.395061728395062,
"grad_norm": 0.18131359376325312,
"learning_rate": 7.322669966775321e-06,
"loss": 0.0095,
"step": 356
},
{
"epoch": 4.407407407407407,
"grad_norm": 0.09723568551494033,
"learning_rate": 7.263370099279173e-06,
"loss": 0.0027,
"step": 357
},
{
"epoch": 4.419753086419753,
"grad_norm": 0.1672781410819625,
"learning_rate": 7.204174074032562e-06,
"loss": 0.0066,
"step": 358
},
{
"epoch": 4.432098765432099,
"grad_norm": 0.1258158648501038,
"learning_rate": 7.1450841372465806e-06,
"loss": 0.0056,
"step": 359
},
{
"epoch": 4.444444444444445,
"grad_norm": 0.11731253934039346,
"learning_rate": 7.086102531106755e-06,
"loss": 0.0031,
"step": 360
},
{
"epoch": 4.45679012345679,
"grad_norm": 0.21138242950668157,
"learning_rate": 7.027231493687974e-06,
"loss": 0.0058,
"step": 361
},
{
"epoch": 4.469135802469136,
"grad_norm": 0.12057312296375491,
"learning_rate": 6.968473258869566e-06,
"loss": 0.0055,
"step": 362
},
{
"epoch": 4.481481481481482,
"grad_norm": 0.18470894070091085,
"learning_rate": 6.909830056250527e-06,
"loss": 0.0061,
"step": 363
},
{
"epoch": 4.493827160493828,
"grad_norm": 0.10365823512721326,
"learning_rate": 6.851304111064923e-06,
"loss": 0.004,
"step": 364
},
{
"epoch": 4.506172839506172,
"grad_norm": 0.14863072835909422,
"learning_rate": 6.7928976440974504e-06,
"loss": 0.0064,
"step": 365
},
{
"epoch": 4.518518518518518,
"grad_norm": 0.11216394972543642,
"learning_rate": 6.734612871599169e-06,
"loss": 0.0047,
"step": 366
},
{
"epoch": 4.530864197530864,
"grad_norm": 0.1769379177433203,
"learning_rate": 6.6764520052034054e-06,
"loss": 0.0071,
"step": 367
},
{
"epoch": 4.54320987654321,
"grad_norm": 0.11523498585679909,
"learning_rate": 6.618417251841829e-06,
"loss": 0.0052,
"step": 368
},
{
"epoch": 4.555555555555555,
"grad_norm": 0.213990728294252,
"learning_rate": 6.560510813660719e-06,
"loss": 0.01,
"step": 369
},
{
"epoch": 4.567901234567901,
"grad_norm": 0.22242402616595255,
"learning_rate": 6.502734887937389e-06,
"loss": 0.0091,
"step": 370
},
{
"epoch": 4.580246913580247,
"grad_norm": 0.22736754834624315,
"learning_rate": 6.44509166699682e-06,
"loss": 0.0098,
"step": 371
},
{
"epoch": 4.592592592592593,
"grad_norm": 0.12867942855007472,
"learning_rate": 6.387583338128471e-06,
"loss": 0.0035,
"step": 372
},
{
"epoch": 4.604938271604938,
"grad_norm": 0.12725911519226457,
"learning_rate": 6.33021208350328e-06,
"loss": 0.0064,
"step": 373
},
{
"epoch": 4.617283950617284,
"grad_norm": 0.17702466948892448,
"learning_rate": 6.27298008009086e-06,
"loss": 0.0054,
"step": 374
},
{
"epoch": 4.62962962962963,
"grad_norm": 0.09649965912500313,
"learning_rate": 6.215889499576898e-06,
"loss": 0.0074,
"step": 375
},
{
"epoch": 4.6419753086419755,
"grad_norm": 0.1616613279100784,
"learning_rate": 6.158942508280743e-06,
"loss": 0.0069,
"step": 376
},
{
"epoch": 4.654320987654321,
"grad_norm": 0.18968833143117372,
"learning_rate": 6.102141267073207e-06,
"loss": 0.0077,
"step": 377
},
{
"epoch": 4.666666666666667,
"grad_norm": 0.16248020648691364,
"learning_rate": 6.0454879312945755e-06,
"loss": 0.0078,
"step": 378
},
{
"epoch": 4.679012345679013,
"grad_norm": 0.12051760198559061,
"learning_rate": 5.988984650672813e-06,
"loss": 0.0034,
"step": 379
},
{
"epoch": 4.6913580246913575,
"grad_norm": 0.13200347199182663,
"learning_rate": 5.932633569242e-06,
"loss": 0.0098,
"step": 380
},
{
"epoch": 4.703703703703704,
"grad_norm": 0.16681661954858018,
"learning_rate": 5.876436825260967e-06,
"loss": 0.0041,
"step": 381
},
{
"epoch": 4.716049382716049,
"grad_norm": 0.14006551980072618,
"learning_rate": 5.82039655113217e-06,
"loss": 0.0058,
"step": 382
},
{
"epoch": 4.728395061728395,
"grad_norm": 0.15271237139456825,
"learning_rate": 5.764514873320761e-06,
"loss": 0.0069,
"step": 383
},
{
"epoch": 4.7407407407407405,
"grad_norm": 0.09090612510274507,
"learning_rate": 5.708793912273911e-06,
"loss": 0.003,
"step": 384
},
{
"epoch": 4.753086419753086,
"grad_norm": 0.2582667643739463,
"learning_rate": 5.653235782340351e-06,
"loss": 0.0048,
"step": 385
},
{
"epoch": 4.765432098765432,
"grad_norm": 0.11419791645573754,
"learning_rate": 5.597842591690128e-06,
"loss": 0.0029,
"step": 386
},
{
"epoch": 4.777777777777778,
"grad_norm": 0.12658265868721003,
"learning_rate": 5.542616442234618e-06,
"loss": 0.0052,
"step": 387
},
{
"epoch": 4.790123456790123,
"grad_norm": 0.11780861249847288,
"learning_rate": 5.487559429546772e-06,
"loss": 0.003,
"step": 388
},
{
"epoch": 4.802469135802469,
"grad_norm": 0.10635775400249556,
"learning_rate": 5.432673642781595e-06,
"loss": 0.0056,
"step": 389
},
{
"epoch": 4.814814814814815,
"grad_norm": 0.1886486359438476,
"learning_rate": 5.3779611645968696e-06,
"loss": 0.0035,
"step": 390
},
{
"epoch": 4.827160493827161,
"grad_norm": 0.11101625407249367,
"learning_rate": 5.3234240710741335e-06,
"loss": 0.003,
"step": 391
},
{
"epoch": 4.839506172839506,
"grad_norm": 0.20857170942236092,
"learning_rate": 5.269064431639901e-06,
"loss": 0.0058,
"step": 392
},
{
"epoch": 4.851851851851852,
"grad_norm": 0.14690806172755955,
"learning_rate": 5.214884308987136e-06,
"loss": 0.0046,
"step": 393
},
{
"epoch": 4.864197530864198,
"grad_norm": 0.11451036979746576,
"learning_rate": 5.160885758996985e-06,
"loss": 0.0033,
"step": 394
},
{
"epoch": 4.8765432098765435,
"grad_norm": 0.16454312350788597,
"learning_rate": 5.107070830660765e-06,
"loss": 0.0033,
"step": 395
},
{
"epoch": 4.888888888888889,
"grad_norm": 0.0788817718822425,
"learning_rate": 5.053441566002214e-06,
"loss": 0.0027,
"step": 396
},
{
"epoch": 4.901234567901234,
"grad_norm": 0.14669951303785775,
"learning_rate": 5.000000000000003e-06,
"loss": 0.0081,
"step": 397
},
{
"epoch": 4.91358024691358,
"grad_norm": 0.10678859264163215,
"learning_rate": 4.946748160510522e-06,
"loss": 0.0026,
"step": 398
},
{
"epoch": 4.925925925925926,
"grad_norm": 0.13706317316845099,
"learning_rate": 4.893688068190933e-06,
"loss": 0.011,
"step": 399
},
{
"epoch": 4.938271604938271,
"grad_norm": 0.17403711057746862,
"learning_rate": 4.8408217364224886e-06,
"loss": 0.0135,
"step": 400
},
{
"epoch": 4.950617283950617,
"grad_norm": 0.11398641090194171,
"learning_rate": 4.788151171234149e-06,
"loss": 0.0027,
"step": 401
},
{
"epoch": 4.962962962962963,
"grad_norm": 0.13576944911004699,
"learning_rate": 4.7356783712264405e-06,
"loss": 0.0076,
"step": 402
},
{
"epoch": 4.9753086419753085,
"grad_norm": 0.12391596280841162,
"learning_rate": 4.683405327495638e-06,
"loss": 0.0039,
"step": 403
},
{
"epoch": 4.987654320987654,
"grad_norm": 0.14003588815325335,
"learning_rate": 4.631334023558199e-06,
"loss": 0.0036,
"step": 404
},
{
"epoch": 5.0,
"grad_norm": 0.13760549101766917,
"learning_rate": 4.579466435275506e-06,
"loss": 0.0044,
"step": 405
},
{
"epoch": 5.012345679012346,
"grad_norm": 0.10108142455180544,
"learning_rate": 4.5278045307788885e-06,
"loss": 0.0029,
"step": 406
},
{
"epoch": 5.0246913580246915,
"grad_norm": 0.06135525071329738,
"learning_rate": 4.476350270394942e-06,
"loss": 0.0017,
"step": 407
},
{
"epoch": 5.037037037037037,
"grad_norm": 0.1357779809507268,
"learning_rate": 4.425105606571145e-06,
"loss": 0.0035,
"step": 408
},
{
"epoch": 5.049382716049383,
"grad_norm": 0.09775066607270783,
"learning_rate": 4.374072483801769e-06,
"loss": 0.0026,
"step": 409
},
{
"epoch": 5.061728395061729,
"grad_norm": 0.0990513642880167,
"learning_rate": 4.323252838554099e-06,
"loss": 0.0024,
"step": 410
},
{
"epoch": 5.074074074074074,
"grad_norm": 0.07747603969017622,
"learning_rate": 4.272648599194948e-06,
"loss": 0.0016,
"step": 411
},
{
"epoch": 5.08641975308642,
"grad_norm": 0.14776525482891759,
"learning_rate": 4.222261685917489e-06,
"loss": 0.0061,
"step": 412
},
{
"epoch": 5.098765432098766,
"grad_norm": 0.06839289353699521,
"learning_rate": 4.1720940106683915e-06,
"loss": 0.0017,
"step": 413
},
{
"epoch": 5.111111111111111,
"grad_norm": 0.08384861829329324,
"learning_rate": 4.12214747707527e-06,
"loss": 0.0029,
"step": 414
},
{
"epoch": 5.1234567901234565,
"grad_norm": 0.08959984954999654,
"learning_rate": 4.0724239803744524e-06,
"loss": 0.0058,
"step": 415
},
{
"epoch": 5.135802469135802,
"grad_norm": 0.12415321940213553,
"learning_rate": 4.022925407339064e-06,
"loss": 0.0046,
"step": 416
},
{
"epoch": 5.148148148148148,
"grad_norm": 0.08785424848498224,
"learning_rate": 3.973653636207437e-06,
"loss": 0.0028,
"step": 417
},
{
"epoch": 5.160493827160494,
"grad_norm": 0.08347388278874737,
"learning_rate": 3.924610536611833e-06,
"loss": 0.0035,
"step": 418
},
{
"epoch": 5.172839506172839,
"grad_norm": 0.14158253470926024,
"learning_rate": 3.875797969507502e-06,
"loss": 0.0029,
"step": 419
},
{
"epoch": 5.185185185185185,
"grad_norm": 0.07618847346132784,
"learning_rate": 3.827217787102072e-06,
"loss": 0.0021,
"step": 420
},
{
"epoch": 5.197530864197531,
"grad_norm": 0.0799051798869917,
"learning_rate": 3.7788718327852625e-06,
"loss": 0.0021,
"step": 421
},
{
"epoch": 5.209876543209877,
"grad_norm": 0.12744475006620873,
"learning_rate": 3.730761941058938e-06,
"loss": 0.0048,
"step": 422
},
{
"epoch": 5.222222222222222,
"grad_norm": 0.1367296903886182,
"learning_rate": 3.6828899374674933e-06,
"loss": 0.003,
"step": 423
},
{
"epoch": 5.234567901234568,
"grad_norm": 0.12613896676405553,
"learning_rate": 3.635257638528589e-06,
"loss": 0.0028,
"step": 424
},
{
"epoch": 5.246913580246914,
"grad_norm": 0.08185887424793617,
"learning_rate": 3.587866851664219e-06,
"loss": 0.0021,
"step": 425
},
{
"epoch": 5.2592592592592595,
"grad_norm": 0.05307305165318999,
"learning_rate": 3.540719375132129e-06,
"loss": 0.0016,
"step": 426
},
{
"epoch": 5.271604938271605,
"grad_norm": 0.13804574516617765,
"learning_rate": 3.493816997957582e-06,
"loss": 0.0047,
"step": 427
},
{
"epoch": 5.283950617283951,
"grad_norm": 0.07491487304908827,
"learning_rate": 3.447161499865467e-06,
"loss": 0.0018,
"step": 428
},
{
"epoch": 5.296296296296296,
"grad_norm": 0.08625813878349398,
"learning_rate": 3.4007546512127764e-06,
"loss": 0.0019,
"step": 429
},
{
"epoch": 5.308641975308642,
"grad_norm": 0.06179766978879838,
"learning_rate": 3.3545982129214227e-06,
"loss": 0.0015,
"step": 430
},
{
"epoch": 5.320987654320987,
"grad_norm": 0.15978150579410316,
"learning_rate": 3.308693936411421e-06,
"loss": 0.0043,
"step": 431
},
{
"epoch": 5.333333333333333,
"grad_norm": 0.1469342970511695,
"learning_rate": 3.2630435635344283e-06,
"loss": 0.0035,
"step": 432
},
{
"epoch": 5.345679012345679,
"grad_norm": 0.14475265179014074,
"learning_rate": 3.21764882650766e-06,
"loss": 0.004,
"step": 433
},
{
"epoch": 5.3580246913580245,
"grad_norm": 0.1308223994783395,
"learning_rate": 3.1725114478481458e-06,
"loss": 0.0026,
"step": 434
},
{
"epoch": 5.37037037037037,
"grad_norm": 0.13456024267263844,
"learning_rate": 3.1276331403073733e-06,
"loss": 0.0113,
"step": 435
},
{
"epoch": 5.382716049382716,
"grad_norm": 0.08889181174562952,
"learning_rate": 3.0830156068063e-06,
"loss": 0.0016,
"step": 436
},
{
"epoch": 5.395061728395062,
"grad_norm": 0.10350317432315419,
"learning_rate": 3.0386605403707347e-06,
"loss": 0.0029,
"step": 437
},
{
"epoch": 5.407407407407407,
"grad_norm": 0.11255331179888899,
"learning_rate": 2.9945696240670905e-06,
"loss": 0.007,
"step": 438
},
{
"epoch": 5.419753086419753,
"grad_norm": 0.12613776990357034,
"learning_rate": 2.95074453093853e-06,
"loss": 0.0037,
"step": 439
},
{
"epoch": 5.432098765432099,
"grad_norm": 0.13211120615930835,
"learning_rate": 2.907186923941466e-06,
"loss": 0.0029,
"step": 440
},
{
"epoch": 5.444444444444445,
"grad_norm": 0.13009133491131789,
"learning_rate": 2.8638984558824777e-06,
"loss": 0.0056,
"step": 441
},
{
"epoch": 5.45679012345679,
"grad_norm": 0.11165738765343762,
"learning_rate": 2.820880769355582e-06,
"loss": 0.0023,
"step": 442
},
{
"epoch": 5.469135802469136,
"grad_norm": 0.09955128366405679,
"learning_rate": 2.778135496679908e-06,
"loss": 0.0043,
"step": 443
},
{
"epoch": 5.481481481481482,
"grad_norm": 0.13140680037758684,
"learning_rate": 2.7356642598377604e-06,
"loss": 0.005,
"step": 444
},
{
"epoch": 5.493827160493828,
"grad_norm": 0.11819363282576739,
"learning_rate": 2.6934686704130698e-06,
"loss": 0.0021,
"step": 445
},
{
"epoch": 5.506172839506172,
"grad_norm": 0.08748806809387456,
"learning_rate": 2.6515503295302446e-06,
"loss": 0.0014,
"step": 446
},
{
"epoch": 5.518518518518518,
"grad_norm": 0.10665634756766484,
"learning_rate": 2.6099108277934105e-06,
"loss": 0.0038,
"step": 447
},
{
"epoch": 5.530864197530864,
"grad_norm": 0.06465287847372936,
"learning_rate": 2.5685517452260566e-06,
"loss": 0.0043,
"step": 448
},
{
"epoch": 5.54320987654321,
"grad_norm": 0.06355150302844088,
"learning_rate": 2.527474651211089e-06,
"loss": 0.0016,
"step": 449
},
{
"epoch": 5.555555555555555,
"grad_norm": 0.14706952848764682,
"learning_rate": 2.4866811044312667e-06,
"loss": 0.0095,
"step": 450
},
{
"epoch": 5.567901234567901,
"grad_norm": 0.11732857265877357,
"learning_rate": 2.4461726528100615e-06,
"loss": 0.0062,
"step": 451
},
{
"epoch": 5.580246913580247,
"grad_norm": 0.09196481051355447,
"learning_rate": 2.405950833452928e-06,
"loss": 0.0043,
"step": 452
},
{
"epoch": 5.592592592592593,
"grad_norm": 0.16089432829098174,
"learning_rate": 2.3660171725889703e-06,
"loss": 0.005,
"step": 453
},
{
"epoch": 5.604938271604938,
"grad_norm": 0.09752738474918766,
"learning_rate": 2.32637318551303e-06,
"loss": 0.003,
"step": 454
},
{
"epoch": 5.617283950617284,
"grad_norm": 0.111669117303663,
"learning_rate": 2.287020376528193e-06,
"loss": 0.0023,
"step": 455
},
{
"epoch": 5.62962962962963,
"grad_norm": 0.06950187494180007,
"learning_rate": 2.2479602388887013e-06,
"loss": 0.0024,
"step": 456
},
{
"epoch": 5.6419753086419755,
"grad_norm": 0.09005092942064802,
"learning_rate": 2.209194254743295e-06,
"loss": 0.0041,
"step": 457
},
{
"epoch": 5.654320987654321,
"grad_norm": 0.058271649974103654,
"learning_rate": 2.170723895078972e-06,
"loss": 0.0013,
"step": 458
},
{
"epoch": 5.666666666666667,
"grad_norm": 0.05801255129790246,
"learning_rate": 2.132550619665168e-06,
"loss": 0.0015,
"step": 459
},
{
"epoch": 5.679012345679013,
"grad_norm": 0.05173255822131142,
"learning_rate": 2.0946758769983666e-06,
"loss": 0.0015,
"step": 460
},
{
"epoch": 5.6913580246913575,
"grad_norm": 0.12763685917437947,
"learning_rate": 2.05710110424714e-06,
"loss": 0.0022,
"step": 461
},
{
"epoch": 5.703703703703704,
"grad_norm": 0.13571202690989434,
"learning_rate": 2.019827727197605e-06,
"loss": 0.0031,
"step": 462
},
{
"epoch": 5.716049382716049,
"grad_norm": 0.08858787712719375,
"learning_rate": 1.982857160199334e-06,
"loss": 0.0025,
"step": 463
},
{
"epoch": 5.728395061728395,
"grad_norm": 0.09041339350659952,
"learning_rate": 1.946190806111674e-06,
"loss": 0.0022,
"step": 464
},
{
"epoch": 5.7407407407407405,
"grad_norm": 0.12427642574112911,
"learning_rate": 1.9098300562505266e-06,
"loss": 0.0026,
"step": 465
},
{
"epoch": 5.753086419753086,
"grad_norm": 0.07753348116821808,
"learning_rate": 1.873776290335545e-06,
"loss": 0.0024,
"step": 466
},
{
"epoch": 5.765432098765432,
"grad_norm": 0.10239535590693546,
"learning_rate": 1.8380308764377841e-06,
"loss": 0.0023,
"step": 467
},
{
"epoch": 5.777777777777778,
"grad_norm": 0.0932480254649128,
"learning_rate": 1.80259517092779e-06,
"loss": 0.0026,
"step": 468
},
{
"epoch": 5.790123456790123,
"grad_norm": 0.07678126376324187,
"learning_rate": 1.767470518424129e-06,
"loss": 0.0023,
"step": 469
},
{
"epoch": 5.802469135802469,
"grad_norm": 0.07579256484399811,
"learning_rate": 1.7326582517423662e-06,
"loss": 0.0014,
"step": 470
},
{
"epoch": 5.814814814814815,
"grad_norm": 0.08198420550789928,
"learning_rate": 1.6981596918444953e-06,
"loss": 0.0019,
"step": 471
},
{
"epoch": 5.827160493827161,
"grad_norm": 0.06306638866343868,
"learning_rate": 1.663976147788806e-06,
"loss": 0.0013,
"step": 472
},
{
"epoch": 5.839506172839506,
"grad_norm": 0.07973362090991411,
"learning_rate": 1.6301089166802232e-06,
"loss": 0.0017,
"step": 473
},
{
"epoch": 5.851851851851852,
"grad_norm": 0.10612340130126985,
"learning_rate": 1.596559283621074e-06,
"loss": 0.0023,
"step": 474
},
{
"epoch": 5.864197530864198,
"grad_norm": 0.060264533571981124,
"learning_rate": 1.5633285216623384e-06,
"loss": 0.0016,
"step": 475
},
{
"epoch": 5.8765432098765435,
"grad_norm": 0.06817593452948047,
"learning_rate": 1.5304178917553302e-06,
"loss": 0.0012,
"step": 476
},
{
"epoch": 5.888888888888889,
"grad_norm": 0.09699227562887566,
"learning_rate": 1.4978286427038602e-06,
"loss": 0.003,
"step": 477
},
{
"epoch": 5.901234567901234,
"grad_norm": 0.08289137794935784,
"learning_rate": 1.4655620111168422e-06,
"loss": 0.002,
"step": 478
},
{
"epoch": 5.91358024691358,
"grad_norm": 0.09286549417981105,
"learning_rate": 1.4336192213613742e-06,
"loss": 0.0059,
"step": 479
},
{
"epoch": 5.925925925925926,
"grad_norm": 0.11098773394965465,
"learning_rate": 1.4020014855162755e-06,
"loss": 0.0053,
"step": 480
},
{
"epoch": 5.938271604938271,
"grad_norm": 0.09542493224972246,
"learning_rate": 1.3707100033261035e-06,
"loss": 0.0031,
"step": 481
},
{
"epoch": 5.950617283950617,
"grad_norm": 0.07639864399541678,
"learning_rate": 1.339745962155613e-06,
"loss": 0.0022,
"step": 482
},
{
"epoch": 5.962962962962963,
"grad_norm": 0.08849988543559398,
"learning_rate": 1.3091105369447166e-06,
"loss": 0.0017,
"step": 483
},
{
"epoch": 5.9753086419753085,
"grad_norm": 0.0884132771726154,
"learning_rate": 1.2788048901638917e-06,
"loss": 0.0017,
"step": 484
},
{
"epoch": 5.987654320987654,
"grad_norm": 0.06292245984188637,
"learning_rate": 1.2488301717700735e-06,
"loss": 0.0016,
"step": 485
},
{
"epoch": 6.0,
"grad_norm": 0.062090787373349644,
"learning_rate": 1.2191875191630209e-06,
"loss": 0.0017,
"step": 486
},
{
"epoch": 6.012345679012346,
"grad_norm": 0.04437033735955514,
"learning_rate": 1.1898780571421554e-06,
"loss": 0.0013,
"step": 487
},
{
"epoch": 6.0246913580246915,
"grad_norm": 0.055690581108790035,
"learning_rate": 1.1609028978638804e-06,
"loss": 0.0015,
"step": 488
},
{
"epoch": 6.037037037037037,
"grad_norm": 0.06322304615189499,
"learning_rate": 1.132263140799381e-06,
"loss": 0.0016,
"step": 489
},
{
"epoch": 6.049382716049383,
"grad_norm": 0.06141880723579853,
"learning_rate": 1.1039598726929046e-06,
"loss": 0.0019,
"step": 490
},
{
"epoch": 6.061728395061729,
"grad_norm": 0.0559934719425724,
"learning_rate": 1.0759941675205221e-06,
"loss": 0.0019,
"step": 491
},
{
"epoch": 6.074074074074074,
"grad_norm": 0.052119641486773,
"learning_rate": 1.0483670864493777e-06,
"loss": 0.0041,
"step": 492
},
{
"epoch": 6.08641975308642,
"grad_norm": 0.052574867923764465,
"learning_rate": 1.0210796777974196e-06,
"loss": 0.0016,
"step": 493
},
{
"epoch": 6.098765432098766,
"grad_norm": 0.04551032766348937,
"learning_rate": 9.94132976993627e-07,
"loss": 0.0015,
"step": 494
},
{
"epoch": 6.111111111111111,
"grad_norm": 0.07300563736192743,
"learning_rate": 9.675280065387117e-07,
"loss": 0.0026,
"step": 495
},
{
"epoch": 6.1234567901234565,
"grad_norm": 0.0589388985855795,
"learning_rate": 9.412657759663279e-07,
"loss": 0.002,
"step": 496
},
{
"epoch": 6.135802469135802,
"grad_norm": 0.036832313525668736,
"learning_rate": 9.153472818047627e-07,
"loss": 0.001,
"step": 497
},
{
"epoch": 6.148148148148148,
"grad_norm": 0.08579156187113851,
"learning_rate": 8.897735075391156e-07,
"loss": 0.0019,
"step": 498
},
{
"epoch": 6.160493827160494,
"grad_norm": 0.03254562268330724,
"learning_rate": 8.645454235739903e-07,
"loss": 0.001,
"step": 499
},
{
"epoch": 6.172839506172839,
"grad_norm": 0.06668010013885942,
"learning_rate": 8.39663987196665e-07,
"loss": 0.0036,
"step": 500
},
{
"epoch": 6.185185185185185,
"grad_norm": 0.05254242210245364,
"learning_rate": 8.151301425407699e-07,
"loss": 0.0052,
"step": 501
},
{
"epoch": 6.197530864197531,
"grad_norm": 0.06269798774876521,
"learning_rate": 7.909448205504633e-07,
"loss": 0.0017,
"step": 502
},
{
"epoch": 6.209876543209877,
"grad_norm": 0.05212716006751509,
"learning_rate": 7.671089389451059e-07,
"loss": 0.0014,
"step": 503
},
{
"epoch": 6.222222222222222,
"grad_norm": 0.06723279411740171,
"learning_rate": 7.43623402184438e-07,
"loss": 0.0046,
"step": 504
},
{
"epoch": 6.234567901234568,
"grad_norm": 0.053338908065449955,
"learning_rate": 7.204891014342552e-07,
"loss": 0.0015,
"step": 505
},
{
"epoch": 6.246913580246914,
"grad_norm": 0.06851159604452965,
"learning_rate": 6.977069145325987e-07,
"loss": 0.0019,
"step": 506
},
{
"epoch": 6.2592592592592595,
"grad_norm": 0.03228477145799883,
"learning_rate": 6.752777059564431e-07,
"loss": 0.0012,
"step": 507
},
{
"epoch": 6.271604938271605,
"grad_norm": 0.037103749830438106,
"learning_rate": 6.532023267888954e-07,
"loss": 0.0009,
"step": 508
},
{
"epoch": 6.283950617283951,
"grad_norm": 0.06563663547083265,
"learning_rate": 6.314816146868951e-07,
"loss": 0.0017,
"step": 509
},
{
"epoch": 6.296296296296296,
"grad_norm": 0.07016302273033481,
"learning_rate": 6.101163938494359e-07,
"loss": 0.0035,
"step": 510
},
{
"epoch": 6.308641975308642,
"grad_norm": 0.06182335435231687,
"learning_rate": 5.891074749862857e-07,
"loss": 0.0012,
"step": 511
},
{
"epoch": 6.320987654320987,
"grad_norm": 0.05596013279096978,
"learning_rate": 5.684556552872256e-07,
"loss": 0.0033,
"step": 512
},
{
"epoch": 6.333333333333333,
"grad_norm": 0.05534288287471249,
"learning_rate": 5.481617183918053e-07,
"loss": 0.0014,
"step": 513
},
{
"epoch": 6.345679012345679,
"grad_norm": 0.04118593026135544,
"learning_rate": 5.282264343595977e-07,
"loss": 0.0013,
"step": 514
},
{
"epoch": 6.3580246913580245,
"grad_norm": 0.04719174098780563,
"learning_rate": 5.086505596409885e-07,
"loss": 0.0012,
"step": 515
},
{
"epoch": 6.37037037037037,
"grad_norm": 0.06236635038173847,
"learning_rate": 4.894348370484648e-07,
"loss": 0.0019,
"step": 516
},
{
"epoch": 6.382716049382716,
"grad_norm": 0.05738720109912046,
"learning_rate": 4.7057999572843516e-07,
"loss": 0.0015,
"step": 517
},
{
"epoch": 6.395061728395062,
"grad_norm": 0.06748476210640128,
"learning_rate": 4.520867511335569e-07,
"loss": 0.0053,
"step": 518
},
{
"epoch": 6.407407407407407,
"grad_norm": 0.05317337463002061,
"learning_rate": 4.3395580499559276e-07,
"loss": 0.0016,
"step": 519
},
{
"epoch": 6.419753086419753,
"grad_norm": 0.042798403768389934,
"learning_rate": 4.161878452987778e-07,
"loss": 0.0014,
"step": 520
},
{
"epoch": 6.432098765432099,
"grad_norm": 0.03849912145724003,
"learning_rate": 3.9878354625371927e-07,
"loss": 0.001,
"step": 521
},
{
"epoch": 6.444444444444445,
"grad_norm": 0.06507943189963111,
"learning_rate": 3.817435682718096e-07,
"loss": 0.0016,
"step": 522
},
{
"epoch": 6.45679012345679,
"grad_norm": 0.06834437894921451,
"learning_rate": 3.650685579401692e-07,
"loss": 0.0014,
"step": 523
},
{
"epoch": 6.469135802469136,
"grad_norm": 0.068135936362266,
"learning_rate": 3.4875914799710975e-07,
"loss": 0.0037,
"step": 524
},
{
"epoch": 6.481481481481482,
"grad_norm": 0.0769225470497784,
"learning_rate": 3.328159573081258e-07,
"loss": 0.0025,
"step": 525
},
{
"epoch": 6.493827160493828,
"grad_norm": 0.04510279405686483,
"learning_rate": 3.1723959084241195e-07,
"loss": 0.0013,
"step": 526
},
{
"epoch": 6.506172839506172,
"grad_norm": 0.06253123572490814,
"learning_rate": 3.020306396499062e-07,
"loss": 0.0018,
"step": 527
},
{
"epoch": 6.518518518518518,
"grad_norm": 0.05607017364278352,
"learning_rate": 2.871896808388608e-07,
"loss": 0.0015,
"step": 528
},
{
"epoch": 6.530864197530864,
"grad_norm": 0.06565892652904905,
"learning_rate": 2.727172775539522e-07,
"loss": 0.0015,
"step": 529
},
{
"epoch": 6.54320987654321,
"grad_norm": 0.06779883467508055,
"learning_rate": 2.5861397895489914e-07,
"loss": 0.0054,
"step": 530
},
{
"epoch": 6.555555555555555,
"grad_norm": 0.0625439341793089,
"learning_rate": 2.44880320195634e-07,
"loss": 0.0014,
"step": 531
},
{
"epoch": 6.567901234567901,
"grad_norm": 0.046645244461245204,
"learning_rate": 2.315168224039932e-07,
"loss": 0.0011,
"step": 532
},
{
"epoch": 6.580246913580247,
"grad_norm": 0.05642062644236279,
"learning_rate": 2.1852399266194312e-07,
"loss": 0.002,
"step": 533
},
{
"epoch": 6.592592592592593,
"grad_norm": 0.053426947443764354,
"learning_rate": 2.0590232398634114e-07,
"loss": 0.0035,
"step": 534
},
{
"epoch": 6.604938271604938,
"grad_norm": 0.02804527022906818,
"learning_rate": 1.9365229531022267e-07,
"loss": 0.0009,
"step": 535
},
{
"epoch": 6.617283950617284,
"grad_norm": 0.04988184506603678,
"learning_rate": 1.817743714646314e-07,
"loss": 0.0012,
"step": 536
},
{
"epoch": 6.62962962962963,
"grad_norm": 0.03572163754383024,
"learning_rate": 1.7026900316098217e-07,
"loss": 0.0012,
"step": 537
},
{
"epoch": 6.6419753086419755,
"grad_norm": 0.040421555953721704,
"learning_rate": 1.5913662697395583e-07,
"loss": 0.0011,
"step": 538
},
{
"epoch": 6.654320987654321,
"grad_norm": 0.0854680374793536,
"learning_rate": 1.4837766532493469e-07,
"loss": 0.0015,
"step": 539
},
{
"epoch": 6.666666666666667,
"grad_norm": 0.0622890200616722,
"learning_rate": 1.3799252646597428e-07,
"loss": 0.0013,
"step": 540
},
{
"epoch": 6.679012345679013,
"grad_norm": 0.04487325280230035,
"learning_rate": 1.2798160446431006e-07,
"loss": 0.001,
"step": 541
},
{
"epoch": 6.6913580246913575,
"grad_norm": 0.06827865424951797,
"learning_rate": 1.1834527918740624e-07,
"loss": 0.0024,
"step": 542
},
{
"epoch": 6.703703703703704,
"grad_norm": 0.05828814401863449,
"learning_rate": 1.0908391628854042e-07,
"loss": 0.002,
"step": 543
},
{
"epoch": 6.716049382716049,
"grad_norm": 0.04071748268828138,
"learning_rate": 1.0019786719293046e-07,
"loss": 0.0012,
"step": 544
},
{
"epoch": 6.728395061728395,
"grad_norm": 0.0457370541103018,
"learning_rate": 9.168746908439718e-08,
"loss": 0.0013,
"step": 545
},
{
"epoch": 6.7407407407407405,
"grad_norm": 0.03663041381389654,
"learning_rate": 8.355304489257254e-08,
"loss": 0.0011,
"step": 546
},
{
"epoch": 6.753086419753086,
"grad_norm": 0.07030810856213271,
"learning_rate": 7.579490328064265e-08,
"loss": 0.0013,
"step": 547
},
{
"epoch": 6.765432098765432,
"grad_norm": 0.07476184707674007,
"learning_rate": 6.841333863363831e-08,
"loss": 0.0013,
"step": 548
},
{
"epoch": 6.777777777777778,
"grad_norm": 0.039877687798977456,
"learning_rate": 6.140863104726391e-08,
"loss": 0.0008,
"step": 549
},
{
"epoch": 6.790123456790123,
"grad_norm": 0.0692536388183646,
"learning_rate": 5.4781046317267103e-08,
"loss": 0.0074,
"step": 550
},
{
"epoch": 6.802469135802469,
"grad_norm": 0.05771681097265902,
"learning_rate": 4.85308359293557e-08,
"loss": 0.0034,
"step": 551
},
{
"epoch": 6.814814814814815,
"grad_norm": 0.07405017087059966,
"learning_rate": 4.2658237049655325e-08,
"loss": 0.0011,
"step": 552
},
{
"epoch": 6.827160493827161,
"grad_norm": 0.03321213886254247,
"learning_rate": 3.716347251570551e-08,
"loss": 0.0011,
"step": 553
},
{
"epoch": 6.839506172839506,
"grad_norm": 0.061503207253362814,
"learning_rate": 3.204675082800979e-08,
"loss": 0.0014,
"step": 554
},
{
"epoch": 6.851851851851852,
"grad_norm": 0.04618914106100694,
"learning_rate": 2.7308266142119788e-08,
"loss": 0.0012,
"step": 555
},
{
"epoch": 6.864197530864198,
"grad_norm": 0.06693954396383099,
"learning_rate": 2.294819826127115e-08,
"loss": 0.002,
"step": 556
},
{
"epoch": 6.8765432098765435,
"grad_norm": 0.1037551779412373,
"learning_rate": 1.896671262955896e-08,
"loss": 0.002,
"step": 557
},
{
"epoch": 6.888888888888889,
"grad_norm": 0.049967199152869485,
"learning_rate": 1.5363960325660565e-08,
"loss": 0.0023,
"step": 558
},
{
"epoch": 6.901234567901234,
"grad_norm": 0.0840544069774677,
"learning_rate": 1.2140078057101269e-08,
"loss": 0.0016,
"step": 559
},
{
"epoch": 6.91358024691358,
"grad_norm": 0.06330250298740656,
"learning_rate": 9.29518815506847e-09,
"loss": 0.0021,
"step": 560
},
{
"epoch": 6.925925925925926,
"grad_norm": 0.05729095992108939,
"learning_rate": 6.82939856977094e-09,
"loss": 0.0021,
"step": 561
},
{
"epoch": 6.938271604938271,
"grad_norm": 0.053351401675753425,
"learning_rate": 4.74280286634099e-09,
"loss": 0.0014,
"step": 562
},
{
"epoch": 6.950617283950617,
"grad_norm": 0.08262341994499242,
"learning_rate": 3.0354802212839705e-09,
"loss": 0.0044,
"step": 563
},
{
"epoch": 6.962962962962963,
"grad_norm": 0.058488381808320514,
"learning_rate": 1.7074954194729044e-09,
"loss": 0.0012,
"step": 564
},
{
"epoch": 6.9753086419753085,
"grad_norm": 0.08520368149220187,
"learning_rate": 7.588988516937789e-10,
"loss": 0.0021,
"step": 565
},
{
"epoch": 6.987654320987654,
"grad_norm": 0.05967553760304721,
"learning_rate": 1.897265127315251e-10,
"loss": 0.0029,
"step": 566
},
{
"epoch": 7.0,
"grad_norm": 0.06556025137721225,
"learning_rate": 0.0,
"loss": 0.0019,
"step": 567
},
{
"epoch": 7.0,
"step": 567,
"total_flos": 1.448974662653444e+18,
"train_loss": 0.046748905014567,
"train_runtime": 15715.0912,
"train_samples_per_second": 3.463,
"train_steps_per_second": 0.036
}
],
"logging_steps": 1,
"max_steps": 567,
"num_input_tokens_seen": 0,
"num_train_epochs": 7,
"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": 1.448974662653444e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}