HT-duplicate-Qwen-40k / trainer_state.json
ShourenWSR's picture
Upload folder using huggingface_hub
5ffba58 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 2502,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0012,
"grad_norm": 5.4833939764880055,
"learning_rate": 0.0,
"loss": 0.7607,
"step": 1
},
{
"epoch": 0.0024,
"grad_norm": 5.645497868805024,
"learning_rate": 3.984063745019921e-08,
"loss": 0.7813,
"step": 2
},
{
"epoch": 0.0036,
"grad_norm": 6.379959917142346,
"learning_rate": 7.968127490039842e-08,
"loss": 0.831,
"step": 3
},
{
"epoch": 0.0048,
"grad_norm": 5.8802628972123365,
"learning_rate": 1.195219123505976e-07,
"loss": 0.8058,
"step": 4
},
{
"epoch": 0.006,
"grad_norm": 6.3940729810393915,
"learning_rate": 1.5936254980079683e-07,
"loss": 0.8678,
"step": 5
},
{
"epoch": 0.0072,
"grad_norm": 5.766068577494406,
"learning_rate": 1.9920318725099604e-07,
"loss": 0.7789,
"step": 6
},
{
"epoch": 0.0084,
"grad_norm": 6.045841004732535,
"learning_rate": 2.390438247011952e-07,
"loss": 0.7945,
"step": 7
},
{
"epoch": 0.0096,
"grad_norm": 5.904260263294003,
"learning_rate": 2.7888446215139444e-07,
"loss": 0.7843,
"step": 8
},
{
"epoch": 0.0108,
"grad_norm": 5.8665611721938005,
"learning_rate": 3.1872509960159367e-07,
"loss": 0.8259,
"step": 9
},
{
"epoch": 0.012,
"grad_norm": 5.927624645803215,
"learning_rate": 3.585657370517929e-07,
"loss": 0.802,
"step": 10
},
{
"epoch": 0.0132,
"grad_norm": 5.2336347881583345,
"learning_rate": 3.9840637450199207e-07,
"loss": 0.7326,
"step": 11
},
{
"epoch": 0.0144,
"grad_norm": 5.744302914470887,
"learning_rate": 4.382470119521913e-07,
"loss": 0.8034,
"step": 12
},
{
"epoch": 0.0156,
"grad_norm": 5.569314249586872,
"learning_rate": 4.780876494023904e-07,
"loss": 0.8058,
"step": 13
},
{
"epoch": 0.0168,
"grad_norm": 5.279680827326481,
"learning_rate": 5.179282868525898e-07,
"loss": 0.7743,
"step": 14
},
{
"epoch": 0.018,
"grad_norm": 5.386497992251826,
"learning_rate": 5.577689243027889e-07,
"loss": 0.7688,
"step": 15
},
{
"epoch": 0.0192,
"grad_norm": 4.254944734918593,
"learning_rate": 5.976095617529881e-07,
"loss": 0.7054,
"step": 16
},
{
"epoch": 0.0204,
"grad_norm": 4.523233221330768,
"learning_rate": 6.374501992031873e-07,
"loss": 0.7534,
"step": 17
},
{
"epoch": 0.0216,
"grad_norm": 4.1987378310333545,
"learning_rate": 6.772908366533865e-07,
"loss": 0.708,
"step": 18
},
{
"epoch": 0.0228,
"grad_norm": 4.574293688793641,
"learning_rate": 7.171314741035858e-07,
"loss": 0.7916,
"step": 19
},
{
"epoch": 0.024,
"grad_norm": 4.3308716519686525,
"learning_rate": 7.569721115537849e-07,
"loss": 0.7431,
"step": 20
},
{
"epoch": 0.0252,
"grad_norm": 3.7190318716723603,
"learning_rate": 7.968127490039841e-07,
"loss": 0.7258,
"step": 21
},
{
"epoch": 0.0264,
"grad_norm": 2.543036752039916,
"learning_rate": 8.366533864541833e-07,
"loss": 0.7224,
"step": 22
},
{
"epoch": 0.0276,
"grad_norm": 2.4950258096616382,
"learning_rate": 8.764940239043826e-07,
"loss": 0.6736,
"step": 23
},
{
"epoch": 0.0288,
"grad_norm": 2.5069650630469984,
"learning_rate": 9.163346613545817e-07,
"loss": 0.6854,
"step": 24
},
{
"epoch": 0.03,
"grad_norm": 2.4088611292545825,
"learning_rate": 9.561752988047808e-07,
"loss": 0.7204,
"step": 25
},
{
"epoch": 0.0312,
"grad_norm": 2.186249677635318,
"learning_rate": 9.9601593625498e-07,
"loss": 0.6998,
"step": 26
},
{
"epoch": 0.0324,
"grad_norm": 2.1736221592657285,
"learning_rate": 1.0358565737051795e-06,
"loss": 0.7414,
"step": 27
},
{
"epoch": 0.0336,
"grad_norm": 2.1914465779108747,
"learning_rate": 1.0756972111553785e-06,
"loss": 0.6722,
"step": 28
},
{
"epoch": 0.0348,
"grad_norm": 1.6670960514305742,
"learning_rate": 1.1155378486055778e-06,
"loss": 0.6422,
"step": 29
},
{
"epoch": 0.036,
"grad_norm": 2.077075223596808,
"learning_rate": 1.155378486055777e-06,
"loss": 0.6731,
"step": 30
},
{
"epoch": 0.0372,
"grad_norm": 2.4897778527200183,
"learning_rate": 1.1952191235059762e-06,
"loss": 0.6803,
"step": 31
},
{
"epoch": 0.0384,
"grad_norm": 2.4711412460186373,
"learning_rate": 1.2350597609561754e-06,
"loss": 0.6559,
"step": 32
},
{
"epoch": 0.0396,
"grad_norm": 2.700674504910902,
"learning_rate": 1.2749003984063747e-06,
"loss": 0.6834,
"step": 33
},
{
"epoch": 0.0408,
"grad_norm": 2.682606066253942,
"learning_rate": 1.3147410358565737e-06,
"loss": 0.6516,
"step": 34
},
{
"epoch": 0.042,
"grad_norm": 2.1676293521182077,
"learning_rate": 1.354581673306773e-06,
"loss": 0.579,
"step": 35
},
{
"epoch": 0.0432,
"grad_norm": 2.173873959031573,
"learning_rate": 1.3944223107569721e-06,
"loss": 0.6436,
"step": 36
},
{
"epoch": 0.0444,
"grad_norm": 2.151838164329406,
"learning_rate": 1.4342629482071716e-06,
"loss": 0.698,
"step": 37
},
{
"epoch": 0.0456,
"grad_norm": 1.9621644411146546,
"learning_rate": 1.4741035856573708e-06,
"loss": 0.6488,
"step": 38
},
{
"epoch": 0.0468,
"grad_norm": 1.5368598635090958,
"learning_rate": 1.5139442231075698e-06,
"loss": 0.6302,
"step": 39
},
{
"epoch": 0.048,
"grad_norm": 1.2190139458412592,
"learning_rate": 1.553784860557769e-06,
"loss": 0.619,
"step": 40
},
{
"epoch": 0.0492,
"grad_norm": 1.165215618150115,
"learning_rate": 1.5936254980079683e-06,
"loss": 0.6194,
"step": 41
},
{
"epoch": 0.0504,
"grad_norm": 1.1142120578252954,
"learning_rate": 1.6334661354581673e-06,
"loss": 0.6337,
"step": 42
},
{
"epoch": 0.0516,
"grad_norm": 1.2057442448808577,
"learning_rate": 1.6733067729083665e-06,
"loss": 0.6082,
"step": 43
},
{
"epoch": 0.0528,
"grad_norm": 1.1247636014697693,
"learning_rate": 1.7131474103585658e-06,
"loss": 0.5601,
"step": 44
},
{
"epoch": 0.054,
"grad_norm": 1.41137183436536,
"learning_rate": 1.7529880478087652e-06,
"loss": 0.6557,
"step": 45
},
{
"epoch": 0.0552,
"grad_norm": 1.052690589137755,
"learning_rate": 1.7928286852589644e-06,
"loss": 0.5787,
"step": 46
},
{
"epoch": 0.0564,
"grad_norm": 1.0601978966012295,
"learning_rate": 1.8326693227091634e-06,
"loss": 0.5694,
"step": 47
},
{
"epoch": 0.0576,
"grad_norm": 1.1070630189694626,
"learning_rate": 1.8725099601593627e-06,
"loss": 0.5738,
"step": 48
},
{
"epoch": 0.0588,
"grad_norm": 0.9889388769001813,
"learning_rate": 1.9123505976095617e-06,
"loss": 0.6186,
"step": 49
},
{
"epoch": 0.06,
"grad_norm": 1.034868436810796,
"learning_rate": 1.952191235059761e-06,
"loss": 0.5675,
"step": 50
},
{
"epoch": 0.0612,
"grad_norm": 0.9562269599632713,
"learning_rate": 1.99203187250996e-06,
"loss": 0.5942,
"step": 51
},
{
"epoch": 0.0624,
"grad_norm": 0.9684662774229134,
"learning_rate": 2.03187250996016e-06,
"loss": 0.6069,
"step": 52
},
{
"epoch": 0.0636,
"grad_norm": 0.8597378190067636,
"learning_rate": 2.071713147410359e-06,
"loss": 0.5945,
"step": 53
},
{
"epoch": 0.0648,
"grad_norm": 0.8824670909191019,
"learning_rate": 2.111553784860558e-06,
"loss": 0.5827,
"step": 54
},
{
"epoch": 0.066,
"grad_norm": 0.8654040011891053,
"learning_rate": 2.151394422310757e-06,
"loss": 0.5595,
"step": 55
},
{
"epoch": 0.0672,
"grad_norm": 0.8364907260530519,
"learning_rate": 2.1912350597609563e-06,
"loss": 0.5796,
"step": 56
},
{
"epoch": 0.0684,
"grad_norm": 1.1546546198663707,
"learning_rate": 2.2310756972111555e-06,
"loss": 0.5941,
"step": 57
},
{
"epoch": 0.0696,
"grad_norm": 0.9420669315033214,
"learning_rate": 2.2709163346613547e-06,
"loss": 0.5198,
"step": 58
},
{
"epoch": 0.0708,
"grad_norm": 0.8528290907483634,
"learning_rate": 2.310756972111554e-06,
"loss": 0.552,
"step": 59
},
{
"epoch": 0.072,
"grad_norm": 0.687703387092168,
"learning_rate": 2.350597609561753e-06,
"loss": 0.5594,
"step": 60
},
{
"epoch": 0.0732,
"grad_norm": 0.8167294442985146,
"learning_rate": 2.3904382470119524e-06,
"loss": 0.5207,
"step": 61
},
{
"epoch": 0.0744,
"grad_norm": 0.7156708048117644,
"learning_rate": 2.4302788844621517e-06,
"loss": 0.5407,
"step": 62
},
{
"epoch": 0.0756,
"grad_norm": 0.7154050872369027,
"learning_rate": 2.470119521912351e-06,
"loss": 0.5025,
"step": 63
},
{
"epoch": 0.0768,
"grad_norm": 0.7943440935347835,
"learning_rate": 2.50996015936255e-06,
"loss": 0.5001,
"step": 64
},
{
"epoch": 0.078,
"grad_norm": 0.7903112896611016,
"learning_rate": 2.5498007968127493e-06,
"loss": 0.5137,
"step": 65
},
{
"epoch": 0.0792,
"grad_norm": 0.8102722270709948,
"learning_rate": 2.589641434262948e-06,
"loss": 0.556,
"step": 66
},
{
"epoch": 0.0804,
"grad_norm": 0.7763350160286278,
"learning_rate": 2.6294820717131474e-06,
"loss": 0.5618,
"step": 67
},
{
"epoch": 0.0816,
"grad_norm": 0.682413072872721,
"learning_rate": 2.6693227091633466e-06,
"loss": 0.4881,
"step": 68
},
{
"epoch": 0.0828,
"grad_norm": 0.737478801790203,
"learning_rate": 2.709163346613546e-06,
"loss": 0.5233,
"step": 69
},
{
"epoch": 0.084,
"grad_norm": 0.6670536824914544,
"learning_rate": 2.749003984063745e-06,
"loss": 0.5001,
"step": 70
},
{
"epoch": 0.0852,
"grad_norm": 0.7172377448602175,
"learning_rate": 2.7888446215139443e-06,
"loss": 0.4963,
"step": 71
},
{
"epoch": 0.0864,
"grad_norm": 0.7354871896590247,
"learning_rate": 2.828685258964144e-06,
"loss": 0.5183,
"step": 72
},
{
"epoch": 0.0876,
"grad_norm": 0.7366911418089935,
"learning_rate": 2.868525896414343e-06,
"loss": 0.5217,
"step": 73
},
{
"epoch": 0.0888,
"grad_norm": 0.7134400568775913,
"learning_rate": 2.9083665338645424e-06,
"loss": 0.4876,
"step": 74
},
{
"epoch": 0.09,
"grad_norm": 0.6927096104174592,
"learning_rate": 2.9482071713147416e-06,
"loss": 0.4966,
"step": 75
},
{
"epoch": 0.0912,
"grad_norm": 0.8230762404086284,
"learning_rate": 2.9880478087649404e-06,
"loss": 0.5208,
"step": 76
},
{
"epoch": 0.0924,
"grad_norm": 0.6292358040686521,
"learning_rate": 3.0278884462151397e-06,
"loss": 0.4882,
"step": 77
},
{
"epoch": 0.0936,
"grad_norm": 0.6147701154442085,
"learning_rate": 3.067729083665339e-06,
"loss": 0.5072,
"step": 78
},
{
"epoch": 0.0948,
"grad_norm": 0.676809508051408,
"learning_rate": 3.107569721115538e-06,
"loss": 0.5037,
"step": 79
},
{
"epoch": 0.096,
"grad_norm": 0.725489569911828,
"learning_rate": 3.1474103585657373e-06,
"loss": 0.5643,
"step": 80
},
{
"epoch": 0.0972,
"grad_norm": 0.7145147926010281,
"learning_rate": 3.1872509960159366e-06,
"loss": 0.497,
"step": 81
},
{
"epoch": 0.0984,
"grad_norm": 0.7170590419899483,
"learning_rate": 3.227091633466136e-06,
"loss": 0.5332,
"step": 82
},
{
"epoch": 0.0996,
"grad_norm": 0.7150216583981881,
"learning_rate": 3.2669322709163346e-06,
"loss": 0.529,
"step": 83
},
{
"epoch": 0.1008,
"grad_norm": 0.8613846664442866,
"learning_rate": 3.306772908366534e-06,
"loss": 0.5657,
"step": 84
},
{
"epoch": 0.102,
"grad_norm": 0.6494754319315506,
"learning_rate": 3.346613545816733e-06,
"loss": 0.5012,
"step": 85
},
{
"epoch": 0.1032,
"grad_norm": 0.6471011282897726,
"learning_rate": 3.3864541832669323e-06,
"loss": 0.5433,
"step": 86
},
{
"epoch": 0.1044,
"grad_norm": 0.750892718766989,
"learning_rate": 3.4262948207171315e-06,
"loss": 0.492,
"step": 87
},
{
"epoch": 0.1056,
"grad_norm": 0.687548849485851,
"learning_rate": 3.466135458167331e-06,
"loss": 0.5056,
"step": 88
},
{
"epoch": 0.1068,
"grad_norm": 0.5756351359995551,
"learning_rate": 3.5059760956175304e-06,
"loss": 0.5009,
"step": 89
},
{
"epoch": 0.108,
"grad_norm": 0.553648661965167,
"learning_rate": 3.5458167330677296e-06,
"loss": 0.5012,
"step": 90
},
{
"epoch": 0.1092,
"grad_norm": 0.6913182105350174,
"learning_rate": 3.585657370517929e-06,
"loss": 0.5025,
"step": 91
},
{
"epoch": 0.1104,
"grad_norm": 0.6951766377383839,
"learning_rate": 3.625498007968128e-06,
"loss": 0.4856,
"step": 92
},
{
"epoch": 0.1116,
"grad_norm": 0.8777499137224892,
"learning_rate": 3.665338645418327e-06,
"loss": 0.5467,
"step": 93
},
{
"epoch": 0.1128,
"grad_norm": 0.6508644020236586,
"learning_rate": 3.705179282868526e-06,
"loss": 0.5167,
"step": 94
},
{
"epoch": 0.114,
"grad_norm": 0.7808006683075185,
"learning_rate": 3.7450199203187254e-06,
"loss": 0.5321,
"step": 95
},
{
"epoch": 0.1152,
"grad_norm": 0.7388512378176182,
"learning_rate": 3.7848605577689246e-06,
"loss": 0.4805,
"step": 96
},
{
"epoch": 0.1164,
"grad_norm": 0.6408885474401719,
"learning_rate": 3.824701195219123e-06,
"loss": 0.5082,
"step": 97
},
{
"epoch": 0.1176,
"grad_norm": 0.69317399782124,
"learning_rate": 3.864541832669323e-06,
"loss": 0.4837,
"step": 98
},
{
"epoch": 0.1188,
"grad_norm": 0.6576371657994416,
"learning_rate": 3.904382470119522e-06,
"loss": 0.5152,
"step": 99
},
{
"epoch": 0.12,
"grad_norm": 0.6828846265642615,
"learning_rate": 3.9442231075697215e-06,
"loss": 0.4694,
"step": 100
},
{
"epoch": 0.1212,
"grad_norm": 0.7013083213096467,
"learning_rate": 3.98406374501992e-06,
"loss": 0.4717,
"step": 101
},
{
"epoch": 0.1224,
"grad_norm": 0.7085053688033561,
"learning_rate": 4.02390438247012e-06,
"loss": 0.5022,
"step": 102
},
{
"epoch": 0.1236,
"grad_norm": 0.622698153537725,
"learning_rate": 4.06374501992032e-06,
"loss": 0.472,
"step": 103
},
{
"epoch": 0.1248,
"grad_norm": 0.6203547767857722,
"learning_rate": 4.103585657370518e-06,
"loss": 0.5088,
"step": 104
},
{
"epoch": 0.126,
"grad_norm": 0.650054802792842,
"learning_rate": 4.143426294820718e-06,
"loss": 0.4799,
"step": 105
},
{
"epoch": 0.1272,
"grad_norm": 0.8174829915460686,
"learning_rate": 4.183266932270917e-06,
"loss": 0.5355,
"step": 106
},
{
"epoch": 0.1284,
"grad_norm": 0.7598310150274463,
"learning_rate": 4.223107569721116e-06,
"loss": 0.4735,
"step": 107
},
{
"epoch": 0.1296,
"grad_norm": 0.6175762629568949,
"learning_rate": 4.262948207171315e-06,
"loss": 0.4794,
"step": 108
},
{
"epoch": 0.1308,
"grad_norm": 0.6900127333389381,
"learning_rate": 4.302788844621514e-06,
"loss": 0.5021,
"step": 109
},
{
"epoch": 0.132,
"grad_norm": 0.7055276817230187,
"learning_rate": 4.342629482071714e-06,
"loss": 0.4629,
"step": 110
},
{
"epoch": 0.1332,
"grad_norm": 0.7842646328067162,
"learning_rate": 4.382470119521913e-06,
"loss": 0.488,
"step": 111
},
{
"epoch": 0.1344,
"grad_norm": 0.5941563107721001,
"learning_rate": 4.422310756972112e-06,
"loss": 0.4486,
"step": 112
},
{
"epoch": 0.1356,
"grad_norm": 0.7237315324309976,
"learning_rate": 4.462151394422311e-06,
"loss": 0.4986,
"step": 113
},
{
"epoch": 0.1368,
"grad_norm": 0.7257207423897334,
"learning_rate": 4.50199203187251e-06,
"loss": 0.5096,
"step": 114
},
{
"epoch": 0.138,
"grad_norm": 0.7465709953170436,
"learning_rate": 4.5418326693227095e-06,
"loss": 0.5142,
"step": 115
},
{
"epoch": 0.1392,
"grad_norm": 0.732912113525707,
"learning_rate": 4.581673306772908e-06,
"loss": 0.461,
"step": 116
},
{
"epoch": 0.1404,
"grad_norm": 0.7210355960852634,
"learning_rate": 4.621513944223108e-06,
"loss": 0.4912,
"step": 117
},
{
"epoch": 0.1416,
"grad_norm": 0.7445245458464089,
"learning_rate": 4.661354581673307e-06,
"loss": 0.4772,
"step": 118
},
{
"epoch": 0.1428,
"grad_norm": 0.7799623601896731,
"learning_rate": 4.701195219123506e-06,
"loss": 0.4767,
"step": 119
},
{
"epoch": 0.144,
"grad_norm": 0.7129943430352418,
"learning_rate": 4.741035856573706e-06,
"loss": 0.4656,
"step": 120
},
{
"epoch": 0.1452,
"grad_norm": 0.7461185079641656,
"learning_rate": 4.780876494023905e-06,
"loss": 0.4931,
"step": 121
},
{
"epoch": 0.1464,
"grad_norm": 0.7611752287702312,
"learning_rate": 4.8207171314741045e-06,
"loss": 0.5384,
"step": 122
},
{
"epoch": 0.1476,
"grad_norm": 0.7308153552924613,
"learning_rate": 4.860557768924303e-06,
"loss": 0.5044,
"step": 123
},
{
"epoch": 0.1488,
"grad_norm": 0.7504274125405789,
"learning_rate": 4.900398406374502e-06,
"loss": 0.4739,
"step": 124
},
{
"epoch": 0.15,
"grad_norm": 0.770948215484845,
"learning_rate": 4.940239043824702e-06,
"loss": 0.4838,
"step": 125
},
{
"epoch": 0.1512,
"grad_norm": 0.6552840020750029,
"learning_rate": 4.980079681274901e-06,
"loss": 0.4546,
"step": 126
},
{
"epoch": 0.1524,
"grad_norm": 0.6499796185675982,
"learning_rate": 5.0199203187251e-06,
"loss": 0.4281,
"step": 127
},
{
"epoch": 0.1536,
"grad_norm": 0.7061114170956655,
"learning_rate": 5.059760956175299e-06,
"loss": 0.4866,
"step": 128
},
{
"epoch": 0.1548,
"grad_norm": 0.6691189853551326,
"learning_rate": 5.099601593625499e-06,
"loss": 0.4656,
"step": 129
},
{
"epoch": 0.156,
"grad_norm": 0.6054265360566154,
"learning_rate": 5.1394422310756975e-06,
"loss": 0.4613,
"step": 130
},
{
"epoch": 0.1572,
"grad_norm": 0.7027909113103062,
"learning_rate": 5.179282868525896e-06,
"loss": 0.4839,
"step": 131
},
{
"epoch": 0.1584,
"grad_norm": 0.7587486957676393,
"learning_rate": 5.219123505976096e-06,
"loss": 0.4712,
"step": 132
},
{
"epoch": 0.1596,
"grad_norm": 0.6686724879335956,
"learning_rate": 5.258964143426295e-06,
"loss": 0.4807,
"step": 133
},
{
"epoch": 0.1608,
"grad_norm": 0.7646309087614517,
"learning_rate": 5.298804780876494e-06,
"loss": 0.5157,
"step": 134
},
{
"epoch": 0.162,
"grad_norm": 0.7313825683191312,
"learning_rate": 5.338645418326693e-06,
"loss": 0.4933,
"step": 135
},
{
"epoch": 0.1632,
"grad_norm": 0.6627767030401844,
"learning_rate": 5.378486055776893e-06,
"loss": 0.4536,
"step": 136
},
{
"epoch": 0.1644,
"grad_norm": 0.6888832810971692,
"learning_rate": 5.418326693227092e-06,
"loss": 0.529,
"step": 137
},
{
"epoch": 0.1656,
"grad_norm": 0.6727192064732628,
"learning_rate": 5.4581673306772905e-06,
"loss": 0.4668,
"step": 138
},
{
"epoch": 0.1668,
"grad_norm": 0.6712098985581779,
"learning_rate": 5.49800796812749e-06,
"loss": 0.5003,
"step": 139
},
{
"epoch": 0.168,
"grad_norm": 0.7449187354193221,
"learning_rate": 5.537848605577689e-06,
"loss": 0.4908,
"step": 140
},
{
"epoch": 0.1692,
"grad_norm": 0.6855809836685066,
"learning_rate": 5.577689243027889e-06,
"loss": 0.4957,
"step": 141
},
{
"epoch": 0.1704,
"grad_norm": 0.7057372872081618,
"learning_rate": 5.617529880478087e-06,
"loss": 0.4927,
"step": 142
},
{
"epoch": 0.1716,
"grad_norm": 0.7145902018121971,
"learning_rate": 5.657370517928288e-06,
"loss": 0.4798,
"step": 143
},
{
"epoch": 0.1728,
"grad_norm": 0.7178330938670102,
"learning_rate": 5.697211155378487e-06,
"loss": 0.487,
"step": 144
},
{
"epoch": 0.174,
"grad_norm": 0.6176066325572566,
"learning_rate": 5.737051792828686e-06,
"loss": 0.4484,
"step": 145
},
{
"epoch": 0.1752,
"grad_norm": 0.699828650914944,
"learning_rate": 5.776892430278885e-06,
"loss": 0.5061,
"step": 146
},
{
"epoch": 0.1764,
"grad_norm": 0.6256805259332396,
"learning_rate": 5.816733067729085e-06,
"loss": 0.4718,
"step": 147
},
{
"epoch": 0.1776,
"grad_norm": 0.671096546707992,
"learning_rate": 5.856573705179284e-06,
"loss": 0.4496,
"step": 148
},
{
"epoch": 0.1788,
"grad_norm": 0.6365351919676361,
"learning_rate": 5.896414342629483e-06,
"loss": 0.4771,
"step": 149
},
{
"epoch": 0.18,
"grad_norm": 0.7462898845468449,
"learning_rate": 5.936254980079682e-06,
"loss": 0.4875,
"step": 150
},
{
"epoch": 0.1812,
"grad_norm": 0.7734601052023893,
"learning_rate": 5.976095617529881e-06,
"loss": 0.4913,
"step": 151
},
{
"epoch": 0.1824,
"grad_norm": 0.6552628757396356,
"learning_rate": 6.0159362549800805e-06,
"loss": 0.4287,
"step": 152
},
{
"epoch": 0.1836,
"grad_norm": 0.7160128628436617,
"learning_rate": 6.055776892430279e-06,
"loss": 0.5023,
"step": 153
},
{
"epoch": 0.1848,
"grad_norm": 0.6975627744114324,
"learning_rate": 6.095617529880479e-06,
"loss": 0.5021,
"step": 154
},
{
"epoch": 0.186,
"grad_norm": 0.6958244479341551,
"learning_rate": 6.135458167330678e-06,
"loss": 0.5021,
"step": 155
},
{
"epoch": 0.1872,
"grad_norm": 0.7162813897115381,
"learning_rate": 6.1752988047808774e-06,
"loss": 0.4672,
"step": 156
},
{
"epoch": 0.1884,
"grad_norm": 0.7789415750652668,
"learning_rate": 6.215139442231076e-06,
"loss": 0.4896,
"step": 157
},
{
"epoch": 0.1896,
"grad_norm": 0.7168463214778908,
"learning_rate": 6.254980079681275e-06,
"loss": 0.4777,
"step": 158
},
{
"epoch": 0.1908,
"grad_norm": 0.8135516887242896,
"learning_rate": 6.294820717131475e-06,
"loss": 0.4925,
"step": 159
},
{
"epoch": 0.192,
"grad_norm": 0.6803879928943569,
"learning_rate": 6.3346613545816735e-06,
"loss": 0.4055,
"step": 160
},
{
"epoch": 0.1932,
"grad_norm": 0.7986039712293413,
"learning_rate": 6.374501992031873e-06,
"loss": 0.4986,
"step": 161
},
{
"epoch": 0.1944,
"grad_norm": 0.820290402644595,
"learning_rate": 6.414342629482072e-06,
"loss": 0.5198,
"step": 162
},
{
"epoch": 0.1956,
"grad_norm": 0.7151437249829041,
"learning_rate": 6.454183266932272e-06,
"loss": 0.4781,
"step": 163
},
{
"epoch": 0.1968,
"grad_norm": 0.7216410394435034,
"learning_rate": 6.49402390438247e-06,
"loss": 0.4695,
"step": 164
},
{
"epoch": 0.198,
"grad_norm": 0.811009148225086,
"learning_rate": 6.533864541832669e-06,
"loss": 0.5092,
"step": 165
},
{
"epoch": 0.1992,
"grad_norm": 0.7295087084663571,
"learning_rate": 6.573705179282869e-06,
"loss": 0.443,
"step": 166
},
{
"epoch": 0.2004,
"grad_norm": 0.7762881643704093,
"learning_rate": 6.613545816733068e-06,
"loss": 0.463,
"step": 167
},
{
"epoch": 0.2016,
"grad_norm": 0.6598351418548212,
"learning_rate": 6.653386454183267e-06,
"loss": 0.4849,
"step": 168
},
{
"epoch": 0.2028,
"grad_norm": 0.6153969557744298,
"learning_rate": 6.693227091633466e-06,
"loss": 0.4856,
"step": 169
},
{
"epoch": 0.204,
"grad_norm": 0.7406116382206025,
"learning_rate": 6.733067729083666e-06,
"loss": 0.5247,
"step": 170
},
{
"epoch": 0.2052,
"grad_norm": 0.7368196241047612,
"learning_rate": 6.772908366533865e-06,
"loss": 0.4725,
"step": 171
},
{
"epoch": 0.2064,
"grad_norm": 0.8195250297003582,
"learning_rate": 6.812749003984063e-06,
"loss": 0.4776,
"step": 172
},
{
"epoch": 0.2076,
"grad_norm": 0.7902293780066731,
"learning_rate": 6.852589641434263e-06,
"loss": 0.4758,
"step": 173
},
{
"epoch": 0.2088,
"grad_norm": 0.6707040821942747,
"learning_rate": 6.8924302788844635e-06,
"loss": 0.4627,
"step": 174
},
{
"epoch": 0.21,
"grad_norm": 0.7130948005327684,
"learning_rate": 6.932270916334662e-06,
"loss": 0.5399,
"step": 175
},
{
"epoch": 0.2112,
"grad_norm": 0.6547820529997068,
"learning_rate": 6.972111553784862e-06,
"loss": 0.4766,
"step": 176
},
{
"epoch": 0.2124,
"grad_norm": 0.7146947814036898,
"learning_rate": 7.011952191235061e-06,
"loss": 0.4411,
"step": 177
},
{
"epoch": 0.2136,
"grad_norm": 0.7396734814075536,
"learning_rate": 7.05179282868526e-06,
"loss": 0.4819,
"step": 178
},
{
"epoch": 0.2148,
"grad_norm": 0.9240207511689131,
"learning_rate": 7.091633466135459e-06,
"loss": 0.5158,
"step": 179
},
{
"epoch": 0.216,
"grad_norm": 0.6587494748649532,
"learning_rate": 7.131474103585658e-06,
"loss": 0.4513,
"step": 180
},
{
"epoch": 0.2172,
"grad_norm": 0.7451455583479791,
"learning_rate": 7.171314741035858e-06,
"loss": 0.5298,
"step": 181
},
{
"epoch": 0.2184,
"grad_norm": 0.7183108926930426,
"learning_rate": 7.2111553784860565e-06,
"loss": 0.455,
"step": 182
},
{
"epoch": 0.2196,
"grad_norm": 0.7294845173799966,
"learning_rate": 7.250996015936256e-06,
"loss": 0.5282,
"step": 183
},
{
"epoch": 0.2208,
"grad_norm": 0.7829206050862241,
"learning_rate": 7.290836653386455e-06,
"loss": 0.4999,
"step": 184
},
{
"epoch": 0.222,
"grad_norm": 0.8144601373674129,
"learning_rate": 7.330677290836654e-06,
"loss": 0.4479,
"step": 185
},
{
"epoch": 0.2232,
"grad_norm": 0.686434544081989,
"learning_rate": 7.3705179282868534e-06,
"loss": 0.4779,
"step": 186
},
{
"epoch": 0.2244,
"grad_norm": 0.7662036672164863,
"learning_rate": 7.410358565737052e-06,
"loss": 0.4406,
"step": 187
},
{
"epoch": 0.2256,
"grad_norm": 0.7372149806305216,
"learning_rate": 7.450199203187252e-06,
"loss": 0.4739,
"step": 188
},
{
"epoch": 0.2268,
"grad_norm": 0.7018904889059249,
"learning_rate": 7.490039840637451e-06,
"loss": 0.4739,
"step": 189
},
{
"epoch": 0.228,
"grad_norm": 0.8034958475921105,
"learning_rate": 7.52988047808765e-06,
"loss": 0.4871,
"step": 190
},
{
"epoch": 0.2292,
"grad_norm": 0.6664145984053934,
"learning_rate": 7.569721115537849e-06,
"loss": 0.4785,
"step": 191
},
{
"epoch": 0.2304,
"grad_norm": 0.8349490725943484,
"learning_rate": 7.609561752988048e-06,
"loss": 0.4728,
"step": 192
},
{
"epoch": 0.2316,
"grad_norm": 0.8371769404608206,
"learning_rate": 7.649402390438247e-06,
"loss": 0.49,
"step": 193
},
{
"epoch": 0.2328,
"grad_norm": 0.6234603463023238,
"learning_rate": 7.689243027888446e-06,
"loss": 0.4656,
"step": 194
},
{
"epoch": 0.234,
"grad_norm": 0.7514666066051947,
"learning_rate": 7.729083665338646e-06,
"loss": 0.4555,
"step": 195
},
{
"epoch": 0.2352,
"grad_norm": 0.7598762484166676,
"learning_rate": 7.768924302788846e-06,
"loss": 0.522,
"step": 196
},
{
"epoch": 0.2364,
"grad_norm": 0.6424752977780204,
"learning_rate": 7.808764940239044e-06,
"loss": 0.4565,
"step": 197
},
{
"epoch": 0.2376,
"grad_norm": 0.755381522376294,
"learning_rate": 7.848605577689243e-06,
"loss": 0.5013,
"step": 198
},
{
"epoch": 0.2388,
"grad_norm": 0.6978248462713278,
"learning_rate": 7.888446215139443e-06,
"loss": 0.4662,
"step": 199
},
{
"epoch": 0.24,
"grad_norm": 0.5877441481282527,
"learning_rate": 7.928286852589641e-06,
"loss": 0.4795,
"step": 200
},
{
"epoch": 0.2412,
"grad_norm": 0.6401561354692661,
"learning_rate": 7.96812749003984e-06,
"loss": 0.4533,
"step": 201
},
{
"epoch": 0.2424,
"grad_norm": 0.7088942657140168,
"learning_rate": 8.00796812749004e-06,
"loss": 0.4473,
"step": 202
},
{
"epoch": 0.2436,
"grad_norm": 0.7601768924017626,
"learning_rate": 8.04780876494024e-06,
"loss": 0.4763,
"step": 203
},
{
"epoch": 0.2448,
"grad_norm": 0.7125101036178078,
"learning_rate": 8.087649402390438e-06,
"loss": 0.4672,
"step": 204
},
{
"epoch": 0.246,
"grad_norm": 0.8043890918009915,
"learning_rate": 8.12749003984064e-06,
"loss": 0.437,
"step": 205
},
{
"epoch": 0.2472,
"grad_norm": 0.6323381607535531,
"learning_rate": 8.167330677290837e-06,
"loss": 0.4562,
"step": 206
},
{
"epoch": 0.2484,
"grad_norm": 0.6219753280994653,
"learning_rate": 8.207171314741037e-06,
"loss": 0.5002,
"step": 207
},
{
"epoch": 0.2496,
"grad_norm": 0.6933892431644402,
"learning_rate": 8.247011952191236e-06,
"loss": 0.4716,
"step": 208
},
{
"epoch": 0.2508,
"grad_norm": 0.7369573747691623,
"learning_rate": 8.286852589641436e-06,
"loss": 0.4511,
"step": 209
},
{
"epoch": 0.252,
"grad_norm": 0.7463770460556426,
"learning_rate": 8.326693227091634e-06,
"loss": 0.5064,
"step": 210
},
{
"epoch": 0.2532,
"grad_norm": 0.7244915397995428,
"learning_rate": 8.366533864541834e-06,
"loss": 0.4673,
"step": 211
},
{
"epoch": 0.2544,
"grad_norm": 0.8064801832473215,
"learning_rate": 8.406374501992033e-06,
"loss": 0.5211,
"step": 212
},
{
"epoch": 0.2556,
"grad_norm": 0.7652570679960387,
"learning_rate": 8.446215139442231e-06,
"loss": 0.5202,
"step": 213
},
{
"epoch": 0.2568,
"grad_norm": 0.7589679248842585,
"learning_rate": 8.486055776892431e-06,
"loss": 0.4553,
"step": 214
},
{
"epoch": 0.258,
"grad_norm": 0.7966754694324576,
"learning_rate": 8.52589641434263e-06,
"loss": 0.5175,
"step": 215
},
{
"epoch": 0.2592,
"grad_norm": 0.7377325050844753,
"learning_rate": 8.56573705179283e-06,
"loss": 0.4496,
"step": 216
},
{
"epoch": 0.2604,
"grad_norm": 0.7749878316688397,
"learning_rate": 8.605577689243028e-06,
"loss": 0.4658,
"step": 217
},
{
"epoch": 0.2616,
"grad_norm": 0.6787616893433324,
"learning_rate": 8.645418326693228e-06,
"loss": 0.4414,
"step": 218
},
{
"epoch": 0.2628,
"grad_norm": 0.7222976534056557,
"learning_rate": 8.685258964143428e-06,
"loss": 0.4508,
"step": 219
},
{
"epoch": 0.264,
"grad_norm": 0.8075552569283245,
"learning_rate": 8.725099601593626e-06,
"loss": 0.4243,
"step": 220
},
{
"epoch": 0.2652,
"grad_norm": 0.6491598307437855,
"learning_rate": 8.764940239043825e-06,
"loss": 0.4997,
"step": 221
},
{
"epoch": 0.2664,
"grad_norm": 0.6844778520577768,
"learning_rate": 8.804780876494025e-06,
"loss": 0.4858,
"step": 222
},
{
"epoch": 0.2676,
"grad_norm": 0.7299226965380259,
"learning_rate": 8.844621513944224e-06,
"loss": 0.4667,
"step": 223
},
{
"epoch": 0.2688,
"grad_norm": 0.7229205916891049,
"learning_rate": 8.884462151394422e-06,
"loss": 0.4917,
"step": 224
},
{
"epoch": 0.27,
"grad_norm": 0.748705120455487,
"learning_rate": 8.924302788844622e-06,
"loss": 0.4817,
"step": 225
},
{
"epoch": 0.2712,
"grad_norm": 0.677971001063236,
"learning_rate": 8.964143426294822e-06,
"loss": 0.4709,
"step": 226
},
{
"epoch": 0.2724,
"grad_norm": 0.7616943137332174,
"learning_rate": 9.00398406374502e-06,
"loss": 0.4808,
"step": 227
},
{
"epoch": 0.2736,
"grad_norm": 0.7143523129028485,
"learning_rate": 9.04382470119522e-06,
"loss": 0.412,
"step": 228
},
{
"epoch": 0.2748,
"grad_norm": 0.8915146206619335,
"learning_rate": 9.083665338645419e-06,
"loss": 0.5016,
"step": 229
},
{
"epoch": 0.276,
"grad_norm": 0.7424475381270349,
"learning_rate": 9.123505976095619e-06,
"loss": 0.4243,
"step": 230
},
{
"epoch": 0.2772,
"grad_norm": 0.7559501710877834,
"learning_rate": 9.163346613545817e-06,
"loss": 0.458,
"step": 231
},
{
"epoch": 0.2784,
"grad_norm": 0.7142866700220379,
"learning_rate": 9.203187250996016e-06,
"loss": 0.4185,
"step": 232
},
{
"epoch": 0.2796,
"grad_norm": 0.7452270504572424,
"learning_rate": 9.243027888446216e-06,
"loss": 0.4892,
"step": 233
},
{
"epoch": 0.2808,
"grad_norm": 0.8804844254437856,
"learning_rate": 9.282868525896414e-06,
"loss": 0.5244,
"step": 234
},
{
"epoch": 0.282,
"grad_norm": 0.8102009036568012,
"learning_rate": 9.322709163346614e-06,
"loss": 0.5066,
"step": 235
},
{
"epoch": 0.2832,
"grad_norm": 0.7933963116447116,
"learning_rate": 9.362549800796813e-06,
"loss": 0.4986,
"step": 236
},
{
"epoch": 0.2844,
"grad_norm": 0.7483634628646818,
"learning_rate": 9.402390438247013e-06,
"loss": 0.5034,
"step": 237
},
{
"epoch": 0.2856,
"grad_norm": 0.8277784663677117,
"learning_rate": 9.442231075697212e-06,
"loss": 0.5524,
"step": 238
},
{
"epoch": 0.2868,
"grad_norm": 0.6631534504892171,
"learning_rate": 9.482071713147412e-06,
"loss": 0.4541,
"step": 239
},
{
"epoch": 0.288,
"grad_norm": 0.78921647058052,
"learning_rate": 9.52191235059761e-06,
"loss": 0.4768,
"step": 240
},
{
"epoch": 0.2892,
"grad_norm": 0.8301076778286669,
"learning_rate": 9.56175298804781e-06,
"loss": 0.4848,
"step": 241
},
{
"epoch": 0.2904,
"grad_norm": 0.7697646502273703,
"learning_rate": 9.60159362549801e-06,
"loss": 0.4503,
"step": 242
},
{
"epoch": 0.2916,
"grad_norm": 0.6985190325570851,
"learning_rate": 9.641434262948209e-06,
"loss": 0.4828,
"step": 243
},
{
"epoch": 0.2928,
"grad_norm": 0.7025564878021328,
"learning_rate": 9.681274900398407e-06,
"loss": 0.438,
"step": 244
},
{
"epoch": 0.294,
"grad_norm": 0.7620206097667034,
"learning_rate": 9.721115537848607e-06,
"loss": 0.43,
"step": 245
},
{
"epoch": 0.2952,
"grad_norm": 0.7424593029148787,
"learning_rate": 9.760956175298806e-06,
"loss": 0.4403,
"step": 246
},
{
"epoch": 0.2964,
"grad_norm": 0.7360561757895007,
"learning_rate": 9.800796812749004e-06,
"loss": 0.4622,
"step": 247
},
{
"epoch": 0.2976,
"grad_norm": 0.8243384149245737,
"learning_rate": 9.840637450199204e-06,
"loss": 0.4751,
"step": 248
},
{
"epoch": 0.2988,
"grad_norm": 0.8199288515333606,
"learning_rate": 9.880478087649404e-06,
"loss": 0.462,
"step": 249
},
{
"epoch": 0.3,
"grad_norm": 0.7221117049284725,
"learning_rate": 9.920318725099603e-06,
"loss": 0.458,
"step": 250
},
{
"epoch": 0.3012,
"grad_norm": 0.7821380388196548,
"learning_rate": 9.960159362549801e-06,
"loss": 0.471,
"step": 251
},
{
"epoch": 0.3024,
"grad_norm": 0.726002114022043,
"learning_rate": 1e-05,
"loss": 0.4695,
"step": 252
},
{
"epoch": 0.3036,
"grad_norm": 0.8286938369018219,
"learning_rate": 9.999995130451526e-06,
"loss": 0.4656,
"step": 253
},
{
"epoch": 0.3048,
"grad_norm": 0.7564878742577376,
"learning_rate": 9.999980521815582e-06,
"loss": 0.4948,
"step": 254
},
{
"epoch": 0.306,
"grad_norm": 0.7000729873027478,
"learning_rate": 9.999956174120626e-06,
"loss": 0.4332,
"step": 255
},
{
"epoch": 0.3072,
"grad_norm": 0.8115383428570657,
"learning_rate": 9.999922087414084e-06,
"loss": 0.4798,
"step": 256
},
{
"epoch": 0.3084,
"grad_norm": 0.7624880545561682,
"learning_rate": 9.99987826176235e-06,
"loss": 0.4566,
"step": 257
},
{
"epoch": 0.3096,
"grad_norm": 0.6533963701995729,
"learning_rate": 9.999824697250786e-06,
"loss": 0.4639,
"step": 258
},
{
"epoch": 0.3108,
"grad_norm": 0.787105570960542,
"learning_rate": 9.999761393983728e-06,
"loss": 0.4788,
"step": 259
},
{
"epoch": 0.312,
"grad_norm": 0.7293361363854345,
"learning_rate": 9.999688352084482e-06,
"loss": 0.4786,
"step": 260
},
{
"epoch": 0.3132,
"grad_norm": 0.6894092015944603,
"learning_rate": 9.999605571695317e-06,
"loss": 0.4621,
"step": 261
},
{
"epoch": 0.3144,
"grad_norm": 0.7674479288861019,
"learning_rate": 9.999513052977473e-06,
"loss": 0.4585,
"step": 262
},
{
"epoch": 0.3156,
"grad_norm": 0.6382731142387786,
"learning_rate": 9.999410796111163e-06,
"loss": 0.4536,
"step": 263
},
{
"epoch": 0.3168,
"grad_norm": 0.662827100843487,
"learning_rate": 9.999298801295564e-06,
"loss": 0.453,
"step": 264
},
{
"epoch": 0.318,
"grad_norm": 0.6997882828244627,
"learning_rate": 9.99917706874882e-06,
"loss": 0.4207,
"step": 265
},
{
"epoch": 0.3192,
"grad_norm": 0.6673329177111227,
"learning_rate": 9.999045598708047e-06,
"loss": 0.4303,
"step": 266
},
{
"epoch": 0.3204,
"grad_norm": 0.7751069968264359,
"learning_rate": 9.998904391429323e-06,
"loss": 0.4857,
"step": 267
},
{
"epoch": 0.3216,
"grad_norm": 0.6924789628677505,
"learning_rate": 9.998753447187693e-06,
"loss": 0.4352,
"step": 268
},
{
"epoch": 0.3228,
"grad_norm": 0.6780594398982913,
"learning_rate": 9.998592766277173e-06,
"loss": 0.4944,
"step": 269
},
{
"epoch": 0.324,
"grad_norm": 0.665053499034254,
"learning_rate": 9.998422349010736e-06,
"loss": 0.4452,
"step": 270
},
{
"epoch": 0.3252,
"grad_norm": 0.6702681281204724,
"learning_rate": 9.998242195720327e-06,
"loss": 0.4325,
"step": 271
},
{
"epoch": 0.3264,
"grad_norm": 0.726150711484855,
"learning_rate": 9.998052306756852e-06,
"loss": 0.5118,
"step": 272
},
{
"epoch": 0.3276,
"grad_norm": 0.6717001483141275,
"learning_rate": 9.997852682490179e-06,
"loss": 0.5304,
"step": 273
},
{
"epoch": 0.3288,
"grad_norm": 0.7846459341475974,
"learning_rate": 9.997643323309139e-06,
"loss": 0.4412,
"step": 274
},
{
"epoch": 0.33,
"grad_norm": 0.6081818827878402,
"learning_rate": 9.997424229621529e-06,
"loss": 0.4469,
"step": 275
},
{
"epoch": 0.3312,
"grad_norm": 0.7262395716436079,
"learning_rate": 9.997195401854102e-06,
"loss": 0.4597,
"step": 276
},
{
"epoch": 0.3324,
"grad_norm": 0.7600492053225586,
"learning_rate": 9.996956840452573e-06,
"loss": 0.4595,
"step": 277
},
{
"epoch": 0.3336,
"grad_norm": 0.8932083636988103,
"learning_rate": 9.996708545881617e-06,
"loss": 0.5023,
"step": 278
},
{
"epoch": 0.3348,
"grad_norm": 0.8005056158540009,
"learning_rate": 9.996450518624868e-06,
"loss": 0.5398,
"step": 279
},
{
"epoch": 0.336,
"grad_norm": 0.6480206111944329,
"learning_rate": 9.996182759184916e-06,
"loss": 0.4722,
"step": 280
},
{
"epoch": 0.3372,
"grad_norm": 0.7959325904926813,
"learning_rate": 9.995905268083306e-06,
"loss": 0.4481,
"step": 281
},
{
"epoch": 0.3384,
"grad_norm": 0.7204153062584008,
"learning_rate": 9.995618045860545e-06,
"loss": 0.4842,
"step": 282
},
{
"epoch": 0.3396,
"grad_norm": 0.72854290136139,
"learning_rate": 9.995321093076085e-06,
"loss": 0.4987,
"step": 283
},
{
"epoch": 0.3408,
"grad_norm": 0.897424392822921,
"learning_rate": 9.995014410308336e-06,
"loss": 0.4771,
"step": 284
},
{
"epoch": 0.342,
"grad_norm": 0.6147300343692838,
"learning_rate": 9.994697998154668e-06,
"loss": 0.4521,
"step": 285
},
{
"epoch": 0.3432,
"grad_norm": 0.6948770085824113,
"learning_rate": 9.994371857231388e-06,
"loss": 0.4669,
"step": 286
},
{
"epoch": 0.3444,
"grad_norm": 0.6152586871071891,
"learning_rate": 9.99403598817376e-06,
"loss": 0.4563,
"step": 287
},
{
"epoch": 0.3456,
"grad_norm": 0.6314128897864619,
"learning_rate": 9.993690391636e-06,
"loss": 0.4731,
"step": 288
},
{
"epoch": 0.3468,
"grad_norm": 0.7829857719356524,
"learning_rate": 9.993335068291264e-06,
"loss": 0.4622,
"step": 289
},
{
"epoch": 0.348,
"grad_norm": 0.6737044799083258,
"learning_rate": 9.99297001883166e-06,
"loss": 0.4813,
"step": 290
},
{
"epoch": 0.3492,
"grad_norm": 0.5917541108922811,
"learning_rate": 9.992595243968238e-06,
"loss": 0.4081,
"step": 291
},
{
"epoch": 0.3504,
"grad_norm": 0.6643962397169046,
"learning_rate": 9.99221074443099e-06,
"loss": 0.4286,
"step": 292
},
{
"epoch": 0.3516,
"grad_norm": 0.7065581838104028,
"learning_rate": 9.991816520968853e-06,
"loss": 0.4936,
"step": 293
},
{
"epoch": 0.3528,
"grad_norm": 0.75814172453645,
"learning_rate": 9.991412574349704e-06,
"loss": 0.4056,
"step": 294
},
{
"epoch": 0.354,
"grad_norm": 0.8194643924909449,
"learning_rate": 9.990998905360357e-06,
"loss": 0.5801,
"step": 295
},
{
"epoch": 0.3552,
"grad_norm": 0.763485244925637,
"learning_rate": 9.990575514806563e-06,
"loss": 0.5059,
"step": 296
},
{
"epoch": 0.3564,
"grad_norm": 0.7735447262387536,
"learning_rate": 9.990142403513012e-06,
"loss": 0.4766,
"step": 297
},
{
"epoch": 0.3576,
"grad_norm": 0.7264381504889462,
"learning_rate": 9.989699572323328e-06,
"loss": 0.4333,
"step": 298
},
{
"epoch": 0.3588,
"grad_norm": 0.7271709281599659,
"learning_rate": 9.989247022100065e-06,
"loss": 0.4666,
"step": 299
},
{
"epoch": 0.36,
"grad_norm": 0.8211659316661323,
"learning_rate": 9.988784753724707e-06,
"loss": 0.5099,
"step": 300
},
{
"epoch": 0.3612,
"grad_norm": 0.659254969233651,
"learning_rate": 9.988312768097673e-06,
"loss": 0.4643,
"step": 301
},
{
"epoch": 0.3624,
"grad_norm": 0.7484170942453734,
"learning_rate": 9.987831066138302e-06,
"loss": 0.5009,
"step": 302
},
{
"epoch": 0.3636,
"grad_norm": 0.7864892723792284,
"learning_rate": 9.987339648784866e-06,
"loss": 0.4203,
"step": 303
},
{
"epoch": 0.3648,
"grad_norm": 0.6463979301616607,
"learning_rate": 9.986838516994555e-06,
"loss": 0.4561,
"step": 304
},
{
"epoch": 0.366,
"grad_norm": 0.6423447978300654,
"learning_rate": 9.986327671743484e-06,
"loss": 0.4574,
"step": 305
},
{
"epoch": 0.3672,
"grad_norm": 0.7542345067080994,
"learning_rate": 9.985807114026684e-06,
"loss": 0.4578,
"step": 306
},
{
"epoch": 0.3684,
"grad_norm": 0.7439657684658759,
"learning_rate": 9.985276844858114e-06,
"loss": 0.4514,
"step": 307
},
{
"epoch": 0.3696,
"grad_norm": 0.6667505301910438,
"learning_rate": 9.984736865270637e-06,
"loss": 0.4599,
"step": 308
},
{
"epoch": 0.3708,
"grad_norm": 0.8943884014425674,
"learning_rate": 9.984187176316038e-06,
"loss": 0.4623,
"step": 309
},
{
"epoch": 0.372,
"grad_norm": 0.696770872305506,
"learning_rate": 9.983627779065012e-06,
"loss": 0.4571,
"step": 310
},
{
"epoch": 0.3732,
"grad_norm": 0.7386939862481439,
"learning_rate": 9.983058674607164e-06,
"loss": 0.5217,
"step": 311
},
{
"epoch": 0.3744,
"grad_norm": 0.696077560396179,
"learning_rate": 9.982479864051005e-06,
"loss": 0.4162,
"step": 312
},
{
"epoch": 0.3756,
"grad_norm": 0.8530728224344589,
"learning_rate": 9.981891348523955e-06,
"loss": 0.4728,
"step": 313
},
{
"epoch": 0.3768,
"grad_norm": 0.771830568256366,
"learning_rate": 9.981293129172334e-06,
"loss": 0.485,
"step": 314
},
{
"epoch": 0.378,
"grad_norm": 0.7063337203590513,
"learning_rate": 9.980685207161368e-06,
"loss": 0.4422,
"step": 315
},
{
"epoch": 0.3792,
"grad_norm": 0.7224325864884874,
"learning_rate": 9.980067583675177e-06,
"loss": 0.4256,
"step": 316
},
{
"epoch": 0.3804,
"grad_norm": 0.6246003822583702,
"learning_rate": 9.979440259916782e-06,
"loss": 0.4678,
"step": 317
},
{
"epoch": 0.3816,
"grad_norm": 0.7559983887815412,
"learning_rate": 9.978803237108095e-06,
"loss": 0.4484,
"step": 318
},
{
"epoch": 0.3828,
"grad_norm": 0.6978566836204545,
"learning_rate": 9.97815651648992e-06,
"loss": 0.4573,
"step": 319
},
{
"epoch": 0.384,
"grad_norm": 0.6182832620711831,
"learning_rate": 9.977500099321956e-06,
"loss": 0.4565,
"step": 320
},
{
"epoch": 0.3852,
"grad_norm": 0.7203458615299193,
"learning_rate": 9.97683398688278e-06,
"loss": 0.471,
"step": 321
},
{
"epoch": 0.3864,
"grad_norm": 0.6931652914141438,
"learning_rate": 9.976158180469866e-06,
"loss": 0.4371,
"step": 322
},
{
"epoch": 0.3876,
"grad_norm": 0.7575323963763007,
"learning_rate": 9.975472681399556e-06,
"loss": 0.4439,
"step": 323
},
{
"epoch": 0.3888,
"grad_norm": 0.5994997176503296,
"learning_rate": 9.97477749100708e-06,
"loss": 0.4709,
"step": 324
},
{
"epoch": 0.39,
"grad_norm": 1.2959886351973398,
"learning_rate": 9.974072610646543e-06,
"loss": 0.496,
"step": 325
},
{
"epoch": 0.3912,
"grad_norm": 0.7430734453560549,
"learning_rate": 9.973358041690926e-06,
"loss": 0.4611,
"step": 326
},
{
"epoch": 0.3924,
"grad_norm": 0.5893403359291437,
"learning_rate": 9.972633785532082e-06,
"loss": 0.4519,
"step": 327
},
{
"epoch": 0.3936,
"grad_norm": 0.641216294197375,
"learning_rate": 9.971899843580728e-06,
"loss": 0.4632,
"step": 328
},
{
"epoch": 0.3948,
"grad_norm": 0.7207213873714433,
"learning_rate": 9.971156217266451e-06,
"loss": 0.4196,
"step": 329
},
{
"epoch": 0.396,
"grad_norm": 0.6968100390508554,
"learning_rate": 9.970402908037703e-06,
"loss": 0.4658,
"step": 330
},
{
"epoch": 0.3972,
"grad_norm": 0.6672738370323194,
"learning_rate": 9.96963991736179e-06,
"loss": 0.5153,
"step": 331
},
{
"epoch": 0.3984,
"grad_norm": 0.738487994232809,
"learning_rate": 9.968867246724882e-06,
"loss": 0.4853,
"step": 332
},
{
"epoch": 0.3996,
"grad_norm": 0.6741190624842869,
"learning_rate": 9.968084897632004e-06,
"loss": 0.4555,
"step": 333
},
{
"epoch": 0.4008,
"grad_norm": 0.6061071531934571,
"learning_rate": 9.96729287160703e-06,
"loss": 0.4528,
"step": 334
},
{
"epoch": 0.402,
"grad_norm": 0.668424285013486,
"learning_rate": 9.966491170192682e-06,
"loss": 0.4664,
"step": 335
},
{
"epoch": 0.4032,
"grad_norm": 0.71365063515845,
"learning_rate": 9.96567979495053e-06,
"loss": 0.4182,
"step": 336
},
{
"epoch": 0.4044,
"grad_norm": 0.6537179787219397,
"learning_rate": 9.964858747460989e-06,
"loss": 0.4339,
"step": 337
},
{
"epoch": 0.4056,
"grad_norm": 0.6696665849925841,
"learning_rate": 9.964028029323305e-06,
"loss": 0.4289,
"step": 338
},
{
"epoch": 0.4068,
"grad_norm": 0.6317580155063051,
"learning_rate": 9.963187642155573e-06,
"loss": 0.4945,
"step": 339
},
{
"epoch": 0.408,
"grad_norm": 0.6204662379721577,
"learning_rate": 9.962337587594713e-06,
"loss": 0.4567,
"step": 340
},
{
"epoch": 0.4092,
"grad_norm": 0.790028299585896,
"learning_rate": 9.961477867296479e-06,
"loss": 0.4402,
"step": 341
},
{
"epoch": 0.4104,
"grad_norm": 0.7057465563218057,
"learning_rate": 9.96060848293545e-06,
"loss": 0.4367,
"step": 342
},
{
"epoch": 0.4116,
"grad_norm": 0.7048538659143536,
"learning_rate": 9.959729436205027e-06,
"loss": 0.4263,
"step": 343
},
{
"epoch": 0.4128,
"grad_norm": 0.7532569063688792,
"learning_rate": 9.95884072881744e-06,
"loss": 0.4515,
"step": 344
},
{
"epoch": 0.414,
"grad_norm": 0.6432955870583814,
"learning_rate": 9.957942362503728e-06,
"loss": 0.4401,
"step": 345
},
{
"epoch": 0.4152,
"grad_norm": 0.8190892141495807,
"learning_rate": 9.957034339013742e-06,
"loss": 0.4806,
"step": 346
},
{
"epoch": 0.4164,
"grad_norm": 0.6171924188310567,
"learning_rate": 9.956116660116155e-06,
"loss": 0.4543,
"step": 347
},
{
"epoch": 0.4176,
"grad_norm": 0.7490952693283688,
"learning_rate": 9.955189327598435e-06,
"loss": 0.4771,
"step": 348
},
{
"epoch": 0.4188,
"grad_norm": 0.6659631701706887,
"learning_rate": 9.954252343266859e-06,
"loss": 0.4746,
"step": 349
},
{
"epoch": 0.42,
"grad_norm": 0.7006208304084274,
"learning_rate": 9.953305708946504e-06,
"loss": 0.4549,
"step": 350
},
{
"epoch": 0.4212,
"grad_norm": 0.7723073626680784,
"learning_rate": 9.952349426481243e-06,
"loss": 0.451,
"step": 351
},
{
"epoch": 0.4224,
"grad_norm": 0.6455592166376946,
"learning_rate": 9.95138349773374e-06,
"loss": 0.4469,
"step": 352
},
{
"epoch": 0.4236,
"grad_norm": 0.6168164821410606,
"learning_rate": 9.95040792458545e-06,
"loss": 0.446,
"step": 353
},
{
"epoch": 0.4248,
"grad_norm": 0.7884092477255151,
"learning_rate": 9.949422708936616e-06,
"loss": 0.422,
"step": 354
},
{
"epoch": 0.426,
"grad_norm": 0.698668525342129,
"learning_rate": 9.948427852706257e-06,
"loss": 0.4477,
"step": 355
},
{
"epoch": 0.4272,
"grad_norm": 0.6442795235670659,
"learning_rate": 9.947423357832176e-06,
"loss": 0.4609,
"step": 356
},
{
"epoch": 0.4284,
"grad_norm": 0.642341192268614,
"learning_rate": 9.946409226270945e-06,
"loss": 0.4454,
"step": 357
},
{
"epoch": 0.4296,
"grad_norm": 0.8390627888315508,
"learning_rate": 9.945385459997909e-06,
"loss": 0.4801,
"step": 358
},
{
"epoch": 0.4308,
"grad_norm": 0.6449530595213682,
"learning_rate": 9.944352061007182e-06,
"loss": 0.4386,
"step": 359
},
{
"epoch": 0.432,
"grad_norm": 0.7703127556763665,
"learning_rate": 9.943309031311637e-06,
"loss": 0.4715,
"step": 360
},
{
"epoch": 0.4332,
"grad_norm": 0.6509249606252259,
"learning_rate": 9.942256372942909e-06,
"loss": 0.4483,
"step": 361
},
{
"epoch": 0.4344,
"grad_norm": 0.7656766131252136,
"learning_rate": 9.941194087951384e-06,
"loss": 0.4601,
"step": 362
},
{
"epoch": 0.4356,
"grad_norm": 0.6898891431076177,
"learning_rate": 9.940122178406205e-06,
"loss": 0.4399,
"step": 363
},
{
"epoch": 0.4368,
"grad_norm": 0.7159579906071247,
"learning_rate": 9.939040646395252e-06,
"loss": 0.476,
"step": 364
},
{
"epoch": 0.438,
"grad_norm": 0.627792587615608,
"learning_rate": 9.93794949402516e-06,
"loss": 0.4547,
"step": 365
},
{
"epoch": 0.4392,
"grad_norm": 0.645333115151413,
"learning_rate": 9.936848723421295e-06,
"loss": 0.4094,
"step": 366
},
{
"epoch": 0.4404,
"grad_norm": 0.7175839194918624,
"learning_rate": 9.93573833672776e-06,
"loss": 0.4823,
"step": 367
},
{
"epoch": 0.4416,
"grad_norm": 0.6214300099183283,
"learning_rate": 9.934618336107385e-06,
"loss": 0.4408,
"step": 368
},
{
"epoch": 0.4428,
"grad_norm": 0.5518265976503514,
"learning_rate": 9.933488723741731e-06,
"loss": 0.4457,
"step": 369
},
{
"epoch": 0.444,
"grad_norm": 0.6409718617060514,
"learning_rate": 9.932349501831077e-06,
"loss": 0.4173,
"step": 370
},
{
"epoch": 0.4452,
"grad_norm": 0.7792630397126719,
"learning_rate": 9.931200672594425e-06,
"loss": 0.4454,
"step": 371
},
{
"epoch": 0.4464,
"grad_norm": 0.6344199240027713,
"learning_rate": 9.930042238269485e-06,
"loss": 0.4566,
"step": 372
},
{
"epoch": 0.4476,
"grad_norm": 0.6594867191224307,
"learning_rate": 9.928874201112677e-06,
"loss": 0.4321,
"step": 373
},
{
"epoch": 0.4488,
"grad_norm": 0.6791155653346912,
"learning_rate": 9.927696563399127e-06,
"loss": 0.4569,
"step": 374
},
{
"epoch": 0.45,
"grad_norm": 0.6886786049932618,
"learning_rate": 9.926509327422661e-06,
"loss": 0.4628,
"step": 375
},
{
"epoch": 0.4512,
"grad_norm": 0.6660330432227839,
"learning_rate": 9.9253124954958e-06,
"loss": 0.4295,
"step": 376
},
{
"epoch": 0.4524,
"grad_norm": 0.6814134075798867,
"learning_rate": 9.924106069949756e-06,
"loss": 0.4327,
"step": 377
},
{
"epoch": 0.4536,
"grad_norm": 0.5875291506684054,
"learning_rate": 9.922890053134428e-06,
"loss": 0.3928,
"step": 378
},
{
"epoch": 0.4548,
"grad_norm": 0.6922382672621923,
"learning_rate": 9.9216644474184e-06,
"loss": 0.4442,
"step": 379
},
{
"epoch": 0.456,
"grad_norm": 0.771249714605511,
"learning_rate": 9.920429255188926e-06,
"loss": 0.4714,
"step": 380
},
{
"epoch": 0.4572,
"grad_norm": 0.676431519188336,
"learning_rate": 9.91918447885194e-06,
"loss": 0.4332,
"step": 381
},
{
"epoch": 0.4584,
"grad_norm": 0.6945908407546308,
"learning_rate": 9.91793012083204e-06,
"loss": 0.4321,
"step": 382
},
{
"epoch": 0.4596,
"grad_norm": 0.7552120817204823,
"learning_rate": 9.916666183572492e-06,
"loss": 0.474,
"step": 383
},
{
"epoch": 0.4608,
"grad_norm": 0.6574840581158748,
"learning_rate": 9.915392669535214e-06,
"loss": 0.486,
"step": 384
},
{
"epoch": 0.462,
"grad_norm": 0.6484938793376102,
"learning_rate": 9.914109581200785e-06,
"loss": 0.4083,
"step": 385
},
{
"epoch": 0.4632,
"grad_norm": 0.6996866528066688,
"learning_rate": 9.912816921068424e-06,
"loss": 0.4566,
"step": 386
},
{
"epoch": 0.4644,
"grad_norm": 0.7430007400183946,
"learning_rate": 9.911514691656003e-06,
"loss": 0.4219,
"step": 387
},
{
"epoch": 0.4656,
"grad_norm": 0.6289704944260223,
"learning_rate": 9.910202895500031e-06,
"loss": 0.4275,
"step": 388
},
{
"epoch": 0.4668,
"grad_norm": 0.7107427907702816,
"learning_rate": 9.908881535155647e-06,
"loss": 0.4258,
"step": 389
},
{
"epoch": 0.468,
"grad_norm": 0.6733732649843523,
"learning_rate": 9.907550613196624e-06,
"loss": 0.4277,
"step": 390
},
{
"epoch": 0.4692,
"grad_norm": 0.7252156396982233,
"learning_rate": 9.906210132215357e-06,
"loss": 0.439,
"step": 391
},
{
"epoch": 0.4704,
"grad_norm": 0.7527036675475188,
"learning_rate": 9.904860094822861e-06,
"loss": 0.4545,
"step": 392
},
{
"epoch": 0.4716,
"grad_norm": 0.7048228902693239,
"learning_rate": 9.903500503648766e-06,
"loss": 0.4374,
"step": 393
},
{
"epoch": 0.4728,
"grad_norm": 0.5448140600277945,
"learning_rate": 9.902131361341307e-06,
"loss": 0.4112,
"step": 394
},
{
"epoch": 0.474,
"grad_norm": 0.8345035569254616,
"learning_rate": 9.900752670567331e-06,
"loss": 0.4775,
"step": 395
},
{
"epoch": 0.4752,
"grad_norm": 0.6443454053956338,
"learning_rate": 9.899364434012273e-06,
"loss": 0.4229,
"step": 396
},
{
"epoch": 0.4764,
"grad_norm": 0.6230088120223173,
"learning_rate": 9.897966654380172e-06,
"loss": 0.4615,
"step": 397
},
{
"epoch": 0.4776,
"grad_norm": 0.6088046653465384,
"learning_rate": 9.896559334393644e-06,
"loss": 0.4317,
"step": 398
},
{
"epoch": 0.4788,
"grad_norm": 0.7279533322927584,
"learning_rate": 9.895142476793902e-06,
"loss": 0.4514,
"step": 399
},
{
"epoch": 0.48,
"grad_norm": 0.7745953818595627,
"learning_rate": 9.893716084340723e-06,
"loss": 0.458,
"step": 400
},
{
"epoch": 0.4812,
"grad_norm": 0.6405817762556505,
"learning_rate": 9.892280159812465e-06,
"loss": 0.4883,
"step": 401
},
{
"epoch": 0.4824,
"grad_norm": 0.7685486452534169,
"learning_rate": 9.890834706006048e-06,
"loss": 0.4661,
"step": 402
},
{
"epoch": 0.4836,
"grad_norm": 0.7003098986825237,
"learning_rate": 9.889379725736953e-06,
"loss": 0.4714,
"step": 403
},
{
"epoch": 0.4848,
"grad_norm": 0.6254581564414184,
"learning_rate": 9.887915221839223e-06,
"loss": 0.4878,
"step": 404
},
{
"epoch": 0.486,
"grad_norm": 0.6311252500153688,
"learning_rate": 9.886441197165446e-06,
"loss": 0.4539,
"step": 405
},
{
"epoch": 0.4872,
"grad_norm": 0.7820317979670909,
"learning_rate": 9.884957654586753e-06,
"loss": 0.4518,
"step": 406
},
{
"epoch": 0.4884,
"grad_norm": 0.6004530584614425,
"learning_rate": 9.88346459699282e-06,
"loss": 0.4094,
"step": 407
},
{
"epoch": 0.4896,
"grad_norm": 0.6654376969151062,
"learning_rate": 9.881962027291855e-06,
"loss": 0.4348,
"step": 408
},
{
"epoch": 0.4908,
"grad_norm": 0.6679415371212997,
"learning_rate": 9.880449948410587e-06,
"loss": 0.4545,
"step": 409
},
{
"epoch": 0.492,
"grad_norm": 0.5547515670600671,
"learning_rate": 9.878928363294275e-06,
"loss": 0.4536,
"step": 410
},
{
"epoch": 0.4932,
"grad_norm": 0.7125897617236266,
"learning_rate": 9.877397274906694e-06,
"loss": 0.4473,
"step": 411
},
{
"epoch": 0.4944,
"grad_norm": 0.8192773426910452,
"learning_rate": 9.875856686230125e-06,
"loss": 0.4249,
"step": 412
},
{
"epoch": 0.4956,
"grad_norm": 0.6649787950976529,
"learning_rate": 9.87430660026536e-06,
"loss": 0.4264,
"step": 413
},
{
"epoch": 0.4968,
"grad_norm": 0.7064920470637832,
"learning_rate": 9.872747020031682e-06,
"loss": 0.4363,
"step": 414
},
{
"epoch": 0.498,
"grad_norm": 0.827612732425241,
"learning_rate": 9.871177948566875e-06,
"loss": 0.461,
"step": 415
},
{
"epoch": 0.4992,
"grad_norm": 0.5894682708442801,
"learning_rate": 9.869599388927204e-06,
"loss": 0.4355,
"step": 416
},
{
"epoch": 0.5004,
"grad_norm": 0.7652566876221197,
"learning_rate": 9.868011344187421e-06,
"loss": 0.4847,
"step": 417
},
{
"epoch": 0.5016,
"grad_norm": 0.651422164836644,
"learning_rate": 9.866413817440748e-06,
"loss": 0.4502,
"step": 418
},
{
"epoch": 0.5028,
"grad_norm": 0.5806540058251656,
"learning_rate": 9.864806811798881e-06,
"loss": 0.4249,
"step": 419
},
{
"epoch": 0.504,
"grad_norm": 0.674941136160933,
"learning_rate": 9.863190330391974e-06,
"loss": 0.4216,
"step": 420
},
{
"epoch": 0.5052,
"grad_norm": 0.680982850770216,
"learning_rate": 9.861564376368645e-06,
"loss": 0.4387,
"step": 421
},
{
"epoch": 0.5064,
"grad_norm": 0.6080701465778402,
"learning_rate": 9.859928952895952e-06,
"loss": 0.4404,
"step": 422
},
{
"epoch": 0.5076,
"grad_norm": 0.6543859823085602,
"learning_rate": 9.858284063159411e-06,
"loss": 0.4346,
"step": 423
},
{
"epoch": 0.5088,
"grad_norm": 0.7544442628769478,
"learning_rate": 9.856629710362966e-06,
"loss": 0.5297,
"step": 424
},
{
"epoch": 0.51,
"grad_norm": 0.6970683012667609,
"learning_rate": 9.854965897729001e-06,
"loss": 0.4346,
"step": 425
},
{
"epoch": 0.5112,
"grad_norm": 0.701699016606531,
"learning_rate": 9.853292628498319e-06,
"loss": 0.4832,
"step": 426
},
{
"epoch": 0.5124,
"grad_norm": 0.6531688259436925,
"learning_rate": 9.851609905930149e-06,
"loss": 0.4374,
"step": 427
},
{
"epoch": 0.5136,
"grad_norm": 0.6459875412165775,
"learning_rate": 9.849917733302128e-06,
"loss": 0.4342,
"step": 428
},
{
"epoch": 0.5148,
"grad_norm": 0.7824619238063619,
"learning_rate": 9.848216113910306e-06,
"loss": 0.4755,
"step": 429
},
{
"epoch": 0.516,
"grad_norm": 0.7068114804561866,
"learning_rate": 9.846505051069126e-06,
"loss": 0.4359,
"step": 430
},
{
"epoch": 0.5172,
"grad_norm": 0.620420764465856,
"learning_rate": 9.844784548111433e-06,
"loss": 0.4773,
"step": 431
},
{
"epoch": 0.5184,
"grad_norm": 0.7342607747683098,
"learning_rate": 9.843054608388455e-06,
"loss": 0.4684,
"step": 432
},
{
"epoch": 0.5196,
"grad_norm": 0.6222896944551798,
"learning_rate": 9.8413152352698e-06,
"loss": 0.4572,
"step": 433
},
{
"epoch": 0.5208,
"grad_norm": 0.6870261388789632,
"learning_rate": 9.839566432143459e-06,
"loss": 0.4877,
"step": 434
},
{
"epoch": 0.522,
"grad_norm": 0.708094366948388,
"learning_rate": 9.837808202415778e-06,
"loss": 0.4594,
"step": 435
},
{
"epoch": 0.5232,
"grad_norm": 0.7280627126935877,
"learning_rate": 9.836040549511472e-06,
"loss": 0.3944,
"step": 436
},
{
"epoch": 0.5244,
"grad_norm": 0.6534507741530826,
"learning_rate": 9.83426347687361e-06,
"loss": 0.4493,
"step": 437
},
{
"epoch": 0.5256,
"grad_norm": 0.6739264885892525,
"learning_rate": 9.832476987963613e-06,
"loss": 0.4679,
"step": 438
},
{
"epoch": 0.5268,
"grad_norm": 0.7234520229667384,
"learning_rate": 9.830681086261234e-06,
"loss": 0.4702,
"step": 439
},
{
"epoch": 0.528,
"grad_norm": 0.7828665277714815,
"learning_rate": 9.828875775264564e-06,
"loss": 0.4368,
"step": 440
},
{
"epoch": 0.5292,
"grad_norm": 0.6490715970064068,
"learning_rate": 9.827061058490027e-06,
"loss": 0.4551,
"step": 441
},
{
"epoch": 0.5304,
"grad_norm": 0.6224193032887139,
"learning_rate": 9.82523693947236e-06,
"loss": 0.4118,
"step": 442
},
{
"epoch": 0.5316,
"grad_norm": 0.7860434173772851,
"learning_rate": 9.82340342176462e-06,
"loss": 0.4743,
"step": 443
},
{
"epoch": 0.5328,
"grad_norm": 0.6563230618839756,
"learning_rate": 9.821560508938167e-06,
"loss": 0.4464,
"step": 444
},
{
"epoch": 0.534,
"grad_norm": 0.6357292241353918,
"learning_rate": 9.819708204582664e-06,
"loss": 0.4408,
"step": 445
},
{
"epoch": 0.5352,
"grad_norm": 0.6339997044952623,
"learning_rate": 9.817846512306062e-06,
"loss": 0.4349,
"step": 446
},
{
"epoch": 0.5364,
"grad_norm": 0.6905877592322843,
"learning_rate": 9.815975435734604e-06,
"loss": 0.4519,
"step": 447
},
{
"epoch": 0.5376,
"grad_norm": 0.6436677936539605,
"learning_rate": 9.814094978512808e-06,
"loss": 0.4515,
"step": 448
},
{
"epoch": 0.5388,
"grad_norm": 0.736711209053276,
"learning_rate": 9.812205144303466e-06,
"loss": 0.4405,
"step": 449
},
{
"epoch": 0.54,
"grad_norm": 0.570921962141912,
"learning_rate": 9.810305936787633e-06,
"loss": 0.431,
"step": 450
},
{
"epoch": 0.5412,
"grad_norm": 0.7409880194062697,
"learning_rate": 9.808397359664624e-06,
"loss": 0.4526,
"step": 451
},
{
"epoch": 0.5424,
"grad_norm": 0.7175975384895951,
"learning_rate": 9.806479416652e-06,
"loss": 0.485,
"step": 452
},
{
"epoch": 0.5436,
"grad_norm": 0.6253351575776439,
"learning_rate": 9.804552111485568e-06,
"loss": 0.431,
"step": 453
},
{
"epoch": 0.5448,
"grad_norm": 0.6802528057657675,
"learning_rate": 9.80261544791937e-06,
"loss": 0.49,
"step": 454
},
{
"epoch": 0.546,
"grad_norm": 0.6486533789851607,
"learning_rate": 9.80066942972568e-06,
"loss": 0.4338,
"step": 455
},
{
"epoch": 0.5472,
"grad_norm": 0.7020644490479623,
"learning_rate": 9.798714060694988e-06,
"loss": 0.4304,
"step": 456
},
{
"epoch": 0.5484,
"grad_norm": 0.6211547222543867,
"learning_rate": 9.796749344635996e-06,
"loss": 0.4478,
"step": 457
},
{
"epoch": 0.5496,
"grad_norm": 0.7757775131158096,
"learning_rate": 9.794775285375623e-06,
"loss": 0.5034,
"step": 458
},
{
"epoch": 0.5508,
"grad_norm": 0.6482029929415141,
"learning_rate": 9.792791886758976e-06,
"loss": 0.4419,
"step": 459
},
{
"epoch": 0.552,
"grad_norm": 0.6538846576233442,
"learning_rate": 9.790799152649356e-06,
"loss": 0.4192,
"step": 460
},
{
"epoch": 0.5532,
"grad_norm": 0.6824916183880225,
"learning_rate": 9.788797086928252e-06,
"loss": 0.434,
"step": 461
},
{
"epoch": 0.5544,
"grad_norm": 0.5985857501737958,
"learning_rate": 9.786785693495327e-06,
"loss": 0.4472,
"step": 462
},
{
"epoch": 0.5556,
"grad_norm": 0.5360509289546986,
"learning_rate": 9.784764976268408e-06,
"loss": 0.4622,
"step": 463
},
{
"epoch": 0.5568,
"grad_norm": 0.5953489959775399,
"learning_rate": 9.78273493918349e-06,
"loss": 0.4202,
"step": 464
},
{
"epoch": 0.558,
"grad_norm": 0.6533019600801288,
"learning_rate": 9.780695586194719e-06,
"loss": 0.4935,
"step": 465
},
{
"epoch": 0.5592,
"grad_norm": 0.6660863843497694,
"learning_rate": 9.778646921274385e-06,
"loss": 0.4377,
"step": 466
},
{
"epoch": 0.5604,
"grad_norm": 0.6203187626727608,
"learning_rate": 9.776588948412917e-06,
"loss": 0.4375,
"step": 467
},
{
"epoch": 0.5616,
"grad_norm": 0.6898987375410509,
"learning_rate": 9.774521671618877e-06,
"loss": 0.4353,
"step": 468
},
{
"epoch": 0.5628,
"grad_norm": 0.6234056463809782,
"learning_rate": 9.772445094918944e-06,
"loss": 0.4783,
"step": 469
},
{
"epoch": 0.564,
"grad_norm": 0.49658034146463864,
"learning_rate": 9.770359222357914e-06,
"loss": 0.3945,
"step": 470
},
{
"epoch": 0.5652,
"grad_norm": 0.533972255693177,
"learning_rate": 9.768264057998693e-06,
"loss": 0.4071,
"step": 471
},
{
"epoch": 0.5664,
"grad_norm": 0.6337264376454038,
"learning_rate": 9.766159605922282e-06,
"loss": 0.4484,
"step": 472
},
{
"epoch": 0.5676,
"grad_norm": 0.6491123411466448,
"learning_rate": 9.764045870227772e-06,
"loss": 0.4221,
"step": 473
},
{
"epoch": 0.5688,
"grad_norm": 0.6154543475969276,
"learning_rate": 9.761922855032339e-06,
"loss": 0.4246,
"step": 474
},
{
"epoch": 0.57,
"grad_norm": 0.7818066340168961,
"learning_rate": 9.759790564471233e-06,
"loss": 0.439,
"step": 475
},
{
"epoch": 0.5712,
"grad_norm": 0.6369517648681066,
"learning_rate": 9.757649002697771e-06,
"loss": 0.4417,
"step": 476
},
{
"epoch": 0.5724,
"grad_norm": 0.6329140386406038,
"learning_rate": 9.755498173883331e-06,
"loss": 0.4695,
"step": 477
},
{
"epoch": 0.5736,
"grad_norm": 0.650969631762846,
"learning_rate": 9.753338082217334e-06,
"loss": 0.4413,
"step": 478
},
{
"epoch": 0.5748,
"grad_norm": 0.6647141567277757,
"learning_rate": 9.751168731907253e-06,
"loss": 0.4302,
"step": 479
},
{
"epoch": 0.576,
"grad_norm": 0.5711206546908786,
"learning_rate": 9.748990127178589e-06,
"loss": 0.4357,
"step": 480
},
{
"epoch": 0.5772,
"grad_norm": 0.7567029621367047,
"learning_rate": 9.746802272274868e-06,
"loss": 0.4505,
"step": 481
},
{
"epoch": 0.5784,
"grad_norm": 0.7223332864810444,
"learning_rate": 9.74460517145764e-06,
"loss": 0.5013,
"step": 482
},
{
"epoch": 0.5796,
"grad_norm": 0.7025765949826398,
"learning_rate": 9.74239882900646e-06,
"loss": 0.4562,
"step": 483
},
{
"epoch": 0.5808,
"grad_norm": 0.599300464196571,
"learning_rate": 9.740183249218883e-06,
"loss": 0.454,
"step": 484
},
{
"epoch": 0.582,
"grad_norm": 0.5782068939359957,
"learning_rate": 9.737958436410459e-06,
"loss": 0.4423,
"step": 485
},
{
"epoch": 0.5832,
"grad_norm": 0.6703880092235986,
"learning_rate": 9.73572439491472e-06,
"loss": 0.4431,
"step": 486
},
{
"epoch": 0.5844,
"grad_norm": 0.6345048946696713,
"learning_rate": 9.73348112908318e-06,
"loss": 0.4652,
"step": 487
},
{
"epoch": 0.5856,
"grad_norm": 0.6233074474085821,
"learning_rate": 9.73122864328531e-06,
"loss": 0.4039,
"step": 488
},
{
"epoch": 0.5868,
"grad_norm": 0.7197278707098507,
"learning_rate": 9.72896694190855e-06,
"loss": 0.4603,
"step": 489
},
{
"epoch": 0.588,
"grad_norm": 0.8162957949734152,
"learning_rate": 9.726696029358283e-06,
"loss": 0.4793,
"step": 490
},
{
"epoch": 0.5892,
"grad_norm": 0.601510162216323,
"learning_rate": 9.724415910057839e-06,
"loss": 0.4162,
"step": 491
},
{
"epoch": 0.5904,
"grad_norm": 0.5806154181617217,
"learning_rate": 9.722126588448473e-06,
"loss": 0.3978,
"step": 492
},
{
"epoch": 0.5916,
"grad_norm": 0.8352938422444114,
"learning_rate": 9.719828068989378e-06,
"loss": 0.4814,
"step": 493
},
{
"epoch": 0.5928,
"grad_norm": 0.6607407237224796,
"learning_rate": 9.717520356157648e-06,
"loss": 0.4352,
"step": 494
},
{
"epoch": 0.594,
"grad_norm": 0.6769495331597323,
"learning_rate": 9.715203454448297e-06,
"loss": 0.4888,
"step": 495
},
{
"epoch": 0.5952,
"grad_norm": 0.582792575005291,
"learning_rate": 9.712877368374226e-06,
"loss": 0.4169,
"step": 496
},
{
"epoch": 0.5964,
"grad_norm": 0.6969663052525739,
"learning_rate": 9.710542102466229e-06,
"loss": 0.427,
"step": 497
},
{
"epoch": 0.5976,
"grad_norm": 0.6940068268557692,
"learning_rate": 9.708197661272989e-06,
"loss": 0.4432,
"step": 498
},
{
"epoch": 0.5988,
"grad_norm": 0.7417084803159525,
"learning_rate": 9.70584404936105e-06,
"loss": 0.4439,
"step": 499
},
{
"epoch": 0.6,
"grad_norm": 0.609730769080397,
"learning_rate": 9.703481271314823e-06,
"loss": 0.4365,
"step": 500
},
{
"epoch": 0.6012,
"grad_norm": 0.777232315860157,
"learning_rate": 9.701109331736573e-06,
"loss": 0.4083,
"step": 501
},
{
"epoch": 0.6024,
"grad_norm": 0.7618896885321611,
"learning_rate": 9.69872823524641e-06,
"loss": 0.4805,
"step": 502
},
{
"epoch": 0.6036,
"grad_norm": 0.5887187702529323,
"learning_rate": 9.69633798648228e-06,
"loss": 0.4099,
"step": 503
},
{
"epoch": 0.6048,
"grad_norm": 0.6133967037667464,
"learning_rate": 9.693938590099958e-06,
"loss": 0.4899,
"step": 504
},
{
"epoch": 0.606,
"grad_norm": 0.7578058357125181,
"learning_rate": 9.691530050773031e-06,
"loss": 0.4358,
"step": 505
},
{
"epoch": 0.6072,
"grad_norm": 0.6773190619603114,
"learning_rate": 9.689112373192899e-06,
"loss": 0.4971,
"step": 506
},
{
"epoch": 0.6084,
"grad_norm": 0.5854080621024346,
"learning_rate": 9.686685562068765e-06,
"loss": 0.4166,
"step": 507
},
{
"epoch": 0.6096,
"grad_norm": 0.6400199675000283,
"learning_rate": 9.684249622127616e-06,
"loss": 0.4471,
"step": 508
},
{
"epoch": 0.6108,
"grad_norm": 0.7258315288657347,
"learning_rate": 9.681804558114222e-06,
"loss": 0.435,
"step": 509
},
{
"epoch": 0.612,
"grad_norm": 0.6272130773016463,
"learning_rate": 9.67935037479113e-06,
"loss": 0.4332,
"step": 510
},
{
"epoch": 0.6132,
"grad_norm": 0.7048388710386582,
"learning_rate": 9.676887076938642e-06,
"loss": 0.4581,
"step": 511
},
{
"epoch": 0.6144,
"grad_norm": 0.641338293198559,
"learning_rate": 9.674414669354819e-06,
"loss": 0.4486,
"step": 512
},
{
"epoch": 0.6156,
"grad_norm": 0.6312876175673993,
"learning_rate": 9.671933156855464e-06,
"loss": 0.439,
"step": 513
},
{
"epoch": 0.6168,
"grad_norm": 0.611370421206214,
"learning_rate": 9.669442544274115e-06,
"loss": 0.411,
"step": 514
},
{
"epoch": 0.618,
"grad_norm": 0.6083285462122053,
"learning_rate": 9.666942836462036e-06,
"loss": 0.4717,
"step": 515
},
{
"epoch": 0.6192,
"grad_norm": 0.6095089710667009,
"learning_rate": 9.664434038288207e-06,
"loss": 0.4442,
"step": 516
},
{
"epoch": 0.6204,
"grad_norm": 0.5498795949636963,
"learning_rate": 9.661916154639312e-06,
"loss": 0.4446,
"step": 517
},
{
"epoch": 0.6216,
"grad_norm": 0.5672758183613847,
"learning_rate": 9.659389190419735e-06,
"loss": 0.4346,
"step": 518
},
{
"epoch": 0.6228,
"grad_norm": 0.5420075396704381,
"learning_rate": 9.656853150551543e-06,
"loss": 0.4343,
"step": 519
},
{
"epoch": 0.624,
"grad_norm": 0.6174137897150176,
"learning_rate": 9.654308039974489e-06,
"loss": 0.4635,
"step": 520
},
{
"epoch": 0.6252,
"grad_norm": 0.6180765800998059,
"learning_rate": 9.651753863645985e-06,
"loss": 0.4549,
"step": 521
},
{
"epoch": 0.6264,
"grad_norm": 0.5533027840877576,
"learning_rate": 9.649190626541105e-06,
"loss": 0.4173,
"step": 522
},
{
"epoch": 0.6276,
"grad_norm": 0.6897757672749916,
"learning_rate": 9.646618333652574e-06,
"loss": 0.4569,
"step": 523
},
{
"epoch": 0.6288,
"grad_norm": 0.6526298629600158,
"learning_rate": 9.644036989990753e-06,
"loss": 0.4641,
"step": 524
},
{
"epoch": 0.63,
"grad_norm": 0.6251092549491414,
"learning_rate": 9.641446600583632e-06,
"loss": 0.4122,
"step": 525
},
{
"epoch": 0.6312,
"grad_norm": 0.7292256985425808,
"learning_rate": 9.638847170476824e-06,
"loss": 0.4506,
"step": 526
},
{
"epoch": 0.6324,
"grad_norm": 0.7109600305730553,
"learning_rate": 9.636238704733547e-06,
"loss": 0.4456,
"step": 527
},
{
"epoch": 0.6336,
"grad_norm": 0.6486870015904621,
"learning_rate": 9.633621208434623e-06,
"loss": 0.4099,
"step": 528
},
{
"epoch": 0.6348,
"grad_norm": 0.7260733786982925,
"learning_rate": 9.630994686678462e-06,
"loss": 0.4787,
"step": 529
},
{
"epoch": 0.636,
"grad_norm": 0.6972415633425663,
"learning_rate": 9.628359144581052e-06,
"loss": 0.4376,
"step": 530
},
{
"epoch": 0.6372,
"grad_norm": 0.6575368501965961,
"learning_rate": 9.625714587275954e-06,
"loss": 0.4236,
"step": 531
},
{
"epoch": 0.6384,
"grad_norm": 0.70852769192471,
"learning_rate": 9.623061019914291e-06,
"loss": 0.4051,
"step": 532
},
{
"epoch": 0.6396,
"grad_norm": 0.5920331266680952,
"learning_rate": 9.620398447664727e-06,
"loss": 0.4555,
"step": 533
},
{
"epoch": 0.6408,
"grad_norm": 0.6512018747591982,
"learning_rate": 9.617726875713477e-06,
"loss": 0.4562,
"step": 534
},
{
"epoch": 0.642,
"grad_norm": 0.7030860335919872,
"learning_rate": 9.615046309264278e-06,
"loss": 0.4325,
"step": 535
},
{
"epoch": 0.6432,
"grad_norm": 0.6451559251581166,
"learning_rate": 9.612356753538392e-06,
"loss": 0.4534,
"step": 536
},
{
"epoch": 0.6444,
"grad_norm": 0.6555638840456719,
"learning_rate": 9.609658213774584e-06,
"loss": 0.4506,
"step": 537
},
{
"epoch": 0.6456,
"grad_norm": 0.7724431049478658,
"learning_rate": 9.606950695229125e-06,
"loss": 0.4332,
"step": 538
},
{
"epoch": 0.6468,
"grad_norm": 0.6458233985243111,
"learning_rate": 9.60423420317577e-06,
"loss": 0.4548,
"step": 539
},
{
"epoch": 0.648,
"grad_norm": 0.7365226748584056,
"learning_rate": 9.601508742905757e-06,
"loss": 0.4159,
"step": 540
},
{
"epoch": 0.6492,
"grad_norm": 0.6345441906843313,
"learning_rate": 9.59877431972779e-06,
"loss": 0.4575,
"step": 541
},
{
"epoch": 0.6504,
"grad_norm": 0.6608103527291914,
"learning_rate": 9.596030938968028e-06,
"loss": 0.4494,
"step": 542
},
{
"epoch": 0.6516,
"grad_norm": 0.6815011264185148,
"learning_rate": 9.593278605970086e-06,
"loss": 0.4243,
"step": 543
},
{
"epoch": 0.6528,
"grad_norm": 0.6647489003833473,
"learning_rate": 9.590517326095012e-06,
"loss": 0.4215,
"step": 544
},
{
"epoch": 0.654,
"grad_norm": 0.5306957908700147,
"learning_rate": 9.587747104721275e-06,
"loss": 0.4095,
"step": 545
},
{
"epoch": 0.6552,
"grad_norm": 0.623430114444001,
"learning_rate": 9.58496794724477e-06,
"loss": 0.4202,
"step": 546
},
{
"epoch": 0.6564,
"grad_norm": 0.707639118232904,
"learning_rate": 9.582179859078793e-06,
"loss": 0.4321,
"step": 547
},
{
"epoch": 0.6576,
"grad_norm": 0.6187571241832408,
"learning_rate": 9.579382845654038e-06,
"loss": 0.445,
"step": 548
},
{
"epoch": 0.6588,
"grad_norm": 0.5968971585377222,
"learning_rate": 9.576576912418577e-06,
"loss": 0.4513,
"step": 549
},
{
"epoch": 0.66,
"grad_norm": 0.6060109622440814,
"learning_rate": 9.573762064837866e-06,
"loss": 0.4371,
"step": 550
},
{
"epoch": 0.6612,
"grad_norm": 0.64789777702885,
"learning_rate": 9.570938308394717e-06,
"loss": 0.4028,
"step": 551
},
{
"epoch": 0.6624,
"grad_norm": 0.6958071032383644,
"learning_rate": 9.568105648589299e-06,
"loss": 0.4675,
"step": 552
},
{
"epoch": 0.6636,
"grad_norm": 0.5822451096374223,
"learning_rate": 9.56526409093912e-06,
"loss": 0.4172,
"step": 553
},
{
"epoch": 0.6648,
"grad_norm": 0.7601888523954846,
"learning_rate": 9.562413640979024e-06,
"loss": 0.4297,
"step": 554
},
{
"epoch": 0.666,
"grad_norm": 0.6567049323975153,
"learning_rate": 9.55955430426117e-06,
"loss": 0.4847,
"step": 555
},
{
"epoch": 0.6672,
"grad_norm": 0.7131405370806122,
"learning_rate": 9.556686086355032e-06,
"loss": 0.4297,
"step": 556
},
{
"epoch": 0.6684,
"grad_norm": 0.5589981609294136,
"learning_rate": 9.553808992847377e-06,
"loss": 0.4024,
"step": 557
},
{
"epoch": 0.6696,
"grad_norm": 0.5931996857224913,
"learning_rate": 9.550923029342266e-06,
"loss": 0.4292,
"step": 558
},
{
"epoch": 0.6708,
"grad_norm": 0.5971466573228367,
"learning_rate": 9.548028201461034e-06,
"loss": 0.4534,
"step": 559
},
{
"epoch": 0.672,
"grad_norm": 0.6272928947603887,
"learning_rate": 9.545124514842284e-06,
"loss": 0.4836,
"step": 560
},
{
"epoch": 0.6732,
"grad_norm": 0.5489911002674158,
"learning_rate": 9.542211975141871e-06,
"loss": 0.4315,
"step": 561
},
{
"epoch": 0.6744,
"grad_norm": 0.6087915426456199,
"learning_rate": 9.5392905880329e-06,
"loss": 0.4049,
"step": 562
},
{
"epoch": 0.6756,
"grad_norm": 0.7304570976047295,
"learning_rate": 9.5363603592057e-06,
"loss": 0.472,
"step": 563
},
{
"epoch": 0.6768,
"grad_norm": 0.5930708966353481,
"learning_rate": 9.53342129436783e-06,
"loss": 0.4363,
"step": 564
},
{
"epoch": 0.678,
"grad_norm": 0.7121202398856948,
"learning_rate": 9.530473399244061e-06,
"loss": 0.4731,
"step": 565
},
{
"epoch": 0.6792,
"grad_norm": 0.6019249811320259,
"learning_rate": 9.527516679576353e-06,
"loss": 0.4384,
"step": 566
},
{
"epoch": 0.6804,
"grad_norm": 0.6200731562962547,
"learning_rate": 9.52455114112387e-06,
"loss": 0.4585,
"step": 567
},
{
"epoch": 0.6816,
"grad_norm": 0.6093242073566054,
"learning_rate": 9.52157678966294e-06,
"loss": 0.3909,
"step": 568
},
{
"epoch": 0.6828,
"grad_norm": 0.5730476216253085,
"learning_rate": 9.518593630987063e-06,
"loss": 0.3917,
"step": 569
},
{
"epoch": 0.684,
"grad_norm": 0.631650579006344,
"learning_rate": 9.515601670906895e-06,
"loss": 0.4188,
"step": 570
},
{
"epoch": 0.6852,
"grad_norm": 0.5969798975988179,
"learning_rate": 9.512600915250232e-06,
"loss": 0.4429,
"step": 571
},
{
"epoch": 0.6864,
"grad_norm": 0.7154797074011033,
"learning_rate": 9.509591369862007e-06,
"loss": 0.4747,
"step": 572
},
{
"epoch": 0.6876,
"grad_norm": 0.631301195463087,
"learning_rate": 9.506573040604268e-06,
"loss": 0.4594,
"step": 573
},
{
"epoch": 0.6888,
"grad_norm": 0.6430321252881295,
"learning_rate": 9.503545933356175e-06,
"loss": 0.413,
"step": 574
},
{
"epoch": 0.69,
"grad_norm": 0.7152067265164149,
"learning_rate": 9.500510054013989e-06,
"loss": 0.4309,
"step": 575
},
{
"epoch": 0.6912,
"grad_norm": 0.7111876494885973,
"learning_rate": 9.49746540849105e-06,
"loss": 0.4497,
"step": 576
},
{
"epoch": 0.6924,
"grad_norm": 0.6212907351663844,
"learning_rate": 9.494412002717784e-06,
"loss": 0.4646,
"step": 577
},
{
"epoch": 0.6936,
"grad_norm": 0.730034488610609,
"learning_rate": 9.49134984264167e-06,
"loss": 0.4624,
"step": 578
},
{
"epoch": 0.6948,
"grad_norm": 0.5959601487634612,
"learning_rate": 9.488278934227242e-06,
"loss": 0.4428,
"step": 579
},
{
"epoch": 0.696,
"grad_norm": 0.7697382736559194,
"learning_rate": 9.485199283456078e-06,
"loss": 0.4148,
"step": 580
},
{
"epoch": 0.6972,
"grad_norm": 0.6914659012199768,
"learning_rate": 9.48211089632678e-06,
"loss": 0.4493,
"step": 581
},
{
"epoch": 0.6984,
"grad_norm": 0.6626411891434971,
"learning_rate": 9.479013778854966e-06,
"loss": 0.4443,
"step": 582
},
{
"epoch": 0.6996,
"grad_norm": 0.633938170597033,
"learning_rate": 9.475907937073265e-06,
"loss": 0.3769,
"step": 583
},
{
"epoch": 0.7008,
"grad_norm": 0.7858253823595117,
"learning_rate": 9.472793377031293e-06,
"loss": 0.4217,
"step": 584
},
{
"epoch": 0.702,
"grad_norm": 0.6575698835345714,
"learning_rate": 9.469670104795655e-06,
"loss": 0.3947,
"step": 585
},
{
"epoch": 0.7032,
"grad_norm": 0.6759835247423801,
"learning_rate": 9.466538126449915e-06,
"loss": 0.4487,
"step": 586
},
{
"epoch": 0.7044,
"grad_norm": 0.7224436257813325,
"learning_rate": 9.463397448094605e-06,
"loss": 0.4517,
"step": 587
},
{
"epoch": 0.7056,
"grad_norm": 0.7738499274833498,
"learning_rate": 9.460248075847199e-06,
"loss": 0.4717,
"step": 588
},
{
"epoch": 0.7068,
"grad_norm": 0.6335072145475474,
"learning_rate": 9.457090015842104e-06,
"loss": 0.4718,
"step": 589
},
{
"epoch": 0.708,
"grad_norm": 0.5694752122818949,
"learning_rate": 9.453923274230653e-06,
"loss": 0.4133,
"step": 590
},
{
"epoch": 0.7092,
"grad_norm": 0.6712010029402053,
"learning_rate": 9.450747857181084e-06,
"loss": 0.4704,
"step": 591
},
{
"epoch": 0.7104,
"grad_norm": 0.5758341569803189,
"learning_rate": 9.447563770878535e-06,
"loss": 0.4155,
"step": 592
},
{
"epoch": 0.7116,
"grad_norm": 0.5917081772026239,
"learning_rate": 9.444371021525036e-06,
"loss": 0.4439,
"step": 593
},
{
"epoch": 0.7128,
"grad_norm": 0.7370046552121895,
"learning_rate": 9.441169615339482e-06,
"loss": 0.4448,
"step": 594
},
{
"epoch": 0.714,
"grad_norm": 0.5867359493833982,
"learning_rate": 9.437959558557635e-06,
"loss": 0.4492,
"step": 595
},
{
"epoch": 0.7152,
"grad_norm": 0.5916408949328779,
"learning_rate": 9.434740857432105e-06,
"loss": 0.4659,
"step": 596
},
{
"epoch": 0.7164,
"grad_norm": 0.6755568158069253,
"learning_rate": 9.431513518232343e-06,
"loss": 0.4589,
"step": 597
},
{
"epoch": 0.7176,
"grad_norm": 0.6870614012337154,
"learning_rate": 9.42827754724462e-06,
"loss": 0.429,
"step": 598
},
{
"epoch": 0.7188,
"grad_norm": 0.5993193342124471,
"learning_rate": 9.425032950772025e-06,
"loss": 0.422,
"step": 599
},
{
"epoch": 0.72,
"grad_norm": 0.7078484042302691,
"learning_rate": 9.421779735134446e-06,
"loss": 0.4191,
"step": 600
},
{
"epoch": 0.7212,
"grad_norm": 0.599792674357655,
"learning_rate": 9.418517906668556e-06,
"loss": 0.4043,
"step": 601
},
{
"epoch": 0.7224,
"grad_norm": 0.6845611310806173,
"learning_rate": 9.415247471727813e-06,
"loss": 0.4655,
"step": 602
},
{
"epoch": 0.7236,
"grad_norm": 0.6932010017308168,
"learning_rate": 9.41196843668243e-06,
"loss": 0.4682,
"step": 603
},
{
"epoch": 0.7248,
"grad_norm": 0.5990590128220633,
"learning_rate": 9.408680807919377e-06,
"loss": 0.4526,
"step": 604
},
{
"epoch": 0.726,
"grad_norm": 0.5695053632896584,
"learning_rate": 9.405384591842358e-06,
"loss": 0.4438,
"step": 605
},
{
"epoch": 0.7272,
"grad_norm": 0.615343410887285,
"learning_rate": 9.402079794871812e-06,
"loss": 0.3965,
"step": 606
},
{
"epoch": 0.7284,
"grad_norm": 0.6294619092300135,
"learning_rate": 9.398766423444883e-06,
"loss": 0.4019,
"step": 607
},
{
"epoch": 0.7296,
"grad_norm": 0.6684737765394154,
"learning_rate": 9.39544448401542e-06,
"loss": 0.4529,
"step": 608
},
{
"epoch": 0.7308,
"grad_norm": 0.6732998913927993,
"learning_rate": 9.392113983053958e-06,
"loss": 0.4552,
"step": 609
},
{
"epoch": 0.732,
"grad_norm": 0.6195247143898496,
"learning_rate": 9.38877492704772e-06,
"loss": 0.4395,
"step": 610
},
{
"epoch": 0.7332,
"grad_norm": 0.5872506201056751,
"learning_rate": 9.385427322500575e-06,
"loss": 0.4559,
"step": 611
},
{
"epoch": 0.7344,
"grad_norm": 0.5536526539474482,
"learning_rate": 9.382071175933058e-06,
"loss": 0.4194,
"step": 612
},
{
"epoch": 0.7356,
"grad_norm": 0.7582072512128472,
"learning_rate": 9.378706493882335e-06,
"loss": 0.4653,
"step": 613
},
{
"epoch": 0.7368,
"grad_norm": 0.6614836961213285,
"learning_rate": 9.375333282902198e-06,
"loss": 0.3983,
"step": 614
},
{
"epoch": 0.738,
"grad_norm": 0.5175199737865152,
"learning_rate": 9.37195154956305e-06,
"loss": 0.4068,
"step": 615
},
{
"epoch": 0.7392,
"grad_norm": 0.5792350511810672,
"learning_rate": 9.368561300451902e-06,
"loss": 0.4449,
"step": 616
},
{
"epoch": 0.7404,
"grad_norm": 0.7508510167859355,
"learning_rate": 9.365162542172346e-06,
"loss": 0.4546,
"step": 617
},
{
"epoch": 0.7416,
"grad_norm": 0.614188583793761,
"learning_rate": 9.361755281344547e-06,
"loss": 0.4308,
"step": 618
},
{
"epoch": 0.7428,
"grad_norm": 0.5853450714355288,
"learning_rate": 9.358339524605233e-06,
"loss": 0.4072,
"step": 619
},
{
"epoch": 0.744,
"grad_norm": 0.6606079599644,
"learning_rate": 9.354915278607685e-06,
"loss": 0.4253,
"step": 620
},
{
"epoch": 0.7452,
"grad_norm": 0.5495365865735293,
"learning_rate": 9.351482550021713e-06,
"loss": 0.3939,
"step": 621
},
{
"epoch": 0.7464,
"grad_norm": 0.5874360289468493,
"learning_rate": 9.348041345533653e-06,
"loss": 0.3773,
"step": 622
},
{
"epoch": 0.7476,
"grad_norm": 0.5717129934949987,
"learning_rate": 9.34459167184635e-06,
"loss": 0.4145,
"step": 623
},
{
"epoch": 0.7488,
"grad_norm": 0.6136224031869074,
"learning_rate": 9.341133535679145e-06,
"loss": 0.4236,
"step": 624
},
{
"epoch": 0.75,
"grad_norm": 0.6951415308062633,
"learning_rate": 9.337666943767863e-06,
"loss": 0.4306,
"step": 625
},
{
"epoch": 0.7512,
"grad_norm": 0.6009861219564681,
"learning_rate": 9.334191902864799e-06,
"loss": 0.4254,
"step": 626
},
{
"epoch": 0.7524,
"grad_norm": 0.6379389773803311,
"learning_rate": 9.330708419738704e-06,
"loss": 0.4147,
"step": 627
},
{
"epoch": 0.7536,
"grad_norm": 0.6669128113033436,
"learning_rate": 9.327216501174775e-06,
"loss": 0.3833,
"step": 628
},
{
"epoch": 0.7548,
"grad_norm": 0.7104418292195664,
"learning_rate": 9.323716153974639e-06,
"loss": 0.4726,
"step": 629
},
{
"epoch": 0.756,
"grad_norm": 0.654440647321272,
"learning_rate": 9.320207384956339e-06,
"loss": 0.4296,
"step": 630
},
{
"epoch": 0.7572,
"grad_norm": 0.6619054881639291,
"learning_rate": 9.316690200954324e-06,
"loss": 0.4188,
"step": 631
},
{
"epoch": 0.7584,
"grad_norm": 0.6841333972952182,
"learning_rate": 9.313164608819434e-06,
"loss": 0.4478,
"step": 632
},
{
"epoch": 0.7596,
"grad_norm": 0.5710359981446784,
"learning_rate": 9.309630615418884e-06,
"loss": 0.4287,
"step": 633
},
{
"epoch": 0.7608,
"grad_norm": 0.6989870896838141,
"learning_rate": 9.306088227636257e-06,
"loss": 0.4468,
"step": 634
},
{
"epoch": 0.762,
"grad_norm": 0.6612144448025201,
"learning_rate": 9.302537452371482e-06,
"loss": 0.425,
"step": 635
},
{
"epoch": 0.7632,
"grad_norm": 0.6969697409579375,
"learning_rate": 9.298978296540829e-06,
"loss": 0.4517,
"step": 636
},
{
"epoch": 0.7644,
"grad_norm": 0.7919369251237935,
"learning_rate": 9.295410767076891e-06,
"loss": 0.4607,
"step": 637
},
{
"epoch": 0.7656,
"grad_norm": 0.5458427570437363,
"learning_rate": 9.291834870928573e-06,
"loss": 0.4031,
"step": 638
},
{
"epoch": 0.7668,
"grad_norm": 0.6381806148795917,
"learning_rate": 9.288250615061073e-06,
"loss": 0.4361,
"step": 639
},
{
"epoch": 0.768,
"grad_norm": 0.5567043441199502,
"learning_rate": 9.284658006455871e-06,
"loss": 0.4447,
"step": 640
},
{
"epoch": 0.7692,
"grad_norm": 0.6820050719757548,
"learning_rate": 9.281057052110725e-06,
"loss": 0.4319,
"step": 641
},
{
"epoch": 0.7704,
"grad_norm": 0.6191273927150611,
"learning_rate": 9.27744775903964e-06,
"loss": 0.429,
"step": 642
},
{
"epoch": 0.7716,
"grad_norm": 0.6025505058454256,
"learning_rate": 9.27383013427287e-06,
"loss": 0.4405,
"step": 643
},
{
"epoch": 0.7728,
"grad_norm": 0.5645308546428628,
"learning_rate": 9.270204184856893e-06,
"loss": 0.4113,
"step": 644
},
{
"epoch": 0.774,
"grad_norm": 0.6987379901636854,
"learning_rate": 9.266569917854403e-06,
"loss": 0.4446,
"step": 645
},
{
"epoch": 0.7752,
"grad_norm": 0.567658261872694,
"learning_rate": 9.262927340344296e-06,
"loss": 0.4271,
"step": 646
},
{
"epoch": 0.7764,
"grad_norm": 0.6148385805143911,
"learning_rate": 9.259276459421655e-06,
"loss": 0.4583,
"step": 647
},
{
"epoch": 0.7776,
"grad_norm": 0.5384768224567317,
"learning_rate": 9.255617282197739e-06,
"loss": 0.4199,
"step": 648
},
{
"epoch": 0.7788,
"grad_norm": 0.5664509797096036,
"learning_rate": 9.25194981579996e-06,
"loss": 0.4072,
"step": 649
},
{
"epoch": 0.78,
"grad_norm": 0.6345199889121595,
"learning_rate": 9.248274067371886e-06,
"loss": 0.4061,
"step": 650
},
{
"epoch": 0.7812,
"grad_norm": 0.668319016739173,
"learning_rate": 9.244590044073205e-06,
"loss": 0.4665,
"step": 651
},
{
"epoch": 0.7824,
"grad_norm": 0.617592082191704,
"learning_rate": 9.240897753079734e-06,
"loss": 0.4726,
"step": 652
},
{
"epoch": 0.7836,
"grad_norm": 0.528719226788771,
"learning_rate": 9.237197201583386e-06,
"loss": 0.437,
"step": 653
},
{
"epoch": 0.7848,
"grad_norm": 0.5521191630194192,
"learning_rate": 9.233488396792167e-06,
"loss": 0.4632,
"step": 654
},
{
"epoch": 0.786,
"grad_norm": 0.5494094848077772,
"learning_rate": 9.22977134593016e-06,
"loss": 0.4233,
"step": 655
},
{
"epoch": 0.7872,
"grad_norm": 0.6498032978943626,
"learning_rate": 9.226046056237508e-06,
"loss": 0.4354,
"step": 656
},
{
"epoch": 0.7884,
"grad_norm": 0.565056383687927,
"learning_rate": 9.222312534970403e-06,
"loss": 0.4424,
"step": 657
},
{
"epoch": 0.7896,
"grad_norm": 0.6451984327966027,
"learning_rate": 9.218570789401071e-06,
"loss": 0.4216,
"step": 658
},
{
"epoch": 0.7908,
"grad_norm": 0.6069367627302308,
"learning_rate": 9.214820826817754e-06,
"loss": 0.4254,
"step": 659
},
{
"epoch": 0.792,
"grad_norm": 0.5691604792064953,
"learning_rate": 9.211062654524705e-06,
"loss": 0.4563,
"step": 660
},
{
"epoch": 0.7932,
"grad_norm": 0.5776804270149507,
"learning_rate": 9.207296279842162e-06,
"loss": 0.4283,
"step": 661
},
{
"epoch": 0.7944,
"grad_norm": 0.6999394766859208,
"learning_rate": 9.203521710106344e-06,
"loss": 0.4602,
"step": 662
},
{
"epoch": 0.7956,
"grad_norm": 0.6438802975614829,
"learning_rate": 9.199738952669431e-06,
"loss": 0.4393,
"step": 663
},
{
"epoch": 0.7968,
"grad_norm": 0.5992918054119178,
"learning_rate": 9.195948014899551e-06,
"loss": 0.4283,
"step": 664
},
{
"epoch": 0.798,
"grad_norm": 0.6202168032693136,
"learning_rate": 9.192148904180769e-06,
"loss": 0.4188,
"step": 665
},
{
"epoch": 0.7992,
"grad_norm": 0.7288678745681955,
"learning_rate": 9.188341627913061e-06,
"loss": 0.4312,
"step": 666
},
{
"epoch": 0.8004,
"grad_norm": 0.5808399736514862,
"learning_rate": 9.184526193512318e-06,
"loss": 0.4176,
"step": 667
},
{
"epoch": 0.8016,
"grad_norm": 0.5787103392680376,
"learning_rate": 9.180702608410314e-06,
"loss": 0.4474,
"step": 668
},
{
"epoch": 0.8028,
"grad_norm": 0.5832821317833403,
"learning_rate": 9.176870880054704e-06,
"loss": 0.3817,
"step": 669
},
{
"epoch": 0.804,
"grad_norm": 0.6566019896631167,
"learning_rate": 9.173031015909005e-06,
"loss": 0.4272,
"step": 670
},
{
"epoch": 0.8052,
"grad_norm": 0.6120582912743191,
"learning_rate": 9.169183023452574e-06,
"loss": 0.4293,
"step": 671
},
{
"epoch": 0.8064,
"grad_norm": 0.6226293645151404,
"learning_rate": 9.165326910180608e-06,
"loss": 0.384,
"step": 672
},
{
"epoch": 0.8076,
"grad_norm": 0.5943686082952457,
"learning_rate": 9.161462683604118e-06,
"loss": 0.4525,
"step": 673
},
{
"epoch": 0.8088,
"grad_norm": 0.6657427880686746,
"learning_rate": 9.157590351249923e-06,
"loss": 0.3887,
"step": 674
},
{
"epoch": 0.81,
"grad_norm": 0.6213739366625514,
"learning_rate": 9.153709920660624e-06,
"loss": 0.4349,
"step": 675
},
{
"epoch": 0.8112,
"grad_norm": 0.5478996164341824,
"learning_rate": 9.149821399394597e-06,
"loss": 0.4224,
"step": 676
},
{
"epoch": 0.8124,
"grad_norm": 0.6161518824983402,
"learning_rate": 9.145924795025984e-06,
"loss": 0.4267,
"step": 677
},
{
"epoch": 0.8136,
"grad_norm": 0.653475446858111,
"learning_rate": 9.142020115144662e-06,
"loss": 0.4256,
"step": 678
},
{
"epoch": 0.8148,
"grad_norm": 0.7258034801778298,
"learning_rate": 9.138107367356247e-06,
"loss": 0.4782,
"step": 679
},
{
"epoch": 0.816,
"grad_norm": 0.5964600131823877,
"learning_rate": 9.134186559282058e-06,
"loss": 0.417,
"step": 680
},
{
"epoch": 0.8172,
"grad_norm": 0.6665321408235314,
"learning_rate": 9.130257698559129e-06,
"loss": 0.4563,
"step": 681
},
{
"epoch": 0.8184,
"grad_norm": 0.6353290941165448,
"learning_rate": 9.126320792840165e-06,
"loss": 0.4097,
"step": 682
},
{
"epoch": 0.8196,
"grad_norm": 0.5550868353040516,
"learning_rate": 9.12237584979355e-06,
"loss": 0.4417,
"step": 683
},
{
"epoch": 0.8208,
"grad_norm": 0.6341329284133028,
"learning_rate": 9.11842287710332e-06,
"loss": 0.4447,
"step": 684
},
{
"epoch": 0.822,
"grad_norm": 0.6840045072161562,
"learning_rate": 9.114461882469154e-06,
"loss": 0.4093,
"step": 685
},
{
"epoch": 0.8232,
"grad_norm": 0.666702761146514,
"learning_rate": 9.110492873606351e-06,
"loss": 0.4351,
"step": 686
},
{
"epoch": 0.8244,
"grad_norm": 0.6770837060218068,
"learning_rate": 9.106515858245825e-06,
"loss": 0.4331,
"step": 687
},
{
"epoch": 0.8256,
"grad_norm": 0.6138322525456078,
"learning_rate": 9.102530844134084e-06,
"loss": 0.4153,
"step": 688
},
{
"epoch": 0.8268,
"grad_norm": 0.679029102207529,
"learning_rate": 9.098537839033213e-06,
"loss": 0.4697,
"step": 689
},
{
"epoch": 0.828,
"grad_norm": 0.5730739203701445,
"learning_rate": 9.094536850720867e-06,
"loss": 0.4137,
"step": 690
},
{
"epoch": 0.8292,
"grad_norm": 0.676913973932885,
"learning_rate": 9.090527886990249e-06,
"loss": 0.4377,
"step": 691
},
{
"epoch": 0.8304,
"grad_norm": 0.589745753186363,
"learning_rate": 9.086510955650095e-06,
"loss": 0.439,
"step": 692
},
{
"epoch": 0.8316,
"grad_norm": 0.6562093394558783,
"learning_rate": 9.082486064524663e-06,
"loss": 0.4702,
"step": 693
},
{
"epoch": 0.8328,
"grad_norm": 0.7014623843789226,
"learning_rate": 9.078453221453714e-06,
"loss": 0.4241,
"step": 694
},
{
"epoch": 0.834,
"grad_norm": 0.7211409839747195,
"learning_rate": 9.074412434292496e-06,
"loss": 0.466,
"step": 695
},
{
"epoch": 0.8352,
"grad_norm": 0.6327658360652593,
"learning_rate": 9.070363710911736e-06,
"loss": 0.4038,
"step": 696
},
{
"epoch": 0.8364,
"grad_norm": 0.7117499494134616,
"learning_rate": 9.066307059197612e-06,
"loss": 0.4542,
"step": 697
},
{
"epoch": 0.8376,
"grad_norm": 0.6779934892161844,
"learning_rate": 9.062242487051752e-06,
"loss": 0.4527,
"step": 698
},
{
"epoch": 0.8388,
"grad_norm": 0.6132662026238554,
"learning_rate": 9.058170002391205e-06,
"loss": 0.3867,
"step": 699
},
{
"epoch": 0.84,
"grad_norm": 0.6328253145414058,
"learning_rate": 9.05408961314844e-06,
"loss": 0.4552,
"step": 700
},
{
"epoch": 0.8412,
"grad_norm": 0.6093284392860376,
"learning_rate": 9.050001327271314e-06,
"loss": 0.3813,
"step": 701
},
{
"epoch": 0.8424,
"grad_norm": 0.5175203763355374,
"learning_rate": 9.045905152723074e-06,
"loss": 0.389,
"step": 702
},
{
"epoch": 0.8436,
"grad_norm": 0.5842103018209932,
"learning_rate": 9.041801097482323e-06,
"loss": 0.4245,
"step": 703
},
{
"epoch": 0.8448,
"grad_norm": 0.599455910511483,
"learning_rate": 9.037689169543024e-06,
"loss": 0.47,
"step": 704
},
{
"epoch": 0.846,
"grad_norm": 0.5465215060060533,
"learning_rate": 9.033569376914467e-06,
"loss": 0.4238,
"step": 705
},
{
"epoch": 0.8472,
"grad_norm": 0.5470742118288643,
"learning_rate": 9.029441727621267e-06,
"loss": 0.4145,
"step": 706
},
{
"epoch": 0.8484,
"grad_norm": 0.5969139676060078,
"learning_rate": 9.025306229703334e-06,
"loss": 0.4382,
"step": 707
},
{
"epoch": 0.8496,
"grad_norm": 0.5840979426058325,
"learning_rate": 9.021162891215879e-06,
"loss": 0.3982,
"step": 708
},
{
"epoch": 0.8508,
"grad_norm": 0.6069802881120857,
"learning_rate": 9.017011720229368e-06,
"loss": 0.4569,
"step": 709
},
{
"epoch": 0.852,
"grad_norm": 0.603696871584329,
"learning_rate": 9.012852724829539e-06,
"loss": 0.473,
"step": 710
},
{
"epoch": 0.8532,
"grad_norm": 0.6423008232878531,
"learning_rate": 9.008685913117361e-06,
"loss": 0.4572,
"step": 711
},
{
"epoch": 0.8544,
"grad_norm": 0.5264477044024208,
"learning_rate": 9.00451129320903e-06,
"loss": 0.4183,
"step": 712
},
{
"epoch": 0.8556,
"grad_norm": 0.5893466589452381,
"learning_rate": 9.000328873235955e-06,
"loss": 0.4322,
"step": 713
},
{
"epoch": 0.8568,
"grad_norm": 0.6272192483642774,
"learning_rate": 8.996138661344734e-06,
"loss": 0.4471,
"step": 714
},
{
"epoch": 0.858,
"grad_norm": 0.6509496086878485,
"learning_rate": 8.99194066569714e-06,
"loss": 0.4292,
"step": 715
},
{
"epoch": 0.8592,
"grad_norm": 0.5828069494646737,
"learning_rate": 8.987734894470111e-06,
"loss": 0.4054,
"step": 716
},
{
"epoch": 0.8604,
"grad_norm": 0.6121311166005416,
"learning_rate": 8.983521355855731e-06,
"loss": 0.4549,
"step": 717
},
{
"epoch": 0.8616,
"grad_norm": 0.5311697585062636,
"learning_rate": 8.979300058061214e-06,
"loss": 0.4543,
"step": 718
},
{
"epoch": 0.8628,
"grad_norm": 0.5954597656920868,
"learning_rate": 8.97507100930888e-06,
"loss": 0.4127,
"step": 719
},
{
"epoch": 0.864,
"grad_norm": 0.6603824524906432,
"learning_rate": 8.97083421783616e-06,
"loss": 0.4689,
"step": 720
},
{
"epoch": 0.8652,
"grad_norm": 0.5873463266140205,
"learning_rate": 8.96658969189555e-06,
"loss": 0.4581,
"step": 721
},
{
"epoch": 0.8664,
"grad_norm": 0.6302692465732681,
"learning_rate": 8.962337439754627e-06,
"loss": 0.3957,
"step": 722
},
{
"epoch": 0.8676,
"grad_norm": 0.5660832560318835,
"learning_rate": 8.958077469696007e-06,
"loss": 0.4216,
"step": 723
},
{
"epoch": 0.8688,
"grad_norm": 0.6057306547058764,
"learning_rate": 8.953809790017342e-06,
"loss": 0.4203,
"step": 724
},
{
"epoch": 0.87,
"grad_norm": 0.5791265869794112,
"learning_rate": 8.949534409031305e-06,
"loss": 0.4125,
"step": 725
},
{
"epoch": 0.8712,
"grad_norm": 0.6074661846992012,
"learning_rate": 8.94525133506556e-06,
"loss": 0.425,
"step": 726
},
{
"epoch": 0.8724,
"grad_norm": 0.6041559413723928,
"learning_rate": 8.940960576462763e-06,
"loss": 0.4132,
"step": 727
},
{
"epoch": 0.8736,
"grad_norm": 0.5119184403985256,
"learning_rate": 8.93666214158054e-06,
"loss": 0.4173,
"step": 728
},
{
"epoch": 0.8748,
"grad_norm": 0.6184156720209869,
"learning_rate": 8.932356038791465e-06,
"loss": 0.47,
"step": 729
},
{
"epoch": 0.876,
"grad_norm": 0.6445492952041065,
"learning_rate": 8.928042276483048e-06,
"loss": 0.442,
"step": 730
},
{
"epoch": 0.8772,
"grad_norm": 0.560388170961086,
"learning_rate": 8.923720863057718e-06,
"loss": 0.4378,
"step": 731
},
{
"epoch": 0.8784,
"grad_norm": 0.6295368599127688,
"learning_rate": 8.919391806932807e-06,
"loss": 0.4294,
"step": 732
},
{
"epoch": 0.8796,
"grad_norm": 0.6154129494849543,
"learning_rate": 8.915055116540538e-06,
"loss": 0.4734,
"step": 733
},
{
"epoch": 0.8808,
"grad_norm": 0.6775101992228028,
"learning_rate": 8.910710800327996e-06,
"loss": 0.4273,
"step": 734
},
{
"epoch": 0.882,
"grad_norm": 0.5810481242782447,
"learning_rate": 8.906358866757128e-06,
"loss": 0.4419,
"step": 735
},
{
"epoch": 0.8832,
"grad_norm": 0.6474016395952771,
"learning_rate": 8.901999324304713e-06,
"loss": 0.4576,
"step": 736
},
{
"epoch": 0.8844,
"grad_norm": 0.5687417809242015,
"learning_rate": 8.897632181462354e-06,
"loss": 0.3953,
"step": 737
},
{
"epoch": 0.8856,
"grad_norm": 0.626015511992478,
"learning_rate": 8.893257446736455e-06,
"loss": 0.458,
"step": 738
},
{
"epoch": 0.8868,
"grad_norm": 0.5936874146722684,
"learning_rate": 8.888875128648208e-06,
"loss": 0.4139,
"step": 739
},
{
"epoch": 0.888,
"grad_norm": 0.5445910669585022,
"learning_rate": 8.884485235733579e-06,
"loss": 0.4083,
"step": 740
},
{
"epoch": 0.8892,
"grad_norm": 0.5871241121470941,
"learning_rate": 8.880087776543287e-06,
"loss": 0.4161,
"step": 741
},
{
"epoch": 0.8904,
"grad_norm": 0.5861156757463913,
"learning_rate": 8.875682759642786e-06,
"loss": 0.4572,
"step": 742
},
{
"epoch": 0.8916,
"grad_norm": 0.6524639315283577,
"learning_rate": 8.871270193612254e-06,
"loss": 0.415,
"step": 743
},
{
"epoch": 0.8928,
"grad_norm": 0.5579578541929966,
"learning_rate": 8.866850087046574e-06,
"loss": 0.4599,
"step": 744
},
{
"epoch": 0.894,
"grad_norm": 0.5983372699604969,
"learning_rate": 8.862422448555317e-06,
"loss": 0.4685,
"step": 745
},
{
"epoch": 0.8952,
"grad_norm": 0.5313257499086619,
"learning_rate": 8.857987286762718e-06,
"loss": 0.4482,
"step": 746
},
{
"epoch": 0.8964,
"grad_norm": 0.5741224075650853,
"learning_rate": 8.853544610307675e-06,
"loss": 0.4593,
"step": 747
},
{
"epoch": 0.8976,
"grad_norm": 0.5180035885365164,
"learning_rate": 8.84909442784372e-06,
"loss": 0.4393,
"step": 748
},
{
"epoch": 0.8988,
"grad_norm": 0.6267083133045676,
"learning_rate": 8.844636748038999e-06,
"loss": 0.4835,
"step": 749
},
{
"epoch": 0.9,
"grad_norm": 0.5831683863999229,
"learning_rate": 8.840171579576273e-06,
"loss": 0.4201,
"step": 750
},
{
"epoch": 0.9012,
"grad_norm": 0.5405806665890253,
"learning_rate": 8.83569893115288e-06,
"loss": 0.4114,
"step": 751
},
{
"epoch": 0.9024,
"grad_norm": 0.5885634994888672,
"learning_rate": 8.831218811480735e-06,
"loss": 0.4217,
"step": 752
},
{
"epoch": 0.9036,
"grad_norm": 0.5630756675541398,
"learning_rate": 8.8267312292863e-06,
"loss": 0.4205,
"step": 753
},
{
"epoch": 0.9048,
"grad_norm": 0.5589821186819856,
"learning_rate": 8.822236193310574e-06,
"loss": 0.3911,
"step": 754
},
{
"epoch": 0.906,
"grad_norm": 0.5377971370223851,
"learning_rate": 8.817733712309078e-06,
"loss": 0.416,
"step": 755
},
{
"epoch": 0.9072,
"grad_norm": 0.6594746319910847,
"learning_rate": 8.813223795051828e-06,
"loss": 0.449,
"step": 756
},
{
"epoch": 0.9084,
"grad_norm": 0.6337830271660636,
"learning_rate": 8.80870645032333e-06,
"loss": 0.4733,
"step": 757
},
{
"epoch": 0.9096,
"grad_norm": 0.5096341057306893,
"learning_rate": 8.804181686922555e-06,
"loss": 0.3898,
"step": 758
},
{
"epoch": 0.9108,
"grad_norm": 0.5551424785674842,
"learning_rate": 8.799649513662926e-06,
"loss": 0.4354,
"step": 759
},
{
"epoch": 0.912,
"grad_norm": 0.5430810494161301,
"learning_rate": 8.795109939372298e-06,
"loss": 0.4314,
"step": 760
},
{
"epoch": 0.9132,
"grad_norm": 0.7303510387293701,
"learning_rate": 8.79056297289294e-06,
"loss": 0.4608,
"step": 761
},
{
"epoch": 0.9144,
"grad_norm": 0.58979590436832,
"learning_rate": 8.786008623081526e-06,
"loss": 0.408,
"step": 762
},
{
"epoch": 0.9156,
"grad_norm": 0.6431267478376745,
"learning_rate": 8.781446898809101e-06,
"loss": 0.4256,
"step": 763
},
{
"epoch": 0.9168,
"grad_norm": 0.6800621940281214,
"learning_rate": 8.776877808961082e-06,
"loss": 0.4125,
"step": 764
},
{
"epoch": 0.918,
"grad_norm": 0.6078542654953756,
"learning_rate": 8.772301362437233e-06,
"loss": 0.4037,
"step": 765
},
{
"epoch": 0.9192,
"grad_norm": 0.6667156303614546,
"learning_rate": 8.767717568151643e-06,
"loss": 0.451,
"step": 766
},
{
"epoch": 0.9204,
"grad_norm": 0.6067648015386979,
"learning_rate": 8.763126435032717e-06,
"loss": 0.4565,
"step": 767
},
{
"epoch": 0.9216,
"grad_norm": 0.6192295222939654,
"learning_rate": 8.758527972023151e-06,
"loss": 0.4849,
"step": 768
},
{
"epoch": 0.9228,
"grad_norm": 0.6777113466817958,
"learning_rate": 8.753922188079923e-06,
"loss": 0.4293,
"step": 769
},
{
"epoch": 0.924,
"grad_norm": 0.5627832110347925,
"learning_rate": 8.749309092174267e-06,
"loss": 0.4395,
"step": 770
},
{
"epoch": 0.9252,
"grad_norm": 0.5973915994189181,
"learning_rate": 8.744688693291658e-06,
"loss": 0.4326,
"step": 771
},
{
"epoch": 0.9264,
"grad_norm": 0.617631648752449,
"learning_rate": 8.740061000431805e-06,
"loss": 0.431,
"step": 772
},
{
"epoch": 0.9276,
"grad_norm": 0.6257907130189103,
"learning_rate": 8.735426022608611e-06,
"loss": 0.4335,
"step": 773
},
{
"epoch": 0.9288,
"grad_norm": 0.7140143377367802,
"learning_rate": 8.73078376885018e-06,
"loss": 0.4372,
"step": 774
},
{
"epoch": 0.93,
"grad_norm": 0.59771651363443,
"learning_rate": 8.726134248198782e-06,
"loss": 0.4215,
"step": 775
},
{
"epoch": 0.9312,
"grad_norm": 0.5943798731936335,
"learning_rate": 8.721477469710845e-06,
"loss": 0.4462,
"step": 776
},
{
"epoch": 0.9324,
"grad_norm": 0.5461113902503175,
"learning_rate": 8.71681344245693e-06,
"loss": 0.4113,
"step": 777
},
{
"epoch": 0.9336,
"grad_norm": 0.515942973208093,
"learning_rate": 8.712142175521723e-06,
"loss": 0.4043,
"step": 778
},
{
"epoch": 0.9348,
"grad_norm": 0.5451413502419838,
"learning_rate": 8.707463678004004e-06,
"loss": 0.3996,
"step": 779
},
{
"epoch": 0.936,
"grad_norm": 0.5670643504057882,
"learning_rate": 8.702777959016647e-06,
"loss": 0.408,
"step": 780
},
{
"epoch": 0.9372,
"grad_norm": 0.6228540869849157,
"learning_rate": 8.698085027686581e-06,
"loss": 0.4175,
"step": 781
},
{
"epoch": 0.9384,
"grad_norm": 0.664468192479229,
"learning_rate": 8.69338489315479e-06,
"loss": 0.4633,
"step": 782
},
{
"epoch": 0.9396,
"grad_norm": 0.6484118822851579,
"learning_rate": 8.68867756457629e-06,
"loss": 0.4073,
"step": 783
},
{
"epoch": 0.9408,
"grad_norm": 0.6014610628440286,
"learning_rate": 8.683963051120103e-06,
"loss": 0.4577,
"step": 784
},
{
"epoch": 0.942,
"grad_norm": 0.6898835243028868,
"learning_rate": 8.679241361969252e-06,
"loss": 0.4383,
"step": 785
},
{
"epoch": 0.9432,
"grad_norm": 0.6872534552572807,
"learning_rate": 8.674512506320733e-06,
"loss": 0.4585,
"step": 786
},
{
"epoch": 0.9444,
"grad_norm": 0.5229387300614929,
"learning_rate": 8.669776493385506e-06,
"loss": 0.4234,
"step": 787
},
{
"epoch": 0.9456,
"grad_norm": 0.5528374158454885,
"learning_rate": 8.665033332388466e-06,
"loss": 0.4342,
"step": 788
},
{
"epoch": 0.9468,
"grad_norm": 0.6163928252539012,
"learning_rate": 8.660283032568435e-06,
"loss": 0.4114,
"step": 789
},
{
"epoch": 0.948,
"grad_norm": 0.6580057062480736,
"learning_rate": 8.655525603178137e-06,
"loss": 0.444,
"step": 790
},
{
"epoch": 0.9492,
"grad_norm": 0.5104550641112756,
"learning_rate": 8.650761053484188e-06,
"loss": 0.3754,
"step": 791
},
{
"epoch": 0.9504,
"grad_norm": 0.6206032119821056,
"learning_rate": 8.645989392767068e-06,
"loss": 0.4312,
"step": 792
},
{
"epoch": 0.9516,
"grad_norm": 0.619139263946645,
"learning_rate": 8.641210630321115e-06,
"loss": 0.4523,
"step": 793
},
{
"epoch": 0.9528,
"grad_norm": 0.6111017449033144,
"learning_rate": 8.636424775454489e-06,
"loss": 0.4259,
"step": 794
},
{
"epoch": 0.954,
"grad_norm": 0.6401410230121702,
"learning_rate": 8.631631837489173e-06,
"loss": 0.439,
"step": 795
},
{
"epoch": 0.9552,
"grad_norm": 0.5181884328548102,
"learning_rate": 8.626831825760946e-06,
"loss": 0.436,
"step": 796
},
{
"epoch": 0.9564,
"grad_norm": 0.5334649831223063,
"learning_rate": 8.622024749619363e-06,
"loss": 0.3736,
"step": 797
},
{
"epoch": 0.9576,
"grad_norm": 0.5454287713548887,
"learning_rate": 8.61721061842774e-06,
"loss": 0.393,
"step": 798
},
{
"epoch": 0.9588,
"grad_norm": 0.6269409723008791,
"learning_rate": 8.612389441563136e-06,
"loss": 0.4473,
"step": 799
},
{
"epoch": 0.96,
"grad_norm": 0.594447646809085,
"learning_rate": 8.60756122841633e-06,
"loss": 0.4438,
"step": 800
},
{
"epoch": 0.9612,
"grad_norm": 0.7102600986895227,
"learning_rate": 8.602725988391814e-06,
"loss": 0.4139,
"step": 801
},
{
"epoch": 0.9624,
"grad_norm": 0.5254806738684399,
"learning_rate": 8.597883730907757e-06,
"loss": 0.4293,
"step": 802
},
{
"epoch": 0.9636,
"grad_norm": 0.6746355278803491,
"learning_rate": 8.593034465396007e-06,
"loss": 0.4331,
"step": 803
},
{
"epoch": 0.9648,
"grad_norm": 0.5759430041057655,
"learning_rate": 8.588178201302052e-06,
"loss": 0.4639,
"step": 804
},
{
"epoch": 0.966,
"grad_norm": 0.4887389359434623,
"learning_rate": 8.583314948085023e-06,
"loss": 0.4314,
"step": 805
},
{
"epoch": 0.9672,
"grad_norm": 0.6240592097801354,
"learning_rate": 8.578444715217652e-06,
"loss": 0.4693,
"step": 806
},
{
"epoch": 0.9684,
"grad_norm": 0.5906677636695606,
"learning_rate": 8.57356751218628e-06,
"loss": 0.4128,
"step": 807
},
{
"epoch": 0.9696,
"grad_norm": 0.5522058124741028,
"learning_rate": 8.568683348490817e-06,
"loss": 0.423,
"step": 808
},
{
"epoch": 0.9708,
"grad_norm": 0.519899853308098,
"learning_rate": 8.563792233644725e-06,
"loss": 0.3736,
"step": 809
},
{
"epoch": 0.972,
"grad_norm": 0.6138196751505098,
"learning_rate": 8.558894177175019e-06,
"loss": 0.4196,
"step": 810
},
{
"epoch": 0.9732,
"grad_norm": 0.5873280535324105,
"learning_rate": 8.553989188622228e-06,
"loss": 0.4282,
"step": 811
},
{
"epoch": 0.9744,
"grad_norm": 0.5976713589531968,
"learning_rate": 8.549077277540379e-06,
"loss": 0.4445,
"step": 812
},
{
"epoch": 0.9756,
"grad_norm": 0.6502292535424105,
"learning_rate": 8.544158453496992e-06,
"loss": 0.4108,
"step": 813
},
{
"epoch": 0.9768,
"grad_norm": 0.578412995801094,
"learning_rate": 8.539232726073046e-06,
"loss": 0.4226,
"step": 814
},
{
"epoch": 0.978,
"grad_norm": 0.5707868572285869,
"learning_rate": 8.53430010486297e-06,
"loss": 0.4008,
"step": 815
},
{
"epoch": 0.9792,
"grad_norm": 0.5852269887281469,
"learning_rate": 8.529360599474616e-06,
"loss": 0.4257,
"step": 816
},
{
"epoch": 0.9804,
"grad_norm": 0.609048557208369,
"learning_rate": 8.524414219529253e-06,
"loss": 0.4436,
"step": 817
},
{
"epoch": 0.9816,
"grad_norm": 0.5406348011212092,
"learning_rate": 8.519460974661533e-06,
"loss": 0.4204,
"step": 818
},
{
"epoch": 0.9828,
"grad_norm": 0.5677734036143112,
"learning_rate": 8.514500874519483e-06,
"loss": 0.4348,
"step": 819
},
{
"epoch": 0.984,
"grad_norm": 0.5388294590032593,
"learning_rate": 8.509533928764482e-06,
"loss": 0.4105,
"step": 820
},
{
"epoch": 0.9852,
"grad_norm": 0.519324821092748,
"learning_rate": 8.50456014707124e-06,
"loss": 0.3875,
"step": 821
},
{
"epoch": 0.9864,
"grad_norm": 0.6125545198699792,
"learning_rate": 8.499579539127794e-06,
"loss": 0.4901,
"step": 822
},
{
"epoch": 0.9876,
"grad_norm": 0.6206780765007368,
"learning_rate": 8.494592114635458e-06,
"loss": 0.4046,
"step": 823
},
{
"epoch": 0.9888,
"grad_norm": 0.6222065962769607,
"learning_rate": 8.489597883308844e-06,
"loss": 0.4488,
"step": 824
},
{
"epoch": 0.99,
"grad_norm": 0.5329723921756123,
"learning_rate": 8.484596854875806e-06,
"loss": 0.4141,
"step": 825
},
{
"epoch": 0.9912,
"grad_norm": 0.5993982737276878,
"learning_rate": 8.479589039077446e-06,
"loss": 0.4308,
"step": 826
},
{
"epoch": 0.9924,
"grad_norm": 0.5584722292572113,
"learning_rate": 8.474574445668085e-06,
"loss": 0.4568,
"step": 827
},
{
"epoch": 0.9936,
"grad_norm": 0.6091566316507482,
"learning_rate": 8.469553084415247e-06,
"loss": 0.4556,
"step": 828
},
{
"epoch": 0.9948,
"grad_norm": 0.6233664933735439,
"learning_rate": 8.464524965099632e-06,
"loss": 0.4206,
"step": 829
},
{
"epoch": 0.996,
"grad_norm": 0.5293099738021775,
"learning_rate": 8.459490097515114e-06,
"loss": 0.4071,
"step": 830
},
{
"epoch": 0.9972,
"grad_norm": 0.5350545129837626,
"learning_rate": 8.454448491468702e-06,
"loss": 0.461,
"step": 831
},
{
"epoch": 0.9984,
"grad_norm": 0.5326198305245626,
"learning_rate": 8.449400156780536e-06,
"loss": 0.4713,
"step": 832
},
{
"epoch": 0.9996,
"grad_norm": 0.5901616953038845,
"learning_rate": 8.444345103283858e-06,
"loss": 0.4088,
"step": 833
},
{
"epoch": 1.0,
"grad_norm": 0.5901616953038845,
"learning_rate": 8.439283340825002e-06,
"loss": 0.4155,
"step": 834
},
{
"epoch": 1.0012,
"grad_norm": 1.1368981411522645,
"learning_rate": 8.434214879263365e-06,
"loss": 0.3407,
"step": 835
},
{
"epoch": 1.0024,
"grad_norm": 0.5681339768011925,
"learning_rate": 8.429139728471395e-06,
"loss": 0.3837,
"step": 836
},
{
"epoch": 1.0036,
"grad_norm": 0.60245776707601,
"learning_rate": 8.424057898334569e-06,
"loss": 0.3865,
"step": 837
},
{
"epoch": 1.0048,
"grad_norm": 0.5452067685083533,
"learning_rate": 8.418969398751375e-06,
"loss": 0.3761,
"step": 838
},
{
"epoch": 1.006,
"grad_norm": 0.5432481480755952,
"learning_rate": 8.413874239633291e-06,
"loss": 0.3312,
"step": 839
},
{
"epoch": 1.0072,
"grad_norm": 0.5850088244625832,
"learning_rate": 8.408772430904768e-06,
"loss": 0.3936,
"step": 840
},
{
"epoch": 1.0084,
"grad_norm": 0.6134958489823171,
"learning_rate": 8.403663982503205e-06,
"loss": 0.3783,
"step": 841
},
{
"epoch": 1.0096,
"grad_norm": 0.7310393810023733,
"learning_rate": 8.398548904378938e-06,
"loss": 0.3633,
"step": 842
},
{
"epoch": 1.0108,
"grad_norm": 0.7594188380669021,
"learning_rate": 8.393427206495217e-06,
"loss": 0.3487,
"step": 843
},
{
"epoch": 1.012,
"grad_norm": 0.5430393582609394,
"learning_rate": 8.388298898828182e-06,
"loss": 0.3709,
"step": 844
},
{
"epoch": 1.0132,
"grad_norm": 0.5470009445017966,
"learning_rate": 8.383163991366852e-06,
"loss": 0.3646,
"step": 845
},
{
"epoch": 1.0144,
"grad_norm": 0.6955376746178842,
"learning_rate": 8.378022494113099e-06,
"loss": 0.3942,
"step": 846
},
{
"epoch": 1.0156,
"grad_norm": 0.5443192423069556,
"learning_rate": 8.372874417081632e-06,
"loss": 0.3858,
"step": 847
},
{
"epoch": 1.0168,
"grad_norm": 0.5383937902578093,
"learning_rate": 8.367719770299972e-06,
"loss": 0.3424,
"step": 848
},
{
"epoch": 1.018,
"grad_norm": 0.5753603289032625,
"learning_rate": 8.36255856380844e-06,
"loss": 0.3775,
"step": 849
},
{
"epoch": 1.0192,
"grad_norm": 0.5817346658130989,
"learning_rate": 8.35739080766014e-06,
"loss": 0.3656,
"step": 850
},
{
"epoch": 1.0204,
"grad_norm": 0.5106554756161691,
"learning_rate": 8.352216511920921e-06,
"loss": 0.3676,
"step": 851
},
{
"epoch": 1.0216,
"grad_norm": 0.7331147307012948,
"learning_rate": 8.34703568666938e-06,
"loss": 0.416,
"step": 852
},
{
"epoch": 1.0228,
"grad_norm": 0.6003405960110348,
"learning_rate": 8.341848341996828e-06,
"loss": 0.3731,
"step": 853
},
{
"epoch": 1.024,
"grad_norm": 0.6343146490156437,
"learning_rate": 8.336654488007277e-06,
"loss": 0.3696,
"step": 854
},
{
"epoch": 1.0252,
"grad_norm": 0.6076533649109559,
"learning_rate": 8.331454134817414e-06,
"loss": 0.3635,
"step": 855
},
{
"epoch": 1.0264,
"grad_norm": 0.5107426303437482,
"learning_rate": 8.326247292556588e-06,
"loss": 0.365,
"step": 856
},
{
"epoch": 1.0276,
"grad_norm": 0.6148391495624019,
"learning_rate": 8.321033971366788e-06,
"loss": 0.3838,
"step": 857
},
{
"epoch": 1.0288,
"grad_norm": 0.5880138628410642,
"learning_rate": 8.315814181402623e-06,
"loss": 0.3796,
"step": 858
},
{
"epoch": 1.03,
"grad_norm": 0.587112143817172,
"learning_rate": 8.310587932831302e-06,
"loss": 0.3873,
"step": 859
},
{
"epoch": 1.0312,
"grad_norm": 0.6083837513556604,
"learning_rate": 8.305355235832611e-06,
"loss": 0.3788,
"step": 860
},
{
"epoch": 1.0324,
"grad_norm": 0.611985092855698,
"learning_rate": 8.300116100598899e-06,
"loss": 0.3979,
"step": 861
},
{
"epoch": 1.0336,
"grad_norm": 0.5127321006506305,
"learning_rate": 8.294870537335054e-06,
"loss": 0.3646,
"step": 862
},
{
"epoch": 1.0348,
"grad_norm": 0.6133913722569919,
"learning_rate": 8.28961855625849e-06,
"loss": 0.3477,
"step": 863
},
{
"epoch": 1.036,
"grad_norm": 0.5450221570823442,
"learning_rate": 8.284360167599113e-06,
"loss": 0.3767,
"step": 864
},
{
"epoch": 1.0372,
"grad_norm": 0.5590817088934886,
"learning_rate": 8.279095381599318e-06,
"loss": 0.4125,
"step": 865
},
{
"epoch": 1.0384,
"grad_norm": 0.6700002859155857,
"learning_rate": 8.273824208513956e-06,
"loss": 0.3979,
"step": 866
},
{
"epoch": 1.0396,
"grad_norm": 0.5752277444681116,
"learning_rate": 8.268546658610319e-06,
"loss": 0.3758,
"step": 867
},
{
"epoch": 1.0408,
"grad_norm": 0.468366324162763,
"learning_rate": 8.26326274216812e-06,
"loss": 0.3532,
"step": 868
},
{
"epoch": 1.042,
"grad_norm": 0.5970315093052967,
"learning_rate": 8.257972469479478e-06,
"loss": 0.3935,
"step": 869
},
{
"epoch": 1.0432,
"grad_norm": 0.5782502505738331,
"learning_rate": 8.252675850848886e-06,
"loss": 0.4055,
"step": 870
},
{
"epoch": 1.0444,
"grad_norm": 0.6631298992386631,
"learning_rate": 8.247372896593203e-06,
"loss": 0.4402,
"step": 871
},
{
"epoch": 1.0456,
"grad_norm": 0.6584560435617298,
"learning_rate": 8.24206361704162e-06,
"loss": 0.3878,
"step": 872
},
{
"epoch": 1.0468,
"grad_norm": 0.5480440560061248,
"learning_rate": 8.236748022535662e-06,
"loss": 0.3733,
"step": 873
},
{
"epoch": 1.048,
"grad_norm": 0.486477480838575,
"learning_rate": 8.231426123429143e-06,
"loss": 0.3764,
"step": 874
},
{
"epoch": 1.0492,
"grad_norm": 0.6142299872341441,
"learning_rate": 8.226097930088162e-06,
"loss": 0.3728,
"step": 875
},
{
"epoch": 1.0504,
"grad_norm": 0.5858378638083889,
"learning_rate": 8.220763452891078e-06,
"loss": 0.3613,
"step": 876
},
{
"epoch": 1.0516,
"grad_norm": 0.6139402721613479,
"learning_rate": 8.215422702228487e-06,
"loss": 0.3953,
"step": 877
},
{
"epoch": 1.0528,
"grad_norm": 0.4753434037158642,
"learning_rate": 8.210075688503209e-06,
"loss": 0.3758,
"step": 878
},
{
"epoch": 1.054,
"grad_norm": 0.6564325725197425,
"learning_rate": 8.20472242213026e-06,
"loss": 0.3856,
"step": 879
},
{
"epoch": 1.0552,
"grad_norm": 0.5887890156582547,
"learning_rate": 8.199362913536837e-06,
"loss": 0.3843,
"step": 880
},
{
"epoch": 1.0564,
"grad_norm": 1.364183705833157,
"learning_rate": 8.193997173162293e-06,
"loss": 0.409,
"step": 881
},
{
"epoch": 1.0576,
"grad_norm": 0.6165333073646311,
"learning_rate": 8.188625211458123e-06,
"loss": 0.382,
"step": 882
},
{
"epoch": 1.0588,
"grad_norm": 0.6041705166135537,
"learning_rate": 8.183247038887937e-06,
"loss": 0.3749,
"step": 883
},
{
"epoch": 1.06,
"grad_norm": 0.5109914794108413,
"learning_rate": 8.177862665927445e-06,
"loss": 0.4021,
"step": 884
},
{
"epoch": 1.0612,
"grad_norm": 0.742930894603376,
"learning_rate": 8.17247210306443e-06,
"loss": 0.3833,
"step": 885
},
{
"epoch": 1.0624,
"grad_norm": 0.5932978751229475,
"learning_rate": 8.167075360798739e-06,
"loss": 0.421,
"step": 886
},
{
"epoch": 1.0636,
"grad_norm": 0.6758659142255161,
"learning_rate": 8.161672449642248e-06,
"loss": 0.414,
"step": 887
},
{
"epoch": 1.0648,
"grad_norm": 0.4904299584222583,
"learning_rate": 8.156263380118855e-06,
"loss": 0.3795,
"step": 888
},
{
"epoch": 1.066,
"grad_norm": 0.6513253017015571,
"learning_rate": 8.150848162764448e-06,
"loss": 0.3342,
"step": 889
},
{
"epoch": 1.0672,
"grad_norm": 0.5473246180694088,
"learning_rate": 8.145426808126894e-06,
"loss": 0.3682,
"step": 890
},
{
"epoch": 1.0684,
"grad_norm": 0.6249870992905973,
"learning_rate": 8.139999326766011e-06,
"loss": 0.392,
"step": 891
},
{
"epoch": 1.0695999999999999,
"grad_norm": 0.6798970371707683,
"learning_rate": 8.134565729253554e-06,
"loss": 0.3877,
"step": 892
},
{
"epoch": 1.0708,
"grad_norm": 0.6216883250382137,
"learning_rate": 8.129126026173189e-06,
"loss": 0.3796,
"step": 893
},
{
"epoch": 1.072,
"grad_norm": 0.5807715354692247,
"learning_rate": 8.123680228120474e-06,
"loss": 0.3777,
"step": 894
},
{
"epoch": 1.0732,
"grad_norm": 0.5740983667800384,
"learning_rate": 8.118228345702843e-06,
"loss": 0.3837,
"step": 895
},
{
"epoch": 1.0744,
"grad_norm": 0.5956483364277361,
"learning_rate": 8.112770389539574e-06,
"loss": 0.4024,
"step": 896
},
{
"epoch": 1.0756000000000001,
"grad_norm": 0.6140423109873733,
"learning_rate": 8.107306370261785e-06,
"loss": 0.3918,
"step": 897
},
{
"epoch": 1.0768,
"grad_norm": 0.568523839966626,
"learning_rate": 8.101836298512396e-06,
"loss": 0.3879,
"step": 898
},
{
"epoch": 1.078,
"grad_norm": 0.6580373875828848,
"learning_rate": 8.096360184946117e-06,
"loss": 0.3843,
"step": 899
},
{
"epoch": 1.0792,
"grad_norm": 0.7056683024436424,
"learning_rate": 8.09087804022943e-06,
"loss": 0.3941,
"step": 900
},
{
"epoch": 1.0804,
"grad_norm": 0.6575712990687944,
"learning_rate": 8.085389875040566e-06,
"loss": 0.3931,
"step": 901
},
{
"epoch": 1.0816,
"grad_norm": 0.5930671138814431,
"learning_rate": 8.079895700069473e-06,
"loss": 0.3731,
"step": 902
},
{
"epoch": 1.0828,
"grad_norm": 0.6630649504900762,
"learning_rate": 8.074395526017816e-06,
"loss": 0.406,
"step": 903
},
{
"epoch": 1.084,
"grad_norm": 0.6261787992626612,
"learning_rate": 8.06888936359894e-06,
"loss": 0.3594,
"step": 904
},
{
"epoch": 1.0852,
"grad_norm": 0.7502706494959577,
"learning_rate": 8.063377223537853e-06,
"loss": 0.4491,
"step": 905
},
{
"epoch": 1.0864,
"grad_norm": 0.5534701644541706,
"learning_rate": 8.057859116571213e-06,
"loss": 0.3703,
"step": 906
},
{
"epoch": 1.0876,
"grad_norm": 0.49799763923502727,
"learning_rate": 8.05233505344729e-06,
"loss": 0.3543,
"step": 907
},
{
"epoch": 1.0888,
"grad_norm": 1.0019839872604202,
"learning_rate": 8.046805044925964e-06,
"loss": 0.4528,
"step": 908
},
{
"epoch": 1.09,
"grad_norm": 0.5461445023349752,
"learning_rate": 8.041269101778694e-06,
"loss": 0.3737,
"step": 909
},
{
"epoch": 1.0912,
"grad_norm": 0.628865585347978,
"learning_rate": 8.035727234788496e-06,
"loss": 0.4015,
"step": 910
},
{
"epoch": 1.0924,
"grad_norm": 0.6079959467058422,
"learning_rate": 8.030179454749925e-06,
"loss": 0.3825,
"step": 911
},
{
"epoch": 1.0936,
"grad_norm": 0.5305429021193003,
"learning_rate": 8.024625772469055e-06,
"loss": 0.3451,
"step": 912
},
{
"epoch": 1.0948,
"grad_norm": 0.7032844908658168,
"learning_rate": 8.019066198763458e-06,
"loss": 0.3684,
"step": 913
},
{
"epoch": 1.096,
"grad_norm": 0.6134951245718815,
"learning_rate": 8.013500744462177e-06,
"loss": 0.3723,
"step": 914
},
{
"epoch": 1.0972,
"grad_norm": 0.6064703933463657,
"learning_rate": 8.007929420405714e-06,
"loss": 0.3598,
"step": 915
},
{
"epoch": 1.0984,
"grad_norm": 0.5129566311438788,
"learning_rate": 8.002352237446e-06,
"loss": 0.3657,
"step": 916
},
{
"epoch": 1.0996,
"grad_norm": 0.7027224551032671,
"learning_rate": 7.996769206446383e-06,
"loss": 0.3788,
"step": 917
},
{
"epoch": 1.1008,
"grad_norm": 0.5759047752415848,
"learning_rate": 7.991180338281594e-06,
"loss": 0.407,
"step": 918
},
{
"epoch": 1.102,
"grad_norm": 0.6444902877949182,
"learning_rate": 7.985585643837743e-06,
"loss": 0.3525,
"step": 919
},
{
"epoch": 1.1032,
"grad_norm": 0.5481881054619812,
"learning_rate": 7.979985134012285e-06,
"loss": 0.3501,
"step": 920
},
{
"epoch": 1.1044,
"grad_norm": 0.733075166199247,
"learning_rate": 7.974378819713998e-06,
"loss": 0.3672,
"step": 921
},
{
"epoch": 1.1056,
"grad_norm": 0.5614221050399015,
"learning_rate": 7.968766711862971e-06,
"loss": 0.4116,
"step": 922
},
{
"epoch": 1.1068,
"grad_norm": 0.5800899794679014,
"learning_rate": 7.963148821390578e-06,
"loss": 0.3827,
"step": 923
},
{
"epoch": 1.108,
"grad_norm": 0.6622601415747164,
"learning_rate": 7.957525159239454e-06,
"loss": 0.3893,
"step": 924
},
{
"epoch": 1.1092,
"grad_norm": 0.8333817588459229,
"learning_rate": 7.951895736363477e-06,
"loss": 0.3963,
"step": 925
},
{
"epoch": 1.1104,
"grad_norm": 0.5559197241751965,
"learning_rate": 7.946260563727746e-06,
"loss": 0.4014,
"step": 926
},
{
"epoch": 1.1116,
"grad_norm": 0.5241069356216917,
"learning_rate": 7.940619652308562e-06,
"loss": 0.3773,
"step": 927
},
{
"epoch": 1.1128,
"grad_norm": 0.620024828062137,
"learning_rate": 7.934973013093397e-06,
"loss": 0.3507,
"step": 928
},
{
"epoch": 1.114,
"grad_norm": 0.5232386711797898,
"learning_rate": 7.929320657080886e-06,
"loss": 0.4054,
"step": 929
},
{
"epoch": 1.1152,
"grad_norm": 0.5683648527722198,
"learning_rate": 7.923662595280799e-06,
"loss": 0.3905,
"step": 930
},
{
"epoch": 1.1164,
"grad_norm": 0.6032191740051801,
"learning_rate": 7.917998838714019e-06,
"loss": 0.4023,
"step": 931
},
{
"epoch": 1.1176,
"grad_norm": 0.6257951900611615,
"learning_rate": 7.91232939841252e-06,
"loss": 0.3718,
"step": 932
},
{
"epoch": 1.1188,
"grad_norm": 0.5722970175151099,
"learning_rate": 7.906654285419347e-06,
"loss": 0.3609,
"step": 933
},
{
"epoch": 1.12,
"grad_norm": 0.8443882209730433,
"learning_rate": 7.900973510788595e-06,
"loss": 0.4205,
"step": 934
},
{
"epoch": 1.1212,
"grad_norm": 0.6536057882938714,
"learning_rate": 7.895287085585386e-06,
"loss": 0.4226,
"step": 935
},
{
"epoch": 1.1224,
"grad_norm": 0.6275687443932316,
"learning_rate": 7.889595020885853e-06,
"loss": 0.3833,
"step": 936
},
{
"epoch": 1.1236,
"grad_norm": 0.6708387846357547,
"learning_rate": 7.883897327777108e-06,
"loss": 0.3703,
"step": 937
},
{
"epoch": 1.1248,
"grad_norm": 0.557455511462114,
"learning_rate": 7.878194017357229e-06,
"loss": 0.3761,
"step": 938
},
{
"epoch": 1.126,
"grad_norm": 0.6825974121921209,
"learning_rate": 7.87248510073523e-06,
"loss": 0.3611,
"step": 939
},
{
"epoch": 1.1272,
"grad_norm": 0.6790308408992597,
"learning_rate": 7.866770589031057e-06,
"loss": 0.3928,
"step": 940
},
{
"epoch": 1.1284,
"grad_norm": 0.6309280591439027,
"learning_rate": 7.86105049337554e-06,
"loss": 0.3772,
"step": 941
},
{
"epoch": 1.1296,
"grad_norm": 0.48897990635157607,
"learning_rate": 7.855324824910395e-06,
"loss": 0.3747,
"step": 942
},
{
"epoch": 1.1308,
"grad_norm": 0.7680633972491443,
"learning_rate": 7.849593594788192e-06,
"loss": 0.3887,
"step": 943
},
{
"epoch": 1.1320000000000001,
"grad_norm": 0.700393461346577,
"learning_rate": 7.843856814172329e-06,
"loss": 0.4446,
"step": 944
},
{
"epoch": 1.1332,
"grad_norm": 0.6410912266603231,
"learning_rate": 7.83811449423702e-06,
"loss": 0.3871,
"step": 945
},
{
"epoch": 1.1344,
"grad_norm": 0.576018077563102,
"learning_rate": 7.832366646167268e-06,
"loss": 0.3697,
"step": 946
},
{
"epoch": 1.1356,
"grad_norm": 0.5828135175960664,
"learning_rate": 7.82661328115884e-06,
"loss": 0.3619,
"step": 947
},
{
"epoch": 1.1368,
"grad_norm": 0.6198683619402525,
"learning_rate": 7.820854410418255e-06,
"loss": 0.3734,
"step": 948
},
{
"epoch": 1.138,
"grad_norm": 0.5457064722309137,
"learning_rate": 7.815090045162752e-06,
"loss": 0.3895,
"step": 949
},
{
"epoch": 1.1392,
"grad_norm": 0.6253352691670119,
"learning_rate": 7.809320196620272e-06,
"loss": 0.3516,
"step": 950
},
{
"epoch": 1.1404,
"grad_norm": 0.46672264103831795,
"learning_rate": 7.80354487602944e-06,
"loss": 0.3616,
"step": 951
},
{
"epoch": 1.1416,
"grad_norm": 0.56828089422725,
"learning_rate": 7.797764094639537e-06,
"loss": 0.4059,
"step": 952
},
{
"epoch": 1.1428,
"grad_norm": 0.7154646105300466,
"learning_rate": 7.79197786371048e-06,
"loss": 0.4187,
"step": 953
},
{
"epoch": 1.144,
"grad_norm": 0.46637500730882037,
"learning_rate": 7.786186194512802e-06,
"loss": 0.3673,
"step": 954
},
{
"epoch": 1.1452,
"grad_norm": 0.6072905735893732,
"learning_rate": 7.780389098327629e-06,
"loss": 0.3819,
"step": 955
},
{
"epoch": 1.1464,
"grad_norm": 0.6043532507563107,
"learning_rate": 7.774586586446658e-06,
"loss": 0.3882,
"step": 956
},
{
"epoch": 1.1476,
"grad_norm": 0.5917199850834677,
"learning_rate": 7.768778670172135e-06,
"loss": 0.4069,
"step": 957
},
{
"epoch": 1.1488,
"grad_norm": 0.5379580116927375,
"learning_rate": 7.762965360816828e-06,
"loss": 0.3949,
"step": 958
},
{
"epoch": 1.15,
"grad_norm": 0.6510481738337858,
"learning_rate": 7.757146669704016e-06,
"loss": 0.3803,
"step": 959
},
{
"epoch": 1.1512,
"grad_norm": 0.5541510150757831,
"learning_rate": 7.75132260816746e-06,
"loss": 0.3924,
"step": 960
},
{
"epoch": 1.1524,
"grad_norm": 0.5678044715290077,
"learning_rate": 7.745493187551378e-06,
"loss": 0.3413,
"step": 961
},
{
"epoch": 1.1536,
"grad_norm": 0.5214786220289697,
"learning_rate": 7.739658419210429e-06,
"loss": 0.3447,
"step": 962
},
{
"epoch": 1.1548,
"grad_norm": 0.5553831003825472,
"learning_rate": 7.733818314509689e-06,
"loss": 0.3828,
"step": 963
},
{
"epoch": 1.156,
"grad_norm": 0.5394612049022063,
"learning_rate": 7.727972884824625e-06,
"loss": 0.3932,
"step": 964
},
{
"epoch": 1.1572,
"grad_norm": 0.5718005170648759,
"learning_rate": 7.72212214154108e-06,
"loss": 0.3618,
"step": 965
},
{
"epoch": 1.1584,
"grad_norm": 0.5460985344820174,
"learning_rate": 7.716266096055243e-06,
"loss": 0.3664,
"step": 966
},
{
"epoch": 1.1596,
"grad_norm": 0.5277059242679342,
"learning_rate": 7.710404759773637e-06,
"loss": 0.3593,
"step": 967
},
{
"epoch": 1.1608,
"grad_norm": 0.4587152815171526,
"learning_rate": 7.704538144113082e-06,
"loss": 0.387,
"step": 968
},
{
"epoch": 1.162,
"grad_norm": 0.5779733417720779,
"learning_rate": 7.698666260500688e-06,
"loss": 0.431,
"step": 969
},
{
"epoch": 1.1632,
"grad_norm": 0.6005822867829625,
"learning_rate": 7.692789120373824e-06,
"loss": 0.363,
"step": 970
},
{
"epoch": 1.1644,
"grad_norm": 0.5935009091208785,
"learning_rate": 7.686906735180099e-06,
"loss": 0.3753,
"step": 971
},
{
"epoch": 1.1656,
"grad_norm": 0.5669008223119324,
"learning_rate": 7.681019116377331e-06,
"loss": 0.3478,
"step": 972
},
{
"epoch": 1.1668,
"grad_norm": 0.5573181478984236,
"learning_rate": 7.675126275433545e-06,
"loss": 0.3835,
"step": 973
},
{
"epoch": 1.168,
"grad_norm": 0.5473700764940981,
"learning_rate": 7.669228223826926e-06,
"loss": 0.384,
"step": 974
},
{
"epoch": 1.1692,
"grad_norm": 0.5107739763185454,
"learning_rate": 7.663324973045818e-06,
"loss": 0.3832,
"step": 975
},
{
"epoch": 1.1703999999999999,
"grad_norm": 0.6102163681788227,
"learning_rate": 7.657416534588683e-06,
"loss": 0.3549,
"step": 976
},
{
"epoch": 1.1716,
"grad_norm": 0.5532021885497728,
"learning_rate": 7.651502919964092e-06,
"loss": 0.3714,
"step": 977
},
{
"epoch": 1.1728,
"grad_norm": 0.5185588376669081,
"learning_rate": 7.645584140690702e-06,
"loss": 0.378,
"step": 978
},
{
"epoch": 1.174,
"grad_norm": 0.5389544442206081,
"learning_rate": 7.639660208297221e-06,
"loss": 0.3629,
"step": 979
},
{
"epoch": 1.1752,
"grad_norm": 0.5796224788639072,
"learning_rate": 7.633731134322404e-06,
"loss": 0.3561,
"step": 980
},
{
"epoch": 1.1764000000000001,
"grad_norm": 0.5724657571915197,
"learning_rate": 7.6277969303150155e-06,
"loss": 0.413,
"step": 981
},
{
"epoch": 1.1776,
"grad_norm": 0.5651233783215182,
"learning_rate": 7.6218576078338115e-06,
"loss": 0.3978,
"step": 982
},
{
"epoch": 1.1788,
"grad_norm": 0.704877889693732,
"learning_rate": 7.615913178447518e-06,
"loss": 0.4037,
"step": 983
},
{
"epoch": 1.18,
"grad_norm": 0.5952367164096336,
"learning_rate": 7.609963653734814e-06,
"loss": 0.3816,
"step": 984
},
{
"epoch": 1.1812,
"grad_norm": 0.5324116366363447,
"learning_rate": 7.604009045284295e-06,
"loss": 0.3808,
"step": 985
},
{
"epoch": 1.1824,
"grad_norm": 0.6910948988868738,
"learning_rate": 7.598049364694466e-06,
"loss": 0.4147,
"step": 986
},
{
"epoch": 1.1836,
"grad_norm": 0.5296354520478062,
"learning_rate": 7.592084623573708e-06,
"loss": 0.3543,
"step": 987
},
{
"epoch": 1.1848,
"grad_norm": 0.5566407528272007,
"learning_rate": 7.586114833540257e-06,
"loss": 0.353,
"step": 988
},
{
"epoch": 1.186,
"grad_norm": 0.6015296281547915,
"learning_rate": 7.58014000622219e-06,
"loss": 0.3453,
"step": 989
},
{
"epoch": 1.1872,
"grad_norm": 0.6228830366597151,
"learning_rate": 7.574160153257386e-06,
"loss": 0.3881,
"step": 990
},
{
"epoch": 1.1884000000000001,
"grad_norm": 0.6334169507119521,
"learning_rate": 7.568175286293522e-06,
"loss": 0.392,
"step": 991
},
{
"epoch": 1.1896,
"grad_norm": 0.6236183748838798,
"learning_rate": 7.562185416988039e-06,
"loss": 0.3918,
"step": 992
},
{
"epoch": 1.1908,
"grad_norm": 0.7794773133283003,
"learning_rate": 7.556190557008116e-06,
"loss": 0.4051,
"step": 993
},
{
"epoch": 1.192,
"grad_norm": 0.7274930809023531,
"learning_rate": 7.550190718030663e-06,
"loss": 0.3829,
"step": 994
},
{
"epoch": 1.1932,
"grad_norm": 0.5707170114861568,
"learning_rate": 7.54418591174228e-06,
"loss": 0.3346,
"step": 995
},
{
"epoch": 1.1944,
"grad_norm": 0.6243157977136253,
"learning_rate": 7.5381761498392435e-06,
"loss": 0.3908,
"step": 996
},
{
"epoch": 1.1956,
"grad_norm": 0.4674660238025524,
"learning_rate": 7.532161444027488e-06,
"loss": 0.3703,
"step": 997
},
{
"epoch": 1.1968,
"grad_norm": 0.6270586839157299,
"learning_rate": 7.526141806022571e-06,
"loss": 0.3858,
"step": 998
},
{
"epoch": 1.198,
"grad_norm": 0.6317796196938065,
"learning_rate": 7.520117247549661e-06,
"loss": 0.3841,
"step": 999
},
{
"epoch": 1.1992,
"grad_norm": 0.6180551084846545,
"learning_rate": 7.514087780343511e-06,
"loss": 0.3466,
"step": 1000
},
{
"epoch": 1.2004,
"grad_norm": 0.543272841474728,
"learning_rate": 7.508053416148433e-06,
"loss": 0.4188,
"step": 1001
},
{
"epoch": 1.2016,
"grad_norm": 0.7036612686305553,
"learning_rate": 7.502014166718279e-06,
"loss": 0.3691,
"step": 1002
},
{
"epoch": 1.2028,
"grad_norm": 0.6190104178533024,
"learning_rate": 7.495970043816416e-06,
"loss": 0.3798,
"step": 1003
},
{
"epoch": 1.204,
"grad_norm": 0.6451355580120752,
"learning_rate": 7.489921059215703e-06,
"loss": 0.3796,
"step": 1004
},
{
"epoch": 1.2052,
"grad_norm": 0.5804373693313714,
"learning_rate": 7.483867224698471e-06,
"loss": 0.4045,
"step": 1005
},
{
"epoch": 1.2064,
"grad_norm": 0.5248356312926055,
"learning_rate": 7.477808552056496e-06,
"loss": 0.3557,
"step": 1006
},
{
"epoch": 1.2076,
"grad_norm": 0.4393039138466907,
"learning_rate": 7.471745053090976e-06,
"loss": 0.363,
"step": 1007
},
{
"epoch": 1.2088,
"grad_norm": 0.5104233947455193,
"learning_rate": 7.465676739612514e-06,
"loss": 0.3686,
"step": 1008
},
{
"epoch": 1.21,
"grad_norm": 0.5918547266203569,
"learning_rate": 7.45960362344109e-06,
"loss": 0.4017,
"step": 1009
},
{
"epoch": 1.2112,
"grad_norm": 0.6442980718898021,
"learning_rate": 7.4535257164060324e-06,
"loss": 0.3556,
"step": 1010
},
{
"epoch": 1.2124,
"grad_norm": 0.539978003071063,
"learning_rate": 7.447443030346011e-06,
"loss": 0.3552,
"step": 1011
},
{
"epoch": 1.2136,
"grad_norm": 0.5167136332720779,
"learning_rate": 7.441355577108998e-06,
"loss": 0.3099,
"step": 1012
},
{
"epoch": 1.2147999999999999,
"grad_norm": 0.547961404686605,
"learning_rate": 7.4352633685522535e-06,
"loss": 0.3788,
"step": 1013
},
{
"epoch": 1.216,
"grad_norm": 0.5739201487486998,
"learning_rate": 7.4291664165422985e-06,
"loss": 0.3749,
"step": 1014
},
{
"epoch": 1.2172,
"grad_norm": 0.6136292174850035,
"learning_rate": 7.423064732954895e-06,
"loss": 0.3518,
"step": 1015
},
{
"epoch": 1.2184,
"grad_norm": 0.6084805729019309,
"learning_rate": 7.4169583296750194e-06,
"loss": 0.367,
"step": 1016
},
{
"epoch": 1.2196,
"grad_norm": 0.6012984604908796,
"learning_rate": 7.410847218596846e-06,
"loss": 0.3727,
"step": 1017
},
{
"epoch": 1.2208,
"grad_norm": 0.5630545476779787,
"learning_rate": 7.40473141162371e-06,
"loss": 0.3787,
"step": 1018
},
{
"epoch": 1.222,
"grad_norm": 0.5608030101506133,
"learning_rate": 7.398610920668102e-06,
"loss": 0.3708,
"step": 1019
},
{
"epoch": 1.2232,
"grad_norm": 0.658915854539811,
"learning_rate": 7.392485757651634e-06,
"loss": 0.3602,
"step": 1020
},
{
"epoch": 1.2244,
"grad_norm": 0.5656184716996064,
"learning_rate": 7.386355934505015e-06,
"loss": 0.3878,
"step": 1021
},
{
"epoch": 1.2256,
"grad_norm": 0.598806868596272,
"learning_rate": 7.380221463168036e-06,
"loss": 0.3574,
"step": 1022
},
{
"epoch": 1.2268,
"grad_norm": 0.6170648083398085,
"learning_rate": 7.374082355589536e-06,
"loss": 0.4131,
"step": 1023
},
{
"epoch": 1.228,
"grad_norm": 0.6838329966789305,
"learning_rate": 7.367938623727389e-06,
"loss": 0.3419,
"step": 1024
},
{
"epoch": 1.2292,
"grad_norm": 0.6205390501504368,
"learning_rate": 7.361790279548476e-06,
"loss": 0.4052,
"step": 1025
},
{
"epoch": 1.2304,
"grad_norm": 0.6888179942809703,
"learning_rate": 7.35563733502866e-06,
"loss": 0.4047,
"step": 1026
},
{
"epoch": 1.2316,
"grad_norm": 0.4939596619401398,
"learning_rate": 7.3494798021527665e-06,
"loss": 0.3842,
"step": 1027
},
{
"epoch": 1.2328000000000001,
"grad_norm": 0.7400721543380634,
"learning_rate": 7.3433176929145574e-06,
"loss": 0.3556,
"step": 1028
},
{
"epoch": 1.234,
"grad_norm": 0.6168688148164273,
"learning_rate": 7.337151019316708e-06,
"loss": 0.4099,
"step": 1029
},
{
"epoch": 1.2352,
"grad_norm": 0.8003583257346282,
"learning_rate": 7.330979793370784e-06,
"loss": 0.3968,
"step": 1030
},
{
"epoch": 1.2364,
"grad_norm": 0.47326526074914016,
"learning_rate": 7.324804027097221e-06,
"loss": 0.3813,
"step": 1031
},
{
"epoch": 1.2376,
"grad_norm": 0.5472467555562863,
"learning_rate": 7.318623732525294e-06,
"loss": 0.3877,
"step": 1032
},
{
"epoch": 1.2388,
"grad_norm": 0.6594200360662721,
"learning_rate": 7.312438921693101e-06,
"loss": 0.3954,
"step": 1033
},
{
"epoch": 1.24,
"grad_norm": 0.6448566701959111,
"learning_rate": 7.30624960664754e-06,
"loss": 0.3618,
"step": 1034
},
{
"epoch": 1.2412,
"grad_norm": 0.5416473187116394,
"learning_rate": 7.300055799444273e-06,
"loss": 0.3671,
"step": 1035
},
{
"epoch": 1.2424,
"grad_norm": 0.5737841757940133,
"learning_rate": 7.293857512147723e-06,
"loss": 0.3956,
"step": 1036
},
{
"epoch": 1.2436,
"grad_norm": 0.7038531065019307,
"learning_rate": 7.287654756831031e-06,
"loss": 0.3735,
"step": 1037
},
{
"epoch": 1.2448,
"grad_norm": 0.5648836478829721,
"learning_rate": 7.2814475455760445e-06,
"loss": 0.337,
"step": 1038
},
{
"epoch": 1.246,
"grad_norm": 0.5734291395414816,
"learning_rate": 7.275235890473291e-06,
"loss": 0.3759,
"step": 1039
},
{
"epoch": 1.2472,
"grad_norm": 0.535389124757447,
"learning_rate": 7.269019803621953e-06,
"loss": 0.3489,
"step": 1040
},
{
"epoch": 1.2484,
"grad_norm": 0.6260068951532562,
"learning_rate": 7.262799297129843e-06,
"loss": 0.356,
"step": 1041
},
{
"epoch": 1.2496,
"grad_norm": 0.5626553670499461,
"learning_rate": 7.256574383113386e-06,
"loss": 0.3432,
"step": 1042
},
{
"epoch": 1.2508,
"grad_norm": 0.5979399572770037,
"learning_rate": 7.25034507369759e-06,
"loss": 0.3963,
"step": 1043
},
{
"epoch": 1.252,
"grad_norm": 0.7627246495578786,
"learning_rate": 7.244111381016024e-06,
"loss": 0.3589,
"step": 1044
},
{
"epoch": 1.2532,
"grad_norm": 0.5924793390470331,
"learning_rate": 7.237873317210796e-06,
"loss": 0.3863,
"step": 1045
},
{
"epoch": 1.2544,
"grad_norm": 0.5618731305329973,
"learning_rate": 7.231630894432527e-06,
"loss": 0.3686,
"step": 1046
},
{
"epoch": 1.2556,
"grad_norm": 0.5484996166547617,
"learning_rate": 7.22538412484033e-06,
"loss": 0.3638,
"step": 1047
},
{
"epoch": 1.2568,
"grad_norm": 0.9948159562952937,
"learning_rate": 7.219133020601783e-06,
"loss": 0.4089,
"step": 1048
},
{
"epoch": 1.258,
"grad_norm": 0.6944598116037035,
"learning_rate": 7.2128775938929095e-06,
"loss": 0.3984,
"step": 1049
},
{
"epoch": 1.2591999999999999,
"grad_norm": 0.6757195579443828,
"learning_rate": 7.206617856898149e-06,
"loss": 0.3778,
"step": 1050
},
{
"epoch": 1.2604,
"grad_norm": 0.5920958068986326,
"learning_rate": 7.20035382181034e-06,
"loss": 0.3819,
"step": 1051
},
{
"epoch": 1.2616,
"grad_norm": 0.6417635133827952,
"learning_rate": 7.194085500830691e-06,
"loss": 0.403,
"step": 1052
},
{
"epoch": 1.2628,
"grad_norm": 0.5568650988903112,
"learning_rate": 7.1878129061687595e-06,
"loss": 0.3743,
"step": 1053
},
{
"epoch": 1.264,
"grad_norm": 0.6533788297396388,
"learning_rate": 7.181536050042427e-06,
"loss": 0.3706,
"step": 1054
},
{
"epoch": 1.2652,
"grad_norm": 0.5362349579796641,
"learning_rate": 7.175254944677874e-06,
"loss": 0.3644,
"step": 1055
},
{
"epoch": 1.2664,
"grad_norm": 0.5692213887579612,
"learning_rate": 7.1689696023095625e-06,
"loss": 0.3808,
"step": 1056
},
{
"epoch": 1.2676,
"grad_norm": 0.5795212902620743,
"learning_rate": 7.162680035180201e-06,
"loss": 0.3998,
"step": 1057
},
{
"epoch": 1.2688,
"grad_norm": 0.5209493310515746,
"learning_rate": 7.156386255540732e-06,
"loss": 0.3532,
"step": 1058
},
{
"epoch": 1.27,
"grad_norm": 0.5424829545715086,
"learning_rate": 7.150088275650302e-06,
"loss": 0.392,
"step": 1059
},
{
"epoch": 1.2711999999999999,
"grad_norm": 0.6464063223655275,
"learning_rate": 7.143786107776236e-06,
"loss": 0.3422,
"step": 1060
},
{
"epoch": 1.2724,
"grad_norm": 0.4813528214550613,
"learning_rate": 7.137479764194022e-06,
"loss": 0.3491,
"step": 1061
},
{
"epoch": 1.2736,
"grad_norm": 0.5800825959591446,
"learning_rate": 7.131169257187276e-06,
"loss": 0.377,
"step": 1062
},
{
"epoch": 1.2748,
"grad_norm": 0.4796442134420188,
"learning_rate": 7.1248545990477256e-06,
"loss": 0.4,
"step": 1063
},
{
"epoch": 1.276,
"grad_norm": 0.5534864415940866,
"learning_rate": 7.1185358020751875e-06,
"loss": 0.3917,
"step": 1064
},
{
"epoch": 1.2772000000000001,
"grad_norm": 0.5579998901657407,
"learning_rate": 7.112212878577533e-06,
"loss": 0.4256,
"step": 1065
},
{
"epoch": 1.2784,
"grad_norm": 0.6136435635336623,
"learning_rate": 7.1058858408706765e-06,
"loss": 0.358,
"step": 1066
},
{
"epoch": 1.2796,
"grad_norm": 0.5559228232636972,
"learning_rate": 7.099554701278547e-06,
"loss": 0.3811,
"step": 1067
},
{
"epoch": 1.2808,
"grad_norm": 0.5429673734481284,
"learning_rate": 7.093219472133059e-06,
"loss": 0.3952,
"step": 1068
},
{
"epoch": 1.282,
"grad_norm": 0.5013586384163251,
"learning_rate": 7.086880165774093e-06,
"loss": 0.3772,
"step": 1069
},
{
"epoch": 1.2832,
"grad_norm": 0.586078174565493,
"learning_rate": 7.080536794549477e-06,
"loss": 0.3987,
"step": 1070
},
{
"epoch": 1.2844,
"grad_norm": 0.5979682949821524,
"learning_rate": 7.0741893708149475e-06,
"loss": 0.3832,
"step": 1071
},
{
"epoch": 1.2856,
"grad_norm": 0.5450858560697013,
"learning_rate": 7.067837906934143e-06,
"loss": 0.3889,
"step": 1072
},
{
"epoch": 1.2868,
"grad_norm": 0.5251321928301862,
"learning_rate": 7.061482415278569e-06,
"loss": 0.3332,
"step": 1073
},
{
"epoch": 1.288,
"grad_norm": 0.5441331999246979,
"learning_rate": 7.055122908227571e-06,
"loss": 0.3732,
"step": 1074
},
{
"epoch": 1.2892000000000001,
"grad_norm": 0.623583931716984,
"learning_rate": 7.0487593981683246e-06,
"loss": 0.4046,
"step": 1075
},
{
"epoch": 1.2904,
"grad_norm": 0.5914739695772331,
"learning_rate": 7.042391897495795e-06,
"loss": 0.3659,
"step": 1076
},
{
"epoch": 1.2916,
"grad_norm": 0.4592194257354245,
"learning_rate": 7.036020418612724e-06,
"loss": 0.3283,
"step": 1077
},
{
"epoch": 1.2928,
"grad_norm": 0.5090800599047531,
"learning_rate": 7.029644973929604e-06,
"loss": 0.3703,
"step": 1078
},
{
"epoch": 1.294,
"grad_norm": 0.48246491923077817,
"learning_rate": 7.023265575864648e-06,
"loss": 0.4136,
"step": 1079
},
{
"epoch": 1.2952,
"grad_norm": 0.5743016554708543,
"learning_rate": 7.016882236843769e-06,
"loss": 0.3541,
"step": 1080
},
{
"epoch": 1.2964,
"grad_norm": 0.6693297930478648,
"learning_rate": 7.0104949693005645e-06,
"loss": 0.3687,
"step": 1081
},
{
"epoch": 1.2976,
"grad_norm": 0.45548226688407856,
"learning_rate": 7.00410378567627e-06,
"loss": 0.395,
"step": 1082
},
{
"epoch": 1.2988,
"grad_norm": 0.58335044226178,
"learning_rate": 6.997708698419765e-06,
"loss": 0.3464,
"step": 1083
},
{
"epoch": 1.3,
"grad_norm": 0.513898806820525,
"learning_rate": 6.99130971998752e-06,
"loss": 0.3728,
"step": 1084
},
{
"epoch": 1.3012000000000001,
"grad_norm": 0.6606013764400083,
"learning_rate": 6.98490686284359e-06,
"loss": 0.4032,
"step": 1085
},
{
"epoch": 1.3024,
"grad_norm": 0.5495677779568732,
"learning_rate": 6.978500139459583e-06,
"loss": 0.3893,
"step": 1086
},
{
"epoch": 1.3035999999999999,
"grad_norm": 0.6833083037662147,
"learning_rate": 6.972089562314644e-06,
"loss": 0.4113,
"step": 1087
},
{
"epoch": 1.3048,
"grad_norm": 0.5759574990861633,
"learning_rate": 6.9656751438954115e-06,
"loss": 0.3968,
"step": 1088
},
{
"epoch": 1.306,
"grad_norm": 0.5115832604155884,
"learning_rate": 6.959256896696021e-06,
"loss": 0.432,
"step": 1089
},
{
"epoch": 1.3072,
"grad_norm": 0.7911985158075923,
"learning_rate": 6.952834833218056e-06,
"loss": 0.3883,
"step": 1090
},
{
"epoch": 1.3084,
"grad_norm": 0.4993447860656908,
"learning_rate": 6.946408965970536e-06,
"loss": 0.3802,
"step": 1091
},
{
"epoch": 1.3096,
"grad_norm": 0.4400466240534706,
"learning_rate": 6.939979307469892e-06,
"loss": 0.3579,
"step": 1092
},
{
"epoch": 1.3108,
"grad_norm": 0.5685894710046548,
"learning_rate": 6.933545870239933e-06,
"loss": 0.3482,
"step": 1093
},
{
"epoch": 1.312,
"grad_norm": 0.583902836607969,
"learning_rate": 6.927108666811837e-06,
"loss": 0.3556,
"step": 1094
},
{
"epoch": 1.3132,
"grad_norm": 0.6469186054837391,
"learning_rate": 6.920667709724113e-06,
"loss": 0.4031,
"step": 1095
},
{
"epoch": 1.3144,
"grad_norm": 0.5135019797765831,
"learning_rate": 6.914223011522581e-06,
"loss": 0.3853,
"step": 1096
},
{
"epoch": 1.3155999999999999,
"grad_norm": 0.6132864607455859,
"learning_rate": 6.90777458476035e-06,
"loss": 0.3856,
"step": 1097
},
{
"epoch": 1.3168,
"grad_norm": 0.5260723911564891,
"learning_rate": 6.901322441997791e-06,
"loss": 0.3722,
"step": 1098
},
{
"epoch": 1.318,
"grad_norm": 0.5769135122339621,
"learning_rate": 6.894866595802509e-06,
"loss": 0.3929,
"step": 1099
},
{
"epoch": 1.3192,
"grad_norm": 0.5217068431943552,
"learning_rate": 6.888407058749331e-06,
"loss": 0.4039,
"step": 1100
},
{
"epoch": 1.3204,
"grad_norm": 0.5882564291044068,
"learning_rate": 6.881943843420268e-06,
"loss": 0.3685,
"step": 1101
},
{
"epoch": 1.3216,
"grad_norm": 0.5397540009109018,
"learning_rate": 6.875476962404495e-06,
"loss": 0.412,
"step": 1102
},
{
"epoch": 1.3228,
"grad_norm": 0.5426493873301287,
"learning_rate": 6.869006428298328e-06,
"loss": 0.374,
"step": 1103
},
{
"epoch": 1.324,
"grad_norm": 0.6918593386156987,
"learning_rate": 6.862532253705199e-06,
"loss": 0.3746,
"step": 1104
},
{
"epoch": 1.3252,
"grad_norm": 0.6622425191632842,
"learning_rate": 6.85605445123563e-06,
"loss": 0.3446,
"step": 1105
},
{
"epoch": 1.3264,
"grad_norm": 0.49566249352496705,
"learning_rate": 6.849573033507213e-06,
"loss": 0.3509,
"step": 1106
},
{
"epoch": 1.3276,
"grad_norm": 0.6281029533094057,
"learning_rate": 6.843088013144575e-06,
"loss": 0.3732,
"step": 1107
},
{
"epoch": 1.3288,
"grad_norm": 0.5339005742427865,
"learning_rate": 6.8365994027793695e-06,
"loss": 0.384,
"step": 1108
},
{
"epoch": 1.33,
"grad_norm": 0.6084133196447026,
"learning_rate": 6.830107215050232e-06,
"loss": 0.3865,
"step": 1109
},
{
"epoch": 1.3312,
"grad_norm": 0.5362040240692425,
"learning_rate": 6.823611462602777e-06,
"loss": 0.3574,
"step": 1110
},
{
"epoch": 1.3324,
"grad_norm": 0.5946541050024513,
"learning_rate": 6.817112158089554e-06,
"loss": 0.3923,
"step": 1111
},
{
"epoch": 1.3336000000000001,
"grad_norm": 0.6799270385641254,
"learning_rate": 6.8106093141700336e-06,
"loss": 0.3851,
"step": 1112
},
{
"epoch": 1.3348,
"grad_norm": 0.5447366160367654,
"learning_rate": 6.804102943510583e-06,
"loss": 0.3994,
"step": 1113
},
{
"epoch": 1.336,
"grad_norm": 0.7605032504216259,
"learning_rate": 6.797593058784437e-06,
"loss": 0.3779,
"step": 1114
},
{
"epoch": 1.3372,
"grad_norm": 0.44318202194361933,
"learning_rate": 6.791079672671677e-06,
"loss": 0.388,
"step": 1115
},
{
"epoch": 1.3384,
"grad_norm": 0.6594966278078546,
"learning_rate": 6.784562797859198e-06,
"loss": 0.3966,
"step": 1116
},
{
"epoch": 1.3396,
"grad_norm": 0.7278372919818764,
"learning_rate": 6.7780424470407004e-06,
"loss": 0.3981,
"step": 1117
},
{
"epoch": 1.3408,
"grad_norm": 0.6394652933854147,
"learning_rate": 6.771518632916645e-06,
"loss": 0.4393,
"step": 1118
},
{
"epoch": 1.342,
"grad_norm": 0.5205892552379503,
"learning_rate": 6.7649913681942455e-06,
"loss": 0.3487,
"step": 1119
},
{
"epoch": 1.3432,
"grad_norm": 0.6518308231968244,
"learning_rate": 6.758460665587437e-06,
"loss": 0.3918,
"step": 1120
},
{
"epoch": 1.3444,
"grad_norm": 0.5518171765062463,
"learning_rate": 6.751926537816846e-06,
"loss": 0.3874,
"step": 1121
},
{
"epoch": 1.3456000000000001,
"grad_norm": 0.5680320375122355,
"learning_rate": 6.745388997609774e-06,
"loss": 0.3748,
"step": 1122
},
{
"epoch": 1.3468,
"grad_norm": 0.5430526903558245,
"learning_rate": 6.738848057700169e-06,
"loss": 0.3537,
"step": 1123
},
{
"epoch": 1.3479999999999999,
"grad_norm": 0.5320208214764721,
"learning_rate": 6.732303730828601e-06,
"loss": 0.4052,
"step": 1124
},
{
"epoch": 1.3492,
"grad_norm": 0.6325336919288119,
"learning_rate": 6.725756029742234e-06,
"loss": 0.358,
"step": 1125
},
{
"epoch": 1.3504,
"grad_norm": 0.685515799326995,
"learning_rate": 6.7192049671948115e-06,
"loss": 0.3678,
"step": 1126
},
{
"epoch": 1.3516,
"grad_norm": 0.41920896936731294,
"learning_rate": 6.712650555946616e-06,
"loss": 0.3517,
"step": 1127
},
{
"epoch": 1.3528,
"grad_norm": 0.5881032868010501,
"learning_rate": 6.706092808764459e-06,
"loss": 0.3582,
"step": 1128
},
{
"epoch": 1.354,
"grad_norm": 0.6483636072262811,
"learning_rate": 6.699531738421648e-06,
"loss": 0.36,
"step": 1129
},
{
"epoch": 1.3552,
"grad_norm": 0.5061302602252388,
"learning_rate": 6.692967357697961e-06,
"loss": 0.3615,
"step": 1130
},
{
"epoch": 1.3564,
"grad_norm": 0.6545204627775875,
"learning_rate": 6.6863996793796286e-06,
"loss": 0.3633,
"step": 1131
},
{
"epoch": 1.3576,
"grad_norm": 0.44178673336282853,
"learning_rate": 6.6798287162593e-06,
"loss": 0.3496,
"step": 1132
},
{
"epoch": 1.3588,
"grad_norm": 0.5475017826386849,
"learning_rate": 6.6732544811360255e-06,
"loss": 0.3429,
"step": 1133
},
{
"epoch": 1.3599999999999999,
"grad_norm": 0.6564333403411471,
"learning_rate": 6.666676986815227e-06,
"loss": 0.3973,
"step": 1134
},
{
"epoch": 1.3612,
"grad_norm": 0.5747932244479428,
"learning_rate": 6.660096246108677e-06,
"loss": 0.3733,
"step": 1135
},
{
"epoch": 1.3624,
"grad_norm": 0.505681834990622,
"learning_rate": 6.653512271834468e-06,
"loss": 0.3694,
"step": 1136
},
{
"epoch": 1.3636,
"grad_norm": 0.5302728720800068,
"learning_rate": 6.646925076816994e-06,
"loss": 0.3804,
"step": 1137
},
{
"epoch": 1.3648,
"grad_norm": 0.6137108947228674,
"learning_rate": 6.640334673886921e-06,
"loss": 0.3591,
"step": 1138
},
{
"epoch": 1.366,
"grad_norm": 0.49332652913525415,
"learning_rate": 6.633741075881163e-06,
"loss": 0.3868,
"step": 1139
},
{
"epoch": 1.3672,
"grad_norm": 0.6555162595538057,
"learning_rate": 6.627144295642859e-06,
"loss": 0.3583,
"step": 1140
},
{
"epoch": 1.3684,
"grad_norm": 0.7103276751363496,
"learning_rate": 6.6205443460213445e-06,
"loss": 0.3558,
"step": 1141
},
{
"epoch": 1.3696,
"grad_norm": 0.5512752934225792,
"learning_rate": 6.613941239872129e-06,
"loss": 0.3512,
"step": 1142
},
{
"epoch": 1.3708,
"grad_norm": 0.502186227760979,
"learning_rate": 6.607334990056873e-06,
"loss": 0.3604,
"step": 1143
},
{
"epoch": 1.3719999999999999,
"grad_norm": 0.6927880455165364,
"learning_rate": 6.600725609443356e-06,
"loss": 0.4005,
"step": 1144
},
{
"epoch": 1.3732,
"grad_norm": 0.44954852300084996,
"learning_rate": 6.594113110905458e-06,
"loss": 0.3637,
"step": 1145
},
{
"epoch": 1.3744,
"grad_norm": 0.554225409355741,
"learning_rate": 6.587497507323132e-06,
"loss": 0.3559,
"step": 1146
},
{
"epoch": 1.3756,
"grad_norm": 0.4711345253979762,
"learning_rate": 6.580878811582379e-06,
"loss": 0.3701,
"step": 1147
},
{
"epoch": 1.3768,
"grad_norm": 0.4722647339207153,
"learning_rate": 6.574257036575224e-06,
"loss": 0.3597,
"step": 1148
},
{
"epoch": 1.3780000000000001,
"grad_norm": 0.5324128787221746,
"learning_rate": 6.567632195199686e-06,
"loss": 0.3456,
"step": 1149
},
{
"epoch": 1.3792,
"grad_norm": 0.588026631619211,
"learning_rate": 6.5610043003597615e-06,
"loss": 0.3792,
"step": 1150
},
{
"epoch": 1.3804,
"grad_norm": 1.0073668433242489,
"learning_rate": 6.554373364965392e-06,
"loss": 0.3638,
"step": 1151
},
{
"epoch": 1.3816,
"grad_norm": 0.6643904374505395,
"learning_rate": 6.547739401932443e-06,
"loss": 0.3575,
"step": 1152
},
{
"epoch": 1.3828,
"grad_norm": 0.5292843395135677,
"learning_rate": 6.541102424182676e-06,
"loss": 0.3811,
"step": 1153
},
{
"epoch": 1.384,
"grad_norm": 0.5932814470820361,
"learning_rate": 6.5344624446437234e-06,
"loss": 0.3827,
"step": 1154
},
{
"epoch": 1.3852,
"grad_norm": 0.4782195713849791,
"learning_rate": 6.527819476249066e-06,
"loss": 0.3692,
"step": 1155
},
{
"epoch": 1.3864,
"grad_norm": 0.6540604048296909,
"learning_rate": 6.521173531938011e-06,
"loss": 0.3641,
"step": 1156
},
{
"epoch": 1.3876,
"grad_norm": 0.4800644925234897,
"learning_rate": 6.514524624655654e-06,
"loss": 0.3472,
"step": 1157
},
{
"epoch": 1.3888,
"grad_norm": 0.5509070760157515,
"learning_rate": 6.507872767352863e-06,
"loss": 0.3396,
"step": 1158
},
{
"epoch": 1.3900000000000001,
"grad_norm": 0.5622639222865433,
"learning_rate": 6.50121797298626e-06,
"loss": 0.3962,
"step": 1159
},
{
"epoch": 1.3912,
"grad_norm": 0.5231156502515467,
"learning_rate": 6.494560254518179e-06,
"loss": 0.3661,
"step": 1160
},
{
"epoch": 1.3924,
"grad_norm": 0.5081529588369621,
"learning_rate": 6.487899624916654e-06,
"loss": 0.3483,
"step": 1161
},
{
"epoch": 1.3936,
"grad_norm": 0.5391745904936082,
"learning_rate": 6.481236097155389e-06,
"loss": 0.3793,
"step": 1162
},
{
"epoch": 1.3948,
"grad_norm": 0.6615024814667997,
"learning_rate": 6.4745696842137305e-06,
"loss": 0.4145,
"step": 1163
},
{
"epoch": 1.396,
"grad_norm": 0.6356413469309103,
"learning_rate": 6.467900399076651e-06,
"loss": 0.3496,
"step": 1164
},
{
"epoch": 1.3972,
"grad_norm": 0.5487989053399741,
"learning_rate": 6.461228254734711e-06,
"loss": 0.3639,
"step": 1165
},
{
"epoch": 1.3984,
"grad_norm": 0.5278545281830197,
"learning_rate": 6.454553264184041e-06,
"loss": 0.3506,
"step": 1166
},
{
"epoch": 1.3996,
"grad_norm": 0.5829669117942872,
"learning_rate": 6.447875440426319e-06,
"loss": 0.3787,
"step": 1167
},
{
"epoch": 1.4008,
"grad_norm": 0.5682179378418761,
"learning_rate": 6.441194796468739e-06,
"loss": 0.3667,
"step": 1168
},
{
"epoch": 1.4020000000000001,
"grad_norm": 0.4888397389782733,
"learning_rate": 6.434511345323988e-06,
"loss": 0.3634,
"step": 1169
},
{
"epoch": 1.4032,
"grad_norm": 0.530620756337448,
"learning_rate": 6.427825100010225e-06,
"loss": 0.3964,
"step": 1170
},
{
"epoch": 1.4043999999999999,
"grad_norm": 0.6363987713172962,
"learning_rate": 6.421136073551047e-06,
"loss": 0.3841,
"step": 1171
},
{
"epoch": 1.4056,
"grad_norm": 0.5568572793629121,
"learning_rate": 6.414444278975465e-06,
"loss": 0.3633,
"step": 1172
},
{
"epoch": 1.4068,
"grad_norm": 0.49553189194984,
"learning_rate": 6.407749729317892e-06,
"loss": 0.3976,
"step": 1173
},
{
"epoch": 1.408,
"grad_norm": 0.5728937899487624,
"learning_rate": 6.401052437618098e-06,
"loss": 0.387,
"step": 1174
},
{
"epoch": 1.4092,
"grad_norm": 0.6518243725832273,
"learning_rate": 6.394352416921201e-06,
"loss": 0.3849,
"step": 1175
},
{
"epoch": 1.4104,
"grad_norm": 0.6043382361141673,
"learning_rate": 6.387649680277629e-06,
"loss": 0.4052,
"step": 1176
},
{
"epoch": 1.4116,
"grad_norm": 0.5404652461010379,
"learning_rate": 6.380944240743101e-06,
"loss": 0.4085,
"step": 1177
},
{
"epoch": 1.4128,
"grad_norm": 0.579388596697836,
"learning_rate": 6.374236111378605e-06,
"loss": 0.3551,
"step": 1178
},
{
"epoch": 1.414,
"grad_norm": 0.5523484176312922,
"learning_rate": 6.3675253052503645e-06,
"loss": 0.3907,
"step": 1179
},
{
"epoch": 1.4152,
"grad_norm": 0.5561093182555912,
"learning_rate": 6.360811835429817e-06,
"loss": 0.4324,
"step": 1180
},
{
"epoch": 1.4163999999999999,
"grad_norm": 0.5535367271687733,
"learning_rate": 6.35409571499359e-06,
"loss": 0.3714,
"step": 1181
},
{
"epoch": 1.4176,
"grad_norm": 0.6083376911119112,
"learning_rate": 6.347376957023471e-06,
"loss": 0.3501,
"step": 1182
},
{
"epoch": 1.4188,
"grad_norm": 0.4654771508132885,
"learning_rate": 6.340655574606391e-06,
"loss": 0.4048,
"step": 1183
},
{
"epoch": 1.42,
"grad_norm": 0.7318252472693728,
"learning_rate": 6.333931580834387e-06,
"loss": 0.4087,
"step": 1184
},
{
"epoch": 1.4212,
"grad_norm": 0.5945086209964745,
"learning_rate": 6.327204988804583e-06,
"loss": 0.3741,
"step": 1185
},
{
"epoch": 1.4224,
"grad_norm": 0.6152246104938062,
"learning_rate": 6.320475811619167e-06,
"loss": 0.3861,
"step": 1186
},
{
"epoch": 1.4236,
"grad_norm": 0.545937458552413,
"learning_rate": 6.313744062385363e-06,
"loss": 0.3716,
"step": 1187
},
{
"epoch": 1.4248,
"grad_norm": 0.4799280381039969,
"learning_rate": 6.307009754215397e-06,
"loss": 0.3337,
"step": 1188
},
{
"epoch": 1.426,
"grad_norm": 0.5334617138994628,
"learning_rate": 6.30027290022649e-06,
"loss": 0.3791,
"step": 1189
},
{
"epoch": 1.4272,
"grad_norm": 0.5466526534649941,
"learning_rate": 6.2935335135408135e-06,
"loss": 0.3937,
"step": 1190
},
{
"epoch": 1.4284,
"grad_norm": 0.5904387210760845,
"learning_rate": 6.286791607285478e-06,
"loss": 0.3826,
"step": 1191
},
{
"epoch": 1.4296,
"grad_norm": 0.4768564178143852,
"learning_rate": 6.2800471945925e-06,
"loss": 0.3774,
"step": 1192
},
{
"epoch": 1.4308,
"grad_norm": 0.5895377413883869,
"learning_rate": 6.2733002885987734e-06,
"loss": 0.378,
"step": 1193
},
{
"epoch": 1.432,
"grad_norm": 0.691532734331932,
"learning_rate": 6.2665509024460554e-06,
"loss": 0.3676,
"step": 1194
},
{
"epoch": 1.4332,
"grad_norm": 0.6050439284597378,
"learning_rate": 6.259799049280932e-06,
"loss": 0.3932,
"step": 1195
},
{
"epoch": 1.4344000000000001,
"grad_norm": 0.5905694488707581,
"learning_rate": 6.253044742254791e-06,
"loss": 0.3585,
"step": 1196
},
{
"epoch": 1.4356,
"grad_norm": 0.5868666557800802,
"learning_rate": 6.246287994523805e-06,
"loss": 0.3688,
"step": 1197
},
{
"epoch": 1.4368,
"grad_norm": 0.4676587208954629,
"learning_rate": 6.239528819248898e-06,
"loss": 0.3975,
"step": 1198
},
{
"epoch": 1.438,
"grad_norm": 0.5699774072418529,
"learning_rate": 6.232767229595719e-06,
"loss": 0.382,
"step": 1199
},
{
"epoch": 1.4392,
"grad_norm": 0.5678736625348831,
"learning_rate": 6.226003238734628e-06,
"loss": 0.3743,
"step": 1200
},
{
"epoch": 1.4404,
"grad_norm": 0.6577077891365315,
"learning_rate": 6.219236859840656e-06,
"loss": 0.361,
"step": 1201
},
{
"epoch": 1.4416,
"grad_norm": 0.5749292581736202,
"learning_rate": 6.2124681060934866e-06,
"loss": 0.358,
"step": 1202
},
{
"epoch": 1.4428,
"grad_norm": 0.590931833530579,
"learning_rate": 6.205696990677431e-06,
"loss": 0.3669,
"step": 1203
},
{
"epoch": 1.444,
"grad_norm": 0.5397996359057199,
"learning_rate": 6.1989235267813964e-06,
"loss": 0.3555,
"step": 1204
},
{
"epoch": 1.4452,
"grad_norm": 0.5743513481789311,
"learning_rate": 6.192147727598869e-06,
"loss": 0.421,
"step": 1205
},
{
"epoch": 1.4464000000000001,
"grad_norm": 0.505200072259294,
"learning_rate": 6.185369606327882e-06,
"loss": 0.359,
"step": 1206
},
{
"epoch": 1.4476,
"grad_norm": 0.5937106634500684,
"learning_rate": 6.178589176170991e-06,
"loss": 0.3997,
"step": 1207
},
{
"epoch": 1.4487999999999999,
"grad_norm": 0.5052831659088494,
"learning_rate": 6.171806450335248e-06,
"loss": 0.4214,
"step": 1208
},
{
"epoch": 1.45,
"grad_norm": 0.5918053967047753,
"learning_rate": 6.16502144203218e-06,
"loss": 0.4002,
"step": 1209
},
{
"epoch": 1.4512,
"grad_norm": 0.5443590614450066,
"learning_rate": 6.1582341644777575e-06,
"loss": 0.378,
"step": 1210
},
{
"epoch": 1.4524,
"grad_norm": 0.42555020008139566,
"learning_rate": 6.151444630892372e-06,
"loss": 0.3693,
"step": 1211
},
{
"epoch": 1.4536,
"grad_norm": 0.720412574975775,
"learning_rate": 6.144652854500806e-06,
"loss": 0.3938,
"step": 1212
},
{
"epoch": 1.4548,
"grad_norm": 0.4724926217608286,
"learning_rate": 6.1378588485322145e-06,
"loss": 0.3823,
"step": 1213
},
{
"epoch": 1.456,
"grad_norm": 0.5802120718021628,
"learning_rate": 6.131062626220094e-06,
"loss": 0.3461,
"step": 1214
},
{
"epoch": 1.4572,
"grad_norm": 0.5480554564104493,
"learning_rate": 6.124264200802259e-06,
"loss": 0.4336,
"step": 1215
},
{
"epoch": 1.4584,
"grad_norm": 0.6443331793045611,
"learning_rate": 6.117463585520813e-06,
"loss": 0.3841,
"step": 1216
},
{
"epoch": 1.4596,
"grad_norm": 0.711519411508646,
"learning_rate": 6.110660793622127e-06,
"loss": 0.4117,
"step": 1217
},
{
"epoch": 1.4607999999999999,
"grad_norm": 0.45866436123572735,
"learning_rate": 6.103855838356813e-06,
"loss": 0.3785,
"step": 1218
},
{
"epoch": 1.462,
"grad_norm": 0.5538669654532735,
"learning_rate": 6.097048732979691e-06,
"loss": 0.4143,
"step": 1219
},
{
"epoch": 1.4632,
"grad_norm": 0.5571894823619101,
"learning_rate": 6.090239490749775e-06,
"loss": 0.3716,
"step": 1220
},
{
"epoch": 1.4644,
"grad_norm": 0.5089942930991793,
"learning_rate": 6.083428124930239e-06,
"loss": 0.3669,
"step": 1221
},
{
"epoch": 1.4656,
"grad_norm": 0.6702683351874047,
"learning_rate": 6.076614648788392e-06,
"loss": 0.372,
"step": 1222
},
{
"epoch": 1.4668,
"grad_norm": 0.5794656181086777,
"learning_rate": 6.069799075595658e-06,
"loss": 0.3874,
"step": 1223
},
{
"epoch": 1.468,
"grad_norm": 0.5468499297531119,
"learning_rate": 6.062981418627539e-06,
"loss": 0.3573,
"step": 1224
},
{
"epoch": 1.4692,
"grad_norm": 0.5438151205713972,
"learning_rate": 6.0561616911636025e-06,
"loss": 0.3545,
"step": 1225
},
{
"epoch": 1.4704,
"grad_norm": 0.4578732396242083,
"learning_rate": 6.049339906487443e-06,
"loss": 0.3697,
"step": 1226
},
{
"epoch": 1.4716,
"grad_norm": 0.5676454535315087,
"learning_rate": 6.042516077886669e-06,
"loss": 0.3554,
"step": 1227
},
{
"epoch": 1.4727999999999999,
"grad_norm": 0.46462267233651716,
"learning_rate": 6.035690218652861e-06,
"loss": 0.4058,
"step": 1228
},
{
"epoch": 1.474,
"grad_norm": 0.63335218585527,
"learning_rate": 6.028862342081564e-06,
"loss": 0.3613,
"step": 1229
},
{
"epoch": 1.4752,
"grad_norm": 0.5392835735551683,
"learning_rate": 6.022032461472247e-06,
"loss": 0.3428,
"step": 1230
},
{
"epoch": 1.4764,
"grad_norm": 0.5238389880445473,
"learning_rate": 6.015200590128284e-06,
"loss": 0.3897,
"step": 1231
},
{
"epoch": 1.4776,
"grad_norm": 0.6032296929764723,
"learning_rate": 6.008366741356926e-06,
"loss": 0.3808,
"step": 1232
},
{
"epoch": 1.4788000000000001,
"grad_norm": 0.5117870671101838,
"learning_rate": 6.001530928469277e-06,
"loss": 0.361,
"step": 1233
},
{
"epoch": 1.48,
"grad_norm": 0.5560766032265853,
"learning_rate": 5.9946931647802645e-06,
"loss": 0.3973,
"step": 1234
},
{
"epoch": 1.4812,
"grad_norm": 0.44978913649649527,
"learning_rate": 5.987853463608618e-06,
"loss": 0.333,
"step": 1235
},
{
"epoch": 1.4824,
"grad_norm": 0.557611973248781,
"learning_rate": 5.981011838276841e-06,
"loss": 0.4082,
"step": 1236
},
{
"epoch": 1.4836,
"grad_norm": 0.4976072856197252,
"learning_rate": 5.974168302111183e-06,
"loss": 0.3798,
"step": 1237
},
{
"epoch": 1.4848,
"grad_norm": 0.5403150413190907,
"learning_rate": 5.967322868441616e-06,
"loss": 0.3709,
"step": 1238
},
{
"epoch": 1.486,
"grad_norm": 0.5377355735149714,
"learning_rate": 5.9604755506018105e-06,
"loss": 0.36,
"step": 1239
},
{
"epoch": 1.4872,
"grad_norm": 0.526622078545486,
"learning_rate": 5.953626361929102e-06,
"loss": 0.3785,
"step": 1240
},
{
"epoch": 1.4884,
"grad_norm": 0.7084647792247337,
"learning_rate": 5.946775315764476e-06,
"loss": 0.3774,
"step": 1241
},
{
"epoch": 1.4896,
"grad_norm": 0.6306590790537886,
"learning_rate": 5.939922425452531e-06,
"loss": 0.3807,
"step": 1242
},
{
"epoch": 1.4908000000000001,
"grad_norm": 0.5364226942594238,
"learning_rate": 5.93306770434146e-06,
"loss": 0.3713,
"step": 1243
},
{
"epoch": 1.492,
"grad_norm": 0.5430047001467209,
"learning_rate": 5.926211165783021e-06,
"loss": 0.4211,
"step": 1244
},
{
"epoch": 1.4932,
"grad_norm": 0.44567637661850384,
"learning_rate": 5.919352823132515e-06,
"loss": 0.3768,
"step": 1245
},
{
"epoch": 1.4944,
"grad_norm": 0.5882164379790376,
"learning_rate": 5.912492689748753e-06,
"loss": 0.3286,
"step": 1246
},
{
"epoch": 1.4956,
"grad_norm": 0.6135999835439079,
"learning_rate": 5.905630778994036e-06,
"loss": 0.4097,
"step": 1247
},
{
"epoch": 1.4968,
"grad_norm": 0.49445597722429835,
"learning_rate": 5.898767104234128e-06,
"loss": 0.3304,
"step": 1248
},
{
"epoch": 1.498,
"grad_norm": 0.7115505037101663,
"learning_rate": 5.891901678838227e-06,
"loss": 0.3749,
"step": 1249
},
{
"epoch": 1.4992,
"grad_norm": 0.5715925192085276,
"learning_rate": 5.88503451617894e-06,
"loss": 0.3844,
"step": 1250
},
{
"epoch": 1.5004,
"grad_norm": 0.5748426571859289,
"learning_rate": 5.878165629632262e-06,
"loss": 0.3686,
"step": 1251
},
{
"epoch": 1.5016,
"grad_norm": 0.4827515224905166,
"learning_rate": 5.8712950325775416e-06,
"loss": 0.3284,
"step": 1252
},
{
"epoch": 1.5028000000000001,
"grad_norm": 0.5379573591619449,
"learning_rate": 5.864422738397465e-06,
"loss": 0.3576,
"step": 1253
},
{
"epoch": 1.504,
"grad_norm": 0.5425509909802786,
"learning_rate": 5.857548760478015e-06,
"loss": 0.402,
"step": 1254
},
{
"epoch": 1.5051999999999999,
"grad_norm": 0.4720264422543957,
"learning_rate": 5.85067311220846e-06,
"loss": 0.3801,
"step": 1255
},
{
"epoch": 1.5064,
"grad_norm": 0.6050179214602931,
"learning_rate": 5.843795806981325e-06,
"loss": 0.3527,
"step": 1256
},
{
"epoch": 1.5076,
"grad_norm": 0.5509117101205769,
"learning_rate": 5.836916858192353e-06,
"loss": 0.3828,
"step": 1257
},
{
"epoch": 1.5088,
"grad_norm": 0.5304817046842542,
"learning_rate": 5.830036279240497e-06,
"loss": 0.3736,
"step": 1258
},
{
"epoch": 1.51,
"grad_norm": 0.5683967543341494,
"learning_rate": 5.823154083527884e-06,
"loss": 0.3534,
"step": 1259
},
{
"epoch": 1.5112,
"grad_norm": 0.5114190307881069,
"learning_rate": 5.816270284459783e-06,
"loss": 0.3914,
"step": 1260
},
{
"epoch": 1.5124,
"grad_norm": 0.6098036240699997,
"learning_rate": 5.809384895444594e-06,
"loss": 0.3745,
"step": 1261
},
{
"epoch": 1.5135999999999998,
"grad_norm": 0.5408809868247507,
"learning_rate": 5.802497929893813e-06,
"loss": 0.3909,
"step": 1262
},
{
"epoch": 1.5148000000000001,
"grad_norm": 0.5848579955188301,
"learning_rate": 5.795609401222001e-06,
"loss": 0.3784,
"step": 1263
},
{
"epoch": 1.516,
"grad_norm": 0.6277636788518177,
"learning_rate": 5.7887193228467735e-06,
"loss": 0.3578,
"step": 1264
},
{
"epoch": 1.5171999999999999,
"grad_norm": 0.4668198224626426,
"learning_rate": 5.781827708188753e-06,
"loss": 0.3718,
"step": 1265
},
{
"epoch": 1.5184,
"grad_norm": 0.49948272590176307,
"learning_rate": 5.774934570671562e-06,
"loss": 0.3824,
"step": 1266
},
{
"epoch": 1.5196,
"grad_norm": 0.5680796315248136,
"learning_rate": 5.768039923721791e-06,
"loss": 0.3506,
"step": 1267
},
{
"epoch": 1.5208,
"grad_norm": 0.5448127204775214,
"learning_rate": 5.761143780768962e-06,
"loss": 0.413,
"step": 1268
},
{
"epoch": 1.522,
"grad_norm": 0.420450639517775,
"learning_rate": 5.7542461552455165e-06,
"loss": 0.3798,
"step": 1269
},
{
"epoch": 1.5232,
"grad_norm": 0.6781585974401322,
"learning_rate": 5.747347060586787e-06,
"loss": 0.3766,
"step": 1270
},
{
"epoch": 1.5244,
"grad_norm": 0.5288110386339663,
"learning_rate": 5.740446510230959e-06,
"loss": 0.375,
"step": 1271
},
{
"epoch": 1.5255999999999998,
"grad_norm": 0.46475178742662154,
"learning_rate": 5.733544517619064e-06,
"loss": 0.3924,
"step": 1272
},
{
"epoch": 1.5268000000000002,
"grad_norm": 0.6725064950199741,
"learning_rate": 5.726641096194932e-06,
"loss": 0.38,
"step": 1273
},
{
"epoch": 1.528,
"grad_norm": 0.551028126958213,
"learning_rate": 5.719736259405182e-06,
"loss": 0.3854,
"step": 1274
},
{
"epoch": 1.5292,
"grad_norm": 0.5247630115155179,
"learning_rate": 5.7128300206991925e-06,
"loss": 0.3703,
"step": 1275
},
{
"epoch": 1.5304,
"grad_norm": 0.6214543397495828,
"learning_rate": 5.705922393529065e-06,
"loss": 0.3604,
"step": 1276
},
{
"epoch": 1.5316,
"grad_norm": 0.5078068129567174,
"learning_rate": 5.69901339134961e-06,
"loss": 0.3787,
"step": 1277
},
{
"epoch": 1.5328,
"grad_norm": 0.6304781199299307,
"learning_rate": 5.692103027618321e-06,
"loss": 0.3531,
"step": 1278
},
{
"epoch": 1.534,
"grad_norm": 0.5860745608692262,
"learning_rate": 5.685191315795331e-06,
"loss": 0.3994,
"step": 1279
},
{
"epoch": 1.5352000000000001,
"grad_norm": 0.4848169895854493,
"learning_rate": 5.678278269343411e-06,
"loss": 0.3657,
"step": 1280
},
{
"epoch": 1.5364,
"grad_norm": 0.529871229005269,
"learning_rate": 5.671363901727927e-06,
"loss": 0.3635,
"step": 1281
},
{
"epoch": 1.5375999999999999,
"grad_norm": 0.4786421442271292,
"learning_rate": 5.664448226416815e-06,
"loss": 0.3311,
"step": 1282
},
{
"epoch": 1.5388,
"grad_norm": 0.6114626793926188,
"learning_rate": 5.657531256880565e-06,
"loss": 0.3952,
"step": 1283
},
{
"epoch": 1.54,
"grad_norm": 0.5310963476303867,
"learning_rate": 5.650613006592185e-06,
"loss": 0.3822,
"step": 1284
},
{
"epoch": 1.5412,
"grad_norm": 0.5258123472073719,
"learning_rate": 5.643693489027172e-06,
"loss": 0.3509,
"step": 1285
},
{
"epoch": 1.5424,
"grad_norm": 0.6791198162986184,
"learning_rate": 5.636772717663501e-06,
"loss": 0.4113,
"step": 1286
},
{
"epoch": 1.5436,
"grad_norm": 0.4858757606096244,
"learning_rate": 5.629850705981584e-06,
"loss": 0.3503,
"step": 1287
},
{
"epoch": 1.5448,
"grad_norm": 0.5987150490357991,
"learning_rate": 5.622927467464247e-06,
"loss": 0.3589,
"step": 1288
},
{
"epoch": 1.546,
"grad_norm": 0.4802188661179889,
"learning_rate": 5.6160030155967116e-06,
"loss": 0.3454,
"step": 1289
},
{
"epoch": 1.5472000000000001,
"grad_norm": 0.4341244942793535,
"learning_rate": 5.609077363866555e-06,
"loss": 0.3805,
"step": 1290
},
{
"epoch": 1.5484,
"grad_norm": 0.5557467999785081,
"learning_rate": 5.602150525763701e-06,
"loss": 0.3963,
"step": 1291
},
{
"epoch": 1.5495999999999999,
"grad_norm": 0.7604236726580861,
"learning_rate": 5.595222514780379e-06,
"loss": 0.3446,
"step": 1292
},
{
"epoch": 1.5508,
"grad_norm": 0.6097142049223084,
"learning_rate": 5.588293344411097e-06,
"loss": 0.4047,
"step": 1293
},
{
"epoch": 1.552,
"grad_norm": 0.4521389519523142,
"learning_rate": 5.581363028152633e-06,
"loss": 0.366,
"step": 1294
},
{
"epoch": 1.5532,
"grad_norm": 0.5914221922011543,
"learning_rate": 5.574431579503991e-06,
"loss": 0.3948,
"step": 1295
},
{
"epoch": 1.5544,
"grad_norm": 0.5527344528002235,
"learning_rate": 5.5674990119663794e-06,
"loss": 0.3724,
"step": 1296
},
{
"epoch": 1.5556,
"grad_norm": 0.5580619134209653,
"learning_rate": 5.560565339043188e-06,
"loss": 0.3792,
"step": 1297
},
{
"epoch": 1.5568,
"grad_norm": 0.48215355759585776,
"learning_rate": 5.55363057423996e-06,
"loss": 0.4127,
"step": 1298
},
{
"epoch": 1.558,
"grad_norm": 0.5646877501201418,
"learning_rate": 5.546694731064364e-06,
"loss": 0.3939,
"step": 1299
},
{
"epoch": 1.5592000000000001,
"grad_norm": 0.6052210861484988,
"learning_rate": 5.5397578230261715e-06,
"loss": 0.3869,
"step": 1300
},
{
"epoch": 1.5604,
"grad_norm": 0.534027609323964,
"learning_rate": 5.532819863637223e-06,
"loss": 0.3497,
"step": 1301
},
{
"epoch": 1.5615999999999999,
"grad_norm": 0.626114651009411,
"learning_rate": 5.525880866411414e-06,
"loss": 0.3998,
"step": 1302
},
{
"epoch": 1.5628,
"grad_norm": 0.519083288023901,
"learning_rate": 5.5189408448646565e-06,
"loss": 0.3872,
"step": 1303
},
{
"epoch": 1.564,
"grad_norm": 0.5358073451310125,
"learning_rate": 5.511999812514857e-06,
"loss": 0.3685,
"step": 1304
},
{
"epoch": 1.5652,
"grad_norm": 0.5220660598380811,
"learning_rate": 5.505057782881896e-06,
"loss": 0.4071,
"step": 1305
},
{
"epoch": 1.5664,
"grad_norm": 0.682306549572413,
"learning_rate": 5.4981147694875924e-06,
"loss": 0.387,
"step": 1306
},
{
"epoch": 1.5676,
"grad_norm": 0.6555366532053696,
"learning_rate": 5.491170785855681e-06,
"loss": 0.3655,
"step": 1307
},
{
"epoch": 1.5688,
"grad_norm": 0.5995526776767848,
"learning_rate": 5.484225845511791e-06,
"loss": 0.3934,
"step": 1308
},
{
"epoch": 1.5699999999999998,
"grad_norm": 0.5699166744963786,
"learning_rate": 5.477279961983408e-06,
"loss": 0.3453,
"step": 1309
},
{
"epoch": 1.5712000000000002,
"grad_norm": 0.5562303554423684,
"learning_rate": 5.470333148799862e-06,
"loss": 0.3652,
"step": 1310
},
{
"epoch": 1.5724,
"grad_norm": 0.6507510125155322,
"learning_rate": 5.463385419492288e-06,
"loss": 0.3701,
"step": 1311
},
{
"epoch": 1.5735999999999999,
"grad_norm": 0.4851458447128553,
"learning_rate": 5.456436787593609e-06,
"loss": 0.3719,
"step": 1312
},
{
"epoch": 1.5748,
"grad_norm": 0.5566503858408686,
"learning_rate": 5.449487266638504e-06,
"loss": 0.3216,
"step": 1313
},
{
"epoch": 1.576,
"grad_norm": 0.5554972580250097,
"learning_rate": 5.442536870163386e-06,
"loss": 0.3979,
"step": 1314
},
{
"epoch": 1.5772,
"grad_norm": 0.5212955621171509,
"learning_rate": 5.43558561170637e-06,
"loss": 0.387,
"step": 1315
},
{
"epoch": 1.5784,
"grad_norm": 0.5507781499962543,
"learning_rate": 5.428633504807253e-06,
"loss": 0.3552,
"step": 1316
},
{
"epoch": 1.5796000000000001,
"grad_norm": 0.5329563540861513,
"learning_rate": 5.421680563007486e-06,
"loss": 0.386,
"step": 1317
},
{
"epoch": 1.5808,
"grad_norm": 0.4940099612080407,
"learning_rate": 5.414726799850141e-06,
"loss": 0.3462,
"step": 1318
},
{
"epoch": 1.5819999999999999,
"grad_norm": 0.5206713014005246,
"learning_rate": 5.407772228879894e-06,
"loss": 0.3404,
"step": 1319
},
{
"epoch": 1.5832000000000002,
"grad_norm": 0.6097320195459147,
"learning_rate": 5.400816863642991e-06,
"loss": 0.3484,
"step": 1320
},
{
"epoch": 1.5844,
"grad_norm": 0.6273168866546157,
"learning_rate": 5.393860717687231e-06,
"loss": 0.3628,
"step": 1321
},
{
"epoch": 1.5856,
"grad_norm": 0.5175627550499874,
"learning_rate": 5.3869038045619275e-06,
"loss": 0.3299,
"step": 1322
},
{
"epoch": 1.5868,
"grad_norm": 0.60322548886763,
"learning_rate": 5.379946137817891e-06,
"loss": 0.3763,
"step": 1323
},
{
"epoch": 1.588,
"grad_norm": 0.5767326804956401,
"learning_rate": 5.3729877310073985e-06,
"loss": 0.3913,
"step": 1324
},
{
"epoch": 1.5892,
"grad_norm": 0.5596287781864292,
"learning_rate": 5.366028597684173e-06,
"loss": 0.3901,
"step": 1325
},
{
"epoch": 1.5904,
"grad_norm": 0.6024156487106819,
"learning_rate": 5.359068751403347e-06,
"loss": 0.3751,
"step": 1326
},
{
"epoch": 1.5916000000000001,
"grad_norm": 0.6279038062933642,
"learning_rate": 5.352108205721445e-06,
"loss": 0.3699,
"step": 1327
},
{
"epoch": 1.5928,
"grad_norm": 0.5857494852870978,
"learning_rate": 5.345146974196351e-06,
"loss": 0.4159,
"step": 1328
},
{
"epoch": 1.5939999999999999,
"grad_norm": 0.5754425791232705,
"learning_rate": 5.338185070387289e-06,
"loss": 0.4033,
"step": 1329
},
{
"epoch": 1.5952,
"grad_norm": 0.6918790398986856,
"learning_rate": 5.3312225078547895e-06,
"loss": 0.4049,
"step": 1330
},
{
"epoch": 1.5964,
"grad_norm": 0.5294408996060997,
"learning_rate": 5.324259300160667e-06,
"loss": 0.3429,
"step": 1331
},
{
"epoch": 1.5976,
"grad_norm": 0.5030371247094101,
"learning_rate": 5.31729546086799e-06,
"loss": 0.3506,
"step": 1332
},
{
"epoch": 1.5988,
"grad_norm": 0.654689724453025,
"learning_rate": 5.310331003541065e-06,
"loss": 0.3985,
"step": 1333
},
{
"epoch": 1.6,
"grad_norm": 0.5706743127335765,
"learning_rate": 5.303365941745392e-06,
"loss": 0.3893,
"step": 1334
},
{
"epoch": 1.6012,
"grad_norm": 0.5909680552053695,
"learning_rate": 5.296400289047655e-06,
"loss": 0.3921,
"step": 1335
},
{
"epoch": 1.6024,
"grad_norm": 0.4716929907938838,
"learning_rate": 5.289434059015689e-06,
"loss": 0.386,
"step": 1336
},
{
"epoch": 1.6036000000000001,
"grad_norm": 0.5622232383432192,
"learning_rate": 5.28246726521845e-06,
"loss": 0.3646,
"step": 1337
},
{
"epoch": 1.6048,
"grad_norm": 0.5464132459639033,
"learning_rate": 5.275499921225994e-06,
"loss": 0.3588,
"step": 1338
},
{
"epoch": 1.6059999999999999,
"grad_norm": 0.5217117536373536,
"learning_rate": 5.26853204060945e-06,
"loss": 0.35,
"step": 1339
},
{
"epoch": 1.6072,
"grad_norm": 0.5463678279611617,
"learning_rate": 5.26156363694099e-06,
"loss": 0.3876,
"step": 1340
},
{
"epoch": 1.6084,
"grad_norm": 0.5414386696177665,
"learning_rate": 5.2545947237938055e-06,
"loss": 0.3581,
"step": 1341
},
{
"epoch": 1.6096,
"grad_norm": 0.6073251400646295,
"learning_rate": 5.247625314742083e-06,
"loss": 0.3593,
"step": 1342
},
{
"epoch": 1.6108,
"grad_norm": 0.5760004674528303,
"learning_rate": 5.240655423360969e-06,
"loss": 0.3964,
"step": 1343
},
{
"epoch": 1.612,
"grad_norm": 0.4860576503060455,
"learning_rate": 5.233685063226557e-06,
"loss": 0.3882,
"step": 1344
},
{
"epoch": 1.6132,
"grad_norm": 0.5384288035807419,
"learning_rate": 5.226714247915846e-06,
"loss": 0.3959,
"step": 1345
},
{
"epoch": 1.6143999999999998,
"grad_norm": 0.5693824402850227,
"learning_rate": 5.219742991006728e-06,
"loss": 0.3835,
"step": 1346
},
{
"epoch": 1.6156000000000001,
"grad_norm": 0.46937121460922504,
"learning_rate": 5.212771306077951e-06,
"loss": 0.3564,
"step": 1347
},
{
"epoch": 1.6168,
"grad_norm": 0.5617653190002595,
"learning_rate": 5.205799206709097e-06,
"loss": 0.3821,
"step": 1348
},
{
"epoch": 1.6179999999999999,
"grad_norm": 0.49040910836464563,
"learning_rate": 5.198826706480558e-06,
"loss": 0.3627,
"step": 1349
},
{
"epoch": 1.6192,
"grad_norm": 0.4856910298870825,
"learning_rate": 5.191853818973505e-06,
"loss": 0.389,
"step": 1350
},
{
"epoch": 1.6204,
"grad_norm": 0.4983582654746984,
"learning_rate": 5.184880557769865e-06,
"loss": 0.3417,
"step": 1351
},
{
"epoch": 1.6216,
"grad_norm": 0.5804818829476863,
"learning_rate": 5.177906936452287e-06,
"loss": 0.4007,
"step": 1352
},
{
"epoch": 1.6228,
"grad_norm": 0.5023000093685094,
"learning_rate": 5.170932968604131e-06,
"loss": 0.3831,
"step": 1353
},
{
"epoch": 1.624,
"grad_norm": 0.5644778366930446,
"learning_rate": 5.163958667809422e-06,
"loss": 0.3495,
"step": 1354
},
{
"epoch": 1.6252,
"grad_norm": 0.5385981194161137,
"learning_rate": 5.156984047652841e-06,
"loss": 0.3814,
"step": 1355
},
{
"epoch": 1.6263999999999998,
"grad_norm": 0.6001189538653291,
"learning_rate": 5.15000912171969e-06,
"loss": 0.4016,
"step": 1356
},
{
"epoch": 1.6276000000000002,
"grad_norm": 0.6470783016801794,
"learning_rate": 5.143033903595862e-06,
"loss": 0.4066,
"step": 1357
},
{
"epoch": 1.6288,
"grad_norm": 0.4072511431074058,
"learning_rate": 5.1360584068678225e-06,
"loss": 0.4364,
"step": 1358
},
{
"epoch": 1.63,
"grad_norm": 0.6029007981337849,
"learning_rate": 5.129082645122579e-06,
"loss": 0.3365,
"step": 1359
},
{
"epoch": 1.6312,
"grad_norm": 0.6437529016436582,
"learning_rate": 5.1221066319476576e-06,
"loss": 0.3935,
"step": 1360
},
{
"epoch": 1.6324,
"grad_norm": 0.5753512318024624,
"learning_rate": 5.115130380931071e-06,
"loss": 0.3722,
"step": 1361
},
{
"epoch": 1.6336,
"grad_norm": 0.4681832965615783,
"learning_rate": 5.108153905661296e-06,
"loss": 0.3874,
"step": 1362
},
{
"epoch": 1.6348,
"grad_norm": 0.5281516586208013,
"learning_rate": 5.101177219727245e-06,
"loss": 0.3853,
"step": 1363
},
{
"epoch": 1.6360000000000001,
"grad_norm": 0.6024168559501076,
"learning_rate": 5.094200336718246e-06,
"loss": 0.3763,
"step": 1364
},
{
"epoch": 1.6372,
"grad_norm": 0.5560166408755494,
"learning_rate": 5.087223270224003e-06,
"loss": 0.3936,
"step": 1365
},
{
"epoch": 1.6383999999999999,
"grad_norm": 0.6592848051247064,
"learning_rate": 5.080246033834581e-06,
"loss": 0.3893,
"step": 1366
},
{
"epoch": 1.6396,
"grad_norm": 0.4941210653374417,
"learning_rate": 5.0732686411403816e-06,
"loss": 0.3843,
"step": 1367
},
{
"epoch": 1.6408,
"grad_norm": 0.6710539393196384,
"learning_rate": 5.066291105732102e-06,
"loss": 0.3698,
"step": 1368
},
{
"epoch": 1.642,
"grad_norm": 0.5142664088562426,
"learning_rate": 5.05931344120072e-06,
"loss": 0.3857,
"step": 1369
},
{
"epoch": 1.6432,
"grad_norm": 0.570957765911436,
"learning_rate": 5.052335661137467e-06,
"loss": 0.3841,
"step": 1370
},
{
"epoch": 1.6444,
"grad_norm": 0.4699235231843021,
"learning_rate": 5.0453577791337984e-06,
"loss": 0.3912,
"step": 1371
},
{
"epoch": 1.6456,
"grad_norm": 0.5922972031744552,
"learning_rate": 5.03837980878137e-06,
"loss": 0.3878,
"step": 1372
},
{
"epoch": 1.6468,
"grad_norm": 0.6480156287733289,
"learning_rate": 5.031401763672003e-06,
"loss": 0.377,
"step": 1373
},
{
"epoch": 1.6480000000000001,
"grad_norm": 0.4834097534494832,
"learning_rate": 5.024423657397674e-06,
"loss": 0.377,
"step": 1374
},
{
"epoch": 1.6492,
"grad_norm": 0.5244756334982461,
"learning_rate": 5.017445503550471e-06,
"loss": 0.4217,
"step": 1375
},
{
"epoch": 1.6503999999999999,
"grad_norm": 0.5615837738343601,
"learning_rate": 5.010467315722578e-06,
"loss": 0.321,
"step": 1376
},
{
"epoch": 1.6516,
"grad_norm": 0.5615280581267553,
"learning_rate": 5.003489107506243e-06,
"loss": 0.3599,
"step": 1377
},
{
"epoch": 1.6528,
"grad_norm": 0.5122115198363286,
"learning_rate": 4.996510892493758e-06,
"loss": 0.3617,
"step": 1378
},
{
"epoch": 1.654,
"grad_norm": 0.5966845770519522,
"learning_rate": 4.989532684277424e-06,
"loss": 0.3752,
"step": 1379
},
{
"epoch": 1.6552,
"grad_norm": 0.5385391339983795,
"learning_rate": 4.98255449644953e-06,
"loss": 0.3981,
"step": 1380
},
{
"epoch": 1.6564,
"grad_norm": 0.5139696119449357,
"learning_rate": 4.975576342602329e-06,
"loss": 0.3696,
"step": 1381
},
{
"epoch": 1.6576,
"grad_norm": 0.5604633683105653,
"learning_rate": 4.968598236327998e-06,
"loss": 0.3773,
"step": 1382
},
{
"epoch": 1.6588,
"grad_norm": 0.5294822576550577,
"learning_rate": 4.961620191218632e-06,
"loss": 0.3739,
"step": 1383
},
{
"epoch": 1.6600000000000001,
"grad_norm": 0.5271341601650227,
"learning_rate": 4.954642220866202e-06,
"loss": 0.3947,
"step": 1384
},
{
"epoch": 1.6612,
"grad_norm": 0.5474409392635282,
"learning_rate": 4.947664338862534e-06,
"loss": 0.3611,
"step": 1385
},
{
"epoch": 1.6623999999999999,
"grad_norm": 0.5244053338973764,
"learning_rate": 4.940686558799283e-06,
"loss": 0.3728,
"step": 1386
},
{
"epoch": 1.6636,
"grad_norm": 0.5832859120381986,
"learning_rate": 4.933708894267901e-06,
"loss": 0.3439,
"step": 1387
},
{
"epoch": 1.6648,
"grad_norm": 0.48313733002500353,
"learning_rate": 4.92673135885962e-06,
"loss": 0.3421,
"step": 1388
},
{
"epoch": 1.666,
"grad_norm": 0.6160310969315205,
"learning_rate": 4.919753966165419e-06,
"loss": 0.3726,
"step": 1389
},
{
"epoch": 1.6672,
"grad_norm": 0.47547459852065205,
"learning_rate": 4.912776729775999e-06,
"loss": 0.3735,
"step": 1390
},
{
"epoch": 1.6684,
"grad_norm": 0.6780958925182213,
"learning_rate": 4.905799663281756e-06,
"loss": 0.4067,
"step": 1391
},
{
"epoch": 1.6696,
"grad_norm": 0.47505858645110066,
"learning_rate": 4.898822780272757e-06,
"loss": 0.3789,
"step": 1392
},
{
"epoch": 1.6707999999999998,
"grad_norm": 0.5961933662672453,
"learning_rate": 4.8918460943387065e-06,
"loss": 0.398,
"step": 1393
},
{
"epoch": 1.6720000000000002,
"grad_norm": 0.44935560844126476,
"learning_rate": 4.8848696190689295e-06,
"loss": 0.3787,
"step": 1394
},
{
"epoch": 1.6732,
"grad_norm": 1.1909559435485608,
"learning_rate": 4.877893368052343e-06,
"loss": 0.39,
"step": 1395
},
{
"epoch": 1.6743999999999999,
"grad_norm": 0.7290903045539241,
"learning_rate": 4.870917354877421e-06,
"loss": 0.3748,
"step": 1396
},
{
"epoch": 1.6756,
"grad_norm": 0.6363422553601993,
"learning_rate": 4.863941593132179e-06,
"loss": 0.3666,
"step": 1397
},
{
"epoch": 1.6768,
"grad_norm": 0.6144068562709365,
"learning_rate": 4.856966096404141e-06,
"loss": 0.365,
"step": 1398
},
{
"epoch": 1.678,
"grad_norm": 0.528262360343204,
"learning_rate": 4.849990878280313e-06,
"loss": 0.3758,
"step": 1399
},
{
"epoch": 1.6792,
"grad_norm": 0.5048194620723434,
"learning_rate": 4.843015952347159e-06,
"loss": 0.4263,
"step": 1400
},
{
"epoch": 1.6804000000000001,
"grad_norm": 0.5970202904903041,
"learning_rate": 4.8360413321905786e-06,
"loss": 0.3829,
"step": 1401
},
{
"epoch": 1.6816,
"grad_norm": 0.5466644316823482,
"learning_rate": 4.829067031395871e-06,
"loss": 0.3727,
"step": 1402
},
{
"epoch": 1.6827999999999999,
"grad_norm": 0.5151765615919326,
"learning_rate": 4.822093063547715e-06,
"loss": 0.3847,
"step": 1403
},
{
"epoch": 1.6840000000000002,
"grad_norm": 0.556770360004758,
"learning_rate": 4.815119442230138e-06,
"loss": 0.3767,
"step": 1404
},
{
"epoch": 1.6852,
"grad_norm": 0.5566673757348145,
"learning_rate": 4.8081461810264955e-06,
"loss": 0.3777,
"step": 1405
},
{
"epoch": 1.6864,
"grad_norm": 0.49635429488969113,
"learning_rate": 4.801173293519442e-06,
"loss": 0.376,
"step": 1406
},
{
"epoch": 1.6876,
"grad_norm": 0.5192887943818861,
"learning_rate": 4.7942007932909046e-06,
"loss": 0.3903,
"step": 1407
},
{
"epoch": 1.6888,
"grad_norm": 0.5653881468056294,
"learning_rate": 4.7872286939220516e-06,
"loss": 0.3811,
"step": 1408
},
{
"epoch": 1.69,
"grad_norm": 0.48983125503166525,
"learning_rate": 4.7802570089932746e-06,
"loss": 0.3838,
"step": 1409
},
{
"epoch": 1.6912,
"grad_norm": 0.6022472711625542,
"learning_rate": 4.773285752084154e-06,
"loss": 0.3712,
"step": 1410
},
{
"epoch": 1.6924000000000001,
"grad_norm": 0.4592209495815929,
"learning_rate": 4.766314936773445e-06,
"loss": 0.372,
"step": 1411
},
{
"epoch": 1.6936,
"grad_norm": 0.5988832880945443,
"learning_rate": 4.7593445766390315e-06,
"loss": 0.3655,
"step": 1412
},
{
"epoch": 1.6947999999999999,
"grad_norm": 0.45588967216207166,
"learning_rate": 4.752374685257919e-06,
"loss": 0.3643,
"step": 1413
},
{
"epoch": 1.696,
"grad_norm": 0.5556541460465428,
"learning_rate": 4.745405276206196e-06,
"loss": 0.3915,
"step": 1414
},
{
"epoch": 1.6972,
"grad_norm": 0.5012126046135608,
"learning_rate": 4.738436363059013e-06,
"loss": 0.379,
"step": 1415
},
{
"epoch": 1.6984,
"grad_norm": 0.5041937531710121,
"learning_rate": 4.731467959390552e-06,
"loss": 0.3421,
"step": 1416
},
{
"epoch": 1.6996,
"grad_norm": 0.5283604855128771,
"learning_rate": 4.724500078774008e-06,
"loss": 0.337,
"step": 1417
},
{
"epoch": 1.7008,
"grad_norm": 0.6133511645731402,
"learning_rate": 4.717532734781552e-06,
"loss": 0.3779,
"step": 1418
},
{
"epoch": 1.702,
"grad_norm": 0.507800303653185,
"learning_rate": 4.710565940984313e-06,
"loss": 0.3598,
"step": 1419
},
{
"epoch": 1.7032,
"grad_norm": 0.5426917965409874,
"learning_rate": 4.703599710952347e-06,
"loss": 0.3441,
"step": 1420
},
{
"epoch": 1.7044000000000001,
"grad_norm": 0.543149879989818,
"learning_rate": 4.6966340582546085e-06,
"loss": 0.3569,
"step": 1421
},
{
"epoch": 1.7056,
"grad_norm": 0.5079384765827253,
"learning_rate": 4.689668996458937e-06,
"loss": 0.345,
"step": 1422
},
{
"epoch": 1.7067999999999999,
"grad_norm": 0.5717820988124686,
"learning_rate": 4.682704539132011e-06,
"loss": 0.3777,
"step": 1423
},
{
"epoch": 1.708,
"grad_norm": 0.6581756912070847,
"learning_rate": 4.6757406998393354e-06,
"loss": 0.3802,
"step": 1424
},
{
"epoch": 1.7092,
"grad_norm": 0.4871198567708227,
"learning_rate": 4.668777492145212e-06,
"loss": 0.3737,
"step": 1425
},
{
"epoch": 1.7104,
"grad_norm": 0.5836916748404222,
"learning_rate": 4.661814929612713e-06,
"loss": 0.3505,
"step": 1426
},
{
"epoch": 1.7116,
"grad_norm": 0.5377148867740239,
"learning_rate": 4.654853025803649e-06,
"loss": 0.3593,
"step": 1427
},
{
"epoch": 1.7128,
"grad_norm": 0.554213026309025,
"learning_rate": 4.6478917942785575e-06,
"loss": 0.3546,
"step": 1428
},
{
"epoch": 1.714,
"grad_norm": 0.46820629569907074,
"learning_rate": 4.640931248596655e-06,
"loss": 0.3324,
"step": 1429
},
{
"epoch": 1.7151999999999998,
"grad_norm": 0.5097593929717572,
"learning_rate": 4.633971402315828e-06,
"loss": 0.3443,
"step": 1430
},
{
"epoch": 1.7164000000000001,
"grad_norm": 0.39050241505881494,
"learning_rate": 4.627012268992603e-06,
"loss": 0.3414,
"step": 1431
},
{
"epoch": 1.7176,
"grad_norm": 0.5964374941992685,
"learning_rate": 4.62005386218211e-06,
"loss": 0.3849,
"step": 1432
},
{
"epoch": 1.7187999999999999,
"grad_norm": 0.5484383135611258,
"learning_rate": 4.613096195438074e-06,
"loss": 0.4054,
"step": 1433
},
{
"epoch": 1.72,
"grad_norm": 0.6148189012721375,
"learning_rate": 4.60613928231277e-06,
"loss": 0.3755,
"step": 1434
},
{
"epoch": 1.7212,
"grad_norm": 0.5326369467247358,
"learning_rate": 4.59918313635701e-06,
"loss": 0.3271,
"step": 1435
},
{
"epoch": 1.7224,
"grad_norm": 0.5273601456958985,
"learning_rate": 4.592227771120108e-06,
"loss": 0.3817,
"step": 1436
},
{
"epoch": 1.7236,
"grad_norm": 0.4203737234869034,
"learning_rate": 4.585273200149859e-06,
"loss": 0.3974,
"step": 1437
},
{
"epoch": 1.7248,
"grad_norm": 0.4936440580649403,
"learning_rate": 4.578319436992515e-06,
"loss": 0.3354,
"step": 1438
},
{
"epoch": 1.726,
"grad_norm": 0.5130638369711625,
"learning_rate": 4.5713664951927475e-06,
"loss": 0.3403,
"step": 1439
},
{
"epoch": 1.7271999999999998,
"grad_norm": 0.6258616009702679,
"learning_rate": 4.5644143882936316e-06,
"loss": 0.3865,
"step": 1440
},
{
"epoch": 1.7284000000000002,
"grad_norm": 0.5885108055193322,
"learning_rate": 4.5574631298366165e-06,
"loss": 0.3858,
"step": 1441
},
{
"epoch": 1.7296,
"grad_norm": 0.45377890152896866,
"learning_rate": 4.550512733361499e-06,
"loss": 0.3625,
"step": 1442
},
{
"epoch": 1.7308,
"grad_norm": 0.52449993022665,
"learning_rate": 4.543563212406392e-06,
"loss": 0.3985,
"step": 1443
},
{
"epoch": 1.732,
"grad_norm": 0.5485173263452691,
"learning_rate": 4.536614580507714e-06,
"loss": 0.3235,
"step": 1444
},
{
"epoch": 1.7332,
"grad_norm": 0.5660979450438521,
"learning_rate": 4.52966685120014e-06,
"loss": 0.3796,
"step": 1445
},
{
"epoch": 1.7344,
"grad_norm": 0.538975607453001,
"learning_rate": 4.5227200380165925e-06,
"loss": 0.3837,
"step": 1446
},
{
"epoch": 1.7356,
"grad_norm": 0.5331842010157929,
"learning_rate": 4.51577415448821e-06,
"loss": 0.359,
"step": 1447
},
{
"epoch": 1.7368000000000001,
"grad_norm": 0.54645760044252,
"learning_rate": 4.508829214144318e-06,
"loss": 0.3623,
"step": 1448
},
{
"epoch": 1.738,
"grad_norm": 0.46325458435024974,
"learning_rate": 4.5018852305124075e-06,
"loss": 0.3622,
"step": 1449
},
{
"epoch": 1.7391999999999999,
"grad_norm": 0.5145168772489319,
"learning_rate": 4.494942217118105e-06,
"loss": 0.3722,
"step": 1450
},
{
"epoch": 1.7404,
"grad_norm": 0.39840795787353445,
"learning_rate": 4.488000187485144e-06,
"loss": 0.388,
"step": 1451
},
{
"epoch": 1.7416,
"grad_norm": 0.49935920190729366,
"learning_rate": 4.481059155135346e-06,
"loss": 0.3801,
"step": 1452
},
{
"epoch": 1.7428,
"grad_norm": 0.6174089378652856,
"learning_rate": 4.474119133588588e-06,
"loss": 0.3663,
"step": 1453
},
{
"epoch": 1.744,
"grad_norm": 0.5238711669514375,
"learning_rate": 4.4671801363627776e-06,
"loss": 0.3329,
"step": 1454
},
{
"epoch": 1.7452,
"grad_norm": 0.522612750946382,
"learning_rate": 4.460242176973829e-06,
"loss": 0.3578,
"step": 1455
},
{
"epoch": 1.7464,
"grad_norm": 0.4683774342022167,
"learning_rate": 4.453305268935637e-06,
"loss": 0.3571,
"step": 1456
},
{
"epoch": 1.7476,
"grad_norm": 0.5671416556577483,
"learning_rate": 4.446369425760042e-06,
"loss": 0.3724,
"step": 1457
},
{
"epoch": 1.7488000000000001,
"grad_norm": 0.5141170253368055,
"learning_rate": 4.439434660956814e-06,
"loss": 0.3694,
"step": 1458
},
{
"epoch": 1.75,
"grad_norm": 0.4756332978389349,
"learning_rate": 4.432500988033621e-06,
"loss": 0.3665,
"step": 1459
},
{
"epoch": 1.7511999999999999,
"grad_norm": 0.5755242591889931,
"learning_rate": 4.42556842049601e-06,
"loss": 0.3672,
"step": 1460
},
{
"epoch": 1.7524,
"grad_norm": 0.507611616514393,
"learning_rate": 4.418636971847367e-06,
"loss": 0.3578,
"step": 1461
},
{
"epoch": 1.7536,
"grad_norm": 0.570791682381513,
"learning_rate": 4.4117066555889045e-06,
"loss": 0.3762,
"step": 1462
},
{
"epoch": 1.7548,
"grad_norm": 0.5423061322282628,
"learning_rate": 4.404777485219624e-06,
"loss": 0.3295,
"step": 1463
},
{
"epoch": 1.756,
"grad_norm": 0.6360241213438484,
"learning_rate": 4.397849474236299e-06,
"loss": 0.3811,
"step": 1464
},
{
"epoch": 1.7572,
"grad_norm": 0.5389604778367516,
"learning_rate": 4.390922636133444e-06,
"loss": 0.39,
"step": 1465
},
{
"epoch": 1.7584,
"grad_norm": 0.6798118345646637,
"learning_rate": 4.38399698440329e-06,
"loss": 0.3853,
"step": 1466
},
{
"epoch": 1.7596,
"grad_norm": 0.53508198471987,
"learning_rate": 4.377072532535756e-06,
"loss": 0.4439,
"step": 1467
},
{
"epoch": 1.7608000000000001,
"grad_norm": 0.6128605125037492,
"learning_rate": 4.370149294018419e-06,
"loss": 0.3696,
"step": 1468
},
{
"epoch": 1.762,
"grad_norm": 0.9725373567491395,
"learning_rate": 4.3632272823365004e-06,
"loss": 0.4337,
"step": 1469
},
{
"epoch": 1.7631999999999999,
"grad_norm": 0.7151413691087543,
"learning_rate": 4.356306510972829e-06,
"loss": 0.3852,
"step": 1470
},
{
"epoch": 1.7644,
"grad_norm": 0.6052197334436037,
"learning_rate": 4.349386993407817e-06,
"loss": 0.3937,
"step": 1471
},
{
"epoch": 1.7656,
"grad_norm": 0.5187059957562955,
"learning_rate": 4.342468743119436e-06,
"loss": 0.3747,
"step": 1472
},
{
"epoch": 1.7668,
"grad_norm": 0.567039630276201,
"learning_rate": 4.335551773583186e-06,
"loss": 0.4003,
"step": 1473
},
{
"epoch": 1.768,
"grad_norm": 0.6097264557850262,
"learning_rate": 4.328636098272075e-06,
"loss": 0.353,
"step": 1474
},
{
"epoch": 1.7692,
"grad_norm": 0.5345926269351398,
"learning_rate": 4.32172173065659e-06,
"loss": 0.3525,
"step": 1475
},
{
"epoch": 1.7704,
"grad_norm": 0.4797111307565443,
"learning_rate": 4.31480868420467e-06,
"loss": 0.4078,
"step": 1476
},
{
"epoch": 1.7715999999999998,
"grad_norm": 0.7058092031192786,
"learning_rate": 4.307896972381681e-06,
"loss": 0.3951,
"step": 1477
},
{
"epoch": 1.7728000000000002,
"grad_norm": 0.5919098471895498,
"learning_rate": 4.3009866086503905e-06,
"loss": 0.3946,
"step": 1478
},
{
"epoch": 1.774,
"grad_norm": 0.544218809026354,
"learning_rate": 4.294077606470937e-06,
"loss": 0.3591,
"step": 1479
},
{
"epoch": 1.7752,
"grad_norm": 0.5234209998058871,
"learning_rate": 4.28716997930081e-06,
"loss": 0.3682,
"step": 1480
},
{
"epoch": 1.7764,
"grad_norm": 0.6673345442623971,
"learning_rate": 4.2802637405948175e-06,
"loss": 0.3723,
"step": 1481
},
{
"epoch": 1.7776,
"grad_norm": 0.5063519336134552,
"learning_rate": 4.273358903805069e-06,
"loss": 0.376,
"step": 1482
},
{
"epoch": 1.7788,
"grad_norm": 0.6427479164034344,
"learning_rate": 4.266455482380938e-06,
"loss": 0.3697,
"step": 1483
},
{
"epoch": 1.78,
"grad_norm": 0.6492453092818349,
"learning_rate": 4.2595534897690415e-06,
"loss": 0.3645,
"step": 1484
},
{
"epoch": 1.7812000000000001,
"grad_norm": 0.5198773636033195,
"learning_rate": 4.252652939413215e-06,
"loss": 0.3657,
"step": 1485
},
{
"epoch": 1.7824,
"grad_norm": 0.5506908323597158,
"learning_rate": 4.245753844754484e-06,
"loss": 0.3639,
"step": 1486
},
{
"epoch": 1.7835999999999999,
"grad_norm": 0.5937767470670442,
"learning_rate": 4.23885621923104e-06,
"loss": 0.354,
"step": 1487
},
{
"epoch": 1.7848000000000002,
"grad_norm": 0.5679193278077739,
"learning_rate": 4.231960076278211e-06,
"loss": 0.3855,
"step": 1488
},
{
"epoch": 1.786,
"grad_norm": 0.4961986169959603,
"learning_rate": 4.225065429328439e-06,
"loss": 0.3803,
"step": 1489
},
{
"epoch": 1.7872,
"grad_norm": 0.5751271293242203,
"learning_rate": 4.218172291811249e-06,
"loss": 0.3588,
"step": 1490
},
{
"epoch": 1.7884,
"grad_norm": 0.4650715545650898,
"learning_rate": 4.211280677153228e-06,
"loss": 0.3868,
"step": 1491
},
{
"epoch": 1.7896,
"grad_norm": 0.4609837338355205,
"learning_rate": 4.204390598777999e-06,
"loss": 0.4046,
"step": 1492
},
{
"epoch": 1.7908,
"grad_norm": 0.674095634826924,
"learning_rate": 4.1975020701061884e-06,
"loss": 0.3631,
"step": 1493
},
{
"epoch": 1.792,
"grad_norm": 0.5176761419170193,
"learning_rate": 4.190615104555407e-06,
"loss": 0.3636,
"step": 1494
},
{
"epoch": 1.7932000000000001,
"grad_norm": 0.6213097687855804,
"learning_rate": 4.1837297155402204e-06,
"loss": 0.3924,
"step": 1495
},
{
"epoch": 1.7944,
"grad_norm": 0.550061903460295,
"learning_rate": 4.17684591647212e-06,
"loss": 0.3697,
"step": 1496
},
{
"epoch": 1.7955999999999999,
"grad_norm": 0.5631316141718825,
"learning_rate": 4.1699637207595035e-06,
"loss": 0.3983,
"step": 1497
},
{
"epoch": 1.7968,
"grad_norm": 0.4676009983905216,
"learning_rate": 4.163083141807648e-06,
"loss": 0.3777,
"step": 1498
},
{
"epoch": 1.798,
"grad_norm": 0.5819571057130736,
"learning_rate": 4.156204193018677e-06,
"loss": 0.3556,
"step": 1499
},
{
"epoch": 1.7992,
"grad_norm": 0.525961103299972,
"learning_rate": 4.149326887791541e-06,
"loss": 0.3669,
"step": 1500
},
{
"epoch": 1.8004,
"grad_norm": 0.5537939950528783,
"learning_rate": 4.142451239521988e-06,
"loss": 0.3548,
"step": 1501
},
{
"epoch": 1.8016,
"grad_norm": 0.4036885983445749,
"learning_rate": 4.135577261602537e-06,
"loss": 0.3397,
"step": 1502
},
{
"epoch": 1.8028,
"grad_norm": 0.47808838294530726,
"learning_rate": 4.128704967422458e-06,
"loss": 0.3701,
"step": 1503
},
{
"epoch": 1.804,
"grad_norm": 0.5715714749581858,
"learning_rate": 4.1218343703677385e-06,
"loss": 0.3816,
"step": 1504
},
{
"epoch": 1.8052000000000001,
"grad_norm": 0.5827261024498809,
"learning_rate": 4.114965483821061e-06,
"loss": 0.3827,
"step": 1505
},
{
"epoch": 1.8064,
"grad_norm": 0.8838156305247361,
"learning_rate": 4.108098321161776e-06,
"loss": 0.385,
"step": 1506
},
{
"epoch": 1.8075999999999999,
"grad_norm": 0.4359612260958472,
"learning_rate": 4.101232895765875e-06,
"loss": 0.3544,
"step": 1507
},
{
"epoch": 1.8088,
"grad_norm": 0.5361847284685747,
"learning_rate": 4.094369221005965e-06,
"loss": 0.3654,
"step": 1508
},
{
"epoch": 1.81,
"grad_norm": 0.5756106529204161,
"learning_rate": 4.0875073102512485e-06,
"loss": 0.4157,
"step": 1509
},
{
"epoch": 1.8112,
"grad_norm": 0.573089376292654,
"learning_rate": 4.080647176867486e-06,
"loss": 0.3692,
"step": 1510
},
{
"epoch": 1.8124,
"grad_norm": 0.5705072972686631,
"learning_rate": 4.07378883421698e-06,
"loss": 0.3711,
"step": 1511
},
{
"epoch": 1.8136,
"grad_norm": 0.482441385829138,
"learning_rate": 4.066932295658543e-06,
"loss": 0.3568,
"step": 1512
},
{
"epoch": 1.8148,
"grad_norm": 0.4209533458740963,
"learning_rate": 4.06007757454747e-06,
"loss": 0.393,
"step": 1513
},
{
"epoch": 1.8159999999999998,
"grad_norm": 0.5236483171737902,
"learning_rate": 4.053224684235526e-06,
"loss": 0.3611,
"step": 1514
},
{
"epoch": 1.8172000000000001,
"grad_norm": 0.5351363903713013,
"learning_rate": 4.0463736380708986e-06,
"loss": 0.3706,
"step": 1515
},
{
"epoch": 1.8184,
"grad_norm": 0.5425221056948407,
"learning_rate": 4.039524449398191e-06,
"loss": 0.3601,
"step": 1516
},
{
"epoch": 1.8195999999999999,
"grad_norm": 0.5048292023028578,
"learning_rate": 4.032677131558386e-06,
"loss": 0.3854,
"step": 1517
},
{
"epoch": 1.8208,
"grad_norm": 0.4737232612699287,
"learning_rate": 4.025831697888817e-06,
"loss": 0.398,
"step": 1518
},
{
"epoch": 1.822,
"grad_norm": 0.6024946408130546,
"learning_rate": 4.01898816172316e-06,
"loss": 0.3237,
"step": 1519
},
{
"epoch": 1.8232,
"grad_norm": 0.4552328457278779,
"learning_rate": 4.012146536391383e-06,
"loss": 0.3623,
"step": 1520
},
{
"epoch": 1.8244,
"grad_norm": 0.5315194274609365,
"learning_rate": 4.005306835219737e-06,
"loss": 0.3841,
"step": 1521
},
{
"epoch": 1.8256000000000001,
"grad_norm": 0.7239466583222977,
"learning_rate": 3.998469071530725e-06,
"loss": 0.3804,
"step": 1522
},
{
"epoch": 1.8268,
"grad_norm": 0.492142871435618,
"learning_rate": 3.991633258643077e-06,
"loss": 0.3465,
"step": 1523
},
{
"epoch": 1.8279999999999998,
"grad_norm": 0.5303185304197421,
"learning_rate": 3.9847994098717166e-06,
"loss": 0.3673,
"step": 1524
},
{
"epoch": 1.8292000000000002,
"grad_norm": 0.478005077740389,
"learning_rate": 3.9779675385277545e-06,
"loss": 0.4136,
"step": 1525
},
{
"epoch": 1.8304,
"grad_norm": 0.4635938541288215,
"learning_rate": 3.971137657918437e-06,
"loss": 0.3532,
"step": 1526
},
{
"epoch": 1.8316,
"grad_norm": 0.5448930511722193,
"learning_rate": 3.96430978134714e-06,
"loss": 0.3784,
"step": 1527
},
{
"epoch": 1.8328,
"grad_norm": 0.5951035174657143,
"learning_rate": 3.957483922113334e-06,
"loss": 0.3702,
"step": 1528
},
{
"epoch": 1.834,
"grad_norm": 0.4767784439251168,
"learning_rate": 3.950660093512556e-06,
"loss": 0.4134,
"step": 1529
},
{
"epoch": 1.8352,
"grad_norm": 0.5436932780193476,
"learning_rate": 3.943838308836398e-06,
"loss": 0.346,
"step": 1530
},
{
"epoch": 1.8364,
"grad_norm": 0.5243606584297386,
"learning_rate": 3.937018581372462e-06,
"loss": 0.3653,
"step": 1531
},
{
"epoch": 1.8376000000000001,
"grad_norm": 0.4947097270033976,
"learning_rate": 3.9302009244043435e-06,
"loss": 0.3237,
"step": 1532
},
{
"epoch": 1.8388,
"grad_norm": 0.5388902364359281,
"learning_rate": 3.923385351211609e-06,
"loss": 0.3879,
"step": 1533
},
{
"epoch": 1.8399999999999999,
"grad_norm": 0.6355454986812967,
"learning_rate": 3.916571875069764e-06,
"loss": 0.3569,
"step": 1534
},
{
"epoch": 1.8412,
"grad_norm": 0.5443889556984081,
"learning_rate": 3.909760509250225e-06,
"loss": 0.3843,
"step": 1535
},
{
"epoch": 1.8424,
"grad_norm": 0.505390080729345,
"learning_rate": 3.902951267020311e-06,
"loss": 0.4259,
"step": 1536
},
{
"epoch": 1.8436,
"grad_norm": 0.5215460521041646,
"learning_rate": 3.896144161643189e-06,
"loss": 0.3681,
"step": 1537
},
{
"epoch": 1.8448,
"grad_norm": 0.5296599874059905,
"learning_rate": 3.8893392063778736e-06,
"loss": 0.3621,
"step": 1538
},
{
"epoch": 1.846,
"grad_norm": 0.5461805110697693,
"learning_rate": 3.882536414479189e-06,
"loss": 0.3366,
"step": 1539
},
{
"epoch": 1.8472,
"grad_norm": 0.5108889914041715,
"learning_rate": 3.8757357991977415e-06,
"loss": 0.3724,
"step": 1540
},
{
"epoch": 1.8484,
"grad_norm": 0.5420603294921492,
"learning_rate": 3.868937373779907e-06,
"loss": 0.3711,
"step": 1541
},
{
"epoch": 1.8496000000000001,
"grad_norm": 0.48916721532093305,
"learning_rate": 3.862141151467787e-06,
"loss": 0.3728,
"step": 1542
},
{
"epoch": 1.8508,
"grad_norm": 0.48465691505041003,
"learning_rate": 3.855347145499197e-06,
"loss": 0.3519,
"step": 1543
},
{
"epoch": 1.8519999999999999,
"grad_norm": 0.5754386422023029,
"learning_rate": 3.848555369107631e-06,
"loss": 0.421,
"step": 1544
},
{
"epoch": 1.8532,
"grad_norm": 0.5317947339959318,
"learning_rate": 3.841765835522242e-06,
"loss": 0.3663,
"step": 1545
},
{
"epoch": 1.8544,
"grad_norm": 0.4955907576641392,
"learning_rate": 3.83497855796782e-06,
"loss": 0.355,
"step": 1546
},
{
"epoch": 1.8556,
"grad_norm": 0.6458398468766117,
"learning_rate": 3.8281935496647526e-06,
"loss": 0.3743,
"step": 1547
},
{
"epoch": 1.8568,
"grad_norm": 0.46667706268297304,
"learning_rate": 3.821410823829011e-06,
"loss": 0.4003,
"step": 1548
},
{
"epoch": 1.858,
"grad_norm": 0.5520866583909263,
"learning_rate": 3.8146303936721197e-06,
"loss": 0.4399,
"step": 1549
},
{
"epoch": 1.8592,
"grad_norm": 0.5773823022211833,
"learning_rate": 3.8078522724011324e-06,
"loss": 0.3507,
"step": 1550
},
{
"epoch": 1.8604,
"grad_norm": 0.5223004050075711,
"learning_rate": 3.8010764732186044e-06,
"loss": 0.3768,
"step": 1551
},
{
"epoch": 1.8616000000000001,
"grad_norm": 0.6233856499906449,
"learning_rate": 3.79430300932257e-06,
"loss": 0.3945,
"step": 1552
},
{
"epoch": 1.8628,
"grad_norm": 0.5769446082685447,
"learning_rate": 3.7875318939065147e-06,
"loss": 0.3796,
"step": 1553
},
{
"epoch": 1.8639999999999999,
"grad_norm": 0.4976458511268623,
"learning_rate": 3.7807631401593455e-06,
"loss": 0.3901,
"step": 1554
},
{
"epoch": 1.8652,
"grad_norm": 0.5581927189349054,
"learning_rate": 3.773996761265373e-06,
"loss": 0.3695,
"step": 1555
},
{
"epoch": 1.8664,
"grad_norm": 0.5680626951405663,
"learning_rate": 3.767232770404281e-06,
"loss": 0.344,
"step": 1556
},
{
"epoch": 1.8676,
"grad_norm": 0.6465253624638309,
"learning_rate": 3.7604711807511034e-06,
"loss": 0.4288,
"step": 1557
},
{
"epoch": 1.8688,
"grad_norm": 0.5578386611760368,
"learning_rate": 3.753712005476197e-06,
"loss": 0.3603,
"step": 1558
},
{
"epoch": 1.87,
"grad_norm": 0.499801410275945,
"learning_rate": 3.74695525774521e-06,
"loss": 0.3668,
"step": 1559
},
{
"epoch": 1.8712,
"grad_norm": 0.4790306329461315,
"learning_rate": 3.7402009507190696e-06,
"loss": 0.3946,
"step": 1560
},
{
"epoch": 1.8723999999999998,
"grad_norm": 0.5682465666500219,
"learning_rate": 3.733449097553945e-06,
"loss": 0.3891,
"step": 1561
},
{
"epoch": 1.8736000000000002,
"grad_norm": 0.4774980812732164,
"learning_rate": 3.7266997114012265e-06,
"loss": 0.3472,
"step": 1562
},
{
"epoch": 1.8748,
"grad_norm": 0.4619452981736414,
"learning_rate": 3.7199528054075005e-06,
"loss": 0.3827,
"step": 1563
},
{
"epoch": 1.876,
"grad_norm": 0.5389963332916776,
"learning_rate": 3.713208392714523e-06,
"loss": 0.3717,
"step": 1564
},
{
"epoch": 1.8772,
"grad_norm": 0.5186510437822639,
"learning_rate": 3.7064664864591878e-06,
"loss": 0.3886,
"step": 1565
},
{
"epoch": 1.8784,
"grad_norm": 0.4926466183813456,
"learning_rate": 3.6997270997735122e-06,
"loss": 0.3468,
"step": 1566
},
{
"epoch": 1.8796,
"grad_norm": 0.49498331518620853,
"learning_rate": 3.6929902457846034e-06,
"loss": 0.3705,
"step": 1567
},
{
"epoch": 1.8808,
"grad_norm": 0.5425259501099292,
"learning_rate": 3.6862559376146388e-06,
"loss": 0.3747,
"step": 1568
},
{
"epoch": 1.8820000000000001,
"grad_norm": 0.5609809814213781,
"learning_rate": 3.6795241883808342e-06,
"loss": 0.3532,
"step": 1569
},
{
"epoch": 1.8832,
"grad_norm": 0.5328793062157196,
"learning_rate": 3.6727950111954186e-06,
"loss": 0.3862,
"step": 1570
},
{
"epoch": 1.8843999999999999,
"grad_norm": 0.544774371543934,
"learning_rate": 3.6660684191656155e-06,
"loss": 0.3842,
"step": 1571
},
{
"epoch": 1.8856000000000002,
"grad_norm": 0.5647936712841308,
"learning_rate": 3.6593444253936094e-06,
"loss": 0.3604,
"step": 1572
},
{
"epoch": 1.8868,
"grad_norm": 0.5021335213225804,
"learning_rate": 3.652623042976529e-06,
"loss": 0.3821,
"step": 1573
},
{
"epoch": 1.888,
"grad_norm": 0.4692959275922926,
"learning_rate": 3.645904285006412e-06,
"loss": 0.3727,
"step": 1574
},
{
"epoch": 1.8892,
"grad_norm": 0.5394960915353251,
"learning_rate": 3.6391881645701854e-06,
"loss": 0.3679,
"step": 1575
},
{
"epoch": 1.8904,
"grad_norm": 0.5215317490147278,
"learning_rate": 3.632474694749638e-06,
"loss": 0.367,
"step": 1576
},
{
"epoch": 1.8916,
"grad_norm": 0.47546096708845664,
"learning_rate": 3.625763888621397e-06,
"loss": 0.3649,
"step": 1577
},
{
"epoch": 1.8928,
"grad_norm": 0.5621601110474697,
"learning_rate": 3.6190557592569e-06,
"loss": 0.3321,
"step": 1578
},
{
"epoch": 1.8940000000000001,
"grad_norm": 0.5028797468518021,
"learning_rate": 3.612350319722372e-06,
"loss": 0.3547,
"step": 1579
},
{
"epoch": 1.8952,
"grad_norm": 0.5196444646365543,
"learning_rate": 3.6056475830787997e-06,
"loss": 0.3401,
"step": 1580
},
{
"epoch": 1.8963999999999999,
"grad_norm": 0.5638831380664279,
"learning_rate": 3.5989475623819025e-06,
"loss": 0.4,
"step": 1581
},
{
"epoch": 1.8976,
"grad_norm": 0.5338161038634314,
"learning_rate": 3.5922502706821094e-06,
"loss": 0.3743,
"step": 1582
},
{
"epoch": 1.8988,
"grad_norm": 0.5183504121139286,
"learning_rate": 3.585555721024535e-06,
"loss": 0.3736,
"step": 1583
},
{
"epoch": 1.9,
"grad_norm": 0.5441996684359627,
"learning_rate": 3.578863926448955e-06,
"loss": 0.3638,
"step": 1584
},
{
"epoch": 1.9012,
"grad_norm": 0.532009924270318,
"learning_rate": 3.5721748999897753e-06,
"loss": 0.3809,
"step": 1585
},
{
"epoch": 1.9024,
"grad_norm": 0.4888811306933541,
"learning_rate": 3.5654886546760125e-06,
"loss": 0.3885,
"step": 1586
},
{
"epoch": 1.9036,
"grad_norm": 0.5412619197000217,
"learning_rate": 3.558805203531263e-06,
"loss": 0.3502,
"step": 1587
},
{
"epoch": 1.9048,
"grad_norm": 0.5738922874471647,
"learning_rate": 3.5521245595736837e-06,
"loss": 0.3849,
"step": 1588
},
{
"epoch": 1.9060000000000001,
"grad_norm": 0.459414943658132,
"learning_rate": 3.5454467358159606e-06,
"loss": 0.3954,
"step": 1589
},
{
"epoch": 1.9072,
"grad_norm": 0.5942278277436939,
"learning_rate": 3.5387717452652914e-06,
"loss": 0.3591,
"step": 1590
},
{
"epoch": 1.9083999999999999,
"grad_norm": 0.6053235629917005,
"learning_rate": 3.53209960092335e-06,
"loss": 0.3679,
"step": 1591
},
{
"epoch": 1.9096,
"grad_norm": 0.5589122111952849,
"learning_rate": 3.5254303157862707e-06,
"loss": 0.3635,
"step": 1592
},
{
"epoch": 1.9108,
"grad_norm": 0.4688125407210059,
"learning_rate": 3.5187639028446136e-06,
"loss": 0.3652,
"step": 1593
},
{
"epoch": 1.912,
"grad_norm": 0.5608055017754532,
"learning_rate": 3.512100375083347e-06,
"loss": 0.3985,
"step": 1594
},
{
"epoch": 1.9132,
"grad_norm": 0.5716758221470768,
"learning_rate": 3.5054397454818224e-06,
"loss": 0.3888,
"step": 1595
},
{
"epoch": 1.9144,
"grad_norm": 0.5971907346915089,
"learning_rate": 3.498782027013742e-06,
"loss": 0.3564,
"step": 1596
},
{
"epoch": 1.9156,
"grad_norm": 0.43530261136331444,
"learning_rate": 3.4921272326471388e-06,
"loss": 0.3481,
"step": 1597
},
{
"epoch": 1.9167999999999998,
"grad_norm": 0.5258196952161901,
"learning_rate": 3.4854753753443494e-06,
"loss": 0.3846,
"step": 1598
},
{
"epoch": 1.9180000000000001,
"grad_norm": 0.5371980957790986,
"learning_rate": 3.47882646806199e-06,
"loss": 0.3688,
"step": 1599
},
{
"epoch": 1.9192,
"grad_norm": 0.5651026783522477,
"learning_rate": 3.472180523750933e-06,
"loss": 0.3746,
"step": 1600
},
{
"epoch": 1.9203999999999999,
"grad_norm": 0.48795009081539864,
"learning_rate": 3.4655375553562774e-06,
"loss": 0.3798,
"step": 1601
},
{
"epoch": 1.9216,
"grad_norm": 0.5500703230461617,
"learning_rate": 3.458897575817326e-06,
"loss": 0.3748,
"step": 1602
},
{
"epoch": 1.9228,
"grad_norm": 0.48505267826496845,
"learning_rate": 3.4522605980675593e-06,
"loss": 0.3722,
"step": 1603
},
{
"epoch": 1.924,
"grad_norm": 0.5497801258468328,
"learning_rate": 3.44562663503461e-06,
"loss": 0.3491,
"step": 1604
},
{
"epoch": 1.9252,
"grad_norm": 0.5234204994472285,
"learning_rate": 3.43899569964024e-06,
"loss": 0.3658,
"step": 1605
},
{
"epoch": 1.9264000000000001,
"grad_norm": 0.5696552141865382,
"learning_rate": 3.432367804800316e-06,
"loss": 0.3639,
"step": 1606
},
{
"epoch": 1.9276,
"grad_norm": 0.5257832715215263,
"learning_rate": 3.4257429634247783e-06,
"loss": 0.372,
"step": 1607
},
{
"epoch": 1.9287999999999998,
"grad_norm": 0.5737981603168226,
"learning_rate": 3.419121188417622e-06,
"loss": 0.4145,
"step": 1608
},
{
"epoch": 1.9300000000000002,
"grad_norm": 0.5764943601824868,
"learning_rate": 3.41250249267687e-06,
"loss": 0.3691,
"step": 1609
},
{
"epoch": 1.9312,
"grad_norm": 0.5491686888513345,
"learning_rate": 3.4058868890945425e-06,
"loss": 0.3662,
"step": 1610
},
{
"epoch": 1.9324,
"grad_norm": 0.6044451235894281,
"learning_rate": 3.3992743905566453e-06,
"loss": 0.3584,
"step": 1611
},
{
"epoch": 1.9336,
"grad_norm": 0.4198217864035794,
"learning_rate": 3.3926650099431286e-06,
"loss": 0.3525,
"step": 1612
},
{
"epoch": 1.9348,
"grad_norm": 0.5120087990944092,
"learning_rate": 3.3860587601278715e-06,
"loss": 0.3701,
"step": 1613
},
{
"epoch": 1.936,
"grad_norm": 0.6721308523835157,
"learning_rate": 3.3794556539786584e-06,
"loss": 0.3343,
"step": 1614
},
{
"epoch": 1.9372,
"grad_norm": 0.5970580286538677,
"learning_rate": 3.372855704357144e-06,
"loss": 0.3936,
"step": 1615
},
{
"epoch": 1.9384000000000001,
"grad_norm": 0.5137829253898193,
"learning_rate": 3.3662589241188382e-06,
"loss": 0.3556,
"step": 1616
},
{
"epoch": 1.9396,
"grad_norm": 0.531855958226858,
"learning_rate": 3.3596653261130806e-06,
"loss": 0.3589,
"step": 1617
},
{
"epoch": 1.9407999999999999,
"grad_norm": 0.5946587085895932,
"learning_rate": 3.3530749231830073e-06,
"loss": 0.329,
"step": 1618
},
{
"epoch": 1.942,
"grad_norm": 0.4686111871670008,
"learning_rate": 3.3464877281655335e-06,
"loss": 0.3747,
"step": 1619
},
{
"epoch": 1.9432,
"grad_norm": 0.5992858573364381,
"learning_rate": 3.339903753891326e-06,
"loss": 0.3973,
"step": 1620
},
{
"epoch": 1.9444,
"grad_norm": 0.5396983308324597,
"learning_rate": 3.333323013184773e-06,
"loss": 0.329,
"step": 1621
},
{
"epoch": 1.9456,
"grad_norm": 0.4964847650360278,
"learning_rate": 3.326745518863976e-06,
"loss": 0.3878,
"step": 1622
},
{
"epoch": 1.9468,
"grad_norm": 0.6389901728616871,
"learning_rate": 3.320171283740702e-06,
"loss": 0.3713,
"step": 1623
},
{
"epoch": 1.948,
"grad_norm": 0.49670306633769595,
"learning_rate": 3.3136003206203727e-06,
"loss": 0.3551,
"step": 1624
},
{
"epoch": 1.9492,
"grad_norm": 0.5582835213645627,
"learning_rate": 3.307032642302041e-06,
"loss": 0.3608,
"step": 1625
},
{
"epoch": 1.9504000000000001,
"grad_norm": 0.5196610105352837,
"learning_rate": 3.3004682615783524e-06,
"loss": 0.4153,
"step": 1626
},
{
"epoch": 1.9516,
"grad_norm": 0.5771469058077887,
"learning_rate": 3.2939071912355424e-06,
"loss": 0.4018,
"step": 1627
},
{
"epoch": 1.9527999999999999,
"grad_norm": 0.49003266342707796,
"learning_rate": 3.2873494440533856e-06,
"loss": 0.3513,
"step": 1628
},
{
"epoch": 1.954,
"grad_norm": 0.46639523883379724,
"learning_rate": 3.2807950328051906e-06,
"loss": 0.3605,
"step": 1629
},
{
"epoch": 1.9552,
"grad_norm": 0.6535942240293555,
"learning_rate": 3.2742439702577665e-06,
"loss": 0.3778,
"step": 1630
},
{
"epoch": 1.9564,
"grad_norm": 0.48640850782934464,
"learning_rate": 3.267696269171402e-06,
"loss": 0.3624,
"step": 1631
},
{
"epoch": 1.9576,
"grad_norm": 0.48986749881878144,
"learning_rate": 3.2611519422998308e-06,
"loss": 0.3612,
"step": 1632
},
{
"epoch": 1.9588,
"grad_norm": 0.48744895083408596,
"learning_rate": 3.254611002390227e-06,
"loss": 0.3887,
"step": 1633
},
{
"epoch": 1.96,
"grad_norm": 0.6344603647518389,
"learning_rate": 3.248073462183155e-06,
"loss": 0.343,
"step": 1634
},
{
"epoch": 1.9612,
"grad_norm": 0.5455560601200089,
"learning_rate": 3.2415393344125647e-06,
"loss": 0.3784,
"step": 1635
},
{
"epoch": 1.9624000000000001,
"grad_norm": 0.551024447947303,
"learning_rate": 3.235008631805755e-06,
"loss": 0.3725,
"step": 1636
},
{
"epoch": 1.9636,
"grad_norm": 0.5820297969549334,
"learning_rate": 3.228481367083356e-06,
"loss": 0.3958,
"step": 1637
},
{
"epoch": 1.9647999999999999,
"grad_norm": 0.5127265706843908,
"learning_rate": 3.2219575529593017e-06,
"loss": 0.4105,
"step": 1638
},
{
"epoch": 1.966,
"grad_norm": 0.6449585442393768,
"learning_rate": 3.215437202140803e-06,
"loss": 0.3753,
"step": 1639
},
{
"epoch": 1.9672,
"grad_norm": 0.5258365554255892,
"learning_rate": 3.2089203273283253e-06,
"loss": 0.3682,
"step": 1640
},
{
"epoch": 1.9684,
"grad_norm": 0.6624073726742211,
"learning_rate": 3.2024069412155632e-06,
"loss": 0.3743,
"step": 1641
},
{
"epoch": 1.9696,
"grad_norm": 0.4454142223170098,
"learning_rate": 3.1958970564894187e-06,
"loss": 0.3807,
"step": 1642
},
{
"epoch": 1.9708,
"grad_norm": 0.5076196288677864,
"learning_rate": 3.189390685829967e-06,
"loss": 0.3751,
"step": 1643
},
{
"epoch": 1.972,
"grad_norm": 0.6237584283328298,
"learning_rate": 3.182887841910448e-06,
"loss": 0.3547,
"step": 1644
},
{
"epoch": 1.9731999999999998,
"grad_norm": 0.49749450545562557,
"learning_rate": 3.1763885373972246e-06,
"loss": 0.3614,
"step": 1645
},
{
"epoch": 1.9744000000000002,
"grad_norm": 0.5197430897510873,
"learning_rate": 3.1698927849497683e-06,
"loss": 0.3128,
"step": 1646
},
{
"epoch": 1.9756,
"grad_norm": 0.4808955813148,
"learning_rate": 3.1634005972206326e-06,
"loss": 0.3484,
"step": 1647
},
{
"epoch": 1.9768,
"grad_norm": 0.5607402476639658,
"learning_rate": 3.156911986855425e-06,
"loss": 0.3803,
"step": 1648
},
{
"epoch": 1.978,
"grad_norm": 0.5236498380407515,
"learning_rate": 3.150426966492788e-06,
"loss": 0.3778,
"step": 1649
},
{
"epoch": 1.9792,
"grad_norm": 0.6068813245345871,
"learning_rate": 3.143945548764371e-06,
"loss": 0.367,
"step": 1650
},
{
"epoch": 1.9804,
"grad_norm": 0.6489137877474729,
"learning_rate": 3.137467746294803e-06,
"loss": 0.3533,
"step": 1651
},
{
"epoch": 1.9816,
"grad_norm": 0.5096114548543674,
"learning_rate": 3.130993571701674e-06,
"loss": 0.3736,
"step": 1652
},
{
"epoch": 1.9828000000000001,
"grad_norm": 0.5145598601879202,
"learning_rate": 3.124523037595506e-06,
"loss": 0.3733,
"step": 1653
},
{
"epoch": 1.984,
"grad_norm": 0.47189446289024184,
"learning_rate": 3.1180561565797323e-06,
"loss": 0.3892,
"step": 1654
},
{
"epoch": 1.9851999999999999,
"grad_norm": 0.6762021822264397,
"learning_rate": 3.1115929412506698e-06,
"loss": 0.3443,
"step": 1655
},
{
"epoch": 1.9864000000000002,
"grad_norm": 0.48938422415105837,
"learning_rate": 3.1051334041974923e-06,
"loss": 0.3891,
"step": 1656
},
{
"epoch": 1.9876,
"grad_norm": 0.5313458718096313,
"learning_rate": 3.0986775580022122e-06,
"loss": 0.3361,
"step": 1657
},
{
"epoch": 1.9888,
"grad_norm": 0.5134417300426585,
"learning_rate": 3.092225415239652e-06,
"loss": 0.3499,
"step": 1658
},
{
"epoch": 1.99,
"grad_norm": 0.5303924683671959,
"learning_rate": 3.0857769884774192e-06,
"loss": 0.3514,
"step": 1659
},
{
"epoch": 1.9912,
"grad_norm": 0.4818336976271982,
"learning_rate": 3.079332290275887e-06,
"loss": 0.4273,
"step": 1660
},
{
"epoch": 1.9924,
"grad_norm": 0.5734803663648053,
"learning_rate": 3.0728913331881638e-06,
"loss": 0.3691,
"step": 1661
},
{
"epoch": 1.9936,
"grad_norm": 0.5192783415445636,
"learning_rate": 3.0664541297600682e-06,
"loss": 0.3576,
"step": 1662
},
{
"epoch": 1.9948000000000001,
"grad_norm": 0.5898979211568474,
"learning_rate": 3.0600206925301114e-06,
"loss": 0.3545,
"step": 1663
},
{
"epoch": 1.996,
"grad_norm": 0.6277347416397628,
"learning_rate": 3.053591034029465e-06,
"loss": 0.4102,
"step": 1664
},
{
"epoch": 1.9971999999999999,
"grad_norm": 0.4769980692535198,
"learning_rate": 3.0471651667819447e-06,
"loss": 0.3512,
"step": 1665
},
{
"epoch": 1.9984,
"grad_norm": 0.5481061372674426,
"learning_rate": 3.0407431033039795e-06,
"loss": 0.343,
"step": 1666
},
{
"epoch": 1.9996,
"grad_norm": 0.49844107285522393,
"learning_rate": 3.03432485610459e-06,
"loss": 0.378,
"step": 1667
},
{
"epoch": 2.0,
"grad_norm": 0.49844107285522393,
"learning_rate": 3.0279104376853592e-06,
"loss": 0.3941,
"step": 1668
},
{
"epoch": 2.0012,
"grad_norm": 1.0591897389977591,
"learning_rate": 3.0214998605404165e-06,
"loss": 0.297,
"step": 1669
},
{
"epoch": 2.0024,
"grad_norm": 0.5624930162629377,
"learning_rate": 3.0150931371564107e-06,
"loss": 0.332,
"step": 1670
},
{
"epoch": 2.0036,
"grad_norm": 0.5333759652290109,
"learning_rate": 3.0086902800124806e-06,
"loss": 0.3298,
"step": 1671
},
{
"epoch": 2.0048,
"grad_norm": 0.6392975587937063,
"learning_rate": 3.0022913015802363e-06,
"loss": 0.2996,
"step": 1672
},
{
"epoch": 2.006,
"grad_norm": 0.6357461041838159,
"learning_rate": 2.99589621432373e-06,
"loss": 0.3202,
"step": 1673
},
{
"epoch": 2.0072,
"grad_norm": 0.4796703489749423,
"learning_rate": 2.9895050306994385e-06,
"loss": 0.333,
"step": 1674
},
{
"epoch": 2.0084,
"grad_norm": 0.5361479466387894,
"learning_rate": 2.9831177631562306e-06,
"loss": 0.3002,
"step": 1675
},
{
"epoch": 2.0096,
"grad_norm": 0.7273485065920107,
"learning_rate": 2.9767344241353535e-06,
"loss": 0.3374,
"step": 1676
},
{
"epoch": 2.0108,
"grad_norm": 0.610498600023043,
"learning_rate": 2.9703550260703974e-06,
"loss": 0.2796,
"step": 1677
},
{
"epoch": 2.012,
"grad_norm": 0.6240915148913934,
"learning_rate": 2.9639795813872773e-06,
"loss": 0.3453,
"step": 1678
},
{
"epoch": 2.0132,
"grad_norm": 0.5664530610060867,
"learning_rate": 2.9576081025042068e-06,
"loss": 0.3245,
"step": 1679
},
{
"epoch": 2.0144,
"grad_norm": 0.5028610712097004,
"learning_rate": 2.9512406018316763e-06,
"loss": 0.3709,
"step": 1680
},
{
"epoch": 2.0156,
"grad_norm": 0.6410080913126401,
"learning_rate": 2.9448770917724296e-06,
"loss": 0.3355,
"step": 1681
},
{
"epoch": 2.0168,
"grad_norm": 0.5406313865257344,
"learning_rate": 2.9385175847214325e-06,
"loss": 0.3464,
"step": 1682
},
{
"epoch": 2.018,
"grad_norm": 0.5824184255775667,
"learning_rate": 2.9321620930658578e-06,
"loss": 0.3288,
"step": 1683
},
{
"epoch": 2.0192,
"grad_norm": 0.5210606649765747,
"learning_rate": 2.925810629185054e-06,
"loss": 0.3252,
"step": 1684
},
{
"epoch": 2.0204,
"grad_norm": 0.6983919226474768,
"learning_rate": 2.919463205450526e-06,
"loss": 0.2967,
"step": 1685
},
{
"epoch": 2.0216,
"grad_norm": 0.4058001224876323,
"learning_rate": 2.9131198342259065e-06,
"loss": 0.286,
"step": 1686
},
{
"epoch": 2.0228,
"grad_norm": 0.5599493682522717,
"learning_rate": 2.9067805278669425e-06,
"loss": 0.3139,
"step": 1687
},
{
"epoch": 2.024,
"grad_norm": 0.48325874794700446,
"learning_rate": 2.900445298721455e-06,
"loss": 0.3179,
"step": 1688
},
{
"epoch": 2.0252,
"grad_norm": 0.5193881973767888,
"learning_rate": 2.894114159129324e-06,
"loss": 0.3273,
"step": 1689
},
{
"epoch": 2.0264,
"grad_norm": 0.5058550742725104,
"learning_rate": 2.8877871214224694e-06,
"loss": 0.3554,
"step": 1690
},
{
"epoch": 2.0276,
"grad_norm": 0.6095816753403872,
"learning_rate": 2.881464197924814e-06,
"loss": 0.3258,
"step": 1691
},
{
"epoch": 2.0288,
"grad_norm": 0.6214600375192152,
"learning_rate": 2.875145400952274e-06,
"loss": 0.2933,
"step": 1692
},
{
"epoch": 2.03,
"grad_norm": 0.4814523589044724,
"learning_rate": 2.868830742812726e-06,
"loss": 0.3466,
"step": 1693
},
{
"epoch": 2.0312,
"grad_norm": 0.5850847089967712,
"learning_rate": 2.8625202358059806e-06,
"loss": 0.3137,
"step": 1694
},
{
"epoch": 2.0324,
"grad_norm": 0.5066166407228644,
"learning_rate": 2.8562138922237648e-06,
"loss": 0.3253,
"step": 1695
},
{
"epoch": 2.0336,
"grad_norm": 0.4972535078378977,
"learning_rate": 2.8499117243496986e-06,
"loss": 0.317,
"step": 1696
},
{
"epoch": 2.0348,
"grad_norm": 0.46441502204129653,
"learning_rate": 2.8436137444592694e-06,
"loss": 0.3207,
"step": 1697
},
{
"epoch": 2.036,
"grad_norm": 0.5061917060410045,
"learning_rate": 2.837319964819801e-06,
"loss": 0.3188,
"step": 1698
},
{
"epoch": 2.0372,
"grad_norm": 0.5331265053707761,
"learning_rate": 2.8310303976904396e-06,
"loss": 0.3357,
"step": 1699
},
{
"epoch": 2.0384,
"grad_norm": 0.48253564003344795,
"learning_rate": 2.824745055322128e-06,
"loss": 0.3327,
"step": 1700
},
{
"epoch": 2.0396,
"grad_norm": 0.561262170115495,
"learning_rate": 2.818463949957575e-06,
"loss": 0.3304,
"step": 1701
},
{
"epoch": 2.0408,
"grad_norm": 0.517301545564998,
"learning_rate": 2.8121870938312413e-06,
"loss": 0.325,
"step": 1702
},
{
"epoch": 2.042,
"grad_norm": 0.5013079295176773,
"learning_rate": 2.80591449916931e-06,
"loss": 0.2929,
"step": 1703
},
{
"epoch": 2.0432,
"grad_norm": 0.5906583350314853,
"learning_rate": 2.7996461781896624e-06,
"loss": 0.3455,
"step": 1704
},
{
"epoch": 2.0444,
"grad_norm": 0.5648511168991517,
"learning_rate": 2.7933821431018523e-06,
"loss": 0.3446,
"step": 1705
},
{
"epoch": 2.0456,
"grad_norm": 0.5106440892824704,
"learning_rate": 2.7871224061070935e-06,
"loss": 0.3181,
"step": 1706
},
{
"epoch": 2.0468,
"grad_norm": 0.59545898809765,
"learning_rate": 2.780866979398218e-06,
"loss": 0.3606,
"step": 1707
},
{
"epoch": 2.048,
"grad_norm": 0.5652481430453722,
"learning_rate": 2.77461587515967e-06,
"loss": 0.3228,
"step": 1708
},
{
"epoch": 2.0492,
"grad_norm": 0.4548590670825555,
"learning_rate": 2.7683691055674745e-06,
"loss": 0.3187,
"step": 1709
},
{
"epoch": 2.0504,
"grad_norm": 0.537889637661287,
"learning_rate": 2.7621266827892062e-06,
"loss": 0.3285,
"step": 1710
},
{
"epoch": 2.0516,
"grad_norm": 0.48534951318330044,
"learning_rate": 2.755888618983977e-06,
"loss": 0.3347,
"step": 1711
},
{
"epoch": 2.0528,
"grad_norm": 0.4562443608158437,
"learning_rate": 2.749654926302412e-06,
"loss": 0.3029,
"step": 1712
},
{
"epoch": 2.054,
"grad_norm": 0.5451371452941084,
"learning_rate": 2.743425616886615e-06,
"loss": 0.3251,
"step": 1713
},
{
"epoch": 2.0552,
"grad_norm": 0.5001312557188845,
"learning_rate": 2.737200702870157e-06,
"loss": 0.2942,
"step": 1714
},
{
"epoch": 2.0564,
"grad_norm": 0.5597264970045684,
"learning_rate": 2.7309801963780485e-06,
"loss": 0.3253,
"step": 1715
},
{
"epoch": 2.0576,
"grad_norm": 0.4875406465840606,
"learning_rate": 2.724764109526711e-06,
"loss": 0.3359,
"step": 1716
},
{
"epoch": 2.0588,
"grad_norm": 0.4521934668461847,
"learning_rate": 2.7185524544239567e-06,
"loss": 0.2898,
"step": 1717
},
{
"epoch": 2.06,
"grad_norm": 0.5129446705571528,
"learning_rate": 2.71234524316897e-06,
"loss": 0.3297,
"step": 1718
},
{
"epoch": 2.0612,
"grad_norm": 0.6356511930660489,
"learning_rate": 2.706142487852279e-06,
"loss": 0.3566,
"step": 1719
},
{
"epoch": 2.0624,
"grad_norm": 0.6062536311557942,
"learning_rate": 2.699944200555727e-06,
"loss": 0.3328,
"step": 1720
},
{
"epoch": 2.0636,
"grad_norm": 0.5755634538397862,
"learning_rate": 2.693750393352462e-06,
"loss": 0.3231,
"step": 1721
},
{
"epoch": 2.0648,
"grad_norm": 0.41540127708606395,
"learning_rate": 2.6875610783069007e-06,
"loss": 0.313,
"step": 1722
},
{
"epoch": 2.066,
"grad_norm": 0.5649731253952338,
"learning_rate": 2.681376267474707e-06,
"loss": 0.331,
"step": 1723
},
{
"epoch": 2.0672,
"grad_norm": 0.42456286519904973,
"learning_rate": 2.67519597290278e-06,
"loss": 0.294,
"step": 1724
},
{
"epoch": 2.0684,
"grad_norm": 0.644090239706362,
"learning_rate": 2.669020206629217e-06,
"loss": 0.345,
"step": 1725
},
{
"epoch": 2.0696,
"grad_norm": 0.6154771529796125,
"learning_rate": 2.6628489806832947e-06,
"loss": 0.3091,
"step": 1726
},
{
"epoch": 2.0708,
"grad_norm": 0.5171610309392253,
"learning_rate": 2.6566823070854442e-06,
"loss": 0.3147,
"step": 1727
},
{
"epoch": 2.072,
"grad_norm": 0.4790305699062261,
"learning_rate": 2.650520197847235e-06,
"loss": 0.3149,
"step": 1728
},
{
"epoch": 2.0732,
"grad_norm": 0.5807330206595483,
"learning_rate": 2.6443626649713407e-06,
"loss": 0.3275,
"step": 1729
},
{
"epoch": 2.0744,
"grad_norm": 0.5596362638288411,
"learning_rate": 2.6382097204515246e-06,
"loss": 0.2882,
"step": 1730
},
{
"epoch": 2.0756,
"grad_norm": 0.4542833127022029,
"learning_rate": 2.6320613762726123e-06,
"loss": 0.3388,
"step": 1731
},
{
"epoch": 2.0768,
"grad_norm": 0.6155612626173607,
"learning_rate": 2.625917644410467e-06,
"loss": 0.3177,
"step": 1732
},
{
"epoch": 2.078,
"grad_norm": 0.5610949947795919,
"learning_rate": 2.6197785368319663e-06,
"loss": 0.3191,
"step": 1733
},
{
"epoch": 2.0792,
"grad_norm": 0.5195379523395367,
"learning_rate": 2.613644065494985e-06,
"loss": 0.2999,
"step": 1734
},
{
"epoch": 2.0804,
"grad_norm": 0.5130953063709173,
"learning_rate": 2.6075142423483675e-06,
"loss": 0.3326,
"step": 1735
},
{
"epoch": 2.0816,
"grad_norm": 0.5438357579709007,
"learning_rate": 2.6013890793318972e-06,
"loss": 0.321,
"step": 1736
},
{
"epoch": 2.0828,
"grad_norm": 0.42650077102042744,
"learning_rate": 2.5952685883762918e-06,
"loss": 0.3227,
"step": 1737
},
{
"epoch": 2.084,
"grad_norm": 0.6695148645027172,
"learning_rate": 2.589152781403158e-06,
"loss": 0.3475,
"step": 1738
},
{
"epoch": 2.0852,
"grad_norm": 0.5432827453027909,
"learning_rate": 2.583041670324982e-06,
"loss": 0.3244,
"step": 1739
},
{
"epoch": 2.0864,
"grad_norm": 0.4764183872286641,
"learning_rate": 2.5769352670451058e-06,
"loss": 0.3045,
"step": 1740
},
{
"epoch": 2.0876,
"grad_norm": 0.5671290448812345,
"learning_rate": 2.5708335834577035e-06,
"loss": 0.3315,
"step": 1741
},
{
"epoch": 2.0888,
"grad_norm": 0.44226466754051197,
"learning_rate": 2.5647366314477473e-06,
"loss": 0.2882,
"step": 1742
},
{
"epoch": 2.09,
"grad_norm": 0.5551492977171031,
"learning_rate": 2.5586444228910036e-06,
"loss": 0.3308,
"step": 1743
},
{
"epoch": 2.0912,
"grad_norm": 0.5878933682398272,
"learning_rate": 2.5525569696539916e-06,
"loss": 0.3083,
"step": 1744
},
{
"epoch": 2.0924,
"grad_norm": 0.4629754288702419,
"learning_rate": 2.546474283593969e-06,
"loss": 0.3,
"step": 1745
},
{
"epoch": 2.0936,
"grad_norm": 0.49341087104009557,
"learning_rate": 2.540396376558912e-06,
"loss": 0.3259,
"step": 1746
},
{
"epoch": 2.0948,
"grad_norm": 0.5455999623520852,
"learning_rate": 2.5343232603874868e-06,
"loss": 0.3105,
"step": 1747
},
{
"epoch": 2.096,
"grad_norm": 0.5184574680106683,
"learning_rate": 2.5282549469090246e-06,
"loss": 0.3274,
"step": 1748
},
{
"epoch": 2.0972,
"grad_norm": 0.48886013939692935,
"learning_rate": 2.522191447943506e-06,
"loss": 0.3222,
"step": 1749
},
{
"epoch": 2.0984,
"grad_norm": 0.5339952790495294,
"learning_rate": 2.5161327753015297e-06,
"loss": 0.326,
"step": 1750
},
{
"epoch": 2.0996,
"grad_norm": 0.5769798704745386,
"learning_rate": 2.5100789407842985e-06,
"loss": 0.3131,
"step": 1751
},
{
"epoch": 2.1008,
"grad_norm": 0.471327477630944,
"learning_rate": 2.5040299561835846e-06,
"loss": 0.3158,
"step": 1752
},
{
"epoch": 2.102,
"grad_norm": 0.5026221605877529,
"learning_rate": 2.4979858332817225e-06,
"loss": 0.3197,
"step": 1753
},
{
"epoch": 2.1032,
"grad_norm": 0.5372165990104663,
"learning_rate": 2.4919465838515687e-06,
"loss": 0.3145,
"step": 1754
},
{
"epoch": 2.1044,
"grad_norm": 0.6641614157675277,
"learning_rate": 2.48591221965649e-06,
"loss": 0.3084,
"step": 1755
},
{
"epoch": 2.1056,
"grad_norm": 0.45435517949191667,
"learning_rate": 2.479882752450339e-06,
"loss": 0.3048,
"step": 1756
},
{
"epoch": 2.1068,
"grad_norm": 0.45788975083727507,
"learning_rate": 2.4738581939774303e-06,
"loss": 0.3236,
"step": 1757
},
{
"epoch": 2.108,
"grad_norm": 0.6053906560507565,
"learning_rate": 2.4678385559725125e-06,
"loss": 0.3097,
"step": 1758
},
{
"epoch": 2.1092,
"grad_norm": 0.46864880799396397,
"learning_rate": 2.4618238501607577e-06,
"loss": 0.2929,
"step": 1759
},
{
"epoch": 2.1104,
"grad_norm": 0.5358136543962301,
"learning_rate": 2.455814088257723e-06,
"loss": 0.3018,
"step": 1760
},
{
"epoch": 2.1116,
"grad_norm": 0.5084290417208595,
"learning_rate": 2.4498092819693364e-06,
"loss": 0.3451,
"step": 1761
},
{
"epoch": 2.1128,
"grad_norm": 0.5376300173544492,
"learning_rate": 2.443809442991884e-06,
"loss": 0.3202,
"step": 1762
},
{
"epoch": 2.114,
"grad_norm": 0.49013479713273206,
"learning_rate": 2.4378145830119637e-06,
"loss": 0.2757,
"step": 1763
},
{
"epoch": 2.1152,
"grad_norm": 0.47235239540072455,
"learning_rate": 2.4318247137064788e-06,
"loss": 0.3227,
"step": 1764
},
{
"epoch": 2.1164,
"grad_norm": 0.5754905338883051,
"learning_rate": 2.425839846742616e-06,
"loss": 0.3213,
"step": 1765
},
{
"epoch": 2.1176,
"grad_norm": 0.5328513695540388,
"learning_rate": 2.4198599937778138e-06,
"loss": 0.3453,
"step": 1766
},
{
"epoch": 2.1188,
"grad_norm": 0.49451716774508825,
"learning_rate": 2.4138851664597424e-06,
"loss": 0.3448,
"step": 1767
},
{
"epoch": 2.12,
"grad_norm": 0.5848833881296942,
"learning_rate": 2.407915376426293e-06,
"loss": 0.3209,
"step": 1768
},
{
"epoch": 2.1212,
"grad_norm": 0.4748565112996034,
"learning_rate": 2.401950635305535e-06,
"loss": 0.3322,
"step": 1769
},
{
"epoch": 2.1224,
"grad_norm": 0.5764982528591845,
"learning_rate": 2.395990954715705e-06,
"loss": 0.3169,
"step": 1770
},
{
"epoch": 2.1236,
"grad_norm": 0.5233034231797605,
"learning_rate": 2.390036346265188e-06,
"loss": 0.2888,
"step": 1771
},
{
"epoch": 2.1248,
"grad_norm": 0.4229446066382141,
"learning_rate": 2.3840868215524824e-06,
"loss": 0.3386,
"step": 1772
},
{
"epoch": 2.126,
"grad_norm": 0.49839777559135007,
"learning_rate": 2.378142392166191e-06,
"loss": 0.3024,
"step": 1773
},
{
"epoch": 2.1272,
"grad_norm": 0.5854598988817233,
"learning_rate": 2.3722030696849857e-06,
"loss": 0.3043,
"step": 1774
},
{
"epoch": 2.1284,
"grad_norm": 0.5251285279619012,
"learning_rate": 2.3662688656775973e-06,
"loss": 0.3055,
"step": 1775
},
{
"epoch": 2.1296,
"grad_norm": 0.5299135615265957,
"learning_rate": 2.3603397917027787e-06,
"loss": 0.3272,
"step": 1776
},
{
"epoch": 2.1308,
"grad_norm": 0.44624661103843993,
"learning_rate": 2.3544158593092986e-06,
"loss": 0.3231,
"step": 1777
},
{
"epoch": 2.132,
"grad_norm": 0.4238463305096519,
"learning_rate": 2.3484970800359087e-06,
"loss": 0.2789,
"step": 1778
},
{
"epoch": 2.1332,
"grad_norm": 0.49592857839548543,
"learning_rate": 2.34258346541132e-06,
"loss": 0.3207,
"step": 1779
},
{
"epoch": 2.1344,
"grad_norm": 0.5602302444148246,
"learning_rate": 2.3366750269541833e-06,
"loss": 0.2812,
"step": 1780
},
{
"epoch": 2.1356,
"grad_norm": 0.496040693843303,
"learning_rate": 2.3307717761730745e-06,
"loss": 0.3372,
"step": 1781
},
{
"epoch": 2.1368,
"grad_norm": 0.5368866392412842,
"learning_rate": 2.3248737245664575e-06,
"loss": 0.3225,
"step": 1782
},
{
"epoch": 2.138,
"grad_norm": 0.4372922000394104,
"learning_rate": 2.318980883622668e-06,
"loss": 0.3308,
"step": 1783
},
{
"epoch": 2.1391999999999998,
"grad_norm": 0.5800037494382269,
"learning_rate": 2.313093264819903e-06,
"loss": 0.3219,
"step": 1784
},
{
"epoch": 2.1404,
"grad_norm": 0.5100182459831669,
"learning_rate": 2.3072108796261766e-06,
"loss": 0.3631,
"step": 1785
},
{
"epoch": 2.1416,
"grad_norm": 0.637238814394478,
"learning_rate": 2.301333739499312e-06,
"loss": 0.2987,
"step": 1786
},
{
"epoch": 2.1428,
"grad_norm": 0.5994611543419464,
"learning_rate": 2.2954618558869194e-06,
"loss": 0.3471,
"step": 1787
},
{
"epoch": 2.144,
"grad_norm": 0.43867780095707637,
"learning_rate": 2.2895952402263642e-06,
"loss": 0.3036,
"step": 1788
},
{
"epoch": 2.1452,
"grad_norm": 0.5083667260496417,
"learning_rate": 2.283733903944756e-06,
"loss": 0.3079,
"step": 1789
},
{
"epoch": 2.1464,
"grad_norm": 0.5130542668182789,
"learning_rate": 2.2778778584589214e-06,
"loss": 0.2861,
"step": 1790
},
{
"epoch": 2.1476,
"grad_norm": 0.514032456923874,
"learning_rate": 2.272027115175377e-06,
"loss": 0.3508,
"step": 1791
},
{
"epoch": 2.1488,
"grad_norm": 0.640022764728223,
"learning_rate": 2.2661816854903117e-06,
"loss": 0.3207,
"step": 1792
},
{
"epoch": 2.15,
"grad_norm": 0.45744824436218107,
"learning_rate": 2.2603415807895718e-06,
"loss": 0.3376,
"step": 1793
},
{
"epoch": 2.1512000000000002,
"grad_norm": 0.455358581238065,
"learning_rate": 2.254506812448622e-06,
"loss": 0.2976,
"step": 1794
},
{
"epoch": 2.1524,
"grad_norm": 0.5327275147447389,
"learning_rate": 2.2486773918325394e-06,
"loss": 0.3117,
"step": 1795
},
{
"epoch": 2.1536,
"grad_norm": 0.5451791706243986,
"learning_rate": 2.242853330295984e-06,
"loss": 0.3648,
"step": 1796
},
{
"epoch": 2.1548,
"grad_norm": 0.5600203400053401,
"learning_rate": 2.2370346391831737e-06,
"loss": 0.2985,
"step": 1797
},
{
"epoch": 2.156,
"grad_norm": 0.4796107945013445,
"learning_rate": 2.231221329827867e-06,
"loss": 0.3204,
"step": 1798
},
{
"epoch": 2.1572,
"grad_norm": 0.5343319911045897,
"learning_rate": 2.225413413553341e-06,
"loss": 0.3552,
"step": 1799
},
{
"epoch": 2.1584,
"grad_norm": 0.5936500640441726,
"learning_rate": 2.219610901672371e-06,
"loss": 0.2796,
"step": 1800
},
{
"epoch": 2.1596,
"grad_norm": 0.5103626788162533,
"learning_rate": 2.2138138054871993e-06,
"loss": 0.3185,
"step": 1801
},
{
"epoch": 2.1608,
"grad_norm": 0.46686532203197323,
"learning_rate": 2.208022136289521e-06,
"loss": 0.3319,
"step": 1802
},
{
"epoch": 2.162,
"grad_norm": 0.5469832397157314,
"learning_rate": 2.2022359053604654e-06,
"loss": 0.3374,
"step": 1803
},
{
"epoch": 2.1632,
"grad_norm": 0.6159816821332904,
"learning_rate": 2.1964551239705604e-06,
"loss": 0.35,
"step": 1804
},
{
"epoch": 2.1644,
"grad_norm": 0.5124786815177615,
"learning_rate": 2.1906798033797276e-06,
"loss": 0.3526,
"step": 1805
},
{
"epoch": 2.1656,
"grad_norm": 0.5140149212193701,
"learning_rate": 2.1849099548372492e-06,
"loss": 0.273,
"step": 1806
},
{
"epoch": 2.1668,
"grad_norm": 0.5062178694235752,
"learning_rate": 2.179145589581747e-06,
"loss": 0.3373,
"step": 1807
},
{
"epoch": 2.168,
"grad_norm": 0.49234615502579876,
"learning_rate": 2.1733867188411606e-06,
"loss": 0.3294,
"step": 1808
},
{
"epoch": 2.1692,
"grad_norm": 0.48285204019246003,
"learning_rate": 2.167633353832734e-06,
"loss": 0.3167,
"step": 1809
},
{
"epoch": 2.1704,
"grad_norm": 0.653391977737391,
"learning_rate": 2.1618855057629804e-06,
"loss": 0.3188,
"step": 1810
},
{
"epoch": 2.1716,
"grad_norm": 0.568596597442444,
"learning_rate": 2.156143185827671e-06,
"loss": 0.304,
"step": 1811
},
{
"epoch": 2.1728,
"grad_norm": 0.43840364040639196,
"learning_rate": 2.1504064052118095e-06,
"loss": 0.344,
"step": 1812
},
{
"epoch": 2.174,
"grad_norm": 0.5678920344689531,
"learning_rate": 2.144675175089606e-06,
"loss": 0.3455,
"step": 1813
},
{
"epoch": 2.1752,
"grad_norm": 0.4808681831796438,
"learning_rate": 2.1389495066244613e-06,
"loss": 0.3101,
"step": 1814
},
{
"epoch": 2.1764,
"grad_norm": 0.45444323442625073,
"learning_rate": 2.1332294109689446e-06,
"loss": 0.328,
"step": 1815
},
{
"epoch": 2.1776,
"grad_norm": 0.5508433787565589,
"learning_rate": 2.127514899264771e-06,
"loss": 0.3674,
"step": 1816
},
{
"epoch": 2.1788,
"grad_norm": 0.5532777904919081,
"learning_rate": 2.1218059826427727e-06,
"loss": 0.3107,
"step": 1817
},
{
"epoch": 2.18,
"grad_norm": 0.5284704056608788,
"learning_rate": 2.1161026722228932e-06,
"loss": 0.3064,
"step": 1818
},
{
"epoch": 2.1812,
"grad_norm": 0.5138803356821838,
"learning_rate": 2.110404979114149e-06,
"loss": 0.2991,
"step": 1819
},
{
"epoch": 2.1824,
"grad_norm": 0.6064872980601652,
"learning_rate": 2.104712914414615e-06,
"loss": 0.3046,
"step": 1820
},
{
"epoch": 2.1836,
"grad_norm": 0.5686013271119306,
"learning_rate": 2.0990264892114067e-06,
"loss": 0.3063,
"step": 1821
},
{
"epoch": 2.1848,
"grad_norm": 0.504330189065762,
"learning_rate": 2.093345714580656e-06,
"loss": 0.3626,
"step": 1822
},
{
"epoch": 2.186,
"grad_norm": 0.532977215227692,
"learning_rate": 2.0876706015874816e-06,
"loss": 0.3321,
"step": 1823
},
{
"epoch": 2.1872,
"grad_norm": 0.5065964923161952,
"learning_rate": 2.0820011612859825e-06,
"loss": 0.2779,
"step": 1824
},
{
"epoch": 2.1884,
"grad_norm": 0.5214169367813709,
"learning_rate": 2.076337404719203e-06,
"loss": 0.3354,
"step": 1825
},
{
"epoch": 2.1896,
"grad_norm": 0.540875044686939,
"learning_rate": 2.0706793429191156e-06,
"loss": 0.3246,
"step": 1826
},
{
"epoch": 2.1908,
"grad_norm": 0.5290340427211486,
"learning_rate": 2.0650269869066048e-06,
"loss": 0.3128,
"step": 1827
},
{
"epoch": 2.192,
"grad_norm": 0.4212512063016084,
"learning_rate": 2.0593803476914407e-06,
"loss": 0.317,
"step": 1828
},
{
"epoch": 2.1932,
"grad_norm": 0.564012613928127,
"learning_rate": 2.053739436272256e-06,
"loss": 0.3003,
"step": 1829
},
{
"epoch": 2.1944,
"grad_norm": 0.5006215016427719,
"learning_rate": 2.0481042636365243e-06,
"loss": 0.2887,
"step": 1830
},
{
"epoch": 2.1955999999999998,
"grad_norm": 0.5975039999001123,
"learning_rate": 2.0424748407605468e-06,
"loss": 0.3312,
"step": 1831
},
{
"epoch": 2.1968,
"grad_norm": 0.419444036679051,
"learning_rate": 2.036851178609423e-06,
"loss": 0.3395,
"step": 1832
},
{
"epoch": 2.198,
"grad_norm": 0.5296721376266754,
"learning_rate": 2.0312332881370294e-06,
"loss": 0.3052,
"step": 1833
},
{
"epoch": 2.1992,
"grad_norm": 0.49181852359041606,
"learning_rate": 2.0256211802860044e-06,
"loss": 0.3139,
"step": 1834
},
{
"epoch": 2.2004,
"grad_norm": 0.43499705360276436,
"learning_rate": 2.0200148659877185e-06,
"loss": 0.3025,
"step": 1835
},
{
"epoch": 2.2016,
"grad_norm": 0.521684040854764,
"learning_rate": 2.014414356162258e-06,
"loss": 0.3216,
"step": 1836
},
{
"epoch": 2.2028,
"grad_norm": 0.5246897918039743,
"learning_rate": 2.0088196617184065e-06,
"loss": 0.3255,
"step": 1837
},
{
"epoch": 2.204,
"grad_norm": 0.6951126296040991,
"learning_rate": 2.00323079355362e-06,
"loss": 0.3179,
"step": 1838
},
{
"epoch": 2.2052,
"grad_norm": 0.4991834103517274,
"learning_rate": 1.997647762554e-06,
"loss": 0.3121,
"step": 1839
},
{
"epoch": 2.2064,
"grad_norm": 0.6435612169425046,
"learning_rate": 1.992070579594288e-06,
"loss": 0.3215,
"step": 1840
},
{
"epoch": 2.2076000000000002,
"grad_norm": 0.5550473315339096,
"learning_rate": 1.9864992555378256e-06,
"loss": 0.3068,
"step": 1841
},
{
"epoch": 2.2088,
"grad_norm": 0.519085929027261,
"learning_rate": 1.9809338012365438e-06,
"loss": 0.3147,
"step": 1842
},
{
"epoch": 2.21,
"grad_norm": 0.5398378021114274,
"learning_rate": 1.9753742275309456e-06,
"loss": 0.3109,
"step": 1843
},
{
"epoch": 2.2112,
"grad_norm": 0.496031217404602,
"learning_rate": 1.9698205452500772e-06,
"loss": 0.3309,
"step": 1844
},
{
"epoch": 2.2124,
"grad_norm": 0.5091006413823043,
"learning_rate": 1.9642727652115056e-06,
"loss": 0.3353,
"step": 1845
},
{
"epoch": 2.2136,
"grad_norm": 0.5127551039153834,
"learning_rate": 1.9587308982213077e-06,
"loss": 0.3128,
"step": 1846
},
{
"epoch": 2.2148,
"grad_norm": 0.4801485456191148,
"learning_rate": 1.953194955074038e-06,
"loss": 0.3157,
"step": 1847
},
{
"epoch": 2.216,
"grad_norm": 0.5294673058010643,
"learning_rate": 1.9476649465527116e-06,
"loss": 0.3341,
"step": 1848
},
{
"epoch": 2.2172,
"grad_norm": 0.5316094839936794,
"learning_rate": 1.942140883428788e-06,
"loss": 0.2717,
"step": 1849
},
{
"epoch": 2.2184,
"grad_norm": 0.5556388765318808,
"learning_rate": 1.936622776462147e-06,
"loss": 0.3404,
"step": 1850
},
{
"epoch": 2.2196,
"grad_norm": 0.503236246918105,
"learning_rate": 1.931110636401062e-06,
"loss": 0.3213,
"step": 1851
},
{
"epoch": 2.2208,
"grad_norm": 0.5690711638502806,
"learning_rate": 1.925604473982185e-06,
"loss": 0.3517,
"step": 1852
},
{
"epoch": 2.222,
"grad_norm": 0.5101343946978072,
"learning_rate": 1.9201042999305276e-06,
"loss": 0.3432,
"step": 1853
},
{
"epoch": 2.2232,
"grad_norm": 0.604211155077798,
"learning_rate": 1.914610124959437e-06,
"loss": 0.3373,
"step": 1854
},
{
"epoch": 2.2244,
"grad_norm": 0.4884119407257996,
"learning_rate": 1.9091219597705694e-06,
"loss": 0.32,
"step": 1855
},
{
"epoch": 2.2256,
"grad_norm": 0.47970146777377404,
"learning_rate": 1.9036398150538842e-06,
"loss": 0.2938,
"step": 1856
},
{
"epoch": 2.2268,
"grad_norm": 0.4769871704809297,
"learning_rate": 1.898163701487607e-06,
"loss": 0.3154,
"step": 1857
},
{
"epoch": 2.228,
"grad_norm": 0.5381238546223847,
"learning_rate": 1.8926936297382148e-06,
"loss": 0.3169,
"step": 1858
},
{
"epoch": 2.2292,
"grad_norm": 0.3847813549375159,
"learning_rate": 1.8872296104604255e-06,
"loss": 0.3497,
"step": 1859
},
{
"epoch": 2.2304,
"grad_norm": 0.5054743631613623,
"learning_rate": 1.8817716542971593e-06,
"loss": 0.3585,
"step": 1860
},
{
"epoch": 2.2316,
"grad_norm": 0.5095273965254722,
"learning_rate": 1.8763197718795262e-06,
"loss": 0.3253,
"step": 1861
},
{
"epoch": 2.2328,
"grad_norm": 0.5372747138212405,
"learning_rate": 1.8708739738268133e-06,
"loss": 0.3198,
"step": 1862
},
{
"epoch": 2.234,
"grad_norm": 0.4273289084664602,
"learning_rate": 1.865434270746449e-06,
"loss": 0.299,
"step": 1863
},
{
"epoch": 2.2352,
"grad_norm": 0.5057411248655699,
"learning_rate": 1.8600006732339892e-06,
"loss": 0.3223,
"step": 1864
},
{
"epoch": 2.2364,
"grad_norm": 0.5329286312340235,
"learning_rate": 1.8545731918731074e-06,
"loss": 0.3095,
"step": 1865
},
{
"epoch": 2.2376,
"grad_norm": 0.42512261208802066,
"learning_rate": 1.8491518372355538e-06,
"loss": 0.3344,
"step": 1866
},
{
"epoch": 2.2388,
"grad_norm": 0.5144880443849901,
"learning_rate": 1.8437366198811463e-06,
"loss": 0.3155,
"step": 1867
},
{
"epoch": 2.24,
"grad_norm": 0.546562593617825,
"learning_rate": 1.838327550357753e-06,
"loss": 0.3274,
"step": 1868
},
{
"epoch": 2.2412,
"grad_norm": 0.46168655023050265,
"learning_rate": 1.8329246392012622e-06,
"loss": 0.3095,
"step": 1869
},
{
"epoch": 2.2424,
"grad_norm": 0.5873141909299401,
"learning_rate": 1.8275278969355714e-06,
"loss": 0.3193,
"step": 1870
},
{
"epoch": 2.2436,
"grad_norm": 0.5126551069055825,
"learning_rate": 1.8221373340725568e-06,
"loss": 0.332,
"step": 1871
},
{
"epoch": 2.2448,
"grad_norm": 0.473975809996108,
"learning_rate": 1.8167529611120648e-06,
"loss": 0.3331,
"step": 1872
},
{
"epoch": 2.246,
"grad_norm": 0.5973549893034205,
"learning_rate": 1.811374788541878e-06,
"loss": 0.3499,
"step": 1873
},
{
"epoch": 2.2472,
"grad_norm": 0.4937805794383019,
"learning_rate": 1.8060028268377088e-06,
"loss": 0.3591,
"step": 1874
},
{
"epoch": 2.2484,
"grad_norm": 0.5520047912990497,
"learning_rate": 1.8006370864631644e-06,
"loss": 0.3091,
"step": 1875
},
{
"epoch": 2.2496,
"grad_norm": 0.5280603840791102,
"learning_rate": 1.7952775778697418e-06,
"loss": 0.3822,
"step": 1876
},
{
"epoch": 2.2508,
"grad_norm": 0.4747984575549771,
"learning_rate": 1.7899243114967918e-06,
"loss": 0.3414,
"step": 1877
},
{
"epoch": 2.252,
"grad_norm": 0.5008025728171192,
"learning_rate": 1.7845772977715148e-06,
"loss": 0.3217,
"step": 1878
},
{
"epoch": 2.2532,
"grad_norm": 0.5460878686864432,
"learning_rate": 1.7792365471089252e-06,
"loss": 0.3116,
"step": 1879
},
{
"epoch": 2.2544,
"grad_norm": 0.4706181673801985,
"learning_rate": 1.773902069911838e-06,
"loss": 0.3054,
"step": 1880
},
{
"epoch": 2.2556,
"grad_norm": 0.5659208409724967,
"learning_rate": 1.7685738765708576e-06,
"loss": 0.3284,
"step": 1881
},
{
"epoch": 2.2568,
"grad_norm": 0.5914803959669342,
"learning_rate": 1.7632519774643391e-06,
"loss": 0.3247,
"step": 1882
},
{
"epoch": 2.258,
"grad_norm": 0.4904554286879897,
"learning_rate": 1.7579363829583794e-06,
"loss": 0.3047,
"step": 1883
},
{
"epoch": 2.2592,
"grad_norm": 0.5227748841136812,
"learning_rate": 1.7526271034067993e-06,
"loss": 0.3024,
"step": 1884
},
{
"epoch": 2.2604,
"grad_norm": 0.5318646011263115,
"learning_rate": 1.7473241491511139e-06,
"loss": 0.3072,
"step": 1885
},
{
"epoch": 2.2616,
"grad_norm": 0.5068337164319224,
"learning_rate": 1.7420275305205214e-06,
"loss": 0.3037,
"step": 1886
},
{
"epoch": 2.2628,
"grad_norm": 0.6220942329692476,
"learning_rate": 1.7367372578318797e-06,
"loss": 0.3223,
"step": 1887
},
{
"epoch": 2.2640000000000002,
"grad_norm": 0.5531572497923152,
"learning_rate": 1.7314533413896833e-06,
"loss": 0.3059,
"step": 1888
},
{
"epoch": 2.2652,
"grad_norm": 0.5635377817414549,
"learning_rate": 1.7261757914860456e-06,
"loss": 0.3399,
"step": 1889
},
{
"epoch": 2.2664,
"grad_norm": 0.5515908215511565,
"learning_rate": 1.720904618400684e-06,
"loss": 0.3133,
"step": 1890
},
{
"epoch": 2.2676,
"grad_norm": 0.4711116326295099,
"learning_rate": 1.7156398324008871e-06,
"loss": 0.3488,
"step": 1891
},
{
"epoch": 2.2688,
"grad_norm": 0.5169505814378155,
"learning_rate": 1.7103814437415105e-06,
"loss": 0.3119,
"step": 1892
},
{
"epoch": 2.27,
"grad_norm": 0.4808802832107604,
"learning_rate": 1.7051294626649462e-06,
"loss": 0.2876,
"step": 1893
},
{
"epoch": 2.2712,
"grad_norm": 0.4965800820258115,
"learning_rate": 1.6998838994011041e-06,
"loss": 0.3081,
"step": 1894
},
{
"epoch": 2.2724,
"grad_norm": 0.5765108347910463,
"learning_rate": 1.6946447641673907e-06,
"loss": 0.2977,
"step": 1895
},
{
"epoch": 2.2736,
"grad_norm": 0.5747764419760193,
"learning_rate": 1.6894120671686986e-06,
"loss": 0.3082,
"step": 1896
},
{
"epoch": 2.2748,
"grad_norm": 0.4979801826768883,
"learning_rate": 1.6841858185973775e-06,
"loss": 0.307,
"step": 1897
},
{
"epoch": 2.276,
"grad_norm": 0.5668296391608482,
"learning_rate": 1.6789660286332132e-06,
"loss": 0.349,
"step": 1898
},
{
"epoch": 2.2772,
"grad_norm": 0.5273497007825191,
"learning_rate": 1.6737527074434135e-06,
"loss": 0.3298,
"step": 1899
},
{
"epoch": 2.2784,
"grad_norm": 0.5205699456340483,
"learning_rate": 1.6685458651825892e-06,
"loss": 0.3269,
"step": 1900
},
{
"epoch": 2.2796,
"grad_norm": 0.5603073017723463,
"learning_rate": 1.6633455119927256e-06,
"loss": 0.3002,
"step": 1901
},
{
"epoch": 2.2808,
"grad_norm": 0.49085075677697326,
"learning_rate": 1.658151658003172e-06,
"loss": 0.3231,
"step": 1902
},
{
"epoch": 2.282,
"grad_norm": 0.47995130663114505,
"learning_rate": 1.6529643133306212e-06,
"loss": 0.3055,
"step": 1903
},
{
"epoch": 2.2832,
"grad_norm": 0.4808506099960846,
"learning_rate": 1.647783488079081e-06,
"loss": 0.3307,
"step": 1904
},
{
"epoch": 2.2843999999999998,
"grad_norm": 0.6235221866733798,
"learning_rate": 1.6426091923398619e-06,
"loss": 0.3389,
"step": 1905
},
{
"epoch": 2.2856,
"grad_norm": 0.47445141356388904,
"learning_rate": 1.6374414361915613e-06,
"loss": 0.2944,
"step": 1906
},
{
"epoch": 2.2868,
"grad_norm": 0.4774585184186103,
"learning_rate": 1.6322802297000306e-06,
"loss": 0.3438,
"step": 1907
},
{
"epoch": 2.288,
"grad_norm": 0.5375942289912291,
"learning_rate": 1.6271255829183702e-06,
"loss": 0.3348,
"step": 1908
},
{
"epoch": 2.2892,
"grad_norm": 0.40086256227082395,
"learning_rate": 1.6219775058869019e-06,
"loss": 0.3448,
"step": 1909
},
{
"epoch": 2.2904,
"grad_norm": 0.5172335820787956,
"learning_rate": 1.6168360086331498e-06,
"loss": 0.3233,
"step": 1910
},
{
"epoch": 2.2916,
"grad_norm": 0.5207471934020876,
"learning_rate": 1.6117011011718188e-06,
"loss": 0.3255,
"step": 1911
},
{
"epoch": 2.2928,
"grad_norm": 0.4878094924131688,
"learning_rate": 1.6065727935047837e-06,
"loss": 0.3286,
"step": 1912
},
{
"epoch": 2.294,
"grad_norm": 0.5950357137095884,
"learning_rate": 1.6014510956210632e-06,
"loss": 0.3283,
"step": 1913
},
{
"epoch": 2.2952,
"grad_norm": 0.48086928299992526,
"learning_rate": 1.5963360174967956e-06,
"loss": 0.3503,
"step": 1914
},
{
"epoch": 2.2964,
"grad_norm": 0.5892110607046107,
"learning_rate": 1.5912275690952339e-06,
"loss": 0.3245,
"step": 1915
},
{
"epoch": 2.2976,
"grad_norm": 0.5028560710697326,
"learning_rate": 1.5861257603667106e-06,
"loss": 0.3065,
"step": 1916
},
{
"epoch": 2.2988,
"grad_norm": 0.5189757571086189,
"learning_rate": 1.581030601248626e-06,
"loss": 0.3135,
"step": 1917
},
{
"epoch": 2.3,
"grad_norm": 0.4491773427040121,
"learning_rate": 1.5759421016654314e-06,
"loss": 0.318,
"step": 1918
},
{
"epoch": 2.3012,
"grad_norm": 0.4874058563272879,
"learning_rate": 1.570860271528607e-06,
"loss": 0.3001,
"step": 1919
},
{
"epoch": 2.3024,
"grad_norm": 0.4852758364086213,
"learning_rate": 1.5657851207366359e-06,
"loss": 0.3135,
"step": 1920
},
{
"epoch": 2.3036,
"grad_norm": 0.5759656835583162,
"learning_rate": 1.5607166591749995e-06,
"loss": 0.3452,
"step": 1921
},
{
"epoch": 2.3048,
"grad_norm": 0.5207395067658368,
"learning_rate": 1.555654896716144e-06,
"loss": 0.32,
"step": 1922
},
{
"epoch": 2.306,
"grad_norm": 0.452747411117712,
"learning_rate": 1.5505998432194658e-06,
"loss": 0.3405,
"step": 1923
},
{
"epoch": 2.3072,
"grad_norm": 0.5412198304899934,
"learning_rate": 1.5455515085312984e-06,
"loss": 0.3322,
"step": 1924
},
{
"epoch": 2.3084,
"grad_norm": 0.49544095265814697,
"learning_rate": 1.5405099024848874e-06,
"loss": 0.3124,
"step": 1925
},
{
"epoch": 2.3096,
"grad_norm": 0.5183316214030841,
"learning_rate": 1.5354750349003694e-06,
"loss": 0.3495,
"step": 1926
},
{
"epoch": 2.3108,
"grad_norm": 0.5463151196073748,
"learning_rate": 1.5304469155847556e-06,
"loss": 0.3434,
"step": 1927
},
{
"epoch": 2.312,
"grad_norm": 0.5516143512305192,
"learning_rate": 1.5254255543319168e-06,
"loss": 0.3168,
"step": 1928
},
{
"epoch": 2.3132,
"grad_norm": 0.45573917604459735,
"learning_rate": 1.5204109609225553e-06,
"loss": 0.3195,
"step": 1929
},
{
"epoch": 2.3144,
"grad_norm": 0.5469056634720862,
"learning_rate": 1.5154031451241952e-06,
"loss": 0.3411,
"step": 1930
},
{
"epoch": 2.3156,
"grad_norm": 0.5158849784855729,
"learning_rate": 1.5104021166911582e-06,
"loss": 0.3165,
"step": 1931
},
{
"epoch": 2.3168,
"grad_norm": 0.5544014080581716,
"learning_rate": 1.5054078853645432e-06,
"loss": 0.3089,
"step": 1932
},
{
"epoch": 2.318,
"grad_norm": 0.4524564910062869,
"learning_rate": 1.5004204608722088e-06,
"loss": 0.3152,
"step": 1933
},
{
"epoch": 2.3192,
"grad_norm": 0.502269133931021,
"learning_rate": 1.495439852928759e-06,
"loss": 0.295,
"step": 1934
},
{
"epoch": 2.3204000000000002,
"grad_norm": 0.4800443904019531,
"learning_rate": 1.4904660712355207e-06,
"loss": 0.3223,
"step": 1935
},
{
"epoch": 2.3216,
"grad_norm": 0.4711817206662842,
"learning_rate": 1.4854991254805179e-06,
"loss": 0.3182,
"step": 1936
},
{
"epoch": 2.3228,
"grad_norm": 0.5204593299602398,
"learning_rate": 1.4805390253384683e-06,
"loss": 0.3432,
"step": 1937
},
{
"epoch": 2.324,
"grad_norm": 0.5634364961018137,
"learning_rate": 1.4755857804707485e-06,
"loss": 0.3003,
"step": 1938
},
{
"epoch": 2.3252,
"grad_norm": 0.4498885389670753,
"learning_rate": 1.4706394005253838e-06,
"loss": 0.3033,
"step": 1939
},
{
"epoch": 2.3264,
"grad_norm": 0.5581196261806362,
"learning_rate": 1.465699895137031e-06,
"loss": 0.3033,
"step": 1940
},
{
"epoch": 2.3276,
"grad_norm": 0.4408958316492097,
"learning_rate": 1.4607672739269552e-06,
"loss": 0.3073,
"step": 1941
},
{
"epoch": 2.3288,
"grad_norm": 0.6240652317829313,
"learning_rate": 1.455841546503009e-06,
"loss": 0.3215,
"step": 1942
},
{
"epoch": 2.33,
"grad_norm": 0.5492997699864376,
"learning_rate": 1.450922722459623e-06,
"loss": 0.3279,
"step": 1943
},
{
"epoch": 2.3312,
"grad_norm": 0.49666520199305403,
"learning_rate": 1.446010811377776e-06,
"loss": 0.332,
"step": 1944
},
{
"epoch": 2.3324,
"grad_norm": 0.5916923146968442,
"learning_rate": 1.4411058228249824e-06,
"loss": 0.3188,
"step": 1945
},
{
"epoch": 2.3336,
"grad_norm": 0.5692144923101109,
"learning_rate": 1.4362077663552754e-06,
"loss": 0.3205,
"step": 1946
},
{
"epoch": 2.3348,
"grad_norm": 0.5244770882576342,
"learning_rate": 1.4313166515091863e-06,
"loss": 0.3567,
"step": 1947
},
{
"epoch": 2.336,
"grad_norm": 0.4985866219770014,
"learning_rate": 1.4264324878137204e-06,
"loss": 0.2955,
"step": 1948
},
{
"epoch": 2.3372,
"grad_norm": 0.5124823393836703,
"learning_rate": 1.421555284782349e-06,
"loss": 0.3044,
"step": 1949
},
{
"epoch": 2.3384,
"grad_norm": 0.5408319376191771,
"learning_rate": 1.4166850519149794e-06,
"loss": 0.3171,
"step": 1950
},
{
"epoch": 2.3396,
"grad_norm": 0.5398033571111165,
"learning_rate": 1.41182179869795e-06,
"loss": 0.3416,
"step": 1951
},
{
"epoch": 2.3407999999999998,
"grad_norm": 0.7408763803046092,
"learning_rate": 1.406965534603995e-06,
"loss": 0.2988,
"step": 1952
},
{
"epoch": 2.342,
"grad_norm": 0.48957085058491534,
"learning_rate": 1.4021162690922441e-06,
"loss": 0.2904,
"step": 1953
},
{
"epoch": 2.3432,
"grad_norm": 0.4800797062711741,
"learning_rate": 1.397274011608189e-06,
"loss": 0.3394,
"step": 1954
},
{
"epoch": 2.3444,
"grad_norm": 0.504636425789252,
"learning_rate": 1.3924387715836706e-06,
"loss": 0.3306,
"step": 1955
},
{
"epoch": 2.3456,
"grad_norm": 0.5273255459333012,
"learning_rate": 1.3876105584368653e-06,
"loss": 0.3111,
"step": 1956
},
{
"epoch": 2.3468,
"grad_norm": 0.46242283582128896,
"learning_rate": 1.3827893815722614e-06,
"loss": 0.3308,
"step": 1957
},
{
"epoch": 2.348,
"grad_norm": 0.5268676502741816,
"learning_rate": 1.3779752503806375e-06,
"loss": 0.3434,
"step": 1958
},
{
"epoch": 2.3492,
"grad_norm": 0.5821097945236201,
"learning_rate": 1.3731681742390558e-06,
"loss": 0.3084,
"step": 1959
},
{
"epoch": 2.3504,
"grad_norm": 0.5469587875056272,
"learning_rate": 1.368368162510829e-06,
"loss": 0.302,
"step": 1960
},
{
"epoch": 2.3516,
"grad_norm": 0.4634982197304083,
"learning_rate": 1.363575224545512e-06,
"loss": 0.3182,
"step": 1961
},
{
"epoch": 2.3528000000000002,
"grad_norm": 0.5274355185700257,
"learning_rate": 1.3587893696788868e-06,
"loss": 0.2801,
"step": 1962
},
{
"epoch": 2.354,
"grad_norm": 0.41061222873885617,
"learning_rate": 1.3540106072329323e-06,
"loss": 0.3271,
"step": 1963
},
{
"epoch": 2.3552,
"grad_norm": 0.5577455688560952,
"learning_rate": 1.349238946515813e-06,
"loss": 0.3324,
"step": 1964
},
{
"epoch": 2.3564,
"grad_norm": 0.5086040856439964,
"learning_rate": 1.344474396821865e-06,
"loss": 0.3383,
"step": 1965
},
{
"epoch": 2.3576,
"grad_norm": 0.45513590551982486,
"learning_rate": 1.3397169674315668e-06,
"loss": 0.3029,
"step": 1966
},
{
"epoch": 2.3588,
"grad_norm": 0.5032377873503844,
"learning_rate": 1.3349666676115358e-06,
"loss": 0.3161,
"step": 1967
},
{
"epoch": 2.36,
"grad_norm": 0.5429418662908604,
"learning_rate": 1.3302235066144948e-06,
"loss": 0.3173,
"step": 1968
},
{
"epoch": 2.3612,
"grad_norm": 0.582433350815691,
"learning_rate": 1.3254874936792672e-06,
"loss": 0.3203,
"step": 1969
},
{
"epoch": 2.3624,
"grad_norm": 0.5061141475840918,
"learning_rate": 1.3207586380307486e-06,
"loss": 0.3133,
"step": 1970
},
{
"epoch": 2.3636,
"grad_norm": 0.38935634371222433,
"learning_rate": 1.3160369488798984e-06,
"loss": 0.289,
"step": 1971
},
{
"epoch": 2.3648,
"grad_norm": 0.49040182272898925,
"learning_rate": 1.3113224354237113e-06,
"loss": 0.2872,
"step": 1972
},
{
"epoch": 2.366,
"grad_norm": 0.49662477301690094,
"learning_rate": 1.306615106845211e-06,
"loss": 0.3028,
"step": 1973
},
{
"epoch": 2.3672,
"grad_norm": 0.5614718485198186,
"learning_rate": 1.30191497231342e-06,
"loss": 0.2845,
"step": 1974
},
{
"epoch": 2.3684,
"grad_norm": 0.5474441188846769,
"learning_rate": 1.2972220409833552e-06,
"loss": 0.3182,
"step": 1975
},
{
"epoch": 2.3696,
"grad_norm": 0.44863953590190264,
"learning_rate": 1.2925363219959958e-06,
"loss": 0.3233,
"step": 1976
},
{
"epoch": 2.3708,
"grad_norm": 0.5859727221445215,
"learning_rate": 1.2878578244782775e-06,
"loss": 0.338,
"step": 1977
},
{
"epoch": 2.372,
"grad_norm": 0.518472285200413,
"learning_rate": 1.2831865575430702e-06,
"loss": 0.3103,
"step": 1978
},
{
"epoch": 2.3731999999999998,
"grad_norm": 0.4486954376925882,
"learning_rate": 1.2785225302891568e-06,
"loss": 0.3036,
"step": 1979
},
{
"epoch": 2.3744,
"grad_norm": 0.5332520800656768,
"learning_rate": 1.2738657518012188e-06,
"loss": 0.3455,
"step": 1980
},
{
"epoch": 2.3756,
"grad_norm": 0.5180503448026015,
"learning_rate": 1.2692162311498219e-06,
"loss": 0.3315,
"step": 1981
},
{
"epoch": 2.3768000000000002,
"grad_norm": 0.5503669228226378,
"learning_rate": 1.2645739773913911e-06,
"loss": 0.3106,
"step": 1982
},
{
"epoch": 2.378,
"grad_norm": 0.5572704229738278,
"learning_rate": 1.259938999568196e-06,
"loss": 0.3235,
"step": 1983
},
{
"epoch": 2.3792,
"grad_norm": 0.49223457619635097,
"learning_rate": 1.2553113067083417e-06,
"loss": 0.323,
"step": 1984
},
{
"epoch": 2.3804,
"grad_norm": 0.5298160573023049,
"learning_rate": 1.2506909078257357e-06,
"loss": 0.3113,
"step": 1985
},
{
"epoch": 2.3816,
"grad_norm": 0.4739126163553427,
"learning_rate": 1.2460778119200778e-06,
"loss": 0.3063,
"step": 1986
},
{
"epoch": 2.3828,
"grad_norm": 0.5756760439695214,
"learning_rate": 1.24147202797685e-06,
"loss": 0.29,
"step": 1987
},
{
"epoch": 2.384,
"grad_norm": 0.47841470066730546,
"learning_rate": 1.236873564967284e-06,
"loss": 0.3334,
"step": 1988
},
{
"epoch": 2.3852,
"grad_norm": 0.5372964333839603,
"learning_rate": 1.2322824318483568e-06,
"loss": 0.3182,
"step": 1989
},
{
"epoch": 2.3864,
"grad_norm": 0.5360611358567483,
"learning_rate": 1.227698637562768e-06,
"loss": 0.3028,
"step": 1990
},
{
"epoch": 2.3876,
"grad_norm": 0.546051270415823,
"learning_rate": 1.2231221910389196e-06,
"loss": 0.3365,
"step": 1991
},
{
"epoch": 2.3888,
"grad_norm": 0.5355902634690551,
"learning_rate": 1.2185531011909008e-06,
"loss": 0.3134,
"step": 1992
},
{
"epoch": 2.39,
"grad_norm": 0.5667030431688215,
"learning_rate": 1.2139913769184757e-06,
"loss": 0.3551,
"step": 1993
},
{
"epoch": 2.3912,
"grad_norm": 0.5586223985170625,
"learning_rate": 1.2094370271070599e-06,
"loss": 0.2898,
"step": 1994
},
{
"epoch": 2.3924,
"grad_norm": 0.49938487817710525,
"learning_rate": 1.2048900606277036e-06,
"loss": 0.3374,
"step": 1995
},
{
"epoch": 2.3936,
"grad_norm": 0.5172811558524574,
"learning_rate": 1.2003504863370746e-06,
"loss": 0.3239,
"step": 1996
},
{
"epoch": 2.3948,
"grad_norm": 0.4448576698489582,
"learning_rate": 1.195818313077447e-06,
"loss": 0.2964,
"step": 1997
},
{
"epoch": 2.396,
"grad_norm": 0.5067804987087742,
"learning_rate": 1.1912935496766719e-06,
"loss": 0.3149,
"step": 1998
},
{
"epoch": 2.3971999999999998,
"grad_norm": 0.5289972726643701,
"learning_rate": 1.186776204948173e-06,
"loss": 0.3197,
"step": 1999
},
{
"epoch": 2.3984,
"grad_norm": 0.48058445208828016,
"learning_rate": 1.182266287690924e-06,
"loss": 0.3183,
"step": 2000
},
{
"epoch": 2.3996,
"grad_norm": 0.5330873826333208,
"learning_rate": 1.177763806689427e-06,
"loss": 0.2727,
"step": 2001
},
{
"epoch": 2.4008,
"grad_norm": 0.4467421647008868,
"learning_rate": 1.173268770713701e-06,
"loss": 0.3235,
"step": 2002
},
{
"epoch": 2.402,
"grad_norm": 0.5189215617155751,
"learning_rate": 1.1687811885192662e-06,
"loss": 0.3345,
"step": 2003
},
{
"epoch": 2.4032,
"grad_norm": 0.5236509042131071,
"learning_rate": 1.16430106884712e-06,
"loss": 0.2923,
"step": 2004
},
{
"epoch": 2.4044,
"grad_norm": 0.4174001768905615,
"learning_rate": 1.159828420423728e-06,
"loss": 0.3329,
"step": 2005
},
{
"epoch": 2.4056,
"grad_norm": 0.48556976851826156,
"learning_rate": 1.1553632519610025e-06,
"loss": 0.3181,
"step": 2006
},
{
"epoch": 2.4068,
"grad_norm": 0.5333581260392112,
"learning_rate": 1.1509055721562839e-06,
"loss": 0.3558,
"step": 2007
},
{
"epoch": 2.408,
"grad_norm": 0.5504819445821775,
"learning_rate": 1.1464553896923264e-06,
"loss": 0.3366,
"step": 2008
},
{
"epoch": 2.4092000000000002,
"grad_norm": 0.6575472559330863,
"learning_rate": 1.1420127132372839e-06,
"loss": 0.3023,
"step": 2009
},
{
"epoch": 2.4104,
"grad_norm": 0.5224772296793633,
"learning_rate": 1.1375775514446846e-06,
"loss": 0.3487,
"step": 2010
},
{
"epoch": 2.4116,
"grad_norm": 0.5362750622696038,
"learning_rate": 1.1331499129534252e-06,
"loss": 0.3985,
"step": 2011
},
{
"epoch": 2.4128,
"grad_norm": 0.6480867835273206,
"learning_rate": 1.128729806387746e-06,
"loss": 0.3078,
"step": 2012
},
{
"epoch": 2.414,
"grad_norm": 0.510972758166725,
"learning_rate": 1.124317240357216e-06,
"loss": 0.374,
"step": 2013
},
{
"epoch": 2.4152,
"grad_norm": 0.4353416689772134,
"learning_rate": 1.119912223456715e-06,
"loss": 0.3434,
"step": 2014
},
{
"epoch": 2.4164,
"grad_norm": 0.4992041891223879,
"learning_rate": 1.1155147642664217e-06,
"loss": 0.2941,
"step": 2015
},
{
"epoch": 2.4176,
"grad_norm": 0.4981271993131891,
"learning_rate": 1.1111248713517935e-06,
"loss": 0.3399,
"step": 2016
},
{
"epoch": 2.4188,
"grad_norm": 0.5315501345593295,
"learning_rate": 1.1067425532635463e-06,
"loss": 0.3059,
"step": 2017
},
{
"epoch": 2.42,
"grad_norm": 0.5272127156791281,
"learning_rate": 1.1023678185376474e-06,
"loss": 0.3252,
"step": 2018
},
{
"epoch": 2.4212,
"grad_norm": 0.5509389015682687,
"learning_rate": 1.0980006756952882e-06,
"loss": 0.3637,
"step": 2019
},
{
"epoch": 2.4224,
"grad_norm": 0.5577993072539054,
"learning_rate": 1.0936411332428732e-06,
"loss": 0.2761,
"step": 2020
},
{
"epoch": 2.4236,
"grad_norm": 0.47620274555170333,
"learning_rate": 1.089289199672004e-06,
"loss": 0.3058,
"step": 2021
},
{
"epoch": 2.4248,
"grad_norm": 0.4872285887781436,
"learning_rate": 1.084944883459464e-06,
"loss": 0.3357,
"step": 2022
},
{
"epoch": 2.426,
"grad_norm": 0.6092064550019151,
"learning_rate": 1.0806081930671947e-06,
"loss": 0.3375,
"step": 2023
},
{
"epoch": 2.4272,
"grad_norm": 0.45105808930779195,
"learning_rate": 1.0762791369422838e-06,
"loss": 0.3343,
"step": 2024
},
{
"epoch": 2.4284,
"grad_norm": 0.5673808224464912,
"learning_rate": 1.0719577235169537e-06,
"loss": 0.3471,
"step": 2025
},
{
"epoch": 2.4295999999999998,
"grad_norm": 0.39845689327313377,
"learning_rate": 1.0676439612085353e-06,
"loss": 0.2865,
"step": 2026
},
{
"epoch": 2.4308,
"grad_norm": 0.4922044208177292,
"learning_rate": 1.0633378584194593e-06,
"loss": 0.3436,
"step": 2027
},
{
"epoch": 2.432,
"grad_norm": 0.48469669953570305,
"learning_rate": 1.059039423537237e-06,
"loss": 0.3211,
"step": 2028
},
{
"epoch": 2.4332,
"grad_norm": 0.47164823939079553,
"learning_rate": 1.054748664934443e-06,
"loss": 0.3072,
"step": 2029
},
{
"epoch": 2.4344,
"grad_norm": 0.5096766564767695,
"learning_rate": 1.0504655909686978e-06,
"loss": 0.2921,
"step": 2030
},
{
"epoch": 2.4356,
"grad_norm": 0.5908518033772256,
"learning_rate": 1.0461902099826577e-06,
"loss": 0.2869,
"step": 2031
},
{
"epoch": 2.4368,
"grad_norm": 0.434773091611653,
"learning_rate": 1.0419225303039943e-06,
"loss": 0.3247,
"step": 2032
},
{
"epoch": 2.438,
"grad_norm": 0.46572147262123414,
"learning_rate": 1.0376625602453733e-06,
"loss": 0.2892,
"step": 2033
},
{
"epoch": 2.4392,
"grad_norm": 0.46387952563383317,
"learning_rate": 1.0334103081044504e-06,
"loss": 0.2989,
"step": 2034
},
{
"epoch": 2.4404,
"grad_norm": 0.431710967434786,
"learning_rate": 1.0291657821638435e-06,
"loss": 0.3158,
"step": 2035
},
{
"epoch": 2.4416,
"grad_norm": 0.5109671398031306,
"learning_rate": 1.024928990691121e-06,
"loss": 0.3335,
"step": 2036
},
{
"epoch": 2.4428,
"grad_norm": 0.5032323659686736,
"learning_rate": 1.0206999419387881e-06,
"loss": 0.3047,
"step": 2037
},
{
"epoch": 2.444,
"grad_norm": 0.6420893900320955,
"learning_rate": 1.0164786441442698e-06,
"loss": 0.3465,
"step": 2038
},
{
"epoch": 2.4452,
"grad_norm": 0.45454109064614695,
"learning_rate": 1.0122651055298898e-06,
"loss": 0.3011,
"step": 2039
},
{
"epoch": 2.4464,
"grad_norm": 0.5729786871162398,
"learning_rate": 1.0080593343028621e-06,
"loss": 0.32,
"step": 2040
},
{
"epoch": 2.4476,
"grad_norm": 0.44189500277710014,
"learning_rate": 1.0038613386552687e-06,
"loss": 0.345,
"step": 2041
},
{
"epoch": 2.4488,
"grad_norm": 0.5957486603332705,
"learning_rate": 9.996711267640451e-07,
"loss": 0.3576,
"step": 2042
},
{
"epoch": 2.45,
"grad_norm": 0.43553588307583163,
"learning_rate": 9.95488706790969e-07,
"loss": 0.2917,
"step": 2043
},
{
"epoch": 2.4512,
"grad_norm": 0.4703442233300642,
"learning_rate": 9.913140868826405e-07,
"loss": 0.2682,
"step": 2044
},
{
"epoch": 2.4524,
"grad_norm": 0.5954017228399653,
"learning_rate": 9.871472751704625e-07,
"loss": 0.3111,
"step": 2045
},
{
"epoch": 2.4536,
"grad_norm": 0.4958235772248686,
"learning_rate": 9.829882797706336e-07,
"loss": 0.3283,
"step": 2046
},
{
"epoch": 2.4548,
"grad_norm": 0.5337507765326563,
"learning_rate": 9.788371087841236e-07,
"loss": 0.3177,
"step": 2047
},
{
"epoch": 2.456,
"grad_norm": 0.45094450039509193,
"learning_rate": 9.74693770296667e-07,
"loss": 0.3107,
"step": 2048
},
{
"epoch": 2.4572,
"grad_norm": 0.47640902644388333,
"learning_rate": 9.705582723787348e-07,
"loss": 0.3105,
"step": 2049
},
{
"epoch": 2.4584,
"grad_norm": 0.5551540280130849,
"learning_rate": 9.664306230855342e-07,
"loss": 0.3121,
"step": 2050
},
{
"epoch": 2.4596,
"grad_norm": 0.5296141106730923,
"learning_rate": 9.623108304569783e-07,
"loss": 0.3084,
"step": 2051
},
{
"epoch": 2.4608,
"grad_norm": 0.4503517127076986,
"learning_rate": 9.58198902517678e-07,
"loss": 0.2762,
"step": 2052
},
{
"epoch": 2.462,
"grad_norm": 0.49176682948014766,
"learning_rate": 9.540948472769278e-07,
"loss": 0.3189,
"step": 2053
},
{
"epoch": 2.4632,
"grad_norm": 0.5479259942541452,
"learning_rate": 9.499986727286869e-07,
"loss": 0.334,
"step": 2054
},
{
"epoch": 2.4644,
"grad_norm": 0.48701373761857364,
"learning_rate": 9.459103868515618e-07,
"loss": 0.2987,
"step": 2055
},
{
"epoch": 2.4656000000000002,
"grad_norm": 0.5904824064092714,
"learning_rate": 9.418299976087964e-07,
"loss": 0.314,
"step": 2056
},
{
"epoch": 2.4668,
"grad_norm": 0.5119778019908812,
"learning_rate": 9.377575129482513e-07,
"loss": 0.299,
"step": 2057
},
{
"epoch": 2.468,
"grad_norm": 0.5037314727759037,
"learning_rate": 9.336929408023887e-07,
"loss": 0.3288,
"step": 2058
},
{
"epoch": 2.4692,
"grad_norm": 0.45996754162030723,
"learning_rate": 9.29636289088266e-07,
"loss": 0.3327,
"step": 2059
},
{
"epoch": 2.4704,
"grad_norm": 0.5661953382786764,
"learning_rate": 9.255875657075053e-07,
"loss": 0.3223,
"step": 2060
},
{
"epoch": 2.4716,
"grad_norm": 0.39589800407491577,
"learning_rate": 9.215467785462873e-07,
"loss": 0.3084,
"step": 2061
},
{
"epoch": 2.4728,
"grad_norm": 0.4707799591645335,
"learning_rate": 9.175139354753382e-07,
"loss": 0.3077,
"step": 2062
},
{
"epoch": 2.474,
"grad_norm": 0.5658672639000812,
"learning_rate": 9.134890443499068e-07,
"loss": 0.3339,
"step": 2063
},
{
"epoch": 2.4752,
"grad_norm": 0.5072807402522693,
"learning_rate": 9.094721130097517e-07,
"loss": 0.3212,
"step": 2064
},
{
"epoch": 2.4764,
"grad_norm": 0.5364063819378891,
"learning_rate": 9.054631492791344e-07,
"loss": 0.3138,
"step": 2065
},
{
"epoch": 2.4776,
"grad_norm": 0.7490696104573091,
"learning_rate": 9.014621609667896e-07,
"loss": 0.343,
"step": 2066
},
{
"epoch": 2.4788,
"grad_norm": 0.5640758366462366,
"learning_rate": 8.974691558659187e-07,
"loss": 0.3108,
"step": 2067
},
{
"epoch": 2.48,
"grad_norm": 0.5379307007062987,
"learning_rate": 8.934841417541767e-07,
"loss": 0.3131,
"step": 2068
},
{
"epoch": 2.4812,
"grad_norm": 0.6205649567389456,
"learning_rate": 8.8950712639365e-07,
"loss": 0.3468,
"step": 2069
},
{
"epoch": 2.4824,
"grad_norm": 0.4531821748984028,
"learning_rate": 8.855381175308475e-07,
"loss": 0.326,
"step": 2070
},
{
"epoch": 2.4836,
"grad_norm": 0.5276488030607475,
"learning_rate": 8.815771228966796e-07,
"loss": 0.3581,
"step": 2071
},
{
"epoch": 2.4848,
"grad_norm": 0.5484540605204118,
"learning_rate": 8.776241502064508e-07,
"loss": 0.3439,
"step": 2072
},
{
"epoch": 2.4859999999999998,
"grad_norm": 0.5539951976069447,
"learning_rate": 8.736792071598355e-07,
"loss": 0.3129,
"step": 2073
},
{
"epoch": 2.4872,
"grad_norm": 0.43894121359141824,
"learning_rate": 8.697423014408718e-07,
"loss": 0.3137,
"step": 2074
},
{
"epoch": 2.4884,
"grad_norm": 0.500242458968423,
"learning_rate": 8.658134407179419e-07,
"loss": 0.2873,
"step": 2075
},
{
"epoch": 2.4896,
"grad_norm": 0.430036684418572,
"learning_rate": 8.61892632643756e-07,
"loss": 0.3149,
"step": 2076
},
{
"epoch": 2.4908,
"grad_norm": 0.5133263436266973,
"learning_rate": 8.579798848553389e-07,
"loss": 0.3572,
"step": 2077
},
{
"epoch": 2.492,
"grad_norm": 0.48514597886695165,
"learning_rate": 8.540752049740181e-07,
"loss": 0.3061,
"step": 2078
},
{
"epoch": 2.4932,
"grad_norm": 0.56251703182634,
"learning_rate": 8.501786006054047e-07,
"loss": 0.3312,
"step": 2079
},
{
"epoch": 2.4944,
"grad_norm": 0.5474216931160197,
"learning_rate": 8.462900793393775e-07,
"loss": 0.3163,
"step": 2080
},
{
"epoch": 2.4956,
"grad_norm": 0.43448486787857776,
"learning_rate": 8.424096487500777e-07,
"loss": 0.3229,
"step": 2081
},
{
"epoch": 2.4968,
"grad_norm": 0.47897062850634076,
"learning_rate": 8.385373163958821e-07,
"loss": 0.3588,
"step": 2082
},
{
"epoch": 2.498,
"grad_norm": 0.47095621511292696,
"learning_rate": 8.346730898193928e-07,
"loss": 0.3344,
"step": 2083
},
{
"epoch": 2.4992,
"grad_norm": 0.43858239978498487,
"learning_rate": 8.308169765474278e-07,
"loss": 0.343,
"step": 2084
},
{
"epoch": 2.5004,
"grad_norm": 0.5939243267036994,
"learning_rate": 8.269689840909967e-07,
"loss": 0.3472,
"step": 2085
},
{
"epoch": 2.5016,
"grad_norm": 0.5357699409881242,
"learning_rate": 8.231291199452956e-07,
"loss": 0.3001,
"step": 2086
},
{
"epoch": 2.5028,
"grad_norm": 0.5250204364175717,
"learning_rate": 8.192973915896868e-07,
"loss": 0.3625,
"step": 2087
},
{
"epoch": 2.504,
"grad_norm": 0.4470480707133538,
"learning_rate": 8.154738064876843e-07,
"loss": 0.3383,
"step": 2088
},
{
"epoch": 2.5052,
"grad_norm": 0.5450729961924072,
"learning_rate": 8.116583720869398e-07,
"loss": 0.3155,
"step": 2089
},
{
"epoch": 2.5064,
"grad_norm": 0.47490875408006045,
"learning_rate": 8.078510958192337e-07,
"loss": 0.2995,
"step": 2090
},
{
"epoch": 2.5076,
"grad_norm": 0.5372639556416108,
"learning_rate": 8.040519851004492e-07,
"loss": 0.3129,
"step": 2091
},
{
"epoch": 2.5088,
"grad_norm": 0.46359050540817753,
"learning_rate": 8.002610473305688e-07,
"loss": 0.3292,
"step": 2092
},
{
"epoch": 2.51,
"grad_norm": 0.4796323163330126,
"learning_rate": 7.964782898936569e-07,
"loss": 0.3354,
"step": 2093
},
{
"epoch": 2.5112,
"grad_norm": 0.5208612693160106,
"learning_rate": 7.927037201578397e-07,
"loss": 0.3309,
"step": 2094
},
{
"epoch": 2.5124,
"grad_norm": 0.47838743726112654,
"learning_rate": 7.889373454752964e-07,
"loss": 0.3281,
"step": 2095
},
{
"epoch": 2.5136,
"grad_norm": 0.5320773076821897,
"learning_rate": 7.851791731822461e-07,
"loss": 0.3542,
"step": 2096
},
{
"epoch": 2.5148,
"grad_norm": 0.49702028419375943,
"learning_rate": 7.814292105989308e-07,
"loss": 0.3168,
"step": 2097
},
{
"epoch": 2.516,
"grad_norm": 0.5788902384801881,
"learning_rate": 7.776874650295984e-07,
"loss": 0.3037,
"step": 2098
},
{
"epoch": 2.5172,
"grad_norm": 0.43347682146682476,
"learning_rate": 7.739539437624933e-07,
"loss": 0.3132,
"step": 2099
},
{
"epoch": 2.5183999999999997,
"grad_norm": 0.5138448794184438,
"learning_rate": 7.702286540698417e-07,
"loss": 0.3049,
"step": 2100
},
{
"epoch": 2.5196,
"grad_norm": 0.47716781325402874,
"learning_rate": 7.665116032078346e-07,
"loss": 0.2931,
"step": 2101
},
{
"epoch": 2.5208,
"grad_norm": 0.3987044585261873,
"learning_rate": 7.628027984166153e-07,
"loss": 0.344,
"step": 2102
},
{
"epoch": 2.5220000000000002,
"grad_norm": 0.6009158876364648,
"learning_rate": 7.591022469202675e-07,
"loss": 0.3317,
"step": 2103
},
{
"epoch": 2.5232,
"grad_norm": 0.5084696612857849,
"learning_rate": 7.554099559267964e-07,
"loss": 0.298,
"step": 2104
},
{
"epoch": 2.5244,
"grad_norm": 0.5239081316589891,
"learning_rate": 7.517259326281157e-07,
"loss": 0.3244,
"step": 2105
},
{
"epoch": 2.5256,
"grad_norm": 0.4583710076541313,
"learning_rate": 7.480501842000404e-07,
"loss": 0.2939,
"step": 2106
},
{
"epoch": 2.5268,
"grad_norm": 0.4979847876868833,
"learning_rate": 7.443827178022628e-07,
"loss": 0.3132,
"step": 2107
},
{
"epoch": 2.528,
"grad_norm": 0.5304581631546377,
"learning_rate": 7.407235405783453e-07,
"loss": 0.287,
"step": 2108
},
{
"epoch": 2.5292,
"grad_norm": 0.5560997292562039,
"learning_rate": 7.370726596557059e-07,
"loss": 0.3329,
"step": 2109
},
{
"epoch": 2.5304,
"grad_norm": 0.5419001893194091,
"learning_rate": 7.334300821455998e-07,
"loss": 0.3251,
"step": 2110
},
{
"epoch": 2.5316,
"grad_norm": 0.4762163668078043,
"learning_rate": 7.297958151431094e-07,
"loss": 0.313,
"step": 2111
},
{
"epoch": 2.5328,
"grad_norm": 0.5471030007881764,
"learning_rate": 7.26169865727131e-07,
"loss": 0.3096,
"step": 2112
},
{
"epoch": 2.534,
"grad_norm": 0.4859920497046962,
"learning_rate": 7.225522409603608e-07,
"loss": 0.307,
"step": 2113
},
{
"epoch": 2.5352,
"grad_norm": 0.46802572251033253,
"learning_rate": 7.189429478892762e-07,
"loss": 0.3221,
"step": 2114
},
{
"epoch": 2.5364,
"grad_norm": 0.5019363613852679,
"learning_rate": 7.153419935441303e-07,
"loss": 0.3353,
"step": 2115
},
{
"epoch": 2.5376,
"grad_norm": 0.5306887115275352,
"learning_rate": 7.117493849389306e-07,
"loss": 0.3121,
"step": 2116
},
{
"epoch": 2.5388,
"grad_norm": 0.4171241804844174,
"learning_rate": 7.081651290714287e-07,
"loss": 0.3156,
"step": 2117
},
{
"epoch": 2.54,
"grad_norm": 0.5238443225363039,
"learning_rate": 7.045892329231086e-07,
"loss": 0.3228,
"step": 2118
},
{
"epoch": 2.5412,
"grad_norm": 0.5814868599249048,
"learning_rate": 7.010217034591721e-07,
"loss": 0.3493,
"step": 2119
},
{
"epoch": 2.5423999999999998,
"grad_norm": 0.5437026734195148,
"learning_rate": 6.974625476285191e-07,
"loss": 0.3385,
"step": 2120
},
{
"epoch": 2.5436,
"grad_norm": 0.6691095297956318,
"learning_rate": 6.93911772363745e-07,
"loss": 0.3557,
"step": 2121
},
{
"epoch": 2.5448,
"grad_norm": 0.5019767886763458,
"learning_rate": 6.903693845811176e-07,
"loss": 0.3436,
"step": 2122
},
{
"epoch": 2.5460000000000003,
"grad_norm": 0.48356310492003735,
"learning_rate": 6.86835391180567e-07,
"loss": 0.3358,
"step": 2123
},
{
"epoch": 2.5472,
"grad_norm": 0.5094244777532211,
"learning_rate": 6.833097990456761e-07,
"loss": 0.3039,
"step": 2124
},
{
"epoch": 2.5484,
"grad_norm": 0.459217234761272,
"learning_rate": 6.797926150436618e-07,
"loss": 0.2956,
"step": 2125
},
{
"epoch": 2.5496,
"grad_norm": 0.4472215759109841,
"learning_rate": 6.762838460253629e-07,
"loss": 0.2976,
"step": 2126
},
{
"epoch": 2.5507999999999997,
"grad_norm": 0.5277980736149517,
"learning_rate": 6.727834988252258e-07,
"loss": 0.326,
"step": 2127
},
{
"epoch": 2.552,
"grad_norm": 0.5853107231529425,
"learning_rate": 6.692915802612965e-07,
"loss": 0.278,
"step": 2128
},
{
"epoch": 2.5532,
"grad_norm": 0.43077774106221967,
"learning_rate": 6.658080971352026e-07,
"loss": 0.3112,
"step": 2129
},
{
"epoch": 2.5544000000000002,
"grad_norm": 0.5099625444460486,
"learning_rate": 6.623330562321378e-07,
"loss": 0.3486,
"step": 2130
},
{
"epoch": 2.5556,
"grad_norm": 0.5749327138836842,
"learning_rate": 6.588664643208559e-07,
"loss": 0.3064,
"step": 2131
},
{
"epoch": 2.5568,
"grad_norm": 0.47739858914285005,
"learning_rate": 6.554083281536516e-07,
"loss": 0.3224,
"step": 2132
},
{
"epoch": 2.558,
"grad_norm": 0.5140359142056293,
"learning_rate": 6.519586544663481e-07,
"loss": 0.2938,
"step": 2133
},
{
"epoch": 2.5592,
"grad_norm": 0.456904859506897,
"learning_rate": 6.485174499782876e-07,
"loss": 0.3425,
"step": 2134
},
{
"epoch": 2.5604,
"grad_norm": 0.5122662847914252,
"learning_rate": 6.450847213923162e-07,
"loss": 0.3162,
"step": 2135
},
{
"epoch": 2.5616,
"grad_norm": 0.5141807606724946,
"learning_rate": 6.416604753947675e-07,
"loss": 0.3396,
"step": 2136
},
{
"epoch": 2.5628,
"grad_norm": 0.5342014748603048,
"learning_rate": 6.382447186554553e-07,
"loss": 0.3093,
"step": 2137
},
{
"epoch": 2.564,
"grad_norm": 0.4997597150358832,
"learning_rate": 6.348374578276567e-07,
"loss": 0.3159,
"step": 2138
},
{
"epoch": 2.5652,
"grad_norm": 0.5884891445466817,
"learning_rate": 6.314386995480987e-07,
"loss": 0.3148,
"step": 2139
},
{
"epoch": 2.5664,
"grad_norm": 0.5441867009250919,
"learning_rate": 6.280484504369505e-07,
"loss": 0.3186,
"step": 2140
},
{
"epoch": 2.5676,
"grad_norm": 0.4516493539423153,
"learning_rate": 6.246667170978049e-07,
"loss": 0.3612,
"step": 2141
},
{
"epoch": 2.5688,
"grad_norm": 0.5084051121542065,
"learning_rate": 6.212935061176667e-07,
"loss": 0.2953,
"step": 2142
},
{
"epoch": 2.57,
"grad_norm": 0.5877296146705292,
"learning_rate": 6.179288240669429e-07,
"loss": 0.3559,
"step": 2143
},
{
"epoch": 2.5712,
"grad_norm": 0.5558192685394937,
"learning_rate": 6.14572677499426e-07,
"loss": 0.3191,
"step": 2144
},
{
"epoch": 2.5724,
"grad_norm": 0.5012538516660345,
"learning_rate": 6.112250729522823e-07,
"loss": 0.2866,
"step": 2145
},
{
"epoch": 2.5736,
"grad_norm": 0.5217614598335951,
"learning_rate": 6.078860169460416e-07,
"loss": 0.2985,
"step": 2146
},
{
"epoch": 2.5747999999999998,
"grad_norm": 0.4496361197136127,
"learning_rate": 6.045555159845828e-07,
"loss": 0.3082,
"step": 2147
},
{
"epoch": 2.576,
"grad_norm": 0.5288177340496564,
"learning_rate": 6.012335765551186e-07,
"loss": 0.3197,
"step": 2148
},
{
"epoch": 2.5772,
"grad_norm": 0.5440414858731661,
"learning_rate": 5.979202051281891e-07,
"loss": 0.351,
"step": 2149
},
{
"epoch": 2.5784000000000002,
"grad_norm": 0.5361965970453445,
"learning_rate": 5.946154081576411e-07,
"loss": 0.3013,
"step": 2150
},
{
"epoch": 2.5796,
"grad_norm": 0.4975790271997051,
"learning_rate": 5.913191920806244e-07,
"loss": 0.3108,
"step": 2151
},
{
"epoch": 2.5808,
"grad_norm": 0.6761947196756368,
"learning_rate": 5.880315633175704e-07,
"loss": 0.3161,
"step": 2152
},
{
"epoch": 2.582,
"grad_norm": 0.5129758578390546,
"learning_rate": 5.847525282721883e-07,
"loss": 0.3235,
"step": 2153
},
{
"epoch": 2.5832,
"grad_norm": 0.5057608403560111,
"learning_rate": 5.814820933314446e-07,
"loss": 0.3091,
"step": 2154
},
{
"epoch": 2.5844,
"grad_norm": 0.5027515285537514,
"learning_rate": 5.78220264865555e-07,
"loss": 0.3229,
"step": 2155
},
{
"epoch": 2.5856,
"grad_norm": 0.4960339355194685,
"learning_rate": 5.749670492279757e-07,
"loss": 0.3336,
"step": 2156
},
{
"epoch": 2.5868,
"grad_norm": 0.5695628614608653,
"learning_rate": 5.717224527553811e-07,
"loss": 0.3282,
"step": 2157
},
{
"epoch": 2.588,
"grad_norm": 0.45658740631386857,
"learning_rate": 5.684864817676583e-07,
"loss": 0.3212,
"step": 2158
},
{
"epoch": 2.5892,
"grad_norm": 0.4714820843044785,
"learning_rate": 5.65259142567896e-07,
"loss": 0.3287,
"step": 2159
},
{
"epoch": 2.5904,
"grad_norm": 0.4845192136259272,
"learning_rate": 5.620404414423674e-07,
"loss": 0.3383,
"step": 2160
},
{
"epoch": 2.5916,
"grad_norm": 0.5099050915537449,
"learning_rate": 5.588303846605187e-07,
"loss": 0.3468,
"step": 2161
},
{
"epoch": 2.5928,
"grad_norm": 0.5348537812273993,
"learning_rate": 5.556289784749653e-07,
"loss": 0.2881,
"step": 2162
},
{
"epoch": 2.594,
"grad_norm": 0.48728381677988253,
"learning_rate": 5.524362291214652e-07,
"loss": 0.2991,
"step": 2163
},
{
"epoch": 2.5952,
"grad_norm": 0.46169914830110426,
"learning_rate": 5.492521428189179e-07,
"loss": 0.3198,
"step": 2164
},
{
"epoch": 2.5964,
"grad_norm": 0.4922362789271303,
"learning_rate": 5.460767257693489e-07,
"loss": 0.2959,
"step": 2165
},
{
"epoch": 2.5976,
"grad_norm": 0.6088704547745182,
"learning_rate": 5.429099841578966e-07,
"loss": 0.3211,
"step": 2166
},
{
"epoch": 2.5987999999999998,
"grad_norm": 0.5110850313521915,
"learning_rate": 5.397519241528026e-07,
"loss": 0.3377,
"step": 2167
},
{
"epoch": 2.6,
"grad_norm": 0.5724649889513147,
"learning_rate": 5.366025519053958e-07,
"loss": 0.3144,
"step": 2168
},
{
"epoch": 2.6012,
"grad_norm": 0.4709112775705695,
"learning_rate": 5.334618735500868e-07,
"loss": 0.3076,
"step": 2169
},
{
"epoch": 2.6024000000000003,
"grad_norm": 0.5228928468545195,
"learning_rate": 5.303298952043473e-07,
"loss": 0.3291,
"step": 2170
},
{
"epoch": 2.6036,
"grad_norm": 0.5740694244625357,
"learning_rate": 5.272066229687078e-07,
"loss": 0.3356,
"step": 2171
},
{
"epoch": 2.6048,
"grad_norm": 0.4470914388889558,
"learning_rate": 5.24092062926736e-07,
"loss": 0.3396,
"step": 2172
},
{
"epoch": 2.606,
"grad_norm": 0.5051172194547626,
"learning_rate": 5.209862211450351e-07,
"loss": 0.3434,
"step": 2173
},
{
"epoch": 2.6071999999999997,
"grad_norm": 0.49743096722575747,
"learning_rate": 5.17889103673222e-07,
"loss": 0.314,
"step": 2174
},
{
"epoch": 2.6084,
"grad_norm": 0.42926620449426334,
"learning_rate": 5.148007165439234e-07,
"loss": 0.3041,
"step": 2175
},
{
"epoch": 2.6096,
"grad_norm": 0.5297863445354456,
"learning_rate": 5.117210657727589e-07,
"loss": 0.3153,
"step": 2176
},
{
"epoch": 2.6108000000000002,
"grad_norm": 0.5751473879165507,
"learning_rate": 5.086501573583302e-07,
"loss": 0.3164,
"step": 2177
},
{
"epoch": 2.612,
"grad_norm": 0.49512626673277416,
"learning_rate": 5.055879972822164e-07,
"loss": 0.309,
"step": 2178
},
{
"epoch": 2.6132,
"grad_norm": 0.45728478789306254,
"learning_rate": 5.025345915089497e-07,
"loss": 0.3019,
"step": 2179
},
{
"epoch": 2.6144,
"grad_norm": 0.6727936896653166,
"learning_rate": 4.994899459860125e-07,
"loss": 0.3238,
"step": 2180
},
{
"epoch": 2.6156,
"grad_norm": 0.5716215767054986,
"learning_rate": 4.964540666438261e-07,
"loss": 0.2843,
"step": 2181
},
{
"epoch": 2.6168,
"grad_norm": 0.5098447409370542,
"learning_rate": 4.934269593957336e-07,
"loss": 0.3484,
"step": 2182
},
{
"epoch": 2.618,
"grad_norm": 0.5093018487176464,
"learning_rate": 4.90408630137994e-07,
"loss": 0.3431,
"step": 2183
},
{
"epoch": 2.6192,
"grad_norm": 0.4840635587660347,
"learning_rate": 4.873990847497684e-07,
"loss": 0.3126,
"step": 2184
},
{
"epoch": 2.6204,
"grad_norm": 0.5082242061141022,
"learning_rate": 4.843983290931064e-07,
"loss": 0.2969,
"step": 2185
},
{
"epoch": 2.6216,
"grad_norm": 0.530976128386319,
"learning_rate": 4.814063690129378e-07,
"loss": 0.343,
"step": 2186
},
{
"epoch": 2.6228,
"grad_norm": 0.6016390612143504,
"learning_rate": 4.784232103370617e-07,
"loss": 0.3294,
"step": 2187
},
{
"epoch": 2.624,
"grad_norm": 0.556223638359762,
"learning_rate": 4.7544885887613136e-07,
"loss": 0.3379,
"step": 2188
},
{
"epoch": 2.6252,
"grad_norm": 0.54268125037195,
"learning_rate": 4.724833204236462e-07,
"loss": 0.3331,
"step": 2189
},
{
"epoch": 2.6264,
"grad_norm": 0.48122211227489187,
"learning_rate": 4.695266007559407e-07,
"loss": 0.3071,
"step": 2190
},
{
"epoch": 2.6276,
"grad_norm": 0.45238477752729517,
"learning_rate": 4.6657870563217076e-07,
"loss": 0.3218,
"step": 2191
},
{
"epoch": 2.6288,
"grad_norm": 0.5379409443968398,
"learning_rate": 4.6363964079430166e-07,
"loss": 0.3314,
"step": 2192
},
{
"epoch": 2.63,
"grad_norm": 0.49881724473585415,
"learning_rate": 4.6070941196710186e-07,
"loss": 0.318,
"step": 2193
},
{
"epoch": 2.6311999999999998,
"grad_norm": 0.5073400286081319,
"learning_rate": 4.5778802485812956e-07,
"loss": 0.3141,
"step": 2194
},
{
"epoch": 2.6324,
"grad_norm": 0.4494142417800747,
"learning_rate": 4.548754851577175e-07,
"loss": 0.2828,
"step": 2195
},
{
"epoch": 2.6336,
"grad_norm": 0.4241794577035244,
"learning_rate": 4.5197179853896654e-07,
"loss": 0.2998,
"step": 2196
},
{
"epoch": 2.6348000000000003,
"grad_norm": 0.5113141807236002,
"learning_rate": 4.4907697065773523e-07,
"loss": 0.3286,
"step": 2197
},
{
"epoch": 2.636,
"grad_norm": 0.5055727152160692,
"learning_rate": 4.4619100715262374e-07,
"loss": 0.3036,
"step": 2198
},
{
"epoch": 2.6372,
"grad_norm": 0.44911584465721455,
"learning_rate": 4.4331391364496934e-07,
"loss": 0.3037,
"step": 2199
},
{
"epoch": 2.6384,
"grad_norm": 0.45094950311107046,
"learning_rate": 4.404456957388309e-07,
"loss": 0.3316,
"step": 2200
},
{
"epoch": 2.6395999999999997,
"grad_norm": 0.41455331969132403,
"learning_rate": 4.375863590209778e-07,
"loss": 0.323,
"step": 2201
},
{
"epoch": 2.6408,
"grad_norm": 0.43075888630028597,
"learning_rate": 4.3473590906088046e-07,
"loss": 0.3196,
"step": 2202
},
{
"epoch": 2.642,
"grad_norm": 0.5686147750139869,
"learning_rate": 4.3189435141070324e-07,
"loss": 0.3349,
"step": 2203
},
{
"epoch": 2.6432,
"grad_norm": 0.5689753507315711,
"learning_rate": 4.2906169160528424e-07,
"loss": 0.3011,
"step": 2204
},
{
"epoch": 2.6444,
"grad_norm": 0.5472134263806879,
"learning_rate": 4.262379351621354e-07,
"loss": 0.2809,
"step": 2205
},
{
"epoch": 2.6456,
"grad_norm": 0.5637048283007405,
"learning_rate": 4.2342308758142437e-07,
"loss": 0.2791,
"step": 2206
},
{
"epoch": 2.6468,
"grad_norm": 0.5038051098339148,
"learning_rate": 4.2061715434596475e-07,
"loss": 0.3214,
"step": 2207
},
{
"epoch": 2.648,
"grad_norm": 0.5582257157785934,
"learning_rate": 4.1782014092120735e-07,
"loss": 0.3465,
"step": 2208
},
{
"epoch": 2.6492,
"grad_norm": 0.4816975269500765,
"learning_rate": 4.150320527552304e-07,
"loss": 0.3351,
"step": 2209
},
{
"epoch": 2.6504,
"grad_norm": 0.6144407863119755,
"learning_rate": 4.122528952787258e-07,
"loss": 0.303,
"step": 2210
},
{
"epoch": 2.6516,
"grad_norm": 0.4690793878545897,
"learning_rate": 4.0948267390498953e-07,
"loss": 0.295,
"step": 2211
},
{
"epoch": 2.6528,
"grad_norm": 0.45771893866529245,
"learning_rate": 4.067213940299136e-07,
"loss": 0.2892,
"step": 2212
},
{
"epoch": 2.654,
"grad_norm": 0.5107864430619176,
"learning_rate": 4.0396906103197244e-07,
"loss": 0.3355,
"step": 2213
},
{
"epoch": 2.6552,
"grad_norm": 0.48459788169311224,
"learning_rate": 4.01225680272212e-07,
"loss": 0.3564,
"step": 2214
},
{
"epoch": 2.6564,
"grad_norm": 0.5223513258810136,
"learning_rate": 3.984912570942434e-07,
"loss": 0.3575,
"step": 2215
},
{
"epoch": 2.6576,
"grad_norm": 0.4748513216901385,
"learning_rate": 3.9576579682423066e-07,
"loss": 0.31,
"step": 2216
},
{
"epoch": 2.6588000000000003,
"grad_norm": 0.49092324085793265,
"learning_rate": 3.930493047708761e-07,
"loss": 0.3273,
"step": 2217
},
{
"epoch": 2.66,
"grad_norm": 0.5086589663869817,
"learning_rate": 3.903417862254172e-07,
"loss": 0.3037,
"step": 2218
},
{
"epoch": 2.6612,
"grad_norm": 0.5944392528245183,
"learning_rate": 3.876432464616103e-07,
"loss": 0.3102,
"step": 2219
},
{
"epoch": 2.6624,
"grad_norm": 0.43997132818776513,
"learning_rate": 3.8495369073572266e-07,
"loss": 0.2987,
"step": 2220
},
{
"epoch": 2.6635999999999997,
"grad_norm": 0.4479899231191285,
"learning_rate": 3.822731242865235e-07,
"loss": 0.307,
"step": 2221
},
{
"epoch": 2.6648,
"grad_norm": 0.49879617773556645,
"learning_rate": 3.7960155233527364e-07,
"loss": 0.2972,
"step": 2222
},
{
"epoch": 2.666,
"grad_norm": 0.4096782707041188,
"learning_rate": 3.7693898008571205e-07,
"loss": 0.3173,
"step": 2223
},
{
"epoch": 2.6672000000000002,
"grad_norm": 0.5370525570342256,
"learning_rate": 3.742854127240464e-07,
"loss": 0.3189,
"step": 2224
},
{
"epoch": 2.6684,
"grad_norm": 0.6088932928091623,
"learning_rate": 3.7164085541894937e-07,
"loss": 0.3435,
"step": 2225
},
{
"epoch": 2.6696,
"grad_norm": 0.5113014285512854,
"learning_rate": 3.690053133215399e-07,
"loss": 0.2883,
"step": 2226
},
{
"epoch": 2.6708,
"grad_norm": 0.5238135484660226,
"learning_rate": 3.663787915653777e-07,
"loss": 0.3635,
"step": 2227
},
{
"epoch": 2.672,
"grad_norm": 0.4478739081822928,
"learning_rate": 3.6376129526645376e-07,
"loss": 0.347,
"step": 2228
},
{
"epoch": 2.6732,
"grad_norm": 0.443318474124737,
"learning_rate": 3.6115282952317807e-07,
"loss": 0.3447,
"step": 2229
},
{
"epoch": 2.6744,
"grad_norm": 0.5298718770182731,
"learning_rate": 3.5855339941636867e-07,
"loss": 0.2945,
"step": 2230
},
{
"epoch": 2.6756,
"grad_norm": 0.5045580008727675,
"learning_rate": 3.5596301000924815e-07,
"loss": 0.3105,
"step": 2231
},
{
"epoch": 2.6768,
"grad_norm": 0.5605568435967557,
"learning_rate": 3.533816663474271e-07,
"loss": 0.297,
"step": 2232
},
{
"epoch": 2.678,
"grad_norm": 0.5631677023915352,
"learning_rate": 3.508093734588952e-07,
"loss": 0.3222,
"step": 2233
},
{
"epoch": 2.6792,
"grad_norm": 0.49016251569886216,
"learning_rate": 3.482461363540163e-07,
"loss": 0.2997,
"step": 2234
},
{
"epoch": 2.6804,
"grad_norm": 0.570641230845762,
"learning_rate": 3.456919600255126e-07,
"loss": 0.3159,
"step": 2235
},
{
"epoch": 2.6816,
"grad_norm": 0.5818993208808508,
"learning_rate": 3.4314684944845747e-07,
"loss": 0.2889,
"step": 2236
},
{
"epoch": 2.6828,
"grad_norm": 0.4895819240155071,
"learning_rate": 3.406108095802668e-07,
"loss": 0.3207,
"step": 2237
},
{
"epoch": 2.684,
"grad_norm": 0.4467165658918849,
"learning_rate": 3.3808384536068997e-07,
"loss": 0.3007,
"step": 2238
},
{
"epoch": 2.6852,
"grad_norm": 0.44346210912182843,
"learning_rate": 3.3556596171179455e-07,
"loss": 0.3345,
"step": 2239
},
{
"epoch": 2.6864,
"grad_norm": 0.4527063516873698,
"learning_rate": 3.3305716353796537e-07,
"loss": 0.3217,
"step": 2240
},
{
"epoch": 2.6875999999999998,
"grad_norm": 0.591329001450134,
"learning_rate": 3.305574557258867e-07,
"loss": 0.308,
"step": 2241
},
{
"epoch": 2.6888,
"grad_norm": 0.5253134786452384,
"learning_rate": 3.2806684314453774e-07,
"loss": 0.2952,
"step": 2242
},
{
"epoch": 2.69,
"grad_norm": 0.5211911228534106,
"learning_rate": 3.255853306451823e-07,
"loss": 0.3346,
"step": 2243
},
{
"epoch": 2.6912000000000003,
"grad_norm": 0.4965699355946691,
"learning_rate": 3.2311292306135944e-07,
"loss": 0.2861,
"step": 2244
},
{
"epoch": 2.6924,
"grad_norm": 0.4527295772355629,
"learning_rate": 3.2064962520887146e-07,
"loss": 0.3208,
"step": 2245
},
{
"epoch": 2.6936,
"grad_norm": 0.7191617099605793,
"learning_rate": 3.18195441885778e-07,
"loss": 0.3842,
"step": 2246
},
{
"epoch": 2.6948,
"grad_norm": 0.4660456167342279,
"learning_rate": 3.157503778723847e-07,
"loss": 0.3154,
"step": 2247
},
{
"epoch": 2.6959999999999997,
"grad_norm": 0.4581700106444103,
"learning_rate": 3.1331443793123585e-07,
"loss": 0.316,
"step": 2248
},
{
"epoch": 2.6972,
"grad_norm": 0.4910588872563203,
"learning_rate": 3.108876268071009e-07,
"loss": 0.3142,
"step": 2249
},
{
"epoch": 2.6984,
"grad_norm": 0.4636219243535772,
"learning_rate": 3.0846994922697104e-07,
"loss": 0.347,
"step": 2250
},
{
"epoch": 2.6996,
"grad_norm": 0.5034421859299472,
"learning_rate": 3.060614099000442e-07,
"loss": 0.3028,
"step": 2251
},
{
"epoch": 2.7008,
"grad_norm": 0.4912027303121192,
"learning_rate": 3.0366201351771983e-07,
"loss": 0.3035,
"step": 2252
},
{
"epoch": 2.702,
"grad_norm": 0.4533175883327802,
"learning_rate": 3.0127176475359065e-07,
"loss": 0.3072,
"step": 2253
},
{
"epoch": 2.7032,
"grad_norm": 0.5440528983007439,
"learning_rate": 2.988906682634285e-07,
"loss": 0.3108,
"step": 2254
},
{
"epoch": 2.7044,
"grad_norm": 0.4642521095472613,
"learning_rate": 2.965187286851784e-07,
"loss": 0.3107,
"step": 2255
},
{
"epoch": 2.7056,
"grad_norm": 0.5045589770404426,
"learning_rate": 2.941559506389513e-07,
"loss": 0.322,
"step": 2256
},
{
"epoch": 2.7068,
"grad_norm": 0.45835153688747926,
"learning_rate": 2.9180233872701247e-07,
"loss": 0.2856,
"step": 2257
},
{
"epoch": 2.708,
"grad_norm": 0.7022876521237986,
"learning_rate": 2.894578975337703e-07,
"loss": 0.3423,
"step": 2258
},
{
"epoch": 2.7092,
"grad_norm": 0.48858267358117635,
"learning_rate": 2.8712263162577636e-07,
"loss": 0.3113,
"step": 2259
},
{
"epoch": 2.7104,
"grad_norm": 0.5026542424606422,
"learning_rate": 2.8479654555170546e-07,
"loss": 0.3298,
"step": 2260
},
{
"epoch": 2.7116,
"grad_norm": 0.4570485297158199,
"learning_rate": 2.8247964384235214e-07,
"loss": 0.2936,
"step": 2261
},
{
"epoch": 2.7128,
"grad_norm": 0.5256313228851077,
"learning_rate": 2.8017193101062377e-07,
"loss": 0.2875,
"step": 2262
},
{
"epoch": 2.714,
"grad_norm": 0.5508455978649375,
"learning_rate": 2.778734115515269e-07,
"loss": 0.3706,
"step": 2263
},
{
"epoch": 2.7152,
"grad_norm": 0.6383900095161533,
"learning_rate": 2.755840899421636e-07,
"loss": 0.32,
"step": 2264
},
{
"epoch": 2.7164,
"grad_norm": 0.5310429090489197,
"learning_rate": 2.7330397064171787e-07,
"loss": 0.3448,
"step": 2265
},
{
"epoch": 2.7176,
"grad_norm": 0.5272757528343431,
"learning_rate": 2.7103305809145106e-07,
"loss": 0.321,
"step": 2266
},
{
"epoch": 2.7188,
"grad_norm": 0.48216498913473155,
"learning_rate": 2.687713567146899e-07,
"loss": 0.3286,
"step": 2267
},
{
"epoch": 2.7199999999999998,
"grad_norm": 0.5025142667561859,
"learning_rate": 2.665188709168215e-07,
"loss": 0.303,
"step": 2268
},
{
"epoch": 2.7212,
"grad_norm": 0.47857628655183854,
"learning_rate": 2.642756050852796e-07,
"loss": 0.2973,
"step": 2269
},
{
"epoch": 2.7224,
"grad_norm": 0.6016413542553345,
"learning_rate": 2.620415635895429e-07,
"loss": 0.3043,
"step": 2270
},
{
"epoch": 2.7236000000000002,
"grad_norm": 0.5393265361428493,
"learning_rate": 2.5981675078111835e-07,
"loss": 0.2915,
"step": 2271
},
{
"epoch": 2.7248,
"grad_norm": 0.41790215327308255,
"learning_rate": 2.5760117099354163e-07,
"loss": 0.3091,
"step": 2272
},
{
"epoch": 2.726,
"grad_norm": 0.4533408612597167,
"learning_rate": 2.5539482854236076e-07,
"loss": 0.3261,
"step": 2273
},
{
"epoch": 2.7272,
"grad_norm": 0.5594124000073417,
"learning_rate": 2.531977277251324e-07,
"loss": 0.3247,
"step": 2274
},
{
"epoch": 2.7284,
"grad_norm": 0.4721945408146146,
"learning_rate": 2.510098728214133e-07,
"loss": 0.3185,
"step": 2275
},
{
"epoch": 2.7296,
"grad_norm": 0.4774126492232413,
"learning_rate": 2.488312680927485e-07,
"loss": 0.3366,
"step": 2276
},
{
"epoch": 2.7308,
"grad_norm": 0.43569948109138307,
"learning_rate": 2.466619177826668e-07,
"loss": 0.3292,
"step": 2277
},
{
"epoch": 2.732,
"grad_norm": 0.4611430979498518,
"learning_rate": 2.4450182611667096e-07,
"loss": 0.2826,
"step": 2278
},
{
"epoch": 2.7332,
"grad_norm": 0.5292796289010596,
"learning_rate": 2.423509973022292e-07,
"loss": 0.3274,
"step": 2279
},
{
"epoch": 2.7344,
"grad_norm": 0.4888164532692469,
"learning_rate": 2.4020943552876706e-07,
"loss": 0.3298,
"step": 2280
},
{
"epoch": 2.7356,
"grad_norm": 0.5713745001080968,
"learning_rate": 2.3807714496766165e-07,
"loss": 0.3049,
"step": 2281
},
{
"epoch": 2.7368,
"grad_norm": 0.478309599150092,
"learning_rate": 2.3595412977222897e-07,
"loss": 0.3028,
"step": 2282
},
{
"epoch": 2.738,
"grad_norm": 0.5217959378930856,
"learning_rate": 2.3384039407771896e-07,
"loss": 0.2874,
"step": 2283
},
{
"epoch": 2.7392,
"grad_norm": 0.44818891342230366,
"learning_rate": 2.317359420013071e-07,
"loss": 0.2947,
"step": 2284
},
{
"epoch": 2.7404,
"grad_norm": 0.617465212236017,
"learning_rate": 2.2964077764208615e-07,
"loss": 0.3485,
"step": 2285
},
{
"epoch": 2.7416,
"grad_norm": 0.5941416952735622,
"learning_rate": 2.2755490508105716e-07,
"loss": 0.322,
"step": 2286
},
{
"epoch": 2.7428,
"grad_norm": 0.42877695545329736,
"learning_rate": 2.254783283811246e-07,
"loss": 0.3339,
"step": 2287
},
{
"epoch": 2.7439999999999998,
"grad_norm": 0.5975641375977941,
"learning_rate": 2.2341105158708408e-07,
"loss": 0.3219,
"step": 2288
},
{
"epoch": 2.7452,
"grad_norm": 0.5071505659295147,
"learning_rate": 2.2135307872561628e-07,
"loss": 0.3069,
"step": 2289
},
{
"epoch": 2.7464,
"grad_norm": 0.47164314091927906,
"learning_rate": 2.1930441380528243e-07,
"loss": 0.33,
"step": 2290
},
{
"epoch": 2.7476000000000003,
"grad_norm": 0.5271951285578713,
"learning_rate": 2.172650608165111e-07,
"loss": 0.3162,
"step": 2291
},
{
"epoch": 2.7488,
"grad_norm": 0.4448132879912682,
"learning_rate": 2.1523502373159367e-07,
"loss": 0.3224,
"step": 2292
},
{
"epoch": 2.75,
"grad_norm": 0.46226885551679747,
"learning_rate": 2.1321430650467546e-07,
"loss": 0.2991,
"step": 2293
},
{
"epoch": 2.7512,
"grad_norm": 0.5606191502800733,
"learning_rate": 2.112029130717491e-07,
"loss": 0.3334,
"step": 2294
},
{
"epoch": 2.7523999999999997,
"grad_norm": 0.5385580493671908,
"learning_rate": 2.092008473506446e-07,
"loss": 0.3071,
"step": 2295
},
{
"epoch": 2.7536,
"grad_norm": 0.45254035994089326,
"learning_rate": 2.072081132410253e-07,
"loss": 0.3224,
"step": 2296
},
{
"epoch": 2.7548,
"grad_norm": 0.49436523245287795,
"learning_rate": 2.0522471462437798e-07,
"loss": 0.3204,
"step": 2297
},
{
"epoch": 2.7560000000000002,
"grad_norm": 0.5125106319731159,
"learning_rate": 2.0325065536400456e-07,
"loss": 0.3419,
"step": 2298
},
{
"epoch": 2.7572,
"grad_norm": 0.48631037797606286,
"learning_rate": 2.0128593930501427e-07,
"loss": 0.2907,
"step": 2299
},
{
"epoch": 2.7584,
"grad_norm": 0.5155305501751236,
"learning_rate": 1.9933057027432147e-07,
"loss": 0.3314,
"step": 2300
},
{
"epoch": 2.7596,
"grad_norm": 0.5173988364156298,
"learning_rate": 1.9738455208063055e-07,
"loss": 0.3238,
"step": 2301
},
{
"epoch": 2.7608,
"grad_norm": 0.4803287920371691,
"learning_rate": 1.9544788851443342e-07,
"loss": 0.3081,
"step": 2302
},
{
"epoch": 2.762,
"grad_norm": 0.41561006841126175,
"learning_rate": 1.9352058334800195e-07,
"loss": 0.3041,
"step": 2303
},
{
"epoch": 2.7632,
"grad_norm": 0.43105371379486856,
"learning_rate": 1.9160264033537824e-07,
"loss": 0.304,
"step": 2304
},
{
"epoch": 2.7644,
"grad_norm": 0.6000309915607545,
"learning_rate": 1.8969406321236727e-07,
"loss": 0.3206,
"step": 2305
},
{
"epoch": 2.7656,
"grad_norm": 0.5067718716132091,
"learning_rate": 1.8779485569653422e-07,
"loss": 0.3411,
"step": 2306
},
{
"epoch": 2.7668,
"grad_norm": 0.5875424125802807,
"learning_rate": 1.859050214871927e-07,
"loss": 0.3384,
"step": 2307
},
{
"epoch": 2.768,
"grad_norm": 0.4612732891476999,
"learning_rate": 1.8402456426539706e-07,
"loss": 0.2919,
"step": 2308
},
{
"epoch": 2.7692,
"grad_norm": 0.5583069625711948,
"learning_rate": 1.8215348769393904e-07,
"loss": 0.3176,
"step": 2309
},
{
"epoch": 2.7704,
"grad_norm": 0.5530704471556647,
"learning_rate": 1.8029179541733833e-07,
"loss": 0.3048,
"step": 2310
},
{
"epoch": 2.7716,
"grad_norm": 0.5222524590044774,
"learning_rate": 1.7843949106183368e-07,
"loss": 0.34,
"step": 2311
},
{
"epoch": 2.7728,
"grad_norm": 0.44633840212479664,
"learning_rate": 1.7659657823538067e-07,
"loss": 0.3181,
"step": 2312
},
{
"epoch": 2.774,
"grad_norm": 0.4671319697242977,
"learning_rate": 1.747630605276407e-07,
"loss": 0.3485,
"step": 2313
},
{
"epoch": 2.7752,
"grad_norm": 0.4397303650706271,
"learning_rate": 1.7293894150997414e-07,
"loss": 0.3032,
"step": 2314
},
{
"epoch": 2.7763999999999998,
"grad_norm": 0.5338150922235967,
"learning_rate": 1.711242247354372e-07,
"loss": 0.2934,
"step": 2315
},
{
"epoch": 2.7776,
"grad_norm": 0.4409628237565337,
"learning_rate": 1.6931891373876852e-07,
"loss": 0.3235,
"step": 2316
},
{
"epoch": 2.7788,
"grad_norm": 0.5834993749132399,
"learning_rate": 1.6752301203638854e-07,
"loss": 0.3328,
"step": 2317
},
{
"epoch": 2.7800000000000002,
"grad_norm": 0.5345933957862313,
"learning_rate": 1.657365231263891e-07,
"loss": 0.3268,
"step": 2318
},
{
"epoch": 2.7812,
"grad_norm": 0.4421569088458628,
"learning_rate": 1.6395945048852947e-07,
"loss": 0.3335,
"step": 2319
},
{
"epoch": 2.7824,
"grad_norm": 0.5036343755740437,
"learning_rate": 1.6219179758422465e-07,
"loss": 0.2935,
"step": 2320
},
{
"epoch": 2.7836,
"grad_norm": 0.4498029958875161,
"learning_rate": 1.6043356785654273e-07,
"loss": 0.3304,
"step": 2321
},
{
"epoch": 2.7848,
"grad_norm": 0.44670818158992454,
"learning_rate": 1.5868476473019922e-07,
"loss": 0.3237,
"step": 2322
},
{
"epoch": 2.786,
"grad_norm": 0.5601966592047272,
"learning_rate": 1.5694539161154598e-07,
"loss": 0.334,
"step": 2323
},
{
"epoch": 2.7872,
"grad_norm": 0.6091250972218841,
"learning_rate": 1.5521545188856734e-07,
"loss": 0.357,
"step": 2324
},
{
"epoch": 2.7884,
"grad_norm": 0.5192268009841875,
"learning_rate": 1.5349494893087514e-07,
"loss": 0.3098,
"step": 2325
},
{
"epoch": 2.7896,
"grad_norm": 0.446567573109484,
"learning_rate": 1.517838860896964e-07,
"loss": 0.3099,
"step": 2326
},
{
"epoch": 2.7908,
"grad_norm": 0.5095120428916953,
"learning_rate": 1.500822666978735e-07,
"loss": 0.3343,
"step": 2327
},
{
"epoch": 2.792,
"grad_norm": 0.48823817262313274,
"learning_rate": 1.4839009406985295e-07,
"loss": 0.3166,
"step": 2328
},
{
"epoch": 2.7932,
"grad_norm": 0.49617621664603856,
"learning_rate": 1.4670737150168257e-07,
"loss": 0.3257,
"step": 2329
},
{
"epoch": 2.7944,
"grad_norm": 0.6698791671979474,
"learning_rate": 1.4503410227100057e-07,
"loss": 0.3605,
"step": 2330
},
{
"epoch": 2.7956,
"grad_norm": 0.5626220748746263,
"learning_rate": 1.433702896370348e-07,
"loss": 0.3195,
"step": 2331
},
{
"epoch": 2.7968,
"grad_norm": 0.4228810671781407,
"learning_rate": 1.417159368405907e-07,
"loss": 0.3549,
"step": 2332
},
{
"epoch": 2.798,
"grad_norm": 0.5091371276500047,
"learning_rate": 1.4007104710404838e-07,
"loss": 0.3254,
"step": 2333
},
{
"epoch": 2.7992,
"grad_norm": 0.451129778887307,
"learning_rate": 1.384356236313572e-07,
"loss": 0.3034,
"step": 2334
},
{
"epoch": 2.8004,
"grad_norm": 0.5403192642000542,
"learning_rate": 1.3680966960802623e-07,
"loss": 0.3344,
"step": 2335
},
{
"epoch": 2.8016,
"grad_norm": 0.4428088174158261,
"learning_rate": 1.3519318820111983e-07,
"loss": 0.3385,
"step": 2336
},
{
"epoch": 2.8028,
"grad_norm": 0.464586004978293,
"learning_rate": 1.3358618255925214e-07,
"loss": 0.3036,
"step": 2337
},
{
"epoch": 2.8040000000000003,
"grad_norm": 0.43094893731579953,
"learning_rate": 1.3198865581258046e-07,
"loss": 0.3176,
"step": 2338
},
{
"epoch": 2.8052,
"grad_norm": 0.4410368975259025,
"learning_rate": 1.3040061107279679e-07,
"loss": 0.3685,
"step": 2339
},
{
"epoch": 2.8064,
"grad_norm": 0.5689204693655754,
"learning_rate": 1.2882205143312676e-07,
"loss": 0.309,
"step": 2340
},
{
"epoch": 2.8076,
"grad_norm": 0.5638340896622204,
"learning_rate": 1.272529799683192e-07,
"loss": 0.3798,
"step": 2341
},
{
"epoch": 2.8087999999999997,
"grad_norm": 0.5782087446727083,
"learning_rate": 1.2569339973464155e-07,
"loss": 0.3366,
"step": 2342
},
{
"epoch": 2.81,
"grad_norm": 0.499291269992539,
"learning_rate": 1.2414331376987555e-07,
"loss": 0.3406,
"step": 2343
},
{
"epoch": 2.8112,
"grad_norm": 0.4415314530087141,
"learning_rate": 1.2260272509330707e-07,
"loss": 0.3048,
"step": 2344
},
{
"epoch": 2.8124000000000002,
"grad_norm": 0.4904567272284453,
"learning_rate": 1.2107163670572574e-07,
"loss": 0.3074,
"step": 2345
},
{
"epoch": 2.8136,
"grad_norm": 0.4415238455362824,
"learning_rate": 1.195500515894149e-07,
"loss": 0.3227,
"step": 2346
},
{
"epoch": 2.8148,
"grad_norm": 0.5376510127249622,
"learning_rate": 1.1803797270814765e-07,
"loss": 0.3177,
"step": 2347
},
{
"epoch": 2.816,
"grad_norm": 0.4527948890890255,
"learning_rate": 1.165354030071808e-07,
"loss": 0.3297,
"step": 2348
},
{
"epoch": 2.8172,
"grad_norm": 0.4646289088914466,
"learning_rate": 1.1504234541324765e-07,
"loss": 0.3037,
"step": 2349
},
{
"epoch": 2.8184,
"grad_norm": 0.5920759085126542,
"learning_rate": 1.1355880283455523e-07,
"loss": 0.3043,
"step": 2350
},
{
"epoch": 2.8196,
"grad_norm": 0.4049496950901706,
"learning_rate": 1.1208477816077756e-07,
"loss": 0.3159,
"step": 2351
},
{
"epoch": 2.8208,
"grad_norm": 0.48697259926343966,
"learning_rate": 1.1062027426304744e-07,
"loss": 0.3032,
"step": 2352
},
{
"epoch": 2.822,
"grad_norm": 0.4821516548964115,
"learning_rate": 1.091652939939547e-07,
"loss": 0.3165,
"step": 2353
},
{
"epoch": 2.8232,
"grad_norm": 0.5198051132803483,
"learning_rate": 1.0771984018753733e-07,
"loss": 0.3152,
"step": 2354
},
{
"epoch": 2.8244,
"grad_norm": 0.4585361524417428,
"learning_rate": 1.0628391565927765e-07,
"loss": 0.3115,
"step": 2355
},
{
"epoch": 2.8256,
"grad_norm": 0.4859294764163727,
"learning_rate": 1.0485752320609944e-07,
"loss": 0.3138,
"step": 2356
},
{
"epoch": 2.8268,
"grad_norm": 0.4775270643408919,
"learning_rate": 1.0344066560635635e-07,
"loss": 0.3279,
"step": 2357
},
{
"epoch": 2.828,
"grad_norm": 0.4647411286051379,
"learning_rate": 1.0203334561983025e-07,
"loss": 0.3003,
"step": 2358
},
{
"epoch": 2.8292,
"grad_norm": 0.4844392351864473,
"learning_rate": 1.0063556598772839e-07,
"loss": 0.3155,
"step": 2359
},
{
"epoch": 2.8304,
"grad_norm": 0.605604539251082,
"learning_rate": 9.924732943267068e-08,
"loss": 0.3248,
"step": 2360
},
{
"epoch": 2.8316,
"grad_norm": 0.5032824623072444,
"learning_rate": 9.7868638658693e-08,
"loss": 0.3094,
"step": 2361
},
{
"epoch": 2.8327999999999998,
"grad_norm": 0.499371598929128,
"learning_rate": 9.6499496351235e-08,
"loss": 0.286,
"step": 2362
},
{
"epoch": 2.834,
"grad_norm": 0.5373737210650458,
"learning_rate": 9.513990517713955e-08,
"loss": 0.3104,
"step": 2363
},
{
"epoch": 2.8352,
"grad_norm": 0.4510168069438221,
"learning_rate": 9.378986778464327e-08,
"loss": 0.3522,
"step": 2364
},
{
"epoch": 2.8364000000000003,
"grad_norm": 0.4934286179082433,
"learning_rate": 9.244938680337656e-08,
"loss": 0.2772,
"step": 2365
},
{
"epoch": 2.8376,
"grad_norm": 0.45055787521651935,
"learning_rate": 9.111846484435361e-08,
"loss": 0.3121,
"step": 2366
},
{
"epoch": 2.8388,
"grad_norm": 0.48992248863843435,
"learning_rate": 8.979710449997014e-08,
"loss": 0.3474,
"step": 2367
},
{
"epoch": 2.84,
"grad_norm": 0.5360710799868117,
"learning_rate": 8.848530834399683e-08,
"loss": 0.2979,
"step": 2368
},
{
"epoch": 2.8411999999999997,
"grad_norm": 0.5385843863790583,
"learning_rate": 8.718307893157696e-08,
"loss": 0.3061,
"step": 2369
},
{
"epoch": 2.8424,
"grad_norm": 0.4817960470762318,
"learning_rate": 8.589041879921711e-08,
"loss": 0.329,
"step": 2370
},
{
"epoch": 2.8436,
"grad_norm": 0.5981766823358456,
"learning_rate": 8.460733046478653e-08,
"loss": 0.3174,
"step": 2371
},
{
"epoch": 2.8448,
"grad_norm": 0.4538696419025796,
"learning_rate": 8.333381642750882e-08,
"loss": 0.3161,
"step": 2372
},
{
"epoch": 2.846,
"grad_norm": 0.505003014137601,
"learning_rate": 8.206987916796027e-08,
"loss": 0.3087,
"step": 2373
},
{
"epoch": 2.8472,
"grad_norm": 0.42719544721616665,
"learning_rate": 8.081552114806101e-08,
"loss": 0.323,
"step": 2374
},
{
"epoch": 2.8484,
"grad_norm": 0.4898473682925847,
"learning_rate": 7.957074481107551e-08,
"loss": 0.3049,
"step": 2375
},
{
"epoch": 2.8496,
"grad_norm": 0.43235343770639867,
"learning_rate": 7.833555258160208e-08,
"loss": 0.2999,
"step": 2376
},
{
"epoch": 2.8508,
"grad_norm": 0.510727061012366,
"learning_rate": 7.710994686557172e-08,
"loss": 0.2989,
"step": 2377
},
{
"epoch": 2.852,
"grad_norm": 0.5846787914082995,
"learning_rate": 7.589393005024482e-08,
"loss": 0.3349,
"step": 2378
},
{
"epoch": 2.8532,
"grad_norm": 0.49638533272937807,
"learning_rate": 7.468750450420114e-08,
"loss": 0.3358,
"step": 2379
},
{
"epoch": 2.8544,
"grad_norm": 0.501970908690895,
"learning_rate": 7.349067257733989e-08,
"loss": 0.2965,
"step": 2380
},
{
"epoch": 2.8556,
"grad_norm": 0.4181659544610093,
"learning_rate": 7.230343660087402e-08,
"loss": 0.2911,
"step": 2381
},
{
"epoch": 2.8568,
"grad_norm": 0.4714363547285763,
"learning_rate": 7.11257988873243e-08,
"loss": 0.3049,
"step": 2382
},
{
"epoch": 2.858,
"grad_norm": 0.4457501963987575,
"learning_rate": 6.995776173051583e-08,
"loss": 0.3196,
"step": 2383
},
{
"epoch": 2.8592,
"grad_norm": 0.6019596004834906,
"learning_rate": 6.879932740557538e-08,
"loss": 0.2975,
"step": 2384
},
{
"epoch": 2.8604000000000003,
"grad_norm": 0.54148366874017,
"learning_rate": 6.7650498168923e-08,
"loss": 0.3173,
"step": 2385
},
{
"epoch": 2.8616,
"grad_norm": 0.4357959192650805,
"learning_rate": 6.651127625827037e-08,
"loss": 0.3565,
"step": 2386
},
{
"epoch": 2.8628,
"grad_norm": 0.5406915990260817,
"learning_rate": 6.538166389261635e-08,
"loss": 0.3188,
"step": 2387
},
{
"epoch": 2.864,
"grad_norm": 0.43937973467304026,
"learning_rate": 6.426166327224148e-08,
"loss": 0.3133,
"step": 2388
},
{
"epoch": 2.8651999999999997,
"grad_norm": 0.43446611545236996,
"learning_rate": 6.315127657870513e-08,
"loss": 0.3116,
"step": 2389
},
{
"epoch": 2.8664,
"grad_norm": 0.5007274102404864,
"learning_rate": 6.205050597483997e-08,
"loss": 0.3516,
"step": 2390
},
{
"epoch": 2.8676,
"grad_norm": 0.5235249201937633,
"learning_rate": 6.095935360474814e-08,
"loss": 0.3195,
"step": 2391
},
{
"epoch": 2.8688000000000002,
"grad_norm": 0.4295429354477048,
"learning_rate": 5.98778215937973e-08,
"loss": 0.3198,
"step": 2392
},
{
"epoch": 2.87,
"grad_norm": 0.43532397576400567,
"learning_rate": 5.880591204861674e-08,
"loss": 0.3065,
"step": 2393
},
{
"epoch": 2.8712,
"grad_norm": 0.45294078857264364,
"learning_rate": 5.7743627057092463e-08,
"loss": 0.3134,
"step": 2394
},
{
"epoch": 2.8724,
"grad_norm": 0.5769728088585871,
"learning_rate": 5.669096868836377e-08,
"loss": 0.293,
"step": 2395
},
{
"epoch": 2.8736,
"grad_norm": 0.5111669242404633,
"learning_rate": 5.564793899281884e-08,
"loss": 0.322,
"step": 2396
},
{
"epoch": 2.8748,
"grad_norm": 0.48760200574794743,
"learning_rate": 5.461454000209199e-08,
"loss": 0.2901,
"step": 2397
},
{
"epoch": 2.876,
"grad_norm": 0.6298759028578274,
"learning_rate": 5.3590773729056965e-08,
"loss": 0.2865,
"step": 2398
},
{
"epoch": 2.8772,
"grad_norm": 0.4076819774768756,
"learning_rate": 5.257664216782532e-08,
"loss": 0.3175,
"step": 2399
},
{
"epoch": 2.8784,
"grad_norm": 0.5230996839581467,
"learning_rate": 5.157214729374305e-08,
"loss": 0.3008,
"step": 2400
},
{
"epoch": 2.8796,
"grad_norm": 0.4980150312966219,
"learning_rate": 5.057729106338505e-08,
"loss": 0.3097,
"step": 2401
},
{
"epoch": 2.8808,
"grad_norm": 0.6269231740041662,
"learning_rate": 4.959207541455013e-08,
"loss": 0.301,
"step": 2402
},
{
"epoch": 2.882,
"grad_norm": 0.5509563063387615,
"learning_rate": 4.8616502266261026e-08,
"loss": 0.3222,
"step": 2403
},
{
"epoch": 2.8832,
"grad_norm": 0.4777034613305951,
"learning_rate": 4.765057351875879e-08,
"loss": 0.3455,
"step": 2404
},
{
"epoch": 2.8844,
"grad_norm": 0.6229122328651276,
"learning_rate": 4.6694291053496766e-08,
"loss": 0.3427,
"step": 2405
},
{
"epoch": 2.8856,
"grad_norm": 0.4363010937071919,
"learning_rate": 4.5747656733142184e-08,
"loss": 0.3079,
"step": 2406
},
{
"epoch": 2.8868,
"grad_norm": 0.5164317618387799,
"learning_rate": 4.481067240156678e-08,
"loss": 0.3577,
"step": 2407
},
{
"epoch": 2.888,
"grad_norm": 0.5310453003938295,
"learning_rate": 4.3883339883846186e-08,
"loss": 0.3092,
"step": 2408
},
{
"epoch": 2.8891999999999998,
"grad_norm": 0.47591299716717905,
"learning_rate": 4.296566098625776e-08,
"loss": 0.289,
"step": 2409
},
{
"epoch": 2.8904,
"grad_norm": 0.4057762248455579,
"learning_rate": 4.2057637496273896e-08,
"loss": 0.3078,
"step": 2410
},
{
"epoch": 2.8916,
"grad_norm": 0.5170309933348114,
"learning_rate": 4.115927118256036e-08,
"loss": 0.3055,
"step": 2411
},
{
"epoch": 2.8928000000000003,
"grad_norm": 0.5785863146042344,
"learning_rate": 4.02705637949724e-08,
"loss": 0.2945,
"step": 2412
},
{
"epoch": 2.894,
"grad_norm": 0.5280703827425812,
"learning_rate": 3.939151706455146e-08,
"loss": 0.3231,
"step": 2413
},
{
"epoch": 2.8952,
"grad_norm": 0.48114304647455475,
"learning_rate": 3.8522132703521784e-08,
"loss": 0.3291,
"step": 2414
},
{
"epoch": 2.8964,
"grad_norm": 0.6179466081930308,
"learning_rate": 3.7662412405286567e-08,
"loss": 0.2838,
"step": 2415
},
{
"epoch": 2.8975999999999997,
"grad_norm": 0.553574524175213,
"learning_rate": 3.6812357844427385e-08,
"loss": 0.3424,
"step": 2416
},
{
"epoch": 2.8988,
"grad_norm": 0.476372373712313,
"learning_rate": 3.597197067669533e-08,
"loss": 0.2967,
"step": 2417
},
{
"epoch": 2.9,
"grad_norm": 0.4589191775819069,
"learning_rate": 3.514125253901324e-08,
"loss": 0.289,
"step": 2418
},
{
"epoch": 2.9012000000000002,
"grad_norm": 0.5025480183641806,
"learning_rate": 3.432020504947064e-08,
"loss": 0.3634,
"step": 2419
},
{
"epoch": 2.9024,
"grad_norm": 0.5378007119550171,
"learning_rate": 3.350882980731884e-08,
"loss": 0.3292,
"step": 2420
},
{
"epoch": 2.9036,
"grad_norm": 0.5132483847780795,
"learning_rate": 3.27071283929703e-08,
"loss": 0.2929,
"step": 2421
},
{
"epoch": 2.9048,
"grad_norm": 0.5017330891104314,
"learning_rate": 3.191510236799589e-08,
"loss": 0.3086,
"step": 2422
},
{
"epoch": 2.906,
"grad_norm": 0.6054142245943596,
"learning_rate": 3.113275327511767e-08,
"loss": 0.301,
"step": 2423
},
{
"epoch": 2.9072,
"grad_norm": 0.5009794571004275,
"learning_rate": 3.0360082638211666e-08,
"loss": 0.3447,
"step": 2424
},
{
"epoch": 2.9084,
"grad_norm": 0.4911659041607946,
"learning_rate": 2.959709196229954e-08,
"loss": 0.2892,
"step": 2425
},
{
"epoch": 2.9096,
"grad_norm": 0.4597986261621221,
"learning_rate": 2.8843782733549706e-08,
"loss": 0.3198,
"step": 2426
},
{
"epoch": 2.9108,
"grad_norm": 0.5628532674537106,
"learning_rate": 2.8100156419272885e-08,
"loss": 0.2947,
"step": 2427
},
{
"epoch": 2.912,
"grad_norm": 0.4282700586751546,
"learning_rate": 2.7366214467919318e-08,
"loss": 0.3177,
"step": 2428
},
{
"epoch": 2.9132,
"grad_norm": 0.4618771875772092,
"learning_rate": 2.664195830907379e-08,
"loss": 0.319,
"step": 2429
},
{
"epoch": 2.9144,
"grad_norm": 0.5395700391467462,
"learning_rate": 2.5927389353457842e-08,
"loss": 0.3321,
"step": 2430
},
{
"epoch": 2.9156,
"grad_norm": 0.463777655214732,
"learning_rate": 2.5222508992922e-08,
"loss": 0.3089,
"step": 2431
},
{
"epoch": 2.9168,
"grad_norm": 0.5016833316461257,
"learning_rate": 2.4527318600446324e-08,
"loss": 0.3221,
"step": 2432
},
{
"epoch": 2.918,
"grad_norm": 0.658339893908631,
"learning_rate": 2.3841819530135424e-08,
"loss": 0.2819,
"step": 2433
},
{
"epoch": 2.9192,
"grad_norm": 0.5588249527983352,
"learning_rate": 2.3166013117218998e-08,
"loss": 0.3116,
"step": 2434
},
{
"epoch": 2.9204,
"grad_norm": 0.5610785433460539,
"learning_rate": 2.249990067804464e-08,
"loss": 0.3453,
"step": 2435
},
{
"epoch": 2.9215999999999998,
"grad_norm": 0.5814747100855794,
"learning_rate": 2.1843483510080032e-08,
"loss": 0.3171,
"step": 2436
},
{
"epoch": 2.9228,
"grad_norm": 0.44736991991141273,
"learning_rate": 2.119676289190631e-08,
"loss": 0.3127,
"step": 2437
},
{
"epoch": 2.924,
"grad_norm": 0.43175452101413864,
"learning_rate": 2.0559740083219147e-08,
"loss": 0.2804,
"step": 2438
},
{
"epoch": 2.9252000000000002,
"grad_norm": 0.4487740244420664,
"learning_rate": 1.9932416324823235e-08,
"loss": 0.3243,
"step": 2439
},
{
"epoch": 2.9264,
"grad_norm": 0.5800148752493892,
"learning_rate": 1.93147928386328e-08,
"loss": 0.3212,
"step": 2440
},
{
"epoch": 2.9276,
"grad_norm": 0.47301517237011714,
"learning_rate": 1.8706870827666646e-08,
"loss": 0.3203,
"step": 2441
},
{
"epoch": 2.9288,
"grad_norm": 0.47724066371088975,
"learning_rate": 1.8108651476046457e-08,
"loss": 0.3107,
"step": 2442
},
{
"epoch": 2.93,
"grad_norm": 0.4757237945362611,
"learning_rate": 1.7520135948996263e-08,
"loss": 0.3048,
"step": 2443
},
{
"epoch": 2.9312,
"grad_norm": 0.49605027064755813,
"learning_rate": 1.6941325392837437e-08,
"loss": 0.2888,
"step": 2444
},
{
"epoch": 2.9324,
"grad_norm": 0.5628634219339166,
"learning_rate": 1.6372220934988693e-08,
"loss": 0.3292,
"step": 2445
},
{
"epoch": 2.9336,
"grad_norm": 0.5974348122702033,
"learning_rate": 1.5812823683962198e-08,
"loss": 0.3527,
"step": 2446
},
{
"epoch": 2.9348,
"grad_norm": 0.42311587911880083,
"learning_rate": 1.5263134729363582e-08,
"loss": 0.3057,
"step": 2447
},
{
"epoch": 2.936,
"grad_norm": 0.4424085149778813,
"learning_rate": 1.472315514188749e-08,
"loss": 0.3196,
"step": 2448
},
{
"epoch": 2.9372,
"grad_norm": 0.5206514274509301,
"learning_rate": 1.4192885973315918e-08,
"loss": 0.3227,
"step": 2449
},
{
"epoch": 2.9384,
"grad_norm": 0.38928868836839525,
"learning_rate": 1.3672328256518208e-08,
"loss": 0.2786,
"step": 2450
},
{
"epoch": 2.9396,
"grad_norm": 0.49822036281929044,
"learning_rate": 1.3161483005446618e-08,
"loss": 0.3139,
"step": 2451
},
{
"epoch": 2.9408,
"grad_norm": 0.5862817429516328,
"learning_rate": 1.2660351215135203e-08,
"loss": 0.3477,
"step": 2452
},
{
"epoch": 2.942,
"grad_norm": 0.41117931769988564,
"learning_rate": 1.2168933861698151e-08,
"loss": 0.2959,
"step": 2453
},
{
"epoch": 2.9432,
"grad_norm": 0.5331845701928941,
"learning_rate": 1.1687231902328122e-08,
"loss": 0.329,
"step": 2454
},
{
"epoch": 2.9444,
"grad_norm": 0.4131289314624644,
"learning_rate": 1.1215246275292913e-08,
"loss": 0.3148,
"step": 2455
},
{
"epoch": 2.9455999999999998,
"grad_norm": 0.5709882695339584,
"learning_rate": 1.0752977899936013e-08,
"loss": 0.3669,
"step": 2456
},
{
"epoch": 2.9468,
"grad_norm": 0.5383496373093002,
"learning_rate": 1.0300427676672164e-08,
"loss": 0.3207,
"step": 2457
},
{
"epoch": 2.948,
"grad_norm": 0.5036985474975768,
"learning_rate": 9.85759648698792e-09,
"loss": 0.3518,
"step": 2458
},
{
"epoch": 2.9492000000000003,
"grad_norm": 0.48604138385546897,
"learning_rate": 9.42448519343775e-09,
"loss": 0.3003,
"step": 2459
},
{
"epoch": 2.9504,
"grad_norm": 0.46063161763653415,
"learning_rate": 9.001094639645158e-09,
"loss": 0.3145,
"step": 2460
},
{
"epoch": 2.9516,
"grad_norm": 0.44888976548593773,
"learning_rate": 8.587425650297688e-09,
"loss": 0.3134,
"step": 2461
},
{
"epoch": 2.9528,
"grad_norm": 0.5443734200582069,
"learning_rate": 8.183479031148022e-09,
"loss": 0.3055,
"step": 2462
},
{
"epoch": 2.9539999999999997,
"grad_norm": 0.46207920486485327,
"learning_rate": 7.789255569011223e-09,
"loss": 0.3055,
"step": 2463
},
{
"epoch": 2.9552,
"grad_norm": 0.5590105689639684,
"learning_rate": 7.40475603176416e-09,
"loss": 0.3282,
"step": 2464
},
{
"epoch": 2.9564,
"grad_norm": 0.4728662799676889,
"learning_rate": 7.029981168341082e-09,
"loss": 0.3237,
"step": 2465
},
{
"epoch": 2.9576000000000002,
"grad_norm": 0.6122492473730149,
"learning_rate": 6.664931708736943e-09,
"loss": 0.3634,
"step": 2466
},
{
"epoch": 2.9588,
"grad_norm": 0.48871890909775306,
"learning_rate": 6.309608364001296e-09,
"loss": 0.3274,
"step": 2467
},
{
"epoch": 2.96,
"grad_norm": 0.4736736566376079,
"learning_rate": 5.9640118262399575e-09,
"loss": 0.3647,
"step": 2468
},
{
"epoch": 2.9612,
"grad_norm": 0.7702392956065972,
"learning_rate": 5.628142768613343e-09,
"loss": 0.3343,
"step": 2469
},
{
"epoch": 2.9624,
"grad_norm": 0.5110995784521956,
"learning_rate": 5.302001845333138e-09,
"loss": 0.3469,
"step": 2470
},
{
"epoch": 2.9636,
"grad_norm": 0.6271313615816718,
"learning_rate": 4.9855896916634065e-09,
"loss": 0.3011,
"step": 2471
},
{
"epoch": 2.9648,
"grad_norm": 0.5049024790218767,
"learning_rate": 4.678906923916704e-09,
"loss": 0.3349,
"step": 2472
},
{
"epoch": 2.966,
"grad_norm": 0.5083737036919284,
"learning_rate": 4.381954139457411e-09,
"loss": 0.3188,
"step": 2473
},
{
"epoch": 2.9672,
"grad_norm": 0.4439220103812538,
"learning_rate": 4.094731916693962e-09,
"loss": 0.303,
"step": 2474
},
{
"epoch": 2.9684,
"grad_norm": 0.5349117810636819,
"learning_rate": 3.817240815084944e-09,
"loss": 0.3106,
"step": 2475
},
{
"epoch": 2.9696,
"grad_norm": 0.4796397451321217,
"learning_rate": 3.5494813751324466e-09,
"loss": 0.3208,
"step": 2476
},
{
"epoch": 2.9708,
"grad_norm": 0.49657074475370644,
"learning_rate": 3.291454118383164e-09,
"loss": 0.3001,
"step": 2477
},
{
"epoch": 2.972,
"grad_norm": 0.45887033574992925,
"learning_rate": 3.043159547427843e-09,
"loss": 0.297,
"step": 2478
},
{
"epoch": 2.9732,
"grad_norm": 0.4654686233837408,
"learning_rate": 2.804598145899062e-09,
"loss": 0.3224,
"step": 2479
},
{
"epoch": 2.9744,
"grad_norm": 0.5123518073183415,
"learning_rate": 2.575770378472342e-09,
"loss": 0.3298,
"step": 2480
},
{
"epoch": 2.9756,
"grad_norm": 0.5616833089679418,
"learning_rate": 2.3566766908622586e-09,
"loss": 0.3039,
"step": 2481
},
{
"epoch": 2.9768,
"grad_norm": 0.5781294294873015,
"learning_rate": 2.1473175098229993e-09,
"loss": 0.3141,
"step": 2482
},
{
"epoch": 2.9779999999999998,
"grad_norm": 0.45192639710917176,
"learning_rate": 1.9476932431500286e-09,
"loss": 0.3087,
"step": 2483
},
{
"epoch": 2.9792,
"grad_norm": 0.42422671390989036,
"learning_rate": 1.7578042796739803e-09,
"loss": 0.3399,
"step": 2484
},
{
"epoch": 2.9804,
"grad_norm": 0.4340043445979178,
"learning_rate": 1.5776509892645453e-09,
"loss": 0.3088,
"step": 2485
},
{
"epoch": 2.9816000000000003,
"grad_norm": 0.4425405316221765,
"learning_rate": 1.4072337228282496e-09,
"loss": 0.3216,
"step": 2486
},
{
"epoch": 2.9828,
"grad_norm": 0.4987650680285554,
"learning_rate": 1.2465528123073445e-09,
"loss": 0.2955,
"step": 2487
},
{
"epoch": 2.984,
"grad_norm": 0.43270212291705107,
"learning_rate": 1.0956085706781416e-09,
"loss": 0.3315,
"step": 2488
},
{
"epoch": 2.9852,
"grad_norm": 0.5028265173730337,
"learning_rate": 9.54401291953788e-10,
"loss": 0.3101,
"step": 2489
},
{
"epoch": 2.9864,
"grad_norm": 0.5165685400479998,
"learning_rate": 8.229312511803811e-10,
"loss": 0.2875,
"step": 2490
},
{
"epoch": 2.9876,
"grad_norm": 0.5184185637940487,
"learning_rate": 7.011987044369673e-10,
"loss": 0.3225,
"step": 2491
},
{
"epoch": 2.9888,
"grad_norm": 0.6063108943109075,
"learning_rate": 5.892038888377638e-10,
"loss": 0.3134,
"step": 2492
},
{
"epoch": 2.99,
"grad_norm": 0.4981802545627052,
"learning_rate": 4.869470225277174e-10,
"loss": 0.3149,
"step": 2493
},
{
"epoch": 2.9912,
"grad_norm": 0.47237623616462293,
"learning_rate": 3.9442830468472414e-10,
"loss": 0.3044,
"step": 2494
},
{
"epoch": 2.9924,
"grad_norm": 0.5143523787824885,
"learning_rate": 3.1164791551907545e-10,
"loss": 0.3246,
"step": 2495
},
{
"epoch": 2.9936,
"grad_norm": 0.44405248487108334,
"learning_rate": 2.386060162717918e-10,
"loss": 0.2675,
"step": 2496
},
{
"epoch": 2.9948,
"grad_norm": 0.4952936205896277,
"learning_rate": 1.7530274921462308e-10,
"loss": 0.3095,
"step": 2497
},
{
"epoch": 2.996,
"grad_norm": 0.5220038590190723,
"learning_rate": 1.21738237651714e-10,
"loss": 0.2753,
"step": 2498
},
{
"epoch": 2.9972,
"grad_norm": 0.5220070903249548,
"learning_rate": 7.791258591682837e-11,
"loss": 0.3656,
"step": 2499
},
{
"epoch": 2.9984,
"grad_norm": 0.51190349865152,
"learning_rate": 4.382587937445948e-11,
"loss": 0.3388,
"step": 2500
},
{
"epoch": 2.9996,
"grad_norm": 0.4599694574699737,
"learning_rate": 1.947818441927485e-11,
"loss": 0.317,
"step": 2501
},
{
"epoch": 3.0,
"grad_norm": 0.9361995803938117,
"learning_rate": 4.8695484761163145e-12,
"loss": 0.374,
"step": 2502
},
{
"epoch": 3.0,
"step": 2502,
"total_flos": 898946847588352.0,
"train_loss": 0.38813807606268275,
"train_runtime": 64430.8646,
"train_samples_per_second": 1.862,
"train_steps_per_second": 0.039
}
],
"logging_steps": 1,
"max_steps": 2502,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 4000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 898946847588352.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}