HT-duplicate-Qwen-20k / trainer_state.json
ShourenWSR's picture
Upload folder using huggingface_hub
c7c9be8 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": 5.437974782280815,
"learning_rate": 0.0,
"loss": 0.7193,
"step": 1
},
{
"epoch": 0.0024,
"grad_norm": 6.842162085432246,
"learning_rate": 3.984063745019921e-08,
"loss": 0.9452,
"step": 2
},
{
"epoch": 0.0036,
"grad_norm": 6.355658561135521,
"learning_rate": 7.968127490039842e-08,
"loss": 0.8366,
"step": 3
},
{
"epoch": 0.0048,
"grad_norm": 5.924836611406062,
"learning_rate": 1.195219123505976e-07,
"loss": 0.8188,
"step": 4
},
{
"epoch": 0.006,
"grad_norm": 5.326313244153661,
"learning_rate": 1.5936254980079683e-07,
"loss": 0.8408,
"step": 5
},
{
"epoch": 0.0072,
"grad_norm": 6.325730629066542,
"learning_rate": 1.9920318725099604e-07,
"loss": 0.8177,
"step": 6
},
{
"epoch": 0.0084,
"grad_norm": 6.252775759714414,
"learning_rate": 2.390438247011952e-07,
"loss": 0.8224,
"step": 7
},
{
"epoch": 0.0096,
"grad_norm": 5.795867645434717,
"learning_rate": 2.7888446215139444e-07,
"loss": 0.7766,
"step": 8
},
{
"epoch": 0.0108,
"grad_norm": 5.2853858751488305,
"learning_rate": 3.1872509960159367e-07,
"loss": 0.7183,
"step": 9
},
{
"epoch": 0.012,
"grad_norm": 6.056253155406832,
"learning_rate": 3.585657370517929e-07,
"loss": 0.8337,
"step": 10
},
{
"epoch": 0.0132,
"grad_norm": 6.301648473946036,
"learning_rate": 3.9840637450199207e-07,
"loss": 0.8744,
"step": 11
},
{
"epoch": 0.0144,
"grad_norm": 5.7316860083810495,
"learning_rate": 4.382470119521913e-07,
"loss": 0.7591,
"step": 12
},
{
"epoch": 0.0156,
"grad_norm": 5.049538453325047,
"learning_rate": 4.780876494023904e-07,
"loss": 0.7252,
"step": 13
},
{
"epoch": 0.0168,
"grad_norm": 6.077750133392651,
"learning_rate": 5.179282868525898e-07,
"loss": 0.8345,
"step": 14
},
{
"epoch": 0.018,
"grad_norm": 6.203537482181997,
"learning_rate": 5.577689243027889e-07,
"loss": 0.825,
"step": 15
},
{
"epoch": 0.0192,
"grad_norm": 4.842573626700455,
"learning_rate": 5.976095617529881e-07,
"loss": 0.7521,
"step": 16
},
{
"epoch": 0.0204,
"grad_norm": 4.790770839272536,
"learning_rate": 6.374501992031873e-07,
"loss": 0.7521,
"step": 17
},
{
"epoch": 0.0216,
"grad_norm": 4.837716860428012,
"learning_rate": 6.772908366533865e-07,
"loss": 0.7732,
"step": 18
},
{
"epoch": 0.0228,
"grad_norm": 4.184149256799008,
"learning_rate": 7.171314741035858e-07,
"loss": 0.7206,
"step": 19
},
{
"epoch": 0.024,
"grad_norm": 3.9986348904040243,
"learning_rate": 7.569721115537849e-07,
"loss": 0.6698,
"step": 20
},
{
"epoch": 0.0252,
"grad_norm": 4.248473754128453,
"learning_rate": 7.968127490039841e-07,
"loss": 0.7471,
"step": 21
},
{
"epoch": 0.0264,
"grad_norm": 2.91819353570462,
"learning_rate": 8.366533864541833e-07,
"loss": 0.7136,
"step": 22
},
{
"epoch": 0.0276,
"grad_norm": 2.847210925319652,
"learning_rate": 8.764940239043826e-07,
"loss": 0.7068,
"step": 23
},
{
"epoch": 0.0288,
"grad_norm": 2.602789823390448,
"learning_rate": 9.163346613545817e-07,
"loss": 0.7163,
"step": 24
},
{
"epoch": 0.03,
"grad_norm": 2.847371756412202,
"learning_rate": 9.561752988047808e-07,
"loss": 0.7686,
"step": 25
},
{
"epoch": 0.0312,
"grad_norm": 2.4224905492215862,
"learning_rate": 9.9601593625498e-07,
"loss": 0.7027,
"step": 26
},
{
"epoch": 0.0324,
"grad_norm": 2.3435862696930987,
"learning_rate": 1.0358565737051795e-06,
"loss": 0.6542,
"step": 27
},
{
"epoch": 0.0336,
"grad_norm": 2.3628391267754387,
"learning_rate": 1.0756972111553785e-06,
"loss": 0.6905,
"step": 28
},
{
"epoch": 0.0348,
"grad_norm": 1.9204427730897262,
"learning_rate": 1.1155378486055778e-06,
"loss": 0.6851,
"step": 29
},
{
"epoch": 0.036,
"grad_norm": 1.7690807873252774,
"learning_rate": 1.155378486055777e-06,
"loss": 0.6097,
"step": 30
},
{
"epoch": 0.0372,
"grad_norm": 2.67530126623312,
"learning_rate": 1.1952191235059762e-06,
"loss": 0.7184,
"step": 31
},
{
"epoch": 0.0384,
"grad_norm": 2.8296733572520822,
"learning_rate": 1.2350597609561754e-06,
"loss": 0.6422,
"step": 32
},
{
"epoch": 0.0396,
"grad_norm": 2.8085381123819815,
"learning_rate": 1.2749003984063747e-06,
"loss": 0.7202,
"step": 33
},
{
"epoch": 0.0408,
"grad_norm": 2.4328010899445625,
"learning_rate": 1.3147410358565737e-06,
"loss": 0.6412,
"step": 34
},
{
"epoch": 0.042,
"grad_norm": 2.6938356497763496,
"learning_rate": 1.354581673306773e-06,
"loss": 0.6266,
"step": 35
},
{
"epoch": 0.0432,
"grad_norm": 2.2564777858540164,
"learning_rate": 1.3944223107569721e-06,
"loss": 0.6138,
"step": 36
},
{
"epoch": 0.0444,
"grad_norm": 2.504817899421148,
"learning_rate": 1.4342629482071716e-06,
"loss": 0.6767,
"step": 37
},
{
"epoch": 0.0456,
"grad_norm": 1.9560766406484427,
"learning_rate": 1.4741035856573708e-06,
"loss": 0.6381,
"step": 38
},
{
"epoch": 0.0468,
"grad_norm": 1.78023857220181,
"learning_rate": 1.5139442231075698e-06,
"loss": 0.634,
"step": 39
},
{
"epoch": 0.048,
"grad_norm": 1.5097231299163532,
"learning_rate": 1.553784860557769e-06,
"loss": 0.5882,
"step": 40
},
{
"epoch": 0.0492,
"grad_norm": 1.371228632904813,
"learning_rate": 1.5936254980079683e-06,
"loss": 0.6358,
"step": 41
},
{
"epoch": 0.0504,
"grad_norm": 1.486399285202626,
"learning_rate": 1.6334661354581673e-06,
"loss": 0.6333,
"step": 42
},
{
"epoch": 0.0516,
"grad_norm": 1.2874409401054354,
"learning_rate": 1.6733067729083665e-06,
"loss": 0.5843,
"step": 43
},
{
"epoch": 0.0528,
"grad_norm": 1.195525415079638,
"learning_rate": 1.7131474103585658e-06,
"loss": 0.6319,
"step": 44
},
{
"epoch": 0.054,
"grad_norm": 1.405789824170085,
"learning_rate": 1.7529880478087652e-06,
"loss": 0.5645,
"step": 45
},
{
"epoch": 0.0552,
"grad_norm": 1.3753941175120579,
"learning_rate": 1.7928286852589644e-06,
"loss": 0.614,
"step": 46
},
{
"epoch": 0.0564,
"grad_norm": 1.3013061588150556,
"learning_rate": 1.8326693227091634e-06,
"loss": 0.5316,
"step": 47
},
{
"epoch": 0.0576,
"grad_norm": 1.4421672498239801,
"learning_rate": 1.8725099601593627e-06,
"loss": 0.5921,
"step": 48
},
{
"epoch": 0.0588,
"grad_norm": 1.112202213113257,
"learning_rate": 1.9123505976095617e-06,
"loss": 0.5765,
"step": 49
},
{
"epoch": 0.06,
"grad_norm": 1.1529005363886367,
"learning_rate": 1.952191235059761e-06,
"loss": 0.573,
"step": 50
},
{
"epoch": 0.0612,
"grad_norm": 1.2138463753222286,
"learning_rate": 1.99203187250996e-06,
"loss": 0.5148,
"step": 51
},
{
"epoch": 0.0624,
"grad_norm": 1.3526729764280512,
"learning_rate": 2.03187250996016e-06,
"loss": 0.5849,
"step": 52
},
{
"epoch": 0.0636,
"grad_norm": 1.2617132921797698,
"learning_rate": 2.071713147410359e-06,
"loss": 0.5952,
"step": 53
},
{
"epoch": 0.0648,
"grad_norm": 1.0741247836377654,
"learning_rate": 2.111553784860558e-06,
"loss": 0.5346,
"step": 54
},
{
"epoch": 0.066,
"grad_norm": 1.2585682641918436,
"learning_rate": 2.151394422310757e-06,
"loss": 0.6219,
"step": 55
},
{
"epoch": 0.0672,
"grad_norm": 1.203560701408079,
"learning_rate": 2.1912350597609563e-06,
"loss": 0.6504,
"step": 56
},
{
"epoch": 0.0684,
"grad_norm": 0.9598854345956231,
"learning_rate": 2.2310756972111555e-06,
"loss": 0.6001,
"step": 57
},
{
"epoch": 0.0696,
"grad_norm": 1.0992564559566185,
"learning_rate": 2.2709163346613547e-06,
"loss": 0.5439,
"step": 58
},
{
"epoch": 0.0708,
"grad_norm": 1.156500393692863,
"learning_rate": 2.310756972111554e-06,
"loss": 0.5627,
"step": 59
},
{
"epoch": 0.072,
"grad_norm": 1.1790175581989504,
"learning_rate": 2.350597609561753e-06,
"loss": 0.5678,
"step": 60
},
{
"epoch": 0.0732,
"grad_norm": 0.9676854085466849,
"learning_rate": 2.3904382470119524e-06,
"loss": 0.5236,
"step": 61
},
{
"epoch": 0.0744,
"grad_norm": 1.0396113612763562,
"learning_rate": 2.4302788844621517e-06,
"loss": 0.5578,
"step": 62
},
{
"epoch": 0.0756,
"grad_norm": 0.9204603674207228,
"learning_rate": 2.470119521912351e-06,
"loss": 0.5177,
"step": 63
},
{
"epoch": 0.0768,
"grad_norm": 1.0032428688781194,
"learning_rate": 2.50996015936255e-06,
"loss": 0.5235,
"step": 64
},
{
"epoch": 0.078,
"grad_norm": 1.022877400084754,
"learning_rate": 2.5498007968127493e-06,
"loss": 0.5132,
"step": 65
},
{
"epoch": 0.0792,
"grad_norm": 1.2562167071607413,
"learning_rate": 2.589641434262948e-06,
"loss": 0.5778,
"step": 66
},
{
"epoch": 0.0804,
"grad_norm": 1.1233370853472004,
"learning_rate": 2.6294820717131474e-06,
"loss": 0.6101,
"step": 67
},
{
"epoch": 0.0816,
"grad_norm": 1.248221719735639,
"learning_rate": 2.6693227091633466e-06,
"loss": 0.6065,
"step": 68
},
{
"epoch": 0.0828,
"grad_norm": 1.0800371225868943,
"learning_rate": 2.709163346613546e-06,
"loss": 0.579,
"step": 69
},
{
"epoch": 0.084,
"grad_norm": 1.092352445738113,
"learning_rate": 2.749003984063745e-06,
"loss": 0.468,
"step": 70
},
{
"epoch": 0.0852,
"grad_norm": 1.0140929679449753,
"learning_rate": 2.7888446215139443e-06,
"loss": 0.5602,
"step": 71
},
{
"epoch": 0.0864,
"grad_norm": 0.957577673497774,
"learning_rate": 2.828685258964144e-06,
"loss": 0.5571,
"step": 72
},
{
"epoch": 0.0876,
"grad_norm": 0.8888559325056431,
"learning_rate": 2.868525896414343e-06,
"loss": 0.4821,
"step": 73
},
{
"epoch": 0.0888,
"grad_norm": 1.1322533413758695,
"learning_rate": 2.9083665338645424e-06,
"loss": 0.5592,
"step": 74
},
{
"epoch": 0.09,
"grad_norm": 0.97763957238552,
"learning_rate": 2.9482071713147416e-06,
"loss": 0.5186,
"step": 75
},
{
"epoch": 0.0912,
"grad_norm": 1.0356924543500423,
"learning_rate": 2.9880478087649404e-06,
"loss": 0.5695,
"step": 76
},
{
"epoch": 0.0924,
"grad_norm": 0.9962947546355103,
"learning_rate": 3.0278884462151397e-06,
"loss": 0.5127,
"step": 77
},
{
"epoch": 0.0936,
"grad_norm": 0.8575404806126316,
"learning_rate": 3.067729083665339e-06,
"loss": 0.5279,
"step": 78
},
{
"epoch": 0.0948,
"grad_norm": 0.9138166070865791,
"learning_rate": 3.107569721115538e-06,
"loss": 0.5,
"step": 79
},
{
"epoch": 0.096,
"grad_norm": 0.819971935800434,
"learning_rate": 3.1474103585657373e-06,
"loss": 0.5036,
"step": 80
},
{
"epoch": 0.0972,
"grad_norm": 0.8834307354764792,
"learning_rate": 3.1872509960159366e-06,
"loss": 0.4711,
"step": 81
},
{
"epoch": 0.0984,
"grad_norm": 0.7373766208496033,
"learning_rate": 3.227091633466136e-06,
"loss": 0.4737,
"step": 82
},
{
"epoch": 0.0996,
"grad_norm": 0.9650390271824048,
"learning_rate": 3.2669322709163346e-06,
"loss": 0.5063,
"step": 83
},
{
"epoch": 0.1008,
"grad_norm": 0.8768753423774562,
"learning_rate": 3.306772908366534e-06,
"loss": 0.5038,
"step": 84
},
{
"epoch": 0.102,
"grad_norm": 0.8127574478371596,
"learning_rate": 3.346613545816733e-06,
"loss": 0.4746,
"step": 85
},
{
"epoch": 0.1032,
"grad_norm": 0.9474942304164558,
"learning_rate": 3.3864541832669323e-06,
"loss": 0.583,
"step": 86
},
{
"epoch": 0.1044,
"grad_norm": 1.0013443483518105,
"learning_rate": 3.4262948207171315e-06,
"loss": 0.5385,
"step": 87
},
{
"epoch": 0.1056,
"grad_norm": 1.1511380035131105,
"learning_rate": 3.466135458167331e-06,
"loss": 0.5463,
"step": 88
},
{
"epoch": 0.1068,
"grad_norm": 1.0592663941691445,
"learning_rate": 3.5059760956175304e-06,
"loss": 0.5385,
"step": 89
},
{
"epoch": 0.108,
"grad_norm": 0.9901664394198704,
"learning_rate": 3.5458167330677296e-06,
"loss": 0.5121,
"step": 90
},
{
"epoch": 0.1092,
"grad_norm": 0.9238068417645418,
"learning_rate": 3.585657370517929e-06,
"loss": 0.5045,
"step": 91
},
{
"epoch": 0.1104,
"grad_norm": 0.8450994487843391,
"learning_rate": 3.625498007968128e-06,
"loss": 0.4761,
"step": 92
},
{
"epoch": 0.1116,
"grad_norm": 0.814074816355805,
"learning_rate": 3.665338645418327e-06,
"loss": 0.5261,
"step": 93
},
{
"epoch": 0.1128,
"grad_norm": 0.9960814212413449,
"learning_rate": 3.705179282868526e-06,
"loss": 0.5257,
"step": 94
},
{
"epoch": 0.114,
"grad_norm": 0.8783708128322982,
"learning_rate": 3.7450199203187254e-06,
"loss": 0.5519,
"step": 95
},
{
"epoch": 0.1152,
"grad_norm": 1.0483005034604667,
"learning_rate": 3.7848605577689246e-06,
"loss": 0.5437,
"step": 96
},
{
"epoch": 0.1164,
"grad_norm": 0.8179308482072307,
"learning_rate": 3.824701195219123e-06,
"loss": 0.4938,
"step": 97
},
{
"epoch": 0.1176,
"grad_norm": 1.3018552667116559,
"learning_rate": 3.864541832669323e-06,
"loss": 0.5001,
"step": 98
},
{
"epoch": 0.1188,
"grad_norm": 0.853404559727024,
"learning_rate": 3.904382470119522e-06,
"loss": 0.4855,
"step": 99
},
{
"epoch": 0.12,
"grad_norm": 1.1350677836850158,
"learning_rate": 3.9442231075697215e-06,
"loss": 0.4694,
"step": 100
},
{
"epoch": 0.1212,
"grad_norm": 0.9540558978625329,
"learning_rate": 3.98406374501992e-06,
"loss": 0.495,
"step": 101
},
{
"epoch": 0.1224,
"grad_norm": 0.9937439793131955,
"learning_rate": 4.02390438247012e-06,
"loss": 0.5322,
"step": 102
},
{
"epoch": 0.1236,
"grad_norm": 1.1286337927934704,
"learning_rate": 4.06374501992032e-06,
"loss": 0.512,
"step": 103
},
{
"epoch": 0.1248,
"grad_norm": 1.0093146016482792,
"learning_rate": 4.103585657370518e-06,
"loss": 0.5127,
"step": 104
},
{
"epoch": 0.126,
"grad_norm": 0.920887089744768,
"learning_rate": 4.143426294820718e-06,
"loss": 0.5274,
"step": 105
},
{
"epoch": 0.1272,
"grad_norm": 1.2952853616257172,
"learning_rate": 4.183266932270917e-06,
"loss": 0.5246,
"step": 106
},
{
"epoch": 0.1284,
"grad_norm": 0.9820669915194236,
"learning_rate": 4.223107569721116e-06,
"loss": 0.442,
"step": 107
},
{
"epoch": 0.1296,
"grad_norm": 1.0295626367503317,
"learning_rate": 4.262948207171315e-06,
"loss": 0.4313,
"step": 108
},
{
"epoch": 0.1308,
"grad_norm": 1.055611088395656,
"learning_rate": 4.302788844621514e-06,
"loss": 0.5222,
"step": 109
},
{
"epoch": 0.132,
"grad_norm": 0.9831334707901427,
"learning_rate": 4.342629482071714e-06,
"loss": 0.4894,
"step": 110
},
{
"epoch": 0.1332,
"grad_norm": 1.0260654582039745,
"learning_rate": 4.382470119521913e-06,
"loss": 0.5642,
"step": 111
},
{
"epoch": 0.1344,
"grad_norm": 1.0488559822828725,
"learning_rate": 4.422310756972112e-06,
"loss": 0.4928,
"step": 112
},
{
"epoch": 0.1356,
"grad_norm": 1.0325450455053031,
"learning_rate": 4.462151394422311e-06,
"loss": 0.4719,
"step": 113
},
{
"epoch": 0.1368,
"grad_norm": 0.9827648807531677,
"learning_rate": 4.50199203187251e-06,
"loss": 0.5766,
"step": 114
},
{
"epoch": 0.138,
"grad_norm": 1.1061657182058315,
"learning_rate": 4.5418326693227095e-06,
"loss": 0.4424,
"step": 115
},
{
"epoch": 0.1392,
"grad_norm": 0.9815737377997215,
"learning_rate": 4.581673306772908e-06,
"loss": 0.4805,
"step": 116
},
{
"epoch": 0.1404,
"grad_norm": 1.01891912549681,
"learning_rate": 4.621513944223108e-06,
"loss": 0.5733,
"step": 117
},
{
"epoch": 0.1416,
"grad_norm": 0.8267290631611671,
"learning_rate": 4.661354581673307e-06,
"loss": 0.4911,
"step": 118
},
{
"epoch": 0.1428,
"grad_norm": 0.9807807425460487,
"learning_rate": 4.701195219123506e-06,
"loss": 0.4921,
"step": 119
},
{
"epoch": 0.144,
"grad_norm": 0.8101086486760266,
"learning_rate": 4.741035856573706e-06,
"loss": 0.5527,
"step": 120
},
{
"epoch": 0.1452,
"grad_norm": 0.9289448915016966,
"learning_rate": 4.780876494023905e-06,
"loss": 0.5266,
"step": 121
},
{
"epoch": 0.1464,
"grad_norm": 0.8038797872589569,
"learning_rate": 4.8207171314741045e-06,
"loss": 0.5156,
"step": 122
},
{
"epoch": 0.1476,
"grad_norm": 1.2718932508165854,
"learning_rate": 4.860557768924303e-06,
"loss": 0.5226,
"step": 123
},
{
"epoch": 0.1488,
"grad_norm": 0.9742510348163138,
"learning_rate": 4.900398406374502e-06,
"loss": 0.4415,
"step": 124
},
{
"epoch": 0.15,
"grad_norm": 1.0071410763993498,
"learning_rate": 4.940239043824702e-06,
"loss": 0.5229,
"step": 125
},
{
"epoch": 0.1512,
"grad_norm": 0.9210403095698111,
"learning_rate": 4.980079681274901e-06,
"loss": 0.4895,
"step": 126
},
{
"epoch": 0.1524,
"grad_norm": 1.0197489386006973,
"learning_rate": 5.0199203187251e-06,
"loss": 0.4999,
"step": 127
},
{
"epoch": 0.1536,
"grad_norm": 1.0848479407295935,
"learning_rate": 5.059760956175299e-06,
"loss": 0.4447,
"step": 128
},
{
"epoch": 0.1548,
"grad_norm": 0.9836478213462632,
"learning_rate": 5.099601593625499e-06,
"loss": 0.4416,
"step": 129
},
{
"epoch": 0.156,
"grad_norm": 1.0409878583249208,
"learning_rate": 5.1394422310756975e-06,
"loss": 0.4946,
"step": 130
},
{
"epoch": 0.1572,
"grad_norm": 0.8537658082457844,
"learning_rate": 5.179282868525896e-06,
"loss": 0.4785,
"step": 131
},
{
"epoch": 0.1584,
"grad_norm": 1.0207875966184468,
"learning_rate": 5.219123505976096e-06,
"loss": 0.4434,
"step": 132
},
{
"epoch": 0.1596,
"grad_norm": 1.1386506118389828,
"learning_rate": 5.258964143426295e-06,
"loss": 0.4666,
"step": 133
},
{
"epoch": 0.1608,
"grad_norm": 0.9400780815977201,
"learning_rate": 5.298804780876494e-06,
"loss": 0.5284,
"step": 134
},
{
"epoch": 0.162,
"grad_norm": 0.8761474725977585,
"learning_rate": 5.338645418326693e-06,
"loss": 0.4993,
"step": 135
},
{
"epoch": 0.1632,
"grad_norm": 0.935219722443438,
"learning_rate": 5.378486055776893e-06,
"loss": 0.5012,
"step": 136
},
{
"epoch": 0.1644,
"grad_norm": 1.166475858975654,
"learning_rate": 5.418326693227092e-06,
"loss": 0.4532,
"step": 137
},
{
"epoch": 0.1656,
"grad_norm": 1.0948869351052124,
"learning_rate": 5.4581673306772905e-06,
"loss": 0.5031,
"step": 138
},
{
"epoch": 0.1668,
"grad_norm": 1.1261555457798458,
"learning_rate": 5.49800796812749e-06,
"loss": 0.4681,
"step": 139
},
{
"epoch": 0.168,
"grad_norm": 1.0377050843870503,
"learning_rate": 5.537848605577689e-06,
"loss": 0.4222,
"step": 140
},
{
"epoch": 0.1692,
"grad_norm": 1.083876204769456,
"learning_rate": 5.577689243027889e-06,
"loss": 0.6036,
"step": 141
},
{
"epoch": 0.1704,
"grad_norm": 1.161601324214863,
"learning_rate": 5.617529880478087e-06,
"loss": 0.4748,
"step": 142
},
{
"epoch": 0.1716,
"grad_norm": 0.9644121510158614,
"learning_rate": 5.657370517928288e-06,
"loss": 0.464,
"step": 143
},
{
"epoch": 0.1728,
"grad_norm": 1.026333692733931,
"learning_rate": 5.697211155378487e-06,
"loss": 0.5446,
"step": 144
},
{
"epoch": 0.174,
"grad_norm": 0.933927631669845,
"learning_rate": 5.737051792828686e-06,
"loss": 0.4277,
"step": 145
},
{
"epoch": 0.1752,
"grad_norm": 0.9567998944756527,
"learning_rate": 5.776892430278885e-06,
"loss": 0.4483,
"step": 146
},
{
"epoch": 0.1764,
"grad_norm": 1.0564276318065147,
"learning_rate": 5.816733067729085e-06,
"loss": 0.444,
"step": 147
},
{
"epoch": 0.1776,
"grad_norm": 1.3401247869309183,
"learning_rate": 5.856573705179284e-06,
"loss": 0.5238,
"step": 148
},
{
"epoch": 0.1788,
"grad_norm": 1.487752140632628,
"learning_rate": 5.896414342629483e-06,
"loss": 0.5296,
"step": 149
},
{
"epoch": 0.18,
"grad_norm": 1.2204963749620594,
"learning_rate": 5.936254980079682e-06,
"loss": 0.4779,
"step": 150
},
{
"epoch": 0.1812,
"grad_norm": 0.9990671790853447,
"learning_rate": 5.976095617529881e-06,
"loss": 0.4559,
"step": 151
},
{
"epoch": 0.1824,
"grad_norm": 0.8683645887802781,
"learning_rate": 6.0159362549800805e-06,
"loss": 0.4104,
"step": 152
},
{
"epoch": 0.1836,
"grad_norm": 1.0245078867577397,
"learning_rate": 6.055776892430279e-06,
"loss": 0.4584,
"step": 153
},
{
"epoch": 0.1848,
"grad_norm": 0.940334237564819,
"learning_rate": 6.095617529880479e-06,
"loss": 0.499,
"step": 154
},
{
"epoch": 0.186,
"grad_norm": 0.9842610908234986,
"learning_rate": 6.135458167330678e-06,
"loss": 0.489,
"step": 155
},
{
"epoch": 0.1872,
"grad_norm": 0.9082060907473806,
"learning_rate": 6.1752988047808774e-06,
"loss": 0.4781,
"step": 156
},
{
"epoch": 0.1884,
"grad_norm": 0.812392878584094,
"learning_rate": 6.215139442231076e-06,
"loss": 0.4114,
"step": 157
},
{
"epoch": 0.1896,
"grad_norm": 1.1042867399753167,
"learning_rate": 6.254980079681275e-06,
"loss": 0.4341,
"step": 158
},
{
"epoch": 0.1908,
"grad_norm": 0.9561224712673747,
"learning_rate": 6.294820717131475e-06,
"loss": 0.5047,
"step": 159
},
{
"epoch": 0.192,
"grad_norm": 1.007172356759588,
"learning_rate": 6.3346613545816735e-06,
"loss": 0.5179,
"step": 160
},
{
"epoch": 0.1932,
"grad_norm": 1.0244919606819907,
"learning_rate": 6.374501992031873e-06,
"loss": 0.5316,
"step": 161
},
{
"epoch": 0.1944,
"grad_norm": 0.9706434974525322,
"learning_rate": 6.414342629482072e-06,
"loss": 0.4617,
"step": 162
},
{
"epoch": 0.1956,
"grad_norm": 0.9953503774369005,
"learning_rate": 6.454183266932272e-06,
"loss": 0.5084,
"step": 163
},
{
"epoch": 0.1968,
"grad_norm": 0.8766032809615142,
"learning_rate": 6.49402390438247e-06,
"loss": 0.4697,
"step": 164
},
{
"epoch": 0.198,
"grad_norm": 1.1905157324799795,
"learning_rate": 6.533864541832669e-06,
"loss": 0.5502,
"step": 165
},
{
"epoch": 0.1992,
"grad_norm": 0.958416513970278,
"learning_rate": 6.573705179282869e-06,
"loss": 0.5253,
"step": 166
},
{
"epoch": 0.2004,
"grad_norm": 1.1155638678555497,
"learning_rate": 6.613545816733068e-06,
"loss": 0.4682,
"step": 167
},
{
"epoch": 0.2016,
"grad_norm": 0.9621075654768388,
"learning_rate": 6.653386454183267e-06,
"loss": 0.4064,
"step": 168
},
{
"epoch": 0.2028,
"grad_norm": 0.8784910371087611,
"learning_rate": 6.693227091633466e-06,
"loss": 0.5153,
"step": 169
},
{
"epoch": 0.204,
"grad_norm": 1.190711658875052,
"learning_rate": 6.733067729083666e-06,
"loss": 0.5322,
"step": 170
},
{
"epoch": 0.2052,
"grad_norm": 1.0540666789051163,
"learning_rate": 6.772908366533865e-06,
"loss": 0.4932,
"step": 171
},
{
"epoch": 0.2064,
"grad_norm": 1.1838420019651616,
"learning_rate": 6.812749003984063e-06,
"loss": 0.4631,
"step": 172
},
{
"epoch": 0.2076,
"grad_norm": 1.0334725601559618,
"learning_rate": 6.852589641434263e-06,
"loss": 0.4608,
"step": 173
},
{
"epoch": 0.2088,
"grad_norm": 0.9482754462649715,
"learning_rate": 6.8924302788844635e-06,
"loss": 0.469,
"step": 174
},
{
"epoch": 0.21,
"grad_norm": 1.190921034868312,
"learning_rate": 6.932270916334662e-06,
"loss": 0.5589,
"step": 175
},
{
"epoch": 0.2112,
"grad_norm": 0.8871824092291994,
"learning_rate": 6.972111553784862e-06,
"loss": 0.4792,
"step": 176
},
{
"epoch": 0.2124,
"grad_norm": 1.0216949578139023,
"learning_rate": 7.011952191235061e-06,
"loss": 0.4325,
"step": 177
},
{
"epoch": 0.2136,
"grad_norm": 1.2127812921072645,
"learning_rate": 7.05179282868526e-06,
"loss": 0.586,
"step": 178
},
{
"epoch": 0.2148,
"grad_norm": 1.110029722356259,
"learning_rate": 7.091633466135459e-06,
"loss": 0.483,
"step": 179
},
{
"epoch": 0.216,
"grad_norm": 1.128565737183673,
"learning_rate": 7.131474103585658e-06,
"loss": 0.4842,
"step": 180
},
{
"epoch": 0.2172,
"grad_norm": 1.46117876490334,
"learning_rate": 7.171314741035858e-06,
"loss": 0.5312,
"step": 181
},
{
"epoch": 0.2184,
"grad_norm": 1.1872425310629424,
"learning_rate": 7.2111553784860565e-06,
"loss": 0.4361,
"step": 182
},
{
"epoch": 0.2196,
"grad_norm": 1.1698656302535893,
"learning_rate": 7.250996015936256e-06,
"loss": 0.4783,
"step": 183
},
{
"epoch": 0.2208,
"grad_norm": 1.0150624039148788,
"learning_rate": 7.290836653386455e-06,
"loss": 0.497,
"step": 184
},
{
"epoch": 0.222,
"grad_norm": 0.9228982859724498,
"learning_rate": 7.330677290836654e-06,
"loss": 0.4952,
"step": 185
},
{
"epoch": 0.2232,
"grad_norm": 0.941207184408332,
"learning_rate": 7.3705179282868534e-06,
"loss": 0.4754,
"step": 186
},
{
"epoch": 0.2244,
"grad_norm": 1.1183304118846102,
"learning_rate": 7.410358565737052e-06,
"loss": 0.4919,
"step": 187
},
{
"epoch": 0.2256,
"grad_norm": 0.9987437659822236,
"learning_rate": 7.450199203187252e-06,
"loss": 0.4436,
"step": 188
},
{
"epoch": 0.2268,
"grad_norm": 0.9241833200197873,
"learning_rate": 7.490039840637451e-06,
"loss": 0.5543,
"step": 189
},
{
"epoch": 0.228,
"grad_norm": 0.9374311693390385,
"learning_rate": 7.52988047808765e-06,
"loss": 0.4838,
"step": 190
},
{
"epoch": 0.2292,
"grad_norm": 0.8146043198808711,
"learning_rate": 7.569721115537849e-06,
"loss": 0.4625,
"step": 191
},
{
"epoch": 0.2304,
"grad_norm": 0.7656278828764158,
"learning_rate": 7.609561752988048e-06,
"loss": 0.4248,
"step": 192
},
{
"epoch": 0.2316,
"grad_norm": 1.128826007675778,
"learning_rate": 7.649402390438247e-06,
"loss": 0.5531,
"step": 193
},
{
"epoch": 0.2328,
"grad_norm": 0.9452753045477225,
"learning_rate": 7.689243027888446e-06,
"loss": 0.4604,
"step": 194
},
{
"epoch": 0.234,
"grad_norm": 0.8049486032505343,
"learning_rate": 7.729083665338646e-06,
"loss": 0.5026,
"step": 195
},
{
"epoch": 0.2352,
"grad_norm": 0.9586072562891299,
"learning_rate": 7.768924302788846e-06,
"loss": 0.4996,
"step": 196
},
{
"epoch": 0.2364,
"grad_norm": 1.1366388906466776,
"learning_rate": 7.808764940239044e-06,
"loss": 0.5105,
"step": 197
},
{
"epoch": 0.2376,
"grad_norm": 1.0121989369433997,
"learning_rate": 7.848605577689243e-06,
"loss": 0.5118,
"step": 198
},
{
"epoch": 0.2388,
"grad_norm": 0.9669380250949908,
"learning_rate": 7.888446215139443e-06,
"loss": 0.4475,
"step": 199
},
{
"epoch": 0.24,
"grad_norm": 1.0256407184793639,
"learning_rate": 7.928286852589641e-06,
"loss": 0.5351,
"step": 200
},
{
"epoch": 0.2412,
"grad_norm": 1.0529584162008458,
"learning_rate": 7.96812749003984e-06,
"loss": 0.473,
"step": 201
},
{
"epoch": 0.2424,
"grad_norm": 1.0438117017814197,
"learning_rate": 8.00796812749004e-06,
"loss": 0.4658,
"step": 202
},
{
"epoch": 0.2436,
"grad_norm": 1.0283344876022071,
"learning_rate": 8.04780876494024e-06,
"loss": 0.5109,
"step": 203
},
{
"epoch": 0.2448,
"grad_norm": 1.0231258744249236,
"learning_rate": 8.087649402390438e-06,
"loss": 0.5246,
"step": 204
},
{
"epoch": 0.246,
"grad_norm": 0.9756134167371139,
"learning_rate": 8.12749003984064e-06,
"loss": 0.4984,
"step": 205
},
{
"epoch": 0.2472,
"grad_norm": 1.0715343260283263,
"learning_rate": 8.167330677290837e-06,
"loss": 0.501,
"step": 206
},
{
"epoch": 0.2484,
"grad_norm": 1.086384368967051,
"learning_rate": 8.207171314741037e-06,
"loss": 0.4435,
"step": 207
},
{
"epoch": 0.2496,
"grad_norm": 1.1137719854778647,
"learning_rate": 8.247011952191236e-06,
"loss": 0.44,
"step": 208
},
{
"epoch": 0.2508,
"grad_norm": 0.9838363470254422,
"learning_rate": 8.286852589641436e-06,
"loss": 0.5166,
"step": 209
},
{
"epoch": 0.252,
"grad_norm": 0.9602419266621367,
"learning_rate": 8.326693227091634e-06,
"loss": 0.4741,
"step": 210
},
{
"epoch": 0.2532,
"grad_norm": 0.8395148133598535,
"learning_rate": 8.366533864541834e-06,
"loss": 0.454,
"step": 211
},
{
"epoch": 0.2544,
"grad_norm": 0.9164271806280002,
"learning_rate": 8.406374501992033e-06,
"loss": 0.4909,
"step": 212
},
{
"epoch": 0.2556,
"grad_norm": 0.9049844723019004,
"learning_rate": 8.446215139442231e-06,
"loss": 0.477,
"step": 213
},
{
"epoch": 0.2568,
"grad_norm": 0.8408694127731915,
"learning_rate": 8.486055776892431e-06,
"loss": 0.419,
"step": 214
},
{
"epoch": 0.258,
"grad_norm": 0.9766777397033436,
"learning_rate": 8.52589641434263e-06,
"loss": 0.4625,
"step": 215
},
{
"epoch": 0.2592,
"grad_norm": 1.1085535870973129,
"learning_rate": 8.56573705179283e-06,
"loss": 0.5384,
"step": 216
},
{
"epoch": 0.2604,
"grad_norm": 0.8160112134494882,
"learning_rate": 8.605577689243028e-06,
"loss": 0.4694,
"step": 217
},
{
"epoch": 0.2616,
"grad_norm": 1.1451243810549683,
"learning_rate": 8.645418326693228e-06,
"loss": 0.4807,
"step": 218
},
{
"epoch": 0.2628,
"grad_norm": 1.090428581808537,
"learning_rate": 8.685258964143428e-06,
"loss": 0.456,
"step": 219
},
{
"epoch": 0.264,
"grad_norm": 1.0154934939843443,
"learning_rate": 8.725099601593626e-06,
"loss": 0.5481,
"step": 220
},
{
"epoch": 0.2652,
"grad_norm": 0.8198534439829566,
"learning_rate": 8.764940239043825e-06,
"loss": 0.4453,
"step": 221
},
{
"epoch": 0.2664,
"grad_norm": 1.0676274447962364,
"learning_rate": 8.804780876494025e-06,
"loss": 0.4187,
"step": 222
},
{
"epoch": 0.2676,
"grad_norm": 1.130012075559157,
"learning_rate": 8.844621513944224e-06,
"loss": 0.4873,
"step": 223
},
{
"epoch": 0.2688,
"grad_norm": 1.1473958212079667,
"learning_rate": 8.884462151394422e-06,
"loss": 0.5696,
"step": 224
},
{
"epoch": 0.27,
"grad_norm": 0.9260191266124003,
"learning_rate": 8.924302788844622e-06,
"loss": 0.4382,
"step": 225
},
{
"epoch": 0.2712,
"grad_norm": 1.1693781431863157,
"learning_rate": 8.964143426294822e-06,
"loss": 0.5012,
"step": 226
},
{
"epoch": 0.2724,
"grad_norm": 1.1230201384797203,
"learning_rate": 9.00398406374502e-06,
"loss": 0.4727,
"step": 227
},
{
"epoch": 0.2736,
"grad_norm": 0.9472213827242388,
"learning_rate": 9.04382470119522e-06,
"loss": 0.4082,
"step": 228
},
{
"epoch": 0.2748,
"grad_norm": 1.044558161975557,
"learning_rate": 9.083665338645419e-06,
"loss": 0.4437,
"step": 229
},
{
"epoch": 0.276,
"grad_norm": 1.098130409085039,
"learning_rate": 9.123505976095619e-06,
"loss": 0.4173,
"step": 230
},
{
"epoch": 0.2772,
"grad_norm": 0.9302248173198375,
"learning_rate": 9.163346613545817e-06,
"loss": 0.4538,
"step": 231
},
{
"epoch": 0.2784,
"grad_norm": 0.9626649931300868,
"learning_rate": 9.203187250996016e-06,
"loss": 0.4418,
"step": 232
},
{
"epoch": 0.2796,
"grad_norm": 1.2184374842385355,
"learning_rate": 9.243027888446216e-06,
"loss": 0.4868,
"step": 233
},
{
"epoch": 0.2808,
"grad_norm": 0.7728172135214796,
"learning_rate": 9.282868525896414e-06,
"loss": 0.4569,
"step": 234
},
{
"epoch": 0.282,
"grad_norm": 0.9177873012496948,
"learning_rate": 9.322709163346614e-06,
"loss": 0.4575,
"step": 235
},
{
"epoch": 0.2832,
"grad_norm": 1.049520358219107,
"learning_rate": 9.362549800796813e-06,
"loss": 0.4618,
"step": 236
},
{
"epoch": 0.2844,
"grad_norm": 0.8285617950751487,
"learning_rate": 9.402390438247013e-06,
"loss": 0.4642,
"step": 237
},
{
"epoch": 0.2856,
"grad_norm": 1.183175992991948,
"learning_rate": 9.442231075697212e-06,
"loss": 0.458,
"step": 238
},
{
"epoch": 0.2868,
"grad_norm": 1.018409411700616,
"learning_rate": 9.482071713147412e-06,
"loss": 0.4819,
"step": 239
},
{
"epoch": 0.288,
"grad_norm": 0.8532545129981404,
"learning_rate": 9.52191235059761e-06,
"loss": 0.5176,
"step": 240
},
{
"epoch": 0.2892,
"grad_norm": 0.8404446731506587,
"learning_rate": 9.56175298804781e-06,
"loss": 0.462,
"step": 241
},
{
"epoch": 0.2904,
"grad_norm": 1.0358347911226142,
"learning_rate": 9.60159362549801e-06,
"loss": 0.4609,
"step": 242
},
{
"epoch": 0.2916,
"grad_norm": 1.0343957133634967,
"learning_rate": 9.641434262948209e-06,
"loss": 0.5185,
"step": 243
},
{
"epoch": 0.2928,
"grad_norm": 0.7748419181745317,
"learning_rate": 9.681274900398407e-06,
"loss": 0.4418,
"step": 244
},
{
"epoch": 0.294,
"grad_norm": 0.8460472875219072,
"learning_rate": 9.721115537848607e-06,
"loss": 0.4639,
"step": 245
},
{
"epoch": 0.2952,
"grad_norm": 1.027485373889621,
"learning_rate": 9.760956175298806e-06,
"loss": 0.4633,
"step": 246
},
{
"epoch": 0.2964,
"grad_norm": 0.900139823743846,
"learning_rate": 9.800796812749004e-06,
"loss": 0.4196,
"step": 247
},
{
"epoch": 0.2976,
"grad_norm": 0.8781151740392895,
"learning_rate": 9.840637450199204e-06,
"loss": 0.5465,
"step": 248
},
{
"epoch": 0.2988,
"grad_norm": 1.109368473283032,
"learning_rate": 9.880478087649404e-06,
"loss": 0.4387,
"step": 249
},
{
"epoch": 0.3,
"grad_norm": 1.0803696061688328,
"learning_rate": 9.920318725099603e-06,
"loss": 0.513,
"step": 250
},
{
"epoch": 0.3012,
"grad_norm": 0.9475983561137107,
"learning_rate": 9.960159362549801e-06,
"loss": 0.4625,
"step": 251
},
{
"epoch": 0.3024,
"grad_norm": 0.8298243967718946,
"learning_rate": 1e-05,
"loss": 0.47,
"step": 252
},
{
"epoch": 0.3036,
"grad_norm": 0.9802047998729689,
"learning_rate": 9.999995130451526e-06,
"loss": 0.481,
"step": 253
},
{
"epoch": 0.3048,
"grad_norm": 0.8355585972438275,
"learning_rate": 9.999980521815582e-06,
"loss": 0.4438,
"step": 254
},
{
"epoch": 0.306,
"grad_norm": 1.0489889975088451,
"learning_rate": 9.999956174120626e-06,
"loss": 0.5313,
"step": 255
},
{
"epoch": 0.3072,
"grad_norm": 1.1247332135696382,
"learning_rate": 9.999922087414084e-06,
"loss": 0.4788,
"step": 256
},
{
"epoch": 0.3084,
"grad_norm": 0.9819570634903716,
"learning_rate": 9.99987826176235e-06,
"loss": 0.4065,
"step": 257
},
{
"epoch": 0.3096,
"grad_norm": 1.0341154408969762,
"learning_rate": 9.999824697250786e-06,
"loss": 0.513,
"step": 258
},
{
"epoch": 0.3108,
"grad_norm": 0.9740682734444778,
"learning_rate": 9.999761393983728e-06,
"loss": 0.4716,
"step": 259
},
{
"epoch": 0.312,
"grad_norm": 1.1642335154157055,
"learning_rate": 9.999688352084482e-06,
"loss": 0.4723,
"step": 260
},
{
"epoch": 0.3132,
"grad_norm": 0.9069086441787787,
"learning_rate": 9.999605571695317e-06,
"loss": 0.4985,
"step": 261
},
{
"epoch": 0.3144,
"grad_norm": 0.962115569667121,
"learning_rate": 9.999513052977473e-06,
"loss": 0.4966,
"step": 262
},
{
"epoch": 0.3156,
"grad_norm": 0.9077231826429085,
"learning_rate": 9.999410796111163e-06,
"loss": 0.5006,
"step": 263
},
{
"epoch": 0.3168,
"grad_norm": 1.0157550623266656,
"learning_rate": 9.999298801295564e-06,
"loss": 0.5198,
"step": 264
},
{
"epoch": 0.318,
"grad_norm": 0.9350038143677769,
"learning_rate": 9.99917706874882e-06,
"loss": 0.4471,
"step": 265
},
{
"epoch": 0.3192,
"grad_norm": 0.8375521847483756,
"learning_rate": 9.999045598708047e-06,
"loss": 0.4351,
"step": 266
},
{
"epoch": 0.3204,
"grad_norm": 0.798580128924615,
"learning_rate": 9.998904391429323e-06,
"loss": 0.4237,
"step": 267
},
{
"epoch": 0.3216,
"grad_norm": 0.8832440965280086,
"learning_rate": 9.998753447187693e-06,
"loss": 0.4228,
"step": 268
},
{
"epoch": 0.3228,
"grad_norm": 0.8961034052704294,
"learning_rate": 9.998592766277173e-06,
"loss": 0.4767,
"step": 269
},
{
"epoch": 0.324,
"grad_norm": 0.8646844483264865,
"learning_rate": 9.998422349010736e-06,
"loss": 0.4986,
"step": 270
},
{
"epoch": 0.3252,
"grad_norm": 1.151118269437645,
"learning_rate": 9.998242195720327e-06,
"loss": 0.5016,
"step": 271
},
{
"epoch": 0.3264,
"grad_norm": 0.8915310112974539,
"learning_rate": 9.998052306756852e-06,
"loss": 0.4966,
"step": 272
},
{
"epoch": 0.3276,
"grad_norm": 0.8757554549566282,
"learning_rate": 9.997852682490179e-06,
"loss": 0.4508,
"step": 273
},
{
"epoch": 0.3288,
"grad_norm": 0.9442946080979946,
"learning_rate": 9.997643323309139e-06,
"loss": 0.4219,
"step": 274
},
{
"epoch": 0.33,
"grad_norm": 0.9413572016402217,
"learning_rate": 9.997424229621529e-06,
"loss": 0.4493,
"step": 275
},
{
"epoch": 0.3312,
"grad_norm": 0.8790862737060983,
"learning_rate": 9.997195401854102e-06,
"loss": 0.4771,
"step": 276
},
{
"epoch": 0.3324,
"grad_norm": 1.0268975875458013,
"learning_rate": 9.996956840452573e-06,
"loss": 0.5493,
"step": 277
},
{
"epoch": 0.3336,
"grad_norm": 0.8832942779366042,
"learning_rate": 9.996708545881617e-06,
"loss": 0.4761,
"step": 278
},
{
"epoch": 0.3348,
"grad_norm": 0.9912645329934423,
"learning_rate": 9.996450518624868e-06,
"loss": 0.4701,
"step": 279
},
{
"epoch": 0.336,
"grad_norm": 1.0085936157365127,
"learning_rate": 9.996182759184916e-06,
"loss": 0.4741,
"step": 280
},
{
"epoch": 0.3372,
"grad_norm": 1.013422863495019,
"learning_rate": 9.995905268083306e-06,
"loss": 0.4988,
"step": 281
},
{
"epoch": 0.3384,
"grad_norm": 0.866527586917797,
"learning_rate": 9.995618045860545e-06,
"loss": 0.4562,
"step": 282
},
{
"epoch": 0.3396,
"grad_norm": 0.9700582331668275,
"learning_rate": 9.995321093076085e-06,
"loss": 0.4336,
"step": 283
},
{
"epoch": 0.3408,
"grad_norm": 0.8698427456680538,
"learning_rate": 9.995014410308336e-06,
"loss": 0.4126,
"step": 284
},
{
"epoch": 0.342,
"grad_norm": 0.9534346497836448,
"learning_rate": 9.994697998154668e-06,
"loss": 0.504,
"step": 285
},
{
"epoch": 0.3432,
"grad_norm": 0.9646900677345498,
"learning_rate": 9.994371857231388e-06,
"loss": 0.4582,
"step": 286
},
{
"epoch": 0.3444,
"grad_norm": 0.9223840792356871,
"learning_rate": 9.99403598817376e-06,
"loss": 0.5262,
"step": 287
},
{
"epoch": 0.3456,
"grad_norm": 1.0081493455502901,
"learning_rate": 9.993690391636e-06,
"loss": 0.4813,
"step": 288
},
{
"epoch": 0.3468,
"grad_norm": 0.9992409004966168,
"learning_rate": 9.993335068291264e-06,
"loss": 0.515,
"step": 289
},
{
"epoch": 0.348,
"grad_norm": 0.8462474144190385,
"learning_rate": 9.99297001883166e-06,
"loss": 0.4404,
"step": 290
},
{
"epoch": 0.3492,
"grad_norm": 0.8616449462761342,
"learning_rate": 9.992595243968238e-06,
"loss": 0.4786,
"step": 291
},
{
"epoch": 0.3504,
"grad_norm": 0.9471837465631915,
"learning_rate": 9.99221074443099e-06,
"loss": 0.479,
"step": 292
},
{
"epoch": 0.3516,
"grad_norm": 0.8743700744526802,
"learning_rate": 9.991816520968853e-06,
"loss": 0.4454,
"step": 293
},
{
"epoch": 0.3528,
"grad_norm": 0.7676446635862832,
"learning_rate": 9.991412574349704e-06,
"loss": 0.4532,
"step": 294
},
{
"epoch": 0.354,
"grad_norm": 1.0201505599804612,
"learning_rate": 9.990998905360357e-06,
"loss": 0.4913,
"step": 295
},
{
"epoch": 0.3552,
"grad_norm": 1.0576326058460073,
"learning_rate": 9.990575514806563e-06,
"loss": 0.4885,
"step": 296
},
{
"epoch": 0.3564,
"grad_norm": 1.031107624397378,
"learning_rate": 9.990142403513012e-06,
"loss": 0.4937,
"step": 297
},
{
"epoch": 0.3576,
"grad_norm": 0.8897862962894296,
"learning_rate": 9.989699572323328e-06,
"loss": 0.4664,
"step": 298
},
{
"epoch": 0.3588,
"grad_norm": 1.0773473735035723,
"learning_rate": 9.989247022100065e-06,
"loss": 0.4415,
"step": 299
},
{
"epoch": 0.36,
"grad_norm": 1.1848180508681543,
"learning_rate": 9.988784753724707e-06,
"loss": 0.5319,
"step": 300
},
{
"epoch": 0.3612,
"grad_norm": 0.9093930543027307,
"learning_rate": 9.988312768097673e-06,
"loss": 0.4317,
"step": 301
},
{
"epoch": 0.3624,
"grad_norm": 0.9963815159424825,
"learning_rate": 9.987831066138302e-06,
"loss": 0.4583,
"step": 302
},
{
"epoch": 0.3636,
"grad_norm": 1.2085638960635496,
"learning_rate": 9.987339648784866e-06,
"loss": 0.45,
"step": 303
},
{
"epoch": 0.3648,
"grad_norm": 0.9578529266803061,
"learning_rate": 9.986838516994555e-06,
"loss": 0.5038,
"step": 304
},
{
"epoch": 0.366,
"grad_norm": 0.8577004084077766,
"learning_rate": 9.986327671743484e-06,
"loss": 0.5205,
"step": 305
},
{
"epoch": 0.3672,
"grad_norm": 1.0237114741259223,
"learning_rate": 9.985807114026684e-06,
"loss": 0.4508,
"step": 306
},
{
"epoch": 0.3684,
"grad_norm": 0.8740436491455652,
"learning_rate": 9.985276844858114e-06,
"loss": 0.4672,
"step": 307
},
{
"epoch": 0.3696,
"grad_norm": 0.8660842498383028,
"learning_rate": 9.984736865270637e-06,
"loss": 0.4901,
"step": 308
},
{
"epoch": 0.3708,
"grad_norm": 0.9469961741141615,
"learning_rate": 9.984187176316038e-06,
"loss": 0.4913,
"step": 309
},
{
"epoch": 0.372,
"grad_norm": 0.9657654476424753,
"learning_rate": 9.983627779065012e-06,
"loss": 0.4727,
"step": 310
},
{
"epoch": 0.3732,
"grad_norm": 0.839270519941747,
"learning_rate": 9.983058674607164e-06,
"loss": 0.4057,
"step": 311
},
{
"epoch": 0.3744,
"grad_norm": 0.9053595463161098,
"learning_rate": 9.982479864051005e-06,
"loss": 0.443,
"step": 312
},
{
"epoch": 0.3756,
"grad_norm": 0.7689233396635476,
"learning_rate": 9.981891348523955e-06,
"loss": 0.4215,
"step": 313
},
{
"epoch": 0.3768,
"grad_norm": 0.9515452931016123,
"learning_rate": 9.981293129172334e-06,
"loss": 0.5077,
"step": 314
},
{
"epoch": 0.378,
"grad_norm": 0.8607767291168189,
"learning_rate": 9.980685207161368e-06,
"loss": 0.4155,
"step": 315
},
{
"epoch": 0.3792,
"grad_norm": 0.8413932910410457,
"learning_rate": 9.980067583675177e-06,
"loss": 0.4455,
"step": 316
},
{
"epoch": 0.3804,
"grad_norm": 0.965276634022141,
"learning_rate": 9.979440259916782e-06,
"loss": 0.4834,
"step": 317
},
{
"epoch": 0.3816,
"grad_norm": 0.7046459043630207,
"learning_rate": 9.978803237108095e-06,
"loss": 0.4379,
"step": 318
},
{
"epoch": 0.3828,
"grad_norm": 0.8618286224443213,
"learning_rate": 9.97815651648992e-06,
"loss": 0.4576,
"step": 319
},
{
"epoch": 0.384,
"grad_norm": 0.9109576569890303,
"learning_rate": 9.977500099321956e-06,
"loss": 0.4517,
"step": 320
},
{
"epoch": 0.3852,
"grad_norm": 0.9474873810835672,
"learning_rate": 9.97683398688278e-06,
"loss": 0.5149,
"step": 321
},
{
"epoch": 0.3864,
"grad_norm": 0.7624043208484511,
"learning_rate": 9.976158180469866e-06,
"loss": 0.4487,
"step": 322
},
{
"epoch": 0.3876,
"grad_norm": 0.8460362849367942,
"learning_rate": 9.975472681399556e-06,
"loss": 0.4713,
"step": 323
},
{
"epoch": 0.3888,
"grad_norm": 1.038112445901252,
"learning_rate": 9.97477749100708e-06,
"loss": 0.5146,
"step": 324
},
{
"epoch": 0.39,
"grad_norm": 1.0190921318705397,
"learning_rate": 9.974072610646543e-06,
"loss": 0.586,
"step": 325
},
{
"epoch": 0.3912,
"grad_norm": 1.0475822675448425,
"learning_rate": 9.973358041690926e-06,
"loss": 0.4653,
"step": 326
},
{
"epoch": 0.3924,
"grad_norm": 0.8532726396811824,
"learning_rate": 9.972633785532082e-06,
"loss": 0.4938,
"step": 327
},
{
"epoch": 0.3936,
"grad_norm": 1.1211294020539635,
"learning_rate": 9.971899843580728e-06,
"loss": 0.5093,
"step": 328
},
{
"epoch": 0.3948,
"grad_norm": 1.1363015372007326,
"learning_rate": 9.971156217266451e-06,
"loss": 0.5422,
"step": 329
},
{
"epoch": 0.396,
"grad_norm": 1.000798629239088,
"learning_rate": 9.970402908037703e-06,
"loss": 0.4696,
"step": 330
},
{
"epoch": 0.3972,
"grad_norm": 0.78748028117645,
"learning_rate": 9.96963991736179e-06,
"loss": 0.4869,
"step": 331
},
{
"epoch": 0.3984,
"grad_norm": 0.942485231080613,
"learning_rate": 9.968867246724882e-06,
"loss": 0.5079,
"step": 332
},
{
"epoch": 0.3996,
"grad_norm": 1.0287551298226354,
"learning_rate": 9.968084897632004e-06,
"loss": 0.5142,
"step": 333
},
{
"epoch": 0.4008,
"grad_norm": 0.9872779665108361,
"learning_rate": 9.96729287160703e-06,
"loss": 0.4374,
"step": 334
},
{
"epoch": 0.402,
"grad_norm": 1.1624783271341648,
"learning_rate": 9.966491170192682e-06,
"loss": 0.5254,
"step": 335
},
{
"epoch": 0.4032,
"grad_norm": 0.8365759849717022,
"learning_rate": 9.96567979495053e-06,
"loss": 0.4567,
"step": 336
},
{
"epoch": 0.4044,
"grad_norm": 1.0653040783134162,
"learning_rate": 9.964858747460989e-06,
"loss": 0.5021,
"step": 337
},
{
"epoch": 0.4056,
"grad_norm": 0.8795827347981251,
"learning_rate": 9.964028029323305e-06,
"loss": 0.4669,
"step": 338
},
{
"epoch": 0.4068,
"grad_norm": 0.9764081723632619,
"learning_rate": 9.963187642155573e-06,
"loss": 0.458,
"step": 339
},
{
"epoch": 0.408,
"grad_norm": 1.268512596914498,
"learning_rate": 9.962337587594713e-06,
"loss": 0.4769,
"step": 340
},
{
"epoch": 0.4092,
"grad_norm": 0.8797506892824893,
"learning_rate": 9.961477867296479e-06,
"loss": 0.496,
"step": 341
},
{
"epoch": 0.4104,
"grad_norm": 0.8137373777345258,
"learning_rate": 9.96060848293545e-06,
"loss": 0.4039,
"step": 342
},
{
"epoch": 0.4116,
"grad_norm": 0.8480878578413085,
"learning_rate": 9.959729436205027e-06,
"loss": 0.4683,
"step": 343
},
{
"epoch": 0.4128,
"grad_norm": 0.778038446541413,
"learning_rate": 9.95884072881744e-06,
"loss": 0.4536,
"step": 344
},
{
"epoch": 0.414,
"grad_norm": 0.9490708765568948,
"learning_rate": 9.957942362503728e-06,
"loss": 0.4998,
"step": 345
},
{
"epoch": 0.4152,
"grad_norm": 0.8158835780779223,
"learning_rate": 9.957034339013742e-06,
"loss": 0.4078,
"step": 346
},
{
"epoch": 0.4164,
"grad_norm": 0.8623210709547612,
"learning_rate": 9.956116660116155e-06,
"loss": 0.4198,
"step": 347
},
{
"epoch": 0.4176,
"grad_norm": 0.9056186505862733,
"learning_rate": 9.955189327598435e-06,
"loss": 0.4872,
"step": 348
},
{
"epoch": 0.4188,
"grad_norm": 0.842893478860219,
"learning_rate": 9.954252343266859e-06,
"loss": 0.4617,
"step": 349
},
{
"epoch": 0.42,
"grad_norm": 0.7995733969846445,
"learning_rate": 9.953305708946504e-06,
"loss": 0.4307,
"step": 350
},
{
"epoch": 0.4212,
"grad_norm": 0.9984313943413776,
"learning_rate": 9.952349426481243e-06,
"loss": 0.4566,
"step": 351
},
{
"epoch": 0.4224,
"grad_norm": 0.8079192183166884,
"learning_rate": 9.95138349773374e-06,
"loss": 0.4305,
"step": 352
},
{
"epoch": 0.4236,
"grad_norm": 0.8491639695660568,
"learning_rate": 9.95040792458545e-06,
"loss": 0.4398,
"step": 353
},
{
"epoch": 0.4248,
"grad_norm": 0.8314620677825669,
"learning_rate": 9.949422708936616e-06,
"loss": 0.4085,
"step": 354
},
{
"epoch": 0.426,
"grad_norm": 0.9507807612036341,
"learning_rate": 9.948427852706257e-06,
"loss": 0.4441,
"step": 355
},
{
"epoch": 0.4272,
"grad_norm": 0.9136961892823259,
"learning_rate": 9.947423357832176e-06,
"loss": 0.5081,
"step": 356
},
{
"epoch": 0.4284,
"grad_norm": 0.7700820238121457,
"learning_rate": 9.946409226270945e-06,
"loss": 0.4526,
"step": 357
},
{
"epoch": 0.4296,
"grad_norm": 0.7859423040833289,
"learning_rate": 9.945385459997909e-06,
"loss": 0.4677,
"step": 358
},
{
"epoch": 0.4308,
"grad_norm": 0.7835528051143813,
"learning_rate": 9.944352061007182e-06,
"loss": 0.4151,
"step": 359
},
{
"epoch": 0.432,
"grad_norm": 0.864416474334718,
"learning_rate": 9.943309031311637e-06,
"loss": 0.4872,
"step": 360
},
{
"epoch": 0.4332,
"grad_norm": 0.7686942846185059,
"learning_rate": 9.942256372942909e-06,
"loss": 0.4551,
"step": 361
},
{
"epoch": 0.4344,
"grad_norm": 0.8694734977140924,
"learning_rate": 9.941194087951384e-06,
"loss": 0.4449,
"step": 362
},
{
"epoch": 0.4356,
"grad_norm": 0.8104097029892385,
"learning_rate": 9.940122178406205e-06,
"loss": 0.4673,
"step": 363
},
{
"epoch": 0.4368,
"grad_norm": 0.8707924072369511,
"learning_rate": 9.939040646395252e-06,
"loss": 0.4532,
"step": 364
},
{
"epoch": 0.438,
"grad_norm": 0.8595081525636566,
"learning_rate": 9.93794949402516e-06,
"loss": 0.4201,
"step": 365
},
{
"epoch": 0.4392,
"grad_norm": 0.9806070577367421,
"learning_rate": 9.936848723421295e-06,
"loss": 0.4721,
"step": 366
},
{
"epoch": 0.4404,
"grad_norm": 0.758164324102052,
"learning_rate": 9.93573833672776e-06,
"loss": 0.4517,
"step": 367
},
{
"epoch": 0.4416,
"grad_norm": 0.90886061761394,
"learning_rate": 9.934618336107385e-06,
"loss": 0.4884,
"step": 368
},
{
"epoch": 0.4428,
"grad_norm": 0.9885712038297171,
"learning_rate": 9.933488723741731e-06,
"loss": 0.4177,
"step": 369
},
{
"epoch": 0.444,
"grad_norm": 0.9636121279366031,
"learning_rate": 9.932349501831077e-06,
"loss": 0.4701,
"step": 370
},
{
"epoch": 0.4452,
"grad_norm": 1.2408149979719083,
"learning_rate": 9.931200672594425e-06,
"loss": 0.5063,
"step": 371
},
{
"epoch": 0.4464,
"grad_norm": 0.9545959723028793,
"learning_rate": 9.930042238269485e-06,
"loss": 0.5072,
"step": 372
},
{
"epoch": 0.4476,
"grad_norm": 0.9336132868520455,
"learning_rate": 9.928874201112677e-06,
"loss": 0.4912,
"step": 373
},
{
"epoch": 0.4488,
"grad_norm": 0.7955564966751042,
"learning_rate": 9.927696563399127e-06,
"loss": 0.4715,
"step": 374
},
{
"epoch": 0.45,
"grad_norm": 0.8018658923789644,
"learning_rate": 9.926509327422661e-06,
"loss": 0.4285,
"step": 375
},
{
"epoch": 0.4512,
"grad_norm": 1.0288453066662198,
"learning_rate": 9.9253124954958e-06,
"loss": 0.4925,
"step": 376
},
{
"epoch": 0.4524,
"grad_norm": 1.0681890627681496,
"learning_rate": 9.924106069949756e-06,
"loss": 0.4843,
"step": 377
},
{
"epoch": 0.4536,
"grad_norm": 0.7320792725923697,
"learning_rate": 9.922890053134428e-06,
"loss": 0.4583,
"step": 378
},
{
"epoch": 0.4548,
"grad_norm": 0.8379947920124755,
"learning_rate": 9.9216644474184e-06,
"loss": 0.4513,
"step": 379
},
{
"epoch": 0.456,
"grad_norm": 0.8715853024137559,
"learning_rate": 9.920429255188926e-06,
"loss": 0.4586,
"step": 380
},
{
"epoch": 0.4572,
"grad_norm": 1.2622393420710876,
"learning_rate": 9.91918447885194e-06,
"loss": 0.4977,
"step": 381
},
{
"epoch": 0.4584,
"grad_norm": 0.8692479203153781,
"learning_rate": 9.91793012083204e-06,
"loss": 0.5141,
"step": 382
},
{
"epoch": 0.4596,
"grad_norm": 0.8599555052373915,
"learning_rate": 9.916666183572492e-06,
"loss": 0.4912,
"step": 383
},
{
"epoch": 0.4608,
"grad_norm": 0.9778615164218931,
"learning_rate": 9.915392669535214e-06,
"loss": 0.5339,
"step": 384
},
{
"epoch": 0.462,
"grad_norm": 0.7795255502684801,
"learning_rate": 9.914109581200785e-06,
"loss": 0.4339,
"step": 385
},
{
"epoch": 0.4632,
"grad_norm": 0.9825211053839761,
"learning_rate": 9.912816921068424e-06,
"loss": 0.439,
"step": 386
},
{
"epoch": 0.4644,
"grad_norm": 0.8211846175748321,
"learning_rate": 9.911514691656003e-06,
"loss": 0.4747,
"step": 387
},
{
"epoch": 0.4656,
"grad_norm": 0.7878975021728168,
"learning_rate": 9.910202895500031e-06,
"loss": 0.5001,
"step": 388
},
{
"epoch": 0.4668,
"grad_norm": 0.8527301678481147,
"learning_rate": 9.908881535155647e-06,
"loss": 0.404,
"step": 389
},
{
"epoch": 0.468,
"grad_norm": 0.9360561585687912,
"learning_rate": 9.907550613196624e-06,
"loss": 0.4959,
"step": 390
},
{
"epoch": 0.4692,
"grad_norm": 0.9367957286530252,
"learning_rate": 9.906210132215357e-06,
"loss": 0.5093,
"step": 391
},
{
"epoch": 0.4704,
"grad_norm": 0.8728936475093994,
"learning_rate": 9.904860094822861e-06,
"loss": 0.5376,
"step": 392
},
{
"epoch": 0.4716,
"grad_norm": 0.9550595296430375,
"learning_rate": 9.903500503648766e-06,
"loss": 0.4713,
"step": 393
},
{
"epoch": 0.4728,
"grad_norm": 0.7710994745300122,
"learning_rate": 9.902131361341307e-06,
"loss": 0.4836,
"step": 394
},
{
"epoch": 0.474,
"grad_norm": 0.896162425259607,
"learning_rate": 9.900752670567331e-06,
"loss": 0.5204,
"step": 395
},
{
"epoch": 0.4752,
"grad_norm": 0.8320678318150154,
"learning_rate": 9.899364434012273e-06,
"loss": 0.4621,
"step": 396
},
{
"epoch": 0.4764,
"grad_norm": 0.9272571324804332,
"learning_rate": 9.897966654380172e-06,
"loss": 0.3989,
"step": 397
},
{
"epoch": 0.4776,
"grad_norm": 0.9174421636995335,
"learning_rate": 9.896559334393644e-06,
"loss": 0.5076,
"step": 398
},
{
"epoch": 0.4788,
"grad_norm": 0.8496410754928007,
"learning_rate": 9.895142476793902e-06,
"loss": 0.4391,
"step": 399
},
{
"epoch": 0.48,
"grad_norm": 0.8391425755500477,
"learning_rate": 9.893716084340723e-06,
"loss": 0.3964,
"step": 400
},
{
"epoch": 0.4812,
"grad_norm": 0.7638166928680227,
"learning_rate": 9.892280159812465e-06,
"loss": 0.4944,
"step": 401
},
{
"epoch": 0.4824,
"grad_norm": 0.7795228066516753,
"learning_rate": 9.890834706006048e-06,
"loss": 0.4353,
"step": 402
},
{
"epoch": 0.4836,
"grad_norm": 0.8208167346023207,
"learning_rate": 9.889379725736953e-06,
"loss": 0.4352,
"step": 403
},
{
"epoch": 0.4848,
"grad_norm": 0.7666377006541595,
"learning_rate": 9.887915221839223e-06,
"loss": 0.4906,
"step": 404
},
{
"epoch": 0.486,
"grad_norm": 0.8301640360717246,
"learning_rate": 9.886441197165446e-06,
"loss": 0.3794,
"step": 405
},
{
"epoch": 0.4872,
"grad_norm": 1.0321521606215063,
"learning_rate": 9.884957654586753e-06,
"loss": 0.4361,
"step": 406
},
{
"epoch": 0.4884,
"grad_norm": 0.8930164424269926,
"learning_rate": 9.88346459699282e-06,
"loss": 0.519,
"step": 407
},
{
"epoch": 0.4896,
"grad_norm": 0.8501569310548959,
"learning_rate": 9.881962027291855e-06,
"loss": 0.4789,
"step": 408
},
{
"epoch": 0.4908,
"grad_norm": 1.049714574811071,
"learning_rate": 9.880449948410587e-06,
"loss": 0.4879,
"step": 409
},
{
"epoch": 0.492,
"grad_norm": 0.9659215863167773,
"learning_rate": 9.878928363294275e-06,
"loss": 0.471,
"step": 410
},
{
"epoch": 0.4932,
"grad_norm": 0.9105799377075012,
"learning_rate": 9.877397274906694e-06,
"loss": 0.4928,
"step": 411
},
{
"epoch": 0.4944,
"grad_norm": 0.8481837747385731,
"learning_rate": 9.875856686230125e-06,
"loss": 0.4543,
"step": 412
},
{
"epoch": 0.4956,
"grad_norm": 0.964977127374614,
"learning_rate": 9.87430660026536e-06,
"loss": 0.4682,
"step": 413
},
{
"epoch": 0.4968,
"grad_norm": 1.092841281098979,
"learning_rate": 9.872747020031682e-06,
"loss": 0.5114,
"step": 414
},
{
"epoch": 0.498,
"grad_norm": 0.8001988945952248,
"learning_rate": 9.871177948566875e-06,
"loss": 0.5291,
"step": 415
},
{
"epoch": 0.4992,
"grad_norm": 0.9381856642234143,
"learning_rate": 9.869599388927204e-06,
"loss": 0.4498,
"step": 416
},
{
"epoch": 0.5004,
"grad_norm": 0.7828682972558573,
"learning_rate": 9.868011344187421e-06,
"loss": 0.4302,
"step": 417
},
{
"epoch": 0.5016,
"grad_norm": 0.8743072078786475,
"learning_rate": 9.866413817440748e-06,
"loss": 0.4983,
"step": 418
},
{
"epoch": 0.5028,
"grad_norm": 1.0403618866420057,
"learning_rate": 9.864806811798881e-06,
"loss": 0.5031,
"step": 419
},
{
"epoch": 0.504,
"grad_norm": 1.1308159263900466,
"learning_rate": 9.863190330391974e-06,
"loss": 0.5263,
"step": 420
},
{
"epoch": 0.5052,
"grad_norm": 1.0871377233410793,
"learning_rate": 9.861564376368645e-06,
"loss": 0.4771,
"step": 421
},
{
"epoch": 0.5064,
"grad_norm": 0.8758458921411397,
"learning_rate": 9.859928952895952e-06,
"loss": 0.5038,
"step": 422
},
{
"epoch": 0.5076,
"grad_norm": 0.8803012622675811,
"learning_rate": 9.858284063159411e-06,
"loss": 0.3888,
"step": 423
},
{
"epoch": 0.5088,
"grad_norm": 0.924024372993349,
"learning_rate": 9.856629710362966e-06,
"loss": 0.4793,
"step": 424
},
{
"epoch": 0.51,
"grad_norm": 0.894217409582383,
"learning_rate": 9.854965897729001e-06,
"loss": 0.4719,
"step": 425
},
{
"epoch": 0.5112,
"grad_norm": 0.9132775120417949,
"learning_rate": 9.853292628498319e-06,
"loss": 0.4813,
"step": 426
},
{
"epoch": 0.5124,
"grad_norm": 0.6668420908679299,
"learning_rate": 9.851609905930149e-06,
"loss": 0.4514,
"step": 427
},
{
"epoch": 0.5136,
"grad_norm": 0.7646476159587744,
"learning_rate": 9.849917733302128e-06,
"loss": 0.4603,
"step": 428
},
{
"epoch": 0.5148,
"grad_norm": 1.3275635093426787,
"learning_rate": 9.848216113910306e-06,
"loss": 0.4934,
"step": 429
},
{
"epoch": 0.516,
"grad_norm": 0.9145885078128915,
"learning_rate": 9.846505051069126e-06,
"loss": 0.5174,
"step": 430
},
{
"epoch": 0.5172,
"grad_norm": 0.8726124444086926,
"learning_rate": 9.844784548111433e-06,
"loss": 0.4327,
"step": 431
},
{
"epoch": 0.5184,
"grad_norm": 0.7882521012863235,
"learning_rate": 9.843054608388455e-06,
"loss": 0.4865,
"step": 432
},
{
"epoch": 0.5196,
"grad_norm": 0.8764732506443571,
"learning_rate": 9.8413152352698e-06,
"loss": 0.5382,
"step": 433
},
{
"epoch": 0.5208,
"grad_norm": 0.7120641796177193,
"learning_rate": 9.839566432143459e-06,
"loss": 0.5142,
"step": 434
},
{
"epoch": 0.522,
"grad_norm": 0.9620462760463923,
"learning_rate": 9.837808202415778e-06,
"loss": 0.4487,
"step": 435
},
{
"epoch": 0.5232,
"grad_norm": 0.8980048407973095,
"learning_rate": 9.836040549511472e-06,
"loss": 0.4178,
"step": 436
},
{
"epoch": 0.5244,
"grad_norm": 0.954246833457193,
"learning_rate": 9.83426347687361e-06,
"loss": 0.4439,
"step": 437
},
{
"epoch": 0.5256,
"grad_norm": 0.8536424327469211,
"learning_rate": 9.832476987963613e-06,
"loss": 0.4424,
"step": 438
},
{
"epoch": 0.5268,
"grad_norm": 0.8768458563852505,
"learning_rate": 9.830681086261234e-06,
"loss": 0.4765,
"step": 439
},
{
"epoch": 0.528,
"grad_norm": 0.7566866786615243,
"learning_rate": 9.828875775264564e-06,
"loss": 0.4953,
"step": 440
},
{
"epoch": 0.5292,
"grad_norm": 0.8927572846850163,
"learning_rate": 9.827061058490027e-06,
"loss": 0.4739,
"step": 441
},
{
"epoch": 0.5304,
"grad_norm": 0.8382316002314895,
"learning_rate": 9.82523693947236e-06,
"loss": 0.4542,
"step": 442
},
{
"epoch": 0.5316,
"grad_norm": 0.8503050287875985,
"learning_rate": 9.82340342176462e-06,
"loss": 0.4792,
"step": 443
},
{
"epoch": 0.5328,
"grad_norm": 0.9501603464890747,
"learning_rate": 9.821560508938167e-06,
"loss": 0.4827,
"step": 444
},
{
"epoch": 0.534,
"grad_norm": 0.7638285486053017,
"learning_rate": 9.819708204582664e-06,
"loss": 0.4433,
"step": 445
},
{
"epoch": 0.5352,
"grad_norm": 0.8634957469463319,
"learning_rate": 9.817846512306062e-06,
"loss": 0.4353,
"step": 446
},
{
"epoch": 0.5364,
"grad_norm": 1.0055531567280331,
"learning_rate": 9.815975435734604e-06,
"loss": 0.4321,
"step": 447
},
{
"epoch": 0.5376,
"grad_norm": 0.7705572831856009,
"learning_rate": 9.814094978512808e-06,
"loss": 0.4405,
"step": 448
},
{
"epoch": 0.5388,
"grad_norm": 0.7652507059142534,
"learning_rate": 9.812205144303466e-06,
"loss": 0.4501,
"step": 449
},
{
"epoch": 0.54,
"grad_norm": 0.7730287600694025,
"learning_rate": 9.810305936787633e-06,
"loss": 0.4775,
"step": 450
},
{
"epoch": 0.5412,
"grad_norm": 1.0573732911792568,
"learning_rate": 9.808397359664624e-06,
"loss": 0.5107,
"step": 451
},
{
"epoch": 0.5424,
"grad_norm": 1.137577052687185,
"learning_rate": 9.806479416652e-06,
"loss": 0.531,
"step": 452
},
{
"epoch": 0.5436,
"grad_norm": 0.8443248475212453,
"learning_rate": 9.804552111485568e-06,
"loss": 0.4604,
"step": 453
},
{
"epoch": 0.5448,
"grad_norm": 0.9230650826630182,
"learning_rate": 9.80261544791937e-06,
"loss": 0.4162,
"step": 454
},
{
"epoch": 0.546,
"grad_norm": 0.8732669712397506,
"learning_rate": 9.80066942972568e-06,
"loss": 0.4923,
"step": 455
},
{
"epoch": 0.5472,
"grad_norm": 0.7815058647852078,
"learning_rate": 9.798714060694988e-06,
"loss": 0.4532,
"step": 456
},
{
"epoch": 0.5484,
"grad_norm": 0.8376369606217267,
"learning_rate": 9.796749344635996e-06,
"loss": 0.5454,
"step": 457
},
{
"epoch": 0.5496,
"grad_norm": 0.8056415899178225,
"learning_rate": 9.794775285375623e-06,
"loss": 0.4936,
"step": 458
},
{
"epoch": 0.5508,
"grad_norm": 0.7640085447749293,
"learning_rate": 9.792791886758976e-06,
"loss": 0.4985,
"step": 459
},
{
"epoch": 0.552,
"grad_norm": 0.8804025086503511,
"learning_rate": 9.790799152649356e-06,
"loss": 0.4478,
"step": 460
},
{
"epoch": 0.5532,
"grad_norm": 1.0614572700644085,
"learning_rate": 9.788797086928252e-06,
"loss": 0.5212,
"step": 461
},
{
"epoch": 0.5544,
"grad_norm": 0.8158317114744413,
"learning_rate": 9.786785693495327e-06,
"loss": 0.4627,
"step": 462
},
{
"epoch": 0.5556,
"grad_norm": 0.9054451205358116,
"learning_rate": 9.784764976268408e-06,
"loss": 0.4336,
"step": 463
},
{
"epoch": 0.5568,
"grad_norm": 0.7703709320981571,
"learning_rate": 9.78273493918349e-06,
"loss": 0.4552,
"step": 464
},
{
"epoch": 0.558,
"grad_norm": 0.7854298319990675,
"learning_rate": 9.780695586194719e-06,
"loss": 0.4111,
"step": 465
},
{
"epoch": 0.5592,
"grad_norm": 0.8678198196581406,
"learning_rate": 9.778646921274385e-06,
"loss": 0.5151,
"step": 466
},
{
"epoch": 0.5604,
"grad_norm": 1.051192074400555,
"learning_rate": 9.776588948412917e-06,
"loss": 0.4649,
"step": 467
},
{
"epoch": 0.5616,
"grad_norm": 0.8160374421028391,
"learning_rate": 9.774521671618877e-06,
"loss": 0.4223,
"step": 468
},
{
"epoch": 0.5628,
"grad_norm": 0.753218682967093,
"learning_rate": 9.772445094918944e-06,
"loss": 0.4681,
"step": 469
},
{
"epoch": 0.564,
"grad_norm": 1.008099051320608,
"learning_rate": 9.770359222357914e-06,
"loss": 0.4884,
"step": 470
},
{
"epoch": 0.5652,
"grad_norm": 1.0866946653854257,
"learning_rate": 9.768264057998693e-06,
"loss": 0.5049,
"step": 471
},
{
"epoch": 0.5664,
"grad_norm": 0.7865059482552546,
"learning_rate": 9.766159605922282e-06,
"loss": 0.463,
"step": 472
},
{
"epoch": 0.5676,
"grad_norm": 0.8768201396345472,
"learning_rate": 9.764045870227772e-06,
"loss": 0.4694,
"step": 473
},
{
"epoch": 0.5688,
"grad_norm": 0.8701093274536077,
"learning_rate": 9.761922855032339e-06,
"loss": 0.4522,
"step": 474
},
{
"epoch": 0.57,
"grad_norm": 0.905099561930582,
"learning_rate": 9.759790564471233e-06,
"loss": 0.5436,
"step": 475
},
{
"epoch": 0.5712,
"grad_norm": 0.6658167181559317,
"learning_rate": 9.757649002697771e-06,
"loss": 0.465,
"step": 476
},
{
"epoch": 0.5724,
"grad_norm": 0.7561842827343221,
"learning_rate": 9.755498173883331e-06,
"loss": 0.4355,
"step": 477
},
{
"epoch": 0.5736,
"grad_norm": 0.6981271326060572,
"learning_rate": 9.753338082217334e-06,
"loss": 0.4115,
"step": 478
},
{
"epoch": 0.5748,
"grad_norm": 1.2049605632105427,
"learning_rate": 9.751168731907253e-06,
"loss": 0.4356,
"step": 479
},
{
"epoch": 0.576,
"grad_norm": 0.7706199009101948,
"learning_rate": 9.748990127178589e-06,
"loss": 0.4028,
"step": 480
},
{
"epoch": 0.5772,
"grad_norm": 0.8005815506782505,
"learning_rate": 9.746802272274868e-06,
"loss": 0.3993,
"step": 481
},
{
"epoch": 0.5784,
"grad_norm": 0.8289466513914655,
"learning_rate": 9.74460517145764e-06,
"loss": 0.3998,
"step": 482
},
{
"epoch": 0.5796,
"grad_norm": 0.8323901396285354,
"learning_rate": 9.74239882900646e-06,
"loss": 0.4585,
"step": 483
},
{
"epoch": 0.5808,
"grad_norm": 0.8815737098727866,
"learning_rate": 9.740183249218883e-06,
"loss": 0.4598,
"step": 484
},
{
"epoch": 0.582,
"grad_norm": 0.9638739296811456,
"learning_rate": 9.737958436410459e-06,
"loss": 0.4282,
"step": 485
},
{
"epoch": 0.5832,
"grad_norm": 0.798472826540791,
"learning_rate": 9.73572439491472e-06,
"loss": 0.4331,
"step": 486
},
{
"epoch": 0.5844,
"grad_norm": 0.8605379841208811,
"learning_rate": 9.73348112908318e-06,
"loss": 0.5131,
"step": 487
},
{
"epoch": 0.5856,
"grad_norm": 0.9331830038285185,
"learning_rate": 9.73122864328531e-06,
"loss": 0.4496,
"step": 488
},
{
"epoch": 0.5868,
"grad_norm": 0.8583259388782629,
"learning_rate": 9.72896694190855e-06,
"loss": 0.399,
"step": 489
},
{
"epoch": 0.588,
"grad_norm": 0.9010560534496752,
"learning_rate": 9.726696029358283e-06,
"loss": 0.4596,
"step": 490
},
{
"epoch": 0.5892,
"grad_norm": 0.8092770600902328,
"learning_rate": 9.724415910057839e-06,
"loss": 0.4299,
"step": 491
},
{
"epoch": 0.5904,
"grad_norm": 0.8244413868757882,
"learning_rate": 9.722126588448473e-06,
"loss": 0.4654,
"step": 492
},
{
"epoch": 0.5916,
"grad_norm": 0.8442814286132542,
"learning_rate": 9.719828068989378e-06,
"loss": 0.4945,
"step": 493
},
{
"epoch": 0.5928,
"grad_norm": 0.9614696357156148,
"learning_rate": 9.717520356157648e-06,
"loss": 0.4436,
"step": 494
},
{
"epoch": 0.594,
"grad_norm": 0.9675809936647133,
"learning_rate": 9.715203454448297e-06,
"loss": 0.4999,
"step": 495
},
{
"epoch": 0.5952,
"grad_norm": 1.2402076561586621,
"learning_rate": 9.712877368374226e-06,
"loss": 0.553,
"step": 496
},
{
"epoch": 0.5964,
"grad_norm": 0.8843890466720776,
"learning_rate": 9.710542102466229e-06,
"loss": 0.4371,
"step": 497
},
{
"epoch": 0.5976,
"grad_norm": 0.8665947349201416,
"learning_rate": 9.708197661272989e-06,
"loss": 0.4043,
"step": 498
},
{
"epoch": 0.5988,
"grad_norm": 0.7284981915181952,
"learning_rate": 9.70584404936105e-06,
"loss": 0.4331,
"step": 499
},
{
"epoch": 0.6,
"grad_norm": 1.0188109205653166,
"learning_rate": 9.703481271314823e-06,
"loss": 0.4796,
"step": 500
},
{
"epoch": 0.6012,
"grad_norm": 0.8465617271896644,
"learning_rate": 9.701109331736573e-06,
"loss": 0.4496,
"step": 501
},
{
"epoch": 0.6024,
"grad_norm": 0.7940264568338631,
"learning_rate": 9.69872823524641e-06,
"loss": 0.4176,
"step": 502
},
{
"epoch": 0.6036,
"grad_norm": 0.8852630887117046,
"learning_rate": 9.69633798648228e-06,
"loss": 0.4874,
"step": 503
},
{
"epoch": 0.6048,
"grad_norm": 0.791374244793118,
"learning_rate": 9.693938590099958e-06,
"loss": 0.47,
"step": 504
},
{
"epoch": 0.606,
"grad_norm": 0.7953029508282652,
"learning_rate": 9.691530050773031e-06,
"loss": 0.4682,
"step": 505
},
{
"epoch": 0.6072,
"grad_norm": 0.9851070544127377,
"learning_rate": 9.689112373192899e-06,
"loss": 0.5085,
"step": 506
},
{
"epoch": 0.6084,
"grad_norm": 0.8460424349869347,
"learning_rate": 9.686685562068765e-06,
"loss": 0.4701,
"step": 507
},
{
"epoch": 0.6096,
"grad_norm": 0.7708516312962796,
"learning_rate": 9.684249622127616e-06,
"loss": 0.4429,
"step": 508
},
{
"epoch": 0.6108,
"grad_norm": 0.8494792959251225,
"learning_rate": 9.681804558114222e-06,
"loss": 0.4734,
"step": 509
},
{
"epoch": 0.612,
"grad_norm": 0.8632619314086427,
"learning_rate": 9.67935037479113e-06,
"loss": 0.4662,
"step": 510
},
{
"epoch": 0.6132,
"grad_norm": 0.8945814963642842,
"learning_rate": 9.676887076938642e-06,
"loss": 0.439,
"step": 511
},
{
"epoch": 0.6144,
"grad_norm": 0.9063294220289309,
"learning_rate": 9.674414669354819e-06,
"loss": 0.4986,
"step": 512
},
{
"epoch": 0.6156,
"grad_norm": 0.7687738333210964,
"learning_rate": 9.671933156855464e-06,
"loss": 0.3897,
"step": 513
},
{
"epoch": 0.6168,
"grad_norm": 0.9848961550281078,
"learning_rate": 9.669442544274115e-06,
"loss": 0.507,
"step": 514
},
{
"epoch": 0.618,
"grad_norm": 0.8110647606361835,
"learning_rate": 9.666942836462036e-06,
"loss": 0.5043,
"step": 515
},
{
"epoch": 0.6192,
"grad_norm": 0.9310530662281385,
"learning_rate": 9.664434038288207e-06,
"loss": 0.4175,
"step": 516
},
{
"epoch": 0.6204,
"grad_norm": 1.2368255686318934,
"learning_rate": 9.661916154639312e-06,
"loss": 0.5639,
"step": 517
},
{
"epoch": 0.6216,
"grad_norm": 1.2405507744293867,
"learning_rate": 9.659389190419735e-06,
"loss": 0.5179,
"step": 518
},
{
"epoch": 0.6228,
"grad_norm": 0.8362989226721922,
"learning_rate": 9.656853150551543e-06,
"loss": 0.405,
"step": 519
},
{
"epoch": 0.624,
"grad_norm": 0.7313768468631252,
"learning_rate": 9.654308039974489e-06,
"loss": 0.4317,
"step": 520
},
{
"epoch": 0.6252,
"grad_norm": 0.86655086386136,
"learning_rate": 9.651753863645985e-06,
"loss": 0.4822,
"step": 521
},
{
"epoch": 0.6264,
"grad_norm": 0.8257425724573098,
"learning_rate": 9.649190626541105e-06,
"loss": 0.4274,
"step": 522
},
{
"epoch": 0.6276,
"grad_norm": 0.8388377939442883,
"learning_rate": 9.646618333652574e-06,
"loss": 0.4564,
"step": 523
},
{
"epoch": 0.6288,
"grad_norm": 0.8175099717686145,
"learning_rate": 9.644036989990753e-06,
"loss": 0.4019,
"step": 524
},
{
"epoch": 0.63,
"grad_norm": 0.8319866581759247,
"learning_rate": 9.641446600583632e-06,
"loss": 0.439,
"step": 525
},
{
"epoch": 0.6312,
"grad_norm": 1.100420077921984,
"learning_rate": 9.638847170476824e-06,
"loss": 0.4961,
"step": 526
},
{
"epoch": 0.6324,
"grad_norm": 0.7877922912913198,
"learning_rate": 9.636238704733547e-06,
"loss": 0.443,
"step": 527
},
{
"epoch": 0.6336,
"grad_norm": 0.8308514380989223,
"learning_rate": 9.633621208434623e-06,
"loss": 0.4466,
"step": 528
},
{
"epoch": 0.6348,
"grad_norm": 0.9218970682062584,
"learning_rate": 9.630994686678462e-06,
"loss": 0.4731,
"step": 529
},
{
"epoch": 0.636,
"grad_norm": 0.8097304059512906,
"learning_rate": 9.628359144581052e-06,
"loss": 0.4728,
"step": 530
},
{
"epoch": 0.6372,
"grad_norm": 0.9592825946475372,
"learning_rate": 9.625714587275954e-06,
"loss": 0.421,
"step": 531
},
{
"epoch": 0.6384,
"grad_norm": 0.873449582523449,
"learning_rate": 9.623061019914291e-06,
"loss": 0.4814,
"step": 532
},
{
"epoch": 0.6396,
"grad_norm": 0.8501407538713629,
"learning_rate": 9.620398447664727e-06,
"loss": 0.5609,
"step": 533
},
{
"epoch": 0.6408,
"grad_norm": 0.9048679490384988,
"learning_rate": 9.617726875713477e-06,
"loss": 0.4133,
"step": 534
},
{
"epoch": 0.642,
"grad_norm": 0.7923682229261191,
"learning_rate": 9.615046309264278e-06,
"loss": 0.4026,
"step": 535
},
{
"epoch": 0.6432,
"grad_norm": 0.7809490508266329,
"learning_rate": 9.612356753538392e-06,
"loss": 0.4234,
"step": 536
},
{
"epoch": 0.6444,
"grad_norm": 0.7985767224770942,
"learning_rate": 9.609658213774584e-06,
"loss": 0.4708,
"step": 537
},
{
"epoch": 0.6456,
"grad_norm": 0.8761561307790414,
"learning_rate": 9.606950695229125e-06,
"loss": 0.5306,
"step": 538
},
{
"epoch": 0.6468,
"grad_norm": 0.7585802344161217,
"learning_rate": 9.60423420317577e-06,
"loss": 0.4531,
"step": 539
},
{
"epoch": 0.648,
"grad_norm": 0.8905933420909701,
"learning_rate": 9.601508742905757e-06,
"loss": 0.4002,
"step": 540
},
{
"epoch": 0.6492,
"grad_norm": 0.9724386030738104,
"learning_rate": 9.59877431972779e-06,
"loss": 0.4966,
"step": 541
},
{
"epoch": 0.6504,
"grad_norm": 0.8751955273217047,
"learning_rate": 9.596030938968028e-06,
"loss": 0.4746,
"step": 542
},
{
"epoch": 0.6516,
"grad_norm": 0.7299891768444063,
"learning_rate": 9.593278605970086e-06,
"loss": 0.4249,
"step": 543
},
{
"epoch": 0.6528,
"grad_norm": 0.7068084391401543,
"learning_rate": 9.590517326095012e-06,
"loss": 0.4118,
"step": 544
},
{
"epoch": 0.654,
"grad_norm": 0.8557712578491251,
"learning_rate": 9.587747104721275e-06,
"loss": 0.4339,
"step": 545
},
{
"epoch": 0.6552,
"grad_norm": 0.8752067820993749,
"learning_rate": 9.58496794724477e-06,
"loss": 0.4049,
"step": 546
},
{
"epoch": 0.6564,
"grad_norm": 0.8179081653801735,
"learning_rate": 9.582179859078793e-06,
"loss": 0.441,
"step": 547
},
{
"epoch": 0.6576,
"grad_norm": 0.6584765695490294,
"learning_rate": 9.579382845654038e-06,
"loss": 0.416,
"step": 548
},
{
"epoch": 0.6588,
"grad_norm": 0.9063077125051442,
"learning_rate": 9.576576912418577e-06,
"loss": 0.4149,
"step": 549
},
{
"epoch": 0.66,
"grad_norm": 0.8937595468965278,
"learning_rate": 9.573762064837866e-06,
"loss": 0.4787,
"step": 550
},
{
"epoch": 0.6612,
"grad_norm": 0.9097189989089333,
"learning_rate": 9.570938308394717e-06,
"loss": 0.4669,
"step": 551
},
{
"epoch": 0.6624,
"grad_norm": 0.8109319347387726,
"learning_rate": 9.568105648589299e-06,
"loss": 0.4413,
"step": 552
},
{
"epoch": 0.6636,
"grad_norm": 0.765424773172657,
"learning_rate": 9.56526409093912e-06,
"loss": 0.426,
"step": 553
},
{
"epoch": 0.6648,
"grad_norm": 0.6799203060226254,
"learning_rate": 9.562413640979024e-06,
"loss": 0.4334,
"step": 554
},
{
"epoch": 0.666,
"grad_norm": 0.8480144369811539,
"learning_rate": 9.55955430426117e-06,
"loss": 0.4219,
"step": 555
},
{
"epoch": 0.6672,
"grad_norm": 0.8668772179984662,
"learning_rate": 9.556686086355032e-06,
"loss": 0.4521,
"step": 556
},
{
"epoch": 0.6684,
"grad_norm": 0.8440426799201499,
"learning_rate": 9.553808992847377e-06,
"loss": 0.4724,
"step": 557
},
{
"epoch": 0.6696,
"grad_norm": 0.9572942740691547,
"learning_rate": 9.550923029342266e-06,
"loss": 0.3938,
"step": 558
},
{
"epoch": 0.6708,
"grad_norm": 0.8178565887490106,
"learning_rate": 9.548028201461034e-06,
"loss": 0.4201,
"step": 559
},
{
"epoch": 0.672,
"grad_norm": 0.9522012401904189,
"learning_rate": 9.545124514842284e-06,
"loss": 0.4003,
"step": 560
},
{
"epoch": 0.6732,
"grad_norm": 0.84262613763804,
"learning_rate": 9.542211975141871e-06,
"loss": 0.4834,
"step": 561
},
{
"epoch": 0.6744,
"grad_norm": 0.9966853997150754,
"learning_rate": 9.5392905880329e-06,
"loss": 0.496,
"step": 562
},
{
"epoch": 0.6756,
"grad_norm": 0.8665881469636052,
"learning_rate": 9.5363603592057e-06,
"loss": 0.4076,
"step": 563
},
{
"epoch": 0.6768,
"grad_norm": 0.9080397147860061,
"learning_rate": 9.53342129436783e-06,
"loss": 0.5048,
"step": 564
},
{
"epoch": 0.678,
"grad_norm": 0.7232428325178791,
"learning_rate": 9.530473399244061e-06,
"loss": 0.4023,
"step": 565
},
{
"epoch": 0.6792,
"grad_norm": 0.8121137452018021,
"learning_rate": 9.527516679576353e-06,
"loss": 0.5177,
"step": 566
},
{
"epoch": 0.6804,
"grad_norm": 0.7576792965271204,
"learning_rate": 9.52455114112387e-06,
"loss": 0.4386,
"step": 567
},
{
"epoch": 0.6816,
"grad_norm": 0.8460228706679173,
"learning_rate": 9.52157678966294e-06,
"loss": 0.4409,
"step": 568
},
{
"epoch": 0.6828,
"grad_norm": 0.7290604397653926,
"learning_rate": 9.518593630987063e-06,
"loss": 0.4074,
"step": 569
},
{
"epoch": 0.684,
"grad_norm": 0.867874380519241,
"learning_rate": 9.515601670906895e-06,
"loss": 0.481,
"step": 570
},
{
"epoch": 0.6852,
"grad_norm": 0.8380334936861836,
"learning_rate": 9.512600915250232e-06,
"loss": 0.4481,
"step": 571
},
{
"epoch": 0.6864,
"grad_norm": 0.9439474955268691,
"learning_rate": 9.509591369862007e-06,
"loss": 0.4581,
"step": 572
},
{
"epoch": 0.6876,
"grad_norm": 0.8510358046008238,
"learning_rate": 9.506573040604268e-06,
"loss": 0.5491,
"step": 573
},
{
"epoch": 0.6888,
"grad_norm": 0.8716638926801934,
"learning_rate": 9.503545933356175e-06,
"loss": 0.4467,
"step": 574
},
{
"epoch": 0.69,
"grad_norm": 0.7016827958387665,
"learning_rate": 9.500510054013989e-06,
"loss": 0.4655,
"step": 575
},
{
"epoch": 0.6912,
"grad_norm": 0.8505749634895948,
"learning_rate": 9.49746540849105e-06,
"loss": 0.4504,
"step": 576
},
{
"epoch": 0.6924,
"grad_norm": 0.8885463062081087,
"learning_rate": 9.494412002717784e-06,
"loss": 0.4726,
"step": 577
},
{
"epoch": 0.6936,
"grad_norm": 0.8370354492558915,
"learning_rate": 9.49134984264167e-06,
"loss": 0.489,
"step": 578
},
{
"epoch": 0.6948,
"grad_norm": 0.824451255130345,
"learning_rate": 9.488278934227242e-06,
"loss": 0.4186,
"step": 579
},
{
"epoch": 0.696,
"grad_norm": 1.0403779667457187,
"learning_rate": 9.485199283456078e-06,
"loss": 0.5544,
"step": 580
},
{
"epoch": 0.6972,
"grad_norm": 1.196819474643937,
"learning_rate": 9.48211089632678e-06,
"loss": 0.5264,
"step": 581
},
{
"epoch": 0.6984,
"grad_norm": 0.7398780521487462,
"learning_rate": 9.479013778854966e-06,
"loss": 0.4173,
"step": 582
},
{
"epoch": 0.6996,
"grad_norm": 0.8282472691974359,
"learning_rate": 9.475907937073265e-06,
"loss": 0.4328,
"step": 583
},
{
"epoch": 0.7008,
"grad_norm": 0.7525988619408324,
"learning_rate": 9.472793377031293e-06,
"loss": 0.392,
"step": 584
},
{
"epoch": 0.702,
"grad_norm": 0.844311963204182,
"learning_rate": 9.469670104795655e-06,
"loss": 0.4326,
"step": 585
},
{
"epoch": 0.7032,
"grad_norm": 1.030525344385058,
"learning_rate": 9.466538126449915e-06,
"loss": 0.4567,
"step": 586
},
{
"epoch": 0.7044,
"grad_norm": 0.8893073246001602,
"learning_rate": 9.463397448094605e-06,
"loss": 0.4906,
"step": 587
},
{
"epoch": 0.7056,
"grad_norm": 1.0461892051463009,
"learning_rate": 9.460248075847199e-06,
"loss": 0.4725,
"step": 588
},
{
"epoch": 0.7068,
"grad_norm": 0.6655516219605128,
"learning_rate": 9.457090015842104e-06,
"loss": 0.4085,
"step": 589
},
{
"epoch": 0.708,
"grad_norm": 0.8641268293544789,
"learning_rate": 9.453923274230653e-06,
"loss": 0.5296,
"step": 590
},
{
"epoch": 0.7092,
"grad_norm": 0.7851147985892276,
"learning_rate": 9.450747857181084e-06,
"loss": 0.4206,
"step": 591
},
{
"epoch": 0.7104,
"grad_norm": 0.7919636097736487,
"learning_rate": 9.447563770878535e-06,
"loss": 0.4796,
"step": 592
},
{
"epoch": 0.7116,
"grad_norm": 0.9225959902934244,
"learning_rate": 9.444371021525036e-06,
"loss": 0.4837,
"step": 593
},
{
"epoch": 0.7128,
"grad_norm": 0.8329302616550331,
"learning_rate": 9.441169615339482e-06,
"loss": 0.4461,
"step": 594
},
{
"epoch": 0.714,
"grad_norm": 0.8626008969422386,
"learning_rate": 9.437959558557635e-06,
"loss": 0.4877,
"step": 595
},
{
"epoch": 0.7152,
"grad_norm": 0.7801338445106245,
"learning_rate": 9.434740857432105e-06,
"loss": 0.4631,
"step": 596
},
{
"epoch": 0.7164,
"grad_norm": 0.7041792494885234,
"learning_rate": 9.431513518232343e-06,
"loss": 0.5038,
"step": 597
},
{
"epoch": 0.7176,
"grad_norm": 0.8370546579638394,
"learning_rate": 9.42827754724462e-06,
"loss": 0.521,
"step": 598
},
{
"epoch": 0.7188,
"grad_norm": 0.8851017458473013,
"learning_rate": 9.425032950772025e-06,
"loss": 0.4953,
"step": 599
},
{
"epoch": 0.72,
"grad_norm": 0.7261755552262745,
"learning_rate": 9.421779735134446e-06,
"loss": 0.4644,
"step": 600
},
{
"epoch": 0.7212,
"grad_norm": 0.833113244180154,
"learning_rate": 9.418517906668556e-06,
"loss": 0.4301,
"step": 601
},
{
"epoch": 0.7224,
"grad_norm": 0.8123579317360561,
"learning_rate": 9.415247471727813e-06,
"loss": 0.4532,
"step": 602
},
{
"epoch": 0.7236,
"grad_norm": 0.9254006183829643,
"learning_rate": 9.41196843668243e-06,
"loss": 0.3969,
"step": 603
},
{
"epoch": 0.7248,
"grad_norm": 0.7598450149562775,
"learning_rate": 9.408680807919377e-06,
"loss": 0.4526,
"step": 604
},
{
"epoch": 0.726,
"grad_norm": 0.8602868132994694,
"learning_rate": 9.405384591842358e-06,
"loss": 0.4409,
"step": 605
},
{
"epoch": 0.7272,
"grad_norm": 0.7680723946357879,
"learning_rate": 9.402079794871812e-06,
"loss": 0.4257,
"step": 606
},
{
"epoch": 0.7284,
"grad_norm": 0.7857914941863341,
"learning_rate": 9.398766423444883e-06,
"loss": 0.4758,
"step": 607
},
{
"epoch": 0.7296,
"grad_norm": 0.71534460072212,
"learning_rate": 9.39544448401542e-06,
"loss": 0.3807,
"step": 608
},
{
"epoch": 0.7308,
"grad_norm": 0.8424889141205625,
"learning_rate": 9.392113983053958e-06,
"loss": 0.443,
"step": 609
},
{
"epoch": 0.732,
"grad_norm": 0.8318687931397658,
"learning_rate": 9.38877492704772e-06,
"loss": 0.4799,
"step": 610
},
{
"epoch": 0.7332,
"grad_norm": 0.9716940713554754,
"learning_rate": 9.385427322500575e-06,
"loss": 0.4785,
"step": 611
},
{
"epoch": 0.7344,
"grad_norm": 0.7137240142457815,
"learning_rate": 9.382071175933058e-06,
"loss": 0.441,
"step": 612
},
{
"epoch": 0.7356,
"grad_norm": 0.658664967675943,
"learning_rate": 9.378706493882335e-06,
"loss": 0.4061,
"step": 613
},
{
"epoch": 0.7368,
"grad_norm": 0.7020418045580746,
"learning_rate": 9.375333282902198e-06,
"loss": 0.4335,
"step": 614
},
{
"epoch": 0.738,
"grad_norm": 0.7328967170173417,
"learning_rate": 9.37195154956305e-06,
"loss": 0.4771,
"step": 615
},
{
"epoch": 0.7392,
"grad_norm": 0.6930246579135281,
"learning_rate": 9.368561300451902e-06,
"loss": 0.4426,
"step": 616
},
{
"epoch": 0.7404,
"grad_norm": 0.726781587229524,
"learning_rate": 9.365162542172346e-06,
"loss": 0.4146,
"step": 617
},
{
"epoch": 0.7416,
"grad_norm": 0.7646894875759652,
"learning_rate": 9.361755281344547e-06,
"loss": 0.4675,
"step": 618
},
{
"epoch": 0.7428,
"grad_norm": 0.8084338423932803,
"learning_rate": 9.358339524605233e-06,
"loss": 0.4995,
"step": 619
},
{
"epoch": 0.744,
"grad_norm": 0.8423715813271311,
"learning_rate": 9.354915278607685e-06,
"loss": 0.4671,
"step": 620
},
{
"epoch": 0.7452,
"grad_norm": 0.7786096715964355,
"learning_rate": 9.351482550021713e-06,
"loss": 0.4694,
"step": 621
},
{
"epoch": 0.7464,
"grad_norm": 0.8485325679103202,
"learning_rate": 9.348041345533653e-06,
"loss": 0.4701,
"step": 622
},
{
"epoch": 0.7476,
"grad_norm": 0.7897073054765453,
"learning_rate": 9.34459167184635e-06,
"loss": 0.4526,
"step": 623
},
{
"epoch": 0.7488,
"grad_norm": 0.7791609167522182,
"learning_rate": 9.341133535679145e-06,
"loss": 0.4389,
"step": 624
},
{
"epoch": 0.75,
"grad_norm": 0.8385746148159599,
"learning_rate": 9.337666943767863e-06,
"loss": 0.4809,
"step": 625
},
{
"epoch": 0.7512,
"grad_norm": 0.8701758180529376,
"learning_rate": 9.334191902864799e-06,
"loss": 0.4642,
"step": 626
},
{
"epoch": 0.7524,
"grad_norm": 0.9063038976065121,
"learning_rate": 9.330708419738704e-06,
"loss": 0.4408,
"step": 627
},
{
"epoch": 0.7536,
"grad_norm": 0.7426344629840347,
"learning_rate": 9.327216501174775e-06,
"loss": 0.4211,
"step": 628
},
{
"epoch": 0.7548,
"grad_norm": 0.9518115258555072,
"learning_rate": 9.323716153974639e-06,
"loss": 0.4844,
"step": 629
},
{
"epoch": 0.756,
"grad_norm": 0.7567169290603653,
"learning_rate": 9.320207384956339e-06,
"loss": 0.4272,
"step": 630
},
{
"epoch": 0.7572,
"grad_norm": 0.8887537265325719,
"learning_rate": 9.316690200954324e-06,
"loss": 0.4841,
"step": 631
},
{
"epoch": 0.7584,
"grad_norm": 0.7424613156242852,
"learning_rate": 9.313164608819434e-06,
"loss": 0.392,
"step": 632
},
{
"epoch": 0.7596,
"grad_norm": 0.7407562187997445,
"learning_rate": 9.309630615418884e-06,
"loss": 0.4284,
"step": 633
},
{
"epoch": 0.7608,
"grad_norm": 0.7049275491005882,
"learning_rate": 9.306088227636257e-06,
"loss": 0.37,
"step": 634
},
{
"epoch": 0.762,
"grad_norm": 0.8983219957021749,
"learning_rate": 9.302537452371482e-06,
"loss": 0.5145,
"step": 635
},
{
"epoch": 0.7632,
"grad_norm": 0.7732582063747871,
"learning_rate": 9.298978296540829e-06,
"loss": 0.4898,
"step": 636
},
{
"epoch": 0.7644,
"grad_norm": 0.6960427635945086,
"learning_rate": 9.295410767076891e-06,
"loss": 0.4365,
"step": 637
},
{
"epoch": 0.7656,
"grad_norm": 0.9232132043730165,
"learning_rate": 9.291834870928573e-06,
"loss": 0.3868,
"step": 638
},
{
"epoch": 0.7668,
"grad_norm": 0.977583252845986,
"learning_rate": 9.288250615061073e-06,
"loss": 0.4959,
"step": 639
},
{
"epoch": 0.768,
"grad_norm": 0.7569579063571902,
"learning_rate": 9.284658006455871e-06,
"loss": 0.4184,
"step": 640
},
{
"epoch": 0.7692,
"grad_norm": 0.8712490284709843,
"learning_rate": 9.281057052110725e-06,
"loss": 0.466,
"step": 641
},
{
"epoch": 0.7704,
"grad_norm": 0.7336677539698077,
"learning_rate": 9.27744775903964e-06,
"loss": 0.3985,
"step": 642
},
{
"epoch": 0.7716,
"grad_norm": 0.7512324693868481,
"learning_rate": 9.27383013427287e-06,
"loss": 0.4428,
"step": 643
},
{
"epoch": 0.7728,
"grad_norm": 0.8761608818670628,
"learning_rate": 9.270204184856893e-06,
"loss": 0.4774,
"step": 644
},
{
"epoch": 0.774,
"grad_norm": 0.9497880783403919,
"learning_rate": 9.266569917854403e-06,
"loss": 0.4021,
"step": 645
},
{
"epoch": 0.7752,
"grad_norm": 0.9374654621386037,
"learning_rate": 9.262927340344296e-06,
"loss": 0.4226,
"step": 646
},
{
"epoch": 0.7764,
"grad_norm": 0.8399638732234506,
"learning_rate": 9.259276459421655e-06,
"loss": 0.4516,
"step": 647
},
{
"epoch": 0.7776,
"grad_norm": 0.704510557230354,
"learning_rate": 9.255617282197739e-06,
"loss": 0.4423,
"step": 648
},
{
"epoch": 0.7788,
"grad_norm": 0.754216249205254,
"learning_rate": 9.25194981579996e-06,
"loss": 0.4184,
"step": 649
},
{
"epoch": 0.78,
"grad_norm": 0.6911568569721259,
"learning_rate": 9.248274067371886e-06,
"loss": 0.4102,
"step": 650
},
{
"epoch": 0.7812,
"grad_norm": 0.7650799492235767,
"learning_rate": 9.244590044073205e-06,
"loss": 0.4303,
"step": 651
},
{
"epoch": 0.7824,
"grad_norm": 0.9085476691426189,
"learning_rate": 9.240897753079734e-06,
"loss": 0.5205,
"step": 652
},
{
"epoch": 0.7836,
"grad_norm": 0.785540709092756,
"learning_rate": 9.237197201583386e-06,
"loss": 0.456,
"step": 653
},
{
"epoch": 0.7848,
"grad_norm": 0.7343138726108955,
"learning_rate": 9.233488396792167e-06,
"loss": 0.5097,
"step": 654
},
{
"epoch": 0.786,
"grad_norm": 1.3779949749524147,
"learning_rate": 9.22977134593016e-06,
"loss": 0.4708,
"step": 655
},
{
"epoch": 0.7872,
"grad_norm": 0.8690484175601422,
"learning_rate": 9.226046056237508e-06,
"loss": 0.4304,
"step": 656
},
{
"epoch": 0.7884,
"grad_norm": 0.9158843219614342,
"learning_rate": 9.222312534970403e-06,
"loss": 0.4595,
"step": 657
},
{
"epoch": 0.7896,
"grad_norm": 0.6919122756813919,
"learning_rate": 9.218570789401071e-06,
"loss": 0.4243,
"step": 658
},
{
"epoch": 0.7908,
"grad_norm": 0.8788773595045949,
"learning_rate": 9.214820826817754e-06,
"loss": 0.4363,
"step": 659
},
{
"epoch": 0.792,
"grad_norm": 0.63491094153797,
"learning_rate": 9.211062654524705e-06,
"loss": 0.4208,
"step": 660
},
{
"epoch": 0.7932,
"grad_norm": 0.8636907209737108,
"learning_rate": 9.207296279842162e-06,
"loss": 0.4742,
"step": 661
},
{
"epoch": 0.7944,
"grad_norm": 0.9226410118804076,
"learning_rate": 9.203521710106344e-06,
"loss": 0.5275,
"step": 662
},
{
"epoch": 0.7956,
"grad_norm": 0.8403310287900696,
"learning_rate": 9.199738952669431e-06,
"loss": 0.4145,
"step": 663
},
{
"epoch": 0.7968,
"grad_norm": 0.9606119059881773,
"learning_rate": 9.195948014899551e-06,
"loss": 0.5175,
"step": 664
},
{
"epoch": 0.798,
"grad_norm": 0.7041643785619085,
"learning_rate": 9.192148904180769e-06,
"loss": 0.4123,
"step": 665
},
{
"epoch": 0.7992,
"grad_norm": 0.7581826560841832,
"learning_rate": 9.188341627913061e-06,
"loss": 0.4416,
"step": 666
},
{
"epoch": 0.8004,
"grad_norm": 0.707258990475582,
"learning_rate": 9.184526193512318e-06,
"loss": 0.4812,
"step": 667
},
{
"epoch": 0.8016,
"grad_norm": 0.7755477006760664,
"learning_rate": 9.180702608410314e-06,
"loss": 0.5233,
"step": 668
},
{
"epoch": 0.8028,
"grad_norm": 0.7169049921428903,
"learning_rate": 9.176870880054704e-06,
"loss": 0.4285,
"step": 669
},
{
"epoch": 0.804,
"grad_norm": 0.8702739374820536,
"learning_rate": 9.173031015909005e-06,
"loss": 0.4109,
"step": 670
},
{
"epoch": 0.8052,
"grad_norm": 0.9161468570355206,
"learning_rate": 9.169183023452574e-06,
"loss": 0.4466,
"step": 671
},
{
"epoch": 0.8064,
"grad_norm": 0.7287650925856214,
"learning_rate": 9.165326910180608e-06,
"loss": 0.421,
"step": 672
},
{
"epoch": 0.8076,
"grad_norm": 0.7834687441827469,
"learning_rate": 9.161462683604118e-06,
"loss": 0.4834,
"step": 673
},
{
"epoch": 0.8088,
"grad_norm": 0.75715014591666,
"learning_rate": 9.157590351249923e-06,
"loss": 0.4375,
"step": 674
},
{
"epoch": 0.81,
"grad_norm": 0.8622813172893878,
"learning_rate": 9.153709920660624e-06,
"loss": 0.5109,
"step": 675
},
{
"epoch": 0.8112,
"grad_norm": 0.7279308430823479,
"learning_rate": 9.149821399394597e-06,
"loss": 0.4466,
"step": 676
},
{
"epoch": 0.8124,
"grad_norm": 0.8562824166679606,
"learning_rate": 9.145924795025984e-06,
"loss": 0.4356,
"step": 677
},
{
"epoch": 0.8136,
"grad_norm": 0.8302991343321063,
"learning_rate": 9.142020115144662e-06,
"loss": 0.4529,
"step": 678
},
{
"epoch": 0.8148,
"grad_norm": 0.9480287200225315,
"learning_rate": 9.138107367356247e-06,
"loss": 0.4499,
"step": 679
},
{
"epoch": 0.816,
"grad_norm": 0.872132079650296,
"learning_rate": 9.134186559282058e-06,
"loss": 0.5024,
"step": 680
},
{
"epoch": 0.8172,
"grad_norm": 0.9040333141523084,
"learning_rate": 9.130257698559129e-06,
"loss": 0.523,
"step": 681
},
{
"epoch": 0.8184,
"grad_norm": 0.7166259100225029,
"learning_rate": 9.126320792840165e-06,
"loss": 0.4493,
"step": 682
},
{
"epoch": 0.8196,
"grad_norm": 0.8289629385565003,
"learning_rate": 9.12237584979355e-06,
"loss": 0.4546,
"step": 683
},
{
"epoch": 0.8208,
"grad_norm": 0.7612186856506001,
"learning_rate": 9.11842287710332e-06,
"loss": 0.3795,
"step": 684
},
{
"epoch": 0.822,
"grad_norm": 0.8260809132764367,
"learning_rate": 9.114461882469154e-06,
"loss": 0.4751,
"step": 685
},
{
"epoch": 0.8232,
"grad_norm": 0.7607704484154403,
"learning_rate": 9.110492873606351e-06,
"loss": 0.4479,
"step": 686
},
{
"epoch": 0.8244,
"grad_norm": 0.67434711675443,
"learning_rate": 9.106515858245825e-06,
"loss": 0.4588,
"step": 687
},
{
"epoch": 0.8256,
"grad_norm": 0.751651803794908,
"learning_rate": 9.102530844134084e-06,
"loss": 0.4621,
"step": 688
},
{
"epoch": 0.8268,
"grad_norm": 0.7533973452205044,
"learning_rate": 9.098537839033213e-06,
"loss": 0.3787,
"step": 689
},
{
"epoch": 0.828,
"grad_norm": 0.7359046650472385,
"learning_rate": 9.094536850720867e-06,
"loss": 0.4557,
"step": 690
},
{
"epoch": 0.8292,
"grad_norm": 0.6636718963076734,
"learning_rate": 9.090527886990249e-06,
"loss": 0.4573,
"step": 691
},
{
"epoch": 0.8304,
"grad_norm": 0.8518716452638908,
"learning_rate": 9.086510955650095e-06,
"loss": 0.4816,
"step": 692
},
{
"epoch": 0.8316,
"grad_norm": 0.8076208220855162,
"learning_rate": 9.082486064524663e-06,
"loss": 0.421,
"step": 693
},
{
"epoch": 0.8328,
"grad_norm": 0.7633456160283818,
"learning_rate": 9.078453221453714e-06,
"loss": 0.4427,
"step": 694
},
{
"epoch": 0.834,
"grad_norm": 0.7059905997543798,
"learning_rate": 9.074412434292496e-06,
"loss": 0.4592,
"step": 695
},
{
"epoch": 0.8352,
"grad_norm": 0.7422159483319694,
"learning_rate": 9.070363710911736e-06,
"loss": 0.4372,
"step": 696
},
{
"epoch": 0.8364,
"grad_norm": 0.7947299119876228,
"learning_rate": 9.066307059197612e-06,
"loss": 0.4394,
"step": 697
},
{
"epoch": 0.8376,
"grad_norm": 0.6723133040449767,
"learning_rate": 9.062242487051752e-06,
"loss": 0.453,
"step": 698
},
{
"epoch": 0.8388,
"grad_norm": 0.7075368523925276,
"learning_rate": 9.058170002391205e-06,
"loss": 0.5002,
"step": 699
},
{
"epoch": 0.84,
"grad_norm": 0.9650100321752244,
"learning_rate": 9.05408961314844e-06,
"loss": 0.4816,
"step": 700
},
{
"epoch": 0.8412,
"grad_norm": 0.8975824071243218,
"learning_rate": 9.050001327271314e-06,
"loss": 0.4029,
"step": 701
},
{
"epoch": 0.8424,
"grad_norm": 0.842213023519483,
"learning_rate": 9.045905152723074e-06,
"loss": 0.4097,
"step": 702
},
{
"epoch": 0.8436,
"grad_norm": 0.8615557723977405,
"learning_rate": 9.041801097482323e-06,
"loss": 0.4547,
"step": 703
},
{
"epoch": 0.8448,
"grad_norm": 0.8311482594403901,
"learning_rate": 9.037689169543024e-06,
"loss": 0.4568,
"step": 704
},
{
"epoch": 0.846,
"grad_norm": 0.8021777578541367,
"learning_rate": 9.033569376914467e-06,
"loss": 0.4745,
"step": 705
},
{
"epoch": 0.8472,
"grad_norm": 0.9622011367891243,
"learning_rate": 9.029441727621267e-06,
"loss": 0.5235,
"step": 706
},
{
"epoch": 0.8484,
"grad_norm": 0.8459070160726208,
"learning_rate": 9.025306229703334e-06,
"loss": 0.4512,
"step": 707
},
{
"epoch": 0.8496,
"grad_norm": 0.7189305576630535,
"learning_rate": 9.021162891215879e-06,
"loss": 0.4336,
"step": 708
},
{
"epoch": 0.8508,
"grad_norm": 0.8006231018288971,
"learning_rate": 9.017011720229368e-06,
"loss": 0.4152,
"step": 709
},
{
"epoch": 0.852,
"grad_norm": 0.9082006627153408,
"learning_rate": 9.012852724829539e-06,
"loss": 0.4381,
"step": 710
},
{
"epoch": 0.8532,
"grad_norm": 0.9079781918934876,
"learning_rate": 9.008685913117361e-06,
"loss": 0.455,
"step": 711
},
{
"epoch": 0.8544,
"grad_norm": 0.7804859267351731,
"learning_rate": 9.00451129320903e-06,
"loss": 0.4295,
"step": 712
},
{
"epoch": 0.8556,
"grad_norm": 0.7716566862252554,
"learning_rate": 9.000328873235955e-06,
"loss": 0.4328,
"step": 713
},
{
"epoch": 0.8568,
"grad_norm": 0.9795345408731057,
"learning_rate": 8.996138661344734e-06,
"loss": 0.4899,
"step": 714
},
{
"epoch": 0.858,
"grad_norm": 0.759806729522717,
"learning_rate": 8.99194066569714e-06,
"loss": 0.44,
"step": 715
},
{
"epoch": 0.8592,
"grad_norm": 0.7348851219507503,
"learning_rate": 8.987734894470111e-06,
"loss": 0.4736,
"step": 716
},
{
"epoch": 0.8604,
"grad_norm": 0.6959669568259862,
"learning_rate": 8.983521355855731e-06,
"loss": 0.4279,
"step": 717
},
{
"epoch": 0.8616,
"grad_norm": 0.7344395011734224,
"learning_rate": 8.979300058061214e-06,
"loss": 0.4571,
"step": 718
},
{
"epoch": 0.8628,
"grad_norm": 0.9456494602197333,
"learning_rate": 8.97507100930888e-06,
"loss": 0.4735,
"step": 719
},
{
"epoch": 0.864,
"grad_norm": 0.8070567787932608,
"learning_rate": 8.97083421783616e-06,
"loss": 0.4175,
"step": 720
},
{
"epoch": 0.8652,
"grad_norm": 0.8195085409148622,
"learning_rate": 8.96658969189555e-06,
"loss": 0.4118,
"step": 721
},
{
"epoch": 0.8664,
"grad_norm": 0.7896064069771342,
"learning_rate": 8.962337439754627e-06,
"loss": 0.425,
"step": 722
},
{
"epoch": 0.8676,
"grad_norm": 0.7728208976761178,
"learning_rate": 8.958077469696007e-06,
"loss": 0.4531,
"step": 723
},
{
"epoch": 0.8688,
"grad_norm": 0.895340618574441,
"learning_rate": 8.953809790017342e-06,
"loss": 0.4707,
"step": 724
},
{
"epoch": 0.87,
"grad_norm": 0.7011741889258672,
"learning_rate": 8.949534409031305e-06,
"loss": 0.3925,
"step": 725
},
{
"epoch": 0.8712,
"grad_norm": 0.8852929403864704,
"learning_rate": 8.94525133506556e-06,
"loss": 0.4204,
"step": 726
},
{
"epoch": 0.8724,
"grad_norm": 0.8058335521207806,
"learning_rate": 8.940960576462763e-06,
"loss": 0.4343,
"step": 727
},
{
"epoch": 0.8736,
"grad_norm": 0.7621175602504676,
"learning_rate": 8.93666214158054e-06,
"loss": 0.4008,
"step": 728
},
{
"epoch": 0.8748,
"grad_norm": 0.9612977755491162,
"learning_rate": 8.932356038791465e-06,
"loss": 0.4099,
"step": 729
},
{
"epoch": 0.876,
"grad_norm": 0.8450073091718325,
"learning_rate": 8.928042276483048e-06,
"loss": 0.4985,
"step": 730
},
{
"epoch": 0.8772,
"grad_norm": 0.8157891758241009,
"learning_rate": 8.923720863057718e-06,
"loss": 0.4965,
"step": 731
},
{
"epoch": 0.8784,
"grad_norm": 0.8180975972183695,
"learning_rate": 8.919391806932807e-06,
"loss": 0.4249,
"step": 732
},
{
"epoch": 0.8796,
"grad_norm": 0.9467271821467541,
"learning_rate": 8.915055116540538e-06,
"loss": 0.4207,
"step": 733
},
{
"epoch": 0.8808,
"grad_norm": 0.7522014566421825,
"learning_rate": 8.910710800327996e-06,
"loss": 0.4549,
"step": 734
},
{
"epoch": 0.882,
"grad_norm": 1.2761327328702223,
"learning_rate": 8.906358866757128e-06,
"loss": 0.5353,
"step": 735
},
{
"epoch": 0.8832,
"grad_norm": 0.7148150818400383,
"learning_rate": 8.901999324304713e-06,
"loss": 0.4704,
"step": 736
},
{
"epoch": 0.8844,
"grad_norm": 1.0179982776078635,
"learning_rate": 8.897632181462354e-06,
"loss": 0.4808,
"step": 737
},
{
"epoch": 0.8856,
"grad_norm": 0.8270771447737706,
"learning_rate": 8.893257446736455e-06,
"loss": 0.4477,
"step": 738
},
{
"epoch": 0.8868,
"grad_norm": 0.8437748960319309,
"learning_rate": 8.888875128648208e-06,
"loss": 0.4206,
"step": 739
},
{
"epoch": 0.888,
"grad_norm": 0.7807849290893369,
"learning_rate": 8.884485235733579e-06,
"loss": 0.426,
"step": 740
},
{
"epoch": 0.8892,
"grad_norm": 0.9146029621918532,
"learning_rate": 8.880087776543287e-06,
"loss": 0.4435,
"step": 741
},
{
"epoch": 0.8904,
"grad_norm": 0.7871883711445494,
"learning_rate": 8.875682759642786e-06,
"loss": 0.4344,
"step": 742
},
{
"epoch": 0.8916,
"grad_norm": 0.9072170440487901,
"learning_rate": 8.871270193612254e-06,
"loss": 0.5023,
"step": 743
},
{
"epoch": 0.8928,
"grad_norm": 0.9412266984437514,
"learning_rate": 8.866850087046574e-06,
"loss": 0.4749,
"step": 744
},
{
"epoch": 0.894,
"grad_norm": 0.7675021859245141,
"learning_rate": 8.862422448555317e-06,
"loss": 0.4027,
"step": 745
},
{
"epoch": 0.8952,
"grad_norm": 0.8098327079204004,
"learning_rate": 8.857987286762718e-06,
"loss": 0.4368,
"step": 746
},
{
"epoch": 0.8964,
"grad_norm": 0.6839255528943358,
"learning_rate": 8.853544610307675e-06,
"loss": 0.4539,
"step": 747
},
{
"epoch": 0.8976,
"grad_norm": 0.7692983753407266,
"learning_rate": 8.84909442784372e-06,
"loss": 0.4808,
"step": 748
},
{
"epoch": 0.8988,
"grad_norm": 0.6911496204641635,
"learning_rate": 8.844636748038999e-06,
"loss": 0.4088,
"step": 749
},
{
"epoch": 0.9,
"grad_norm": 0.8586187046509839,
"learning_rate": 8.840171579576273e-06,
"loss": 0.3956,
"step": 750
},
{
"epoch": 0.9012,
"grad_norm": 0.6803110924696615,
"learning_rate": 8.83569893115288e-06,
"loss": 0.4076,
"step": 751
},
{
"epoch": 0.9024,
"grad_norm": 0.6334781012948423,
"learning_rate": 8.831218811480735e-06,
"loss": 0.4283,
"step": 752
},
{
"epoch": 0.9036,
"grad_norm": 0.7844223477850635,
"learning_rate": 8.8267312292863e-06,
"loss": 0.4661,
"step": 753
},
{
"epoch": 0.9048,
"grad_norm": 0.7871248775682352,
"learning_rate": 8.822236193310574e-06,
"loss": 0.4761,
"step": 754
},
{
"epoch": 0.906,
"grad_norm": 0.6925562184918242,
"learning_rate": 8.817733712309078e-06,
"loss": 0.4527,
"step": 755
},
{
"epoch": 0.9072,
"grad_norm": 0.8023499905921604,
"learning_rate": 8.813223795051828e-06,
"loss": 0.435,
"step": 756
},
{
"epoch": 0.9084,
"grad_norm": 0.7892762326178507,
"learning_rate": 8.80870645032333e-06,
"loss": 0.4276,
"step": 757
},
{
"epoch": 0.9096,
"grad_norm": 0.707043257744133,
"learning_rate": 8.804181686922555e-06,
"loss": 0.4519,
"step": 758
},
{
"epoch": 0.9108,
"grad_norm": 0.7077790320460181,
"learning_rate": 8.799649513662926e-06,
"loss": 0.416,
"step": 759
},
{
"epoch": 0.912,
"grad_norm": 0.706592714129424,
"learning_rate": 8.795109939372298e-06,
"loss": 0.4139,
"step": 760
},
{
"epoch": 0.9132,
"grad_norm": 0.6856678766495268,
"learning_rate": 8.79056297289294e-06,
"loss": 0.4161,
"step": 761
},
{
"epoch": 0.9144,
"grad_norm": 0.9716186804926186,
"learning_rate": 8.786008623081526e-06,
"loss": 0.5152,
"step": 762
},
{
"epoch": 0.9156,
"grad_norm": 0.7488004182023753,
"learning_rate": 8.781446898809101e-06,
"loss": 0.4617,
"step": 763
},
{
"epoch": 0.9168,
"grad_norm": 0.7682878693341557,
"learning_rate": 8.776877808961082e-06,
"loss": 0.453,
"step": 764
},
{
"epoch": 0.918,
"grad_norm": 0.8050534112825974,
"learning_rate": 8.772301362437233e-06,
"loss": 0.4459,
"step": 765
},
{
"epoch": 0.9192,
"grad_norm": 0.6859092247712775,
"learning_rate": 8.767717568151643e-06,
"loss": 0.4509,
"step": 766
},
{
"epoch": 0.9204,
"grad_norm": 0.9727580813082001,
"learning_rate": 8.763126435032717e-06,
"loss": 0.4199,
"step": 767
},
{
"epoch": 0.9216,
"grad_norm": 0.8787332548702836,
"learning_rate": 8.758527972023151e-06,
"loss": 0.484,
"step": 768
},
{
"epoch": 0.9228,
"grad_norm": 0.7920747238023574,
"learning_rate": 8.753922188079923e-06,
"loss": 0.4962,
"step": 769
},
{
"epoch": 0.924,
"grad_norm": 0.6963883700581066,
"learning_rate": 8.749309092174267e-06,
"loss": 0.4824,
"step": 770
},
{
"epoch": 0.9252,
"grad_norm": 0.6792243383756048,
"learning_rate": 8.744688693291658e-06,
"loss": 0.4057,
"step": 771
},
{
"epoch": 0.9264,
"grad_norm": 0.8008957238261251,
"learning_rate": 8.740061000431805e-06,
"loss": 0.4584,
"step": 772
},
{
"epoch": 0.9276,
"grad_norm": 0.7795832445708237,
"learning_rate": 8.735426022608611e-06,
"loss": 0.5347,
"step": 773
},
{
"epoch": 0.9288,
"grad_norm": 0.7501877669496043,
"learning_rate": 8.73078376885018e-06,
"loss": 0.4848,
"step": 774
},
{
"epoch": 0.93,
"grad_norm": 0.7767405533895025,
"learning_rate": 8.726134248198782e-06,
"loss": 0.4396,
"step": 775
},
{
"epoch": 0.9312,
"grad_norm": 0.6054931704153814,
"learning_rate": 8.721477469710845e-06,
"loss": 0.4491,
"step": 776
},
{
"epoch": 0.9324,
"grad_norm": 0.7075404481192714,
"learning_rate": 8.71681344245693e-06,
"loss": 0.4545,
"step": 777
},
{
"epoch": 0.9336,
"grad_norm": 0.7823362946504085,
"learning_rate": 8.712142175521723e-06,
"loss": 0.4245,
"step": 778
},
{
"epoch": 0.9348,
"grad_norm": 0.6039517823728174,
"learning_rate": 8.707463678004004e-06,
"loss": 0.3998,
"step": 779
},
{
"epoch": 0.936,
"grad_norm": 0.6364272325200467,
"learning_rate": 8.702777959016647e-06,
"loss": 0.4301,
"step": 780
},
{
"epoch": 0.9372,
"grad_norm": 0.774892437596723,
"learning_rate": 8.698085027686581e-06,
"loss": 0.3696,
"step": 781
},
{
"epoch": 0.9384,
"grad_norm": 0.7564527774207315,
"learning_rate": 8.69338489315479e-06,
"loss": 0.456,
"step": 782
},
{
"epoch": 0.9396,
"grad_norm": 0.8978506363473915,
"learning_rate": 8.68867756457629e-06,
"loss": 0.4672,
"step": 783
},
{
"epoch": 0.9408,
"grad_norm": 0.672936602862321,
"learning_rate": 8.683963051120103e-06,
"loss": 0.4886,
"step": 784
},
{
"epoch": 0.942,
"grad_norm": 0.6768775531426905,
"learning_rate": 8.679241361969252e-06,
"loss": 0.4405,
"step": 785
},
{
"epoch": 0.9432,
"grad_norm": 0.652412006700396,
"learning_rate": 8.674512506320733e-06,
"loss": 0.4726,
"step": 786
},
{
"epoch": 0.9444,
"grad_norm": 0.7270605169576491,
"learning_rate": 8.669776493385506e-06,
"loss": 0.4386,
"step": 787
},
{
"epoch": 0.9456,
"grad_norm": 0.7002592292412853,
"learning_rate": 8.665033332388466e-06,
"loss": 0.3826,
"step": 788
},
{
"epoch": 0.9468,
"grad_norm": 0.7845264245420243,
"learning_rate": 8.660283032568435e-06,
"loss": 0.4136,
"step": 789
},
{
"epoch": 0.948,
"grad_norm": 0.6933764713694855,
"learning_rate": 8.655525603178137e-06,
"loss": 0.4187,
"step": 790
},
{
"epoch": 0.9492,
"grad_norm": 0.9363045420708548,
"learning_rate": 8.650761053484188e-06,
"loss": 0.4263,
"step": 791
},
{
"epoch": 0.9504,
"grad_norm": 0.7857164587564591,
"learning_rate": 8.645989392767068e-06,
"loss": 0.3899,
"step": 792
},
{
"epoch": 0.9516,
"grad_norm": 0.7338618714454378,
"learning_rate": 8.641210630321115e-06,
"loss": 0.4194,
"step": 793
},
{
"epoch": 0.9528,
"grad_norm": 0.6570136164168231,
"learning_rate": 8.636424775454489e-06,
"loss": 0.4201,
"step": 794
},
{
"epoch": 0.954,
"grad_norm": 0.757552087109251,
"learning_rate": 8.631631837489173e-06,
"loss": 0.4451,
"step": 795
},
{
"epoch": 0.9552,
"grad_norm": 0.7625964343972887,
"learning_rate": 8.626831825760946e-06,
"loss": 0.4198,
"step": 796
},
{
"epoch": 0.9564,
"grad_norm": 0.6471543714799735,
"learning_rate": 8.622024749619363e-06,
"loss": 0.4274,
"step": 797
},
{
"epoch": 0.9576,
"grad_norm": 0.6764295287928739,
"learning_rate": 8.61721061842774e-06,
"loss": 0.3786,
"step": 798
},
{
"epoch": 0.9588,
"grad_norm": 0.6814866200769416,
"learning_rate": 8.612389441563136e-06,
"loss": 0.4418,
"step": 799
},
{
"epoch": 0.96,
"grad_norm": 0.7102732908783905,
"learning_rate": 8.60756122841633e-06,
"loss": 0.3992,
"step": 800
},
{
"epoch": 0.9612,
"grad_norm": 0.8429308481781861,
"learning_rate": 8.602725988391814e-06,
"loss": 0.4549,
"step": 801
},
{
"epoch": 0.9624,
"grad_norm": 0.7525976244504,
"learning_rate": 8.597883730907757e-06,
"loss": 0.4942,
"step": 802
},
{
"epoch": 0.9636,
"grad_norm": 0.7840793386213482,
"learning_rate": 8.593034465396007e-06,
"loss": 0.4836,
"step": 803
},
{
"epoch": 0.9648,
"grad_norm": 0.726118627041448,
"learning_rate": 8.588178201302052e-06,
"loss": 0.4678,
"step": 804
},
{
"epoch": 0.966,
"grad_norm": 0.7767177356037591,
"learning_rate": 8.583314948085023e-06,
"loss": 0.4674,
"step": 805
},
{
"epoch": 0.9672,
"grad_norm": 0.7702275250048074,
"learning_rate": 8.578444715217652e-06,
"loss": 0.4308,
"step": 806
},
{
"epoch": 0.9684,
"grad_norm": 0.7152790548636399,
"learning_rate": 8.57356751218628e-06,
"loss": 0.4144,
"step": 807
},
{
"epoch": 0.9696,
"grad_norm": 0.6730617848520981,
"learning_rate": 8.568683348490817e-06,
"loss": 0.4392,
"step": 808
},
{
"epoch": 0.9708,
"grad_norm": 0.665276094832332,
"learning_rate": 8.563792233644725e-06,
"loss": 0.4399,
"step": 809
},
{
"epoch": 0.972,
"grad_norm": 1.1761722368465695,
"learning_rate": 8.558894177175019e-06,
"loss": 0.4452,
"step": 810
},
{
"epoch": 0.9732,
"grad_norm": 0.7983623852179724,
"learning_rate": 8.553989188622228e-06,
"loss": 0.4157,
"step": 811
},
{
"epoch": 0.9744,
"grad_norm": 1.024867236324754,
"learning_rate": 8.549077277540379e-06,
"loss": 0.5099,
"step": 812
},
{
"epoch": 0.9756,
"grad_norm": 0.7682157910381594,
"learning_rate": 8.544158453496992e-06,
"loss": 0.4959,
"step": 813
},
{
"epoch": 0.9768,
"grad_norm": 0.6072339538424284,
"learning_rate": 8.539232726073046e-06,
"loss": 0.4352,
"step": 814
},
{
"epoch": 0.978,
"grad_norm": 0.909991143327054,
"learning_rate": 8.53430010486297e-06,
"loss": 0.4701,
"step": 815
},
{
"epoch": 0.9792,
"grad_norm": 0.6794612630945962,
"learning_rate": 8.529360599474616e-06,
"loss": 0.4541,
"step": 816
},
{
"epoch": 0.9804,
"grad_norm": 0.8363760842681671,
"learning_rate": 8.524414219529253e-06,
"loss": 0.4145,
"step": 817
},
{
"epoch": 0.9816,
"grad_norm": 0.6625972208715569,
"learning_rate": 8.519460974661533e-06,
"loss": 0.4652,
"step": 818
},
{
"epoch": 0.9828,
"grad_norm": 0.8035739035628665,
"learning_rate": 8.514500874519483e-06,
"loss": 0.4327,
"step": 819
},
{
"epoch": 0.984,
"grad_norm": 0.8557992609946545,
"learning_rate": 8.509533928764482e-06,
"loss": 0.4527,
"step": 820
},
{
"epoch": 0.9852,
"grad_norm": 0.7682340506332006,
"learning_rate": 8.50456014707124e-06,
"loss": 0.4207,
"step": 821
},
{
"epoch": 0.9864,
"grad_norm": 0.8644013670699227,
"learning_rate": 8.499579539127794e-06,
"loss": 0.4396,
"step": 822
},
{
"epoch": 0.9876,
"grad_norm": 0.9119007953684399,
"learning_rate": 8.494592114635458e-06,
"loss": 0.4993,
"step": 823
},
{
"epoch": 0.9888,
"grad_norm": 0.7201456816586235,
"learning_rate": 8.489597883308844e-06,
"loss": 0.4497,
"step": 824
},
{
"epoch": 0.99,
"grad_norm": 0.7106934368349643,
"learning_rate": 8.484596854875806e-06,
"loss": 0.3507,
"step": 825
},
{
"epoch": 0.9912,
"grad_norm": 0.9062702238280697,
"learning_rate": 8.479589039077446e-06,
"loss": 0.4629,
"step": 826
},
{
"epoch": 0.9924,
"grad_norm": 0.7934706949393131,
"learning_rate": 8.474574445668085e-06,
"loss": 0.4325,
"step": 827
},
{
"epoch": 0.9936,
"grad_norm": 0.7064770575629856,
"learning_rate": 8.469553084415247e-06,
"loss": 0.4417,
"step": 828
},
{
"epoch": 0.9948,
"grad_norm": 0.6688362759618923,
"learning_rate": 8.464524965099632e-06,
"loss": 0.4243,
"step": 829
},
{
"epoch": 0.996,
"grad_norm": 0.6866198670164394,
"learning_rate": 8.459490097515114e-06,
"loss": 0.4245,
"step": 830
},
{
"epoch": 0.9972,
"grad_norm": 0.6501030430481411,
"learning_rate": 8.454448491468702e-06,
"loss": 0.4612,
"step": 831
},
{
"epoch": 0.9984,
"grad_norm": 0.7041966461529617,
"learning_rate": 8.449400156780536e-06,
"loss": 0.3998,
"step": 832
},
{
"epoch": 0.9996,
"grad_norm": 0.7582455313143206,
"learning_rate": 8.444345103283858e-06,
"loss": 0.4371,
"step": 833
},
{
"epoch": 1.0,
"grad_norm": 0.7582455313143206,
"learning_rate": 8.439283340825002e-06,
"loss": 0.4449,
"step": 834
},
{
"epoch": 1.0012,
"grad_norm": 1.3777662145607583,
"learning_rate": 8.434214879263365e-06,
"loss": 0.3808,
"step": 835
},
{
"epoch": 1.0024,
"grad_norm": 0.6552793619897616,
"learning_rate": 8.429139728471395e-06,
"loss": 0.3874,
"step": 836
},
{
"epoch": 1.0036,
"grad_norm": 0.7721683031610606,
"learning_rate": 8.424057898334569e-06,
"loss": 0.4222,
"step": 837
},
{
"epoch": 1.0048,
"grad_norm": 0.7901796974883859,
"learning_rate": 8.418969398751375e-06,
"loss": 0.3331,
"step": 838
},
{
"epoch": 1.006,
"grad_norm": 1.295138039979136,
"learning_rate": 8.413874239633291e-06,
"loss": 0.4794,
"step": 839
},
{
"epoch": 1.0072,
"grad_norm": 0.7798890432700465,
"learning_rate": 8.408772430904768e-06,
"loss": 0.3663,
"step": 840
},
{
"epoch": 1.0084,
"grad_norm": 0.7223442901375939,
"learning_rate": 8.403663982503205e-06,
"loss": 0.3217,
"step": 841
},
{
"epoch": 1.0096,
"grad_norm": 0.7677398507322479,
"learning_rate": 8.398548904378938e-06,
"loss": 0.3773,
"step": 842
},
{
"epoch": 1.0108,
"grad_norm": 0.6770086411024532,
"learning_rate": 8.393427206495217e-06,
"loss": 0.4132,
"step": 843
},
{
"epoch": 1.012,
"grad_norm": 0.7366836146983722,
"learning_rate": 8.388298898828182e-06,
"loss": 0.3483,
"step": 844
},
{
"epoch": 1.0132,
"grad_norm": 0.7549776784109077,
"learning_rate": 8.383163991366852e-06,
"loss": 0.3779,
"step": 845
},
{
"epoch": 1.0144,
"grad_norm": 0.8402938742202115,
"learning_rate": 8.378022494113099e-06,
"loss": 0.3693,
"step": 846
},
{
"epoch": 1.0156,
"grad_norm": 0.7431482496839323,
"learning_rate": 8.372874417081632e-06,
"loss": 0.3772,
"step": 847
},
{
"epoch": 1.0168,
"grad_norm": 0.6813142961596346,
"learning_rate": 8.367719770299972e-06,
"loss": 0.3293,
"step": 848
},
{
"epoch": 1.018,
"grad_norm": 0.9184379288572385,
"learning_rate": 8.36255856380844e-06,
"loss": 0.4232,
"step": 849
},
{
"epoch": 1.0192,
"grad_norm": 0.9503013532310471,
"learning_rate": 8.35739080766014e-06,
"loss": 0.4095,
"step": 850
},
{
"epoch": 1.0204,
"grad_norm": 0.706327616455585,
"learning_rate": 8.352216511920921e-06,
"loss": 0.3282,
"step": 851
},
{
"epoch": 1.0216,
"grad_norm": 0.8639937245121996,
"learning_rate": 8.34703568666938e-06,
"loss": 0.4198,
"step": 852
},
{
"epoch": 1.0228,
"grad_norm": 0.837281919072702,
"learning_rate": 8.341848341996828e-06,
"loss": 0.3358,
"step": 853
},
{
"epoch": 1.024,
"grad_norm": 0.7972733610969217,
"learning_rate": 8.336654488007277e-06,
"loss": 0.3583,
"step": 854
},
{
"epoch": 1.0252,
"grad_norm": 0.6832864480749161,
"learning_rate": 8.331454134817414e-06,
"loss": 0.3746,
"step": 855
},
{
"epoch": 1.0264,
"grad_norm": 0.7598527564999485,
"learning_rate": 8.326247292556588e-06,
"loss": 0.3857,
"step": 856
},
{
"epoch": 1.0276,
"grad_norm": 0.6783204414231828,
"learning_rate": 8.321033971366788e-06,
"loss": 0.3993,
"step": 857
},
{
"epoch": 1.0288,
"grad_norm": 0.9629500829291088,
"learning_rate": 8.315814181402623e-06,
"loss": 0.3597,
"step": 858
},
{
"epoch": 1.03,
"grad_norm": 0.687124591753614,
"learning_rate": 8.310587932831302e-06,
"loss": 0.3489,
"step": 859
},
{
"epoch": 1.0312,
"grad_norm": 0.7180773352873208,
"learning_rate": 8.305355235832611e-06,
"loss": 0.3621,
"step": 860
},
{
"epoch": 1.0324,
"grad_norm": 0.7825413723359373,
"learning_rate": 8.300116100598899e-06,
"loss": 0.3663,
"step": 861
},
{
"epoch": 1.0336,
"grad_norm": 0.7554472710568854,
"learning_rate": 8.294870537335054e-06,
"loss": 0.3981,
"step": 862
},
{
"epoch": 1.0348,
"grad_norm": 0.7356426890693817,
"learning_rate": 8.28961855625849e-06,
"loss": 0.4206,
"step": 863
},
{
"epoch": 1.036,
"grad_norm": 0.8304853837712959,
"learning_rate": 8.284360167599113e-06,
"loss": 0.3485,
"step": 864
},
{
"epoch": 1.0372,
"grad_norm": 0.6539060870686277,
"learning_rate": 8.279095381599318e-06,
"loss": 0.3701,
"step": 865
},
{
"epoch": 1.0384,
"grad_norm": 0.7561070366531779,
"learning_rate": 8.273824208513956e-06,
"loss": 0.3755,
"step": 866
},
{
"epoch": 1.0396,
"grad_norm": 0.757334484047741,
"learning_rate": 8.268546658610319e-06,
"loss": 0.3722,
"step": 867
},
{
"epoch": 1.0408,
"grad_norm": 0.894560312791748,
"learning_rate": 8.26326274216812e-06,
"loss": 0.3849,
"step": 868
},
{
"epoch": 1.042,
"grad_norm": 0.5697497554313866,
"learning_rate": 8.257972469479478e-06,
"loss": 0.3453,
"step": 869
},
{
"epoch": 1.0432,
"grad_norm": 0.9749600651239334,
"learning_rate": 8.252675850848886e-06,
"loss": 0.3578,
"step": 870
},
{
"epoch": 1.0444,
"grad_norm": 0.7507200897447335,
"learning_rate": 8.247372896593203e-06,
"loss": 0.379,
"step": 871
},
{
"epoch": 1.0456,
"grad_norm": 0.893450965343529,
"learning_rate": 8.24206361704162e-06,
"loss": 0.3426,
"step": 872
},
{
"epoch": 1.0468,
"grad_norm": 0.7269828232564797,
"learning_rate": 8.236748022535662e-06,
"loss": 0.3873,
"step": 873
},
{
"epoch": 1.048,
"grad_norm": 0.9341687200065399,
"learning_rate": 8.231426123429143e-06,
"loss": 0.3038,
"step": 874
},
{
"epoch": 1.0492,
"grad_norm": 0.7643089837933243,
"learning_rate": 8.226097930088162e-06,
"loss": 0.3788,
"step": 875
},
{
"epoch": 1.0504,
"grad_norm": 0.5123148691896001,
"learning_rate": 8.220763452891078e-06,
"loss": 0.3567,
"step": 876
},
{
"epoch": 1.0516,
"grad_norm": 2.0357282371436813,
"learning_rate": 8.215422702228487e-06,
"loss": 0.4119,
"step": 877
},
{
"epoch": 1.0528,
"grad_norm": 0.9189582618508174,
"learning_rate": 8.210075688503209e-06,
"loss": 0.3749,
"step": 878
},
{
"epoch": 1.054,
"grad_norm": 0.7031535703503601,
"learning_rate": 8.20472242213026e-06,
"loss": 0.3569,
"step": 879
},
{
"epoch": 1.0552,
"grad_norm": 0.7997531489114871,
"learning_rate": 8.199362913536837e-06,
"loss": 0.3646,
"step": 880
},
{
"epoch": 1.0564,
"grad_norm": 0.8725154646600939,
"learning_rate": 8.193997173162293e-06,
"loss": 0.416,
"step": 881
},
{
"epoch": 1.0576,
"grad_norm": 0.7367272925778001,
"learning_rate": 8.188625211458123e-06,
"loss": 0.3792,
"step": 882
},
{
"epoch": 1.0588,
"grad_norm": 0.6505462909235469,
"learning_rate": 8.183247038887937e-06,
"loss": 0.342,
"step": 883
},
{
"epoch": 1.06,
"grad_norm": 0.6476368829876405,
"learning_rate": 8.177862665927445e-06,
"loss": 0.3242,
"step": 884
},
{
"epoch": 1.0612,
"grad_norm": 0.8877106916074848,
"learning_rate": 8.17247210306443e-06,
"loss": 0.354,
"step": 885
},
{
"epoch": 1.0624,
"grad_norm": 0.6380472379852485,
"learning_rate": 8.167075360798739e-06,
"loss": 0.328,
"step": 886
},
{
"epoch": 1.0636,
"grad_norm": 0.8862724315486167,
"learning_rate": 8.161672449642248e-06,
"loss": 0.3628,
"step": 887
},
{
"epoch": 1.0648,
"grad_norm": 0.7377537748637433,
"learning_rate": 8.156263380118855e-06,
"loss": 0.3619,
"step": 888
},
{
"epoch": 1.066,
"grad_norm": 0.6822197713598283,
"learning_rate": 8.150848162764448e-06,
"loss": 0.3923,
"step": 889
},
{
"epoch": 1.0672,
"grad_norm": 0.8220997932552981,
"learning_rate": 8.145426808126894e-06,
"loss": 0.3486,
"step": 890
},
{
"epoch": 1.0684,
"grad_norm": 0.5355937415921455,
"learning_rate": 8.139999326766011e-06,
"loss": 0.3451,
"step": 891
},
{
"epoch": 1.0695999999999999,
"grad_norm": 0.7834937666398496,
"learning_rate": 8.134565729253554e-06,
"loss": 0.3689,
"step": 892
},
{
"epoch": 1.0708,
"grad_norm": 0.7597858860482266,
"learning_rate": 8.129126026173189e-06,
"loss": 0.3905,
"step": 893
},
{
"epoch": 1.072,
"grad_norm": 0.7406928582865676,
"learning_rate": 8.123680228120474e-06,
"loss": 0.3014,
"step": 894
},
{
"epoch": 1.0732,
"grad_norm": 0.6463658520145176,
"learning_rate": 8.118228345702843e-06,
"loss": 0.3662,
"step": 895
},
{
"epoch": 1.0744,
"grad_norm": 0.8408344899522053,
"learning_rate": 8.112770389539574e-06,
"loss": 0.3545,
"step": 896
},
{
"epoch": 1.0756000000000001,
"grad_norm": 0.696011207356119,
"learning_rate": 8.107306370261785e-06,
"loss": 0.3866,
"step": 897
},
{
"epoch": 1.0768,
"grad_norm": 0.7783025494532597,
"learning_rate": 8.101836298512396e-06,
"loss": 0.4342,
"step": 898
},
{
"epoch": 1.078,
"grad_norm": 0.6030438835633921,
"learning_rate": 8.096360184946117e-06,
"loss": 0.3377,
"step": 899
},
{
"epoch": 1.0792,
"grad_norm": 1.0395976776741278,
"learning_rate": 8.09087804022943e-06,
"loss": 0.3743,
"step": 900
},
{
"epoch": 1.0804,
"grad_norm": 0.7737318685316028,
"learning_rate": 8.085389875040566e-06,
"loss": 0.3466,
"step": 901
},
{
"epoch": 1.0816,
"grad_norm": 0.7116520436889562,
"learning_rate": 8.079895700069473e-06,
"loss": 0.3084,
"step": 902
},
{
"epoch": 1.0828,
"grad_norm": 0.7421644748489488,
"learning_rate": 8.074395526017816e-06,
"loss": 0.3785,
"step": 903
},
{
"epoch": 1.084,
"grad_norm": 0.6344278169147585,
"learning_rate": 8.06888936359894e-06,
"loss": 0.3125,
"step": 904
},
{
"epoch": 1.0852,
"grad_norm": 0.8616517995143195,
"learning_rate": 8.063377223537853e-06,
"loss": 0.3604,
"step": 905
},
{
"epoch": 1.0864,
"grad_norm": 0.6090908216471619,
"learning_rate": 8.057859116571213e-06,
"loss": 0.3431,
"step": 906
},
{
"epoch": 1.0876,
"grad_norm": 0.8105799295766426,
"learning_rate": 8.05233505344729e-06,
"loss": 0.3358,
"step": 907
},
{
"epoch": 1.0888,
"grad_norm": 0.7633234279200679,
"learning_rate": 8.046805044925964e-06,
"loss": 0.3592,
"step": 908
},
{
"epoch": 1.09,
"grad_norm": 0.875690753628073,
"learning_rate": 8.041269101778694e-06,
"loss": 0.3637,
"step": 909
},
{
"epoch": 1.0912,
"grad_norm": 0.7986273966314485,
"learning_rate": 8.035727234788496e-06,
"loss": 0.3911,
"step": 910
},
{
"epoch": 1.0924,
"grad_norm": 0.8831359870671331,
"learning_rate": 8.030179454749925e-06,
"loss": 0.3677,
"step": 911
},
{
"epoch": 1.0936,
"grad_norm": 0.6726442718649948,
"learning_rate": 8.024625772469055e-06,
"loss": 0.3768,
"step": 912
},
{
"epoch": 1.0948,
"grad_norm": 0.9720338949271197,
"learning_rate": 8.019066198763458e-06,
"loss": 0.3432,
"step": 913
},
{
"epoch": 1.096,
"grad_norm": 0.7992110168120549,
"learning_rate": 8.013500744462177e-06,
"loss": 0.4028,
"step": 914
},
{
"epoch": 1.0972,
"grad_norm": 0.7006362092537348,
"learning_rate": 8.007929420405714e-06,
"loss": 0.3646,
"step": 915
},
{
"epoch": 1.0984,
"grad_norm": 0.5586773766078817,
"learning_rate": 8.002352237446e-06,
"loss": 0.3892,
"step": 916
},
{
"epoch": 1.0996,
"grad_norm": 1.0409424864568004,
"learning_rate": 7.996769206446383e-06,
"loss": 0.3905,
"step": 917
},
{
"epoch": 1.1008,
"grad_norm": 0.8248014883506193,
"learning_rate": 7.991180338281594e-06,
"loss": 0.3905,
"step": 918
},
{
"epoch": 1.102,
"grad_norm": 0.7343758222600562,
"learning_rate": 7.985585643837743e-06,
"loss": 0.3657,
"step": 919
},
{
"epoch": 1.1032,
"grad_norm": 0.7148968944688652,
"learning_rate": 7.979985134012285e-06,
"loss": 0.3645,
"step": 920
},
{
"epoch": 1.1044,
"grad_norm": 0.8294595195956319,
"learning_rate": 7.974378819713998e-06,
"loss": 0.3995,
"step": 921
},
{
"epoch": 1.1056,
"grad_norm": 0.7160511453877545,
"learning_rate": 7.968766711862971e-06,
"loss": 0.3506,
"step": 922
},
{
"epoch": 1.1068,
"grad_norm": 0.7425920740808283,
"learning_rate": 7.963148821390578e-06,
"loss": 0.338,
"step": 923
},
{
"epoch": 1.108,
"grad_norm": 0.692178490648977,
"learning_rate": 7.957525159239454e-06,
"loss": 0.377,
"step": 924
},
{
"epoch": 1.1092,
"grad_norm": 0.8622253860472193,
"learning_rate": 7.951895736363477e-06,
"loss": 0.4014,
"step": 925
},
{
"epoch": 1.1104,
"grad_norm": 0.7535759250779376,
"learning_rate": 7.946260563727746e-06,
"loss": 0.3795,
"step": 926
},
{
"epoch": 1.1116,
"grad_norm": 0.7324990837319098,
"learning_rate": 7.940619652308562e-06,
"loss": 0.3407,
"step": 927
},
{
"epoch": 1.1128,
"grad_norm": 0.7527243880681903,
"learning_rate": 7.934973013093397e-06,
"loss": 0.3598,
"step": 928
},
{
"epoch": 1.114,
"grad_norm": 0.8394115559036807,
"learning_rate": 7.929320657080886e-06,
"loss": 0.3745,
"step": 929
},
{
"epoch": 1.1152,
"grad_norm": 0.6165830522355202,
"learning_rate": 7.923662595280799e-06,
"loss": 0.3549,
"step": 930
},
{
"epoch": 1.1164,
"grad_norm": 0.762619468721415,
"learning_rate": 7.917998838714019e-06,
"loss": 0.3423,
"step": 931
},
{
"epoch": 1.1176,
"grad_norm": 0.7652801538363777,
"learning_rate": 7.91232939841252e-06,
"loss": 0.3492,
"step": 932
},
{
"epoch": 1.1188,
"grad_norm": 0.6784394820410108,
"learning_rate": 7.906654285419347e-06,
"loss": 0.3433,
"step": 933
},
{
"epoch": 1.12,
"grad_norm": 0.6982002674418744,
"learning_rate": 7.900973510788595e-06,
"loss": 0.3688,
"step": 934
},
{
"epoch": 1.1212,
"grad_norm": 0.9251096323635646,
"learning_rate": 7.895287085585386e-06,
"loss": 0.4247,
"step": 935
},
{
"epoch": 1.1224,
"grad_norm": 0.877416561743421,
"learning_rate": 7.889595020885853e-06,
"loss": 0.4146,
"step": 936
},
{
"epoch": 1.1236,
"grad_norm": 0.6984285688241576,
"learning_rate": 7.883897327777108e-06,
"loss": 0.3634,
"step": 937
},
{
"epoch": 1.1248,
"grad_norm": 0.8761100122571216,
"learning_rate": 7.878194017357229e-06,
"loss": 0.3986,
"step": 938
},
{
"epoch": 1.126,
"grad_norm": 0.7243543325390691,
"learning_rate": 7.87248510073523e-06,
"loss": 0.3325,
"step": 939
},
{
"epoch": 1.1272,
"grad_norm": 0.6250124444820903,
"learning_rate": 7.866770589031057e-06,
"loss": 0.4431,
"step": 940
},
{
"epoch": 1.1284,
"grad_norm": 0.8805812499454114,
"learning_rate": 7.86105049337554e-06,
"loss": 0.3557,
"step": 941
},
{
"epoch": 1.1296,
"grad_norm": 0.6302915585189326,
"learning_rate": 7.855324824910395e-06,
"loss": 0.4267,
"step": 942
},
{
"epoch": 1.1308,
"grad_norm": 0.7015190809739693,
"learning_rate": 7.849593594788192e-06,
"loss": 0.4228,
"step": 943
},
{
"epoch": 1.1320000000000001,
"grad_norm": 0.775451531650632,
"learning_rate": 7.843856814172329e-06,
"loss": 0.3422,
"step": 944
},
{
"epoch": 1.1332,
"grad_norm": 0.8334911078844376,
"learning_rate": 7.83811449423702e-06,
"loss": 0.3539,
"step": 945
},
{
"epoch": 1.1344,
"grad_norm": 0.757729801632078,
"learning_rate": 7.832366646167268e-06,
"loss": 0.3544,
"step": 946
},
{
"epoch": 1.1356,
"grad_norm": 0.7681428421174293,
"learning_rate": 7.82661328115884e-06,
"loss": 0.413,
"step": 947
},
{
"epoch": 1.1368,
"grad_norm": 0.6838953591612161,
"learning_rate": 7.820854410418255e-06,
"loss": 0.3599,
"step": 948
},
{
"epoch": 1.138,
"grad_norm": 0.9066474638075733,
"learning_rate": 7.815090045162752e-06,
"loss": 0.3774,
"step": 949
},
{
"epoch": 1.1392,
"grad_norm": 0.7908967804949574,
"learning_rate": 7.809320196620272e-06,
"loss": 0.3646,
"step": 950
},
{
"epoch": 1.1404,
"grad_norm": 0.6182784154311053,
"learning_rate": 7.80354487602944e-06,
"loss": 0.3533,
"step": 951
},
{
"epoch": 1.1416,
"grad_norm": 1.0205664502771374,
"learning_rate": 7.797764094639537e-06,
"loss": 0.3777,
"step": 952
},
{
"epoch": 1.1428,
"grad_norm": 0.5789152521085267,
"learning_rate": 7.79197786371048e-06,
"loss": 0.4182,
"step": 953
},
{
"epoch": 1.144,
"grad_norm": 0.9881699821333535,
"learning_rate": 7.786186194512802e-06,
"loss": 0.3794,
"step": 954
},
{
"epoch": 1.1452,
"grad_norm": 0.6822429338415211,
"learning_rate": 7.780389098327629e-06,
"loss": 0.3745,
"step": 955
},
{
"epoch": 1.1464,
"grad_norm": 0.7169865436400005,
"learning_rate": 7.774586586446658e-06,
"loss": 0.412,
"step": 956
},
{
"epoch": 1.1476,
"grad_norm": 0.8996819932603994,
"learning_rate": 7.768778670172135e-06,
"loss": 0.3632,
"step": 957
},
{
"epoch": 1.1488,
"grad_norm": 0.7077271913000908,
"learning_rate": 7.762965360816828e-06,
"loss": 0.3706,
"step": 958
},
{
"epoch": 1.15,
"grad_norm": 0.6717679245396442,
"learning_rate": 7.757146669704016e-06,
"loss": 0.3983,
"step": 959
},
{
"epoch": 1.1512,
"grad_norm": 0.7595547234923705,
"learning_rate": 7.75132260816746e-06,
"loss": 0.3809,
"step": 960
},
{
"epoch": 1.1524,
"grad_norm": 0.6842119722788328,
"learning_rate": 7.745493187551378e-06,
"loss": 0.3507,
"step": 961
},
{
"epoch": 1.1536,
"grad_norm": 0.9977268654136809,
"learning_rate": 7.739658419210429e-06,
"loss": 0.3961,
"step": 962
},
{
"epoch": 1.1548,
"grad_norm": 0.9090092035939734,
"learning_rate": 7.733818314509689e-06,
"loss": 0.37,
"step": 963
},
{
"epoch": 1.156,
"grad_norm": 0.6350409507767489,
"learning_rate": 7.727972884824625e-06,
"loss": 0.3137,
"step": 964
},
{
"epoch": 1.1572,
"grad_norm": 0.7803418255211289,
"learning_rate": 7.72212214154108e-06,
"loss": 0.3904,
"step": 965
},
{
"epoch": 1.1584,
"grad_norm": 0.8301047838769273,
"learning_rate": 7.716266096055243e-06,
"loss": 0.3757,
"step": 966
},
{
"epoch": 1.1596,
"grad_norm": 0.6809074767870678,
"learning_rate": 7.710404759773637e-06,
"loss": 0.3914,
"step": 967
},
{
"epoch": 1.1608,
"grad_norm": 0.6868939438344647,
"learning_rate": 7.704538144113082e-06,
"loss": 0.3676,
"step": 968
},
{
"epoch": 1.162,
"grad_norm": 0.9322935418044663,
"learning_rate": 7.698666260500688e-06,
"loss": 0.3685,
"step": 969
},
{
"epoch": 1.1632,
"grad_norm": 0.7536006898712513,
"learning_rate": 7.692789120373824e-06,
"loss": 0.3551,
"step": 970
},
{
"epoch": 1.1644,
"grad_norm": 0.6768285920987577,
"learning_rate": 7.686906735180099e-06,
"loss": 0.3788,
"step": 971
},
{
"epoch": 1.1656,
"grad_norm": 0.9541975514032539,
"learning_rate": 7.681019116377331e-06,
"loss": 0.358,
"step": 972
},
{
"epoch": 1.1668,
"grad_norm": 0.7117880138942121,
"learning_rate": 7.675126275433545e-06,
"loss": 0.3404,
"step": 973
},
{
"epoch": 1.168,
"grad_norm": 0.8113100058600907,
"learning_rate": 7.669228223826926e-06,
"loss": 0.3717,
"step": 974
},
{
"epoch": 1.1692,
"grad_norm": 0.7371901006858377,
"learning_rate": 7.663324973045818e-06,
"loss": 0.3657,
"step": 975
},
{
"epoch": 1.1703999999999999,
"grad_norm": 0.8445637192778912,
"learning_rate": 7.657416534588683e-06,
"loss": 0.3768,
"step": 976
},
{
"epoch": 1.1716,
"grad_norm": 0.7832976634412572,
"learning_rate": 7.651502919964092e-06,
"loss": 0.3751,
"step": 977
},
{
"epoch": 1.1728,
"grad_norm": 0.7545720260541044,
"learning_rate": 7.645584140690702e-06,
"loss": 0.3471,
"step": 978
},
{
"epoch": 1.174,
"grad_norm": 0.757542312561081,
"learning_rate": 7.639660208297221e-06,
"loss": 0.3883,
"step": 979
},
{
"epoch": 1.1752,
"grad_norm": 0.778912506969338,
"learning_rate": 7.633731134322404e-06,
"loss": 0.3485,
"step": 980
},
{
"epoch": 1.1764000000000001,
"grad_norm": 0.7215134613512876,
"learning_rate": 7.6277969303150155e-06,
"loss": 0.3708,
"step": 981
},
{
"epoch": 1.1776,
"grad_norm": 0.8914328197424854,
"learning_rate": 7.6218576078338115e-06,
"loss": 0.3852,
"step": 982
},
{
"epoch": 1.1788,
"grad_norm": 0.8011192388900438,
"learning_rate": 7.615913178447518e-06,
"loss": 0.3886,
"step": 983
},
{
"epoch": 1.18,
"grad_norm": 0.6869306341837543,
"learning_rate": 7.609963653734814e-06,
"loss": 0.4328,
"step": 984
},
{
"epoch": 1.1812,
"grad_norm": 0.7341665596541775,
"learning_rate": 7.604009045284295e-06,
"loss": 0.3694,
"step": 985
},
{
"epoch": 1.1824,
"grad_norm": 0.6894314122362156,
"learning_rate": 7.598049364694466e-06,
"loss": 0.3524,
"step": 986
},
{
"epoch": 1.1836,
"grad_norm": 0.6768969609611811,
"learning_rate": 7.592084623573708e-06,
"loss": 0.3577,
"step": 987
},
{
"epoch": 1.1848,
"grad_norm": 1.1101884884575084,
"learning_rate": 7.586114833540257e-06,
"loss": 0.4519,
"step": 988
},
{
"epoch": 1.186,
"grad_norm": 0.7598191816817575,
"learning_rate": 7.58014000622219e-06,
"loss": 0.3627,
"step": 989
},
{
"epoch": 1.1872,
"grad_norm": 0.8389663840916158,
"learning_rate": 7.574160153257386e-06,
"loss": 0.4267,
"step": 990
},
{
"epoch": 1.1884000000000001,
"grad_norm": 0.9156287377535559,
"learning_rate": 7.568175286293522e-06,
"loss": 0.3308,
"step": 991
},
{
"epoch": 1.1896,
"grad_norm": 0.6640757313900264,
"learning_rate": 7.562185416988039e-06,
"loss": 0.3444,
"step": 992
},
{
"epoch": 1.1908,
"grad_norm": 0.599847264307754,
"learning_rate": 7.556190557008116e-06,
"loss": 0.3806,
"step": 993
},
{
"epoch": 1.192,
"grad_norm": 0.9616461215258202,
"learning_rate": 7.550190718030663e-06,
"loss": 0.3541,
"step": 994
},
{
"epoch": 1.1932,
"grad_norm": 0.8441776250740627,
"learning_rate": 7.54418591174228e-06,
"loss": 0.3863,
"step": 995
},
{
"epoch": 1.1944,
"grad_norm": 0.7066696703759916,
"learning_rate": 7.5381761498392435e-06,
"loss": 0.4136,
"step": 996
},
{
"epoch": 1.1956,
"grad_norm": 1.4471979541043973,
"learning_rate": 7.532161444027488e-06,
"loss": 0.4392,
"step": 997
},
{
"epoch": 1.1968,
"grad_norm": 0.8434073371370987,
"learning_rate": 7.526141806022571e-06,
"loss": 0.3423,
"step": 998
},
{
"epoch": 1.198,
"grad_norm": 0.6112849971278351,
"learning_rate": 7.520117247549661e-06,
"loss": 0.3622,
"step": 999
},
{
"epoch": 1.1992,
"grad_norm": 0.7179746120306058,
"learning_rate": 7.514087780343511e-06,
"loss": 0.3864,
"step": 1000
},
{
"epoch": 1.2004,
"grad_norm": 0.8003679297677685,
"learning_rate": 7.508053416148433e-06,
"loss": 0.3938,
"step": 1001
},
{
"epoch": 1.2016,
"grad_norm": 0.7140930713904701,
"learning_rate": 7.502014166718279e-06,
"loss": 0.3758,
"step": 1002
},
{
"epoch": 1.2028,
"grad_norm": 0.7539111868842224,
"learning_rate": 7.495970043816416e-06,
"loss": 0.372,
"step": 1003
},
{
"epoch": 1.204,
"grad_norm": 0.8013811040128463,
"learning_rate": 7.489921059215703e-06,
"loss": 0.4206,
"step": 1004
},
{
"epoch": 1.2052,
"grad_norm": 0.7880722218778009,
"learning_rate": 7.483867224698471e-06,
"loss": 0.3165,
"step": 1005
},
{
"epoch": 1.2064,
"grad_norm": 0.6192450445561207,
"learning_rate": 7.477808552056496e-06,
"loss": 0.3177,
"step": 1006
},
{
"epoch": 1.2076,
"grad_norm": 0.7471083579197801,
"learning_rate": 7.471745053090976e-06,
"loss": 0.3722,
"step": 1007
},
{
"epoch": 1.2088,
"grad_norm": 0.8928963977684785,
"learning_rate": 7.465676739612514e-06,
"loss": 0.3289,
"step": 1008
},
{
"epoch": 1.21,
"grad_norm": 0.6719186610987602,
"learning_rate": 7.45960362344109e-06,
"loss": 0.3977,
"step": 1009
},
{
"epoch": 1.2112,
"grad_norm": 0.6957395336276705,
"learning_rate": 7.4535257164060324e-06,
"loss": 0.3825,
"step": 1010
},
{
"epoch": 1.2124,
"grad_norm": 1.0590337261075518,
"learning_rate": 7.447443030346011e-06,
"loss": 0.4605,
"step": 1011
},
{
"epoch": 1.2136,
"grad_norm": 0.8773029628812706,
"learning_rate": 7.441355577108998e-06,
"loss": 0.3983,
"step": 1012
},
{
"epoch": 1.2147999999999999,
"grad_norm": 0.6001443478790164,
"learning_rate": 7.4352633685522535e-06,
"loss": 0.3108,
"step": 1013
},
{
"epoch": 1.216,
"grad_norm": 0.7531112806271716,
"learning_rate": 7.4291664165422985e-06,
"loss": 0.3833,
"step": 1014
},
{
"epoch": 1.2172,
"grad_norm": 0.959745691511132,
"learning_rate": 7.423064732954895e-06,
"loss": 0.3533,
"step": 1015
},
{
"epoch": 1.2184,
"grad_norm": 0.7785347741402869,
"learning_rate": 7.4169583296750194e-06,
"loss": 0.3917,
"step": 1016
},
{
"epoch": 1.2196,
"grad_norm": 0.6892992957726527,
"learning_rate": 7.410847218596846e-06,
"loss": 0.3154,
"step": 1017
},
{
"epoch": 1.2208,
"grad_norm": 0.813936607896147,
"learning_rate": 7.40473141162371e-06,
"loss": 0.3171,
"step": 1018
},
{
"epoch": 1.222,
"grad_norm": 0.6367611959400395,
"learning_rate": 7.398610920668102e-06,
"loss": 0.3607,
"step": 1019
},
{
"epoch": 1.2232,
"grad_norm": 0.7242092613300886,
"learning_rate": 7.392485757651634e-06,
"loss": 0.4003,
"step": 1020
},
{
"epoch": 1.2244,
"grad_norm": 1.032522782894471,
"learning_rate": 7.386355934505015e-06,
"loss": 0.3612,
"step": 1021
},
{
"epoch": 1.2256,
"grad_norm": 0.6890815561750541,
"learning_rate": 7.380221463168036e-06,
"loss": 0.351,
"step": 1022
},
{
"epoch": 1.2268,
"grad_norm": 0.9477402344785505,
"learning_rate": 7.374082355589536e-06,
"loss": 0.3858,
"step": 1023
},
{
"epoch": 1.228,
"grad_norm": 0.7603612537684211,
"learning_rate": 7.367938623727389e-06,
"loss": 0.3959,
"step": 1024
},
{
"epoch": 1.2292,
"grad_norm": 0.8307597985316346,
"learning_rate": 7.361790279548476e-06,
"loss": 0.3928,
"step": 1025
},
{
"epoch": 1.2304,
"grad_norm": 0.5788443583312952,
"learning_rate": 7.35563733502866e-06,
"loss": 0.3961,
"step": 1026
},
{
"epoch": 1.2316,
"grad_norm": 0.9461768302713365,
"learning_rate": 7.3494798021527665e-06,
"loss": 0.3584,
"step": 1027
},
{
"epoch": 1.2328000000000001,
"grad_norm": 0.586991002483031,
"learning_rate": 7.3433176929145574e-06,
"loss": 0.3526,
"step": 1028
},
{
"epoch": 1.234,
"grad_norm": 0.7686992286555594,
"learning_rate": 7.337151019316708e-06,
"loss": 0.3833,
"step": 1029
},
{
"epoch": 1.2352,
"grad_norm": 0.7564581079574358,
"learning_rate": 7.330979793370784e-06,
"loss": 0.3652,
"step": 1030
},
{
"epoch": 1.2364,
"grad_norm": 0.7112625971796706,
"learning_rate": 7.324804027097221e-06,
"loss": 0.4124,
"step": 1031
},
{
"epoch": 1.2376,
"grad_norm": 0.9995057993365307,
"learning_rate": 7.318623732525294e-06,
"loss": 0.3662,
"step": 1032
},
{
"epoch": 1.2388,
"grad_norm": 0.7273695409452402,
"learning_rate": 7.312438921693101e-06,
"loss": 0.3799,
"step": 1033
},
{
"epoch": 1.24,
"grad_norm": 0.8687180839158705,
"learning_rate": 7.30624960664754e-06,
"loss": 0.3349,
"step": 1034
},
{
"epoch": 1.2412,
"grad_norm": 0.6256905250880682,
"learning_rate": 7.300055799444273e-06,
"loss": 0.3282,
"step": 1035
},
{
"epoch": 1.2424,
"grad_norm": 0.5248093367290244,
"learning_rate": 7.293857512147723e-06,
"loss": 0.3702,
"step": 1036
},
{
"epoch": 1.2436,
"grad_norm": 1.0915920320931536,
"learning_rate": 7.287654756831031e-06,
"loss": 0.3686,
"step": 1037
},
{
"epoch": 1.2448,
"grad_norm": 0.8692470921961543,
"learning_rate": 7.2814475455760445e-06,
"loss": 0.369,
"step": 1038
},
{
"epoch": 1.246,
"grad_norm": 0.6675243209613981,
"learning_rate": 7.275235890473291e-06,
"loss": 0.3327,
"step": 1039
},
{
"epoch": 1.2472,
"grad_norm": 0.6500840918683873,
"learning_rate": 7.269019803621953e-06,
"loss": 0.3751,
"step": 1040
},
{
"epoch": 1.2484,
"grad_norm": 0.780241307081768,
"learning_rate": 7.262799297129843e-06,
"loss": 0.3441,
"step": 1041
},
{
"epoch": 1.2496,
"grad_norm": 0.6612649498896949,
"learning_rate": 7.256574383113386e-06,
"loss": 0.377,
"step": 1042
},
{
"epoch": 1.2508,
"grad_norm": 0.8111323710996902,
"learning_rate": 7.25034507369759e-06,
"loss": 0.3299,
"step": 1043
},
{
"epoch": 1.252,
"grad_norm": 0.6714520519346864,
"learning_rate": 7.244111381016024e-06,
"loss": 0.3413,
"step": 1044
},
{
"epoch": 1.2532,
"grad_norm": 0.66181182409563,
"learning_rate": 7.237873317210796e-06,
"loss": 0.4043,
"step": 1045
},
{
"epoch": 1.2544,
"grad_norm": 0.9769864480464788,
"learning_rate": 7.231630894432527e-06,
"loss": 0.3618,
"step": 1046
},
{
"epoch": 1.2556,
"grad_norm": 0.577658378024529,
"learning_rate": 7.22538412484033e-06,
"loss": 0.3619,
"step": 1047
},
{
"epoch": 1.2568,
"grad_norm": 0.6250995601800438,
"learning_rate": 7.219133020601783e-06,
"loss": 0.4185,
"step": 1048
},
{
"epoch": 1.258,
"grad_norm": 0.7979134651038914,
"learning_rate": 7.2128775938929095e-06,
"loss": 0.344,
"step": 1049
},
{
"epoch": 1.2591999999999999,
"grad_norm": 0.7108585387713297,
"learning_rate": 7.206617856898149e-06,
"loss": 0.3666,
"step": 1050
},
{
"epoch": 1.2604,
"grad_norm": 0.6246919871951097,
"learning_rate": 7.20035382181034e-06,
"loss": 0.382,
"step": 1051
},
{
"epoch": 1.2616,
"grad_norm": 0.6327129752654392,
"learning_rate": 7.194085500830691e-06,
"loss": 0.3617,
"step": 1052
},
{
"epoch": 1.2628,
"grad_norm": 0.6975567230302412,
"learning_rate": 7.1878129061687595e-06,
"loss": 0.3386,
"step": 1053
},
{
"epoch": 1.264,
"grad_norm": 0.6616189218429642,
"learning_rate": 7.181536050042427e-06,
"loss": 0.3718,
"step": 1054
},
{
"epoch": 1.2652,
"grad_norm": 0.7994721787011563,
"learning_rate": 7.175254944677874e-06,
"loss": 0.3551,
"step": 1055
},
{
"epoch": 1.2664,
"grad_norm": 0.8629057719678181,
"learning_rate": 7.1689696023095625e-06,
"loss": 0.3953,
"step": 1056
},
{
"epoch": 1.2676,
"grad_norm": 0.718329677595062,
"learning_rate": 7.162680035180201e-06,
"loss": 0.4033,
"step": 1057
},
{
"epoch": 1.2688,
"grad_norm": 0.6354682624892466,
"learning_rate": 7.156386255540732e-06,
"loss": 0.3714,
"step": 1058
},
{
"epoch": 1.27,
"grad_norm": 0.6668276424646945,
"learning_rate": 7.150088275650302e-06,
"loss": 0.3203,
"step": 1059
},
{
"epoch": 1.2711999999999999,
"grad_norm": 0.72521665225341,
"learning_rate": 7.143786107776236e-06,
"loss": 0.4102,
"step": 1060
},
{
"epoch": 1.2724,
"grad_norm": 0.8584665375207425,
"learning_rate": 7.137479764194022e-06,
"loss": 0.3413,
"step": 1061
},
{
"epoch": 1.2736,
"grad_norm": 0.6672551792108226,
"learning_rate": 7.131169257187276e-06,
"loss": 0.3959,
"step": 1062
},
{
"epoch": 1.2748,
"grad_norm": 0.9345308578804161,
"learning_rate": 7.1248545990477256e-06,
"loss": 0.388,
"step": 1063
},
{
"epoch": 1.276,
"grad_norm": 0.7103934594557835,
"learning_rate": 7.1185358020751875e-06,
"loss": 0.4021,
"step": 1064
},
{
"epoch": 1.2772000000000001,
"grad_norm": 0.6527878994822974,
"learning_rate": 7.112212878577533e-06,
"loss": 0.32,
"step": 1065
},
{
"epoch": 1.2784,
"grad_norm": 0.7080606252592918,
"learning_rate": 7.1058858408706765e-06,
"loss": 0.3271,
"step": 1066
},
{
"epoch": 1.2796,
"grad_norm": 0.7057398404822831,
"learning_rate": 7.099554701278547e-06,
"loss": 0.3673,
"step": 1067
},
{
"epoch": 1.2808,
"grad_norm": 0.6031692999066002,
"learning_rate": 7.093219472133059e-06,
"loss": 0.3746,
"step": 1068
},
{
"epoch": 1.282,
"grad_norm": 0.7287826765215238,
"learning_rate": 7.086880165774093e-06,
"loss": 0.3295,
"step": 1069
},
{
"epoch": 1.2832,
"grad_norm": 0.7287225834790275,
"learning_rate": 7.080536794549477e-06,
"loss": 0.3855,
"step": 1070
},
{
"epoch": 1.2844,
"grad_norm": 0.7300805167657453,
"learning_rate": 7.0741893708149475e-06,
"loss": 0.4258,
"step": 1071
},
{
"epoch": 1.2856,
"grad_norm": 0.7351527393515873,
"learning_rate": 7.067837906934143e-06,
"loss": 0.3191,
"step": 1072
},
{
"epoch": 1.2868,
"grad_norm": 1.0547881711203557,
"learning_rate": 7.061482415278569e-06,
"loss": 0.4248,
"step": 1073
},
{
"epoch": 1.288,
"grad_norm": 0.8237056579888817,
"learning_rate": 7.055122908227571e-06,
"loss": 0.4068,
"step": 1074
},
{
"epoch": 1.2892000000000001,
"grad_norm": 0.5982682434538817,
"learning_rate": 7.0487593981683246e-06,
"loss": 0.3775,
"step": 1075
},
{
"epoch": 1.2904,
"grad_norm": 0.9285595624077707,
"learning_rate": 7.042391897495795e-06,
"loss": 0.3245,
"step": 1076
},
{
"epoch": 1.2916,
"grad_norm": 0.6974399732840618,
"learning_rate": 7.036020418612724e-06,
"loss": 0.3272,
"step": 1077
},
{
"epoch": 1.2928,
"grad_norm": 0.7805848019900014,
"learning_rate": 7.029644973929604e-06,
"loss": 0.4609,
"step": 1078
},
{
"epoch": 1.294,
"grad_norm": 1.8389823402347827,
"learning_rate": 7.023265575864648e-06,
"loss": 0.333,
"step": 1079
},
{
"epoch": 1.2952,
"grad_norm": 0.6698304708182644,
"learning_rate": 7.016882236843769e-06,
"loss": 0.3761,
"step": 1080
},
{
"epoch": 1.2964,
"grad_norm": 0.624940912268203,
"learning_rate": 7.0104949693005645e-06,
"loss": 0.3563,
"step": 1081
},
{
"epoch": 1.2976,
"grad_norm": 0.7306299665325151,
"learning_rate": 7.00410378567627e-06,
"loss": 0.3585,
"step": 1082
},
{
"epoch": 1.2988,
"grad_norm": 0.7506136762816767,
"learning_rate": 6.997708698419765e-06,
"loss": 0.3945,
"step": 1083
},
{
"epoch": 1.3,
"grad_norm": 0.9178015253144609,
"learning_rate": 6.99130971998752e-06,
"loss": 0.4142,
"step": 1084
},
{
"epoch": 1.3012000000000001,
"grad_norm": 0.6892935272327999,
"learning_rate": 6.98490686284359e-06,
"loss": 0.4072,
"step": 1085
},
{
"epoch": 1.3024,
"grad_norm": 0.7328169467436497,
"learning_rate": 6.978500139459583e-06,
"loss": 0.3496,
"step": 1086
},
{
"epoch": 1.3035999999999999,
"grad_norm": 0.8116153114287361,
"learning_rate": 6.972089562314644e-06,
"loss": 0.3497,
"step": 1087
},
{
"epoch": 1.3048,
"grad_norm": 0.7816556737585988,
"learning_rate": 6.9656751438954115e-06,
"loss": 0.3785,
"step": 1088
},
{
"epoch": 1.306,
"grad_norm": 0.90761307388172,
"learning_rate": 6.959256896696021e-06,
"loss": 0.3568,
"step": 1089
},
{
"epoch": 1.3072,
"grad_norm": 0.7094597042480794,
"learning_rate": 6.952834833218056e-06,
"loss": 0.3982,
"step": 1090
},
{
"epoch": 1.3084,
"grad_norm": 1.0870954344589703,
"learning_rate": 6.946408965970536e-06,
"loss": 0.3143,
"step": 1091
},
{
"epoch": 1.3096,
"grad_norm": 0.8050583154276475,
"learning_rate": 6.939979307469892e-06,
"loss": 0.4171,
"step": 1092
},
{
"epoch": 1.3108,
"grad_norm": 0.9534302722278813,
"learning_rate": 6.933545870239933e-06,
"loss": 0.3715,
"step": 1093
},
{
"epoch": 1.312,
"grad_norm": 0.8014578632311612,
"learning_rate": 6.927108666811837e-06,
"loss": 0.4182,
"step": 1094
},
{
"epoch": 1.3132,
"grad_norm": 0.7484275002727592,
"learning_rate": 6.920667709724113e-06,
"loss": 0.3575,
"step": 1095
},
{
"epoch": 1.3144,
"grad_norm": 0.6297533396897512,
"learning_rate": 6.914223011522581e-06,
"loss": 0.4103,
"step": 1096
},
{
"epoch": 1.3155999999999999,
"grad_norm": 0.7203161071228373,
"learning_rate": 6.90777458476035e-06,
"loss": 0.3903,
"step": 1097
},
{
"epoch": 1.3168,
"grad_norm": 0.8764206309358192,
"learning_rate": 6.901322441997791e-06,
"loss": 0.3761,
"step": 1098
},
{
"epoch": 1.318,
"grad_norm": 0.6961315086355244,
"learning_rate": 6.894866595802509e-06,
"loss": 0.3299,
"step": 1099
},
{
"epoch": 1.3192,
"grad_norm": 0.6722312583926777,
"learning_rate": 6.888407058749331e-06,
"loss": 0.3707,
"step": 1100
},
{
"epoch": 1.3204,
"grad_norm": 1.0779275473959296,
"learning_rate": 6.881943843420268e-06,
"loss": 0.38,
"step": 1101
},
{
"epoch": 1.3216,
"grad_norm": 0.7318789471743649,
"learning_rate": 6.875476962404495e-06,
"loss": 0.3833,
"step": 1102
},
{
"epoch": 1.3228,
"grad_norm": 0.7649222564607404,
"learning_rate": 6.869006428298328e-06,
"loss": 0.375,
"step": 1103
},
{
"epoch": 1.324,
"grad_norm": 0.7714850987988394,
"learning_rate": 6.862532253705199e-06,
"loss": 0.3496,
"step": 1104
},
{
"epoch": 1.3252,
"grad_norm": 0.8190511364844618,
"learning_rate": 6.85605445123563e-06,
"loss": 0.3653,
"step": 1105
},
{
"epoch": 1.3264,
"grad_norm": 0.7049362980205519,
"learning_rate": 6.849573033507213e-06,
"loss": 0.3726,
"step": 1106
},
{
"epoch": 1.3276,
"grad_norm": 0.6662089369343379,
"learning_rate": 6.843088013144575e-06,
"loss": 0.3819,
"step": 1107
},
{
"epoch": 1.3288,
"grad_norm": 0.7412349636048636,
"learning_rate": 6.8365994027793695e-06,
"loss": 0.3407,
"step": 1108
},
{
"epoch": 1.33,
"grad_norm": 0.7062736054277113,
"learning_rate": 6.830107215050232e-06,
"loss": 0.3765,
"step": 1109
},
{
"epoch": 1.3312,
"grad_norm": 0.6640389645030439,
"learning_rate": 6.823611462602777e-06,
"loss": 0.3124,
"step": 1110
},
{
"epoch": 1.3324,
"grad_norm": 0.7540101680669742,
"learning_rate": 6.817112158089554e-06,
"loss": 0.3789,
"step": 1111
},
{
"epoch": 1.3336000000000001,
"grad_norm": 0.7979794931157069,
"learning_rate": 6.8106093141700336e-06,
"loss": 0.3941,
"step": 1112
},
{
"epoch": 1.3348,
"grad_norm": 0.537595104414093,
"learning_rate": 6.804102943510583e-06,
"loss": 0.3814,
"step": 1113
},
{
"epoch": 1.336,
"grad_norm": 0.8221564550672978,
"learning_rate": 6.797593058784437e-06,
"loss": 0.363,
"step": 1114
},
{
"epoch": 1.3372,
"grad_norm": 0.5595789690725599,
"learning_rate": 6.791079672671677e-06,
"loss": 0.3577,
"step": 1115
},
{
"epoch": 1.3384,
"grad_norm": 0.7027718537242316,
"learning_rate": 6.784562797859198e-06,
"loss": 0.3607,
"step": 1116
},
{
"epoch": 1.3396,
"grad_norm": 0.6880247878227063,
"learning_rate": 6.7780424470407004e-06,
"loss": 0.3886,
"step": 1117
},
{
"epoch": 1.3408,
"grad_norm": 0.6649841649934168,
"learning_rate": 6.771518632916645e-06,
"loss": 0.3524,
"step": 1118
},
{
"epoch": 1.342,
"grad_norm": 0.6062278650745143,
"learning_rate": 6.7649913681942455e-06,
"loss": 0.37,
"step": 1119
},
{
"epoch": 1.3432,
"grad_norm": 0.7648535017093944,
"learning_rate": 6.758460665587437e-06,
"loss": 0.3828,
"step": 1120
},
{
"epoch": 1.3444,
"grad_norm": 0.7353130385761478,
"learning_rate": 6.751926537816846e-06,
"loss": 0.3802,
"step": 1121
},
{
"epoch": 1.3456000000000001,
"grad_norm": 0.6829839541629846,
"learning_rate": 6.745388997609774e-06,
"loss": 0.3468,
"step": 1122
},
{
"epoch": 1.3468,
"grad_norm": 0.6626085923078854,
"learning_rate": 6.738848057700169e-06,
"loss": 0.3391,
"step": 1123
},
{
"epoch": 1.3479999999999999,
"grad_norm": 0.6493361092019838,
"learning_rate": 6.732303730828601e-06,
"loss": 0.3323,
"step": 1124
},
{
"epoch": 1.3492,
"grad_norm": 0.7136963584973568,
"learning_rate": 6.725756029742234e-06,
"loss": 0.4035,
"step": 1125
},
{
"epoch": 1.3504,
"grad_norm": 1.2553206696003358,
"learning_rate": 6.7192049671948115e-06,
"loss": 0.4359,
"step": 1126
},
{
"epoch": 1.3516,
"grad_norm": 0.7485848482212277,
"learning_rate": 6.712650555946616e-06,
"loss": 0.3622,
"step": 1127
},
{
"epoch": 1.3528,
"grad_norm": 0.6524261880658018,
"learning_rate": 6.706092808764459e-06,
"loss": 0.3359,
"step": 1128
},
{
"epoch": 1.354,
"grad_norm": 0.7518626791397351,
"learning_rate": 6.699531738421648e-06,
"loss": 0.3037,
"step": 1129
},
{
"epoch": 1.3552,
"grad_norm": 0.651976318525,
"learning_rate": 6.692967357697961e-06,
"loss": 0.3753,
"step": 1130
},
{
"epoch": 1.3564,
"grad_norm": 0.6783473399798922,
"learning_rate": 6.6863996793796286e-06,
"loss": 0.2768,
"step": 1131
},
{
"epoch": 1.3576,
"grad_norm": 0.6457224804377565,
"learning_rate": 6.6798287162593e-06,
"loss": 0.3633,
"step": 1132
},
{
"epoch": 1.3588,
"grad_norm": 0.8346817618056587,
"learning_rate": 6.6732544811360255e-06,
"loss": 0.3956,
"step": 1133
},
{
"epoch": 1.3599999999999999,
"grad_norm": 0.7587086980457096,
"learning_rate": 6.666676986815227e-06,
"loss": 0.4225,
"step": 1134
},
{
"epoch": 1.3612,
"grad_norm": 0.7615916857787209,
"learning_rate": 6.660096246108677e-06,
"loss": 0.3726,
"step": 1135
},
{
"epoch": 1.3624,
"grad_norm": 0.7143833797662076,
"learning_rate": 6.653512271834468e-06,
"loss": 0.3416,
"step": 1136
},
{
"epoch": 1.3636,
"grad_norm": 0.6710865184361733,
"learning_rate": 6.646925076816994e-06,
"loss": 0.3524,
"step": 1137
},
{
"epoch": 1.3648,
"grad_norm": 0.889853447367123,
"learning_rate": 6.640334673886921e-06,
"loss": 0.4504,
"step": 1138
},
{
"epoch": 1.366,
"grad_norm": 0.8467403347287523,
"learning_rate": 6.633741075881163e-06,
"loss": 0.3715,
"step": 1139
},
{
"epoch": 1.3672,
"grad_norm": 0.7031728267698215,
"learning_rate": 6.627144295642859e-06,
"loss": 0.4213,
"step": 1140
},
{
"epoch": 1.3684,
"grad_norm": 0.9311379749888988,
"learning_rate": 6.6205443460213445e-06,
"loss": 0.3254,
"step": 1141
},
{
"epoch": 1.3696,
"grad_norm": 0.726873127221127,
"learning_rate": 6.613941239872129e-06,
"loss": 0.3835,
"step": 1142
},
{
"epoch": 1.3708,
"grad_norm": 0.7192891691066637,
"learning_rate": 6.607334990056873e-06,
"loss": 0.3485,
"step": 1143
},
{
"epoch": 1.3719999999999999,
"grad_norm": 0.9213884417729308,
"learning_rate": 6.600725609443356e-06,
"loss": 0.376,
"step": 1144
},
{
"epoch": 1.3732,
"grad_norm": 0.9218707865555437,
"learning_rate": 6.594113110905458e-06,
"loss": 0.385,
"step": 1145
},
{
"epoch": 1.3744,
"grad_norm": 0.7571640804730355,
"learning_rate": 6.587497507323132e-06,
"loss": 0.3988,
"step": 1146
},
{
"epoch": 1.3756,
"grad_norm": 0.7543854326432067,
"learning_rate": 6.580878811582379e-06,
"loss": 0.3567,
"step": 1147
},
{
"epoch": 1.3768,
"grad_norm": 0.7364183138914716,
"learning_rate": 6.574257036575224e-06,
"loss": 0.3662,
"step": 1148
},
{
"epoch": 1.3780000000000001,
"grad_norm": 0.6457768212905267,
"learning_rate": 6.567632195199686e-06,
"loss": 0.3516,
"step": 1149
},
{
"epoch": 1.3792,
"grad_norm": 0.8561178152745103,
"learning_rate": 6.5610043003597615e-06,
"loss": 0.349,
"step": 1150
},
{
"epoch": 1.3804,
"grad_norm": 0.6761235934877868,
"learning_rate": 6.554373364965392e-06,
"loss": 0.331,
"step": 1151
},
{
"epoch": 1.3816,
"grad_norm": 0.7438758337342172,
"learning_rate": 6.547739401932443e-06,
"loss": 0.3594,
"step": 1152
},
{
"epoch": 1.3828,
"grad_norm": 0.8202372420467647,
"learning_rate": 6.541102424182676e-06,
"loss": 0.3758,
"step": 1153
},
{
"epoch": 1.384,
"grad_norm": 0.7678936566930825,
"learning_rate": 6.5344624446437234e-06,
"loss": 0.4775,
"step": 1154
},
{
"epoch": 1.3852,
"grad_norm": 1.1565403867631232,
"learning_rate": 6.527819476249066e-06,
"loss": 0.3591,
"step": 1155
},
{
"epoch": 1.3864,
"grad_norm": 0.8388398477850556,
"learning_rate": 6.521173531938011e-06,
"loss": 0.3717,
"step": 1156
},
{
"epoch": 1.3876,
"grad_norm": 0.7738271351254377,
"learning_rate": 6.514524624655654e-06,
"loss": 0.433,
"step": 1157
},
{
"epoch": 1.3888,
"grad_norm": 0.8550307182745539,
"learning_rate": 6.507872767352863e-06,
"loss": 0.4124,
"step": 1158
},
{
"epoch": 1.3900000000000001,
"grad_norm": 0.6737280107790758,
"learning_rate": 6.50121797298626e-06,
"loss": 0.3541,
"step": 1159
},
{
"epoch": 1.3912,
"grad_norm": 0.6772025974463309,
"learning_rate": 6.494560254518179e-06,
"loss": 0.3252,
"step": 1160
},
{
"epoch": 1.3924,
"grad_norm": 0.8492388641554832,
"learning_rate": 6.487899624916654e-06,
"loss": 0.3584,
"step": 1161
},
{
"epoch": 1.3936,
"grad_norm": 0.7008085006126243,
"learning_rate": 6.481236097155389e-06,
"loss": 0.3817,
"step": 1162
},
{
"epoch": 1.3948,
"grad_norm": 0.7244399295585411,
"learning_rate": 6.4745696842137305e-06,
"loss": 0.4264,
"step": 1163
},
{
"epoch": 1.396,
"grad_norm": 1.0335788875610432,
"learning_rate": 6.467900399076651e-06,
"loss": 0.3547,
"step": 1164
},
{
"epoch": 1.3972,
"grad_norm": 0.7211174327038158,
"learning_rate": 6.461228254734711e-06,
"loss": 0.4411,
"step": 1165
},
{
"epoch": 1.3984,
"grad_norm": 0.921793662794543,
"learning_rate": 6.454553264184041e-06,
"loss": 0.3671,
"step": 1166
},
{
"epoch": 1.3996,
"grad_norm": 0.7715514480584313,
"learning_rate": 6.447875440426319e-06,
"loss": 0.3568,
"step": 1167
},
{
"epoch": 1.4008,
"grad_norm": 0.5972952828385716,
"learning_rate": 6.441194796468739e-06,
"loss": 0.326,
"step": 1168
},
{
"epoch": 1.4020000000000001,
"grad_norm": 0.7311666664328393,
"learning_rate": 6.434511345323988e-06,
"loss": 0.4186,
"step": 1169
},
{
"epoch": 1.4032,
"grad_norm": 0.908174979421926,
"learning_rate": 6.427825100010225e-06,
"loss": 0.4534,
"step": 1170
},
{
"epoch": 1.4043999999999999,
"grad_norm": 0.7605400342757017,
"learning_rate": 6.421136073551047e-06,
"loss": 0.3514,
"step": 1171
},
{
"epoch": 1.4056,
"grad_norm": 0.6356396617385787,
"learning_rate": 6.414444278975465e-06,
"loss": 0.3188,
"step": 1172
},
{
"epoch": 1.4068,
"grad_norm": 0.7310432226921798,
"learning_rate": 6.407749729317892e-06,
"loss": 0.3996,
"step": 1173
},
{
"epoch": 1.408,
"grad_norm": 0.7770188845351006,
"learning_rate": 6.401052437618098e-06,
"loss": 0.3812,
"step": 1174
},
{
"epoch": 1.4092,
"grad_norm": 0.9504399553587416,
"learning_rate": 6.394352416921201e-06,
"loss": 0.3195,
"step": 1175
},
{
"epoch": 1.4104,
"grad_norm": 0.8543460333823263,
"learning_rate": 6.387649680277629e-06,
"loss": 0.3688,
"step": 1176
},
{
"epoch": 1.4116,
"grad_norm": 0.7226057362290352,
"learning_rate": 6.380944240743101e-06,
"loss": 0.4119,
"step": 1177
},
{
"epoch": 1.4128,
"grad_norm": 0.8326667012927984,
"learning_rate": 6.374236111378605e-06,
"loss": 0.3671,
"step": 1178
},
{
"epoch": 1.414,
"grad_norm": 0.5712159115534434,
"learning_rate": 6.3675253052503645e-06,
"loss": 0.3576,
"step": 1179
},
{
"epoch": 1.4152,
"grad_norm": 0.7708010441035443,
"learning_rate": 6.360811835429817e-06,
"loss": 0.3495,
"step": 1180
},
{
"epoch": 1.4163999999999999,
"grad_norm": 0.6088997224976647,
"learning_rate": 6.35409571499359e-06,
"loss": 0.376,
"step": 1181
},
{
"epoch": 1.4176,
"grad_norm": 1.2022858169692507,
"learning_rate": 6.347376957023471e-06,
"loss": 0.426,
"step": 1182
},
{
"epoch": 1.4188,
"grad_norm": 0.7263998904585773,
"learning_rate": 6.340655574606391e-06,
"loss": 0.3611,
"step": 1183
},
{
"epoch": 1.42,
"grad_norm": 0.6383366517977046,
"learning_rate": 6.333931580834387e-06,
"loss": 0.3115,
"step": 1184
},
{
"epoch": 1.4212,
"grad_norm": 0.6786880192165294,
"learning_rate": 6.327204988804583e-06,
"loss": 0.33,
"step": 1185
},
{
"epoch": 1.4224,
"grad_norm": 0.7557481689617757,
"learning_rate": 6.320475811619167e-06,
"loss": 0.4193,
"step": 1186
},
{
"epoch": 1.4236,
"grad_norm": 1.0002494568284233,
"learning_rate": 6.313744062385363e-06,
"loss": 0.4159,
"step": 1187
},
{
"epoch": 1.4248,
"grad_norm": 0.6528818406473923,
"learning_rate": 6.307009754215397e-06,
"loss": 0.3959,
"step": 1188
},
{
"epoch": 1.426,
"grad_norm": 0.6077503970269207,
"learning_rate": 6.30027290022649e-06,
"loss": 0.369,
"step": 1189
},
{
"epoch": 1.4272,
"grad_norm": 0.7968142625176655,
"learning_rate": 6.2935335135408135e-06,
"loss": 0.3264,
"step": 1190
},
{
"epoch": 1.4284,
"grad_norm": 0.6438791237513256,
"learning_rate": 6.286791607285478e-06,
"loss": 0.3692,
"step": 1191
},
{
"epoch": 1.4296,
"grad_norm": 0.6319703290915275,
"learning_rate": 6.2800471945925e-06,
"loss": 0.3781,
"step": 1192
},
{
"epoch": 1.4308,
"grad_norm": 0.7292699874403744,
"learning_rate": 6.2733002885987734e-06,
"loss": 0.3472,
"step": 1193
},
{
"epoch": 1.432,
"grad_norm": 0.716537548653806,
"learning_rate": 6.2665509024460554e-06,
"loss": 0.3346,
"step": 1194
},
{
"epoch": 1.4332,
"grad_norm": 0.7254538839223775,
"learning_rate": 6.259799049280932e-06,
"loss": 0.3865,
"step": 1195
},
{
"epoch": 1.4344000000000001,
"grad_norm": 0.4354079145592185,
"learning_rate": 6.253044742254791e-06,
"loss": 0.4157,
"step": 1196
},
{
"epoch": 1.4356,
"grad_norm": 0.8093613815799889,
"learning_rate": 6.246287994523805e-06,
"loss": 0.4188,
"step": 1197
},
{
"epoch": 1.4368,
"grad_norm": 0.9464374458053828,
"learning_rate": 6.239528819248898e-06,
"loss": 0.4217,
"step": 1198
},
{
"epoch": 1.438,
"grad_norm": 0.7155029613997761,
"learning_rate": 6.232767229595719e-06,
"loss": 0.31,
"step": 1199
},
{
"epoch": 1.4392,
"grad_norm": 0.632484255893594,
"learning_rate": 6.226003238734628e-06,
"loss": 0.3858,
"step": 1200
},
{
"epoch": 1.4404,
"grad_norm": 0.784997490089277,
"learning_rate": 6.219236859840656e-06,
"loss": 0.3806,
"step": 1201
},
{
"epoch": 1.4416,
"grad_norm": 0.6632492070166573,
"learning_rate": 6.2124681060934866e-06,
"loss": 0.3595,
"step": 1202
},
{
"epoch": 1.4428,
"grad_norm": 0.9291193454809028,
"learning_rate": 6.205696990677431e-06,
"loss": 0.3787,
"step": 1203
},
{
"epoch": 1.444,
"grad_norm": 0.8715013278507021,
"learning_rate": 6.1989235267813964e-06,
"loss": 0.3594,
"step": 1204
},
{
"epoch": 1.4452,
"grad_norm": 1.0347436368864233,
"learning_rate": 6.192147727598869e-06,
"loss": 0.3959,
"step": 1205
},
{
"epoch": 1.4464000000000001,
"grad_norm": 0.744098627925034,
"learning_rate": 6.185369606327882e-06,
"loss": 0.3486,
"step": 1206
},
{
"epoch": 1.4476,
"grad_norm": 0.6242531215774465,
"learning_rate": 6.178589176170991e-06,
"loss": 0.3741,
"step": 1207
},
{
"epoch": 1.4487999999999999,
"grad_norm": 0.7480709567517083,
"learning_rate": 6.171806450335248e-06,
"loss": 0.3543,
"step": 1208
},
{
"epoch": 1.45,
"grad_norm": 0.5376687673209407,
"learning_rate": 6.16502144203218e-06,
"loss": 0.3457,
"step": 1209
},
{
"epoch": 1.4512,
"grad_norm": 0.70114689334537,
"learning_rate": 6.1582341644777575e-06,
"loss": 0.3573,
"step": 1210
},
{
"epoch": 1.4524,
"grad_norm": 0.6571979928634565,
"learning_rate": 6.151444630892372e-06,
"loss": 0.3396,
"step": 1211
},
{
"epoch": 1.4536,
"grad_norm": 0.6920493085794216,
"learning_rate": 6.144652854500806e-06,
"loss": 0.3357,
"step": 1212
},
{
"epoch": 1.4548,
"grad_norm": 0.7033776653927278,
"learning_rate": 6.1378588485322145e-06,
"loss": 0.3989,
"step": 1213
},
{
"epoch": 1.456,
"grad_norm": 0.7047126683096677,
"learning_rate": 6.131062626220094e-06,
"loss": 0.4078,
"step": 1214
},
{
"epoch": 1.4572,
"grad_norm": 0.718060966900929,
"learning_rate": 6.124264200802259e-06,
"loss": 0.3825,
"step": 1215
},
{
"epoch": 1.4584,
"grad_norm": 0.7077756480360878,
"learning_rate": 6.117463585520813e-06,
"loss": 0.3193,
"step": 1216
},
{
"epoch": 1.4596,
"grad_norm": 0.6788770127221928,
"learning_rate": 6.110660793622127e-06,
"loss": 0.3528,
"step": 1217
},
{
"epoch": 1.4607999999999999,
"grad_norm": 0.7456684378554623,
"learning_rate": 6.103855838356813e-06,
"loss": 0.3292,
"step": 1218
},
{
"epoch": 1.462,
"grad_norm": 0.672512093293969,
"learning_rate": 6.097048732979691e-06,
"loss": 0.3654,
"step": 1219
},
{
"epoch": 1.4632,
"grad_norm": 0.6972026547412433,
"learning_rate": 6.090239490749775e-06,
"loss": 0.3361,
"step": 1220
},
{
"epoch": 1.4644,
"grad_norm": 0.5634157281843526,
"learning_rate": 6.083428124930239e-06,
"loss": 0.4133,
"step": 1221
},
{
"epoch": 1.4656,
"grad_norm": 0.592462973921223,
"learning_rate": 6.076614648788392e-06,
"loss": 0.3324,
"step": 1222
},
{
"epoch": 1.4668,
"grad_norm": 0.8683972178415108,
"learning_rate": 6.069799075595658e-06,
"loss": 0.3678,
"step": 1223
},
{
"epoch": 1.468,
"grad_norm": 0.6534194685112802,
"learning_rate": 6.062981418627539e-06,
"loss": 0.3763,
"step": 1224
},
{
"epoch": 1.4692,
"grad_norm": 0.8389581294277907,
"learning_rate": 6.0561616911636025e-06,
"loss": 0.3634,
"step": 1225
},
{
"epoch": 1.4704,
"grad_norm": 0.9384932293025154,
"learning_rate": 6.049339906487443e-06,
"loss": 0.3842,
"step": 1226
},
{
"epoch": 1.4716,
"grad_norm": 0.5040664792367159,
"learning_rate": 6.042516077886669e-06,
"loss": 0.3355,
"step": 1227
},
{
"epoch": 1.4727999999999999,
"grad_norm": 0.7858403545858635,
"learning_rate": 6.035690218652861e-06,
"loss": 0.3696,
"step": 1228
},
{
"epoch": 1.474,
"grad_norm": 0.7760025405232468,
"learning_rate": 6.028862342081564e-06,
"loss": 0.3617,
"step": 1229
},
{
"epoch": 1.4752,
"grad_norm": 0.7066784275878706,
"learning_rate": 6.022032461472247e-06,
"loss": 0.3101,
"step": 1230
},
{
"epoch": 1.4764,
"grad_norm": 0.8143122926090074,
"learning_rate": 6.015200590128284e-06,
"loss": 0.3757,
"step": 1231
},
{
"epoch": 1.4776,
"grad_norm": 0.6254043126892769,
"learning_rate": 6.008366741356926e-06,
"loss": 0.3349,
"step": 1232
},
{
"epoch": 1.4788000000000001,
"grad_norm": 0.8669731221726169,
"learning_rate": 6.001530928469277e-06,
"loss": 0.3851,
"step": 1233
},
{
"epoch": 1.48,
"grad_norm": 0.7402131492192554,
"learning_rate": 5.9946931647802645e-06,
"loss": 0.3697,
"step": 1234
},
{
"epoch": 1.4812,
"grad_norm": 0.6059249425521024,
"learning_rate": 5.987853463608618e-06,
"loss": 0.338,
"step": 1235
},
{
"epoch": 1.4824,
"grad_norm": 0.6633417872009122,
"learning_rate": 5.981011838276841e-06,
"loss": 0.3662,
"step": 1236
},
{
"epoch": 1.4836,
"grad_norm": 0.685218150111119,
"learning_rate": 5.974168302111183e-06,
"loss": 0.3394,
"step": 1237
},
{
"epoch": 1.4848,
"grad_norm": 0.9291193453346511,
"learning_rate": 5.967322868441616e-06,
"loss": 0.3642,
"step": 1238
},
{
"epoch": 1.486,
"grad_norm": 0.6374719345191574,
"learning_rate": 5.9604755506018105e-06,
"loss": 0.3531,
"step": 1239
},
{
"epoch": 1.4872,
"grad_norm": 0.9044363297242491,
"learning_rate": 5.953626361929102e-06,
"loss": 0.3592,
"step": 1240
},
{
"epoch": 1.4884,
"grad_norm": 0.7307660286250603,
"learning_rate": 5.946775315764476e-06,
"loss": 0.3814,
"step": 1241
},
{
"epoch": 1.4896,
"grad_norm": 0.8527614187957432,
"learning_rate": 5.939922425452531e-06,
"loss": 0.3598,
"step": 1242
},
{
"epoch": 1.4908000000000001,
"grad_norm": 0.7253464321900835,
"learning_rate": 5.93306770434146e-06,
"loss": 0.3622,
"step": 1243
},
{
"epoch": 1.492,
"grad_norm": 0.64036599467131,
"learning_rate": 5.926211165783021e-06,
"loss": 0.3526,
"step": 1244
},
{
"epoch": 1.4932,
"grad_norm": 0.6456050888480263,
"learning_rate": 5.919352823132515e-06,
"loss": 0.3775,
"step": 1245
},
{
"epoch": 1.4944,
"grad_norm": 0.7712672346741274,
"learning_rate": 5.912492689748753e-06,
"loss": 0.3046,
"step": 1246
},
{
"epoch": 1.4956,
"grad_norm": 0.9424246281851408,
"learning_rate": 5.905630778994036e-06,
"loss": 0.3891,
"step": 1247
},
{
"epoch": 1.4968,
"grad_norm": 0.6938655105975465,
"learning_rate": 5.898767104234128e-06,
"loss": 0.3762,
"step": 1248
},
{
"epoch": 1.498,
"grad_norm": 0.8990847386268502,
"learning_rate": 5.891901678838227e-06,
"loss": 0.357,
"step": 1249
},
{
"epoch": 1.4992,
"grad_norm": 0.6317877683756148,
"learning_rate": 5.88503451617894e-06,
"loss": 0.3691,
"step": 1250
},
{
"epoch": 1.5004,
"grad_norm": 0.8237321327678571,
"learning_rate": 5.878165629632262e-06,
"loss": 0.4498,
"step": 1251
},
{
"epoch": 1.5016,
"grad_norm": 0.6789451743297871,
"learning_rate": 5.8712950325775416e-06,
"loss": 0.412,
"step": 1252
},
{
"epoch": 1.5028000000000001,
"grad_norm": 1.0067602921409051,
"learning_rate": 5.864422738397465e-06,
"loss": 0.3534,
"step": 1253
},
{
"epoch": 1.504,
"grad_norm": 0.7396124619343574,
"learning_rate": 5.857548760478015e-06,
"loss": 0.3721,
"step": 1254
},
{
"epoch": 1.5051999999999999,
"grad_norm": 0.7295177627786567,
"learning_rate": 5.85067311220846e-06,
"loss": 0.3528,
"step": 1255
},
{
"epoch": 1.5064,
"grad_norm": 0.6303626223682874,
"learning_rate": 5.843795806981325e-06,
"loss": 0.3454,
"step": 1256
},
{
"epoch": 1.5076,
"grad_norm": 0.7694375021355077,
"learning_rate": 5.836916858192353e-06,
"loss": 0.3733,
"step": 1257
},
{
"epoch": 1.5088,
"grad_norm": 0.8762992140212132,
"learning_rate": 5.830036279240497e-06,
"loss": 0.3588,
"step": 1258
},
{
"epoch": 1.51,
"grad_norm": 0.6223279354472662,
"learning_rate": 5.823154083527884e-06,
"loss": 0.4157,
"step": 1259
},
{
"epoch": 1.5112,
"grad_norm": 0.7729986465156738,
"learning_rate": 5.816270284459783e-06,
"loss": 0.4234,
"step": 1260
},
{
"epoch": 1.5124,
"grad_norm": 0.6420126441012368,
"learning_rate": 5.809384895444594e-06,
"loss": 0.3195,
"step": 1261
},
{
"epoch": 1.5135999999999998,
"grad_norm": 0.9579203581327315,
"learning_rate": 5.802497929893813e-06,
"loss": 0.4,
"step": 1262
},
{
"epoch": 1.5148000000000001,
"grad_norm": 0.7875086609555818,
"learning_rate": 5.795609401222001e-06,
"loss": 0.3846,
"step": 1263
},
{
"epoch": 1.516,
"grad_norm": 0.6143968174103723,
"learning_rate": 5.7887193228467735e-06,
"loss": 0.3892,
"step": 1264
},
{
"epoch": 1.5171999999999999,
"grad_norm": 0.7361354880197091,
"learning_rate": 5.781827708188753e-06,
"loss": 0.3367,
"step": 1265
},
{
"epoch": 1.5184,
"grad_norm": 0.8462867730893091,
"learning_rate": 5.774934570671562e-06,
"loss": 0.3804,
"step": 1266
},
{
"epoch": 1.5196,
"grad_norm": 0.6487685981561768,
"learning_rate": 5.768039923721791e-06,
"loss": 0.3255,
"step": 1267
},
{
"epoch": 1.5208,
"grad_norm": 0.742673108375301,
"learning_rate": 5.761143780768962e-06,
"loss": 0.4054,
"step": 1268
},
{
"epoch": 1.522,
"grad_norm": 0.9098834920225984,
"learning_rate": 5.7542461552455165e-06,
"loss": 0.3707,
"step": 1269
},
{
"epoch": 1.5232,
"grad_norm": 0.7389053135905564,
"learning_rate": 5.747347060586787e-06,
"loss": 0.2892,
"step": 1270
},
{
"epoch": 1.5244,
"grad_norm": 0.9350575271225507,
"learning_rate": 5.740446510230959e-06,
"loss": 0.3456,
"step": 1271
},
{
"epoch": 1.5255999999999998,
"grad_norm": 0.6145579007188433,
"learning_rate": 5.733544517619064e-06,
"loss": 0.3551,
"step": 1272
},
{
"epoch": 1.5268000000000002,
"grad_norm": 0.8581826213907329,
"learning_rate": 5.726641096194932e-06,
"loss": 0.4128,
"step": 1273
},
{
"epoch": 1.528,
"grad_norm": 0.7045167582080664,
"learning_rate": 5.719736259405182e-06,
"loss": 0.327,
"step": 1274
},
{
"epoch": 1.5292,
"grad_norm": 0.8774011198713177,
"learning_rate": 5.7128300206991925e-06,
"loss": 0.3712,
"step": 1275
},
{
"epoch": 1.5304,
"grad_norm": 0.7622915175298381,
"learning_rate": 5.705922393529065e-06,
"loss": 0.3722,
"step": 1276
},
{
"epoch": 1.5316,
"grad_norm": 0.652088157634716,
"learning_rate": 5.69901339134961e-06,
"loss": 0.3693,
"step": 1277
},
{
"epoch": 1.5328,
"grad_norm": 0.8528148500687448,
"learning_rate": 5.692103027618321e-06,
"loss": 0.4055,
"step": 1278
},
{
"epoch": 1.534,
"grad_norm": 0.7717162478145243,
"learning_rate": 5.685191315795331e-06,
"loss": 0.2973,
"step": 1279
},
{
"epoch": 1.5352000000000001,
"grad_norm": 0.7609402684231305,
"learning_rate": 5.678278269343411e-06,
"loss": 0.399,
"step": 1280
},
{
"epoch": 1.5364,
"grad_norm": 0.8111626966854636,
"learning_rate": 5.671363901727927e-06,
"loss": 0.3509,
"step": 1281
},
{
"epoch": 1.5375999999999999,
"grad_norm": 0.6569605004103629,
"learning_rate": 5.664448226416815e-06,
"loss": 0.5169,
"step": 1282
},
{
"epoch": 1.5388,
"grad_norm": 2.692037016176556,
"learning_rate": 5.657531256880565e-06,
"loss": 0.363,
"step": 1283
},
{
"epoch": 1.54,
"grad_norm": 0.8745751748912841,
"learning_rate": 5.650613006592185e-06,
"loss": 0.3516,
"step": 1284
},
{
"epoch": 1.5412,
"grad_norm": 0.6577865817614307,
"learning_rate": 5.643693489027172e-06,
"loss": 0.3445,
"step": 1285
},
{
"epoch": 1.5424,
"grad_norm": 0.7367412223227412,
"learning_rate": 5.636772717663501e-06,
"loss": 0.3841,
"step": 1286
},
{
"epoch": 1.5436,
"grad_norm": 0.5415417472443241,
"learning_rate": 5.629850705981584e-06,
"loss": 0.4081,
"step": 1287
},
{
"epoch": 1.5448,
"grad_norm": 0.8365252798699305,
"learning_rate": 5.622927467464247e-06,
"loss": 0.3341,
"step": 1288
},
{
"epoch": 1.546,
"grad_norm": 0.8411628521673272,
"learning_rate": 5.6160030155967116e-06,
"loss": 0.3683,
"step": 1289
},
{
"epoch": 1.5472000000000001,
"grad_norm": 0.6151940224392339,
"learning_rate": 5.609077363866555e-06,
"loss": 0.4015,
"step": 1290
},
{
"epoch": 1.5484,
"grad_norm": 0.5222505561323959,
"learning_rate": 5.602150525763701e-06,
"loss": 0.4017,
"step": 1291
},
{
"epoch": 1.5495999999999999,
"grad_norm": 0.7883036369562284,
"learning_rate": 5.595222514780379e-06,
"loss": 0.3641,
"step": 1292
},
{
"epoch": 1.5508,
"grad_norm": 0.6860084374667428,
"learning_rate": 5.588293344411097e-06,
"loss": 0.3847,
"step": 1293
},
{
"epoch": 1.552,
"grad_norm": 0.8567832458258418,
"learning_rate": 5.581363028152633e-06,
"loss": 0.4141,
"step": 1294
},
{
"epoch": 1.5532,
"grad_norm": 0.5880792191950815,
"learning_rate": 5.574431579503991e-06,
"loss": 0.3646,
"step": 1295
},
{
"epoch": 1.5544,
"grad_norm": 0.7028669215647632,
"learning_rate": 5.5674990119663794e-06,
"loss": 0.378,
"step": 1296
},
{
"epoch": 1.5556,
"grad_norm": 0.8831268166424471,
"learning_rate": 5.560565339043188e-06,
"loss": 0.3499,
"step": 1297
},
{
"epoch": 1.5568,
"grad_norm": 0.7115024955227263,
"learning_rate": 5.55363057423996e-06,
"loss": 0.4081,
"step": 1298
},
{
"epoch": 1.558,
"grad_norm": 0.625357106348798,
"learning_rate": 5.546694731064364e-06,
"loss": 0.3936,
"step": 1299
},
{
"epoch": 1.5592000000000001,
"grad_norm": 0.6699079235290315,
"learning_rate": 5.5397578230261715e-06,
"loss": 0.339,
"step": 1300
},
{
"epoch": 1.5604,
"grad_norm": 0.6793779504011647,
"learning_rate": 5.532819863637223e-06,
"loss": 0.3724,
"step": 1301
},
{
"epoch": 1.5615999999999999,
"grad_norm": 0.888758385802367,
"learning_rate": 5.525880866411414e-06,
"loss": 0.3867,
"step": 1302
},
{
"epoch": 1.5628,
"grad_norm": 0.7159500214806575,
"learning_rate": 5.5189408448646565e-06,
"loss": 0.3485,
"step": 1303
},
{
"epoch": 1.564,
"grad_norm": 0.7909499320665346,
"learning_rate": 5.511999812514857e-06,
"loss": 0.327,
"step": 1304
},
{
"epoch": 1.5652,
"grad_norm": 0.4607495008234559,
"learning_rate": 5.505057782881896e-06,
"loss": 0.3798,
"step": 1305
},
{
"epoch": 1.5664,
"grad_norm": 0.8163148927950353,
"learning_rate": 5.4981147694875924e-06,
"loss": 0.3787,
"step": 1306
},
{
"epoch": 1.5676,
"grad_norm": 0.7148455511113876,
"learning_rate": 5.491170785855681e-06,
"loss": 0.3389,
"step": 1307
},
{
"epoch": 1.5688,
"grad_norm": 0.6141228071063224,
"learning_rate": 5.484225845511791e-06,
"loss": 0.3861,
"step": 1308
},
{
"epoch": 1.5699999999999998,
"grad_norm": 0.601331262844595,
"learning_rate": 5.477279961983408e-06,
"loss": 0.3651,
"step": 1309
},
{
"epoch": 1.5712000000000002,
"grad_norm": 0.6450400088944953,
"learning_rate": 5.470333148799862e-06,
"loss": 0.3452,
"step": 1310
},
{
"epoch": 1.5724,
"grad_norm": 0.7955061887079758,
"learning_rate": 5.463385419492288e-06,
"loss": 0.3712,
"step": 1311
},
{
"epoch": 1.5735999999999999,
"grad_norm": 1.2772596624595498,
"learning_rate": 5.456436787593609e-06,
"loss": 0.3929,
"step": 1312
},
{
"epoch": 1.5748,
"grad_norm": 0.7739516241543868,
"learning_rate": 5.449487266638504e-06,
"loss": 0.4012,
"step": 1313
},
{
"epoch": 1.576,
"grad_norm": 0.6022161909924404,
"learning_rate": 5.442536870163386e-06,
"loss": 0.3103,
"step": 1314
},
{
"epoch": 1.5772,
"grad_norm": 0.6439186606077979,
"learning_rate": 5.43558561170637e-06,
"loss": 0.3786,
"step": 1315
},
{
"epoch": 1.5784,
"grad_norm": 0.8250939800848203,
"learning_rate": 5.428633504807253e-06,
"loss": 0.3551,
"step": 1316
},
{
"epoch": 1.5796000000000001,
"grad_norm": 0.6149072174639775,
"learning_rate": 5.421680563007486e-06,
"loss": 0.4054,
"step": 1317
},
{
"epoch": 1.5808,
"grad_norm": 0.7852175317750469,
"learning_rate": 5.414726799850141e-06,
"loss": 0.3669,
"step": 1318
},
{
"epoch": 1.5819999999999999,
"grad_norm": 0.6545875908685911,
"learning_rate": 5.407772228879894e-06,
"loss": 0.4013,
"step": 1319
},
{
"epoch": 1.5832000000000002,
"grad_norm": 0.7612748767791189,
"learning_rate": 5.400816863642991e-06,
"loss": 0.325,
"step": 1320
},
{
"epoch": 1.5844,
"grad_norm": 0.6019136398497016,
"learning_rate": 5.393860717687231e-06,
"loss": 0.3575,
"step": 1321
},
{
"epoch": 1.5856,
"grad_norm": 0.751828664610803,
"learning_rate": 5.3869038045619275e-06,
"loss": 0.3703,
"step": 1322
},
{
"epoch": 1.5868,
"grad_norm": 0.5466919973651415,
"learning_rate": 5.379946137817891e-06,
"loss": 0.3466,
"step": 1323
},
{
"epoch": 1.588,
"grad_norm": 0.7651533715672375,
"learning_rate": 5.3729877310073985e-06,
"loss": 0.3489,
"step": 1324
},
{
"epoch": 1.5892,
"grad_norm": 0.7136578977851556,
"learning_rate": 5.366028597684173e-06,
"loss": 0.3705,
"step": 1325
},
{
"epoch": 1.5904,
"grad_norm": 0.6754255929933941,
"learning_rate": 5.359068751403347e-06,
"loss": 0.3992,
"step": 1326
},
{
"epoch": 1.5916000000000001,
"grad_norm": 0.5967884238527579,
"learning_rate": 5.352108205721445e-06,
"loss": 0.3769,
"step": 1327
},
{
"epoch": 1.5928,
"grad_norm": 0.6625218756445377,
"learning_rate": 5.345146974196351e-06,
"loss": 0.3604,
"step": 1328
},
{
"epoch": 1.5939999999999999,
"grad_norm": 0.5525601106316149,
"learning_rate": 5.338185070387289e-06,
"loss": 0.3782,
"step": 1329
},
{
"epoch": 1.5952,
"grad_norm": 0.8075972134495009,
"learning_rate": 5.3312225078547895e-06,
"loss": 0.3431,
"step": 1330
},
{
"epoch": 1.5964,
"grad_norm": 0.7363190685398948,
"learning_rate": 5.324259300160667e-06,
"loss": 0.3784,
"step": 1331
},
{
"epoch": 1.5976,
"grad_norm": 1.0162314265692152,
"learning_rate": 5.31729546086799e-06,
"loss": 0.4074,
"step": 1332
},
{
"epoch": 1.5988,
"grad_norm": 0.9038605756327065,
"learning_rate": 5.310331003541065e-06,
"loss": 0.3224,
"step": 1333
},
{
"epoch": 1.6,
"grad_norm": 0.8016074933345667,
"learning_rate": 5.303365941745392e-06,
"loss": 0.3876,
"step": 1334
},
{
"epoch": 1.6012,
"grad_norm": 0.6869688947389407,
"learning_rate": 5.296400289047655e-06,
"loss": 0.322,
"step": 1335
},
{
"epoch": 1.6024,
"grad_norm": 0.736992618588669,
"learning_rate": 5.289434059015689e-06,
"loss": 0.3105,
"step": 1336
},
{
"epoch": 1.6036000000000001,
"grad_norm": 0.7230652080797938,
"learning_rate": 5.28246726521845e-06,
"loss": 0.3997,
"step": 1337
},
{
"epoch": 1.6048,
"grad_norm": 0.7897683623542725,
"learning_rate": 5.275499921225994e-06,
"loss": 0.4276,
"step": 1338
},
{
"epoch": 1.6059999999999999,
"grad_norm": 0.817642765995061,
"learning_rate": 5.26853204060945e-06,
"loss": 0.3465,
"step": 1339
},
{
"epoch": 1.6072,
"grad_norm": 0.737469496929238,
"learning_rate": 5.26156363694099e-06,
"loss": 0.4482,
"step": 1340
},
{
"epoch": 1.6084,
"grad_norm": 0.6806819235177738,
"learning_rate": 5.2545947237938055e-06,
"loss": 0.4062,
"step": 1341
},
{
"epoch": 1.6096,
"grad_norm": 0.6979969705507644,
"learning_rate": 5.247625314742083e-06,
"loss": 0.3362,
"step": 1342
},
{
"epoch": 1.6108,
"grad_norm": 0.6527971309581354,
"learning_rate": 5.240655423360969e-06,
"loss": 0.3565,
"step": 1343
},
{
"epoch": 1.612,
"grad_norm": 0.771662973773792,
"learning_rate": 5.233685063226557e-06,
"loss": 0.3642,
"step": 1344
},
{
"epoch": 1.6132,
"grad_norm": 0.6659901238090452,
"learning_rate": 5.226714247915846e-06,
"loss": 0.357,
"step": 1345
},
{
"epoch": 1.6143999999999998,
"grad_norm": 0.6358511845478301,
"learning_rate": 5.219742991006728e-06,
"loss": 0.3564,
"step": 1346
},
{
"epoch": 1.6156000000000001,
"grad_norm": 0.7450366230668033,
"learning_rate": 5.212771306077951e-06,
"loss": 0.359,
"step": 1347
},
{
"epoch": 1.6168,
"grad_norm": 0.709296611637657,
"learning_rate": 5.205799206709097e-06,
"loss": 0.3774,
"step": 1348
},
{
"epoch": 1.6179999999999999,
"grad_norm": 0.6401026124118969,
"learning_rate": 5.198826706480558e-06,
"loss": 0.302,
"step": 1349
},
{
"epoch": 1.6192,
"grad_norm": 0.7675390425760013,
"learning_rate": 5.191853818973505e-06,
"loss": 0.3858,
"step": 1350
},
{
"epoch": 1.6204,
"grad_norm": 0.5886281977686693,
"learning_rate": 5.184880557769865e-06,
"loss": 0.3826,
"step": 1351
},
{
"epoch": 1.6216,
"grad_norm": 0.6769994607493972,
"learning_rate": 5.177906936452287e-06,
"loss": 0.3295,
"step": 1352
},
{
"epoch": 1.6228,
"grad_norm": 0.6113722372496796,
"learning_rate": 5.170932968604131e-06,
"loss": 0.366,
"step": 1353
},
{
"epoch": 1.624,
"grad_norm": 0.7396570654175462,
"learning_rate": 5.163958667809422e-06,
"loss": 0.4019,
"step": 1354
},
{
"epoch": 1.6252,
"grad_norm": 0.6521291484789515,
"learning_rate": 5.156984047652841e-06,
"loss": 0.4099,
"step": 1355
},
{
"epoch": 1.6263999999999998,
"grad_norm": 0.8958020047136015,
"learning_rate": 5.15000912171969e-06,
"loss": 0.3966,
"step": 1356
},
{
"epoch": 1.6276000000000002,
"grad_norm": 0.879582337792914,
"learning_rate": 5.143033903595862e-06,
"loss": 0.4172,
"step": 1357
},
{
"epoch": 1.6288,
"grad_norm": 0.700375484097157,
"learning_rate": 5.1360584068678225e-06,
"loss": 0.3885,
"step": 1358
},
{
"epoch": 1.63,
"grad_norm": 0.8058101827380855,
"learning_rate": 5.129082645122579e-06,
"loss": 0.3788,
"step": 1359
},
{
"epoch": 1.6312,
"grad_norm": 0.6578361914431299,
"learning_rate": 5.1221066319476576e-06,
"loss": 0.3847,
"step": 1360
},
{
"epoch": 1.6324,
"grad_norm": 0.7881748267670962,
"learning_rate": 5.115130380931071e-06,
"loss": 0.38,
"step": 1361
},
{
"epoch": 1.6336,
"grad_norm": 0.5761385873542854,
"learning_rate": 5.108153905661296e-06,
"loss": 0.3167,
"step": 1362
},
{
"epoch": 1.6348,
"grad_norm": 0.9319738396303264,
"learning_rate": 5.101177219727245e-06,
"loss": 0.3921,
"step": 1363
},
{
"epoch": 1.6360000000000001,
"grad_norm": 0.7893703206512759,
"learning_rate": 5.094200336718246e-06,
"loss": 0.334,
"step": 1364
},
{
"epoch": 1.6372,
"grad_norm": 0.6819178057708699,
"learning_rate": 5.087223270224003e-06,
"loss": 0.4034,
"step": 1365
},
{
"epoch": 1.6383999999999999,
"grad_norm": 0.7678335705916015,
"learning_rate": 5.080246033834581e-06,
"loss": 0.3627,
"step": 1366
},
{
"epoch": 1.6396,
"grad_norm": 0.7494864620899846,
"learning_rate": 5.0732686411403816e-06,
"loss": 0.3725,
"step": 1367
},
{
"epoch": 1.6408,
"grad_norm": 0.7456937979731961,
"learning_rate": 5.066291105732102e-06,
"loss": 0.4032,
"step": 1368
},
{
"epoch": 1.642,
"grad_norm": 0.7685810215209159,
"learning_rate": 5.05931344120072e-06,
"loss": 0.3662,
"step": 1369
},
{
"epoch": 1.6432,
"grad_norm": 0.6504737497768305,
"learning_rate": 5.052335661137467e-06,
"loss": 0.315,
"step": 1370
},
{
"epoch": 1.6444,
"grad_norm": 0.7378308874781397,
"learning_rate": 5.0453577791337984e-06,
"loss": 0.3833,
"step": 1371
},
{
"epoch": 1.6456,
"grad_norm": 0.8427001115316601,
"learning_rate": 5.03837980878137e-06,
"loss": 0.3915,
"step": 1372
},
{
"epoch": 1.6468,
"grad_norm": 0.5979989584037766,
"learning_rate": 5.031401763672003e-06,
"loss": 0.3413,
"step": 1373
},
{
"epoch": 1.6480000000000001,
"grad_norm": 0.6861333717486952,
"learning_rate": 5.024423657397674e-06,
"loss": 0.383,
"step": 1374
},
{
"epoch": 1.6492,
"grad_norm": 0.7851294131352125,
"learning_rate": 5.017445503550471e-06,
"loss": 0.395,
"step": 1375
},
{
"epoch": 1.6503999999999999,
"grad_norm": 0.6766395055112632,
"learning_rate": 5.010467315722578e-06,
"loss": 0.3698,
"step": 1376
},
{
"epoch": 1.6516,
"grad_norm": 0.5996727514737863,
"learning_rate": 5.003489107506243e-06,
"loss": 0.3914,
"step": 1377
},
{
"epoch": 1.6528,
"grad_norm": 0.7955546836213152,
"learning_rate": 4.996510892493758e-06,
"loss": 0.3966,
"step": 1378
},
{
"epoch": 1.654,
"grad_norm": 0.7409444831765519,
"learning_rate": 4.989532684277424e-06,
"loss": 0.3631,
"step": 1379
},
{
"epoch": 1.6552,
"grad_norm": 0.7035553055837129,
"learning_rate": 4.98255449644953e-06,
"loss": 0.4131,
"step": 1380
},
{
"epoch": 1.6564,
"grad_norm": 0.681132103408971,
"learning_rate": 4.975576342602329e-06,
"loss": 0.3266,
"step": 1381
},
{
"epoch": 1.6576,
"grad_norm": 1.0848955988922409,
"learning_rate": 4.968598236327998e-06,
"loss": 0.4427,
"step": 1382
},
{
"epoch": 1.6588,
"grad_norm": 0.7166106280583338,
"learning_rate": 4.961620191218632e-06,
"loss": 0.3577,
"step": 1383
},
{
"epoch": 1.6600000000000001,
"grad_norm": 0.6122974924663737,
"learning_rate": 4.954642220866202e-06,
"loss": 0.382,
"step": 1384
},
{
"epoch": 1.6612,
"grad_norm": 0.6520938532133623,
"learning_rate": 4.947664338862534e-06,
"loss": 0.3559,
"step": 1385
},
{
"epoch": 1.6623999999999999,
"grad_norm": 0.7136239871003786,
"learning_rate": 4.940686558799283e-06,
"loss": 0.4046,
"step": 1386
},
{
"epoch": 1.6636,
"grad_norm": 0.785789019777447,
"learning_rate": 4.933708894267901e-06,
"loss": 0.3396,
"step": 1387
},
{
"epoch": 1.6648,
"grad_norm": 0.7626115469594512,
"learning_rate": 4.92673135885962e-06,
"loss": 0.3636,
"step": 1388
},
{
"epoch": 1.666,
"grad_norm": 0.6469070517996743,
"learning_rate": 4.919753966165419e-06,
"loss": 0.3521,
"step": 1389
},
{
"epoch": 1.6672,
"grad_norm": 0.7963611340930559,
"learning_rate": 4.912776729775999e-06,
"loss": 0.3522,
"step": 1390
},
{
"epoch": 1.6684,
"grad_norm": 0.6460224959117941,
"learning_rate": 4.905799663281756e-06,
"loss": 0.3328,
"step": 1391
},
{
"epoch": 1.6696,
"grad_norm": 0.7048922937726232,
"learning_rate": 4.898822780272757e-06,
"loss": 0.3388,
"step": 1392
},
{
"epoch": 1.6707999999999998,
"grad_norm": 0.7515975538504189,
"learning_rate": 4.8918460943387065e-06,
"loss": 0.3921,
"step": 1393
},
{
"epoch": 1.6720000000000002,
"grad_norm": 0.6406704202097733,
"learning_rate": 4.8848696190689295e-06,
"loss": 0.344,
"step": 1394
},
{
"epoch": 1.6732,
"grad_norm": 0.6312727883936049,
"learning_rate": 4.877893368052343e-06,
"loss": 0.3482,
"step": 1395
},
{
"epoch": 1.6743999999999999,
"grad_norm": 0.8654209562011752,
"learning_rate": 4.870917354877421e-06,
"loss": 0.3903,
"step": 1396
},
{
"epoch": 1.6756,
"grad_norm": 0.8411597897318203,
"learning_rate": 4.863941593132179e-06,
"loss": 0.3583,
"step": 1397
},
{
"epoch": 1.6768,
"grad_norm": 0.7487153315710853,
"learning_rate": 4.856966096404141e-06,
"loss": 0.3444,
"step": 1398
},
{
"epoch": 1.678,
"grad_norm": 0.7568031425351381,
"learning_rate": 4.849990878280313e-06,
"loss": 0.3951,
"step": 1399
},
{
"epoch": 1.6792,
"grad_norm": 0.9589155216529329,
"learning_rate": 4.843015952347159e-06,
"loss": 0.348,
"step": 1400
},
{
"epoch": 1.6804000000000001,
"grad_norm": 0.7229233028786465,
"learning_rate": 4.8360413321905786e-06,
"loss": 0.3513,
"step": 1401
},
{
"epoch": 1.6816,
"grad_norm": 0.6917315259059171,
"learning_rate": 4.829067031395871e-06,
"loss": 0.3799,
"step": 1402
},
{
"epoch": 1.6827999999999999,
"grad_norm": 0.6828166393431353,
"learning_rate": 4.822093063547715e-06,
"loss": 0.3351,
"step": 1403
},
{
"epoch": 1.6840000000000002,
"grad_norm": 0.7246651099848298,
"learning_rate": 4.815119442230138e-06,
"loss": 0.3329,
"step": 1404
},
{
"epoch": 1.6852,
"grad_norm": 0.5756593586596187,
"learning_rate": 4.8081461810264955e-06,
"loss": 0.3409,
"step": 1405
},
{
"epoch": 1.6864,
"grad_norm": 0.6686483136444735,
"learning_rate": 4.801173293519442e-06,
"loss": 0.3743,
"step": 1406
},
{
"epoch": 1.6876,
"grad_norm": 0.6650460788846222,
"learning_rate": 4.7942007932909046e-06,
"loss": 0.3352,
"step": 1407
},
{
"epoch": 1.6888,
"grad_norm": 0.7225937900874725,
"learning_rate": 4.7872286939220516e-06,
"loss": 0.3289,
"step": 1408
},
{
"epoch": 1.69,
"grad_norm": 0.737752665480001,
"learning_rate": 4.7802570089932746e-06,
"loss": 0.339,
"step": 1409
},
{
"epoch": 1.6912,
"grad_norm": 0.7038717446915606,
"learning_rate": 4.773285752084154e-06,
"loss": 0.3864,
"step": 1410
},
{
"epoch": 1.6924000000000001,
"grad_norm": 0.5939184879985222,
"learning_rate": 4.766314936773445e-06,
"loss": 0.318,
"step": 1411
},
{
"epoch": 1.6936,
"grad_norm": 0.6644420951801195,
"learning_rate": 4.7593445766390315e-06,
"loss": 0.3585,
"step": 1412
},
{
"epoch": 1.6947999999999999,
"grad_norm": 0.9420481590361012,
"learning_rate": 4.752374685257919e-06,
"loss": 0.3271,
"step": 1413
},
{
"epoch": 1.696,
"grad_norm": 0.6777632632308106,
"learning_rate": 4.745405276206196e-06,
"loss": 0.3258,
"step": 1414
},
{
"epoch": 1.6972,
"grad_norm": 0.657525224784251,
"learning_rate": 4.738436363059013e-06,
"loss": 0.3356,
"step": 1415
},
{
"epoch": 1.6984,
"grad_norm": 0.5874234736529492,
"learning_rate": 4.731467959390552e-06,
"loss": 0.3562,
"step": 1416
},
{
"epoch": 1.6996,
"grad_norm": 0.7620154664518827,
"learning_rate": 4.724500078774008e-06,
"loss": 0.3614,
"step": 1417
},
{
"epoch": 1.7008,
"grad_norm": 0.5992724172831685,
"learning_rate": 4.717532734781552e-06,
"loss": 0.3851,
"step": 1418
},
{
"epoch": 1.702,
"grad_norm": 0.8184583302397535,
"learning_rate": 4.710565940984313e-06,
"loss": 0.3681,
"step": 1419
},
{
"epoch": 1.7032,
"grad_norm": 0.6108780793219214,
"learning_rate": 4.703599710952347e-06,
"loss": 0.3883,
"step": 1420
},
{
"epoch": 1.7044000000000001,
"grad_norm": 0.5976324802264569,
"learning_rate": 4.6966340582546085e-06,
"loss": 0.3582,
"step": 1421
},
{
"epoch": 1.7056,
"grad_norm": 0.6687774593648351,
"learning_rate": 4.689668996458937e-06,
"loss": 0.3518,
"step": 1422
},
{
"epoch": 1.7067999999999999,
"grad_norm": 0.47736020784333477,
"learning_rate": 4.682704539132011e-06,
"loss": 0.3857,
"step": 1423
},
{
"epoch": 1.708,
"grad_norm": 0.8395451811956361,
"learning_rate": 4.6757406998393354e-06,
"loss": 0.3906,
"step": 1424
},
{
"epoch": 1.7092,
"grad_norm": 0.7567347352959016,
"learning_rate": 4.668777492145212e-06,
"loss": 0.3541,
"step": 1425
},
{
"epoch": 1.7104,
"grad_norm": 1.0464877944862874,
"learning_rate": 4.661814929612713e-06,
"loss": 0.389,
"step": 1426
},
{
"epoch": 1.7116,
"grad_norm": 0.5592055085354183,
"learning_rate": 4.654853025803649e-06,
"loss": 0.3873,
"step": 1427
},
{
"epoch": 1.7128,
"grad_norm": 0.5864226590622897,
"learning_rate": 4.6478917942785575e-06,
"loss": 0.377,
"step": 1428
},
{
"epoch": 1.714,
"grad_norm": 0.7859640111174085,
"learning_rate": 4.640931248596655e-06,
"loss": 0.3692,
"step": 1429
},
{
"epoch": 1.7151999999999998,
"grad_norm": 0.6632690676240329,
"learning_rate": 4.633971402315828e-06,
"loss": 0.3304,
"step": 1430
},
{
"epoch": 1.7164000000000001,
"grad_norm": 0.820703665781513,
"learning_rate": 4.627012268992603e-06,
"loss": 0.398,
"step": 1431
},
{
"epoch": 1.7176,
"grad_norm": 0.7199281377980384,
"learning_rate": 4.62005386218211e-06,
"loss": 0.3115,
"step": 1432
},
{
"epoch": 1.7187999999999999,
"grad_norm": 0.7421137967521952,
"learning_rate": 4.613096195438074e-06,
"loss": 0.4041,
"step": 1433
},
{
"epoch": 1.72,
"grad_norm": 0.7749842831199167,
"learning_rate": 4.60613928231277e-06,
"loss": 0.326,
"step": 1434
},
{
"epoch": 1.7212,
"grad_norm": 0.6145845838695677,
"learning_rate": 4.59918313635701e-06,
"loss": 0.3897,
"step": 1435
},
{
"epoch": 1.7224,
"grad_norm": 0.743154744435818,
"learning_rate": 4.592227771120108e-06,
"loss": 0.3095,
"step": 1436
},
{
"epoch": 1.7236,
"grad_norm": 0.8816826248157816,
"learning_rate": 4.585273200149859e-06,
"loss": 0.3404,
"step": 1437
},
{
"epoch": 1.7248,
"grad_norm": 0.7057427593004695,
"learning_rate": 4.578319436992515e-06,
"loss": 0.3977,
"step": 1438
},
{
"epoch": 1.726,
"grad_norm": 0.7742683261993301,
"learning_rate": 4.5713664951927475e-06,
"loss": 0.3555,
"step": 1439
},
{
"epoch": 1.7271999999999998,
"grad_norm": 0.7137221637383818,
"learning_rate": 4.5644143882936316e-06,
"loss": 0.3944,
"step": 1440
},
{
"epoch": 1.7284000000000002,
"grad_norm": 0.8217962695771334,
"learning_rate": 4.5574631298366165e-06,
"loss": 0.3602,
"step": 1441
},
{
"epoch": 1.7296,
"grad_norm": 0.7743088864846196,
"learning_rate": 4.550512733361499e-06,
"loss": 0.398,
"step": 1442
},
{
"epoch": 1.7308,
"grad_norm": 0.5526900661634894,
"learning_rate": 4.543563212406392e-06,
"loss": 0.3113,
"step": 1443
},
{
"epoch": 1.732,
"grad_norm": 0.7923314311381595,
"learning_rate": 4.536614580507714e-06,
"loss": 0.3367,
"step": 1444
},
{
"epoch": 1.7332,
"grad_norm": 0.7014593716808332,
"learning_rate": 4.52966685120014e-06,
"loss": 0.3645,
"step": 1445
},
{
"epoch": 1.7344,
"grad_norm": 1.0153075080959297,
"learning_rate": 4.5227200380165925e-06,
"loss": 0.358,
"step": 1446
},
{
"epoch": 1.7356,
"grad_norm": 0.818893645099761,
"learning_rate": 4.51577415448821e-06,
"loss": 0.37,
"step": 1447
},
{
"epoch": 1.7368000000000001,
"grad_norm": 0.527618842943089,
"learning_rate": 4.508829214144318e-06,
"loss": 0.3825,
"step": 1448
},
{
"epoch": 1.738,
"grad_norm": 0.7768621157948593,
"learning_rate": 4.5018852305124075e-06,
"loss": 0.3296,
"step": 1449
},
{
"epoch": 1.7391999999999999,
"grad_norm": 0.6883502786416549,
"learning_rate": 4.494942217118105e-06,
"loss": 0.4146,
"step": 1450
},
{
"epoch": 1.7404,
"grad_norm": 0.6744864874963413,
"learning_rate": 4.488000187485144e-06,
"loss": 0.3815,
"step": 1451
},
{
"epoch": 1.7416,
"grad_norm": 0.9535350825299096,
"learning_rate": 4.481059155135346e-06,
"loss": 0.3755,
"step": 1452
},
{
"epoch": 1.7428,
"grad_norm": 0.6208040428555441,
"learning_rate": 4.474119133588588e-06,
"loss": 0.3529,
"step": 1453
},
{
"epoch": 1.744,
"grad_norm": 0.7058976078379278,
"learning_rate": 4.4671801363627776e-06,
"loss": 0.3387,
"step": 1454
},
{
"epoch": 1.7452,
"grad_norm": 0.7750488051713403,
"learning_rate": 4.460242176973829e-06,
"loss": 0.351,
"step": 1455
},
{
"epoch": 1.7464,
"grad_norm": 0.5983821680867452,
"learning_rate": 4.453305268935637e-06,
"loss": 0.3497,
"step": 1456
},
{
"epoch": 1.7476,
"grad_norm": 0.7510057031903162,
"learning_rate": 4.446369425760042e-06,
"loss": 0.3876,
"step": 1457
},
{
"epoch": 1.7488000000000001,
"grad_norm": 0.6617397556445918,
"learning_rate": 4.439434660956814e-06,
"loss": 0.3762,
"step": 1458
},
{
"epoch": 1.75,
"grad_norm": 0.7757478994129973,
"learning_rate": 4.432500988033621e-06,
"loss": 0.3048,
"step": 1459
},
{
"epoch": 1.7511999999999999,
"grad_norm": 0.7476953780010415,
"learning_rate": 4.42556842049601e-06,
"loss": 0.3863,
"step": 1460
},
{
"epoch": 1.7524,
"grad_norm": 0.6540368477466787,
"learning_rate": 4.418636971847367e-06,
"loss": 0.3167,
"step": 1461
},
{
"epoch": 1.7536,
"grad_norm": 0.674098340810802,
"learning_rate": 4.4117066555889045e-06,
"loss": 0.4289,
"step": 1462
},
{
"epoch": 1.7548,
"grad_norm": 0.8481165177301736,
"learning_rate": 4.404777485219624e-06,
"loss": 0.3472,
"step": 1463
},
{
"epoch": 1.756,
"grad_norm": 0.5938945169957544,
"learning_rate": 4.397849474236299e-06,
"loss": 0.3751,
"step": 1464
},
{
"epoch": 1.7572,
"grad_norm": 0.9562487802524786,
"learning_rate": 4.390922636133444e-06,
"loss": 0.3892,
"step": 1465
},
{
"epoch": 1.7584,
"grad_norm": 0.681978775046848,
"learning_rate": 4.38399698440329e-06,
"loss": 0.3819,
"step": 1466
},
{
"epoch": 1.7596,
"grad_norm": 0.6752561331579562,
"learning_rate": 4.377072532535756e-06,
"loss": 0.3762,
"step": 1467
},
{
"epoch": 1.7608000000000001,
"grad_norm": 0.8525984911089448,
"learning_rate": 4.370149294018419e-06,
"loss": 0.3729,
"step": 1468
},
{
"epoch": 1.762,
"grad_norm": 0.7148334740342769,
"learning_rate": 4.3632272823365004e-06,
"loss": 0.399,
"step": 1469
},
{
"epoch": 1.7631999999999999,
"grad_norm": 0.8716895233990396,
"learning_rate": 4.356306510972829e-06,
"loss": 0.345,
"step": 1470
},
{
"epoch": 1.7644,
"grad_norm": 0.7322168094991977,
"learning_rate": 4.349386993407817e-06,
"loss": 0.4121,
"step": 1471
},
{
"epoch": 1.7656,
"grad_norm": 0.5317993879946495,
"learning_rate": 4.342468743119436e-06,
"loss": 0.3541,
"step": 1472
},
{
"epoch": 1.7668,
"grad_norm": 1.0024619158505292,
"learning_rate": 4.335551773583186e-06,
"loss": 0.3707,
"step": 1473
},
{
"epoch": 1.768,
"grad_norm": 1.0647757271297378,
"learning_rate": 4.328636098272075e-06,
"loss": 0.4299,
"step": 1474
},
{
"epoch": 1.7692,
"grad_norm": 0.8106812515865629,
"learning_rate": 4.32172173065659e-06,
"loss": 0.3781,
"step": 1475
},
{
"epoch": 1.7704,
"grad_norm": 0.8369877982622608,
"learning_rate": 4.31480868420467e-06,
"loss": 0.3565,
"step": 1476
},
{
"epoch": 1.7715999999999998,
"grad_norm": 0.6394932977170336,
"learning_rate": 4.307896972381681e-06,
"loss": 0.3484,
"step": 1477
},
{
"epoch": 1.7728000000000002,
"grad_norm": 0.7418775156967402,
"learning_rate": 4.3009866086503905e-06,
"loss": 0.4252,
"step": 1478
},
{
"epoch": 1.774,
"grad_norm": 0.7223212169696627,
"learning_rate": 4.294077606470937e-06,
"loss": 0.3599,
"step": 1479
},
{
"epoch": 1.7752,
"grad_norm": 0.8004568736049867,
"learning_rate": 4.28716997930081e-06,
"loss": 0.3517,
"step": 1480
},
{
"epoch": 1.7764,
"grad_norm": 0.8484728472633192,
"learning_rate": 4.2802637405948175e-06,
"loss": 0.4238,
"step": 1481
},
{
"epoch": 1.7776,
"grad_norm": 0.607913184181001,
"learning_rate": 4.273358903805069e-06,
"loss": 0.3693,
"step": 1482
},
{
"epoch": 1.7788,
"grad_norm": 0.8758008146560339,
"learning_rate": 4.266455482380938e-06,
"loss": 0.3522,
"step": 1483
},
{
"epoch": 1.78,
"grad_norm": 0.5759063891025465,
"learning_rate": 4.2595534897690415e-06,
"loss": 0.3472,
"step": 1484
},
{
"epoch": 1.7812000000000001,
"grad_norm": 0.6347159704640251,
"learning_rate": 4.252652939413215e-06,
"loss": 0.3487,
"step": 1485
},
{
"epoch": 1.7824,
"grad_norm": 0.677342052319476,
"learning_rate": 4.245753844754484e-06,
"loss": 0.4214,
"step": 1486
},
{
"epoch": 1.7835999999999999,
"grad_norm": 0.6838550073892636,
"learning_rate": 4.23885621923104e-06,
"loss": 0.3386,
"step": 1487
},
{
"epoch": 1.7848000000000002,
"grad_norm": 0.6131430282976505,
"learning_rate": 4.231960076278211e-06,
"loss": 0.352,
"step": 1488
},
{
"epoch": 1.786,
"grad_norm": 0.751027609414722,
"learning_rate": 4.225065429328439e-06,
"loss": 0.4463,
"step": 1489
},
{
"epoch": 1.7872,
"grad_norm": 0.8277241937054599,
"learning_rate": 4.218172291811249e-06,
"loss": 0.3498,
"step": 1490
},
{
"epoch": 1.7884,
"grad_norm": 0.6373401861937386,
"learning_rate": 4.211280677153228e-06,
"loss": 0.3395,
"step": 1491
},
{
"epoch": 1.7896,
"grad_norm": 0.6002576266039981,
"learning_rate": 4.204390598777999e-06,
"loss": 0.3378,
"step": 1492
},
{
"epoch": 1.7908,
"grad_norm": 0.6746388999526817,
"learning_rate": 4.1975020701061884e-06,
"loss": 0.3459,
"step": 1493
},
{
"epoch": 1.792,
"grad_norm": 0.5436355959217024,
"learning_rate": 4.190615104555407e-06,
"loss": 0.3616,
"step": 1494
},
{
"epoch": 1.7932000000000001,
"grad_norm": 0.7583442870565927,
"learning_rate": 4.1837297155402204e-06,
"loss": 0.3425,
"step": 1495
},
{
"epoch": 1.7944,
"grad_norm": 0.5627756529431762,
"learning_rate": 4.17684591647212e-06,
"loss": 0.3731,
"step": 1496
},
{
"epoch": 1.7955999999999999,
"grad_norm": 0.9448888708327425,
"learning_rate": 4.1699637207595035e-06,
"loss": 0.3742,
"step": 1497
},
{
"epoch": 1.7968,
"grad_norm": 0.7991708909582393,
"learning_rate": 4.163083141807648e-06,
"loss": 0.387,
"step": 1498
},
{
"epoch": 1.798,
"grad_norm": 0.6183380003326223,
"learning_rate": 4.156204193018677e-06,
"loss": 0.3676,
"step": 1499
},
{
"epoch": 1.7992,
"grad_norm": 0.7687359165224887,
"learning_rate": 4.149326887791541e-06,
"loss": 0.389,
"step": 1500
},
{
"epoch": 1.8004,
"grad_norm": 0.7403620518621874,
"learning_rate": 4.142451239521988e-06,
"loss": 0.3662,
"step": 1501
},
{
"epoch": 1.8016,
"grad_norm": 0.6844611928262976,
"learning_rate": 4.135577261602537e-06,
"loss": 0.4145,
"step": 1502
},
{
"epoch": 1.8028,
"grad_norm": 0.6869038162637987,
"learning_rate": 4.128704967422458e-06,
"loss": 0.3596,
"step": 1503
},
{
"epoch": 1.804,
"grad_norm": 0.6923943431820913,
"learning_rate": 4.1218343703677385e-06,
"loss": 0.3179,
"step": 1504
},
{
"epoch": 1.8052000000000001,
"grad_norm": 0.6865769031215199,
"learning_rate": 4.114965483821061e-06,
"loss": 0.3717,
"step": 1505
},
{
"epoch": 1.8064,
"grad_norm": 0.746561997171133,
"learning_rate": 4.108098321161776e-06,
"loss": 0.38,
"step": 1506
},
{
"epoch": 1.8075999999999999,
"grad_norm": 0.5859730876952545,
"learning_rate": 4.101232895765875e-06,
"loss": 0.3552,
"step": 1507
},
{
"epoch": 1.8088,
"grad_norm": 0.7508628014795502,
"learning_rate": 4.094369221005965e-06,
"loss": 0.3751,
"step": 1508
},
{
"epoch": 1.81,
"grad_norm": 0.6233073628845128,
"learning_rate": 4.0875073102512485e-06,
"loss": 0.3273,
"step": 1509
},
{
"epoch": 1.8112,
"grad_norm": 0.556944519289841,
"learning_rate": 4.080647176867486e-06,
"loss": 0.3278,
"step": 1510
},
{
"epoch": 1.8124,
"grad_norm": 0.7942591692701104,
"learning_rate": 4.07378883421698e-06,
"loss": 0.378,
"step": 1511
},
{
"epoch": 1.8136,
"grad_norm": 0.629890291962485,
"learning_rate": 4.066932295658543e-06,
"loss": 0.3641,
"step": 1512
},
{
"epoch": 1.8148,
"grad_norm": 0.7296741775656937,
"learning_rate": 4.06007757454747e-06,
"loss": 0.3744,
"step": 1513
},
{
"epoch": 1.8159999999999998,
"grad_norm": 0.7359348678382316,
"learning_rate": 4.053224684235526e-06,
"loss": 0.3639,
"step": 1514
},
{
"epoch": 1.8172000000000001,
"grad_norm": 0.6930398969323774,
"learning_rate": 4.0463736380708986e-06,
"loss": 0.3794,
"step": 1515
},
{
"epoch": 1.8184,
"grad_norm": 0.7403747346350105,
"learning_rate": 4.039524449398191e-06,
"loss": 0.3583,
"step": 1516
},
{
"epoch": 1.8195999999999999,
"grad_norm": 0.8452723109566541,
"learning_rate": 4.032677131558386e-06,
"loss": 0.338,
"step": 1517
},
{
"epoch": 1.8208,
"grad_norm": 0.9045427389143957,
"learning_rate": 4.025831697888817e-06,
"loss": 0.3448,
"step": 1518
},
{
"epoch": 1.822,
"grad_norm": 0.671646040056322,
"learning_rate": 4.01898816172316e-06,
"loss": 0.4044,
"step": 1519
},
{
"epoch": 1.8232,
"grad_norm": 0.5988353471135299,
"learning_rate": 4.012146536391383e-06,
"loss": 0.3769,
"step": 1520
},
{
"epoch": 1.8244,
"grad_norm": 0.6523671498218497,
"learning_rate": 4.005306835219737e-06,
"loss": 0.3459,
"step": 1521
},
{
"epoch": 1.8256000000000001,
"grad_norm": 0.7326969470082931,
"learning_rate": 3.998469071530725e-06,
"loss": 0.3702,
"step": 1522
},
{
"epoch": 1.8268,
"grad_norm": 0.909958077061063,
"learning_rate": 3.991633258643077e-06,
"loss": 0.3561,
"step": 1523
},
{
"epoch": 1.8279999999999998,
"grad_norm": 0.7180926275621504,
"learning_rate": 3.9847994098717166e-06,
"loss": 0.3564,
"step": 1524
},
{
"epoch": 1.8292000000000002,
"grad_norm": 0.6296581283445997,
"learning_rate": 3.9779675385277545e-06,
"loss": 0.3155,
"step": 1525
},
{
"epoch": 1.8304,
"grad_norm": 0.8036529554932723,
"learning_rate": 3.971137657918437e-06,
"loss": 0.3455,
"step": 1526
},
{
"epoch": 1.8316,
"grad_norm": 1.1415048348559849,
"learning_rate": 3.96430978134714e-06,
"loss": 0.4757,
"step": 1527
},
{
"epoch": 1.8328,
"grad_norm": 0.7591043961061273,
"learning_rate": 3.957483922113334e-06,
"loss": 0.3651,
"step": 1528
},
{
"epoch": 1.834,
"grad_norm": 0.7377445089549338,
"learning_rate": 3.950660093512556e-06,
"loss": 0.3283,
"step": 1529
},
{
"epoch": 1.8352,
"grad_norm": 0.7860879053915407,
"learning_rate": 3.943838308836398e-06,
"loss": 0.3859,
"step": 1530
},
{
"epoch": 1.8364,
"grad_norm": 0.6804394158279877,
"learning_rate": 3.937018581372462e-06,
"loss": 0.3526,
"step": 1531
},
{
"epoch": 1.8376000000000001,
"grad_norm": 0.587817308609732,
"learning_rate": 3.9302009244043435e-06,
"loss": 0.3824,
"step": 1532
},
{
"epoch": 1.8388,
"grad_norm": 0.7427178916174038,
"learning_rate": 3.923385351211609e-06,
"loss": 0.3381,
"step": 1533
},
{
"epoch": 1.8399999999999999,
"grad_norm": 0.9474352077967609,
"learning_rate": 3.916571875069764e-06,
"loss": 0.3566,
"step": 1534
},
{
"epoch": 1.8412,
"grad_norm": 0.7388042350648729,
"learning_rate": 3.909760509250225e-06,
"loss": 0.3811,
"step": 1535
},
{
"epoch": 1.8424,
"grad_norm": 0.6807413930744459,
"learning_rate": 3.902951267020311e-06,
"loss": 0.3655,
"step": 1536
},
{
"epoch": 1.8436,
"grad_norm": 0.6683853559058284,
"learning_rate": 3.896144161643189e-06,
"loss": 0.3828,
"step": 1537
},
{
"epoch": 1.8448,
"grad_norm": 0.7860808942164064,
"learning_rate": 3.8893392063778736e-06,
"loss": 0.3849,
"step": 1538
},
{
"epoch": 1.846,
"grad_norm": 0.6725210663439518,
"learning_rate": 3.882536414479189e-06,
"loss": 0.333,
"step": 1539
},
{
"epoch": 1.8472,
"grad_norm": 0.7413488179901978,
"learning_rate": 3.8757357991977415e-06,
"loss": 0.3676,
"step": 1540
},
{
"epoch": 1.8484,
"grad_norm": 0.6430251510115813,
"learning_rate": 3.868937373779907e-06,
"loss": 0.3428,
"step": 1541
},
{
"epoch": 1.8496000000000001,
"grad_norm": 0.7023798155484271,
"learning_rate": 3.862141151467787e-06,
"loss": 0.3643,
"step": 1542
},
{
"epoch": 1.8508,
"grad_norm": 0.7116517259740174,
"learning_rate": 3.855347145499197e-06,
"loss": 0.3078,
"step": 1543
},
{
"epoch": 1.8519999999999999,
"grad_norm": 0.6185131996350404,
"learning_rate": 3.848555369107631e-06,
"loss": 0.3467,
"step": 1544
},
{
"epoch": 1.8532,
"grad_norm": 0.8015280083653907,
"learning_rate": 3.841765835522242e-06,
"loss": 0.349,
"step": 1545
},
{
"epoch": 1.8544,
"grad_norm": 0.6652141729051564,
"learning_rate": 3.83497855796782e-06,
"loss": 0.3907,
"step": 1546
},
{
"epoch": 1.8556,
"grad_norm": 0.616377265139706,
"learning_rate": 3.8281935496647526e-06,
"loss": 0.3853,
"step": 1547
},
{
"epoch": 1.8568,
"grad_norm": 0.8259455933431535,
"learning_rate": 3.821410823829011e-06,
"loss": 0.3635,
"step": 1548
},
{
"epoch": 1.858,
"grad_norm": 0.718324438943734,
"learning_rate": 3.8146303936721197e-06,
"loss": 0.3805,
"step": 1549
},
{
"epoch": 1.8592,
"grad_norm": 0.6194288299372507,
"learning_rate": 3.8078522724011324e-06,
"loss": 0.3361,
"step": 1550
},
{
"epoch": 1.8604,
"grad_norm": 0.6332641021781915,
"learning_rate": 3.8010764732186044e-06,
"loss": 0.3381,
"step": 1551
},
{
"epoch": 1.8616000000000001,
"grad_norm": 0.8677716893089377,
"learning_rate": 3.79430300932257e-06,
"loss": 0.3873,
"step": 1552
},
{
"epoch": 1.8628,
"grad_norm": 0.5982998743808187,
"learning_rate": 3.7875318939065147e-06,
"loss": 0.3256,
"step": 1553
},
{
"epoch": 1.8639999999999999,
"grad_norm": 0.7793279668125604,
"learning_rate": 3.7807631401593455e-06,
"loss": 0.3684,
"step": 1554
},
{
"epoch": 1.8652,
"grad_norm": 0.651590387780422,
"learning_rate": 3.773996761265373e-06,
"loss": 0.3711,
"step": 1555
},
{
"epoch": 1.8664,
"grad_norm": 0.7993431411401498,
"learning_rate": 3.767232770404281e-06,
"loss": 0.3578,
"step": 1556
},
{
"epoch": 1.8676,
"grad_norm": 0.5761264076246854,
"learning_rate": 3.7604711807511034e-06,
"loss": 0.3732,
"step": 1557
},
{
"epoch": 1.8688,
"grad_norm": 0.7003767667069869,
"learning_rate": 3.753712005476197e-06,
"loss": 0.4024,
"step": 1558
},
{
"epoch": 1.87,
"grad_norm": 0.7058666971733841,
"learning_rate": 3.74695525774521e-06,
"loss": 0.3498,
"step": 1559
},
{
"epoch": 1.8712,
"grad_norm": 0.7189682207683492,
"learning_rate": 3.7402009507190696e-06,
"loss": 0.3218,
"step": 1560
},
{
"epoch": 1.8723999999999998,
"grad_norm": 0.5337879684275478,
"learning_rate": 3.733449097553945e-06,
"loss": 0.439,
"step": 1561
},
{
"epoch": 1.8736000000000002,
"grad_norm": 0.891294939196006,
"learning_rate": 3.7266997114012265e-06,
"loss": 0.3866,
"step": 1562
},
{
"epoch": 1.8748,
"grad_norm": 0.8057324955390353,
"learning_rate": 3.7199528054075005e-06,
"loss": 0.3233,
"step": 1563
},
{
"epoch": 1.876,
"grad_norm": 1.1066717790385456,
"learning_rate": 3.713208392714523e-06,
"loss": 0.3866,
"step": 1564
},
{
"epoch": 1.8772,
"grad_norm": 0.6602731940770842,
"learning_rate": 3.7064664864591878e-06,
"loss": 0.3617,
"step": 1565
},
{
"epoch": 1.8784,
"grad_norm": 0.7006848148246145,
"learning_rate": 3.6997270997735122e-06,
"loss": 0.3564,
"step": 1566
},
{
"epoch": 1.8796,
"grad_norm": 0.8563944650223395,
"learning_rate": 3.6929902457846034e-06,
"loss": 0.3799,
"step": 1567
},
{
"epoch": 1.8808,
"grad_norm": 0.6402192204036325,
"learning_rate": 3.6862559376146388e-06,
"loss": 0.3804,
"step": 1568
},
{
"epoch": 1.8820000000000001,
"grad_norm": 0.7113793407203027,
"learning_rate": 3.6795241883808342e-06,
"loss": 0.3743,
"step": 1569
},
{
"epoch": 1.8832,
"grad_norm": 0.6890839912593231,
"learning_rate": 3.6727950111954186e-06,
"loss": 0.3053,
"step": 1570
},
{
"epoch": 1.8843999999999999,
"grad_norm": 0.7390803689338293,
"learning_rate": 3.6660684191656155e-06,
"loss": 0.3405,
"step": 1571
},
{
"epoch": 1.8856000000000002,
"grad_norm": 0.5817835403281171,
"learning_rate": 3.6593444253936094e-06,
"loss": 0.3486,
"step": 1572
},
{
"epoch": 1.8868,
"grad_norm": 0.5782087520175341,
"learning_rate": 3.652623042976529e-06,
"loss": 0.328,
"step": 1573
},
{
"epoch": 1.888,
"grad_norm": 0.7920901688832652,
"learning_rate": 3.645904285006412e-06,
"loss": 0.3153,
"step": 1574
},
{
"epoch": 1.8892,
"grad_norm": 0.5895347762630812,
"learning_rate": 3.6391881645701854e-06,
"loss": 0.4006,
"step": 1575
},
{
"epoch": 1.8904,
"grad_norm": 0.7300113456869172,
"learning_rate": 3.632474694749638e-06,
"loss": 0.3488,
"step": 1576
},
{
"epoch": 1.8916,
"grad_norm": 0.6605896135094141,
"learning_rate": 3.625763888621397e-06,
"loss": 0.3945,
"step": 1577
},
{
"epoch": 1.8928,
"grad_norm": 0.7282823077547287,
"learning_rate": 3.6190557592569e-06,
"loss": 0.4831,
"step": 1578
},
{
"epoch": 1.8940000000000001,
"grad_norm": 1.6543218663167587,
"learning_rate": 3.612350319722372e-06,
"loss": 0.3871,
"step": 1579
},
{
"epoch": 1.8952,
"grad_norm": 0.7838986093814664,
"learning_rate": 3.6056475830787997e-06,
"loss": 0.3772,
"step": 1580
},
{
"epoch": 1.8963999999999999,
"grad_norm": 0.7136533269568321,
"learning_rate": 3.5989475623819025e-06,
"loss": 0.4076,
"step": 1581
},
{
"epoch": 1.8976,
"grad_norm": 0.6621113724768938,
"learning_rate": 3.5922502706821094e-06,
"loss": 0.3222,
"step": 1582
},
{
"epoch": 1.8988,
"grad_norm": 0.7623719646798188,
"learning_rate": 3.585555721024535e-06,
"loss": 0.4217,
"step": 1583
},
{
"epoch": 1.9,
"grad_norm": 0.6067388023801122,
"learning_rate": 3.578863926448955e-06,
"loss": 0.3764,
"step": 1584
},
{
"epoch": 1.9012,
"grad_norm": 0.7933161956403216,
"learning_rate": 3.5721748999897753e-06,
"loss": 0.4196,
"step": 1585
},
{
"epoch": 1.9024,
"grad_norm": 0.7805196481485149,
"learning_rate": 3.5654886546760125e-06,
"loss": 0.379,
"step": 1586
},
{
"epoch": 1.9036,
"grad_norm": 0.7630466646531989,
"learning_rate": 3.558805203531263e-06,
"loss": 0.4131,
"step": 1587
},
{
"epoch": 1.9048,
"grad_norm": 0.9169940134773841,
"learning_rate": 3.5521245595736837e-06,
"loss": 0.3523,
"step": 1588
},
{
"epoch": 1.9060000000000001,
"grad_norm": 0.8885746079698462,
"learning_rate": 3.5454467358159606e-06,
"loss": 0.3838,
"step": 1589
},
{
"epoch": 1.9072,
"grad_norm": 0.6627789396413086,
"learning_rate": 3.5387717452652914e-06,
"loss": 0.3985,
"step": 1590
},
{
"epoch": 1.9083999999999999,
"grad_norm": 0.8317845367021418,
"learning_rate": 3.53209960092335e-06,
"loss": 0.3818,
"step": 1591
},
{
"epoch": 1.9096,
"grad_norm": 0.6410293135263554,
"learning_rate": 3.5254303157862707e-06,
"loss": 0.3634,
"step": 1592
},
{
"epoch": 1.9108,
"grad_norm": 0.8954052810140045,
"learning_rate": 3.5187639028446136e-06,
"loss": 0.3094,
"step": 1593
},
{
"epoch": 1.912,
"grad_norm": 0.7393614071381539,
"learning_rate": 3.512100375083347e-06,
"loss": 0.3639,
"step": 1594
},
{
"epoch": 1.9132,
"grad_norm": 0.7717204389121924,
"learning_rate": 3.5054397454818224e-06,
"loss": 0.3913,
"step": 1595
},
{
"epoch": 1.9144,
"grad_norm": 0.7600370750174577,
"learning_rate": 3.498782027013742e-06,
"loss": 0.4318,
"step": 1596
},
{
"epoch": 1.9156,
"grad_norm": 0.7503240513049503,
"learning_rate": 3.4921272326471388e-06,
"loss": 0.3377,
"step": 1597
},
{
"epoch": 1.9167999999999998,
"grad_norm": 0.6450288898115434,
"learning_rate": 3.4854753753443494e-06,
"loss": 0.3705,
"step": 1598
},
{
"epoch": 1.9180000000000001,
"grad_norm": 0.7747218266983955,
"learning_rate": 3.47882646806199e-06,
"loss": 0.3863,
"step": 1599
},
{
"epoch": 1.9192,
"grad_norm": 0.6457257398279262,
"learning_rate": 3.472180523750933e-06,
"loss": 0.3943,
"step": 1600
},
{
"epoch": 1.9203999999999999,
"grad_norm": 0.7131269781103766,
"learning_rate": 3.4655375553562774e-06,
"loss": 0.3542,
"step": 1601
},
{
"epoch": 1.9216,
"grad_norm": 0.6748953517543854,
"learning_rate": 3.458897575817326e-06,
"loss": 0.3161,
"step": 1602
},
{
"epoch": 1.9228,
"grad_norm": 0.7399438756617545,
"learning_rate": 3.4522605980675593e-06,
"loss": 0.3485,
"step": 1603
},
{
"epoch": 1.924,
"grad_norm": 0.619884406014249,
"learning_rate": 3.44562663503461e-06,
"loss": 0.3406,
"step": 1604
},
{
"epoch": 1.9252,
"grad_norm": 0.6476524748637171,
"learning_rate": 3.43899569964024e-06,
"loss": 0.3623,
"step": 1605
},
{
"epoch": 1.9264000000000001,
"grad_norm": 0.6247829130227588,
"learning_rate": 3.432367804800316e-06,
"loss": 0.3694,
"step": 1606
},
{
"epoch": 1.9276,
"grad_norm": 0.821390907257284,
"learning_rate": 3.4257429634247783e-06,
"loss": 0.3622,
"step": 1607
},
{
"epoch": 1.9287999999999998,
"grad_norm": 0.7050949139255335,
"learning_rate": 3.419121188417622e-06,
"loss": 0.3635,
"step": 1608
},
{
"epoch": 1.9300000000000002,
"grad_norm": 0.6526955802973464,
"learning_rate": 3.41250249267687e-06,
"loss": 0.3462,
"step": 1609
},
{
"epoch": 1.9312,
"grad_norm": 0.7305052511211385,
"learning_rate": 3.4058868890945425e-06,
"loss": 0.3198,
"step": 1610
},
{
"epoch": 1.9324,
"grad_norm": 0.6085547972809887,
"learning_rate": 3.3992743905566453e-06,
"loss": 0.3806,
"step": 1611
},
{
"epoch": 1.9336,
"grad_norm": 0.663194797129961,
"learning_rate": 3.3926650099431286e-06,
"loss": 0.3276,
"step": 1612
},
{
"epoch": 1.9348,
"grad_norm": 0.8403883340998779,
"learning_rate": 3.3860587601278715e-06,
"loss": 0.3676,
"step": 1613
},
{
"epoch": 1.936,
"grad_norm": 0.8529385067632443,
"learning_rate": 3.3794556539786584e-06,
"loss": 0.4565,
"step": 1614
},
{
"epoch": 1.9372,
"grad_norm": 0.5805693398931318,
"learning_rate": 3.372855704357144e-06,
"loss": 0.3587,
"step": 1615
},
{
"epoch": 1.9384000000000001,
"grad_norm": 0.6222832371497026,
"learning_rate": 3.3662589241188382e-06,
"loss": 0.3271,
"step": 1616
},
{
"epoch": 1.9396,
"grad_norm": 0.8006649302914024,
"learning_rate": 3.3596653261130806e-06,
"loss": 0.3662,
"step": 1617
},
{
"epoch": 1.9407999999999999,
"grad_norm": 0.5765540805298704,
"learning_rate": 3.3530749231830073e-06,
"loss": 0.4192,
"step": 1618
},
{
"epoch": 1.942,
"grad_norm": 0.9124936821677964,
"learning_rate": 3.3464877281655335e-06,
"loss": 0.3275,
"step": 1619
},
{
"epoch": 1.9432,
"grad_norm": 0.7794169491834637,
"learning_rate": 3.339903753891326e-06,
"loss": 0.3842,
"step": 1620
},
{
"epoch": 1.9444,
"grad_norm": 0.8034589505524509,
"learning_rate": 3.333323013184773e-06,
"loss": 0.3524,
"step": 1621
},
{
"epoch": 1.9456,
"grad_norm": 0.8610924503776118,
"learning_rate": 3.326745518863976e-06,
"loss": 0.385,
"step": 1622
},
{
"epoch": 1.9468,
"grad_norm": 0.619723323865224,
"learning_rate": 3.320171283740702e-06,
"loss": 0.367,
"step": 1623
},
{
"epoch": 1.948,
"grad_norm": 0.7737064623998821,
"learning_rate": 3.3136003206203727e-06,
"loss": 0.381,
"step": 1624
},
{
"epoch": 1.9492,
"grad_norm": 0.6322629898004767,
"learning_rate": 3.307032642302041e-06,
"loss": 0.3249,
"step": 1625
},
{
"epoch": 1.9504000000000001,
"grad_norm": 0.650785964292608,
"learning_rate": 3.3004682615783524e-06,
"loss": 0.3973,
"step": 1626
},
{
"epoch": 1.9516,
"grad_norm": 0.666129955026848,
"learning_rate": 3.2939071912355424e-06,
"loss": 0.3419,
"step": 1627
},
{
"epoch": 1.9527999999999999,
"grad_norm": 0.7715160567504751,
"learning_rate": 3.2873494440533856e-06,
"loss": 0.3993,
"step": 1628
},
{
"epoch": 1.954,
"grad_norm": 0.781613606855936,
"learning_rate": 3.2807950328051906e-06,
"loss": 0.3925,
"step": 1629
},
{
"epoch": 1.9552,
"grad_norm": 0.8246150930922845,
"learning_rate": 3.2742439702577665e-06,
"loss": 0.3757,
"step": 1630
},
{
"epoch": 1.9564,
"grad_norm": 0.7556531833436847,
"learning_rate": 3.267696269171402e-06,
"loss": 0.3805,
"step": 1631
},
{
"epoch": 1.9576,
"grad_norm": 0.8569962946796378,
"learning_rate": 3.2611519422998308e-06,
"loss": 0.3095,
"step": 1632
},
{
"epoch": 1.9588,
"grad_norm": 0.6532405042390842,
"learning_rate": 3.254611002390227e-06,
"loss": 0.3347,
"step": 1633
},
{
"epoch": 1.96,
"grad_norm": 0.7052828213359866,
"learning_rate": 3.248073462183155e-06,
"loss": 0.3311,
"step": 1634
},
{
"epoch": 1.9612,
"grad_norm": 0.7560298909851522,
"learning_rate": 3.2415393344125647e-06,
"loss": 0.3898,
"step": 1635
},
{
"epoch": 1.9624000000000001,
"grad_norm": 0.762152105408301,
"learning_rate": 3.235008631805755e-06,
"loss": 0.4001,
"step": 1636
},
{
"epoch": 1.9636,
"grad_norm": 0.6503101793385246,
"learning_rate": 3.228481367083356e-06,
"loss": 0.3758,
"step": 1637
},
{
"epoch": 1.9647999999999999,
"grad_norm": 0.9270107254023242,
"learning_rate": 3.2219575529593017e-06,
"loss": 0.3106,
"step": 1638
},
{
"epoch": 1.966,
"grad_norm": 0.7788450071786787,
"learning_rate": 3.215437202140803e-06,
"loss": 0.389,
"step": 1639
},
{
"epoch": 1.9672,
"grad_norm": 0.702106950356252,
"learning_rate": 3.2089203273283253e-06,
"loss": 0.3667,
"step": 1640
},
{
"epoch": 1.9684,
"grad_norm": 0.8392518843220431,
"learning_rate": 3.2024069412155632e-06,
"loss": 0.3664,
"step": 1641
},
{
"epoch": 1.9696,
"grad_norm": 0.7795564137950384,
"learning_rate": 3.1958970564894187e-06,
"loss": 0.3744,
"step": 1642
},
{
"epoch": 1.9708,
"grad_norm": 0.7324435322162108,
"learning_rate": 3.189390685829967e-06,
"loss": 0.3845,
"step": 1643
},
{
"epoch": 1.972,
"grad_norm": 0.6918841535363706,
"learning_rate": 3.182887841910448e-06,
"loss": 0.3702,
"step": 1644
},
{
"epoch": 1.9731999999999998,
"grad_norm": 0.7584187301414655,
"learning_rate": 3.1763885373972246e-06,
"loss": 0.3401,
"step": 1645
},
{
"epoch": 1.9744000000000002,
"grad_norm": 0.696194921602405,
"learning_rate": 3.1698927849497683e-06,
"loss": 0.3633,
"step": 1646
},
{
"epoch": 1.9756,
"grad_norm": 0.7091014598795434,
"learning_rate": 3.1634005972206326e-06,
"loss": 0.41,
"step": 1647
},
{
"epoch": 1.9768,
"grad_norm": 0.5921181380832845,
"learning_rate": 3.156911986855425e-06,
"loss": 0.3401,
"step": 1648
},
{
"epoch": 1.978,
"grad_norm": 0.742554525953543,
"learning_rate": 3.150426966492788e-06,
"loss": 0.3532,
"step": 1649
},
{
"epoch": 1.9792,
"grad_norm": 0.85996580914657,
"learning_rate": 3.143945548764371e-06,
"loss": 0.3715,
"step": 1650
},
{
"epoch": 1.9804,
"grad_norm": 0.6095687507079254,
"learning_rate": 3.137467746294803e-06,
"loss": 0.4024,
"step": 1651
},
{
"epoch": 1.9816,
"grad_norm": 0.6718934302506764,
"learning_rate": 3.130993571701674e-06,
"loss": 0.3394,
"step": 1652
},
{
"epoch": 1.9828000000000001,
"grad_norm": 0.6700246083900112,
"learning_rate": 3.124523037595506e-06,
"loss": 0.2728,
"step": 1653
},
{
"epoch": 1.984,
"grad_norm": 0.702099731141579,
"learning_rate": 3.1180561565797323e-06,
"loss": 0.3374,
"step": 1654
},
{
"epoch": 1.9851999999999999,
"grad_norm": 0.8933267279000984,
"learning_rate": 3.1115929412506698e-06,
"loss": 0.3776,
"step": 1655
},
{
"epoch": 1.9864000000000002,
"grad_norm": 0.5241277443240402,
"learning_rate": 3.1051334041974923e-06,
"loss": 0.4039,
"step": 1656
},
{
"epoch": 1.9876,
"grad_norm": 0.7381972311949766,
"learning_rate": 3.0986775580022122e-06,
"loss": 0.4262,
"step": 1657
},
{
"epoch": 1.9888,
"grad_norm": 0.6075028876491545,
"learning_rate": 3.092225415239652e-06,
"loss": 0.3298,
"step": 1658
},
{
"epoch": 1.99,
"grad_norm": 0.7165526147804472,
"learning_rate": 3.0857769884774192e-06,
"loss": 0.3583,
"step": 1659
},
{
"epoch": 1.9912,
"grad_norm": 0.8198449632290874,
"learning_rate": 3.079332290275887e-06,
"loss": 0.3411,
"step": 1660
},
{
"epoch": 1.9924,
"grad_norm": 0.7238280300687973,
"learning_rate": 3.0728913331881638e-06,
"loss": 0.3241,
"step": 1661
},
{
"epoch": 1.9936,
"grad_norm": 0.684262838887691,
"learning_rate": 3.0664541297600682e-06,
"loss": 0.3212,
"step": 1662
},
{
"epoch": 1.9948000000000001,
"grad_norm": 0.768576388568188,
"learning_rate": 3.0600206925301114e-06,
"loss": 0.3568,
"step": 1663
},
{
"epoch": 1.996,
"grad_norm": 0.7111529284017039,
"learning_rate": 3.053591034029465e-06,
"loss": 0.3473,
"step": 1664
},
{
"epoch": 1.9971999999999999,
"grad_norm": 0.6375303422643437,
"learning_rate": 3.0471651667819447e-06,
"loss": 0.3114,
"step": 1665
},
{
"epoch": 1.9984,
"grad_norm": 0.5598727106338746,
"learning_rate": 3.0407431033039795e-06,
"loss": 0.3479,
"step": 1666
},
{
"epoch": 1.9996,
"grad_norm": 0.6711270230573488,
"learning_rate": 3.03432485610459e-06,
"loss": 0.3774,
"step": 1667
},
{
"epoch": 2.0,
"grad_norm": 0.6711270230573488,
"learning_rate": 3.0279104376853592e-06,
"loss": 0.4298,
"step": 1668
},
{
"epoch": 2.0012,
"grad_norm": 2.088369452122096,
"learning_rate": 3.0214998605404165e-06,
"loss": 0.2852,
"step": 1669
},
{
"epoch": 2.0024,
"grad_norm": 0.5561564449411506,
"learning_rate": 3.0150931371564107e-06,
"loss": 0.2591,
"step": 1670
},
{
"epoch": 2.0036,
"grad_norm": 0.7440791884682637,
"learning_rate": 3.0086902800124806e-06,
"loss": 0.2979,
"step": 1671
},
{
"epoch": 2.0048,
"grad_norm": 0.6378315440468416,
"learning_rate": 3.0022913015802363e-06,
"loss": 0.2947,
"step": 1672
},
{
"epoch": 2.006,
"grad_norm": 0.6507512061325199,
"learning_rate": 2.99589621432373e-06,
"loss": 0.2914,
"step": 1673
},
{
"epoch": 2.0072,
"grad_norm": 0.6159474928102585,
"learning_rate": 2.9895050306994385e-06,
"loss": 0.2776,
"step": 1674
},
{
"epoch": 2.0084,
"grad_norm": 0.9573641975416705,
"learning_rate": 2.9831177631562306e-06,
"loss": 0.2918,
"step": 1675
},
{
"epoch": 2.0096,
"grad_norm": 0.6883010316359878,
"learning_rate": 2.9767344241353535e-06,
"loss": 0.327,
"step": 1676
},
{
"epoch": 2.0108,
"grad_norm": 0.8738388152711194,
"learning_rate": 2.9703550260703974e-06,
"loss": 0.2676,
"step": 1677
},
{
"epoch": 2.012,
"grad_norm": 0.8364767292160958,
"learning_rate": 2.9639795813872773e-06,
"loss": 0.3148,
"step": 1678
},
{
"epoch": 2.0132,
"grad_norm": 0.8932434141483901,
"learning_rate": 2.9576081025042068e-06,
"loss": 0.2948,
"step": 1679
},
{
"epoch": 2.0144,
"grad_norm": 0.8044203548025202,
"learning_rate": 2.9512406018316763e-06,
"loss": 0.2835,
"step": 1680
},
{
"epoch": 2.0156,
"grad_norm": 0.6953800843065969,
"learning_rate": 2.9448770917724296e-06,
"loss": 0.2691,
"step": 1681
},
{
"epoch": 2.0168,
"grad_norm": 0.7027206845080346,
"learning_rate": 2.9385175847214325e-06,
"loss": 0.3092,
"step": 1682
},
{
"epoch": 2.018,
"grad_norm": 0.7252506485023554,
"learning_rate": 2.9321620930658578e-06,
"loss": 0.3099,
"step": 1683
},
{
"epoch": 2.0192,
"grad_norm": 0.8515277646662316,
"learning_rate": 2.925810629185054e-06,
"loss": 0.2676,
"step": 1684
},
{
"epoch": 2.0204,
"grad_norm": 0.718041615039499,
"learning_rate": 2.919463205450526e-06,
"loss": 0.2791,
"step": 1685
},
{
"epoch": 2.0216,
"grad_norm": 0.7265318730736446,
"learning_rate": 2.9131198342259065e-06,
"loss": 0.2438,
"step": 1686
},
{
"epoch": 2.0228,
"grad_norm": 0.6818182762937451,
"learning_rate": 2.9067805278669425e-06,
"loss": 0.3098,
"step": 1687
},
{
"epoch": 2.024,
"grad_norm": 0.7197065218941066,
"learning_rate": 2.900445298721455e-06,
"loss": 0.2713,
"step": 1688
},
{
"epoch": 2.0252,
"grad_norm": 0.7526478629550866,
"learning_rate": 2.894114159129324e-06,
"loss": 0.3372,
"step": 1689
},
{
"epoch": 2.0264,
"grad_norm": 0.8350459183661488,
"learning_rate": 2.8877871214224694e-06,
"loss": 0.3121,
"step": 1690
},
{
"epoch": 2.0276,
"grad_norm": 0.6298488241393917,
"learning_rate": 2.881464197924814e-06,
"loss": 0.3561,
"step": 1691
},
{
"epoch": 2.0288,
"grad_norm": 1.1309515846393525,
"learning_rate": 2.875145400952274e-06,
"loss": 0.2881,
"step": 1692
},
{
"epoch": 2.03,
"grad_norm": 0.8515842432298777,
"learning_rate": 2.868830742812726e-06,
"loss": 0.2732,
"step": 1693
},
{
"epoch": 2.0312,
"grad_norm": 0.7006011006382655,
"learning_rate": 2.8625202358059806e-06,
"loss": 0.2453,
"step": 1694
},
{
"epoch": 2.0324,
"grad_norm": 0.6979609734294181,
"learning_rate": 2.8562138922237648e-06,
"loss": 0.2696,
"step": 1695
},
{
"epoch": 2.0336,
"grad_norm": 0.6810615498598592,
"learning_rate": 2.8499117243496986e-06,
"loss": 0.2684,
"step": 1696
},
{
"epoch": 2.0348,
"grad_norm": 0.6756424643798261,
"learning_rate": 2.8436137444592694e-06,
"loss": 0.3272,
"step": 1697
},
{
"epoch": 2.036,
"grad_norm": 0.7791266925165868,
"learning_rate": 2.837319964819801e-06,
"loss": 0.2684,
"step": 1698
},
{
"epoch": 2.0372,
"grad_norm": 0.7500073993712527,
"learning_rate": 2.8310303976904396e-06,
"loss": 0.2873,
"step": 1699
},
{
"epoch": 2.0384,
"grad_norm": 0.8026970313609865,
"learning_rate": 2.824745055322128e-06,
"loss": 0.3128,
"step": 1700
},
{
"epoch": 2.0396,
"grad_norm": 0.5896306621015587,
"learning_rate": 2.818463949957575e-06,
"loss": 0.2719,
"step": 1701
},
{
"epoch": 2.0408,
"grad_norm": 0.7481018871116664,
"learning_rate": 2.8121870938312413e-06,
"loss": 0.2317,
"step": 1702
},
{
"epoch": 2.042,
"grad_norm": 0.6014083478843647,
"learning_rate": 2.80591449916931e-06,
"loss": 0.2893,
"step": 1703
},
{
"epoch": 2.0432,
"grad_norm": 0.7081711057341615,
"learning_rate": 2.7996461781896624e-06,
"loss": 0.2812,
"step": 1704
},
{
"epoch": 2.0444,
"grad_norm": 0.7282248248728276,
"learning_rate": 2.7933821431018523e-06,
"loss": 0.2932,
"step": 1705
},
{
"epoch": 2.0456,
"grad_norm": 0.6193604749941026,
"learning_rate": 2.7871224061070935e-06,
"loss": 0.2918,
"step": 1706
},
{
"epoch": 2.0468,
"grad_norm": 0.6938516722296971,
"learning_rate": 2.780866979398218e-06,
"loss": 0.2983,
"step": 1707
},
{
"epoch": 2.048,
"grad_norm": 0.7700462439975865,
"learning_rate": 2.77461587515967e-06,
"loss": 0.2592,
"step": 1708
},
{
"epoch": 2.0492,
"grad_norm": 0.8302766598987845,
"learning_rate": 2.7683691055674745e-06,
"loss": 0.2806,
"step": 1709
},
{
"epoch": 2.0504,
"grad_norm": 0.5823216453276686,
"learning_rate": 2.7621266827892062e-06,
"loss": 0.3313,
"step": 1710
},
{
"epoch": 2.0516,
"grad_norm": 0.79299123397595,
"learning_rate": 2.755888618983977e-06,
"loss": 0.263,
"step": 1711
},
{
"epoch": 2.0528,
"grad_norm": 0.4918042847721973,
"learning_rate": 2.749654926302412e-06,
"loss": 0.2563,
"step": 1712
},
{
"epoch": 2.054,
"grad_norm": 0.8321914344703125,
"learning_rate": 2.743425616886615e-06,
"loss": 0.3057,
"step": 1713
},
{
"epoch": 2.0552,
"grad_norm": 0.8125758226854091,
"learning_rate": 2.737200702870157e-06,
"loss": 0.322,
"step": 1714
},
{
"epoch": 2.0564,
"grad_norm": 0.5827419593278992,
"learning_rate": 2.7309801963780485e-06,
"loss": 0.3061,
"step": 1715
},
{
"epoch": 2.0576,
"grad_norm": 0.7329105490389991,
"learning_rate": 2.724764109526711e-06,
"loss": 0.2817,
"step": 1716
},
{
"epoch": 2.0588,
"grad_norm": 0.5891476434904772,
"learning_rate": 2.7185524544239567e-06,
"loss": 0.2979,
"step": 1717
},
{
"epoch": 2.06,
"grad_norm": 0.7231651638442215,
"learning_rate": 2.71234524316897e-06,
"loss": 0.2625,
"step": 1718
},
{
"epoch": 2.0612,
"grad_norm": 0.5829736091670309,
"learning_rate": 2.706142487852279e-06,
"loss": 0.2972,
"step": 1719
},
{
"epoch": 2.0624,
"grad_norm": 0.6510641374838887,
"learning_rate": 2.699944200555727e-06,
"loss": 0.2826,
"step": 1720
},
{
"epoch": 2.0636,
"grad_norm": 0.7805456437827156,
"learning_rate": 2.693750393352462e-06,
"loss": 0.2935,
"step": 1721
},
{
"epoch": 2.0648,
"grad_norm": 0.7234539030166994,
"learning_rate": 2.6875610783069007e-06,
"loss": 0.2802,
"step": 1722
},
{
"epoch": 2.066,
"grad_norm": 0.6518428326297813,
"learning_rate": 2.681376267474707e-06,
"loss": 0.276,
"step": 1723
},
{
"epoch": 2.0672,
"grad_norm": 0.8118603417654962,
"learning_rate": 2.67519597290278e-06,
"loss": 0.2765,
"step": 1724
},
{
"epoch": 2.0684,
"grad_norm": 0.620459560194058,
"learning_rate": 2.669020206629217e-06,
"loss": 0.3143,
"step": 1725
},
{
"epoch": 2.0696,
"grad_norm": 0.8203183116848812,
"learning_rate": 2.6628489806832947e-06,
"loss": 0.2779,
"step": 1726
},
{
"epoch": 2.0708,
"grad_norm": 0.701101587558589,
"learning_rate": 2.6566823070854442e-06,
"loss": 0.2817,
"step": 1727
},
{
"epoch": 2.072,
"grad_norm": 0.7777053700979897,
"learning_rate": 2.650520197847235e-06,
"loss": 0.3399,
"step": 1728
},
{
"epoch": 2.0732,
"grad_norm": 0.7803521268234632,
"learning_rate": 2.6443626649713407e-06,
"loss": 0.306,
"step": 1729
},
{
"epoch": 2.0744,
"grad_norm": 0.7405853173809038,
"learning_rate": 2.6382097204515246e-06,
"loss": 0.2955,
"step": 1730
},
{
"epoch": 2.0756,
"grad_norm": 0.7181807623900187,
"learning_rate": 2.6320613762726123e-06,
"loss": 0.2708,
"step": 1731
},
{
"epoch": 2.0768,
"grad_norm": 0.6306618390660236,
"learning_rate": 2.625917644410467e-06,
"loss": 0.3078,
"step": 1732
},
{
"epoch": 2.078,
"grad_norm": 0.5694293811138348,
"learning_rate": 2.6197785368319663e-06,
"loss": 0.3046,
"step": 1733
},
{
"epoch": 2.0792,
"grad_norm": 0.7199508535949585,
"learning_rate": 2.613644065494985e-06,
"loss": 0.3065,
"step": 1734
},
{
"epoch": 2.0804,
"grad_norm": 0.7411352973204174,
"learning_rate": 2.6075142423483675e-06,
"loss": 0.3014,
"step": 1735
},
{
"epoch": 2.0816,
"grad_norm": 0.6526901038864603,
"learning_rate": 2.6013890793318972e-06,
"loss": 0.3008,
"step": 1736
},
{
"epoch": 2.0828,
"grad_norm": 0.8585255946691019,
"learning_rate": 2.5952685883762918e-06,
"loss": 0.2627,
"step": 1737
},
{
"epoch": 2.084,
"grad_norm": 0.76562343351442,
"learning_rate": 2.589152781403158e-06,
"loss": 0.2706,
"step": 1738
},
{
"epoch": 2.0852,
"grad_norm": 0.8244010900941466,
"learning_rate": 2.583041670324982e-06,
"loss": 0.2608,
"step": 1739
},
{
"epoch": 2.0864,
"grad_norm": 0.8633880182633827,
"learning_rate": 2.5769352670451058e-06,
"loss": 0.2872,
"step": 1740
},
{
"epoch": 2.0876,
"grad_norm": 0.7509328824199477,
"learning_rate": 2.5708335834577035e-06,
"loss": 0.2528,
"step": 1741
},
{
"epoch": 2.0888,
"grad_norm": 0.640204504112653,
"learning_rate": 2.5647366314477473e-06,
"loss": 0.2541,
"step": 1742
},
{
"epoch": 2.09,
"grad_norm": 0.737327852708041,
"learning_rate": 2.5586444228910036e-06,
"loss": 0.241,
"step": 1743
},
{
"epoch": 2.0912,
"grad_norm": 0.6924468824743977,
"learning_rate": 2.5525569696539916e-06,
"loss": 0.2935,
"step": 1744
},
{
"epoch": 2.0924,
"grad_norm": 0.751568838354342,
"learning_rate": 2.546474283593969e-06,
"loss": 0.2469,
"step": 1745
},
{
"epoch": 2.0936,
"grad_norm": 0.6208807893184197,
"learning_rate": 2.540396376558912e-06,
"loss": 0.3044,
"step": 1746
},
{
"epoch": 2.0948,
"grad_norm": 0.8013228913513426,
"learning_rate": 2.5343232603874868e-06,
"loss": 0.3169,
"step": 1747
},
{
"epoch": 2.096,
"grad_norm": 0.7152792486363544,
"learning_rate": 2.5282549469090246e-06,
"loss": 0.2967,
"step": 1748
},
{
"epoch": 2.0972,
"grad_norm": 0.7146544884255634,
"learning_rate": 2.522191447943506e-06,
"loss": 0.3025,
"step": 1749
},
{
"epoch": 2.0984,
"grad_norm": 0.7449341525237713,
"learning_rate": 2.5161327753015297e-06,
"loss": 0.2906,
"step": 1750
},
{
"epoch": 2.0996,
"grad_norm": 1.1237173554790165,
"learning_rate": 2.5100789407842985e-06,
"loss": 0.2932,
"step": 1751
},
{
"epoch": 2.1008,
"grad_norm": 0.9051098497011659,
"learning_rate": 2.5040299561835846e-06,
"loss": 0.2742,
"step": 1752
},
{
"epoch": 2.102,
"grad_norm": 0.655795428124948,
"learning_rate": 2.4979858332817225e-06,
"loss": 0.2649,
"step": 1753
},
{
"epoch": 2.1032,
"grad_norm": 0.6472420685217467,
"learning_rate": 2.4919465838515687e-06,
"loss": 0.2525,
"step": 1754
},
{
"epoch": 2.1044,
"grad_norm": 0.7912575799269113,
"learning_rate": 2.48591221965649e-06,
"loss": 0.2716,
"step": 1755
},
{
"epoch": 2.1056,
"grad_norm": 0.7260798235944638,
"learning_rate": 2.479882752450339e-06,
"loss": 0.277,
"step": 1756
},
{
"epoch": 2.1068,
"grad_norm": 0.725874816511404,
"learning_rate": 2.4738581939774303e-06,
"loss": 0.2971,
"step": 1757
},
{
"epoch": 2.108,
"grad_norm": 0.6861462665757562,
"learning_rate": 2.4678385559725125e-06,
"loss": 0.2586,
"step": 1758
},
{
"epoch": 2.1092,
"grad_norm": 0.6360990779836201,
"learning_rate": 2.4618238501607577e-06,
"loss": 0.2876,
"step": 1759
},
{
"epoch": 2.1104,
"grad_norm": 0.7175549317243431,
"learning_rate": 2.455814088257723e-06,
"loss": 0.2923,
"step": 1760
},
{
"epoch": 2.1116,
"grad_norm": 0.8692093619211555,
"learning_rate": 2.4498092819693364e-06,
"loss": 0.3034,
"step": 1761
},
{
"epoch": 2.1128,
"grad_norm": 0.6315391457039741,
"learning_rate": 2.443809442991884e-06,
"loss": 0.2607,
"step": 1762
},
{
"epoch": 2.114,
"grad_norm": 0.6613238955831777,
"learning_rate": 2.4378145830119637e-06,
"loss": 0.303,
"step": 1763
},
{
"epoch": 2.1152,
"grad_norm": 0.60882360407893,
"learning_rate": 2.4318247137064788e-06,
"loss": 0.2763,
"step": 1764
},
{
"epoch": 2.1164,
"grad_norm": 0.7581308041292392,
"learning_rate": 2.425839846742616e-06,
"loss": 0.2757,
"step": 1765
},
{
"epoch": 2.1176,
"grad_norm": 0.5852519523930492,
"learning_rate": 2.4198599937778138e-06,
"loss": 0.2558,
"step": 1766
},
{
"epoch": 2.1188,
"grad_norm": 0.6410924648095844,
"learning_rate": 2.4138851664597424e-06,
"loss": 0.3076,
"step": 1767
},
{
"epoch": 2.12,
"grad_norm": 0.7220533600955701,
"learning_rate": 2.407915376426293e-06,
"loss": 0.3543,
"step": 1768
},
{
"epoch": 2.1212,
"grad_norm": 0.9925534477896854,
"learning_rate": 2.401950635305535e-06,
"loss": 0.2822,
"step": 1769
},
{
"epoch": 2.1224,
"grad_norm": 0.6010082568458979,
"learning_rate": 2.395990954715705e-06,
"loss": 0.3,
"step": 1770
},
{
"epoch": 2.1236,
"grad_norm": 0.6772771268050389,
"learning_rate": 2.390036346265188e-06,
"loss": 0.2682,
"step": 1771
},
{
"epoch": 2.1248,
"grad_norm": 0.883085647229646,
"learning_rate": 2.3840868215524824e-06,
"loss": 0.2711,
"step": 1772
},
{
"epoch": 2.126,
"grad_norm": 0.6991509288620459,
"learning_rate": 2.378142392166191e-06,
"loss": 0.2775,
"step": 1773
},
{
"epoch": 2.1272,
"grad_norm": 0.7394418188851275,
"learning_rate": 2.3722030696849857e-06,
"loss": 0.2272,
"step": 1774
},
{
"epoch": 2.1284,
"grad_norm": 0.6767374928736871,
"learning_rate": 2.3662688656775973e-06,
"loss": 0.282,
"step": 1775
},
{
"epoch": 2.1296,
"grad_norm": 0.7938466899804603,
"learning_rate": 2.3603397917027787e-06,
"loss": 0.2505,
"step": 1776
},
{
"epoch": 2.1308,
"grad_norm": 0.5508815681221602,
"learning_rate": 2.3544158593092986e-06,
"loss": 0.2311,
"step": 1777
},
{
"epoch": 2.132,
"grad_norm": 0.7402173195257308,
"learning_rate": 2.3484970800359087e-06,
"loss": 0.321,
"step": 1778
},
{
"epoch": 2.1332,
"grad_norm": 0.788018919475665,
"learning_rate": 2.34258346541132e-06,
"loss": 0.2751,
"step": 1779
},
{
"epoch": 2.1344,
"grad_norm": 0.7614342441818248,
"learning_rate": 2.3366750269541833e-06,
"loss": 0.3374,
"step": 1780
},
{
"epoch": 2.1356,
"grad_norm": 0.7013037268831918,
"learning_rate": 2.3307717761730745e-06,
"loss": 0.2683,
"step": 1781
},
{
"epoch": 2.1368,
"grad_norm": 0.7765438931188983,
"learning_rate": 2.3248737245664575e-06,
"loss": 0.3049,
"step": 1782
},
{
"epoch": 2.138,
"grad_norm": 0.5891414203800416,
"learning_rate": 2.318980883622668e-06,
"loss": 0.2637,
"step": 1783
},
{
"epoch": 2.1391999999999998,
"grad_norm": 0.8316964675465408,
"learning_rate": 2.313093264819903e-06,
"loss": 0.3101,
"step": 1784
},
{
"epoch": 2.1404,
"grad_norm": 0.6828643746261911,
"learning_rate": 2.3072108796261766e-06,
"loss": 0.3074,
"step": 1785
},
{
"epoch": 2.1416,
"grad_norm": 0.5496333788415517,
"learning_rate": 2.301333739499312e-06,
"loss": 0.2626,
"step": 1786
},
{
"epoch": 2.1428,
"grad_norm": 0.7086988626220481,
"learning_rate": 2.2954618558869194e-06,
"loss": 0.3075,
"step": 1787
},
{
"epoch": 2.144,
"grad_norm": 0.8133278905796454,
"learning_rate": 2.2895952402263642e-06,
"loss": 0.2301,
"step": 1788
},
{
"epoch": 2.1452,
"grad_norm": 0.8154510835501877,
"learning_rate": 2.283733903944756e-06,
"loss": 0.2967,
"step": 1789
},
{
"epoch": 2.1464,
"grad_norm": 0.6612967304831779,
"learning_rate": 2.2778778584589214e-06,
"loss": 0.2829,
"step": 1790
},
{
"epoch": 2.1476,
"grad_norm": 0.5242174256478419,
"learning_rate": 2.272027115175377e-06,
"loss": 0.2547,
"step": 1791
},
{
"epoch": 2.1488,
"grad_norm": 0.7600242351372234,
"learning_rate": 2.2661816854903117e-06,
"loss": 0.2702,
"step": 1792
},
{
"epoch": 2.15,
"grad_norm": 0.5906405354069276,
"learning_rate": 2.2603415807895718e-06,
"loss": 0.2633,
"step": 1793
},
{
"epoch": 2.1512000000000002,
"grad_norm": 0.6371494504277992,
"learning_rate": 2.254506812448622e-06,
"loss": 0.2885,
"step": 1794
},
{
"epoch": 2.1524,
"grad_norm": 0.7179442422892502,
"learning_rate": 2.2486773918325394e-06,
"loss": 0.3211,
"step": 1795
},
{
"epoch": 2.1536,
"grad_norm": 0.7906465435219664,
"learning_rate": 2.242853330295984e-06,
"loss": 0.3202,
"step": 1796
},
{
"epoch": 2.1548,
"grad_norm": 0.7021915009307176,
"learning_rate": 2.2370346391831737e-06,
"loss": 0.2871,
"step": 1797
},
{
"epoch": 2.156,
"grad_norm": 0.5871223870432859,
"learning_rate": 2.231221329827867e-06,
"loss": 0.3026,
"step": 1798
},
{
"epoch": 2.1572,
"grad_norm": 0.7154693930814482,
"learning_rate": 2.225413413553341e-06,
"loss": 0.317,
"step": 1799
},
{
"epoch": 2.1584,
"grad_norm": 0.5566077554453305,
"learning_rate": 2.219610901672371e-06,
"loss": 0.2601,
"step": 1800
},
{
"epoch": 2.1596,
"grad_norm": 0.6578125790406493,
"learning_rate": 2.2138138054871993e-06,
"loss": 0.2983,
"step": 1801
},
{
"epoch": 2.1608,
"grad_norm": 0.7562626622528975,
"learning_rate": 2.208022136289521e-06,
"loss": 0.2663,
"step": 1802
},
{
"epoch": 2.162,
"grad_norm": 0.8441121082239703,
"learning_rate": 2.2022359053604654e-06,
"loss": 0.3195,
"step": 1803
},
{
"epoch": 2.1632,
"grad_norm": 0.8816430714434005,
"learning_rate": 2.1964551239705604e-06,
"loss": 0.3354,
"step": 1804
},
{
"epoch": 2.1644,
"grad_norm": 0.6338643614201181,
"learning_rate": 2.1906798033797276e-06,
"loss": 0.3156,
"step": 1805
},
{
"epoch": 2.1656,
"grad_norm": 0.7635069452199187,
"learning_rate": 2.1849099548372492e-06,
"loss": 0.2368,
"step": 1806
},
{
"epoch": 2.1668,
"grad_norm": 0.7316591764576956,
"learning_rate": 2.179145589581747e-06,
"loss": 0.2926,
"step": 1807
},
{
"epoch": 2.168,
"grad_norm": 0.7051700121997925,
"learning_rate": 2.1733867188411606e-06,
"loss": 0.2823,
"step": 1808
},
{
"epoch": 2.1692,
"grad_norm": 0.709484279758578,
"learning_rate": 2.167633353832734e-06,
"loss": 0.2645,
"step": 1809
},
{
"epoch": 2.1704,
"grad_norm": 0.7197156077723833,
"learning_rate": 2.1618855057629804e-06,
"loss": 0.271,
"step": 1810
},
{
"epoch": 2.1716,
"grad_norm": 0.8340895779020381,
"learning_rate": 2.156143185827671e-06,
"loss": 0.264,
"step": 1811
},
{
"epoch": 2.1728,
"grad_norm": 0.6944784966795656,
"learning_rate": 2.1504064052118095e-06,
"loss": 0.2973,
"step": 1812
},
{
"epoch": 2.174,
"grad_norm": 0.670276707335882,
"learning_rate": 2.144675175089606e-06,
"loss": 0.3105,
"step": 1813
},
{
"epoch": 2.1752,
"grad_norm": 0.8391449179253878,
"learning_rate": 2.1389495066244613e-06,
"loss": 0.2808,
"step": 1814
},
{
"epoch": 2.1764,
"grad_norm": 0.8351881849061288,
"learning_rate": 2.1332294109689446e-06,
"loss": 0.3319,
"step": 1815
},
{
"epoch": 2.1776,
"grad_norm": 0.7438241895896566,
"learning_rate": 2.127514899264771e-06,
"loss": 0.2851,
"step": 1816
},
{
"epoch": 2.1788,
"grad_norm": 0.6772561771249624,
"learning_rate": 2.1218059826427727e-06,
"loss": 0.3204,
"step": 1817
},
{
"epoch": 2.18,
"grad_norm": 0.8526883989017642,
"learning_rate": 2.1161026722228932e-06,
"loss": 0.3187,
"step": 1818
},
{
"epoch": 2.1812,
"grad_norm": 0.6241125173815714,
"learning_rate": 2.110404979114149e-06,
"loss": 0.3176,
"step": 1819
},
{
"epoch": 2.1824,
"grad_norm": 0.6812650572627527,
"learning_rate": 2.104712914414615e-06,
"loss": 0.3095,
"step": 1820
},
{
"epoch": 2.1836,
"grad_norm": 0.6656341317851194,
"learning_rate": 2.0990264892114067e-06,
"loss": 0.3297,
"step": 1821
},
{
"epoch": 2.1848,
"grad_norm": 0.7611078703441717,
"learning_rate": 2.093345714580656e-06,
"loss": 0.2431,
"step": 1822
},
{
"epoch": 2.186,
"grad_norm": 0.6473924522377008,
"learning_rate": 2.0876706015874816e-06,
"loss": 0.2876,
"step": 1823
},
{
"epoch": 2.1872,
"grad_norm": 0.7487061668174932,
"learning_rate": 2.0820011612859825e-06,
"loss": 0.2562,
"step": 1824
},
{
"epoch": 2.1884,
"grad_norm": 0.5865184392303474,
"learning_rate": 2.076337404719203e-06,
"loss": 0.2771,
"step": 1825
},
{
"epoch": 2.1896,
"grad_norm": 0.6771666864570925,
"learning_rate": 2.0706793429191156e-06,
"loss": 0.2598,
"step": 1826
},
{
"epoch": 2.1908,
"grad_norm": 0.656708483933799,
"learning_rate": 2.0650269869066048e-06,
"loss": 0.2835,
"step": 1827
},
{
"epoch": 2.192,
"grad_norm": 0.8168599494819258,
"learning_rate": 2.0593803476914407e-06,
"loss": 0.2844,
"step": 1828
},
{
"epoch": 2.1932,
"grad_norm": 0.6889662817500772,
"learning_rate": 2.053739436272256e-06,
"loss": 0.3048,
"step": 1829
},
{
"epoch": 2.1944,
"grad_norm": 0.9175631703446326,
"learning_rate": 2.0481042636365243e-06,
"loss": 0.2857,
"step": 1830
},
{
"epoch": 2.1955999999999998,
"grad_norm": 0.8891259675906453,
"learning_rate": 2.0424748407605468e-06,
"loss": 0.3049,
"step": 1831
},
{
"epoch": 2.1968,
"grad_norm": 0.6110155526268235,
"learning_rate": 2.036851178609423e-06,
"loss": 0.2505,
"step": 1832
},
{
"epoch": 2.198,
"grad_norm": 0.73796003339032,
"learning_rate": 2.0312332881370294e-06,
"loss": 0.2914,
"step": 1833
},
{
"epoch": 2.1992,
"grad_norm": 0.6218594994549229,
"learning_rate": 2.0256211802860044e-06,
"loss": 0.2569,
"step": 1834
},
{
"epoch": 2.2004,
"grad_norm": 0.6849603502355964,
"learning_rate": 2.0200148659877185e-06,
"loss": 0.2698,
"step": 1835
},
{
"epoch": 2.2016,
"grad_norm": 0.6429947313388805,
"learning_rate": 2.014414356162258e-06,
"loss": 0.2705,
"step": 1836
},
{
"epoch": 2.2028,
"grad_norm": 0.8274500284226253,
"learning_rate": 2.0088196617184065e-06,
"loss": 0.2803,
"step": 1837
},
{
"epoch": 2.204,
"grad_norm": 0.6397435814669205,
"learning_rate": 2.00323079355362e-06,
"loss": 0.305,
"step": 1838
},
{
"epoch": 2.2052,
"grad_norm": 0.7610089715152695,
"learning_rate": 1.997647762554e-06,
"loss": 0.3498,
"step": 1839
},
{
"epoch": 2.2064,
"grad_norm": 0.5943590401830786,
"learning_rate": 1.992070579594288e-06,
"loss": 0.3143,
"step": 1840
},
{
"epoch": 2.2076000000000002,
"grad_norm": 0.8433484497599082,
"learning_rate": 1.9864992555378256e-06,
"loss": 0.2632,
"step": 1841
},
{
"epoch": 2.2088,
"grad_norm": 0.6036340932811731,
"learning_rate": 1.9809338012365438e-06,
"loss": 0.3074,
"step": 1842
},
{
"epoch": 2.21,
"grad_norm": 0.647884703677156,
"learning_rate": 1.9753742275309456e-06,
"loss": 0.2395,
"step": 1843
},
{
"epoch": 2.2112,
"grad_norm": 0.6185760283233283,
"learning_rate": 1.9698205452500772e-06,
"loss": 0.3169,
"step": 1844
},
{
"epoch": 2.2124,
"grad_norm": 0.9099157814535219,
"learning_rate": 1.9642727652115056e-06,
"loss": 0.2483,
"step": 1845
},
{
"epoch": 2.2136,
"grad_norm": 0.8361158735964174,
"learning_rate": 1.9587308982213077e-06,
"loss": 0.2702,
"step": 1846
},
{
"epoch": 2.2148,
"grad_norm": 0.6086207877959438,
"learning_rate": 1.953194955074038e-06,
"loss": 0.2739,
"step": 1847
},
{
"epoch": 2.216,
"grad_norm": 0.5403614640498644,
"learning_rate": 1.9476649465527116e-06,
"loss": 0.2796,
"step": 1848
},
{
"epoch": 2.2172,
"grad_norm": 0.7504018798795475,
"learning_rate": 1.942140883428788e-06,
"loss": 0.277,
"step": 1849
},
{
"epoch": 2.2184,
"grad_norm": 0.6379662110895271,
"learning_rate": 1.936622776462147e-06,
"loss": 0.3248,
"step": 1850
},
{
"epoch": 2.2196,
"grad_norm": 0.8542354755351312,
"learning_rate": 1.931110636401062e-06,
"loss": 0.2867,
"step": 1851
},
{
"epoch": 2.2208,
"grad_norm": 0.7181598277022851,
"learning_rate": 1.925604473982185e-06,
"loss": 0.2979,
"step": 1852
},
{
"epoch": 2.222,
"grad_norm": 0.647888141789453,
"learning_rate": 1.9201042999305276e-06,
"loss": 0.2694,
"step": 1853
},
{
"epoch": 2.2232,
"grad_norm": 0.7805346416274643,
"learning_rate": 1.914610124959437e-06,
"loss": 0.3243,
"step": 1854
},
{
"epoch": 2.2244,
"grad_norm": 0.7353350119958244,
"learning_rate": 1.9091219597705694e-06,
"loss": 0.2751,
"step": 1855
},
{
"epoch": 2.2256,
"grad_norm": 0.6593436345241795,
"learning_rate": 1.9036398150538842e-06,
"loss": 0.3018,
"step": 1856
},
{
"epoch": 2.2268,
"grad_norm": 0.6859886533762563,
"learning_rate": 1.898163701487607e-06,
"loss": 0.2859,
"step": 1857
},
{
"epoch": 2.228,
"grad_norm": 0.5951955127891551,
"learning_rate": 1.8926936297382148e-06,
"loss": 0.2828,
"step": 1858
},
{
"epoch": 2.2292,
"grad_norm": 0.8274440824574081,
"learning_rate": 1.8872296104604255e-06,
"loss": 0.3149,
"step": 1859
},
{
"epoch": 2.2304,
"grad_norm": 0.7207270015387346,
"learning_rate": 1.8817716542971593e-06,
"loss": 0.2613,
"step": 1860
},
{
"epoch": 2.2316,
"grad_norm": 0.5987868956518425,
"learning_rate": 1.8763197718795262e-06,
"loss": 0.3082,
"step": 1861
},
{
"epoch": 2.2328,
"grad_norm": 0.5349292112782891,
"learning_rate": 1.8708739738268133e-06,
"loss": 0.2406,
"step": 1862
},
{
"epoch": 2.234,
"grad_norm": 0.5856463634698124,
"learning_rate": 1.865434270746449e-06,
"loss": 0.2995,
"step": 1863
},
{
"epoch": 2.2352,
"grad_norm": 0.8113351613082591,
"learning_rate": 1.8600006732339892e-06,
"loss": 0.2701,
"step": 1864
},
{
"epoch": 2.2364,
"grad_norm": 0.7182990837688242,
"learning_rate": 1.8545731918731074e-06,
"loss": 0.2888,
"step": 1865
},
{
"epoch": 2.2376,
"grad_norm": 0.6683437338135383,
"learning_rate": 1.8491518372355538e-06,
"loss": 0.3004,
"step": 1866
},
{
"epoch": 2.2388,
"grad_norm": 0.6599850741932629,
"learning_rate": 1.8437366198811463e-06,
"loss": 0.2447,
"step": 1867
},
{
"epoch": 2.24,
"grad_norm": 0.7966684127956547,
"learning_rate": 1.838327550357753e-06,
"loss": 0.2733,
"step": 1868
},
{
"epoch": 2.2412,
"grad_norm": 0.6347808892892448,
"learning_rate": 1.8329246392012622e-06,
"loss": 0.2998,
"step": 1869
},
{
"epoch": 2.2424,
"grad_norm": 0.7727908627108967,
"learning_rate": 1.8275278969355714e-06,
"loss": 0.2899,
"step": 1870
},
{
"epoch": 2.2436,
"grad_norm": 0.7056919720061404,
"learning_rate": 1.8221373340725568e-06,
"loss": 0.2798,
"step": 1871
},
{
"epoch": 2.2448,
"grad_norm": 0.7652901526102439,
"learning_rate": 1.8167529611120648e-06,
"loss": 0.2552,
"step": 1872
},
{
"epoch": 2.246,
"grad_norm": 0.6658323896736708,
"learning_rate": 1.811374788541878e-06,
"loss": 0.3065,
"step": 1873
},
{
"epoch": 2.2472,
"grad_norm": 0.6756337631673197,
"learning_rate": 1.8060028268377088e-06,
"loss": 0.3146,
"step": 1874
},
{
"epoch": 2.2484,
"grad_norm": 0.7313270034785085,
"learning_rate": 1.8006370864631644e-06,
"loss": 0.2822,
"step": 1875
},
{
"epoch": 2.2496,
"grad_norm": 0.773726426466683,
"learning_rate": 1.7952775778697418e-06,
"loss": 0.2935,
"step": 1876
},
{
"epoch": 2.2508,
"grad_norm": 0.7534988748324787,
"learning_rate": 1.7899243114967918e-06,
"loss": 0.2735,
"step": 1877
},
{
"epoch": 2.252,
"grad_norm": 0.7938988074048295,
"learning_rate": 1.7845772977715148e-06,
"loss": 0.283,
"step": 1878
},
{
"epoch": 2.2532,
"grad_norm": 0.8083844884771482,
"learning_rate": 1.7792365471089252e-06,
"loss": 0.2874,
"step": 1879
},
{
"epoch": 2.2544,
"grad_norm": 0.6791262156568534,
"learning_rate": 1.773902069911838e-06,
"loss": 0.283,
"step": 1880
},
{
"epoch": 2.2556,
"grad_norm": 0.7901693510364795,
"learning_rate": 1.7685738765708576e-06,
"loss": 0.3211,
"step": 1881
},
{
"epoch": 2.2568,
"grad_norm": 0.893363798816337,
"learning_rate": 1.7632519774643391e-06,
"loss": 0.2882,
"step": 1882
},
{
"epoch": 2.258,
"grad_norm": 0.8670923261885365,
"learning_rate": 1.7579363829583794e-06,
"loss": 0.3109,
"step": 1883
},
{
"epoch": 2.2592,
"grad_norm": 0.6675878413132018,
"learning_rate": 1.7526271034067993e-06,
"loss": 0.3041,
"step": 1884
},
{
"epoch": 2.2604,
"grad_norm": 0.7257865155768792,
"learning_rate": 1.7473241491511139e-06,
"loss": 0.2708,
"step": 1885
},
{
"epoch": 2.2616,
"grad_norm": 0.7823832867140552,
"learning_rate": 1.7420275305205214e-06,
"loss": 0.3344,
"step": 1886
},
{
"epoch": 2.2628,
"grad_norm": 0.8889890716505195,
"learning_rate": 1.7367372578318797e-06,
"loss": 0.315,
"step": 1887
},
{
"epoch": 2.2640000000000002,
"grad_norm": 0.6901081847170341,
"learning_rate": 1.7314533413896833e-06,
"loss": 0.2849,
"step": 1888
},
{
"epoch": 2.2652,
"grad_norm": 0.6840560171759852,
"learning_rate": 1.7261757914860456e-06,
"loss": 0.2789,
"step": 1889
},
{
"epoch": 2.2664,
"grad_norm": 0.9294459322393186,
"learning_rate": 1.720904618400684e-06,
"loss": 0.2518,
"step": 1890
},
{
"epoch": 2.2676,
"grad_norm": 0.7256231058411816,
"learning_rate": 1.7156398324008871e-06,
"loss": 0.3139,
"step": 1891
},
{
"epoch": 2.2688,
"grad_norm": 0.6595550983659683,
"learning_rate": 1.7103814437415105e-06,
"loss": 0.2973,
"step": 1892
},
{
"epoch": 2.27,
"grad_norm": 0.7848010198737841,
"learning_rate": 1.7051294626649462e-06,
"loss": 0.3323,
"step": 1893
},
{
"epoch": 2.2712,
"grad_norm": 0.7011947209303638,
"learning_rate": 1.6998838994011041e-06,
"loss": 0.3126,
"step": 1894
},
{
"epoch": 2.2724,
"grad_norm": 0.7584979584742341,
"learning_rate": 1.6946447641673907e-06,
"loss": 0.2961,
"step": 1895
},
{
"epoch": 2.2736,
"grad_norm": 0.8700419166998146,
"learning_rate": 1.6894120671686986e-06,
"loss": 0.2667,
"step": 1896
},
{
"epoch": 2.2748,
"grad_norm": 0.7342578995193595,
"learning_rate": 1.6841858185973775e-06,
"loss": 0.3064,
"step": 1897
},
{
"epoch": 2.276,
"grad_norm": 0.6625055327325216,
"learning_rate": 1.6789660286332132e-06,
"loss": 0.3123,
"step": 1898
},
{
"epoch": 2.2772,
"grad_norm": 0.8631997705494986,
"learning_rate": 1.6737527074434135e-06,
"loss": 0.2708,
"step": 1899
},
{
"epoch": 2.2784,
"grad_norm": 0.7428492853450093,
"learning_rate": 1.6685458651825892e-06,
"loss": 0.2716,
"step": 1900
},
{
"epoch": 2.2796,
"grad_norm": 0.7957024527866422,
"learning_rate": 1.6633455119927256e-06,
"loss": 0.271,
"step": 1901
},
{
"epoch": 2.2808,
"grad_norm": 0.4581419890819959,
"learning_rate": 1.658151658003172e-06,
"loss": 0.244,
"step": 1902
},
{
"epoch": 2.282,
"grad_norm": 0.7620101015250027,
"learning_rate": 1.6529643133306212e-06,
"loss": 0.2799,
"step": 1903
},
{
"epoch": 2.2832,
"grad_norm": 0.8465395698220026,
"learning_rate": 1.647783488079081e-06,
"loss": 0.314,
"step": 1904
},
{
"epoch": 2.2843999999999998,
"grad_norm": 0.6311371766338576,
"learning_rate": 1.6426091923398619e-06,
"loss": 0.2577,
"step": 1905
},
{
"epoch": 2.2856,
"grad_norm": 0.7919059517850136,
"learning_rate": 1.6374414361915613e-06,
"loss": 0.2761,
"step": 1906
},
{
"epoch": 2.2868,
"grad_norm": 0.673990577847982,
"learning_rate": 1.6322802297000306e-06,
"loss": 0.2914,
"step": 1907
},
{
"epoch": 2.288,
"grad_norm": 0.5642478939526163,
"learning_rate": 1.6271255829183702e-06,
"loss": 0.3147,
"step": 1908
},
{
"epoch": 2.2892,
"grad_norm": 1.0378407321023961,
"learning_rate": 1.6219775058869019e-06,
"loss": 0.2694,
"step": 1909
},
{
"epoch": 2.2904,
"grad_norm": 0.6633672664252135,
"learning_rate": 1.6168360086331498e-06,
"loss": 0.3239,
"step": 1910
},
{
"epoch": 2.2916,
"grad_norm": 0.9311767871994348,
"learning_rate": 1.6117011011718188e-06,
"loss": 0.2931,
"step": 1911
},
{
"epoch": 2.2928,
"grad_norm": 0.6499635767024712,
"learning_rate": 1.6065727935047837e-06,
"loss": 0.2627,
"step": 1912
},
{
"epoch": 2.294,
"grad_norm": 0.6976767341801269,
"learning_rate": 1.6014510956210632e-06,
"loss": 0.2723,
"step": 1913
},
{
"epoch": 2.2952,
"grad_norm": 0.6480574877139175,
"learning_rate": 1.5963360174967956e-06,
"loss": 0.2625,
"step": 1914
},
{
"epoch": 2.2964,
"grad_norm": 0.5193166847468137,
"learning_rate": 1.5912275690952339e-06,
"loss": 0.2752,
"step": 1915
},
{
"epoch": 2.2976,
"grad_norm": 0.9880632678614614,
"learning_rate": 1.5861257603667106e-06,
"loss": 0.3063,
"step": 1916
},
{
"epoch": 2.2988,
"grad_norm": 0.6232836881233889,
"learning_rate": 1.581030601248626e-06,
"loss": 0.2619,
"step": 1917
},
{
"epoch": 2.3,
"grad_norm": 0.7112786281924534,
"learning_rate": 1.5759421016654314e-06,
"loss": 0.2707,
"step": 1918
},
{
"epoch": 2.3012,
"grad_norm": 0.6819932418504141,
"learning_rate": 1.570860271528607e-06,
"loss": 0.2867,
"step": 1919
},
{
"epoch": 2.3024,
"grad_norm": 0.6003576139061961,
"learning_rate": 1.5657851207366359e-06,
"loss": 0.2929,
"step": 1920
},
{
"epoch": 2.3036,
"grad_norm": 0.6947627829064076,
"learning_rate": 1.5607166591749995e-06,
"loss": 0.2656,
"step": 1921
},
{
"epoch": 2.3048,
"grad_norm": 0.6857549212204418,
"learning_rate": 1.555654896716144e-06,
"loss": 0.254,
"step": 1922
},
{
"epoch": 2.306,
"grad_norm": 0.6900890058761181,
"learning_rate": 1.5505998432194658e-06,
"loss": 0.2679,
"step": 1923
},
{
"epoch": 2.3072,
"grad_norm": 0.7382586820968852,
"learning_rate": 1.5455515085312984e-06,
"loss": 0.2627,
"step": 1924
},
{
"epoch": 2.3084,
"grad_norm": 0.7767762490963626,
"learning_rate": 1.5405099024848874e-06,
"loss": 0.298,
"step": 1925
},
{
"epoch": 2.3096,
"grad_norm": 0.6252185095990261,
"learning_rate": 1.5354750349003694e-06,
"loss": 0.3028,
"step": 1926
},
{
"epoch": 2.3108,
"grad_norm": 0.6665246763076556,
"learning_rate": 1.5304469155847556e-06,
"loss": 0.2437,
"step": 1927
},
{
"epoch": 2.312,
"grad_norm": 0.8548779889182841,
"learning_rate": 1.5254255543319168e-06,
"loss": 0.2813,
"step": 1928
},
{
"epoch": 2.3132,
"grad_norm": 0.6664193785215454,
"learning_rate": 1.5204109609225553e-06,
"loss": 0.3154,
"step": 1929
},
{
"epoch": 2.3144,
"grad_norm": 0.8821706046881854,
"learning_rate": 1.5154031451241952e-06,
"loss": 0.2914,
"step": 1930
},
{
"epoch": 2.3156,
"grad_norm": 0.6296342955721203,
"learning_rate": 1.5104021166911582e-06,
"loss": 0.2577,
"step": 1931
},
{
"epoch": 2.3168,
"grad_norm": 0.6345640643435636,
"learning_rate": 1.5054078853645432e-06,
"loss": 0.3166,
"step": 1932
},
{
"epoch": 2.318,
"grad_norm": 0.7169567485669481,
"learning_rate": 1.5004204608722088e-06,
"loss": 0.2843,
"step": 1933
},
{
"epoch": 2.3192,
"grad_norm": 0.6509071318888697,
"learning_rate": 1.495439852928759e-06,
"loss": 0.3006,
"step": 1934
},
{
"epoch": 2.3204000000000002,
"grad_norm": 0.8148382854447535,
"learning_rate": 1.4904660712355207e-06,
"loss": 0.2572,
"step": 1935
},
{
"epoch": 2.3216,
"grad_norm": 0.7569136780175667,
"learning_rate": 1.4854991254805179e-06,
"loss": 0.2862,
"step": 1936
},
{
"epoch": 2.3228,
"grad_norm": 0.7273179120649447,
"learning_rate": 1.4805390253384683e-06,
"loss": 0.3164,
"step": 1937
},
{
"epoch": 2.324,
"grad_norm": 0.6319665719458061,
"learning_rate": 1.4755857804707485e-06,
"loss": 0.2566,
"step": 1938
},
{
"epoch": 2.3252,
"grad_norm": 1.0384474187927952,
"learning_rate": 1.4706394005253838e-06,
"loss": 0.3157,
"step": 1939
},
{
"epoch": 2.3264,
"grad_norm": 0.7059273483025585,
"learning_rate": 1.465699895137031e-06,
"loss": 0.2832,
"step": 1940
},
{
"epoch": 2.3276,
"grad_norm": 0.6621089291386077,
"learning_rate": 1.4607672739269552e-06,
"loss": 0.2625,
"step": 1941
},
{
"epoch": 2.3288,
"grad_norm": 0.6277983545311215,
"learning_rate": 1.455841546503009e-06,
"loss": 0.2828,
"step": 1942
},
{
"epoch": 2.33,
"grad_norm": 0.7619292179064054,
"learning_rate": 1.450922722459623e-06,
"loss": 0.3014,
"step": 1943
},
{
"epoch": 2.3312,
"grad_norm": 0.7567263028275791,
"learning_rate": 1.446010811377776e-06,
"loss": 0.2593,
"step": 1944
},
{
"epoch": 2.3324,
"grad_norm": 0.734405816026238,
"learning_rate": 1.4411058228249824e-06,
"loss": 0.2549,
"step": 1945
},
{
"epoch": 2.3336,
"grad_norm": 0.7379103797300406,
"learning_rate": 1.4362077663552754e-06,
"loss": 0.2643,
"step": 1946
},
{
"epoch": 2.3348,
"grad_norm": 0.7077193481421317,
"learning_rate": 1.4313166515091863e-06,
"loss": 0.2789,
"step": 1947
},
{
"epoch": 2.336,
"grad_norm": 1.0498393111561082,
"learning_rate": 1.4264324878137204e-06,
"loss": 0.276,
"step": 1948
},
{
"epoch": 2.3372,
"grad_norm": 0.5860394123173929,
"learning_rate": 1.421555284782349e-06,
"loss": 0.2847,
"step": 1949
},
{
"epoch": 2.3384,
"grad_norm": 0.621941365141143,
"learning_rate": 1.4166850519149794e-06,
"loss": 0.28,
"step": 1950
},
{
"epoch": 2.3396,
"grad_norm": 0.6376731544752134,
"learning_rate": 1.41182179869795e-06,
"loss": 0.2783,
"step": 1951
},
{
"epoch": 2.3407999999999998,
"grad_norm": 0.8364388218860074,
"learning_rate": 1.406965534603995e-06,
"loss": 0.3154,
"step": 1952
},
{
"epoch": 2.342,
"grad_norm": 0.6550151033568089,
"learning_rate": 1.4021162690922441e-06,
"loss": 0.2526,
"step": 1953
},
{
"epoch": 2.3432,
"grad_norm": 0.649414510416142,
"learning_rate": 1.397274011608189e-06,
"loss": 0.2754,
"step": 1954
},
{
"epoch": 2.3444,
"grad_norm": 0.589693248381491,
"learning_rate": 1.3924387715836706e-06,
"loss": 0.2491,
"step": 1955
},
{
"epoch": 2.3456,
"grad_norm": 0.8047434513027613,
"learning_rate": 1.3876105584368653e-06,
"loss": 0.2889,
"step": 1956
},
{
"epoch": 2.3468,
"grad_norm": 0.48218905825243197,
"learning_rate": 1.3827893815722614e-06,
"loss": 0.2536,
"step": 1957
},
{
"epoch": 2.348,
"grad_norm": 0.6357509907948024,
"learning_rate": 1.3779752503806375e-06,
"loss": 0.2316,
"step": 1958
},
{
"epoch": 2.3492,
"grad_norm": 0.8108875927633574,
"learning_rate": 1.3731681742390558e-06,
"loss": 0.2525,
"step": 1959
},
{
"epoch": 2.3504,
"grad_norm": 0.7037602443503415,
"learning_rate": 1.368368162510829e-06,
"loss": 0.289,
"step": 1960
},
{
"epoch": 2.3516,
"grad_norm": 0.7739474880274559,
"learning_rate": 1.363575224545512e-06,
"loss": 0.3668,
"step": 1961
},
{
"epoch": 2.3528000000000002,
"grad_norm": 0.9076400276831817,
"learning_rate": 1.3587893696788868e-06,
"loss": 0.2787,
"step": 1962
},
{
"epoch": 2.354,
"grad_norm": 0.6227070548651138,
"learning_rate": 1.3540106072329323e-06,
"loss": 0.3123,
"step": 1963
},
{
"epoch": 2.3552,
"grad_norm": 0.8796791668598759,
"learning_rate": 1.349238946515813e-06,
"loss": 0.3307,
"step": 1964
},
{
"epoch": 2.3564,
"grad_norm": 0.6218747279651277,
"learning_rate": 1.344474396821865e-06,
"loss": 0.3133,
"step": 1965
},
{
"epoch": 2.3576,
"grad_norm": 0.905080923193005,
"learning_rate": 1.3397169674315668e-06,
"loss": 0.3189,
"step": 1966
},
{
"epoch": 2.3588,
"grad_norm": 0.6463226917186492,
"learning_rate": 1.3349666676115358e-06,
"loss": 0.3353,
"step": 1967
},
{
"epoch": 2.36,
"grad_norm": 0.6541872884530091,
"learning_rate": 1.3302235066144948e-06,
"loss": 0.2865,
"step": 1968
},
{
"epoch": 2.3612,
"grad_norm": 0.8372226528374508,
"learning_rate": 1.3254874936792672e-06,
"loss": 0.2549,
"step": 1969
},
{
"epoch": 2.3624,
"grad_norm": 0.8054755370594131,
"learning_rate": 1.3207586380307486e-06,
"loss": 0.3126,
"step": 1970
},
{
"epoch": 2.3636,
"grad_norm": 0.8405395878830167,
"learning_rate": 1.3160369488798984e-06,
"loss": 0.2506,
"step": 1971
},
{
"epoch": 2.3648,
"grad_norm": 1.0329887770740844,
"learning_rate": 1.3113224354237113e-06,
"loss": 0.3138,
"step": 1972
},
{
"epoch": 2.366,
"grad_norm": 0.5937700397369654,
"learning_rate": 1.306615106845211e-06,
"loss": 0.3056,
"step": 1973
},
{
"epoch": 2.3672,
"grad_norm": 0.8082980475262066,
"learning_rate": 1.30191497231342e-06,
"loss": 0.3055,
"step": 1974
},
{
"epoch": 2.3684,
"grad_norm": 0.812801536809135,
"learning_rate": 1.2972220409833552e-06,
"loss": 0.263,
"step": 1975
},
{
"epoch": 2.3696,
"grad_norm": 1.2201434471974846,
"learning_rate": 1.2925363219959958e-06,
"loss": 0.323,
"step": 1976
},
{
"epoch": 2.3708,
"grad_norm": 0.6392942054287046,
"learning_rate": 1.2878578244782775e-06,
"loss": 0.2815,
"step": 1977
},
{
"epoch": 2.372,
"grad_norm": 0.67801591623462,
"learning_rate": 1.2831865575430702e-06,
"loss": 0.2575,
"step": 1978
},
{
"epoch": 2.3731999999999998,
"grad_norm": 0.5174853545586786,
"learning_rate": 1.2785225302891568e-06,
"loss": 0.252,
"step": 1979
},
{
"epoch": 2.3744,
"grad_norm": 0.8619288184637666,
"learning_rate": 1.2738657518012188e-06,
"loss": 0.2918,
"step": 1980
},
{
"epoch": 2.3756,
"grad_norm": 0.6797498832794855,
"learning_rate": 1.2692162311498219e-06,
"loss": 0.3012,
"step": 1981
},
{
"epoch": 2.3768000000000002,
"grad_norm": 0.569674999830851,
"learning_rate": 1.2645739773913911e-06,
"loss": 0.2822,
"step": 1982
},
{
"epoch": 2.378,
"grad_norm": 0.6999831720903421,
"learning_rate": 1.259938999568196e-06,
"loss": 0.2759,
"step": 1983
},
{
"epoch": 2.3792,
"grad_norm": 0.6751324677088292,
"learning_rate": 1.2553113067083417e-06,
"loss": 0.3047,
"step": 1984
},
{
"epoch": 2.3804,
"grad_norm": 0.6251905499803202,
"learning_rate": 1.2506909078257357e-06,
"loss": 0.313,
"step": 1985
},
{
"epoch": 2.3816,
"grad_norm": 0.7518640766189668,
"learning_rate": 1.2460778119200778e-06,
"loss": 0.3075,
"step": 1986
},
{
"epoch": 2.3828,
"grad_norm": 0.7834689820257049,
"learning_rate": 1.24147202797685e-06,
"loss": 0.2658,
"step": 1987
},
{
"epoch": 2.384,
"grad_norm": 0.6172343974604968,
"learning_rate": 1.236873564967284e-06,
"loss": 0.2544,
"step": 1988
},
{
"epoch": 2.3852,
"grad_norm": 0.6803854143699946,
"learning_rate": 1.2322824318483568e-06,
"loss": 0.2999,
"step": 1989
},
{
"epoch": 2.3864,
"grad_norm": 0.8399502551831883,
"learning_rate": 1.227698637562768e-06,
"loss": 0.2966,
"step": 1990
},
{
"epoch": 2.3876,
"grad_norm": 0.6327179320629635,
"learning_rate": 1.2231221910389196e-06,
"loss": 0.2652,
"step": 1991
},
{
"epoch": 2.3888,
"grad_norm": 0.5762942388963672,
"learning_rate": 1.2185531011909008e-06,
"loss": 0.2736,
"step": 1992
},
{
"epoch": 2.39,
"grad_norm": 0.6733186897217389,
"learning_rate": 1.2139913769184757e-06,
"loss": 0.2903,
"step": 1993
},
{
"epoch": 2.3912,
"grad_norm": 0.8662934583246499,
"learning_rate": 1.2094370271070599e-06,
"loss": 0.2405,
"step": 1994
},
{
"epoch": 2.3924,
"grad_norm": 0.7067638717213011,
"learning_rate": 1.2048900606277036e-06,
"loss": 0.2638,
"step": 1995
},
{
"epoch": 2.3936,
"grad_norm": 0.5658655563921311,
"learning_rate": 1.2003504863370746e-06,
"loss": 0.3214,
"step": 1996
},
{
"epoch": 2.3948,
"grad_norm": 0.8571904375989193,
"learning_rate": 1.195818313077447e-06,
"loss": 0.2836,
"step": 1997
},
{
"epoch": 2.396,
"grad_norm": 0.6389422877351538,
"learning_rate": 1.1912935496766719e-06,
"loss": 0.3409,
"step": 1998
},
{
"epoch": 2.3971999999999998,
"grad_norm": 0.8688551612486514,
"learning_rate": 1.186776204948173e-06,
"loss": 0.2815,
"step": 1999
},
{
"epoch": 2.3984,
"grad_norm": 0.8445469852739739,
"learning_rate": 1.182266287690924e-06,
"loss": 0.2593,
"step": 2000
},
{
"epoch": 2.3996,
"grad_norm": 0.5779515338289782,
"learning_rate": 1.177763806689427e-06,
"loss": 0.3476,
"step": 2001
},
{
"epoch": 2.4008,
"grad_norm": 0.7445104336110696,
"learning_rate": 1.173268770713701e-06,
"loss": 0.2857,
"step": 2002
},
{
"epoch": 2.402,
"grad_norm": 0.8145480406714878,
"learning_rate": 1.1687811885192662e-06,
"loss": 0.2666,
"step": 2003
},
{
"epoch": 2.4032,
"grad_norm": 0.5915872831485786,
"learning_rate": 1.16430106884712e-06,
"loss": 0.2695,
"step": 2004
},
{
"epoch": 2.4044,
"grad_norm": 0.7490718525844583,
"learning_rate": 1.159828420423728e-06,
"loss": 0.2828,
"step": 2005
},
{
"epoch": 2.4056,
"grad_norm": 0.5262397078241221,
"learning_rate": 1.1553632519610025e-06,
"loss": 0.3214,
"step": 2006
},
{
"epoch": 2.4068,
"grad_norm": 0.8523398376907122,
"learning_rate": 1.1509055721562839e-06,
"loss": 0.2618,
"step": 2007
},
{
"epoch": 2.408,
"grad_norm": 0.7369560457966171,
"learning_rate": 1.1464553896923264e-06,
"loss": 0.2856,
"step": 2008
},
{
"epoch": 2.4092000000000002,
"grad_norm": 0.569409159019725,
"learning_rate": 1.1420127132372839e-06,
"loss": 0.2944,
"step": 2009
},
{
"epoch": 2.4104,
"grad_norm": 0.7811538679253816,
"learning_rate": 1.1375775514446846e-06,
"loss": 0.2885,
"step": 2010
},
{
"epoch": 2.4116,
"grad_norm": 0.7555226696391547,
"learning_rate": 1.1331499129534252e-06,
"loss": 0.2785,
"step": 2011
},
{
"epoch": 2.4128,
"grad_norm": 0.7825402630380734,
"learning_rate": 1.128729806387746e-06,
"loss": 0.2752,
"step": 2012
},
{
"epoch": 2.414,
"grad_norm": 0.6701124740770267,
"learning_rate": 1.124317240357216e-06,
"loss": 0.3275,
"step": 2013
},
{
"epoch": 2.4152,
"grad_norm": 0.8297310356676278,
"learning_rate": 1.119912223456715e-06,
"loss": 0.2743,
"step": 2014
},
{
"epoch": 2.4164,
"grad_norm": 0.8909057488955949,
"learning_rate": 1.1155147642664217e-06,
"loss": 0.217,
"step": 2015
},
{
"epoch": 2.4176,
"grad_norm": 0.669707464447291,
"learning_rate": 1.1111248713517935e-06,
"loss": 0.2859,
"step": 2016
},
{
"epoch": 2.4188,
"grad_norm": 0.6861740430825445,
"learning_rate": 1.1067425532635463e-06,
"loss": 0.3392,
"step": 2017
},
{
"epoch": 2.42,
"grad_norm": 0.6950841573092722,
"learning_rate": 1.1023678185376474e-06,
"loss": 0.2807,
"step": 2018
},
{
"epoch": 2.4212,
"grad_norm": 0.5401800599781725,
"learning_rate": 1.0980006756952882e-06,
"loss": 0.2795,
"step": 2019
},
{
"epoch": 2.4224,
"grad_norm": 0.7589595024931296,
"learning_rate": 1.0936411332428732e-06,
"loss": 0.2947,
"step": 2020
},
{
"epoch": 2.4236,
"grad_norm": 0.8375346540653694,
"learning_rate": 1.089289199672004e-06,
"loss": 0.255,
"step": 2021
},
{
"epoch": 2.4248,
"grad_norm": 0.7110966509065856,
"learning_rate": 1.084944883459464e-06,
"loss": 0.2463,
"step": 2022
},
{
"epoch": 2.426,
"grad_norm": 0.6843419130025422,
"learning_rate": 1.0806081930671947e-06,
"loss": 0.28,
"step": 2023
},
{
"epoch": 2.4272,
"grad_norm": 0.7358091967950351,
"learning_rate": 1.0762791369422838e-06,
"loss": 0.2678,
"step": 2024
},
{
"epoch": 2.4284,
"grad_norm": 0.7170805544237613,
"learning_rate": 1.0719577235169537e-06,
"loss": 0.2923,
"step": 2025
},
{
"epoch": 2.4295999999999998,
"grad_norm": 0.6449025487261146,
"learning_rate": 1.0676439612085353e-06,
"loss": 0.2956,
"step": 2026
},
{
"epoch": 2.4308,
"grad_norm": 0.6131605076859993,
"learning_rate": 1.0633378584194593e-06,
"loss": 0.2719,
"step": 2027
},
{
"epoch": 2.432,
"grad_norm": 0.6854558731769252,
"learning_rate": 1.059039423537237e-06,
"loss": 0.2751,
"step": 2028
},
{
"epoch": 2.4332,
"grad_norm": 0.7441942074784614,
"learning_rate": 1.054748664934443e-06,
"loss": 0.2678,
"step": 2029
},
{
"epoch": 2.4344,
"grad_norm": 0.6154652130644275,
"learning_rate": 1.0504655909686978e-06,
"loss": 0.3044,
"step": 2030
},
{
"epoch": 2.4356,
"grad_norm": 0.7769916562713324,
"learning_rate": 1.0461902099826577e-06,
"loss": 0.2886,
"step": 2031
},
{
"epoch": 2.4368,
"grad_norm": 0.7028161424135243,
"learning_rate": 1.0419225303039943e-06,
"loss": 0.2935,
"step": 2032
},
{
"epoch": 2.438,
"grad_norm": 0.7455961272815085,
"learning_rate": 1.0376625602453733e-06,
"loss": 0.2942,
"step": 2033
},
{
"epoch": 2.4392,
"grad_norm": 0.7693200515865598,
"learning_rate": 1.0334103081044504e-06,
"loss": 0.3024,
"step": 2034
},
{
"epoch": 2.4404,
"grad_norm": 0.834870555733633,
"learning_rate": 1.0291657821638435e-06,
"loss": 0.3076,
"step": 2035
},
{
"epoch": 2.4416,
"grad_norm": 0.5863974693483013,
"learning_rate": 1.024928990691121e-06,
"loss": 0.2655,
"step": 2036
},
{
"epoch": 2.4428,
"grad_norm": 0.8923941410337408,
"learning_rate": 1.0206999419387881e-06,
"loss": 0.2877,
"step": 2037
},
{
"epoch": 2.444,
"grad_norm": 0.5152051594999155,
"learning_rate": 1.0164786441442698e-06,
"loss": 0.2588,
"step": 2038
},
{
"epoch": 2.4452,
"grad_norm": 0.705328176983332,
"learning_rate": 1.0122651055298898e-06,
"loss": 0.2828,
"step": 2039
},
{
"epoch": 2.4464,
"grad_norm": 0.6922311972593358,
"learning_rate": 1.0080593343028621e-06,
"loss": 0.2962,
"step": 2040
},
{
"epoch": 2.4476,
"grad_norm": 0.6946956800765861,
"learning_rate": 1.0038613386552687e-06,
"loss": 0.3236,
"step": 2041
},
{
"epoch": 2.4488,
"grad_norm": 0.7170957195319527,
"learning_rate": 9.996711267640451e-07,
"loss": 0.286,
"step": 2042
},
{
"epoch": 2.45,
"grad_norm": 0.8082056123277392,
"learning_rate": 9.95488706790969e-07,
"loss": 0.2909,
"step": 2043
},
{
"epoch": 2.4512,
"grad_norm": 0.6440378623344876,
"learning_rate": 9.913140868826405e-07,
"loss": 0.3071,
"step": 2044
},
{
"epoch": 2.4524,
"grad_norm": 0.656154819170875,
"learning_rate": 9.871472751704625e-07,
"loss": 0.2895,
"step": 2045
},
{
"epoch": 2.4536,
"grad_norm": 0.7038432455716981,
"learning_rate": 9.829882797706336e-07,
"loss": 0.2759,
"step": 2046
},
{
"epoch": 2.4548,
"grad_norm": 0.602369968610495,
"learning_rate": 9.788371087841236e-07,
"loss": 0.2815,
"step": 2047
},
{
"epoch": 2.456,
"grad_norm": 0.7131028794621941,
"learning_rate": 9.74693770296667e-07,
"loss": 0.2484,
"step": 2048
},
{
"epoch": 2.4572,
"grad_norm": 0.7807557505486811,
"learning_rate": 9.705582723787348e-07,
"loss": 0.3032,
"step": 2049
},
{
"epoch": 2.4584,
"grad_norm": 0.5963541440355288,
"learning_rate": 9.664306230855342e-07,
"loss": 0.2775,
"step": 2050
},
{
"epoch": 2.4596,
"grad_norm": 0.7353952102035798,
"learning_rate": 9.623108304569783e-07,
"loss": 0.2962,
"step": 2051
},
{
"epoch": 2.4608,
"grad_norm": 0.5806118102123087,
"learning_rate": 9.58198902517678e-07,
"loss": 0.2751,
"step": 2052
},
{
"epoch": 2.462,
"grad_norm": 0.6487802104846568,
"learning_rate": 9.540948472769278e-07,
"loss": 0.2668,
"step": 2053
},
{
"epoch": 2.4632,
"grad_norm": 0.6544901046541113,
"learning_rate": 9.499986727286869e-07,
"loss": 0.3344,
"step": 2054
},
{
"epoch": 2.4644,
"grad_norm": 0.5902441028069662,
"learning_rate": 9.459103868515618e-07,
"loss": 0.2946,
"step": 2055
},
{
"epoch": 2.4656000000000002,
"grad_norm": 0.6430741589709079,
"learning_rate": 9.418299976087964e-07,
"loss": 0.2912,
"step": 2056
},
{
"epoch": 2.4668,
"grad_norm": 0.6757650668894376,
"learning_rate": 9.377575129482513e-07,
"loss": 0.347,
"step": 2057
},
{
"epoch": 2.468,
"grad_norm": 0.6955861650212085,
"learning_rate": 9.336929408023887e-07,
"loss": 0.2478,
"step": 2058
},
{
"epoch": 2.4692,
"grad_norm": 0.6909760911131515,
"learning_rate": 9.29636289088266e-07,
"loss": 0.2757,
"step": 2059
},
{
"epoch": 2.4704,
"grad_norm": 0.7247381989416801,
"learning_rate": 9.255875657075053e-07,
"loss": 0.2743,
"step": 2060
},
{
"epoch": 2.4716,
"grad_norm": 0.6104150276815804,
"learning_rate": 9.215467785462873e-07,
"loss": 0.2687,
"step": 2061
},
{
"epoch": 2.4728,
"grad_norm": 0.8875531380952814,
"learning_rate": 9.175139354753382e-07,
"loss": 0.3125,
"step": 2062
},
{
"epoch": 2.474,
"grad_norm": 0.587012704415787,
"learning_rate": 9.134890443499068e-07,
"loss": 0.3126,
"step": 2063
},
{
"epoch": 2.4752,
"grad_norm": 0.8960276131342865,
"learning_rate": 9.094721130097517e-07,
"loss": 0.3236,
"step": 2064
},
{
"epoch": 2.4764,
"grad_norm": 0.6677066237398112,
"learning_rate": 9.054631492791344e-07,
"loss": 0.323,
"step": 2065
},
{
"epoch": 2.4776,
"grad_norm": 0.8138869091824592,
"learning_rate": 9.014621609667896e-07,
"loss": 0.3101,
"step": 2066
},
{
"epoch": 2.4788,
"grad_norm": 0.695994033710393,
"learning_rate": 8.974691558659187e-07,
"loss": 0.292,
"step": 2067
},
{
"epoch": 2.48,
"grad_norm": 0.7464338985888993,
"learning_rate": 8.934841417541767e-07,
"loss": 0.3177,
"step": 2068
},
{
"epoch": 2.4812,
"grad_norm": 0.7179201273389993,
"learning_rate": 8.8950712639365e-07,
"loss": 0.2573,
"step": 2069
},
{
"epoch": 2.4824,
"grad_norm": 0.6239073751181373,
"learning_rate": 8.855381175308475e-07,
"loss": 0.2779,
"step": 2070
},
{
"epoch": 2.4836,
"grad_norm": 0.6705852211266631,
"learning_rate": 8.815771228966796e-07,
"loss": 0.229,
"step": 2071
},
{
"epoch": 2.4848,
"grad_norm": 0.6010437210481082,
"learning_rate": 8.776241502064508e-07,
"loss": 0.2696,
"step": 2072
},
{
"epoch": 2.4859999999999998,
"grad_norm": 0.817722650833114,
"learning_rate": 8.736792071598355e-07,
"loss": 0.2859,
"step": 2073
},
{
"epoch": 2.4872,
"grad_norm": 0.9246803995226833,
"learning_rate": 8.697423014408718e-07,
"loss": 0.2592,
"step": 2074
},
{
"epoch": 2.4884,
"grad_norm": 0.8396928399289096,
"learning_rate": 8.658134407179419e-07,
"loss": 0.2967,
"step": 2075
},
{
"epoch": 2.4896,
"grad_norm": 0.7110631181607171,
"learning_rate": 8.61892632643756e-07,
"loss": 0.2873,
"step": 2076
},
{
"epoch": 2.4908,
"grad_norm": 0.8163799837176345,
"learning_rate": 8.579798848553389e-07,
"loss": 0.2754,
"step": 2077
},
{
"epoch": 2.492,
"grad_norm": 0.6107344595413731,
"learning_rate": 8.540752049740181e-07,
"loss": 0.2875,
"step": 2078
},
{
"epoch": 2.4932,
"grad_norm": 0.87485954809143,
"learning_rate": 8.501786006054047e-07,
"loss": 0.2912,
"step": 2079
},
{
"epoch": 2.4944,
"grad_norm": 0.8521492816584161,
"learning_rate": 8.462900793393775e-07,
"loss": 0.3043,
"step": 2080
},
{
"epoch": 2.4956,
"grad_norm": 0.7035639248420411,
"learning_rate": 8.424096487500777e-07,
"loss": 0.2475,
"step": 2081
},
{
"epoch": 2.4968,
"grad_norm": 0.826618071389074,
"learning_rate": 8.385373163958821e-07,
"loss": 0.2918,
"step": 2082
},
{
"epoch": 2.498,
"grad_norm": 0.8282368986244178,
"learning_rate": 8.346730898193928e-07,
"loss": 0.2815,
"step": 2083
},
{
"epoch": 2.4992,
"grad_norm": 0.665650025083666,
"learning_rate": 8.308169765474278e-07,
"loss": 0.3127,
"step": 2084
},
{
"epoch": 2.5004,
"grad_norm": 0.7289215759473803,
"learning_rate": 8.269689840909967e-07,
"loss": 0.3162,
"step": 2085
},
{
"epoch": 2.5016,
"grad_norm": 0.6702416711812764,
"learning_rate": 8.231291199452956e-07,
"loss": 0.2893,
"step": 2086
},
{
"epoch": 2.5028,
"grad_norm": 0.5605293279080651,
"learning_rate": 8.192973915896868e-07,
"loss": 0.2782,
"step": 2087
},
{
"epoch": 2.504,
"grad_norm": 0.714618007562953,
"learning_rate": 8.154738064876843e-07,
"loss": 0.2804,
"step": 2088
},
{
"epoch": 2.5052,
"grad_norm": 0.5311763098335279,
"learning_rate": 8.116583720869398e-07,
"loss": 0.2665,
"step": 2089
},
{
"epoch": 2.5064,
"grad_norm": 0.9287741342433332,
"learning_rate": 8.078510958192337e-07,
"loss": 0.3006,
"step": 2090
},
{
"epoch": 2.5076,
"grad_norm": 0.6342183332153583,
"learning_rate": 8.040519851004492e-07,
"loss": 0.2702,
"step": 2091
},
{
"epoch": 2.5088,
"grad_norm": 0.7689707111081208,
"learning_rate": 8.002610473305688e-07,
"loss": 0.3224,
"step": 2092
},
{
"epoch": 2.51,
"grad_norm": 0.6874244083233649,
"learning_rate": 7.964782898936569e-07,
"loss": 0.299,
"step": 2093
},
{
"epoch": 2.5112,
"grad_norm": 0.6829556628493936,
"learning_rate": 7.927037201578397e-07,
"loss": 0.3013,
"step": 2094
},
{
"epoch": 2.5124,
"grad_norm": 0.7694618330376841,
"learning_rate": 7.889373454752964e-07,
"loss": 0.2938,
"step": 2095
},
{
"epoch": 2.5136,
"grad_norm": 1.0815187990972266,
"learning_rate": 7.851791731822461e-07,
"loss": 0.3559,
"step": 2096
},
{
"epoch": 2.5148,
"grad_norm": 0.7905559868072847,
"learning_rate": 7.814292105989308e-07,
"loss": 0.2904,
"step": 2097
},
{
"epoch": 2.516,
"grad_norm": 0.511129588733436,
"learning_rate": 7.776874650295984e-07,
"loss": 0.2768,
"step": 2098
},
{
"epoch": 2.5172,
"grad_norm": 0.6878681993518705,
"learning_rate": 7.739539437624933e-07,
"loss": 0.3416,
"step": 2099
},
{
"epoch": 2.5183999999999997,
"grad_norm": 0.7005193806252948,
"learning_rate": 7.702286540698417e-07,
"loss": 0.3089,
"step": 2100
},
{
"epoch": 2.5196,
"grad_norm": 0.5784867465602536,
"learning_rate": 7.665116032078346e-07,
"loss": 0.2538,
"step": 2101
},
{
"epoch": 2.5208,
"grad_norm": 0.7252611591775111,
"learning_rate": 7.628027984166153e-07,
"loss": 0.3329,
"step": 2102
},
{
"epoch": 2.5220000000000002,
"grad_norm": 0.786482466324177,
"learning_rate": 7.591022469202675e-07,
"loss": 0.2889,
"step": 2103
},
{
"epoch": 2.5232,
"grad_norm": 0.8603045927226652,
"learning_rate": 7.554099559267964e-07,
"loss": 0.2555,
"step": 2104
},
{
"epoch": 2.5244,
"grad_norm": 0.608935363934432,
"learning_rate": 7.517259326281157e-07,
"loss": 0.2608,
"step": 2105
},
{
"epoch": 2.5256,
"grad_norm": 0.5571172645613067,
"learning_rate": 7.480501842000404e-07,
"loss": 0.2788,
"step": 2106
},
{
"epoch": 2.5268,
"grad_norm": 0.8302854826402234,
"learning_rate": 7.443827178022628e-07,
"loss": 0.2764,
"step": 2107
},
{
"epoch": 2.528,
"grad_norm": 0.8337048250086003,
"learning_rate": 7.407235405783453e-07,
"loss": 0.2681,
"step": 2108
},
{
"epoch": 2.5292,
"grad_norm": 0.7374860754412593,
"learning_rate": 7.370726596557059e-07,
"loss": 0.2973,
"step": 2109
},
{
"epoch": 2.5304,
"grad_norm": 0.7886324400096725,
"learning_rate": 7.334300821455998e-07,
"loss": 0.2564,
"step": 2110
},
{
"epoch": 2.5316,
"grad_norm": 0.7424108790092626,
"learning_rate": 7.297958151431094e-07,
"loss": 0.2635,
"step": 2111
},
{
"epoch": 2.5328,
"grad_norm": 0.7153949891460527,
"learning_rate": 7.26169865727131e-07,
"loss": 0.2598,
"step": 2112
},
{
"epoch": 2.534,
"grad_norm": 0.7752067483688443,
"learning_rate": 7.225522409603608e-07,
"loss": 0.3398,
"step": 2113
},
{
"epoch": 2.5352,
"grad_norm": 1.367675746527483,
"learning_rate": 7.189429478892762e-07,
"loss": 0.2424,
"step": 2114
},
{
"epoch": 2.5364,
"grad_norm": 0.6708852547390795,
"learning_rate": 7.153419935441303e-07,
"loss": 0.2924,
"step": 2115
},
{
"epoch": 2.5376,
"grad_norm": 0.9838915779196031,
"learning_rate": 7.117493849389306e-07,
"loss": 0.271,
"step": 2116
},
{
"epoch": 2.5388,
"grad_norm": 0.7581981061836748,
"learning_rate": 7.081651290714287e-07,
"loss": 0.2439,
"step": 2117
},
{
"epoch": 2.54,
"grad_norm": 0.7967360482223734,
"learning_rate": 7.045892329231086e-07,
"loss": 0.325,
"step": 2118
},
{
"epoch": 2.5412,
"grad_norm": 0.7014415044804616,
"learning_rate": 7.010217034591721e-07,
"loss": 0.2711,
"step": 2119
},
{
"epoch": 2.5423999999999998,
"grad_norm": 0.6558484128618423,
"learning_rate": 6.974625476285191e-07,
"loss": 0.2646,
"step": 2120
},
{
"epoch": 2.5436,
"grad_norm": 0.6518053531301943,
"learning_rate": 6.93911772363745e-07,
"loss": 0.2458,
"step": 2121
},
{
"epoch": 2.5448,
"grad_norm": 0.7137521066073853,
"learning_rate": 6.903693845811176e-07,
"loss": 0.2769,
"step": 2122
},
{
"epoch": 2.5460000000000003,
"grad_norm": 0.7084103756513347,
"learning_rate": 6.86835391180567e-07,
"loss": 0.338,
"step": 2123
},
{
"epoch": 2.5472,
"grad_norm": 1.066975427306229,
"learning_rate": 6.833097990456761e-07,
"loss": 0.2848,
"step": 2124
},
{
"epoch": 2.5484,
"grad_norm": 0.47758746992771056,
"learning_rate": 6.797926150436618e-07,
"loss": 0.2586,
"step": 2125
},
{
"epoch": 2.5496,
"grad_norm": 0.6807923304856558,
"learning_rate": 6.762838460253629e-07,
"loss": 0.3039,
"step": 2126
},
{
"epoch": 2.5507999999999997,
"grad_norm": 0.6827502634864607,
"learning_rate": 6.727834988252258e-07,
"loss": 0.3051,
"step": 2127
},
{
"epoch": 2.552,
"grad_norm": 0.7046804202397908,
"learning_rate": 6.692915802612965e-07,
"loss": 0.2588,
"step": 2128
},
{
"epoch": 2.5532,
"grad_norm": 0.733217068273347,
"learning_rate": 6.658080971352026e-07,
"loss": 0.2528,
"step": 2129
},
{
"epoch": 2.5544000000000002,
"grad_norm": 0.6006628658563726,
"learning_rate": 6.623330562321378e-07,
"loss": 0.272,
"step": 2130
},
{
"epoch": 2.5556,
"grad_norm": 0.6272540434789801,
"learning_rate": 6.588664643208559e-07,
"loss": 0.2952,
"step": 2131
},
{
"epoch": 2.5568,
"grad_norm": 0.8141180419426902,
"learning_rate": 6.554083281536516e-07,
"loss": 0.2715,
"step": 2132
},
{
"epoch": 2.558,
"grad_norm": 0.7769241321788415,
"learning_rate": 6.519586544663481e-07,
"loss": 0.2683,
"step": 2133
},
{
"epoch": 2.5592,
"grad_norm": 0.5174357795010261,
"learning_rate": 6.485174499782876e-07,
"loss": 0.2716,
"step": 2134
},
{
"epoch": 2.5604,
"grad_norm": 0.7234307376621996,
"learning_rate": 6.450847213923162e-07,
"loss": 0.3115,
"step": 2135
},
{
"epoch": 2.5616,
"grad_norm": 0.7424538116702596,
"learning_rate": 6.416604753947675e-07,
"loss": 0.2603,
"step": 2136
},
{
"epoch": 2.5628,
"grad_norm": 0.6683056783089876,
"learning_rate": 6.382447186554553e-07,
"loss": 0.3238,
"step": 2137
},
{
"epoch": 2.564,
"grad_norm": 0.676343364968695,
"learning_rate": 6.348374578276567e-07,
"loss": 0.2905,
"step": 2138
},
{
"epoch": 2.5652,
"grad_norm": 0.6307265316127199,
"learning_rate": 6.314386995480987e-07,
"loss": 0.2682,
"step": 2139
},
{
"epoch": 2.5664,
"grad_norm": 0.8635575798619177,
"learning_rate": 6.280484504369505e-07,
"loss": 0.2974,
"step": 2140
},
{
"epoch": 2.5676,
"grad_norm": 0.7504884771213327,
"learning_rate": 6.246667170978049e-07,
"loss": 0.2431,
"step": 2141
},
{
"epoch": 2.5688,
"grad_norm": 0.8606529987249782,
"learning_rate": 6.212935061176667e-07,
"loss": 0.3264,
"step": 2142
},
{
"epoch": 2.57,
"grad_norm": 0.6350489914515207,
"learning_rate": 6.179288240669429e-07,
"loss": 0.2755,
"step": 2143
},
{
"epoch": 2.5712,
"grad_norm": 0.8439792959064886,
"learning_rate": 6.14572677499426e-07,
"loss": 0.2983,
"step": 2144
},
{
"epoch": 2.5724,
"grad_norm": 0.6287145474274123,
"learning_rate": 6.112250729522823e-07,
"loss": 0.3086,
"step": 2145
},
{
"epoch": 2.5736,
"grad_norm": 0.7568883079381424,
"learning_rate": 6.078860169460416e-07,
"loss": 0.2454,
"step": 2146
},
{
"epoch": 2.5747999999999998,
"grad_norm": 0.6250612919674088,
"learning_rate": 6.045555159845828e-07,
"loss": 0.2795,
"step": 2147
},
{
"epoch": 2.576,
"grad_norm": 0.6854826144067064,
"learning_rate": 6.012335765551186e-07,
"loss": 0.2548,
"step": 2148
},
{
"epoch": 2.5772,
"grad_norm": 0.5111642669118037,
"learning_rate": 5.979202051281891e-07,
"loss": 0.2826,
"step": 2149
},
{
"epoch": 2.5784000000000002,
"grad_norm": 0.7042512406408615,
"learning_rate": 5.946154081576411e-07,
"loss": 0.2726,
"step": 2150
},
{
"epoch": 2.5796,
"grad_norm": 0.5982087292956106,
"learning_rate": 5.913191920806244e-07,
"loss": 0.2745,
"step": 2151
},
{
"epoch": 2.5808,
"grad_norm": 0.7319660972515959,
"learning_rate": 5.880315633175704e-07,
"loss": 0.232,
"step": 2152
},
{
"epoch": 2.582,
"grad_norm": 0.7712191788339121,
"learning_rate": 5.847525282721883e-07,
"loss": 0.2866,
"step": 2153
},
{
"epoch": 2.5832,
"grad_norm": 0.6453817082802199,
"learning_rate": 5.814820933314446e-07,
"loss": 0.2917,
"step": 2154
},
{
"epoch": 2.5844,
"grad_norm": 0.7094775199167409,
"learning_rate": 5.78220264865555e-07,
"loss": 0.2888,
"step": 2155
},
{
"epoch": 2.5856,
"grad_norm": 0.6224811816839296,
"learning_rate": 5.749670492279757e-07,
"loss": 0.2618,
"step": 2156
},
{
"epoch": 2.5868,
"grad_norm": 0.8421430068153629,
"learning_rate": 5.717224527553811e-07,
"loss": 0.2799,
"step": 2157
},
{
"epoch": 2.588,
"grad_norm": 0.6842192341216612,
"learning_rate": 5.684864817676583e-07,
"loss": 0.2838,
"step": 2158
},
{
"epoch": 2.5892,
"grad_norm": 0.6714060045291238,
"learning_rate": 5.65259142567896e-07,
"loss": 0.3015,
"step": 2159
},
{
"epoch": 2.5904,
"grad_norm": 0.731969589610579,
"learning_rate": 5.620404414423674e-07,
"loss": 0.2948,
"step": 2160
},
{
"epoch": 2.5916,
"grad_norm": 0.5845977981467594,
"learning_rate": 5.588303846605187e-07,
"loss": 0.2949,
"step": 2161
},
{
"epoch": 2.5928,
"grad_norm": 0.7847005106402112,
"learning_rate": 5.556289784749653e-07,
"loss": 0.3116,
"step": 2162
},
{
"epoch": 2.594,
"grad_norm": 0.6126285560627935,
"learning_rate": 5.524362291214652e-07,
"loss": 0.3027,
"step": 2163
},
{
"epoch": 2.5952,
"grad_norm": 0.6671501696747408,
"learning_rate": 5.492521428189179e-07,
"loss": 0.2818,
"step": 2164
},
{
"epoch": 2.5964,
"grad_norm": 0.5939835318567993,
"learning_rate": 5.460767257693489e-07,
"loss": 0.2702,
"step": 2165
},
{
"epoch": 2.5976,
"grad_norm": 0.6897079960416239,
"learning_rate": 5.429099841578966e-07,
"loss": 0.2687,
"step": 2166
},
{
"epoch": 2.5987999999999998,
"grad_norm": 0.876116617925504,
"learning_rate": 5.397519241528026e-07,
"loss": 0.312,
"step": 2167
},
{
"epoch": 2.6,
"grad_norm": 0.6775979697669733,
"learning_rate": 5.366025519053958e-07,
"loss": 0.288,
"step": 2168
},
{
"epoch": 2.6012,
"grad_norm": 0.6470250388662834,
"learning_rate": 5.334618735500868e-07,
"loss": 0.3299,
"step": 2169
},
{
"epoch": 2.6024000000000003,
"grad_norm": 0.7676431230066653,
"learning_rate": 5.303298952043473e-07,
"loss": 0.2929,
"step": 2170
},
{
"epoch": 2.6036,
"grad_norm": 0.7488681451965283,
"learning_rate": 5.272066229687078e-07,
"loss": 0.2475,
"step": 2171
},
{
"epoch": 2.6048,
"grad_norm": 0.6816432772590324,
"learning_rate": 5.24092062926736e-07,
"loss": 0.2665,
"step": 2172
},
{
"epoch": 2.606,
"grad_norm": 0.5946012484360149,
"learning_rate": 5.209862211450351e-07,
"loss": 0.2916,
"step": 2173
},
{
"epoch": 2.6071999999999997,
"grad_norm": 0.6089864202764649,
"learning_rate": 5.17889103673222e-07,
"loss": 0.2904,
"step": 2174
},
{
"epoch": 2.6084,
"grad_norm": 0.7317087731871482,
"learning_rate": 5.148007165439234e-07,
"loss": 0.3137,
"step": 2175
},
{
"epoch": 2.6096,
"grad_norm": 0.6934272479110588,
"learning_rate": 5.117210657727589e-07,
"loss": 0.3282,
"step": 2176
},
{
"epoch": 2.6108000000000002,
"grad_norm": 0.6371695418638036,
"learning_rate": 5.086501573583302e-07,
"loss": 0.2578,
"step": 2177
},
{
"epoch": 2.612,
"grad_norm": 0.8456359848619833,
"learning_rate": 5.055879972822164e-07,
"loss": 0.2989,
"step": 2178
},
{
"epoch": 2.6132,
"grad_norm": 1.0056100323867774,
"learning_rate": 5.025345915089497e-07,
"loss": 0.3365,
"step": 2179
},
{
"epoch": 2.6144,
"grad_norm": 0.7147689779039166,
"learning_rate": 4.994899459860125e-07,
"loss": 0.279,
"step": 2180
},
{
"epoch": 2.6156,
"grad_norm": 0.7176392779443701,
"learning_rate": 4.964540666438261e-07,
"loss": 0.323,
"step": 2181
},
{
"epoch": 2.6168,
"grad_norm": 0.9629994282009234,
"learning_rate": 4.934269593957336e-07,
"loss": 0.3033,
"step": 2182
},
{
"epoch": 2.618,
"grad_norm": 0.6355636681573144,
"learning_rate": 4.90408630137994e-07,
"loss": 0.3006,
"step": 2183
},
{
"epoch": 2.6192,
"grad_norm": 0.7836560561304082,
"learning_rate": 4.873990847497684e-07,
"loss": 0.304,
"step": 2184
},
{
"epoch": 2.6204,
"grad_norm": 0.7356429202700248,
"learning_rate": 4.843983290931064e-07,
"loss": 0.2716,
"step": 2185
},
{
"epoch": 2.6216,
"grad_norm": 0.6242720063108813,
"learning_rate": 4.814063690129378e-07,
"loss": 0.2705,
"step": 2186
},
{
"epoch": 2.6228,
"grad_norm": 0.6318555413322213,
"learning_rate": 4.784232103370617e-07,
"loss": 0.2896,
"step": 2187
},
{
"epoch": 2.624,
"grad_norm": 0.7768009294358684,
"learning_rate": 4.7544885887613136e-07,
"loss": 0.2887,
"step": 2188
},
{
"epoch": 2.6252,
"grad_norm": 0.6549837820688543,
"learning_rate": 4.724833204236462e-07,
"loss": 0.2614,
"step": 2189
},
{
"epoch": 2.6264,
"grad_norm": 0.7180138423163781,
"learning_rate": 4.695266007559407e-07,
"loss": 0.294,
"step": 2190
},
{
"epoch": 2.6276,
"grad_norm": 0.5457429131402424,
"learning_rate": 4.6657870563217076e-07,
"loss": 0.3042,
"step": 2191
},
{
"epoch": 2.6288,
"grad_norm": 0.6533262876668459,
"learning_rate": 4.6363964079430166e-07,
"loss": 0.2849,
"step": 2192
},
{
"epoch": 2.63,
"grad_norm": 1.1061093409715972,
"learning_rate": 4.6070941196710186e-07,
"loss": 0.2531,
"step": 2193
},
{
"epoch": 2.6311999999999998,
"grad_norm": 0.761049234718346,
"learning_rate": 4.5778802485812956e-07,
"loss": 0.2899,
"step": 2194
},
{
"epoch": 2.6324,
"grad_norm": 0.7124544768965472,
"learning_rate": 4.548754851577175e-07,
"loss": 0.2539,
"step": 2195
},
{
"epoch": 2.6336,
"grad_norm": 0.7691672861443284,
"learning_rate": 4.5197179853896654e-07,
"loss": 0.2805,
"step": 2196
},
{
"epoch": 2.6348000000000003,
"grad_norm": 0.6557382261310944,
"learning_rate": 4.4907697065773523e-07,
"loss": 0.2366,
"step": 2197
},
{
"epoch": 2.636,
"grad_norm": 0.6750583773439472,
"learning_rate": 4.4619100715262374e-07,
"loss": 0.2787,
"step": 2198
},
{
"epoch": 2.6372,
"grad_norm": 0.5867440035565352,
"learning_rate": 4.4331391364496934e-07,
"loss": 0.2669,
"step": 2199
},
{
"epoch": 2.6384,
"grad_norm": 0.5962110893281645,
"learning_rate": 4.404456957388309e-07,
"loss": 0.2826,
"step": 2200
},
{
"epoch": 2.6395999999999997,
"grad_norm": 0.9200079910478826,
"learning_rate": 4.375863590209778e-07,
"loss": 0.2554,
"step": 2201
},
{
"epoch": 2.6408,
"grad_norm": 0.8170036767330543,
"learning_rate": 4.3473590906088046e-07,
"loss": 0.2774,
"step": 2202
},
{
"epoch": 2.642,
"grad_norm": 0.6635984329833553,
"learning_rate": 4.3189435141070324e-07,
"loss": 0.2954,
"step": 2203
},
{
"epoch": 2.6432,
"grad_norm": 0.6296610540339946,
"learning_rate": 4.2906169160528424e-07,
"loss": 0.2892,
"step": 2204
},
{
"epoch": 2.6444,
"grad_norm": 0.6474922100066525,
"learning_rate": 4.262379351621354e-07,
"loss": 0.2577,
"step": 2205
},
{
"epoch": 2.6456,
"grad_norm": 0.6784660583961697,
"learning_rate": 4.2342308758142437e-07,
"loss": 0.2577,
"step": 2206
},
{
"epoch": 2.6468,
"grad_norm": 0.7801139599825996,
"learning_rate": 4.2061715434596475e-07,
"loss": 0.2396,
"step": 2207
},
{
"epoch": 2.648,
"grad_norm": 0.5259427017853717,
"learning_rate": 4.1782014092120735e-07,
"loss": 0.2532,
"step": 2208
},
{
"epoch": 2.6492,
"grad_norm": 0.6350586490877915,
"learning_rate": 4.150320527552304e-07,
"loss": 0.257,
"step": 2209
},
{
"epoch": 2.6504,
"grad_norm": 0.6130026490729126,
"learning_rate": 4.122528952787258e-07,
"loss": 0.3017,
"step": 2210
},
{
"epoch": 2.6516,
"grad_norm": 0.7017323989635674,
"learning_rate": 4.0948267390498953e-07,
"loss": 0.2563,
"step": 2211
},
{
"epoch": 2.6528,
"grad_norm": 0.5894372215605755,
"learning_rate": 4.067213940299136e-07,
"loss": 0.2252,
"step": 2212
},
{
"epoch": 2.654,
"grad_norm": 0.6798404101983243,
"learning_rate": 4.0396906103197244e-07,
"loss": 0.2658,
"step": 2213
},
{
"epoch": 2.6552,
"grad_norm": 0.7266677492832656,
"learning_rate": 4.01225680272212e-07,
"loss": 0.2894,
"step": 2214
},
{
"epoch": 2.6564,
"grad_norm": 0.6748261060469086,
"learning_rate": 3.984912570942434e-07,
"loss": 0.2506,
"step": 2215
},
{
"epoch": 2.6576,
"grad_norm": 0.7909695243587097,
"learning_rate": 3.9576579682423066e-07,
"loss": 0.3007,
"step": 2216
},
{
"epoch": 2.6588000000000003,
"grad_norm": 0.6598917886250832,
"learning_rate": 3.930493047708761e-07,
"loss": 0.3195,
"step": 2217
},
{
"epoch": 2.66,
"grad_norm": 0.8580375903766323,
"learning_rate": 3.903417862254172e-07,
"loss": 0.2987,
"step": 2218
},
{
"epoch": 2.6612,
"grad_norm": 0.652223835498985,
"learning_rate": 3.876432464616103e-07,
"loss": 0.2765,
"step": 2219
},
{
"epoch": 2.6624,
"grad_norm": 0.8185270286268873,
"learning_rate": 3.8495369073572266e-07,
"loss": 0.2901,
"step": 2220
},
{
"epoch": 2.6635999999999997,
"grad_norm": 0.7780894643825088,
"learning_rate": 3.822731242865235e-07,
"loss": 0.2565,
"step": 2221
},
{
"epoch": 2.6648,
"grad_norm": 0.6279994493544137,
"learning_rate": 3.7960155233527364e-07,
"loss": 0.2701,
"step": 2222
},
{
"epoch": 2.666,
"grad_norm": 0.6506297709555272,
"learning_rate": 3.7693898008571205e-07,
"loss": 0.2752,
"step": 2223
},
{
"epoch": 2.6672000000000002,
"grad_norm": 0.818659019815678,
"learning_rate": 3.742854127240464e-07,
"loss": 0.2826,
"step": 2224
},
{
"epoch": 2.6684,
"grad_norm": 0.5922451942692247,
"learning_rate": 3.7164085541894937e-07,
"loss": 0.2976,
"step": 2225
},
{
"epoch": 2.6696,
"grad_norm": 0.7955622400536568,
"learning_rate": 3.690053133215399e-07,
"loss": 0.2838,
"step": 2226
},
{
"epoch": 2.6708,
"grad_norm": 0.5368513698853616,
"learning_rate": 3.663787915653777e-07,
"loss": 0.3014,
"step": 2227
},
{
"epoch": 2.672,
"grad_norm": 0.7579052868530307,
"learning_rate": 3.6376129526645376e-07,
"loss": 0.3028,
"step": 2228
},
{
"epoch": 2.6732,
"grad_norm": 0.5975556511501016,
"learning_rate": 3.6115282952317807e-07,
"loss": 0.2778,
"step": 2229
},
{
"epoch": 2.6744,
"grad_norm": 0.7923818450533048,
"learning_rate": 3.5855339941636867e-07,
"loss": 0.267,
"step": 2230
},
{
"epoch": 2.6756,
"grad_norm": 0.7421111819007243,
"learning_rate": 3.5596301000924815e-07,
"loss": 0.2692,
"step": 2231
},
{
"epoch": 2.6768,
"grad_norm": 0.5980800345276354,
"learning_rate": 3.533816663474271e-07,
"loss": 0.2709,
"step": 2232
},
{
"epoch": 2.678,
"grad_norm": 0.7143505001582382,
"learning_rate": 3.508093734588952e-07,
"loss": 0.295,
"step": 2233
},
{
"epoch": 2.6792,
"grad_norm": 0.6600003239340894,
"learning_rate": 3.482461363540163e-07,
"loss": 0.2968,
"step": 2234
},
{
"epoch": 2.6804,
"grad_norm": 0.6334104463989448,
"learning_rate": 3.456919600255126e-07,
"loss": 0.2375,
"step": 2235
},
{
"epoch": 2.6816,
"grad_norm": 0.5776171056880808,
"learning_rate": 3.4314684944845747e-07,
"loss": 0.2881,
"step": 2236
},
{
"epoch": 2.6828,
"grad_norm": 0.8875017356030829,
"learning_rate": 3.406108095802668e-07,
"loss": 0.3259,
"step": 2237
},
{
"epoch": 2.684,
"grad_norm": 0.6297377196585013,
"learning_rate": 3.3808384536068997e-07,
"loss": 0.2706,
"step": 2238
},
{
"epoch": 2.6852,
"grad_norm": 0.771205946329643,
"learning_rate": 3.3556596171179455e-07,
"loss": 0.2536,
"step": 2239
},
{
"epoch": 2.6864,
"grad_norm": 0.6003853432991978,
"learning_rate": 3.3305716353796537e-07,
"loss": 0.3022,
"step": 2240
},
{
"epoch": 2.6875999999999998,
"grad_norm": 0.7317807643290035,
"learning_rate": 3.305574557258867e-07,
"loss": 0.2498,
"step": 2241
},
{
"epoch": 2.6888,
"grad_norm": 0.6197617905574234,
"learning_rate": 3.2806684314453774e-07,
"loss": 0.2873,
"step": 2242
},
{
"epoch": 2.69,
"grad_norm": 0.6906426398908749,
"learning_rate": 3.255853306451823e-07,
"loss": 0.312,
"step": 2243
},
{
"epoch": 2.6912000000000003,
"grad_norm": 0.6905670324441344,
"learning_rate": 3.2311292306135944e-07,
"loss": 0.311,
"step": 2244
},
{
"epoch": 2.6924,
"grad_norm": 0.7184930727698079,
"learning_rate": 3.2064962520887146e-07,
"loss": 0.2668,
"step": 2245
},
{
"epoch": 2.6936,
"grad_norm": 0.7951668141713001,
"learning_rate": 3.18195441885778e-07,
"loss": 0.3017,
"step": 2246
},
{
"epoch": 2.6948,
"grad_norm": 0.7339795362953534,
"learning_rate": 3.157503778723847e-07,
"loss": 0.3242,
"step": 2247
},
{
"epoch": 2.6959999999999997,
"grad_norm": 0.5010531238523579,
"learning_rate": 3.1331443793123585e-07,
"loss": 0.2919,
"step": 2248
},
{
"epoch": 2.6972,
"grad_norm": 0.7585631056885037,
"learning_rate": 3.108876268071009e-07,
"loss": 0.2982,
"step": 2249
},
{
"epoch": 2.6984,
"grad_norm": 0.646815926338651,
"learning_rate": 3.0846994922697104e-07,
"loss": 0.2988,
"step": 2250
},
{
"epoch": 2.6996,
"grad_norm": 0.6535934008926059,
"learning_rate": 3.060614099000442e-07,
"loss": 0.3012,
"step": 2251
},
{
"epoch": 2.7008,
"grad_norm": 0.6727430113037003,
"learning_rate": 3.0366201351771983e-07,
"loss": 0.3058,
"step": 2252
},
{
"epoch": 2.702,
"grad_norm": 0.6206405333306073,
"learning_rate": 3.0127176475359065e-07,
"loss": 0.2988,
"step": 2253
},
{
"epoch": 2.7032,
"grad_norm": 0.8409314079244052,
"learning_rate": 2.988906682634285e-07,
"loss": 0.305,
"step": 2254
},
{
"epoch": 2.7044,
"grad_norm": 0.7201826788227337,
"learning_rate": 2.965187286851784e-07,
"loss": 0.3245,
"step": 2255
},
{
"epoch": 2.7056,
"grad_norm": 0.6780390557381377,
"learning_rate": 2.941559506389513e-07,
"loss": 0.275,
"step": 2256
},
{
"epoch": 2.7068,
"grad_norm": 0.6651241710940864,
"learning_rate": 2.9180233872701247e-07,
"loss": 0.3221,
"step": 2257
},
{
"epoch": 2.708,
"grad_norm": 0.6494634593968596,
"learning_rate": 2.894578975337703e-07,
"loss": 0.3083,
"step": 2258
},
{
"epoch": 2.7092,
"grad_norm": 0.5643601731637798,
"learning_rate": 2.8712263162577636e-07,
"loss": 0.2898,
"step": 2259
},
{
"epoch": 2.7104,
"grad_norm": 0.777896547810455,
"learning_rate": 2.8479654555170546e-07,
"loss": 0.2513,
"step": 2260
},
{
"epoch": 2.7116,
"grad_norm": 0.6721559056076929,
"learning_rate": 2.8247964384235214e-07,
"loss": 0.3144,
"step": 2261
},
{
"epoch": 2.7128,
"grad_norm": 0.8162039912547114,
"learning_rate": 2.8017193101062377e-07,
"loss": 0.3143,
"step": 2262
},
{
"epoch": 2.714,
"grad_norm": 0.6406148829352927,
"learning_rate": 2.778734115515269e-07,
"loss": 0.2737,
"step": 2263
},
{
"epoch": 2.7152,
"grad_norm": 0.6218386510338555,
"learning_rate": 2.755840899421636e-07,
"loss": 0.2998,
"step": 2264
},
{
"epoch": 2.7164,
"grad_norm": 0.6583029692987573,
"learning_rate": 2.7330397064171787e-07,
"loss": 0.3132,
"step": 2265
},
{
"epoch": 2.7176,
"grad_norm": 0.592237079602336,
"learning_rate": 2.7103305809145106e-07,
"loss": 0.2965,
"step": 2266
},
{
"epoch": 2.7188,
"grad_norm": 0.7807797383429858,
"learning_rate": 2.687713567146899e-07,
"loss": 0.2783,
"step": 2267
},
{
"epoch": 2.7199999999999998,
"grad_norm": 0.6991497990434121,
"learning_rate": 2.665188709168215e-07,
"loss": 0.2928,
"step": 2268
},
{
"epoch": 2.7212,
"grad_norm": 0.8815279469403849,
"learning_rate": 2.642756050852796e-07,
"loss": 0.2453,
"step": 2269
},
{
"epoch": 2.7224,
"grad_norm": 0.6153693318331473,
"learning_rate": 2.620415635895429e-07,
"loss": 0.2812,
"step": 2270
},
{
"epoch": 2.7236000000000002,
"grad_norm": 0.7663333381472748,
"learning_rate": 2.5981675078111835e-07,
"loss": 0.2839,
"step": 2271
},
{
"epoch": 2.7248,
"grad_norm": 0.8332220572070982,
"learning_rate": 2.5760117099354163e-07,
"loss": 0.2845,
"step": 2272
},
{
"epoch": 2.726,
"grad_norm": 0.7796875138965523,
"learning_rate": 2.5539482854236076e-07,
"loss": 0.2385,
"step": 2273
},
{
"epoch": 2.7272,
"grad_norm": 0.5985793251985193,
"learning_rate": 2.531977277251324e-07,
"loss": 0.343,
"step": 2274
},
{
"epoch": 2.7284,
"grad_norm": 0.7347732993596829,
"learning_rate": 2.510098728214133e-07,
"loss": 0.3047,
"step": 2275
},
{
"epoch": 2.7296,
"grad_norm": 0.6680063570700783,
"learning_rate": 2.488312680927485e-07,
"loss": 0.3072,
"step": 2276
},
{
"epoch": 2.7308,
"grad_norm": 0.5873275344359791,
"learning_rate": 2.466619177826668e-07,
"loss": 0.286,
"step": 2277
},
{
"epoch": 2.732,
"grad_norm": 0.6718246234300642,
"learning_rate": 2.4450182611667096e-07,
"loss": 0.254,
"step": 2278
},
{
"epoch": 2.7332,
"grad_norm": 0.6977410210912166,
"learning_rate": 2.423509973022292e-07,
"loss": 0.2785,
"step": 2279
},
{
"epoch": 2.7344,
"grad_norm": 0.662328876969527,
"learning_rate": 2.4020943552876706e-07,
"loss": 0.3171,
"step": 2280
},
{
"epoch": 2.7356,
"grad_norm": 0.7683060956211476,
"learning_rate": 2.3807714496766165e-07,
"loss": 0.2942,
"step": 2281
},
{
"epoch": 2.7368,
"grad_norm": 0.6715288751708779,
"learning_rate": 2.3595412977222897e-07,
"loss": 0.3223,
"step": 2282
},
{
"epoch": 2.738,
"grad_norm": 0.6344776640266345,
"learning_rate": 2.3384039407771896e-07,
"loss": 0.2608,
"step": 2283
},
{
"epoch": 2.7392,
"grad_norm": 0.6626564910615163,
"learning_rate": 2.317359420013071e-07,
"loss": 0.3168,
"step": 2284
},
{
"epoch": 2.7404,
"grad_norm": 0.8190217769349288,
"learning_rate": 2.2964077764208615e-07,
"loss": 0.2856,
"step": 2285
},
{
"epoch": 2.7416,
"grad_norm": 0.7541826633891097,
"learning_rate": 2.2755490508105716e-07,
"loss": 0.2401,
"step": 2286
},
{
"epoch": 2.7428,
"grad_norm": 0.6510299188293615,
"learning_rate": 2.254783283811246e-07,
"loss": 0.3,
"step": 2287
},
{
"epoch": 2.7439999999999998,
"grad_norm": 0.6677847022487289,
"learning_rate": 2.2341105158708408e-07,
"loss": 0.3085,
"step": 2288
},
{
"epoch": 2.7452,
"grad_norm": 0.5743582434948037,
"learning_rate": 2.2135307872561628e-07,
"loss": 0.2724,
"step": 2289
},
{
"epoch": 2.7464,
"grad_norm": 0.7590871132641382,
"learning_rate": 2.1930441380528243e-07,
"loss": 0.2791,
"step": 2290
},
{
"epoch": 2.7476000000000003,
"grad_norm": 0.8294816101512608,
"learning_rate": 2.172650608165111e-07,
"loss": 0.303,
"step": 2291
},
{
"epoch": 2.7488,
"grad_norm": 0.9098459761432557,
"learning_rate": 2.1523502373159367e-07,
"loss": 0.2801,
"step": 2292
},
{
"epoch": 2.75,
"grad_norm": 0.5521850654699119,
"learning_rate": 2.1321430650467546e-07,
"loss": 0.256,
"step": 2293
},
{
"epoch": 2.7512,
"grad_norm": 0.6473775989198229,
"learning_rate": 2.112029130717491e-07,
"loss": 0.2568,
"step": 2294
},
{
"epoch": 2.7523999999999997,
"grad_norm": 0.8514780480762504,
"learning_rate": 2.092008473506446e-07,
"loss": 0.3022,
"step": 2295
},
{
"epoch": 2.7536,
"grad_norm": 0.6427741363312998,
"learning_rate": 2.072081132410253e-07,
"loss": 0.2881,
"step": 2296
},
{
"epoch": 2.7548,
"grad_norm": 0.9216463208390858,
"learning_rate": 2.0522471462437798e-07,
"loss": 0.2916,
"step": 2297
},
{
"epoch": 2.7560000000000002,
"grad_norm": 0.6957480337115077,
"learning_rate": 2.0325065536400456e-07,
"loss": 0.2524,
"step": 2298
},
{
"epoch": 2.7572,
"grad_norm": 0.671707790066162,
"learning_rate": 2.0128593930501427e-07,
"loss": 0.3181,
"step": 2299
},
{
"epoch": 2.7584,
"grad_norm": 0.6975815314223021,
"learning_rate": 1.9933057027432147e-07,
"loss": 0.2493,
"step": 2300
},
{
"epoch": 2.7596,
"grad_norm": 0.7137263993083056,
"learning_rate": 1.9738455208063055e-07,
"loss": 0.2883,
"step": 2301
},
{
"epoch": 2.7608,
"grad_norm": 0.6793626349750588,
"learning_rate": 1.9544788851443342e-07,
"loss": 0.289,
"step": 2302
},
{
"epoch": 2.762,
"grad_norm": 0.8185591569108523,
"learning_rate": 1.9352058334800195e-07,
"loss": 0.2615,
"step": 2303
},
{
"epoch": 2.7632,
"grad_norm": 0.5612419147357184,
"learning_rate": 1.9160264033537824e-07,
"loss": 0.2849,
"step": 2304
},
{
"epoch": 2.7644,
"grad_norm": 0.8101429433991842,
"learning_rate": 1.8969406321236727e-07,
"loss": 0.3452,
"step": 2305
},
{
"epoch": 2.7656,
"grad_norm": 0.7021056540292714,
"learning_rate": 1.8779485569653422e-07,
"loss": 0.3301,
"step": 2306
},
{
"epoch": 2.7668,
"grad_norm": 0.6872609977267615,
"learning_rate": 1.859050214871927e-07,
"loss": 0.2722,
"step": 2307
},
{
"epoch": 2.768,
"grad_norm": 0.5948268680103144,
"learning_rate": 1.8402456426539706e-07,
"loss": 0.261,
"step": 2308
},
{
"epoch": 2.7692,
"grad_norm": 0.6658381496085037,
"learning_rate": 1.8215348769393904e-07,
"loss": 0.2281,
"step": 2309
},
{
"epoch": 2.7704,
"grad_norm": 0.6386865932665424,
"learning_rate": 1.8029179541733833e-07,
"loss": 0.2536,
"step": 2310
},
{
"epoch": 2.7716,
"grad_norm": 0.5387429872617496,
"learning_rate": 1.7843949106183368e-07,
"loss": 0.2919,
"step": 2311
},
{
"epoch": 2.7728,
"grad_norm": 0.9459633794031159,
"learning_rate": 1.7659657823538067e-07,
"loss": 0.2789,
"step": 2312
},
{
"epoch": 2.774,
"grad_norm": 0.5944872376081571,
"learning_rate": 1.747630605276407e-07,
"loss": 0.2994,
"step": 2313
},
{
"epoch": 2.7752,
"grad_norm": 0.7762137430931023,
"learning_rate": 1.7293894150997414e-07,
"loss": 0.3184,
"step": 2314
},
{
"epoch": 2.7763999999999998,
"grad_norm": 0.7823106215427452,
"learning_rate": 1.711242247354372e-07,
"loss": 0.2705,
"step": 2315
},
{
"epoch": 2.7776,
"grad_norm": 0.7235272866138038,
"learning_rate": 1.6931891373876852e-07,
"loss": 0.2716,
"step": 2316
},
{
"epoch": 2.7788,
"grad_norm": 0.7102357739329513,
"learning_rate": 1.6752301203638854e-07,
"loss": 0.2613,
"step": 2317
},
{
"epoch": 2.7800000000000002,
"grad_norm": 0.6383690636704885,
"learning_rate": 1.657365231263891e-07,
"loss": 0.2504,
"step": 2318
},
{
"epoch": 2.7812,
"grad_norm": 0.7175680261104689,
"learning_rate": 1.6395945048852947e-07,
"loss": 0.275,
"step": 2319
},
{
"epoch": 2.7824,
"grad_norm": 0.6917508934762812,
"learning_rate": 1.6219179758422465e-07,
"loss": 0.3288,
"step": 2320
},
{
"epoch": 2.7836,
"grad_norm": 0.6624840280638272,
"learning_rate": 1.6043356785654273e-07,
"loss": 0.2296,
"step": 2321
},
{
"epoch": 2.7848,
"grad_norm": 0.5625210333311368,
"learning_rate": 1.5868476473019922e-07,
"loss": 0.2452,
"step": 2322
},
{
"epoch": 2.786,
"grad_norm": 0.6645947381125819,
"learning_rate": 1.5694539161154598e-07,
"loss": 0.2637,
"step": 2323
},
{
"epoch": 2.7872,
"grad_norm": 0.858331455628384,
"learning_rate": 1.5521545188856734e-07,
"loss": 0.2962,
"step": 2324
},
{
"epoch": 2.7884,
"grad_norm": 0.5818974130553681,
"learning_rate": 1.5349494893087514e-07,
"loss": 0.2641,
"step": 2325
},
{
"epoch": 2.7896,
"grad_norm": 0.7000708801543115,
"learning_rate": 1.517838860896964e-07,
"loss": 0.2301,
"step": 2326
},
{
"epoch": 2.7908,
"grad_norm": 1.0379115995450328,
"learning_rate": 1.500822666978735e-07,
"loss": 0.2841,
"step": 2327
},
{
"epoch": 2.792,
"grad_norm": 0.6397878998170997,
"learning_rate": 1.4839009406985295e-07,
"loss": 0.314,
"step": 2328
},
{
"epoch": 2.7932,
"grad_norm": 0.8068255262719065,
"learning_rate": 1.4670737150168257e-07,
"loss": 0.2924,
"step": 2329
},
{
"epoch": 2.7944,
"grad_norm": 0.6424419293136353,
"learning_rate": 1.4503410227100057e-07,
"loss": 0.2336,
"step": 2330
},
{
"epoch": 2.7956,
"grad_norm": 0.7329975422448929,
"learning_rate": 1.433702896370348e-07,
"loss": 0.3361,
"step": 2331
},
{
"epoch": 2.7968,
"grad_norm": 1.0198735009965914,
"learning_rate": 1.417159368405907e-07,
"loss": 0.2987,
"step": 2332
},
{
"epoch": 2.798,
"grad_norm": 0.6608444482139405,
"learning_rate": 1.4007104710404838e-07,
"loss": 0.2439,
"step": 2333
},
{
"epoch": 2.7992,
"grad_norm": 0.6921521367521364,
"learning_rate": 1.384356236313572e-07,
"loss": 0.3088,
"step": 2334
},
{
"epoch": 2.8004,
"grad_norm": 0.7790176322863624,
"learning_rate": 1.3680966960802623e-07,
"loss": 0.2489,
"step": 2335
},
{
"epoch": 2.8016,
"grad_norm": 0.7209429373978853,
"learning_rate": 1.3519318820111983e-07,
"loss": 0.2634,
"step": 2336
},
{
"epoch": 2.8028,
"grad_norm": 0.6615965395378771,
"learning_rate": 1.3358618255925214e-07,
"loss": 0.3238,
"step": 2337
},
{
"epoch": 2.8040000000000003,
"grad_norm": 0.9180679333408356,
"learning_rate": 1.3198865581258046e-07,
"loss": 0.3317,
"step": 2338
},
{
"epoch": 2.8052,
"grad_norm": 0.6810292946730327,
"learning_rate": 1.3040061107279679e-07,
"loss": 0.2897,
"step": 2339
},
{
"epoch": 2.8064,
"grad_norm": 0.7288465076135464,
"learning_rate": 1.2882205143312676e-07,
"loss": 0.2791,
"step": 2340
},
{
"epoch": 2.8076,
"grad_norm": 0.7914414156817373,
"learning_rate": 1.272529799683192e-07,
"loss": 0.2884,
"step": 2341
},
{
"epoch": 2.8087999999999997,
"grad_norm": 0.6703048875885049,
"learning_rate": 1.2569339973464155e-07,
"loss": 0.2585,
"step": 2342
},
{
"epoch": 2.81,
"grad_norm": 0.8889565548561289,
"learning_rate": 1.2414331376987555e-07,
"loss": 0.2773,
"step": 2343
},
{
"epoch": 2.8112,
"grad_norm": 0.6810063877533172,
"learning_rate": 1.2260272509330707e-07,
"loss": 0.2979,
"step": 2344
},
{
"epoch": 2.8124000000000002,
"grad_norm": 0.49726926700584206,
"learning_rate": 1.2107163670572574e-07,
"loss": 0.2807,
"step": 2345
},
{
"epoch": 2.8136,
"grad_norm": 0.7706130001010241,
"learning_rate": 1.195500515894149e-07,
"loss": 0.2595,
"step": 2346
},
{
"epoch": 2.8148,
"grad_norm": 0.6764930070719378,
"learning_rate": 1.1803797270814765e-07,
"loss": 0.2958,
"step": 2347
},
{
"epoch": 2.816,
"grad_norm": 0.7914440521735895,
"learning_rate": 1.165354030071808e-07,
"loss": 0.2682,
"step": 2348
},
{
"epoch": 2.8172,
"grad_norm": 0.5844758356134893,
"learning_rate": 1.1504234541324765e-07,
"loss": 0.272,
"step": 2349
},
{
"epoch": 2.8184,
"grad_norm": 0.6821559414934388,
"learning_rate": 1.1355880283455523e-07,
"loss": 0.3261,
"step": 2350
},
{
"epoch": 2.8196,
"grad_norm": 0.6449741263745696,
"learning_rate": 1.1208477816077756e-07,
"loss": 0.252,
"step": 2351
},
{
"epoch": 2.8208,
"grad_norm": 0.6413279535712334,
"learning_rate": 1.1062027426304744e-07,
"loss": 0.3128,
"step": 2352
},
{
"epoch": 2.822,
"grad_norm": 0.6462931871515444,
"learning_rate": 1.091652939939547e-07,
"loss": 0.2674,
"step": 2353
},
{
"epoch": 2.8232,
"grad_norm": 0.7415474436688044,
"learning_rate": 1.0771984018753733e-07,
"loss": 0.2892,
"step": 2354
},
{
"epoch": 2.8244,
"grad_norm": 0.7471896204720297,
"learning_rate": 1.0628391565927765e-07,
"loss": 0.3097,
"step": 2355
},
{
"epoch": 2.8256,
"grad_norm": 0.861822200014857,
"learning_rate": 1.0485752320609944e-07,
"loss": 0.2943,
"step": 2356
},
{
"epoch": 2.8268,
"grad_norm": 0.7250557828908615,
"learning_rate": 1.0344066560635635e-07,
"loss": 0.3169,
"step": 2357
},
{
"epoch": 2.828,
"grad_norm": 0.6128378327254843,
"learning_rate": 1.0203334561983025e-07,
"loss": 0.2524,
"step": 2358
},
{
"epoch": 2.8292,
"grad_norm": 0.7343015833848221,
"learning_rate": 1.0063556598772839e-07,
"loss": 0.2958,
"step": 2359
},
{
"epoch": 2.8304,
"grad_norm": 0.5758133672932962,
"learning_rate": 9.924732943267068e-08,
"loss": 0.2945,
"step": 2360
},
{
"epoch": 2.8316,
"grad_norm": 0.6033292927474737,
"learning_rate": 9.7868638658693e-08,
"loss": 0.3118,
"step": 2361
},
{
"epoch": 2.8327999999999998,
"grad_norm": 0.6912989882455538,
"learning_rate": 9.6499496351235e-08,
"loss": 0.2903,
"step": 2362
},
{
"epoch": 2.834,
"grad_norm": 0.6393428263191321,
"learning_rate": 9.513990517713955e-08,
"loss": 0.2705,
"step": 2363
},
{
"epoch": 2.8352,
"grad_norm": 0.6686098031973033,
"learning_rate": 9.378986778464327e-08,
"loss": 0.271,
"step": 2364
},
{
"epoch": 2.8364000000000003,
"grad_norm": 0.6975075219135892,
"learning_rate": 9.244938680337656e-08,
"loss": 0.3035,
"step": 2365
},
{
"epoch": 2.8376,
"grad_norm": 0.6546988837796381,
"learning_rate": 9.111846484435361e-08,
"loss": 0.2862,
"step": 2366
},
{
"epoch": 2.8388,
"grad_norm": 0.612389491955198,
"learning_rate": 8.979710449997014e-08,
"loss": 0.3098,
"step": 2367
},
{
"epoch": 2.84,
"grad_norm": 0.7429352416598924,
"learning_rate": 8.848530834399683e-08,
"loss": 0.3126,
"step": 2368
},
{
"epoch": 2.8411999999999997,
"grad_norm": 0.3616513942542363,
"learning_rate": 8.718307893157696e-08,
"loss": 0.2729,
"step": 2369
},
{
"epoch": 2.8424,
"grad_norm": 0.7113971645389936,
"learning_rate": 8.589041879921711e-08,
"loss": 0.2972,
"step": 2370
},
{
"epoch": 2.8436,
"grad_norm": 0.807898337941532,
"learning_rate": 8.460733046478653e-08,
"loss": 0.2851,
"step": 2371
},
{
"epoch": 2.8448,
"grad_norm": 0.6397865010267628,
"learning_rate": 8.333381642750882e-08,
"loss": 0.2866,
"step": 2372
},
{
"epoch": 2.846,
"grad_norm": 0.7133194398036705,
"learning_rate": 8.206987916796027e-08,
"loss": 0.2268,
"step": 2373
},
{
"epoch": 2.8472,
"grad_norm": 0.7150674277871135,
"learning_rate": 8.081552114806101e-08,
"loss": 0.2465,
"step": 2374
},
{
"epoch": 2.8484,
"grad_norm": 0.6889406456361211,
"learning_rate": 7.957074481107551e-08,
"loss": 0.2609,
"step": 2375
},
{
"epoch": 2.8496,
"grad_norm": 0.6370237907987255,
"learning_rate": 7.833555258160208e-08,
"loss": 0.2535,
"step": 2376
},
{
"epoch": 2.8508,
"grad_norm": 0.7013618695854411,
"learning_rate": 7.710994686557172e-08,
"loss": 0.261,
"step": 2377
},
{
"epoch": 2.852,
"grad_norm": 0.7399960645156306,
"learning_rate": 7.589393005024482e-08,
"loss": 0.3461,
"step": 2378
},
{
"epoch": 2.8532,
"grad_norm": 0.7312395462597923,
"learning_rate": 7.468750450420114e-08,
"loss": 0.2679,
"step": 2379
},
{
"epoch": 2.8544,
"grad_norm": 0.5617940119614816,
"learning_rate": 7.349067257733989e-08,
"loss": 0.2494,
"step": 2380
},
{
"epoch": 2.8556,
"grad_norm": 0.6321877123905674,
"learning_rate": 7.230343660087402e-08,
"loss": 0.3002,
"step": 2381
},
{
"epoch": 2.8568,
"grad_norm": 0.6313993314228777,
"learning_rate": 7.11257988873243e-08,
"loss": 0.3156,
"step": 2382
},
{
"epoch": 2.858,
"grad_norm": 0.7849447984752222,
"learning_rate": 6.995776173051583e-08,
"loss": 0.2946,
"step": 2383
},
{
"epoch": 2.8592,
"grad_norm": 0.7403893912146318,
"learning_rate": 6.879932740557538e-08,
"loss": 0.3044,
"step": 2384
},
{
"epoch": 2.8604000000000003,
"grad_norm": 0.6609082938586613,
"learning_rate": 6.7650498168923e-08,
"loss": 0.3015,
"step": 2385
},
{
"epoch": 2.8616,
"grad_norm": 0.6567418731537202,
"learning_rate": 6.651127625827037e-08,
"loss": 0.3404,
"step": 2386
},
{
"epoch": 2.8628,
"grad_norm": 0.7553452191492749,
"learning_rate": 6.538166389261635e-08,
"loss": 0.2789,
"step": 2387
},
{
"epoch": 2.864,
"grad_norm": 0.6823200760258538,
"learning_rate": 6.426166327224148e-08,
"loss": 0.2503,
"step": 2388
},
{
"epoch": 2.8651999999999997,
"grad_norm": 0.5500421442223611,
"learning_rate": 6.315127657870513e-08,
"loss": 0.2953,
"step": 2389
},
{
"epoch": 2.8664,
"grad_norm": 0.7496221897843739,
"learning_rate": 6.205050597483997e-08,
"loss": 0.2902,
"step": 2390
},
{
"epoch": 2.8676,
"grad_norm": 1.127038293296292,
"learning_rate": 6.095935360474814e-08,
"loss": 0.2809,
"step": 2391
},
{
"epoch": 2.8688000000000002,
"grad_norm": 0.6591116086797535,
"learning_rate": 5.98778215937973e-08,
"loss": 0.2583,
"step": 2392
},
{
"epoch": 2.87,
"grad_norm": 0.583120567940326,
"learning_rate": 5.880591204861674e-08,
"loss": 0.2644,
"step": 2393
},
{
"epoch": 2.8712,
"grad_norm": 0.6604596298104798,
"learning_rate": 5.7743627057092463e-08,
"loss": 0.3116,
"step": 2394
},
{
"epoch": 2.8724,
"grad_norm": 0.7769207947532905,
"learning_rate": 5.669096868836377e-08,
"loss": 0.2497,
"step": 2395
},
{
"epoch": 2.8736,
"grad_norm": 0.9248624019574861,
"learning_rate": 5.564793899281884e-08,
"loss": 0.2887,
"step": 2396
},
{
"epoch": 2.8748,
"grad_norm": 0.5456766354723961,
"learning_rate": 5.461454000209199e-08,
"loss": 0.292,
"step": 2397
},
{
"epoch": 2.876,
"grad_norm": 0.9849160027752111,
"learning_rate": 5.3590773729056965e-08,
"loss": 0.3095,
"step": 2398
},
{
"epoch": 2.8772,
"grad_norm": 0.6805827747556151,
"learning_rate": 5.257664216782532e-08,
"loss": 0.2908,
"step": 2399
},
{
"epoch": 2.8784,
"grad_norm": 0.589405496766034,
"learning_rate": 5.157214729374305e-08,
"loss": 0.3282,
"step": 2400
},
{
"epoch": 2.8796,
"grad_norm": 0.6498125791445938,
"learning_rate": 5.057729106338505e-08,
"loss": 0.252,
"step": 2401
},
{
"epoch": 2.8808,
"grad_norm": 0.7627357866256731,
"learning_rate": 4.959207541455013e-08,
"loss": 0.2572,
"step": 2402
},
{
"epoch": 2.882,
"grad_norm": 0.7499782302171957,
"learning_rate": 4.8616502266261026e-08,
"loss": 0.3414,
"step": 2403
},
{
"epoch": 2.8832,
"grad_norm": 0.5935478807270599,
"learning_rate": 4.765057351875879e-08,
"loss": 0.2818,
"step": 2404
},
{
"epoch": 2.8844,
"grad_norm": 0.7050390220613826,
"learning_rate": 4.6694291053496766e-08,
"loss": 0.2949,
"step": 2405
},
{
"epoch": 2.8856,
"grad_norm": 0.6812039769395078,
"learning_rate": 4.5747656733142184e-08,
"loss": 0.3044,
"step": 2406
},
{
"epoch": 2.8868,
"grad_norm": 0.7102463841316684,
"learning_rate": 4.481067240156678e-08,
"loss": 0.3068,
"step": 2407
},
{
"epoch": 2.888,
"grad_norm": 0.6279608711854671,
"learning_rate": 4.3883339883846186e-08,
"loss": 0.2761,
"step": 2408
},
{
"epoch": 2.8891999999999998,
"grad_norm": 0.575486697115783,
"learning_rate": 4.296566098625776e-08,
"loss": 0.2684,
"step": 2409
},
{
"epoch": 2.8904,
"grad_norm": 0.7802983495532722,
"learning_rate": 4.2057637496273896e-08,
"loss": 0.2394,
"step": 2410
},
{
"epoch": 2.8916,
"grad_norm": 0.6283978539810603,
"learning_rate": 4.115927118256036e-08,
"loss": 0.2738,
"step": 2411
},
{
"epoch": 2.8928000000000003,
"grad_norm": 0.6327047359666658,
"learning_rate": 4.02705637949724e-08,
"loss": 0.3025,
"step": 2412
},
{
"epoch": 2.894,
"grad_norm": 0.6561048430763371,
"learning_rate": 3.939151706455146e-08,
"loss": 0.2758,
"step": 2413
},
{
"epoch": 2.8952,
"grad_norm": 0.6528496465252244,
"learning_rate": 3.8522132703521784e-08,
"loss": 0.3043,
"step": 2414
},
{
"epoch": 2.8964,
"grad_norm": 0.849232865820924,
"learning_rate": 3.7662412405286567e-08,
"loss": 0.2876,
"step": 2415
},
{
"epoch": 2.8975999999999997,
"grad_norm": 0.7282192441782935,
"learning_rate": 3.6812357844427385e-08,
"loss": 0.3056,
"step": 2416
},
{
"epoch": 2.8988,
"grad_norm": 0.6791149906609802,
"learning_rate": 3.597197067669533e-08,
"loss": 0.2732,
"step": 2417
},
{
"epoch": 2.9,
"grad_norm": 0.8188554436233152,
"learning_rate": 3.514125253901324e-08,
"loss": 0.3211,
"step": 2418
},
{
"epoch": 2.9012000000000002,
"grad_norm": 0.6507474406626096,
"learning_rate": 3.432020504947064e-08,
"loss": 0.2937,
"step": 2419
},
{
"epoch": 2.9024,
"grad_norm": 0.8916293881812993,
"learning_rate": 3.350882980731884e-08,
"loss": 0.2606,
"step": 2420
},
{
"epoch": 2.9036,
"grad_norm": 0.5942990494183431,
"learning_rate": 3.27071283929703e-08,
"loss": 0.33,
"step": 2421
},
{
"epoch": 2.9048,
"grad_norm": 0.7213460626626911,
"learning_rate": 3.191510236799589e-08,
"loss": 0.2795,
"step": 2422
},
{
"epoch": 2.906,
"grad_norm": 0.5789385299781349,
"learning_rate": 3.113275327511767e-08,
"loss": 0.2736,
"step": 2423
},
{
"epoch": 2.9072,
"grad_norm": 0.6917975286895354,
"learning_rate": 3.0360082638211666e-08,
"loss": 0.2783,
"step": 2424
},
{
"epoch": 2.9084,
"grad_norm": 0.7299328053529393,
"learning_rate": 2.959709196229954e-08,
"loss": 0.2614,
"step": 2425
},
{
"epoch": 2.9096,
"grad_norm": 0.6773714806677963,
"learning_rate": 2.8843782733549706e-08,
"loss": 0.3238,
"step": 2426
},
{
"epoch": 2.9108,
"grad_norm": 0.6905036803808762,
"learning_rate": 2.8100156419272885e-08,
"loss": 0.2772,
"step": 2427
},
{
"epoch": 2.912,
"grad_norm": 0.6553806490942691,
"learning_rate": 2.7366214467919318e-08,
"loss": 0.269,
"step": 2428
},
{
"epoch": 2.9132,
"grad_norm": 0.6726854109800227,
"learning_rate": 2.664195830907379e-08,
"loss": 0.3041,
"step": 2429
},
{
"epoch": 2.9144,
"grad_norm": 0.6054605532490988,
"learning_rate": 2.5927389353457842e-08,
"loss": 0.267,
"step": 2430
},
{
"epoch": 2.9156,
"grad_norm": 0.717995725885742,
"learning_rate": 2.5222508992922e-08,
"loss": 0.3037,
"step": 2431
},
{
"epoch": 2.9168,
"grad_norm": 0.5433888766051628,
"learning_rate": 2.4527318600446324e-08,
"loss": 0.3258,
"step": 2432
},
{
"epoch": 2.918,
"grad_norm": 0.67798668109855,
"learning_rate": 2.3841819530135424e-08,
"loss": 0.295,
"step": 2433
},
{
"epoch": 2.9192,
"grad_norm": 0.7891678386574797,
"learning_rate": 2.3166013117218998e-08,
"loss": 0.3064,
"step": 2434
},
{
"epoch": 2.9204,
"grad_norm": 0.732463416028174,
"learning_rate": 2.249990067804464e-08,
"loss": 0.3324,
"step": 2435
},
{
"epoch": 2.9215999999999998,
"grad_norm": 0.8382963993382336,
"learning_rate": 2.1843483510080032e-08,
"loss": 0.28,
"step": 2436
},
{
"epoch": 2.9228,
"grad_norm": 0.7057103073693665,
"learning_rate": 2.119676289190631e-08,
"loss": 0.2772,
"step": 2437
},
{
"epoch": 2.924,
"grad_norm": 0.48362705063006683,
"learning_rate": 2.0559740083219147e-08,
"loss": 0.2823,
"step": 2438
},
{
"epoch": 2.9252000000000002,
"grad_norm": 0.7039048731681169,
"learning_rate": 1.9932416324823235e-08,
"loss": 0.3035,
"step": 2439
},
{
"epoch": 2.9264,
"grad_norm": 0.5243845611819271,
"learning_rate": 1.93147928386328e-08,
"loss": 0.3243,
"step": 2440
},
{
"epoch": 2.9276,
"grad_norm": 0.8872387121012874,
"learning_rate": 1.8706870827666646e-08,
"loss": 0.2317,
"step": 2441
},
{
"epoch": 2.9288,
"grad_norm": 0.7014032992149177,
"learning_rate": 1.8108651476046457e-08,
"loss": 0.2675,
"step": 2442
},
{
"epoch": 2.93,
"grad_norm": 0.6236563832238406,
"learning_rate": 1.7520135948996263e-08,
"loss": 0.2624,
"step": 2443
},
{
"epoch": 2.9312,
"grad_norm": 0.698888049584497,
"learning_rate": 1.6941325392837437e-08,
"loss": 0.2866,
"step": 2444
},
{
"epoch": 2.9324,
"grad_norm": 0.7190676230862082,
"learning_rate": 1.6372220934988693e-08,
"loss": 0.3204,
"step": 2445
},
{
"epoch": 2.9336,
"grad_norm": 0.76136810491595,
"learning_rate": 1.5812823683962198e-08,
"loss": 0.2681,
"step": 2446
},
{
"epoch": 2.9348,
"grad_norm": 0.6810269926813283,
"learning_rate": 1.5263134729363582e-08,
"loss": 0.3078,
"step": 2447
},
{
"epoch": 2.936,
"grad_norm": 0.6669337163184537,
"learning_rate": 1.472315514188749e-08,
"loss": 0.2931,
"step": 2448
},
{
"epoch": 2.9372,
"grad_norm": 0.73735790104459,
"learning_rate": 1.4192885973315918e-08,
"loss": 0.3148,
"step": 2449
},
{
"epoch": 2.9384,
"grad_norm": 0.8463124936729638,
"learning_rate": 1.3672328256518208e-08,
"loss": 0.331,
"step": 2450
},
{
"epoch": 2.9396,
"grad_norm": 0.6604026325928568,
"learning_rate": 1.3161483005446618e-08,
"loss": 0.2817,
"step": 2451
},
{
"epoch": 2.9408,
"grad_norm": 0.6918463572710472,
"learning_rate": 1.2660351215135203e-08,
"loss": 0.2554,
"step": 2452
},
{
"epoch": 2.942,
"grad_norm": 0.7059661030800861,
"learning_rate": 1.2168933861698151e-08,
"loss": 0.3132,
"step": 2453
},
{
"epoch": 2.9432,
"grad_norm": 0.9149530302588724,
"learning_rate": 1.1687231902328122e-08,
"loss": 0.2966,
"step": 2454
},
{
"epoch": 2.9444,
"grad_norm": 0.6258148765230703,
"learning_rate": 1.1215246275292913e-08,
"loss": 0.2661,
"step": 2455
},
{
"epoch": 2.9455999999999998,
"grad_norm": 0.544275280835532,
"learning_rate": 1.0752977899936013e-08,
"loss": 0.323,
"step": 2456
},
{
"epoch": 2.9468,
"grad_norm": 0.8190898139164626,
"learning_rate": 1.0300427676672164e-08,
"loss": 0.2755,
"step": 2457
},
{
"epoch": 2.948,
"grad_norm": 1.2673236737341802,
"learning_rate": 9.85759648698792e-09,
"loss": 0.2402,
"step": 2458
},
{
"epoch": 2.9492000000000003,
"grad_norm": 0.757620109976881,
"learning_rate": 9.42448519343775e-09,
"loss": 0.3351,
"step": 2459
},
{
"epoch": 2.9504,
"grad_norm": 0.5748587248662484,
"learning_rate": 9.001094639645158e-09,
"loss": 0.3023,
"step": 2460
},
{
"epoch": 2.9516,
"grad_norm": 0.6303196268673067,
"learning_rate": 8.587425650297688e-09,
"loss": 0.2815,
"step": 2461
},
{
"epoch": 2.9528,
"grad_norm": 0.7371982820168463,
"learning_rate": 8.183479031148022e-09,
"loss": 0.2532,
"step": 2462
},
{
"epoch": 2.9539999999999997,
"grad_norm": 0.9174186579432821,
"learning_rate": 7.789255569011223e-09,
"loss": 0.2693,
"step": 2463
},
{
"epoch": 2.9552,
"grad_norm": 0.7518097298443339,
"learning_rate": 7.40475603176416e-09,
"loss": 0.2738,
"step": 2464
},
{
"epoch": 2.9564,
"grad_norm": 0.6529633161465678,
"learning_rate": 7.029981168341082e-09,
"loss": 0.2886,
"step": 2465
},
{
"epoch": 2.9576000000000002,
"grad_norm": 0.7378254276544852,
"learning_rate": 6.664931708736943e-09,
"loss": 0.3184,
"step": 2466
},
{
"epoch": 2.9588,
"grad_norm": 0.6663444692808295,
"learning_rate": 6.309608364001296e-09,
"loss": 0.2889,
"step": 2467
},
{
"epoch": 2.96,
"grad_norm": 0.8534779499271332,
"learning_rate": 5.9640118262399575e-09,
"loss": 0.3012,
"step": 2468
},
{
"epoch": 2.9612,
"grad_norm": 0.741203558795986,
"learning_rate": 5.628142768613343e-09,
"loss": 0.272,
"step": 2469
},
{
"epoch": 2.9624,
"grad_norm": 0.6509250283649045,
"learning_rate": 5.302001845333138e-09,
"loss": 0.3225,
"step": 2470
},
{
"epoch": 2.9636,
"grad_norm": 0.6463190122601488,
"learning_rate": 4.9855896916634065e-09,
"loss": 0.2737,
"step": 2471
},
{
"epoch": 2.9648,
"grad_norm": 0.6635977222826027,
"learning_rate": 4.678906923916704e-09,
"loss": 0.2798,
"step": 2472
},
{
"epoch": 2.966,
"grad_norm": 0.7802206196414878,
"learning_rate": 4.381954139457411e-09,
"loss": 0.2823,
"step": 2473
},
{
"epoch": 2.9672,
"grad_norm": 0.61864567960403,
"learning_rate": 4.094731916693962e-09,
"loss": 0.3505,
"step": 2474
},
{
"epoch": 2.9684,
"grad_norm": 0.6965183120808695,
"learning_rate": 3.817240815084944e-09,
"loss": 0.2584,
"step": 2475
},
{
"epoch": 2.9696,
"grad_norm": 0.6046463344841392,
"learning_rate": 3.5494813751324466e-09,
"loss": 0.297,
"step": 2476
},
{
"epoch": 2.9708,
"grad_norm": 0.9234883537003901,
"learning_rate": 3.291454118383164e-09,
"loss": 0.2762,
"step": 2477
},
{
"epoch": 2.972,
"grad_norm": 0.6987785654349705,
"learning_rate": 3.043159547427843e-09,
"loss": 0.274,
"step": 2478
},
{
"epoch": 2.9732,
"grad_norm": 0.6554687217398003,
"learning_rate": 2.804598145899062e-09,
"loss": 0.3073,
"step": 2479
},
{
"epoch": 2.9744,
"grad_norm": 0.6783021010729677,
"learning_rate": 2.575770378472342e-09,
"loss": 0.3242,
"step": 2480
},
{
"epoch": 2.9756,
"grad_norm": 0.6645551867585759,
"learning_rate": 2.3566766908622586e-09,
"loss": 0.2743,
"step": 2481
},
{
"epoch": 2.9768,
"grad_norm": 0.8083638986140238,
"learning_rate": 2.1473175098229993e-09,
"loss": 0.2889,
"step": 2482
},
{
"epoch": 2.9779999999999998,
"grad_norm": 0.8400439805167591,
"learning_rate": 1.9476932431500286e-09,
"loss": 0.2423,
"step": 2483
},
{
"epoch": 2.9792,
"grad_norm": 0.6996377169152231,
"learning_rate": 1.7578042796739803e-09,
"loss": 0.2777,
"step": 2484
},
{
"epoch": 2.9804,
"grad_norm": 0.6374977751204586,
"learning_rate": 1.5776509892645453e-09,
"loss": 0.2944,
"step": 2485
},
{
"epoch": 2.9816000000000003,
"grad_norm": 0.7965767939944651,
"learning_rate": 1.4072337228282496e-09,
"loss": 0.2795,
"step": 2486
},
{
"epoch": 2.9828,
"grad_norm": 0.5942698125217902,
"learning_rate": 1.2465528123073445e-09,
"loss": 0.3073,
"step": 2487
},
{
"epoch": 2.984,
"grad_norm": 0.7911530212423521,
"learning_rate": 1.0956085706781416e-09,
"loss": 0.2838,
"step": 2488
},
{
"epoch": 2.9852,
"grad_norm": 1.0102002020271916,
"learning_rate": 9.54401291953788e-10,
"loss": 0.2624,
"step": 2489
},
{
"epoch": 2.9864,
"grad_norm": 0.6927474640405619,
"learning_rate": 8.229312511803811e-10,
"loss": 0.2627,
"step": 2490
},
{
"epoch": 2.9876,
"grad_norm": 0.725536969478269,
"learning_rate": 7.011987044369673e-10,
"loss": 0.2836,
"step": 2491
},
{
"epoch": 2.9888,
"grad_norm": 0.6477991646840576,
"learning_rate": 5.892038888377638e-10,
"loss": 0.2513,
"step": 2492
},
{
"epoch": 2.99,
"grad_norm": 0.6631131932449185,
"learning_rate": 4.869470225277174e-10,
"loss": 0.2797,
"step": 2493
},
{
"epoch": 2.9912,
"grad_norm": 0.7647956082429362,
"learning_rate": 3.9442830468472414e-10,
"loss": 0.2426,
"step": 2494
},
{
"epoch": 2.9924,
"grad_norm": 0.6713755406980298,
"learning_rate": 3.1164791551907545e-10,
"loss": 0.2559,
"step": 2495
},
{
"epoch": 2.9936,
"grad_norm": 0.7035301008315098,
"learning_rate": 2.386060162717918e-10,
"loss": 0.25,
"step": 2496
},
{
"epoch": 2.9948,
"grad_norm": 0.711419557083978,
"learning_rate": 1.7530274921462308e-10,
"loss": 0.2999,
"step": 2497
},
{
"epoch": 2.996,
"grad_norm": 0.7520518854348565,
"learning_rate": 1.21738237651714e-10,
"loss": 0.2774,
"step": 2498
},
{
"epoch": 2.9972,
"grad_norm": 0.8767599854478709,
"learning_rate": 7.791258591682837e-11,
"loss": 0.3085,
"step": 2499
},
{
"epoch": 2.9984,
"grad_norm": 0.5408305851587399,
"learning_rate": 4.382587937445948e-11,
"loss": 0.2905,
"step": 2500
},
{
"epoch": 2.9996,
"grad_norm": 0.8299389847414689,
"learning_rate": 1.947818441927485e-11,
"loss": 0.3371,
"step": 2501
},
{
"epoch": 3.0,
"grad_norm": 1.931893729681929,
"learning_rate": 4.8695484761163145e-12,
"loss": 0.3254,
"step": 2502
},
{
"epoch": 3.0,
"step": 2502,
"total_flos": 446037403000832.0,
"train_loss": 0.3789820095284475,
"train_runtime": 44495.1254,
"train_samples_per_second": 1.348,
"train_steps_per_second": 0.056
}
],
"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": 446037403000832.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}