mesolitica/Malaysian-Llama-3.1-8B-Instruct
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity
Files
main
Malaysian-Llama-3.1-8B-Inst…/lora-embedding-128-llama3.1-8b-malaysian-8k/checkpoint-2600/trainer_state.json
ModelHub XC a580c21353 初始化项目，由ModelHub XC社区提供模型 
Model: mesolitica/Malaysian-Llama-3.1-8B-Instruct
Source: Original Platform
2026-08-11 15:03:18 +08:00

18235 lines
400 KiB
JSON
Raw Permalink Blame History

{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 0.8974801518812565,
"eval_steps": 500,
"global_step": 2600,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.00034518467380048324,
"grad_norm": 164.25050354003906,
"learning_rate": 0.0,
"loss": 1.9257,
"step": 1
},
{
"epoch": 0.0006903693476009665,
"grad_norm": 116.43001556396484,
"learning_rate": 2e-05,
"loss": 1.8201,
"step": 2
},
{
"epoch": 0.0010355540214014498,
"grad_norm": 111.98278045654297,
"learning_rate": 2e-05,
"loss": 1.9652,
"step": 3
},
{
"epoch": 0.001380738695201933,
"grad_norm": 312.7107238769531,
"learning_rate": 2e-05,
"loss": 1.6971,
"step": 4
},
{
"epoch": 0.0017259233690024164,
"grad_norm": 124.51359558105469,
"learning_rate": 2e-05,
"loss": 1.7228,
"step": 5
},
{
"epoch": 0.0020711080428028996,
"grad_norm": 117.87893676757812,
"learning_rate": 2e-05,
"loss": 1.7254,
"step": 6
},
{
"epoch": 0.002416292716603383,
"grad_norm": 59.12928771972656,
"learning_rate": 2e-05,
"loss": 1.473,
"step": 7
},
{
"epoch": 0.002761477390403866,
"grad_norm": 24.38951873779297,
"learning_rate": 2e-05,
"loss": 1.5216,
"step": 8
},
{
"epoch": 0.0031066620642043494,
"grad_norm": 57.787757873535156,
"learning_rate": 2e-05,
"loss": 1.5563,
"step": 9
},
{
"epoch": 0.0034518467380048328,
"grad_norm": 21.435535430908203,
"learning_rate": 2e-05,
"loss": 1.4052,
"step": 10
},
{
"epoch": 0.0037970314118053157,
"grad_norm": 25.76114273071289,
"learning_rate": 2e-05,
"loss": 1.4289,
"step": 11
},
{
"epoch": 0.004142216085605799,
"grad_norm": 16.697547912597656,
"learning_rate": 2e-05,
"loss": 1.5014,
"step": 12
},
{
"epoch": 0.0044874007594062825,
"grad_norm": 10.946109771728516,
"learning_rate": 2e-05,
"loss": 1.4236,
"step": 13
},
{
"epoch": 0.004832585433206766,
"grad_norm": 9.638275146484375,
"learning_rate": 2e-05,
"loss": 1.4098,
"step": 14
},
{
"epoch": 0.0051777701070072485,
"grad_norm": 17.02400016784668,
"learning_rate": 2e-05,
"loss": 1.4035,
"step": 15
},
{
"epoch": 0.005522954780807732,
"grad_norm": 7.689630508422852,
"learning_rate": 2e-05,
"loss": 1.3273,
"step": 16
},
{
"epoch": 0.005868139454608215,
"grad_norm": 7.028171539306641,
"learning_rate": 2e-05,
"loss": 1.3675,
"step": 17
},
{
"epoch": 0.006213324128408699,
"grad_norm": 8.425003051757812,
"learning_rate": 2e-05,
"loss": 1.3065,
"step": 18
},
{
"epoch": 0.006558508802209182,
"grad_norm": 7.860549449920654,
"learning_rate": 2e-05,
"loss": 1.2914,
"step": 19
},
{
"epoch": 0.0069036934760096655,
"grad_norm": 5.793131351470947,
"learning_rate": 2e-05,
"loss": 1.2639,
"step": 20
},
{
"epoch": 0.007248878149810148,
"grad_norm": 6.066828727722168,
"learning_rate": 2e-05,
"loss": 1.2845,
"step": 21
},
{
"epoch": 0.0075940628236106315,
"grad_norm": 6.9683074951171875,
"learning_rate": 2e-05,
"loss": 1.2839,
"step": 22
},
{
"epoch": 0.007939247497411116,
"grad_norm": 7.026453018188477,
"learning_rate": 2e-05,
"loss": 1.3129,
"step": 23
},
{
"epoch": 0.008284432171211598,
"grad_norm": 6.011166095733643,
"learning_rate": 2e-05,
"loss": 1.2021,
"step": 24
},
{
"epoch": 0.00862961684501208,
"grad_norm": 6.688496112823486,
"learning_rate": 2e-05,
"loss": 1.2988,
"step": 25
},
{
"epoch": 0.008974801518812565,
"grad_norm": 5.536924839019775,
"learning_rate": 2e-05,
"loss": 1.2529,
"step": 26
},
{
"epoch": 0.009319986192613048,
"grad_norm": 6.189913749694824,
"learning_rate": 2e-05,
"loss": 1.2541,
"step": 27
},
{
"epoch": 0.009665170866413532,
"grad_norm": 5.186566352844238,
"learning_rate": 2e-05,
"loss": 1.1685,
"step": 28
},
{
"epoch": 0.010010355540214014,
"grad_norm": 5.111907482147217,
"learning_rate": 2e-05,
"loss": 1.1647,
"step": 29
},
{
"epoch": 0.010355540214014497,
"grad_norm": 7.816215515136719,
"learning_rate": 2e-05,
"loss": 1.1613,
"step": 30
},
{
"epoch": 0.010700724887814981,
"grad_norm": 6.994287014007568,
"learning_rate": 2e-05,
"loss": 1.1798,
"step": 31
},
{
"epoch": 0.011045909561615464,
"grad_norm": 6.679455280303955,
"learning_rate": 2e-05,
"loss": 1.1356,
"step": 32
},
{
"epoch": 0.011391094235415948,
"grad_norm": 5.451405048370361,
"learning_rate": 2e-05,
"loss": 1.1679,
"step": 33
},
{
"epoch": 0.01173627890921643,
"grad_norm": 5.443926811218262,
"learning_rate": 2e-05,
"loss": 1.1729,
"step": 34
},
{
"epoch": 0.012081463583016915,
"grad_norm": 5.605351448059082,
"learning_rate": 2e-05,
"loss": 1.1808,
"step": 35
},
{
"epoch": 0.012426648256817397,
"grad_norm": 5.121994972229004,
"learning_rate": 2e-05,
"loss": 1.1379,
"step": 36
},
{
"epoch": 0.01277183293061788,
"grad_norm": 4.825695037841797,
"learning_rate": 2e-05,
"loss": 1.1291,
"step": 37
},
{
"epoch": 0.013117017604418364,
"grad_norm": 4.905743598937988,
"learning_rate": 2e-05,
"loss": 1.1687,
"step": 38
},
{
"epoch": 0.013462202278218847,
"grad_norm": 5.373997688293457,
"learning_rate": 2e-05,
"loss": 1.2164,
"step": 39
},
{
"epoch": 0.013807386952019331,
"grad_norm": 4.702381134033203,
"learning_rate": 2e-05,
"loss": 1.1769,
"step": 40
},
{
"epoch": 0.014152571625819814,
"grad_norm": 5.017072677612305,
"learning_rate": 2e-05,
"loss": 1.1081,
"step": 41
},
{
"epoch": 0.014497756299620296,
"grad_norm": 4.598694801330566,
"learning_rate": 2e-05,
"loss": 1.0872,
"step": 42
},
{
"epoch": 0.01484294097342078,
"grad_norm": 4.88613224029541,
"learning_rate": 2e-05,
"loss": 1.166,
"step": 43
},
{
"epoch": 0.015188125647221263,
"grad_norm": 4.172844886779785,
"learning_rate": 2e-05,
"loss": 1.1078,
"step": 44
},
{
"epoch": 0.015533310321021747,
"grad_norm": 4.9598612785339355,
"learning_rate": 2e-05,
"loss": 1.1068,
"step": 45
},
{
"epoch": 0.01587849499482223,
"grad_norm": 4.003320693969727,
"learning_rate": 2e-05,
"loss": 1.0951,
"step": 46
},
{
"epoch": 0.016223679668622714,
"grad_norm": 4.30601692199707,
"learning_rate": 2e-05,
"loss": 1.1088,
"step": 47
},
{
"epoch": 0.016568864342423197,
"grad_norm": 4.957517623901367,
"learning_rate": 2e-05,
"loss": 1.1152,
"step": 48
},
{
"epoch": 0.01691404901622368,
"grad_norm": 5.051650047302246,
"learning_rate": 2e-05,
"loss": 1.0514,
"step": 49
},
{
"epoch": 0.01725923369002416,
"grad_norm": 4.508876800537109,
"learning_rate": 2e-05,
"loss": 1.1081,
"step": 50
},
{
"epoch": 0.017604418363824648,
"grad_norm": 4.666370391845703,
"learning_rate": 2e-05,
"loss": 1.0889,
"step": 51
},
{
"epoch": 0.01794960303762513,
"grad_norm": 4.62983512878418,
"learning_rate": 2e-05,
"loss": 1.0561,
"step": 52
},
{
"epoch": 0.018294787711425613,
"grad_norm": 3.8194386959075928,
"learning_rate": 2e-05,
"loss": 1.0314,
"step": 53
},
{
"epoch": 0.018639972385226095,
"grad_norm": 4.9420695304870605,
"learning_rate": 2e-05,
"loss": 1.1218,
"step": 54
},
{
"epoch": 0.018985157059026578,
"grad_norm": 4.9239068031311035,
"learning_rate": 2e-05,
"loss": 1.0647,
"step": 55
},
{
"epoch": 0.019330341732827064,
"grad_norm": 4.308078289031982,
"learning_rate": 2e-05,
"loss": 1.0469,
"step": 56
},
{
"epoch": 0.019675526406627546,
"grad_norm": 4.183582782745361,
"learning_rate": 2e-05,
"loss": 1.101,
"step": 57
},
{
"epoch": 0.02002071108042803,
"grad_norm": 5.259477138519287,
"learning_rate": 2e-05,
"loss": 1.1166,
"step": 58
},
{
"epoch": 0.02036589575422851,
"grad_norm": 4.483434677124023,
"learning_rate": 2e-05,
"loss": 1.0173,
"step": 59
},
{
"epoch": 0.020711080428028994,
"grad_norm": 4.112723350524902,
"learning_rate": 2e-05,
"loss": 1.062,
"step": 60
},
{
"epoch": 0.02105626510182948,
"grad_norm": 4.924528121948242,
"learning_rate": 2e-05,
"loss": 1.0574,
"step": 61
},
{
"epoch": 0.021401449775629963,
"grad_norm": 3.717799663543701,
"learning_rate": 2e-05,
"loss": 1.0268,
"step": 62
},
{
"epoch": 0.021746634449430445,
"grad_norm": 4.5001044273376465,
"learning_rate": 2e-05,
"loss": 1.057,
"step": 63
},
{
"epoch": 0.022091819123230928,
"grad_norm": 4.214874267578125,
"learning_rate": 2e-05,
"loss": 1.1244,
"step": 64
},
{
"epoch": 0.02243700379703141,
"grad_norm": 4.098433494567871,
"learning_rate": 2e-05,
"loss": 1.0685,
"step": 65
},
{
"epoch": 0.022782188470831896,
"grad_norm": 3.9524285793304443,
"learning_rate": 2e-05,
"loss": 1.0642,
"step": 66
},
{
"epoch": 0.02312737314463238,
"grad_norm": 4.906257152557373,
"learning_rate": 2e-05,
"loss": 1.0543,
"step": 67
},
{
"epoch": 0.02347255781843286,
"grad_norm": 5.597499847412109,
"learning_rate": 2e-05,
"loss": 1.0525,
"step": 68
},
{
"epoch": 0.023817742492233344,
"grad_norm": 3.9120569229125977,
"learning_rate": 2e-05,
"loss": 1.0316,
"step": 69
},
{
"epoch": 0.02416292716603383,
"grad_norm": 3.9987313747406006,
"learning_rate": 2e-05,
"loss": 1.0415,
"step": 70
},
{
"epoch": 0.024508111839834312,
"grad_norm": 3.854943037033081,
"learning_rate": 2e-05,
"loss": 1.0162,
"step": 71
},
{
"epoch": 0.024853296513634795,
"grad_norm": 4.38804292678833,
"learning_rate": 2e-05,
"loss": 0.9893,
"step": 72
},
{
"epoch": 0.025198481187435277,
"grad_norm": 3.375614881515503,
"learning_rate": 2e-05,
"loss": 1.0135,
"step": 73
},
{
"epoch": 0.02554366586123576,
"grad_norm": 3.9334664344787598,
"learning_rate": 2e-05,
"loss": 1.0267,
"step": 74
},
{
"epoch": 0.025888850535036246,
"grad_norm": 3.8816137313842773,
"learning_rate": 2e-05,
"loss": 1.0347,
"step": 75
},
{
"epoch": 0.02623403520883673,
"grad_norm": 3.884535551071167,
"learning_rate": 2e-05,
"loss": 0.9914,
"step": 76
},
{
"epoch": 0.02657921988263721,
"grad_norm": 3.833580493927002,
"learning_rate": 2e-05,
"loss": 1.021,
"step": 77
},
{
"epoch": 0.026924404556437694,
"grad_norm": 3.6896729469299316,
"learning_rate": 2e-05,
"loss": 0.9915,
"step": 78
},
{
"epoch": 0.027269589230238176,
"grad_norm": 3.8832285404205322,
"learning_rate": 2e-05,
"loss": 1.0081,
"step": 79
},
{
"epoch": 0.027614773904038662,
"grad_norm": 4.00054931640625,
"learning_rate": 2e-05,
"loss": 0.9723,
"step": 80
},
{
"epoch": 0.027959958577839145,
"grad_norm": 3.9635231494903564,
"learning_rate": 2e-05,
"loss": 1.0677,
"step": 81
},
{
"epoch": 0.028305143251639627,
"grad_norm": 4.25091552734375,
"learning_rate": 2e-05,
"loss": 0.9683,
"step": 82
},
{
"epoch": 0.02865032792544011,
"grad_norm": 4.387684345245361,
"learning_rate": 2e-05,
"loss": 1.0273,
"step": 83
},
{
"epoch": 0.028995512599240592,
"grad_norm": 4.047482013702393,
"learning_rate": 2e-05,
"loss": 1.0372,
"step": 84
},
{
"epoch": 0.029340697273041078,
"grad_norm": 4.226908206939697,
"learning_rate": 2e-05,
"loss": 1.0192,
"step": 85
},
{
"epoch": 0.02968588194684156,
"grad_norm": 4.123228073120117,
"learning_rate": 2e-05,
"loss": 0.9915,
"step": 86
},
{
"epoch": 0.030031066620642043,
"grad_norm": 6.669504165649414,
"learning_rate": 2e-05,
"loss": 0.9531,
"step": 87
},
{
"epoch": 0.030376251294442526,
"grad_norm": 4.364908695220947,
"learning_rate": 2e-05,
"loss": 1.0042,
"step": 88
},
{
"epoch": 0.03072143596824301,
"grad_norm": 3.7988955974578857,
"learning_rate": 2e-05,
"loss": 0.9767,
"step": 89
},
{
"epoch": 0.031066620642043494,
"grad_norm": 3.547370433807373,
"learning_rate": 2e-05,
"loss": 1.0045,
"step": 90
},
{
"epoch": 0.03141180531584398,
"grad_norm": 4.054666042327881,
"learning_rate": 2e-05,
"loss": 0.9679,
"step": 91
},
{
"epoch": 0.03175698998964446,
"grad_norm": 4.089395523071289,
"learning_rate": 2e-05,
"loss": 1.009,
"step": 92
},
{
"epoch": 0.03210217466344494,
"grad_norm": 3.888781785964966,
"learning_rate": 2e-05,
"loss": 0.9993,
"step": 93
},
{
"epoch": 0.03244735933724543,
"grad_norm": 4.337654113769531,
"learning_rate": 2e-05,
"loss": 1.0637,
"step": 94
},
{
"epoch": 0.03279254401104591,
"grad_norm": 3.9866831302642822,
"learning_rate": 2e-05,
"loss": 1.0415,
"step": 95
},
{
"epoch": 0.03313772868484639,
"grad_norm": 3.694230556488037,
"learning_rate": 2e-05,
"loss": 1.0111,
"step": 96
},
{
"epoch": 0.03348291335864688,
"grad_norm": 3.8903889656066895,
"learning_rate": 2e-05,
"loss": 1.0113,
"step": 97
},
{
"epoch": 0.03382809803244736,
"grad_norm": 4.173677444458008,
"learning_rate": 2e-05,
"loss": 0.9969,
"step": 98
},
{
"epoch": 0.034173282706247844,
"grad_norm": 3.679419755935669,
"learning_rate": 2e-05,
"loss": 1.0401,
"step": 99
},
{
"epoch": 0.03451846738004832,
"grad_norm": 4.073668003082275,
"learning_rate": 2e-05,
"loss": 1.0062,
"step": 100
},
{
"epoch": 0.03486365205384881,
"grad_norm": 3.659552574157715,
"learning_rate": 2e-05,
"loss": 0.9113,
"step": 101
},
{
"epoch": 0.035208836727649295,
"grad_norm": 3.581220865249634,
"learning_rate": 2e-05,
"loss": 1.0402,
"step": 102
},
{
"epoch": 0.035554021401449774,
"grad_norm": 3.6875903606414795,
"learning_rate": 2e-05,
"loss": 1.0051,
"step": 103
},
{
"epoch": 0.03589920607525026,
"grad_norm": 3.3204426765441895,
"learning_rate": 2e-05,
"loss": 1.0041,
"step": 104
},
{
"epoch": 0.03624439074905074,
"grad_norm": 3.408184766769409,
"learning_rate": 2e-05,
"loss": 0.9501,
"step": 105
},
{
"epoch": 0.036589575422851225,
"grad_norm": 3.7698254585266113,
"learning_rate": 2e-05,
"loss": 0.9487,
"step": 106
},
{
"epoch": 0.03693476009665171,
"grad_norm": 4.5543694496154785,
"learning_rate": 2e-05,
"loss": 1.0095,
"step": 107
},
{
"epoch": 0.03727994477045219,
"grad_norm": 4.206326961517334,
"learning_rate": 2e-05,
"loss": 0.9307,
"step": 108
},
{
"epoch": 0.037625129444252677,
"grad_norm": 3.7120132446289062,
"learning_rate": 2e-05,
"loss": 0.9017,
"step": 109
},
{
"epoch": 0.037970314118053156,
"grad_norm": 3.6532175540924072,
"learning_rate": 2e-05,
"loss": 0.9544,
"step": 110
},
{
"epoch": 0.03831549879185364,
"grad_norm": 3.4927427768707275,
"learning_rate": 2e-05,
"loss": 0.97,
"step": 111
},
{
"epoch": 0.03866068346565413,
"grad_norm": 4.344557762145996,
"learning_rate": 2e-05,
"loss": 0.9311,
"step": 112
},
{
"epoch": 0.03900586813945461,
"grad_norm": 3.0494396686553955,
"learning_rate": 2e-05,
"loss": 1.0055,
"step": 113
},
{
"epoch": 0.03935105281325509,
"grad_norm": 3.7052459716796875,
"learning_rate": 2e-05,
"loss": 0.9356,
"step": 114
},
{
"epoch": 0.03969623748705557,
"grad_norm": 4.43536376953125,
"learning_rate": 2e-05,
"loss": 0.9695,
"step": 115
},
{
"epoch": 0.04004142216085606,
"grad_norm": 3.6565682888031006,
"learning_rate": 2e-05,
"loss": 1.0503,
"step": 116
},
{
"epoch": 0.040386606834656544,
"grad_norm": 3.5132741928100586,
"learning_rate": 2e-05,
"loss": 0.959,
"step": 117
},
{
"epoch": 0.04073179150845702,
"grad_norm": 4.00861120223999,
"learning_rate": 2e-05,
"loss": 0.8973,
"step": 118
},
{
"epoch": 0.04107697618225751,
"grad_norm": 4.118383407592773,
"learning_rate": 2e-05,
"loss": 0.961,
"step": 119
},
{
"epoch": 0.04142216085605799,
"grad_norm": 3.6368112564086914,
"learning_rate": 2e-05,
"loss": 1.0302,
"step": 120
},
{
"epoch": 0.041767345529858474,
"grad_norm": 3.668741226196289,
"learning_rate": 2e-05,
"loss": 0.9267,
"step": 121
},
{
"epoch": 0.04211253020365896,
"grad_norm": 3.392117500305176,
"learning_rate": 2e-05,
"loss": 1.0133,
"step": 122
},
{
"epoch": 0.04245771487745944,
"grad_norm": 3.939965009689331,
"learning_rate": 2e-05,
"loss": 0.9151,
"step": 123
},
{
"epoch": 0.042802899551259925,
"grad_norm": 3.3222975730895996,
"learning_rate": 2e-05,
"loss": 0.9028,
"step": 124
},
{
"epoch": 0.043148084225060404,
"grad_norm": 3.639970541000366,
"learning_rate": 2e-05,
"loss": 0.9709,
"step": 125
},
{
"epoch": 0.04349326889886089,
"grad_norm": 3.2109615802764893,
"learning_rate": 2e-05,
"loss": 0.9867,
"step": 126
},
{
"epoch": 0.043838453572661376,
"grad_norm": 3.3262574672698975,
"learning_rate": 2e-05,
"loss": 0.964,
"step": 127
},
{
"epoch": 0.044183638246461855,
"grad_norm": 5.754081726074219,
"learning_rate": 2e-05,
"loss": 0.8765,
"step": 128
},
{
"epoch": 0.04452882292026234,
"grad_norm": 3.367676019668579,
"learning_rate": 2e-05,
"loss": 0.9724,
"step": 129
},
{
"epoch": 0.04487400759406282,
"grad_norm": 3.8047783374786377,
"learning_rate": 2e-05,
"loss": 0.9596,
"step": 130
},
{
"epoch": 0.045219192267863306,
"grad_norm": 3.5262019634246826,
"learning_rate": 2e-05,
"loss": 0.8931,
"step": 131
},
{
"epoch": 0.04556437694166379,
"grad_norm": 3.1185085773468018,
"learning_rate": 2e-05,
"loss": 0.891,
"step": 132
},
{
"epoch": 0.04590956161546427,
"grad_norm": 3.795315980911255,
"learning_rate": 2e-05,
"loss": 0.9536,
"step": 133
},
{
"epoch": 0.04625474628926476,
"grad_norm": 3.698105812072754,
"learning_rate": 2e-05,
"loss": 0.918,
"step": 134
},
{
"epoch": 0.04659993096306524,
"grad_norm": 4.450272560119629,
"learning_rate": 2e-05,
"loss": 0.9075,
"step": 135
},
{
"epoch": 0.04694511563686572,
"grad_norm": 3.3076443672180176,
"learning_rate": 2e-05,
"loss": 0.9334,
"step": 136
},
{
"epoch": 0.04729030031066621,
"grad_norm": 3.551239252090454,
"learning_rate": 2e-05,
"loss": 0.9221,
"step": 137
},
{
"epoch": 0.04763548498446669,
"grad_norm": 3.3865537643432617,
"learning_rate": 2e-05,
"loss": 0.96,
"step": 138
},
{
"epoch": 0.047980669658267174,
"grad_norm": 3.477189302444458,
"learning_rate": 2e-05,
"loss": 0.9794,
"step": 139
},
{
"epoch": 0.04832585433206766,
"grad_norm": 3.2932441234588623,
"learning_rate": 2e-05,
"loss": 0.9078,
"step": 140
},
{
"epoch": 0.04867103900586814,
"grad_norm": 3.095069646835327,
"learning_rate": 2e-05,
"loss": 0.9045,
"step": 141
},
{
"epoch": 0.049016223679668625,
"grad_norm": 3.386526107788086,
"learning_rate": 2e-05,
"loss": 0.937,
"step": 142
},
{
"epoch": 0.049361408353469104,
"grad_norm": 3.274214267730713,
"learning_rate": 2e-05,
"loss": 0.9356,
"step": 143
},
{
"epoch": 0.04970659302726959,
"grad_norm": 3.472280263900757,
"learning_rate": 2e-05,
"loss": 0.8798,
"step": 144
},
{
"epoch": 0.050051777701070076,
"grad_norm": 3.4150500297546387,
"learning_rate": 2e-05,
"loss": 1.0118,
"step": 145
},
{
"epoch": 0.050396962374870555,
"grad_norm": 3.7823333740234375,
"learning_rate": 2e-05,
"loss": 0.9675,
"step": 146
},
{
"epoch": 0.05074214704867104,
"grad_norm": 3.460674524307251,
"learning_rate": 2e-05,
"loss": 0.9366,
"step": 147
},
{
"epoch": 0.05108733172247152,
"grad_norm": 4.081386089324951,
"learning_rate": 2e-05,
"loss": 0.9596,
"step": 148
},
{
"epoch": 0.051432516396272006,
"grad_norm": 3.5548624992370605,
"learning_rate": 2e-05,
"loss": 0.9369,
"step": 149
},
{
"epoch": 0.05177770107007249,
"grad_norm": 3.713184118270874,
"learning_rate": 2e-05,
"loss": 0.9645,
"step": 150
},
{
"epoch": 0.05212288574387297,
"grad_norm": 9.342556953430176,
"learning_rate": 2e-05,
"loss": 0.953,
"step": 151
},
{
"epoch": 0.05246807041767346,
"grad_norm": 3.330892562866211,
"learning_rate": 2e-05,
"loss": 0.9851,
"step": 152
},
{
"epoch": 0.052813255091473936,
"grad_norm": 3.2820513248443604,
"learning_rate": 2e-05,
"loss": 0.9881,
"step": 153
},
{
"epoch": 0.05315843976527442,
"grad_norm": 3.08601713180542,
"learning_rate": 2e-05,
"loss": 0.8678,
"step": 154
},
{
"epoch": 0.05350362443907491,
"grad_norm": 3.466398239135742,
"learning_rate": 2e-05,
"loss": 0.9688,
"step": 155
},
{
"epoch": 0.05384880911287539,
"grad_norm": 3.8165998458862305,
"learning_rate": 2e-05,
"loss": 0.9516,
"step": 156
},
{
"epoch": 0.05419399378667587,
"grad_norm": 3.843984365463257,
"learning_rate": 2e-05,
"loss": 0.8846,
"step": 157
},
{
"epoch": 0.05453917846047635,
"grad_norm": 3.6460742950439453,
"learning_rate": 2e-05,
"loss": 0.9738,
"step": 158
},
{
"epoch": 0.05488436313427684,
"grad_norm": 3.6064538955688477,
"learning_rate": 2e-05,
"loss": 0.8974,
"step": 159
},
{
"epoch": 0.055229547808077324,
"grad_norm": 3.8065435886383057,
"learning_rate": 2e-05,
"loss": 0.9804,
"step": 160
},
{
"epoch": 0.0555747324818778,
"grad_norm": 3.020841121673584,
"learning_rate": 2e-05,
"loss": 0.9037,
"step": 161
},
{
"epoch": 0.05591991715567829,
"grad_norm": 3.5063493251800537,
"learning_rate": 2e-05,
"loss": 0.892,
"step": 162
},
{
"epoch": 0.05626510182947877,
"grad_norm": 3.2224860191345215,
"learning_rate": 2e-05,
"loss": 0.8995,
"step": 163
},
{
"epoch": 0.056610286503279254,
"grad_norm": 3.520097494125366,
"learning_rate": 2e-05,
"loss": 0.9791,
"step": 164
},
{
"epoch": 0.05695547117707974,
"grad_norm": 3.435135841369629,
"learning_rate": 2e-05,
"loss": 1.0013,
"step": 165
},
{
"epoch": 0.05730065585088022,
"grad_norm": 3.173973321914673,
"learning_rate": 2e-05,
"loss": 0.9146,
"step": 166
},
{
"epoch": 0.057645840524680705,
"grad_norm": 3.4863386154174805,
"learning_rate": 2e-05,
"loss": 0.9885,
"step": 167
},
{
"epoch": 0.057991025198481184,
"grad_norm": 3.2673957347869873,
"learning_rate": 2e-05,
"loss": 0.9354,
"step": 168
},
{
"epoch": 0.05833620987228167,
"grad_norm": 3.613051652908325,
"learning_rate": 2e-05,
"loss": 0.9169,
"step": 169
},
{
"epoch": 0.058681394546082156,
"grad_norm": 3.6220316886901855,
"learning_rate": 2e-05,
"loss": 0.9123,
"step": 170
},
{
"epoch": 0.059026579219882636,
"grad_norm": 2.9920058250427246,
"learning_rate": 2e-05,
"loss": 0.9228,
"step": 171
},
{
"epoch": 0.05937176389368312,
"grad_norm": 3.1163201332092285,
"learning_rate": 2e-05,
"loss": 0.9196,
"step": 172
},
{
"epoch": 0.0597169485674836,
"grad_norm": 3.5045080184936523,
"learning_rate": 2e-05,
"loss": 0.9711,
"step": 173
},
{
"epoch": 0.06006213324128409,
"grad_norm": 3.3398544788360596,
"learning_rate": 2e-05,
"loss": 0.8714,
"step": 174
},
{
"epoch": 0.06040731791508457,
"grad_norm": 2.9095466136932373,
"learning_rate": 2e-05,
"loss": 0.8746,
"step": 175
},
{
"epoch": 0.06075250258888505,
"grad_norm": 3.1426031589508057,
"learning_rate": 2e-05,
"loss": 0.8826,
"step": 176
},
{
"epoch": 0.06109768726268554,
"grad_norm": 4.104501724243164,
"learning_rate": 2e-05,
"loss": 0.9346,
"step": 177
},
{
"epoch": 0.06144287193648602,
"grad_norm": 2.990579128265381,
"learning_rate": 2e-05,
"loss": 0.9675,
"step": 178
},
{
"epoch": 0.0617880566102865,
"grad_norm": 3.756319284439087,
"learning_rate": 2e-05,
"loss": 0.9229,
"step": 179
},
{
"epoch": 0.06213324128408699,
"grad_norm": 3.5764338970184326,
"learning_rate": 2e-05,
"loss": 0.9925,
"step": 180
},
{
"epoch": 0.06247842595788747,
"grad_norm": 3.068021774291992,
"learning_rate": 2e-05,
"loss": 0.8987,
"step": 181
},
{
"epoch": 0.06282361063168795,
"grad_norm": 3.2142879962921143,
"learning_rate": 2e-05,
"loss": 0.8418,
"step": 182
},
{
"epoch": 0.06316879530548844,
"grad_norm": 3.3960886001586914,
"learning_rate": 2e-05,
"loss": 0.8892,
"step": 183
},
{
"epoch": 0.06351397997928893,
"grad_norm": 4.435737609863281,
"learning_rate": 2e-05,
"loss": 0.9485,
"step": 184
},
{
"epoch": 0.0638591646530894,
"grad_norm": 11.62714958190918,
"learning_rate": 2e-05,
"loss": 0.8685,
"step": 185
},
{
"epoch": 0.06420434932688988,
"grad_norm": 3.0193748474121094,
"learning_rate": 2e-05,
"loss": 0.9131,
"step": 186
},
{
"epoch": 0.06454953400069037,
"grad_norm": 3.1290862560272217,
"learning_rate": 2e-05,
"loss": 0.9776,
"step": 187
},
{
"epoch": 0.06489471867449086,
"grad_norm": 3.153416872024536,
"learning_rate": 2e-05,
"loss": 0.9031,
"step": 188
},
{
"epoch": 0.06523990334829134,
"grad_norm": 3.450758457183838,
"learning_rate": 2e-05,
"loss": 0.8928,
"step": 189
},
{
"epoch": 0.06558508802209181,
"grad_norm": 3.281147003173828,
"learning_rate": 2e-05,
"loss": 1.0054,
"step": 190
},
{
"epoch": 0.0659302726958923,
"grad_norm": 3.1141176223754883,
"learning_rate": 2e-05,
"loss": 0.913,
"step": 191
},
{
"epoch": 0.06627545736969279,
"grad_norm": 2.945939064025879,
"learning_rate": 2e-05,
"loss": 0.895,
"step": 192
},
{
"epoch": 0.06662064204349327,
"grad_norm": 3.141927719116211,
"learning_rate": 2e-05,
"loss": 0.9336,
"step": 193
},
{
"epoch": 0.06696582671729376,
"grad_norm": 3.6074554920196533,
"learning_rate": 2e-05,
"loss": 0.9821,
"step": 194
},
{
"epoch": 0.06731101139109423,
"grad_norm": 2.96323823928833,
"learning_rate": 2e-05,
"loss": 0.9417,
"step": 195
},
{
"epoch": 0.06765619606489472,
"grad_norm": 3.4270036220550537,
"learning_rate": 2e-05,
"loss": 0.9268,
"step": 196
},
{
"epoch": 0.0680013807386952,
"grad_norm": 3.006737232208252,
"learning_rate": 2e-05,
"loss": 0.9125,
"step": 197
},
{
"epoch": 0.06834656541249569,
"grad_norm": 3.3677432537078857,
"learning_rate": 2e-05,
"loss": 0.9403,
"step": 198
},
{
"epoch": 0.06869175008629617,
"grad_norm": 3.6785035133361816,
"learning_rate": 2e-05,
"loss": 0.8477,
"step": 199
},
{
"epoch": 0.06903693476009665,
"grad_norm": 3.1328837871551514,
"learning_rate": 2e-05,
"loss": 0.8866,
"step": 200
},
{
"epoch": 0.06938211943389713,
"grad_norm": 3.178600311279297,
"learning_rate": 2e-05,
"loss": 0.8418,
"step": 201
},
{
"epoch": 0.06972730410769762,
"grad_norm": 3.471315622329712,
"learning_rate": 2e-05,
"loss": 0.8908,
"step": 202
},
{
"epoch": 0.0700724887814981,
"grad_norm": 3.3694980144500732,
"learning_rate": 2e-05,
"loss": 0.8825,
"step": 203
},
{
"epoch": 0.07041767345529859,
"grad_norm": 3.511202812194824,
"learning_rate": 2e-05,
"loss": 0.89,
"step": 204
},
{
"epoch": 0.07076285812909906,
"grad_norm": 3.5508413314819336,
"learning_rate": 2e-05,
"loss": 0.8933,
"step": 205
},
{
"epoch": 0.07110804280289955,
"grad_norm": 2.9084413051605225,
"learning_rate": 2e-05,
"loss": 0.9258,
"step": 206
},
{
"epoch": 0.07145322747670003,
"grad_norm": 3.2880771160125732,
"learning_rate": 2e-05,
"loss": 0.9357,
"step": 207
},
{
"epoch": 0.07179841215050052,
"grad_norm": 3.110895872116089,
"learning_rate": 2e-05,
"loss": 0.8697,
"step": 208
},
{
"epoch": 0.072143596824301,
"grad_norm": 3.059079170227051,
"learning_rate": 2e-05,
"loss": 0.8579,
"step": 209
},
{
"epoch": 0.07248878149810148,
"grad_norm": 2.9385037422180176,
"learning_rate": 2e-05,
"loss": 0.8913,
"step": 210
},
{
"epoch": 0.07283396617190196,
"grad_norm": 2.984720468521118,
"learning_rate": 2e-05,
"loss": 0.9311,
"step": 211
},
{
"epoch": 0.07317915084570245,
"grad_norm": 3.1889357566833496,
"learning_rate": 2e-05,
"loss": 0.931,
"step": 212
},
{
"epoch": 0.07352433551950294,
"grad_norm": 3.020983934402466,
"learning_rate": 2e-05,
"loss": 0.8572,
"step": 213
},
{
"epoch": 0.07386952019330342,
"grad_norm": 3.3382914066314697,
"learning_rate": 2e-05,
"loss": 0.9128,
"step": 214
},
{
"epoch": 0.0742147048671039,
"grad_norm": 3.1657910346984863,
"learning_rate": 2e-05,
"loss": 0.8976,
"step": 215
},
{
"epoch": 0.07455988954090438,
"grad_norm": 3.058868885040283,
"learning_rate": 2e-05,
"loss": 0.8939,
"step": 216
},
{
"epoch": 0.07490507421470487,
"grad_norm": 3.422123908996582,
"learning_rate": 2e-05,
"loss": 0.9378,
"step": 217
},
{
"epoch": 0.07525025888850535,
"grad_norm": 3.1783478260040283,
"learning_rate": 2e-05,
"loss": 0.9485,
"step": 218
},
{
"epoch": 0.07559544356230584,
"grad_norm": 3.0557048320770264,
"learning_rate": 2e-05,
"loss": 0.8256,
"step": 219
},
{
"epoch": 0.07594062823610631,
"grad_norm": 2.9779226779937744,
"learning_rate": 2e-05,
"loss": 0.8713,
"step": 220
},
{
"epoch": 0.0762858129099068,
"grad_norm": 3.3758676052093506,
"learning_rate": 2e-05,
"loss": 0.8843,
"step": 221
},
{
"epoch": 0.07663099758370728,
"grad_norm": 3.1416192054748535,
"learning_rate": 2e-05,
"loss": 0.9064,
"step": 222
},
{
"epoch": 0.07697618225750777,
"grad_norm": 2.9653568267822266,
"learning_rate": 2e-05,
"loss": 0.8296,
"step": 223
},
{
"epoch": 0.07732136693130826,
"grad_norm": 3.2615296840667725,
"learning_rate": 2e-05,
"loss": 0.9057,
"step": 224
},
{
"epoch": 0.07766655160510873,
"grad_norm": 3.3381457328796387,
"learning_rate": 2e-05,
"loss": 0.9088,
"step": 225
},
{
"epoch": 0.07801173627890921,
"grad_norm": 3.4600749015808105,
"learning_rate": 2e-05,
"loss": 0.86,
"step": 226
},
{
"epoch": 0.0783569209527097,
"grad_norm": 3.2292826175689697,
"learning_rate": 2e-05,
"loss": 0.9591,
"step": 227
},
{
"epoch": 0.07870210562651019,
"grad_norm": 3.345545530319214,
"learning_rate": 2e-05,
"loss": 0.8476,
"step": 228
},
{
"epoch": 0.07904729030031067,
"grad_norm": 3.09397292137146,
"learning_rate": 2e-05,
"loss": 0.8652,
"step": 229
},
{
"epoch": 0.07939247497411114,
"grad_norm": 3.004117965698242,
"learning_rate": 2e-05,
"loss": 0.9689,
"step": 230
},
{
"epoch": 0.07973765964791163,
"grad_norm": 3.200100898742676,
"learning_rate": 2e-05,
"loss": 0.8666,
"step": 231
},
{
"epoch": 0.08008284432171212,
"grad_norm": 2.7272403240203857,
"learning_rate": 2e-05,
"loss": 0.8602,
"step": 232
},
{
"epoch": 0.0804280289955126,
"grad_norm": 3.2486329078674316,
"learning_rate": 2e-05,
"loss": 0.8464,
"step": 233
},
{
"epoch": 0.08077321366931309,
"grad_norm": 3.303467035293579,
"learning_rate": 2e-05,
"loss": 0.873,
"step": 234
},
{
"epoch": 0.08111839834311356,
"grad_norm": 3.0120198726654053,
"learning_rate": 2e-05,
"loss": 0.8934,
"step": 235
},
{
"epoch": 0.08146358301691405,
"grad_norm": 2.962071657180786,
"learning_rate": 2e-05,
"loss": 0.8904,
"step": 236
},
{
"epoch": 0.08180876769071453,
"grad_norm": 3.7542247772216797,
"learning_rate": 2e-05,
"loss": 0.9119,
"step": 237
},
{
"epoch": 0.08215395236451502,
"grad_norm": 3.281599521636963,
"learning_rate": 2e-05,
"loss": 0.9107,
"step": 238
},
{
"epoch": 0.0824991370383155,
"grad_norm": 3.1624579429626465,
"learning_rate": 2e-05,
"loss": 0.8766,
"step": 239
},
{
"epoch": 0.08284432171211598,
"grad_norm": 3.3022897243499756,
"learning_rate": 2e-05,
"loss": 0.9303,
"step": 240
},
{
"epoch": 0.08318950638591646,
"grad_norm": 3.1247267723083496,
"learning_rate": 2e-05,
"loss": 0.9125,
"step": 241
},
{
"epoch": 0.08353469105971695,
"grad_norm": 3.2210910320281982,
"learning_rate": 2e-05,
"loss": 0.9221,
"step": 242
},
{
"epoch": 0.08387987573351743,
"grad_norm": 3.1737873554229736,
"learning_rate": 2e-05,
"loss": 0.8178,
"step": 243
},
{
"epoch": 0.08422506040731792,
"grad_norm": 3.2427146434783936,
"learning_rate": 2e-05,
"loss": 0.9566,
"step": 244
},
{
"epoch": 0.08457024508111839,
"grad_norm": 3.4193999767303467,
"learning_rate": 2e-05,
"loss": 0.8557,
"step": 245
},
{
"epoch": 0.08491542975491888,
"grad_norm": 3.4254438877105713,
"learning_rate": 2e-05,
"loss": 0.8968,
"step": 246
},
{
"epoch": 0.08526061442871936,
"grad_norm": 3.159757614135742,
"learning_rate": 2e-05,
"loss": 0.9018,
"step": 247
},
{
"epoch": 0.08560579910251985,
"grad_norm": 3.187760353088379,
"learning_rate": 2e-05,
"loss": 0.9057,
"step": 248
},
{
"epoch": 0.08595098377632034,
"grad_norm": 2.959864616394043,
"learning_rate": 2e-05,
"loss": 0.904,
"step": 249
},
{
"epoch": 0.08629616845012081,
"grad_norm": 3.240973711013794,
"learning_rate": 2e-05,
"loss": 0.9294,
"step": 250
},
{
"epoch": 0.0866413531239213,
"grad_norm": 2.98917555809021,
"learning_rate": 2e-05,
"loss": 0.8764,
"step": 251
},
{
"epoch": 0.08698653779772178,
"grad_norm": 2.8783609867095947,
"learning_rate": 2e-05,
"loss": 0.8941,
"step": 252
},
{
"epoch": 0.08733172247152227,
"grad_norm": 3.0823488235473633,
"learning_rate": 2e-05,
"loss": 0.9319,
"step": 253
},
{
"epoch": 0.08767690714532275,
"grad_norm": 3.0972771644592285,
"learning_rate": 2e-05,
"loss": 0.8236,
"step": 254
},
{
"epoch": 0.08802209181912322,
"grad_norm": 3.005472421646118,
"learning_rate": 2e-05,
"loss": 0.8301,
"step": 255
},
{
"epoch": 0.08836727649292371,
"grad_norm": 3.145705223083496,
"learning_rate": 2e-05,
"loss": 0.866,
"step": 256
},
{
"epoch": 0.0887124611667242,
"grad_norm": 2.7804439067840576,
"learning_rate": 2e-05,
"loss": 0.8594,
"step": 257
},
{
"epoch": 0.08905764584052468,
"grad_norm": 3.092439889907837,
"learning_rate": 2e-05,
"loss": 0.8486,
"step": 258
},
{
"epoch": 0.08940283051432517,
"grad_norm": 3.228419303894043,
"learning_rate": 2e-05,
"loss": 0.8338,
"step": 259
},
{
"epoch": 0.08974801518812564,
"grad_norm": 3.093811273574829,
"learning_rate": 2e-05,
"loss": 0.9424,
"step": 260
},
{
"epoch": 0.09009319986192613,
"grad_norm": 3.0747523307800293,
"learning_rate": 2e-05,
"loss": 0.8944,
"step": 261
},
{
"epoch": 0.09043838453572661,
"grad_norm": 3.186995267868042,
"learning_rate": 2e-05,
"loss": 0.8586,
"step": 262
},
{
"epoch": 0.0907835692095271,
"grad_norm": 3.1833300590515137,
"learning_rate": 2e-05,
"loss": 0.8728,
"step": 263
},
{
"epoch": 0.09112875388332758,
"grad_norm": 3.0900795459747314,
"learning_rate": 2e-05,
"loss": 0.9022,
"step": 264
},
{
"epoch": 0.09147393855712806,
"grad_norm": 3.004206895828247,
"learning_rate": 2e-05,
"loss": 0.9062,
"step": 265
},
{
"epoch": 0.09181912323092854,
"grad_norm": 3.1954448223114014,
"learning_rate": 2e-05,
"loss": 0.8732,
"step": 266
},
{
"epoch": 0.09216430790472903,
"grad_norm": 3.324725866317749,
"learning_rate": 2e-05,
"loss": 0.9214,
"step": 267
},
{
"epoch": 0.09250949257852951,
"grad_norm": 2.8366189002990723,
"learning_rate": 2e-05,
"loss": 0.9153,
"step": 268
},
{
"epoch": 0.09285467725233,
"grad_norm": 3.264113187789917,
"learning_rate": 2e-05,
"loss": 0.8697,
"step": 269
},
{
"epoch": 0.09319986192613049,
"grad_norm": 3.186286211013794,
"learning_rate": 2e-05,
"loss": 0.9348,
"step": 270
},
{
"epoch": 0.09354504659993096,
"grad_norm": 2.8541440963745117,
"learning_rate": 2e-05,
"loss": 0.9067,
"step": 271
},
{
"epoch": 0.09389023127373144,
"grad_norm": 3.074235200881958,
"learning_rate": 2e-05,
"loss": 0.8935,
"step": 272
},
{
"epoch": 0.09423541594753193,
"grad_norm": 3.0192456245422363,
"learning_rate": 2e-05,
"loss": 0.8231,
"step": 273
},
{
"epoch": 0.09458060062133242,
"grad_norm": 2.740384101867676,
"learning_rate": 2e-05,
"loss": 0.8533,
"step": 274
},
{
"epoch": 0.0949257852951329,
"grad_norm": 2.9101722240448,
"learning_rate": 2e-05,
"loss": 0.8529,
"step": 275
},
{
"epoch": 0.09527096996893337,
"grad_norm": 3.137927770614624,
"learning_rate": 2e-05,
"loss": 0.9192,
"step": 276
},
{
"epoch": 0.09561615464273386,
"grad_norm": 3.054185628890991,
"learning_rate": 2e-05,
"loss": 0.8945,
"step": 277
},
{
"epoch": 0.09596133931653435,
"grad_norm": 3.1088597774505615,
"learning_rate": 2e-05,
"loss": 0.9008,
"step": 278
},
{
"epoch": 0.09630652399033483,
"grad_norm": 2.7827138900756836,
"learning_rate": 2e-05,
"loss": 0.8461,
"step": 279
},
{
"epoch": 0.09665170866413532,
"grad_norm": 2.864086389541626,
"learning_rate": 2e-05,
"loss": 0.8751,
"step": 280
},
{
"epoch": 0.09699689333793579,
"grad_norm": 3.1949148178100586,
"learning_rate": 2e-05,
"loss": 0.8377,
"step": 281
},
{
"epoch": 0.09734207801173628,
"grad_norm": 2.7971835136413574,
"learning_rate": 2e-05,
"loss": 0.9062,
"step": 282
},
{
"epoch": 0.09768726268553676,
"grad_norm": 6.014020919799805,
"learning_rate": 2e-05,
"loss": 0.8477,
"step": 283
},
{
"epoch": 0.09803244735933725,
"grad_norm": 3.1988205909729004,
"learning_rate": 2e-05,
"loss": 0.8605,
"step": 284
},
{
"epoch": 0.09837763203313774,
"grad_norm": 3.0080959796905518,
"learning_rate": 2e-05,
"loss": 0.9122,
"step": 285
},
{
"epoch": 0.09872281670693821,
"grad_norm": 2.848585844039917,
"learning_rate": 2e-05,
"loss": 0.861,
"step": 286
},
{
"epoch": 0.0990680013807387,
"grad_norm": 3.069856882095337,
"learning_rate": 2e-05,
"loss": 0.9373,
"step": 287
},
{
"epoch": 0.09941318605453918,
"grad_norm": 3.0742900371551514,
"learning_rate": 2e-05,
"loss": 0.907,
"step": 288
},
{
"epoch": 0.09975837072833967,
"grad_norm": 2.8606960773468018,
"learning_rate": 2e-05,
"loss": 0.8644,
"step": 289
},
{
"epoch": 0.10010355540214015,
"grad_norm": 2.9489126205444336,
"learning_rate": 2e-05,
"loss": 0.8922,
"step": 290
},
{
"epoch": 0.10044874007594062,
"grad_norm": 3.0227949619293213,
"learning_rate": 2e-05,
"loss": 0.9255,
"step": 291
},
{
"epoch": 0.10079392474974111,
"grad_norm": 2.758237361907959,
"learning_rate": 2e-05,
"loss": 0.8339,
"step": 292
},
{
"epoch": 0.1011391094235416,
"grad_norm": 2.8767168521881104,
"learning_rate": 2e-05,
"loss": 0.8697,
"step": 293
},
{
"epoch": 0.10148429409734208,
"grad_norm": 2.8063676357269287,
"learning_rate": 2e-05,
"loss": 0.8628,
"step": 294
},
{
"epoch": 0.10182947877114257,
"grad_norm": 2.9139151573181152,
"learning_rate": 2e-05,
"loss": 0.9157,
"step": 295
},
{
"epoch": 0.10217466344494304,
"grad_norm": 2.857400894165039,
"learning_rate": 2e-05,
"loss": 0.892,
"step": 296
},
{
"epoch": 0.10251984811874353,
"grad_norm": 2.8675410747528076,
"learning_rate": 2e-05,
"loss": 0.8438,
"step": 297
},
{
"epoch": 0.10286503279254401,
"grad_norm": 2.8956315517425537,
"learning_rate": 2e-05,
"loss": 0.847,
"step": 298
},
{
"epoch": 0.1032102174663445,
"grad_norm": 2.743086099624634,
"learning_rate": 2e-05,
"loss": 0.9041,
"step": 299
},
{
"epoch": 0.10355540214014498,
"grad_norm": 3.0366575717926025,
"learning_rate": 2e-05,
"loss": 0.8704,
"step": 300
},
{
"epoch": 0.10390058681394546,
"grad_norm": 3.290945291519165,
"learning_rate": 2e-05,
"loss": 0.7993,
"step": 301
},
{
"epoch": 0.10424577148774594,
"grad_norm": 2.562634229660034,
"learning_rate": 2e-05,
"loss": 0.7964,
"step": 302
},
{
"epoch": 0.10459095616154643,
"grad_norm": 3.005824089050293,
"learning_rate": 2e-05,
"loss": 0.8158,
"step": 303
},
{
"epoch": 0.10493614083534691,
"grad_norm": 2.912410259246826,
"learning_rate": 2e-05,
"loss": 0.8584,
"step": 304
},
{
"epoch": 0.1052813255091474,
"grad_norm": 2.7015678882598877,
"learning_rate": 2e-05,
"loss": 0.8731,
"step": 305
},
{
"epoch": 0.10562651018294787,
"grad_norm": 2.857595682144165,
"learning_rate": 2e-05,
"loss": 0.8809,
"step": 306
},
{
"epoch": 0.10597169485674836,
"grad_norm": 2.7211265563964844,
"learning_rate": 2e-05,
"loss": 0.9234,
"step": 307
},
{
"epoch": 0.10631687953054884,
"grad_norm": 3.018676519393921,
"learning_rate": 2e-05,
"loss": 0.9101,
"step": 308
},
{
"epoch": 0.10666206420434933,
"grad_norm": 2.8703527450561523,
"learning_rate": 2e-05,
"loss": 0.852,
"step": 309
},
{
"epoch": 0.10700724887814982,
"grad_norm": 2.912076473236084,
"learning_rate": 2e-05,
"loss": 0.8749,
"step": 310
},
{
"epoch": 0.10735243355195029,
"grad_norm": 2.652156352996826,
"learning_rate": 2e-05,
"loss": 0.8838,
"step": 311
},
{
"epoch": 0.10769761822575077,
"grad_norm": 2.8902244567871094,
"learning_rate": 2e-05,
"loss": 0.8898,
"step": 312
},
{
"epoch": 0.10804280289955126,
"grad_norm": 2.7108359336853027,
"learning_rate": 2e-05,
"loss": 0.8312,
"step": 313
},
{
"epoch": 0.10838798757335175,
"grad_norm": 2.6468076705932617,
"learning_rate": 2e-05,
"loss": 0.7757,
"step": 314
},
{
"epoch": 0.10873317224715223,
"grad_norm": 2.969503402709961,
"learning_rate": 2e-05,
"loss": 0.8498,
"step": 315
},
{
"epoch": 0.1090783569209527,
"grad_norm": 2.7394843101501465,
"learning_rate": 2e-05,
"loss": 0.9067,
"step": 316
},
{
"epoch": 0.10942354159475319,
"grad_norm": 2.8321640491485596,
"learning_rate": 2e-05,
"loss": 0.8693,
"step": 317
},
{
"epoch": 0.10976872626855368,
"grad_norm": 3.0017828941345215,
"learning_rate": 2e-05,
"loss": 0.8555,
"step": 318
},
{
"epoch": 0.11011391094235416,
"grad_norm": 2.7267954349517822,
"learning_rate": 2e-05,
"loss": 0.8947,
"step": 319
},
{
"epoch": 0.11045909561615465,
"grad_norm": 2.9304072856903076,
"learning_rate": 2e-05,
"loss": 0.9042,
"step": 320
},
{
"epoch": 0.11080428028995512,
"grad_norm": 2.9202630519866943,
"learning_rate": 2e-05,
"loss": 0.8426,
"step": 321
},
{
"epoch": 0.1111494649637556,
"grad_norm": 3.0602047443389893,
"learning_rate": 2e-05,
"loss": 0.8573,
"step": 322
},
{
"epoch": 0.11149464963755609,
"grad_norm": 2.987820863723755,
"learning_rate": 2e-05,
"loss": 0.8695,
"step": 323
},
{
"epoch": 0.11183983431135658,
"grad_norm": 3.158129930496216,
"learning_rate": 2e-05,
"loss": 0.9306,
"step": 324
},
{
"epoch": 0.11218501898515706,
"grad_norm": 2.893429756164551,
"learning_rate": 2e-05,
"loss": 0.9157,
"step": 325
},
{
"epoch": 0.11253020365895754,
"grad_norm": 3.009833335876465,
"learning_rate": 2e-05,
"loss": 0.9162,
"step": 326
},
{
"epoch": 0.11287538833275802,
"grad_norm": 2.768258571624756,
"learning_rate": 2e-05,
"loss": 0.8547,
"step": 327
},
{
"epoch": 0.11322057300655851,
"grad_norm": 3.0964066982269287,
"learning_rate": 2e-05,
"loss": 0.8735,
"step": 328
},
{
"epoch": 0.113565757680359,
"grad_norm": 2.9525063037872314,
"learning_rate": 2e-05,
"loss": 0.8907,
"step": 329
},
{
"epoch": 0.11391094235415948,
"grad_norm": 3.317748785018921,
"learning_rate": 2e-05,
"loss": 0.8673,
"step": 330
},
{
"epoch": 0.11425612702795995,
"grad_norm": 2.934115409851074,
"learning_rate": 2e-05,
"loss": 0.8477,
"step": 331
},
{
"epoch": 0.11460131170176044,
"grad_norm": 2.8080825805664062,
"learning_rate": 2e-05,
"loss": 0.9336,
"step": 332
},
{
"epoch": 0.11494649637556092,
"grad_norm": 2.938957929611206,
"learning_rate": 2e-05,
"loss": 0.8972,
"step": 333
},
{
"epoch": 0.11529168104936141,
"grad_norm": 2.778604745864868,
"learning_rate": 2e-05,
"loss": 0.8013,
"step": 334
},
{
"epoch": 0.1156368657231619,
"grad_norm": 2.779245376586914,
"learning_rate": 2e-05,
"loss": 0.9065,
"step": 335
},
{
"epoch": 0.11598205039696237,
"grad_norm": 2.834568738937378,
"learning_rate": 2e-05,
"loss": 0.8904,
"step": 336
},
{
"epoch": 0.11632723507076285,
"grad_norm": 2.7382054328918457,
"learning_rate": 2e-05,
"loss": 0.8932,
"step": 337
},
{
"epoch": 0.11667241974456334,
"grad_norm": 2.746325731277466,
"learning_rate": 2e-05,
"loss": 0.8365,
"step": 338
},
{
"epoch": 0.11701760441836383,
"grad_norm": 2.750140905380249,
"learning_rate": 2e-05,
"loss": 0.8255,
"step": 339
},
{
"epoch": 0.11736278909216431,
"grad_norm": 3.3976187705993652,
"learning_rate": 2e-05,
"loss": 0.859,
"step": 340
},
{
"epoch": 0.11770797376596479,
"grad_norm": 3.004563093185425,
"learning_rate": 2e-05,
"loss": 0.8895,
"step": 341
},
{
"epoch": 0.11805315843976527,
"grad_norm": 3.343090295791626,
"learning_rate": 2e-05,
"loss": 0.8644,
"step": 342
},
{
"epoch": 0.11839834311356576,
"grad_norm": 2.7037289142608643,
"learning_rate": 2e-05,
"loss": 0.834,
"step": 343
},
{
"epoch": 0.11874352778736624,
"grad_norm": 2.881302833557129,
"learning_rate": 2e-05,
"loss": 0.8343,
"step": 344
},
{
"epoch": 0.11908871246116673,
"grad_norm": 2.9096665382385254,
"learning_rate": 2e-05,
"loss": 0.8559,
"step": 345
},
{
"epoch": 0.1194338971349672,
"grad_norm": 3.109509229660034,
"learning_rate": 2e-05,
"loss": 0.9016,
"step": 346
},
{
"epoch": 0.11977908180876769,
"grad_norm": 2.9444544315338135,
"learning_rate": 2e-05,
"loss": 0.8391,
"step": 347
},
{
"epoch": 0.12012426648256817,
"grad_norm": 3.122678518295288,
"learning_rate": 2e-05,
"loss": 0.8868,
"step": 348
},
{
"epoch": 0.12046945115636866,
"grad_norm": 2.646322250366211,
"learning_rate": 2e-05,
"loss": 0.8174,
"step": 349
},
{
"epoch": 0.12081463583016915,
"grad_norm": 2.9330692291259766,
"learning_rate": 2e-05,
"loss": 0.8805,
"step": 350
},
{
"epoch": 0.12115982050396962,
"grad_norm": 2.9383981227874756,
"learning_rate": 2e-05,
"loss": 0.8618,
"step": 351
},
{
"epoch": 0.1215050051777701,
"grad_norm": 2.9238908290863037,
"learning_rate": 2e-05,
"loss": 0.9407,
"step": 352
},
{
"epoch": 0.12185018985157059,
"grad_norm": 2.8855233192443848,
"learning_rate": 2e-05,
"loss": 0.8713,
"step": 353
},
{
"epoch": 0.12219537452537108,
"grad_norm": 2.822115182876587,
"learning_rate": 2e-05,
"loss": 0.8916,
"step": 354
},
{
"epoch": 0.12254055919917156,
"grad_norm": 2.8500313758850098,
"learning_rate": 2e-05,
"loss": 0.8349,
"step": 355
},
{
"epoch": 0.12288574387297203,
"grad_norm": 3.022473096847534,
"learning_rate": 2e-05,
"loss": 0.8479,
"step": 356
},
{
"epoch": 0.12323092854677252,
"grad_norm": 2.7787365913391113,
"learning_rate": 2e-05,
"loss": 0.8833,
"step": 357
},
{
"epoch": 0.123576113220573,
"grad_norm": 2.9101765155792236,
"learning_rate": 2e-05,
"loss": 0.888,
"step": 358
},
{
"epoch": 0.12392129789437349,
"grad_norm": 2.7252466678619385,
"learning_rate": 2e-05,
"loss": 0.8657,
"step": 359
},
{
"epoch": 0.12426648256817398,
"grad_norm": 2.8830056190490723,
"learning_rate": 2e-05,
"loss": 0.8259,
"step": 360
},
{
"epoch": 0.12461166724197445,
"grad_norm": 2.671375274658203,
"learning_rate": 2e-05,
"loss": 0.8253,
"step": 361
},
{
"epoch": 0.12495685191577494,
"grad_norm": 2.706015110015869,
"learning_rate": 2e-05,
"loss": 0.838,
"step": 362
},
{
"epoch": 0.12530203658957542,
"grad_norm": 2.616062641143799,
"learning_rate": 2e-05,
"loss": 0.8607,
"step": 363
},
{
"epoch": 0.1256472212633759,
"grad_norm": 3.385586738586426,
"learning_rate": 2e-05,
"loss": 0.8603,
"step": 364
},
{
"epoch": 0.1259924059371764,
"grad_norm": 2.9748425483703613,
"learning_rate": 2e-05,
"loss": 0.8802,
"step": 365
},
{
"epoch": 0.12633759061097688,
"grad_norm": 3.009232759475708,
"learning_rate": 2e-05,
"loss": 0.8572,
"step": 366
},
{
"epoch": 0.12668277528477737,
"grad_norm": 2.995032548904419,
"learning_rate": 2e-05,
"loss": 0.8542,
"step": 367
},
{
"epoch": 0.12702795995857785,
"grad_norm": 3.1191444396972656,
"learning_rate": 2e-05,
"loss": 0.9096,
"step": 368
},
{
"epoch": 0.1273731446323783,
"grad_norm": 3.3370883464813232,
"learning_rate": 2e-05,
"loss": 0.8456,
"step": 369
},
{
"epoch": 0.1277183293061788,
"grad_norm": 2.910351276397705,
"learning_rate": 2e-05,
"loss": 0.8415,
"step": 370
},
{
"epoch": 0.12806351397997928,
"grad_norm": 2.381078004837036,
"learning_rate": 2e-05,
"loss": 0.7864,
"step": 371
},
{
"epoch": 0.12840869865377977,
"grad_norm": 2.8764638900756836,
"learning_rate": 2e-05,
"loss": 0.8731,
"step": 372
},
{
"epoch": 0.12875388332758025,
"grad_norm": 2.9746713638305664,
"learning_rate": 2e-05,
"loss": 0.8447,
"step": 373
},
{
"epoch": 0.12909906800138074,
"grad_norm": 2.7619199752807617,
"learning_rate": 2e-05,
"loss": 0.8315,
"step": 374
},
{
"epoch": 0.12944425267518123,
"grad_norm": 2.5940611362457275,
"learning_rate": 2e-05,
"loss": 0.8749,
"step": 375
},
{
"epoch": 0.1297894373489817,
"grad_norm": 2.9887807369232178,
"learning_rate": 2e-05,
"loss": 0.9069,
"step": 376
},
{
"epoch": 0.1301346220227822,
"grad_norm": 3.1408069133758545,
"learning_rate": 2e-05,
"loss": 0.9029,
"step": 377
},
{
"epoch": 0.13047980669658268,
"grad_norm": 2.876115083694458,
"learning_rate": 2e-05,
"loss": 0.9135,
"step": 378
},
{
"epoch": 0.13082499137038314,
"grad_norm": 2.7395379543304443,
"learning_rate": 2e-05,
"loss": 0.8597,
"step": 379
},
{
"epoch": 0.13117017604418363,
"grad_norm": 3.036904811859131,
"learning_rate": 2e-05,
"loss": 0.8886,
"step": 380
},
{
"epoch": 0.13151536071798411,
"grad_norm": 3.047651767730713,
"learning_rate": 2e-05,
"loss": 0.8923,
"step": 381
},
{
"epoch": 0.1318605453917846,
"grad_norm": 2.8202462196350098,
"learning_rate": 2e-05,
"loss": 0.8434,
"step": 382
},
{
"epoch": 0.1322057300655851,
"grad_norm": 2.605100393295288,
"learning_rate": 2e-05,
"loss": 0.8975,
"step": 383
},
{
"epoch": 0.13255091473938557,
"grad_norm": 2.781841516494751,
"learning_rate": 2e-05,
"loss": 0.809,
"step": 384
},
{
"epoch": 0.13289609941318606,
"grad_norm": 2.6456520557403564,
"learning_rate": 2e-05,
"loss": 0.8827,
"step": 385
},
{
"epoch": 0.13324128408698654,
"grad_norm": 10.660346984863281,
"learning_rate": 2e-05,
"loss": 0.8615,
"step": 386
},
{
"epoch": 0.13358646876078703,
"grad_norm": 3.2193210124969482,
"learning_rate": 2e-05,
"loss": 0.8553,
"step": 387
},
{
"epoch": 0.13393165343458752,
"grad_norm": 2.9482877254486084,
"learning_rate": 2e-05,
"loss": 0.7931,
"step": 388
},
{
"epoch": 0.13427683810838797,
"grad_norm": 2.851332664489746,
"learning_rate": 2e-05,
"loss": 0.8291,
"step": 389
},
{
"epoch": 0.13462202278218846,
"grad_norm": 3.519622802734375,
"learning_rate": 2e-05,
"loss": 0.8803,
"step": 390
},
{
"epoch": 0.13496720745598895,
"grad_norm": 2.979022741317749,
"learning_rate": 2e-05,
"loss": 0.9023,
"step": 391
},
{
"epoch": 0.13531239212978943,
"grad_norm": 2.8534867763519287,
"learning_rate": 2e-05,
"loss": 0.8652,
"step": 392
},
{
"epoch": 0.13565757680358992,
"grad_norm": 2.7661948204040527,
"learning_rate": 2e-05,
"loss": 0.8136,
"step": 393
},
{
"epoch": 0.1360027614773904,
"grad_norm": 2.6558003425598145,
"learning_rate": 2e-05,
"loss": 0.8805,
"step": 394
},
{
"epoch": 0.1363479461511909,
"grad_norm": 2.9477477073669434,
"learning_rate": 2e-05,
"loss": 0.8341,
"step": 395
},
{
"epoch": 0.13669313082499138,
"grad_norm": 2.777376174926758,
"learning_rate": 2e-05,
"loss": 0.8435,
"step": 396
},
{
"epoch": 0.13703831549879186,
"grad_norm": 2.947483777999878,
"learning_rate": 2e-05,
"loss": 0.8495,
"step": 397
},
{
"epoch": 0.13738350017259235,
"grad_norm": 2.923473834991455,
"learning_rate": 2e-05,
"loss": 0.8357,
"step": 398
},
{
"epoch": 0.1377286848463928,
"grad_norm": 2.5483436584472656,
"learning_rate": 2e-05,
"loss": 0.7862,
"step": 399
},
{
"epoch": 0.1380738695201933,
"grad_norm": 2.5327706336975098,
"learning_rate": 2e-05,
"loss": 0.8527,
"step": 400
},
{
"epoch": 0.13841905419399378,
"grad_norm": 2.684678792953491,
"learning_rate": 2e-05,
"loss": 0.8465,
"step": 401
},
{
"epoch": 0.13876423886779427,
"grad_norm": 2.74104905128479,
"learning_rate": 2e-05,
"loss": 0.8284,
"step": 402
},
{
"epoch": 0.13910942354159475,
"grad_norm": 3.0023629665374756,
"learning_rate": 2e-05,
"loss": 0.8337,
"step": 403
},
{
"epoch": 0.13945460821539524,
"grad_norm": 2.791071891784668,
"learning_rate": 2e-05,
"loss": 0.7955,
"step": 404
},
{
"epoch": 0.13979979288919572,
"grad_norm": 3.101384401321411,
"learning_rate": 2e-05,
"loss": 0.866,
"step": 405
},
{
"epoch": 0.1401449775629962,
"grad_norm": 2.731903553009033,
"learning_rate": 2e-05,
"loss": 0.8317,
"step": 406
},
{
"epoch": 0.1404901622367967,
"grad_norm": 2.892133951187134,
"learning_rate": 2e-05,
"loss": 0.8903,
"step": 407
},
{
"epoch": 0.14083534691059718,
"grad_norm": 2.6382410526275635,
"learning_rate": 2e-05,
"loss": 0.8085,
"step": 408
},
{
"epoch": 0.14118053158439764,
"grad_norm": 2.5040509700775146,
"learning_rate": 2e-05,
"loss": 0.7556,
"step": 409
},
{
"epoch": 0.14152571625819813,
"grad_norm": 2.659632921218872,
"learning_rate": 2e-05,
"loss": 0.8627,
"step": 410
},
{
"epoch": 0.1418709009319986,
"grad_norm": 2.9481699466705322,
"learning_rate": 2e-05,
"loss": 0.769,
"step": 411
},
{
"epoch": 0.1422160856057991,
"grad_norm": 2.7535393238067627,
"learning_rate": 2e-05,
"loss": 0.8624,
"step": 412
},
{
"epoch": 0.14256127027959958,
"grad_norm": 2.8887505531311035,
"learning_rate": 2e-05,
"loss": 0.8746,
"step": 413
},
{
"epoch": 0.14290645495340007,
"grad_norm": 3.1027655601501465,
"learning_rate": 2e-05,
"loss": 0.8201,
"step": 414
},
{
"epoch": 0.14325163962720056,
"grad_norm": 2.7258031368255615,
"learning_rate": 2e-05,
"loss": 0.8758,
"step": 415
},
{
"epoch": 0.14359682430100104,
"grad_norm": 3.6109793186187744,
"learning_rate": 2e-05,
"loss": 0.8368,
"step": 416
},
{
"epoch": 0.14394200897480153,
"grad_norm": 3.0309035778045654,
"learning_rate": 2e-05,
"loss": 0.8502,
"step": 417
},
{
"epoch": 0.144287193648602,
"grad_norm": 2.7749741077423096,
"learning_rate": 2e-05,
"loss": 0.8239,
"step": 418
},
{
"epoch": 0.14463237832240247,
"grad_norm": 3.057831287384033,
"learning_rate": 2e-05,
"loss": 0.8649,
"step": 419
},
{
"epoch": 0.14497756299620296,
"grad_norm": 2.7731754779815674,
"learning_rate": 2e-05,
"loss": 0.8073,
"step": 420
},
{
"epoch": 0.14532274767000344,
"grad_norm": 2.7278342247009277,
"learning_rate": 2e-05,
"loss": 0.8367,
"step": 421
},
{
"epoch": 0.14566793234380393,
"grad_norm": 2.941387891769409,
"learning_rate": 2e-05,
"loss": 0.8205,
"step": 422
},
{
"epoch": 0.14601311701760442,
"grad_norm": 2.7232892513275146,
"learning_rate": 2e-05,
"loss": 0.8872,
"step": 423
},
{
"epoch": 0.1463583016914049,
"grad_norm": 2.953651189804077,
"learning_rate": 2e-05,
"loss": 0.8523,
"step": 424
},
{
"epoch": 0.1467034863652054,
"grad_norm": 2.7981178760528564,
"learning_rate": 2e-05,
"loss": 0.8528,
"step": 425
},
{
"epoch": 0.14704867103900587,
"grad_norm": 2.7209296226501465,
"learning_rate": 2e-05,
"loss": 0.8183,
"step": 426
},
{
"epoch": 0.14739385571280636,
"grad_norm": 2.716719627380371,
"learning_rate": 2e-05,
"loss": 0.8502,
"step": 427
},
{
"epoch": 0.14773904038660685,
"grad_norm": 2.8264238834381104,
"learning_rate": 2e-05,
"loss": 0.8225,
"step": 428
},
{
"epoch": 0.1480842250604073,
"grad_norm": 2.665504217147827,
"learning_rate": 2e-05,
"loss": 0.7785,
"step": 429
},
{
"epoch": 0.1484294097342078,
"grad_norm": 2.574075698852539,
"learning_rate": 2e-05,
"loss": 0.8745,
"step": 430
},
{
"epoch": 0.14877459440800828,
"grad_norm": 2.8665859699249268,
"learning_rate": 2e-05,
"loss": 0.8567,
"step": 431
},
{
"epoch": 0.14911977908180876,
"grad_norm": 2.5270538330078125,
"learning_rate": 2e-05,
"loss": 0.8125,
"step": 432
},
{
"epoch": 0.14946496375560925,
"grad_norm": 2.569016933441162,
"learning_rate": 2e-05,
"loss": 0.8516,
"step": 433
},
{
"epoch": 0.14981014842940973,
"grad_norm": 2.6712615489959717,
"learning_rate": 2e-05,
"loss": 0.8079,
"step": 434
},
{
"epoch": 0.15015533310321022,
"grad_norm": 2.6932435035705566,
"learning_rate": 2e-05,
"loss": 0.818,
"step": 435
},
{
"epoch": 0.1505005177770107,
"grad_norm": 2.5300912857055664,
"learning_rate": 2e-05,
"loss": 0.867,
"step": 436
},
{
"epoch": 0.1508457024508112,
"grad_norm": 2.4744701385498047,
"learning_rate": 2e-05,
"loss": 0.863,
"step": 437
},
{
"epoch": 0.15119088712461168,
"grad_norm": 4.339775562286377,
"learning_rate": 2e-05,
"loss": 0.8626,
"step": 438
},
{
"epoch": 0.15153607179841216,
"grad_norm": 2.5476183891296387,
"learning_rate": 2e-05,
"loss": 0.8193,
"step": 439
},
{
"epoch": 0.15188125647221262,
"grad_norm": 2.7269837856292725,
"learning_rate": 2e-05,
"loss": 0.8251,
"step": 440
},
{
"epoch": 0.1522264411460131,
"grad_norm": 2.843550205230713,
"learning_rate": 2e-05,
"loss": 0.8724,
"step": 441
},
{
"epoch": 0.1525716258198136,
"grad_norm": 5.800281524658203,
"learning_rate": 2e-05,
"loss": 0.7982,
"step": 442
},
{
"epoch": 0.15291681049361408,
"grad_norm": 2.7126641273498535,
"learning_rate": 2e-05,
"loss": 0.7775,
"step": 443
},
{
"epoch": 0.15326199516741457,
"grad_norm": 2.8936424255371094,
"learning_rate": 2e-05,
"loss": 0.8628,
"step": 444
},
{
"epoch": 0.15360717984121505,
"grad_norm": 2.403440475463867,
"learning_rate": 2e-05,
"loss": 0.8367,
"step": 445
},
{
"epoch": 0.15395236451501554,
"grad_norm": 2.6871049404144287,
"learning_rate": 2e-05,
"loss": 0.7947,
"step": 446
},
{
"epoch": 0.15429754918881602,
"grad_norm": 2.6108877658843994,
"learning_rate": 2e-05,
"loss": 0.868,
"step": 447
},
{
"epoch": 0.1546427338626165,
"grad_norm": 2.7435109615325928,
"learning_rate": 2e-05,
"loss": 0.8498,
"step": 448
},
{
"epoch": 0.154987918536417,
"grad_norm": 2.5271124839782715,
"learning_rate": 2e-05,
"loss": 0.8271,
"step": 449
},
{
"epoch": 0.15533310321021745,
"grad_norm": 2.607501268386841,
"learning_rate": 2e-05,
"loss": 0.8588,
"step": 450
},
{
"epoch": 0.15567828788401794,
"grad_norm": 2.610903024673462,
"learning_rate": 2e-05,
"loss": 0.8411,
"step": 451
},
{
"epoch": 0.15602347255781843,
"grad_norm": 2.6607000827789307,
"learning_rate": 2e-05,
"loss": 0.8085,
"step": 452
},
{
"epoch": 0.1563686572316189,
"grad_norm": 2.6461052894592285,
"learning_rate": 2e-05,
"loss": 0.8041,
"step": 453
},
{
"epoch": 0.1567138419054194,
"grad_norm": 2.7526638507843018,
"learning_rate": 2e-05,
"loss": 0.8848,
"step": 454
},
{
"epoch": 0.15705902657921988,
"grad_norm": 2.546306610107422,
"learning_rate": 2e-05,
"loss": 0.8602,
"step": 455
},
{
"epoch": 0.15740421125302037,
"grad_norm": 2.6657474040985107,
"learning_rate": 2e-05,
"loss": 0.8307,
"step": 456
},
{
"epoch": 0.15774939592682086,
"grad_norm": 2.6293036937713623,
"learning_rate": 2e-05,
"loss": 0.8155,
"step": 457
},
{
"epoch": 0.15809458060062134,
"grad_norm": 3.03344988822937,
"learning_rate": 2e-05,
"loss": 0.819,
"step": 458
},
{
"epoch": 0.15843976527442183,
"grad_norm": 2.7682418823242188,
"learning_rate": 2e-05,
"loss": 0.8674,
"step": 459
},
{
"epoch": 0.1587849499482223,
"grad_norm": 2.441422462463379,
"learning_rate": 2e-05,
"loss": 0.7109,
"step": 460
},
{
"epoch": 0.15913013462202277,
"grad_norm": 3.0306053161621094,
"learning_rate": 2e-05,
"loss": 0.8352,
"step": 461
},
{
"epoch": 0.15947531929582326,
"grad_norm": 3.1089224815368652,
"learning_rate": 2e-05,
"loss": 0.9059,
"step": 462
},
{
"epoch": 0.15982050396962375,
"grad_norm": 2.5796120166778564,
"learning_rate": 2e-05,
"loss": 0.7726,
"step": 463
},
{
"epoch": 0.16016568864342423,
"grad_norm": 2.7144784927368164,
"learning_rate": 2e-05,
"loss": 0.7258,
"step": 464
},
{
"epoch": 0.16051087331722472,
"grad_norm": 2.6572415828704834,
"learning_rate": 2e-05,
"loss": 0.8354,
"step": 465
},
{
"epoch": 0.1608560579910252,
"grad_norm": 2.556910753250122,
"learning_rate": 2e-05,
"loss": 0.8727,
"step": 466
},
{
"epoch": 0.1612012426648257,
"grad_norm": 2.621105194091797,
"learning_rate": 2e-05,
"loss": 0.8326,
"step": 467
},
{
"epoch": 0.16154642733862618,
"grad_norm": 2.4325952529907227,
"learning_rate": 2e-05,
"loss": 0.8102,
"step": 468
},
{
"epoch": 0.16189161201242666,
"grad_norm": 2.616182327270508,
"learning_rate": 2e-05,
"loss": 0.8983,
"step": 469
},
{
"epoch": 0.16223679668622712,
"grad_norm": 2.7066891193389893,
"learning_rate": 2e-05,
"loss": 0.8761,
"step": 470
},
{
"epoch": 0.1625819813600276,
"grad_norm": 2.4946465492248535,
"learning_rate": 2e-05,
"loss": 0.8051,
"step": 471
},
{
"epoch": 0.1629271660338281,
"grad_norm": 2.7673726081848145,
"learning_rate": 2e-05,
"loss": 0.8436,
"step": 472
},
{
"epoch": 0.16327235070762858,
"grad_norm": 2.8104934692382812,
"learning_rate": 2e-05,
"loss": 0.83,
"step": 473
},
{
"epoch": 0.16361753538142906,
"grad_norm": 2.5568485260009766,
"learning_rate": 2e-05,
"loss": 0.8144,
"step": 474
},
{
"epoch": 0.16396272005522955,
"grad_norm": 2.5758121013641357,
"learning_rate": 2e-05,
"loss": 0.838,
"step": 475
},
{
"epoch": 0.16430790472903004,
"grad_norm": 2.492135524749756,
"learning_rate": 2e-05,
"loss": 0.8959,
"step": 476
},
{
"epoch": 0.16465308940283052,
"grad_norm": 2.6613054275512695,
"learning_rate": 2e-05,
"loss": 0.8152,
"step": 477
},
{
"epoch": 0.164998274076631,
"grad_norm": 2.75256609916687,
"learning_rate": 2e-05,
"loss": 0.8267,
"step": 478
},
{
"epoch": 0.1653434587504315,
"grad_norm": 2.8076133728027344,
"learning_rate": 2e-05,
"loss": 0.7922,
"step": 479
},
{
"epoch": 0.16568864342423195,
"grad_norm": 2.722338914871216,
"learning_rate": 2e-05,
"loss": 0.8598,
"step": 480
},
{
"epoch": 0.16603382809803244,
"grad_norm": 2.5735299587249756,
"learning_rate": 2e-05,
"loss": 0.7875,
"step": 481
},
{
"epoch": 0.16637901277183292,
"grad_norm": 2.702031373977661,
"learning_rate": 2e-05,
"loss": 0.8604,
"step": 482
},
{
"epoch": 0.1667241974456334,
"grad_norm": 2.734985589981079,
"learning_rate": 2e-05,
"loss": 0.851,
"step": 483
},
{
"epoch": 0.1670693821194339,
"grad_norm": 2.9320790767669678,
"learning_rate": 2e-05,
"loss": 0.8655,
"step": 484
},
{
"epoch": 0.16741456679323438,
"grad_norm": 2.8295514583587646,
"learning_rate": 2e-05,
"loss": 0.8963,
"step": 485
},
{
"epoch": 0.16775975146703487,
"grad_norm": 2.4725968837738037,
"learning_rate": 2e-05,
"loss": 0.8372,
"step": 486
},
{
"epoch": 0.16810493614083535,
"grad_norm": 2.6364479064941406,
"learning_rate": 2e-05,
"loss": 0.8244,
"step": 487
},
{
"epoch": 0.16845012081463584,
"grad_norm": 3.3929388523101807,
"learning_rate": 2e-05,
"loss": 0.8379,
"step": 488
},
{
"epoch": 0.16879530548843633,
"grad_norm": 2.701646566390991,
"learning_rate": 2e-05,
"loss": 0.8196,
"step": 489
},
{
"epoch": 0.16914049016223678,
"grad_norm": 2.559340476989746,
"learning_rate": 2e-05,
"loss": 0.7844,
"step": 490
},
{
"epoch": 0.16948567483603727,
"grad_norm": 2.644239664077759,
"learning_rate": 2e-05,
"loss": 0.8601,
"step": 491
},
{
"epoch": 0.16983085950983776,
"grad_norm": 2.5983693599700928,
"learning_rate": 2e-05,
"loss": 0.8278,
"step": 492
},
{
"epoch": 0.17017604418363824,
"grad_norm": 2.650519609451294,
"learning_rate": 2e-05,
"loss": 0.8763,
"step": 493
},
{
"epoch": 0.17052122885743873,
"grad_norm": 2.715981960296631,
"learning_rate": 2e-05,
"loss": 0.7283,
"step": 494
},
{
"epoch": 0.17086641353123921,
"grad_norm": 3.0167715549468994,
"learning_rate": 2e-05,
"loss": 0.8507,
"step": 495
},
{
"epoch": 0.1712115982050397,
"grad_norm": 2.9270315170288086,
"learning_rate": 2e-05,
"loss": 0.8593,
"step": 496
},
{
"epoch": 0.17155678287884019,
"grad_norm": 2.5383360385894775,
"learning_rate": 2e-05,
"loss": 0.7877,
"step": 497
},
{
"epoch": 0.17190196755264067,
"grad_norm": 2.868652582168579,
"learning_rate": 2e-05,
"loss": 0.8635,
"step": 498
},
{
"epoch": 0.17224715222644116,
"grad_norm": 2.8719124794006348,
"learning_rate": 2e-05,
"loss": 0.795,
"step": 499
},
{
"epoch": 0.17259233690024162,
"grad_norm": 2.806877613067627,
"learning_rate": 2e-05,
"loss": 0.7998,
"step": 500
},
{
"epoch": 0.1729375215740421,
"grad_norm": 2.63967227935791,
"learning_rate": 2e-05,
"loss": 0.8817,
"step": 501
},
{
"epoch": 0.1732827062478426,
"grad_norm": 2.68904185295105,
"learning_rate": 2e-05,
"loss": 0.7909,
"step": 502
},
{
"epoch": 0.17362789092164307,
"grad_norm": 2.5513813495635986,
"learning_rate": 2e-05,
"loss": 0.8033,
"step": 503
},
{
"epoch": 0.17397307559544356,
"grad_norm": 2.581880569458008,
"learning_rate": 2e-05,
"loss": 0.8586,
"step": 504
},
{
"epoch": 0.17431826026924405,
"grad_norm": 2.901590585708618,
"learning_rate": 2e-05,
"loss": 0.8116,
"step": 505
},
{
"epoch": 0.17466344494304453,
"grad_norm": 2.7014777660369873,
"learning_rate": 2e-05,
"loss": 0.8502,
"step": 506
},
{
"epoch": 0.17500862961684502,
"grad_norm": 2.875823497772217,
"learning_rate": 2e-05,
"loss": 0.7739,
"step": 507
},
{
"epoch": 0.1753538142906455,
"grad_norm": 2.4771320819854736,
"learning_rate": 2e-05,
"loss": 0.8,
"step": 508
},
{
"epoch": 0.175698998964446,
"grad_norm": 2.648470163345337,
"learning_rate": 2e-05,
"loss": 0.8632,
"step": 509
},
{
"epoch": 0.17604418363824645,
"grad_norm": 2.8525655269622803,
"learning_rate": 2e-05,
"loss": 0.8639,
"step": 510
},
{
"epoch": 0.17638936831204693,
"grad_norm": 2.7430825233459473,
"learning_rate": 2e-05,
"loss": 0.9099,
"step": 511
},
{
"epoch": 0.17673455298584742,
"grad_norm": 2.9599390029907227,
"learning_rate": 2e-05,
"loss": 0.9326,
"step": 512
},
{
"epoch": 0.1770797376596479,
"grad_norm": 2.705155611038208,
"learning_rate": 2e-05,
"loss": 0.8808,
"step": 513
},
{
"epoch": 0.1774249223334484,
"grad_norm": 2.507206678390503,
"learning_rate": 2e-05,
"loss": 0.8614,
"step": 514
},
{
"epoch": 0.17777010700724888,
"grad_norm": 2.3883049488067627,
"learning_rate": 2e-05,
"loss": 0.8161,
"step": 515
},
{
"epoch": 0.17811529168104936,
"grad_norm": 2.7002079486846924,
"learning_rate": 2e-05,
"loss": 0.9636,
"step": 516
},
{
"epoch": 0.17846047635484985,
"grad_norm": 2.4914982318878174,
"learning_rate": 2e-05,
"loss": 0.8232,
"step": 517
},
{
"epoch": 0.17880566102865034,
"grad_norm": 2.644822835922241,
"learning_rate": 2e-05,
"loss": 0.8306,
"step": 518
},
{
"epoch": 0.17915084570245082,
"grad_norm": 2.900297164916992,
"learning_rate": 2e-05,
"loss": 0.9079,
"step": 519
},
{
"epoch": 0.17949603037625128,
"grad_norm": 2.8263745307922363,
"learning_rate": 2e-05,
"loss": 0.8189,
"step": 520
},
{
"epoch": 0.17984121505005177,
"grad_norm": 2.6911349296569824,
"learning_rate": 2e-05,
"loss": 0.8398,
"step": 521
},
{
"epoch": 0.18018639972385225,
"grad_norm": 2.7833213806152344,
"learning_rate": 2e-05,
"loss": 0.7453,
"step": 522
},
{
"epoch": 0.18053158439765274,
"grad_norm": 2.680326223373413,
"learning_rate": 2e-05,
"loss": 0.8591,
"step": 523
},
{
"epoch": 0.18087676907145323,
"grad_norm": 2.902534246444702,
"learning_rate": 2e-05,
"loss": 0.9315,
"step": 524
},
{
"epoch": 0.1812219537452537,
"grad_norm": 2.9760384559631348,
"learning_rate": 2e-05,
"loss": 0.8882,
"step": 525
},
{
"epoch": 0.1815671384190542,
"grad_norm": 2.626789093017578,
"learning_rate": 2e-05,
"loss": 0.7857,
"step": 526
},
{
"epoch": 0.18191232309285468,
"grad_norm": 2.6375679969787598,
"learning_rate": 2e-05,
"loss": 0.8132,
"step": 527
},
{
"epoch": 0.18225750776665517,
"grad_norm": 2.9310142993927,
"learning_rate": 2e-05,
"loss": 0.907,
"step": 528
},
{
"epoch": 0.18260269244045566,
"grad_norm": 2.7192530632019043,
"learning_rate": 2e-05,
"loss": 0.8633,
"step": 529
},
{
"epoch": 0.1829478771142561,
"grad_norm": 2.753049612045288,
"learning_rate": 2e-05,
"loss": 0.8589,
"step": 530
},
{
"epoch": 0.1832930617880566,
"grad_norm": 3.2843739986419678,
"learning_rate": 2e-05,
"loss": 0.8325,
"step": 531
},
{
"epoch": 0.18363824646185709,
"grad_norm": 2.6691083908081055,
"learning_rate": 2e-05,
"loss": 0.8381,
"step": 532
},
{
"epoch": 0.18398343113565757,
"grad_norm": 3.0254318714141846,
"learning_rate": 2e-05,
"loss": 0.8816,
"step": 533
},
{
"epoch": 0.18432861580945806,
"grad_norm": 3.193315029144287,
"learning_rate": 2e-05,
"loss": 0.8958,
"step": 534
},
{
"epoch": 0.18467380048325854,
"grad_norm": 2.721979856491089,
"learning_rate": 2e-05,
"loss": 0.7731,
"step": 535
},
{
"epoch": 0.18501898515705903,
"grad_norm": 2.6035878658294678,
"learning_rate": 2e-05,
"loss": 0.797,
"step": 536
},
{
"epoch": 0.18536416983085952,
"grad_norm": 2.6113343238830566,
"learning_rate": 2e-05,
"loss": 0.7576,
"step": 537
},
{
"epoch": 0.18570935450466,
"grad_norm": 2.4984490871429443,
"learning_rate": 2e-05,
"loss": 0.7743,
"step": 538
},
{
"epoch": 0.1860545391784605,
"grad_norm": 2.4978625774383545,
"learning_rate": 2e-05,
"loss": 0.7983,
"step": 539
},
{
"epoch": 0.18639972385226097,
"grad_norm": 2.6827447414398193,
"learning_rate": 2e-05,
"loss": 0.8141,
"step": 540
},
{
"epoch": 0.18674490852606143,
"grad_norm": 2.6190788745880127,
"learning_rate": 2e-05,
"loss": 0.8467,
"step": 541
},
{
"epoch": 0.18709009319986192,
"grad_norm": 2.6752450466156006,
"learning_rate": 2e-05,
"loss": 0.8986,
"step": 542
},
{
"epoch": 0.1874352778736624,
"grad_norm": 2.7006146907806396,
"learning_rate": 2e-05,
"loss": 0.9105,
"step": 543
},
{
"epoch": 0.1877804625474629,
"grad_norm": 2.7405309677124023,
"learning_rate": 2e-05,
"loss": 0.8625,
"step": 544
},
{
"epoch": 0.18812564722126338,
"grad_norm": 3.042073965072632,
"learning_rate": 2e-05,
"loss": 0.8316,
"step": 545
},
{
"epoch": 0.18847083189506386,
"grad_norm": 2.744330406188965,
"learning_rate": 2e-05,
"loss": 0.8137,
"step": 546
},
{
"epoch": 0.18881601656886435,
"grad_norm": 2.7891201972961426,
"learning_rate": 2e-05,
"loss": 0.83,
"step": 547
},
{
"epoch": 0.18916120124266483,
"grad_norm": 3.782231330871582,
"learning_rate": 2e-05,
"loss": 0.8158,
"step": 548
},
{
"epoch": 0.18950638591646532,
"grad_norm": 3.0141708850860596,
"learning_rate": 2e-05,
"loss": 0.9087,
"step": 549
},
{
"epoch": 0.1898515705902658,
"grad_norm": 2.74953031539917,
"learning_rate": 2e-05,
"loss": 0.8793,
"step": 550
},
{
"epoch": 0.19019675526406626,
"grad_norm": 2.6540184020996094,
"learning_rate": 2e-05,
"loss": 0.9037,
"step": 551
},
{
"epoch": 0.19054193993786675,
"grad_norm": 3.051140308380127,
"learning_rate": 2e-05,
"loss": 0.7495,
"step": 552
},
{
"epoch": 0.19088712461166724,
"grad_norm": 2.5354433059692383,
"learning_rate": 2e-05,
"loss": 0.8458,
"step": 553
},
{
"epoch": 0.19123230928546772,
"grad_norm": 2.5842111110687256,
"learning_rate": 2e-05,
"loss": 0.8055,
"step": 554
},
{
"epoch": 0.1915774939592682,
"grad_norm": 2.568885326385498,
"learning_rate": 2e-05,
"loss": 0.8083,
"step": 555
},
{
"epoch": 0.1919226786330687,
"grad_norm": 3.036980152130127,
"learning_rate": 2e-05,
"loss": 0.8081,
"step": 556
},
{
"epoch": 0.19226786330686918,
"grad_norm": 2.852733850479126,
"learning_rate": 2e-05,
"loss": 0.8148,
"step": 557
},
{
"epoch": 0.19261304798066967,
"grad_norm": 2.7637667655944824,
"learning_rate": 2e-05,
"loss": 0.7862,
"step": 558
},
{
"epoch": 0.19295823265447015,
"grad_norm": 2.8737142086029053,
"learning_rate": 2e-05,
"loss": 0.8552,
"step": 559
},
{
"epoch": 0.19330341732827064,
"grad_norm": 2.655191659927368,
"learning_rate": 2e-05,
"loss": 0.8625,
"step": 560
},
{
"epoch": 0.1936486020020711,
"grad_norm": 2.6270878314971924,
"learning_rate": 2e-05,
"loss": 0.8398,
"step": 561
},
{
"epoch": 0.19399378667587158,
"grad_norm": 2.647904872894287,
"learning_rate": 2e-05,
"loss": 0.7879,
"step": 562
},
{
"epoch": 0.19433897134967207,
"grad_norm": 2.548400640487671,
"learning_rate": 2e-05,
"loss": 0.8751,
"step": 563
},
{
"epoch": 0.19468415602347255,
"grad_norm": 2.8312828540802,
"learning_rate": 2e-05,
"loss": 0.7278,
"step": 564
},
{
"epoch": 0.19502934069727304,
"grad_norm": 2.663391351699829,
"learning_rate": 2e-05,
"loss": 0.8448,
"step": 565
},
{
"epoch": 0.19537452537107353,
"grad_norm": 2.9681437015533447,
"learning_rate": 2e-05,
"loss": 0.8456,
"step": 566
},
{
"epoch": 0.195719710044874,
"grad_norm": 2.6108689308166504,
"learning_rate": 2e-05,
"loss": 0.7968,
"step": 567
},
{
"epoch": 0.1960648947186745,
"grad_norm": 2.440091609954834,
"learning_rate": 2e-05,
"loss": 0.8446,
"step": 568
},
{
"epoch": 0.19641007939247498,
"grad_norm": 2.702831268310547,
"learning_rate": 2e-05,
"loss": 0.7967,
"step": 569
},
{
"epoch": 0.19675526406627547,
"grad_norm": 2.943120002746582,
"learning_rate": 2e-05,
"loss": 0.8356,
"step": 570
},
{
"epoch": 0.19710044874007593,
"grad_norm": 3.0097904205322266,
"learning_rate": 2e-05,
"loss": 0.7726,
"step": 571
},
{
"epoch": 0.19744563341387641,
"grad_norm": 2.3800501823425293,
"learning_rate": 2e-05,
"loss": 0.8435,
"step": 572
},
{
"epoch": 0.1977908180876769,
"grad_norm": 2.6139631271362305,
"learning_rate": 2e-05,
"loss": 0.7929,
"step": 573
},
{
"epoch": 0.1981360027614774,
"grad_norm": 2.5904738903045654,
"learning_rate": 2e-05,
"loss": 0.8143,
"step": 574
},
{
"epoch": 0.19848118743527787,
"grad_norm": 2.6801297664642334,
"learning_rate": 2e-05,
"loss": 0.7782,
"step": 575
},
{
"epoch": 0.19882637210907836,
"grad_norm": 2.631114959716797,
"learning_rate": 2e-05,
"loss": 0.775,
"step": 576
},
{
"epoch": 0.19917155678287884,
"grad_norm": 2.3431198596954346,
"learning_rate": 2e-05,
"loss": 0.7601,
"step": 577
},
{
"epoch": 0.19951674145667933,
"grad_norm": 2.488933563232422,
"learning_rate": 2e-05,
"loss": 0.7926,
"step": 578
},
{
"epoch": 0.19986192613047982,
"grad_norm": 2.3383426666259766,
"learning_rate": 2e-05,
"loss": 0.7666,
"step": 579
},
{
"epoch": 0.2002071108042803,
"grad_norm": 2.5152347087860107,
"learning_rate": 2e-05,
"loss": 0.8124,
"step": 580
},
{
"epoch": 0.20055229547808076,
"grad_norm": 2.5689401626586914,
"learning_rate": 2e-05,
"loss": 0.7708,
"step": 581
},
{
"epoch": 0.20089748015188125,
"grad_norm": 2.508896827697754,
"learning_rate": 2e-05,
"loss": 0.8723,
"step": 582
},
{
"epoch": 0.20124266482568173,
"grad_norm": 2.9416592121124268,
"learning_rate": 2e-05,
"loss": 0.7583,
"step": 583
},
{
"epoch": 0.20158784949948222,
"grad_norm": 2.3808095455169678,
"learning_rate": 2e-05,
"loss": 0.7332,
"step": 584
},
{
"epoch": 0.2019330341732827,
"grad_norm": 2.675591230392456,
"learning_rate": 2e-05,
"loss": 0.8669,
"step": 585
},
{
"epoch": 0.2022782188470832,
"grad_norm": 3.0490567684173584,
"learning_rate": 2e-05,
"loss": 0.8451,
"step": 586
},
{
"epoch": 0.20262340352088368,
"grad_norm": 2.47438907623291,
"learning_rate": 2e-05,
"loss": 0.8056,
"step": 587
},
{
"epoch": 0.20296858819468416,
"grad_norm": 2.73553729057312,
"learning_rate": 2e-05,
"loss": 0.8702,
"step": 588
},
{
"epoch": 0.20331377286848465,
"grad_norm": 2.674639940261841,
"learning_rate": 2e-05,
"loss": 0.8189,
"step": 589
},
{
"epoch": 0.20365895754228514,
"grad_norm": 2.494344711303711,
"learning_rate": 2e-05,
"loss": 0.8256,
"step": 590
},
{
"epoch": 0.2040041422160856,
"grad_norm": 2.297337055206299,
"learning_rate": 2e-05,
"loss": 0.779,
"step": 591
},
{
"epoch": 0.20434932688988608,
"grad_norm": 2.652824878692627,
"learning_rate": 2e-05,
"loss": 0.8074,
"step": 592
},
{
"epoch": 0.20469451156368657,
"grad_norm": 2.507140636444092,
"learning_rate": 2e-05,
"loss": 0.8578,
"step": 593
},
{
"epoch": 0.20503969623748705,
"grad_norm": 3.012864351272583,
"learning_rate": 2e-05,
"loss": 0.7699,
"step": 594
},
{
"epoch": 0.20538488091128754,
"grad_norm": 2.511747121810913,
"learning_rate": 2e-05,
"loss": 0.8248,
"step": 595
},
{
"epoch": 0.20573006558508802,
"grad_norm": 2.4231340885162354,
"learning_rate": 2e-05,
"loss": 0.8095,
"step": 596
},
{
"epoch": 0.2060752502588885,
"grad_norm": 2.3656606674194336,
"learning_rate": 2e-05,
"loss": 0.7241,
"step": 597
},
{
"epoch": 0.206420434932689,
"grad_norm": 2.389401912689209,
"learning_rate": 2e-05,
"loss": 0.8047,
"step": 598
},
{
"epoch": 0.20676561960648948,
"grad_norm": 2.439185857772827,
"learning_rate": 2e-05,
"loss": 0.8732,
"step": 599
},
{
"epoch": 0.20711080428028997,
"grad_norm": 2.5809507369995117,
"learning_rate": 2e-05,
"loss": 0.8037,
"step": 600
},
{
"epoch": 0.20745598895409043,
"grad_norm": 2.3989405632019043,
"learning_rate": 2e-05,
"loss": 0.839,
"step": 601
},
{
"epoch": 0.2078011736278909,
"grad_norm": 2.638395071029663,
"learning_rate": 2e-05,
"loss": 0.8044,
"step": 602
},
{
"epoch": 0.2081463583016914,
"grad_norm": 2.566681385040283,
"learning_rate": 2e-05,
"loss": 0.8102,
"step": 603
},
{
"epoch": 0.20849154297549188,
"grad_norm": 2.6014740467071533,
"learning_rate": 2e-05,
"loss": 0.8381,
"step": 604
},
{
"epoch": 0.20883672764929237,
"grad_norm": 2.6535403728485107,
"learning_rate": 2e-05,
"loss": 0.7749,
"step": 605
},
{
"epoch": 0.20918191232309286,
"grad_norm": 2.539785385131836,
"learning_rate": 2e-05,
"loss": 0.8037,
"step": 606
},
{
"epoch": 0.20952709699689334,
"grad_norm": 2.783963203430176,
"learning_rate": 2e-05,
"loss": 0.7472,
"step": 607
},
{
"epoch": 0.20987228167069383,
"grad_norm": 2.6907851696014404,
"learning_rate": 2e-05,
"loss": 0.8709,
"step": 608
},
{
"epoch": 0.2102174663444943,
"grad_norm": 2.713240146636963,
"learning_rate": 2e-05,
"loss": 0.8506,
"step": 609
},
{
"epoch": 0.2105626510182948,
"grad_norm": 2.8039517402648926,
"learning_rate": 2e-05,
"loss": 0.8109,
"step": 610
},
{
"epoch": 0.21090783569209526,
"grad_norm": 2.2727177143096924,
"learning_rate": 2e-05,
"loss": 0.7678,
"step": 611
},
{
"epoch": 0.21125302036589574,
"grad_norm": 2.36690092086792,
"learning_rate": 2e-05,
"loss": 0.8283,
"step": 612
},
{
"epoch": 0.21159820503969623,
"grad_norm": 2.436732530593872,
"learning_rate": 2e-05,
"loss": 0.7969,
"step": 613
},
{
"epoch": 0.21194338971349672,
"grad_norm": 2.499091148376465,
"learning_rate": 2e-05,
"loss": 0.8403,
"step": 614
},
{
"epoch": 0.2122885743872972,
"grad_norm": 3.2683467864990234,
"learning_rate": 2e-05,
"loss": 0.7849,
"step": 615
},
{
"epoch": 0.2126337590610977,
"grad_norm": 2.452500343322754,
"learning_rate": 2e-05,
"loss": 0.7606,
"step": 616
},
{
"epoch": 0.21297894373489817,
"grad_norm": 2.5727200508117676,
"learning_rate": 2e-05,
"loss": 0.8454,
"step": 617
},
{
"epoch": 0.21332412840869866,
"grad_norm": 2.3810341358184814,
"learning_rate": 2e-05,
"loss": 0.7995,
"step": 618
},
{
"epoch": 0.21366931308249915,
"grad_norm": 2.748025417327881,
"learning_rate": 2e-05,
"loss": 0.8918,
"step": 619
},
{
"epoch": 0.21401449775629963,
"grad_norm": 2.6371870040893555,
"learning_rate": 2e-05,
"loss": 0.8178,
"step": 620
},
{
"epoch": 0.2143596824301001,
"grad_norm": 2.570796489715576,
"learning_rate": 2e-05,
"loss": 0.7938,
"step": 621
},
{
"epoch": 0.21470486710390058,
"grad_norm": 2.523061513900757,
"learning_rate": 2e-05,
"loss": 0.8267,
"step": 622
},
{
"epoch": 0.21505005177770106,
"grad_norm": 2.45378041267395,
"learning_rate": 2e-05,
"loss": 0.8556,
"step": 623
},
{
"epoch": 0.21539523645150155,
"grad_norm": 2.567342758178711,
"learning_rate": 2e-05,
"loss": 0.7609,
"step": 624
},
{
"epoch": 0.21574042112530203,
"grad_norm": 2.612757444381714,
"learning_rate": 2e-05,
"loss": 0.8949,
"step": 625
},
{
"epoch": 0.21608560579910252,
"grad_norm": 2.7509217262268066,
"learning_rate": 2e-05,
"loss": 0.8208,
"step": 626
},
{
"epoch": 0.216430790472903,
"grad_norm": 2.6197450160980225,
"learning_rate": 2e-05,
"loss": 0.8406,
"step": 627
},
{
"epoch": 0.2167759751467035,
"grad_norm": 2.7045562267303467,
"learning_rate": 2e-05,
"loss": 0.8764,
"step": 628
},
{
"epoch": 0.21712115982050398,
"grad_norm": 2.5740184783935547,
"learning_rate": 2e-05,
"loss": 0.8362,
"step": 629
},
{
"epoch": 0.21746634449430446,
"grad_norm": 3.1988041400909424,
"learning_rate": 2e-05,
"loss": 0.7933,
"step": 630
},
{
"epoch": 0.21781152916810492,
"grad_norm": 2.2488019466400146,
"learning_rate": 2e-05,
"loss": 0.7828,
"step": 631
},
{
"epoch": 0.2181567138419054,
"grad_norm": 2.442720651626587,
"learning_rate": 2e-05,
"loss": 0.7974,
"step": 632
},
{
"epoch": 0.2185018985157059,
"grad_norm": 2.7867915630340576,
"learning_rate": 2e-05,
"loss": 0.8298,
"step": 633
},
{
"epoch": 0.21884708318950638,
"grad_norm": 2.512460470199585,
"learning_rate": 2e-05,
"loss": 0.7807,
"step": 634
},
{
"epoch": 0.21919226786330687,
"grad_norm": 2.34566330909729,
"learning_rate": 2e-05,
"loss": 0.8103,
"step": 635
},
{
"epoch": 0.21953745253710735,
"grad_norm": 2.3921985626220703,
"learning_rate": 2e-05,
"loss": 0.8099,
"step": 636
},
{
"epoch": 0.21988263721090784,
"grad_norm": 2.790856122970581,
"learning_rate": 2e-05,
"loss": 0.8027,
"step": 637
},
{
"epoch": 0.22022782188470832,
"grad_norm": 2.4337832927703857,
"learning_rate": 2e-05,
"loss": 0.8459,
"step": 638
},
{
"epoch": 0.2205730065585088,
"grad_norm": 2.626878261566162,
"learning_rate": 2e-05,
"loss": 0.822,
"step": 639
},
{
"epoch": 0.2209181912323093,
"grad_norm": 2.411151647567749,
"learning_rate": 2e-05,
"loss": 0.7889,
"step": 640
},
{
"epoch": 0.22126337590610975,
"grad_norm": 2.780468463897705,
"learning_rate": 2e-05,
"loss": 0.8204,
"step": 641
},
{
"epoch": 0.22160856057991024,
"grad_norm": 2.384120464324951,
"learning_rate": 2e-05,
"loss": 0.794,
"step": 642
},
{
"epoch": 0.22195374525371073,
"grad_norm": 2.7003538608551025,
"learning_rate": 2e-05,
"loss": 0.8721,
"step": 643
},
{
"epoch": 0.2222989299275112,
"grad_norm": 2.5736396312713623,
"learning_rate": 2e-05,
"loss": 0.7718,
"step": 644
},
{
"epoch": 0.2226441146013117,
"grad_norm": 2.5534183979034424,
"learning_rate": 2e-05,
"loss": 0.8039,
"step": 645
},
{
"epoch": 0.22298929927511218,
"grad_norm": 2.4901063442230225,
"learning_rate": 2e-05,
"loss": 0.8664,
"step": 646
},
{
"epoch": 0.22333448394891267,
"grad_norm": 2.5854790210723877,
"learning_rate": 2e-05,
"loss": 0.849,
"step": 647
},
{
"epoch": 0.22367966862271316,
"grad_norm": 2.4237279891967773,
"learning_rate": 2e-05,
"loss": 0.7571,
"step": 648
},
{
"epoch": 0.22402485329651364,
"grad_norm": 2.410684108734131,
"learning_rate": 2e-05,
"loss": 0.7899,
"step": 649
},
{
"epoch": 0.22437003797031413,
"grad_norm": 2.4895997047424316,
"learning_rate": 2e-05,
"loss": 0.7699,
"step": 650
},
{
"epoch": 0.22471522264411462,
"grad_norm": 2.3816611766815186,
"learning_rate": 2e-05,
"loss": 0.7872,
"step": 651
},
{
"epoch": 0.22506040731791507,
"grad_norm": 2.417294979095459,
"learning_rate": 2e-05,
"loss": 0.7714,
"step": 652
},
{
"epoch": 0.22540559199171556,
"grad_norm": 2.592925548553467,
"learning_rate": 2e-05,
"loss": 0.817,
"step": 653
},
{
"epoch": 0.22575077666551605,
"grad_norm": 2.4112489223480225,
"learning_rate": 2e-05,
"loss": 0.8018,
"step": 654
},
{
"epoch": 0.22609596133931653,
"grad_norm": 2.340555191040039,
"learning_rate": 2e-05,
"loss": 0.781,
"step": 655
},
{
"epoch": 0.22644114601311702,
"grad_norm": 2.5071909427642822,
"learning_rate": 2e-05,
"loss": 0.8527,
"step": 656
},
{
"epoch": 0.2267863306869175,
"grad_norm": 2.465301275253296,
"learning_rate": 2e-05,
"loss": 0.827,
"step": 657
},
{
"epoch": 0.227131515360718,
"grad_norm": 2.3387086391448975,
"learning_rate": 2e-05,
"loss": 0.7627,
"step": 658
},
{
"epoch": 0.22747670003451848,
"grad_norm": 2.5895376205444336,
"learning_rate": 2e-05,
"loss": 0.7968,
"step": 659
},
{
"epoch": 0.22782188470831896,
"grad_norm": 2.829334020614624,
"learning_rate": 2e-05,
"loss": 0.8498,
"step": 660
},
{
"epoch": 0.22816706938211945,
"grad_norm": 2.4148943424224854,
"learning_rate": 2e-05,
"loss": 0.7726,
"step": 661
},
{
"epoch": 0.2285122540559199,
"grad_norm": 2.487060546875,
"learning_rate": 2e-05,
"loss": 0.7653,
"step": 662
},
{
"epoch": 0.2288574387297204,
"grad_norm": 2.4439961910247803,
"learning_rate": 2e-05,
"loss": 0.84,
"step": 663
},
{
"epoch": 0.22920262340352088,
"grad_norm": 2.485614061355591,
"learning_rate": 2e-05,
"loss": 0.8145,
"step": 664
},
{
"epoch": 0.22954780807732136,
"grad_norm": 2.3690195083618164,
"learning_rate": 2e-05,
"loss": 0.7946,
"step": 665
},
{
"epoch": 0.22989299275112185,
"grad_norm": 2.5638961791992188,
"learning_rate": 2e-05,
"loss": 0.7886,
"step": 666
},
{
"epoch": 0.23023817742492234,
"grad_norm": 2.3436648845672607,
"learning_rate": 2e-05,
"loss": 0.8196,
"step": 667
},
{
"epoch": 0.23058336209872282,
"grad_norm": 2.5709872245788574,
"learning_rate": 2e-05,
"loss": 0.8286,
"step": 668
},
{
"epoch": 0.2309285467725233,
"grad_norm": 2.56705379486084,
"learning_rate": 2e-05,
"loss": 0.8495,
"step": 669
},
{
"epoch": 0.2312737314463238,
"grad_norm": 2.5212414264678955,
"learning_rate": 2e-05,
"loss": 0.7668,
"step": 670
},
{
"epoch": 0.23161891612012428,
"grad_norm": 2.5123586654663086,
"learning_rate": 2e-05,
"loss": 0.7845,
"step": 671
},
{
"epoch": 0.23196410079392474,
"grad_norm": 2.5270276069641113,
"learning_rate": 2e-05,
"loss": 0.8227,
"step": 672
},
{
"epoch": 0.23230928546772522,
"grad_norm": 2.382445812225342,
"learning_rate": 2e-05,
"loss": 0.7476,
"step": 673
},
{
"epoch": 0.2326544701415257,
"grad_norm": 4.354213237762451,
"learning_rate": 2e-05,
"loss": 0.8111,
"step": 674
},
{
"epoch": 0.2329996548153262,
"grad_norm": 2.542351722717285,
"learning_rate": 2e-05,
"loss": 0.8647,
"step": 675
},
{
"epoch": 0.23334483948912668,
"grad_norm": 2.492007255554199,
"learning_rate": 2e-05,
"loss": 0.7812,
"step": 676
},
{
"epoch": 0.23369002416292717,
"grad_norm": 2.8558623790740967,
"learning_rate": 2e-05,
"loss": 0.7819,
"step": 677
},
{
"epoch": 0.23403520883672765,
"grad_norm": 2.424027681350708,
"learning_rate": 2e-05,
"loss": 0.7798,
"step": 678
},
{
"epoch": 0.23438039351052814,
"grad_norm": 2.9818484783172607,
"learning_rate": 2e-05,
"loss": 0.8442,
"step": 679
},
{
"epoch": 0.23472557818432863,
"grad_norm": 2.5655038356781006,
"learning_rate": 2e-05,
"loss": 0.8554,
"step": 680
},
{
"epoch": 0.2350707628581291,
"grad_norm": 2.6559510231018066,
"learning_rate": 2e-05,
"loss": 0.828,
"step": 681
},
{
"epoch": 0.23541594753192957,
"grad_norm": 2.5213375091552734,
"learning_rate": 2e-05,
"loss": 0.8371,
"step": 682
},
{
"epoch": 0.23576113220573006,
"grad_norm": 2.5891947746276855,
"learning_rate": 2e-05,
"loss": 0.8425,
"step": 683
},
{
"epoch": 0.23610631687953054,
"grad_norm": 2.673092842102051,
"learning_rate": 2e-05,
"loss": 0.8405,
"step": 684
},
{
"epoch": 0.23645150155333103,
"grad_norm": 2.362226724624634,
"learning_rate": 2e-05,
"loss": 0.7775,
"step": 685
},
{
"epoch": 0.23679668622713151,
"grad_norm": 2.7180612087249756,
"learning_rate": 2e-05,
"loss": 0.8528,
"step": 686
},
{
"epoch": 0.237141870900932,
"grad_norm": 2.40690541267395,
"learning_rate": 2e-05,
"loss": 0.8133,
"step": 687
},
{
"epoch": 0.2374870555747325,
"grad_norm": 2.5823540687561035,
"learning_rate": 2e-05,
"loss": 0.8635,
"step": 688
},
{
"epoch": 0.23783224024853297,
"grad_norm": 2.387899398803711,
"learning_rate": 2e-05,
"loss": 0.8258,
"step": 689
},
{
"epoch": 0.23817742492233346,
"grad_norm": 3.080235004425049,
"learning_rate": 2e-05,
"loss": 0.8311,
"step": 690
},
{
"epoch": 0.23852260959613394,
"grad_norm": 2.466670274734497,
"learning_rate": 2e-05,
"loss": 0.7914,
"step": 691
},
{
"epoch": 0.2388677942699344,
"grad_norm": 2.4669206142425537,
"learning_rate": 2e-05,
"loss": 0.8205,
"step": 692
},
{
"epoch": 0.2392129789437349,
"grad_norm": 2.385197877883911,
"learning_rate": 2e-05,
"loss": 0.8081,
"step": 693
},
{
"epoch": 0.23955816361753537,
"grad_norm": 2.5215725898742676,
"learning_rate": 2e-05,
"loss": 0.8061,
"step": 694
},
{
"epoch": 0.23990334829133586,
"grad_norm": 2.4044137001037598,
"learning_rate": 2e-05,
"loss": 0.7884,
"step": 695
},
{
"epoch": 0.24024853296513635,
"grad_norm": 2.598735809326172,
"learning_rate": 2e-05,
"loss": 0.8283,
"step": 696
},
{
"epoch": 0.24059371763893683,
"grad_norm": 2.434346914291382,
"learning_rate": 2e-05,
"loss": 0.8366,
"step": 697
},
{
"epoch": 0.24093890231273732,
"grad_norm": 2.321727991104126,
"learning_rate": 2e-05,
"loss": 0.8012,
"step": 698
},
{
"epoch": 0.2412840869865378,
"grad_norm": 2.6814427375793457,
"learning_rate": 2e-05,
"loss": 0.7923,
"step": 699
},
{
"epoch": 0.2416292716603383,
"grad_norm": 2.978170871734619,
"learning_rate": 2e-05,
"loss": 0.8642,
"step": 700
},
{
"epoch": 0.24197445633413878,
"grad_norm": 2.447589635848999,
"learning_rate": 2e-05,
"loss": 0.8947,
"step": 701
},
{
"epoch": 0.24231964100793923,
"grad_norm": 2.542748212814331,
"learning_rate": 2e-05,
"loss": 0.8199,
"step": 702
},
{
"epoch": 0.24266482568173972,
"grad_norm": 2.26539945602417,
"learning_rate": 2e-05,
"loss": 0.8165,
"step": 703
},
{
"epoch": 0.2430100103555402,
"grad_norm": 2.310394287109375,
"learning_rate": 2e-05,
"loss": 0.8057,
"step": 704
},
{
"epoch": 0.2433551950293407,
"grad_norm": 2.352877140045166,
"learning_rate": 2e-05,
"loss": 0.7986,
"step": 705
},
{
"epoch": 0.24370037970314118,
"grad_norm": 2.257847309112549,
"learning_rate": 2e-05,
"loss": 0.7638,
"step": 706
},
{
"epoch": 0.24404556437694166,
"grad_norm": 2.4806249141693115,
"learning_rate": 2e-05,
"loss": 0.8122,
"step": 707
},
{
"epoch": 0.24439074905074215,
"grad_norm": 2.5089046955108643,
"learning_rate": 2e-05,
"loss": 0.7944,
"step": 708
},
{
"epoch": 0.24473593372454264,
"grad_norm": 2.7169389724731445,
"learning_rate": 2e-05,
"loss": 0.8342,
"step": 709
},
{
"epoch": 0.24508111839834312,
"grad_norm": 2.3075709342956543,
"learning_rate": 2e-05,
"loss": 0.84,
"step": 710
},
{
"epoch": 0.2454263030721436,
"grad_norm": 2.3579206466674805,
"learning_rate": 2e-05,
"loss": 0.7753,
"step": 711
},
{
"epoch": 0.24577148774594407,
"grad_norm": 2.247012138366699,
"learning_rate": 2e-05,
"loss": 0.7577,
"step": 712
},
{
"epoch": 0.24611667241974455,
"grad_norm": 2.456413984298706,
"learning_rate": 2e-05,
"loss": 0.766,
"step": 713
},
{
"epoch": 0.24646185709354504,
"grad_norm": 2.334796190261841,
"learning_rate": 2e-05,
"loss": 0.8091,
"step": 714
},
{
"epoch": 0.24680704176734553,
"grad_norm": 2.7342369556427,
"learning_rate": 2e-05,
"loss": 0.7921,
"step": 715
},
{
"epoch": 0.247152226441146,
"grad_norm": 2.421428680419922,
"learning_rate": 2e-05,
"loss": 0.8399,
"step": 716
},
{
"epoch": 0.2474974111149465,
"grad_norm": 2.294358253479004,
"learning_rate": 2e-05,
"loss": 0.8151,
"step": 717
},
{
"epoch": 0.24784259578874698,
"grad_norm": 2.2485649585723877,
"learning_rate": 2e-05,
"loss": 0.8042,
"step": 718
},
{
"epoch": 0.24818778046254747,
"grad_norm": 2.320042610168457,
"learning_rate": 2e-05,
"loss": 0.819,
"step": 719
},
{
"epoch": 0.24853296513634796,
"grad_norm": 2.550518751144409,
"learning_rate": 2e-05,
"loss": 0.8183,
"step": 720
},
{
"epoch": 0.24887814981014844,
"grad_norm": 2.2849254608154297,
"learning_rate": 2e-05,
"loss": 0.776,
"step": 721
},
{
"epoch": 0.2492233344839489,
"grad_norm": 2.62604022026062,
"learning_rate": 2e-05,
"loss": 0.8025,
"step": 722
},
{
"epoch": 0.24956851915774939,
"grad_norm": 2.673774480819702,
"learning_rate": 2e-05,
"loss": 0.835,
"step": 723
},
{
"epoch": 0.24991370383154987,
"grad_norm": 2.6909732818603516,
"learning_rate": 2e-05,
"loss": 0.8302,
"step": 724
},
{
"epoch": 0.2502588885053504,
"grad_norm": 2.456923484802246,
"learning_rate": 2e-05,
"loss": 0.7896,
"step": 725
},
{
"epoch": 0.25060407317915084,
"grad_norm": 2.4879603385925293,
"learning_rate": 2e-05,
"loss": 0.7855,
"step": 726
},
{
"epoch": 0.2509492578529513,
"grad_norm": 2.568673610687256,
"learning_rate": 2e-05,
"loss": 0.8927,
"step": 727
},
{
"epoch": 0.2512944425267518,
"grad_norm": 2.581019878387451,
"learning_rate": 2e-05,
"loss": 0.8248,
"step": 728
},
{
"epoch": 0.2516396272005523,
"grad_norm": 2.6271448135375977,
"learning_rate": 2e-05,
"loss": 0.773,
"step": 729
},
{
"epoch": 0.2519848118743528,
"grad_norm": 2.3511996269226074,
"learning_rate": 2e-05,
"loss": 0.7718,
"step": 730
},
{
"epoch": 0.25232999654815325,
"grad_norm": 2.275012493133545,
"learning_rate": 2e-05,
"loss": 0.7246,
"step": 731
},
{
"epoch": 0.25267518122195376,
"grad_norm": 2.4964468479156494,
"learning_rate": 2e-05,
"loss": 0.7716,
"step": 732
},
{
"epoch": 0.2530203658957542,
"grad_norm": 3.0296764373779297,
"learning_rate": 2e-05,
"loss": 0.805,
"step": 733
},
{
"epoch": 0.25336555056955473,
"grad_norm": 2.7500593662261963,
"learning_rate": 2e-05,
"loss": 0.8182,
"step": 734
},
{
"epoch": 0.2537107352433552,
"grad_norm": 2.582059144973755,
"learning_rate": 2e-05,
"loss": 0.8504,
"step": 735
},
{
"epoch": 0.2540559199171557,
"grad_norm": 2.1892638206481934,
"learning_rate": 2e-05,
"loss": 0.7769,
"step": 736
},
{
"epoch": 0.25440110459095616,
"grad_norm": 2.6324143409729004,
"learning_rate": 2e-05,
"loss": 0.8411,
"step": 737
},
{
"epoch": 0.2547462892647566,
"grad_norm": 2.5749869346618652,
"learning_rate": 2e-05,
"loss": 0.8162,
"step": 738
},
{
"epoch": 0.25509147393855713,
"grad_norm": 2.639956474304199,
"learning_rate": 2e-05,
"loss": 0.7715,
"step": 739
},
{
"epoch": 0.2554366586123576,
"grad_norm": 2.4721555709838867,
"learning_rate": 2e-05,
"loss": 0.7986,
"step": 740
},
{
"epoch": 0.2557818432861581,
"grad_norm": 2.4588711261749268,
"learning_rate": 2e-05,
"loss": 0.8353,
"step": 741
},
{
"epoch": 0.25612702795995856,
"grad_norm": 2.6208677291870117,
"learning_rate": 2e-05,
"loss": 0.841,
"step": 742
},
{
"epoch": 0.2564722126337591,
"grad_norm": 2.3099710941314697,
"learning_rate": 2e-05,
"loss": 0.8081,
"step": 743
},
{
"epoch": 0.25681739730755954,
"grad_norm": 2.3112940788269043,
"learning_rate": 2e-05,
"loss": 0.7497,
"step": 744
},
{
"epoch": 0.25716258198136005,
"grad_norm": 2.443956136703491,
"learning_rate": 2e-05,
"loss": 0.8086,
"step": 745
},
{
"epoch": 0.2575077666551605,
"grad_norm": 2.4621360301971436,
"learning_rate": 2e-05,
"loss": 0.8026,
"step": 746
},
{
"epoch": 0.25785295132896097,
"grad_norm": 2.369255304336548,
"learning_rate": 2e-05,
"loss": 0.7849,
"step": 747
},
{
"epoch": 0.2581981360027615,
"grad_norm": 2.385077476501465,
"learning_rate": 2e-05,
"loss": 0.835,
"step": 748
},
{
"epoch": 0.25854332067656194,
"grad_norm": 2.5307819843292236,
"learning_rate": 2e-05,
"loss": 0.8743,
"step": 749
},
{
"epoch": 0.25888850535036245,
"grad_norm": 2.4265220165252686,
"learning_rate": 2e-05,
"loss": 0.7868,
"step": 750
},
{
"epoch": 0.2592336900241629,
"grad_norm": 2.4824352264404297,
"learning_rate": 2e-05,
"loss": 0.8095,
"step": 751
},
{
"epoch": 0.2595788746979634,
"grad_norm": 2.437535524368286,
"learning_rate": 2e-05,
"loss": 0.8011,
"step": 752
},
{
"epoch": 0.2599240593717639,
"grad_norm": 2.4075124263763428,
"learning_rate": 2e-05,
"loss": 0.7654,
"step": 753
},
{
"epoch": 0.2602692440455644,
"grad_norm": 2.4554355144500732,
"learning_rate": 2e-05,
"loss": 0.7917,
"step": 754
},
{
"epoch": 0.26061442871936485,
"grad_norm": 2.8890931606292725,
"learning_rate": 2e-05,
"loss": 0.8628,
"step": 755
},
{
"epoch": 0.26095961339316537,
"grad_norm": 2.5634958744049072,
"learning_rate": 2e-05,
"loss": 0.8342,
"step": 756
},
{
"epoch": 0.2613047980669658,
"grad_norm": 2.2676279544830322,
"learning_rate": 2e-05,
"loss": 0.7785,
"step": 757
},
{
"epoch": 0.2616499827407663,
"grad_norm": 2.420266628265381,
"learning_rate": 2e-05,
"loss": 0.7597,
"step": 758
},
{
"epoch": 0.2619951674145668,
"grad_norm": 2.5409836769104004,
"learning_rate": 2e-05,
"loss": 0.7625,
"step": 759
},
{
"epoch": 0.26234035208836726,
"grad_norm": 2.6637697219848633,
"learning_rate": 2e-05,
"loss": 0.8447,
"step": 760
},
{
"epoch": 0.26268553676216777,
"grad_norm": 2.5580368041992188,
"learning_rate": 2e-05,
"loss": 0.8483,
"step": 761
},
{
"epoch": 0.26303072143596823,
"grad_norm": 2.1736929416656494,
"learning_rate": 2e-05,
"loss": 0.7846,
"step": 762
},
{
"epoch": 0.26337590610976874,
"grad_norm": 2.432163715362549,
"learning_rate": 2e-05,
"loss": 0.837,
"step": 763
},
{
"epoch": 0.2637210907835692,
"grad_norm": 2.2890450954437256,
"learning_rate": 2e-05,
"loss": 0.7575,
"step": 764
},
{
"epoch": 0.2640662754573697,
"grad_norm": 2.147167921066284,
"learning_rate": 2e-05,
"loss": 0.748,
"step": 765
},
{
"epoch": 0.2644114601311702,
"grad_norm": 2.2331297397613525,
"learning_rate": 2e-05,
"loss": 0.7877,
"step": 766
},
{
"epoch": 0.2647566448049707,
"grad_norm": 2.6834170818328857,
"learning_rate": 2e-05,
"loss": 0.7593,
"step": 767
},
{
"epoch": 0.26510182947877114,
"grad_norm": 2.3029916286468506,
"learning_rate": 2e-05,
"loss": 0.779,
"step": 768
},
{
"epoch": 0.2654470141525716,
"grad_norm": 2.5964062213897705,
"learning_rate": 2e-05,
"loss": 0.7763,
"step": 769
},
{
"epoch": 0.2657921988263721,
"grad_norm": 2.380195140838623,
"learning_rate": 2e-05,
"loss": 0.8466,
"step": 770
},
{
"epoch": 0.2661373835001726,
"grad_norm": 2.3388257026672363,
"learning_rate": 2e-05,
"loss": 0.8285,
"step": 771
},
{
"epoch": 0.2664825681739731,
"grad_norm": 2.679877519607544,
"learning_rate": 2e-05,
"loss": 0.8432,
"step": 772
},
{
"epoch": 0.26682775284777355,
"grad_norm": 2.3452959060668945,
"learning_rate": 2e-05,
"loss": 0.8195,
"step": 773
},
{
"epoch": 0.26717293752157406,
"grad_norm": 2.2456681728363037,
"learning_rate": 2e-05,
"loss": 0.8075,
"step": 774
},
{
"epoch": 0.2675181221953745,
"grad_norm": 3.0491323471069336,
"learning_rate": 2e-05,
"loss": 0.7777,
"step": 775
},
{
"epoch": 0.26786330686917503,
"grad_norm": 2.182220697402954,
"learning_rate": 2e-05,
"loss": 0.8208,
"step": 776
},
{
"epoch": 0.2682084915429755,
"grad_norm": 2.5717360973358154,
"learning_rate": 2e-05,
"loss": 0.8557,
"step": 777
},
{
"epoch": 0.26855367621677595,
"grad_norm": 2.4059348106384277,
"learning_rate": 2e-05,
"loss": 0.7936,
"step": 778
},
{
"epoch": 0.26889886089057646,
"grad_norm": 2.5083248615264893,
"learning_rate": 2e-05,
"loss": 0.8024,
"step": 779
},
{
"epoch": 0.2692440455643769,
"grad_norm": 2.284400463104248,
"learning_rate": 2e-05,
"loss": 0.7757,
"step": 780
},
{
"epoch": 0.26958923023817744,
"grad_norm": 2.477802038192749,
"learning_rate": 2e-05,
"loss": 0.798,
"step": 781
},
{
"epoch": 0.2699344149119779,
"grad_norm": 2.2841451168060303,
"learning_rate": 2e-05,
"loss": 0.705,
"step": 782
},
{
"epoch": 0.2702795995857784,
"grad_norm": 2.7210919857025146,
"learning_rate": 2e-05,
"loss": 0.8506,
"step": 783
},
{
"epoch": 0.27062478425957887,
"grad_norm": 2.205794095993042,
"learning_rate": 2e-05,
"loss": 0.7054,
"step": 784
},
{
"epoch": 0.2709699689333794,
"grad_norm": 2.332463026046753,
"learning_rate": 2e-05,
"loss": 0.8203,
"step": 785
},
{
"epoch": 0.27131515360717984,
"grad_norm": 2.1899123191833496,
"learning_rate": 2e-05,
"loss": 0.7996,
"step": 786
},
{
"epoch": 0.27166033828098035,
"grad_norm": 3.033325433731079,
"learning_rate": 2e-05,
"loss": 0.8023,
"step": 787
},
{
"epoch": 0.2720055229547808,
"grad_norm": 2.5546796321868896,
"learning_rate": 2e-05,
"loss": 0.7844,
"step": 788
},
{
"epoch": 0.27235070762858127,
"grad_norm": 2.4093902111053467,
"learning_rate": 2e-05,
"loss": 0.7127,
"step": 789
},
{
"epoch": 0.2726958923023818,
"grad_norm": 2.6239256858825684,
"learning_rate": 2e-05,
"loss": 0.8067,
"step": 790
},
{
"epoch": 0.27304107697618224,
"grad_norm": 2.5717294216156006,
"learning_rate": 2e-05,
"loss": 0.7939,
"step": 791
},
{
"epoch": 0.27338626164998275,
"grad_norm": 2.6218199729919434,
"learning_rate": 2e-05,
"loss": 0.7864,
"step": 792
},
{
"epoch": 0.2737314463237832,
"grad_norm": 2.2493772506713867,
"learning_rate": 2e-05,
"loss": 0.7935,
"step": 793
},
{
"epoch": 0.2740766309975837,
"grad_norm": 2.1874711513519287,
"learning_rate": 2e-05,
"loss": 0.7678,
"step": 794
},
{
"epoch": 0.2744218156713842,
"grad_norm": 2.2770395278930664,
"learning_rate": 2e-05,
"loss": 0.7615,
"step": 795
},
{
"epoch": 0.2747670003451847,
"grad_norm": 2.453977584838867,
"learning_rate": 2e-05,
"loss": 0.827,
"step": 796
},
{
"epoch": 0.27511218501898516,
"grad_norm": 2.3395955562591553,
"learning_rate": 2e-05,
"loss": 0.856,
"step": 797
},
{
"epoch": 0.2754573696927856,
"grad_norm": 2.4298174381256104,
"learning_rate": 2e-05,
"loss": 0.8347,
"step": 798
},
{
"epoch": 0.27580255436658613,
"grad_norm": 2.3910465240478516,
"learning_rate": 2e-05,
"loss": 0.7876,
"step": 799
},
{
"epoch": 0.2761477390403866,
"grad_norm": 2.2215797901153564,
"learning_rate": 2e-05,
"loss": 0.7799,
"step": 800
},
{
"epoch": 0.2764929237141871,
"grad_norm": 2.430864095687866,
"learning_rate": 2e-05,
"loss": 0.8191,
"step": 801
},
{
"epoch": 0.27683810838798756,
"grad_norm": 2.1841373443603516,
"learning_rate": 2e-05,
"loss": 0.7809,
"step": 802
},
{
"epoch": 0.27718329306178807,
"grad_norm": 2.417025089263916,
"learning_rate": 2e-05,
"loss": 0.7814,
"step": 803
},
{
"epoch": 0.27752847773558853,
"grad_norm": 2.4221138954162598,
"learning_rate": 2e-05,
"loss": 0.7432,
"step": 804
},
{
"epoch": 0.27787366240938904,
"grad_norm": 2.243767261505127,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 805
},
{
"epoch": 0.2782188470831895,
"grad_norm": 2.408363103866577,
"learning_rate": 2e-05,
"loss": 0.7653,
"step": 806
},
{
"epoch": 0.27856403175699,
"grad_norm": 2.5052545070648193,
"learning_rate": 2e-05,
"loss": 0.7543,
"step": 807
},
{
"epoch": 0.2789092164307905,
"grad_norm": 2.40278697013855,
"learning_rate": 2e-05,
"loss": 0.8283,
"step": 808
},
{
"epoch": 0.27925440110459093,
"grad_norm": 2.448979377746582,
"learning_rate": 2e-05,
"loss": 0.8131,
"step": 809
},
{
"epoch": 0.27959958577839145,
"grad_norm": 2.434659004211426,
"learning_rate": 2e-05,
"loss": 0.8619,
"step": 810
},
{
"epoch": 0.2799447704521919,
"grad_norm": 2.2524662017822266,
"learning_rate": 2e-05,
"loss": 0.8179,
"step": 811
},
{
"epoch": 0.2802899551259924,
"grad_norm": 2.5318238735198975,
"learning_rate": 2e-05,
"loss": 0.8369,
"step": 812
},
{
"epoch": 0.2806351397997929,
"grad_norm": 2.411334991455078,
"learning_rate": 2e-05,
"loss": 0.7542,
"step": 813
},
{
"epoch": 0.2809803244735934,
"grad_norm": 2.1707065105438232,
"learning_rate": 2e-05,
"loss": 0.7828,
"step": 814
},
{
"epoch": 0.28132550914739385,
"grad_norm": 2.2191224098205566,
"learning_rate": 2e-05,
"loss": 0.8132,
"step": 815
},
{
"epoch": 0.28167069382119436,
"grad_norm": 2.3489272594451904,
"learning_rate": 2e-05,
"loss": 0.756,
"step": 816
},
{
"epoch": 0.2820158784949948,
"grad_norm": 2.3153269290924072,
"learning_rate": 2e-05,
"loss": 0.78,
"step": 817
},
{
"epoch": 0.2823610631687953,
"grad_norm": 2.785445213317871,
"learning_rate": 2e-05,
"loss": 0.9018,
"step": 818
},
{
"epoch": 0.2827062478425958,
"grad_norm": 2.4912357330322266,
"learning_rate": 2e-05,
"loss": 0.7729,
"step": 819
},
{
"epoch": 0.28305143251639625,
"grad_norm": 2.4810216426849365,
"learning_rate": 2e-05,
"loss": 0.8686,
"step": 820
},
{
"epoch": 0.28339661719019676,
"grad_norm": 2.5625271797180176,
"learning_rate": 2e-05,
"loss": 0.8389,
"step": 821
},
{
"epoch": 0.2837418018639972,
"grad_norm": 2.2283105850219727,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 822
},
{
"epoch": 0.28408698653779774,
"grad_norm": 2.3928627967834473,
"learning_rate": 2e-05,
"loss": 0.7676,
"step": 823
},
{
"epoch": 0.2844321712115982,
"grad_norm": 2.4077234268188477,
"learning_rate": 2e-05,
"loss": 0.8363,
"step": 824
},
{
"epoch": 0.2847773558853987,
"grad_norm": 2.534303665161133,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 825
},
{
"epoch": 0.28512254055919917,
"grad_norm": 2.242882251739502,
"learning_rate": 2e-05,
"loss": 0.751,
"step": 826
},
{
"epoch": 0.2854677252329997,
"grad_norm": 2.318525791168213,
"learning_rate": 2e-05,
"loss": 0.7636,
"step": 827
},
{
"epoch": 0.28581290990680014,
"grad_norm": 2.241232395172119,
"learning_rate": 2e-05,
"loss": 0.7765,
"step": 828
},
{
"epoch": 0.2861580945806006,
"grad_norm": 2.368382453918457,
"learning_rate": 2e-05,
"loss": 0.778,
"step": 829
},
{
"epoch": 0.2865032792544011,
"grad_norm": 3.2637546062469482,
"learning_rate": 2e-05,
"loss": 0.7795,
"step": 830
},
{
"epoch": 0.28684846392820157,
"grad_norm": 2.1969919204711914,
"learning_rate": 2e-05,
"loss": 0.8387,
"step": 831
},
{
"epoch": 0.2871936486020021,
"grad_norm": 2.7861218452453613,
"learning_rate": 2e-05,
"loss": 0.836,
"step": 832
},
{
"epoch": 0.28753883327580254,
"grad_norm": 2.5360143184661865,
"learning_rate": 2e-05,
"loss": 0.807,
"step": 833
},
{
"epoch": 0.28788401794960305,
"grad_norm": 2.3107171058654785,
"learning_rate": 2e-05,
"loss": 0.7693,
"step": 834
},
{
"epoch": 0.2882292026234035,
"grad_norm": 2.3907177448272705,
"learning_rate": 2e-05,
"loss": 0.817,
"step": 835
},
{
"epoch": 0.288574387297204,
"grad_norm": 2.210458517074585,
"learning_rate": 2e-05,
"loss": 0.7167,
"step": 836
},
{
"epoch": 0.2889195719710045,
"grad_norm": 2.457772970199585,
"learning_rate": 2e-05,
"loss": 0.8089,
"step": 837
},
{
"epoch": 0.28926475664480494,
"grad_norm": 2.281331777572632,
"learning_rate": 2e-05,
"loss": 0.8044,
"step": 838
},
{
"epoch": 0.28960994131860546,
"grad_norm": 2.2387568950653076,
"learning_rate": 2e-05,
"loss": 0.7641,
"step": 839
},
{
"epoch": 0.2899551259924059,
"grad_norm": 2.3929879665374756,
"learning_rate": 2e-05,
"loss": 0.8267,
"step": 840
},
{
"epoch": 0.29030031066620643,
"grad_norm": 2.53993558883667,
"learning_rate": 2e-05,
"loss": 0.7637,
"step": 841
},
{
"epoch": 0.2906454953400069,
"grad_norm": 2.3552863597869873,
"learning_rate": 2e-05,
"loss": 0.7518,
"step": 842
},
{
"epoch": 0.2909906800138074,
"grad_norm": 2.438748359680176,
"learning_rate": 2e-05,
"loss": 0.8188,
"step": 843
},
{
"epoch": 0.29133586468760786,
"grad_norm": 2.2611324787139893,
"learning_rate": 2e-05,
"loss": 0.7828,
"step": 844
},
{
"epoch": 0.2916810493614084,
"grad_norm": 2.1586382389068604,
"learning_rate": 2e-05,
"loss": 0.7041,
"step": 845
},
{
"epoch": 0.29202623403520883,
"grad_norm": 2.498196601867676,
"learning_rate": 2e-05,
"loss": 0.8151,
"step": 846
},
{
"epoch": 0.29237141870900935,
"grad_norm": 2.294062852859497,
"learning_rate": 2e-05,
"loss": 0.758,
"step": 847
},
{
"epoch": 0.2927166033828098,
"grad_norm": 2.2366390228271484,
"learning_rate": 2e-05,
"loss": 0.803,
"step": 848
},
{
"epoch": 0.29306178805661026,
"grad_norm": 2.4893128871917725,
"learning_rate": 2e-05,
"loss": 0.7928,
"step": 849
},
{
"epoch": 0.2934069727304108,
"grad_norm": 2.315063238143921,
"learning_rate": 2e-05,
"loss": 0.7889,
"step": 850
},
{
"epoch": 0.29375215740421123,
"grad_norm": 2.363053321838379,
"learning_rate": 2e-05,
"loss": 0.7091,
"step": 851
},
{
"epoch": 0.29409734207801175,
"grad_norm": 2.5700855255126953,
"learning_rate": 2e-05,
"loss": 0.8141,
"step": 852
},
{
"epoch": 0.2944425267518122,
"grad_norm": 2.262510299682617,
"learning_rate": 2e-05,
"loss": 0.8147,
"step": 853
},
{
"epoch": 0.2947877114256127,
"grad_norm": 2.419389486312866,
"learning_rate": 2e-05,
"loss": 0.8083,
"step": 854
},
{
"epoch": 0.2951328960994132,
"grad_norm": 2.3417556285858154,
"learning_rate": 2e-05,
"loss": 0.7007,
"step": 855
},
{
"epoch": 0.2954780807732137,
"grad_norm": 2.3193564414978027,
"learning_rate": 2e-05,
"loss": 0.7122,
"step": 856
},
{
"epoch": 0.29582326544701415,
"grad_norm": 2.9452264308929443,
"learning_rate": 2e-05,
"loss": 0.7895,
"step": 857
},
{
"epoch": 0.2961684501208146,
"grad_norm": 2.6561877727508545,
"learning_rate": 2e-05,
"loss": 0.8272,
"step": 858
},
{
"epoch": 0.2965136347946151,
"grad_norm": 2.652930736541748,
"learning_rate": 2e-05,
"loss": 0.7604,
"step": 859
},
{
"epoch": 0.2968588194684156,
"grad_norm": 2.882368564605713,
"learning_rate": 2e-05,
"loss": 0.845,
"step": 860
},
{
"epoch": 0.2972040041422161,
"grad_norm": 2.3490078449249268,
"learning_rate": 2e-05,
"loss": 0.8426,
"step": 861
},
{
"epoch": 0.29754918881601655,
"grad_norm": 2.719566583633423,
"learning_rate": 2e-05,
"loss": 0.8542,
"step": 862
},
{
"epoch": 0.29789437348981707,
"grad_norm": 2.6340112686157227,
"learning_rate": 2e-05,
"loss": 0.7555,
"step": 863
},
{
"epoch": 0.2982395581636175,
"grad_norm": 2.5312600135803223,
"learning_rate": 2e-05,
"loss": 0.7925,
"step": 864
},
{
"epoch": 0.29858474283741804,
"grad_norm": 2.4707272052764893,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 865
},
{
"epoch": 0.2989299275112185,
"grad_norm": 2.3420259952545166,
"learning_rate": 2e-05,
"loss": 0.7992,
"step": 866
},
{
"epoch": 0.299275112185019,
"grad_norm": 2.4616615772247314,
"learning_rate": 2e-05,
"loss": 0.8436,
"step": 867
},
{
"epoch": 0.29962029685881947,
"grad_norm": 2.380793333053589,
"learning_rate": 2e-05,
"loss": 0.7698,
"step": 868
},
{
"epoch": 0.2999654815326199,
"grad_norm": 2.823423147201538,
"learning_rate": 2e-05,
"loss": 0.8289,
"step": 869
},
{
"epoch": 0.30031066620642044,
"grad_norm": 2.4169535636901855,
"learning_rate": 2e-05,
"loss": 0.8078,
"step": 870
},
{
"epoch": 0.3006558508802209,
"grad_norm": 2.4925384521484375,
"learning_rate": 2e-05,
"loss": 0.7523,
"step": 871
},
{
"epoch": 0.3010010355540214,
"grad_norm": 2.4129271507263184,
"learning_rate": 2e-05,
"loss": 0.7943,
"step": 872
},
{
"epoch": 0.30134622022782187,
"grad_norm": 2.366149663925171,
"learning_rate": 2e-05,
"loss": 0.7477,
"step": 873
},
{
"epoch": 0.3016914049016224,
"grad_norm": 2.176807403564453,
"learning_rate": 2e-05,
"loss": 0.7573,
"step": 874
},
{
"epoch": 0.30203658957542284,
"grad_norm": 2.3971831798553467,
"learning_rate": 2e-05,
"loss": 0.8159,
"step": 875
},
{
"epoch": 0.30238177424922336,
"grad_norm": 2.2808613777160645,
"learning_rate": 2e-05,
"loss": 0.8124,
"step": 876
},
{
"epoch": 0.3027269589230238,
"grad_norm": 2.2993972301483154,
"learning_rate": 2e-05,
"loss": 0.7731,
"step": 877
},
{
"epoch": 0.30307214359682433,
"grad_norm": 2.7519545555114746,
"learning_rate": 2e-05,
"loss": 0.8003,
"step": 878
},
{
"epoch": 0.3034173282706248,
"grad_norm": 2.4404890537261963,
"learning_rate": 2e-05,
"loss": 0.8554,
"step": 879
},
{
"epoch": 0.30376251294442524,
"grad_norm": 2.3009283542633057,
"learning_rate": 2e-05,
"loss": 0.7411,
"step": 880
},
{
"epoch": 0.30410769761822576,
"grad_norm": 2.109697103500366,
"learning_rate": 2e-05,
"loss": 0.8216,
"step": 881
},
{
"epoch": 0.3044528822920262,
"grad_norm": 2.3877744674682617,
"learning_rate": 2e-05,
"loss": 0.8017,
"step": 882
},
{
"epoch": 0.30479806696582673,
"grad_norm": 2.4049599170684814,
"learning_rate": 2e-05,
"loss": 0.8427,
"step": 883
},
{
"epoch": 0.3051432516396272,
"grad_norm": 2.552929162979126,
"learning_rate": 2e-05,
"loss": 0.8331,
"step": 884
},
{
"epoch": 0.3054884363134277,
"grad_norm": 2.3699262142181396,
"learning_rate": 2e-05,
"loss": 0.7913,
"step": 885
},
{
"epoch": 0.30583362098722816,
"grad_norm": 2.175236463546753,
"learning_rate": 2e-05,
"loss": 0.7643,
"step": 886
},
{
"epoch": 0.3061788056610287,
"grad_norm": 2.194126605987549,
"learning_rate": 2e-05,
"loss": 0.7641,
"step": 887
},
{
"epoch": 0.30652399033482913,
"grad_norm": 2.1524200439453125,
"learning_rate": 2e-05,
"loss": 0.7448,
"step": 888
},
{
"epoch": 0.3068691750086296,
"grad_norm": 2.622220516204834,
"learning_rate": 2e-05,
"loss": 0.9229,
"step": 889
},
{
"epoch": 0.3072143596824301,
"grad_norm": 2.2300989627838135,
"learning_rate": 2e-05,
"loss": 0.7906,
"step": 890
},
{
"epoch": 0.30755954435623056,
"grad_norm": 2.328655958175659,
"learning_rate": 2e-05,
"loss": 0.7749,
"step": 891
},
{
"epoch": 0.3079047290300311,
"grad_norm": 2.381244659423828,
"learning_rate": 2e-05,
"loss": 0.761,
"step": 892
},
{
"epoch": 0.30824991370383154,
"grad_norm": 2.3483664989471436,
"learning_rate": 2e-05,
"loss": 0.8149,
"step": 893
},
{
"epoch": 0.30859509837763205,
"grad_norm": 2.3732330799102783,
"learning_rate": 2e-05,
"loss": 0.8124,
"step": 894
},
{
"epoch": 0.3089402830514325,
"grad_norm": 2.227066993713379,
"learning_rate": 2e-05,
"loss": 0.7761,
"step": 895
},
{
"epoch": 0.309285467725233,
"grad_norm": 2.516470432281494,
"learning_rate": 2e-05,
"loss": 0.8069,
"step": 896
},
{
"epoch": 0.3096306523990335,
"grad_norm": 2.320647716522217,
"learning_rate": 2e-05,
"loss": 0.7664,
"step": 897
},
{
"epoch": 0.309975837072834,
"grad_norm": 2.3482866287231445,
"learning_rate": 2e-05,
"loss": 0.7724,
"step": 898
},
{
"epoch": 0.31032102174663445,
"grad_norm": 2.428755521774292,
"learning_rate": 2e-05,
"loss": 0.8333,
"step": 899
},
{
"epoch": 0.3106662064204349,
"grad_norm": 2.6640501022338867,
"learning_rate": 2e-05,
"loss": 0.7432,
"step": 900
},
{
"epoch": 0.3110113910942354,
"grad_norm": 2.2669763565063477,
"learning_rate": 2e-05,
"loss": 0.8351,
"step": 901
},
{
"epoch": 0.3113565757680359,
"grad_norm": 2.370856761932373,
"learning_rate": 2e-05,
"loss": 0.8207,
"step": 902
},
{
"epoch": 0.3117017604418364,
"grad_norm": 2.3243308067321777,
"learning_rate": 2e-05,
"loss": 0.7393,
"step": 903
},
{
"epoch": 0.31204694511563685,
"grad_norm": 2.1420657634735107,
"learning_rate": 2e-05,
"loss": 0.8058,
"step": 904
},
{
"epoch": 0.31239212978943737,
"grad_norm": 2.0886402130126953,
"learning_rate": 2e-05,
"loss": 0.8045,
"step": 905
},
{
"epoch": 0.3127373144632378,
"grad_norm": 2.280564069747925,
"learning_rate": 2e-05,
"loss": 0.7212,
"step": 906
},
{
"epoch": 0.31308249913703834,
"grad_norm": 2.2843098640441895,
"learning_rate": 2e-05,
"loss": 0.7496,
"step": 907
},
{
"epoch": 0.3134276838108388,
"grad_norm": 2.160379409790039,
"learning_rate": 2e-05,
"loss": 0.8435,
"step": 908
},
{
"epoch": 0.31377286848463926,
"grad_norm": 2.27815318107605,
"learning_rate": 2e-05,
"loss": 0.7455,
"step": 909
},
{
"epoch": 0.31411805315843977,
"grad_norm": 2.2776007652282715,
"learning_rate": 2e-05,
"loss": 0.7947,
"step": 910
},
{
"epoch": 0.3144632378322402,
"grad_norm": 2.898167133331299,
"learning_rate": 2e-05,
"loss": 0.6744,
"step": 911
},
{
"epoch": 0.31480842250604074,
"grad_norm": 2.51155424118042,
"learning_rate": 2e-05,
"loss": 0.8273,
"step": 912
},
{
"epoch": 0.3151536071798412,
"grad_norm": 2.273869037628174,
"learning_rate": 2e-05,
"loss": 0.7426,
"step": 913
},
{
"epoch": 0.3154987918536417,
"grad_norm": 2.4767696857452393,
"learning_rate": 2e-05,
"loss": 0.7697,
"step": 914
},
{
"epoch": 0.31584397652744217,
"grad_norm": 2.3529250621795654,
"learning_rate": 2e-05,
"loss": 0.8166,
"step": 915
},
{
"epoch": 0.3161891612012427,
"grad_norm": 2.354522466659546,
"learning_rate": 2e-05,
"loss": 0.7638,
"step": 916
},
{
"epoch": 0.31653434587504314,
"grad_norm": 2.5877065658569336,
"learning_rate": 2e-05,
"loss": 0.8177,
"step": 917
},
{
"epoch": 0.31687953054884366,
"grad_norm": 2.252060890197754,
"learning_rate": 2e-05,
"loss": 0.8002,
"step": 918
},
{
"epoch": 0.3172247152226441,
"grad_norm": 2.7347230911254883,
"learning_rate": 2e-05,
"loss": 0.7618,
"step": 919
},
{
"epoch": 0.3175698998964446,
"grad_norm": 2.3109960556030273,
"learning_rate": 2e-05,
"loss": 0.814,
"step": 920
},
{
"epoch": 0.3179150845702451,
"grad_norm": 2.368884801864624,
"learning_rate": 2e-05,
"loss": 0.8036,
"step": 921
},
{
"epoch": 0.31826026924404555,
"grad_norm": 2.480783700942993,
"learning_rate": 2e-05,
"loss": 0.7088,
"step": 922
},
{
"epoch": 0.31860545391784606,
"grad_norm": 2.4031803607940674,
"learning_rate": 2e-05,
"loss": 0.7326,
"step": 923
},
{
"epoch": 0.3189506385916465,
"grad_norm": 2.323190450668335,
"learning_rate": 2e-05,
"loss": 0.7588,
"step": 924
},
{
"epoch": 0.31929582326544703,
"grad_norm": 2.202821969985962,
"learning_rate": 2e-05,
"loss": 0.7255,
"step": 925
},
{
"epoch": 0.3196410079392475,
"grad_norm": 2.212124824523926,
"learning_rate": 2e-05,
"loss": 0.8272,
"step": 926
},
{
"epoch": 0.319986192613048,
"grad_norm": 2.3351826667785645,
"learning_rate": 2e-05,
"loss": 0.7665,
"step": 927
},
{
"epoch": 0.32033137728684846,
"grad_norm": 2.088829755783081,
"learning_rate": 2e-05,
"loss": 0.7304,
"step": 928
},
{
"epoch": 0.3206765619606489,
"grad_norm": 2.103823661804199,
"learning_rate": 2e-05,
"loss": 0.7687,
"step": 929
},
{
"epoch": 0.32102174663444943,
"grad_norm": 2.1671319007873535,
"learning_rate": 2e-05,
"loss": 0.7811,
"step": 930
},
{
"epoch": 0.3213669313082499,
"grad_norm": 2.0397748947143555,
"learning_rate": 2e-05,
"loss": 0.787,
"step": 931
},
{
"epoch": 0.3217121159820504,
"grad_norm": 2.274035692214966,
"learning_rate": 2e-05,
"loss": 0.8261,
"step": 932
},
{
"epoch": 0.32205730065585086,
"grad_norm": 2.330312967300415,
"learning_rate": 2e-05,
"loss": 0.7617,
"step": 933
},
{
"epoch": 0.3224024853296514,
"grad_norm": 2.260516881942749,
"learning_rate": 2e-05,
"loss": 0.8307,
"step": 934
},
{
"epoch": 0.32274767000345184,
"grad_norm": 2.4335250854492188,
"learning_rate": 2e-05,
"loss": 0.7427,
"step": 935
},
{
"epoch": 0.32309285467725235,
"grad_norm": 2.2778549194335938,
"learning_rate": 2e-05,
"loss": 0.827,
"step": 936
},
{
"epoch": 0.3234380393510528,
"grad_norm": 2.3608310222625732,
"learning_rate": 2e-05,
"loss": 0.7026,
"step": 937
},
{
"epoch": 0.3237832240248533,
"grad_norm": 2.297184705734253,
"learning_rate": 2e-05,
"loss": 0.7688,
"step": 938
},
{
"epoch": 0.3241284086986538,
"grad_norm": 2.3631842136383057,
"learning_rate": 2e-05,
"loss": 0.7836,
"step": 939
},
{
"epoch": 0.32447359337245424,
"grad_norm": 2.25142765045166,
"learning_rate": 2e-05,
"loss": 0.8511,
"step": 940
},
{
"epoch": 0.32481877804625475,
"grad_norm": 2.3377201557159424,
"learning_rate": 2e-05,
"loss": 0.7421,
"step": 941
},
{
"epoch": 0.3251639627200552,
"grad_norm": 2.250669479370117,
"learning_rate": 2e-05,
"loss": 0.6976,
"step": 942
},
{
"epoch": 0.3255091473938557,
"grad_norm": 2.241152286529541,
"learning_rate": 2e-05,
"loss": 0.7744,
"step": 943
},
{
"epoch": 0.3258543320676562,
"grad_norm": 2.282179355621338,
"learning_rate": 2e-05,
"loss": 0.7057,
"step": 944
},
{
"epoch": 0.3261995167414567,
"grad_norm": 2.3644256591796875,
"learning_rate": 2e-05,
"loss": 0.8459,
"step": 945
},
{
"epoch": 0.32654470141525715,
"grad_norm": 2.6764843463897705,
"learning_rate": 2e-05,
"loss": 0.7514,
"step": 946
},
{
"epoch": 0.32688988608905767,
"grad_norm": 2.2294766902923584,
"learning_rate": 2e-05,
"loss": 0.8113,
"step": 947
},
{
"epoch": 0.3272350707628581,
"grad_norm": 2.2124416828155518,
"learning_rate": 2e-05,
"loss": 0.6864,
"step": 948
},
{
"epoch": 0.3275802554366586,
"grad_norm": 2.4242794513702393,
"learning_rate": 2e-05,
"loss": 0.7904,
"step": 949
},
{
"epoch": 0.3279254401104591,
"grad_norm": 2.272757053375244,
"learning_rate": 2e-05,
"loss": 0.7613,
"step": 950
},
{
"epoch": 0.32827062478425956,
"grad_norm": 2.1296377182006836,
"learning_rate": 2e-05,
"loss": 0.778,
"step": 951
},
{
"epoch": 0.32861580945806007,
"grad_norm": 2.3032419681549072,
"learning_rate": 2e-05,
"loss": 0.8262,
"step": 952
},
{
"epoch": 0.32896099413186053,
"grad_norm": 2.254443645477295,
"learning_rate": 2e-05,
"loss": 0.8342,
"step": 953
},
{
"epoch": 0.32930617880566104,
"grad_norm": 2.3815431594848633,
"learning_rate": 2e-05,
"loss": 0.7047,
"step": 954
},
{
"epoch": 0.3296513634794615,
"grad_norm": 2.073943853378296,
"learning_rate": 2e-05,
"loss": 0.7717,
"step": 955
},
{
"epoch": 0.329996548153262,
"grad_norm": 2.1470041275024414,
"learning_rate": 2e-05,
"loss": 0.7661,
"step": 956
},
{
"epoch": 0.3303417328270625,
"grad_norm": 2.346449851989746,
"learning_rate": 2e-05,
"loss": 0.8083,
"step": 957
},
{
"epoch": 0.330686917500863,
"grad_norm": 2.5005104541778564,
"learning_rate": 2e-05,
"loss": 0.7771,
"step": 958
},
{
"epoch": 0.33103210217466345,
"grad_norm": 2.5050013065338135,
"learning_rate": 2e-05,
"loss": 0.8185,
"step": 959
},
{
"epoch": 0.3313772868484639,
"grad_norm": 2.206801652908325,
"learning_rate": 2e-05,
"loss": 0.761,
"step": 960
},
{
"epoch": 0.3317224715222644,
"grad_norm": 2.3278515338897705,
"learning_rate": 2e-05,
"loss": 0.7641,
"step": 961
},
{
"epoch": 0.3320676561960649,
"grad_norm": 2.2485294342041016,
"learning_rate": 2e-05,
"loss": 0.8701,
"step": 962
},
{
"epoch": 0.3324128408698654,
"grad_norm": 2.4933691024780273,
"learning_rate": 2e-05,
"loss": 0.7856,
"step": 963
},
{
"epoch": 0.33275802554366585,
"grad_norm": 2.3348371982574463,
"learning_rate": 2e-05,
"loss": 0.7705,
"step": 964
},
{
"epoch": 0.33310321021746636,
"grad_norm": 2.4476113319396973,
"learning_rate": 2e-05,
"loss": 0.8149,
"step": 965
},
{
"epoch": 0.3334483948912668,
"grad_norm": 2.321218729019165,
"learning_rate": 2e-05,
"loss": 0.8113,
"step": 966
},
{
"epoch": 0.33379357956506733,
"grad_norm": 2.575770378112793,
"learning_rate": 2e-05,
"loss": 0.8143,
"step": 967
},
{
"epoch": 0.3341387642388678,
"grad_norm": 2.182896852493286,
"learning_rate": 2e-05,
"loss": 0.7476,
"step": 968
},
{
"epoch": 0.3344839489126683,
"grad_norm": 2.0320241451263428,
"learning_rate": 2e-05,
"loss": 0.7314,
"step": 969
},
{
"epoch": 0.33482913358646876,
"grad_norm": 2.2373087406158447,
"learning_rate": 2e-05,
"loss": 0.7352,
"step": 970
},
{
"epoch": 0.3351743182602692,
"grad_norm": 2.1493780612945557,
"learning_rate": 2e-05,
"loss": 0.8441,
"step": 971
},
{
"epoch": 0.33551950293406974,
"grad_norm": 3.392732620239258,
"learning_rate": 2e-05,
"loss": 0.7954,
"step": 972
},
{
"epoch": 0.3358646876078702,
"grad_norm": 2.2319388389587402,
"learning_rate": 2e-05,
"loss": 0.8502,
"step": 973
},
{
"epoch": 0.3362098722816707,
"grad_norm": 2.31962251663208,
"learning_rate": 2e-05,
"loss": 0.7509,
"step": 974
},
{
"epoch": 0.33655505695547117,
"grad_norm": 2.3655319213867188,
"learning_rate": 2e-05,
"loss": 0.7724,
"step": 975
},
{
"epoch": 0.3369002416292717,
"grad_norm": 2.3151824474334717,
"learning_rate": 2e-05,
"loss": 0.7318,
"step": 976
},
{
"epoch": 0.33724542630307214,
"grad_norm": 2.2314257621765137,
"learning_rate": 2e-05,
"loss": 0.7646,
"step": 977
},
{
"epoch": 0.33759061097687265,
"grad_norm": 2.140748977661133,
"learning_rate": 2e-05,
"loss": 0.7019,
"step": 978
},
{
"epoch": 0.3379357956506731,
"grad_norm": 2.5134060382843018,
"learning_rate": 2e-05,
"loss": 0.7662,
"step": 979
},
{
"epoch": 0.33828098032447357,
"grad_norm": 2.497497081756592,
"learning_rate": 2e-05,
"loss": 0.7673,
"step": 980
},
{
"epoch": 0.3386261649982741,
"grad_norm": 2.574514627456665,
"learning_rate": 2e-05,
"loss": 0.8181,
"step": 981
},
{
"epoch": 0.33897134967207454,
"grad_norm": 2.1714820861816406,
"learning_rate": 2e-05,
"loss": 0.8158,
"step": 982
},
{
"epoch": 0.33931653434587505,
"grad_norm": 2.1695728302001953,
"learning_rate": 2e-05,
"loss": 0.7305,
"step": 983
},
{
"epoch": 0.3396617190196755,
"grad_norm": 2.3775722980499268,
"learning_rate": 2e-05,
"loss": 0.7287,
"step": 984
},
{
"epoch": 0.340006903693476,
"grad_norm": 2.2514188289642334,
"learning_rate": 2e-05,
"loss": 0.8116,
"step": 985
},
{
"epoch": 0.3403520883672765,
"grad_norm": 2.3686721324920654,
"learning_rate": 2e-05,
"loss": 0.7692,
"step": 986
},
{
"epoch": 0.340697273041077,
"grad_norm": 2.2785122394561768,
"learning_rate": 2e-05,
"loss": 0.7572,
"step": 987
},
{
"epoch": 0.34104245771487746,
"grad_norm": 2.2682762145996094,
"learning_rate": 2e-05,
"loss": 0.7479,
"step": 988
},
{
"epoch": 0.34138764238867797,
"grad_norm": 2.2011334896087646,
"learning_rate": 2e-05,
"loss": 0.7539,
"step": 989
},
{
"epoch": 0.34173282706247843,
"grad_norm": 2.2542169094085693,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 990
},
{
"epoch": 0.3420780117362789,
"grad_norm": 2.443906545639038,
"learning_rate": 2e-05,
"loss": 0.7847,
"step": 991
},
{
"epoch": 0.3424231964100794,
"grad_norm": 2.492189645767212,
"learning_rate": 2e-05,
"loss": 0.7883,
"step": 992
},
{
"epoch": 0.34276838108387986,
"grad_norm": 2.384871006011963,
"learning_rate": 2e-05,
"loss": 0.8122,
"step": 993
},
{
"epoch": 0.34311356575768037,
"grad_norm": 2.2505013942718506,
"learning_rate": 2e-05,
"loss": 0.7727,
"step": 994
},
{
"epoch": 0.34345875043148083,
"grad_norm": 2.2702903747558594,
"learning_rate": 2e-05,
"loss": 0.7939,
"step": 995
},
{
"epoch": 0.34380393510528134,
"grad_norm": 2.3637566566467285,
"learning_rate": 2e-05,
"loss": 0.8628,
"step": 996
},
{
"epoch": 0.3441491197790818,
"grad_norm": 2.1582224369049072,
"learning_rate": 2e-05,
"loss": 0.6891,
"step": 997
},
{
"epoch": 0.3444943044528823,
"grad_norm": 2.3541908264160156,
"learning_rate": 2e-05,
"loss": 0.8119,
"step": 998
},
{
"epoch": 0.3448394891266828,
"grad_norm": 2.225794553756714,
"learning_rate": 2e-05,
"loss": 0.8058,
"step": 999
},
{
"epoch": 0.34518467380048323,
"grad_norm": 2.3808062076568604,
"learning_rate": 2e-05,
"loss": 0.744,
"step": 1000
},
{
"epoch": 0.34552985847428375,
"grad_norm": 2.588101863861084,
"learning_rate": 2e-05,
"loss": 0.8203,
"step": 1001
},
{
"epoch": 0.3458750431480842,
"grad_norm": 2.231476306915283,
"learning_rate": 2e-05,
"loss": 0.8119,
"step": 1002
},
{
"epoch": 0.3462202278218847,
"grad_norm": 2.4350528717041016,
"learning_rate": 2e-05,
"loss": 0.773,
"step": 1003
},
{
"epoch": 0.3465654124956852,
"grad_norm": 2.6913435459136963,
"learning_rate": 2e-05,
"loss": 0.7275,
"step": 1004
},
{
"epoch": 0.3469105971694857,
"grad_norm": 2.163970708847046,
"learning_rate": 2e-05,
"loss": 0.7209,
"step": 1005
},
{
"epoch": 0.34725578184328615,
"grad_norm": 2.2166736125946045,
"learning_rate": 2e-05,
"loss": 0.795,
"step": 1006
},
{
"epoch": 0.34760096651708666,
"grad_norm": 2.3976263999938965,
"learning_rate": 2e-05,
"loss": 0.7924,
"step": 1007
},
{
"epoch": 0.3479461511908871,
"grad_norm": 2.1813788414001465,
"learning_rate": 2e-05,
"loss": 0.7929,
"step": 1008
},
{
"epoch": 0.34829133586468763,
"grad_norm": 2.2161941528320312,
"learning_rate": 2e-05,
"loss": 0.7101,
"step": 1009
},
{
"epoch": 0.3486365205384881,
"grad_norm": 2.2501449584960938,
"learning_rate": 2e-05,
"loss": 0.7653,
"step": 1010
},
{
"epoch": 0.34898170521228855,
"grad_norm": 2.1272668838500977,
"learning_rate": 2e-05,
"loss": 0.7572,
"step": 1011
},
{
"epoch": 0.34932688988608906,
"grad_norm": 2.2361133098602295,
"learning_rate": 2e-05,
"loss": 0.7296,
"step": 1012
},
{
"epoch": 0.3496720745598895,
"grad_norm": 2.0001351833343506,
"learning_rate": 2e-05,
"loss": 0.7929,
"step": 1013
},
{
"epoch": 0.35001725923369004,
"grad_norm": 2.3519656658172607,
"learning_rate": 2e-05,
"loss": 0.7901,
"step": 1014
},
{
"epoch": 0.3503624439074905,
"grad_norm": 2.3613696098327637,
"learning_rate": 2e-05,
"loss": 0.7231,
"step": 1015
},
{
"epoch": 0.350707628581291,
"grad_norm": 2.139997959136963,
"learning_rate": 2e-05,
"loss": 0.7261,
"step": 1016
},
{
"epoch": 0.35105281325509147,
"grad_norm": 14.36734390258789,
"learning_rate": 2e-05,
"loss": 0.8697,
"step": 1017
},
{
"epoch": 0.351397997928892,
"grad_norm": 2.069366455078125,
"learning_rate": 2e-05,
"loss": 0.7451,
"step": 1018
},
{
"epoch": 0.35174318260269244,
"grad_norm": 2.2877488136291504,
"learning_rate": 2e-05,
"loss": 0.8225,
"step": 1019
},
{
"epoch": 0.3520883672764929,
"grad_norm": 3.585495710372925,
"learning_rate": 2e-05,
"loss": 0.7427,
"step": 1020
},
{
"epoch": 0.3524335519502934,
"grad_norm": 5.510530471801758,
"learning_rate": 2e-05,
"loss": 0.7389,
"step": 1021
},
{
"epoch": 0.35277873662409387,
"grad_norm": 5.017457962036133,
"learning_rate": 2e-05,
"loss": 0.8652,
"step": 1022
},
{
"epoch": 0.3531239212978944,
"grad_norm": 2.244147539138794,
"learning_rate": 2e-05,
"loss": 0.7548,
"step": 1023
},
{
"epoch": 0.35346910597169484,
"grad_norm": 2.668184995651245,
"learning_rate": 2e-05,
"loss": 0.7764,
"step": 1024
},
{
"epoch": 0.35381429064549536,
"grad_norm": 2.8163087368011475,
"learning_rate": 2e-05,
"loss": 0.8405,
"step": 1025
},
{
"epoch": 0.3541594753192958,
"grad_norm": 3.4871091842651367,
"learning_rate": 2e-05,
"loss": 0.7922,
"step": 1026
},
{
"epoch": 0.3545046599930963,
"grad_norm": 2.8498363494873047,
"learning_rate": 2e-05,
"loss": 0.8372,
"step": 1027
},
{
"epoch": 0.3548498446668968,
"grad_norm": 2.1822509765625,
"learning_rate": 2e-05,
"loss": 0.7538,
"step": 1028
},
{
"epoch": 0.3551950293406973,
"grad_norm": 2.324803352355957,
"learning_rate": 2e-05,
"loss": 0.7857,
"step": 1029
},
{
"epoch": 0.35554021401449776,
"grad_norm": 2.2336905002593994,
"learning_rate": 2e-05,
"loss": 0.7824,
"step": 1030
},
{
"epoch": 0.3558853986882982,
"grad_norm": 2.2240264415740967,
"learning_rate": 2e-05,
"loss": 0.7381,
"step": 1031
},
{
"epoch": 0.35623058336209873,
"grad_norm": 2.414031982421875,
"learning_rate": 2e-05,
"loss": 0.8329,
"step": 1032
},
{
"epoch": 0.3565757680358992,
"grad_norm": 2.2272632122039795,
"learning_rate": 2e-05,
"loss": 0.8171,
"step": 1033
},
{
"epoch": 0.3569209527096997,
"grad_norm": 2.130920171737671,
"learning_rate": 2e-05,
"loss": 0.7525,
"step": 1034
},
{
"epoch": 0.35726613738350016,
"grad_norm": 2.2277228832244873,
"learning_rate": 2e-05,
"loss": 0.8339,
"step": 1035
},
{
"epoch": 0.3576113220573007,
"grad_norm": 2.4667656421661377,
"learning_rate": 2e-05,
"loss": 0.762,
"step": 1036
},
{
"epoch": 0.35795650673110113,
"grad_norm": 2.354909658432007,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 1037
},
{
"epoch": 0.35830169140490165,
"grad_norm": 2.461592435836792,
"learning_rate": 2e-05,
"loss": 0.7248,
"step": 1038
},
{
"epoch": 0.3586468760787021,
"grad_norm": 2.2202396392822266,
"learning_rate": 2e-05,
"loss": 0.7549,
"step": 1039
},
{
"epoch": 0.35899206075250256,
"grad_norm": 2.314002513885498,
"learning_rate": 2e-05,
"loss": 0.7911,
"step": 1040
},
{
"epoch": 0.3593372454263031,
"grad_norm": 2.5224123001098633,
"learning_rate": 2e-05,
"loss": 0.8159,
"step": 1041
},
{
"epoch": 0.35968243010010353,
"grad_norm": 2.4876797199249268,
"learning_rate": 2e-05,
"loss": 0.8669,
"step": 1042
},
{
"epoch": 0.36002761477390405,
"grad_norm": 2.2885406017303467,
"learning_rate": 2e-05,
"loss": 0.7548,
"step": 1043
},
{
"epoch": 0.3603727994477045,
"grad_norm": 2.188265323638916,
"learning_rate": 2e-05,
"loss": 0.7932,
"step": 1044
},
{
"epoch": 0.360717984121505,
"grad_norm": 2.117037773132324,
"learning_rate": 2e-05,
"loss": 0.7895,
"step": 1045
},
{
"epoch": 0.3610631687953055,
"grad_norm": 2.2959017753601074,
"learning_rate": 2e-05,
"loss": 0.7933,
"step": 1046
},
{
"epoch": 0.361408353469106,
"grad_norm": 2.337918996810913,
"learning_rate": 2e-05,
"loss": 0.7737,
"step": 1047
},
{
"epoch": 0.36175353814290645,
"grad_norm": 2.4764339923858643,
"learning_rate": 2e-05,
"loss": 0.871,
"step": 1048
},
{
"epoch": 0.36209872281670696,
"grad_norm": 2.2121264934539795,
"learning_rate": 2e-05,
"loss": 0.7398,
"step": 1049
},
{
"epoch": 0.3624439074905074,
"grad_norm": 2.2514188289642334,
"learning_rate": 2e-05,
"loss": 0.8016,
"step": 1050
},
{
"epoch": 0.3627890921643079,
"grad_norm": 2.1522557735443115,
"learning_rate": 2e-05,
"loss": 0.7633,
"step": 1051
},
{
"epoch": 0.3631342768381084,
"grad_norm": 2.3215880393981934,
"learning_rate": 2e-05,
"loss": 0.7927,
"step": 1052
},
{
"epoch": 0.36347946151190885,
"grad_norm": 2.335054874420166,
"learning_rate": 2e-05,
"loss": 0.7554,
"step": 1053
},
{
"epoch": 0.36382464618570937,
"grad_norm": 2.3971474170684814,
"learning_rate": 2e-05,
"loss": 0.8281,
"step": 1054
},
{
"epoch": 0.3641698308595098,
"grad_norm": 2.3090717792510986,
"learning_rate": 2e-05,
"loss": 0.7836,
"step": 1055
},
{
"epoch": 0.36451501553331034,
"grad_norm": 2.2745440006256104,
"learning_rate": 2e-05,
"loss": 0.8027,
"step": 1056
},
{
"epoch": 0.3648602002071108,
"grad_norm": 2.216360092163086,
"learning_rate": 2e-05,
"loss": 0.7499,
"step": 1057
},
{
"epoch": 0.3652053848809113,
"grad_norm": 2.138859272003174,
"learning_rate": 2e-05,
"loss": 0.7577,
"step": 1058
},
{
"epoch": 0.36555056955471177,
"grad_norm": 2.1670620441436768,
"learning_rate": 2e-05,
"loss": 0.755,
"step": 1059
},
{
"epoch": 0.3658957542285122,
"grad_norm": 2.2582995891571045,
"learning_rate": 2e-05,
"loss": 0.766,
"step": 1060
},
{
"epoch": 0.36624093890231274,
"grad_norm": 2.2919275760650635,
"learning_rate": 2e-05,
"loss": 0.8076,
"step": 1061
},
{
"epoch": 0.3665861235761132,
"grad_norm": 2.2397828102111816,
"learning_rate": 2e-05,
"loss": 0.8277,
"step": 1062
},
{
"epoch": 0.3669313082499137,
"grad_norm": 2.138180732727051,
"learning_rate": 2e-05,
"loss": 0.7478,
"step": 1063
},
{
"epoch": 0.36727649292371417,
"grad_norm": 2.4074158668518066,
"learning_rate": 2e-05,
"loss": 0.7679,
"step": 1064
},
{
"epoch": 0.3676216775975147,
"grad_norm": 2.173220634460449,
"learning_rate": 2e-05,
"loss": 0.7615,
"step": 1065
},
{
"epoch": 0.36796686227131514,
"grad_norm": 2.0912346839904785,
"learning_rate": 2e-05,
"loss": 0.7531,
"step": 1066
},
{
"epoch": 0.36831204694511566,
"grad_norm": 2.115833044052124,
"learning_rate": 2e-05,
"loss": 0.7378,
"step": 1067
},
{
"epoch": 0.3686572316189161,
"grad_norm": 2.617403984069824,
"learning_rate": 2e-05,
"loss": 0.7591,
"step": 1068
},
{
"epoch": 0.36900241629271663,
"grad_norm": 2.2421016693115234,
"learning_rate": 2e-05,
"loss": 0.8123,
"step": 1069
},
{
"epoch": 0.3693476009665171,
"grad_norm": 2.042734384536743,
"learning_rate": 2e-05,
"loss": 0.7462,
"step": 1070
},
{
"epoch": 0.36969278564031754,
"grad_norm": 1.985018253326416,
"learning_rate": 2e-05,
"loss": 0.7905,
"step": 1071
},
{
"epoch": 0.37003797031411806,
"grad_norm": 2.3043973445892334,
"learning_rate": 2e-05,
"loss": 0.7785,
"step": 1072
},
{
"epoch": 0.3703831549879185,
"grad_norm": 2.2626423835754395,
"learning_rate": 2e-05,
"loss": 0.7671,
"step": 1073
},
{
"epoch": 0.37072833966171903,
"grad_norm": 2.274169921875,
"learning_rate": 2e-05,
"loss": 0.7971,
"step": 1074
},
{
"epoch": 0.3710735243355195,
"grad_norm": 2.2535250186920166,
"learning_rate": 2e-05,
"loss": 0.826,
"step": 1075
},
{
"epoch": 0.37141870900932,
"grad_norm": 2.3453760147094727,
"learning_rate": 2e-05,
"loss": 0.8127,
"step": 1076
},
{
"epoch": 0.37176389368312046,
"grad_norm": 2.190314531326294,
"learning_rate": 2e-05,
"loss": 0.7601,
"step": 1077
},
{
"epoch": 0.372109078356921,
"grad_norm": 2.257371187210083,
"learning_rate": 2e-05,
"loss": 0.7538,
"step": 1078
},
{
"epoch": 0.37245426303072143,
"grad_norm": 2.5159974098205566,
"learning_rate": 2e-05,
"loss": 0.7937,
"step": 1079
},
{
"epoch": 0.37279944770452195,
"grad_norm": 2.3142731189727783,
"learning_rate": 2e-05,
"loss": 0.8356,
"step": 1080
},
{
"epoch": 0.3731446323783224,
"grad_norm": 2.019498109817505,
"learning_rate": 2e-05,
"loss": 0.7626,
"step": 1081
},
{
"epoch": 0.37348981705212286,
"grad_norm": 2.466688394546509,
"learning_rate": 2e-05,
"loss": 0.7942,
"step": 1082
},
{
"epoch": 0.3738350017259234,
"grad_norm": 2.2081542015075684,
"learning_rate": 2e-05,
"loss": 0.7526,
"step": 1083
},
{
"epoch": 0.37418018639972384,
"grad_norm": 2.239201784133911,
"learning_rate": 2e-05,
"loss": 0.7974,
"step": 1084
},
{
"epoch": 0.37452537107352435,
"grad_norm": 2.036984443664551,
"learning_rate": 2e-05,
"loss": 0.6817,
"step": 1085
},
{
"epoch": 0.3748705557473248,
"grad_norm": 2.1549367904663086,
"learning_rate": 2e-05,
"loss": 0.7906,
"step": 1086
},
{
"epoch": 0.3752157404211253,
"grad_norm": 2.041508197784424,
"learning_rate": 2e-05,
"loss": 0.7032,
"step": 1087
},
{
"epoch": 0.3755609250949258,
"grad_norm": 2.2457003593444824,
"learning_rate": 2e-05,
"loss": 0.7936,
"step": 1088
},
{
"epoch": 0.3759061097687263,
"grad_norm": 2.1312613487243652,
"learning_rate": 2e-05,
"loss": 0.8029,
"step": 1089
},
{
"epoch": 0.37625129444252675,
"grad_norm": 2.1522438526153564,
"learning_rate": 2e-05,
"loss": 0.7228,
"step": 1090
},
{
"epoch": 0.3765964791163272,
"grad_norm": 2.3211474418640137,
"learning_rate": 2e-05,
"loss": 0.7812,
"step": 1091
},
{
"epoch": 0.3769416637901277,
"grad_norm": 2.1609480381011963,
"learning_rate": 2e-05,
"loss": 0.8076,
"step": 1092
},
{
"epoch": 0.3772868484639282,
"grad_norm": 2.3022890090942383,
"learning_rate": 2e-05,
"loss": 0.8061,
"step": 1093
},
{
"epoch": 0.3776320331377287,
"grad_norm": 2.2796778678894043,
"learning_rate": 2e-05,
"loss": 0.7379,
"step": 1094
},
{
"epoch": 0.37797721781152915,
"grad_norm": 2.533264636993408,
"learning_rate": 2e-05,
"loss": 0.7544,
"step": 1095
},
{
"epoch": 0.37832240248532967,
"grad_norm": 2.397287130355835,
"learning_rate": 2e-05,
"loss": 0.7679,
"step": 1096
},
{
"epoch": 0.3786675871591301,
"grad_norm": 2.5518314838409424,
"learning_rate": 2e-05,
"loss": 0.8307,
"step": 1097
},
{
"epoch": 0.37901277183293064,
"grad_norm": 2.3980071544647217,
"learning_rate": 2e-05,
"loss": 0.8259,
"step": 1098
},
{
"epoch": 0.3793579565067311,
"grad_norm": 2.284888744354248,
"learning_rate": 2e-05,
"loss": 0.8307,
"step": 1099
},
{
"epoch": 0.3797031411805316,
"grad_norm": 2.1695706844329834,
"learning_rate": 2e-05,
"loss": 0.733,
"step": 1100
},
{
"epoch": 0.38004832585433207,
"grad_norm": 2.2246172428131104,
"learning_rate": 2e-05,
"loss": 0.7539,
"step": 1101
},
{
"epoch": 0.38039351052813253,
"grad_norm": 2.2581632137298584,
"learning_rate": 2e-05,
"loss": 0.7748,
"step": 1102
},
{
"epoch": 0.38073869520193304,
"grad_norm": 2.126983165740967,
"learning_rate": 2e-05,
"loss": 0.7539,
"step": 1103
},
{
"epoch": 0.3810838798757335,
"grad_norm": 2.453768491744995,
"learning_rate": 2e-05,
"loss": 0.7992,
"step": 1104
},
{
"epoch": 0.381429064549534,
"grad_norm": 2.154642343521118,
"learning_rate": 2e-05,
"loss": 0.7887,
"step": 1105
},
{
"epoch": 0.38177424922333447,
"grad_norm": 2.197160482406616,
"learning_rate": 2e-05,
"loss": 0.7536,
"step": 1106
},
{
"epoch": 0.382119433897135,
"grad_norm": 2.257078170776367,
"learning_rate": 2e-05,
"loss": 0.7902,
"step": 1107
},
{
"epoch": 0.38246461857093544,
"grad_norm": 2.5250813961029053,
"learning_rate": 2e-05,
"loss": 0.8121,
"step": 1108
},
{
"epoch": 0.38280980324473596,
"grad_norm": 2.095309019088745,
"learning_rate": 2e-05,
"loss": 0.7935,
"step": 1109
},
{
"epoch": 0.3831549879185364,
"grad_norm": 2.176626682281494,
"learning_rate": 2e-05,
"loss": 0.7112,
"step": 1110
},
{
"epoch": 0.3835001725923369,
"grad_norm": 2.3763976097106934,
"learning_rate": 2e-05,
"loss": 0.7353,
"step": 1111
},
{
"epoch": 0.3838453572661374,
"grad_norm": 2.5461270809173584,
"learning_rate": 2e-05,
"loss": 0.7264,
"step": 1112
},
{
"epoch": 0.38419054193993785,
"grad_norm": 2.195230007171631,
"learning_rate": 2e-05,
"loss": 0.7596,
"step": 1113
},
{
"epoch": 0.38453572661373836,
"grad_norm": 2.244112253189087,
"learning_rate": 2e-05,
"loss": 0.7719,
"step": 1114
},
{
"epoch": 0.3848809112875388,
"grad_norm": 2.1582376956939697,
"learning_rate": 2e-05,
"loss": 0.7692,
"step": 1115
},
{
"epoch": 0.38522609596133933,
"grad_norm": 2.161746025085449,
"learning_rate": 2e-05,
"loss": 0.8285,
"step": 1116
},
{
"epoch": 0.3855712806351398,
"grad_norm": 2.000035047531128,
"learning_rate": 2e-05,
"loss": 0.7802,
"step": 1117
},
{
"epoch": 0.3859164653089403,
"grad_norm": 2.102853298187256,
"learning_rate": 2e-05,
"loss": 0.7239,
"step": 1118
},
{
"epoch": 0.38626164998274076,
"grad_norm": 2.521127223968506,
"learning_rate": 2e-05,
"loss": 0.8367,
"step": 1119
},
{
"epoch": 0.3866068346565413,
"grad_norm": 2.2251970767974854,
"learning_rate": 2e-05,
"loss": 0.7663,
"step": 1120
},
{
"epoch": 0.38695201933034173,
"grad_norm": 2.0934650897979736,
"learning_rate": 2e-05,
"loss": 0.7846,
"step": 1121
},
{
"epoch": 0.3872972040041422,
"grad_norm": 2.161764621734619,
"learning_rate": 2e-05,
"loss": 0.8228,
"step": 1122
},
{
"epoch": 0.3876423886779427,
"grad_norm": 2.1264915466308594,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 1123
},
{
"epoch": 0.38798757335174316,
"grad_norm": 2.1243889331817627,
"learning_rate": 2e-05,
"loss": 0.767,
"step": 1124
},
{
"epoch": 0.3883327580255437,
"grad_norm": 2.3819096088409424,
"learning_rate": 2e-05,
"loss": 0.8094,
"step": 1125
},
{
"epoch": 0.38867794269934414,
"grad_norm": 2.246093988418579,
"learning_rate": 2e-05,
"loss": 0.8032,
"step": 1126
},
{
"epoch": 0.38902312737314465,
"grad_norm": 2.079819679260254,
"learning_rate": 2e-05,
"loss": 0.7928,
"step": 1127
},
{
"epoch": 0.3893683120469451,
"grad_norm": 2.060594320297241,
"learning_rate": 2e-05,
"loss": 0.8272,
"step": 1128
},
{
"epoch": 0.3897134967207456,
"grad_norm": 2.432152509689331,
"learning_rate": 2e-05,
"loss": 0.827,
"step": 1129
},
{
"epoch": 0.3900586813945461,
"grad_norm": 1.996207594871521,
"learning_rate": 2e-05,
"loss": 0.7176,
"step": 1130
},
{
"epoch": 0.39040386606834654,
"grad_norm": 2.36133074760437,
"learning_rate": 2e-05,
"loss": 0.8379,
"step": 1131
},
{
"epoch": 0.39074905074214705,
"grad_norm": 2.1505513191223145,
"learning_rate": 2e-05,
"loss": 0.7577,
"step": 1132
},
{
"epoch": 0.3910942354159475,
"grad_norm": 2.0275981426239014,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 1133
},
{
"epoch": 0.391439420089748,
"grad_norm": 2.166504383087158,
"learning_rate": 2e-05,
"loss": 0.7435,
"step": 1134
},
{
"epoch": 0.3917846047635485,
"grad_norm": 2.1447205543518066,
"learning_rate": 2e-05,
"loss": 0.7538,
"step": 1135
},
{
"epoch": 0.392129789437349,
"grad_norm": 2.2855634689331055,
"learning_rate": 2e-05,
"loss": 0.7259,
"step": 1136
},
{
"epoch": 0.39247497411114945,
"grad_norm": 2.1907403469085693,
"learning_rate": 2e-05,
"loss": 0.7442,
"step": 1137
},
{
"epoch": 0.39282015878494997,
"grad_norm": 1.9817599058151245,
"learning_rate": 2e-05,
"loss": 0.7869,
"step": 1138
},
{
"epoch": 0.3931653434587504,
"grad_norm": 2.035456657409668,
"learning_rate": 2e-05,
"loss": 0.7655,
"step": 1139
},
{
"epoch": 0.39351052813255094,
"grad_norm": 1.9822310209274292,
"learning_rate": 2e-05,
"loss": 0.7048,
"step": 1140
},
{
"epoch": 0.3938557128063514,
"grad_norm": 2.0834767818450928,
"learning_rate": 2e-05,
"loss": 0.7661,
"step": 1141
},
{
"epoch": 0.39420089748015186,
"grad_norm": 2.2362561225891113,
"learning_rate": 2e-05,
"loss": 0.7165,
"step": 1142
},
{
"epoch": 0.39454608215395237,
"grad_norm": 2.341055154800415,
"learning_rate": 2e-05,
"loss": 0.829,
"step": 1143
},
{
"epoch": 0.39489126682775283,
"grad_norm": 2.3132922649383545,
"learning_rate": 2e-05,
"loss": 0.8019,
"step": 1144
},
{
"epoch": 0.39523645150155334,
"grad_norm": 2.2544994354248047,
"learning_rate": 2e-05,
"loss": 0.785,
"step": 1145
},
{
"epoch": 0.3955816361753538,
"grad_norm": 2.6545863151550293,
"learning_rate": 2e-05,
"loss": 0.7846,
"step": 1146
},
{
"epoch": 0.3959268208491543,
"grad_norm": 2.2618069648742676,
"learning_rate": 2e-05,
"loss": 0.7391,
"step": 1147
},
{
"epoch": 0.3962720055229548,
"grad_norm": 2.2470247745513916,
"learning_rate": 2e-05,
"loss": 0.7667,
"step": 1148
},
{
"epoch": 0.3966171901967553,
"grad_norm": 2.351428985595703,
"learning_rate": 2e-05,
"loss": 0.8479,
"step": 1149
},
{
"epoch": 0.39696237487055575,
"grad_norm": 2.3018267154693604,
"learning_rate": 2e-05,
"loss": 0.8787,
"step": 1150
},
{
"epoch": 0.3973075595443562,
"grad_norm": 2.1409080028533936,
"learning_rate": 2e-05,
"loss": 0.7833,
"step": 1151
},
{
"epoch": 0.3976527442181567,
"grad_norm": 2.0390005111694336,
"learning_rate": 2e-05,
"loss": 0.7387,
"step": 1152
},
{
"epoch": 0.3979979288919572,
"grad_norm": 2.13079571723938,
"learning_rate": 2e-05,
"loss": 0.7552,
"step": 1153
},
{
"epoch": 0.3983431135657577,
"grad_norm": 2.46516489982605,
"learning_rate": 2e-05,
"loss": 0.788,
"step": 1154
},
{
"epoch": 0.39868829823955815,
"grad_norm": 2.123201608657837,
"learning_rate": 2e-05,
"loss": 0.744,
"step": 1155
},
{
"epoch": 0.39903348291335866,
"grad_norm": 2.0077736377716064,
"learning_rate": 2e-05,
"loss": 0.785,
"step": 1156
},
{
"epoch": 0.3993786675871591,
"grad_norm": 2.1959493160247803,
"learning_rate": 2e-05,
"loss": 0.673,
"step": 1157
},
{
"epoch": 0.39972385226095963,
"grad_norm": 2.1401984691619873,
"learning_rate": 2e-05,
"loss": 0.7695,
"step": 1158
},
{
"epoch": 0.4000690369347601,
"grad_norm": 2.2160251140594482,
"learning_rate": 2e-05,
"loss": 0.7764,
"step": 1159
},
{
"epoch": 0.4004142216085606,
"grad_norm": 2.3270435333251953,
"learning_rate": 2e-05,
"loss": 0.7837,
"step": 1160
},
{
"epoch": 0.40075940628236106,
"grad_norm": 2.262467861175537,
"learning_rate": 2e-05,
"loss": 0.7679,
"step": 1161
},
{
"epoch": 0.4011045909561615,
"grad_norm": 2.1126415729522705,
"learning_rate": 2e-05,
"loss": 0.7412,
"step": 1162
},
{
"epoch": 0.40144977562996204,
"grad_norm": 2.132660388946533,
"learning_rate": 2e-05,
"loss": 0.6676,
"step": 1163
},
{
"epoch": 0.4017949603037625,
"grad_norm": 2.5451934337615967,
"learning_rate": 2e-05,
"loss": 0.7682,
"step": 1164
},
{
"epoch": 0.402140144977563,
"grad_norm": 2.107463836669922,
"learning_rate": 2e-05,
"loss": 0.828,
"step": 1165
},
{
"epoch": 0.40248532965136347,
"grad_norm": 2.0670745372772217,
"learning_rate": 2e-05,
"loss": 0.7242,
"step": 1166
},
{
"epoch": 0.402830514325164,
"grad_norm": 2.237152099609375,
"learning_rate": 2e-05,
"loss": 0.7851,
"step": 1167
},
{
"epoch": 0.40317569899896444,
"grad_norm": 2.1627330780029297,
"learning_rate": 2e-05,
"loss": 0.8038,
"step": 1168
},
{
"epoch": 0.40352088367276495,
"grad_norm": 2.510139226913452,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 1169
},
{
"epoch": 0.4038660683465654,
"grad_norm": 2.369367837905884,
"learning_rate": 2e-05,
"loss": 0.791,
"step": 1170
},
{
"epoch": 0.40421125302036587,
"grad_norm": 2.2461135387420654,
"learning_rate": 2e-05,
"loss": 0.7888,
"step": 1171
},
{
"epoch": 0.4045564376941664,
"grad_norm": 2.149815559387207,
"learning_rate": 2e-05,
"loss": 0.8311,
"step": 1172
},
{
"epoch": 0.40490162236796684,
"grad_norm": 2.065385580062866,
"learning_rate": 2e-05,
"loss": 0.8055,
"step": 1173
},
{
"epoch": 0.40524680704176735,
"grad_norm": 2.0966150760650635,
"learning_rate": 2e-05,
"loss": 0.7472,
"step": 1174
},
{
"epoch": 0.4055919917155678,
"grad_norm": 2.2043638229370117,
"learning_rate": 2e-05,
"loss": 0.7666,
"step": 1175
},
{
"epoch": 0.4059371763893683,
"grad_norm": 2.3187267780303955,
"learning_rate": 2e-05,
"loss": 0.7634,
"step": 1176
},
{
"epoch": 0.4062823610631688,
"grad_norm": 2.0239498615264893,
"learning_rate": 2e-05,
"loss": 0.7471,
"step": 1177
},
{
"epoch": 0.4066275457369693,
"grad_norm": 2.3455922603607178,
"learning_rate": 2e-05,
"loss": 0.8358,
"step": 1178
},
{
"epoch": 0.40697273041076976,
"grad_norm": 2.2640397548675537,
"learning_rate": 2e-05,
"loss": 0.7376,
"step": 1179
},
{
"epoch": 0.40731791508457027,
"grad_norm": 2.159149408340454,
"learning_rate": 2e-05,
"loss": 0.7787,
"step": 1180
},
{
"epoch": 0.40766309975837073,
"grad_norm": 2.2202343940734863,
"learning_rate": 2e-05,
"loss": 0.7735,
"step": 1181
},
{
"epoch": 0.4080082844321712,
"grad_norm": 2.2971279621124268,
"learning_rate": 2e-05,
"loss": 0.8021,
"step": 1182
},
{
"epoch": 0.4083534691059717,
"grad_norm": 2.1960690021514893,
"learning_rate": 2e-05,
"loss": 0.8189,
"step": 1183
},
{
"epoch": 0.40869865377977216,
"grad_norm": 2.331207036972046,
"learning_rate": 2e-05,
"loss": 0.7844,
"step": 1184
},
{
"epoch": 0.4090438384535727,
"grad_norm": 2.351132392883301,
"learning_rate": 2e-05,
"loss": 0.7574,
"step": 1185
},
{
"epoch": 0.40938902312737313,
"grad_norm": 2.302999258041382,
"learning_rate": 2e-05,
"loss": 0.7218,
"step": 1186
},
{
"epoch": 0.40973420780117364,
"grad_norm": 2.275294542312622,
"learning_rate": 2e-05,
"loss": 0.7474,
"step": 1187
},
{
"epoch": 0.4100793924749741,
"grad_norm": 2.5616424083709717,
"learning_rate": 2e-05,
"loss": 0.7795,
"step": 1188
},
{
"epoch": 0.4104245771487746,
"grad_norm": 2.325176477432251,
"learning_rate": 2e-05,
"loss": 0.8554,
"step": 1189
},
{
"epoch": 0.4107697618225751,
"grad_norm": 2.2610647678375244,
"learning_rate": 2e-05,
"loss": 0.7577,
"step": 1190
},
{
"epoch": 0.4111149464963756,
"grad_norm": 2.1423611640930176,
"learning_rate": 2e-05,
"loss": 0.8232,
"step": 1191
},
{
"epoch": 0.41146013117017605,
"grad_norm": 2.3171980381011963,
"learning_rate": 2e-05,
"loss": 0.7204,
"step": 1192
},
{
"epoch": 0.4118053158439765,
"grad_norm": 2.126913070678711,
"learning_rate": 2e-05,
"loss": 0.7934,
"step": 1193
},
{
"epoch": 0.412150500517777,
"grad_norm": 2.1600658893585205,
"learning_rate": 2e-05,
"loss": 0.8029,
"step": 1194
},
{
"epoch": 0.4124956851915775,
"grad_norm": 2.588409423828125,
"learning_rate": 2e-05,
"loss": 0.8909,
"step": 1195
},
{
"epoch": 0.412840869865378,
"grad_norm": 2.3043060302734375,
"learning_rate": 2e-05,
"loss": 0.8179,
"step": 1196
},
{
"epoch": 0.41318605453917845,
"grad_norm": 2.0503504276275635,
"learning_rate": 2e-05,
"loss": 0.7479,
"step": 1197
},
{
"epoch": 0.41353123921297896,
"grad_norm": 2.470106601715088,
"learning_rate": 2e-05,
"loss": 0.8424,
"step": 1198
},
{
"epoch": 0.4138764238867794,
"grad_norm": 1.9749798774719238,
"learning_rate": 2e-05,
"loss": 0.753,
"step": 1199
},
{
"epoch": 0.41422160856057993,
"grad_norm": 2.1416146755218506,
"learning_rate": 2e-05,
"loss": 0.7811,
"step": 1200
},
{
"epoch": 0.4145667932343804,
"grad_norm": 2.3665664196014404,
"learning_rate": 2e-05,
"loss": 0.78,
"step": 1201
},
{
"epoch": 0.41491197790818085,
"grad_norm": 2.5720670223236084,
"learning_rate": 2e-05,
"loss": 0.7601,
"step": 1202
},
{
"epoch": 0.41525716258198136,
"grad_norm": 2.119982957839966,
"learning_rate": 2e-05,
"loss": 0.7662,
"step": 1203
},
{
"epoch": 0.4156023472557818,
"grad_norm": 2.38934063911438,
"learning_rate": 2e-05,
"loss": 0.8291,
"step": 1204
},
{
"epoch": 0.41594753192958234,
"grad_norm": 1.9148893356323242,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 1205
},
{
"epoch": 0.4162927166033828,
"grad_norm": 2.151456594467163,
"learning_rate": 2e-05,
"loss": 0.731,
"step": 1206
},
{
"epoch": 0.4166379012771833,
"grad_norm": 2.110645294189453,
"learning_rate": 2e-05,
"loss": 0.827,
"step": 1207
},
{
"epoch": 0.41698308595098377,
"grad_norm": 2.2597014904022217,
"learning_rate": 2e-05,
"loss": 0.7929,
"step": 1208
},
{
"epoch": 0.4173282706247843,
"grad_norm": 2.021125316619873,
"learning_rate": 2e-05,
"loss": 0.8108,
"step": 1209
},
{
"epoch": 0.41767345529858474,
"grad_norm": 2.1065304279327393,
"learning_rate": 2e-05,
"loss": 0.7356,
"step": 1210
},
{
"epoch": 0.41801863997238525,
"grad_norm": 2.1394643783569336,
"learning_rate": 2e-05,
"loss": 0.7714,
"step": 1211
},
{
"epoch": 0.4183638246461857,
"grad_norm": 2.0147531032562256,
"learning_rate": 2e-05,
"loss": 0.7801,
"step": 1212
},
{
"epoch": 0.41870900931998617,
"grad_norm": 2.1060707569122314,
"learning_rate": 2e-05,
"loss": 0.7489,
"step": 1213
},
{
"epoch": 0.4190541939937867,
"grad_norm": 2.1238880157470703,
"learning_rate": 2e-05,
"loss": 0.7178,
"step": 1214
},
{
"epoch": 0.41939937866758714,
"grad_norm": 2.2426836490631104,
"learning_rate": 2e-05,
"loss": 0.7457,
"step": 1215
},
{
"epoch": 0.41974456334138766,
"grad_norm": 2.1819989681243896,
"learning_rate": 2e-05,
"loss": 0.7768,
"step": 1216
},
{
"epoch": 0.4200897480151881,
"grad_norm": 2.2598514556884766,
"learning_rate": 2e-05,
"loss": 0.804,
"step": 1217
},
{
"epoch": 0.4204349326889886,
"grad_norm": 2.0700066089630127,
"learning_rate": 2e-05,
"loss": 0.7692,
"step": 1218
},
{
"epoch": 0.4207801173627891,
"grad_norm": 2.286806106567383,
"learning_rate": 2e-05,
"loss": 0.8531,
"step": 1219
},
{
"epoch": 0.4211253020365896,
"grad_norm": 2.0580227375030518,
"learning_rate": 2e-05,
"loss": 0.7557,
"step": 1220
},
{
"epoch": 0.42147048671039006,
"grad_norm": 2.0104353427886963,
"learning_rate": 2e-05,
"loss": 0.8089,
"step": 1221
},
{
"epoch": 0.4218156713841905,
"grad_norm": 2.1290783882141113,
"learning_rate": 2e-05,
"loss": 0.7675,
"step": 1222
},
{
"epoch": 0.42216085605799103,
"grad_norm": 2.1737821102142334,
"learning_rate": 2e-05,
"loss": 0.8123,
"step": 1223
},
{
"epoch": 0.4225060407317915,
"grad_norm": 2.1916322708129883,
"learning_rate": 2e-05,
"loss": 0.7462,
"step": 1224
},
{
"epoch": 0.422851225405592,
"grad_norm": 2.368980884552002,
"learning_rate": 2e-05,
"loss": 0.7675,
"step": 1225
},
{
"epoch": 0.42319641007939246,
"grad_norm": 2.1589972972869873,
"learning_rate": 2e-05,
"loss": 0.7306,
"step": 1226
},
{
"epoch": 0.423541594753193,
"grad_norm": 1.9490817785263062,
"learning_rate": 2e-05,
"loss": 0.7259,
"step": 1227
},
{
"epoch": 0.42388677942699343,
"grad_norm": 2.2712414264678955,
"learning_rate": 2e-05,
"loss": 0.7923,
"step": 1228
},
{
"epoch": 0.42423196410079395,
"grad_norm": 2.217989206314087,
"learning_rate": 2e-05,
"loss": 0.7289,
"step": 1229
},
{
"epoch": 0.4245771487745944,
"grad_norm": 2.227036714553833,
"learning_rate": 2e-05,
"loss": 0.749,
"step": 1230
},
{
"epoch": 0.4249223334483949,
"grad_norm": 2.231045722961426,
"learning_rate": 2e-05,
"loss": 0.7524,
"step": 1231
},
{
"epoch": 0.4252675181221954,
"grad_norm": 9.880687713623047,
"learning_rate": 2e-05,
"loss": 0.8156,
"step": 1232
},
{
"epoch": 0.42561270279599583,
"grad_norm": 2.3977463245391846,
"learning_rate": 2e-05,
"loss": 0.7841,
"step": 1233
},
{
"epoch": 0.42595788746979635,
"grad_norm": 2.0663363933563232,
"learning_rate": 2e-05,
"loss": 0.7864,
"step": 1234
},
{
"epoch": 0.4263030721435968,
"grad_norm": 2.194828987121582,
"learning_rate": 2e-05,
"loss": 0.7467,
"step": 1235
},
{
"epoch": 0.4266482568173973,
"grad_norm": 2.170609951019287,
"learning_rate": 2e-05,
"loss": 0.8234,
"step": 1236
},
{
"epoch": 0.4269934414911978,
"grad_norm": 2.142153739929199,
"learning_rate": 2e-05,
"loss": 0.791,
"step": 1237
},
{
"epoch": 0.4273386261649983,
"grad_norm": 2.1600191593170166,
"learning_rate": 2e-05,
"loss": 0.7976,
"step": 1238
},
{
"epoch": 0.42768381083879875,
"grad_norm": 2.1493258476257324,
"learning_rate": 2e-05,
"loss": 0.8099,
"step": 1239
},
{
"epoch": 0.42802899551259926,
"grad_norm": 2.224118232727051,
"learning_rate": 2e-05,
"loss": 0.767,
"step": 1240
},
{
"epoch": 0.4283741801863997,
"grad_norm": 2.3489747047424316,
"learning_rate": 2e-05,
"loss": 0.7905,
"step": 1241
},
{
"epoch": 0.4287193648602002,
"grad_norm": 2.2355237007141113,
"learning_rate": 2e-05,
"loss": 0.7986,
"step": 1242
},
{
"epoch": 0.4290645495340007,
"grad_norm": 2.3177027702331543,
"learning_rate": 2e-05,
"loss": 0.7496,
"step": 1243
},
{
"epoch": 0.42940973420780115,
"grad_norm": 2.1514225006103516,
"learning_rate": 2e-05,
"loss": 0.7379,
"step": 1244
},
{
"epoch": 0.42975491888160167,
"grad_norm": 2.01289701461792,
"learning_rate": 2e-05,
"loss": 0.7784,
"step": 1245
},
{
"epoch": 0.4301001035554021,
"grad_norm": 2.025200128555298,
"learning_rate": 2e-05,
"loss": 0.7324,
"step": 1246
},
{
"epoch": 0.43044528822920264,
"grad_norm": 1.967621088027954,
"learning_rate": 2e-05,
"loss": 0.7406,
"step": 1247
},
{
"epoch": 0.4307904729030031,
"grad_norm": 2.3146955966949463,
"learning_rate": 2e-05,
"loss": 0.8037,
"step": 1248
},
{
"epoch": 0.4311356575768036,
"grad_norm": 2.5918831825256348,
"learning_rate": 2e-05,
"loss": 0.8372,
"step": 1249
},
{
"epoch": 0.43148084225060407,
"grad_norm": 2.410187244415283,
"learning_rate": 2e-05,
"loss": 0.7838,
"step": 1250
},
{
"epoch": 0.4318260269244046,
"grad_norm": 2.047809600830078,
"learning_rate": 2e-05,
"loss": 0.7852,
"step": 1251
},
{
"epoch": 0.43217121159820504,
"grad_norm": 2.1864817142486572,
"learning_rate": 2e-05,
"loss": 0.7094,
"step": 1252
},
{
"epoch": 0.4325163962720055,
"grad_norm": 2.3229713439941406,
"learning_rate": 2e-05,
"loss": 0.6908,
"step": 1253
},
{
"epoch": 0.432861580945806,
"grad_norm": 2.1283926963806152,
"learning_rate": 2e-05,
"loss": 0.7898,
"step": 1254
},
{
"epoch": 0.43320676561960647,
"grad_norm": 1.9980229139328003,
"learning_rate": 2e-05,
"loss": 0.7341,
"step": 1255
},
{
"epoch": 0.433551950293407,
"grad_norm": 1.9197133779525757,
"learning_rate": 2e-05,
"loss": 0.7348,
"step": 1256
},
{
"epoch": 0.43389713496720744,
"grad_norm": 2.0243780612945557,
"learning_rate": 2e-05,
"loss": 0.7016,
"step": 1257
},
{
"epoch": 0.43424231964100796,
"grad_norm": 2.0702686309814453,
"learning_rate": 2e-05,
"loss": 0.7458,
"step": 1258
},
{
"epoch": 0.4345875043148084,
"grad_norm": 2.033076763153076,
"learning_rate": 2e-05,
"loss": 0.7755,
"step": 1259
},
{
"epoch": 0.43493268898860893,
"grad_norm": 2.0470402240753174,
"learning_rate": 2e-05,
"loss": 0.8167,
"step": 1260
},
{
"epoch": 0.4352778736624094,
"grad_norm": 2.0547688007354736,
"learning_rate": 2e-05,
"loss": 0.7673,
"step": 1261
},
{
"epoch": 0.43562305833620985,
"grad_norm": 2.0047366619110107,
"learning_rate": 2e-05,
"loss": 0.7729,
"step": 1262
},
{
"epoch": 0.43596824301001036,
"grad_norm": 2.285064935684204,
"learning_rate": 2e-05,
"loss": 0.8687,
"step": 1263
},
{
"epoch": 0.4363134276838108,
"grad_norm": 2.1682817935943604,
"learning_rate": 2e-05,
"loss": 0.7681,
"step": 1264
},
{
"epoch": 0.43665861235761133,
"grad_norm": 2.421111583709717,
"learning_rate": 2e-05,
"loss": 0.7594,
"step": 1265
},
{
"epoch": 0.4370037970314118,
"grad_norm": 2.241191864013672,
"learning_rate": 2e-05,
"loss": 0.7542,
"step": 1266
},
{
"epoch": 0.4373489817052123,
"grad_norm": 1.9653596878051758,
"learning_rate": 2e-05,
"loss": 0.7453,
"step": 1267
},
{
"epoch": 0.43769416637901276,
"grad_norm": 2.1906819343566895,
"learning_rate": 2e-05,
"loss": 0.7871,
"step": 1268
},
{
"epoch": 0.4380393510528133,
"grad_norm": 2.295667886734009,
"learning_rate": 2e-05,
"loss": 0.758,
"step": 1269
},
{
"epoch": 0.43838453572661373,
"grad_norm": 2.139810562133789,
"learning_rate": 2e-05,
"loss": 0.7853,
"step": 1270
},
{
"epoch": 0.43872972040041425,
"grad_norm": 2.1784024238586426,
"learning_rate": 2e-05,
"loss": 0.7636,
"step": 1271
},
{
"epoch": 0.4390749050742147,
"grad_norm": 2.303664207458496,
"learning_rate": 2e-05,
"loss": 0.7863,
"step": 1272
},
{
"epoch": 0.43942008974801516,
"grad_norm": 2.2972357273101807,
"learning_rate": 2e-05,
"loss": 0.7367,
"step": 1273
},
{
"epoch": 0.4397652744218157,
"grad_norm": 2.179234504699707,
"learning_rate": 2e-05,
"loss": 0.726,
"step": 1274
},
{
"epoch": 0.44011045909561614,
"grad_norm": 2.252678632736206,
"learning_rate": 2e-05,
"loss": 0.8083,
"step": 1275
},
{
"epoch": 0.44045564376941665,
"grad_norm": 2.1799802780151367,
"learning_rate": 2e-05,
"loss": 0.7281,
"step": 1276
},
{
"epoch": 0.4408008284432171,
"grad_norm": 2.248781204223633,
"learning_rate": 2e-05,
"loss": 0.8112,
"step": 1277
},
{
"epoch": 0.4411460131170176,
"grad_norm": 2.0230889320373535,
"learning_rate": 2e-05,
"loss": 0.7564,
"step": 1278
},
{
"epoch": 0.4414911977908181,
"grad_norm": 2.2543110847473145,
"learning_rate": 2e-05,
"loss": 0.734,
"step": 1279
},
{
"epoch": 0.4418363824646186,
"grad_norm": 2.003783702850342,
"learning_rate": 2e-05,
"loss": 0.7496,
"step": 1280
},
{
"epoch": 0.44218156713841905,
"grad_norm": 2.2942395210266113,
"learning_rate": 2e-05,
"loss": 0.738,
"step": 1281
},
{
"epoch": 0.4425267518122195,
"grad_norm": 2.220264196395874,
"learning_rate": 2e-05,
"loss": 0.7518,
"step": 1282
},
{
"epoch": 0.44287193648602,
"grad_norm": 2.044330596923828,
"learning_rate": 2e-05,
"loss": 0.8117,
"step": 1283
},
{
"epoch": 0.4432171211598205,
"grad_norm": 1.9773883819580078,
"learning_rate": 2e-05,
"loss": 0.7442,
"step": 1284
},
{
"epoch": 0.443562305833621,
"grad_norm": 2.127779483795166,
"learning_rate": 2e-05,
"loss": 0.7243,
"step": 1285
},
{
"epoch": 0.44390749050742145,
"grad_norm": 2.048158645629883,
"learning_rate": 2e-05,
"loss": 0.7283,
"step": 1286
},
{
"epoch": 0.44425267518122197,
"grad_norm": 2.0748143196105957,
"learning_rate": 2e-05,
"loss": 0.75,
"step": 1287
},
{
"epoch": 0.4445978598550224,
"grad_norm": 2.017326831817627,
"learning_rate": 2e-05,
"loss": 0.7573,
"step": 1288
},
{
"epoch": 0.44494304452882294,
"grad_norm": 2.0272696018218994,
"learning_rate": 2e-05,
"loss": 0.7682,
"step": 1289
},
{
"epoch": 0.4452882292026234,
"grad_norm": 2.270801305770874,
"learning_rate": 2e-05,
"loss": 0.8238,
"step": 1290
},
{
"epoch": 0.4456334138764239,
"grad_norm": 2.1245110034942627,
"learning_rate": 2e-05,
"loss": 0.7319,
"step": 1291
},
{
"epoch": 0.44597859855022437,
"grad_norm": 2.217327356338501,
"learning_rate": 2e-05,
"loss": 0.7684,
"step": 1292
},
{
"epoch": 0.44632378322402483,
"grad_norm": 1.9212812185287476,
"learning_rate": 2e-05,
"loss": 0.7474,
"step": 1293
},
{
"epoch": 0.44666896789782534,
"grad_norm": 2.145240545272827,
"learning_rate": 2e-05,
"loss": 0.7361,
"step": 1294
},
{
"epoch": 0.4470141525716258,
"grad_norm": 2.1934456825256348,
"learning_rate": 2e-05,
"loss": 0.7836,
"step": 1295
},
{
"epoch": 0.4473593372454263,
"grad_norm": 2.0999088287353516,
"learning_rate": 2e-05,
"loss": 0.7297,
"step": 1296
},
{
"epoch": 0.44770452191922677,
"grad_norm": 2.095811605453491,
"learning_rate": 2e-05,
"loss": 0.7369,
"step": 1297
},
{
"epoch": 0.4480497065930273,
"grad_norm": 2.0320262908935547,
"learning_rate": 2e-05,
"loss": 0.684,
"step": 1298
},
{
"epoch": 0.44839489126682774,
"grad_norm": 2.2569198608398438,
"learning_rate": 2e-05,
"loss": 0.7681,
"step": 1299
},
{
"epoch": 0.44874007594062826,
"grad_norm": 2.1936898231506348,
"learning_rate": 2e-05,
"loss": 0.7121,
"step": 1300
},
{
"epoch": 0.4490852606144287,
"grad_norm": 2.2229762077331543,
"learning_rate": 2e-05,
"loss": 0.7721,
"step": 1301
},
{
"epoch": 0.44943044528822923,
"grad_norm": 2.2748780250549316,
"learning_rate": 2e-05,
"loss": 0.8173,
"step": 1302
},
{
"epoch": 0.4497756299620297,
"grad_norm": 2.2365429401397705,
"learning_rate": 2e-05,
"loss": 0.7877,
"step": 1303
},
{
"epoch": 0.45012081463583015,
"grad_norm": 2.1983840465545654,
"learning_rate": 2e-05,
"loss": 0.8239,
"step": 1304
},
{
"epoch": 0.45046599930963066,
"grad_norm": 3.0414493083953857,
"learning_rate": 2e-05,
"loss": 0.7786,
"step": 1305
},
{
"epoch": 0.4508111839834311,
"grad_norm": 2.2142982482910156,
"learning_rate": 2e-05,
"loss": 0.6341,
"step": 1306
},
{
"epoch": 0.45115636865723163,
"grad_norm": 2.180875778198242,
"learning_rate": 2e-05,
"loss": 0.7449,
"step": 1307
},
{
"epoch": 0.4515015533310321,
"grad_norm": 1.9446709156036377,
"learning_rate": 2e-05,
"loss": 0.7782,
"step": 1308
},
{
"epoch": 0.4518467380048326,
"grad_norm": 2.031499147415161,
"learning_rate": 2e-05,
"loss": 0.7763,
"step": 1309
},
{
"epoch": 0.45219192267863306,
"grad_norm": 2.098013162612915,
"learning_rate": 2e-05,
"loss": 0.748,
"step": 1310
},
{
"epoch": 0.4525371073524336,
"grad_norm": 2.250537395477295,
"learning_rate": 2e-05,
"loss": 0.8191,
"step": 1311
},
{
"epoch": 0.45288229202623403,
"grad_norm": 1.994511365890503,
"learning_rate": 2e-05,
"loss": 0.7688,
"step": 1312
},
{
"epoch": 0.4532274767000345,
"grad_norm": 2.2545230388641357,
"learning_rate": 2e-05,
"loss": 0.7746,
"step": 1313
},
{
"epoch": 0.453572661373835,
"grad_norm": 2.0760281085968018,
"learning_rate": 2e-05,
"loss": 0.7745,
"step": 1314
},
{
"epoch": 0.45391784604763546,
"grad_norm": 1.8348346948623657,
"learning_rate": 2e-05,
"loss": 0.7273,
"step": 1315
},
{
"epoch": 0.454263030721436,
"grad_norm": 2.121410369873047,
"learning_rate": 2e-05,
"loss": 0.6916,
"step": 1316
},
{
"epoch": 0.45460821539523644,
"grad_norm": 2.1730988025665283,
"learning_rate": 2e-05,
"loss": 0.8226,
"step": 1317
},
{
"epoch": 0.45495340006903695,
"grad_norm": 2.1960904598236084,
"learning_rate": 2e-05,
"loss": 0.817,
"step": 1318
},
{
"epoch": 0.4552985847428374,
"grad_norm": 2.0488274097442627,
"learning_rate": 2e-05,
"loss": 0.7877,
"step": 1319
},
{
"epoch": 0.4556437694166379,
"grad_norm": 2.103205442428589,
"learning_rate": 2e-05,
"loss": 0.7825,
"step": 1320
},
{
"epoch": 0.4559889540904384,
"grad_norm": 2.1127564907073975,
"learning_rate": 2e-05,
"loss": 0.7553,
"step": 1321
},
{
"epoch": 0.4563341387642389,
"grad_norm": 2.159409284591675,
"learning_rate": 2e-05,
"loss": 0.7428,
"step": 1322
},
{
"epoch": 0.45667932343803935,
"grad_norm": 2.2200045585632324,
"learning_rate": 2e-05,
"loss": 0.7883,
"step": 1323
},
{
"epoch": 0.4570245081118398,
"grad_norm": 1.9319753646850586,
"learning_rate": 2e-05,
"loss": 0.7498,
"step": 1324
},
{
"epoch": 0.4573696927856403,
"grad_norm": 2.16727614402771,
"learning_rate": 2e-05,
"loss": 0.752,
"step": 1325
},
{
"epoch": 0.4577148774594408,
"grad_norm": 2.146498680114746,
"learning_rate": 2e-05,
"loss": 0.7823,
"step": 1326
},
{
"epoch": 0.4580600621332413,
"grad_norm": 2.032365560531616,
"learning_rate": 2e-05,
"loss": 0.7309,
"step": 1327
},
{
"epoch": 0.45840524680704176,
"grad_norm": 2.152923345565796,
"learning_rate": 2e-05,
"loss": 0.7884,
"step": 1328
},
{
"epoch": 0.45875043148084227,
"grad_norm": 2.1290228366851807,
"learning_rate": 2e-05,
"loss": 0.7409,
"step": 1329
},
{
"epoch": 0.4590956161546427,
"grad_norm": 2.1132442951202393,
"learning_rate": 2e-05,
"loss": 0.7564,
"step": 1330
},
{
"epoch": 0.45944080082844324,
"grad_norm": 2.1418418884277344,
"learning_rate": 2e-05,
"loss": 0.7248,
"step": 1331
},
{
"epoch": 0.4597859855022437,
"grad_norm": 2.118337631225586,
"learning_rate": 2e-05,
"loss": 0.6854,
"step": 1332
},
{
"epoch": 0.46013117017604416,
"grad_norm": 2.06254506111145,
"learning_rate": 2e-05,
"loss": 0.7487,
"step": 1333
},
{
"epoch": 0.46047635484984467,
"grad_norm": 2.0143659114837646,
"learning_rate": 2e-05,
"loss": 0.7791,
"step": 1334
},
{
"epoch": 0.46082153952364513,
"grad_norm": 2.177813768386841,
"learning_rate": 2e-05,
"loss": 0.8241,
"step": 1335
},
{
"epoch": 0.46116672419744564,
"grad_norm": 2.22977876663208,
"learning_rate": 2e-05,
"loss": 0.743,
"step": 1336
},
{
"epoch": 0.4615119088712461,
"grad_norm": 2.1962244510650635,
"learning_rate": 2e-05,
"loss": 0.733,
"step": 1337
},
{
"epoch": 0.4618570935450466,
"grad_norm": 2.049772024154663,
"learning_rate": 2e-05,
"loss": 0.7629,
"step": 1338
},
{
"epoch": 0.4622022782188471,
"grad_norm": 2.1585311889648438,
"learning_rate": 2e-05,
"loss": 0.6975,
"step": 1339
},
{
"epoch": 0.4625474628926476,
"grad_norm": 1.9191311597824097,
"learning_rate": 2e-05,
"loss": 0.7413,
"step": 1340
},
{
"epoch": 0.46289264756644805,
"grad_norm": 1.9648008346557617,
"learning_rate": 2e-05,
"loss": 0.7347,
"step": 1341
},
{
"epoch": 0.46323783224024856,
"grad_norm": 2.122073173522949,
"learning_rate": 2e-05,
"loss": 0.7718,
"step": 1342
},
{
"epoch": 0.463583016914049,
"grad_norm": 2.2762906551361084,
"learning_rate": 2e-05,
"loss": 0.7777,
"step": 1343
},
{
"epoch": 0.4639282015878495,
"grad_norm": 2.120659351348877,
"learning_rate": 2e-05,
"loss": 0.8117,
"step": 1344
},
{
"epoch": 0.46427338626165,
"grad_norm": 3.4558870792388916,
"learning_rate": 2e-05,
"loss": 0.7135,
"step": 1345
},
{
"epoch": 0.46461857093545045,
"grad_norm": 2.2689363956451416,
"learning_rate": 2e-05,
"loss": 0.779,
"step": 1346
},
{
"epoch": 0.46496375560925096,
"grad_norm": 2.4292638301849365,
"learning_rate": 2e-05,
"loss": 0.7371,
"step": 1347
},
{
"epoch": 0.4653089402830514,
"grad_norm": 2.0308992862701416,
"learning_rate": 2e-05,
"loss": 0.7341,
"step": 1348
},
{
"epoch": 0.46565412495685193,
"grad_norm": 2.185943841934204,
"learning_rate": 2e-05,
"loss": 0.7948,
"step": 1349
},
{
"epoch": 0.4659993096306524,
"grad_norm": 1.9282371997833252,
"learning_rate": 2e-05,
"loss": 0.715,
"step": 1350
},
{
"epoch": 0.4663444943044529,
"grad_norm": 1.8744617700576782,
"learning_rate": 2e-05,
"loss": 0.7102,
"step": 1351
},
{
"epoch": 0.46668967897825336,
"grad_norm": 1.971297025680542,
"learning_rate": 2e-05,
"loss": 0.7981,
"step": 1352
},
{
"epoch": 0.4670348636520538,
"grad_norm": 2.6090900897979736,
"learning_rate": 2e-05,
"loss": 0.7493,
"step": 1353
},
{
"epoch": 0.46738004832585434,
"grad_norm": 2.214540958404541,
"learning_rate": 2e-05,
"loss": 0.759,
"step": 1354
},
{
"epoch": 0.4677252329996548,
"grad_norm": 2.2588086128234863,
"learning_rate": 2e-05,
"loss": 0.8032,
"step": 1355
},
{
"epoch": 0.4680704176734553,
"grad_norm": 2.1242079734802246,
"learning_rate": 2e-05,
"loss": 0.7584,
"step": 1356
},
{
"epoch": 0.46841560234725577,
"grad_norm": 2.1841304302215576,
"learning_rate": 2e-05,
"loss": 0.8329,
"step": 1357
},
{
"epoch": 0.4687607870210563,
"grad_norm": 2.0054054260253906,
"learning_rate": 2e-05,
"loss": 0.8105,
"step": 1358
},
{
"epoch": 0.46910597169485674,
"grad_norm": 1.9585438966751099,
"learning_rate": 2e-05,
"loss": 0.7541,
"step": 1359
},
{
"epoch": 0.46945115636865725,
"grad_norm": 2.04925799369812,
"learning_rate": 2e-05,
"loss": 0.6978,
"step": 1360
},
{
"epoch": 0.4697963410424577,
"grad_norm": 2.042327404022217,
"learning_rate": 2e-05,
"loss": 0.7302,
"step": 1361
},
{
"epoch": 0.4701415257162582,
"grad_norm": 2.0400002002716064,
"learning_rate": 2e-05,
"loss": 0.7405,
"step": 1362
},
{
"epoch": 0.4704867103900587,
"grad_norm": 2.0755996704101562,
"learning_rate": 2e-05,
"loss": 0.764,
"step": 1363
},
{
"epoch": 0.47083189506385914,
"grad_norm": 2.2529706954956055,
"learning_rate": 2e-05,
"loss": 0.8228,
"step": 1364
},
{
"epoch": 0.47117707973765965,
"grad_norm": 2.171063184738159,
"learning_rate": 2e-05,
"loss": 0.7693,
"step": 1365
},
{
"epoch": 0.4715222644114601,
"grad_norm": 2.0561063289642334,
"learning_rate": 2e-05,
"loss": 0.794,
"step": 1366
},
{
"epoch": 0.4718674490852606,
"grad_norm": 1.916958212852478,
"learning_rate": 2e-05,
"loss": 0.6832,
"step": 1367
},
{
"epoch": 0.4722126337590611,
"grad_norm": 1.9005790948867798,
"learning_rate": 2e-05,
"loss": 0.7158,
"step": 1368
},
{
"epoch": 0.4725578184328616,
"grad_norm": 2.0838615894317627,
"learning_rate": 2e-05,
"loss": 0.7773,
"step": 1369
},
{
"epoch": 0.47290300310666206,
"grad_norm": 2.0513100624084473,
"learning_rate": 2e-05,
"loss": 0.7767,
"step": 1370
},
{
"epoch": 0.47324818778046257,
"grad_norm": 2.0236966609954834,
"learning_rate": 2e-05,
"loss": 0.7495,
"step": 1371
},
{
"epoch": 0.47359337245426303,
"grad_norm": 2.035886764526367,
"learning_rate": 2e-05,
"loss": 0.7004,
"step": 1372
},
{
"epoch": 0.4739385571280635,
"grad_norm": 2.147725820541382,
"learning_rate": 2e-05,
"loss": 0.783,
"step": 1373
},
{
"epoch": 0.474283741801864,
"grad_norm": 1.9615451097488403,
"learning_rate": 2e-05,
"loss": 0.7125,
"step": 1374
},
{
"epoch": 0.47462892647566446,
"grad_norm": 2.1239616870880127,
"learning_rate": 2e-05,
"loss": 0.7035,
"step": 1375
},
{
"epoch": 0.474974111149465,
"grad_norm": 1.981346845626831,
"learning_rate": 2e-05,
"loss": 0.802,
"step": 1376
},
{
"epoch": 0.47531929582326543,
"grad_norm": 1.9348574876785278,
"learning_rate": 2e-05,
"loss": 0.7443,
"step": 1377
},
{
"epoch": 0.47566448049706594,
"grad_norm": 2.3827192783355713,
"learning_rate": 2e-05,
"loss": 0.809,
"step": 1378
},
{
"epoch": 0.4760096651708664,
"grad_norm": 1.9881306886672974,
"learning_rate": 2e-05,
"loss": 0.7656,
"step": 1379
},
{
"epoch": 0.4763548498446669,
"grad_norm": 2.1034233570098877,
"learning_rate": 2e-05,
"loss": 0.7444,
"step": 1380
},
{
"epoch": 0.4767000345184674,
"grad_norm": 1.9430773258209229,
"learning_rate": 2e-05,
"loss": 0.7302,
"step": 1381
},
{
"epoch": 0.4770452191922679,
"grad_norm": 1.9969552755355835,
"learning_rate": 2e-05,
"loss": 0.8142,
"step": 1382
},
{
"epoch": 0.47739040386606835,
"grad_norm": 2.323213577270508,
"learning_rate": 2e-05,
"loss": 0.8382,
"step": 1383
},
{
"epoch": 0.4777355885398688,
"grad_norm": 2.0729734897613525,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 1384
},
{
"epoch": 0.4780807732136693,
"grad_norm": 2.2260313034057617,
"learning_rate": 2e-05,
"loss": 0.7522,
"step": 1385
},
{
"epoch": 0.4784259578874698,
"grad_norm": 2.1217124462127686,
"learning_rate": 2e-05,
"loss": 0.7385,
"step": 1386
},
{
"epoch": 0.4787711425612703,
"grad_norm": 2.0635108947753906,
"learning_rate": 2e-05,
"loss": 0.7538,
"step": 1387
},
{
"epoch": 0.47911632723507075,
"grad_norm": 2.06231427192688,
"learning_rate": 2e-05,
"loss": 0.7792,
"step": 1388
},
{
"epoch": 0.47946151190887126,
"grad_norm": 1.9126551151275635,
"learning_rate": 2e-05,
"loss": 0.7306,
"step": 1389
},
{
"epoch": 0.4798066965826717,
"grad_norm": 2.001126527786255,
"learning_rate": 2e-05,
"loss": 0.8227,
"step": 1390
},
{
"epoch": 0.48015188125647223,
"grad_norm": 1.9141910076141357,
"learning_rate": 2e-05,
"loss": 0.6806,
"step": 1391
},
{
"epoch": 0.4804970659302727,
"grad_norm": 2.301952838897705,
"learning_rate": 2e-05,
"loss": 0.7867,
"step": 1392
},
{
"epoch": 0.48084225060407315,
"grad_norm": 2.123697519302368,
"learning_rate": 2e-05,
"loss": 0.7696,
"step": 1393
},
{
"epoch": 0.48118743527787367,
"grad_norm": 1.9873336553573608,
"learning_rate": 2e-05,
"loss": 0.8089,
"step": 1394
},
{
"epoch": 0.4815326199516741,
"grad_norm": 2.0276663303375244,
"learning_rate": 2e-05,
"loss": 0.7757,
"step": 1395
},
{
"epoch": 0.48187780462547464,
"grad_norm": 1.899289608001709,
"learning_rate": 2e-05,
"loss": 0.6837,
"step": 1396
},
{
"epoch": 0.4822229892992751,
"grad_norm": 2.0093820095062256,
"learning_rate": 2e-05,
"loss": 0.7561,
"step": 1397
},
{
"epoch": 0.4825681739730756,
"grad_norm": 2.130183458328247,
"learning_rate": 2e-05,
"loss": 0.7683,
"step": 1398
},
{
"epoch": 0.48291335864687607,
"grad_norm": 2.0682289600372314,
"learning_rate": 2e-05,
"loss": 0.7906,
"step": 1399
},
{
"epoch": 0.4832585433206766,
"grad_norm": 2.005575656890869,
"learning_rate": 2e-05,
"loss": 0.8102,
"step": 1400
},
{
"epoch": 0.48360372799447704,
"grad_norm": 2.1126019954681396,
"learning_rate": 2e-05,
"loss": 0.7492,
"step": 1401
},
{
"epoch": 0.48394891266827755,
"grad_norm": 2.0274546146392822,
"learning_rate": 2e-05,
"loss": 0.7258,
"step": 1402
},
{
"epoch": 0.484294097342078,
"grad_norm": 2.07676100730896,
"learning_rate": 2e-05,
"loss": 0.7311,
"step": 1403
},
{
"epoch": 0.48463928201587847,
"grad_norm": 1.912497878074646,
"learning_rate": 2e-05,
"loss": 0.7401,
"step": 1404
},
{
"epoch": 0.484984466689679,
"grad_norm": 2.320788621902466,
"learning_rate": 2e-05,
"loss": 0.7862,
"step": 1405
},
{
"epoch": 0.48532965136347944,
"grad_norm": 2.176485061645508,
"learning_rate": 2e-05,
"loss": 0.8005,
"step": 1406
},
{
"epoch": 0.48567483603727996,
"grad_norm": 2.018045663833618,
"learning_rate": 2e-05,
"loss": 0.7262,
"step": 1407
},
{
"epoch": 0.4860200207110804,
"grad_norm": 1.7937039136886597,
"learning_rate": 2e-05,
"loss": 0.7897,
"step": 1408
},
{
"epoch": 0.4863652053848809,
"grad_norm": 2.066096544265747,
"learning_rate": 2e-05,
"loss": 0.7262,
"step": 1409
},
{
"epoch": 0.4867103900586814,
"grad_norm": 1.8681124448776245,
"learning_rate": 2e-05,
"loss": 0.7492,
"step": 1410
},
{
"epoch": 0.4870555747324819,
"grad_norm": 2.0489392280578613,
"learning_rate": 2e-05,
"loss": 0.6817,
"step": 1411
},
{
"epoch": 0.48740075940628236,
"grad_norm": 2.0938589572906494,
"learning_rate": 2e-05,
"loss": 0.791,
"step": 1412
},
{
"epoch": 0.48774594408008287,
"grad_norm": 2.265406847000122,
"learning_rate": 2e-05,
"loss": 0.7032,
"step": 1413
},
{
"epoch": 0.48809112875388333,
"grad_norm": 2.1209421157836914,
"learning_rate": 2e-05,
"loss": 0.741,
"step": 1414
},
{
"epoch": 0.4884363134276838,
"grad_norm": 1.9842997789382935,
"learning_rate": 2e-05,
"loss": 0.8134,
"step": 1415
},
{
"epoch": 0.4887814981014843,
"grad_norm": 2.2025301456451416,
"learning_rate": 2e-05,
"loss": 0.7636,
"step": 1416
},
{
"epoch": 0.48912668277528476,
"grad_norm": 1.9891681671142578,
"learning_rate": 2e-05,
"loss": 0.661,
"step": 1417
},
{
"epoch": 0.4894718674490853,
"grad_norm": 1.7080938816070557,
"learning_rate": 2e-05,
"loss": 0.7104,
"step": 1418
},
{
"epoch": 0.48981705212288573,
"grad_norm": 2.358642101287842,
"learning_rate": 2e-05,
"loss": 0.7915,
"step": 1419
},
{
"epoch": 0.49016223679668625,
"grad_norm": 2.174504280090332,
"learning_rate": 2e-05,
"loss": 0.7419,
"step": 1420
},
{
"epoch": 0.4905074214704867,
"grad_norm": 1.925825834274292,
"learning_rate": 2e-05,
"loss": 0.7043,
"step": 1421
},
{
"epoch": 0.4908526061442872,
"grad_norm": 1.9665645360946655,
"learning_rate": 2e-05,
"loss": 0.8022,
"step": 1422
},
{
"epoch": 0.4911977908180877,
"grad_norm": 2.216688394546509,
"learning_rate": 2e-05,
"loss": 0.8466,
"step": 1423
},
{
"epoch": 0.49154297549188813,
"grad_norm": 2.05391263961792,
"learning_rate": 2e-05,
"loss": 0.7482,
"step": 1424
},
{
"epoch": 0.49188816016568865,
"grad_norm": 1.9934406280517578,
"learning_rate": 2e-05,
"loss": 0.7311,
"step": 1425
},
{
"epoch": 0.4922333448394891,
"grad_norm": 1.9874615669250488,
"learning_rate": 2e-05,
"loss": 0.7423,
"step": 1426
},
{
"epoch": 0.4925785295132896,
"grad_norm": 1.9626049995422363,
"learning_rate": 2e-05,
"loss": 0.7714,
"step": 1427
},
{
"epoch": 0.4929237141870901,
"grad_norm": 1.9596810340881348,
"learning_rate": 2e-05,
"loss": 0.736,
"step": 1428
},
{
"epoch": 0.4932688988608906,
"grad_norm": 2.215177059173584,
"learning_rate": 2e-05,
"loss": 0.7713,
"step": 1429
},
{
"epoch": 0.49361408353469105,
"grad_norm": 1.9259142875671387,
"learning_rate": 2e-05,
"loss": 0.7522,
"step": 1430
},
{
"epoch": 0.49395926820849156,
"grad_norm": 2.1035358905792236,
"learning_rate": 2e-05,
"loss": 0.7654,
"step": 1431
},
{
"epoch": 0.494304452882292,
"grad_norm": 2.373173952102661,
"learning_rate": 2e-05,
"loss": 0.7834,
"step": 1432
},
{
"epoch": 0.49464963755609254,
"grad_norm": 2.126002311706543,
"learning_rate": 2e-05,
"loss": 0.7095,
"step": 1433
},
{
"epoch": 0.494994822229893,
"grad_norm": 2.0989010334014893,
"learning_rate": 2e-05,
"loss": 0.7122,
"step": 1434
},
{
"epoch": 0.49534000690369345,
"grad_norm": 2.3958206176757812,
"learning_rate": 2e-05,
"loss": 0.7614,
"step": 1435
},
{
"epoch": 0.49568519157749397,
"grad_norm": 2.403425693511963,
"learning_rate": 2e-05,
"loss": 0.8194,
"step": 1436
},
{
"epoch": 0.4960303762512944,
"grad_norm": 2.9969820976257324,
"learning_rate": 2e-05,
"loss": 0.7663,
"step": 1437
},
{
"epoch": 0.49637556092509494,
"grad_norm": 2.14561128616333,
"learning_rate": 2e-05,
"loss": 0.7278,
"step": 1438
},
{
"epoch": 0.4967207455988954,
"grad_norm": 2.0737359523773193,
"learning_rate": 2e-05,
"loss": 0.7303,
"step": 1439
},
{
"epoch": 0.4970659302726959,
"grad_norm": 2.2937800884246826,
"learning_rate": 2e-05,
"loss": 0.7683,
"step": 1440
},
{
"epoch": 0.49741111494649637,
"grad_norm": 2.048412799835205,
"learning_rate": 2e-05,
"loss": 0.7369,
"step": 1441
},
{
"epoch": 0.4977562996202969,
"grad_norm": 2.1031463146209717,
"learning_rate": 2e-05,
"loss": 0.7732,
"step": 1442
},
{
"epoch": 0.49810148429409734,
"grad_norm": 2.2189128398895264,
"learning_rate": 2e-05,
"loss": 0.7951,
"step": 1443
},
{
"epoch": 0.4984466689678978,
"grad_norm": 2.1492562294006348,
"learning_rate": 2e-05,
"loss": 0.7951,
"step": 1444
},
{
"epoch": 0.4987918536416983,
"grad_norm": 2.0376765727996826,
"learning_rate": 2e-05,
"loss": 0.7864,
"step": 1445
},
{
"epoch": 0.49913703831549877,
"grad_norm": 1.9230624437332153,
"learning_rate": 2e-05,
"loss": 0.6954,
"step": 1446
},
{
"epoch": 0.4994822229892993,
"grad_norm": 1.9471065998077393,
"learning_rate": 2e-05,
"loss": 0.7413,
"step": 1447
},
{
"epoch": 0.49982740766309974,
"grad_norm": 2.1954751014709473,
"learning_rate": 2e-05,
"loss": 0.7738,
"step": 1448
},
{
"epoch": 0.5001725923369003,
"grad_norm": 2.2330517768859863,
"learning_rate": 2e-05,
"loss": 0.7161,
"step": 1449
},
{
"epoch": 0.5005177770107008,
"grad_norm": 2.1354331970214844,
"learning_rate": 2e-05,
"loss": 0.6457,
"step": 1450
},
{
"epoch": 0.5008629616845012,
"grad_norm": 2.0056471824645996,
"learning_rate": 2e-05,
"loss": 0.7372,
"step": 1451
},
{
"epoch": 0.5012081463583017,
"grad_norm": 2.2320990562438965,
"learning_rate": 2e-05,
"loss": 0.8261,
"step": 1452
},
{
"epoch": 0.5015533310321022,
"grad_norm": 1.9749743938446045,
"learning_rate": 2e-05,
"loss": 0.7372,
"step": 1453
},
{
"epoch": 0.5018985157059026,
"grad_norm": 2.635192632675171,
"learning_rate": 2e-05,
"loss": 0.8702,
"step": 1454
},
{
"epoch": 0.5022437003797031,
"grad_norm": 2.126178026199341,
"learning_rate": 2e-05,
"loss": 0.7182,
"step": 1455
},
{
"epoch": 0.5025888850535036,
"grad_norm": 2.3000221252441406,
"learning_rate": 2e-05,
"loss": 0.7966,
"step": 1456
},
{
"epoch": 0.5029340697273041,
"grad_norm": 1.9998804330825806,
"learning_rate": 2e-05,
"loss": 0.7479,
"step": 1457
},
{
"epoch": 0.5032792544011045,
"grad_norm": 2.0375466346740723,
"learning_rate": 2e-05,
"loss": 0.791,
"step": 1458
},
{
"epoch": 0.5036244390749051,
"grad_norm": 2.1413846015930176,
"learning_rate": 2e-05,
"loss": 0.763,
"step": 1459
},
{
"epoch": 0.5039696237487056,
"grad_norm": 1.9989755153656006,
"learning_rate": 2e-05,
"loss": 0.7703,
"step": 1460
},
{
"epoch": 0.5043148084225061,
"grad_norm": 1.9388622045516968,
"learning_rate": 2e-05,
"loss": 0.7679,
"step": 1461
},
{
"epoch": 0.5046599930963065,
"grad_norm": 2.1844089031219482,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 1462
},
{
"epoch": 0.505005177770107,
"grad_norm": 1.9982469081878662,
"learning_rate": 2e-05,
"loss": 0.7939,
"step": 1463
},
{
"epoch": 0.5053503624439075,
"grad_norm": 2.0412251949310303,
"learning_rate": 2e-05,
"loss": 0.7376,
"step": 1464
},
{
"epoch": 0.5056955471177079,
"grad_norm": 2.1272287368774414,
"learning_rate": 2e-05,
"loss": 0.7477,
"step": 1465
},
{
"epoch": 0.5060407317915084,
"grad_norm": 2.094679594039917,
"learning_rate": 2e-05,
"loss": 0.7581,
"step": 1466
},
{
"epoch": 0.506385916465309,
"grad_norm": 2.2969553470611572,
"learning_rate": 2e-05,
"loss": 0.7634,
"step": 1467
},
{
"epoch": 0.5067311011391095,
"grad_norm": 2.0551202297210693,
"learning_rate": 2e-05,
"loss": 0.6759,
"step": 1468
},
{
"epoch": 0.5070762858129099,
"grad_norm": 2.0950334072113037,
"learning_rate": 2e-05,
"loss": 0.8159,
"step": 1469
},
{
"epoch": 0.5074214704867104,
"grad_norm": 2.216313600540161,
"learning_rate": 2e-05,
"loss": 0.7852,
"step": 1470
},
{
"epoch": 0.5077666551605109,
"grad_norm": 2.1019294261932373,
"learning_rate": 2e-05,
"loss": 0.7405,
"step": 1471
},
{
"epoch": 0.5081118398343114,
"grad_norm": 1.9898511171340942,
"learning_rate": 2e-05,
"loss": 0.7703,
"step": 1472
},
{
"epoch": 0.5084570245081118,
"grad_norm": 1.9961392879486084,
"learning_rate": 2e-05,
"loss": 0.729,
"step": 1473
},
{
"epoch": 0.5088022091819123,
"grad_norm": 2.0674960613250732,
"learning_rate": 2e-05,
"loss": 0.7474,
"step": 1474
},
{
"epoch": 0.5091473938557128,
"grad_norm": 1.9780677556991577,
"learning_rate": 2e-05,
"loss": 0.805,
"step": 1475
},
{
"epoch": 0.5094925785295132,
"grad_norm": 1.9984103441238403,
"learning_rate": 2e-05,
"loss": 0.8204,
"step": 1476
},
{
"epoch": 0.5098377632033138,
"grad_norm": 1.9758920669555664,
"learning_rate": 2e-05,
"loss": 0.7423,
"step": 1477
},
{
"epoch": 0.5101829478771143,
"grad_norm": 2.131047010421753,
"learning_rate": 2e-05,
"loss": 0.7335,
"step": 1478
},
{
"epoch": 0.5105281325509148,
"grad_norm": 2.262054443359375,
"learning_rate": 2e-05,
"loss": 0.7478,
"step": 1479
},
{
"epoch": 0.5108733172247152,
"grad_norm": 1.8867930173873901,
"learning_rate": 2e-05,
"loss": 0.7315,
"step": 1480
},
{
"epoch": 0.5112185018985157,
"grad_norm": 2.000673770904541,
"learning_rate": 2e-05,
"loss": 0.7262,
"step": 1481
},
{
"epoch": 0.5115636865723162,
"grad_norm": 1.9304447174072266,
"learning_rate": 2e-05,
"loss": 0.7361,
"step": 1482
},
{
"epoch": 0.5119088712461167,
"grad_norm": 2.049893856048584,
"learning_rate": 2e-05,
"loss": 0.745,
"step": 1483
},
{
"epoch": 0.5122540559199171,
"grad_norm": 2.0151312351226807,
"learning_rate": 2e-05,
"loss": 0.8003,
"step": 1484
},
{
"epoch": 0.5125992405937176,
"grad_norm": 2.1705594062805176,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 1485
},
{
"epoch": 0.5129444252675182,
"grad_norm": 2.0189003944396973,
"learning_rate": 2e-05,
"loss": 0.8213,
"step": 1486
},
{
"epoch": 0.5132896099413186,
"grad_norm": 2.0302422046661377,
"learning_rate": 2e-05,
"loss": 0.7287,
"step": 1487
},
{
"epoch": 0.5136347946151191,
"grad_norm": 2.0635666847229004,
"learning_rate": 2e-05,
"loss": 0.7931,
"step": 1488
},
{
"epoch": 0.5139799792889196,
"grad_norm": 2.0531816482543945,
"learning_rate": 2e-05,
"loss": 0.766,
"step": 1489
},
{
"epoch": 0.5143251639627201,
"grad_norm": 1.9699041843414307,
"learning_rate": 2e-05,
"loss": 0.6789,
"step": 1490
},
{
"epoch": 0.5146703486365205,
"grad_norm": 2.2556989192962646,
"learning_rate": 2e-05,
"loss": 0.7945,
"step": 1491
},
{
"epoch": 0.515015533310321,
"grad_norm": 2.1396570205688477,
"learning_rate": 2e-05,
"loss": 0.7395,
"step": 1492
},
{
"epoch": 0.5153607179841215,
"grad_norm": 2.188671588897705,
"learning_rate": 2e-05,
"loss": 0.7786,
"step": 1493
},
{
"epoch": 0.5157059026579219,
"grad_norm": 2.048292875289917,
"learning_rate": 2e-05,
"loss": 0.7548,
"step": 1494
},
{
"epoch": 0.5160510873317224,
"grad_norm": 2.00166392326355,
"learning_rate": 2e-05,
"loss": 0.7535,
"step": 1495
},
{
"epoch": 0.516396272005523,
"grad_norm": 2.247016429901123,
"learning_rate": 2e-05,
"loss": 0.7897,
"step": 1496
},
{
"epoch": 0.5167414566793235,
"grad_norm": 2.0391135215759277,
"learning_rate": 2e-05,
"loss": 0.7455,
"step": 1497
},
{
"epoch": 0.5170866413531239,
"grad_norm": 2.0756640434265137,
"learning_rate": 2e-05,
"loss": 0.7534,
"step": 1498
},
{
"epoch": 0.5174318260269244,
"grad_norm": 2.005875587463379,
"learning_rate": 2e-05,
"loss": 0.7393,
"step": 1499
},
{
"epoch": 0.5177770107007249,
"grad_norm": 2.9611847400665283,
"learning_rate": 2e-05,
"loss": 0.7754,
"step": 1500
},
{
"epoch": 0.5181221953745254,
"grad_norm": 2.0369226932525635,
"learning_rate": 2e-05,
"loss": 0.76,
"step": 1501
},
{
"epoch": 0.5184673800483258,
"grad_norm": 2.136617422103882,
"learning_rate": 2e-05,
"loss": 0.7442,
"step": 1502
},
{
"epoch": 0.5188125647221263,
"grad_norm": 2.0811007022857666,
"learning_rate": 2e-05,
"loss": 0.7554,
"step": 1503
},
{
"epoch": 0.5191577493959268,
"grad_norm": 1.9961743354797363,
"learning_rate": 2e-05,
"loss": 0.8295,
"step": 1504
},
{
"epoch": 0.5195029340697273,
"grad_norm": 2.0538458824157715,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 1505
},
{
"epoch": 0.5198481187435278,
"grad_norm": 1.9358627796173096,
"learning_rate": 2e-05,
"loss": 0.739,
"step": 1506
},
{
"epoch": 0.5201933034173283,
"grad_norm": 2.0105397701263428,
"learning_rate": 2e-05,
"loss": 0.7654,
"step": 1507
},
{
"epoch": 0.5205384880911288,
"grad_norm": 1.9926280975341797,
"learning_rate": 2e-05,
"loss": 0.6894,
"step": 1508
},
{
"epoch": 0.5208836727649292,
"grad_norm": 2.034719705581665,
"learning_rate": 2e-05,
"loss": 0.7308,
"step": 1509
},
{
"epoch": 0.5212288574387297,
"grad_norm": 2.001704454421997,
"learning_rate": 2e-05,
"loss": 0.767,
"step": 1510
},
{
"epoch": 0.5215740421125302,
"grad_norm": 1.9315989017486572,
"learning_rate": 2e-05,
"loss": 0.7711,
"step": 1511
},
{
"epoch": 0.5219192267863307,
"grad_norm": 1.8887486457824707,
"learning_rate": 2e-05,
"loss": 0.6943,
"step": 1512
},
{
"epoch": 0.5222644114601311,
"grad_norm": 2.2250192165374756,
"learning_rate": 2e-05,
"loss": 0.686,
"step": 1513
},
{
"epoch": 0.5226095961339317,
"grad_norm": 1.9821523427963257,
"learning_rate": 2e-05,
"loss": 0.7435,
"step": 1514
},
{
"epoch": 0.5229547808077322,
"grad_norm": 1.9626295566558838,
"learning_rate": 2e-05,
"loss": 0.7665,
"step": 1515
},
{
"epoch": 0.5232999654815326,
"grad_norm": 1.9082125425338745,
"learning_rate": 2e-05,
"loss": 0.7459,
"step": 1516
},
{
"epoch": 0.5236451501553331,
"grad_norm": 2.0444540977478027,
"learning_rate": 2e-05,
"loss": 0.8075,
"step": 1517
},
{
"epoch": 0.5239903348291336,
"grad_norm": 2.2874510288238525,
"learning_rate": 2e-05,
"loss": 0.6978,
"step": 1518
},
{
"epoch": 0.5243355195029341,
"grad_norm": 2.4831151962280273,
"learning_rate": 2e-05,
"loss": 0.768,
"step": 1519
},
{
"epoch": 0.5246807041767345,
"grad_norm": 1.9155309200286865,
"learning_rate": 2e-05,
"loss": 0.7397,
"step": 1520
},
{
"epoch": 0.525025888850535,
"grad_norm": 2.0274205207824707,
"learning_rate": 2e-05,
"loss": 0.6859,
"step": 1521
},
{
"epoch": 0.5253710735243355,
"grad_norm": 1.999887466430664,
"learning_rate": 2e-05,
"loss": 0.7105,
"step": 1522
},
{
"epoch": 0.525716258198136,
"grad_norm": 2.0119621753692627,
"learning_rate": 2e-05,
"loss": 0.6969,
"step": 1523
},
{
"epoch": 0.5260614428719365,
"grad_norm": 2.017012119293213,
"learning_rate": 2e-05,
"loss": 0.8121,
"step": 1524
},
{
"epoch": 0.526406627545737,
"grad_norm": 2.2585651874542236,
"learning_rate": 2e-05,
"loss": 0.7834,
"step": 1525
},
{
"epoch": 0.5267518122195375,
"grad_norm": 2.0236425399780273,
"learning_rate": 2e-05,
"loss": 0.7249,
"step": 1526
},
{
"epoch": 0.5270969968933379,
"grad_norm": 2.310391664505005,
"learning_rate": 2e-05,
"loss": 0.794,
"step": 1527
},
{
"epoch": 0.5274421815671384,
"grad_norm": 2.076503038406372,
"learning_rate": 2e-05,
"loss": 0.7216,
"step": 1528
},
{
"epoch": 0.5277873662409389,
"grad_norm": 2.0014805793762207,
"learning_rate": 2e-05,
"loss": 0.7194,
"step": 1529
},
{
"epoch": 0.5281325509147394,
"grad_norm": 2.0097439289093018,
"learning_rate": 2e-05,
"loss": 0.7366,
"step": 1530
},
{
"epoch": 0.5284777355885398,
"grad_norm": 1.9401037693023682,
"learning_rate": 2e-05,
"loss": 0.6921,
"step": 1531
},
{
"epoch": 0.5288229202623403,
"grad_norm": 2.037564277648926,
"learning_rate": 2e-05,
"loss": 0.7337,
"step": 1532
},
{
"epoch": 0.5291681049361409,
"grad_norm": 2.024232864379883,
"learning_rate": 2e-05,
"loss": 0.7869,
"step": 1533
},
{
"epoch": 0.5295132896099414,
"grad_norm": 2.170314073562622,
"learning_rate": 2e-05,
"loss": 0.7656,
"step": 1534
},
{
"epoch": 0.5298584742837418,
"grad_norm": 1.9077699184417725,
"learning_rate": 2e-05,
"loss": 0.7441,
"step": 1535
},
{
"epoch": 0.5302036589575423,
"grad_norm": 1.995130181312561,
"learning_rate": 2e-05,
"loss": 0.7264,
"step": 1536
},
{
"epoch": 0.5305488436313428,
"grad_norm": 2.122166633605957,
"learning_rate": 2e-05,
"loss": 0.7467,
"step": 1537
},
{
"epoch": 0.5308940283051432,
"grad_norm": 1.928505301475525,
"learning_rate": 2e-05,
"loss": 0.759,
"step": 1538
},
{
"epoch": 0.5312392129789437,
"grad_norm": 2.259756565093994,
"learning_rate": 2e-05,
"loss": 0.7654,
"step": 1539
},
{
"epoch": 0.5315843976527442,
"grad_norm": 2.067192554473877,
"learning_rate": 2e-05,
"loss": 0.7518,
"step": 1540
},
{
"epoch": 0.5319295823265447,
"grad_norm": 1.9541715383529663,
"learning_rate": 2e-05,
"loss": 0.7354,
"step": 1541
},
{
"epoch": 0.5322747670003452,
"grad_norm": 2.145857334136963,
"learning_rate": 2e-05,
"loss": 0.6887,
"step": 1542
},
{
"epoch": 0.5326199516741457,
"grad_norm": 1.9784005880355835,
"learning_rate": 2e-05,
"loss": 0.7734,
"step": 1543
},
{
"epoch": 0.5329651363479462,
"grad_norm": 2.2178709506988525,
"learning_rate": 2e-05,
"loss": 0.7385,
"step": 1544
},
{
"epoch": 0.5333103210217466,
"grad_norm": 1.983673334121704,
"learning_rate": 2e-05,
"loss": 0.774,
"step": 1545
},
{
"epoch": 0.5336555056955471,
"grad_norm": 1.9952528476715088,
"learning_rate": 2e-05,
"loss": 0.7436,
"step": 1546
},
{
"epoch": 0.5340006903693476,
"grad_norm": 2.0169942378997803,
"learning_rate": 2e-05,
"loss": 0.7037,
"step": 1547
},
{
"epoch": 0.5343458750431481,
"grad_norm": 2.069014310836792,
"learning_rate": 2e-05,
"loss": 0.7819,
"step": 1548
},
{
"epoch": 0.5346910597169485,
"grad_norm": 2.130476951599121,
"learning_rate": 2e-05,
"loss": 0.7435,
"step": 1549
},
{
"epoch": 0.535036244390749,
"grad_norm": 1.9785997867584229,
"learning_rate": 2e-05,
"loss": 0.816,
"step": 1550
},
{
"epoch": 0.5353814290645496,
"grad_norm": 1.9749222993850708,
"learning_rate": 2e-05,
"loss": 0.705,
"step": 1551
},
{
"epoch": 0.5357266137383501,
"grad_norm": 1.9402345418930054,
"learning_rate": 2e-05,
"loss": 0.7154,
"step": 1552
},
{
"epoch": 0.5360717984121505,
"grad_norm": 2.0092172622680664,
"learning_rate": 2e-05,
"loss": 0.8105,
"step": 1553
},
{
"epoch": 0.536416983085951,
"grad_norm": 2.191633462905884,
"learning_rate": 2e-05,
"loss": 0.7407,
"step": 1554
},
{
"epoch": 0.5367621677597515,
"grad_norm": 1.9947285652160645,
"learning_rate": 2e-05,
"loss": 0.7649,
"step": 1555
},
{
"epoch": 0.5371073524335519,
"grad_norm": 2.1279420852661133,
"learning_rate": 2e-05,
"loss": 0.7469,
"step": 1556
},
{
"epoch": 0.5374525371073524,
"grad_norm": 2.353172540664673,
"learning_rate": 2e-05,
"loss": 0.8125,
"step": 1557
},
{
"epoch": 0.5377977217811529,
"grad_norm": 2.147198438644409,
"learning_rate": 2e-05,
"loss": 0.7344,
"step": 1558
},
{
"epoch": 0.5381429064549534,
"grad_norm": 2.1066479682922363,
"learning_rate": 2e-05,
"loss": 0.7672,
"step": 1559
},
{
"epoch": 0.5384880911287538,
"grad_norm": 2.157383441925049,
"learning_rate": 2e-05,
"loss": 0.7361,
"step": 1560
},
{
"epoch": 0.5388332758025544,
"grad_norm": 2.15088152885437,
"learning_rate": 2e-05,
"loss": 0.741,
"step": 1561
},
{
"epoch": 0.5391784604763549,
"grad_norm": 2.0658955574035645,
"learning_rate": 2e-05,
"loss": 0.6977,
"step": 1562
},
{
"epoch": 0.5395236451501554,
"grad_norm": 1.9369417428970337,
"learning_rate": 2e-05,
"loss": 0.6814,
"step": 1563
},
{
"epoch": 0.5398688298239558,
"grad_norm": 1.9364086389541626,
"learning_rate": 2e-05,
"loss": 0.724,
"step": 1564
},
{
"epoch": 0.5402140144977563,
"grad_norm": 2.0376758575439453,
"learning_rate": 2e-05,
"loss": 0.7336,
"step": 1565
},
{
"epoch": 0.5405591991715568,
"grad_norm": 2.0366671085357666,
"learning_rate": 2e-05,
"loss": 0.7441,
"step": 1566
},
{
"epoch": 0.5409043838453572,
"grad_norm": 2.0837013721466064,
"learning_rate": 2e-05,
"loss": 0.7092,
"step": 1567
},
{
"epoch": 0.5412495685191577,
"grad_norm": 2.155195474624634,
"learning_rate": 2e-05,
"loss": 0.7876,
"step": 1568
},
{
"epoch": 0.5415947531929582,
"grad_norm": 2.9427101612091064,
"learning_rate": 2e-05,
"loss": 0.7363,
"step": 1569
},
{
"epoch": 0.5419399378667588,
"grad_norm": 2.0746378898620605,
"learning_rate": 2e-05,
"loss": 0.8186,
"step": 1570
},
{
"epoch": 0.5422851225405592,
"grad_norm": 2.2725448608398438,
"learning_rate": 2e-05,
"loss": 0.7915,
"step": 1571
},
{
"epoch": 0.5426303072143597,
"grad_norm": 1.9780291318893433,
"learning_rate": 2e-05,
"loss": 0.6905,
"step": 1572
},
{
"epoch": 0.5429754918881602,
"grad_norm": 1.725458025932312,
"learning_rate": 2e-05,
"loss": 0.675,
"step": 1573
},
{
"epoch": 0.5433206765619607,
"grad_norm": 1.9048550128936768,
"learning_rate": 2e-05,
"loss": 0.7574,
"step": 1574
},
{
"epoch": 0.5436658612357611,
"grad_norm": 2.066535472869873,
"learning_rate": 2e-05,
"loss": 0.7394,
"step": 1575
},
{
"epoch": 0.5440110459095616,
"grad_norm": 2.005197048187256,
"learning_rate": 2e-05,
"loss": 0.7291,
"step": 1576
},
{
"epoch": 0.5443562305833621,
"grad_norm": 1.9261797666549683,
"learning_rate": 2e-05,
"loss": 0.7259,
"step": 1577
},
{
"epoch": 0.5447014152571625,
"grad_norm": 2.225381851196289,
"learning_rate": 2e-05,
"loss": 0.7296,
"step": 1578
},
{
"epoch": 0.545046599930963,
"grad_norm": 2.1634013652801514,
"learning_rate": 2e-05,
"loss": 0.7334,
"step": 1579
},
{
"epoch": 0.5453917846047636,
"grad_norm": 1.9177619218826294,
"learning_rate": 2e-05,
"loss": 0.7311,
"step": 1580
},
{
"epoch": 0.5457369692785641,
"grad_norm": 1.9547218084335327,
"learning_rate": 2e-05,
"loss": 0.7289,
"step": 1581
},
{
"epoch": 0.5460821539523645,
"grad_norm": 2.0254147052764893,
"learning_rate": 2e-05,
"loss": 0.7309,
"step": 1582
},
{
"epoch": 0.546427338626165,
"grad_norm": 1.9393632411956787,
"learning_rate": 2e-05,
"loss": 0.7163,
"step": 1583
},
{
"epoch": 0.5467725232999655,
"grad_norm": 2.081874132156372,
"learning_rate": 2e-05,
"loss": 0.7336,
"step": 1584
},
{
"epoch": 0.5471177079737659,
"grad_norm": 2.0485341548919678,
"learning_rate": 2e-05,
"loss": 0.7051,
"step": 1585
},
{
"epoch": 0.5474628926475664,
"grad_norm": 1.9539107084274292,
"learning_rate": 2e-05,
"loss": 0.7768,
"step": 1586
},
{
"epoch": 0.5478080773213669,
"grad_norm": 2.1795575618743896,
"learning_rate": 2e-05,
"loss": 0.7287,
"step": 1587
},
{
"epoch": 0.5481532619951675,
"grad_norm": 1.9676514863967896,
"learning_rate": 2e-05,
"loss": 0.7297,
"step": 1588
},
{
"epoch": 0.5484984466689679,
"grad_norm": 2.0641229152679443,
"learning_rate": 2e-05,
"loss": 0.7974,
"step": 1589
},
{
"epoch": 0.5488436313427684,
"grad_norm": 2.012836217880249,
"learning_rate": 2e-05,
"loss": 0.8395,
"step": 1590
},
{
"epoch": 0.5491888160165689,
"grad_norm": 2.2027604579925537,
"learning_rate": 2e-05,
"loss": 0.7407,
"step": 1591
},
{
"epoch": 0.5495340006903694,
"grad_norm": 2.0422706604003906,
"learning_rate": 2e-05,
"loss": 0.7707,
"step": 1592
},
{
"epoch": 0.5498791853641698,
"grad_norm": 2.224719762802124,
"learning_rate": 2e-05,
"loss": 0.7816,
"step": 1593
},
{
"epoch": 0.5502243700379703,
"grad_norm": 2.003119468688965,
"learning_rate": 2e-05,
"loss": 0.7255,
"step": 1594
},
{
"epoch": 0.5505695547117708,
"grad_norm": 1.9763742685317993,
"learning_rate": 2e-05,
"loss": 0.7147,
"step": 1595
},
{
"epoch": 0.5509147393855712,
"grad_norm": 1.9684779644012451,
"learning_rate": 2e-05,
"loss": 0.77,
"step": 1596
},
{
"epoch": 0.5512599240593717,
"grad_norm": 1.6828429698944092,
"learning_rate": 2e-05,
"loss": 0.6757,
"step": 1597
},
{
"epoch": 0.5516051087331723,
"grad_norm": 2.001671075820923,
"learning_rate": 2e-05,
"loss": 0.7069,
"step": 1598
},
{
"epoch": 0.5519502934069728,
"grad_norm": 2.0887136459350586,
"learning_rate": 2e-05,
"loss": 0.7267,
"step": 1599
},
{
"epoch": 0.5522954780807732,
"grad_norm": 2.1112077236175537,
"learning_rate": 2e-05,
"loss": 0.8088,
"step": 1600
},
{
"epoch": 0.5526406627545737,
"grad_norm": 2.1282196044921875,
"learning_rate": 2e-05,
"loss": 0.7983,
"step": 1601
},
{
"epoch": 0.5529858474283742,
"grad_norm": 2.200883388519287,
"learning_rate": 2e-05,
"loss": 0.8408,
"step": 1602
},
{
"epoch": 0.5533310321021747,
"grad_norm": 2.1473004817962646,
"learning_rate": 2e-05,
"loss": 0.7933,
"step": 1603
},
{
"epoch": 0.5536762167759751,
"grad_norm": 2.2632839679718018,
"learning_rate": 2e-05,
"loss": 0.8073,
"step": 1604
},
{
"epoch": 0.5540214014497756,
"grad_norm": 2.306340456008911,
"learning_rate": 2e-05,
"loss": 0.7737,
"step": 1605
},
{
"epoch": 0.5543665861235761,
"grad_norm": 2.2031912803649902,
"learning_rate": 2e-05,
"loss": 0.7601,
"step": 1606
},
{
"epoch": 0.5547117707973765,
"grad_norm": 2.121080160140991,
"learning_rate": 2e-05,
"loss": 0.7729,
"step": 1607
},
{
"epoch": 0.5550569554711771,
"grad_norm": 1.951230764389038,
"learning_rate": 2e-05,
"loss": 0.7555,
"step": 1608
},
{
"epoch": 0.5554021401449776,
"grad_norm": 2.157416343688965,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 1609
},
{
"epoch": 0.5557473248187781,
"grad_norm": 2.0069007873535156,
"learning_rate": 2e-05,
"loss": 0.7634,
"step": 1610
},
{
"epoch": 0.5560925094925785,
"grad_norm": 1.9920408725738525,
"learning_rate": 2e-05,
"loss": 0.7564,
"step": 1611
},
{
"epoch": 0.556437694166379,
"grad_norm": 2.0004312992095947,
"learning_rate": 2e-05,
"loss": 0.7827,
"step": 1612
},
{
"epoch": 0.5567828788401795,
"grad_norm": 1.8371578454971313,
"learning_rate": 2e-05,
"loss": 0.768,
"step": 1613
},
{
"epoch": 0.55712806351398,
"grad_norm": 2.0383615493774414,
"learning_rate": 2e-05,
"loss": 0.7132,
"step": 1614
},
{
"epoch": 0.5574732481877804,
"grad_norm": 1.9911715984344482,
"learning_rate": 2e-05,
"loss": 0.7275,
"step": 1615
},
{
"epoch": 0.557818432861581,
"grad_norm": 2.0321645736694336,
"learning_rate": 2e-05,
"loss": 0.7741,
"step": 1616
},
{
"epoch": 0.5581636175353815,
"grad_norm": 2.201392650604248,
"learning_rate": 2e-05,
"loss": 0.7139,
"step": 1617
},
{
"epoch": 0.5585088022091819,
"grad_norm": 2.305196523666382,
"learning_rate": 2e-05,
"loss": 0.7536,
"step": 1618
},
{
"epoch": 0.5588539868829824,
"grad_norm": 2.0547752380371094,
"learning_rate": 2e-05,
"loss": 0.7634,
"step": 1619
},
{
"epoch": 0.5591991715567829,
"grad_norm": 2.1537587642669678,
"learning_rate": 2e-05,
"loss": 0.7689,
"step": 1620
},
{
"epoch": 0.5595443562305834,
"grad_norm": 2.0345499515533447,
"learning_rate": 2e-05,
"loss": 0.7343,
"step": 1621
},
{
"epoch": 0.5598895409043838,
"grad_norm": 2.0531249046325684,
"learning_rate": 2e-05,
"loss": 0.7926,
"step": 1622
},
{
"epoch": 0.5602347255781843,
"grad_norm": 1.865893006324768,
"learning_rate": 2e-05,
"loss": 0.6914,
"step": 1623
},
{
"epoch": 0.5605799102519848,
"grad_norm": 2.147118330001831,
"learning_rate": 2e-05,
"loss": 0.724,
"step": 1624
},
{
"epoch": 0.5609250949257854,
"grad_norm": 2.1363260746002197,
"learning_rate": 2e-05,
"loss": 0.7638,
"step": 1625
},
{
"epoch": 0.5612702795995858,
"grad_norm": 2.059763193130493,
"learning_rate": 2e-05,
"loss": 0.6898,
"step": 1626
},
{
"epoch": 0.5616154642733863,
"grad_norm": 1.9224729537963867,
"learning_rate": 2e-05,
"loss": 0.7285,
"step": 1627
},
{
"epoch": 0.5619606489471868,
"grad_norm": 1.9039995670318604,
"learning_rate": 2e-05,
"loss": 0.7426,
"step": 1628
},
{
"epoch": 0.5623058336209872,
"grad_norm": 2.0386292934417725,
"learning_rate": 2e-05,
"loss": 0.7187,
"step": 1629
},
{
"epoch": 0.5626510182947877,
"grad_norm": 2.1839492321014404,
"learning_rate": 2e-05,
"loss": 0.717,
"step": 1630
},
{
"epoch": 0.5629962029685882,
"grad_norm": 1.9907827377319336,
"learning_rate": 2e-05,
"loss": 0.8015,
"step": 1631
},
{
"epoch": 0.5633413876423887,
"grad_norm": 1.9858790636062622,
"learning_rate": 2e-05,
"loss": 0.783,
"step": 1632
},
{
"epoch": 0.5636865723161891,
"grad_norm": 2.020822286605835,
"learning_rate": 2e-05,
"loss": 0.7108,
"step": 1633
},
{
"epoch": 0.5640317569899896,
"grad_norm": 2.067755937576294,
"learning_rate": 2e-05,
"loss": 0.8178,
"step": 1634
},
{
"epoch": 0.5643769416637902,
"grad_norm": 2.01557993888855,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 1635
},
{
"epoch": 0.5647221263375906,
"grad_norm": 1.7513538599014282,
"learning_rate": 2e-05,
"loss": 0.685,
"step": 1636
},
{
"epoch": 0.5650673110113911,
"grad_norm": 2.10416316986084,
"learning_rate": 2e-05,
"loss": 0.7789,
"step": 1637
},
{
"epoch": 0.5654124956851916,
"grad_norm": 2.2250373363494873,
"learning_rate": 2e-05,
"loss": 0.7544,
"step": 1638
},
{
"epoch": 0.5657576803589921,
"grad_norm": 1.9085569381713867,
"learning_rate": 2e-05,
"loss": 0.7766,
"step": 1639
},
{
"epoch": 0.5661028650327925,
"grad_norm": 2.5093936920166016,
"learning_rate": 2e-05,
"loss": 0.6989,
"step": 1640
},
{
"epoch": 0.566448049706593,
"grad_norm": 2.0018603801727295,
"learning_rate": 2e-05,
"loss": 0.7504,
"step": 1641
},
{
"epoch": 0.5667932343803935,
"grad_norm": 2.0827620029449463,
"learning_rate": 2e-05,
"loss": 0.8142,
"step": 1642
},
{
"epoch": 0.567138419054194,
"grad_norm": 2.149536609649658,
"learning_rate": 2e-05,
"loss": 0.7936,
"step": 1643
},
{
"epoch": 0.5674836037279944,
"grad_norm": 2.1129469871520996,
"learning_rate": 2e-05,
"loss": 0.7667,
"step": 1644
},
{
"epoch": 0.567828788401795,
"grad_norm": 2.035013198852539,
"learning_rate": 2e-05,
"loss": 0.6809,
"step": 1645
},
{
"epoch": 0.5681739730755955,
"grad_norm": 2.144164562225342,
"learning_rate": 2e-05,
"loss": 0.7808,
"step": 1646
},
{
"epoch": 0.5685191577493959,
"grad_norm": 2.0699079036712646,
"learning_rate": 2e-05,
"loss": 0.7605,
"step": 1647
},
{
"epoch": 0.5688643424231964,
"grad_norm": 2.038463592529297,
"learning_rate": 2e-05,
"loss": 0.7557,
"step": 1648
},
{
"epoch": 0.5692095270969969,
"grad_norm": 2.055161237716675,
"learning_rate": 2e-05,
"loss": 0.7014,
"step": 1649
},
{
"epoch": 0.5695547117707974,
"grad_norm": 2.0253241062164307,
"learning_rate": 2e-05,
"loss": 0.7643,
"step": 1650
},
{
"epoch": 0.5698998964445978,
"grad_norm": 2.076207160949707,
"learning_rate": 2e-05,
"loss": 0.7118,
"step": 1651
},
{
"epoch": 0.5702450811183983,
"grad_norm": 2.239205837249756,
"learning_rate": 2e-05,
"loss": 0.7365,
"step": 1652
},
{
"epoch": 0.5705902657921988,
"grad_norm": 2.0288281440734863,
"learning_rate": 2e-05,
"loss": 0.716,
"step": 1653
},
{
"epoch": 0.5709354504659994,
"grad_norm": 1.860352635383606,
"learning_rate": 2e-05,
"loss": 0.7208,
"step": 1654
},
{
"epoch": 0.5712806351397998,
"grad_norm": 1.8015706539154053,
"learning_rate": 2e-05,
"loss": 0.7034,
"step": 1655
},
{
"epoch": 0.5716258198136003,
"grad_norm": 2.0100910663604736,
"learning_rate": 2e-05,
"loss": 0.7788,
"step": 1656
},
{
"epoch": 0.5719710044874008,
"grad_norm": 2.0733449459075928,
"learning_rate": 2e-05,
"loss": 0.7912,
"step": 1657
},
{
"epoch": 0.5723161891612012,
"grad_norm": 1.8067480325698853,
"learning_rate": 2e-05,
"loss": 0.6478,
"step": 1658
},
{
"epoch": 0.5726613738350017,
"grad_norm": 1.8684823513031006,
"learning_rate": 2e-05,
"loss": 0.6878,
"step": 1659
},
{
"epoch": 0.5730065585088022,
"grad_norm": 2.1666419506073,
"learning_rate": 2e-05,
"loss": 0.7513,
"step": 1660
},
{
"epoch": 0.5733517431826027,
"grad_norm": 2.0782723426818848,
"learning_rate": 2e-05,
"loss": 0.78,
"step": 1661
},
{
"epoch": 0.5736969278564031,
"grad_norm": 2.139180898666382,
"learning_rate": 2e-05,
"loss": 0.7002,
"step": 1662
},
{
"epoch": 0.5740421125302037,
"grad_norm": 1.9298183917999268,
"learning_rate": 2e-05,
"loss": 0.7442,
"step": 1663
},
{
"epoch": 0.5743872972040042,
"grad_norm": 2.006073474884033,
"learning_rate": 2e-05,
"loss": 0.7592,
"step": 1664
},
{
"epoch": 0.5747324818778047,
"grad_norm": 1.9423385858535767,
"learning_rate": 2e-05,
"loss": 0.7451,
"step": 1665
},
{
"epoch": 0.5750776665516051,
"grad_norm": 1.912705898284912,
"learning_rate": 2e-05,
"loss": 0.7387,
"step": 1666
},
{
"epoch": 0.5754228512254056,
"grad_norm": 1.8943840265274048,
"learning_rate": 2e-05,
"loss": 0.7685,
"step": 1667
},
{
"epoch": 0.5757680358992061,
"grad_norm": 2.166865348815918,
"learning_rate": 2e-05,
"loss": 0.788,
"step": 1668
},
{
"epoch": 0.5761132205730065,
"grad_norm": 1.9983549118041992,
"learning_rate": 2e-05,
"loss": 0.7082,
"step": 1669
},
{
"epoch": 0.576458405246807,
"grad_norm": 2.038658380508423,
"learning_rate": 2e-05,
"loss": 0.7364,
"step": 1670
},
{
"epoch": 0.5768035899206075,
"grad_norm": 1.986772894859314,
"learning_rate": 2e-05,
"loss": 0.7583,
"step": 1671
},
{
"epoch": 0.577148774594408,
"grad_norm": 2.411196708679199,
"learning_rate": 2e-05,
"loss": 0.8288,
"step": 1672
},
{
"epoch": 0.5774939592682085,
"grad_norm": 2.085162401199341,
"learning_rate": 2e-05,
"loss": 0.8174,
"step": 1673
},
{
"epoch": 0.577839143942009,
"grad_norm": 1.9370945692062378,
"learning_rate": 2e-05,
"loss": 0.6927,
"step": 1674
},
{
"epoch": 0.5781843286158095,
"grad_norm": 1.9420502185821533,
"learning_rate": 2e-05,
"loss": 0.6648,
"step": 1675
},
{
"epoch": 0.5785295132896099,
"grad_norm": 2.007918119430542,
"learning_rate": 2e-05,
"loss": 0.7475,
"step": 1676
},
{
"epoch": 0.5788746979634104,
"grad_norm": 2.177980899810791,
"learning_rate": 2e-05,
"loss": 0.7581,
"step": 1677
},
{
"epoch": 0.5792198826372109,
"grad_norm": 1.9090574979782104,
"learning_rate": 2e-05,
"loss": 0.788,
"step": 1678
},
{
"epoch": 0.5795650673110114,
"grad_norm": 2.1762993335723877,
"learning_rate": 2e-05,
"loss": 0.7959,
"step": 1679
},
{
"epoch": 0.5799102519848118,
"grad_norm": 1.9556688070297241,
"learning_rate": 2e-05,
"loss": 0.6954,
"step": 1680
},
{
"epoch": 0.5802554366586123,
"grad_norm": 1.808090329170227,
"learning_rate": 2e-05,
"loss": 0.6858,
"step": 1681
},
{
"epoch": 0.5806006213324129,
"grad_norm": 2.1649343967437744,
"learning_rate": 2e-05,
"loss": 0.8071,
"step": 1682
},
{
"epoch": 0.5809458060062134,
"grad_norm": 1.8695378303527832,
"learning_rate": 2e-05,
"loss": 0.6969,
"step": 1683
},
{
"epoch": 0.5812909906800138,
"grad_norm": 2.02304744720459,
"learning_rate": 2e-05,
"loss": 0.77,
"step": 1684
},
{
"epoch": 0.5816361753538143,
"grad_norm": 2.2785298824310303,
"learning_rate": 2e-05,
"loss": 0.8095,
"step": 1685
},
{
"epoch": 0.5819813600276148,
"grad_norm": 1.8864084482192993,
"learning_rate": 2e-05,
"loss": 0.742,
"step": 1686
},
{
"epoch": 0.5823265447014152,
"grad_norm": 2.0156564712524414,
"learning_rate": 2e-05,
"loss": 0.7366,
"step": 1687
},
{
"epoch": 0.5826717293752157,
"grad_norm": 1.9681613445281982,
"learning_rate": 2e-05,
"loss": 0.7202,
"step": 1688
},
{
"epoch": 0.5830169140490162,
"grad_norm": 1.924330234527588,
"learning_rate": 2e-05,
"loss": 0.7693,
"step": 1689
},
{
"epoch": 0.5833620987228167,
"grad_norm": 2.186133623123169,
"learning_rate": 2e-05,
"loss": 0.7721,
"step": 1690
},
{
"epoch": 0.5837072833966171,
"grad_norm": 2.036860704421997,
"learning_rate": 2e-05,
"loss": 0.7897,
"step": 1691
},
{
"epoch": 0.5840524680704177,
"grad_norm": 2.060377359390259,
"learning_rate": 2e-05,
"loss": 0.7381,
"step": 1692
},
{
"epoch": 0.5843976527442182,
"grad_norm": 1.8350690603256226,
"learning_rate": 2e-05,
"loss": 0.7999,
"step": 1693
},
{
"epoch": 0.5847428374180187,
"grad_norm": 1.918983817100525,
"learning_rate": 2e-05,
"loss": 0.7065,
"step": 1694
},
{
"epoch": 0.5850880220918191,
"grad_norm": 2.3777992725372314,
"learning_rate": 2e-05,
"loss": 0.7155,
"step": 1695
},
{
"epoch": 0.5854332067656196,
"grad_norm": 1.9003201723098755,
"learning_rate": 2e-05,
"loss": 0.7461,
"step": 1696
},
{
"epoch": 0.5857783914394201,
"grad_norm": 2.0567221641540527,
"learning_rate": 2e-05,
"loss": 0.8043,
"step": 1697
},
{
"epoch": 0.5861235761132205,
"grad_norm": 1.971789836883545,
"learning_rate": 2e-05,
"loss": 0.7982,
"step": 1698
},
{
"epoch": 0.586468760787021,
"grad_norm": 2.068467378616333,
"learning_rate": 2e-05,
"loss": 0.7839,
"step": 1699
},
{
"epoch": 0.5868139454608216,
"grad_norm": 2.196093797683716,
"learning_rate": 2e-05,
"loss": 0.7204,
"step": 1700
},
{
"epoch": 0.5871591301346221,
"grad_norm": 2.0267608165740967,
"learning_rate": 2e-05,
"loss": 0.7686,
"step": 1701
},
{
"epoch": 0.5875043148084225,
"grad_norm": 1.9164024591445923,
"learning_rate": 2e-05,
"loss": 0.7488,
"step": 1702
},
{
"epoch": 0.587849499482223,
"grad_norm": 1.9293785095214844,
"learning_rate": 2e-05,
"loss": 0.7129,
"step": 1703
},
{
"epoch": 0.5881946841560235,
"grad_norm": 1.9195809364318848,
"learning_rate": 2e-05,
"loss": 0.7141,
"step": 1704
},
{
"epoch": 0.588539868829824,
"grad_norm": 2.0254979133605957,
"learning_rate": 2e-05,
"loss": 0.7567,
"step": 1705
},
{
"epoch": 0.5888850535036244,
"grad_norm": 1.9486969709396362,
"learning_rate": 2e-05,
"loss": 0.7221,
"step": 1706
},
{
"epoch": 0.5892302381774249,
"grad_norm": 2.134180784225464,
"learning_rate": 2e-05,
"loss": 0.7213,
"step": 1707
},
{
"epoch": 0.5895754228512254,
"grad_norm": 2.0742263793945312,
"learning_rate": 2e-05,
"loss": 0.8293,
"step": 1708
},
{
"epoch": 0.5899206075250258,
"grad_norm": 1.7730787992477417,
"learning_rate": 2e-05,
"loss": 0.7447,
"step": 1709
},
{
"epoch": 0.5902657921988264,
"grad_norm": 1.8583852052688599,
"learning_rate": 2e-05,
"loss": 0.6666,
"step": 1710
},
{
"epoch": 0.5906109768726269,
"grad_norm": 2.0180208683013916,
"learning_rate": 2e-05,
"loss": 0.6629,
"step": 1711
},
{
"epoch": 0.5909561615464274,
"grad_norm": 2.239849328994751,
"learning_rate": 2e-05,
"loss": 0.7729,
"step": 1712
},
{
"epoch": 0.5913013462202278,
"grad_norm": 1.933658480644226,
"learning_rate": 2e-05,
"loss": 0.7052,
"step": 1713
},
{
"epoch": 0.5916465308940283,
"grad_norm": 1.8087807893753052,
"learning_rate": 2e-05,
"loss": 0.686,
"step": 1714
},
{
"epoch": 0.5919917155678288,
"grad_norm": 2.1065690517425537,
"learning_rate": 2e-05,
"loss": 0.7558,
"step": 1715
},
{
"epoch": 0.5923369002416292,
"grad_norm": 2.0715813636779785,
"learning_rate": 2e-05,
"loss": 0.7997,
"step": 1716
},
{
"epoch": 0.5926820849154297,
"grad_norm": 2.178509473800659,
"learning_rate": 2e-05,
"loss": 0.7247,
"step": 1717
},
{
"epoch": 0.5930272695892302,
"grad_norm": 1.9442565441131592,
"learning_rate": 2e-05,
"loss": 0.751,
"step": 1718
},
{
"epoch": 0.5933724542630308,
"grad_norm": 2.473721742630005,
"learning_rate": 2e-05,
"loss": 0.7021,
"step": 1719
},
{
"epoch": 0.5937176389368312,
"grad_norm": 2.1822264194488525,
"learning_rate": 2e-05,
"loss": 0.7788,
"step": 1720
},
{
"epoch": 0.5940628236106317,
"grad_norm": 1.9657896757125854,
"learning_rate": 2e-05,
"loss": 0.8364,
"step": 1721
},
{
"epoch": 0.5944080082844322,
"grad_norm": 2.1759588718414307,
"learning_rate": 2e-05,
"loss": 0.7492,
"step": 1722
},
{
"epoch": 0.5947531929582327,
"grad_norm": 2.267892837524414,
"learning_rate": 2e-05,
"loss": 0.7746,
"step": 1723
},
{
"epoch": 0.5950983776320331,
"grad_norm": 2.3498010635375977,
"learning_rate": 2e-05,
"loss": 0.7976,
"step": 1724
},
{
"epoch": 0.5954435623058336,
"grad_norm": 1.9854902029037476,
"learning_rate": 2e-05,
"loss": 0.7015,
"step": 1725
},
{
"epoch": 0.5957887469796341,
"grad_norm": 1.831830620765686,
"learning_rate": 2e-05,
"loss": 0.7616,
"step": 1726
},
{
"epoch": 0.5961339316534345,
"grad_norm": 1.8729759454727173,
"learning_rate": 2e-05,
"loss": 0.7203,
"step": 1727
},
{
"epoch": 0.596479116327235,
"grad_norm": 2.0665640830993652,
"learning_rate": 2e-05,
"loss": 0.8078,
"step": 1728
},
{
"epoch": 0.5968243010010356,
"grad_norm": 1.8335654735565186,
"learning_rate": 2e-05,
"loss": 0.7417,
"step": 1729
},
{
"epoch": 0.5971694856748361,
"grad_norm": 2.133212089538574,
"learning_rate": 2e-05,
"loss": 0.7637,
"step": 1730
},
{
"epoch": 0.5975146703486365,
"grad_norm": 2.0471878051757812,
"learning_rate": 2e-05,
"loss": 0.7602,
"step": 1731
},
{
"epoch": 0.597859855022437,
"grad_norm": 1.9667534828186035,
"learning_rate": 2e-05,
"loss": 0.7237,
"step": 1732
},
{
"epoch": 0.5982050396962375,
"grad_norm": 1.8838579654693604,
"learning_rate": 2e-05,
"loss": 0.7392,
"step": 1733
},
{
"epoch": 0.598550224370038,
"grad_norm": 1.9155827760696411,
"learning_rate": 2e-05,
"loss": 0.747,
"step": 1734
},
{
"epoch": 0.5988954090438384,
"grad_norm": 1.894668459892273,
"learning_rate": 2e-05,
"loss": 0.7584,
"step": 1735
},
{
"epoch": 0.5992405937176389,
"grad_norm": 2.290691614151001,
"learning_rate": 2e-05,
"loss": 0.7439,
"step": 1736
},
{
"epoch": 0.5995857783914395,
"grad_norm": 1.9438388347625732,
"learning_rate": 2e-05,
"loss": 0.6965,
"step": 1737
},
{
"epoch": 0.5999309630652399,
"grad_norm": 1.9882612228393555,
"learning_rate": 2e-05,
"loss": 0.7536,
"step": 1738
},
{
"epoch": 0.6002761477390404,
"grad_norm": 2.268845796585083,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 1739
},
{
"epoch": 0.6006213324128409,
"grad_norm": 2.092918872833252,
"learning_rate": 2e-05,
"loss": 0.7648,
"step": 1740
},
{
"epoch": 0.6009665170866414,
"grad_norm": 1.9001197814941406,
"learning_rate": 2e-05,
"loss": 0.6952,
"step": 1741
},
{
"epoch": 0.6013117017604418,
"grad_norm": 1.882922649383545,
"learning_rate": 2e-05,
"loss": 0.6908,
"step": 1742
},
{
"epoch": 0.6016568864342423,
"grad_norm": 2.2816600799560547,
"learning_rate": 2e-05,
"loss": 0.7897,
"step": 1743
},
{
"epoch": 0.6020020711080428,
"grad_norm": 1.8837263584136963,
"learning_rate": 2e-05,
"loss": 0.7058,
"step": 1744
},
{
"epoch": 0.6023472557818433,
"grad_norm": 1.9763485193252563,
"learning_rate": 2e-05,
"loss": 0.7919,
"step": 1745
},
{
"epoch": 0.6026924404556437,
"grad_norm": 1.9180881977081299,
"learning_rate": 2e-05,
"loss": 0.7787,
"step": 1746
},
{
"epoch": 0.6030376251294443,
"grad_norm": 2.1102733612060547,
"learning_rate": 2e-05,
"loss": 0.6886,
"step": 1747
},
{
"epoch": 0.6033828098032448,
"grad_norm": 2.0356481075286865,
"learning_rate": 2e-05,
"loss": 0.7486,
"step": 1748
},
{
"epoch": 0.6037279944770452,
"grad_norm": 1.8429824113845825,
"learning_rate": 2e-05,
"loss": 0.7076,
"step": 1749
},
{
"epoch": 0.6040731791508457,
"grad_norm": 2.011791706085205,
"learning_rate": 2e-05,
"loss": 0.8403,
"step": 1750
},
{
"epoch": 0.6044183638246462,
"grad_norm": 2.0239579677581787,
"learning_rate": 2e-05,
"loss": 0.7039,
"step": 1751
},
{
"epoch": 0.6047635484984467,
"grad_norm": 2.055365800857544,
"learning_rate": 2e-05,
"loss": 0.7784,
"step": 1752
},
{
"epoch": 0.6051087331722471,
"grad_norm": 1.9595205783843994,
"learning_rate": 2e-05,
"loss": 0.715,
"step": 1753
},
{
"epoch": 0.6054539178460476,
"grad_norm": 1.906576156616211,
"learning_rate": 2e-05,
"loss": 0.7467,
"step": 1754
},
{
"epoch": 0.6057991025198481,
"grad_norm": 1.8687126636505127,
"learning_rate": 2e-05,
"loss": 0.7083,
"step": 1755
},
{
"epoch": 0.6061442871936487,
"grad_norm": 1.9710134267807007,
"learning_rate": 2e-05,
"loss": 0.742,
"step": 1756
},
{
"epoch": 0.6064894718674491,
"grad_norm": 2.020474910736084,
"learning_rate": 2e-05,
"loss": 0.749,
"step": 1757
},
{
"epoch": 0.6068346565412496,
"grad_norm": 1.8899691104888916,
"learning_rate": 2e-05,
"loss": 0.7157,
"step": 1758
},
{
"epoch": 0.6071798412150501,
"grad_norm": 1.9584208726882935,
"learning_rate": 2e-05,
"loss": 0.8034,
"step": 1759
},
{
"epoch": 0.6075250258888505,
"grad_norm": 1.99704909324646,
"learning_rate": 2e-05,
"loss": 0.797,
"step": 1760
},
{
"epoch": 0.607870210562651,
"grad_norm": 1.8569170236587524,
"learning_rate": 2e-05,
"loss": 0.686,
"step": 1761
},
{
"epoch": 0.6082153952364515,
"grad_norm": 1.9338957071304321,
"learning_rate": 2e-05,
"loss": 0.713,
"step": 1762
},
{
"epoch": 0.608560579910252,
"grad_norm": 1.8302687406539917,
"learning_rate": 2e-05,
"loss": 0.7838,
"step": 1763
},
{
"epoch": 0.6089057645840524,
"grad_norm": 1.8911799192428589,
"learning_rate": 2e-05,
"loss": 0.694,
"step": 1764
},
{
"epoch": 0.609250949257853,
"grad_norm": 1.919000267982483,
"learning_rate": 2e-05,
"loss": 0.7398,
"step": 1765
},
{
"epoch": 0.6095961339316535,
"grad_norm": 1.704526662826538,
"learning_rate": 2e-05,
"loss": 0.7244,
"step": 1766
},
{
"epoch": 0.6099413186054539,
"grad_norm": 2.246150493621826,
"learning_rate": 2e-05,
"loss": 0.7638,
"step": 1767
},
{
"epoch": 0.6102865032792544,
"grad_norm": 2.198012590408325,
"learning_rate": 2e-05,
"loss": 0.8096,
"step": 1768
},
{
"epoch": 0.6106316879530549,
"grad_norm": 1.8018826246261597,
"learning_rate": 2e-05,
"loss": 0.7102,
"step": 1769
},
{
"epoch": 0.6109768726268554,
"grad_norm": 1.9078664779663086,
"learning_rate": 2e-05,
"loss": 0.7707,
"step": 1770
},
{
"epoch": 0.6113220573006558,
"grad_norm": 2.0132734775543213,
"learning_rate": 2e-05,
"loss": 0.7412,
"step": 1771
},
{
"epoch": 0.6116672419744563,
"grad_norm": 1.9725755453109741,
"learning_rate": 2e-05,
"loss": 0.6708,
"step": 1772
},
{
"epoch": 0.6120124266482568,
"grad_norm": 1.8628737926483154,
"learning_rate": 2e-05,
"loss": 0.6974,
"step": 1773
},
{
"epoch": 0.6123576113220573,
"grad_norm": 2.0750017166137695,
"learning_rate": 2e-05,
"loss": 0.6955,
"step": 1774
},
{
"epoch": 0.6127027959958578,
"grad_norm": 2.146649122238159,
"learning_rate": 2e-05,
"loss": 0.7738,
"step": 1775
},
{
"epoch": 0.6130479806696583,
"grad_norm": 2.194870710372925,
"learning_rate": 2e-05,
"loss": 0.7436,
"step": 1776
},
{
"epoch": 0.6133931653434588,
"grad_norm": 1.86497962474823,
"learning_rate": 2e-05,
"loss": 0.671,
"step": 1777
},
{
"epoch": 0.6137383500172592,
"grad_norm": 1.817057490348816,
"learning_rate": 2e-05,
"loss": 0.7139,
"step": 1778
},
{
"epoch": 0.6140835346910597,
"grad_norm": 1.77781081199646,
"learning_rate": 2e-05,
"loss": 0.8184,
"step": 1779
},
{
"epoch": 0.6144287193648602,
"grad_norm": 2.260162830352783,
"learning_rate": 2e-05,
"loss": 0.7326,
"step": 1780
},
{
"epoch": 0.6147739040386607,
"grad_norm": 1.9308106899261475,
"learning_rate": 2e-05,
"loss": 0.7825,
"step": 1781
},
{
"epoch": 0.6151190887124611,
"grad_norm": 2.2054190635681152,
"learning_rate": 2e-05,
"loss": 0.7864,
"step": 1782
},
{
"epoch": 0.6154642733862616,
"grad_norm": 1.9252761602401733,
"learning_rate": 2e-05,
"loss": 0.6684,
"step": 1783
},
{
"epoch": 0.6158094580600622,
"grad_norm": 1.8950225114822388,
"learning_rate": 2e-05,
"loss": 0.7536,
"step": 1784
},
{
"epoch": 0.6161546427338627,
"grad_norm": 2.0569279193878174,
"learning_rate": 2e-05,
"loss": 0.7241,
"step": 1785
},
{
"epoch": 0.6164998274076631,
"grad_norm": 1.9371174573898315,
"learning_rate": 2e-05,
"loss": 0.7887,
"step": 1786
},
{
"epoch": 0.6168450120814636,
"grad_norm": 2.2338998317718506,
"learning_rate": 2e-05,
"loss": 0.8006,
"step": 1787
},
{
"epoch": 0.6171901967552641,
"grad_norm": 1.8488763570785522,
"learning_rate": 2e-05,
"loss": 0.6702,
"step": 1788
},
{
"epoch": 0.6175353814290645,
"grad_norm": 2.2544572353363037,
"learning_rate": 2e-05,
"loss": 0.706,
"step": 1789
},
{
"epoch": 0.617880566102865,
"grad_norm": 2.090319871902466,
"learning_rate": 2e-05,
"loss": 0.7558,
"step": 1790
},
{
"epoch": 0.6182257507766655,
"grad_norm": 1.964186429977417,
"learning_rate": 2e-05,
"loss": 0.7239,
"step": 1791
},
{
"epoch": 0.618570935450466,
"grad_norm": 2.0483083724975586,
"learning_rate": 2e-05,
"loss": 0.772,
"step": 1792
},
{
"epoch": 0.6189161201242664,
"grad_norm": 1.9692699909210205,
"learning_rate": 2e-05,
"loss": 0.7027,
"step": 1793
},
{
"epoch": 0.619261304798067,
"grad_norm": 1.870023250579834,
"learning_rate": 2e-05,
"loss": 0.6619,
"step": 1794
},
{
"epoch": 0.6196064894718675,
"grad_norm": 2.014556884765625,
"learning_rate": 2e-05,
"loss": 0.7341,
"step": 1795
},
{
"epoch": 0.619951674145668,
"grad_norm": 2.087446451187134,
"learning_rate": 2e-05,
"loss": 0.7611,
"step": 1796
},
{
"epoch": 0.6202968588194684,
"grad_norm": 1.8687217235565186,
"learning_rate": 2e-05,
"loss": 0.734,
"step": 1797
},
{
"epoch": 0.6206420434932689,
"grad_norm": 2.1333250999450684,
"learning_rate": 2e-05,
"loss": 0.7627,
"step": 1798
},
{
"epoch": 0.6209872281670694,
"grad_norm": 2.1312520503997803,
"learning_rate": 2e-05,
"loss": 0.7873,
"step": 1799
},
{
"epoch": 0.6213324128408698,
"grad_norm": 2.2815170288085938,
"learning_rate": 2e-05,
"loss": 0.7975,
"step": 1800
},
{
"epoch": 0.6216775975146703,
"grad_norm": 1.9498845338821411,
"learning_rate": 2e-05,
"loss": 0.682,
"step": 1801
},
{
"epoch": 0.6220227821884708,
"grad_norm": 1.9391518831253052,
"learning_rate": 2e-05,
"loss": 0.7906,
"step": 1802
},
{
"epoch": 0.6223679668622714,
"grad_norm": 2.0360279083251953,
"learning_rate": 2e-05,
"loss": 0.7266,
"step": 1803
},
{
"epoch": 0.6227131515360718,
"grad_norm": 1.856303095817566,
"learning_rate": 2e-05,
"loss": 0.7342,
"step": 1804
},
{
"epoch": 0.6230583362098723,
"grad_norm": 1.977275013923645,
"learning_rate": 2e-05,
"loss": 0.7155,
"step": 1805
},
{
"epoch": 0.6234035208836728,
"grad_norm": 1.910068154335022,
"learning_rate": 2e-05,
"loss": 0.6858,
"step": 1806
},
{
"epoch": 0.6237487055574732,
"grad_norm": 1.956804633140564,
"learning_rate": 2e-05,
"loss": 0.7133,
"step": 1807
},
{
"epoch": 0.6240938902312737,
"grad_norm": 2.7539641857147217,
"learning_rate": 2e-05,
"loss": 0.8129,
"step": 1808
},
{
"epoch": 0.6244390749050742,
"grad_norm": 2.066072463989258,
"learning_rate": 2e-05,
"loss": 0.7429,
"step": 1809
},
{
"epoch": 0.6247842595788747,
"grad_norm": 2.1080305576324463,
"learning_rate": 2e-05,
"loss": 0.8327,
"step": 1810
},
{
"epoch": 0.6251294442526751,
"grad_norm": 1.8386399745941162,
"learning_rate": 2e-05,
"loss": 0.7471,
"step": 1811
},
{
"epoch": 0.6254746289264757,
"grad_norm": 1.8144325017929077,
"learning_rate": 2e-05,
"loss": 0.7176,
"step": 1812
},
{
"epoch": 0.6258198136002762,
"grad_norm": 1.8652502298355103,
"learning_rate": 2e-05,
"loss": 0.7218,
"step": 1813
},
{
"epoch": 0.6261649982740767,
"grad_norm": 1.791069507598877,
"learning_rate": 2e-05,
"loss": 0.655,
"step": 1814
},
{
"epoch": 0.6265101829478771,
"grad_norm": 1.9747053384780884,
"learning_rate": 2e-05,
"loss": 0.7198,
"step": 1815
},
{
"epoch": 0.6268553676216776,
"grad_norm": 1.8452980518341064,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 1816
},
{
"epoch": 0.6272005522954781,
"grad_norm": 1.998065710067749,
"learning_rate": 2e-05,
"loss": 0.6552,
"step": 1817
},
{
"epoch": 0.6275457369692785,
"grad_norm": 2.0098564624786377,
"learning_rate": 2e-05,
"loss": 0.7543,
"step": 1818
},
{
"epoch": 0.627890921643079,
"grad_norm": 2.0185914039611816,
"learning_rate": 2e-05,
"loss": 0.6491,
"step": 1819
},
{
"epoch": 0.6282361063168795,
"grad_norm": 1.8583019971847534,
"learning_rate": 2e-05,
"loss": 0.7594,
"step": 1820
},
{
"epoch": 0.62858129099068,
"grad_norm": 2.032019853591919,
"learning_rate": 2e-05,
"loss": 0.7946,
"step": 1821
},
{
"epoch": 0.6289264756644805,
"grad_norm": 1.780571699142456,
"learning_rate": 2e-05,
"loss": 0.6687,
"step": 1822
},
{
"epoch": 0.629271660338281,
"grad_norm": 1.90001380443573,
"learning_rate": 2e-05,
"loss": 0.7503,
"step": 1823
},
{
"epoch": 0.6296168450120815,
"grad_norm": 1.8221665620803833,
"learning_rate": 2e-05,
"loss": 0.7005,
"step": 1824
},
{
"epoch": 0.629962029685882,
"grad_norm": 2.047510862350464,
"learning_rate": 2e-05,
"loss": 0.7297,
"step": 1825
},
{
"epoch": 0.6303072143596824,
"grad_norm": 2.1644015312194824,
"learning_rate": 2e-05,
"loss": 0.7539,
"step": 1826
},
{
"epoch": 0.6306523990334829,
"grad_norm": 1.903037428855896,
"learning_rate": 2e-05,
"loss": 0.7411,
"step": 1827
},
{
"epoch": 0.6309975837072834,
"grad_norm": 2.0709774494171143,
"learning_rate": 2e-05,
"loss": 0.7528,
"step": 1828
},
{
"epoch": 0.6313427683810838,
"grad_norm": 1.818328619003296,
"learning_rate": 2e-05,
"loss": 0.7505,
"step": 1829
},
{
"epoch": 0.6316879530548843,
"grad_norm": 2.16007924079895,
"learning_rate": 2e-05,
"loss": 0.7712,
"step": 1830
},
{
"epoch": 0.6320331377286849,
"grad_norm": 2.229264497756958,
"learning_rate": 2e-05,
"loss": 0.6883,
"step": 1831
},
{
"epoch": 0.6323783224024854,
"grad_norm": 2.0821151733398438,
"learning_rate": 2e-05,
"loss": 0.7146,
"step": 1832
},
{
"epoch": 0.6327235070762858,
"grad_norm": 2.144075632095337,
"learning_rate": 2e-05,
"loss": 0.7315,
"step": 1833
},
{
"epoch": 0.6330686917500863,
"grad_norm": 1.9066596031188965,
"learning_rate": 2e-05,
"loss": 0.7179,
"step": 1834
},
{
"epoch": 0.6334138764238868,
"grad_norm": 2.207310438156128,
"learning_rate": 2e-05,
"loss": 0.8625,
"step": 1835
},
{
"epoch": 0.6337590610976873,
"grad_norm": 2.1372647285461426,
"learning_rate": 2e-05,
"loss": 0.7152,
"step": 1836
},
{
"epoch": 0.6341042457714877,
"grad_norm": 2.1546401977539062,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 1837
},
{
"epoch": 0.6344494304452882,
"grad_norm": 1.917305588722229,
"learning_rate": 2e-05,
"loss": 0.6907,
"step": 1838
},
{
"epoch": 0.6347946151190887,
"grad_norm": 2.1244444847106934,
"learning_rate": 2e-05,
"loss": 0.6422,
"step": 1839
},
{
"epoch": 0.6351397997928891,
"grad_norm": 1.8833587169647217,
"learning_rate": 2e-05,
"loss": 0.7405,
"step": 1840
},
{
"epoch": 0.6354849844666897,
"grad_norm": 1.9901988506317139,
"learning_rate": 2e-05,
"loss": 0.6463,
"step": 1841
},
{
"epoch": 0.6358301691404902,
"grad_norm": 1.7179203033447266,
"learning_rate": 2e-05,
"loss": 0.6762,
"step": 1842
},
{
"epoch": 0.6361753538142907,
"grad_norm": 2.0448687076568604,
"learning_rate": 2e-05,
"loss": 0.773,
"step": 1843
},
{
"epoch": 0.6365205384880911,
"grad_norm": 1.9936163425445557,
"learning_rate": 2e-05,
"loss": 0.8002,
"step": 1844
},
{
"epoch": 0.6368657231618916,
"grad_norm": 1.7121373414993286,
"learning_rate": 2e-05,
"loss": 0.7177,
"step": 1845
},
{
"epoch": 0.6372109078356921,
"grad_norm": 2.215914011001587,
"learning_rate": 2e-05,
"loss": 0.7671,
"step": 1846
},
{
"epoch": 0.6375560925094926,
"grad_norm": 1.9892266988754272,
"learning_rate": 2e-05,
"loss": 0.695,
"step": 1847
},
{
"epoch": 0.637901277183293,
"grad_norm": 2.027421236038208,
"learning_rate": 2e-05,
"loss": 0.7789,
"step": 1848
},
{
"epoch": 0.6382464618570935,
"grad_norm": 1.7547681331634521,
"learning_rate": 2e-05,
"loss": 0.7371,
"step": 1849
},
{
"epoch": 0.6385916465308941,
"grad_norm": 1.8186923265457153,
"learning_rate": 2e-05,
"loss": 0.6986,
"step": 1850
},
{
"epoch": 0.6389368312046945,
"grad_norm": 2.215562343597412,
"learning_rate": 2e-05,
"loss": 0.7462,
"step": 1851
},
{
"epoch": 0.639282015878495,
"grad_norm": 2.0459797382354736,
"learning_rate": 2e-05,
"loss": 0.7707,
"step": 1852
},
{
"epoch": 0.6396272005522955,
"grad_norm": 2.03794264793396,
"learning_rate": 2e-05,
"loss": 0.6956,
"step": 1853
},
{
"epoch": 0.639972385226096,
"grad_norm": 1.9094321727752686,
"learning_rate": 2e-05,
"loss": 0.7241,
"step": 1854
},
{
"epoch": 0.6403175698998964,
"grad_norm": 2.1592180728912354,
"learning_rate": 2e-05,
"loss": 0.7627,
"step": 1855
},
{
"epoch": 0.6406627545736969,
"grad_norm": 2.0268704891204834,
"learning_rate": 2e-05,
"loss": 0.7224,
"step": 1856
},
{
"epoch": 0.6410079392474974,
"grad_norm": 1.8396837711334229,
"learning_rate": 2e-05,
"loss": 0.7284,
"step": 1857
},
{
"epoch": 0.6413531239212978,
"grad_norm": 1.9870339632034302,
"learning_rate": 2e-05,
"loss": 0.7563,
"step": 1858
},
{
"epoch": 0.6416983085950984,
"grad_norm": 2.0115363597869873,
"learning_rate": 2e-05,
"loss": 0.7208,
"step": 1859
},
{
"epoch": 0.6420434932688989,
"grad_norm": 1.8832600116729736,
"learning_rate": 2e-05,
"loss": 0.7239,
"step": 1860
},
{
"epoch": 0.6423886779426994,
"grad_norm": 2.019841194152832,
"learning_rate": 2e-05,
"loss": 0.7471,
"step": 1861
},
{
"epoch": 0.6427338626164998,
"grad_norm": 2.1090104579925537,
"learning_rate": 2e-05,
"loss": 0.68,
"step": 1862
},
{
"epoch": 0.6430790472903003,
"grad_norm": 4.953197479248047,
"learning_rate": 2e-05,
"loss": 0.7384,
"step": 1863
},
{
"epoch": 0.6434242319641008,
"grad_norm": 2.010185956954956,
"learning_rate": 2e-05,
"loss": 0.7749,
"step": 1864
},
{
"epoch": 0.6437694166379013,
"grad_norm": 2.1290745735168457,
"learning_rate": 2e-05,
"loss": 0.8058,
"step": 1865
},
{
"epoch": 0.6441146013117017,
"grad_norm": 1.844883918762207,
"learning_rate": 2e-05,
"loss": 0.7414,
"step": 1866
},
{
"epoch": 0.6444597859855022,
"grad_norm": 1.7499961853027344,
"learning_rate": 2e-05,
"loss": 0.6545,
"step": 1867
},
{
"epoch": 0.6448049706593028,
"grad_norm": 1.98123037815094,
"learning_rate": 2e-05,
"loss": 0.7576,
"step": 1868
},
{
"epoch": 0.6451501553331032,
"grad_norm": 1.8479448556900024,
"learning_rate": 2e-05,
"loss": 0.7226,
"step": 1869
},
{
"epoch": 0.6454953400069037,
"grad_norm": 1.8625061511993408,
"learning_rate": 2e-05,
"loss": 0.7184,
"step": 1870
},
{
"epoch": 0.6458405246807042,
"grad_norm": 1.8531122207641602,
"learning_rate": 2e-05,
"loss": 0.7183,
"step": 1871
},
{
"epoch": 0.6461857093545047,
"grad_norm": 1.983819603919983,
"learning_rate": 2e-05,
"loss": 0.7413,
"step": 1872
},
{
"epoch": 0.6465308940283051,
"grad_norm": 2.3842971324920654,
"learning_rate": 2e-05,
"loss": 0.7503,
"step": 1873
},
{
"epoch": 0.6468760787021056,
"grad_norm": 2.144202947616577,
"learning_rate": 2e-05,
"loss": 0.7396,
"step": 1874
},
{
"epoch": 0.6472212633759061,
"grad_norm": 1.8604322671890259,
"learning_rate": 2e-05,
"loss": 0.6854,
"step": 1875
},
{
"epoch": 0.6475664480497066,
"grad_norm": 1.8701947927474976,
"learning_rate": 2e-05,
"loss": 0.7588,
"step": 1876
},
{
"epoch": 0.647911632723507,
"grad_norm": 1.8759455680847168,
"learning_rate": 2e-05,
"loss": 0.7261,
"step": 1877
},
{
"epoch": 0.6482568173973076,
"grad_norm": 1.939672589302063,
"learning_rate": 2e-05,
"loss": 0.7466,
"step": 1878
},
{
"epoch": 0.6486020020711081,
"grad_norm": 1.9740561246871948,
"learning_rate": 2e-05,
"loss": 0.7002,
"step": 1879
},
{
"epoch": 0.6489471867449085,
"grad_norm": 1.8402413129806519,
"learning_rate": 2e-05,
"loss": 0.7037,
"step": 1880
},
{
"epoch": 0.649292371418709,
"grad_norm": 2.111009359359741,
"learning_rate": 2e-05,
"loss": 0.7178,
"step": 1881
},
{
"epoch": 0.6496375560925095,
"grad_norm": 1.8650747537612915,
"learning_rate": 2e-05,
"loss": 0.7488,
"step": 1882
},
{
"epoch": 0.64998274076631,
"grad_norm": 2.1026782989501953,
"learning_rate": 2e-05,
"loss": 0.7038,
"step": 1883
},
{
"epoch": 0.6503279254401104,
"grad_norm": 1.8944039344787598,
"learning_rate": 2e-05,
"loss": 0.7441,
"step": 1884
},
{
"epoch": 0.6506731101139109,
"grad_norm": 2.293912887573242,
"learning_rate": 2e-05,
"loss": 0.7434,
"step": 1885
},
{
"epoch": 0.6510182947877114,
"grad_norm": 2.0031826496124268,
"learning_rate": 2e-05,
"loss": 0.7619,
"step": 1886
},
{
"epoch": 0.651363479461512,
"grad_norm": 2.3530232906341553,
"learning_rate": 2e-05,
"loss": 0.7579,
"step": 1887
},
{
"epoch": 0.6517086641353124,
"grad_norm": 1.8405147790908813,
"learning_rate": 2e-05,
"loss": 0.6814,
"step": 1888
},
{
"epoch": 0.6520538488091129,
"grad_norm": 1.9461179971694946,
"learning_rate": 2e-05,
"loss": 0.7671,
"step": 1889
},
{
"epoch": 0.6523990334829134,
"grad_norm": 1.7684873342514038,
"learning_rate": 2e-05,
"loss": 0.7004,
"step": 1890
},
{
"epoch": 0.6527442181567138,
"grad_norm": 1.8901242017745972,
"learning_rate": 2e-05,
"loss": 0.7023,
"step": 1891
},
{
"epoch": 0.6530894028305143,
"grad_norm": 1.7435083389282227,
"learning_rate": 2e-05,
"loss": 0.7241,
"step": 1892
},
{
"epoch": 0.6534345875043148,
"grad_norm": 2.1555747985839844,
"learning_rate": 2e-05,
"loss": 0.7367,
"step": 1893
},
{
"epoch": 0.6537797721781153,
"grad_norm": 1.8946081399917603,
"learning_rate": 2e-05,
"loss": 0.7442,
"step": 1894
},
{
"epoch": 0.6541249568519157,
"grad_norm": 2.026679039001465,
"learning_rate": 2e-05,
"loss": 0.744,
"step": 1895
},
{
"epoch": 0.6544701415257163,
"grad_norm": 2.0073494911193848,
"learning_rate": 2e-05,
"loss": 0.739,
"step": 1896
},
{
"epoch": 0.6548153261995168,
"grad_norm": 2.209312915802002,
"learning_rate": 2e-05,
"loss": 0.8024,
"step": 1897
},
{
"epoch": 0.6551605108733172,
"grad_norm": 1.9675532579421997,
"learning_rate": 2e-05,
"loss": 0.7477,
"step": 1898
},
{
"epoch": 0.6555056955471177,
"grad_norm": 1.9891093969345093,
"learning_rate": 2e-05,
"loss": 0.6932,
"step": 1899
},
{
"epoch": 0.6558508802209182,
"grad_norm": 1.8822745084762573,
"learning_rate": 2e-05,
"loss": 0.7461,
"step": 1900
},
{
"epoch": 0.6561960648947187,
"grad_norm": 1.8406872749328613,
"learning_rate": 2e-05,
"loss": 0.6554,
"step": 1901
},
{
"epoch": 0.6565412495685191,
"grad_norm": 1.999000906944275,
"learning_rate": 2e-05,
"loss": 0.7744,
"step": 1902
},
{
"epoch": 0.6568864342423196,
"grad_norm": 1.885979413986206,
"learning_rate": 2e-05,
"loss": 0.7196,
"step": 1903
},
{
"epoch": 0.6572316189161201,
"grad_norm": 2.010152816772461,
"learning_rate": 2e-05,
"loss": 0.703,
"step": 1904
},
{
"epoch": 0.6575768035899207,
"grad_norm": 1.9070861339569092,
"learning_rate": 2e-05,
"loss": 0.7323,
"step": 1905
},
{
"epoch": 0.6579219882637211,
"grad_norm": 2.058511734008789,
"learning_rate": 2e-05,
"loss": 0.7564,
"step": 1906
},
{
"epoch": 0.6582671729375216,
"grad_norm": 1.7195355892181396,
"learning_rate": 2e-05,
"loss": 0.7032,
"step": 1907
},
{
"epoch": 0.6586123576113221,
"grad_norm": 1.9448564052581787,
"learning_rate": 2e-05,
"loss": 0.743,
"step": 1908
},
{
"epoch": 0.6589575422851225,
"grad_norm": 1.9529019594192505,
"learning_rate": 2e-05,
"loss": 0.7543,
"step": 1909
},
{
"epoch": 0.659302726958923,
"grad_norm": 2.0683536529541016,
"learning_rate": 2e-05,
"loss": 0.7242,
"step": 1910
},
{
"epoch": 0.6596479116327235,
"grad_norm": 1.9975461959838867,
"learning_rate": 2e-05,
"loss": 0.6788,
"step": 1911
},
{
"epoch": 0.659993096306524,
"grad_norm": 1.8823899030685425,
"learning_rate": 2e-05,
"loss": 0.7409,
"step": 1912
},
{
"epoch": 0.6603382809803244,
"grad_norm": 1.9681503772735596,
"learning_rate": 2e-05,
"loss": 0.7618,
"step": 1913
},
{
"epoch": 0.660683465654125,
"grad_norm": 2.1105501651763916,
"learning_rate": 2e-05,
"loss": 0.7371,
"step": 1914
},
{
"epoch": 0.6610286503279255,
"grad_norm": 1.8724969625473022,
"learning_rate": 2e-05,
"loss": 0.7364,
"step": 1915
},
{
"epoch": 0.661373835001726,
"grad_norm": 1.7453885078430176,
"learning_rate": 2e-05,
"loss": 0.7459,
"step": 1916
},
{
"epoch": 0.6617190196755264,
"grad_norm": 1.86320960521698,
"learning_rate": 2e-05,
"loss": 0.6788,
"step": 1917
},
{
"epoch": 0.6620642043493269,
"grad_norm": 2.0787503719329834,
"learning_rate": 2e-05,
"loss": 0.6537,
"step": 1918
},
{
"epoch": 0.6624093890231274,
"grad_norm": 1.908781886100769,
"learning_rate": 2e-05,
"loss": 0.7056,
"step": 1919
},
{
"epoch": 0.6627545736969278,
"grad_norm": 1.7666840553283691,
"learning_rate": 2e-05,
"loss": 0.6854,
"step": 1920
},
{
"epoch": 0.6630997583707283,
"grad_norm": 2.0884993076324463,
"learning_rate": 2e-05,
"loss": 0.6902,
"step": 1921
},
{
"epoch": 0.6634449430445288,
"grad_norm": 2.226449489593506,
"learning_rate": 2e-05,
"loss": 0.7029,
"step": 1922
},
{
"epoch": 0.6637901277183293,
"grad_norm": 1.8212461471557617,
"learning_rate": 2e-05,
"loss": 0.7956,
"step": 1923
},
{
"epoch": 0.6641353123921298,
"grad_norm": 1.8987208604812622,
"learning_rate": 2e-05,
"loss": 0.7318,
"step": 1924
},
{
"epoch": 0.6644804970659303,
"grad_norm": 2.003673553466797,
"learning_rate": 2e-05,
"loss": 0.7001,
"step": 1925
},
{
"epoch": 0.6648256817397308,
"grad_norm": 2.025902271270752,
"learning_rate": 2e-05,
"loss": 0.7352,
"step": 1926
},
{
"epoch": 0.6651708664135313,
"grad_norm": 1.8917664289474487,
"learning_rate": 2e-05,
"loss": 0.6461,
"step": 1927
},
{
"epoch": 0.6655160510873317,
"grad_norm": 1.92942214012146,
"learning_rate": 2e-05,
"loss": 0.7062,
"step": 1928
},
{
"epoch": 0.6658612357611322,
"grad_norm": 2.1221516132354736,
"learning_rate": 2e-05,
"loss": 0.7176,
"step": 1929
},
{
"epoch": 0.6662064204349327,
"grad_norm": 1.9428044557571411,
"learning_rate": 2e-05,
"loss": 0.6917,
"step": 1930
},
{
"epoch": 0.6665516051087331,
"grad_norm": 1.7395883798599243,
"learning_rate": 2e-05,
"loss": 0.7188,
"step": 1931
},
{
"epoch": 0.6668967897825336,
"grad_norm": 1.9090925455093384,
"learning_rate": 2e-05,
"loss": 0.6685,
"step": 1932
},
{
"epoch": 0.6672419744563342,
"grad_norm": 1.909045696258545,
"learning_rate": 2e-05,
"loss": 0.7461,
"step": 1933
},
{
"epoch": 0.6675871591301347,
"grad_norm": 1.8969789743423462,
"learning_rate": 2e-05,
"loss": 0.7144,
"step": 1934
},
{
"epoch": 0.6679323438039351,
"grad_norm": 1.9943249225616455,
"learning_rate": 2e-05,
"loss": 0.7793,
"step": 1935
},
{
"epoch": 0.6682775284777356,
"grad_norm": 1.7785159349441528,
"learning_rate": 2e-05,
"loss": 0.7046,
"step": 1936
},
{
"epoch": 0.6686227131515361,
"grad_norm": 1.997578740119934,
"learning_rate": 2e-05,
"loss": 0.7546,
"step": 1937
},
{
"epoch": 0.6689678978253366,
"grad_norm": 1.8899143934249878,
"learning_rate": 2e-05,
"loss": 0.7889,
"step": 1938
},
{
"epoch": 0.669313082499137,
"grad_norm": 2.030691385269165,
"learning_rate": 2e-05,
"loss": 0.7349,
"step": 1939
},
{
"epoch": 0.6696582671729375,
"grad_norm": 2.083049774169922,
"learning_rate": 2e-05,
"loss": 0.7519,
"step": 1940
},
{
"epoch": 0.670003451846738,
"grad_norm": 1.8681942224502563,
"learning_rate": 2e-05,
"loss": 0.7124,
"step": 1941
},
{
"epoch": 0.6703486365205384,
"grad_norm": 1.788272738456726,
"learning_rate": 2e-05,
"loss": 0.777,
"step": 1942
},
{
"epoch": 0.670693821194339,
"grad_norm": 1.867202639579773,
"learning_rate": 2e-05,
"loss": 0.7557,
"step": 1943
},
{
"epoch": 0.6710390058681395,
"grad_norm": 1.9626368284225464,
"learning_rate": 2e-05,
"loss": 0.7768,
"step": 1944
},
{
"epoch": 0.67138419054194,
"grad_norm": 1.6680896282196045,
"learning_rate": 2e-05,
"loss": 0.6917,
"step": 1945
},
{
"epoch": 0.6717293752157404,
"grad_norm": 2.0082855224609375,
"learning_rate": 2e-05,
"loss": 0.7013,
"step": 1946
},
{
"epoch": 0.6720745598895409,
"grad_norm": 2.0258872509002686,
"learning_rate": 2e-05,
"loss": 0.7393,
"step": 1947
},
{
"epoch": 0.6724197445633414,
"grad_norm": 1.9214152097702026,
"learning_rate": 2e-05,
"loss": 0.7728,
"step": 1948
},
{
"epoch": 0.6727649292371418,
"grad_norm": 1.9643549919128418,
"learning_rate": 2e-05,
"loss": 0.8084,
"step": 1949
},
{
"epoch": 0.6731101139109423,
"grad_norm": 1.8793781995773315,
"learning_rate": 2e-05,
"loss": 0.6406,
"step": 1950
},
{
"epoch": 0.6734552985847428,
"grad_norm": 1.7508033514022827,
"learning_rate": 2e-05,
"loss": 0.7493,
"step": 1951
},
{
"epoch": 0.6738004832585434,
"grad_norm": 2.0109660625457764,
"learning_rate": 2e-05,
"loss": 0.6838,
"step": 1952
},
{
"epoch": 0.6741456679323438,
"grad_norm": 2.2193634510040283,
"learning_rate": 2e-05,
"loss": 0.8267,
"step": 1953
},
{
"epoch": 0.6744908526061443,
"grad_norm": 1.7887126207351685,
"learning_rate": 2e-05,
"loss": 0.7343,
"step": 1954
},
{
"epoch": 0.6748360372799448,
"grad_norm": 1.8429667949676514,
"learning_rate": 2e-05,
"loss": 0.7461,
"step": 1955
},
{
"epoch": 0.6751812219537453,
"grad_norm": 1.7922366857528687,
"learning_rate": 2e-05,
"loss": 0.7569,
"step": 1956
},
{
"epoch": 0.6755264066275457,
"grad_norm": 1.96672785282135,
"learning_rate": 2e-05,
"loss": 0.7195,
"step": 1957
},
{
"epoch": 0.6758715913013462,
"grad_norm": 2.067754030227661,
"learning_rate": 2e-05,
"loss": 0.7751,
"step": 1958
},
{
"epoch": 0.6762167759751467,
"grad_norm": 1.9873639345169067,
"learning_rate": 2e-05,
"loss": 0.7171,
"step": 1959
},
{
"epoch": 0.6765619606489471,
"grad_norm": 1.8441859483718872,
"learning_rate": 2e-05,
"loss": 0.7682,
"step": 1960
},
{
"epoch": 0.6769071453227476,
"grad_norm": 1.941342830657959,
"learning_rate": 2e-05,
"loss": 0.773,
"step": 1961
},
{
"epoch": 0.6772523299965482,
"grad_norm": 1.7210693359375,
"learning_rate": 2e-05,
"loss": 0.7115,
"step": 1962
},
{
"epoch": 0.6775975146703487,
"grad_norm": 1.7671408653259277,
"learning_rate": 2e-05,
"loss": 0.7045,
"step": 1963
},
{
"epoch": 0.6779426993441491,
"grad_norm": 1.859602451324463,
"learning_rate": 2e-05,
"loss": 0.6997,
"step": 1964
},
{
"epoch": 0.6782878840179496,
"grad_norm": 1.8216410875320435,
"learning_rate": 2e-05,
"loss": 0.7883,
"step": 1965
},
{
"epoch": 0.6786330686917501,
"grad_norm": 1.786430835723877,
"learning_rate": 2e-05,
"loss": 0.7674,
"step": 1966
},
{
"epoch": 0.6789782533655506,
"grad_norm": 1.9019408226013184,
"learning_rate": 2e-05,
"loss": 0.747,
"step": 1967
},
{
"epoch": 0.679323438039351,
"grad_norm": 1.6905081272125244,
"learning_rate": 2e-05,
"loss": 0.7103,
"step": 1968
},
{
"epoch": 0.6796686227131515,
"grad_norm": 2.1167047023773193,
"learning_rate": 2e-05,
"loss": 0.6932,
"step": 1969
},
{
"epoch": 0.680013807386952,
"grad_norm": 2.043133020401001,
"learning_rate": 2e-05,
"loss": 0.6787,
"step": 1970
},
{
"epoch": 0.6803589920607525,
"grad_norm": 1.9253062009811401,
"learning_rate": 2e-05,
"loss": 0.7626,
"step": 1971
},
{
"epoch": 0.680704176734553,
"grad_norm": 9.413606643676758,
"learning_rate": 2e-05,
"loss": 0.7885,
"step": 1972
},
{
"epoch": 0.6810493614083535,
"grad_norm": 1.855717420578003,
"learning_rate": 2e-05,
"loss": 0.6962,
"step": 1973
},
{
"epoch": 0.681394546082154,
"grad_norm": 1.9458452463150024,
"learning_rate": 2e-05,
"loss": 0.7286,
"step": 1974
},
{
"epoch": 0.6817397307559544,
"grad_norm": 2.1786346435546875,
"learning_rate": 2e-05,
"loss": 0.7362,
"step": 1975
},
{
"epoch": 0.6820849154297549,
"grad_norm": 1.9228813648223877,
"learning_rate": 2e-05,
"loss": 0.7438,
"step": 1976
},
{
"epoch": 0.6824301001035554,
"grad_norm": 1.8485450744628906,
"learning_rate": 2e-05,
"loss": 0.7569,
"step": 1977
},
{
"epoch": 0.6827752847773559,
"grad_norm": 1.8340065479278564,
"learning_rate": 2e-05,
"loss": 0.6705,
"step": 1978
},
{
"epoch": 0.6831204694511563,
"grad_norm": 1.8531389236450195,
"learning_rate": 2e-05,
"loss": 0.7137,
"step": 1979
},
{
"epoch": 0.6834656541249569,
"grad_norm": 1.8941481113433838,
"learning_rate": 2e-05,
"loss": 0.6973,
"step": 1980
},
{
"epoch": 0.6838108387987574,
"grad_norm": 1.9881943464279175,
"learning_rate": 2e-05,
"loss": 0.7301,
"step": 1981
},
{
"epoch": 0.6841560234725578,
"grad_norm": 1.9976316690444946,
"learning_rate": 2e-05,
"loss": 0.641,
"step": 1982
},
{
"epoch": 0.6845012081463583,
"grad_norm": 1.9177840948104858,
"learning_rate": 2e-05,
"loss": 0.6988,
"step": 1983
},
{
"epoch": 0.6848463928201588,
"grad_norm": 1.8939712047576904,
"learning_rate": 2e-05,
"loss": 0.746,
"step": 1984
},
{
"epoch": 0.6851915774939593,
"grad_norm": 1.75008225440979,
"learning_rate": 2e-05,
"loss": 0.6892,
"step": 1985
},
{
"epoch": 0.6855367621677597,
"grad_norm": 1.826703667640686,
"learning_rate": 2e-05,
"loss": 0.6953,
"step": 1986
},
{
"epoch": 0.6858819468415602,
"grad_norm": 1.8086583614349365,
"learning_rate": 2e-05,
"loss": 0.7222,
"step": 1987
},
{
"epoch": 0.6862271315153607,
"grad_norm": 1.7769607305526733,
"learning_rate": 2e-05,
"loss": 0.7096,
"step": 1988
},
{
"epoch": 0.6865723161891611,
"grad_norm": 1.791774868965149,
"learning_rate": 2e-05,
"loss": 0.74,
"step": 1989
},
{
"epoch": 0.6869175008629617,
"grad_norm": 1.7862011194229126,
"learning_rate": 2e-05,
"loss": 0.722,
"step": 1990
},
{
"epoch": 0.6872626855367622,
"grad_norm": 1.9089808464050293,
"learning_rate": 2e-05,
"loss": 0.7408,
"step": 1991
},
{
"epoch": 0.6876078702105627,
"grad_norm": 1.8044335842132568,
"learning_rate": 2e-05,
"loss": 0.6796,
"step": 1992
},
{
"epoch": 0.6879530548843631,
"grad_norm": 1.9218417406082153,
"learning_rate": 2e-05,
"loss": 0.657,
"step": 1993
},
{
"epoch": 0.6882982395581636,
"grad_norm": 2.2235755920410156,
"learning_rate": 2e-05,
"loss": 0.7303,
"step": 1994
},
{
"epoch": 0.6886434242319641,
"grad_norm": 1.9212822914123535,
"learning_rate": 2e-05,
"loss": 0.7132,
"step": 1995
},
{
"epoch": 0.6889886089057646,
"grad_norm": 1.8071088790893555,
"learning_rate": 2e-05,
"loss": 0.7945,
"step": 1996
},
{
"epoch": 0.689333793579565,
"grad_norm": 1.9095362424850464,
"learning_rate": 2e-05,
"loss": 0.7727,
"step": 1997
},
{
"epoch": 0.6896789782533655,
"grad_norm": 1.8873735666275024,
"learning_rate": 2e-05,
"loss": 0.7049,
"step": 1998
},
{
"epoch": 0.6900241629271661,
"grad_norm": 2.031343698501587,
"learning_rate": 2e-05,
"loss": 0.7456,
"step": 1999
},
{
"epoch": 0.6903693476009665,
"grad_norm": 1.8298405408859253,
"learning_rate": 2e-05,
"loss": 0.6711,
"step": 2000
},
{
"epoch": 0.690714532274767,
"grad_norm": 1.7381670475006104,
"learning_rate": 2e-05,
"loss": 0.725,
"step": 2001
},
{
"epoch": 0.6910597169485675,
"grad_norm": 1.7420293092727661,
"learning_rate": 2e-05,
"loss": 0.7253,
"step": 2002
},
{
"epoch": 0.691404901622368,
"grad_norm": 1.9419313669204712,
"learning_rate": 2e-05,
"loss": 0.7454,
"step": 2003
},
{
"epoch": 0.6917500862961684,
"grad_norm": 1.751600742340088,
"learning_rate": 2e-05,
"loss": 0.6653,
"step": 2004
},
{
"epoch": 0.6920952709699689,
"grad_norm": 1.7798436880111694,
"learning_rate": 2e-05,
"loss": 0.74,
"step": 2005
},
{
"epoch": 0.6924404556437694,
"grad_norm": 1.9783426523208618,
"learning_rate": 2e-05,
"loss": 0.7706,
"step": 2006
},
{
"epoch": 0.69278564031757,
"grad_norm": 2.567061424255371,
"learning_rate": 2e-05,
"loss": 0.7334,
"step": 2007
},
{
"epoch": 0.6931308249913704,
"grad_norm": 1.967150330543518,
"learning_rate": 2e-05,
"loss": 0.7207,
"step": 2008
},
{
"epoch": 0.6934760096651709,
"grad_norm": 2.13596248626709,
"learning_rate": 2e-05,
"loss": 0.6873,
"step": 2009
},
{
"epoch": 0.6938211943389714,
"grad_norm": 1.868224024772644,
"learning_rate": 2e-05,
"loss": 0.8539,
"step": 2010
},
{
"epoch": 0.6941663790127718,
"grad_norm": 2.0384867191314697,
"learning_rate": 2e-05,
"loss": 0.7786,
"step": 2011
},
{
"epoch": 0.6945115636865723,
"grad_norm": 1.9924743175506592,
"learning_rate": 2e-05,
"loss": 0.7169,
"step": 2012
},
{
"epoch": 0.6948567483603728,
"grad_norm": 2.20979642868042,
"learning_rate": 2e-05,
"loss": 0.7322,
"step": 2013
},
{
"epoch": 0.6952019330341733,
"grad_norm": 2.0181972980499268,
"learning_rate": 2e-05,
"loss": 0.7795,
"step": 2014
},
{
"epoch": 0.6955471177079737,
"grad_norm": 1.788724660873413,
"learning_rate": 2e-05,
"loss": 0.6801,
"step": 2015
},
{
"epoch": 0.6958923023817742,
"grad_norm": 2.113201379776001,
"learning_rate": 2e-05,
"loss": 0.802,
"step": 2016
},
{
"epoch": 0.6962374870555748,
"grad_norm": 1.8842284679412842,
"learning_rate": 2e-05,
"loss": 0.7029,
"step": 2017
},
{
"epoch": 0.6965826717293753,
"grad_norm": 2.102614641189575,
"learning_rate": 2e-05,
"loss": 0.7782,
"step": 2018
},
{
"epoch": 0.6969278564031757,
"grad_norm": 1.9915715456008911,
"learning_rate": 2e-05,
"loss": 0.7777,
"step": 2019
},
{
"epoch": 0.6972730410769762,
"grad_norm": 1.9043521881103516,
"learning_rate": 2e-05,
"loss": 0.7221,
"step": 2020
},
{
"epoch": 0.6976182257507767,
"grad_norm": 2.12050461769104,
"learning_rate": 2e-05,
"loss": 0.6955,
"step": 2021
},
{
"epoch": 0.6979634104245771,
"grad_norm": 1.8248411417007446,
"learning_rate": 2e-05,
"loss": 0.7446,
"step": 2022
},
{
"epoch": 0.6983085950983776,
"grad_norm": 1.9997882843017578,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 2023
},
{
"epoch": 0.6986537797721781,
"grad_norm": 1.9308781623840332,
"learning_rate": 2e-05,
"loss": 0.757,
"step": 2024
},
{
"epoch": 0.6989989644459786,
"grad_norm": 1.934055209159851,
"learning_rate": 2e-05,
"loss": 0.6974,
"step": 2025
},
{
"epoch": 0.699344149119779,
"grad_norm": 1.7758277654647827,
"learning_rate": 2e-05,
"loss": 0.7264,
"step": 2026
},
{
"epoch": 0.6996893337935796,
"grad_norm": 1.9232991933822632,
"learning_rate": 2e-05,
"loss": 0.6958,
"step": 2027
},
{
"epoch": 0.7000345184673801,
"grad_norm": 1.7799192667007446,
"learning_rate": 2e-05,
"loss": 0.6787,
"step": 2028
},
{
"epoch": 0.7003797031411805,
"grad_norm": 1.8326587677001953,
"learning_rate": 2e-05,
"loss": 0.7522,
"step": 2029
},
{
"epoch": 0.700724887814981,
"grad_norm": 1.8313210010528564,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 2030
},
{
"epoch": 0.7010700724887815,
"grad_norm": 2.0515120029449463,
"learning_rate": 2e-05,
"loss": 0.7269,
"step": 2031
},
{
"epoch": 0.701415257162582,
"grad_norm": 1.939740777015686,
"learning_rate": 2e-05,
"loss": 0.7978,
"step": 2032
},
{
"epoch": 0.7017604418363824,
"grad_norm": 1.7957408428192139,
"learning_rate": 2e-05,
"loss": 0.7372,
"step": 2033
},
{
"epoch": 0.7021056265101829,
"grad_norm": 1.8494184017181396,
"learning_rate": 2e-05,
"loss": 0.7411,
"step": 2034
},
{
"epoch": 0.7024508111839834,
"grad_norm": 1.8921087980270386,
"learning_rate": 2e-05,
"loss": 0.7279,
"step": 2035
},
{
"epoch": 0.702795995857784,
"grad_norm": 1.8954356908798218,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 2036
},
{
"epoch": 0.7031411805315844,
"grad_norm": 1.9165939092636108,
"learning_rate": 2e-05,
"loss": 0.7106,
"step": 2037
},
{
"epoch": 0.7034863652053849,
"grad_norm": 1.7581866979599,
"learning_rate": 2e-05,
"loss": 0.7025,
"step": 2038
},
{
"epoch": 0.7038315498791854,
"grad_norm": 2.14585280418396,
"learning_rate": 2e-05,
"loss": 0.8079,
"step": 2039
},
{
"epoch": 0.7041767345529858,
"grad_norm": 1.7623639106750488,
"learning_rate": 2e-05,
"loss": 0.7908,
"step": 2040
},
{
"epoch": 0.7045219192267863,
"grad_norm": 1.8051908016204834,
"learning_rate": 2e-05,
"loss": 0.6696,
"step": 2041
},
{
"epoch": 0.7048671039005868,
"grad_norm": 1.7889599800109863,
"learning_rate": 2e-05,
"loss": 0.7209,
"step": 2042
},
{
"epoch": 0.7052122885743873,
"grad_norm": 1.9818137884140015,
"learning_rate": 2e-05,
"loss": 0.6993,
"step": 2043
},
{
"epoch": 0.7055574732481877,
"grad_norm": 1.819785237312317,
"learning_rate": 2e-05,
"loss": 0.7321,
"step": 2044
},
{
"epoch": 0.7059026579219883,
"grad_norm": 1.739596962928772,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 2045
},
{
"epoch": 0.7062478425957888,
"grad_norm": 1.9742025136947632,
"learning_rate": 2e-05,
"loss": 0.7372,
"step": 2046
},
{
"epoch": 0.7065930272695893,
"grad_norm": 1.6787482500076294,
"learning_rate": 2e-05,
"loss": 0.7119,
"step": 2047
},
{
"epoch": 0.7069382119433897,
"grad_norm": 2.0980167388916016,
"learning_rate": 2e-05,
"loss": 0.7711,
"step": 2048
},
{
"epoch": 0.7072833966171902,
"grad_norm": 1.8267072439193726,
"learning_rate": 2e-05,
"loss": 0.7034,
"step": 2049
},
{
"epoch": 0.7076285812909907,
"grad_norm": 2.0006978511810303,
"learning_rate": 2e-05,
"loss": 0.6688,
"step": 2050
},
{
"epoch": 0.7079737659647911,
"grad_norm": 1.8036727905273438,
"learning_rate": 2e-05,
"loss": 0.7444,
"step": 2051
},
{
"epoch": 0.7083189506385916,
"grad_norm": 2.205792188644409,
"learning_rate": 2e-05,
"loss": 0.7554,
"step": 2052
},
{
"epoch": 0.7086641353123921,
"grad_norm": 2.229905128479004,
"learning_rate": 2e-05,
"loss": 0.8077,
"step": 2053
},
{
"epoch": 0.7090093199861927,
"grad_norm": 1.7052178382873535,
"learning_rate": 2e-05,
"loss": 0.6925,
"step": 2054
},
{
"epoch": 0.7093545046599931,
"grad_norm": 1.8549474477767944,
"learning_rate": 2e-05,
"loss": 0.7352,
"step": 2055
},
{
"epoch": 0.7096996893337936,
"grad_norm": 1.9165446758270264,
"learning_rate": 2e-05,
"loss": 0.7456,
"step": 2056
},
{
"epoch": 0.7100448740075941,
"grad_norm": 1.9475483894348145,
"learning_rate": 2e-05,
"loss": 0.6674,
"step": 2057
},
{
"epoch": 0.7103900586813946,
"grad_norm": 1.9495691061019897,
"learning_rate": 2e-05,
"loss": 0.7915,
"step": 2058
},
{
"epoch": 0.710735243355195,
"grad_norm": 1.741241455078125,
"learning_rate": 2e-05,
"loss": 0.7516,
"step": 2059
},
{
"epoch": 0.7110804280289955,
"grad_norm": 1.7158613204956055,
"learning_rate": 2e-05,
"loss": 0.6747,
"step": 2060
},
{
"epoch": 0.711425612702796,
"grad_norm": 1.9132354259490967,
"learning_rate": 2e-05,
"loss": 0.7196,
"step": 2061
},
{
"epoch": 0.7117707973765964,
"grad_norm": 1.9160149097442627,
"learning_rate": 2e-05,
"loss": 0.8075,
"step": 2062
},
{
"epoch": 0.712115982050397,
"grad_norm": 1.8003803491592407,
"learning_rate": 2e-05,
"loss": 0.7105,
"step": 2063
},
{
"epoch": 0.7124611667241975,
"grad_norm": 1.9001412391662598,
"learning_rate": 2e-05,
"loss": 0.6968,
"step": 2064
},
{
"epoch": 0.712806351397998,
"grad_norm": 1.8029978275299072,
"learning_rate": 2e-05,
"loss": 0.7382,
"step": 2065
},
{
"epoch": 0.7131515360717984,
"grad_norm": 2.009681224822998,
"learning_rate": 2e-05,
"loss": 0.7468,
"step": 2066
},
{
"epoch": 0.7134967207455989,
"grad_norm": 1.6839584112167358,
"learning_rate": 2e-05,
"loss": 0.7051,
"step": 2067
},
{
"epoch": 0.7138419054193994,
"grad_norm": 2.071986675262451,
"learning_rate": 2e-05,
"loss": 0.7745,
"step": 2068
},
{
"epoch": 0.7141870900931999,
"grad_norm": 1.9866576194763184,
"learning_rate": 2e-05,
"loss": 0.747,
"step": 2069
},
{
"epoch": 0.7145322747670003,
"grad_norm": 1.9116586446762085,
"learning_rate": 2e-05,
"loss": 0.7545,
"step": 2070
},
{
"epoch": 0.7148774594408008,
"grad_norm": 2.0839502811431885,
"learning_rate": 2e-05,
"loss": 0.7296,
"step": 2071
},
{
"epoch": 0.7152226441146013,
"grad_norm": 1.7687896490097046,
"learning_rate": 2e-05,
"loss": 0.7593,
"step": 2072
},
{
"epoch": 0.7155678287884017,
"grad_norm": 1.9102952480316162,
"learning_rate": 2e-05,
"loss": 0.7739,
"step": 2073
},
{
"epoch": 0.7159130134622023,
"grad_norm": 1.8381487131118774,
"learning_rate": 2e-05,
"loss": 0.7069,
"step": 2074
},
{
"epoch": 0.7162581981360028,
"grad_norm": 2.54113507270813,
"learning_rate": 2e-05,
"loss": 0.8459,
"step": 2075
},
{
"epoch": 0.7166033828098033,
"grad_norm": 1.825212001800537,
"learning_rate": 2e-05,
"loss": 0.6583,
"step": 2076
},
{
"epoch": 0.7169485674836037,
"grad_norm": 1.848466396331787,
"learning_rate": 2e-05,
"loss": 0.7315,
"step": 2077
},
{
"epoch": 0.7172937521574042,
"grad_norm": 1.8835593461990356,
"learning_rate": 2e-05,
"loss": 0.7375,
"step": 2078
},
{
"epoch": 0.7176389368312047,
"grad_norm": 2.037550926208496,
"learning_rate": 2e-05,
"loss": 0.7731,
"step": 2079
},
{
"epoch": 0.7179841215050051,
"grad_norm": 1.9440861940383911,
"learning_rate": 2e-05,
"loss": 0.7646,
"step": 2080
},
{
"epoch": 0.7183293061788056,
"grad_norm": 1.8228204250335693,
"learning_rate": 2e-05,
"loss": 0.7163,
"step": 2081
},
{
"epoch": 0.7186744908526062,
"grad_norm": 2.002439498901367,
"learning_rate": 2e-05,
"loss": 0.7674,
"step": 2082
},
{
"epoch": 0.7190196755264067,
"grad_norm": 1.762283444404602,
"learning_rate": 2e-05,
"loss": 0.7085,
"step": 2083
},
{
"epoch": 0.7193648602002071,
"grad_norm": 1.9574036598205566,
"learning_rate": 2e-05,
"loss": 0.7417,
"step": 2084
},
{
"epoch": 0.7197100448740076,
"grad_norm": 1.95353364944458,
"learning_rate": 2e-05,
"loss": 0.7189,
"step": 2085
},
{
"epoch": 0.7200552295478081,
"grad_norm": 1.9906678199768066,
"learning_rate": 2e-05,
"loss": 0.7337,
"step": 2086
},
{
"epoch": 0.7204004142216086,
"grad_norm": 1.9747782945632935,
"learning_rate": 2e-05,
"loss": 0.7792,
"step": 2087
},
{
"epoch": 0.720745598895409,
"grad_norm": 1.9326345920562744,
"learning_rate": 2e-05,
"loss": 0.7144,
"step": 2088
},
{
"epoch": 0.7210907835692095,
"grad_norm": 1.884963035583496,
"learning_rate": 2e-05,
"loss": 0.7529,
"step": 2089
},
{
"epoch": 0.72143596824301,
"grad_norm": 1.875533103942871,
"learning_rate": 2e-05,
"loss": 0.7021,
"step": 2090
},
{
"epoch": 0.7217811529168104,
"grad_norm": 1.9843158721923828,
"learning_rate": 2e-05,
"loss": 0.6604,
"step": 2091
},
{
"epoch": 0.722126337590611,
"grad_norm": 1.9331083297729492,
"learning_rate": 2e-05,
"loss": 0.7086,
"step": 2092
},
{
"epoch": 0.7224715222644115,
"grad_norm": 2.4096901416778564,
"learning_rate": 2e-05,
"loss": 0.7865,
"step": 2093
},
{
"epoch": 0.722816706938212,
"grad_norm": 2.2345683574676514,
"learning_rate": 2e-05,
"loss": 0.779,
"step": 2094
},
{
"epoch": 0.7231618916120124,
"grad_norm": 2.060119390487671,
"learning_rate": 2e-05,
"loss": 0.7732,
"step": 2095
},
{
"epoch": 0.7235070762858129,
"grad_norm": 1.8734238147735596,
"learning_rate": 2e-05,
"loss": 0.7592,
"step": 2096
},
{
"epoch": 0.7238522609596134,
"grad_norm": 1.8172612190246582,
"learning_rate": 2e-05,
"loss": 0.6826,
"step": 2097
},
{
"epoch": 0.7241974456334139,
"grad_norm": 1.7159579992294312,
"learning_rate": 2e-05,
"loss": 0.7441,
"step": 2098
},
{
"epoch": 0.7245426303072143,
"grad_norm": 2.074211359024048,
"learning_rate": 2e-05,
"loss": 0.7122,
"step": 2099
},
{
"epoch": 0.7248878149810148,
"grad_norm": 1.7628990411758423,
"learning_rate": 2e-05,
"loss": 0.7273,
"step": 2100
},
{
"epoch": 0.7252329996548154,
"grad_norm": 1.7684333324432373,
"learning_rate": 2e-05,
"loss": 0.7111,
"step": 2101
},
{
"epoch": 0.7255781843286158,
"grad_norm": 1.977342128753662,
"learning_rate": 2e-05,
"loss": 0.6664,
"step": 2102
},
{
"epoch": 0.7259233690024163,
"grad_norm": 1.8608331680297852,
"learning_rate": 2e-05,
"loss": 0.7352,
"step": 2103
},
{
"epoch": 0.7262685536762168,
"grad_norm": 2.0203628540039062,
"learning_rate": 2e-05,
"loss": 0.6897,
"step": 2104
},
{
"epoch": 0.7266137383500173,
"grad_norm": 1.8506054878234863,
"learning_rate": 2e-05,
"loss": 0.6503,
"step": 2105
},
{
"epoch": 0.7269589230238177,
"grad_norm": 1.9112423658370972,
"learning_rate": 2e-05,
"loss": 0.8203,
"step": 2106
},
{
"epoch": 0.7273041076976182,
"grad_norm": 1.9576603174209595,
"learning_rate": 2e-05,
"loss": 0.789,
"step": 2107
},
{
"epoch": 0.7276492923714187,
"grad_norm": 2.0183277130126953,
"learning_rate": 2e-05,
"loss": 0.675,
"step": 2108
},
{
"epoch": 0.7279944770452192,
"grad_norm": 1.8968771696090698,
"learning_rate": 2e-05,
"loss": 0.726,
"step": 2109
},
{
"epoch": 0.7283396617190196,
"grad_norm": 1.9071968793869019,
"learning_rate": 2e-05,
"loss": 0.7008,
"step": 2110
},
{
"epoch": 0.7286848463928202,
"grad_norm": 1.8449773788452148,
"learning_rate": 2e-05,
"loss": 0.7346,
"step": 2111
},
{
"epoch": 0.7290300310666207,
"grad_norm": 1.894066333770752,
"learning_rate": 2e-05,
"loss": 0.6751,
"step": 2112
},
{
"epoch": 0.7293752157404211,
"grad_norm": 1.8265663385391235,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 2113
},
{
"epoch": 0.7297204004142216,
"grad_norm": 1.7281618118286133,
"learning_rate": 2e-05,
"loss": 0.7433,
"step": 2114
},
{
"epoch": 0.7300655850880221,
"grad_norm": 1.9177651405334473,
"learning_rate": 2e-05,
"loss": 0.7905,
"step": 2115
},
{
"epoch": 0.7304107697618226,
"grad_norm": 1.8048851490020752,
"learning_rate": 2e-05,
"loss": 0.7415,
"step": 2116
},
{
"epoch": 0.730755954435623,
"grad_norm": 1.8339844942092896,
"learning_rate": 2e-05,
"loss": 0.7659,
"step": 2117
},
{
"epoch": 0.7311011391094235,
"grad_norm": 1.8510236740112305,
"learning_rate": 2e-05,
"loss": 0.6755,
"step": 2118
},
{
"epoch": 0.731446323783224,
"grad_norm": 1.982960820198059,
"learning_rate": 2e-05,
"loss": 0.7072,
"step": 2119
},
{
"epoch": 0.7317915084570245,
"grad_norm": 1.8617637157440186,
"learning_rate": 2e-05,
"loss": 0.704,
"step": 2120
},
{
"epoch": 0.732136693130825,
"grad_norm": 1.9271409511566162,
"learning_rate": 2e-05,
"loss": 0.7561,
"step": 2121
},
{
"epoch": 0.7324818778046255,
"grad_norm": 2.065410852432251,
"learning_rate": 2e-05,
"loss": 0.7137,
"step": 2122
},
{
"epoch": 0.732827062478426,
"grad_norm": 1.8226072788238525,
"learning_rate": 2e-05,
"loss": 0.6565,
"step": 2123
},
{
"epoch": 0.7331722471522264,
"grad_norm": 1.830252766609192,
"learning_rate": 2e-05,
"loss": 0.7002,
"step": 2124
},
{
"epoch": 0.7335174318260269,
"grad_norm": 1.848960518836975,
"learning_rate": 2e-05,
"loss": 0.7243,
"step": 2125
},
{
"epoch": 0.7338626164998274,
"grad_norm": 2.117652416229248,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 2126
},
{
"epoch": 0.7342078011736279,
"grad_norm": 1.748321294784546,
"learning_rate": 2e-05,
"loss": 0.7248,
"step": 2127
},
{
"epoch": 0.7345529858474283,
"grad_norm": 1.8564791679382324,
"learning_rate": 2e-05,
"loss": 0.7351,
"step": 2128
},
{
"epoch": 0.7348981705212289,
"grad_norm": 2.061476945877075,
"learning_rate": 2e-05,
"loss": 0.7241,
"step": 2129
},
{
"epoch": 0.7352433551950294,
"grad_norm": 2.0772247314453125,
"learning_rate": 2e-05,
"loss": 0.7261,
"step": 2130
},
{
"epoch": 0.7355885398688298,
"grad_norm": 2.0089240074157715,
"learning_rate": 2e-05,
"loss": 0.7524,
"step": 2131
},
{
"epoch": 0.7359337245426303,
"grad_norm": 2.1447134017944336,
"learning_rate": 2e-05,
"loss": 0.7127,
"step": 2132
},
{
"epoch": 0.7362789092164308,
"grad_norm": 1.9239834547042847,
"learning_rate": 2e-05,
"loss": 0.782,
"step": 2133
},
{
"epoch": 0.7366240938902313,
"grad_norm": 2.000593900680542,
"learning_rate": 2e-05,
"loss": 0.7124,
"step": 2134
},
{
"epoch": 0.7369692785640317,
"grad_norm": 1.662696123123169,
"learning_rate": 2e-05,
"loss": 0.729,
"step": 2135
},
{
"epoch": 0.7373144632378322,
"grad_norm": 1.8220033645629883,
"learning_rate": 2e-05,
"loss": 0.6811,
"step": 2136
},
{
"epoch": 0.7376596479116327,
"grad_norm": 1.8143640756607056,
"learning_rate": 2e-05,
"loss": 0.7221,
"step": 2137
},
{
"epoch": 0.7380048325854333,
"grad_norm": 1.7895587682724,
"learning_rate": 2e-05,
"loss": 0.6983,
"step": 2138
},
{
"epoch": 0.7383500172592337,
"grad_norm": 2.052912712097168,
"learning_rate": 2e-05,
"loss": 0.7248,
"step": 2139
},
{
"epoch": 0.7386952019330342,
"grad_norm": 1.8995085954666138,
"learning_rate": 2e-05,
"loss": 0.6959,
"step": 2140
},
{
"epoch": 0.7390403866068347,
"grad_norm": 1.8756906986236572,
"learning_rate": 2e-05,
"loss": 0.6629,
"step": 2141
},
{
"epoch": 0.7393855712806351,
"grad_norm": 1.7140847444534302,
"learning_rate": 2e-05,
"loss": 0.777,
"step": 2142
},
{
"epoch": 0.7397307559544356,
"grad_norm": 2.1346499919891357,
"learning_rate": 2e-05,
"loss": 0.6801,
"step": 2143
},
{
"epoch": 0.7400759406282361,
"grad_norm": 1.8423312902450562,
"learning_rate": 2e-05,
"loss": 0.6968,
"step": 2144
},
{
"epoch": 0.7404211253020366,
"grad_norm": 2.020374298095703,
"learning_rate": 2e-05,
"loss": 0.756,
"step": 2145
},
{
"epoch": 0.740766309975837,
"grad_norm": 1.900222659111023,
"learning_rate": 2e-05,
"loss": 0.7478,
"step": 2146
},
{
"epoch": 0.7411114946496375,
"grad_norm": 1.745725154876709,
"learning_rate": 2e-05,
"loss": 0.727,
"step": 2147
},
{
"epoch": 0.7414566793234381,
"grad_norm": 1.7668520212173462,
"learning_rate": 2e-05,
"loss": 0.7006,
"step": 2148
},
{
"epoch": 0.7418018639972386,
"grad_norm": 1.7557846307754517,
"learning_rate": 2e-05,
"loss": 0.7361,
"step": 2149
},
{
"epoch": 0.742147048671039,
"grad_norm": 1.7820024490356445,
"learning_rate": 2e-05,
"loss": 0.7204,
"step": 2150
},
{
"epoch": 0.7424922333448395,
"grad_norm": 1.8768728971481323,
"learning_rate": 2e-05,
"loss": 0.7632,
"step": 2151
},
{
"epoch": 0.74283741801864,
"grad_norm": 1.9444397687911987,
"learning_rate": 2e-05,
"loss": 0.7004,
"step": 2152
},
{
"epoch": 0.7431826026924404,
"grad_norm": 2.0667853355407715,
"learning_rate": 2e-05,
"loss": 0.726,
"step": 2153
},
{
"epoch": 0.7435277873662409,
"grad_norm": 1.9062163829803467,
"learning_rate": 2e-05,
"loss": 0.7021,
"step": 2154
},
{
"epoch": 0.7438729720400414,
"grad_norm": 1.7155424356460571,
"learning_rate": 2e-05,
"loss": 0.6581,
"step": 2155
},
{
"epoch": 0.744218156713842,
"grad_norm": 2.166463613510132,
"learning_rate": 2e-05,
"loss": 0.7057,
"step": 2156
},
{
"epoch": 0.7445633413876424,
"grad_norm": 1.7151546478271484,
"learning_rate": 2e-05,
"loss": 0.72,
"step": 2157
},
{
"epoch": 0.7449085260614429,
"grad_norm": 3.0908069610595703,
"learning_rate": 2e-05,
"loss": 0.7009,
"step": 2158
},
{
"epoch": 0.7452537107352434,
"grad_norm": 1.8328068256378174,
"learning_rate": 2e-05,
"loss": 0.7571,
"step": 2159
},
{
"epoch": 0.7455988954090439,
"grad_norm": 1.7982745170593262,
"learning_rate": 2e-05,
"loss": 0.7369,
"step": 2160
},
{
"epoch": 0.7459440800828443,
"grad_norm": 1.7473913431167603,
"learning_rate": 2e-05,
"loss": 0.6908,
"step": 2161
},
{
"epoch": 0.7462892647566448,
"grad_norm": 1.8078123331069946,
"learning_rate": 2e-05,
"loss": 0.69,
"step": 2162
},
{
"epoch": 0.7466344494304453,
"grad_norm": 2.3278989791870117,
"learning_rate": 2e-05,
"loss": 0.727,
"step": 2163
},
{
"epoch": 0.7469796341042457,
"grad_norm": 1.96751868724823,
"learning_rate": 2e-05,
"loss": 0.7336,
"step": 2164
},
{
"epoch": 0.7473248187780462,
"grad_norm": 1.8072227239608765,
"learning_rate": 2e-05,
"loss": 0.7284,
"step": 2165
},
{
"epoch": 0.7476700034518468,
"grad_norm": 1.7643908262252808,
"learning_rate": 2e-05,
"loss": 0.6742,
"step": 2166
},
{
"epoch": 0.7480151881256473,
"grad_norm": 1.963390588760376,
"learning_rate": 2e-05,
"loss": 0.8295,
"step": 2167
},
{
"epoch": 0.7483603727994477,
"grad_norm": 1.7388458251953125,
"learning_rate": 2e-05,
"loss": 0.8024,
"step": 2168
},
{
"epoch": 0.7487055574732482,
"grad_norm": 1.847796082496643,
"learning_rate": 2e-05,
"loss": 0.7415,
"step": 2169
},
{
"epoch": 0.7490507421470487,
"grad_norm": 1.9620717763900757,
"learning_rate": 2e-05,
"loss": 0.7552,
"step": 2170
},
{
"epoch": 0.7493959268208491,
"grad_norm": 1.794451355934143,
"learning_rate": 2e-05,
"loss": 0.7155,
"step": 2171
},
{
"epoch": 0.7497411114946496,
"grad_norm": 1.9601366519927979,
"learning_rate": 2e-05,
"loss": 0.6927,
"step": 2172
},
{
"epoch": 0.7500862961684501,
"grad_norm": 2.0080158710479736,
"learning_rate": 2e-05,
"loss": 0.7349,
"step": 2173
},
{
"epoch": 0.7504314808422506,
"grad_norm": 1.9543483257293701,
"learning_rate": 2e-05,
"loss": 0.8098,
"step": 2174
},
{
"epoch": 0.750776665516051,
"grad_norm": 1.780216097831726,
"learning_rate": 2e-05,
"loss": 0.7167,
"step": 2175
},
{
"epoch": 0.7511218501898516,
"grad_norm": 1.9291138648986816,
"learning_rate": 2e-05,
"loss": 0.7834,
"step": 2176
},
{
"epoch": 0.7514670348636521,
"grad_norm": 2.0347471237182617,
"learning_rate": 2e-05,
"loss": 0.8375,
"step": 2177
},
{
"epoch": 0.7518122195374526,
"grad_norm": 1.8207837343215942,
"learning_rate": 2e-05,
"loss": 0.6998,
"step": 2178
},
{
"epoch": 0.752157404211253,
"grad_norm": 1.6159279346466064,
"learning_rate": 2e-05,
"loss": 0.7143,
"step": 2179
},
{
"epoch": 0.7525025888850535,
"grad_norm": 1.7967194318771362,
"learning_rate": 2e-05,
"loss": 0.6478,
"step": 2180
},
{
"epoch": 0.752847773558854,
"grad_norm": 1.9695372581481934,
"learning_rate": 2e-05,
"loss": 0.7426,
"step": 2181
},
{
"epoch": 0.7531929582326544,
"grad_norm": 1.8325481414794922,
"learning_rate": 2e-05,
"loss": 0.7128,
"step": 2182
},
{
"epoch": 0.7535381429064549,
"grad_norm": 1.7993351221084595,
"learning_rate": 2e-05,
"loss": 0.7181,
"step": 2183
},
{
"epoch": 0.7538833275802554,
"grad_norm": 1.794467568397522,
"learning_rate": 2e-05,
"loss": 0.6935,
"step": 2184
},
{
"epoch": 0.754228512254056,
"grad_norm": 1.692400336265564,
"learning_rate": 2e-05,
"loss": 0.6746,
"step": 2185
},
{
"epoch": 0.7545736969278564,
"grad_norm": 1.6721417903900146,
"learning_rate": 2e-05,
"loss": 0.6595,
"step": 2186
},
{
"epoch": 0.7549188816016569,
"grad_norm": 2.0058200359344482,
"learning_rate": 2e-05,
"loss": 0.7855,
"step": 2187
},
{
"epoch": 0.7552640662754574,
"grad_norm": 1.8826192617416382,
"learning_rate": 2e-05,
"loss": 0.7501,
"step": 2188
},
{
"epoch": 0.7556092509492579,
"grad_norm": 1.7434967756271362,
"learning_rate": 2e-05,
"loss": 0.7086,
"step": 2189
},
{
"epoch": 0.7559544356230583,
"grad_norm": 1.928126573562622,
"learning_rate": 2e-05,
"loss": 0.7644,
"step": 2190
},
{
"epoch": 0.7562996202968588,
"grad_norm": 1.6485679149627686,
"learning_rate": 2e-05,
"loss": 0.6727,
"step": 2191
},
{
"epoch": 0.7566448049706593,
"grad_norm": 1.791934609413147,
"learning_rate": 2e-05,
"loss": 0.7738,
"step": 2192
},
{
"epoch": 0.7569899896444597,
"grad_norm": 1.8998874425888062,
"learning_rate": 2e-05,
"loss": 0.6849,
"step": 2193
},
{
"epoch": 0.7573351743182603,
"grad_norm": 1.8552969694137573,
"learning_rate": 2e-05,
"loss": 0.728,
"step": 2194
},
{
"epoch": 0.7576803589920608,
"grad_norm": 1.8946011066436768,
"learning_rate": 2e-05,
"loss": 0.6749,
"step": 2195
},
{
"epoch": 0.7580255436658613,
"grad_norm": 1.903032660484314,
"learning_rate": 2e-05,
"loss": 0.7469,
"step": 2196
},
{
"epoch": 0.7583707283396617,
"grad_norm": 1.7411540746688843,
"learning_rate": 2e-05,
"loss": 0.617,
"step": 2197
},
{
"epoch": 0.7587159130134622,
"grad_norm": 2.280888319015503,
"learning_rate": 2e-05,
"loss": 0.6541,
"step": 2198
},
{
"epoch": 0.7590610976872627,
"grad_norm": 2.096526622772217,
"learning_rate": 2e-05,
"loss": 0.7119,
"step": 2199
},
{
"epoch": 0.7594062823610632,
"grad_norm": 2.057097911834717,
"learning_rate": 2e-05,
"loss": 0.7378,
"step": 2200
},
{
"epoch": 0.7597514670348636,
"grad_norm": 1.788740873336792,
"learning_rate": 2e-05,
"loss": 0.7226,
"step": 2201
},
{
"epoch": 0.7600966517086641,
"grad_norm": 2.155947685241699,
"learning_rate": 2e-05,
"loss": 0.8064,
"step": 2202
},
{
"epoch": 0.7604418363824647,
"grad_norm": 1.9464656114578247,
"learning_rate": 2e-05,
"loss": 0.6913,
"step": 2203
},
{
"epoch": 0.7607870210562651,
"grad_norm": 2.004995346069336,
"learning_rate": 2e-05,
"loss": 0.7272,
"step": 2204
},
{
"epoch": 0.7611322057300656,
"grad_norm": 1.860318899154663,
"learning_rate": 2e-05,
"loss": 0.7596,
"step": 2205
},
{
"epoch": 0.7614773904038661,
"grad_norm": 1.7992146015167236,
"learning_rate": 2e-05,
"loss": 0.802,
"step": 2206
},
{
"epoch": 0.7618225750776666,
"grad_norm": 1.9061779975891113,
"learning_rate": 2e-05,
"loss": 0.7204,
"step": 2207
},
{
"epoch": 0.762167759751467,
"grad_norm": 1.9315762519836426,
"learning_rate": 2e-05,
"loss": 0.6814,
"step": 2208
},
{
"epoch": 0.7625129444252675,
"grad_norm": 1.8451159000396729,
"learning_rate": 2e-05,
"loss": 0.7403,
"step": 2209
},
{
"epoch": 0.762858129099068,
"grad_norm": 1.8756223917007446,
"learning_rate": 2e-05,
"loss": 0.7263,
"step": 2210
},
{
"epoch": 0.7632033137728684,
"grad_norm": 1.8500291109085083,
"learning_rate": 2e-05,
"loss": 0.7417,
"step": 2211
},
{
"epoch": 0.7635484984466689,
"grad_norm": 1.9556981325149536,
"learning_rate": 2e-05,
"loss": 0.7541,
"step": 2212
},
{
"epoch": 0.7638936831204695,
"grad_norm": 3.820652723312378,
"learning_rate": 2e-05,
"loss": 0.7076,
"step": 2213
},
{
"epoch": 0.76423886779427,
"grad_norm": 1.8197849988937378,
"learning_rate": 2e-05,
"loss": 0.6893,
"step": 2214
},
{
"epoch": 0.7645840524680704,
"grad_norm": 2.0452497005462646,
"learning_rate": 2e-05,
"loss": 0.7006,
"step": 2215
},
{
"epoch": 0.7649292371418709,
"grad_norm": 1.7825652360916138,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 2216
},
{
"epoch": 0.7652744218156714,
"grad_norm": 1.984188199043274,
"learning_rate": 2e-05,
"loss": 0.6846,
"step": 2217
},
{
"epoch": 0.7656196064894719,
"grad_norm": 1.9198614358901978,
"learning_rate": 2e-05,
"loss": 0.7089,
"step": 2218
},
{
"epoch": 0.7659647911632723,
"grad_norm": 1.798697590827942,
"learning_rate": 2e-05,
"loss": 0.7457,
"step": 2219
},
{
"epoch": 0.7663099758370728,
"grad_norm": 2.0910837650299072,
"learning_rate": 2e-05,
"loss": 0.7043,
"step": 2220
},
{
"epoch": 0.7666551605108733,
"grad_norm": 1.9757916927337646,
"learning_rate": 2e-05,
"loss": 0.7999,
"step": 2221
},
{
"epoch": 0.7670003451846737,
"grad_norm": 1.9802093505859375,
"learning_rate": 2e-05,
"loss": 0.7817,
"step": 2222
},
{
"epoch": 0.7673455298584743,
"grad_norm": 1.8185251951217651,
"learning_rate": 2e-05,
"loss": 0.7298,
"step": 2223
},
{
"epoch": 0.7676907145322748,
"grad_norm": 1.749416708946228,
"learning_rate": 2e-05,
"loss": 0.6885,
"step": 2224
},
{
"epoch": 0.7680358992060753,
"grad_norm": 1.7374919652938843,
"learning_rate": 2e-05,
"loss": 0.7609,
"step": 2225
},
{
"epoch": 0.7683810838798757,
"grad_norm": 1.8804537057876587,
"learning_rate": 2e-05,
"loss": 0.7233,
"step": 2226
},
{
"epoch": 0.7687262685536762,
"grad_norm": 1.930681824684143,
"learning_rate": 2e-05,
"loss": 0.7446,
"step": 2227
},
{
"epoch": 0.7690714532274767,
"grad_norm": 1.8426949977874756,
"learning_rate": 2e-05,
"loss": 0.6811,
"step": 2228
},
{
"epoch": 0.7694166379012772,
"grad_norm": 1.7996834516525269,
"learning_rate": 2e-05,
"loss": 0.7278,
"step": 2229
},
{
"epoch": 0.7697618225750776,
"grad_norm": 1.9096876382827759,
"learning_rate": 2e-05,
"loss": 0.7477,
"step": 2230
},
{
"epoch": 0.7701070072488782,
"grad_norm": 1.836283564567566,
"learning_rate": 2e-05,
"loss": 0.6844,
"step": 2231
},
{
"epoch": 0.7704521919226787,
"grad_norm": 1.9448199272155762,
"learning_rate": 2e-05,
"loss": 0.7773,
"step": 2232
},
{
"epoch": 0.7707973765964791,
"grad_norm": 1.7022688388824463,
"learning_rate": 2e-05,
"loss": 0.6359,
"step": 2233
},
{
"epoch": 0.7711425612702796,
"grad_norm": 1.9062732458114624,
"learning_rate": 2e-05,
"loss": 0.7169,
"step": 2234
},
{
"epoch": 0.7714877459440801,
"grad_norm": 1.9543297290802002,
"learning_rate": 2e-05,
"loss": 0.708,
"step": 2235
},
{
"epoch": 0.7718329306178806,
"grad_norm": 1.9864168167114258,
"learning_rate": 2e-05,
"loss": 0.6234,
"step": 2236
},
{
"epoch": 0.772178115291681,
"grad_norm": 1.9136667251586914,
"learning_rate": 2e-05,
"loss": 0.7607,
"step": 2237
},
{
"epoch": 0.7725232999654815,
"grad_norm": 1.9810315370559692,
"learning_rate": 2e-05,
"loss": 0.7001,
"step": 2238
},
{
"epoch": 0.772868484639282,
"grad_norm": 1.924910068511963,
"learning_rate": 2e-05,
"loss": 0.6909,
"step": 2239
},
{
"epoch": 0.7732136693130826,
"grad_norm": 2.2119340896606445,
"learning_rate": 2e-05,
"loss": 0.7173,
"step": 2240
},
{
"epoch": 0.773558853986883,
"grad_norm": 1.7677148580551147,
"learning_rate": 2e-05,
"loss": 0.693,
"step": 2241
},
{
"epoch": 0.7739040386606835,
"grad_norm": 1.7973172664642334,
"learning_rate": 2e-05,
"loss": 0.698,
"step": 2242
},
{
"epoch": 0.774249223334484,
"grad_norm": 1.9511065483093262,
"learning_rate": 2e-05,
"loss": 0.7582,
"step": 2243
},
{
"epoch": 0.7745944080082844,
"grad_norm": 1.9313523769378662,
"learning_rate": 2e-05,
"loss": 0.7181,
"step": 2244
},
{
"epoch": 0.7749395926820849,
"grad_norm": 1.9995770454406738,
"learning_rate": 2e-05,
"loss": 0.6944,
"step": 2245
},
{
"epoch": 0.7752847773558854,
"grad_norm": 2.137603998184204,
"learning_rate": 2e-05,
"loss": 0.7036,
"step": 2246
},
{
"epoch": 0.7756299620296859,
"grad_norm": 1.8653441667556763,
"learning_rate": 2e-05,
"loss": 0.7655,
"step": 2247
},
{
"epoch": 0.7759751467034863,
"grad_norm": 2.0550620555877686,
"learning_rate": 2e-05,
"loss": 0.7339,
"step": 2248
},
{
"epoch": 0.7763203313772868,
"grad_norm": 2.1096105575561523,
"learning_rate": 2e-05,
"loss": 0.7104,
"step": 2249
},
{
"epoch": 0.7766655160510874,
"grad_norm": 1.8769968748092651,
"learning_rate": 2e-05,
"loss": 0.7195,
"step": 2250
},
{
"epoch": 0.7770107007248878,
"grad_norm": 1.773937702178955,
"learning_rate": 2e-05,
"loss": 0.7137,
"step": 2251
},
{
"epoch": 0.7773558853986883,
"grad_norm": 2.1285836696624756,
"learning_rate": 2e-05,
"loss": 0.7754,
"step": 2252
},
{
"epoch": 0.7777010700724888,
"grad_norm": 1.8876349925994873,
"learning_rate": 2e-05,
"loss": 0.7614,
"step": 2253
},
{
"epoch": 0.7780462547462893,
"grad_norm": 1.887521743774414,
"learning_rate": 2e-05,
"loss": 0.791,
"step": 2254
},
{
"epoch": 0.7783914394200897,
"grad_norm": 1.7976861000061035,
"learning_rate": 2e-05,
"loss": 0.6787,
"step": 2255
},
{
"epoch": 0.7787366240938902,
"grad_norm": 1.7389079332351685,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 2256
},
{
"epoch": 0.7790818087676907,
"grad_norm": 1.9697335958480835,
"learning_rate": 2e-05,
"loss": 0.6964,
"step": 2257
},
{
"epoch": 0.7794269934414912,
"grad_norm": 1.973560094833374,
"learning_rate": 2e-05,
"loss": 0.7592,
"step": 2258
},
{
"epoch": 0.7797721781152916,
"grad_norm": 1.9148238897323608,
"learning_rate": 2e-05,
"loss": 0.6688,
"step": 2259
},
{
"epoch": 0.7801173627890922,
"grad_norm": 1.8295315504074097,
"learning_rate": 2e-05,
"loss": 0.694,
"step": 2260
},
{
"epoch": 0.7804625474628927,
"grad_norm": 1.8120843172073364,
"learning_rate": 2e-05,
"loss": 0.6739,
"step": 2261
},
{
"epoch": 0.7808077321366931,
"grad_norm": 1.881322979927063,
"learning_rate": 2e-05,
"loss": 0.7371,
"step": 2262
},
{
"epoch": 0.7811529168104936,
"grad_norm": 2.14341402053833,
"learning_rate": 2e-05,
"loss": 0.7169,
"step": 2263
},
{
"epoch": 0.7814981014842941,
"grad_norm": 1.8698093891143799,
"learning_rate": 2e-05,
"loss": 0.7019,
"step": 2264
},
{
"epoch": 0.7818432861580946,
"grad_norm": 1.8601022958755493,
"learning_rate": 2e-05,
"loss": 0.7247,
"step": 2265
},
{
"epoch": 0.782188470831895,
"grad_norm": 1.9398045539855957,
"learning_rate": 2e-05,
"loss": 0.7095,
"step": 2266
},
{
"epoch": 0.7825336555056955,
"grad_norm": 1.8816583156585693,
"learning_rate": 2e-05,
"loss": 0.7139,
"step": 2267
},
{
"epoch": 0.782878840179496,
"grad_norm": 1.7741619348526,
"learning_rate": 2e-05,
"loss": 0.7315,
"step": 2268
},
{
"epoch": 0.7832240248532966,
"grad_norm": 2.233764886856079,
"learning_rate": 2e-05,
"loss": 0.7374,
"step": 2269
},
{
"epoch": 0.783569209527097,
"grad_norm": 1.6425055265426636,
"learning_rate": 2e-05,
"loss": 0.6725,
"step": 2270
},
{
"epoch": 0.7839143942008975,
"grad_norm": 1.720949411392212,
"learning_rate": 2e-05,
"loss": 0.6638,
"step": 2271
},
{
"epoch": 0.784259578874698,
"grad_norm": 1.8727104663848877,
"learning_rate": 2e-05,
"loss": 0.6937,
"step": 2272
},
{
"epoch": 0.7846047635484984,
"grad_norm": 1.89191472530365,
"learning_rate": 2e-05,
"loss": 0.7268,
"step": 2273
},
{
"epoch": 0.7849499482222989,
"grad_norm": 1.763363242149353,
"learning_rate": 2e-05,
"loss": 0.7247,
"step": 2274
},
{
"epoch": 0.7852951328960994,
"grad_norm": 2.358360528945923,
"learning_rate": 2e-05,
"loss": 0.7657,
"step": 2275
},
{
"epoch": 0.7856403175698999,
"grad_norm": 1.8516687154769897,
"learning_rate": 2e-05,
"loss": 0.7159,
"step": 2276
},
{
"epoch": 0.7859855022437003,
"grad_norm": 1.773671269416809,
"learning_rate": 2e-05,
"loss": 0.6797,
"step": 2277
},
{
"epoch": 0.7863306869175009,
"grad_norm": 1.8288607597351074,
"learning_rate": 2e-05,
"loss": 0.6965,
"step": 2278
},
{
"epoch": 0.7866758715913014,
"grad_norm": 1.8652377128601074,
"learning_rate": 2e-05,
"loss": 0.7625,
"step": 2279
},
{
"epoch": 0.7870210562651019,
"grad_norm": 1.91368567943573,
"learning_rate": 2e-05,
"loss": 0.663,
"step": 2280
},
{
"epoch": 0.7873662409389023,
"grad_norm": 1.8277349472045898,
"learning_rate": 2e-05,
"loss": 0.7714,
"step": 2281
},
{
"epoch": 0.7877114256127028,
"grad_norm": 1.8344676494598389,
"learning_rate": 2e-05,
"loss": 0.7044,
"step": 2282
},
{
"epoch": 0.7880566102865033,
"grad_norm": 1.981057047843933,
"learning_rate": 2e-05,
"loss": 0.7696,
"step": 2283
},
{
"epoch": 0.7884017949603037,
"grad_norm": 1.6478639841079712,
"learning_rate": 2e-05,
"loss": 0.6907,
"step": 2284
},
{
"epoch": 0.7887469796341042,
"grad_norm": 1.9392811059951782,
"learning_rate": 2e-05,
"loss": 0.688,
"step": 2285
},
{
"epoch": 0.7890921643079047,
"grad_norm": 2.109785556793213,
"learning_rate": 2e-05,
"loss": 0.732,
"step": 2286
},
{
"epoch": 0.7894373489817053,
"grad_norm": 1.798567295074463,
"learning_rate": 2e-05,
"loss": 0.7151,
"step": 2287
},
{
"epoch": 0.7897825336555057,
"grad_norm": 1.945080041885376,
"learning_rate": 2e-05,
"loss": 0.7357,
"step": 2288
},
{
"epoch": 0.7901277183293062,
"grad_norm": 1.827257752418518,
"learning_rate": 2e-05,
"loss": 0.6932,
"step": 2289
},
{
"epoch": 0.7904729030031067,
"grad_norm": 1.7544363737106323,
"learning_rate": 2e-05,
"loss": 0.7024,
"step": 2290
},
{
"epoch": 0.7908180876769072,
"grad_norm": 1.6462228298187256,
"learning_rate": 2e-05,
"loss": 0.7239,
"step": 2291
},
{
"epoch": 0.7911632723507076,
"grad_norm": 1.832653522491455,
"learning_rate": 2e-05,
"loss": 0.6876,
"step": 2292
},
{
"epoch": 0.7915084570245081,
"grad_norm": 2.081063985824585,
"learning_rate": 2e-05,
"loss": 0.684,
"step": 2293
},
{
"epoch": 0.7918536416983086,
"grad_norm": 1.8453823328018188,
"learning_rate": 2e-05,
"loss": 0.7973,
"step": 2294
},
{
"epoch": 0.792198826372109,
"grad_norm": 2.0634524822235107,
"learning_rate": 2e-05,
"loss": 0.7801,
"step": 2295
},
{
"epoch": 0.7925440110459095,
"grad_norm": 2.011003017425537,
"learning_rate": 2e-05,
"loss": 0.7404,
"step": 2296
},
{
"epoch": 0.7928891957197101,
"grad_norm": 1.7701154947280884,
"learning_rate": 2e-05,
"loss": 0.7106,
"step": 2297
},
{
"epoch": 0.7932343803935106,
"grad_norm": 1.8254767656326294,
"learning_rate": 2e-05,
"loss": 0.7002,
"step": 2298
},
{
"epoch": 0.793579565067311,
"grad_norm": 1.9096616506576538,
"learning_rate": 2e-05,
"loss": 0.7038,
"step": 2299
},
{
"epoch": 0.7939247497411115,
"grad_norm": 2.1505000591278076,
"learning_rate": 2e-05,
"loss": 0.7402,
"step": 2300
},
{
"epoch": 0.794269934414912,
"grad_norm": 1.8653305768966675,
"learning_rate": 2e-05,
"loss": 0.7266,
"step": 2301
},
{
"epoch": 0.7946151190887124,
"grad_norm": 1.8692653179168701,
"learning_rate": 2e-05,
"loss": 0.714,
"step": 2302
},
{
"epoch": 0.7949603037625129,
"grad_norm": 2.019106864929199,
"learning_rate": 2e-05,
"loss": 0.6803,
"step": 2303
},
{
"epoch": 0.7953054884363134,
"grad_norm": 1.9895415306091309,
"learning_rate": 2e-05,
"loss": 0.7817,
"step": 2304
},
{
"epoch": 0.795650673110114,
"grad_norm": 1.913293480873108,
"learning_rate": 2e-05,
"loss": 0.757,
"step": 2305
},
{
"epoch": 0.7959958577839144,
"grad_norm": 1.956986665725708,
"learning_rate": 2e-05,
"loss": 0.6805,
"step": 2306
},
{
"epoch": 0.7963410424577149,
"grad_norm": 1.8960682153701782,
"learning_rate": 2e-05,
"loss": 0.7473,
"step": 2307
},
{
"epoch": 0.7966862271315154,
"grad_norm": 2.017542600631714,
"learning_rate": 2e-05,
"loss": 0.7342,
"step": 2308
},
{
"epoch": 0.7970314118053159,
"grad_norm": 1.8778269290924072,
"learning_rate": 2e-05,
"loss": 0.7086,
"step": 2309
},
{
"epoch": 0.7973765964791163,
"grad_norm": 1.8838396072387695,
"learning_rate": 2e-05,
"loss": 0.7436,
"step": 2310
},
{
"epoch": 0.7977217811529168,
"grad_norm": 1.771117925643921,
"learning_rate": 2e-05,
"loss": 0.6844,
"step": 2311
},
{
"epoch": 0.7980669658267173,
"grad_norm": 1.8715964555740356,
"learning_rate": 2e-05,
"loss": 0.6718,
"step": 2312
},
{
"epoch": 0.7984121505005177,
"grad_norm": 1.750607967376709,
"learning_rate": 2e-05,
"loss": 0.6763,
"step": 2313
},
{
"epoch": 0.7987573351743182,
"grad_norm": 1.7299509048461914,
"learning_rate": 2e-05,
"loss": 0.7107,
"step": 2314
},
{
"epoch": 0.7991025198481188,
"grad_norm": 2.1369569301605225,
"learning_rate": 2e-05,
"loss": 0.6678,
"step": 2315
},
{
"epoch": 0.7994477045219193,
"grad_norm": 1.7312870025634766,
"learning_rate": 2e-05,
"loss": 0.6949,
"step": 2316
},
{
"epoch": 0.7997928891957197,
"grad_norm": 1.7270495891571045,
"learning_rate": 2e-05,
"loss": 0.6667,
"step": 2317
},
{
"epoch": 0.8001380738695202,
"grad_norm": 1.8034775257110596,
"learning_rate": 2e-05,
"loss": 0.6838,
"step": 2318
},
{
"epoch": 0.8004832585433207,
"grad_norm": 1.7729077339172363,
"learning_rate": 2e-05,
"loss": 0.7192,
"step": 2319
},
{
"epoch": 0.8008284432171212,
"grad_norm": 1.698211908340454,
"learning_rate": 2e-05,
"loss": 0.6974,
"step": 2320
},
{
"epoch": 0.8011736278909216,
"grad_norm": 2.0015385150909424,
"learning_rate": 2e-05,
"loss": 0.7715,
"step": 2321
},
{
"epoch": 0.8015188125647221,
"grad_norm": 1.9835069179534912,
"learning_rate": 2e-05,
"loss": 0.7767,
"step": 2322
},
{
"epoch": 0.8018639972385226,
"grad_norm": 1.8354220390319824,
"learning_rate": 2e-05,
"loss": 0.7391,
"step": 2323
},
{
"epoch": 0.802209181912323,
"grad_norm": 1.7107927799224854,
"learning_rate": 2e-05,
"loss": 0.7272,
"step": 2324
},
{
"epoch": 0.8025543665861236,
"grad_norm": 1.6055887937545776,
"learning_rate": 2e-05,
"loss": 0.695,
"step": 2325
},
{
"epoch": 0.8028995512599241,
"grad_norm": 1.9464995861053467,
"learning_rate": 2e-05,
"loss": 0.7281,
"step": 2326
},
{
"epoch": 0.8032447359337246,
"grad_norm": 1.7567113637924194,
"learning_rate": 2e-05,
"loss": 0.7236,
"step": 2327
},
{
"epoch": 0.803589920607525,
"grad_norm": 1.8707444667816162,
"learning_rate": 2e-05,
"loss": 0.7539,
"step": 2328
},
{
"epoch": 0.8039351052813255,
"grad_norm": 1.9063929319381714,
"learning_rate": 2e-05,
"loss": 0.7015,
"step": 2329
},
{
"epoch": 0.804280289955126,
"grad_norm": 1.7227438688278198,
"learning_rate": 2e-05,
"loss": 0.6719,
"step": 2330
},
{
"epoch": 0.8046254746289265,
"grad_norm": 1.9717336893081665,
"learning_rate": 2e-05,
"loss": 0.7043,
"step": 2331
},
{
"epoch": 0.8049706593027269,
"grad_norm": 1.9405338764190674,
"learning_rate": 2e-05,
"loss": 0.6828,
"step": 2332
},
{
"epoch": 0.8053158439765274,
"grad_norm": 1.925032377243042,
"learning_rate": 2e-05,
"loss": 0.7159,
"step": 2333
},
{
"epoch": 0.805661028650328,
"grad_norm": 1.8302090167999268,
"learning_rate": 2e-05,
"loss": 0.7258,
"step": 2334
},
{
"epoch": 0.8060062133241284,
"grad_norm": 1.7830702066421509,
"learning_rate": 2e-05,
"loss": 0.7066,
"step": 2335
},
{
"epoch": 0.8063513979979289,
"grad_norm": 1.7490566968917847,
"learning_rate": 2e-05,
"loss": 0.7307,
"step": 2336
},
{
"epoch": 0.8066965826717294,
"grad_norm": 1.781952142715454,
"learning_rate": 2e-05,
"loss": 0.6805,
"step": 2337
},
{
"epoch": 0.8070417673455299,
"grad_norm": 1.9819201231002808,
"learning_rate": 2e-05,
"loss": 0.7596,
"step": 2338
},
{
"epoch": 0.8073869520193303,
"grad_norm": 1.6554046869277954,
"learning_rate": 2e-05,
"loss": 0.7386,
"step": 2339
},
{
"epoch": 0.8077321366931308,
"grad_norm": 1.8829734325408936,
"learning_rate": 2e-05,
"loss": 0.7307,
"step": 2340
},
{
"epoch": 0.8080773213669313,
"grad_norm": 1.85360586643219,
"learning_rate": 2e-05,
"loss": 0.7367,
"step": 2341
},
{
"epoch": 0.8084225060407317,
"grad_norm": 1.7930670976638794,
"learning_rate": 2e-05,
"loss": 0.7208,
"step": 2342
},
{
"epoch": 0.8087676907145323,
"grad_norm": 2.1271955966949463,
"learning_rate": 2e-05,
"loss": 0.7462,
"step": 2343
},
{
"epoch": 0.8091128753883328,
"grad_norm": 1.815656065940857,
"learning_rate": 2e-05,
"loss": 0.7067,
"step": 2344
},
{
"epoch": 0.8094580600621333,
"grad_norm": 1.7236554622650146,
"learning_rate": 2e-05,
"loss": 0.6933,
"step": 2345
},
{
"epoch": 0.8098032447359337,
"grad_norm": 1.8129690885543823,
"learning_rate": 2e-05,
"loss": 0.704,
"step": 2346
},
{
"epoch": 0.8101484294097342,
"grad_norm": 1.6459614038467407,
"learning_rate": 2e-05,
"loss": 0.7515,
"step": 2347
},
{
"epoch": 0.8104936140835347,
"grad_norm": 2.4178147315979004,
"learning_rate": 2e-05,
"loss": 0.6995,
"step": 2348
},
{
"epoch": 0.8108387987573352,
"grad_norm": 1.986425518989563,
"learning_rate": 2e-05,
"loss": 0.6942,
"step": 2349
},
{
"epoch": 0.8111839834311356,
"grad_norm": 1.9129230976104736,
"learning_rate": 2e-05,
"loss": 0.6921,
"step": 2350
},
{
"epoch": 0.8115291681049361,
"grad_norm": 1.806161880493164,
"learning_rate": 2e-05,
"loss": 0.7123,
"step": 2351
},
{
"epoch": 0.8118743527787367,
"grad_norm": 1.8187730312347412,
"learning_rate": 2e-05,
"loss": 0.6827,
"step": 2352
},
{
"epoch": 0.812219537452537,
"grad_norm": 2.032099485397339,
"learning_rate": 2e-05,
"loss": 0.8005,
"step": 2353
},
{
"epoch": 0.8125647221263376,
"grad_norm": 1.8090797662734985,
"learning_rate": 2e-05,
"loss": 0.7675,
"step": 2354
},
{
"epoch": 0.8129099068001381,
"grad_norm": 1.8130072355270386,
"learning_rate": 2e-05,
"loss": 0.666,
"step": 2355
},
{
"epoch": 0.8132550914739386,
"grad_norm": 1.7446709871292114,
"learning_rate": 2e-05,
"loss": 0.721,
"step": 2356
},
{
"epoch": 0.813600276147739,
"grad_norm": 1.9640055894851685,
"learning_rate": 2e-05,
"loss": 0.7694,
"step": 2357
},
{
"epoch": 0.8139454608215395,
"grad_norm": 1.856859564781189,
"learning_rate": 2e-05,
"loss": 0.6725,
"step": 2358
},
{
"epoch": 0.81429064549534,
"grad_norm": 1.8081461191177368,
"learning_rate": 2e-05,
"loss": 0.744,
"step": 2359
},
{
"epoch": 0.8146358301691405,
"grad_norm": 2.008338689804077,
"learning_rate": 2e-05,
"loss": 0.7483,
"step": 2360
},
{
"epoch": 0.8149810148429409,
"grad_norm": 1.8773784637451172,
"learning_rate": 2e-05,
"loss": 0.7604,
"step": 2361
},
{
"epoch": 0.8153261995167415,
"grad_norm": 1.7969869375228882,
"learning_rate": 2e-05,
"loss": 0.6881,
"step": 2362
},
{
"epoch": 0.815671384190542,
"grad_norm": 2.0645508766174316,
"learning_rate": 2e-05,
"loss": 0.7308,
"step": 2363
},
{
"epoch": 0.8160165688643424,
"grad_norm": 1.7452669143676758,
"learning_rate": 2e-05,
"loss": 0.6609,
"step": 2364
},
{
"epoch": 0.8163617535381429,
"grad_norm": 1.6699029207229614,
"learning_rate": 2e-05,
"loss": 0.6826,
"step": 2365
},
{
"epoch": 0.8167069382119434,
"grad_norm": 1.985536813735962,
"learning_rate": 2e-05,
"loss": 0.6832,
"step": 2366
},
{
"epoch": 0.8170521228857439,
"grad_norm": 1.836500883102417,
"learning_rate": 2e-05,
"loss": 0.7064,
"step": 2367
},
{
"epoch": 0.8173973075595443,
"grad_norm": 2.0566537380218506,
"learning_rate": 2e-05,
"loss": 0.7221,
"step": 2368
},
{
"epoch": 0.8177424922333448,
"grad_norm": 1.7984514236450195,
"learning_rate": 2e-05,
"loss": 0.7483,
"step": 2369
},
{
"epoch": 0.8180876769071453,
"grad_norm": 1.9777190685272217,
"learning_rate": 2e-05,
"loss": 0.7166,
"step": 2370
},
{
"epoch": 0.8184328615809459,
"grad_norm": 1.7557132244110107,
"learning_rate": 2e-05,
"loss": 0.6833,
"step": 2371
},
{
"epoch": 0.8187780462547463,
"grad_norm": 1.8503764867782593,
"learning_rate": 2e-05,
"loss": 0.6779,
"step": 2372
},
{
"epoch": 0.8191232309285468,
"grad_norm": 1.8787211179733276,
"learning_rate": 2e-05,
"loss": 0.7074,
"step": 2373
},
{
"epoch": 0.8194684156023473,
"grad_norm": 1.7992881536483765,
"learning_rate": 2e-05,
"loss": 0.748,
"step": 2374
},
{
"epoch": 0.8198136002761477,
"grad_norm": 1.6995468139648438,
"learning_rate": 2e-05,
"loss": 0.6927,
"step": 2375
},
{
"epoch": 0.8201587849499482,
"grad_norm": 1.8903498649597168,
"learning_rate": 2e-05,
"loss": 0.7561,
"step": 2376
},
{
"epoch": 0.8205039696237487,
"grad_norm": 1.9479565620422363,
"learning_rate": 2e-05,
"loss": 0.7135,
"step": 2377
},
{
"epoch": 0.8208491542975492,
"grad_norm": 1.8155922889709473,
"learning_rate": 2e-05,
"loss": 0.6736,
"step": 2378
},
{
"epoch": 0.8211943389713496,
"grad_norm": 1.8316346406936646,
"learning_rate": 2e-05,
"loss": 0.83,
"step": 2379
},
{
"epoch": 0.8215395236451501,
"grad_norm": 1.8841546773910522,
"learning_rate": 2e-05,
"loss": 0.7051,
"step": 2380
},
{
"epoch": 0.8218847083189507,
"grad_norm": 1.8500205278396606,
"learning_rate": 2e-05,
"loss": 0.6652,
"step": 2381
},
{
"epoch": 0.8222298929927512,
"grad_norm": 1.7064632177352905,
"learning_rate": 2e-05,
"loss": 0.6339,
"step": 2382
},
{
"epoch": 0.8225750776665516,
"grad_norm": 1.72679603099823,
"learning_rate": 2e-05,
"loss": 0.761,
"step": 2383
},
{
"epoch": 0.8229202623403521,
"grad_norm": 1.6895649433135986,
"learning_rate": 2e-05,
"loss": 0.7072,
"step": 2384
},
{
"epoch": 0.8232654470141526,
"grad_norm": 1.7633329629898071,
"learning_rate": 2e-05,
"loss": 0.6919,
"step": 2385
},
{
"epoch": 0.823610631687953,
"grad_norm": 1.6411778926849365,
"learning_rate": 2e-05,
"loss": 0.655,
"step": 2386
},
{
"epoch": 0.8239558163617535,
"grad_norm": 1.9904199838638306,
"learning_rate": 2e-05,
"loss": 0.7078,
"step": 2387
},
{
"epoch": 0.824301001035554,
"grad_norm": 1.835594654083252,
"learning_rate": 2e-05,
"loss": 0.6824,
"step": 2388
},
{
"epoch": 0.8246461857093546,
"grad_norm": 1.8450719118118286,
"learning_rate": 2e-05,
"loss": 0.7161,
"step": 2389
},
{
"epoch": 0.824991370383155,
"grad_norm": 1.7024744749069214,
"learning_rate": 2e-05,
"loss": 0.6781,
"step": 2390
},
{
"epoch": 0.8253365550569555,
"grad_norm": 1.7562062740325928,
"learning_rate": 2e-05,
"loss": 0.66,
"step": 2391
},
{
"epoch": 0.825681739730756,
"grad_norm": 1.8880400657653809,
"learning_rate": 2e-05,
"loss": 0.7011,
"step": 2392
},
{
"epoch": 0.8260269244045564,
"grad_norm": 1.9267442226409912,
"learning_rate": 2e-05,
"loss": 0.7812,
"step": 2393
},
{
"epoch": 0.8263721090783569,
"grad_norm": 2.140005350112915,
"learning_rate": 2e-05,
"loss": 0.7674,
"step": 2394
},
{
"epoch": 0.8267172937521574,
"grad_norm": 1.7910387516021729,
"learning_rate": 2e-05,
"loss": 0.6861,
"step": 2395
},
{
"epoch": 0.8270624784259579,
"grad_norm": 1.8379400968551636,
"learning_rate": 2e-05,
"loss": 0.7572,
"step": 2396
},
{
"epoch": 0.8274076630997583,
"grad_norm": 1.6917898654937744,
"learning_rate": 2e-05,
"loss": 0.6558,
"step": 2397
},
{
"epoch": 0.8277528477735588,
"grad_norm": 1.8333762884140015,
"learning_rate": 2e-05,
"loss": 0.8155,
"step": 2398
},
{
"epoch": 0.8280980324473594,
"grad_norm": 1.7440677881240845,
"learning_rate": 2e-05,
"loss": 0.6876,
"step": 2399
},
{
"epoch": 0.8284432171211599,
"grad_norm": 1.9367988109588623,
"learning_rate": 2e-05,
"loss": 0.727,
"step": 2400
},
{
"epoch": 0.8287884017949603,
"grad_norm": 2.0004966259002686,
"learning_rate": 2e-05,
"loss": 0.7734,
"step": 2401
},
{
"epoch": 0.8291335864687608,
"grad_norm": 1.7503799200057983,
"learning_rate": 2e-05,
"loss": 0.7088,
"step": 2402
},
{
"epoch": 0.8294787711425613,
"grad_norm": 1.93806791305542,
"learning_rate": 2e-05,
"loss": 0.7796,
"step": 2403
},
{
"epoch": 0.8298239558163617,
"grad_norm": 1.7855018377304077,
"learning_rate": 2e-05,
"loss": 0.724,
"step": 2404
},
{
"epoch": 0.8301691404901622,
"grad_norm": 1.6782795190811157,
"learning_rate": 2e-05,
"loss": 0.6318,
"step": 2405
},
{
"epoch": 0.8305143251639627,
"grad_norm": 1.8783369064331055,
"learning_rate": 2e-05,
"loss": 0.7538,
"step": 2406
},
{
"epoch": 0.8308595098377632,
"grad_norm": 1.8626327514648438,
"learning_rate": 2e-05,
"loss": 0.7138,
"step": 2407
},
{
"epoch": 0.8312046945115636,
"grad_norm": 1.8827041387557983,
"learning_rate": 2e-05,
"loss": 0.7384,
"step": 2408
},
{
"epoch": 0.8315498791853642,
"grad_norm": 1.6890876293182373,
"learning_rate": 2e-05,
"loss": 0.775,
"step": 2409
},
{
"epoch": 0.8318950638591647,
"grad_norm": 1.974753499031067,
"learning_rate": 2e-05,
"loss": 0.7012,
"step": 2410
},
{
"epoch": 0.8322402485329652,
"grad_norm": 1.7650669813156128,
"learning_rate": 2e-05,
"loss": 0.7252,
"step": 2411
},
{
"epoch": 0.8325854332067656,
"grad_norm": 1.757652759552002,
"learning_rate": 2e-05,
"loss": 0.6943,
"step": 2412
},
{
"epoch": 0.8329306178805661,
"grad_norm": 1.6527223587036133,
"learning_rate": 2e-05,
"loss": 0.6663,
"step": 2413
},
{
"epoch": 0.8332758025543666,
"grad_norm": 1.8536051511764526,
"learning_rate": 2e-05,
"loss": 0.7196,
"step": 2414
},
{
"epoch": 0.833620987228167,
"grad_norm": 1.9275211095809937,
"learning_rate": 2e-05,
"loss": 0.7224,
"step": 2415
},
{
"epoch": 0.8339661719019675,
"grad_norm": 1.9716124534606934,
"learning_rate": 2e-05,
"loss": 0.7736,
"step": 2416
},
{
"epoch": 0.834311356575768,
"grad_norm": 2.000140428543091,
"learning_rate": 2e-05,
"loss": 0.7289,
"step": 2417
},
{
"epoch": 0.8346565412495686,
"grad_norm": 1.8189802169799805,
"learning_rate": 2e-05,
"loss": 0.7624,
"step": 2418
},
{
"epoch": 0.835001725923369,
"grad_norm": 1.8551133871078491,
"learning_rate": 2e-05,
"loss": 0.7388,
"step": 2419
},
{
"epoch": 0.8353469105971695,
"grad_norm": 1.7845901250839233,
"learning_rate": 2e-05,
"loss": 0.7893,
"step": 2420
},
{
"epoch": 0.83569209527097,
"grad_norm": 1.7457618713378906,
"learning_rate": 2e-05,
"loss": 0.7069,
"step": 2421
},
{
"epoch": 0.8360372799447705,
"grad_norm": 1.9464480876922607,
"learning_rate": 2e-05,
"loss": 0.7782,
"step": 2422
},
{
"epoch": 0.8363824646185709,
"grad_norm": 1.8111666440963745,
"learning_rate": 2e-05,
"loss": 0.7481,
"step": 2423
},
{
"epoch": 0.8367276492923714,
"grad_norm": 1.8378905057907104,
"learning_rate": 2e-05,
"loss": 0.6565,
"step": 2424
},
{
"epoch": 0.8370728339661719,
"grad_norm": 1.7083327770233154,
"learning_rate": 2e-05,
"loss": 0.6508,
"step": 2425
},
{
"epoch": 0.8374180186399723,
"grad_norm": 1.9257265329360962,
"learning_rate": 2e-05,
"loss": 0.701,
"step": 2426
},
{
"epoch": 0.8377632033137729,
"grad_norm": 1.7853341102600098,
"learning_rate": 2e-05,
"loss": 0.6806,
"step": 2427
},
{
"epoch": 0.8381083879875734,
"grad_norm": 1.9096763134002686,
"learning_rate": 2e-05,
"loss": 0.6671,
"step": 2428
},
{
"epoch": 0.8384535726613739,
"grad_norm": 1.919541358947754,
"learning_rate": 2e-05,
"loss": 0.7652,
"step": 2429
},
{
"epoch": 0.8387987573351743,
"grad_norm": 2.1240127086639404,
"learning_rate": 2e-05,
"loss": 0.7161,
"step": 2430
},
{
"epoch": 0.8391439420089748,
"grad_norm": 1.762609601020813,
"learning_rate": 2e-05,
"loss": 0.7203,
"step": 2431
},
{
"epoch": 0.8394891266827753,
"grad_norm": 1.7753833532333374,
"learning_rate": 2e-05,
"loss": 0.8078,
"step": 2432
},
{
"epoch": 0.8398343113565757,
"grad_norm": 1.8390542268753052,
"learning_rate": 2e-05,
"loss": 0.6837,
"step": 2433
},
{
"epoch": 0.8401794960303762,
"grad_norm": 1.7845293283462524,
"learning_rate": 2e-05,
"loss": 0.7843,
"step": 2434
},
{
"epoch": 0.8405246807041767,
"grad_norm": 2.0822293758392334,
"learning_rate": 2e-05,
"loss": 0.7593,
"step": 2435
},
{
"epoch": 0.8408698653779773,
"grad_norm": 1.8664613962173462,
"learning_rate": 2e-05,
"loss": 0.6903,
"step": 2436
},
{
"epoch": 0.8412150500517777,
"grad_norm": 1.8746742010116577,
"learning_rate": 2e-05,
"loss": 0.7576,
"step": 2437
},
{
"epoch": 0.8415602347255782,
"grad_norm": 1.7649412155151367,
"learning_rate": 2e-05,
"loss": 0.6862,
"step": 2438
},
{
"epoch": 0.8419054193993787,
"grad_norm": 1.9906177520751953,
"learning_rate": 2e-05,
"loss": 0.8314,
"step": 2439
},
{
"epoch": 0.8422506040731792,
"grad_norm": 1.7777230739593506,
"learning_rate": 2e-05,
"loss": 0.7238,
"step": 2440
},
{
"epoch": 0.8425957887469796,
"grad_norm": 1.659514307975769,
"learning_rate": 2e-05,
"loss": 0.7618,
"step": 2441
},
{
"epoch": 0.8429409734207801,
"grad_norm": 1.9551982879638672,
"learning_rate": 2e-05,
"loss": 0.6587,
"step": 2442
},
{
"epoch": 0.8432861580945806,
"grad_norm": 1.7735124826431274,
"learning_rate": 2e-05,
"loss": 0.6999,
"step": 2443
},
{
"epoch": 0.843631342768381,
"grad_norm": 1.8620107173919678,
"learning_rate": 2e-05,
"loss": 0.7142,
"step": 2444
},
{
"epoch": 0.8439765274421815,
"grad_norm": 1.7266052961349487,
"learning_rate": 2e-05,
"loss": 0.7,
"step": 2445
},
{
"epoch": 0.8443217121159821,
"grad_norm": 1.9168193340301514,
"learning_rate": 2e-05,
"loss": 0.6519,
"step": 2446
},
{
"epoch": 0.8446668967897826,
"grad_norm": 1.8489551544189453,
"learning_rate": 2e-05,
"loss": 0.7134,
"step": 2447
},
{
"epoch": 0.845012081463583,
"grad_norm": 2.2672884464263916,
"learning_rate": 2e-05,
"loss": 0.6569,
"step": 2448
},
{
"epoch": 0.8453572661373835,
"grad_norm": 1.7721095085144043,
"learning_rate": 2e-05,
"loss": 0.7432,
"step": 2449
},
{
"epoch": 0.845702450811184,
"grad_norm": 1.9601596593856812,
"learning_rate": 2e-05,
"loss": 0.7194,
"step": 2450
},
{
"epoch": 0.8460476354849845,
"grad_norm": 1.8006802797317505,
"learning_rate": 2e-05,
"loss": 0.7103,
"step": 2451
},
{
"epoch": 0.8463928201587849,
"grad_norm": 1.9409044981002808,
"learning_rate": 2e-05,
"loss": 0.7096,
"step": 2452
},
{
"epoch": 0.8467380048325854,
"grad_norm": 1.9961984157562256,
"learning_rate": 2e-05,
"loss": 0.7834,
"step": 2453
},
{
"epoch": 0.847083189506386,
"grad_norm": 1.7055926322937012,
"learning_rate": 2e-05,
"loss": 0.7419,
"step": 2454
},
{
"epoch": 0.8474283741801864,
"grad_norm": 1.880017876625061,
"learning_rate": 2e-05,
"loss": 0.6749,
"step": 2455
},
{
"epoch": 0.8477735588539869,
"grad_norm": 1.8338311910629272,
"learning_rate": 2e-05,
"loss": 0.7131,
"step": 2456
},
{
"epoch": 0.8481187435277874,
"grad_norm": 1.6697379350662231,
"learning_rate": 2e-05,
"loss": 0.6483,
"step": 2457
},
{
"epoch": 0.8484639282015879,
"grad_norm": 1.7563802003860474,
"learning_rate": 2e-05,
"loss": 0.6848,
"step": 2458
},
{
"epoch": 0.8488091128753883,
"grad_norm": 2.157135248184204,
"learning_rate": 2e-05,
"loss": 0.7276,
"step": 2459
},
{
"epoch": 0.8491542975491888,
"grad_norm": 1.7930192947387695,
"learning_rate": 2e-05,
"loss": 0.6646,
"step": 2460
},
{
"epoch": 0.8494994822229893,
"grad_norm": 1.8408598899841309,
"learning_rate": 2e-05,
"loss": 0.7083,
"step": 2461
},
{
"epoch": 0.8498446668967898,
"grad_norm": 1.8233611583709717,
"learning_rate": 2e-05,
"loss": 0.7192,
"step": 2462
},
{
"epoch": 0.8501898515705902,
"grad_norm": 1.7961004972457886,
"learning_rate": 2e-05,
"loss": 0.7171,
"step": 2463
},
{
"epoch": 0.8505350362443908,
"grad_norm": 1.7718796730041504,
"learning_rate": 2e-05,
"loss": 0.7368,
"step": 2464
},
{
"epoch": 0.8508802209181913,
"grad_norm": 1.7284215688705444,
"learning_rate": 2e-05,
"loss": 0.7198,
"step": 2465
},
{
"epoch": 0.8512254055919917,
"grad_norm": 1.8063946962356567,
"learning_rate": 2e-05,
"loss": 0.7494,
"step": 2466
},
{
"epoch": 0.8515705902657922,
"grad_norm": 1.7492866516113281,
"learning_rate": 2e-05,
"loss": 0.6875,
"step": 2467
},
{
"epoch": 0.8519157749395927,
"grad_norm": 1.9428259134292603,
"learning_rate": 2e-05,
"loss": 0.668,
"step": 2468
},
{
"epoch": 0.8522609596133932,
"grad_norm": 1.833752989768982,
"learning_rate": 2e-05,
"loss": 0.7321,
"step": 2469
},
{
"epoch": 0.8526061442871936,
"grad_norm": 1.800809621810913,
"learning_rate": 2e-05,
"loss": 0.6741,
"step": 2470
},
{
"epoch": 0.8529513289609941,
"grad_norm": 1.7168399095535278,
"learning_rate": 2e-05,
"loss": 0.7063,
"step": 2471
},
{
"epoch": 0.8532965136347946,
"grad_norm": 2.2443535327911377,
"learning_rate": 2e-05,
"loss": 0.672,
"step": 2472
},
{
"epoch": 0.8536416983085952,
"grad_norm": 1.9579691886901855,
"learning_rate": 2e-05,
"loss": 0.6751,
"step": 2473
},
{
"epoch": 0.8539868829823956,
"grad_norm": 1.700211524963379,
"learning_rate": 2e-05,
"loss": 0.633,
"step": 2474
},
{
"epoch": 0.8543320676561961,
"grad_norm": 1.7471891641616821,
"learning_rate": 2e-05,
"loss": 0.7341,
"step": 2475
},
{
"epoch": 0.8546772523299966,
"grad_norm": 1.6846410036087036,
"learning_rate": 2e-05,
"loss": 0.7224,
"step": 2476
},
{
"epoch": 0.855022437003797,
"grad_norm": 1.8387013673782349,
"learning_rate": 2e-05,
"loss": 0.7534,
"step": 2477
},
{
"epoch": 0.8553676216775975,
"grad_norm": 1.7786623239517212,
"learning_rate": 2e-05,
"loss": 0.6931,
"step": 2478
},
{
"epoch": 0.855712806351398,
"grad_norm": 1.8883618116378784,
"learning_rate": 2e-05,
"loss": 0.7114,
"step": 2479
},
{
"epoch": 0.8560579910251985,
"grad_norm": 1.9042893648147583,
"learning_rate": 2e-05,
"loss": 0.7052,
"step": 2480
},
{
"epoch": 0.8564031756989989,
"grad_norm": 1.7905224561691284,
"learning_rate": 2e-05,
"loss": 0.7502,
"step": 2481
},
{
"epoch": 0.8567483603727994,
"grad_norm": 1.7157790660858154,
"learning_rate": 2e-05,
"loss": 0.6529,
"step": 2482
},
{
"epoch": 0.8570935450466,
"grad_norm": 1.7419195175170898,
"learning_rate": 2e-05,
"loss": 0.6987,
"step": 2483
},
{
"epoch": 0.8574387297204004,
"grad_norm": 1.7620506286621094,
"learning_rate": 2e-05,
"loss": 0.7196,
"step": 2484
},
{
"epoch": 0.8577839143942009,
"grad_norm": 1.7928342819213867,
"learning_rate": 2e-05,
"loss": 0.7197,
"step": 2485
},
{
"epoch": 0.8581290990680014,
"grad_norm": 1.8531665802001953,
"learning_rate": 2e-05,
"loss": 0.674,
"step": 2486
},
{
"epoch": 0.8584742837418019,
"grad_norm": 1.7814826965332031,
"learning_rate": 2e-05,
"loss": 0.7185,
"step": 2487
},
{
"epoch": 0.8588194684156023,
"grad_norm": 1.9274402856826782,
"learning_rate": 2e-05,
"loss": 0.7293,
"step": 2488
},
{
"epoch": 0.8591646530894028,
"grad_norm": 1.6030443906784058,
"learning_rate": 2e-05,
"loss": 0.687,
"step": 2489
},
{
"epoch": 0.8595098377632033,
"grad_norm": 1.6420092582702637,
"learning_rate": 2e-05,
"loss": 0.6942,
"step": 2490
},
{
"epoch": 0.8598550224370038,
"grad_norm": 1.8998804092407227,
"learning_rate": 2e-05,
"loss": 0.7217,
"step": 2491
},
{
"epoch": 0.8602002071108042,
"grad_norm": 1.916654109954834,
"learning_rate": 2e-05,
"loss": 0.7342,
"step": 2492
},
{
"epoch": 0.8605453917846048,
"grad_norm": 1.9728790521621704,
"learning_rate": 2e-05,
"loss": 0.7587,
"step": 2493
},
{
"epoch": 0.8608905764584053,
"grad_norm": 1.7145063877105713,
"learning_rate": 2e-05,
"loss": 0.6779,
"step": 2494
},
{
"epoch": 0.8612357611322057,
"grad_norm": 1.805377721786499,
"learning_rate": 2e-05,
"loss": 0.7483,
"step": 2495
},
{
"epoch": 0.8615809458060062,
"grad_norm": 1.9442853927612305,
"learning_rate": 2e-05,
"loss": 0.7615,
"step": 2496
},
{
"epoch": 0.8619261304798067,
"grad_norm": 1.7394760847091675,
"learning_rate": 2e-05,
"loss": 0.692,
"step": 2497
},
{
"epoch": 0.8622713151536072,
"grad_norm": 1.9626586437225342,
"learning_rate": 2e-05,
"loss": 0.6606,
"step": 2498
},
{
"epoch": 0.8626164998274076,
"grad_norm": 2.07535457611084,
"learning_rate": 2e-05,
"loss": 0.7399,
"step": 2499
},
{
"epoch": 0.8629616845012081,
"grad_norm": 1.990527629852295,
"learning_rate": 2e-05,
"loss": 0.6598,
"step": 2500
},
{
"epoch": 0.8633068691750087,
"grad_norm": 1.9309748411178589,
"learning_rate": 2e-05,
"loss": 0.656,
"step": 2501
},
{
"epoch": 0.8636520538488092,
"grad_norm": 2.0087356567382812,
"learning_rate": 2e-05,
"loss": 0.7028,
"step": 2502
},
{
"epoch": 0.8639972385226096,
"grad_norm": 1.7293936014175415,
"learning_rate": 2e-05,
"loss": 0.7174,
"step": 2503
},
{
"epoch": 0.8643424231964101,
"grad_norm": 1.6928752660751343,
"learning_rate": 2e-05,
"loss": 0.6439,
"step": 2504
},
{
"epoch": 0.8646876078702106,
"grad_norm": 1.9770276546478271,
"learning_rate": 2e-05,
"loss": 0.7006,
"step": 2505
},
{
"epoch": 0.865032792544011,
"grad_norm": 1.6598477363586426,
"learning_rate": 2e-05,
"loss": 0.6195,
"step": 2506
},
{
"epoch": 0.8653779772178115,
"grad_norm": 1.7316237688064575,
"learning_rate": 2e-05,
"loss": 0.6557,
"step": 2507
},
{
"epoch": 0.865723161891612,
"grad_norm": 1.9381532669067383,
"learning_rate": 2e-05,
"loss": 0.7155,
"step": 2508
},
{
"epoch": 0.8660683465654125,
"grad_norm": 1.7715622186660767,
"learning_rate": 2e-05,
"loss": 0.6868,
"step": 2509
},
{
"epoch": 0.8664135312392129,
"grad_norm": 1.9071362018585205,
"learning_rate": 2e-05,
"loss": 0.7077,
"step": 2510
},
{
"epoch": 0.8667587159130135,
"grad_norm": 1.6320717334747314,
"learning_rate": 2e-05,
"loss": 0.7289,
"step": 2511
},
{
"epoch": 0.867103900586814,
"grad_norm": 1.8416504859924316,
"learning_rate": 2e-05,
"loss": 0.7297,
"step": 2512
},
{
"epoch": 0.8674490852606145,
"grad_norm": 1.995140552520752,
"learning_rate": 2e-05,
"loss": 0.669,
"step": 2513
},
{
"epoch": 0.8677942699344149,
"grad_norm": 2.169330596923828,
"learning_rate": 2e-05,
"loss": 0.7651,
"step": 2514
},
{
"epoch": 0.8681394546082154,
"grad_norm": 2.0589828491210938,
"learning_rate": 2e-05,
"loss": 0.7548,
"step": 2515
},
{
"epoch": 0.8684846392820159,
"grad_norm": 1.9467494487762451,
"learning_rate": 2e-05,
"loss": 0.7386,
"step": 2516
},
{
"epoch": 0.8688298239558163,
"grad_norm": 1.85875403881073,
"learning_rate": 2e-05,
"loss": 0.7341,
"step": 2517
},
{
"epoch": 0.8691750086296168,
"grad_norm": 1.753756046295166,
"learning_rate": 2e-05,
"loss": 0.7565,
"step": 2518
},
{
"epoch": 0.8695201933034173,
"grad_norm": 1.8102861642837524,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 2519
},
{
"epoch": 0.8698653779772179,
"grad_norm": 2.0030734539031982,
"learning_rate": 2e-05,
"loss": 0.7546,
"step": 2520
},
{
"epoch": 0.8702105626510183,
"grad_norm": 1.7215023040771484,
"learning_rate": 2e-05,
"loss": 0.7259,
"step": 2521
},
{
"epoch": 0.8705557473248188,
"grad_norm": 1.8061789274215698,
"learning_rate": 2e-05,
"loss": 0.7069,
"step": 2522
},
{
"epoch": 0.8709009319986193,
"grad_norm": 1.9568002223968506,
"learning_rate": 2e-05,
"loss": 0.7439,
"step": 2523
},
{
"epoch": 0.8712461166724197,
"grad_norm": 1.832311987876892,
"learning_rate": 2e-05,
"loss": 0.7259,
"step": 2524
},
{
"epoch": 0.8715913013462202,
"grad_norm": 1.731987714767456,
"learning_rate": 2e-05,
"loss": 0.7394,
"step": 2525
},
{
"epoch": 0.8719364860200207,
"grad_norm": 1.7261989116668701,
"learning_rate": 2e-05,
"loss": 0.7092,
"step": 2526
},
{
"epoch": 0.8722816706938212,
"grad_norm": 1.8286298513412476,
"learning_rate": 2e-05,
"loss": 0.7318,
"step": 2527
},
{
"epoch": 0.8726268553676216,
"grad_norm": 1.9575196504592896,
"learning_rate": 2e-05,
"loss": 0.7185,
"step": 2528
},
{
"epoch": 0.8729720400414221,
"grad_norm": 1.8130125999450684,
"learning_rate": 2e-05,
"loss": 0.7449,
"step": 2529
},
{
"epoch": 0.8733172247152227,
"grad_norm": 1.8687063455581665,
"learning_rate": 2e-05,
"loss": 0.726,
"step": 2530
},
{
"epoch": 0.8736624093890232,
"grad_norm": 1.8880723714828491,
"learning_rate": 2e-05,
"loss": 0.6982,
"step": 2531
},
{
"epoch": 0.8740075940628236,
"grad_norm": 1.729681134223938,
"learning_rate": 2e-05,
"loss": 0.6618,
"step": 2532
},
{
"epoch": 0.8743527787366241,
"grad_norm": 1.5875263214111328,
"learning_rate": 2e-05,
"loss": 0.7065,
"step": 2533
},
{
"epoch": 0.8746979634104246,
"grad_norm": 1.9049720764160156,
"learning_rate": 2e-05,
"loss": 0.7256,
"step": 2534
},
{
"epoch": 0.875043148084225,
"grad_norm": 1.7123501300811768,
"learning_rate": 2e-05,
"loss": 0.706,
"step": 2535
},
{
"epoch": 0.8753883327580255,
"grad_norm": 1.7586158514022827,
"learning_rate": 2e-05,
"loss": 0.7489,
"step": 2536
},
{
"epoch": 0.875733517431826,
"grad_norm": 2.1135923862457275,
"learning_rate": 2e-05,
"loss": 0.7643,
"step": 2537
},
{
"epoch": 0.8760787021056266,
"grad_norm": 2.838963508605957,
"learning_rate": 2e-05,
"loss": 0.7397,
"step": 2538
},
{
"epoch": 0.876423886779427,
"grad_norm": 1.8444801568984985,
"learning_rate": 2e-05,
"loss": 0.758,
"step": 2539
},
{
"epoch": 0.8767690714532275,
"grad_norm": 1.7966136932373047,
"learning_rate": 2e-05,
"loss": 0.7499,
"step": 2540
},
{
"epoch": 0.877114256127028,
"grad_norm": 1.6917723417282104,
"learning_rate": 2e-05,
"loss": 0.7448,
"step": 2541
},
{
"epoch": 0.8774594408008285,
"grad_norm": 1.8464945554733276,
"learning_rate": 2e-05,
"loss": 0.7082,
"step": 2542
},
{
"epoch": 0.8778046254746289,
"grad_norm": 1.8043209314346313,
"learning_rate": 2e-05,
"loss": 0.7063,
"step": 2543
},
{
"epoch": 0.8781498101484294,
"grad_norm": 2.1217026710510254,
"learning_rate": 2e-05,
"loss": 0.728,
"step": 2544
},
{
"epoch": 0.8784949948222299,
"grad_norm": 1.7280898094177246,
"learning_rate": 2e-05,
"loss": 0.7631,
"step": 2545
},
{
"epoch": 0.8788401794960303,
"grad_norm": 1.723939061164856,
"learning_rate": 2e-05,
"loss": 0.6866,
"step": 2546
},
{
"epoch": 0.8791853641698308,
"grad_norm": 1.756032109260559,
"learning_rate": 2e-05,
"loss": 0.7238,
"step": 2547
},
{
"epoch": 0.8795305488436314,
"grad_norm": 1.917583703994751,
"learning_rate": 2e-05,
"loss": 0.7054,
"step": 2548
},
{
"epoch": 0.8798757335174319,
"grad_norm": 1.8419638872146606,
"learning_rate": 2e-05,
"loss": 0.7167,
"step": 2549
},
{
"epoch": 0.8802209181912323,
"grad_norm": 1.6455121040344238,
"learning_rate": 2e-05,
"loss": 0.7507,
"step": 2550
},
{
"epoch": 0.8805661028650328,
"grad_norm": 1.8877838850021362,
"learning_rate": 2e-05,
"loss": 0.7409,
"step": 2551
},
{
"epoch": 0.8809112875388333,
"grad_norm": 1.8152822256088257,
"learning_rate": 2e-05,
"loss": 0.7266,
"step": 2552
},
{
"epoch": 0.8812564722126338,
"grad_norm": 1.726479172706604,
"learning_rate": 2e-05,
"loss": 0.6547,
"step": 2553
},
{
"epoch": 0.8816016568864342,
"grad_norm": 1.8056279420852661,
"learning_rate": 2e-05,
"loss": 0.6391,
"step": 2554
},
{
"epoch": 0.8819468415602347,
"grad_norm": 1.845010757446289,
"learning_rate": 2e-05,
"loss": 0.709,
"step": 2555
},
{
"epoch": 0.8822920262340352,
"grad_norm": 1.721825122833252,
"learning_rate": 2e-05,
"loss": 0.6798,
"step": 2556
},
{
"epoch": 0.8826372109078356,
"grad_norm": 1.5322935581207275,
"learning_rate": 2e-05,
"loss": 0.6252,
"step": 2557
},
{
"epoch": 0.8829823955816362,
"grad_norm": 1.743082880973816,
"learning_rate": 2e-05,
"loss": 0.7257,
"step": 2558
},
{
"epoch": 0.8833275802554367,
"grad_norm": 1.8050196170806885,
"learning_rate": 2e-05,
"loss": 0.7445,
"step": 2559
},
{
"epoch": 0.8836727649292372,
"grad_norm": 1.8423820734024048,
"learning_rate": 2e-05,
"loss": 0.7122,
"step": 2560
},
{
"epoch": 0.8840179496030376,
"grad_norm": 1.5825133323669434,
"learning_rate": 2e-05,
"loss": 0.6901,
"step": 2561
},
{
"epoch": 0.8843631342768381,
"grad_norm": 1.7943646907806396,
"learning_rate": 2e-05,
"loss": 0.7185,
"step": 2562
},
{
"epoch": 0.8847083189506386,
"grad_norm": 1.935242772102356,
"learning_rate": 2e-05,
"loss": 0.6398,
"step": 2563
},
{
"epoch": 0.885053503624439,
"grad_norm": 1.676263451576233,
"learning_rate": 2e-05,
"loss": 0.694,
"step": 2564
},
{
"epoch": 0.8853986882982395,
"grad_norm": 1.8152910470962524,
"learning_rate": 2e-05,
"loss": 0.6573,
"step": 2565
},
{
"epoch": 0.88574387297204,
"grad_norm": 1.8568143844604492,
"learning_rate": 2e-05,
"loss": 0.6762,
"step": 2566
},
{
"epoch": 0.8860890576458406,
"grad_norm": 1.8765757083892822,
"learning_rate": 2e-05,
"loss": 0.8084,
"step": 2567
},
{
"epoch": 0.886434242319641,
"grad_norm": 1.985735535621643,
"learning_rate": 2e-05,
"loss": 0.6847,
"step": 2568
},
{
"epoch": 0.8867794269934415,
"grad_norm": 1.7089177370071411,
"learning_rate": 2e-05,
"loss": 0.7449,
"step": 2569
},
{
"epoch": 0.887124611667242,
"grad_norm": 2.0181849002838135,
"learning_rate": 2e-05,
"loss": 0.7529,
"step": 2570
},
{
"epoch": 0.8874697963410425,
"grad_norm": 1.876921534538269,
"learning_rate": 2e-05,
"loss": 0.6578,
"step": 2571
},
{
"epoch": 0.8878149810148429,
"grad_norm": 1.816359043121338,
"learning_rate": 2e-05,
"loss": 0.7214,
"step": 2572
},
{
"epoch": 0.8881601656886434,
"grad_norm": 1.7160218954086304,
"learning_rate": 2e-05,
"loss": 0.7195,
"step": 2573
},
{
"epoch": 0.8885053503624439,
"grad_norm": 1.7351136207580566,
"learning_rate": 2e-05,
"loss": 0.7547,
"step": 2574
},
{
"epoch": 0.8888505350362443,
"grad_norm": 1.859092354774475,
"learning_rate": 2e-05,
"loss": 0.7903,
"step": 2575
},
{
"epoch": 0.8891957197100449,
"grad_norm": 1.6283543109893799,
"learning_rate": 2e-05,
"loss": 0.6574,
"step": 2576
},
{
"epoch": 0.8895409043838454,
"grad_norm": 1.7172080278396606,
"learning_rate": 2e-05,
"loss": 0.6916,
"step": 2577
},
{
"epoch": 0.8898860890576459,
"grad_norm": 1.8297404050827026,
"learning_rate": 2e-05,
"loss": 0.7703,
"step": 2578
},
{
"epoch": 0.8902312737314463,
"grad_norm": 1.9139461517333984,
"learning_rate": 2e-05,
"loss": 0.7025,
"step": 2579
},
{
"epoch": 0.8905764584052468,
"grad_norm": 1.8364346027374268,
"learning_rate": 2e-05,
"loss": 0.6869,
"step": 2580
},
{
"epoch": 0.8909216430790473,
"grad_norm": 1.7251884937286377,
"learning_rate": 2e-05,
"loss": 0.759,
"step": 2581
},
{
"epoch": 0.8912668277528478,
"grad_norm": 1.8190256357192993,
"learning_rate": 2e-05,
"loss": 0.7213,
"step": 2582
},
{
"epoch": 0.8916120124266482,
"grad_norm": 2.070838451385498,
"learning_rate": 2e-05,
"loss": 0.7079,
"step": 2583
},
{
"epoch": 0.8919571971004487,
"grad_norm": 3.710937261581421,
"learning_rate": 2e-05,
"loss": 0.7025,
"step": 2584
},
{
"epoch": 0.8923023817742493,
"grad_norm": 1.7923078536987305,
"learning_rate": 2e-05,
"loss": 0.6798,
"step": 2585
},
{
"epoch": 0.8926475664480497,
"grad_norm": 2.0066332817077637,
"learning_rate": 2e-05,
"loss": 0.7289,
"step": 2586
},
{
"epoch": 0.8929927511218502,
"grad_norm": 1.8006290197372437,
"learning_rate": 2e-05,
"loss": 0.7187,
"step": 2587
},
{
"epoch": 0.8933379357956507,
"grad_norm": 1.6795895099639893,
"learning_rate": 2e-05,
"loss": 0.6614,
"step": 2588
},
{
"epoch": 0.8936831204694512,
"grad_norm": 1.816856026649475,
"learning_rate": 2e-05,
"loss": 0.7146,
"step": 2589
},
{
"epoch": 0.8940283051432516,
"grad_norm": 1.7363369464874268,
"learning_rate": 2e-05,
"loss": 0.6716,
"step": 2590
},
{
"epoch": 0.8943734898170521,
"grad_norm": 2.385967254638672,
"learning_rate": 2e-05,
"loss": 0.7122,
"step": 2591
},
{
"epoch": 0.8947186744908526,
"grad_norm": 1.9235140085220337,
"learning_rate": 2e-05,
"loss": 0.743,
"step": 2592
},
{
"epoch": 0.8950638591646531,
"grad_norm": 1.8357746601104736,
"learning_rate": 2e-05,
"loss": 0.6824,
"step": 2593
},
{
"epoch": 0.8954090438384535,
"grad_norm": 1.7374387979507446,
"learning_rate": 2e-05,
"loss": 0.7361,
"step": 2594
},
{
"epoch": 0.8957542285122541,
"grad_norm": 1.7777645587921143,
"learning_rate": 2e-05,
"loss": 0.6165,
"step": 2595
},
{
"epoch": 0.8960994131860546,
"grad_norm": 1.6975550651550293,
"learning_rate": 2e-05,
"loss": 0.6937,
"step": 2596
},
{
"epoch": 0.896444597859855,
"grad_norm": 1.835229516029358,
"learning_rate": 2e-05,
"loss": 0.7348,
"step": 2597
},
{
"epoch": 0.8967897825336555,
"grad_norm": 1.721623420715332,
"learning_rate": 2e-05,
"loss": 0.7285,
"step": 2598
},
{
"epoch": 0.897134967207456,
"grad_norm": 1.7775485515594482,
"learning_rate": 2e-05,
"loss": 0.7425,
"step": 2599
},
{
"epoch": 0.8974801518812565,
"grad_norm": 1.8077950477600098,
"learning_rate": 2e-05,
"loss": 0.7176,
"step": 2600
}
],
"logging_steps": 1.0,
"max_steps": 14485,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 200,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 1633161314304000.0,
"train_batch_size": 12,
"trial_name": null,
"trial_params": null
}
Reference in New Issue View Git Blame Copy Permalink