Files
ScholarCopilot-v1/trainer_state.json
ModelHub XC 1b351fcfec 初始化项目,由ModelHub XC社区提供模型
Model: TIGER-Lab/ScholarCopilot-v1
Source: Original Platform
2026-09-07 17:12:15 +08:00

14034 lines
338 KiB
JSON

{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 2.0116929653611617,
"eval_steps": 500,
"global_step": 2000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.001005846482680581,
"grad_norm": 12582.778327747239,
"learning_rate": 0.0,
"loss": 18.2821,
"step": 1
},
{
"epoch": 0.002011692965361162,
"grad_norm": 8175.9099112773565,
"learning_rate": 1.1164325195357284e-06,
"loss": 9.792,
"step": 2
},
{
"epoch": 0.0030175394480417427,
"grad_norm": 1649.5623048840205,
"learning_rate": 1.7695036780497693e-06,
"loss": 10.0444,
"step": 3
},
{
"epoch": 0.004023385930722324,
"grad_norm": 1686.6179804241558,
"learning_rate": 2.2328650390714567e-06,
"loss": 9.5109,
"step": 4
},
{
"epoch": 0.005029232413402905,
"grad_norm": 5483.012027070882,
"learning_rate": 2.5922760331558917e-06,
"loss": 8.2383,
"step": 5
},
{
"epoch": 0.006035078896083485,
"grad_norm": 1172.816476755199,
"learning_rate": 2.8859361975854976e-06,
"loss": 7.1834,
"step": 6
},
{
"epoch": 0.007040925378764066,
"grad_norm": 657.0193977991333,
"learning_rate": 3.1342223288637992e-06,
"loss": 5.4344,
"step": 7
},
{
"epoch": 0.008046771861444648,
"grad_norm": 372.67224237745364,
"learning_rate": 3.349297558607185e-06,
"loss": 4.6949,
"step": 8
},
{
"epoch": 0.009052618344125228,
"grad_norm": 380.45025528740604,
"learning_rate": 3.5390073560995386e-06,
"loss": 4.5428,
"step": 9
},
{
"epoch": 0.01005846482680581,
"grad_norm": 773.3075300312582,
"learning_rate": 3.7087085526916205e-06,
"loss": 4.2198,
"step": 10
},
{
"epoch": 0.01106431130948639,
"grad_norm": 208.01920169269192,
"learning_rate": 3.862221958156801e-06,
"loss": 4.3316,
"step": 11
},
{
"epoch": 0.01207015779216697,
"grad_norm": 188.8951453419855,
"learning_rate": 4.0023687171212264e-06,
"loss": 4.0229,
"step": 12
},
{
"epoch": 0.013076004274847551,
"grad_norm": 846.972851780641,
"learning_rate": 4.131291237914341e-06,
"loss": 3.7635,
"step": 13
},
{
"epoch": 0.014081850757528132,
"grad_norm": 252.04772708907407,
"learning_rate": 4.250654848399527e-06,
"loss": 3.6946,
"step": 14
},
{
"epoch": 0.015087697240208713,
"grad_norm": 177.74605915923814,
"learning_rate": 4.3617797112056605e-06,
"loss": 3.539,
"step": 15
},
{
"epoch": 0.016093543722889295,
"grad_norm": 157.7895649733611,
"learning_rate": 4.4657300781429134e-06,
"loss": 3.5069,
"step": 16
},
{
"epoch": 0.017099390205569876,
"grad_norm": 3471.9413534733717,
"learning_rate": 4.563376438365783e-06,
"loss": 3.457,
"step": 17
},
{
"epoch": 0.018105236688250457,
"grad_norm": 9271.52283588444,
"learning_rate": 4.6554398756352665e-06,
"loss": 3.4839,
"step": 18
},
{
"epoch": 0.019111083170931038,
"grad_norm": 1707.1301369245853,
"learning_rate": 4.742524416638964e-06,
"loss": 3.4175,
"step": 19
},
{
"epoch": 0.02011692965361162,
"grad_norm": 380.3932734821994,
"learning_rate": 4.825141072227348e-06,
"loss": 3.413,
"step": 20
},
{
"epoch": 0.0211227761362922,
"grad_norm": 412.42448179438645,
"learning_rate": 4.903726006913569e-06,
"loss": 3.3572,
"step": 21
},
{
"epoch": 0.02212862261897278,
"grad_norm": 155.79462929869592,
"learning_rate": 4.97865447769253e-06,
"loss": 3.3687,
"step": 22
},
{
"epoch": 0.02313446910165336,
"grad_norm": 146.80182332367627,
"learning_rate": 5.050251671876699e-06,
"loss": 3.2905,
"step": 23
},
{
"epoch": 0.02414031558433394,
"grad_norm": 152.70556523236607,
"learning_rate": 5.118801236656955e-06,
"loss": 3.3167,
"step": 24
},
{
"epoch": 0.025146162067014522,
"grad_norm": 145.35719764915638,
"learning_rate": 5.184552066311783e-06,
"loss": 3.2275,
"step": 25
},
{
"epoch": 0.026152008549695103,
"grad_norm": 142.30381750870376,
"learning_rate": 5.247723757450069e-06,
"loss": 3.2108,
"step": 26
},
{
"epoch": 0.027157855032375684,
"grad_norm": 190.7780665632596,
"learning_rate": 5.3085110341493074e-06,
"loss": 3.1127,
"step": 27
},
{
"epoch": 0.028163701515056264,
"grad_norm": 140.71233699510987,
"learning_rate": 5.367087367935257e-06,
"loss": 3.2016,
"step": 28
},
{
"epoch": 0.029169547997736845,
"grad_norm": 144.31455234163855,
"learning_rate": 5.423607962246961e-06,
"loss": 3.1428,
"step": 29
},
{
"epoch": 0.030175394480417426,
"grad_norm": 166.83275617027454,
"learning_rate": 5.47821223074139e-06,
"loss": 3.166,
"step": 30
},
{
"epoch": 0.031181240963098007,
"grad_norm": 150.84029040740148,
"learning_rate": 5.531025869079829e-06,
"loss": 3.0718,
"step": 31
},
{
"epoch": 0.03218708744577859,
"grad_norm": 167.4297289699058,
"learning_rate": 5.582162597678642e-06,
"loss": 3.0989,
"step": 32
},
{
"epoch": 0.03319293392845917,
"grad_norm": 196.37733263631793,
"learning_rate": 5.631725636206569e-06,
"loss": 3.0453,
"step": 33
},
{
"epoch": 0.03419878041113975,
"grad_norm": 167.0654826804254,
"learning_rate": 5.679808957901513e-06,
"loss": 3.0685,
"step": 34
},
{
"epoch": 0.03520462689382033,
"grad_norm": 157.79329205975614,
"learning_rate": 5.726498362019691e-06,
"loss": 3.0637,
"step": 35
},
{
"epoch": 0.036210473376500914,
"grad_norm": 145.63751069723227,
"learning_rate": 5.771872395170995e-06,
"loss": 3.1047,
"step": 36
},
{
"epoch": 0.037216319859181494,
"grad_norm": 138.30989338843392,
"learning_rate": 5.816003146393009e-06,
"loss": 3.0059,
"step": 37
},
{
"epoch": 0.038222166341862075,
"grad_norm": 141.08916496720514,
"learning_rate": 5.858956936174693e-06,
"loss": 3.0426,
"step": 38
},
{
"epoch": 0.039228012824542656,
"grad_norm": 150.01969317858723,
"learning_rate": 5.900794915964109e-06,
"loss": 3.0102,
"step": 39
},
{
"epoch": 0.04023385930722324,
"grad_norm": 145.60366402885353,
"learning_rate": 5.941573591763077e-06,
"loss": 2.9662,
"step": 40
},
{
"epoch": 0.04123970578990382,
"grad_norm": 256.030725034778,
"learning_rate": 5.98134528305946e-06,
"loss": 3.015,
"step": 41
},
{
"epoch": 0.0422455522725844,
"grad_norm": 148.18303193116637,
"learning_rate": 6.020158526449297e-06,
"loss": 3.0394,
"step": 42
},
{
"epoch": 0.04325139875526498,
"grad_norm": 401.9892406324886,
"learning_rate": 6.058058431759984e-06,
"loss": 2.9915,
"step": 43
},
{
"epoch": 0.04425724523794556,
"grad_norm": 138.76017347757087,
"learning_rate": 6.095086997228258e-06,
"loss": 2.9636,
"step": 44
},
{
"epoch": 0.04526309172062614,
"grad_norm": 151.8736160799114,
"learning_rate": 6.13128338925543e-06,
"loss": 3.0144,
"step": 45
},
{
"epoch": 0.04626893820330672,
"grad_norm": 776.4808078486112,
"learning_rate": 6.166684191412428e-06,
"loss": 2.9346,
"step": 46
},
{
"epoch": 0.0472747846859873,
"grad_norm": 148.56518730989325,
"learning_rate": 6.201323626663534e-06,
"loss": 2.9401,
"step": 47
},
{
"epoch": 0.04828063116866788,
"grad_norm": 140.9261131197792,
"learning_rate": 6.235233756192683e-06,
"loss": 2.9184,
"step": 48
},
{
"epoch": 0.04928647765134846,
"grad_norm": 140.86675687418395,
"learning_rate": 6.2684446577275984e-06,
"loss": 2.96,
"step": 49
},
{
"epoch": 0.050292324134029044,
"grad_norm": 133.56755452224644,
"learning_rate": 6.300984585847511e-06,
"loss": 2.9206,
"step": 50
},
{
"epoch": 0.051298170616709625,
"grad_norm": 132.93056724366744,
"learning_rate": 6.332880116415553e-06,
"loss": 2.8758,
"step": 51
},
{
"epoch": 0.052304017099390206,
"grad_norm": 138.26029172258959,
"learning_rate": 6.3641562769857975e-06,
"loss": 2.8796,
"step": 52
},
{
"epoch": 0.053309863582070786,
"grad_norm": 123.99660733499722,
"learning_rate": 6.394836664788228e-06,
"loss": 2.9349,
"step": 53
},
{
"epoch": 0.05431571006475137,
"grad_norm": 113.95184567154406,
"learning_rate": 6.424943553685035e-06,
"loss": 2.8738,
"step": 54
},
{
"epoch": 0.05532155654743195,
"grad_norm": 119.01280095728684,
"learning_rate": 6.454497991312694e-06,
"loss": 2.8592,
"step": 55
},
{
"epoch": 0.05632740303011253,
"grad_norm": 106.28111012983979,
"learning_rate": 6.4835198874709856e-06,
"loss": 2.8986,
"step": 56
},
{
"epoch": 0.05733324951279311,
"grad_norm": 105.28212563495431,
"learning_rate": 6.512028094688734e-06,
"loss": 2.8513,
"step": 57
},
{
"epoch": 0.05833909599547369,
"grad_norm": 102.82251214996867,
"learning_rate": 6.540040481782689e-06,
"loss": 2.8233,
"step": 58
},
{
"epoch": 0.05934494247815427,
"grad_norm": 88.93066733289517,
"learning_rate": 6.567574001128382e-06,
"loss": 2.9266,
"step": 59
},
{
"epoch": 0.06035078896083485,
"grad_norm": 82.42620087801528,
"learning_rate": 6.594644750277117e-06,
"loss": 2.8909,
"step": 60
},
{
"epoch": 0.06135663544351543,
"grad_norm": 74.29787586543553,
"learning_rate": 6.621268028479951e-06,
"loss": 2.8756,
"step": 61
},
{
"epoch": 0.06236248192619601,
"grad_norm": 68.17633312180041,
"learning_rate": 6.647458388615557e-06,
"loss": 2.8298,
"step": 62
},
{
"epoch": 0.0633683284088766,
"grad_norm": 61.94312969580986,
"learning_rate": 6.673229684963337e-06,
"loss": 2.9027,
"step": 63
},
{
"epoch": 0.06437417489155718,
"grad_norm": 54.35466540831732,
"learning_rate": 6.69859511721437e-06,
"loss": 2.8293,
"step": 64
},
{
"epoch": 0.06538002137423776,
"grad_norm": 44.173923656121296,
"learning_rate": 6.723567271070232e-06,
"loss": 2.8335,
"step": 65
},
{
"epoch": 0.06638586785691834,
"grad_norm": 41.577563236699476,
"learning_rate": 6.748158155742298e-06,
"loss": 2.8218,
"step": 66
},
{
"epoch": 0.06739171433959892,
"grad_norm": 37.7162349842147,
"learning_rate": 6.7723792386312174e-06,
"loss": 2.7834,
"step": 67
},
{
"epoch": 0.0683975608222795,
"grad_norm": 30.862561284051413,
"learning_rate": 6.796241477437241e-06,
"loss": 2.8077,
"step": 68
},
{
"epoch": 0.06940340730496009,
"grad_norm": 27.233229248978024,
"learning_rate": 6.8197553499264694e-06,
"loss": 2.7832,
"step": 69
},
{
"epoch": 0.07040925378764067,
"grad_norm": 32.95846432991116,
"learning_rate": 6.8429308815554205e-06,
"loss": 2.8153,
"step": 70
},
{
"epoch": 0.07141510027032125,
"grad_norm": 34.002333788304036,
"learning_rate": 6.865777671136202e-06,
"loss": 2.8491,
"step": 71
},
{
"epoch": 0.07242094675300183,
"grad_norm": 26.769475736430746,
"learning_rate": 6.888304914706724e-06,
"loss": 2.865,
"step": 72
},
{
"epoch": 0.07342679323568241,
"grad_norm": 23.983053033180372,
"learning_rate": 6.910521427754546e-06,
"loss": 2.8012,
"step": 73
},
{
"epoch": 0.07443263971836299,
"grad_norm": 24.466547342637735,
"learning_rate": 6.932435665928737e-06,
"loss": 2.7997,
"step": 74
},
{
"epoch": 0.07543848620104357,
"grad_norm": 25.705189713492953,
"learning_rate": 6.954055744361552e-06,
"loss": 2.7941,
"step": 75
},
{
"epoch": 0.07644433268372415,
"grad_norm": 21.916294810974563,
"learning_rate": 6.975389455710421e-06,
"loss": 2.8076,
"step": 76
},
{
"epoch": 0.07745017916640473,
"grad_norm": 24.763966899310763,
"learning_rate": 6.9964442870206006e-06,
"loss": 2.7578,
"step": 77
},
{
"epoch": 0.07845602564908531,
"grad_norm": 22.805095513442417,
"learning_rate": 7.017227435499838e-06,
"loss": 2.72,
"step": 78
},
{
"epoch": 0.07946187213176589,
"grad_norm": 23.044520510111933,
"learning_rate": 7.037745823288182e-06,
"loss": 2.7863,
"step": 79
},
{
"epoch": 0.08046771861444647,
"grad_norm": 26.56601715759224,
"learning_rate": 7.058006111298805e-06,
"loss": 2.7645,
"step": 80
},
{
"epoch": 0.08147356509712705,
"grad_norm": 23.353217623203157,
"learning_rate": 7.078014712199077e-06,
"loss": 2.8215,
"step": 81
},
{
"epoch": 0.08247941157980763,
"grad_norm": 26.391603597567094,
"learning_rate": 7.097777802595188e-06,
"loss": 2.7263,
"step": 82
},
{
"epoch": 0.08348525806248822,
"grad_norm": 23.916408410799217,
"learning_rate": 7.117301334478265e-06,
"loss": 2.7403,
"step": 83
},
{
"epoch": 0.0844911045451688,
"grad_norm": 20.688015864866205,
"learning_rate": 7.136591045985025e-06,
"loss": 2.7666,
"step": 84
},
{
"epoch": 0.08549695102784938,
"grad_norm": 24.511263727618857,
"learning_rate": 7.155652471521676e-06,
"loss": 2.6841,
"step": 85
},
{
"epoch": 0.08650279751052996,
"grad_norm": 17.920694446601882,
"learning_rate": 7.174490951295713e-06,
"loss": 2.7248,
"step": 86
},
{
"epoch": 0.08750864399321054,
"grad_norm": 18.754958778905763,
"learning_rate": 7.19311164029673e-06,
"loss": 2.7673,
"step": 87
},
{
"epoch": 0.08851449047589112,
"grad_norm": 22.81015763762571,
"learning_rate": 7.211519516763988e-06,
"loss": 2.7238,
"step": 88
},
{
"epoch": 0.0895203369585717,
"grad_norm": 21.31760582506125,
"learning_rate": 7.229719390175563e-06,
"loss": 2.7277,
"step": 89
},
{
"epoch": 0.09052618344125228,
"grad_norm": 19.51350169549616,
"learning_rate": 7.247715908791158e-06,
"loss": 2.7123,
"step": 90
},
{
"epoch": 0.09153202992393286,
"grad_norm": 21.820529035616033,
"learning_rate": 7.26551356677814e-06,
"loss": 2.7569,
"step": 91
},
{
"epoch": 0.09253787640661344,
"grad_norm": 20.728510611672373,
"learning_rate": 7.283116710948156e-06,
"loss": 2.7776,
"step": 92
},
{
"epoch": 0.09354372288929402,
"grad_norm": 21.275335070097242,
"learning_rate": 7.300529547129599e-06,
"loss": 2.6706,
"step": 93
},
{
"epoch": 0.0945495693719746,
"grad_norm": 21.20201924741282,
"learning_rate": 7.3177561461992615e-06,
"loss": 2.8248,
"step": 94
},
{
"epoch": 0.09555541585465518,
"grad_norm": 19.354596760258314,
"learning_rate": 7.334800449794856e-06,
"loss": 2.657,
"step": 95
},
{
"epoch": 0.09656126233733577,
"grad_norm": 26.504627668771203,
"learning_rate": 7.351666275728411e-06,
"loss": 2.7424,
"step": 96
},
{
"epoch": 0.09756710882001635,
"grad_norm": 24.920434718375468,
"learning_rate": 7.368357323119185e-06,
"loss": 2.7804,
"step": 97
},
{
"epoch": 0.09857295530269693,
"grad_norm": 24.734865665675084,
"learning_rate": 7.384877177263328e-06,
"loss": 2.6857,
"step": 98
},
{
"epoch": 0.09957880178537751,
"grad_norm": 20.65028932666736,
"learning_rate": 7.40122931425634e-06,
"loss": 2.6713,
"step": 99
},
{
"epoch": 0.10058464826805809,
"grad_norm": 25.244491519043113,
"learning_rate": 7.417417105383241e-06,
"loss": 2.7438,
"step": 100
},
{
"epoch": 0.10159049475073867,
"grad_norm": 20.297890638350957,
"learning_rate": 7.433443821290305e-06,
"loss": 2.6464,
"step": 101
},
{
"epoch": 0.10259634123341925,
"grad_norm": 25.51714016339381,
"learning_rate": 7.44931263595128e-06,
"loss": 2.7024,
"step": 102
},
{
"epoch": 0.10360218771609983,
"grad_norm": 24.087279291502345,
"learning_rate": 7.465026630440138e-06,
"loss": 2.6897,
"step": 103
},
{
"epoch": 0.10460803419878041,
"grad_norm": 31.639294543052827,
"learning_rate": 7.480588796521525e-06,
"loss": 2.6928,
"step": 104
},
{
"epoch": 0.10561388068146099,
"grad_norm": 21.50022167620154,
"learning_rate": 7.49600204006946e-06,
"loss": 2.6821,
"step": 105
},
{
"epoch": 0.10661972716414157,
"grad_norm": 27.095230272732117,
"learning_rate": 7.511269184323955e-06,
"loss": 2.6596,
"step": 106
},
{
"epoch": 0.10762557364682215,
"grad_norm": 24.733709994175797,
"learning_rate": 7.526392972994766e-06,
"loss": 2.7506,
"step": 107
},
{
"epoch": 0.10863142012950273,
"grad_norm": 25.901406399951554,
"learning_rate": 7.541376073220765e-06,
"loss": 2.7288,
"step": 108
},
{
"epoch": 0.10963726661218332,
"grad_norm": 22.812181794557098,
"learning_rate": 7.556221078392927e-06,
"loss": 2.7115,
"step": 109
},
{
"epoch": 0.1106431130948639,
"grad_norm": 22.08834703071957,
"learning_rate": 7.570930510848422e-06,
"loss": 2.7194,
"step": 110
},
{
"epoch": 0.11164895957754448,
"grad_norm": 19.83989319695105,
"learning_rate": 7.585506824442778e-06,
"loss": 2.7086,
"step": 111
},
{
"epoch": 0.11265480606022506,
"grad_norm": 21.193284488575998,
"learning_rate": 7.599952407006712e-06,
"loss": 2.7459,
"step": 112
},
{
"epoch": 0.11366065254290564,
"grad_norm": 20.482415173444885,
"learning_rate": 7.614269582693758e-06,
"loss": 2.7904,
"step": 113
},
{
"epoch": 0.11466649902558622,
"grad_norm": 22.947534201659618,
"learning_rate": 7.628460614224462e-06,
"loss": 2.7392,
"step": 114
},
{
"epoch": 0.1156723455082668,
"grad_norm": 23.010915124299302,
"learning_rate": 7.64252770503259e-06,
"loss": 2.6972,
"step": 115
},
{
"epoch": 0.11667819199094738,
"grad_norm": 19.538353567555912,
"learning_rate": 7.656473001318417e-06,
"loss": 2.7021,
"step": 116
},
{
"epoch": 0.11768403847362796,
"grad_norm": 24.36272861265073,
"learning_rate": 7.670298594013878e-06,
"loss": 2.6831,
"step": 117
},
{
"epoch": 0.11868988495630854,
"grad_norm": 21.199891028702496,
"learning_rate": 7.68400652066411e-06,
"loss": 2.7558,
"step": 118
},
{
"epoch": 0.11969573143898912,
"grad_norm": 19.670104942454504,
"learning_rate": 7.697598767229584e-06,
"loss": 2.672,
"step": 119
},
{
"epoch": 0.1207015779216697,
"grad_norm": 24.51878373945769,
"learning_rate": 7.711077269812845e-06,
"loss": 2.6764,
"step": 120
},
{
"epoch": 0.12170742440435028,
"grad_norm": 21.921013999351562,
"learning_rate": 7.724443916313603e-06,
"loss": 2.6924,
"step": 121
},
{
"epoch": 0.12271327088703086,
"grad_norm": 19.16037093255755,
"learning_rate": 7.737700548015679e-06,
"loss": 2.7193,
"step": 122
},
{
"epoch": 0.12371911736971145,
"grad_norm": 18.980265808721434,
"learning_rate": 7.750848961109229e-06,
"loss": 2.7391,
"step": 123
},
{
"epoch": 0.12472496385239203,
"grad_norm": 18.756909530996253,
"learning_rate": 7.763890908151285e-06,
"loss": 2.7491,
"step": 124
},
{
"epoch": 0.1257308103350726,
"grad_norm": 18.846050730646436,
"learning_rate": 7.776828099467677e-06,
"loss": 2.5827,
"step": 125
},
{
"epoch": 0.1267366568177532,
"grad_norm": 19.381867511680255,
"learning_rate": 7.789662204499067e-06,
"loss": 2.6621,
"step": 126
},
{
"epoch": 0.12774250330043377,
"grad_norm": 18.915258610203843,
"learning_rate": 7.802394853093812e-06,
"loss": 2.6954,
"step": 127
},
{
"epoch": 0.12874834978311436,
"grad_norm": 17.902322301001714,
"learning_rate": 7.815027636750099e-06,
"loss": 2.7114,
"step": 128
},
{
"epoch": 0.12975419626579493,
"grad_norm": 19.86048307331956,
"learning_rate": 7.827562109809753e-06,
"loss": 2.7168,
"step": 129
},
{
"epoch": 0.13076004274847552,
"grad_norm": 16.13146552524454,
"learning_rate": 7.83999979060596e-06,
"loss": 2.673,
"step": 130
},
{
"epoch": 0.1317658892311561,
"grad_norm": 16.823542205733474,
"learning_rate": 7.852342162567001e-06,
"loss": 2.6572,
"step": 131
},
{
"epoch": 0.13277173571383669,
"grad_norm": 17.565391271847194,
"learning_rate": 7.864590675278027e-06,
"loss": 2.603,
"step": 132
},
{
"epoch": 0.13377758219651725,
"grad_norm": 19.184953207489116,
"learning_rate": 7.876746745502763e-06,
"loss": 2.667,
"step": 133
},
{
"epoch": 0.13478342867919785,
"grad_norm": 18.7606191056195,
"learning_rate": 7.888811758166946e-06,
"loss": 2.6638,
"step": 134
},
{
"epoch": 0.13578927516187841,
"grad_norm": 21.533928798672243,
"learning_rate": 7.9007870673052e-06,
"loss": 2.6812,
"step": 135
},
{
"epoch": 0.136795121644559,
"grad_norm": 21.969885173153724,
"learning_rate": 7.912673996972969e-06,
"loss": 2.6645,
"step": 136
},
{
"epoch": 0.13780096812723958,
"grad_norm": 23.026383962425232,
"learning_rate": 7.924473842125055e-06,
"loss": 2.6692,
"step": 137
},
{
"epoch": 0.13880681460992017,
"grad_norm": 25.308590349511185,
"learning_rate": 7.936187869462198e-06,
"loss": 2.7056,
"step": 138
},
{
"epoch": 0.13981266109260074,
"grad_norm": 21.063550016437297,
"learning_rate": 7.947817318247087e-06,
"loss": 2.6697,
"step": 139
},
{
"epoch": 0.14081850757528133,
"grad_norm": 24.867208465313563,
"learning_rate": 7.959363401091148e-06,
"loss": 2.6561,
"step": 140
},
{
"epoch": 0.1418243540579619,
"grad_norm": 19.72568818674115,
"learning_rate": 7.970827304713302e-06,
"loss": 2.6553,
"step": 141
},
{
"epoch": 0.1428302005406425,
"grad_norm": 23.770176802545095,
"learning_rate": 7.98221019067193e-06,
"loss": 2.6591,
"step": 142
},
{
"epoch": 0.14383604702332306,
"grad_norm": 20.067524140862375,
"learning_rate": 7.99351319607114e-06,
"loss": 2.6719,
"step": 143
},
{
"epoch": 0.14484189350600365,
"grad_norm": 31.768261766580125,
"learning_rate": 8.004737434242453e-06,
"loss": 2.6622,
"step": 144
},
{
"epoch": 0.14584773998868422,
"grad_norm": 25.556982955160176,
"learning_rate": 8.015883995402853e-06,
"loss": 2.6526,
"step": 145
},
{
"epoch": 0.14685358647136482,
"grad_norm": 23.847177634398342,
"learning_rate": 8.026953947290275e-06,
"loss": 2.6424,
"step": 146
},
{
"epoch": 0.14785943295404538,
"grad_norm": 27.152595480642916,
"learning_rate": 8.037948335777368e-06,
"loss": 2.6902,
"step": 147
},
{
"epoch": 0.14886527943672598,
"grad_norm": 18.022064635980716,
"learning_rate": 8.048868185464465e-06,
"loss": 2.5969,
"step": 148
},
{
"epoch": 0.14987112591940654,
"grad_norm": 27.024734522745767,
"learning_rate": 8.059714500252588e-06,
"loss": 2.6461,
"step": 149
},
{
"epoch": 0.15087697240208714,
"grad_norm": 16.732763939745368,
"learning_rate": 8.070488263897281e-06,
"loss": 2.6885,
"step": 150
},
{
"epoch": 0.1518828188847677,
"grad_norm": 24.51054903035366,
"learning_rate": 8.081190440544056e-06,
"loss": 2.6653,
"step": 151
},
{
"epoch": 0.1528886653674483,
"grad_norm": 21.596701273369327,
"learning_rate": 8.09182197524615e-06,
"loss": 2.6377,
"step": 152
},
{
"epoch": 0.15389451185012887,
"grad_norm": 25.587667640660342,
"learning_rate": 8.102383794465321e-06,
"loss": 2.6287,
"step": 153
},
{
"epoch": 0.15490035833280946,
"grad_norm": 28.92637297373861,
"learning_rate": 8.112876806556328e-06,
"loss": 2.6187,
"step": 154
},
{
"epoch": 0.15590620481549003,
"grad_norm": 20.936164888279848,
"learning_rate": 8.123301902235721e-06,
"loss": 2.634,
"step": 155
},
{
"epoch": 0.15691205129817062,
"grad_norm": 23.557901999266143,
"learning_rate": 8.133659955035566e-06,
"loss": 2.587,
"step": 156
},
{
"epoch": 0.1579178977808512,
"grad_norm": 21.159821995258497,
"learning_rate": 8.143951821742681e-06,
"loss": 2.6368,
"step": 157
},
{
"epoch": 0.15892374426353179,
"grad_norm": 24.642078265141432,
"learning_rate": 8.154178342823911e-06,
"loss": 2.6336,
"step": 158
},
{
"epoch": 0.15992959074621235,
"grad_norm": 21.423623973524016,
"learning_rate": 8.164340342837997e-06,
"loss": 2.6378,
"step": 159
},
{
"epoch": 0.16093543722889295,
"grad_norm": 20.788847198646504,
"learning_rate": 8.174438630834533e-06,
"loss": 2.5842,
"step": 160
},
{
"epoch": 0.1619412837115735,
"grad_norm": 20.628954050610897,
"learning_rate": 8.184474000740498e-06,
"loss": 2.6697,
"step": 161
},
{
"epoch": 0.1629471301942541,
"grad_norm": 22.39808575634474,
"learning_rate": 8.194447231734805e-06,
"loss": 2.6348,
"step": 162
},
{
"epoch": 0.16395297667693468,
"grad_norm": 24.746404372242566,
"learning_rate": 8.204359088611344e-06,
"loss": 2.6288,
"step": 163
},
{
"epoch": 0.16495882315961527,
"grad_norm": 23.61401081832116,
"learning_rate": 8.214210322130917e-06,
"loss": 2.6606,
"step": 164
},
{
"epoch": 0.16596466964229584,
"grad_norm": 22.15134362505235,
"learning_rate": 8.224001669362461e-06,
"loss": 2.6394,
"step": 165
},
{
"epoch": 0.16697051612497643,
"grad_norm": 24.75566935763776,
"learning_rate": 8.233733854013994e-06,
"loss": 2.6418,
"step": 166
},
{
"epoch": 0.167976362607657,
"grad_norm": 20.403196739056543,
"learning_rate": 8.24340758675358e-06,
"loss": 2.6979,
"step": 167
},
{
"epoch": 0.1689822090903376,
"grad_norm": 21.928902144665866,
"learning_rate": 8.253023565520753e-06,
"loss": 2.6535,
"step": 168
},
{
"epoch": 0.16998805557301816,
"grad_norm": 21.038785162406096,
"learning_rate": 8.262582475828682e-06,
"loss": 2.6418,
"step": 169
},
{
"epoch": 0.17099390205569875,
"grad_norm": 15.673934043236647,
"learning_rate": 8.272084991057405e-06,
"loss": 2.6811,
"step": 170
},
{
"epoch": 0.17199974853837932,
"grad_norm": 22.43245971157093,
"learning_rate": 8.281531772738503e-06,
"loss": 2.6596,
"step": 171
},
{
"epoch": 0.17300559502105992,
"grad_norm": 18.71520704132205,
"learning_rate": 8.290923470831441e-06,
"loss": 2.6451,
"step": 172
},
{
"epoch": 0.17401144150374048,
"grad_norm": 20.616928637039162,
"learning_rate": 8.300260723991916e-06,
"loss": 2.6314,
"step": 173
},
{
"epoch": 0.17501728798642108,
"grad_norm": 18.91356082247342,
"learning_rate": 8.309544159832458e-06,
"loss": 2.6167,
"step": 174
},
{
"epoch": 0.17602313446910164,
"grad_norm": 18.162533499067138,
"learning_rate": 8.318774395175584e-06,
"loss": 2.6243,
"step": 175
},
{
"epoch": 0.17702898095178224,
"grad_norm": 20.50907483416954,
"learning_rate": 8.327952036299713e-06,
"loss": 2.5896,
"step": 176
},
{
"epoch": 0.1780348274344628,
"grad_norm": 19.102928768874477,
"learning_rate": 8.33707767917815e-06,
"loss": 2.6691,
"step": 177
},
{
"epoch": 0.1790406739171434,
"grad_norm": 17.220346981660366,
"learning_rate": 8.34615190971129e-06,
"loss": 2.5311,
"step": 178
},
{
"epoch": 0.18004652039982397,
"grad_norm": 18.526421388087602,
"learning_rate": 8.35517530395237e-06,
"loss": 2.6687,
"step": 179
},
{
"epoch": 0.18105236688250456,
"grad_norm": 18.68714814663516,
"learning_rate": 8.364148428326887e-06,
"loss": 2.6537,
"step": 180
},
{
"epoch": 0.18205821336518513,
"grad_norm": 19.07870561029967,
"learning_rate": 8.373071839845973e-06,
"loss": 2.5946,
"step": 181
},
{
"epoch": 0.18306405984786572,
"grad_norm": 16.902568714349183,
"learning_rate": 8.381946086313868e-06,
"loss": 2.6438,
"step": 182
},
{
"epoch": 0.1840699063305463,
"grad_norm": 19.86628523072272,
"learning_rate": 8.39077170652972e-06,
"loss": 2.6064,
"step": 183
},
{
"epoch": 0.18507575281322688,
"grad_norm": 17.50543397408068,
"learning_rate": 8.399549230483884e-06,
"loss": 2.6389,
"step": 184
},
{
"epoch": 0.18608159929590745,
"grad_norm": 21.975997332922287,
"learning_rate": 8.408279179548899e-06,
"loss": 2.6116,
"step": 185
},
{
"epoch": 0.18708744577858805,
"grad_norm": 20.239413306373145,
"learning_rate": 8.416962066665327e-06,
"loss": 2.6954,
"step": 186
},
{
"epoch": 0.1880932922612686,
"grad_norm": 21.38873332374349,
"learning_rate": 8.425598396522585e-06,
"loss": 2.6121,
"step": 187
},
{
"epoch": 0.1890991387439492,
"grad_norm": 21.590789079297604,
"learning_rate": 8.43418866573499e-06,
"loss": 2.6493,
"step": 188
},
{
"epoch": 0.19010498522662977,
"grad_norm": 20.836744203370017,
"learning_rate": 8.442733363013107e-06,
"loss": 2.627,
"step": 189
},
{
"epoch": 0.19111083170931037,
"grad_norm": 22.441001841694202,
"learning_rate": 8.451232969330584e-06,
"loss": 2.6559,
"step": 190
},
{
"epoch": 0.19211667819199094,
"grad_norm": 23.400210842885595,
"learning_rate": 8.459687958086613e-06,
"loss": 2.6221,
"step": 191
},
{
"epoch": 0.19312252467467153,
"grad_norm": 17.126955017293337,
"learning_rate": 8.468098795264139e-06,
"loss": 2.6686,
"step": 192
},
{
"epoch": 0.1941283711573521,
"grad_norm": 22.286487375080497,
"learning_rate": 8.476465939583975e-06,
"loss": 2.5953,
"step": 193
},
{
"epoch": 0.1951342176400327,
"grad_norm": 22.623393456148232,
"learning_rate": 8.484789842654914e-06,
"loss": 2.645,
"step": 194
},
{
"epoch": 0.19614006412271326,
"grad_norm": 20.01494469131131,
"learning_rate": 8.493070949120002e-06,
"loss": 2.6935,
"step": 195
},
{
"epoch": 0.19714591060539385,
"grad_norm": 20.652229445798564,
"learning_rate": 8.501309696799054e-06,
"loss": 2.5595,
"step": 196
},
{
"epoch": 0.19815175708807442,
"grad_norm": 19.356436859627642,
"learning_rate": 8.509506516827565e-06,
"loss": 2.6037,
"step": 197
},
{
"epoch": 0.19915760357075502,
"grad_norm": 19.375388210507882,
"learning_rate": 8.517661833792069e-06,
"loss": 2.6151,
"step": 198
},
{
"epoch": 0.20016345005343558,
"grad_norm": 19.58164413054047,
"learning_rate": 8.52577606586212e-06,
"loss": 2.646,
"step": 199
},
{
"epoch": 0.20116929653611618,
"grad_norm": 19.53007585706627,
"learning_rate": 8.533849624918969e-06,
"loss": 2.6153,
"step": 200
},
{
"epoch": 0.20217514301879674,
"grad_norm": 18.412307623077677,
"learning_rate": 8.541882916680986e-06,
"loss": 2.6147,
"step": 201
},
{
"epoch": 0.20318098950147734,
"grad_norm": 20.96726260777965,
"learning_rate": 8.549876340826033e-06,
"loss": 2.6719,
"step": 202
},
{
"epoch": 0.2041868359841579,
"grad_norm": 15.753376282394372,
"learning_rate": 8.55783029111076e-06,
"loss": 2.5969,
"step": 203
},
{
"epoch": 0.2051926824668385,
"grad_norm": 19.19212639218277,
"learning_rate": 8.565745155487009e-06,
"loss": 2.6292,
"step": 204
},
{
"epoch": 0.20619852894951907,
"grad_norm": 17.130226605076096,
"learning_rate": 8.573621316215352e-06,
"loss": 2.5878,
"step": 205
},
{
"epoch": 0.20720437543219966,
"grad_norm": 20.09366506036711,
"learning_rate": 8.581459149975866e-06,
"loss": 2.628,
"step": 206
},
{
"epoch": 0.20821022191488023,
"grad_norm": 18.051831505165058,
"learning_rate": 8.589259027976237e-06,
"loss": 2.6775,
"step": 207
},
{
"epoch": 0.20921606839756082,
"grad_norm": 20.88839122977681,
"learning_rate": 8.597021316057254e-06,
"loss": 2.5629,
"step": 208
},
{
"epoch": 0.2102219148802414,
"grad_norm": 20.052445051755896,
"learning_rate": 8.604746374795765e-06,
"loss": 2.6581,
"step": 209
},
{
"epoch": 0.21122776136292198,
"grad_norm": 20.308993784102455,
"learning_rate": 8.612434559605189e-06,
"loss": 2.6295,
"step": 210
},
{
"epoch": 0.21223360784560258,
"grad_norm": 18.750434945903855,
"learning_rate": 8.620086220833631e-06,
"loss": 2.6288,
"step": 211
},
{
"epoch": 0.21323945432828315,
"grad_norm": 17.71083780989216,
"learning_rate": 8.627701703859685e-06,
"loss": 2.6408,
"step": 212
},
{
"epoch": 0.21424530081096374,
"grad_norm": 18.24620720661975,
"learning_rate": 8.63528134918597e-06,
"loss": 2.6134,
"step": 213
},
{
"epoch": 0.2152511472936443,
"grad_norm": 16.69848128026314,
"learning_rate": 8.642825492530494e-06,
"loss": 2.6553,
"step": 214
},
{
"epoch": 0.2162569937763249,
"grad_norm": 17.731749852066827,
"learning_rate": 8.650334464915875e-06,
"loss": 2.6056,
"step": 215
},
{
"epoch": 0.21726284025900547,
"grad_norm": 15.675950483577644,
"learning_rate": 8.657808592756493e-06,
"loss": 2.5621,
"step": 216
},
{
"epoch": 0.21826868674168606,
"grad_norm": 19.14882202694866,
"learning_rate": 8.665248197943628e-06,
"loss": 2.6483,
"step": 217
},
{
"epoch": 0.21927453322436663,
"grad_norm": 18.258290501797685,
"learning_rate": 8.672653597928656e-06,
"loss": 2.6001,
"step": 218
},
{
"epoch": 0.22028037970704722,
"grad_norm": 17.825922963762647,
"learning_rate": 8.680025105804317e-06,
"loss": 2.6576,
"step": 219
},
{
"epoch": 0.2212862261897278,
"grad_norm": 18.489862697446604,
"learning_rate": 8.68736303038415e-06,
"loss": 2.6701,
"step": 220
},
{
"epoch": 0.2222920726724084,
"grad_norm": 20.623353777687228,
"learning_rate": 8.694667676280122e-06,
"loss": 2.66,
"step": 221
},
{
"epoch": 0.22329791915508895,
"grad_norm": 20.140706964863128,
"learning_rate": 8.701939343978507e-06,
"loss": 2.6114,
"step": 222
},
{
"epoch": 0.22430376563776955,
"grad_norm": 21.084508483111602,
"learning_rate": 8.709178329914038e-06,
"loss": 2.6374,
"step": 223
},
{
"epoch": 0.22530961212045011,
"grad_norm": 21.044763301093216,
"learning_rate": 8.716384926542442e-06,
"loss": 2.5886,
"step": 224
},
{
"epoch": 0.2263154586031307,
"grad_norm": 18.074736561415794,
"learning_rate": 8.723559422411321e-06,
"loss": 2.6019,
"step": 225
},
{
"epoch": 0.22732130508581128,
"grad_norm": 18.34158694872784,
"learning_rate": 8.730702102229487e-06,
"loss": 2.6103,
"step": 226
},
{
"epoch": 0.22832715156849187,
"grad_norm": 20.820577289047332,
"learning_rate": 8.737813246934741e-06,
"loss": 2.6454,
"step": 227
},
{
"epoch": 0.22933299805117244,
"grad_norm": 21.668929247510405,
"learning_rate": 8.744893133760191e-06,
"loss": 2.6261,
"step": 228
},
{
"epoch": 0.23033884453385303,
"grad_norm": 19.292963044451643,
"learning_rate": 8.751942036299099e-06,
"loss": 2.5645,
"step": 229
},
{
"epoch": 0.2313446910165336,
"grad_norm": 21.315973049999645,
"learning_rate": 8.758960224568318e-06,
"loss": 2.5653,
"step": 230
},
{
"epoch": 0.2323505374992142,
"grad_norm": 18.390322171060433,
"learning_rate": 8.765947965070369e-06,
"loss": 2.6106,
"step": 231
},
{
"epoch": 0.23335638398189476,
"grad_norm": 18.76067528787121,
"learning_rate": 8.772905520854146e-06,
"loss": 2.6171,
"step": 232
},
{
"epoch": 0.23436223046457536,
"grad_norm": 24.814248745806943,
"learning_rate": 8.779833151574344e-06,
"loss": 2.6466,
"step": 233
},
{
"epoch": 0.23536807694725592,
"grad_norm": 22.919276078065515,
"learning_rate": 8.786731113549606e-06,
"loss": 2.6033,
"step": 234
},
{
"epoch": 0.23637392342993652,
"grad_norm": 18.520466660426873,
"learning_rate": 8.793599659819425e-06,
"loss": 2.606,
"step": 235
},
{
"epoch": 0.23737976991261708,
"grad_norm": 22.348067724347846,
"learning_rate": 8.800439040199838e-06,
"loss": 2.5912,
"step": 236
},
{
"epoch": 0.23838561639529768,
"grad_norm": 19.173263256977286,
"learning_rate": 8.807249501337953e-06,
"loss": 2.6346,
"step": 237
},
{
"epoch": 0.23939146287797824,
"grad_norm": 23.383354298666408,
"learning_rate": 8.814031286765312e-06,
"loss": 2.6386,
"step": 238
},
{
"epoch": 0.24039730936065884,
"grad_norm": 25.729808734175805,
"learning_rate": 8.820784636950157e-06,
"loss": 2.625,
"step": 239
},
{
"epoch": 0.2414031558433394,
"grad_norm": 18.210794583054227,
"learning_rate": 8.827509789348575e-06,
"loss": 2.5976,
"step": 240
},
{
"epoch": 0.24240900232602,
"grad_norm": 17.672902674151604,
"learning_rate": 8.834206978454614e-06,
"loss": 2.627,
"step": 241
},
{
"epoch": 0.24341484880870057,
"grad_norm": 17.827338301226895,
"learning_rate": 8.84087643584933e-06,
"loss": 2.588,
"step": 242
},
{
"epoch": 0.24442069529138116,
"grad_norm": 17.06456531549323,
"learning_rate": 8.847518390248845e-06,
"loss": 2.6316,
"step": 243
},
{
"epoch": 0.24542654177406173,
"grad_norm": 20.713440314623234,
"learning_rate": 8.854133067551409e-06,
"loss": 2.5697,
"step": 244
},
{
"epoch": 0.24643238825674232,
"grad_norm": 19.6435903200954,
"learning_rate": 8.860720690883492e-06,
"loss": 2.5903,
"step": 245
},
{
"epoch": 0.2474382347394229,
"grad_norm": 18.399526685396268,
"learning_rate": 8.867281480644957e-06,
"loss": 2.6169,
"step": 246
},
{
"epoch": 0.24844408122210349,
"grad_norm": 19.663511314871595,
"learning_rate": 8.873815654553305e-06,
"loss": 2.5982,
"step": 247
},
{
"epoch": 0.24944992770478405,
"grad_norm": 22.569466086155717,
"learning_rate": 8.880323427687015e-06,
"loss": 2.5776,
"step": 248
},
{
"epoch": 0.2504557741874646,
"grad_norm": 22.250217638563107,
"learning_rate": 8.886805012528034e-06,
"loss": 2.612,
"step": 249
},
{
"epoch": 0.2514616206701452,
"grad_norm": 20.751848059953005,
"learning_rate": 8.893260619003403e-06,
"loss": 2.5945,
"step": 250
},
{
"epoch": 0.2524674671528258,
"grad_norm": 21.836021029599017,
"learning_rate": 8.899690454526055e-06,
"loss": 2.6152,
"step": 251
},
{
"epoch": 0.2534733136355064,
"grad_norm": 17.93363953654319,
"learning_rate": 8.906094724034795e-06,
"loss": 2.5886,
"step": 252
},
{
"epoch": 0.25447916011818694,
"grad_norm": 20.885718013721352,
"learning_rate": 8.9124736300335e-06,
"loss": 2.5972,
"step": 253
},
{
"epoch": 0.25548500660086754,
"grad_norm": 21.796601989759413,
"learning_rate": 8.91882737262954e-06,
"loss": 2.5547,
"step": 254
},
{
"epoch": 0.25649085308354813,
"grad_norm": 17.58817706954125,
"learning_rate": 8.925156149571445e-06,
"loss": 2.5832,
"step": 255
},
{
"epoch": 0.2574966995662287,
"grad_norm": 21.59816615398544,
"learning_rate": 8.931460156285827e-06,
"loss": 2.5833,
"step": 256
},
{
"epoch": 0.25850254604890927,
"grad_norm": 18.077149818844823,
"learning_rate": 8.937739585913602e-06,
"loss": 2.5771,
"step": 257
},
{
"epoch": 0.25950839253158986,
"grad_norm": 20.454974748323327,
"learning_rate": 8.943994629345481e-06,
"loss": 2.6646,
"step": 258
},
{
"epoch": 0.26051423901427045,
"grad_norm": 19.680650671359444,
"learning_rate": 8.950225475256808e-06,
"loss": 2.6518,
"step": 259
},
{
"epoch": 0.26152008549695105,
"grad_norm": 17.07370495402799,
"learning_rate": 8.956432310141688e-06,
"loss": 2.6054,
"step": 260
},
{
"epoch": 0.2625259319796316,
"grad_norm": 20.44556881953542,
"learning_rate": 8.962615318346499e-06,
"loss": 2.487,
"step": 261
},
{
"epoch": 0.2635317784623122,
"grad_norm": 17.12917889025369,
"learning_rate": 8.968774682102729e-06,
"loss": 2.6147,
"step": 262
},
{
"epoch": 0.2645376249449928,
"grad_norm": 19.621476667862606,
"learning_rate": 8.974910581559217e-06,
"loss": 2.5783,
"step": 263
},
{
"epoch": 0.26554347142767337,
"grad_norm": 16.290476284537352,
"learning_rate": 8.981023194813755e-06,
"loss": 2.666,
"step": 264
},
{
"epoch": 0.2665493179103539,
"grad_norm": 21.622288460717243,
"learning_rate": 8.987112697944119e-06,
"loss": 2.6843,
"step": 265
},
{
"epoch": 0.2675551643930345,
"grad_norm": 17.770496455788983,
"learning_rate": 8.993179265038493e-06,
"loss": 2.6217,
"step": 266
},
{
"epoch": 0.2685610108757151,
"grad_norm": 18.814542670464103,
"learning_rate": 8.999223068225332e-06,
"loss": 2.5714,
"step": 267
},
{
"epoch": 0.2695668573583957,
"grad_norm": 18.786521072518283,
"learning_rate": 9.005244277702674e-06,
"loss": 2.5553,
"step": 268
},
{
"epoch": 0.27057270384107623,
"grad_norm": 21.070488274527378,
"learning_rate": 9.011243061766895e-06,
"loss": 2.6352,
"step": 269
},
{
"epoch": 0.27157855032375683,
"grad_norm": 19.816693607724073,
"learning_rate": 9.017219586840929e-06,
"loss": 2.6066,
"step": 270
},
{
"epoch": 0.2725843968064374,
"grad_norm": 18.96107953875644,
"learning_rate": 9.023174017501975e-06,
"loss": 2.559,
"step": 271
},
{
"epoch": 0.273590243289118,
"grad_norm": 26.54474932655252,
"learning_rate": 9.029106516508698e-06,
"loss": 2.5407,
"step": 272
},
{
"epoch": 0.27459608977179856,
"grad_norm": 18.8523610186751,
"learning_rate": 9.03501724482791e-06,
"loss": 2.5851,
"step": 273
},
{
"epoch": 0.27560193625447915,
"grad_norm": 19.597045879097127,
"learning_rate": 9.040906361660785e-06,
"loss": 2.5512,
"step": 274
},
{
"epoch": 0.27660778273715975,
"grad_norm": 23.74080554612753,
"learning_rate": 9.046774024468585e-06,
"loss": 2.6118,
"step": 275
},
{
"epoch": 0.27761362921984034,
"grad_norm": 22.714669056716087,
"learning_rate": 9.052620388997924e-06,
"loss": 2.5784,
"step": 276
},
{
"epoch": 0.2786194757025209,
"grad_norm": 23.248188625995795,
"learning_rate": 9.058445609305574e-06,
"loss": 2.6503,
"step": 277
},
{
"epoch": 0.2796253221852015,
"grad_norm": 22.625705056604087,
"learning_rate": 9.064249837782815e-06,
"loss": 2.5583,
"step": 278
},
{
"epoch": 0.28063116866788207,
"grad_norm": 24.60212715825136,
"learning_rate": 9.070033225179367e-06,
"loss": 2.6041,
"step": 279
},
{
"epoch": 0.28163701515056266,
"grad_norm": 22.07266263058576,
"learning_rate": 9.075795920626876e-06,
"loss": 2.5903,
"step": 280
},
{
"epoch": 0.2826428616332432,
"grad_norm": 21.88899389143494,
"learning_rate": 9.081538071661991e-06,
"loss": 2.5693,
"step": 281
},
{
"epoch": 0.2836487081159238,
"grad_norm": 20.285327057873072,
"learning_rate": 9.087259824249031e-06,
"loss": 2.5804,
"step": 282
},
{
"epoch": 0.2846545545986044,
"grad_norm": 17.86840355259839,
"learning_rate": 9.092961322802238e-06,
"loss": 2.5799,
"step": 283
},
{
"epoch": 0.285660401081285,
"grad_norm": 16.179377763205867,
"learning_rate": 9.098642710207657e-06,
"loss": 2.6577,
"step": 284
},
{
"epoch": 0.2866662475639655,
"grad_norm": 18.881043699779436,
"learning_rate": 9.104304127844625e-06,
"loss": 2.5943,
"step": 285
},
{
"epoch": 0.2876720940466461,
"grad_norm": 19.32116500725666,
"learning_rate": 9.10994571560687e-06,
"loss": 2.5656,
"step": 286
},
{
"epoch": 0.2886779405293267,
"grad_norm": 17.056212076430626,
"learning_rate": 9.11556761192326e-06,
"loss": 2.5439,
"step": 287
},
{
"epoch": 0.2896837870120073,
"grad_norm": 17.50398402272584,
"learning_rate": 9.12116995377818e-06,
"loss": 2.5772,
"step": 288
},
{
"epoch": 0.29068963349468785,
"grad_norm": 19.554296082869655,
"learning_rate": 9.126752876731566e-06,
"loss": 2.5929,
"step": 289
},
{
"epoch": 0.29169547997736844,
"grad_norm": 21.68217435347834,
"learning_rate": 9.13231651493858e-06,
"loss": 2.5923,
"step": 290
},
{
"epoch": 0.29270132646004904,
"grad_norm": 20.905430617187555,
"learning_rate": 9.137861001168956e-06,
"loss": 2.5802,
"step": 291
},
{
"epoch": 0.29370717294272963,
"grad_norm": 20.198543571392612,
"learning_rate": 9.143386466826003e-06,
"loss": 2.5822,
"step": 292
},
{
"epoch": 0.29471301942541017,
"grad_norm": 23.77871648988695,
"learning_rate": 9.148893041965311e-06,
"loss": 2.608,
"step": 293
},
{
"epoch": 0.29571886590809077,
"grad_norm": 20.48204598393149,
"learning_rate": 9.154380855313096e-06,
"loss": 2.586,
"step": 294
},
{
"epoch": 0.29672471239077136,
"grad_norm": 18.098438212679774,
"learning_rate": 9.159850034284274e-06,
"loss": 2.6228,
"step": 295
},
{
"epoch": 0.29773055887345196,
"grad_norm": 18.784691263545824,
"learning_rate": 9.165300705000193e-06,
"loss": 2.5802,
"step": 296
},
{
"epoch": 0.2987364053561325,
"grad_norm": 18.5638121249419,
"learning_rate": 9.170732992306109e-06,
"loss": 2.5466,
"step": 297
},
{
"epoch": 0.2997422518388131,
"grad_norm": 18.754954662020456,
"learning_rate": 9.176147019788316e-06,
"loss": 2.6204,
"step": 298
},
{
"epoch": 0.3007480983214937,
"grad_norm": 19.634341646921474,
"learning_rate": 9.18154290979104e-06,
"loss": 2.569,
"step": 299
},
{
"epoch": 0.3017539448041743,
"grad_norm": 16.960368671694702,
"learning_rate": 9.18692078343301e-06,
"loss": 2.5591,
"step": 300
},
{
"epoch": 0.3027597912868548,
"grad_norm": 17.789077674900664,
"learning_rate": 9.192280760623784e-06,
"loss": 2.6259,
"step": 301
},
{
"epoch": 0.3037656377695354,
"grad_norm": 17.01187285822207,
"learning_rate": 9.197622960079784e-06,
"loss": 2.6004,
"step": 302
},
{
"epoch": 0.304771484252216,
"grad_norm": 20.164202493945783,
"learning_rate": 9.202947499340073e-06,
"loss": 2.6163,
"step": 303
},
{
"epoch": 0.3057773307348966,
"grad_norm": 17.737721757957956,
"learning_rate": 9.208254494781877e-06,
"loss": 2.5653,
"step": 304
},
{
"epoch": 0.30678317721757714,
"grad_norm": 18.472587797259994,
"learning_rate": 9.213544061635843e-06,
"loss": 2.586,
"step": 305
},
{
"epoch": 0.30778902370025774,
"grad_norm": 15.17012695048484,
"learning_rate": 9.21881631400105e-06,
"loss": 2.5346,
"step": 306
},
{
"epoch": 0.30879487018293833,
"grad_norm": 19.95709941623259,
"learning_rate": 9.224071364859784e-06,
"loss": 2.6546,
"step": 307
},
{
"epoch": 0.3098007166656189,
"grad_norm": 20.924787917082103,
"learning_rate": 9.229309326092056e-06,
"loss": 2.5815,
"step": 308
},
{
"epoch": 0.31080656314829946,
"grad_norm": 17.65260316523874,
"learning_rate": 9.234530308489906e-06,
"loss": 2.4879,
"step": 309
},
{
"epoch": 0.31181240963098006,
"grad_norm": 17.954780984714258,
"learning_rate": 9.239734421771449e-06,
"loss": 2.5751,
"step": 310
},
{
"epoch": 0.31281825611366065,
"grad_norm": 20.309850645532567,
"learning_rate": 9.244921774594723e-06,
"loss": 2.6018,
"step": 311
},
{
"epoch": 0.31382410259634125,
"grad_norm": 19.4937688848126,
"learning_rate": 9.250092474571294e-06,
"loss": 2.5942,
"step": 312
},
{
"epoch": 0.3148299490790218,
"grad_norm": 19.8969123554459,
"learning_rate": 9.255246628279656e-06,
"loss": 2.6084,
"step": 313
},
{
"epoch": 0.3158357955617024,
"grad_norm": 21.311968170601972,
"learning_rate": 9.260384341278409e-06,
"loss": 2.5642,
"step": 314
},
{
"epoch": 0.316841642044383,
"grad_norm": 19.417179511250875,
"learning_rate": 9.26550571811923e-06,
"loss": 2.6014,
"step": 315
},
{
"epoch": 0.31784748852706357,
"grad_norm": 18.3517298148521,
"learning_rate": 9.270610862359639e-06,
"loss": 2.5491,
"step": 316
},
{
"epoch": 0.31885333500974417,
"grad_norm": 18.48869905606377,
"learning_rate": 9.275699876575568e-06,
"loss": 2.6088,
"step": 317
},
{
"epoch": 0.3198591814924247,
"grad_norm": 18.715905721710747,
"learning_rate": 9.280772862373725e-06,
"loss": 2.5831,
"step": 318
},
{
"epoch": 0.3208650279751053,
"grad_norm": 15.934927854149723,
"learning_rate": 9.285829920403762e-06,
"loss": 2.5761,
"step": 319
},
{
"epoch": 0.3218708744577859,
"grad_norm": 20.177416493199058,
"learning_rate": 9.290871150370263e-06,
"loss": 2.5826,
"step": 320
},
{
"epoch": 0.3228767209404665,
"grad_norm": 22.03439353905269,
"learning_rate": 9.295896651044535e-06,
"loss": 2.5901,
"step": 321
},
{
"epoch": 0.323882567423147,
"grad_norm": 21.417786500881242,
"learning_rate": 9.300906520276228e-06,
"loss": 2.5614,
"step": 322
},
{
"epoch": 0.3248884139058276,
"grad_norm": 17.202029957803592,
"learning_rate": 9.305900855004747e-06,
"loss": 2.6167,
"step": 323
},
{
"epoch": 0.3258942603885082,
"grad_norm": 22.29615093251031,
"learning_rate": 9.310879751270533e-06,
"loss": 2.5662,
"step": 324
},
{
"epoch": 0.3269001068711888,
"grad_norm": 17.64208925311755,
"learning_rate": 9.315843304226122e-06,
"loss": 2.5522,
"step": 325
},
{
"epoch": 0.32790595335386935,
"grad_norm": 19.98320678509978,
"learning_rate": 9.320791608147074e-06,
"loss": 2.6221,
"step": 326
},
{
"epoch": 0.32891179983654995,
"grad_norm": 16.13553457377954,
"learning_rate": 9.325724756442696e-06,
"loss": 2.5761,
"step": 327
},
{
"epoch": 0.32991764631923054,
"grad_norm": 17.31786699950618,
"learning_rate": 9.330642841666646e-06,
"loss": 2.5816,
"step": 328
},
{
"epoch": 0.33092349280191113,
"grad_norm": 18.712111337036895,
"learning_rate": 9.335545955527333e-06,
"loss": 2.6028,
"step": 329
},
{
"epoch": 0.3319293392845917,
"grad_norm": 17.48530200895543,
"learning_rate": 9.340434188898189e-06,
"loss": 2.6095,
"step": 330
},
{
"epoch": 0.33293518576727227,
"grad_norm": 20.06501341317573,
"learning_rate": 9.345307631827775e-06,
"loss": 2.5803,
"step": 331
},
{
"epoch": 0.33394103224995286,
"grad_norm": 18.81437868466945,
"learning_rate": 9.35016637354972e-06,
"loss": 2.5395,
"step": 332
},
{
"epoch": 0.33494687873263346,
"grad_norm": 20.253454715167823,
"learning_rate": 9.355010502492547e-06,
"loss": 2.5588,
"step": 333
},
{
"epoch": 0.335952725215314,
"grad_norm": 20.145901278398018,
"learning_rate": 9.359840106289308e-06,
"loss": 2.6307,
"step": 334
},
{
"epoch": 0.3369585716979946,
"grad_norm": 18.35310873936949,
"learning_rate": 9.36465527178711e-06,
"loss": 2.5903,
"step": 335
},
{
"epoch": 0.3379644181806752,
"grad_norm": 19.993529185424425,
"learning_rate": 9.369456085056482e-06,
"loss": 2.5692,
"step": 336
},
{
"epoch": 0.3389702646633558,
"grad_norm": 22.69340117991585,
"learning_rate": 9.374242631400604e-06,
"loss": 2.6053,
"step": 337
},
{
"epoch": 0.3399761111460363,
"grad_norm": 21.68690669161774,
"learning_rate": 9.37901499536441e-06,
"loss": 2.5805,
"step": 338
},
{
"epoch": 0.3409819576287169,
"grad_norm": 17.746874404228528,
"learning_rate": 9.383773260743527e-06,
"loss": 2.5855,
"step": 339
},
{
"epoch": 0.3419878041113975,
"grad_norm": 22.807692121973822,
"learning_rate": 9.388517510593132e-06,
"loss": 2.6305,
"step": 340
},
{
"epoch": 0.3429936505940781,
"grad_norm": 19.00067460688236,
"learning_rate": 9.39324782723663e-06,
"loss": 2.5656,
"step": 341
},
{
"epoch": 0.34399949707675864,
"grad_norm": 20.121711729728055,
"learning_rate": 9.39796429227423e-06,
"loss": 2.5423,
"step": 342
},
{
"epoch": 0.34500534355943924,
"grad_norm": 20.64153176834041,
"learning_rate": 9.402666986591398e-06,
"loss": 2.6087,
"step": 343
},
{
"epoch": 0.34601119004211983,
"grad_norm": 20.27675989420653,
"learning_rate": 9.407355990367169e-06,
"loss": 2.5557,
"step": 344
},
{
"epoch": 0.3470170365248004,
"grad_norm": 17.450991828252263,
"learning_rate": 9.41203138308236e-06,
"loss": 2.5362,
"step": 345
},
{
"epoch": 0.34802288300748097,
"grad_norm": 17.74999896509276,
"learning_rate": 9.416693243527644e-06,
"loss": 2.5383,
"step": 346
},
{
"epoch": 0.34902872949016156,
"grad_norm": 22.025427737463747,
"learning_rate": 9.421341649811525e-06,
"loss": 2.5713,
"step": 347
},
{
"epoch": 0.35003457597284215,
"grad_norm": 21.459704360516515,
"learning_rate": 9.425976679368188e-06,
"loss": 2.5564,
"step": 348
},
{
"epoch": 0.35104042245552275,
"grad_norm": 19.988619711523008,
"learning_rate": 9.43059840896523e-06,
"loss": 2.6236,
"step": 349
},
{
"epoch": 0.3520462689382033,
"grad_norm": 18.045344386817177,
"learning_rate": 9.43520691471131e-06,
"loss": 2.5872,
"step": 350
},
{
"epoch": 0.3530521154208839,
"grad_norm": 22.81555309347901,
"learning_rate": 9.439802272063646e-06,
"loss": 2.5601,
"step": 351
},
{
"epoch": 0.3540579619035645,
"grad_norm": 22.381498108516144,
"learning_rate": 9.444384555835443e-06,
"loss": 2.582,
"step": 352
},
{
"epoch": 0.35506380838624507,
"grad_norm": 19.824711666359768,
"learning_rate": 9.44895384020319e-06,
"loss": 2.5217,
"step": 353
},
{
"epoch": 0.3560696548689256,
"grad_norm": 21.63777305031021,
"learning_rate": 9.453510198713878e-06,
"loss": 2.582,
"step": 354
},
{
"epoch": 0.3570755013516062,
"grad_norm": 18.452715228073657,
"learning_rate": 9.458053704292093e-06,
"loss": 2.5703,
"step": 355
},
{
"epoch": 0.3580813478342868,
"grad_norm": 16.645525833264035,
"learning_rate": 9.462584429247019e-06,
"loss": 2.6185,
"step": 356
},
{
"epoch": 0.3590871943169674,
"grad_norm": 17.993668104121088,
"learning_rate": 9.467102445279352e-06,
"loss": 2.5823,
"step": 357
},
{
"epoch": 0.36009304079964793,
"grad_norm": 16.501004168798346,
"learning_rate": 9.471607823488098e-06,
"loss": 2.5876,
"step": 358
},
{
"epoch": 0.36109888728232853,
"grad_norm": 20.814369689080046,
"learning_rate": 9.476100634377292e-06,
"loss": 2.5592,
"step": 359
},
{
"epoch": 0.3621047337650091,
"grad_norm": 17.443045103109935,
"learning_rate": 9.480580947862615e-06,
"loss": 2.5701,
"step": 360
},
{
"epoch": 0.3631105802476897,
"grad_norm": 21.11025374969233,
"learning_rate": 9.485048833277928e-06,
"loss": 2.6155,
"step": 361
},
{
"epoch": 0.36411642673037026,
"grad_norm": 17.98585113382355,
"learning_rate": 9.489504359381702e-06,
"loss": 2.5601,
"step": 362
},
{
"epoch": 0.36512227321305085,
"grad_norm": 17.523295432105954,
"learning_rate": 9.49394759436337e-06,
"loss": 2.5736,
"step": 363
},
{
"epoch": 0.36612811969573145,
"grad_norm": 21.831805903399736,
"learning_rate": 9.498378605849596e-06,
"loss": 2.5353,
"step": 364
},
{
"epoch": 0.36713396617841204,
"grad_norm": 19.70767544694675,
"learning_rate": 9.502797460910437e-06,
"loss": 2.609,
"step": 365
},
{
"epoch": 0.3681398126610926,
"grad_norm": 23.989716952193756,
"learning_rate": 9.507204226065449e-06,
"loss": 2.5952,
"step": 366
},
{
"epoch": 0.3691456591437732,
"grad_norm": 20.218287412951227,
"learning_rate": 9.511598967289681e-06,
"loss": 2.5784,
"step": 367
},
{
"epoch": 0.37015150562645377,
"grad_norm": 20.932848929843594,
"learning_rate": 9.515981750019612e-06,
"loss": 2.5482,
"step": 368
},
{
"epoch": 0.37115735210913436,
"grad_norm": 24.64683126624515,
"learning_rate": 9.520352639158998e-06,
"loss": 2.5726,
"step": 369
},
{
"epoch": 0.3721631985918149,
"grad_norm": 20.307454841997078,
"learning_rate": 9.524711699084629e-06,
"loss": 2.6273,
"step": 370
},
{
"epoch": 0.3731690450744955,
"grad_norm": 24.994119900117653,
"learning_rate": 9.529058993652026e-06,
"loss": 2.5826,
"step": 371
},
{
"epoch": 0.3741748915571761,
"grad_norm": 21.779145120794347,
"learning_rate": 9.533394586201055e-06,
"loss": 2.5547,
"step": 372
},
{
"epoch": 0.3751807380398567,
"grad_norm": 21.300796801514867,
"learning_rate": 9.537718539561456e-06,
"loss": 2.5621,
"step": 373
},
{
"epoch": 0.3761865845225372,
"grad_norm": 24.15121491274878,
"learning_rate": 9.542030916058314e-06,
"loss": 2.5851,
"step": 374
},
{
"epoch": 0.3771924310052178,
"grad_norm": 15.35693222917725,
"learning_rate": 9.546331777517445e-06,
"loss": 2.5435,
"step": 375
},
{
"epoch": 0.3781982774878984,
"grad_norm": 24.412432857133343,
"learning_rate": 9.550621185270719e-06,
"loss": 2.6019,
"step": 376
},
{
"epoch": 0.379204123970579,
"grad_norm": 25.613210763882368,
"learning_rate": 9.5548992001613e-06,
"loss": 2.5155,
"step": 377
},
{
"epoch": 0.38020997045325955,
"grad_norm": 22.095030230113984,
"learning_rate": 9.559165882548835e-06,
"loss": 2.5302,
"step": 378
},
{
"epoch": 0.38121581693594014,
"grad_norm": 24.99054102758044,
"learning_rate": 9.563421292314553e-06,
"loss": 2.5662,
"step": 379
},
{
"epoch": 0.38222166341862074,
"grad_norm": 18.606776693809763,
"learning_rate": 9.567665488866313e-06,
"loss": 2.5297,
"step": 380
},
{
"epoch": 0.38322750990130133,
"grad_norm": 20.35605026910687,
"learning_rate": 9.571898531143582e-06,
"loss": 2.5562,
"step": 381
},
{
"epoch": 0.38423335638398187,
"grad_norm": 26.209068443572,
"learning_rate": 9.576120477622342e-06,
"loss": 2.5664,
"step": 382
},
{
"epoch": 0.38523920286666247,
"grad_norm": 19.305248037487228,
"learning_rate": 9.580331386319938e-06,
"loss": 2.5937,
"step": 383
},
{
"epoch": 0.38624504934934306,
"grad_norm": 25.685627098810052,
"learning_rate": 9.584531314799869e-06,
"loss": 2.5336,
"step": 384
},
{
"epoch": 0.38725089583202366,
"grad_norm": 20.289828015945567,
"learning_rate": 9.588720320176494e-06,
"loss": 2.5348,
"step": 385
},
{
"epoch": 0.3882567423147042,
"grad_norm": 21.9547365407109,
"learning_rate": 9.592898459119703e-06,
"loss": 2.5505,
"step": 386
},
{
"epoch": 0.3892625887973848,
"grad_norm": 23.474353452783447,
"learning_rate": 9.597065787859523e-06,
"loss": 2.5506,
"step": 387
},
{
"epoch": 0.3902684352800654,
"grad_norm": 19.100670325198177,
"learning_rate": 9.601222362190642e-06,
"loss": 2.5846,
"step": 388
},
{
"epoch": 0.391274281762746,
"grad_norm": 21.84151290352025,
"learning_rate": 9.605368237476904e-06,
"loss": 2.5172,
"step": 389
},
{
"epoch": 0.3922801282454265,
"grad_norm": 17.584086365684314,
"learning_rate": 9.60950346865573e-06,
"loss": 2.5059,
"step": 390
},
{
"epoch": 0.3932859747281071,
"grad_norm": 20.781615247760833,
"learning_rate": 9.613628110242482e-06,
"loss": 2.5477,
"step": 391
},
{
"epoch": 0.3942918212107877,
"grad_norm": 19.38108993879063,
"learning_rate": 9.617742216334784e-06,
"loss": 2.5573,
"step": 392
},
{
"epoch": 0.3952976676934683,
"grad_norm": 19.846965607767764,
"learning_rate": 9.62184584061677e-06,
"loss": 2.5904,
"step": 393
},
{
"epoch": 0.39630351417614884,
"grad_norm": 18.299830732785676,
"learning_rate": 9.625939036363294e-06,
"loss": 2.5537,
"step": 394
},
{
"epoch": 0.39730936065882944,
"grad_norm": 18.937189256347306,
"learning_rate": 9.630021856444075e-06,
"loss": 2.5381,
"step": 395
},
{
"epoch": 0.39831520714151003,
"grad_norm": 15.860685957888077,
"learning_rate": 9.634094353327797e-06,
"loss": 2.525,
"step": 396
},
{
"epoch": 0.3993210536241906,
"grad_norm": 17.579941984547318,
"learning_rate": 9.638156579086155e-06,
"loss": 2.5499,
"step": 397
},
{
"epoch": 0.40032690010687116,
"grad_norm": 19.948100443223687,
"learning_rate": 9.64220858539785e-06,
"loss": 2.5226,
"step": 398
},
{
"epoch": 0.40133274658955176,
"grad_norm": 17.47822858548496,
"learning_rate": 9.646250423552533e-06,
"loss": 2.5239,
"step": 399
},
{
"epoch": 0.40233859307223235,
"grad_norm": 20.927008574523924,
"learning_rate": 9.650282144454697e-06,
"loss": 2.5573,
"step": 400
},
{
"epoch": 0.40334443955491295,
"grad_norm": 19.98864807675079,
"learning_rate": 9.654303798627532e-06,
"loss": 2.6445,
"step": 401
},
{
"epoch": 0.4043502860375935,
"grad_norm": 18.49557219174862,
"learning_rate": 9.658315436216716e-06,
"loss": 2.5924,
"step": 402
},
{
"epoch": 0.4053561325202741,
"grad_norm": 20.118177651660535,
"learning_rate": 9.662317106994168e-06,
"loss": 2.5399,
"step": 403
},
{
"epoch": 0.4063619790029547,
"grad_norm": 18.653675069722883,
"learning_rate": 9.666308860361762e-06,
"loss": 2.6146,
"step": 404
},
{
"epoch": 0.40736782548563527,
"grad_norm": 18.01802568279985,
"learning_rate": 9.670290745354967e-06,
"loss": 2.5572,
"step": 405
},
{
"epoch": 0.4083736719683158,
"grad_norm": 21.29046463917139,
"learning_rate": 9.674262810646488e-06,
"loss": 2.5068,
"step": 406
},
{
"epoch": 0.4093795184509964,
"grad_norm": 18.24276490157925,
"learning_rate": 9.678225104549809e-06,
"loss": 2.6318,
"step": 407
},
{
"epoch": 0.410385364933677,
"grad_norm": 21.252960660783703,
"learning_rate": 9.682177675022738e-06,
"loss": 2.5443,
"step": 408
},
{
"epoch": 0.4113912114163576,
"grad_norm": 18.70132923614444,
"learning_rate": 9.68612056967088e-06,
"loss": 2.5363,
"step": 409
},
{
"epoch": 0.41239705789903813,
"grad_norm": 16.132340101807507,
"learning_rate": 9.69005383575108e-06,
"loss": 2.5217,
"step": 410
},
{
"epoch": 0.4134029043817187,
"grad_norm": 21.346280277371964,
"learning_rate": 9.693977520174825e-06,
"loss": 2.5621,
"step": 411
},
{
"epoch": 0.4144087508643993,
"grad_norm": 16.940635635285602,
"learning_rate": 9.697891669511594e-06,
"loss": 2.5802,
"step": 412
},
{
"epoch": 0.4154145973470799,
"grad_norm": 19.26324991277955,
"learning_rate": 9.701796329992181e-06,
"loss": 2.533,
"step": 413
},
{
"epoch": 0.41642044382976046,
"grad_norm": 22.835729739016102,
"learning_rate": 9.705691547511965e-06,
"loss": 2.5402,
"step": 414
},
{
"epoch": 0.41742629031244105,
"grad_norm": 21.002625565526287,
"learning_rate": 9.709577367634156e-06,
"loss": 2.5658,
"step": 415
},
{
"epoch": 0.41843213679512165,
"grad_norm": 24.50852088816323,
"learning_rate": 9.713453835592982e-06,
"loss": 2.5428,
"step": 416
},
{
"epoch": 0.41943798327780224,
"grad_norm": 21.492809059871774,
"learning_rate": 9.717320996296857e-06,
"loss": 2.5388,
"step": 417
},
{
"epoch": 0.4204438297604828,
"grad_norm": 19.50678145510213,
"learning_rate": 9.721178894331493e-06,
"loss": 2.6011,
"step": 418
},
{
"epoch": 0.4214496762431634,
"grad_norm": 20.37546943411706,
"learning_rate": 9.725027573963e-06,
"loss": 2.5422,
"step": 419
},
{
"epoch": 0.42245552272584397,
"grad_norm": 19.437883948580463,
"learning_rate": 9.728867079140916e-06,
"loss": 2.557,
"step": 420
},
{
"epoch": 0.42346136920852456,
"grad_norm": 19.911077577378947,
"learning_rate": 9.732697453501233e-06,
"loss": 2.5821,
"step": 421
},
{
"epoch": 0.42446721569120516,
"grad_norm": 20.36935689145355,
"learning_rate": 9.736518740369361e-06,
"loss": 2.586,
"step": 422
},
{
"epoch": 0.4254730621738857,
"grad_norm": 17.840560424560636,
"learning_rate": 9.740330982763071e-06,
"loss": 2.6046,
"step": 423
},
{
"epoch": 0.4264789086565663,
"grad_norm": 19.87276506803033,
"learning_rate": 9.744134223395413e-06,
"loss": 2.5743,
"step": 424
},
{
"epoch": 0.4274847551392469,
"grad_norm": 17.6454935838629,
"learning_rate": 9.747928504677567e-06,
"loss": 2.5709,
"step": 425
},
{
"epoch": 0.4284906016219275,
"grad_norm": 17.696687945509535,
"learning_rate": 9.7517138687217e-06,
"loss": 2.6172,
"step": 426
},
{
"epoch": 0.429496448104608,
"grad_norm": 18.402222368633442,
"learning_rate": 9.75549035734375e-06,
"loss": 2.5621,
"step": 427
},
{
"epoch": 0.4305022945872886,
"grad_norm": 20.64172248569233,
"learning_rate": 9.759258012066223e-06,
"loss": 2.5456,
"step": 428
},
{
"epoch": 0.4315081410699692,
"grad_norm": 19.821918707901084,
"learning_rate": 9.76301687412091e-06,
"loss": 2.5792,
"step": 429
},
{
"epoch": 0.4325139875526498,
"grad_norm": 17.51180178485302,
"learning_rate": 9.766766984451605e-06,
"loss": 2.5254,
"step": 430
},
{
"epoch": 0.43351983403533034,
"grad_norm": 18.21037361364691,
"learning_rate": 9.770508383716777e-06,
"loss": 2.5228,
"step": 431
},
{
"epoch": 0.43452568051801094,
"grad_norm": 22.038929359068945,
"learning_rate": 9.774241112292223e-06,
"loss": 2.6225,
"step": 432
},
{
"epoch": 0.43553152700069153,
"grad_norm": 18.73170725911393,
"learning_rate": 9.777965210273664e-06,
"loss": 2.5835,
"step": 433
},
{
"epoch": 0.4365373734833721,
"grad_norm": 23.510705394520055,
"learning_rate": 9.781680717479356e-06,
"loss": 2.5343,
"step": 434
},
{
"epoch": 0.43754321996605267,
"grad_norm": 23.066316139714914,
"learning_rate": 9.785387673452622e-06,
"loss": 2.5647,
"step": 435
},
{
"epoch": 0.43854906644873326,
"grad_norm": 19.07782833604606,
"learning_rate": 9.789086117464384e-06,
"loss": 2.5615,
"step": 436
},
{
"epoch": 0.43955491293141385,
"grad_norm": 16.751538869640772,
"learning_rate": 9.792776088515663e-06,
"loss": 2.6065,
"step": 437
},
{
"epoch": 0.44056075941409445,
"grad_norm": 20.60303276845087,
"learning_rate": 9.796457625340045e-06,
"loss": 2.5397,
"step": 438
},
{
"epoch": 0.441566605896775,
"grad_norm": 18.45324982339736,
"learning_rate": 9.800130766406115e-06,
"loss": 2.6101,
"step": 439
},
{
"epoch": 0.4425724523794556,
"grad_norm": 20.238472969240057,
"learning_rate": 9.803795549919879e-06,
"loss": 2.5714,
"step": 440
},
{
"epoch": 0.4435782988621362,
"grad_norm": 21.366888626724553,
"learning_rate": 9.807452013827138e-06,
"loss": 2.5863,
"step": 441
},
{
"epoch": 0.4445841453448168,
"grad_norm": 23.870758183159623,
"learning_rate": 9.811100195815852e-06,
"loss": 2.5177,
"step": 442
},
{
"epoch": 0.4455899918274973,
"grad_norm": 18.602411032505366,
"learning_rate": 9.814740133318472e-06,
"loss": 2.5838,
"step": 443
},
{
"epoch": 0.4465958383101779,
"grad_norm": 21.016219269082764,
"learning_rate": 9.818371863514235e-06,
"loss": 2.6025,
"step": 444
},
{
"epoch": 0.4476016847928585,
"grad_norm": 17.77671667492051,
"learning_rate": 9.821995423331454e-06,
"loss": 2.5537,
"step": 445
},
{
"epoch": 0.4486075312755391,
"grad_norm": 20.10732838634908,
"learning_rate": 9.825610849449766e-06,
"loss": 2.5838,
"step": 446
},
{
"epoch": 0.44961337775821963,
"grad_norm": 19.0181000365809,
"learning_rate": 9.829218178302358e-06,
"loss": 2.4954,
"step": 447
},
{
"epoch": 0.45061922424090023,
"grad_norm": 19.33953885797201,
"learning_rate": 9.83281744607817e-06,
"loss": 2.5118,
"step": 448
},
{
"epoch": 0.4516250707235808,
"grad_norm": 19.85434731351852,
"learning_rate": 9.83640868872408e-06,
"loss": 2.6326,
"step": 449
},
{
"epoch": 0.4526309172062614,
"grad_norm": 22.249007362186386,
"learning_rate": 9.83999194194705e-06,
"loss": 2.5417,
"step": 450
},
{
"epoch": 0.45363676368894196,
"grad_norm": 22.301275249798355,
"learning_rate": 9.84356724121626e-06,
"loss": 2.6442,
"step": 451
},
{
"epoch": 0.45464261017162255,
"grad_norm": 20.462177665673316,
"learning_rate": 9.847134621765215e-06,
"loss": 2.5007,
"step": 452
},
{
"epoch": 0.45564845665430315,
"grad_norm": 22.380040927423096,
"learning_rate": 9.850694118593826e-06,
"loss": 2.5468,
"step": 453
},
{
"epoch": 0.45665430313698374,
"grad_norm": 19.651208291746624,
"learning_rate": 9.854245766470469e-06,
"loss": 2.5567,
"step": 454
},
{
"epoch": 0.4576601496196643,
"grad_norm": 22.186410502702007,
"learning_rate": 9.857789599934032e-06,
"loss": 2.4932,
"step": 455
},
{
"epoch": 0.4586659961023449,
"grad_norm": 22.549669207667595,
"learning_rate": 9.861325653295919e-06,
"loss": 2.5198,
"step": 456
},
{
"epoch": 0.45967184258502547,
"grad_norm": 18.811742186281815,
"learning_rate": 9.864853960642048e-06,
"loss": 2.5091,
"step": 457
},
{
"epoch": 0.46067768906770606,
"grad_norm": 22.424509982770353,
"learning_rate": 9.868374555834827e-06,
"loss": 2.4765,
"step": 458
},
{
"epoch": 0.4616835355503866,
"grad_norm": 22.679824568488968,
"learning_rate": 9.871887472515092e-06,
"loss": 2.5785,
"step": 459
},
{
"epoch": 0.4626893820330672,
"grad_norm": 21.332414608524278,
"learning_rate": 9.875392744104048e-06,
"loss": 2.5537,
"step": 460
},
{
"epoch": 0.4636952285157478,
"grad_norm": 21.78085628200312,
"learning_rate": 9.878890403805172e-06,
"loss": 2.5728,
"step": 461
},
{
"epoch": 0.4647010749984284,
"grad_norm": 22.86034758007963,
"learning_rate": 9.882380484606098e-06,
"loss": 2.6021,
"step": 462
},
{
"epoch": 0.4657069214811089,
"grad_norm": 22.35062353253106,
"learning_rate": 9.885863019280488e-06,
"loss": 2.5692,
"step": 463
},
{
"epoch": 0.4667127679637895,
"grad_norm": 20.634601317147265,
"learning_rate": 9.889338040389874e-06,
"loss": 2.5236,
"step": 464
},
{
"epoch": 0.4677186144464701,
"grad_norm": 17.922619313281967,
"learning_rate": 9.892805580285489e-06,
"loss": 2.5819,
"step": 465
},
{
"epoch": 0.4687244609291507,
"grad_norm": 21.482681336740185,
"learning_rate": 9.896265671110072e-06,
"loss": 2.5266,
"step": 466
},
{
"epoch": 0.46973030741183125,
"grad_norm": 17.049968410688997,
"learning_rate": 9.899718344799657e-06,
"loss": 2.566,
"step": 467
},
{
"epoch": 0.47073615389451184,
"grad_norm": 21.62677597122425,
"learning_rate": 9.903163633085336e-06,
"loss": 2.5515,
"step": 468
},
{
"epoch": 0.47174200037719244,
"grad_norm": 16.57841641405861,
"learning_rate": 9.906601567495018e-06,
"loss": 2.584,
"step": 469
},
{
"epoch": 0.47274784685987303,
"grad_norm": 18.926113122865047,
"learning_rate": 9.910032179355153e-06,
"loss": 2.4928,
"step": 470
},
{
"epoch": 0.4737536933425536,
"grad_norm": 19.220992348367954,
"learning_rate": 9.91345549979245e-06,
"loss": 2.5615,
"step": 471
},
{
"epoch": 0.47475953982523417,
"grad_norm": 18.388374619911023,
"learning_rate": 9.916871559735566e-06,
"loss": 2.6263,
"step": 472
},
{
"epoch": 0.47576538630791476,
"grad_norm": 20.662610639360757,
"learning_rate": 9.920280389916785e-06,
"loss": 2.5093,
"step": 473
},
{
"epoch": 0.47677123279059536,
"grad_norm": 19.96841955129098,
"learning_rate": 9.92368202087368e-06,
"loss": 2.5829,
"step": 474
},
{
"epoch": 0.4777770792732759,
"grad_norm": 19.651541678361166,
"learning_rate": 9.927076482950749e-06,
"loss": 2.5313,
"step": 475
},
{
"epoch": 0.4787829257559565,
"grad_norm": 20.768470047845234,
"learning_rate": 9.93046380630104e-06,
"loss": 2.5703,
"step": 476
},
{
"epoch": 0.4797887722386371,
"grad_norm": 19.380117148544503,
"learning_rate": 9.933844020887766e-06,
"loss": 2.5438,
"step": 477
},
{
"epoch": 0.4807946187213177,
"grad_norm": 17.478281054077645,
"learning_rate": 9.937217156485885e-06,
"loss": 2.5358,
"step": 478
},
{
"epoch": 0.4818004652039982,
"grad_norm": 16.628643607719937,
"learning_rate": 9.940583242683675e-06,
"loss": 2.5197,
"step": 479
},
{
"epoch": 0.4828063116866788,
"grad_norm": 21.77545239003663,
"learning_rate": 9.943942308884303e-06,
"loss": 2.5079,
"step": 480
},
{
"epoch": 0.4838121581693594,
"grad_norm": 21.324959520122704,
"learning_rate": 9.947294384307348e-06,
"loss": 2.4906,
"step": 481
},
{
"epoch": 0.48481800465204,
"grad_norm": 16.398520699639988,
"learning_rate": 9.950639497990342e-06,
"loss": 2.4824,
"step": 482
},
{
"epoch": 0.48582385113472054,
"grad_norm": 23.604836816022345,
"learning_rate": 9.953977678790266e-06,
"loss": 2.5571,
"step": 483
},
{
"epoch": 0.48682969761740114,
"grad_norm": 18.0895610543458,
"learning_rate": 9.95730895538506e-06,
"loss": 2.5574,
"step": 484
},
{
"epoch": 0.48783554410008173,
"grad_norm": 19.596954645717,
"learning_rate": 9.960633356275078e-06,
"loss": 2.5706,
"step": 485
},
{
"epoch": 0.4888413905827623,
"grad_norm": 17.69568825924647,
"learning_rate": 9.963950909784575e-06,
"loss": 2.5282,
"step": 486
},
{
"epoch": 0.48984723706544286,
"grad_norm": 18.315031047887874,
"learning_rate": 9.96726164406314e-06,
"loss": 2.5438,
"step": 487
},
{
"epoch": 0.49085308354812346,
"grad_norm": 21.065742691153716,
"learning_rate": 9.970565587087136e-06,
"loss": 2.5552,
"step": 488
},
{
"epoch": 0.49185893003080405,
"grad_norm": 17.188059594385642,
"learning_rate": 9.973862766661114e-06,
"loss": 2.5923,
"step": 489
},
{
"epoch": 0.49286477651348465,
"grad_norm": 21.1108953164887,
"learning_rate": 9.977153210419218e-06,
"loss": 2.6044,
"step": 490
},
{
"epoch": 0.4938706229961652,
"grad_norm": 21.36747102982352,
"learning_rate": 9.980436945826581e-06,
"loss": 2.5927,
"step": 491
},
{
"epoch": 0.4948764694788458,
"grad_norm": 19.868155776423375,
"learning_rate": 9.983714000180685e-06,
"loss": 2.5641,
"step": 492
},
{
"epoch": 0.4958823159615264,
"grad_norm": 22.741938569039043,
"learning_rate": 9.986984400612744e-06,
"loss": 2.602,
"step": 493
},
{
"epoch": 0.49688816244420697,
"grad_norm": 20.68752819403972,
"learning_rate": 9.990248174089033e-06,
"loss": 2.579,
"step": 494
},
{
"epoch": 0.4978940089268875,
"grad_norm": 20.135950916639047,
"learning_rate": 9.99350534741223e-06,
"loss": 2.552,
"step": 495
},
{
"epoch": 0.4988998554095681,
"grad_norm": 18.623416371913574,
"learning_rate": 9.996755947222743e-06,
"loss": 2.5241,
"step": 496
},
{
"epoch": 0.4999057018922487,
"grad_norm": 18.983653734389193,
"learning_rate": 1e-05,
"loss": 2.5411,
"step": 497
},
{
"epoch": 0.5009115483749292,
"grad_norm": 17.004668294054074,
"learning_rate": 1e-05,
"loss": 2.5481,
"step": 498
},
{
"epoch": 0.5019173948576099,
"grad_norm": 19.169938319334523,
"learning_rate": 9.997764363961547e-06,
"loss": 2.5168,
"step": 499
},
{
"epoch": 0.5029232413402904,
"grad_norm": 15.781569650968905,
"learning_rate": 9.995528727923096e-06,
"loss": 2.5549,
"step": 500
},
{
"epoch": 0.503929087822971,
"grad_norm": 20.179625887745996,
"learning_rate": 9.993293091884642e-06,
"loss": 2.4907,
"step": 501
},
{
"epoch": 0.5049349343056516,
"grad_norm": 19.05116376715222,
"learning_rate": 9.991057455846189e-06,
"loss": 2.5621,
"step": 502
},
{
"epoch": 0.5059407807883322,
"grad_norm": 21.91655117108733,
"learning_rate": 9.988821819807735e-06,
"loss": 2.5457,
"step": 503
},
{
"epoch": 0.5069466272710128,
"grad_norm": 20.393552472807908,
"learning_rate": 9.986586183769284e-06,
"loss": 2.5379,
"step": 504
},
{
"epoch": 0.5079524737536933,
"grad_norm": 22.422384427517496,
"learning_rate": 9.98435054773083e-06,
"loss": 2.5373,
"step": 505
},
{
"epoch": 0.5089583202363739,
"grad_norm": 15.409773180659693,
"learning_rate": 9.982114911692378e-06,
"loss": 2.4445,
"step": 506
},
{
"epoch": 0.5099641667190545,
"grad_norm": 21.378407108104337,
"learning_rate": 9.979879275653925e-06,
"loss": 2.4991,
"step": 507
},
{
"epoch": 0.5109700132017351,
"grad_norm": 19.128162487040523,
"learning_rate": 9.977643639615471e-06,
"loss": 2.4958,
"step": 508
},
{
"epoch": 0.5119758596844156,
"grad_norm": 19.690175531171707,
"learning_rate": 9.975408003577018e-06,
"loss": 2.5555,
"step": 509
},
{
"epoch": 0.5129817061670963,
"grad_norm": 20.712793838884973,
"learning_rate": 9.973172367538565e-06,
"loss": 2.5536,
"step": 510
},
{
"epoch": 0.5139875526497768,
"grad_norm": 22.092179923348358,
"learning_rate": 9.970936731500113e-06,
"loss": 2.5978,
"step": 511
},
{
"epoch": 0.5149933991324575,
"grad_norm": 18.84515345278468,
"learning_rate": 9.96870109546166e-06,
"loss": 2.5589,
"step": 512
},
{
"epoch": 0.515999245615138,
"grad_norm": 19.568303464632645,
"learning_rate": 9.966465459423208e-06,
"loss": 2.5411,
"step": 513
},
{
"epoch": 0.5170050920978185,
"grad_norm": 21.2416135192592,
"learning_rate": 9.964229823384754e-06,
"loss": 2.5108,
"step": 514
},
{
"epoch": 0.5180109385804992,
"grad_norm": 18.574355716211667,
"learning_rate": 9.9619941873463e-06,
"loss": 2.5255,
"step": 515
},
{
"epoch": 0.5190167850631797,
"grad_norm": 25.877119413501973,
"learning_rate": 9.959758551307847e-06,
"loss": 2.5063,
"step": 516
},
{
"epoch": 0.5200226315458603,
"grad_norm": 21.502468407593128,
"learning_rate": 9.957522915269396e-06,
"loss": 2.5912,
"step": 517
},
{
"epoch": 0.5210284780285409,
"grad_norm": 24.69392944829647,
"learning_rate": 9.955287279230942e-06,
"loss": 2.567,
"step": 518
},
{
"epoch": 0.5220343245112214,
"grad_norm": 23.76268825844637,
"learning_rate": 9.953051643192489e-06,
"loss": 2.4931,
"step": 519
},
{
"epoch": 0.5230401709939021,
"grad_norm": 20.684631513248082,
"learning_rate": 9.950816007154035e-06,
"loss": 2.5733,
"step": 520
},
{
"epoch": 0.5240460174765826,
"grad_norm": 19.227634049258327,
"learning_rate": 9.948580371115584e-06,
"loss": 2.552,
"step": 521
},
{
"epoch": 0.5250518639592632,
"grad_norm": 24.53885448185181,
"learning_rate": 9.94634473507713e-06,
"loss": 2.5831,
"step": 522
},
{
"epoch": 0.5260577104419438,
"grad_norm": 23.124577154789915,
"learning_rate": 9.944109099038677e-06,
"loss": 2.5542,
"step": 523
},
{
"epoch": 0.5270635569246244,
"grad_norm": 21.424296302635216,
"learning_rate": 9.941873463000225e-06,
"loss": 2.5702,
"step": 524
},
{
"epoch": 0.5280694034073049,
"grad_norm": 21.23884958735581,
"learning_rate": 9.939637826961771e-06,
"loss": 2.4847,
"step": 525
},
{
"epoch": 0.5290752498899856,
"grad_norm": 24.491387720457944,
"learning_rate": 9.937402190923318e-06,
"loss": 2.5822,
"step": 526
},
{
"epoch": 0.5300810963726661,
"grad_norm": 19.874916575754323,
"learning_rate": 9.935166554884865e-06,
"loss": 2.5269,
"step": 527
},
{
"epoch": 0.5310869428553467,
"grad_norm": 60.71635494912012,
"learning_rate": 9.932930918846413e-06,
"loss": 2.5605,
"step": 528
},
{
"epoch": 0.5320927893380273,
"grad_norm": 20.456678607924925,
"learning_rate": 9.93069528280796e-06,
"loss": 2.5238,
"step": 529
},
{
"epoch": 0.5330986358207078,
"grad_norm": 18.136503927465057,
"learning_rate": 9.928459646769508e-06,
"loss": 2.5667,
"step": 530
},
{
"epoch": 0.5341044823033885,
"grad_norm": 16.289075770430706,
"learning_rate": 9.926224010731054e-06,
"loss": 2.5032,
"step": 531
},
{
"epoch": 0.535110328786069,
"grad_norm": 18.52577477934986,
"learning_rate": 9.9239883746926e-06,
"loss": 2.5708,
"step": 532
},
{
"epoch": 0.5361161752687497,
"grad_norm": 18.712980125548064,
"learning_rate": 9.921752738654147e-06,
"loss": 2.5339,
"step": 533
},
{
"epoch": 0.5371220217514302,
"grad_norm": 19.748115495810836,
"learning_rate": 9.919517102615694e-06,
"loss": 2.5236,
"step": 534
},
{
"epoch": 0.5381278682341107,
"grad_norm": 16.510969022398314,
"learning_rate": 9.917281466577242e-06,
"loss": 2.5847,
"step": 535
},
{
"epoch": 0.5391337147167914,
"grad_norm": 18.080202479501924,
"learning_rate": 9.915045830538789e-06,
"loss": 2.5728,
"step": 536
},
{
"epoch": 0.5401395611994719,
"grad_norm": 18.47492461965452,
"learning_rate": 9.912810194500337e-06,
"loss": 2.5226,
"step": 537
},
{
"epoch": 0.5411454076821525,
"grad_norm": 18.114656650302823,
"learning_rate": 9.910574558461884e-06,
"loss": 2.5199,
"step": 538
},
{
"epoch": 0.5421512541648331,
"grad_norm": 16.58075410612739,
"learning_rate": 9.90833892242343e-06,
"loss": 2.5481,
"step": 539
},
{
"epoch": 0.5431571006475137,
"grad_norm": 29.273276472080056,
"learning_rate": 9.906103286384977e-06,
"loss": 2.4944,
"step": 540
},
{
"epoch": 0.5441629471301943,
"grad_norm": 18.462379052260196,
"learning_rate": 9.903867650346525e-06,
"loss": 2.5807,
"step": 541
},
{
"epoch": 0.5451687936128748,
"grad_norm": 17.13630248386338,
"learning_rate": 9.901632014308071e-06,
"loss": 2.5622,
"step": 542
},
{
"epoch": 0.5461746400955554,
"grad_norm": 19.468397745227442,
"learning_rate": 9.89939637826962e-06,
"loss": 2.4879,
"step": 543
},
{
"epoch": 0.547180486578236,
"grad_norm": 17.0399957247732,
"learning_rate": 9.897160742231166e-06,
"loss": 2.5107,
"step": 544
},
{
"epoch": 0.5481863330609166,
"grad_norm": 18.31755981883234,
"learning_rate": 9.894925106192713e-06,
"loss": 2.5205,
"step": 545
},
{
"epoch": 0.5491921795435971,
"grad_norm": 17.749662710789632,
"learning_rate": 9.89268947015426e-06,
"loss": 2.5464,
"step": 546
},
{
"epoch": 0.5501980260262778,
"grad_norm": 19.001762577300585,
"learning_rate": 9.890453834115806e-06,
"loss": 2.549,
"step": 547
},
{
"epoch": 0.5512038725089583,
"grad_norm": 16.48812300958239,
"learning_rate": 9.888218198077354e-06,
"loss": 2.5059,
"step": 548
},
{
"epoch": 0.552209718991639,
"grad_norm": 16.712686374839436,
"learning_rate": 9.8859825620389e-06,
"loss": 2.5445,
"step": 549
},
{
"epoch": 0.5532155654743195,
"grad_norm": 17.463415610893698,
"learning_rate": 9.883746926000447e-06,
"loss": 2.5056,
"step": 550
},
{
"epoch": 0.554221411957,
"grad_norm": 17.741113008733752,
"learning_rate": 9.881511289961994e-06,
"loss": 2.5551,
"step": 551
},
{
"epoch": 0.5552272584396807,
"grad_norm": 16.464918946132286,
"learning_rate": 9.879275653923542e-06,
"loss": 2.5507,
"step": 552
},
{
"epoch": 0.5562331049223612,
"grad_norm": 17.12442892090824,
"learning_rate": 9.877040017885089e-06,
"loss": 2.5451,
"step": 553
},
{
"epoch": 0.5572389514050418,
"grad_norm": 16.345309381319424,
"learning_rate": 9.874804381846637e-06,
"loss": 2.4611,
"step": 554
},
{
"epoch": 0.5582447978877224,
"grad_norm": 18.656826976907546,
"learning_rate": 9.872568745808184e-06,
"loss": 2.5097,
"step": 555
},
{
"epoch": 0.559250644370403,
"grad_norm": 15.956055864098987,
"learning_rate": 9.87033310976973e-06,
"loss": 2.5246,
"step": 556
},
{
"epoch": 0.5602564908530836,
"grad_norm": 18.909338459825516,
"learning_rate": 9.868097473731277e-06,
"loss": 2.5912,
"step": 557
},
{
"epoch": 0.5612623373357641,
"grad_norm": 18.11236716491806,
"learning_rate": 9.865861837692823e-06,
"loss": 2.5275,
"step": 558
},
{
"epoch": 0.5622681838184447,
"grad_norm": 19.6612038201225,
"learning_rate": 9.863626201654371e-06,
"loss": 2.5034,
"step": 559
},
{
"epoch": 0.5632740303011253,
"grad_norm": 22.041905995039123,
"learning_rate": 9.861390565615918e-06,
"loss": 2.5134,
"step": 560
},
{
"epoch": 0.5642798767838059,
"grad_norm": 18.943489012480768,
"learning_rate": 9.859154929577466e-06,
"loss": 2.5176,
"step": 561
},
{
"epoch": 0.5652857232664864,
"grad_norm": 16.995644349198606,
"learning_rate": 9.856919293539013e-06,
"loss": 2.5364,
"step": 562
},
{
"epoch": 0.5662915697491671,
"grad_norm": 18.390254920919176,
"learning_rate": 9.85468365750056e-06,
"loss": 2.5314,
"step": 563
},
{
"epoch": 0.5672974162318476,
"grad_norm": 16.152905937236415,
"learning_rate": 9.852448021462106e-06,
"loss": 2.5169,
"step": 564
},
{
"epoch": 0.5683032627145282,
"grad_norm": 18.48631801217191,
"learning_rate": 9.850212385423654e-06,
"loss": 2.5293,
"step": 565
},
{
"epoch": 0.5693091091972088,
"grad_norm": 17.715384087930207,
"learning_rate": 9.8479767493852e-06,
"loss": 2.5817,
"step": 566
},
{
"epoch": 0.5703149556798893,
"grad_norm": 20.06744852384933,
"learning_rate": 9.845741113346749e-06,
"loss": 2.5676,
"step": 567
},
{
"epoch": 0.57132080216257,
"grad_norm": 19.29605365339498,
"learning_rate": 9.843505477308296e-06,
"loss": 2.5231,
"step": 568
},
{
"epoch": 0.5723266486452505,
"grad_norm": 17.150675872187414,
"learning_rate": 9.841269841269842e-06,
"loss": 2.5448,
"step": 569
},
{
"epoch": 0.573332495127931,
"grad_norm": 16.770426880970362,
"learning_rate": 9.839034205231389e-06,
"loss": 2.5418,
"step": 570
},
{
"epoch": 0.5743383416106117,
"grad_norm": 18.415349918817537,
"learning_rate": 9.836798569192935e-06,
"loss": 2.5525,
"step": 571
},
{
"epoch": 0.5753441880932922,
"grad_norm": 17.722420126001705,
"learning_rate": 9.834562933154484e-06,
"loss": 2.5643,
"step": 572
},
{
"epoch": 0.5763500345759729,
"grad_norm": 17.293290429304342,
"learning_rate": 9.83232729711603e-06,
"loss": 2.5361,
"step": 573
},
{
"epoch": 0.5773558810586534,
"grad_norm": 61.680906816389644,
"learning_rate": 9.830091661077578e-06,
"loss": 2.5717,
"step": 574
},
{
"epoch": 0.578361727541334,
"grad_norm": 22.24029178280409,
"learning_rate": 9.827856025039125e-06,
"loss": 2.5664,
"step": 575
},
{
"epoch": 0.5793675740240146,
"grad_norm": 19.513909304611108,
"learning_rate": 9.825620389000671e-06,
"loss": 2.48,
"step": 576
},
{
"epoch": 0.5803734205066952,
"grad_norm": 19.514919116226682,
"learning_rate": 9.823384752962218e-06,
"loss": 2.5554,
"step": 577
},
{
"epoch": 0.5813792669893757,
"grad_norm": 19.359797198083314,
"learning_rate": 9.821149116923766e-06,
"loss": 2.5304,
"step": 578
},
{
"epoch": 0.5823851134720563,
"grad_norm": 17.480705109387433,
"learning_rate": 9.818913480885313e-06,
"loss": 2.5691,
"step": 579
},
{
"epoch": 0.5833909599547369,
"grad_norm": 17.938688964369206,
"learning_rate": 9.816677844846861e-06,
"loss": 2.5095,
"step": 580
},
{
"epoch": 0.5843968064374175,
"grad_norm": 20.7902045613152,
"learning_rate": 9.814442208808408e-06,
"loss": 2.5317,
"step": 581
},
{
"epoch": 0.5854026529200981,
"grad_norm": 16.80739441378905,
"learning_rate": 9.812206572769954e-06,
"loss": 2.5412,
"step": 582
},
{
"epoch": 0.5864084994027786,
"grad_norm": 19.170545572301233,
"learning_rate": 9.8099709367315e-06,
"loss": 2.5462,
"step": 583
},
{
"epoch": 0.5874143458854593,
"grad_norm": 17.819113412411937,
"learning_rate": 9.807735300693047e-06,
"loss": 2.5127,
"step": 584
},
{
"epoch": 0.5884201923681398,
"grad_norm": 22.38069113233498,
"learning_rate": 9.805499664654596e-06,
"loss": 2.5908,
"step": 585
},
{
"epoch": 0.5894260388508203,
"grad_norm": 18.672856432624258,
"learning_rate": 9.803264028616142e-06,
"loss": 2.5348,
"step": 586
},
{
"epoch": 0.590431885333501,
"grad_norm": 18.638852489711283,
"learning_rate": 9.801028392577689e-06,
"loss": 2.5037,
"step": 587
},
{
"epoch": 0.5914377318161815,
"grad_norm": 19.540961306703963,
"learning_rate": 9.798792756539235e-06,
"loss": 2.5536,
"step": 588
},
{
"epoch": 0.5924435782988622,
"grad_norm": 19.04026220736335,
"learning_rate": 9.796557120500783e-06,
"loss": 2.5337,
"step": 589
},
{
"epoch": 0.5934494247815427,
"grad_norm": 17.93807746202833,
"learning_rate": 9.79432148446233e-06,
"loss": 2.4845,
"step": 590
},
{
"epoch": 0.5944552712642233,
"grad_norm": 19.252745829909692,
"learning_rate": 9.792085848423878e-06,
"loss": 2.5767,
"step": 591
},
{
"epoch": 0.5954611177469039,
"grad_norm": 20.02025664672543,
"learning_rate": 9.789850212385425e-06,
"loss": 2.5426,
"step": 592
},
{
"epoch": 0.5964669642295845,
"grad_norm": 18.93799725756257,
"learning_rate": 9.787614576346971e-06,
"loss": 2.4909,
"step": 593
},
{
"epoch": 0.597472810712265,
"grad_norm": 17.119830462848988,
"learning_rate": 9.785378940308518e-06,
"loss": 2.5408,
"step": 594
},
{
"epoch": 0.5984786571949456,
"grad_norm": 19.594325441428992,
"learning_rate": 9.783143304270065e-06,
"loss": 2.5078,
"step": 595
},
{
"epoch": 0.5994845036776262,
"grad_norm": 17.676742045845327,
"learning_rate": 9.780907668231613e-06,
"loss": 2.554,
"step": 596
},
{
"epoch": 0.6004903501603068,
"grad_norm": 18.41536510551728,
"learning_rate": 9.77867203219316e-06,
"loss": 2.5327,
"step": 597
},
{
"epoch": 0.6014961966429874,
"grad_norm": 18.51209433223085,
"learning_rate": 9.776436396154708e-06,
"loss": 2.54,
"step": 598
},
{
"epoch": 0.6025020431256679,
"grad_norm": 16.521318221902543,
"learning_rate": 9.774200760116254e-06,
"loss": 2.569,
"step": 599
},
{
"epoch": 0.6035078896083486,
"grad_norm": 16.079590282276442,
"learning_rate": 9.7719651240778e-06,
"loss": 2.5368,
"step": 600
},
{
"epoch": 0.6045137360910291,
"grad_norm": 20.23619630960961,
"learning_rate": 9.769729488039347e-06,
"loss": 2.5224,
"step": 601
},
{
"epoch": 0.6055195825737096,
"grad_norm": 17.54702826113766,
"learning_rate": 9.767493852000896e-06,
"loss": 2.5813,
"step": 602
},
{
"epoch": 0.6065254290563903,
"grad_norm": 19.13355798517692,
"learning_rate": 9.765258215962442e-06,
"loss": 2.5462,
"step": 603
},
{
"epoch": 0.6075312755390708,
"grad_norm": 16.431191732101087,
"learning_rate": 9.76302257992399e-06,
"loss": 2.5439,
"step": 604
},
{
"epoch": 0.6085371220217515,
"grad_norm": 18.408948713655214,
"learning_rate": 9.760786943885537e-06,
"loss": 2.5589,
"step": 605
},
{
"epoch": 0.609542968504432,
"grad_norm": 17.712239889116066,
"learning_rate": 9.758551307847083e-06,
"loss": 2.476,
"step": 606
},
{
"epoch": 0.6105488149871126,
"grad_norm": 20.013392068620888,
"learning_rate": 9.75631567180863e-06,
"loss": 2.5947,
"step": 607
},
{
"epoch": 0.6115546614697932,
"grad_norm": 20.423121099411897,
"learning_rate": 9.754080035770177e-06,
"loss": 2.611,
"step": 608
},
{
"epoch": 0.6125605079524737,
"grad_norm": 20.404380298452843,
"learning_rate": 9.751844399731725e-06,
"loss": 2.5225,
"step": 609
},
{
"epoch": 0.6135663544351543,
"grad_norm": 23.64833980844967,
"learning_rate": 9.749608763693271e-06,
"loss": 2.4801,
"step": 610
},
{
"epoch": 0.6145722009178349,
"grad_norm": 22.721372237492865,
"learning_rate": 9.74737312765482e-06,
"loss": 2.5072,
"step": 611
},
{
"epoch": 0.6155780474005155,
"grad_norm": 21.485471706831824,
"learning_rate": 9.745137491616366e-06,
"loss": 2.4746,
"step": 612
},
{
"epoch": 0.6165838938831961,
"grad_norm": 18.77121558736787,
"learning_rate": 9.742901855577913e-06,
"loss": 2.5336,
"step": 613
},
{
"epoch": 0.6175897403658767,
"grad_norm": 19.085989559745865,
"learning_rate": 9.74066621953946e-06,
"loss": 2.5171,
"step": 614
},
{
"epoch": 0.6185955868485572,
"grad_norm": 20.870416879493348,
"learning_rate": 9.738430583501008e-06,
"loss": 2.5151,
"step": 615
},
{
"epoch": 0.6196014333312378,
"grad_norm": 21.14090139438579,
"learning_rate": 9.736194947462554e-06,
"loss": 2.5166,
"step": 616
},
{
"epoch": 0.6206072798139184,
"grad_norm": 19.01841862253812,
"learning_rate": 9.7339593114241e-06,
"loss": 2.5681,
"step": 617
},
{
"epoch": 0.6216131262965989,
"grad_norm": 21.22073856960854,
"learning_rate": 9.731723675385647e-06,
"loss": 2.5043,
"step": 618
},
{
"epoch": 0.6226189727792796,
"grad_norm": 17.725436424684155,
"learning_rate": 9.729488039347194e-06,
"loss": 2.5145,
"step": 619
},
{
"epoch": 0.6236248192619601,
"grad_norm": 20.751400818923063,
"learning_rate": 9.727252403308742e-06,
"loss": 2.4691,
"step": 620
},
{
"epoch": 0.6246306657446408,
"grad_norm": 22.573815319781055,
"learning_rate": 9.725016767270289e-06,
"loss": 2.562,
"step": 621
},
{
"epoch": 0.6256365122273213,
"grad_norm": 17.987265540277342,
"learning_rate": 9.722781131231837e-06,
"loss": 2.5302,
"step": 622
},
{
"epoch": 0.6266423587100018,
"grad_norm": 17.03654847709092,
"learning_rate": 9.720545495193383e-06,
"loss": 2.4943,
"step": 623
},
{
"epoch": 0.6276482051926825,
"grad_norm": 21.47031641830582,
"learning_rate": 9.71830985915493e-06,
"loss": 2.5616,
"step": 624
},
{
"epoch": 0.628654051675363,
"grad_norm": 20.301768991814786,
"learning_rate": 9.716074223116477e-06,
"loss": 2.5403,
"step": 625
},
{
"epoch": 0.6296598981580436,
"grad_norm": 20.183945617233114,
"learning_rate": 9.713838587078025e-06,
"loss": 2.4984,
"step": 626
},
{
"epoch": 0.6306657446407242,
"grad_norm": 18.087166072771478,
"learning_rate": 9.711602951039571e-06,
"loss": 2.5457,
"step": 627
},
{
"epoch": 0.6316715911234048,
"grad_norm": 18.497985162549405,
"learning_rate": 9.709367315001118e-06,
"loss": 2.5379,
"step": 628
},
{
"epoch": 0.6326774376060854,
"grad_norm": 25.906471020026725,
"learning_rate": 9.707131678962666e-06,
"loss": 2.5112,
"step": 629
},
{
"epoch": 0.633683284088766,
"grad_norm": 18.851737821574062,
"learning_rate": 9.704896042924213e-06,
"loss": 2.5183,
"step": 630
},
{
"epoch": 0.6346891305714465,
"grad_norm": 19.34370501708874,
"learning_rate": 9.70266040688576e-06,
"loss": 2.528,
"step": 631
},
{
"epoch": 0.6356949770541271,
"grad_norm": 17.46671844948673,
"learning_rate": 9.700424770847306e-06,
"loss": 2.4713,
"step": 632
},
{
"epoch": 0.6367008235368077,
"grad_norm": 15.857669621477335,
"learning_rate": 9.698189134808854e-06,
"loss": 2.5421,
"step": 633
},
{
"epoch": 0.6377066700194883,
"grad_norm": 20.690235507938166,
"learning_rate": 9.6959534987704e-06,
"loss": 2.5477,
"step": 634
},
{
"epoch": 0.6387125165021689,
"grad_norm": 17.75939120748436,
"learning_rate": 9.693717862731949e-06,
"loss": 2.4967,
"step": 635
},
{
"epoch": 0.6397183629848494,
"grad_norm": 20.86052880385186,
"learning_rate": 9.691482226693496e-06,
"loss": 2.5144,
"step": 636
},
{
"epoch": 0.6407242094675301,
"grad_norm": 21.504493399150405,
"learning_rate": 9.689246590655042e-06,
"loss": 2.4755,
"step": 637
},
{
"epoch": 0.6417300559502106,
"grad_norm": 19.2512677338903,
"learning_rate": 9.687010954616589e-06,
"loss": 2.5303,
"step": 638
},
{
"epoch": 0.6427359024328911,
"grad_norm": 20.932342218853474,
"learning_rate": 9.684775318578137e-06,
"loss": 2.4933,
"step": 639
},
{
"epoch": 0.6437417489155718,
"grad_norm": 21.530735790573438,
"learning_rate": 9.682539682539683e-06,
"loss": 2.5472,
"step": 640
},
{
"epoch": 0.6447475953982523,
"grad_norm": 19.3718729621729,
"learning_rate": 9.68030404650123e-06,
"loss": 2.5117,
"step": 641
},
{
"epoch": 0.645753441880933,
"grad_norm": 25.819881974982856,
"learning_rate": 9.678068410462778e-06,
"loss": 2.4887,
"step": 642
},
{
"epoch": 0.6467592883636135,
"grad_norm": 19.999886193848916,
"learning_rate": 9.675832774424325e-06,
"loss": 2.5082,
"step": 643
},
{
"epoch": 0.647765134846294,
"grad_norm": 21.357372449932193,
"learning_rate": 9.673597138385871e-06,
"loss": 2.4743,
"step": 644
},
{
"epoch": 0.6487709813289747,
"grad_norm": 24.86568125849384,
"learning_rate": 9.671361502347418e-06,
"loss": 2.5181,
"step": 645
},
{
"epoch": 0.6497768278116552,
"grad_norm": 19.99327736605139,
"learning_rate": 9.669125866308966e-06,
"loss": 2.5238,
"step": 646
},
{
"epoch": 0.6507826742943358,
"grad_norm": 22.801398555696743,
"learning_rate": 9.666890230270513e-06,
"loss": 2.5067,
"step": 647
},
{
"epoch": 0.6517885207770164,
"grad_norm": 19.838552809919648,
"learning_rate": 9.664654594232061e-06,
"loss": 2.4869,
"step": 648
},
{
"epoch": 0.652794367259697,
"grad_norm": 17.032352556256118,
"learning_rate": 9.662418958193608e-06,
"loss": 2.5488,
"step": 649
},
{
"epoch": 0.6538002137423776,
"grad_norm": 22.057043086256787,
"learning_rate": 9.660183322155154e-06,
"loss": 2.5637,
"step": 650
},
{
"epoch": 0.6548060602250582,
"grad_norm": 19.929726327790743,
"learning_rate": 9.6579476861167e-06,
"loss": 2.5389,
"step": 651
},
{
"epoch": 0.6558119067077387,
"grad_norm": 18.58065737479155,
"learning_rate": 9.655712050078247e-06,
"loss": 2.4967,
"step": 652
},
{
"epoch": 0.6568177531904194,
"grad_norm": 17.549376553583564,
"learning_rate": 9.653476414039795e-06,
"loss": 2.476,
"step": 653
},
{
"epoch": 0.6578235996730999,
"grad_norm": 20.240240936791928,
"learning_rate": 9.651240778001342e-06,
"loss": 2.5597,
"step": 654
},
{
"epoch": 0.6588294461557804,
"grad_norm": 21.625962244270248,
"learning_rate": 9.649005141962889e-06,
"loss": 2.4762,
"step": 655
},
{
"epoch": 0.6598352926384611,
"grad_norm": 21.173956358162684,
"learning_rate": 9.646769505924435e-06,
"loss": 2.5553,
"step": 656
},
{
"epoch": 0.6608411391211416,
"grad_norm": 18.05355685245098,
"learning_rate": 9.644533869885983e-06,
"loss": 2.5135,
"step": 657
},
{
"epoch": 0.6618469856038223,
"grad_norm": 21.37207774541577,
"learning_rate": 9.64229823384753e-06,
"loss": 2.5101,
"step": 658
},
{
"epoch": 0.6628528320865028,
"grad_norm": 16.35779599381775,
"learning_rate": 9.640062597809078e-06,
"loss": 2.4552,
"step": 659
},
{
"epoch": 0.6638586785691833,
"grad_norm": 17.570702404376508,
"learning_rate": 9.637826961770625e-06,
"loss": 2.5121,
"step": 660
},
{
"epoch": 0.664864525051864,
"grad_norm": 18.06740688331599,
"learning_rate": 9.635591325732171e-06,
"loss": 2.4734,
"step": 661
},
{
"epoch": 0.6658703715345445,
"grad_norm": 17.64673278928389,
"learning_rate": 9.633355689693718e-06,
"loss": 2.5011,
"step": 662
},
{
"epoch": 0.6668762180172251,
"grad_norm": 16.843625112610805,
"learning_rate": 9.631120053655266e-06,
"loss": 2.5188,
"step": 663
},
{
"epoch": 0.6678820644999057,
"grad_norm": 16.859967248687216,
"learning_rate": 9.628884417616813e-06,
"loss": 2.5267,
"step": 664
},
{
"epoch": 0.6688879109825863,
"grad_norm": 17.160716469928374,
"learning_rate": 9.62664878157836e-06,
"loss": 2.5312,
"step": 665
},
{
"epoch": 0.6698937574652669,
"grad_norm": 18.81047677543676,
"learning_rate": 9.624413145539908e-06,
"loss": 2.514,
"step": 666
},
{
"epoch": 0.6708996039479475,
"grad_norm": 17.03092392882562,
"learning_rate": 9.622177509501454e-06,
"loss": 2.5182,
"step": 667
},
{
"epoch": 0.671905450430628,
"grad_norm": 22.031292359972962,
"learning_rate": 9.619941873463e-06,
"loss": 2.4933,
"step": 668
},
{
"epoch": 0.6729112969133086,
"grad_norm": 15.894396840819647,
"learning_rate": 9.617706237424547e-06,
"loss": 2.4675,
"step": 669
},
{
"epoch": 0.6739171433959892,
"grad_norm": 16.78761067447768,
"learning_rate": 9.615470601386095e-06,
"loss": 2.5581,
"step": 670
},
{
"epoch": 0.6749229898786697,
"grad_norm": 17.818343229735852,
"learning_rate": 9.613234965347642e-06,
"loss": 2.5065,
"step": 671
},
{
"epoch": 0.6759288363613504,
"grad_norm": 17.871155303653666,
"learning_rate": 9.61099932930919e-06,
"loss": 2.511,
"step": 672
},
{
"epoch": 0.6769346828440309,
"grad_norm": 19.612335334446225,
"learning_rate": 9.608763693270737e-06,
"loss": 2.5465,
"step": 673
},
{
"epoch": 0.6779405293267116,
"grad_norm": 20.97735200183468,
"learning_rate": 9.606528057232283e-06,
"loss": 2.5658,
"step": 674
},
{
"epoch": 0.6789463758093921,
"grad_norm": 23.948339016531275,
"learning_rate": 9.60429242119383e-06,
"loss": 2.5125,
"step": 675
},
{
"epoch": 0.6799522222920726,
"grad_norm": 20.0467714662731,
"learning_rate": 9.602056785155377e-06,
"loss": 2.5191,
"step": 676
},
{
"epoch": 0.6809580687747533,
"grad_norm": 19.760663718030475,
"learning_rate": 9.599821149116925e-06,
"loss": 2.5257,
"step": 677
},
{
"epoch": 0.6819639152574338,
"grad_norm": 17.66847669368681,
"learning_rate": 9.597585513078471e-06,
"loss": 2.4735,
"step": 678
},
{
"epoch": 0.6829697617401144,
"grad_norm": 20.304213959727797,
"learning_rate": 9.59534987704002e-06,
"loss": 2.5135,
"step": 679
},
{
"epoch": 0.683975608222795,
"grad_norm": 16.874889043585817,
"learning_rate": 9.593114241001566e-06,
"loss": 2.5537,
"step": 680
},
{
"epoch": 0.6849814547054756,
"grad_norm": 17.391916117706774,
"learning_rate": 9.590878604963113e-06,
"loss": 2.5516,
"step": 681
},
{
"epoch": 0.6859873011881562,
"grad_norm": 16.97825687844495,
"learning_rate": 9.58864296892466e-06,
"loss": 2.5205,
"step": 682
},
{
"epoch": 0.6869931476708367,
"grad_norm": 17.746056983891144,
"learning_rate": 9.586407332886208e-06,
"loss": 2.4779,
"step": 683
},
{
"epoch": 0.6879989941535173,
"grad_norm": 21.91146086074379,
"learning_rate": 9.584171696847754e-06,
"loss": 2.5713,
"step": 684
},
{
"epoch": 0.6890048406361979,
"grad_norm": 22.158541573895086,
"learning_rate": 9.581936060809302e-06,
"loss": 2.4892,
"step": 685
},
{
"epoch": 0.6900106871188785,
"grad_norm": 17.263290599415026,
"learning_rate": 9.579700424770847e-06,
"loss": 2.4782,
"step": 686
},
{
"epoch": 0.691016533601559,
"grad_norm": 18.949281510467618,
"learning_rate": 9.577464788732394e-06,
"loss": 2.4675,
"step": 687
},
{
"epoch": 0.6920223800842397,
"grad_norm": 21.84072077538254,
"learning_rate": 9.575229152693942e-06,
"loss": 2.5219,
"step": 688
},
{
"epoch": 0.6930282265669202,
"grad_norm": 22.45437820067143,
"learning_rate": 9.572993516655489e-06,
"loss": 2.5034,
"step": 689
},
{
"epoch": 0.6940340730496009,
"grad_norm": 18.766091949144645,
"learning_rate": 9.570757880617037e-06,
"loss": 2.5564,
"step": 690
},
{
"epoch": 0.6950399195322814,
"grad_norm": 21.409093756509865,
"learning_rate": 9.568522244578583e-06,
"loss": 2.537,
"step": 691
},
{
"epoch": 0.6960457660149619,
"grad_norm": 24.96254769951716,
"learning_rate": 9.56628660854013e-06,
"loss": 2.5363,
"step": 692
},
{
"epoch": 0.6970516124976426,
"grad_norm": 18.60658988302085,
"learning_rate": 9.564050972501677e-06,
"loss": 2.5549,
"step": 693
},
{
"epoch": 0.6980574589803231,
"grad_norm": 18.613218112954755,
"learning_rate": 9.561815336463225e-06,
"loss": 2.4987,
"step": 694
},
{
"epoch": 0.6990633054630037,
"grad_norm": 22.66026247971019,
"learning_rate": 9.559579700424771e-06,
"loss": 2.5338,
"step": 695
},
{
"epoch": 0.7000691519456843,
"grad_norm": 17.804777772482662,
"learning_rate": 9.55734406438632e-06,
"loss": 2.5652,
"step": 696
},
{
"epoch": 0.7010749984283648,
"grad_norm": 19.80514523411522,
"learning_rate": 9.555108428347866e-06,
"loss": 2.501,
"step": 697
},
{
"epoch": 0.7020808449110455,
"grad_norm": 18.920232209491584,
"learning_rate": 9.552872792309413e-06,
"loss": 2.4585,
"step": 698
},
{
"epoch": 0.703086691393726,
"grad_norm": 20.69586994307614,
"learning_rate": 9.55063715627096e-06,
"loss": 2.5144,
"step": 699
},
{
"epoch": 0.7040925378764066,
"grad_norm": 20.291013143429755,
"learning_rate": 9.548401520232506e-06,
"loss": 2.4781,
"step": 700
},
{
"epoch": 0.7050983843590872,
"grad_norm": 18.869940922356157,
"learning_rate": 9.546165884194054e-06,
"loss": 2.5295,
"step": 701
},
{
"epoch": 0.7061042308417678,
"grad_norm": 20.896359569542287,
"learning_rate": 9.5439302481556e-06,
"loss": 2.5302,
"step": 702
},
{
"epoch": 0.7071100773244483,
"grad_norm": 20.082650895679876,
"learning_rate": 9.541694612117149e-06,
"loss": 2.5929,
"step": 703
},
{
"epoch": 0.708115923807129,
"grad_norm": 20.467778705563482,
"learning_rate": 9.539458976078695e-06,
"loss": 2.5059,
"step": 704
},
{
"epoch": 0.7091217702898095,
"grad_norm": 15.681829621018034,
"learning_rate": 9.537223340040242e-06,
"loss": 2.5129,
"step": 705
},
{
"epoch": 0.7101276167724901,
"grad_norm": 20.133542563502154,
"learning_rate": 9.534987704001789e-06,
"loss": 2.5555,
"step": 706
},
{
"epoch": 0.7111334632551707,
"grad_norm": 21.145027241531068,
"learning_rate": 9.532752067963337e-06,
"loss": 2.5357,
"step": 707
},
{
"epoch": 0.7121393097378512,
"grad_norm": 17.53550654089157,
"learning_rate": 9.530516431924883e-06,
"loss": 2.4717,
"step": 708
},
{
"epoch": 0.7131451562205319,
"grad_norm": 20.65677074422556,
"learning_rate": 9.528280795886432e-06,
"loss": 2.513,
"step": 709
},
{
"epoch": 0.7141510027032124,
"grad_norm": 21.89801877836519,
"learning_rate": 9.526045159847978e-06,
"loss": 2.5101,
"step": 710
},
{
"epoch": 0.715156849185893,
"grad_norm": 21.26634341268915,
"learning_rate": 9.523809523809525e-06,
"loss": 2.5282,
"step": 711
},
{
"epoch": 0.7161626956685736,
"grad_norm": 19.627607255985847,
"learning_rate": 9.521573887771071e-06,
"loss": 2.3997,
"step": 712
},
{
"epoch": 0.7171685421512541,
"grad_norm": 17.518544399386524,
"learning_rate": 9.519338251732618e-06,
"loss": 2.502,
"step": 713
},
{
"epoch": 0.7181743886339348,
"grad_norm": 20.25482589759302,
"learning_rate": 9.517102615694166e-06,
"loss": 2.5121,
"step": 714
},
{
"epoch": 0.7191802351166153,
"grad_norm": 20.615452010565395,
"learning_rate": 9.514866979655713e-06,
"loss": 2.5037,
"step": 715
},
{
"epoch": 0.7201860815992959,
"grad_norm": 16.66160935051711,
"learning_rate": 9.512631343617261e-06,
"loss": 2.5062,
"step": 716
},
{
"epoch": 0.7211919280819765,
"grad_norm": 19.14572503732978,
"learning_rate": 9.510395707578807e-06,
"loss": 2.4963,
"step": 717
},
{
"epoch": 0.7221977745646571,
"grad_norm": 19.59250241873381,
"learning_rate": 9.508160071540354e-06,
"loss": 2.4692,
"step": 718
},
{
"epoch": 0.7232036210473376,
"grad_norm": 17.65339946945576,
"learning_rate": 9.5059244355019e-06,
"loss": 2.4886,
"step": 719
},
{
"epoch": 0.7242094675300182,
"grad_norm": 17.059073253679152,
"learning_rate": 9.503688799463449e-06,
"loss": 2.5175,
"step": 720
},
{
"epoch": 0.7252153140126988,
"grad_norm": 16.074504376623388,
"learning_rate": 9.501453163424995e-06,
"loss": 2.5276,
"step": 721
},
{
"epoch": 0.7262211604953794,
"grad_norm": 19.279211211820908,
"learning_rate": 9.499217527386542e-06,
"loss": 2.5111,
"step": 722
},
{
"epoch": 0.72722700697806,
"grad_norm": 18.124284123662274,
"learning_rate": 9.496981891348089e-06,
"loss": 2.5347,
"step": 723
},
{
"epoch": 0.7282328534607405,
"grad_norm": 17.712027375920236,
"learning_rate": 9.494746255309635e-06,
"loss": 2.5076,
"step": 724
},
{
"epoch": 0.7292386999434212,
"grad_norm": 19.28157010260716,
"learning_rate": 9.492510619271183e-06,
"loss": 2.5033,
"step": 725
},
{
"epoch": 0.7302445464261017,
"grad_norm": 19.61185471148692,
"learning_rate": 9.49027498323273e-06,
"loss": 2.4904,
"step": 726
},
{
"epoch": 0.7312503929087822,
"grad_norm": 17.408335083465722,
"learning_rate": 9.488039347194278e-06,
"loss": 2.5597,
"step": 727
},
{
"epoch": 0.7322562393914629,
"grad_norm": 17.951454457586962,
"learning_rate": 9.485803711155825e-06,
"loss": 2.5006,
"step": 728
},
{
"epoch": 0.7332620858741434,
"grad_norm": 20.671307219350894,
"learning_rate": 9.483568075117371e-06,
"loss": 2.5391,
"step": 729
},
{
"epoch": 0.7342679323568241,
"grad_norm": 25.277098409010296,
"learning_rate": 9.481332439078918e-06,
"loss": 2.5179,
"step": 730
},
{
"epoch": 0.7352737788395046,
"grad_norm": 24.093563555760248,
"learning_rate": 9.479096803040466e-06,
"loss": 2.5197,
"step": 731
},
{
"epoch": 0.7362796253221852,
"grad_norm": 19.069594753082438,
"learning_rate": 9.476861167002013e-06,
"loss": 2.5314,
"step": 732
},
{
"epoch": 0.7372854718048658,
"grad_norm": 25.159064679345477,
"learning_rate": 9.474625530963561e-06,
"loss": 2.5633,
"step": 733
},
{
"epoch": 0.7382913182875463,
"grad_norm": 19.255862388290833,
"learning_rate": 9.472389894925107e-06,
"loss": 2.5535,
"step": 734
},
{
"epoch": 0.7392971647702269,
"grad_norm": 20.15846805260267,
"learning_rate": 9.470154258886654e-06,
"loss": 2.5167,
"step": 735
},
{
"epoch": 0.7403030112529075,
"grad_norm": 17.05788614121942,
"learning_rate": 9.4679186228482e-06,
"loss": 2.5146,
"step": 736
},
{
"epoch": 0.7413088577355881,
"grad_norm": 20.346251298426235,
"learning_rate": 9.465682986809747e-06,
"loss": 2.4956,
"step": 737
},
{
"epoch": 0.7423147042182687,
"grad_norm": 17.004794004812634,
"learning_rate": 9.463447350771295e-06,
"loss": 2.5642,
"step": 738
},
{
"epoch": 0.7433205507009493,
"grad_norm": 20.394427656456934,
"learning_rate": 9.461211714732842e-06,
"loss": 2.495,
"step": 739
},
{
"epoch": 0.7443263971836298,
"grad_norm": 19.281712523889723,
"learning_rate": 9.45897607869439e-06,
"loss": 2.4582,
"step": 740
},
{
"epoch": 0.7453322436663105,
"grad_norm": 20.066929363942144,
"learning_rate": 9.456740442655937e-06,
"loss": 2.4908,
"step": 741
},
{
"epoch": 0.746338090148991,
"grad_norm": 18.593791504643672,
"learning_rate": 9.454504806617483e-06,
"loss": 2.5306,
"step": 742
},
{
"epoch": 0.7473439366316716,
"grad_norm": 19.254184430311664,
"learning_rate": 9.45226917057903e-06,
"loss": 2.5237,
"step": 743
},
{
"epoch": 0.7483497831143522,
"grad_norm": 20.26496937620349,
"learning_rate": 9.450033534540578e-06,
"loss": 2.5351,
"step": 744
},
{
"epoch": 0.7493556295970327,
"grad_norm": 19.52286145145994,
"learning_rate": 9.447797898502125e-06,
"loss": 2.507,
"step": 745
},
{
"epoch": 0.7503614760797134,
"grad_norm": 17.098379884990067,
"learning_rate": 9.445562262463671e-06,
"loss": 2.4855,
"step": 746
},
{
"epoch": 0.7513673225623939,
"grad_norm": 19.88240078373313,
"learning_rate": 9.44332662642522e-06,
"loss": 2.4897,
"step": 747
},
{
"epoch": 0.7523731690450745,
"grad_norm": 18.15902628291973,
"learning_rate": 9.441090990386766e-06,
"loss": 2.4605,
"step": 748
},
{
"epoch": 0.7533790155277551,
"grad_norm": 18.379420609505242,
"learning_rate": 9.438855354348313e-06,
"loss": 2.5071,
"step": 749
},
{
"epoch": 0.7543848620104356,
"grad_norm": 19.759355915890186,
"learning_rate": 9.43661971830986e-06,
"loss": 2.5374,
"step": 750
},
{
"epoch": 0.7553907084931163,
"grad_norm": 20.846426367756592,
"learning_rate": 9.434384082271407e-06,
"loss": 2.4438,
"step": 751
},
{
"epoch": 0.7563965549757968,
"grad_norm": 18.569949093964635,
"learning_rate": 9.432148446232954e-06,
"loss": 2.4316,
"step": 752
},
{
"epoch": 0.7574024014584774,
"grad_norm": 18.825641211217608,
"learning_rate": 9.429912810194502e-06,
"loss": 2.4787,
"step": 753
},
{
"epoch": 0.758408247941158,
"grad_norm": 18.251254587229408,
"learning_rate": 9.427677174156049e-06,
"loss": 2.5326,
"step": 754
},
{
"epoch": 0.7594140944238386,
"grad_norm": 19.148609800011037,
"learning_rate": 9.425441538117595e-06,
"loss": 2.4675,
"step": 755
},
{
"epoch": 0.7604199409065191,
"grad_norm": 19.62748223343742,
"learning_rate": 9.423205902079142e-06,
"loss": 2.4738,
"step": 756
},
{
"epoch": 0.7614257873891997,
"grad_norm": 17.275490176829774,
"learning_rate": 9.42097026604069e-06,
"loss": 2.4384,
"step": 757
},
{
"epoch": 0.7624316338718803,
"grad_norm": 20.315078913702827,
"learning_rate": 9.418734630002237e-06,
"loss": 2.4824,
"step": 758
},
{
"epoch": 0.7634374803545609,
"grad_norm": 20.45538876387583,
"learning_rate": 9.416498993963783e-06,
"loss": 2.5446,
"step": 759
},
{
"epoch": 0.7644433268372415,
"grad_norm": 18.494779020290373,
"learning_rate": 9.41426335792533e-06,
"loss": 2.4971,
"step": 760
},
{
"epoch": 0.765449173319922,
"grad_norm": 17.79268503808769,
"learning_rate": 9.412027721886876e-06,
"loss": 2.497,
"step": 761
},
{
"epoch": 0.7664550198026027,
"grad_norm": 17.514485728139903,
"learning_rate": 9.409792085848425e-06,
"loss": 2.4943,
"step": 762
},
{
"epoch": 0.7674608662852832,
"grad_norm": 17.08742932206588,
"learning_rate": 9.407556449809971e-06,
"loss": 2.5622,
"step": 763
},
{
"epoch": 0.7684667127679637,
"grad_norm": 19.17619661894787,
"learning_rate": 9.40532081377152e-06,
"loss": 2.5019,
"step": 764
},
{
"epoch": 0.7694725592506444,
"grad_norm": 20.149404401707567,
"learning_rate": 9.403085177733066e-06,
"loss": 2.5323,
"step": 765
},
{
"epoch": 0.7704784057333249,
"grad_norm": 19.522637904808622,
"learning_rate": 9.400849541694613e-06,
"loss": 2.5068,
"step": 766
},
{
"epoch": 0.7714842522160056,
"grad_norm": 25.35421241795729,
"learning_rate": 9.39861390565616e-06,
"loss": 2.5455,
"step": 767
},
{
"epoch": 0.7724900986986861,
"grad_norm": 21.894940996547597,
"learning_rate": 9.396378269617707e-06,
"loss": 2.4854,
"step": 768
},
{
"epoch": 0.7734959451813667,
"grad_norm": 17.243790817037375,
"learning_rate": 9.394142633579254e-06,
"loss": 2.4934,
"step": 769
},
{
"epoch": 0.7745017916640473,
"grad_norm": 20.051709239288467,
"learning_rate": 9.3919069975408e-06,
"loss": 2.4467,
"step": 770
},
{
"epoch": 0.7755076381467279,
"grad_norm": 19.090236927656047,
"learning_rate": 9.389671361502349e-06,
"loss": 2.5965,
"step": 771
},
{
"epoch": 0.7765134846294084,
"grad_norm": 21.659867988546495,
"learning_rate": 9.387435725463895e-06,
"loss": 2.4831,
"step": 772
},
{
"epoch": 0.777519331112089,
"grad_norm": 22.068674111225377,
"learning_rate": 9.385200089425442e-06,
"loss": 2.4838,
"step": 773
},
{
"epoch": 0.7785251775947696,
"grad_norm": 20.808316291400047,
"learning_rate": 9.382964453386989e-06,
"loss": 2.5204,
"step": 774
},
{
"epoch": 0.7795310240774502,
"grad_norm": 25.226727704140117,
"learning_rate": 9.380728817348537e-06,
"loss": 2.4974,
"step": 775
},
{
"epoch": 0.7805368705601308,
"grad_norm": 27.240953523535676,
"learning_rate": 9.378493181310083e-06,
"loss": 2.5006,
"step": 776
},
{
"epoch": 0.7815427170428113,
"grad_norm": 20.276309928518156,
"learning_rate": 9.376257545271632e-06,
"loss": 2.4803,
"step": 777
},
{
"epoch": 0.782548563525492,
"grad_norm": 20.57402971691398,
"learning_rate": 9.374021909233178e-06,
"loss": 2.5833,
"step": 778
},
{
"epoch": 0.7835544100081725,
"grad_norm": 19.581526916692166,
"learning_rate": 9.371786273194725e-06,
"loss": 2.5137,
"step": 779
},
{
"epoch": 0.784560256490853,
"grad_norm": 17.050469507343866,
"learning_rate": 9.369550637156271e-06,
"loss": 2.4201,
"step": 780
},
{
"epoch": 0.7855661029735337,
"grad_norm": 20.7042448692395,
"learning_rate": 9.36731500111782e-06,
"loss": 2.4744,
"step": 781
},
{
"epoch": 0.7865719494562142,
"grad_norm": 20.411291333400605,
"learning_rate": 9.365079365079366e-06,
"loss": 2.5243,
"step": 782
},
{
"epoch": 0.7875777959388949,
"grad_norm": 18.757539515639955,
"learning_rate": 9.362843729040913e-06,
"loss": 2.4651,
"step": 783
},
{
"epoch": 0.7885836424215754,
"grad_norm": 18.517065201543332,
"learning_rate": 9.360608093002461e-06,
"loss": 2.5381,
"step": 784
},
{
"epoch": 0.789589488904256,
"grad_norm": 15.998281316407619,
"learning_rate": 9.358372456964007e-06,
"loss": 2.4589,
"step": 785
},
{
"epoch": 0.7905953353869366,
"grad_norm": 19.321999707357065,
"learning_rate": 9.356136820925554e-06,
"loss": 2.5073,
"step": 786
},
{
"epoch": 0.7916011818696171,
"grad_norm": 17.3471900742081,
"learning_rate": 9.3539011848871e-06,
"loss": 2.5422,
"step": 787
},
{
"epoch": 0.7926070283522977,
"grad_norm": 18.923681325265438,
"learning_rate": 9.351665548848649e-06,
"loss": 2.5227,
"step": 788
},
{
"epoch": 0.7936128748349783,
"grad_norm": 15.265177415266567,
"learning_rate": 9.349429912810195e-06,
"loss": 2.4649,
"step": 789
},
{
"epoch": 0.7946187213176589,
"grad_norm": 16.530611307751233,
"learning_rate": 9.347194276771742e-06,
"loss": 2.4888,
"step": 790
},
{
"epoch": 0.7956245678003395,
"grad_norm": 17.95202578745719,
"learning_rate": 9.344958640733288e-06,
"loss": 2.5187,
"step": 791
},
{
"epoch": 0.7966304142830201,
"grad_norm": 16.57212408050738,
"learning_rate": 9.342723004694837e-06,
"loss": 2.4524,
"step": 792
},
{
"epoch": 0.7976362607657006,
"grad_norm": 17.77570862737382,
"learning_rate": 9.340487368656383e-06,
"loss": 2.5125,
"step": 793
},
{
"epoch": 0.7986421072483812,
"grad_norm": 17.09498814803331,
"learning_rate": 9.33825173261793e-06,
"loss": 2.5238,
"step": 794
},
{
"epoch": 0.7996479537310618,
"grad_norm": 16.637506590364993,
"learning_rate": 9.336016096579478e-06,
"loss": 2.5148,
"step": 795
},
{
"epoch": 0.8006538002137423,
"grad_norm": 17.260042591014948,
"learning_rate": 9.333780460541025e-06,
"loss": 2.5784,
"step": 796
},
{
"epoch": 0.801659646696423,
"grad_norm": 16.262798660851523,
"learning_rate": 9.331544824502571e-06,
"loss": 2.4777,
"step": 797
},
{
"epoch": 0.8026654931791035,
"grad_norm": 16.23096951371673,
"learning_rate": 9.329309188464118e-06,
"loss": 2.457,
"step": 798
},
{
"epoch": 0.8036713396617842,
"grad_norm": 19.520160951426668,
"learning_rate": 9.327073552425666e-06,
"loss": 2.4871,
"step": 799
},
{
"epoch": 0.8046771861444647,
"grad_norm": 19.03726995483599,
"learning_rate": 9.324837916387213e-06,
"loss": 2.4776,
"step": 800
},
{
"epoch": 0.8056830326271452,
"grad_norm": 15.94571703867295,
"learning_rate": 9.322602280348761e-06,
"loss": 2.5302,
"step": 801
},
{
"epoch": 0.8066888791098259,
"grad_norm": 19.562207263305492,
"learning_rate": 9.320366644310307e-06,
"loss": 2.4405,
"step": 802
},
{
"epoch": 0.8076947255925064,
"grad_norm": 18.15166430685149,
"learning_rate": 9.318131008271854e-06,
"loss": 2.499,
"step": 803
},
{
"epoch": 0.808700572075187,
"grad_norm": 17.669181056591288,
"learning_rate": 9.3158953722334e-06,
"loss": 2.4982,
"step": 804
},
{
"epoch": 0.8097064185578676,
"grad_norm": 16.841416364196867,
"learning_rate": 9.313659736194947e-06,
"loss": 2.4725,
"step": 805
},
{
"epoch": 0.8107122650405482,
"grad_norm": 18.08362901347937,
"learning_rate": 9.311424100156495e-06,
"loss": 2.5331,
"step": 806
},
{
"epoch": 0.8117181115232288,
"grad_norm": 17.783480259954455,
"learning_rate": 9.309188464118042e-06,
"loss": 2.4996,
"step": 807
},
{
"epoch": 0.8127239580059094,
"grad_norm": 16.589835867359405,
"learning_rate": 9.30695282807959e-06,
"loss": 2.4701,
"step": 808
},
{
"epoch": 0.8137298044885899,
"grad_norm": 17.676330608733828,
"learning_rate": 9.304717192041137e-06,
"loss": 2.552,
"step": 809
},
{
"epoch": 0.8147356509712705,
"grad_norm": 15.886159845740128,
"learning_rate": 9.302481556002683e-06,
"loss": 2.4703,
"step": 810
},
{
"epoch": 0.8157414974539511,
"grad_norm": 16.05722427471681,
"learning_rate": 9.30024591996423e-06,
"loss": 2.4908,
"step": 811
},
{
"epoch": 0.8167473439366316,
"grad_norm": 17.36535231267961,
"learning_rate": 9.298010283925778e-06,
"loss": 2.503,
"step": 812
},
{
"epoch": 0.8177531904193123,
"grad_norm": 16.14465414079727,
"learning_rate": 9.295774647887325e-06,
"loss": 2.5195,
"step": 813
},
{
"epoch": 0.8187590369019928,
"grad_norm": 17.089258509215345,
"learning_rate": 9.293539011848873e-06,
"loss": 2.4592,
"step": 814
},
{
"epoch": 0.8197648833846735,
"grad_norm": 18.955704714718312,
"learning_rate": 9.29130337581042e-06,
"loss": 2.4564,
"step": 815
},
{
"epoch": 0.820770729867354,
"grad_norm": 17.309458597444493,
"learning_rate": 9.289067739771966e-06,
"loss": 2.503,
"step": 816
},
{
"epoch": 0.8217765763500345,
"grad_norm": 15.089049062269016,
"learning_rate": 9.286832103733513e-06,
"loss": 2.5347,
"step": 817
},
{
"epoch": 0.8227824228327152,
"grad_norm": 19.012802639782954,
"learning_rate": 9.284596467695059e-06,
"loss": 2.4771,
"step": 818
},
{
"epoch": 0.8237882693153957,
"grad_norm": 18.156252128847964,
"learning_rate": 9.282360831656607e-06,
"loss": 2.4973,
"step": 819
},
{
"epoch": 0.8247941157980763,
"grad_norm": 18.40893210604742,
"learning_rate": 9.280125195618154e-06,
"loss": 2.5712,
"step": 820
},
{
"epoch": 0.8257999622807569,
"grad_norm": 16.399612728431613,
"learning_rate": 9.277889559579702e-06,
"loss": 2.4728,
"step": 821
},
{
"epoch": 0.8268058087634375,
"grad_norm": 18.67911638017806,
"learning_rate": 9.275653923541249e-06,
"loss": 2.5125,
"step": 822
},
{
"epoch": 0.8278116552461181,
"grad_norm": 17.322713351462784,
"learning_rate": 9.273418287502795e-06,
"loss": 2.4476,
"step": 823
},
{
"epoch": 0.8288175017287986,
"grad_norm": 18.716866313005497,
"learning_rate": 9.271182651464342e-06,
"loss": 2.4796,
"step": 824
},
{
"epoch": 0.8298233482114792,
"grad_norm": 18.646291489781458,
"learning_rate": 9.26894701542589e-06,
"loss": 2.4908,
"step": 825
},
{
"epoch": 0.8308291946941598,
"grad_norm": 17.84229204379419,
"learning_rate": 9.266711379387437e-06,
"loss": 2.4703,
"step": 826
},
{
"epoch": 0.8318350411768404,
"grad_norm": 18.64370946639297,
"learning_rate": 9.264475743348983e-06,
"loss": 2.4739,
"step": 827
},
{
"epoch": 0.8328408876595209,
"grad_norm": 18.348032755338412,
"learning_rate": 9.26224010731053e-06,
"loss": 2.497,
"step": 828
},
{
"epoch": 0.8338467341422016,
"grad_norm": 16.841822742120723,
"learning_rate": 9.260004471272076e-06,
"loss": 2.4935,
"step": 829
},
{
"epoch": 0.8348525806248821,
"grad_norm": 19.438358318778544,
"learning_rate": 9.257768835233625e-06,
"loss": 2.5429,
"step": 830
},
{
"epoch": 0.8358584271075628,
"grad_norm": 18.878343814431652,
"learning_rate": 9.255533199195171e-06,
"loss": 2.4938,
"step": 831
},
{
"epoch": 0.8368642735902433,
"grad_norm": 15.766988224446346,
"learning_rate": 9.25329756315672e-06,
"loss": 2.4681,
"step": 832
},
{
"epoch": 0.8378701200729238,
"grad_norm": 17.799674012352813,
"learning_rate": 9.251061927118266e-06,
"loss": 2.5213,
"step": 833
},
{
"epoch": 0.8388759665556045,
"grad_norm": 19.73080921535841,
"learning_rate": 9.248826291079813e-06,
"loss": 2.4887,
"step": 834
},
{
"epoch": 0.839881813038285,
"grad_norm": 21.784930874046488,
"learning_rate": 9.246590655041359e-06,
"loss": 2.5385,
"step": 835
},
{
"epoch": 0.8408876595209656,
"grad_norm": 16.909577276298187,
"learning_rate": 9.244355019002907e-06,
"loss": 2.5363,
"step": 836
},
{
"epoch": 0.8418935060036462,
"grad_norm": 17.585758273866936,
"learning_rate": 9.242119382964454e-06,
"loss": 2.5038,
"step": 837
},
{
"epoch": 0.8428993524863267,
"grad_norm": 21.26661156378665,
"learning_rate": 9.239883746926002e-06,
"loss": 2.4911,
"step": 838
},
{
"epoch": 0.8439051989690074,
"grad_norm": 15.501969039845058,
"learning_rate": 9.237648110887549e-06,
"loss": 2.5092,
"step": 839
},
{
"epoch": 0.8449110454516879,
"grad_norm": 19.35818051179853,
"learning_rate": 9.235412474849095e-06,
"loss": 2.4803,
"step": 840
},
{
"epoch": 0.8459168919343685,
"grad_norm": 16.535060755036255,
"learning_rate": 9.233176838810642e-06,
"loss": 2.4715,
"step": 841
},
{
"epoch": 0.8469227384170491,
"grad_norm": 19.610807091356513,
"learning_rate": 9.230941202772188e-06,
"loss": 2.5063,
"step": 842
},
{
"epoch": 0.8479285848997297,
"grad_norm": 16.2227194450381,
"learning_rate": 9.228705566733737e-06,
"loss": 2.4309,
"step": 843
},
{
"epoch": 0.8489344313824103,
"grad_norm": 18.871725206678025,
"learning_rate": 9.226469930695283e-06,
"loss": 2.5185,
"step": 844
},
{
"epoch": 0.8499402778650909,
"grad_norm": 19.80150197641624,
"learning_rate": 9.224234294656832e-06,
"loss": 2.4366,
"step": 845
},
{
"epoch": 0.8509461243477714,
"grad_norm": 19.82523926255123,
"learning_rate": 9.221998658618378e-06,
"loss": 2.4862,
"step": 846
},
{
"epoch": 0.851951970830452,
"grad_norm": 17.005619743686427,
"learning_rate": 9.219763022579925e-06,
"loss": 2.4198,
"step": 847
},
{
"epoch": 0.8529578173131326,
"grad_norm": 20.636823632802646,
"learning_rate": 9.217527386541471e-06,
"loss": 2.4665,
"step": 848
},
{
"epoch": 0.8539636637958131,
"grad_norm": 19.016174344021117,
"learning_rate": 9.21529175050302e-06,
"loss": 2.4749,
"step": 849
},
{
"epoch": 0.8549695102784938,
"grad_norm": 18.41397658496216,
"learning_rate": 9.213056114464566e-06,
"loss": 2.4358,
"step": 850
},
{
"epoch": 0.8559753567611743,
"grad_norm": 17.46790390111685,
"learning_rate": 9.210820478426114e-06,
"loss": 2.4979,
"step": 851
},
{
"epoch": 0.856981203243855,
"grad_norm": 18.488587583895956,
"learning_rate": 9.20858484238766e-06,
"loss": 2.4829,
"step": 852
},
{
"epoch": 0.8579870497265355,
"grad_norm": 24.74417215175763,
"learning_rate": 9.206349206349207e-06,
"loss": 2.5312,
"step": 853
},
{
"epoch": 0.858992896209216,
"grad_norm": 19.62898918524025,
"learning_rate": 9.204113570310754e-06,
"loss": 2.4736,
"step": 854
},
{
"epoch": 0.8599987426918967,
"grad_norm": 20.709975695069204,
"learning_rate": 9.2018779342723e-06,
"loss": 2.4419,
"step": 855
},
{
"epoch": 0.8610045891745772,
"grad_norm": 23.505683041265875,
"learning_rate": 9.199642298233849e-06,
"loss": 2.52,
"step": 856
},
{
"epoch": 0.8620104356572578,
"grad_norm": 23.061385212409167,
"learning_rate": 9.197406662195395e-06,
"loss": 2.5224,
"step": 857
},
{
"epoch": 0.8630162821399384,
"grad_norm": 19.435142582973928,
"learning_rate": 9.195171026156942e-06,
"loss": 2.5116,
"step": 858
},
{
"epoch": 0.864022128622619,
"grad_norm": 20.935679867064998,
"learning_rate": 9.192935390118488e-06,
"loss": 2.5244,
"step": 859
},
{
"epoch": 0.8650279751052996,
"grad_norm": 19.541576135091336,
"learning_rate": 9.190699754080037e-06,
"loss": 2.5302,
"step": 860
},
{
"epoch": 0.8660338215879801,
"grad_norm": 18.688832058440994,
"learning_rate": 9.188464118041583e-06,
"loss": 2.4724,
"step": 861
},
{
"epoch": 0.8670396680706607,
"grad_norm": 19.95676951865818,
"learning_rate": 9.186228482003131e-06,
"loss": 2.4672,
"step": 862
},
{
"epoch": 0.8680455145533413,
"grad_norm": 19.390108841449873,
"learning_rate": 9.183992845964678e-06,
"loss": 2.4413,
"step": 863
},
{
"epoch": 0.8690513610360219,
"grad_norm": 16.19229969510003,
"learning_rate": 9.181757209926225e-06,
"loss": 2.4524,
"step": 864
},
{
"epoch": 0.8700572075187024,
"grad_norm": 18.60982850694698,
"learning_rate": 9.179521573887771e-06,
"loss": 2.5208,
"step": 865
},
{
"epoch": 0.8710630540013831,
"grad_norm": 17.614897460064757,
"learning_rate": 9.177285937849318e-06,
"loss": 2.462,
"step": 866
},
{
"epoch": 0.8720689004840636,
"grad_norm": 15.123819842436388,
"learning_rate": 9.175050301810866e-06,
"loss": 2.4727,
"step": 867
},
{
"epoch": 0.8730747469667443,
"grad_norm": 17.31041379575615,
"learning_rate": 9.172814665772413e-06,
"loss": 2.5213,
"step": 868
},
{
"epoch": 0.8740805934494248,
"grad_norm": 19.055356941197203,
"learning_rate": 9.17057902973396e-06,
"loss": 2.4818,
"step": 869
},
{
"epoch": 0.8750864399321053,
"grad_norm": 18.077774764214077,
"learning_rate": 9.168343393695507e-06,
"loss": 2.4637,
"step": 870
},
{
"epoch": 0.876092286414786,
"grad_norm": 21.143295726131285,
"learning_rate": 9.166107757657054e-06,
"loss": 2.4704,
"step": 871
},
{
"epoch": 0.8770981328974665,
"grad_norm": 17.942475112504635,
"learning_rate": 9.1638721216186e-06,
"loss": 2.4841,
"step": 872
},
{
"epoch": 0.8781039793801471,
"grad_norm": 19.18496687136911,
"learning_rate": 9.161636485580149e-06,
"loss": 2.4917,
"step": 873
},
{
"epoch": 0.8791098258628277,
"grad_norm": 17.588233448036576,
"learning_rate": 9.159400849541695e-06,
"loss": 2.5656,
"step": 874
},
{
"epoch": 0.8801156723455082,
"grad_norm": 20.044382673740458,
"learning_rate": 9.157165213503244e-06,
"loss": 2.5114,
"step": 875
},
{
"epoch": 0.8811215188281889,
"grad_norm": 17.914129095672784,
"learning_rate": 9.15492957746479e-06,
"loss": 2.5063,
"step": 876
},
{
"epoch": 0.8821273653108694,
"grad_norm": 20.358962481262786,
"learning_rate": 9.152693941426337e-06,
"loss": 2.4819,
"step": 877
},
{
"epoch": 0.88313321179355,
"grad_norm": 20.562144083555342,
"learning_rate": 9.150458305387883e-06,
"loss": 2.5522,
"step": 878
},
{
"epoch": 0.8841390582762306,
"grad_norm": 17.559901617392246,
"learning_rate": 9.14822266934943e-06,
"loss": 2.4994,
"step": 879
},
{
"epoch": 0.8851449047589112,
"grad_norm": 19.353815367284025,
"learning_rate": 9.145987033310978e-06,
"loss": 2.4995,
"step": 880
},
{
"epoch": 0.8861507512415917,
"grad_norm": 20.337197517860805,
"learning_rate": 9.143751397272525e-06,
"loss": 2.5063,
"step": 881
},
{
"epoch": 0.8871565977242724,
"grad_norm": 20.21527535771532,
"learning_rate": 9.141515761234073e-06,
"loss": 2.5583,
"step": 882
},
{
"epoch": 0.8881624442069529,
"grad_norm": 21.947674142946916,
"learning_rate": 9.13928012519562e-06,
"loss": 2.4932,
"step": 883
},
{
"epoch": 0.8891682906896335,
"grad_norm": 20.241067388653406,
"learning_rate": 9.137044489157166e-06,
"loss": 2.4566,
"step": 884
},
{
"epoch": 0.8901741371723141,
"grad_norm": 18.94219764805125,
"learning_rate": 9.134808853118713e-06,
"loss": 2.5193,
"step": 885
},
{
"epoch": 0.8911799836549946,
"grad_norm": 18.21141857163298,
"learning_rate": 9.13257321708026e-06,
"loss": 2.5036,
"step": 886
},
{
"epoch": 0.8921858301376753,
"grad_norm": 18.879260967660596,
"learning_rate": 9.130337581041807e-06,
"loss": 2.4581,
"step": 887
},
{
"epoch": 0.8931916766203558,
"grad_norm": 15.904893478954735,
"learning_rate": 9.128101945003354e-06,
"loss": 2.5704,
"step": 888
},
{
"epoch": 0.8941975231030364,
"grad_norm": 17.719412197108404,
"learning_rate": 9.125866308964902e-06,
"loss": 2.4746,
"step": 889
},
{
"epoch": 0.895203369585717,
"grad_norm": 17.337500587360978,
"learning_rate": 9.123630672926449e-06,
"loss": 2.4891,
"step": 890
},
{
"epoch": 0.8962092160683975,
"grad_norm": 17.40269334016562,
"learning_rate": 9.121395036887995e-06,
"loss": 2.5213,
"step": 891
},
{
"epoch": 0.8972150625510782,
"grad_norm": 16.687508641627172,
"learning_rate": 9.119159400849542e-06,
"loss": 2.5013,
"step": 892
},
{
"epoch": 0.8982209090337587,
"grad_norm": 18.68430097414652,
"learning_rate": 9.11692376481109e-06,
"loss": 2.5189,
"step": 893
},
{
"epoch": 0.8992267555164393,
"grad_norm": 19.021895555033808,
"learning_rate": 9.114688128772637e-06,
"loss": 2.5146,
"step": 894
},
{
"epoch": 0.9002326019991199,
"grad_norm": 19.67201740058453,
"learning_rate": 9.112452492734183e-06,
"loss": 2.4956,
"step": 895
},
{
"epoch": 0.9012384484818005,
"grad_norm": 21.161941542262337,
"learning_rate": 9.11021685669573e-06,
"loss": 2.4697,
"step": 896
},
{
"epoch": 0.902244294964481,
"grad_norm": 21.536270522176043,
"learning_rate": 9.107981220657278e-06,
"loss": 2.511,
"step": 897
},
{
"epoch": 0.9032501414471616,
"grad_norm": 18.722958300928553,
"learning_rate": 9.105745584618825e-06,
"loss": 2.5587,
"step": 898
},
{
"epoch": 0.9042559879298422,
"grad_norm": 21.90000929319424,
"learning_rate": 9.103509948580373e-06,
"loss": 2.4839,
"step": 899
},
{
"epoch": 0.9052618344125228,
"grad_norm": 20.7546999739206,
"learning_rate": 9.10127431254192e-06,
"loss": 2.4831,
"step": 900
},
{
"epoch": 0.9062676808952034,
"grad_norm": 16.69773588405377,
"learning_rate": 9.099038676503466e-06,
"loss": 2.432,
"step": 901
},
{
"epoch": 0.9072735273778839,
"grad_norm": 17.697663909486884,
"learning_rate": 9.096803040465013e-06,
"loss": 2.4746,
"step": 902
},
{
"epoch": 0.9082793738605646,
"grad_norm": 21.136014663688808,
"learning_rate": 9.094567404426559e-06,
"loss": 2.5125,
"step": 903
},
{
"epoch": 0.9092852203432451,
"grad_norm": 15.80770427361095,
"learning_rate": 9.092331768388107e-06,
"loss": 2.4444,
"step": 904
},
{
"epoch": 0.9102910668259256,
"grad_norm": 17.630039598712568,
"learning_rate": 9.090096132349654e-06,
"loss": 2.5108,
"step": 905
},
{
"epoch": 0.9112969133086063,
"grad_norm": 18.196870410777603,
"learning_rate": 9.087860496311202e-06,
"loss": 2.4715,
"step": 906
},
{
"epoch": 0.9123027597912868,
"grad_norm": 15.471661891634048,
"learning_rate": 9.085624860272749e-06,
"loss": 2.4712,
"step": 907
},
{
"epoch": 0.9133086062739675,
"grad_norm": 20.275945900358217,
"learning_rate": 9.083389224234295e-06,
"loss": 2.4799,
"step": 908
},
{
"epoch": 0.914314452756648,
"grad_norm": 22.54244758097815,
"learning_rate": 9.081153588195842e-06,
"loss": 2.5023,
"step": 909
},
{
"epoch": 0.9153202992393286,
"grad_norm": 17.139623998998083,
"learning_rate": 9.07891795215739e-06,
"loss": 2.4485,
"step": 910
},
{
"epoch": 0.9163261457220092,
"grad_norm": 21.09355687447774,
"learning_rate": 9.076682316118937e-06,
"loss": 2.4921,
"step": 911
},
{
"epoch": 0.9173319922046898,
"grad_norm": 20.30710748947842,
"learning_rate": 9.074446680080483e-06,
"loss": 2.513,
"step": 912
},
{
"epoch": 0.9183378386873703,
"grad_norm": 17.780778251179097,
"learning_rate": 9.072211044042031e-06,
"loss": 2.478,
"step": 913
},
{
"epoch": 0.9193436851700509,
"grad_norm": 16.811738212001085,
"learning_rate": 9.069975408003578e-06,
"loss": 2.5019,
"step": 914
},
{
"epoch": 0.9203495316527315,
"grad_norm": 18.395521000320436,
"learning_rate": 9.067739771965125e-06,
"loss": 2.4809,
"step": 915
},
{
"epoch": 0.9213553781354121,
"grad_norm": 20.01473796386837,
"learning_rate": 9.065504135926671e-06,
"loss": 2.5079,
"step": 916
},
{
"epoch": 0.9223612246180927,
"grad_norm": 19.503234127759303,
"learning_rate": 9.06326849988822e-06,
"loss": 2.5152,
"step": 917
},
{
"epoch": 0.9233670711007732,
"grad_norm": 17.988337162833112,
"learning_rate": 9.061032863849766e-06,
"loss": 2.5215,
"step": 918
},
{
"epoch": 0.9243729175834539,
"grad_norm": 21.476525188151253,
"learning_rate": 9.058797227811314e-06,
"loss": 2.5268,
"step": 919
},
{
"epoch": 0.9253787640661344,
"grad_norm": 17.452253827591644,
"learning_rate": 9.05656159177286e-06,
"loss": 2.5168,
"step": 920
},
{
"epoch": 0.9263846105488149,
"grad_norm": 20.067667819078714,
"learning_rate": 9.054325955734407e-06,
"loss": 2.4566,
"step": 921
},
{
"epoch": 0.9273904570314956,
"grad_norm": 19.69466816278088,
"learning_rate": 9.052090319695954e-06,
"loss": 2.4962,
"step": 922
},
{
"epoch": 0.9283963035141761,
"grad_norm": 18.525210255914946,
"learning_rate": 9.0498546836575e-06,
"loss": 2.4585,
"step": 923
},
{
"epoch": 0.9294021499968568,
"grad_norm": 19.444686243782936,
"learning_rate": 9.047619047619049e-06,
"loss": 2.4733,
"step": 924
},
{
"epoch": 0.9304079964795373,
"grad_norm": 17.956521398490462,
"learning_rate": 9.045383411580595e-06,
"loss": 2.4444,
"step": 925
},
{
"epoch": 0.9314138429622179,
"grad_norm": 18.761525002564817,
"learning_rate": 9.043147775542142e-06,
"loss": 2.5182,
"step": 926
},
{
"epoch": 0.9324196894448985,
"grad_norm": 18.517956347635913,
"learning_rate": 9.040912139503688e-06,
"loss": 2.478,
"step": 927
},
{
"epoch": 0.933425535927579,
"grad_norm": 16.722299555521936,
"learning_rate": 9.038676503465237e-06,
"loss": 2.4512,
"step": 928
},
{
"epoch": 0.9344313824102596,
"grad_norm": 16.348937736219405,
"learning_rate": 9.036440867426783e-06,
"loss": 2.489,
"step": 929
},
{
"epoch": 0.9354372288929402,
"grad_norm": 17.128346726995666,
"learning_rate": 9.034205231388331e-06,
"loss": 2.4643,
"step": 930
},
{
"epoch": 0.9364430753756208,
"grad_norm": 15.251872588593749,
"learning_rate": 9.031969595349878e-06,
"loss": 2.5159,
"step": 931
},
{
"epoch": 0.9374489218583014,
"grad_norm": 15.953362658854287,
"learning_rate": 9.029733959311425e-06,
"loss": 2.461,
"step": 932
},
{
"epoch": 0.938454768340982,
"grad_norm": 17.979068884310784,
"learning_rate": 9.027498323272971e-06,
"loss": 2.5227,
"step": 933
},
{
"epoch": 0.9394606148236625,
"grad_norm": 18.551703087669928,
"learning_rate": 9.02526268723452e-06,
"loss": 2.5409,
"step": 934
},
{
"epoch": 0.9404664613063431,
"grad_norm": 19.45169292949665,
"learning_rate": 9.023027051196066e-06,
"loss": 2.5253,
"step": 935
},
{
"epoch": 0.9414723077890237,
"grad_norm": 15.139349601071542,
"learning_rate": 9.020791415157612e-06,
"loss": 2.4184,
"step": 936
},
{
"epoch": 0.9424781542717042,
"grad_norm": 17.231348952189535,
"learning_rate": 9.01855577911916e-06,
"loss": 2.4608,
"step": 937
},
{
"epoch": 0.9434840007543849,
"grad_norm": 17.017076413253353,
"learning_rate": 9.016320143080707e-06,
"loss": 2.4757,
"step": 938
},
{
"epoch": 0.9444898472370654,
"grad_norm": 17.14787644558583,
"learning_rate": 9.014084507042254e-06,
"loss": 2.5353,
"step": 939
},
{
"epoch": 0.9454956937197461,
"grad_norm": 17.737202649119144,
"learning_rate": 9.0118488710038e-06,
"loss": 2.5025,
"step": 940
},
{
"epoch": 0.9465015402024266,
"grad_norm": 17.792474997732924,
"learning_rate": 9.009613234965349e-06,
"loss": 2.5015,
"step": 941
},
{
"epoch": 0.9475073866851071,
"grad_norm": 15.940518080280272,
"learning_rate": 9.007377598926895e-06,
"loss": 2.5094,
"step": 942
},
{
"epoch": 0.9485132331677878,
"grad_norm": 18.487826228534622,
"learning_rate": 9.005141962888443e-06,
"loss": 2.4741,
"step": 943
},
{
"epoch": 0.9495190796504683,
"grad_norm": 17.216862294112154,
"learning_rate": 9.00290632684999e-06,
"loss": 2.4211,
"step": 944
},
{
"epoch": 0.9505249261331489,
"grad_norm": 15.551142433908357,
"learning_rate": 9.000670690811537e-06,
"loss": 2.4976,
"step": 945
},
{
"epoch": 0.9515307726158295,
"grad_norm": 16.559861003507038,
"learning_rate": 8.998435054773083e-06,
"loss": 2.4588,
"step": 946
},
{
"epoch": 0.9525366190985101,
"grad_norm": 17.63172975324414,
"learning_rate": 8.99619941873463e-06,
"loss": 2.4329,
"step": 947
},
{
"epoch": 0.9535424655811907,
"grad_norm": 18.622917952836307,
"learning_rate": 8.993963782696178e-06,
"loss": 2.4807,
"step": 948
},
{
"epoch": 0.9545483120638713,
"grad_norm": 16.338398101232766,
"learning_rate": 8.991728146657725e-06,
"loss": 2.4793,
"step": 949
},
{
"epoch": 0.9555541585465518,
"grad_norm": 21.981541610522587,
"learning_rate": 8.989492510619273e-06,
"loss": 2.504,
"step": 950
},
{
"epoch": 0.9565600050292324,
"grad_norm": 17.581038921848826,
"learning_rate": 8.98725687458082e-06,
"loss": 2.4469,
"step": 951
},
{
"epoch": 0.957565851511913,
"grad_norm": 20.139070495761704,
"learning_rate": 8.985021238542366e-06,
"loss": 2.4517,
"step": 952
},
{
"epoch": 0.9585716979945936,
"grad_norm": 23.595785077951778,
"learning_rate": 8.982785602503912e-06,
"loss": 2.443,
"step": 953
},
{
"epoch": 0.9595775444772742,
"grad_norm": 17.029891375528152,
"learning_rate": 8.98054996646546e-06,
"loss": 2.5467,
"step": 954
},
{
"epoch": 0.9605833909599547,
"grad_norm": 25.44050321518066,
"learning_rate": 8.978314330427007e-06,
"loss": 2.4655,
"step": 955
},
{
"epoch": 0.9615892374426354,
"grad_norm": 27.67074335823461,
"learning_rate": 8.976078694388556e-06,
"loss": 2.5033,
"step": 956
},
{
"epoch": 0.9625950839253159,
"grad_norm": 20.2426136586843,
"learning_rate": 8.973843058350102e-06,
"loss": 2.4978,
"step": 957
},
{
"epoch": 0.9636009304079964,
"grad_norm": 18.658850559045447,
"learning_rate": 8.971607422311649e-06,
"loss": 2.436,
"step": 958
},
{
"epoch": 0.9646067768906771,
"grad_norm": 22.04603622451962,
"learning_rate": 8.969371786273195e-06,
"loss": 2.4269,
"step": 959
},
{
"epoch": 0.9656126233733576,
"grad_norm": 21.75932909465551,
"learning_rate": 8.967136150234742e-06,
"loss": 2.4801,
"step": 960
},
{
"epoch": 0.9666184698560383,
"grad_norm": 19.527187135283782,
"learning_rate": 8.96490051419629e-06,
"loss": 2.5017,
"step": 961
},
{
"epoch": 0.9676243163387188,
"grad_norm": 23.29560371501392,
"learning_rate": 8.962664878157837e-06,
"loss": 2.4738,
"step": 962
},
{
"epoch": 0.9686301628213994,
"grad_norm": 22.706509828966826,
"learning_rate": 8.960429242119383e-06,
"loss": 2.4983,
"step": 963
},
{
"epoch": 0.96963600930408,
"grad_norm": 19.302250360711678,
"learning_rate": 8.95819360608093e-06,
"loss": 2.5015,
"step": 964
},
{
"epoch": 0.9706418557867605,
"grad_norm": 20.274377821577197,
"learning_rate": 8.955957970042478e-06,
"loss": 2.527,
"step": 965
},
{
"epoch": 0.9716477022694411,
"grad_norm": 22.146231245864477,
"learning_rate": 8.953722334004025e-06,
"loss": 2.4754,
"step": 966
},
{
"epoch": 0.9726535487521217,
"grad_norm": 19.283732586891457,
"learning_rate": 8.951486697965573e-06,
"loss": 2.5287,
"step": 967
},
{
"epoch": 0.9736593952348023,
"grad_norm": 16.697804743440038,
"learning_rate": 8.94925106192712e-06,
"loss": 2.446,
"step": 968
},
{
"epoch": 0.9746652417174829,
"grad_norm": 18.750802946568303,
"learning_rate": 8.947015425888666e-06,
"loss": 2.5017,
"step": 969
},
{
"epoch": 0.9756710882001635,
"grad_norm": 19.391455727776027,
"learning_rate": 8.944779789850212e-06,
"loss": 2.4437,
"step": 970
},
{
"epoch": 0.976676934682844,
"grad_norm": 20.49457113262894,
"learning_rate": 8.942544153811759e-06,
"loss": 2.4844,
"step": 971
},
{
"epoch": 0.9776827811655247,
"grad_norm": 16.8204200261022,
"learning_rate": 8.940308517773307e-06,
"loss": 2.5061,
"step": 972
},
{
"epoch": 0.9786886276482052,
"grad_norm": 20.60414960158546,
"learning_rate": 8.938072881734854e-06,
"loss": 2.4707,
"step": 973
},
{
"epoch": 0.9796944741308857,
"grad_norm": 20.536955242053217,
"learning_rate": 8.935837245696402e-06,
"loss": 2.4961,
"step": 974
},
{
"epoch": 0.9807003206135664,
"grad_norm": 19.711651025802677,
"learning_rate": 8.933601609657949e-06,
"loss": 2.5085,
"step": 975
},
{
"epoch": 0.9817061670962469,
"grad_norm": 22.207382958127546,
"learning_rate": 8.931365973619495e-06,
"loss": 2.4681,
"step": 976
},
{
"epoch": 0.9827120135789276,
"grad_norm": 20.071766709389117,
"learning_rate": 8.929130337581042e-06,
"loss": 2.4954,
"step": 977
},
{
"epoch": 0.9837178600616081,
"grad_norm": 16.47582841812017,
"learning_rate": 8.92689470154259e-06,
"loss": 2.5156,
"step": 978
},
{
"epoch": 0.9847237065442886,
"grad_norm": 16.271580188410653,
"learning_rate": 8.924659065504137e-06,
"loss": 2.5552,
"step": 979
},
{
"epoch": 0.9857295530269693,
"grad_norm": 17.119654997142348,
"learning_rate": 8.922423429465685e-06,
"loss": 2.5,
"step": 980
},
{
"epoch": 0.9867353995096498,
"grad_norm": 17.172594870254994,
"learning_rate": 8.920187793427231e-06,
"loss": 2.5154,
"step": 981
},
{
"epoch": 0.9877412459923304,
"grad_norm": 15.830398642073979,
"learning_rate": 8.917952157388778e-06,
"loss": 2.4756,
"step": 982
},
{
"epoch": 0.988747092475011,
"grad_norm": 18.032949559237505,
"learning_rate": 8.915716521350325e-06,
"loss": 2.4683,
"step": 983
},
{
"epoch": 0.9897529389576916,
"grad_norm": 21.19714163524062,
"learning_rate": 8.913480885311871e-06,
"loss": 2.5453,
"step": 984
},
{
"epoch": 0.9907587854403722,
"grad_norm": 17.49779963037731,
"learning_rate": 8.91124524927342e-06,
"loss": 2.5045,
"step": 985
},
{
"epoch": 0.9917646319230528,
"grad_norm": 16.740921716038336,
"learning_rate": 8.909009613234966e-06,
"loss": 2.4452,
"step": 986
},
{
"epoch": 0.9927704784057333,
"grad_norm": 17.19468353475136,
"learning_rate": 8.906773977196514e-06,
"loss": 2.51,
"step": 987
},
{
"epoch": 0.9937763248884139,
"grad_norm": 19.231917401802384,
"learning_rate": 8.90453834115806e-06,
"loss": 2.4869,
"step": 988
},
{
"epoch": 0.9947821713710945,
"grad_norm": 16.664642597713588,
"learning_rate": 8.902302705119607e-06,
"loss": 2.5429,
"step": 989
},
{
"epoch": 0.995788017853775,
"grad_norm": 17.163669929741783,
"learning_rate": 8.900067069081154e-06,
"loss": 2.4918,
"step": 990
},
{
"epoch": 0.9967938643364557,
"grad_norm": 18.533245352747766,
"learning_rate": 8.897831433042702e-06,
"loss": 2.4572,
"step": 991
},
{
"epoch": 0.9977997108191362,
"grad_norm": 19.22814464766119,
"learning_rate": 8.895595797004249e-06,
"loss": 2.4701,
"step": 992
},
{
"epoch": 0.9988055573018169,
"grad_norm": 17.838257655579174,
"learning_rate": 8.893360160965795e-06,
"loss": 2.5084,
"step": 993
},
{
"epoch": 0.9998114037844974,
"grad_norm": 19.097020194522255,
"learning_rate": 8.891124524927342e-06,
"loss": 2.5596,
"step": 994
},
{
"epoch": 1.000817250267178,
"grad_norm": 18.87435671359403,
"learning_rate": 8.888888888888888e-06,
"loss": 2.3563,
"step": 995
},
{
"epoch": 1.0018230967498585,
"grad_norm": 17.90502272682222,
"learning_rate": 8.886653252850437e-06,
"loss": 2.2951,
"step": 996
},
{
"epoch": 1.0028289432325392,
"grad_norm": 16.91198265807751,
"learning_rate": 8.884417616811983e-06,
"loss": 2.3775,
"step": 997
},
{
"epoch": 1.0038347897152198,
"grad_norm": 15.620591946147508,
"learning_rate": 8.882181980773531e-06,
"loss": 2.2319,
"step": 998
},
{
"epoch": 1.0048406361979003,
"grad_norm": 16.42892213916386,
"learning_rate": 8.879946344735078e-06,
"loss": 2.3308,
"step": 999
},
{
"epoch": 1.0058464826805809,
"grad_norm": 18.59757875443186,
"learning_rate": 8.877710708696624e-06,
"loss": 2.2971,
"step": 1000
},
{
"epoch": 1.0068523291632614,
"grad_norm": 16.42491496826214,
"learning_rate": 8.875475072658171e-06,
"loss": 2.2728,
"step": 1001
},
{
"epoch": 1.007858175645942,
"grad_norm": 20.155015522173446,
"learning_rate": 8.87323943661972e-06,
"loss": 2.3029,
"step": 1002
},
{
"epoch": 1.0088640221286227,
"grad_norm": 18.16790024429524,
"learning_rate": 8.871003800581266e-06,
"loss": 2.3385,
"step": 1003
},
{
"epoch": 1.0098698686113032,
"grad_norm": 19.559520755472036,
"learning_rate": 8.868768164542814e-06,
"loss": 2.3133,
"step": 1004
},
{
"epoch": 1.0108757150939838,
"grad_norm": 19.597847391813175,
"learning_rate": 8.86653252850436e-06,
"loss": 2.3219,
"step": 1005
},
{
"epoch": 1.0118815615766643,
"grad_norm": 18.564284872046343,
"learning_rate": 8.864296892465907e-06,
"loss": 2.2791,
"step": 1006
},
{
"epoch": 1.0128874080593449,
"grad_norm": 19.507875257609435,
"learning_rate": 8.862061256427454e-06,
"loss": 2.3247,
"step": 1007
},
{
"epoch": 1.0138932545420256,
"grad_norm": 17.356664585211536,
"learning_rate": 8.859825620389e-06,
"loss": 2.2976,
"step": 1008
},
{
"epoch": 1.0148991010247062,
"grad_norm": 17.685890921074805,
"learning_rate": 8.857589984350549e-06,
"loss": 2.334,
"step": 1009
},
{
"epoch": 1.0159049475073867,
"grad_norm": 18.007613178096328,
"learning_rate": 8.855354348312095e-06,
"loss": 2.3276,
"step": 1010
},
{
"epoch": 1.0169107939900672,
"grad_norm": 21.714385969059393,
"learning_rate": 8.853118712273643e-06,
"loss": 2.3058,
"step": 1011
},
{
"epoch": 1.0179166404727478,
"grad_norm": 17.571687311154825,
"learning_rate": 8.85088307623519e-06,
"loss": 2.3071,
"step": 1012
},
{
"epoch": 1.0189224869554285,
"grad_norm": 18.709912154375786,
"learning_rate": 8.848647440196737e-06,
"loss": 2.3123,
"step": 1013
},
{
"epoch": 1.019928333438109,
"grad_norm": 16.602238512927716,
"learning_rate": 8.846411804158283e-06,
"loss": 2.2744,
"step": 1014
},
{
"epoch": 1.0209341799207896,
"grad_norm": 19.883495863934453,
"learning_rate": 8.844176168119831e-06,
"loss": 2.2654,
"step": 1015
},
{
"epoch": 1.0219400264034701,
"grad_norm": 16.96892661890455,
"learning_rate": 8.841940532081378e-06,
"loss": 2.3024,
"step": 1016
},
{
"epoch": 1.0229458728861507,
"grad_norm": 17.414681329439077,
"learning_rate": 8.839704896042926e-06,
"loss": 2.3003,
"step": 1017
},
{
"epoch": 1.0239517193688312,
"grad_norm": 18.858230864626055,
"learning_rate": 8.837469260004473e-06,
"loss": 2.2919,
"step": 1018
},
{
"epoch": 1.024957565851512,
"grad_norm": 16.699976095125997,
"learning_rate": 8.83523362396602e-06,
"loss": 2.213,
"step": 1019
},
{
"epoch": 1.0259634123341925,
"grad_norm": 15.54260552775944,
"learning_rate": 8.832997987927566e-06,
"loss": 2.3101,
"step": 1020
},
{
"epoch": 1.026969258816873,
"grad_norm": 20.651042832781755,
"learning_rate": 8.830762351889112e-06,
"loss": 2.312,
"step": 1021
},
{
"epoch": 1.0279751052995536,
"grad_norm": 16.94510148062415,
"learning_rate": 8.82852671585066e-06,
"loss": 2.3125,
"step": 1022
},
{
"epoch": 1.0289809517822341,
"grad_norm": 15.84034483396051,
"learning_rate": 8.826291079812207e-06,
"loss": 2.2957,
"step": 1023
},
{
"epoch": 1.029986798264915,
"grad_norm": 16.94753969310773,
"learning_rate": 8.824055443773755e-06,
"loss": 2.3357,
"step": 1024
},
{
"epoch": 1.0309926447475954,
"grad_norm": 16.41312755901395,
"learning_rate": 8.821819807735302e-06,
"loss": 2.2873,
"step": 1025
},
{
"epoch": 1.031998491230276,
"grad_norm": 16.14597452506802,
"learning_rate": 8.819584171696849e-06,
"loss": 2.2386,
"step": 1026
},
{
"epoch": 1.0330043377129565,
"grad_norm": 16.81133489039435,
"learning_rate": 8.817348535658395e-06,
"loss": 2.3233,
"step": 1027
},
{
"epoch": 1.034010184195637,
"grad_norm": 17.11584396129041,
"learning_rate": 8.815112899619943e-06,
"loss": 2.3202,
"step": 1028
},
{
"epoch": 1.0350160306783178,
"grad_norm": 18.735494496930798,
"learning_rate": 8.81287726358149e-06,
"loss": 2.2851,
"step": 1029
},
{
"epoch": 1.0360218771609984,
"grad_norm": 17.886553532549136,
"learning_rate": 8.810641627543037e-06,
"loss": 2.3147,
"step": 1030
},
{
"epoch": 1.037027723643679,
"grad_norm": 23.272350800324688,
"learning_rate": 8.808405991504583e-06,
"loss": 2.2582,
"step": 1031
},
{
"epoch": 1.0380335701263594,
"grad_norm": 20.591526948640638,
"learning_rate": 8.80617035546613e-06,
"loss": 2.2953,
"step": 1032
},
{
"epoch": 1.03903941660904,
"grad_norm": 19.400124919183924,
"learning_rate": 8.803934719427678e-06,
"loss": 2.196,
"step": 1033
},
{
"epoch": 1.0400452630917205,
"grad_norm": 21.820764887623767,
"learning_rate": 8.801699083389224e-06,
"loss": 2.3064,
"step": 1034
},
{
"epoch": 1.0410511095744013,
"grad_norm": 18.083478880895772,
"learning_rate": 8.799463447350773e-06,
"loss": 2.2953,
"step": 1035
},
{
"epoch": 1.0420569560570818,
"grad_norm": 18.954547842365145,
"learning_rate": 8.79722781131232e-06,
"loss": 2.3336,
"step": 1036
},
{
"epoch": 1.0430628025397624,
"grad_norm": 16.39042328254077,
"learning_rate": 8.794992175273866e-06,
"loss": 2.2782,
"step": 1037
},
{
"epoch": 1.044068649022443,
"grad_norm": 19.630757954483688,
"learning_rate": 8.792756539235412e-06,
"loss": 2.2513,
"step": 1038
},
{
"epoch": 1.0450744955051234,
"grad_norm": 18.040721411714422,
"learning_rate": 8.79052090319696e-06,
"loss": 2.3292,
"step": 1039
},
{
"epoch": 1.0460803419878042,
"grad_norm": 18.652583945177508,
"learning_rate": 8.788285267158507e-06,
"loss": 2.2952,
"step": 1040
},
{
"epoch": 1.0470861884704847,
"grad_norm": 20.924815906038443,
"learning_rate": 8.786049631120054e-06,
"loss": 2.3122,
"step": 1041
},
{
"epoch": 1.0480920349531653,
"grad_norm": 17.843040001830474,
"learning_rate": 8.783813995081602e-06,
"loss": 2.3082,
"step": 1042
},
{
"epoch": 1.0490978814358458,
"grad_norm": 16.320339567013974,
"learning_rate": 8.781578359043149e-06,
"loss": 2.2638,
"step": 1043
},
{
"epoch": 1.0501037279185264,
"grad_norm": 17.50310854731714,
"learning_rate": 8.779342723004695e-06,
"loss": 2.2987,
"step": 1044
},
{
"epoch": 1.0511095744012071,
"grad_norm": 21.22193429459719,
"learning_rate": 8.777107086966242e-06,
"loss": 2.3096,
"step": 1045
},
{
"epoch": 1.0521154208838877,
"grad_norm": 17.836388406412574,
"learning_rate": 8.77487145092779e-06,
"loss": 2.2738,
"step": 1046
},
{
"epoch": 1.0531212673665682,
"grad_norm": 19.783838723215364,
"learning_rate": 8.772635814889337e-06,
"loss": 2.3063,
"step": 1047
},
{
"epoch": 1.0541271138492487,
"grad_norm": 18.523579033127074,
"learning_rate": 8.770400178850885e-06,
"loss": 2.3036,
"step": 1048
},
{
"epoch": 1.0551329603319293,
"grad_norm": 15.720657777619268,
"learning_rate": 8.768164542812431e-06,
"loss": 2.3232,
"step": 1049
},
{
"epoch": 1.05613880681461,
"grad_norm": 18.90213796026752,
"learning_rate": 8.765928906773978e-06,
"loss": 2.2807,
"step": 1050
},
{
"epoch": 1.0571446532972906,
"grad_norm": 17.276748437659204,
"learning_rate": 8.763693270735524e-06,
"loss": 2.3144,
"step": 1051
},
{
"epoch": 1.058150499779971,
"grad_norm": 17.070528576683216,
"learning_rate": 8.761457634697073e-06,
"loss": 2.3109,
"step": 1052
},
{
"epoch": 1.0591563462626516,
"grad_norm": 14.745669379798551,
"learning_rate": 8.75922199865862e-06,
"loss": 2.324,
"step": 1053
},
{
"epoch": 1.0601621927453322,
"grad_norm": 17.757207171938163,
"learning_rate": 8.756986362620166e-06,
"loss": 2.2951,
"step": 1054
},
{
"epoch": 1.0611680392280127,
"grad_norm": 16.553773697289195,
"learning_rate": 8.754750726581714e-06,
"loss": 2.3451,
"step": 1055
},
{
"epoch": 1.0621738857106935,
"grad_norm": 18.039386395828323,
"learning_rate": 8.75251509054326e-06,
"loss": 2.3159,
"step": 1056
},
{
"epoch": 1.063179732193374,
"grad_norm": 17.969055109206348,
"learning_rate": 8.750279454504807e-06,
"loss": 2.2989,
"step": 1057
},
{
"epoch": 1.0641855786760546,
"grad_norm": 16.96062690053515,
"learning_rate": 8.748043818466354e-06,
"loss": 2.3199,
"step": 1058
},
{
"epoch": 1.065191425158735,
"grad_norm": 17.158342934397627,
"learning_rate": 8.745808182427902e-06,
"loss": 2.2504,
"step": 1059
},
{
"epoch": 1.0661972716414156,
"grad_norm": 16.784646638835156,
"learning_rate": 8.743572546389449e-06,
"loss": 2.3059,
"step": 1060
},
{
"epoch": 1.0672031181240964,
"grad_norm": 19.418038825614527,
"learning_rate": 8.741336910350995e-06,
"loss": 2.2612,
"step": 1061
},
{
"epoch": 1.068208964606777,
"grad_norm": 17.538394248187252,
"learning_rate": 8.739101274312542e-06,
"loss": 2.2833,
"step": 1062
},
{
"epoch": 1.0692148110894575,
"grad_norm": 17.040190678679068,
"learning_rate": 8.73686563827409e-06,
"loss": 2.2854,
"step": 1063
},
{
"epoch": 1.070220657572138,
"grad_norm": 17.16856813140221,
"learning_rate": 8.734630002235636e-06,
"loss": 2.2879,
"step": 1064
},
{
"epoch": 1.0712265040548186,
"grad_norm": 19.04157743002483,
"learning_rate": 8.732394366197183e-06,
"loss": 2.3107,
"step": 1065
},
{
"epoch": 1.072232350537499,
"grad_norm": 16.92014717159457,
"learning_rate": 8.730158730158731e-06,
"loss": 2.2602,
"step": 1066
},
{
"epoch": 1.0732381970201799,
"grad_norm": 17.33657041438704,
"learning_rate": 8.727923094120278e-06,
"loss": 2.2437,
"step": 1067
},
{
"epoch": 1.0742440435028604,
"grad_norm": 16.117892266244965,
"learning_rate": 8.725687458081824e-06,
"loss": 2.2922,
"step": 1068
},
{
"epoch": 1.075249889985541,
"grad_norm": 18.39743693476391,
"learning_rate": 8.723451822043371e-06,
"loss": 2.243,
"step": 1069
},
{
"epoch": 1.0762557364682215,
"grad_norm": 17.51753883567021,
"learning_rate": 8.72121618600492e-06,
"loss": 2.257,
"step": 1070
},
{
"epoch": 1.077261582950902,
"grad_norm": 19.177939170772685,
"learning_rate": 8.718980549966466e-06,
"loss": 2.3098,
"step": 1071
},
{
"epoch": 1.0782674294335828,
"grad_norm": 17.40174902530237,
"learning_rate": 8.716744913928014e-06,
"loss": 2.2857,
"step": 1072
},
{
"epoch": 1.0792732759162633,
"grad_norm": 20.080758725973578,
"learning_rate": 8.71450927788956e-06,
"loss": 2.2396,
"step": 1073
},
{
"epoch": 1.0802791223989439,
"grad_norm": 19.41196881225004,
"learning_rate": 8.712273641851107e-06,
"loss": 2.2848,
"step": 1074
},
{
"epoch": 1.0812849688816244,
"grad_norm": 18.2994680792408,
"learning_rate": 8.710038005812654e-06,
"loss": 2.2964,
"step": 1075
},
{
"epoch": 1.082290815364305,
"grad_norm": 17.693788846108696,
"learning_rate": 8.707802369774202e-06,
"loss": 2.2845,
"step": 1076
},
{
"epoch": 1.0832966618469857,
"grad_norm": 17.568965785873733,
"learning_rate": 8.705566733735749e-06,
"loss": 2.3162,
"step": 1077
},
{
"epoch": 1.0843025083296662,
"grad_norm": 17.32480439026697,
"learning_rate": 8.703331097697295e-06,
"loss": 2.2887,
"step": 1078
},
{
"epoch": 1.0853083548123468,
"grad_norm": 17.40755024517668,
"learning_rate": 8.701095461658843e-06,
"loss": 2.2616,
"step": 1079
},
{
"epoch": 1.0863142012950273,
"grad_norm": 17.507757383473695,
"learning_rate": 8.69885982562039e-06,
"loss": 2.2929,
"step": 1080
},
{
"epoch": 1.0873200477777079,
"grad_norm": 16.714998883542382,
"learning_rate": 8.696624189581936e-06,
"loss": 2.3106,
"step": 1081
},
{
"epoch": 1.0883258942603886,
"grad_norm": 17.82449763559636,
"learning_rate": 8.694388553543483e-06,
"loss": 2.3194,
"step": 1082
},
{
"epoch": 1.0893317407430692,
"grad_norm": 17.767082137079235,
"learning_rate": 8.692152917505031e-06,
"loss": 2.2838,
"step": 1083
},
{
"epoch": 1.0903375872257497,
"grad_norm": 18.537351379451447,
"learning_rate": 8.689917281466578e-06,
"loss": 2.2932,
"step": 1084
},
{
"epoch": 1.0913434337084302,
"grad_norm": 22.891169562818433,
"learning_rate": 8.687681645428126e-06,
"loss": 2.2773,
"step": 1085
},
{
"epoch": 1.0923492801911108,
"grad_norm": 17.011404027669325,
"learning_rate": 8.685446009389673e-06,
"loss": 2.2595,
"step": 1086
},
{
"epoch": 1.0933551266737913,
"grad_norm": 18.034931522205362,
"learning_rate": 8.68321037335122e-06,
"loss": 2.2591,
"step": 1087
},
{
"epoch": 1.094360973156472,
"grad_norm": 20.060986740107193,
"learning_rate": 8.680974737312766e-06,
"loss": 2.3045,
"step": 1088
},
{
"epoch": 1.0953668196391526,
"grad_norm": 18.99277712623433,
"learning_rate": 8.678739101274312e-06,
"loss": 2.2483,
"step": 1089
},
{
"epoch": 1.0963726661218332,
"grad_norm": 19.532669255280176,
"learning_rate": 8.67650346523586e-06,
"loss": 2.3214,
"step": 1090
},
{
"epoch": 1.0973785126045137,
"grad_norm": 17.28354302513203,
"learning_rate": 8.674267829197407e-06,
"loss": 2.2861,
"step": 1091
},
{
"epoch": 1.0983843590871942,
"grad_norm": 18.82097945985951,
"learning_rate": 8.672032193158955e-06,
"loss": 2.2928,
"step": 1092
},
{
"epoch": 1.099390205569875,
"grad_norm": 19.849603926806605,
"learning_rate": 8.669796557120502e-06,
"loss": 2.2887,
"step": 1093
},
{
"epoch": 1.1003960520525555,
"grad_norm": 17.784606582580825,
"learning_rate": 8.667560921082049e-06,
"loss": 2.2639,
"step": 1094
},
{
"epoch": 1.101401898535236,
"grad_norm": 23.615341983574755,
"learning_rate": 8.665325285043595e-06,
"loss": 2.2542,
"step": 1095
},
{
"epoch": 1.1024077450179166,
"grad_norm": 24.102960822027146,
"learning_rate": 8.663089649005143e-06,
"loss": 2.282,
"step": 1096
},
{
"epoch": 1.1034135915005971,
"grad_norm": 19.590423010500995,
"learning_rate": 8.66085401296669e-06,
"loss": 2.2569,
"step": 1097
},
{
"epoch": 1.1044194379832777,
"grad_norm": 19.718244556833877,
"learning_rate": 8.658618376928236e-06,
"loss": 2.3158,
"step": 1098
},
{
"epoch": 1.1054252844659584,
"grad_norm": 20.811154993006742,
"learning_rate": 8.656382740889783e-06,
"loss": 2.2856,
"step": 1099
},
{
"epoch": 1.106431130948639,
"grad_norm": 20.793341346867432,
"learning_rate": 8.65414710485133e-06,
"loss": 2.3054,
"step": 1100
},
{
"epoch": 1.1074369774313195,
"grad_norm": 15.838546783941315,
"learning_rate": 8.651911468812878e-06,
"loss": 2.2624,
"step": 1101
},
{
"epoch": 1.108442823914,
"grad_norm": 17.47078271491607,
"learning_rate": 8.649675832774424e-06,
"loss": 2.3208,
"step": 1102
},
{
"epoch": 1.1094486703966806,
"grad_norm": 18.48848617655423,
"learning_rate": 8.647440196735973e-06,
"loss": 2.3271,
"step": 1103
},
{
"epoch": 1.1104545168793614,
"grad_norm": 17.91893806068612,
"learning_rate": 8.64520456069752e-06,
"loss": 2.3012,
"step": 1104
},
{
"epoch": 1.111460363362042,
"grad_norm": 17.987110558744828,
"learning_rate": 8.642968924659066e-06,
"loss": 2.2325,
"step": 1105
},
{
"epoch": 1.1124662098447224,
"grad_norm": 20.347612749596323,
"learning_rate": 8.640733288620612e-06,
"loss": 2.2983,
"step": 1106
},
{
"epoch": 1.113472056327403,
"grad_norm": 16.31907596745269,
"learning_rate": 8.63849765258216e-06,
"loss": 2.2954,
"step": 1107
},
{
"epoch": 1.1144779028100835,
"grad_norm": 22.37021243709992,
"learning_rate": 8.636262016543707e-06,
"loss": 2.3291,
"step": 1108
},
{
"epoch": 1.1154837492927643,
"grad_norm": 19.44131371349581,
"learning_rate": 8.634026380505255e-06,
"loss": 2.286,
"step": 1109
},
{
"epoch": 1.1164895957754448,
"grad_norm": 19.948588951628864,
"learning_rate": 8.631790744466802e-06,
"loss": 2.3175,
"step": 1110
},
{
"epoch": 1.1174954422581254,
"grad_norm": 21.1765882245939,
"learning_rate": 8.629555108428349e-06,
"loss": 2.2921,
"step": 1111
},
{
"epoch": 1.118501288740806,
"grad_norm": 21.7341138660393,
"learning_rate": 8.627319472389895e-06,
"loss": 2.2662,
"step": 1112
},
{
"epoch": 1.1195071352234864,
"grad_norm": 22.332431460942793,
"learning_rate": 8.625083836351442e-06,
"loss": 2.3016,
"step": 1113
},
{
"epoch": 1.1205129817061672,
"grad_norm": 18.438975206999558,
"learning_rate": 8.62284820031299e-06,
"loss": 2.2919,
"step": 1114
},
{
"epoch": 1.1215188281888477,
"grad_norm": 23.38811510430387,
"learning_rate": 8.620612564274536e-06,
"loss": 2.2608,
"step": 1115
},
{
"epoch": 1.1225246746715283,
"grad_norm": 16.69490854261764,
"learning_rate": 8.618376928236085e-06,
"loss": 2.3022,
"step": 1116
},
{
"epoch": 1.1235305211542088,
"grad_norm": 21.455740370412343,
"learning_rate": 8.616141292197631e-06,
"loss": 2.2514,
"step": 1117
},
{
"epoch": 1.1245363676368894,
"grad_norm": 19.19448951730175,
"learning_rate": 8.613905656159178e-06,
"loss": 2.2973,
"step": 1118
},
{
"epoch": 1.12554221411957,
"grad_norm": 17.753641330646218,
"learning_rate": 8.611670020120724e-06,
"loss": 2.2801,
"step": 1119
},
{
"epoch": 1.1265480606022507,
"grad_norm": 17.94463995944704,
"learning_rate": 8.609434384082273e-06,
"loss": 2.3101,
"step": 1120
},
{
"epoch": 1.1275539070849312,
"grad_norm": 19.05898650124038,
"learning_rate": 8.60719874804382e-06,
"loss": 2.2527,
"step": 1121
},
{
"epoch": 1.1285597535676117,
"grad_norm": 22.476374958104035,
"learning_rate": 8.604963112005367e-06,
"loss": 2.2742,
"step": 1122
},
{
"epoch": 1.1295656000502923,
"grad_norm": 21.12202310587608,
"learning_rate": 8.602727475966914e-06,
"loss": 2.2671,
"step": 1123
},
{
"epoch": 1.1305714465329728,
"grad_norm": 20.158353323317353,
"learning_rate": 8.60049183992846e-06,
"loss": 2.289,
"step": 1124
},
{
"epoch": 1.1315772930156536,
"grad_norm": 20.297204120696264,
"learning_rate": 8.598256203890007e-06,
"loss": 2.3296,
"step": 1125
},
{
"epoch": 1.1325831394983341,
"grad_norm": 20.087346285671085,
"learning_rate": 8.596020567851554e-06,
"loss": 2.3248,
"step": 1126
},
{
"epoch": 1.1335889859810147,
"grad_norm": 19.115449262812746,
"learning_rate": 8.593784931813102e-06,
"loss": 2.3214,
"step": 1127
},
{
"epoch": 1.1345948324636952,
"grad_norm": 23.077884519254027,
"learning_rate": 8.591549295774648e-06,
"loss": 2.2687,
"step": 1128
},
{
"epoch": 1.1356006789463757,
"grad_norm": 17.696502624622784,
"learning_rate": 8.589313659736197e-06,
"loss": 2.2715,
"step": 1129
},
{
"epoch": 1.1366065254290563,
"grad_norm": 21.595594846158107,
"learning_rate": 8.587078023697742e-06,
"loss": 2.309,
"step": 1130
},
{
"epoch": 1.137612371911737,
"grad_norm": 17.836107574605045,
"learning_rate": 8.58484238765929e-06,
"loss": 2.2824,
"step": 1131
},
{
"epoch": 1.1386182183944176,
"grad_norm": 17.17116187740332,
"learning_rate": 8.582606751620836e-06,
"loss": 2.2881,
"step": 1132
},
{
"epoch": 1.139624064877098,
"grad_norm": 20.83470927331054,
"learning_rate": 8.580371115582385e-06,
"loss": 2.2674,
"step": 1133
},
{
"epoch": 1.1406299113597786,
"grad_norm": 17.37662446021717,
"learning_rate": 8.578135479543931e-06,
"loss": 2.407,
"step": 1134
},
{
"epoch": 1.1416357578424594,
"grad_norm": 18.80680126914326,
"learning_rate": 8.575899843505478e-06,
"loss": 2.3063,
"step": 1135
},
{
"epoch": 1.14264160432514,
"grad_norm": 24.64418636365614,
"learning_rate": 8.573664207467024e-06,
"loss": 2.3152,
"step": 1136
},
{
"epoch": 1.1436474508078205,
"grad_norm": 22.46054687463966,
"learning_rate": 8.571428571428571e-06,
"loss": 2.2916,
"step": 1137
},
{
"epoch": 1.144653297290501,
"grad_norm": 18.10782918295246,
"learning_rate": 8.56919293539012e-06,
"loss": 2.3343,
"step": 1138
},
{
"epoch": 1.1456591437731816,
"grad_norm": 21.905910472556673,
"learning_rate": 8.566957299351666e-06,
"loss": 2.335,
"step": 1139
},
{
"epoch": 1.146664990255862,
"grad_norm": 18.97800330506959,
"learning_rate": 8.564721663313214e-06,
"loss": 2.2896,
"step": 1140
},
{
"epoch": 1.1476708367385429,
"grad_norm": 17.11716507880201,
"learning_rate": 8.56248602727476e-06,
"loss": 2.3289,
"step": 1141
},
{
"epoch": 1.1486766832212234,
"grad_norm": 19.833371608515453,
"learning_rate": 8.560250391236307e-06,
"loss": 2.2488,
"step": 1142
},
{
"epoch": 1.149682529703904,
"grad_norm": 18.23043329147317,
"learning_rate": 8.558014755197854e-06,
"loss": 2.3525,
"step": 1143
},
{
"epoch": 1.1506883761865845,
"grad_norm": 18.5255330515471,
"learning_rate": 8.555779119159402e-06,
"loss": 2.2927,
"step": 1144
},
{
"epoch": 1.151694222669265,
"grad_norm": 17.61637600476979,
"learning_rate": 8.553543483120948e-06,
"loss": 2.2995,
"step": 1145
},
{
"epoch": 1.1527000691519458,
"grad_norm": 20.086175588883503,
"learning_rate": 8.551307847082497e-06,
"loss": 2.284,
"step": 1146
},
{
"epoch": 1.1537059156346263,
"grad_norm": 18.77432452018519,
"learning_rate": 8.549072211044043e-06,
"loss": 2.3082,
"step": 1147
},
{
"epoch": 1.1547117621173069,
"grad_norm": 17.008905022341153,
"learning_rate": 8.54683657500559e-06,
"loss": 2.2956,
"step": 1148
},
{
"epoch": 1.1557176085999874,
"grad_norm": 19.103735106985567,
"learning_rate": 8.544600938967136e-06,
"loss": 2.3223,
"step": 1149
},
{
"epoch": 1.156723455082668,
"grad_norm": 16.68086748430005,
"learning_rate": 8.542365302928683e-06,
"loss": 2.2468,
"step": 1150
},
{
"epoch": 1.1577293015653485,
"grad_norm": 16.688314349646213,
"learning_rate": 8.540129666890231e-06,
"loss": 2.2702,
"step": 1151
},
{
"epoch": 1.1587351480480292,
"grad_norm": 19.553630980497505,
"learning_rate": 8.537894030851778e-06,
"loss": 2.3577,
"step": 1152
},
{
"epoch": 1.1597409945307098,
"grad_norm": 17.578996133346863,
"learning_rate": 8.535658394813326e-06,
"loss": 2.3081,
"step": 1153
},
{
"epoch": 1.1607468410133903,
"grad_norm": 15.714436666924248,
"learning_rate": 8.533422758774873e-06,
"loss": 2.3318,
"step": 1154
},
{
"epoch": 1.1617526874960709,
"grad_norm": 19.104515106493235,
"learning_rate": 8.53118712273642e-06,
"loss": 2.2628,
"step": 1155
},
{
"epoch": 1.1627585339787514,
"grad_norm": 17.380121847851086,
"learning_rate": 8.528951486697966e-06,
"loss": 2.3115,
"step": 1156
},
{
"epoch": 1.1637643804614322,
"grad_norm": 19.36741749113208,
"learning_rate": 8.526715850659514e-06,
"loss": 2.3205,
"step": 1157
},
{
"epoch": 1.1647702269441127,
"grad_norm": 17.21266356569488,
"learning_rate": 8.52448021462106e-06,
"loss": 2.2991,
"step": 1158
},
{
"epoch": 1.1657760734267932,
"grad_norm": 17.08889702038637,
"learning_rate": 8.522244578582607e-06,
"loss": 2.2846,
"step": 1159
},
{
"epoch": 1.1667819199094738,
"grad_norm": 18.747684738568807,
"learning_rate": 8.520008942544155e-06,
"loss": 2.2952,
"step": 1160
},
{
"epoch": 1.1677877663921543,
"grad_norm": 16.235071815437617,
"learning_rate": 8.517773306505702e-06,
"loss": 2.2974,
"step": 1161
},
{
"epoch": 1.168793612874835,
"grad_norm": 16.03374356738149,
"learning_rate": 8.515537670467248e-06,
"loss": 2.3409,
"step": 1162
},
{
"epoch": 1.1697994593575156,
"grad_norm": 17.165622436072415,
"learning_rate": 8.513302034428795e-06,
"loss": 2.3109,
"step": 1163
},
{
"epoch": 1.1708053058401962,
"grad_norm": 16.091528197994307,
"learning_rate": 8.511066398390343e-06,
"loss": 2.2301,
"step": 1164
},
{
"epoch": 1.1718111523228767,
"grad_norm": 18.69712187447156,
"learning_rate": 8.50883076235189e-06,
"loss": 2.2939,
"step": 1165
},
{
"epoch": 1.1728169988055572,
"grad_norm": 18.730608245623188,
"learning_rate": 8.506595126313436e-06,
"loss": 2.3378,
"step": 1166
},
{
"epoch": 1.173822845288238,
"grad_norm": 17.226123777170987,
"learning_rate": 8.504359490274983e-06,
"loss": 2.27,
"step": 1167
},
{
"epoch": 1.1748286917709185,
"grad_norm": 17.729146331886696,
"learning_rate": 8.502123854236531e-06,
"loss": 2.3039,
"step": 1168
},
{
"epoch": 1.175834538253599,
"grad_norm": 20.300657659255517,
"learning_rate": 8.499888218198078e-06,
"loss": 2.2737,
"step": 1169
},
{
"epoch": 1.1768403847362796,
"grad_norm": 17.202357183552945,
"learning_rate": 8.497652582159626e-06,
"loss": 2.2941,
"step": 1170
},
{
"epoch": 1.1778462312189601,
"grad_norm": 18.644941015288847,
"learning_rate": 8.495416946121173e-06,
"loss": 2.2561,
"step": 1171
},
{
"epoch": 1.1788520777016407,
"grad_norm": 21.554516823425732,
"learning_rate": 8.493181310082719e-06,
"loss": 2.3076,
"step": 1172
},
{
"epoch": 1.1798579241843214,
"grad_norm": 21.910009042782725,
"learning_rate": 8.490945674044266e-06,
"loss": 2.2471,
"step": 1173
},
{
"epoch": 1.180863770667002,
"grad_norm": 19.125671110079075,
"learning_rate": 8.488710038005812e-06,
"loss": 2.2931,
"step": 1174
},
{
"epoch": 1.1818696171496825,
"grad_norm": 21.123339715589903,
"learning_rate": 8.48647440196736e-06,
"loss": 2.2847,
"step": 1175
},
{
"epoch": 1.182875463632363,
"grad_norm": 20.86724409603151,
"learning_rate": 8.484238765928907e-06,
"loss": 2.2892,
"step": 1176
},
{
"epoch": 1.1838813101150436,
"grad_norm": 17.136272592006435,
"learning_rate": 8.482003129890455e-06,
"loss": 2.2977,
"step": 1177
},
{
"epoch": 1.1848871565977244,
"grad_norm": 22.212885659542902,
"learning_rate": 8.479767493852002e-06,
"loss": 2.2907,
"step": 1178
},
{
"epoch": 1.185893003080405,
"grad_norm": 20.58646373096641,
"learning_rate": 8.477531857813548e-06,
"loss": 2.3163,
"step": 1179
},
{
"epoch": 1.1868988495630854,
"grad_norm": 17.647532371256542,
"learning_rate": 8.475296221775095e-06,
"loss": 2.3357,
"step": 1180
},
{
"epoch": 1.187904696045766,
"grad_norm": 20.60417039306096,
"learning_rate": 8.473060585736643e-06,
"loss": 2.294,
"step": 1181
},
{
"epoch": 1.1889105425284465,
"grad_norm": 19.29806031230101,
"learning_rate": 8.47082494969819e-06,
"loss": 2.2862,
"step": 1182
},
{
"epoch": 1.189916389011127,
"grad_norm": 16.151911685974117,
"learning_rate": 8.468589313659736e-06,
"loss": 2.2804,
"step": 1183
},
{
"epoch": 1.1909222354938078,
"grad_norm": 21.45097022316625,
"learning_rate": 8.466353677621285e-06,
"loss": 2.2598,
"step": 1184
},
{
"epoch": 1.1919280819764884,
"grad_norm": 19.395288120502105,
"learning_rate": 8.464118041582831e-06,
"loss": 2.2408,
"step": 1185
},
{
"epoch": 1.192933928459169,
"grad_norm": 15.936274586244203,
"learning_rate": 8.461882405544378e-06,
"loss": 2.2585,
"step": 1186
},
{
"epoch": 1.1939397749418494,
"grad_norm": 19.374911345804087,
"learning_rate": 8.459646769505924e-06,
"loss": 2.3039,
"step": 1187
},
{
"epoch": 1.1949456214245302,
"grad_norm": 19.076779151255614,
"learning_rate": 8.457411133467473e-06,
"loss": 2.2606,
"step": 1188
},
{
"epoch": 1.1959514679072107,
"grad_norm": 17.225837525556503,
"learning_rate": 8.455175497429019e-06,
"loss": 2.2772,
"step": 1189
},
{
"epoch": 1.1969573143898913,
"grad_norm": 19.275586433570965,
"learning_rate": 8.452939861390567e-06,
"loss": 2.2688,
"step": 1190
},
{
"epoch": 1.1979631608725718,
"grad_norm": 17.44383544746351,
"learning_rate": 8.450704225352114e-06,
"loss": 2.3338,
"step": 1191
},
{
"epoch": 1.1989690073552524,
"grad_norm": 17.17032187743969,
"learning_rate": 8.44846858931366e-06,
"loss": 2.3047,
"step": 1192
},
{
"epoch": 1.199974853837933,
"grad_norm": 19.481154732160245,
"learning_rate": 8.446232953275207e-06,
"loss": 2.2649,
"step": 1193
},
{
"epoch": 1.2009807003206137,
"grad_norm": 16.256685811166456,
"learning_rate": 8.443997317236755e-06,
"loss": 2.271,
"step": 1194
},
{
"epoch": 1.2019865468032942,
"grad_norm": 20.98511432429012,
"learning_rate": 8.441761681198302e-06,
"loss": 2.3088,
"step": 1195
},
{
"epoch": 1.2029923932859747,
"grad_norm": 18.56499615168342,
"learning_rate": 8.439526045159848e-06,
"loss": 2.2783,
"step": 1196
},
{
"epoch": 1.2039982397686553,
"grad_norm": 18.22411495162368,
"learning_rate": 8.437290409121397e-06,
"loss": 2.3375,
"step": 1197
},
{
"epoch": 1.2050040862513358,
"grad_norm": 17.292427441101047,
"learning_rate": 8.435054773082943e-06,
"loss": 2.2597,
"step": 1198
},
{
"epoch": 1.2060099327340166,
"grad_norm": 16.76514795076355,
"learning_rate": 8.43281913704449e-06,
"loss": 2.3185,
"step": 1199
},
{
"epoch": 1.2070157792166971,
"grad_norm": 17.189964690173653,
"learning_rate": 8.430583501006036e-06,
"loss": 2.2584,
"step": 1200
},
{
"epoch": 1.2080216256993777,
"grad_norm": 17.145210093973088,
"learning_rate": 8.428347864967585e-06,
"loss": 2.277,
"step": 1201
},
{
"epoch": 1.2090274721820582,
"grad_norm": 17.108566452416703,
"learning_rate": 8.426112228929131e-06,
"loss": 2.3085,
"step": 1202
},
{
"epoch": 1.2100333186647387,
"grad_norm": 17.953870673763248,
"learning_rate": 8.423876592890678e-06,
"loss": 2.3264,
"step": 1203
},
{
"epoch": 1.2110391651474193,
"grad_norm": 16.519478559734274,
"learning_rate": 8.421640956852224e-06,
"loss": 2.2744,
"step": 1204
},
{
"epoch": 1.2120450116301,
"grad_norm": 16.011151811895708,
"learning_rate": 8.419405320813773e-06,
"loss": 2.2456,
"step": 1205
},
{
"epoch": 1.2130508581127806,
"grad_norm": 18.155498400723467,
"learning_rate": 8.417169684775319e-06,
"loss": 2.3052,
"step": 1206
},
{
"epoch": 1.2140567045954611,
"grad_norm": 17.308785337003986,
"learning_rate": 8.414934048736866e-06,
"loss": 2.252,
"step": 1207
},
{
"epoch": 1.2150625510781417,
"grad_norm": 18.73502765302484,
"learning_rate": 8.412698412698414e-06,
"loss": 2.3404,
"step": 1208
},
{
"epoch": 1.2160683975608222,
"grad_norm": 18.73042774767471,
"learning_rate": 8.41046277665996e-06,
"loss": 2.298,
"step": 1209
},
{
"epoch": 1.217074244043503,
"grad_norm": 21.319913888122183,
"learning_rate": 8.408227140621507e-06,
"loss": 2.3036,
"step": 1210
},
{
"epoch": 1.2180800905261835,
"grad_norm": 19.28450061796885,
"learning_rate": 8.405991504583054e-06,
"loss": 2.2989,
"step": 1211
},
{
"epoch": 1.219085937008864,
"grad_norm": 19.759425495511334,
"learning_rate": 8.403755868544602e-06,
"loss": 2.3084,
"step": 1212
},
{
"epoch": 1.2200917834915446,
"grad_norm": 18.4648546273213,
"learning_rate": 8.401520232506148e-06,
"loss": 2.252,
"step": 1213
},
{
"epoch": 1.221097629974225,
"grad_norm": 22.537770266805197,
"learning_rate": 8.399284596467697e-06,
"loss": 2.3585,
"step": 1214
},
{
"epoch": 1.2221034764569056,
"grad_norm": 19.213301457694612,
"learning_rate": 8.397048960429243e-06,
"loss": 2.3562,
"step": 1215
},
{
"epoch": 1.2231093229395864,
"grad_norm": 16.149206566975227,
"learning_rate": 8.39481332439079e-06,
"loss": 2.2679,
"step": 1216
},
{
"epoch": 1.224115169422267,
"grad_norm": 18.60654610752786,
"learning_rate": 8.392577688352336e-06,
"loss": 2.3134,
"step": 1217
},
{
"epoch": 1.2251210159049475,
"grad_norm": 17.255750821902858,
"learning_rate": 8.390342052313883e-06,
"loss": 2.2861,
"step": 1218
},
{
"epoch": 1.226126862387628,
"grad_norm": 19.432852436736177,
"learning_rate": 8.388106416275431e-06,
"loss": 2.3494,
"step": 1219
},
{
"epoch": 1.2271327088703088,
"grad_norm": 16.724437009029767,
"learning_rate": 8.385870780236978e-06,
"loss": 2.3124,
"step": 1220
},
{
"epoch": 1.2281385553529893,
"grad_norm": 18.018889584155048,
"learning_rate": 8.383635144198526e-06,
"loss": 2.2987,
"step": 1221
},
{
"epoch": 1.2291444018356699,
"grad_norm": 19.397694587440313,
"learning_rate": 8.381399508160073e-06,
"loss": 2.3054,
"step": 1222
},
{
"epoch": 1.2301502483183504,
"grad_norm": 20.03040184726134,
"learning_rate": 8.379163872121619e-06,
"loss": 2.3023,
"step": 1223
},
{
"epoch": 1.231156094801031,
"grad_norm": 16.056037010914075,
"learning_rate": 8.376928236083166e-06,
"loss": 2.305,
"step": 1224
},
{
"epoch": 1.2321619412837115,
"grad_norm": 16.191308119121445,
"learning_rate": 8.374692600044714e-06,
"loss": 2.2859,
"step": 1225
},
{
"epoch": 1.2331677877663922,
"grad_norm": 17.60184761068015,
"learning_rate": 8.37245696400626e-06,
"loss": 2.2909,
"step": 1226
},
{
"epoch": 1.2341736342490728,
"grad_norm": 17.947870646096533,
"learning_rate": 8.370221327967809e-06,
"loss": 2.3172,
"step": 1227
},
{
"epoch": 1.2351794807317533,
"grad_norm": 16.50472984871475,
"learning_rate": 8.367985691929355e-06,
"loss": 2.3022,
"step": 1228
},
{
"epoch": 1.2361853272144339,
"grad_norm": 19.25983526255059,
"learning_rate": 8.365750055890902e-06,
"loss": 2.3106,
"step": 1229
},
{
"epoch": 1.2371911736971144,
"grad_norm": 19.60196250915634,
"learning_rate": 8.363514419852448e-06,
"loss": 2.2812,
"step": 1230
},
{
"epoch": 1.2381970201797952,
"grad_norm": 16.96407219723983,
"learning_rate": 8.361278783813995e-06,
"loss": 2.282,
"step": 1231
},
{
"epoch": 1.2392028666624757,
"grad_norm": 18.456480602482316,
"learning_rate": 8.359043147775543e-06,
"loss": 2.2883,
"step": 1232
},
{
"epoch": 1.2402087131451562,
"grad_norm": 20.712291216636736,
"learning_rate": 8.35680751173709e-06,
"loss": 2.3425,
"step": 1233
},
{
"epoch": 1.2412145596278368,
"grad_norm": 17.536161583264953,
"learning_rate": 8.354571875698636e-06,
"loss": 2.3433,
"step": 1234
},
{
"epoch": 1.2422204061105173,
"grad_norm": 17.866349249921104,
"learning_rate": 8.352336239660183e-06,
"loss": 2.2708,
"step": 1235
},
{
"epoch": 1.2432262525931979,
"grad_norm": 16.304401450446928,
"learning_rate": 8.350100603621731e-06,
"loss": 2.32,
"step": 1236
},
{
"epoch": 1.2442320990758786,
"grad_norm": 22.642132927378558,
"learning_rate": 8.347864967583278e-06,
"loss": 2.295,
"step": 1237
},
{
"epoch": 1.2452379455585592,
"grad_norm": 18.725267470668236,
"learning_rate": 8.345629331544826e-06,
"loss": 2.2752,
"step": 1238
},
{
"epoch": 1.2462437920412397,
"grad_norm": 16.894777761669875,
"learning_rate": 8.343393695506373e-06,
"loss": 2.2637,
"step": 1239
},
{
"epoch": 1.2472496385239202,
"grad_norm": 17.51083240524946,
"learning_rate": 8.341158059467919e-06,
"loss": 2.3437,
"step": 1240
},
{
"epoch": 1.2482554850066008,
"grad_norm": 16.689242298902055,
"learning_rate": 8.338922423429466e-06,
"loss": 2.275,
"step": 1241
},
{
"epoch": 1.2492613314892815,
"grad_norm": 17.72606774782188,
"learning_rate": 8.336686787391012e-06,
"loss": 2.3099,
"step": 1242
},
{
"epoch": 1.250267177971962,
"grad_norm": 18.255072739724646,
"learning_rate": 8.33445115135256e-06,
"loss": 2.3232,
"step": 1243
},
{
"epoch": 1.2512730244546426,
"grad_norm": 22.898243346681394,
"learning_rate": 8.332215515314107e-06,
"loss": 2.2872,
"step": 1244
},
{
"epoch": 1.2522788709373232,
"grad_norm": 17.303411031287375,
"learning_rate": 8.329979879275655e-06,
"loss": 2.3112,
"step": 1245
},
{
"epoch": 1.2532847174200037,
"grad_norm": 23.83708370176368,
"learning_rate": 8.327744243237202e-06,
"loss": 2.3457,
"step": 1246
},
{
"epoch": 1.2542905639026842,
"grad_norm": 22.44312354316094,
"learning_rate": 8.325508607198748e-06,
"loss": 2.2677,
"step": 1247
},
{
"epoch": 1.255296410385365,
"grad_norm": 20.673549066823906,
"learning_rate": 8.323272971160295e-06,
"loss": 2.3068,
"step": 1248
},
{
"epoch": 1.2563022568680455,
"grad_norm": 23.828249133774097,
"learning_rate": 8.321037335121843e-06,
"loss": 2.3217,
"step": 1249
},
{
"epoch": 1.257308103350726,
"grad_norm": 17.021661742629366,
"learning_rate": 8.31880169908339e-06,
"loss": 2.3293,
"step": 1250
},
{
"epoch": 1.2583139498334066,
"grad_norm": 24.068454583708398,
"learning_rate": 8.316566063044938e-06,
"loss": 2.2811,
"step": 1251
},
{
"epoch": 1.2593197963160874,
"grad_norm": 19.943471018011905,
"learning_rate": 8.314330427006485e-06,
"loss": 2.2987,
"step": 1252
},
{
"epoch": 1.260325642798768,
"grad_norm": 18.75794377417748,
"learning_rate": 8.312094790968031e-06,
"loss": 2.2931,
"step": 1253
},
{
"epoch": 1.2613314892814484,
"grad_norm": 24.490075886483915,
"learning_rate": 8.309859154929578e-06,
"loss": 2.2655,
"step": 1254
},
{
"epoch": 1.262337335764129,
"grad_norm": 18.731052471036953,
"learning_rate": 8.307623518891124e-06,
"loss": 2.3088,
"step": 1255
},
{
"epoch": 1.2633431822468095,
"grad_norm": 18.791512493395082,
"learning_rate": 8.305387882852673e-06,
"loss": 2.2572,
"step": 1256
},
{
"epoch": 1.26434902872949,
"grad_norm": 19.889671056381214,
"learning_rate": 8.303152246814219e-06,
"loss": 2.2602,
"step": 1257
},
{
"epoch": 1.2653548752121706,
"grad_norm": 21.002808896258866,
"learning_rate": 8.300916610775767e-06,
"loss": 2.335,
"step": 1258
},
{
"epoch": 1.2663607216948514,
"grad_norm": 20.225422380521106,
"learning_rate": 8.298680974737314e-06,
"loss": 2.2879,
"step": 1259
},
{
"epoch": 1.267366568177532,
"grad_norm": 23.050568206596104,
"learning_rate": 8.29644533869886e-06,
"loss": 2.2804,
"step": 1260
},
{
"epoch": 1.2683724146602124,
"grad_norm": 18.593076086047805,
"learning_rate": 8.294209702660407e-06,
"loss": 2.2453,
"step": 1261
},
{
"epoch": 1.269378261142893,
"grad_norm": 22.977584836766557,
"learning_rate": 8.291974066621955e-06,
"loss": 2.2711,
"step": 1262
},
{
"epoch": 1.2703841076255737,
"grad_norm": 20.608645625472292,
"learning_rate": 8.289738430583502e-06,
"loss": 2.2741,
"step": 1263
},
{
"epoch": 1.2713899541082543,
"grad_norm": 18.303856024466825,
"learning_rate": 8.28750279454505e-06,
"loss": 2.3507,
"step": 1264
},
{
"epoch": 1.2723958005909348,
"grad_norm": 19.347928838050038,
"learning_rate": 8.285267158506597e-06,
"loss": 2.2831,
"step": 1265
},
{
"epoch": 1.2734016470736154,
"grad_norm": 18.219999067252687,
"learning_rate": 8.283031522468143e-06,
"loss": 2.3249,
"step": 1266
},
{
"epoch": 1.274407493556296,
"grad_norm": 20.180292897382635,
"learning_rate": 8.28079588642969e-06,
"loss": 2.2624,
"step": 1267
},
{
"epoch": 1.2754133400389764,
"grad_norm": 15.374220912213122,
"learning_rate": 8.278560250391236e-06,
"loss": 2.2866,
"step": 1268
},
{
"epoch": 1.2764191865216572,
"grad_norm": 23.30607569252281,
"learning_rate": 8.276324614352785e-06,
"loss": 2.2837,
"step": 1269
},
{
"epoch": 1.2774250330043377,
"grad_norm": 17.137243897208975,
"learning_rate": 8.274088978314331e-06,
"loss": 2.3242,
"step": 1270
},
{
"epoch": 1.2784308794870183,
"grad_norm": 20.389855351910253,
"learning_rate": 8.271853342275878e-06,
"loss": 2.2921,
"step": 1271
},
{
"epoch": 1.2794367259696988,
"grad_norm": 20.884450516450137,
"learning_rate": 8.269617706237424e-06,
"loss": 2.3104,
"step": 1272
},
{
"epoch": 1.2804425724523796,
"grad_norm": 18.511079724079504,
"learning_rate": 8.267382070198972e-06,
"loss": 2.3011,
"step": 1273
},
{
"epoch": 1.2814484189350601,
"grad_norm": 18.1304178064368,
"learning_rate": 8.265146434160519e-06,
"loss": 2.2874,
"step": 1274
},
{
"epoch": 1.2824542654177407,
"grad_norm": 19.385152320988567,
"learning_rate": 8.262910798122067e-06,
"loss": 2.291,
"step": 1275
},
{
"epoch": 1.2834601119004212,
"grad_norm": 17.1541489649596,
"learning_rate": 8.260675162083614e-06,
"loss": 2.3131,
"step": 1276
},
{
"epoch": 1.2844659583831017,
"grad_norm": 18.36182480840744,
"learning_rate": 8.25843952604516e-06,
"loss": 2.2889,
"step": 1277
},
{
"epoch": 1.2854718048657823,
"grad_norm": 20.076674814801525,
"learning_rate": 8.256203890006707e-06,
"loss": 2.2674,
"step": 1278
},
{
"epoch": 1.2864776513484628,
"grad_norm": 16.417572435734545,
"learning_rate": 8.253968253968254e-06,
"loss": 2.3015,
"step": 1279
},
{
"epoch": 1.2874834978311436,
"grad_norm": 17.885444314453963,
"learning_rate": 8.251732617929802e-06,
"loss": 2.2964,
"step": 1280
},
{
"epoch": 1.2884893443138241,
"grad_norm": 16.89635113816944,
"learning_rate": 8.249496981891348e-06,
"loss": 2.2685,
"step": 1281
},
{
"epoch": 1.2894951907965047,
"grad_norm": 19.13993347028726,
"learning_rate": 8.247261345852897e-06,
"loss": 2.2531,
"step": 1282
},
{
"epoch": 1.2905010372791852,
"grad_norm": 19.69722538251037,
"learning_rate": 8.245025709814443e-06,
"loss": 2.3154,
"step": 1283
},
{
"epoch": 1.291506883761866,
"grad_norm": 18.408000700271728,
"learning_rate": 8.24279007377599e-06,
"loss": 2.2434,
"step": 1284
},
{
"epoch": 1.2925127302445465,
"grad_norm": 16.97299347011639,
"learning_rate": 8.240554437737536e-06,
"loss": 2.2538,
"step": 1285
},
{
"epoch": 1.293518576727227,
"grad_norm": 18.751891645294307,
"learning_rate": 8.238318801699085e-06,
"loss": 2.2749,
"step": 1286
},
{
"epoch": 1.2945244232099076,
"grad_norm": 18.046756082355557,
"learning_rate": 8.236083165660631e-06,
"loss": 2.2806,
"step": 1287
},
{
"epoch": 1.295530269692588,
"grad_norm": 17.456125883815147,
"learning_rate": 8.23384752962218e-06,
"loss": 2.3182,
"step": 1288
},
{
"epoch": 1.2965361161752686,
"grad_norm": 16.18488358180136,
"learning_rate": 8.231611893583726e-06,
"loss": 2.2946,
"step": 1289
},
{
"epoch": 1.2975419626579494,
"grad_norm": 15.440372250390611,
"learning_rate": 8.229376257545272e-06,
"loss": 2.3041,
"step": 1290
},
{
"epoch": 1.29854780914063,
"grad_norm": 20.329005492708532,
"learning_rate": 8.227140621506819e-06,
"loss": 2.3064,
"step": 1291
},
{
"epoch": 1.2995536556233105,
"grad_norm": 17.776054130149582,
"learning_rate": 8.224904985468366e-06,
"loss": 2.3201,
"step": 1292
},
{
"epoch": 1.300559502105991,
"grad_norm": 16.816280204724453,
"learning_rate": 8.222669349429914e-06,
"loss": 2.3327,
"step": 1293
},
{
"epoch": 1.3015653485886718,
"grad_norm": 17.568973501750424,
"learning_rate": 8.22043371339146e-06,
"loss": 2.2911,
"step": 1294
},
{
"epoch": 1.3025711950713523,
"grad_norm": 18.696170316078923,
"learning_rate": 8.218198077353009e-06,
"loss": 2.3178,
"step": 1295
},
{
"epoch": 1.3035770415540329,
"grad_norm": 17.548421228757203,
"learning_rate": 8.215962441314555e-06,
"loss": 2.2683,
"step": 1296
},
{
"epoch": 1.3045828880367134,
"grad_norm": 16.252199195405446,
"learning_rate": 8.213726805276102e-06,
"loss": 2.3371,
"step": 1297
},
{
"epoch": 1.305588734519394,
"grad_norm": 20.124397384897307,
"learning_rate": 8.211491169237648e-06,
"loss": 2.3059,
"step": 1298
},
{
"epoch": 1.3065945810020745,
"grad_norm": 21.440072268620945,
"learning_rate": 8.209255533199197e-06,
"loss": 2.2513,
"step": 1299
},
{
"epoch": 1.307600427484755,
"grad_norm": 17.556298612821745,
"learning_rate": 8.207019897160743e-06,
"loss": 2.3107,
"step": 1300
},
{
"epoch": 1.3086062739674358,
"grad_norm": 18.785195273395882,
"learning_rate": 8.20478426112229e-06,
"loss": 2.3165,
"step": 1301
},
{
"epoch": 1.3096121204501163,
"grad_norm": 18.79014461793555,
"learning_rate": 8.202548625083836e-06,
"loss": 2.2382,
"step": 1302
},
{
"epoch": 1.3106179669327969,
"grad_norm": 15.792332282171454,
"learning_rate": 8.200312989045383e-06,
"loss": 2.3201,
"step": 1303
},
{
"epoch": 1.3116238134154774,
"grad_norm": 21.41599796197624,
"learning_rate": 8.198077353006931e-06,
"loss": 2.2499,
"step": 1304
},
{
"epoch": 1.3126296598981582,
"grad_norm": 18.563952098598246,
"learning_rate": 8.195841716968478e-06,
"loss": 2.2913,
"step": 1305
},
{
"epoch": 1.3136355063808387,
"grad_norm": 18.28162901925981,
"learning_rate": 8.193606080930026e-06,
"loss": 2.3062,
"step": 1306
},
{
"epoch": 1.3146413528635192,
"grad_norm": 20.42790763943371,
"learning_rate": 8.191370444891572e-06,
"loss": 2.3028,
"step": 1307
},
{
"epoch": 1.3156471993461998,
"grad_norm": 18.648482980230007,
"learning_rate": 8.189134808853119e-06,
"loss": 2.3107,
"step": 1308
},
{
"epoch": 1.3166530458288803,
"grad_norm": 17.67553108046889,
"learning_rate": 8.186899172814666e-06,
"loss": 2.3137,
"step": 1309
},
{
"epoch": 1.3176588923115609,
"grad_norm": 24.623618881587323,
"learning_rate": 8.184663536776214e-06,
"loss": 2.2853,
"step": 1310
},
{
"epoch": 1.3186647387942414,
"grad_norm": 22.602779570132835,
"learning_rate": 8.18242790073776e-06,
"loss": 2.3202,
"step": 1311
},
{
"epoch": 1.3196705852769222,
"grad_norm": 19.920055666801737,
"learning_rate": 8.180192264699309e-06,
"loss": 2.2867,
"step": 1312
},
{
"epoch": 1.3206764317596027,
"grad_norm": 20.698995378939966,
"learning_rate": 8.177956628660855e-06,
"loss": 2.3015,
"step": 1313
},
{
"epoch": 1.3216822782422832,
"grad_norm": 19.49105745285178,
"learning_rate": 8.175720992622402e-06,
"loss": 2.2992,
"step": 1314
},
{
"epoch": 1.3226881247249638,
"grad_norm": 19.980718568169483,
"learning_rate": 8.173485356583948e-06,
"loss": 2.2762,
"step": 1315
},
{
"epoch": 1.3236939712076445,
"grad_norm": 20.994595344065768,
"learning_rate": 8.171249720545495e-06,
"loss": 2.3046,
"step": 1316
},
{
"epoch": 1.324699817690325,
"grad_norm": 18.561189947529662,
"learning_rate": 8.169014084507043e-06,
"loss": 2.2812,
"step": 1317
},
{
"epoch": 1.3257056641730056,
"grad_norm": 20.345641276000574,
"learning_rate": 8.16677844846859e-06,
"loss": 2.2895,
"step": 1318
},
{
"epoch": 1.3267115106556862,
"grad_norm": 20.8726976500452,
"learning_rate": 8.164542812430138e-06,
"loss": 2.3045,
"step": 1319
},
{
"epoch": 1.3277173571383667,
"grad_norm": 17.641123989553606,
"learning_rate": 8.162307176391685e-06,
"loss": 2.2865,
"step": 1320
},
{
"epoch": 1.3287232036210472,
"grad_norm": 18.27076606433099,
"learning_rate": 8.160071540353231e-06,
"loss": 2.3093,
"step": 1321
},
{
"epoch": 1.329729050103728,
"grad_norm": 19.810653599016497,
"learning_rate": 8.157835904314778e-06,
"loss": 2.2917,
"step": 1322
},
{
"epoch": 1.3307348965864085,
"grad_norm": 19.629184686847974,
"learning_rate": 8.155600268276326e-06,
"loss": 2.3546,
"step": 1323
},
{
"epoch": 1.331740743069089,
"grad_norm": 21.380275384171682,
"learning_rate": 8.153364632237872e-06,
"loss": 2.2434,
"step": 1324
},
{
"epoch": 1.3327465895517696,
"grad_norm": 21.66339792232214,
"learning_rate": 8.151128996199419e-06,
"loss": 2.2751,
"step": 1325
},
{
"epoch": 1.3337524360344504,
"grad_norm": 19.021364106448544,
"learning_rate": 8.148893360160967e-06,
"loss": 2.3185,
"step": 1326
},
{
"epoch": 1.334758282517131,
"grad_norm": 19.811645988757654,
"learning_rate": 8.146657724122514e-06,
"loss": 2.3236,
"step": 1327
},
{
"epoch": 1.3357641289998115,
"grad_norm": 18.294063874329982,
"learning_rate": 8.14442208808406e-06,
"loss": 2.3302,
"step": 1328
},
{
"epoch": 1.336769975482492,
"grad_norm": 17.294912142975505,
"learning_rate": 8.142186452045607e-06,
"loss": 2.2985,
"step": 1329
},
{
"epoch": 1.3377758219651725,
"grad_norm": 16.126934879929436,
"learning_rate": 8.139950816007155e-06,
"loss": 2.3015,
"step": 1330
},
{
"epoch": 1.338781668447853,
"grad_norm": 16.16225564905107,
"learning_rate": 8.137715179968702e-06,
"loss": 2.2449,
"step": 1331
},
{
"epoch": 1.3397875149305336,
"grad_norm": 19.658951752756792,
"learning_rate": 8.13547954393025e-06,
"loss": 2.3109,
"step": 1332
},
{
"epoch": 1.3407933614132144,
"grad_norm": 19.389529788632824,
"learning_rate": 8.133243907891797e-06,
"loss": 2.3006,
"step": 1333
},
{
"epoch": 1.341799207895895,
"grad_norm": 19.107880682491288,
"learning_rate": 8.131008271853343e-06,
"loss": 2.2806,
"step": 1334
},
{
"epoch": 1.3428050543785754,
"grad_norm": 18.50200106508036,
"learning_rate": 8.12877263581489e-06,
"loss": 2.2711,
"step": 1335
},
{
"epoch": 1.343810900861256,
"grad_norm": 17.17908472204596,
"learning_rate": 8.126536999776436e-06,
"loss": 2.3178,
"step": 1336
},
{
"epoch": 1.3448167473439367,
"grad_norm": 19.957877071610238,
"learning_rate": 8.124301363737984e-06,
"loss": 2.2612,
"step": 1337
},
{
"epoch": 1.3458225938266173,
"grad_norm": 21.013195814959914,
"learning_rate": 8.122065727699531e-06,
"loss": 2.3239,
"step": 1338
},
{
"epoch": 1.3468284403092978,
"grad_norm": 18.71020075051148,
"learning_rate": 8.119830091661078e-06,
"loss": 2.2599,
"step": 1339
},
{
"epoch": 1.3478342867919784,
"grad_norm": 19.25262006006824,
"learning_rate": 8.117594455622624e-06,
"loss": 2.2847,
"step": 1340
},
{
"epoch": 1.348840133274659,
"grad_norm": 21.55320098193352,
"learning_rate": 8.115358819584172e-06,
"loss": 2.3184,
"step": 1341
},
{
"epoch": 1.3498459797573394,
"grad_norm": 18.3535039370996,
"learning_rate": 8.113123183545719e-06,
"loss": 2.3066,
"step": 1342
},
{
"epoch": 1.35085182624002,
"grad_norm": 18.416153435479284,
"learning_rate": 8.110887547507267e-06,
"loss": 2.2912,
"step": 1343
},
{
"epoch": 1.3518576727227007,
"grad_norm": 15.017770702208837,
"learning_rate": 8.108651911468814e-06,
"loss": 2.332,
"step": 1344
},
{
"epoch": 1.3528635192053813,
"grad_norm": 16.926121576064208,
"learning_rate": 8.10641627543036e-06,
"loss": 2.3573,
"step": 1345
},
{
"epoch": 1.3538693656880618,
"grad_norm": 17.416574528957202,
"learning_rate": 8.104180639391907e-06,
"loss": 2.2611,
"step": 1346
},
{
"epoch": 1.3548752121707424,
"grad_norm": 16.300702078394863,
"learning_rate": 8.101945003353455e-06,
"loss": 2.2578,
"step": 1347
},
{
"epoch": 1.3558810586534231,
"grad_norm": 19.977823623903497,
"learning_rate": 8.099709367315002e-06,
"loss": 2.3109,
"step": 1348
},
{
"epoch": 1.3568869051361037,
"grad_norm": 18.642195493227934,
"learning_rate": 8.097473731276548e-06,
"loss": 2.2904,
"step": 1349
},
{
"epoch": 1.3578927516187842,
"grad_norm": 18.72968581315017,
"learning_rate": 8.095238095238097e-06,
"loss": 2.3308,
"step": 1350
},
{
"epoch": 1.3588985981014647,
"grad_norm": 17.103859568421928,
"learning_rate": 8.093002459199643e-06,
"loss": 2.3178,
"step": 1351
},
{
"epoch": 1.3599044445841453,
"grad_norm": 17.298230445169914,
"learning_rate": 8.09076682316119e-06,
"loss": 2.2466,
"step": 1352
},
{
"epoch": 1.3609102910668258,
"grad_norm": 17.79454611645911,
"learning_rate": 8.088531187122736e-06,
"loss": 2.2856,
"step": 1353
},
{
"epoch": 1.3619161375495066,
"grad_norm": 18.984923023468426,
"learning_rate": 8.086295551084284e-06,
"loss": 2.3291,
"step": 1354
},
{
"epoch": 1.3629219840321871,
"grad_norm": 17.768320917083503,
"learning_rate": 8.084059915045831e-06,
"loss": 2.3978,
"step": 1355
},
{
"epoch": 1.3639278305148677,
"grad_norm": 17.375971136653813,
"learning_rate": 8.08182427900738e-06,
"loss": 2.3041,
"step": 1356
},
{
"epoch": 1.3649336769975482,
"grad_norm": 16.302797771846613,
"learning_rate": 8.079588642968926e-06,
"loss": 2.2716,
"step": 1357
},
{
"epoch": 1.365939523480229,
"grad_norm": 19.4372043283243,
"learning_rate": 8.077353006930472e-06,
"loss": 2.2259,
"step": 1358
},
{
"epoch": 1.3669453699629095,
"grad_norm": 17.519273113608005,
"learning_rate": 8.075117370892019e-06,
"loss": 2.3083,
"step": 1359
},
{
"epoch": 1.36795121644559,
"grad_norm": 18.847836982555016,
"learning_rate": 8.072881734853566e-06,
"loss": 2.2929,
"step": 1360
},
{
"epoch": 1.3689570629282706,
"grad_norm": 18.80960331014159,
"learning_rate": 8.070646098815114e-06,
"loss": 2.344,
"step": 1361
},
{
"epoch": 1.3699629094109511,
"grad_norm": 18.54062907853193,
"learning_rate": 8.06841046277666e-06,
"loss": 2.291,
"step": 1362
},
{
"epoch": 1.3709687558936317,
"grad_norm": 21.020876962064065,
"learning_rate": 8.066174826738209e-06,
"loss": 2.2943,
"step": 1363
},
{
"epoch": 1.3719746023763122,
"grad_norm": 19.634452718735282,
"learning_rate": 8.063939190699755e-06,
"loss": 2.3096,
"step": 1364
},
{
"epoch": 1.372980448858993,
"grad_norm": 23.75241532805294,
"learning_rate": 8.061703554661302e-06,
"loss": 2.2756,
"step": 1365
},
{
"epoch": 1.3739862953416735,
"grad_norm": 23.184088534634178,
"learning_rate": 8.059467918622848e-06,
"loss": 2.3122,
"step": 1366
},
{
"epoch": 1.374992141824354,
"grad_norm": 21.683222338914778,
"learning_rate": 8.057232282584397e-06,
"loss": 2.2555,
"step": 1367
},
{
"epoch": 1.3759979883070346,
"grad_norm": 17.538940009868362,
"learning_rate": 8.054996646545943e-06,
"loss": 2.2821,
"step": 1368
},
{
"epoch": 1.3770038347897153,
"grad_norm": 20.589385488446872,
"learning_rate": 8.05276101050749e-06,
"loss": 2.3005,
"step": 1369
},
{
"epoch": 1.3780096812723959,
"grad_norm": 18.228248922770465,
"learning_rate": 8.050525374469036e-06,
"loss": 2.2467,
"step": 1370
},
{
"epoch": 1.3790155277550764,
"grad_norm": 18.150099944925707,
"learning_rate": 8.048289738430584e-06,
"loss": 2.3107,
"step": 1371
},
{
"epoch": 1.380021374237757,
"grad_norm": 17.600424138374375,
"learning_rate": 8.046054102392131e-06,
"loss": 2.2301,
"step": 1372
},
{
"epoch": 1.3810272207204375,
"grad_norm": 19.973595794510803,
"learning_rate": 8.043818466353678e-06,
"loss": 2.326,
"step": 1373
},
{
"epoch": 1.382033067203118,
"grad_norm": 18.030406121961228,
"learning_rate": 8.041582830315226e-06,
"loss": 2.299,
"step": 1374
},
{
"epoch": 1.3830389136857988,
"grad_norm": 17.130081737007906,
"learning_rate": 8.039347194276772e-06,
"loss": 2.2898,
"step": 1375
},
{
"epoch": 1.3840447601684793,
"grad_norm": 16.71810323163366,
"learning_rate": 8.037111558238319e-06,
"loss": 2.2864,
"step": 1376
},
{
"epoch": 1.3850506066511599,
"grad_norm": 16.76624558405387,
"learning_rate": 8.034875922199866e-06,
"loss": 2.2581,
"step": 1377
},
{
"epoch": 1.3860564531338404,
"grad_norm": 16.552359135070123,
"learning_rate": 8.032640286161414e-06,
"loss": 2.2802,
"step": 1378
},
{
"epoch": 1.3870622996165212,
"grad_norm": 19.929742012690642,
"learning_rate": 8.03040465012296e-06,
"loss": 2.2913,
"step": 1379
},
{
"epoch": 1.3880681460992017,
"grad_norm": 17.513117588876412,
"learning_rate": 8.028169014084509e-06,
"loss": 2.2613,
"step": 1380
},
{
"epoch": 1.3890739925818822,
"grad_norm": 15.22832720058821,
"learning_rate": 8.025933378046055e-06,
"loss": 2.2937,
"step": 1381
},
{
"epoch": 1.3900798390645628,
"grad_norm": 17.218042427984003,
"learning_rate": 8.023697742007602e-06,
"loss": 2.2741,
"step": 1382
},
{
"epoch": 1.3910856855472433,
"grad_norm": 16.58871727809983,
"learning_rate": 8.021462105969148e-06,
"loss": 2.3226,
"step": 1383
},
{
"epoch": 1.3920915320299239,
"grad_norm": 16.06444662928072,
"learning_rate": 8.019226469930695e-06,
"loss": 2.3746,
"step": 1384
},
{
"epoch": 1.3930973785126044,
"grad_norm": 17.70240118364787,
"learning_rate": 8.016990833892243e-06,
"loss": 2.28,
"step": 1385
},
{
"epoch": 1.3941032249952852,
"grad_norm": 19.892962329697724,
"learning_rate": 8.01475519785379e-06,
"loss": 2.2277,
"step": 1386
},
{
"epoch": 1.3951090714779657,
"grad_norm": 17.631779390893282,
"learning_rate": 8.012519561815338e-06,
"loss": 2.2761,
"step": 1387
},
{
"epoch": 1.3961149179606462,
"grad_norm": 16.853865251692838,
"learning_rate": 8.010283925776884e-06,
"loss": 2.3478,
"step": 1388
},
{
"epoch": 1.3971207644433268,
"grad_norm": 17.66032083738247,
"learning_rate": 8.008048289738431e-06,
"loss": 2.2631,
"step": 1389
},
{
"epoch": 1.3981266109260075,
"grad_norm": 17.17318060438319,
"learning_rate": 8.005812653699978e-06,
"loss": 2.3152,
"step": 1390
},
{
"epoch": 1.399132457408688,
"grad_norm": 18.074483339002846,
"learning_rate": 8.003577017661526e-06,
"loss": 2.3181,
"step": 1391
},
{
"epoch": 1.4001383038913686,
"grad_norm": 20.343979364622978,
"learning_rate": 8.001341381623072e-06,
"loss": 2.3134,
"step": 1392
},
{
"epoch": 1.4011441503740492,
"grad_norm": 18.554688640383343,
"learning_rate": 7.99910574558462e-06,
"loss": 2.2422,
"step": 1393
},
{
"epoch": 1.4021499968567297,
"grad_norm": 16.381818752704735,
"learning_rate": 7.996870109546167e-06,
"loss": 2.2659,
"step": 1394
},
{
"epoch": 1.4031558433394102,
"grad_norm": 17.031748772581285,
"learning_rate": 7.994634473507714e-06,
"loss": 2.2616,
"step": 1395
},
{
"epoch": 1.4041616898220908,
"grad_norm": 16.023918356128732,
"learning_rate": 7.99239883746926e-06,
"loss": 2.2754,
"step": 1396
},
{
"epoch": 1.4051675363047715,
"grad_norm": 16.43655076009244,
"learning_rate": 7.990163201430807e-06,
"loss": 2.2738,
"step": 1397
},
{
"epoch": 1.406173382787452,
"grad_norm": 18.001464549575672,
"learning_rate": 7.987927565392355e-06,
"loss": 2.2448,
"step": 1398
},
{
"epoch": 1.4071792292701326,
"grad_norm": 15.548429987816085,
"learning_rate": 7.985691929353902e-06,
"loss": 2.2493,
"step": 1399
},
{
"epoch": 1.4081850757528132,
"grad_norm": 16.789719108176232,
"learning_rate": 7.98345629331545e-06,
"loss": 2.2504,
"step": 1400
},
{
"epoch": 1.409190922235494,
"grad_norm": 17.761135848676386,
"learning_rate": 7.981220657276996e-06,
"loss": 2.3021,
"step": 1401
},
{
"epoch": 1.4101967687181745,
"grad_norm": 16.397080182416595,
"learning_rate": 7.978985021238543e-06,
"loss": 2.2884,
"step": 1402
},
{
"epoch": 1.411202615200855,
"grad_norm": 21.89243360012097,
"learning_rate": 7.97674938520009e-06,
"loss": 2.2874,
"step": 1403
},
{
"epoch": 1.4122084616835355,
"grad_norm": 18.531696638767684,
"learning_rate": 7.974513749161638e-06,
"loss": 2.2476,
"step": 1404
},
{
"epoch": 1.413214308166216,
"grad_norm": 18.655323089753995,
"learning_rate": 7.972278113123184e-06,
"loss": 2.2807,
"step": 1405
},
{
"epoch": 1.4142201546488966,
"grad_norm": 19.63140503292154,
"learning_rate": 7.970042477084731e-06,
"loss": 2.3598,
"step": 1406
},
{
"epoch": 1.4152260011315774,
"grad_norm": 17.511016093681576,
"learning_rate": 7.967806841046278e-06,
"loss": 2.2915,
"step": 1407
},
{
"epoch": 1.416231847614258,
"grad_norm": 17.705733739636496,
"learning_rate": 7.965571205007824e-06,
"loss": 2.2991,
"step": 1408
},
{
"epoch": 1.4172376940969384,
"grad_norm": 17.07262553982362,
"learning_rate": 7.963335568969372e-06,
"loss": 2.2565,
"step": 1409
},
{
"epoch": 1.418243540579619,
"grad_norm": 21.707460747915643,
"learning_rate": 7.961099932930919e-06,
"loss": 2.2577,
"step": 1410
},
{
"epoch": 1.4192493870622997,
"grad_norm": 20.173807801461937,
"learning_rate": 7.958864296892467e-06,
"loss": 2.272,
"step": 1411
},
{
"epoch": 1.4202552335449803,
"grad_norm": 16.95295828244757,
"learning_rate": 7.956628660854014e-06,
"loss": 2.2635,
"step": 1412
},
{
"epoch": 1.4212610800276608,
"grad_norm": 23.519098966208954,
"learning_rate": 7.95439302481556e-06,
"loss": 2.3345,
"step": 1413
},
{
"epoch": 1.4222669265103414,
"grad_norm": 24.33680951481455,
"learning_rate": 7.952157388777107e-06,
"loss": 2.2901,
"step": 1414
},
{
"epoch": 1.423272772993022,
"grad_norm": 19.232446660438725,
"learning_rate": 7.949921752738655e-06,
"loss": 2.2884,
"step": 1415
},
{
"epoch": 1.4242786194757024,
"grad_norm": 22.943036836976418,
"learning_rate": 7.947686116700202e-06,
"loss": 2.3096,
"step": 1416
},
{
"epoch": 1.425284465958383,
"grad_norm": 26.73875559733533,
"learning_rate": 7.94545048066175e-06,
"loss": 2.3043,
"step": 1417
},
{
"epoch": 1.4262903124410637,
"grad_norm": 19.798386270865898,
"learning_rate": 7.943214844623296e-06,
"loss": 2.2578,
"step": 1418
},
{
"epoch": 1.4272961589237443,
"grad_norm": 19.153287956087972,
"learning_rate": 7.940979208584843e-06,
"loss": 2.2363,
"step": 1419
},
{
"epoch": 1.4283020054064248,
"grad_norm": 20.88307415888858,
"learning_rate": 7.93874357254639e-06,
"loss": 2.3176,
"step": 1420
},
{
"epoch": 1.4293078518891054,
"grad_norm": 19.113742813769317,
"learning_rate": 7.936507936507936e-06,
"loss": 2.2732,
"step": 1421
},
{
"epoch": 1.4303136983717861,
"grad_norm": 19.37026732661173,
"learning_rate": 7.934272300469484e-06,
"loss": 2.3011,
"step": 1422
},
{
"epoch": 1.4313195448544667,
"grad_norm": 18.49949387415047,
"learning_rate": 7.932036664431031e-06,
"loss": 2.2737,
"step": 1423
},
{
"epoch": 1.4323253913371472,
"grad_norm": 19.959543414268822,
"learning_rate": 7.92980102839258e-06,
"loss": 2.3153,
"step": 1424
},
{
"epoch": 1.4333312378198277,
"grad_norm": 28.318795043762517,
"learning_rate": 7.927565392354126e-06,
"loss": 2.3026,
"step": 1425
},
{
"epoch": 1.4343370843025083,
"grad_norm": 28.12273607964509,
"learning_rate": 7.925329756315672e-06,
"loss": 2.2637,
"step": 1426
},
{
"epoch": 1.4353429307851888,
"grad_norm": 17.778700492210334,
"learning_rate": 7.923094120277219e-06,
"loss": 2.3389,
"step": 1427
},
{
"epoch": 1.4363487772678694,
"grad_norm": 20.872414384196436,
"learning_rate": 7.920858484238767e-06,
"loss": 2.3155,
"step": 1428
},
{
"epoch": 1.4373546237505501,
"grad_norm": 25.94655838699598,
"learning_rate": 7.918622848200314e-06,
"loss": 2.2919,
"step": 1429
},
{
"epoch": 1.4383604702332307,
"grad_norm": 18.553279573237166,
"learning_rate": 7.916387212161862e-06,
"loss": 2.3057,
"step": 1430
},
{
"epoch": 1.4393663167159112,
"grad_norm": 20.94861910235284,
"learning_rate": 7.914151576123409e-06,
"loss": 2.3307,
"step": 1431
},
{
"epoch": 1.4403721631985917,
"grad_norm": 23.493662347236377,
"learning_rate": 7.911915940084955e-06,
"loss": 2.2584,
"step": 1432
},
{
"epoch": 1.4413780096812725,
"grad_norm": 17.882426706437784,
"learning_rate": 7.909680304046502e-06,
"loss": 2.2838,
"step": 1433
},
{
"epoch": 1.442383856163953,
"grad_norm": 19.72887454840581,
"learning_rate": 7.907444668008048e-06,
"loss": 2.2449,
"step": 1434
},
{
"epoch": 1.4433897026466336,
"grad_norm": 18.291268174889137,
"learning_rate": 7.905209031969596e-06,
"loss": 2.3025,
"step": 1435
},
{
"epoch": 1.4443955491293141,
"grad_norm": 20.22369948274378,
"learning_rate": 7.902973395931143e-06,
"loss": 2.306,
"step": 1436
},
{
"epoch": 1.4454013956119947,
"grad_norm": 18.7791697107874,
"learning_rate": 7.90073775989269e-06,
"loss": 2.2995,
"step": 1437
},
{
"epoch": 1.4464072420946752,
"grad_norm": 19.201215197139422,
"learning_rate": 7.898502123854236e-06,
"loss": 2.2909,
"step": 1438
},
{
"epoch": 1.447413088577356,
"grad_norm": 16.628374989794644,
"learning_rate": 7.896266487815784e-06,
"loss": 2.262,
"step": 1439
},
{
"epoch": 1.4484189350600365,
"grad_norm": 16.560349526712507,
"learning_rate": 7.894030851777331e-06,
"loss": 2.2959,
"step": 1440
},
{
"epoch": 1.449424781542717,
"grad_norm": 18.58682928866191,
"learning_rate": 7.89179521573888e-06,
"loss": 2.3068,
"step": 1441
},
{
"epoch": 1.4504306280253976,
"grad_norm": 18.162928908222984,
"learning_rate": 7.889559579700426e-06,
"loss": 2.3476,
"step": 1442
},
{
"epoch": 1.4514364745080783,
"grad_norm": 17.10078352883004,
"learning_rate": 7.887323943661972e-06,
"loss": 2.3223,
"step": 1443
},
{
"epoch": 1.4524423209907589,
"grad_norm": 14.894578538621019,
"learning_rate": 7.885088307623519e-06,
"loss": 2.3014,
"step": 1444
},
{
"epoch": 1.4534481674734394,
"grad_norm": 19.416479931616557,
"learning_rate": 7.882852671585065e-06,
"loss": 2.3405,
"step": 1445
},
{
"epoch": 1.45445401395612,
"grad_norm": 19.271151748168545,
"learning_rate": 7.880617035546614e-06,
"loss": 2.2826,
"step": 1446
},
{
"epoch": 1.4554598604388005,
"grad_norm": 20.01304026489289,
"learning_rate": 7.87838139950816e-06,
"loss": 2.2408,
"step": 1447
},
{
"epoch": 1.456465706921481,
"grad_norm": 18.516319943704005,
"learning_rate": 7.876145763469709e-06,
"loss": 2.3402,
"step": 1448
},
{
"epoch": 1.4574715534041616,
"grad_norm": 19.933211770601265,
"learning_rate": 7.873910127431255e-06,
"loss": 2.3071,
"step": 1449
},
{
"epoch": 1.4584773998868423,
"grad_norm": 19.876724976053293,
"learning_rate": 7.871674491392802e-06,
"loss": 2.2551,
"step": 1450
},
{
"epoch": 1.4594832463695229,
"grad_norm": 20.107325835555148,
"learning_rate": 7.869438855354348e-06,
"loss": 2.2941,
"step": 1451
},
{
"epoch": 1.4604890928522034,
"grad_norm": 21.42891005758536,
"learning_rate": 7.867203219315896e-06,
"loss": 2.3477,
"step": 1452
},
{
"epoch": 1.461494939334884,
"grad_norm": 15.536307850085874,
"learning_rate": 7.864967583277443e-06,
"loss": 2.3091,
"step": 1453
},
{
"epoch": 1.4625007858175647,
"grad_norm": 21.50933449834089,
"learning_rate": 7.86273194723899e-06,
"loss": 2.2951,
"step": 1454
},
{
"epoch": 1.4635066323002452,
"grad_norm": 21.12008364466888,
"learning_rate": 7.860496311200538e-06,
"loss": 2.318,
"step": 1455
},
{
"epoch": 1.4645124787829258,
"grad_norm": 18.665952854109474,
"learning_rate": 7.858260675162084e-06,
"loss": 2.2878,
"step": 1456
},
{
"epoch": 1.4655183252656063,
"grad_norm": 21.961368925244503,
"learning_rate": 7.856025039123631e-06,
"loss": 2.291,
"step": 1457
},
{
"epoch": 1.4665241717482869,
"grad_norm": 17.713270839538062,
"learning_rate": 7.853789403085178e-06,
"loss": 2.2357,
"step": 1458
},
{
"epoch": 1.4675300182309674,
"grad_norm": 20.24530939385799,
"learning_rate": 7.851553767046726e-06,
"loss": 2.2657,
"step": 1459
},
{
"epoch": 1.468535864713648,
"grad_norm": 19.141291616606033,
"learning_rate": 7.849318131008272e-06,
"loss": 2.3313,
"step": 1460
},
{
"epoch": 1.4695417111963287,
"grad_norm": 19.037031944583642,
"learning_rate": 7.84708249496982e-06,
"loss": 2.3352,
"step": 1461
},
{
"epoch": 1.4705475576790092,
"grad_norm": 18.07282339133904,
"learning_rate": 7.844846858931367e-06,
"loss": 2.3196,
"step": 1462
},
{
"epoch": 1.4715534041616898,
"grad_norm": 19.871435309000702,
"learning_rate": 7.842611222892914e-06,
"loss": 2.2901,
"step": 1463
},
{
"epoch": 1.4725592506443703,
"grad_norm": 17.65462170569615,
"learning_rate": 7.84037558685446e-06,
"loss": 2.3058,
"step": 1464
},
{
"epoch": 1.473565097127051,
"grad_norm": 17.577295712228253,
"learning_rate": 7.838139950816009e-06,
"loss": 2.2996,
"step": 1465
},
{
"epoch": 1.4745709436097316,
"grad_norm": 17.72683927987532,
"learning_rate": 7.835904314777555e-06,
"loss": 2.3471,
"step": 1466
},
{
"epoch": 1.4755767900924122,
"grad_norm": 16.453599601296077,
"learning_rate": 7.833668678739102e-06,
"loss": 2.271,
"step": 1467
},
{
"epoch": 1.4765826365750927,
"grad_norm": 15.182795987382685,
"learning_rate": 7.83143304270065e-06,
"loss": 2.3045,
"step": 1468
},
{
"epoch": 1.4775884830577732,
"grad_norm": 15.254890201984024,
"learning_rate": 7.829197406662196e-06,
"loss": 2.2917,
"step": 1469
},
{
"epoch": 1.4785943295404538,
"grad_norm": 16.617690368093108,
"learning_rate": 7.826961770623743e-06,
"loss": 2.2954,
"step": 1470
},
{
"epoch": 1.4796001760231345,
"grad_norm": 17.407836062399724,
"learning_rate": 7.82472613458529e-06,
"loss": 2.2826,
"step": 1471
},
{
"epoch": 1.480606022505815,
"grad_norm": 16.60301913849178,
"learning_rate": 7.822490498546838e-06,
"loss": 2.2914,
"step": 1472
},
{
"epoch": 1.4816118689884956,
"grad_norm": 17.802071741821546,
"learning_rate": 7.820254862508384e-06,
"loss": 2.2856,
"step": 1473
},
{
"epoch": 1.4826177154711762,
"grad_norm": 23.12527459981009,
"learning_rate": 7.818019226469931e-06,
"loss": 2.3058,
"step": 1474
},
{
"epoch": 1.483623561953857,
"grad_norm": 16.6587149986682,
"learning_rate": 7.815783590431477e-06,
"loss": 2.3052,
"step": 1475
},
{
"epoch": 1.4846294084365375,
"grad_norm": 16.928897564820495,
"learning_rate": 7.813547954393026e-06,
"loss": 2.289,
"step": 1476
},
{
"epoch": 1.485635254919218,
"grad_norm": 16.34399216078306,
"learning_rate": 7.811312318354572e-06,
"loss": 2.291,
"step": 1477
},
{
"epoch": 1.4866411014018985,
"grad_norm": 17.973201861561957,
"learning_rate": 7.809076682316119e-06,
"loss": 2.3251,
"step": 1478
},
{
"epoch": 1.487646947884579,
"grad_norm": 17.39086510043571,
"learning_rate": 7.806841046277667e-06,
"loss": 2.2327,
"step": 1479
},
{
"epoch": 1.4886527943672596,
"grad_norm": 17.920711958288884,
"learning_rate": 7.804605410239214e-06,
"loss": 2.3026,
"step": 1480
},
{
"epoch": 1.4896586408499402,
"grad_norm": 20.359023819061278,
"learning_rate": 7.80236977420076e-06,
"loss": 2.2502,
"step": 1481
},
{
"epoch": 1.490664487332621,
"grad_norm": 16.90484054607315,
"learning_rate": 7.800134138162307e-06,
"loss": 2.3013,
"step": 1482
},
{
"epoch": 1.4916703338153015,
"grad_norm": 21.90767780268189,
"learning_rate": 7.797898502123855e-06,
"loss": 2.2774,
"step": 1483
},
{
"epoch": 1.492676180297982,
"grad_norm": 18.721112310966856,
"learning_rate": 7.795662866085402e-06,
"loss": 2.2839,
"step": 1484
},
{
"epoch": 1.4936820267806625,
"grad_norm": 17.911089882318574,
"learning_rate": 7.79342723004695e-06,
"loss": 2.2604,
"step": 1485
},
{
"epoch": 1.4946878732633433,
"grad_norm": 22.772704071372782,
"learning_rate": 7.791191594008496e-06,
"loss": 2.2407,
"step": 1486
},
{
"epoch": 1.4956937197460238,
"grad_norm": 17.543471809793463,
"learning_rate": 7.788955957970043e-06,
"loss": 2.2881,
"step": 1487
},
{
"epoch": 1.4966995662287044,
"grad_norm": 20.867966397052506,
"learning_rate": 7.78672032193159e-06,
"loss": 2.2543,
"step": 1488
},
{
"epoch": 1.497705412711385,
"grad_norm": 22.221623052043494,
"learning_rate": 7.784484685893138e-06,
"loss": 2.2677,
"step": 1489
},
{
"epoch": 1.4987112591940654,
"grad_norm": 17.112705665868003,
"learning_rate": 7.782249049854684e-06,
"loss": 2.2572,
"step": 1490
},
{
"epoch": 1.499717105676746,
"grad_norm": 19.690025494932556,
"learning_rate": 7.780013413816231e-06,
"loss": 2.285,
"step": 1491
},
{
"epoch": 1.5007229521594265,
"grad_norm": 17.9695282699989,
"learning_rate": 7.77777777777778e-06,
"loss": 2.3181,
"step": 1492
},
{
"epoch": 1.5017287986421073,
"grad_norm": 19.36781062267981,
"learning_rate": 7.775542141739326e-06,
"loss": 2.3131,
"step": 1493
},
{
"epoch": 1.5027346451247878,
"grad_norm": 20.125903914963455,
"learning_rate": 7.773306505700872e-06,
"loss": 2.2431,
"step": 1494
},
{
"epoch": 1.5037404916074684,
"grad_norm": 16.253820796223152,
"learning_rate": 7.771070869662419e-06,
"loss": 2.2616,
"step": 1495
},
{
"epoch": 1.5047463380901491,
"grad_norm": 23.8382579830196,
"learning_rate": 7.768835233623967e-06,
"loss": 2.304,
"step": 1496
},
{
"epoch": 1.5057521845728297,
"grad_norm": 18.49099241609707,
"learning_rate": 7.766599597585514e-06,
"loss": 2.232,
"step": 1497
},
{
"epoch": 1.5067580310555102,
"grad_norm": 20.838284123192146,
"learning_rate": 7.764363961547062e-06,
"loss": 2.318,
"step": 1498
},
{
"epoch": 1.5077638775381907,
"grad_norm": 22.883153520829786,
"learning_rate": 7.762128325508608e-06,
"loss": 2.282,
"step": 1499
},
{
"epoch": 1.5087697240208713,
"grad_norm": 16.55878728484685,
"learning_rate": 7.759892689470155e-06,
"loss": 2.2729,
"step": 1500
},
{
"epoch": 1.5097755705035518,
"grad_norm": 20.53854426478302,
"learning_rate": 7.757657053431702e-06,
"loss": 2.3073,
"step": 1501
},
{
"epoch": 1.5107814169862324,
"grad_norm": 22.087387133590592,
"learning_rate": 7.755421417393248e-06,
"loss": 2.2862,
"step": 1502
},
{
"epoch": 1.511787263468913,
"grad_norm": 19.819179809534393,
"learning_rate": 7.753185781354796e-06,
"loss": 2.2867,
"step": 1503
},
{
"epoch": 1.5127931099515937,
"grad_norm": 18.2453326516021,
"learning_rate": 7.750950145316343e-06,
"loss": 2.2911,
"step": 1504
},
{
"epoch": 1.5137989564342742,
"grad_norm": 16.93405386428347,
"learning_rate": 7.748714509277891e-06,
"loss": 2.3279,
"step": 1505
},
{
"epoch": 1.514804802916955,
"grad_norm": 22.014071115433765,
"learning_rate": 7.746478873239436e-06,
"loss": 2.303,
"step": 1506
},
{
"epoch": 1.5158106493996355,
"grad_norm": 17.916438730856672,
"learning_rate": 7.744243237200984e-06,
"loss": 2.2677,
"step": 1507
},
{
"epoch": 1.516816495882316,
"grad_norm": 20.27092818009331,
"learning_rate": 7.742007601162531e-06,
"loss": 2.3191,
"step": 1508
},
{
"epoch": 1.5178223423649966,
"grad_norm": 17.669526243256236,
"learning_rate": 7.739771965124079e-06,
"loss": 2.2871,
"step": 1509
},
{
"epoch": 1.5188281888476771,
"grad_norm": 15.088838347448466,
"learning_rate": 7.737536329085626e-06,
"loss": 2.2643,
"step": 1510
},
{
"epoch": 1.5198340353303577,
"grad_norm": 20.41515148594282,
"learning_rate": 7.735300693047172e-06,
"loss": 2.3322,
"step": 1511
},
{
"epoch": 1.5208398818130382,
"grad_norm": 16.05840910527969,
"learning_rate": 7.733065057008719e-06,
"loss": 2.2845,
"step": 1512
},
{
"epoch": 1.5218457282957187,
"grad_norm": 18.893387044686634,
"learning_rate": 7.730829420970265e-06,
"loss": 2.3403,
"step": 1513
},
{
"epoch": 1.5228515747783995,
"grad_norm": 17.915120537065008,
"learning_rate": 7.728593784931814e-06,
"loss": 2.2217,
"step": 1514
},
{
"epoch": 1.52385742126108,
"grad_norm": 17.371245044359178,
"learning_rate": 7.72635814889336e-06,
"loss": 2.2844,
"step": 1515
},
{
"epoch": 1.5248632677437606,
"grad_norm": 20.607204737742613,
"learning_rate": 7.724122512854908e-06,
"loss": 2.2285,
"step": 1516
},
{
"epoch": 1.5258691142264413,
"grad_norm": 15.979393180675775,
"learning_rate": 7.721886876816455e-06,
"loss": 2.2931,
"step": 1517
},
{
"epoch": 1.5268749607091219,
"grad_norm": 19.852453533351184,
"learning_rate": 7.719651240778002e-06,
"loss": 2.3007,
"step": 1518
},
{
"epoch": 1.5278808071918024,
"grad_norm": 19.50268486882533,
"learning_rate": 7.717415604739548e-06,
"loss": 2.2704,
"step": 1519
},
{
"epoch": 1.528886653674483,
"grad_norm": 17.849737130119692,
"learning_rate": 7.715179968701096e-06,
"loss": 2.3057,
"step": 1520
},
{
"epoch": 1.5298925001571635,
"grad_norm": 19.714501629790885,
"learning_rate": 7.712944332662643e-06,
"loss": 2.2569,
"step": 1521
},
{
"epoch": 1.530898346639844,
"grad_norm": 20.09140238122812,
"learning_rate": 7.710708696624191e-06,
"loss": 2.2741,
"step": 1522
},
{
"epoch": 1.5319041931225246,
"grad_norm": 19.05465132384539,
"learning_rate": 7.708473060585738e-06,
"loss": 2.2615,
"step": 1523
},
{
"epoch": 1.532910039605205,
"grad_norm": 16.699820322420052,
"learning_rate": 7.706237424547284e-06,
"loss": 2.2729,
"step": 1524
},
{
"epoch": 1.5339158860878859,
"grad_norm": 18.145399056940995,
"learning_rate": 7.704001788508831e-06,
"loss": 2.3133,
"step": 1525
},
{
"epoch": 1.5349217325705664,
"grad_norm": 17.27467185746975,
"learning_rate": 7.701766152470377e-06,
"loss": 2.3199,
"step": 1526
},
{
"epoch": 1.535927579053247,
"grad_norm": 17.261437605040133,
"learning_rate": 7.699530516431926e-06,
"loss": 2.314,
"step": 1527
},
{
"epoch": 1.5369334255359277,
"grad_norm": 18.016550773848497,
"learning_rate": 7.697294880393472e-06,
"loss": 2.2802,
"step": 1528
},
{
"epoch": 1.5379392720186082,
"grad_norm": 15.645657281375113,
"learning_rate": 7.69505924435502e-06,
"loss": 2.2203,
"step": 1529
},
{
"epoch": 1.5389451185012888,
"grad_norm": 17.11336773058832,
"learning_rate": 7.692823608316567e-06,
"loss": 2.3103,
"step": 1530
},
{
"epoch": 1.5399509649839693,
"grad_norm": 19.886367565353304,
"learning_rate": 7.690587972278114e-06,
"loss": 2.262,
"step": 1531
},
{
"epoch": 1.5409568114666499,
"grad_norm": 15.994878268078955,
"learning_rate": 7.68835233623966e-06,
"loss": 2.3215,
"step": 1532
},
{
"epoch": 1.5419626579493304,
"grad_norm": 17.74213070501031,
"learning_rate": 7.686116700201208e-06,
"loss": 2.3512,
"step": 1533
},
{
"epoch": 1.542968504432011,
"grad_norm": 16.83461984011088,
"learning_rate": 7.683881064162755e-06,
"loss": 2.3289,
"step": 1534
},
{
"epoch": 1.5439743509146915,
"grad_norm": 18.24770376850407,
"learning_rate": 7.681645428124303e-06,
"loss": 2.3014,
"step": 1535
},
{
"epoch": 1.5449801973973722,
"grad_norm": 17.227560803417234,
"learning_rate": 7.67940979208585e-06,
"loss": 2.3434,
"step": 1536
},
{
"epoch": 1.5459860438800528,
"grad_norm": 14.279290910690355,
"learning_rate": 7.677174156047396e-06,
"loss": 2.2851,
"step": 1537
},
{
"epoch": 1.5469918903627335,
"grad_norm": 17.359903771378765,
"learning_rate": 7.674938520008943e-06,
"loss": 2.286,
"step": 1538
},
{
"epoch": 1.547997736845414,
"grad_norm": 19.484495307853788,
"learning_rate": 7.67270288397049e-06,
"loss": 2.351,
"step": 1539
},
{
"epoch": 1.5490035833280946,
"grad_norm": 16.96042326031945,
"learning_rate": 7.670467247932038e-06,
"loss": 2.2583,
"step": 1540
},
{
"epoch": 1.5500094298107752,
"grad_norm": 15.621909656589937,
"learning_rate": 7.668231611893584e-06,
"loss": 2.2972,
"step": 1541
},
{
"epoch": 1.5510152762934557,
"grad_norm": 18.6003281206334,
"learning_rate": 7.665995975855131e-06,
"loss": 2.3043,
"step": 1542
},
{
"epoch": 1.5520211227761362,
"grad_norm": 18.128664424443144,
"learning_rate": 7.663760339816677e-06,
"loss": 2.3357,
"step": 1543
},
{
"epoch": 1.5530269692588168,
"grad_norm": 15.940286885757834,
"learning_rate": 7.661524703778226e-06,
"loss": 2.3372,
"step": 1544
},
{
"epoch": 1.5540328157414973,
"grad_norm": 17.025874797911026,
"learning_rate": 7.659289067739772e-06,
"loss": 2.2931,
"step": 1545
},
{
"epoch": 1.555038662224178,
"grad_norm": 19.790708605098256,
"learning_rate": 7.65705343170132e-06,
"loss": 2.2587,
"step": 1546
},
{
"epoch": 1.5560445087068586,
"grad_norm": 18.918091015148942,
"learning_rate": 7.654817795662867e-06,
"loss": 2.3327,
"step": 1547
},
{
"epoch": 1.5570503551895392,
"grad_norm": 17.33264530456271,
"learning_rate": 7.652582159624414e-06,
"loss": 2.2949,
"step": 1548
},
{
"epoch": 1.55805620167222,
"grad_norm": 23.167025477343103,
"learning_rate": 7.65034652358596e-06,
"loss": 2.2936,
"step": 1549
},
{
"epoch": 1.5590620481549005,
"grad_norm": 17.69559287598297,
"learning_rate": 7.648110887547507e-06,
"loss": 2.2675,
"step": 1550
},
{
"epoch": 1.560067894637581,
"grad_norm": 18.572405920081234,
"learning_rate": 7.645875251509055e-06,
"loss": 2.3261,
"step": 1551
},
{
"epoch": 1.5610737411202615,
"grad_norm": 21.371575498912534,
"learning_rate": 7.643639615470602e-06,
"loss": 2.2848,
"step": 1552
},
{
"epoch": 1.562079587602942,
"grad_norm": 16.137328309550046,
"learning_rate": 7.64140397943215e-06,
"loss": 2.3007,
"step": 1553
},
{
"epoch": 1.5630854340856226,
"grad_norm": 15.887326170285341,
"learning_rate": 7.639168343393696e-06,
"loss": 2.2878,
"step": 1554
},
{
"epoch": 1.5640912805683032,
"grad_norm": 19.70258894896291,
"learning_rate": 7.636932707355243e-06,
"loss": 2.2965,
"step": 1555
},
{
"epoch": 1.5650971270509837,
"grad_norm": 17.077639987037216,
"learning_rate": 7.63469707131679e-06,
"loss": 2.3249,
"step": 1556
},
{
"epoch": 1.5661029735336645,
"grad_norm": 18.204144244300757,
"learning_rate": 7.632461435278338e-06,
"loss": 2.2854,
"step": 1557
},
{
"epoch": 1.567108820016345,
"grad_norm": 17.39808010530622,
"learning_rate": 7.630225799239884e-06,
"loss": 2.3217,
"step": 1558
},
{
"epoch": 1.5681146664990255,
"grad_norm": 18.207352425342457,
"learning_rate": 7.627990163201432e-06,
"loss": 2.3129,
"step": 1559
},
{
"epoch": 1.5691205129817063,
"grad_norm": 16.717726119150125,
"learning_rate": 7.625754527162978e-06,
"loss": 2.3432,
"step": 1560
},
{
"epoch": 1.5701263594643868,
"grad_norm": 18.753395192538807,
"learning_rate": 7.623518891124525e-06,
"loss": 2.2698,
"step": 1561
},
{
"epoch": 1.5711322059470674,
"grad_norm": 16.586849349807938,
"learning_rate": 7.621283255086073e-06,
"loss": 2.3282,
"step": 1562
},
{
"epoch": 1.572138052429748,
"grad_norm": 16.322106428815346,
"learning_rate": 7.61904761904762e-06,
"loss": 2.243,
"step": 1563
},
{
"epoch": 1.5731438989124285,
"grad_norm": 16.818011065400956,
"learning_rate": 7.616811983009167e-06,
"loss": 2.347,
"step": 1564
},
{
"epoch": 1.574149745395109,
"grad_norm": 17.639483723555664,
"learning_rate": 7.614576346970714e-06,
"loss": 2.2921,
"step": 1565
},
{
"epoch": 1.5751555918777895,
"grad_norm": 19.123730656020907,
"learning_rate": 7.612340710932261e-06,
"loss": 2.2522,
"step": 1566
},
{
"epoch": 1.57616143836047,
"grad_norm": 16.981957712953395,
"learning_rate": 7.6101050748938076e-06,
"loss": 2.3308,
"step": 1567
},
{
"epoch": 1.5771672848431508,
"grad_norm": 18.721902223665907,
"learning_rate": 7.607869438855355e-06,
"loss": 2.3076,
"step": 1568
},
{
"epoch": 1.5781731313258314,
"grad_norm": 15.591415803272543,
"learning_rate": 7.6056338028169015e-06,
"loss": 2.3147,
"step": 1569
},
{
"epoch": 1.5791789778085121,
"grad_norm": 17.251802483159004,
"learning_rate": 7.60339816677845e-06,
"loss": 2.3193,
"step": 1570
},
{
"epoch": 1.5801848242911927,
"grad_norm": 15.26863125866677,
"learning_rate": 7.601162530739996e-06,
"loss": 2.2885,
"step": 1571
},
{
"epoch": 1.5811906707738732,
"grad_norm": 15.85702751098647,
"learning_rate": 7.598926894701543e-06,
"loss": 2.2848,
"step": 1572
},
{
"epoch": 1.5821965172565537,
"grad_norm": 18.336826845804772,
"learning_rate": 7.59669125866309e-06,
"loss": 2.3111,
"step": 1573
},
{
"epoch": 1.5832023637392343,
"grad_norm": 17.863224481239193,
"learning_rate": 7.594455622624637e-06,
"loss": 2.3371,
"step": 1574
},
{
"epoch": 1.5842082102219148,
"grad_norm": 17.341866010483724,
"learning_rate": 7.592219986586184e-06,
"loss": 2.3378,
"step": 1575
},
{
"epoch": 1.5852140567045954,
"grad_norm": 17.635446248556168,
"learning_rate": 7.589984350547731e-06,
"loss": 2.2879,
"step": 1576
},
{
"epoch": 1.586219903187276,
"grad_norm": 15.391015775079701,
"learning_rate": 7.587748714509279e-06,
"loss": 2.2639,
"step": 1577
},
{
"epoch": 1.5872257496699567,
"grad_norm": 16.813917700341563,
"learning_rate": 7.585513078470826e-06,
"loss": 2.2648,
"step": 1578
},
{
"epoch": 1.5882315961526372,
"grad_norm": 18.37044213143786,
"learning_rate": 7.583277442432373e-06,
"loss": 2.2342,
"step": 1579
},
{
"epoch": 1.5892374426353177,
"grad_norm": 16.773237086377012,
"learning_rate": 7.58104180639392e-06,
"loss": 2.2814,
"step": 1580
},
{
"epoch": 1.5902432891179985,
"grad_norm": 16.966975249230583,
"learning_rate": 7.578806170355467e-06,
"loss": 2.306,
"step": 1581
},
{
"epoch": 1.591249135600679,
"grad_norm": 19.29591516919744,
"learning_rate": 7.576570534317014e-06,
"loss": 2.2499,
"step": 1582
},
{
"epoch": 1.5922549820833596,
"grad_norm": 18.323197040030927,
"learning_rate": 7.574334898278561e-06,
"loss": 2.2762,
"step": 1583
},
{
"epoch": 1.5932608285660401,
"grad_norm": 16.83769559840373,
"learning_rate": 7.5720992622401075e-06,
"loss": 2.3096,
"step": 1584
},
{
"epoch": 1.5942666750487207,
"grad_norm": 19.1920754770466,
"learning_rate": 7.569863626201654e-06,
"loss": 2.2841,
"step": 1585
},
{
"epoch": 1.5952725215314012,
"grad_norm": 17.194661471609137,
"learning_rate": 7.567627990163202e-06,
"loss": 2.3223,
"step": 1586
},
{
"epoch": 1.5962783680140817,
"grad_norm": 15.766621571000169,
"learning_rate": 7.565392354124749e-06,
"loss": 2.2976,
"step": 1587
},
{
"epoch": 1.5972842144967623,
"grad_norm": 18.95605767159444,
"learning_rate": 7.563156718086296e-06,
"loss": 2.3344,
"step": 1588
},
{
"epoch": 1.598290060979443,
"grad_norm": 17.318139886192373,
"learning_rate": 7.560921082047843e-06,
"loss": 2.2738,
"step": 1589
},
{
"epoch": 1.5992959074621236,
"grad_norm": 17.65570490189223,
"learning_rate": 7.55868544600939e-06,
"loss": 2.3123,
"step": 1590
},
{
"epoch": 1.6003017539448043,
"grad_norm": 16.523811335663304,
"learning_rate": 7.556449809970937e-06,
"loss": 2.2588,
"step": 1591
},
{
"epoch": 1.6013076004274849,
"grad_norm": 17.58715421753379,
"learning_rate": 7.554214173932485e-06,
"loss": 2.2696,
"step": 1592
},
{
"epoch": 1.6023134469101654,
"grad_norm": 15.615437328188383,
"learning_rate": 7.551978537894032e-06,
"loss": 2.2478,
"step": 1593
},
{
"epoch": 1.603319293392846,
"grad_norm": 16.20342586210378,
"learning_rate": 7.549742901855579e-06,
"loss": 2.2727,
"step": 1594
},
{
"epoch": 1.6043251398755265,
"grad_norm": 19.09460593004032,
"learning_rate": 7.547507265817126e-06,
"loss": 2.3034,
"step": 1595
},
{
"epoch": 1.605330986358207,
"grad_norm": 19.305798240899804,
"learning_rate": 7.545271629778672e-06,
"loss": 2.3049,
"step": 1596
},
{
"epoch": 1.6063368328408876,
"grad_norm": 20.287377727477924,
"learning_rate": 7.54303599374022e-06,
"loss": 2.3243,
"step": 1597
},
{
"epoch": 1.6073426793235681,
"grad_norm": 16.070930894706038,
"learning_rate": 7.540800357701766e-06,
"loss": 2.3084,
"step": 1598
},
{
"epoch": 1.6083485258062489,
"grad_norm": 20.07386770790435,
"learning_rate": 7.538564721663314e-06,
"loss": 2.3058,
"step": 1599
},
{
"epoch": 1.6093543722889294,
"grad_norm": 19.88535982593183,
"learning_rate": 7.53632908562486e-06,
"loss": 2.2875,
"step": 1600
},
{
"epoch": 1.61036021877161,
"grad_norm": 16.693701256020397,
"learning_rate": 7.534093449586408e-06,
"loss": 2.2908,
"step": 1601
},
{
"epoch": 1.6113660652542907,
"grad_norm": 22.38511771234477,
"learning_rate": 7.531857813547955e-06,
"loss": 2.3289,
"step": 1602
},
{
"epoch": 1.6123719117369713,
"grad_norm": 19.256280771639375,
"learning_rate": 7.529622177509502e-06,
"loss": 2.2923,
"step": 1603
},
{
"epoch": 1.6133777582196518,
"grad_norm": 15.807615153380606,
"learning_rate": 7.527386541471049e-06,
"loss": 2.3151,
"step": 1604
},
{
"epoch": 1.6143836047023323,
"grad_norm": 17.14795329603459,
"learning_rate": 7.525150905432596e-06,
"loss": 2.2766,
"step": 1605
},
{
"epoch": 1.6153894511850129,
"grad_norm": 19.0314082816804,
"learning_rate": 7.522915269394143e-06,
"loss": 2.2283,
"step": 1606
},
{
"epoch": 1.6163952976676934,
"grad_norm": 16.827496274558783,
"learning_rate": 7.520679633355691e-06,
"loss": 2.2872,
"step": 1607
},
{
"epoch": 1.617401144150374,
"grad_norm": 18.415640161477818,
"learning_rate": 7.518443997317238e-06,
"loss": 2.2646,
"step": 1608
},
{
"epoch": 1.6184069906330545,
"grad_norm": 18.884538591397305,
"learning_rate": 7.516208361278784e-06,
"loss": 2.2652,
"step": 1609
},
{
"epoch": 1.6194128371157352,
"grad_norm": 19.099860731773862,
"learning_rate": 7.513972725240332e-06,
"loss": 2.2819,
"step": 1610
},
{
"epoch": 1.6204186835984158,
"grad_norm": 18.518647433605977,
"learning_rate": 7.511737089201878e-06,
"loss": 2.2982,
"step": 1611
},
{
"epoch": 1.6214245300810963,
"grad_norm": 18.662477998814204,
"learning_rate": 7.509501453163426e-06,
"loss": 2.2893,
"step": 1612
},
{
"epoch": 1.622430376563777,
"grad_norm": 17.23152071581131,
"learning_rate": 7.507265817124972e-06,
"loss": 2.2361,
"step": 1613
},
{
"epoch": 1.6234362230464576,
"grad_norm": 20.286786912700784,
"learning_rate": 7.5050301810865204e-06,
"loss": 2.2675,
"step": 1614
},
{
"epoch": 1.6244420695291382,
"grad_norm": 18.447265051511966,
"learning_rate": 7.502794545048067e-06,
"loss": 2.3422,
"step": 1615
},
{
"epoch": 1.6254479160118187,
"grad_norm": 16.137844374119336,
"learning_rate": 7.500558909009614e-06,
"loss": 2.2929,
"step": 1616
},
{
"epoch": 1.6264537624944992,
"grad_norm": 16.644876712805715,
"learning_rate": 7.498323272971161e-06,
"loss": 2.2395,
"step": 1617
},
{
"epoch": 1.6274596089771798,
"grad_norm": 16.375471679331596,
"learning_rate": 7.496087636932708e-06,
"loss": 2.2761,
"step": 1618
},
{
"epoch": 1.6284654554598603,
"grad_norm": 14.94674470892123,
"learning_rate": 7.493852000894255e-06,
"loss": 2.3036,
"step": 1619
},
{
"epoch": 1.6294713019425409,
"grad_norm": 15.81743617520014,
"learning_rate": 7.4916163648558015e-06,
"loss": 2.2889,
"step": 1620
},
{
"epoch": 1.6304771484252216,
"grad_norm": 17.416368848556772,
"learning_rate": 7.489380728817349e-06,
"loss": 2.3139,
"step": 1621
},
{
"epoch": 1.6314829949079022,
"grad_norm": 16.010398020295742,
"learning_rate": 7.4871450927788954e-06,
"loss": 2.2733,
"step": 1622
},
{
"epoch": 1.632488841390583,
"grad_norm": 17.340450107758183,
"learning_rate": 7.484909456740444e-06,
"loss": 2.2623,
"step": 1623
},
{
"epoch": 1.6334946878732635,
"grad_norm": 17.510561385062868,
"learning_rate": 7.48267382070199e-06,
"loss": 2.3156,
"step": 1624
},
{
"epoch": 1.634500534355944,
"grad_norm": 16.704347034836946,
"learning_rate": 7.480438184663538e-06,
"loss": 2.2747,
"step": 1625
},
{
"epoch": 1.6355063808386245,
"grad_norm": 16.57972190196149,
"learning_rate": 7.478202548625084e-06,
"loss": 2.272,
"step": 1626
},
{
"epoch": 1.636512227321305,
"grad_norm": 15.418220828290822,
"learning_rate": 7.475966912586632e-06,
"loss": 2.2657,
"step": 1627
},
{
"epoch": 1.6375180738039856,
"grad_norm": 17.242284011573524,
"learning_rate": 7.473731276548178e-06,
"loss": 2.2876,
"step": 1628
},
{
"epoch": 1.6385239202866662,
"grad_norm": 17.120144649327322,
"learning_rate": 7.4714956405097264e-06,
"loss": 2.2536,
"step": 1629
},
{
"epoch": 1.6395297667693467,
"grad_norm": 15.648808334665938,
"learning_rate": 7.469260004471273e-06,
"loss": 2.3287,
"step": 1630
},
{
"epoch": 1.6405356132520275,
"grad_norm": 15.567489200950083,
"learning_rate": 7.46702436843282e-06,
"loss": 2.2619,
"step": 1631
},
{
"epoch": 1.641541459734708,
"grad_norm": 19.00407342375508,
"learning_rate": 7.464788732394367e-06,
"loss": 2.302,
"step": 1632
},
{
"epoch": 1.6425473062173885,
"grad_norm": 17.781602484456865,
"learning_rate": 7.4625530963559135e-06,
"loss": 2.3069,
"step": 1633
},
{
"epoch": 1.6435531527000693,
"grad_norm": 15.920283011983388,
"learning_rate": 7.460317460317461e-06,
"loss": 2.2918,
"step": 1634
},
{
"epoch": 1.6445589991827498,
"grad_norm": 17.792057329372298,
"learning_rate": 7.4580818242790075e-06,
"loss": 2.3679,
"step": 1635
},
{
"epoch": 1.6455648456654304,
"grad_norm": 21.676130512651092,
"learning_rate": 7.455846188240555e-06,
"loss": 2.283,
"step": 1636
},
{
"epoch": 1.646570692148111,
"grad_norm": 16.514698135707608,
"learning_rate": 7.4536105522021015e-06,
"loss": 2.2325,
"step": 1637
},
{
"epoch": 1.6475765386307915,
"grad_norm": 21.39786786668378,
"learning_rate": 7.45137491616365e-06,
"loss": 2.2731,
"step": 1638
},
{
"epoch": 1.648582385113472,
"grad_norm": 18.00833184983677,
"learning_rate": 7.449139280125196e-06,
"loss": 2.2486,
"step": 1639
},
{
"epoch": 1.6495882315961525,
"grad_norm": 15.846871390888415,
"learning_rate": 7.446903644086744e-06,
"loss": 2.2691,
"step": 1640
},
{
"epoch": 1.650594078078833,
"grad_norm": 22.000826235719195,
"learning_rate": 7.44466800804829e-06,
"loss": 2.2222,
"step": 1641
},
{
"epoch": 1.6515999245615138,
"grad_norm": 19.455691706197182,
"learning_rate": 7.442432372009838e-06,
"loss": 2.328,
"step": 1642
},
{
"epoch": 1.6526057710441944,
"grad_norm": 17.94871421325755,
"learning_rate": 7.440196735971384e-06,
"loss": 2.3053,
"step": 1643
},
{
"epoch": 1.653611617526875,
"grad_norm": 17.9376483521117,
"learning_rate": 7.437961099932931e-06,
"loss": 2.302,
"step": 1644
},
{
"epoch": 1.6546174640095557,
"grad_norm": 16.4389373413972,
"learning_rate": 7.435725463894479e-06,
"loss": 2.2397,
"step": 1645
},
{
"epoch": 1.6556233104922362,
"grad_norm": 16.99132641098515,
"learning_rate": 7.433489827856026e-06,
"loss": 2.256,
"step": 1646
},
{
"epoch": 1.6566291569749168,
"grad_norm": 17.542599768403043,
"learning_rate": 7.431254191817573e-06,
"loss": 2.2331,
"step": 1647
},
{
"epoch": 1.6576350034575973,
"grad_norm": 17.626876401215032,
"learning_rate": 7.4290185557791196e-06,
"loss": 2.3472,
"step": 1648
},
{
"epoch": 1.6586408499402778,
"grad_norm": 16.872425273361493,
"learning_rate": 7.426782919740667e-06,
"loss": 2.3248,
"step": 1649
},
{
"epoch": 1.6596466964229584,
"grad_norm": 17.162326734465445,
"learning_rate": 7.4245472837022135e-06,
"loss": 2.2991,
"step": 1650
},
{
"epoch": 1.660652542905639,
"grad_norm": 19.75105766718973,
"learning_rate": 7.422311647663761e-06,
"loss": 2.2993,
"step": 1651
},
{
"epoch": 1.6616583893883194,
"grad_norm": 15.972163852419595,
"learning_rate": 7.4200760116253075e-06,
"loss": 2.2432,
"step": 1652
},
{
"epoch": 1.6626642358710002,
"grad_norm": 17.600365845305994,
"learning_rate": 7.417840375586856e-06,
"loss": 2.2461,
"step": 1653
},
{
"epoch": 1.6636700823536807,
"grad_norm": 21.500517437821884,
"learning_rate": 7.415604739548402e-06,
"loss": 2.2542,
"step": 1654
},
{
"epoch": 1.6646759288363615,
"grad_norm": 17.514071400959207,
"learning_rate": 7.413369103509949e-06,
"loss": 2.2928,
"step": 1655
},
{
"epoch": 1.665681775319042,
"grad_norm": 16.516886183339697,
"learning_rate": 7.411133467471496e-06,
"loss": 2.3187,
"step": 1656
},
{
"epoch": 1.6666876218017226,
"grad_norm": 16.518222183924482,
"learning_rate": 7.408897831433043e-06,
"loss": 2.2754,
"step": 1657
},
{
"epoch": 1.6676934682844031,
"grad_norm": 19.814913536195757,
"learning_rate": 7.40666219539459e-06,
"loss": 2.2452,
"step": 1658
},
{
"epoch": 1.6686993147670837,
"grad_norm": 18.775262311804575,
"learning_rate": 7.404426559356137e-06,
"loss": 2.2773,
"step": 1659
},
{
"epoch": 1.6697051612497642,
"grad_norm": 17.397581620828824,
"learning_rate": 7.402190923317685e-06,
"loss": 2.3,
"step": 1660
},
{
"epoch": 1.6707110077324447,
"grad_norm": 16.93196801056379,
"learning_rate": 7.399955287279232e-06,
"loss": 2.2719,
"step": 1661
},
{
"epoch": 1.6717168542151253,
"grad_norm": 16.61964235609142,
"learning_rate": 7.397719651240779e-06,
"loss": 2.2854,
"step": 1662
},
{
"epoch": 1.672722700697806,
"grad_norm": 17.253963452306547,
"learning_rate": 7.3954840152023256e-06,
"loss": 2.2387,
"step": 1663
},
{
"epoch": 1.6737285471804866,
"grad_norm": 18.0575343659353,
"learning_rate": 7.393248379163873e-06,
"loss": 2.3135,
"step": 1664
},
{
"epoch": 1.6747343936631671,
"grad_norm": 18.694711885768108,
"learning_rate": 7.3910127431254195e-06,
"loss": 2.2813,
"step": 1665
},
{
"epoch": 1.6757402401458479,
"grad_norm": 18.46894024543316,
"learning_rate": 7.388777107086968e-06,
"loss": 2.3289,
"step": 1666
},
{
"epoch": 1.6767460866285284,
"grad_norm": 18.075846886633336,
"learning_rate": 7.386541471048514e-06,
"loss": 2.2717,
"step": 1667
},
{
"epoch": 1.677751933111209,
"grad_norm": 16.580635934759897,
"learning_rate": 7.38430583501006e-06,
"loss": 2.3578,
"step": 1668
},
{
"epoch": 1.6787577795938895,
"grad_norm": 19.80057502326656,
"learning_rate": 7.382070198971608e-06,
"loss": 2.2358,
"step": 1669
},
{
"epoch": 1.67976362607657,
"grad_norm": 22.324624123091027,
"learning_rate": 7.379834562933155e-06,
"loss": 2.2999,
"step": 1670
},
{
"epoch": 1.6807694725592506,
"grad_norm": 17.171457668822228,
"learning_rate": 7.377598926894702e-06,
"loss": 2.2892,
"step": 1671
},
{
"epoch": 1.6817753190419311,
"grad_norm": 17.459691170633143,
"learning_rate": 7.375363290856249e-06,
"loss": 2.2847,
"step": 1672
},
{
"epoch": 1.6827811655246117,
"grad_norm": 24.349415622986733,
"learning_rate": 7.373127654817796e-06,
"loss": 2.3483,
"step": 1673
},
{
"epoch": 1.6837870120072924,
"grad_norm": 17.782930821929234,
"learning_rate": 7.370892018779343e-06,
"loss": 2.2817,
"step": 1674
},
{
"epoch": 1.684792858489973,
"grad_norm": 17.193230183146444,
"learning_rate": 7.368656382740891e-06,
"loss": 2.307,
"step": 1675
},
{
"epoch": 1.6857987049726535,
"grad_norm": 15.99168958034795,
"learning_rate": 7.366420746702438e-06,
"loss": 2.2668,
"step": 1676
},
{
"epoch": 1.6868045514553343,
"grad_norm": 16.455265608987492,
"learning_rate": 7.364185110663985e-06,
"loss": 2.2757,
"step": 1677
},
{
"epoch": 1.6878103979380148,
"grad_norm": 16.482362943706256,
"learning_rate": 7.361949474625532e-06,
"loss": 2.3429,
"step": 1678
},
{
"epoch": 1.6888162444206953,
"grad_norm": 15.595423288163719,
"learning_rate": 7.359713838587078e-06,
"loss": 2.3647,
"step": 1679
},
{
"epoch": 1.6898220909033759,
"grad_norm": 16.95527168452514,
"learning_rate": 7.3574782025486256e-06,
"loss": 2.2992,
"step": 1680
},
{
"epoch": 1.6908279373860564,
"grad_norm": 18.179540636982864,
"learning_rate": 7.355242566510172e-06,
"loss": 2.2744,
"step": 1681
},
{
"epoch": 1.691833783868737,
"grad_norm": 16.78451255952181,
"learning_rate": 7.35300693047172e-06,
"loss": 2.3529,
"step": 1682
},
{
"epoch": 1.6928396303514175,
"grad_norm": 16.309874549368278,
"learning_rate": 7.350771294433267e-06,
"loss": 2.2944,
"step": 1683
},
{
"epoch": 1.693845476834098,
"grad_norm": 17.405116712862426,
"learning_rate": 7.348535658394814e-06,
"loss": 2.276,
"step": 1684
},
{
"epoch": 1.6948513233167788,
"grad_norm": 17.55523954163725,
"learning_rate": 7.346300022356361e-06,
"loss": 2.2905,
"step": 1685
},
{
"epoch": 1.6958571697994593,
"grad_norm": 15.709914778295607,
"learning_rate": 7.344064386317908e-06,
"loss": 2.2726,
"step": 1686
},
{
"epoch": 1.69686301628214,
"grad_norm": 17.876951517815804,
"learning_rate": 7.341828750279455e-06,
"loss": 2.2789,
"step": 1687
},
{
"epoch": 1.6978688627648206,
"grad_norm": 18.745107840949895,
"learning_rate": 7.339593114241002e-06,
"loss": 2.2937,
"step": 1688
},
{
"epoch": 1.6988747092475012,
"grad_norm": 16.75408815841215,
"learning_rate": 7.337357478202549e-06,
"loss": 2.2737,
"step": 1689
},
{
"epoch": 1.6998805557301817,
"grad_norm": 16.50796612604206,
"learning_rate": 7.335121842164097e-06,
"loss": 2.2846,
"step": 1690
},
{
"epoch": 1.7008864022128622,
"grad_norm": 18.436456290854704,
"learning_rate": 7.332886206125644e-06,
"loss": 2.3165,
"step": 1691
},
{
"epoch": 1.7018922486955428,
"grad_norm": 15.577438332324887,
"learning_rate": 7.33065057008719e-06,
"loss": 2.2849,
"step": 1692
},
{
"epoch": 1.7028980951782233,
"grad_norm": 19.36486387739182,
"learning_rate": 7.328414934048738e-06,
"loss": 2.3413,
"step": 1693
},
{
"epoch": 1.7039039416609039,
"grad_norm": 17.7840086570158,
"learning_rate": 7.326179298010284e-06,
"loss": 2.2739,
"step": 1694
},
{
"epoch": 1.7049097881435846,
"grad_norm": 19.96875615564056,
"learning_rate": 7.3239436619718316e-06,
"loss": 2.2725,
"step": 1695
},
{
"epoch": 1.7059156346262652,
"grad_norm": 18.895880230471608,
"learning_rate": 7.321708025933378e-06,
"loss": 2.2451,
"step": 1696
},
{
"epoch": 1.7069214811089457,
"grad_norm": 18.494732987067664,
"learning_rate": 7.319472389894926e-06,
"loss": 2.3039,
"step": 1697
},
{
"epoch": 1.7079273275916265,
"grad_norm": 19.548118249517298,
"learning_rate": 7.317236753856473e-06,
"loss": 2.2646,
"step": 1698
},
{
"epoch": 1.708933174074307,
"grad_norm": 17.483404102959916,
"learning_rate": 7.31500111781802e-06,
"loss": 2.2878,
"step": 1699
},
{
"epoch": 1.7099390205569875,
"grad_norm": 16.087083784695995,
"learning_rate": 7.312765481779567e-06,
"loss": 2.2983,
"step": 1700
},
{
"epoch": 1.710944867039668,
"grad_norm": 17.286227907129362,
"learning_rate": 7.310529845741114e-06,
"loss": 2.2769,
"step": 1701
},
{
"epoch": 1.7119507135223486,
"grad_norm": 18.121685621444016,
"learning_rate": 7.308294209702661e-06,
"loss": 2.3164,
"step": 1702
},
{
"epoch": 1.7129565600050292,
"grad_norm": 16.93381419733215,
"learning_rate": 7.3060585736642074e-06,
"loss": 2.2904,
"step": 1703
},
{
"epoch": 1.7139624064877097,
"grad_norm": 17.108344692926316,
"learning_rate": 7.303822937625755e-06,
"loss": 2.2702,
"step": 1704
},
{
"epoch": 1.7149682529703902,
"grad_norm": 16.904685203090065,
"learning_rate": 7.301587301587301e-06,
"loss": 2.3053,
"step": 1705
},
{
"epoch": 1.715974099453071,
"grad_norm": 15.891046973165288,
"learning_rate": 7.29935166554885e-06,
"loss": 2.2788,
"step": 1706
},
{
"epoch": 1.7169799459357515,
"grad_norm": 18.961009088854595,
"learning_rate": 7.297116029510396e-06,
"loss": 2.2647,
"step": 1707
},
{
"epoch": 1.7179857924184323,
"grad_norm": 16.77002633978821,
"learning_rate": 7.294880393471944e-06,
"loss": 2.2649,
"step": 1708
},
{
"epoch": 1.7189916389011128,
"grad_norm": 19.15137573907095,
"learning_rate": 7.29264475743349e-06,
"loss": 2.2612,
"step": 1709
},
{
"epoch": 1.7199974853837934,
"grad_norm": 19.365000785728903,
"learning_rate": 7.290409121395038e-06,
"loss": 2.2832,
"step": 1710
},
{
"epoch": 1.721003331866474,
"grad_norm": 18.550525488582906,
"learning_rate": 7.288173485356584e-06,
"loss": 2.2698,
"step": 1711
},
{
"epoch": 1.7220091783491545,
"grad_norm": 19.045700127374538,
"learning_rate": 7.285937849318132e-06,
"loss": 2.2863,
"step": 1712
},
{
"epoch": 1.723015024831835,
"grad_norm": 16.589908676820695,
"learning_rate": 7.283702213279679e-06,
"loss": 2.3521,
"step": 1713
},
{
"epoch": 1.7240208713145155,
"grad_norm": 18.730596467816195,
"learning_rate": 7.2814665772412255e-06,
"loss": 2.3367,
"step": 1714
},
{
"epoch": 1.725026717797196,
"grad_norm": 21.02491227040751,
"learning_rate": 7.279230941202773e-06,
"loss": 2.2994,
"step": 1715
},
{
"epoch": 1.7260325642798768,
"grad_norm": 18.90894267852598,
"learning_rate": 7.2769953051643195e-06,
"loss": 2.2979,
"step": 1716
},
{
"epoch": 1.7270384107625574,
"grad_norm": 16.90454659841854,
"learning_rate": 7.274759669125867e-06,
"loss": 2.2959,
"step": 1717
},
{
"epoch": 1.728044257245238,
"grad_norm": 17.165608822801428,
"learning_rate": 7.2725240330874135e-06,
"loss": 2.3404,
"step": 1718
},
{
"epoch": 1.7290501037279187,
"grad_norm": 19.372474192014376,
"learning_rate": 7.270288397048961e-06,
"loss": 2.3036,
"step": 1719
},
{
"epoch": 1.7300559502105992,
"grad_norm": 18.182535106181874,
"learning_rate": 7.268052761010507e-06,
"loss": 2.3857,
"step": 1720
},
{
"epoch": 1.7310617966932798,
"grad_norm": 17.114519608407026,
"learning_rate": 7.265817124972056e-06,
"loss": 2.2924,
"step": 1721
},
{
"epoch": 1.7320676431759603,
"grad_norm": 16.53904728868979,
"learning_rate": 7.263581488933602e-06,
"loss": 2.3151,
"step": 1722
},
{
"epoch": 1.7330734896586408,
"grad_norm": 18.726040653410717,
"learning_rate": 7.26134585289515e-06,
"loss": 2.2739,
"step": 1723
},
{
"epoch": 1.7340793361413214,
"grad_norm": 17.341102661687714,
"learning_rate": 7.259110216856696e-06,
"loss": 2.2504,
"step": 1724
},
{
"epoch": 1.735085182624002,
"grad_norm": 16.656227143427788,
"learning_rate": 7.256874580818244e-06,
"loss": 2.2073,
"step": 1725
},
{
"epoch": 1.7360910291066824,
"grad_norm": 15.628860790894302,
"learning_rate": 7.25463894477979e-06,
"loss": 2.2826,
"step": 1726
},
{
"epoch": 1.7370968755893632,
"grad_norm": 16.375635653933628,
"learning_rate": 7.252403308741337e-06,
"loss": 2.2749,
"step": 1727
},
{
"epoch": 1.7381027220720437,
"grad_norm": 16.301987152653233,
"learning_rate": 7.250167672702885e-06,
"loss": 2.3258,
"step": 1728
},
{
"epoch": 1.7391085685547243,
"grad_norm": 18.317871380977778,
"learning_rate": 7.2479320366644315e-06,
"loss": 2.2517,
"step": 1729
},
{
"epoch": 1.740114415037405,
"grad_norm": 20.755179037495648,
"learning_rate": 7.245696400625979e-06,
"loss": 2.2881,
"step": 1730
},
{
"epoch": 1.7411202615200856,
"grad_norm": 20.223197943734334,
"learning_rate": 7.2434607645875255e-06,
"loss": 2.2787,
"step": 1731
},
{
"epoch": 1.7421261080027661,
"grad_norm": 18.66793302042133,
"learning_rate": 7.241225128549073e-06,
"loss": 2.2898,
"step": 1732
},
{
"epoch": 1.7431319544854467,
"grad_norm": 18.28608426011941,
"learning_rate": 7.2389894925106195e-06,
"loss": 2.3038,
"step": 1733
},
{
"epoch": 1.7441378009681272,
"grad_norm": 20.186986215013036,
"learning_rate": 7.236753856472168e-06,
"loss": 2.2693,
"step": 1734
},
{
"epoch": 1.7451436474508077,
"grad_norm": 21.940252910209793,
"learning_rate": 7.234518220433714e-06,
"loss": 2.3508,
"step": 1735
},
{
"epoch": 1.7461494939334883,
"grad_norm": 16.54537436899838,
"learning_rate": 7.232282584395262e-06,
"loss": 2.2544,
"step": 1736
},
{
"epoch": 1.7471553404161688,
"grad_norm": 19.29470763757215,
"learning_rate": 7.230046948356808e-06,
"loss": 2.2422,
"step": 1737
},
{
"epoch": 1.7481611868988496,
"grad_norm": 19.38854319430838,
"learning_rate": 7.227811312318355e-06,
"loss": 2.3005,
"step": 1738
},
{
"epoch": 1.7491670333815301,
"grad_norm": 17.357463170532714,
"learning_rate": 7.225575676279902e-06,
"loss": 2.2972,
"step": 1739
},
{
"epoch": 1.7501728798642109,
"grad_norm": 22.087290432789963,
"learning_rate": 7.223340040241449e-06,
"loss": 2.2539,
"step": 1740
},
{
"epoch": 1.7511787263468914,
"grad_norm": 17.76858889717604,
"learning_rate": 7.221104404202996e-06,
"loss": 2.2746,
"step": 1741
},
{
"epoch": 1.752184572829572,
"grad_norm": 18.116519033951814,
"learning_rate": 7.218868768164543e-06,
"loss": 2.2829,
"step": 1742
},
{
"epoch": 1.7531904193122525,
"grad_norm": 20.309504486994047,
"learning_rate": 7.216633132126091e-06,
"loss": 2.3595,
"step": 1743
},
{
"epoch": 1.754196265794933,
"grad_norm": 14.93742846213375,
"learning_rate": 7.2143974960876376e-06,
"loss": 2.2892,
"step": 1744
},
{
"epoch": 1.7552021122776136,
"grad_norm": 19.603616852280457,
"learning_rate": 7.212161860049185e-06,
"loss": 2.3173,
"step": 1745
},
{
"epoch": 1.7562079587602941,
"grad_norm": 18.398664249151416,
"learning_rate": 7.2099262240107315e-06,
"loss": 2.344,
"step": 1746
},
{
"epoch": 1.7572138052429747,
"grad_norm": 18.49824732739366,
"learning_rate": 7.207690587972279e-06,
"loss": 2.272,
"step": 1747
},
{
"epoch": 1.7582196517256554,
"grad_norm": 18.83743911457428,
"learning_rate": 7.2054549519338255e-06,
"loss": 2.2819,
"step": 1748
},
{
"epoch": 1.759225498208336,
"grad_norm": 16.68493207885488,
"learning_rate": 7.203219315895374e-06,
"loss": 2.2825,
"step": 1749
},
{
"epoch": 1.7602313446910165,
"grad_norm": 17.39084821792995,
"learning_rate": 7.20098367985692e-06,
"loss": 2.2607,
"step": 1750
},
{
"epoch": 1.7612371911736973,
"grad_norm": 16.26128535383905,
"learning_rate": 7.198748043818467e-06,
"loss": 2.3199,
"step": 1751
},
{
"epoch": 1.7622430376563778,
"grad_norm": 18.49950263120119,
"learning_rate": 7.196512407780014e-06,
"loss": 2.2978,
"step": 1752
},
{
"epoch": 1.7632488841390583,
"grad_norm": 16.270202570964813,
"learning_rate": 7.194276771741561e-06,
"loss": 2.2844,
"step": 1753
},
{
"epoch": 1.7642547306217389,
"grad_norm": 17.054025826733476,
"learning_rate": 7.192041135703108e-06,
"loss": 2.3269,
"step": 1754
},
{
"epoch": 1.7652605771044194,
"grad_norm": 18.66246669846856,
"learning_rate": 7.189805499664655e-06,
"loss": 2.2251,
"step": 1755
},
{
"epoch": 1.7662664235871,
"grad_norm": 20.331441036286925,
"learning_rate": 7.187569863626202e-06,
"loss": 2.2769,
"step": 1756
},
{
"epoch": 1.7672722700697805,
"grad_norm": 17.734428888276117,
"learning_rate": 7.185334227587749e-06,
"loss": 2.3053,
"step": 1757
},
{
"epoch": 1.768278116552461,
"grad_norm": 18.036348076403122,
"learning_rate": 7.183098591549297e-06,
"loss": 2.248,
"step": 1758
},
{
"epoch": 1.7692839630351418,
"grad_norm": 17.82023633035522,
"learning_rate": 7.180862955510844e-06,
"loss": 2.3191,
"step": 1759
},
{
"epoch": 1.7702898095178223,
"grad_norm": 16.823322658814927,
"learning_rate": 7.178627319472391e-06,
"loss": 2.3104,
"step": 1760
},
{
"epoch": 1.7712956560005029,
"grad_norm": 18.89933972226583,
"learning_rate": 7.1763916834339375e-06,
"loss": 2.2202,
"step": 1761
},
{
"epoch": 1.7723015024831836,
"grad_norm": 19.642607646851978,
"learning_rate": 7.174156047395484e-06,
"loss": 2.2942,
"step": 1762
},
{
"epoch": 1.7733073489658642,
"grad_norm": 18.254196769125357,
"learning_rate": 7.1719204113570315e-06,
"loss": 2.2822,
"step": 1763
},
{
"epoch": 1.7743131954485447,
"grad_norm": 18.216364417776393,
"learning_rate": 7.169684775318578e-06,
"loss": 2.283,
"step": 1764
},
{
"epoch": 1.7753190419312253,
"grad_norm": 18.098869609136802,
"learning_rate": 7.167449139280126e-06,
"loss": 2.3405,
"step": 1765
},
{
"epoch": 1.7763248884139058,
"grad_norm": 17.066412960592956,
"learning_rate": 7.165213503241673e-06,
"loss": 2.2829,
"step": 1766
},
{
"epoch": 1.7773307348965863,
"grad_norm": 16.025984713835253,
"learning_rate": 7.16297786720322e-06,
"loss": 2.3012,
"step": 1767
},
{
"epoch": 1.7783365813792669,
"grad_norm": 19.686451528761513,
"learning_rate": 7.160742231164767e-06,
"loss": 2.2475,
"step": 1768
},
{
"epoch": 1.7793424278619474,
"grad_norm": 17.114132490787796,
"learning_rate": 7.158506595126314e-06,
"loss": 2.2603,
"step": 1769
},
{
"epoch": 1.7803482743446282,
"grad_norm": 16.788287849009663,
"learning_rate": 7.156270959087861e-06,
"loss": 2.2784,
"step": 1770
},
{
"epoch": 1.7813541208273087,
"grad_norm": 14.660268269237488,
"learning_rate": 7.154035323049408e-06,
"loss": 2.2773,
"step": 1771
},
{
"epoch": 1.7823599673099895,
"grad_norm": 17.33012795252929,
"learning_rate": 7.151799687010955e-06,
"loss": 2.3154,
"step": 1772
},
{
"epoch": 1.78336581379267,
"grad_norm": 16.183451941924876,
"learning_rate": 7.149564050972501e-06,
"loss": 2.2714,
"step": 1773
},
{
"epoch": 1.7843716602753505,
"grad_norm": 17.502249863651574,
"learning_rate": 7.14732841493405e-06,
"loss": 2.2987,
"step": 1774
},
{
"epoch": 1.785377506758031,
"grad_norm": 16.267689850773287,
"learning_rate": 7.145092778895596e-06,
"loss": 2.2764,
"step": 1775
},
{
"epoch": 1.7863833532407116,
"grad_norm": 17.185209369525268,
"learning_rate": 7.1428571428571436e-06,
"loss": 2.3588,
"step": 1776
},
{
"epoch": 1.7873891997233922,
"grad_norm": 20.347299173020325,
"learning_rate": 7.14062150681869e-06,
"loss": 2.264,
"step": 1777
},
{
"epoch": 1.7883950462060727,
"grad_norm": 15.444704186788066,
"learning_rate": 7.1383858707802375e-06,
"loss": 2.3126,
"step": 1778
},
{
"epoch": 1.7894008926887532,
"grad_norm": 18.125816182173008,
"learning_rate": 7.136150234741784e-06,
"loss": 2.2479,
"step": 1779
},
{
"epoch": 1.790406739171434,
"grad_norm": 16.508046759412878,
"learning_rate": 7.133914598703332e-06,
"loss": 2.2783,
"step": 1780
},
{
"epoch": 1.7914125856541145,
"grad_norm": 17.160063240198856,
"learning_rate": 7.131678962664879e-06,
"loss": 2.2799,
"step": 1781
},
{
"epoch": 1.792418432136795,
"grad_norm": 18.21495266317122,
"learning_rate": 7.129443326626426e-06,
"loss": 2.2865,
"step": 1782
},
{
"epoch": 1.7934242786194758,
"grad_norm": 17.027319498991748,
"learning_rate": 7.127207690587973e-06,
"loss": 2.2368,
"step": 1783
},
{
"epoch": 1.7944301251021564,
"grad_norm": 16.229008671664687,
"learning_rate": 7.12497205454952e-06,
"loss": 2.2556,
"step": 1784
},
{
"epoch": 1.795435971584837,
"grad_norm": 18.62974074603085,
"learning_rate": 7.122736418511067e-06,
"loss": 2.3042,
"step": 1785
},
{
"epoch": 1.7964418180675175,
"grad_norm": 14.934846550712225,
"learning_rate": 7.120500782472613e-06,
"loss": 2.262,
"step": 1786
},
{
"epoch": 1.797447664550198,
"grad_norm": 17.7843294903222,
"learning_rate": 7.118265146434161e-06,
"loss": 2.2989,
"step": 1787
},
{
"epoch": 1.7984535110328785,
"grad_norm": 16.043759669850164,
"learning_rate": 7.116029510395707e-06,
"loss": 2.2866,
"step": 1788
},
{
"epoch": 1.799459357515559,
"grad_norm": 15.572175597039415,
"learning_rate": 7.113793874357256e-06,
"loss": 2.2915,
"step": 1789
},
{
"epoch": 1.8004652039982396,
"grad_norm": 18.262520340255662,
"learning_rate": 7.111558238318802e-06,
"loss": 2.2982,
"step": 1790
},
{
"epoch": 1.8014710504809204,
"grad_norm": 14.664239377399232,
"learning_rate": 7.10932260228035e-06,
"loss": 2.2688,
"step": 1791
},
{
"epoch": 1.802476896963601,
"grad_norm": 15.788192217263258,
"learning_rate": 7.107086966241896e-06,
"loss": 2.2728,
"step": 1792
},
{
"epoch": 1.8034827434462817,
"grad_norm": 17.241784049935667,
"learning_rate": 7.1048513302034435e-06,
"loss": 2.2513,
"step": 1793
},
{
"epoch": 1.8044885899289622,
"grad_norm": 18.319754527885618,
"learning_rate": 7.10261569416499e-06,
"loss": 2.2543,
"step": 1794
},
{
"epoch": 1.8054944364116428,
"grad_norm": 16.093204298174406,
"learning_rate": 7.100380058126538e-06,
"loss": 2.2525,
"step": 1795
},
{
"epoch": 1.8065002828943233,
"grad_norm": 16.992116261097483,
"learning_rate": 7.098144422088085e-06,
"loss": 2.2809,
"step": 1796
},
{
"epoch": 1.8075061293770038,
"grad_norm": 19.250314664557777,
"learning_rate": 7.0959087860496315e-06,
"loss": 2.2567,
"step": 1797
},
{
"epoch": 1.8085119758596844,
"grad_norm": 15.994902899251477,
"learning_rate": 7.093673150011179e-06,
"loss": 2.2965,
"step": 1798
},
{
"epoch": 1.809517822342365,
"grad_norm": 18.988591404386828,
"learning_rate": 7.0914375139727254e-06,
"loss": 2.2487,
"step": 1799
},
{
"epoch": 1.8105236688250455,
"grad_norm": 19.11487194496845,
"learning_rate": 7.089201877934273e-06,
"loss": 2.2785,
"step": 1800
},
{
"epoch": 1.8115295153077262,
"grad_norm": 19.275348750335183,
"learning_rate": 7.086966241895819e-06,
"loss": 2.3006,
"step": 1801
},
{
"epoch": 1.8125353617904068,
"grad_norm": 18.60674521941389,
"learning_rate": 7.084730605857368e-06,
"loss": 2.3034,
"step": 1802
},
{
"epoch": 1.8135412082730873,
"grad_norm": 16.767075586894872,
"learning_rate": 7.082494969818914e-06,
"loss": 2.3288,
"step": 1803
},
{
"epoch": 1.814547054755768,
"grad_norm": 20.504218280601204,
"learning_rate": 7.080259333780462e-06,
"loss": 2.3001,
"step": 1804
},
{
"epoch": 1.8155529012384486,
"grad_norm": 20.98661191621691,
"learning_rate": 7.078023697742008e-06,
"loss": 2.2785,
"step": 1805
},
{
"epoch": 1.8165587477211291,
"grad_norm": 17.72071568357678,
"learning_rate": 7.075788061703556e-06,
"loss": 2.2838,
"step": 1806
},
{
"epoch": 1.8175645942038097,
"grad_norm": 17.78877227393904,
"learning_rate": 7.073552425665102e-06,
"loss": 2.2929,
"step": 1807
},
{
"epoch": 1.8185704406864902,
"grad_norm": 19.519678375167413,
"learning_rate": 7.0713167896266496e-06,
"loss": 2.2756,
"step": 1808
},
{
"epoch": 1.8195762871691707,
"grad_norm": 16.776782550396653,
"learning_rate": 7.069081153588196e-06,
"loss": 2.2822,
"step": 1809
},
{
"epoch": 1.8205821336518513,
"grad_norm": 17.04682649251197,
"learning_rate": 7.066845517549743e-06,
"loss": 2.2467,
"step": 1810
},
{
"epoch": 1.8215879801345318,
"grad_norm": 21.9081731265325,
"learning_rate": 7.064609881511291e-06,
"loss": 2.2639,
"step": 1811
},
{
"epoch": 1.8225938266172126,
"grad_norm": 21.67762882011947,
"learning_rate": 7.0623742454728375e-06,
"loss": 2.3243,
"step": 1812
},
{
"epoch": 1.8235996730998931,
"grad_norm": 14.755210379226378,
"learning_rate": 7.060138609434385e-06,
"loss": 2.2856,
"step": 1813
},
{
"epoch": 1.8246055195825737,
"grad_norm": 20.578781544696486,
"learning_rate": 7.0579029733959315e-06,
"loss": 2.2983,
"step": 1814
},
{
"epoch": 1.8256113660652544,
"grad_norm": 21.197669630731394,
"learning_rate": 7.055667337357479e-06,
"loss": 2.2639,
"step": 1815
},
{
"epoch": 1.826617212547935,
"grad_norm": 18.485968760798357,
"learning_rate": 7.0534317013190254e-06,
"loss": 2.3215,
"step": 1816
},
{
"epoch": 1.8276230590306155,
"grad_norm": 18.363711510600858,
"learning_rate": 7.051196065280574e-06,
"loss": 2.3173,
"step": 1817
},
{
"epoch": 1.828628905513296,
"grad_norm": 22.856385538734763,
"learning_rate": 7.04896042924212e-06,
"loss": 2.2784,
"step": 1818
},
{
"epoch": 1.8296347519959766,
"grad_norm": 17.64647189595528,
"learning_rate": 7.046724793203668e-06,
"loss": 2.1983,
"step": 1819
},
{
"epoch": 1.8306405984786571,
"grad_norm": 19.867650267325015,
"learning_rate": 7.044489157165214e-06,
"loss": 2.3085,
"step": 1820
},
{
"epoch": 1.8316464449613377,
"grad_norm": 18.776829497143535,
"learning_rate": 7.042253521126761e-06,
"loss": 2.2682,
"step": 1821
},
{
"epoch": 1.8326522914440182,
"grad_norm": 16.021203164915136,
"learning_rate": 7.040017885088308e-06,
"loss": 2.2734,
"step": 1822
},
{
"epoch": 1.833658137926699,
"grad_norm": 16.417856249676,
"learning_rate": 7.037782249049855e-06,
"loss": 2.2676,
"step": 1823
},
{
"epoch": 1.8346639844093795,
"grad_norm": 16.722581854094724,
"learning_rate": 7.035546613011402e-06,
"loss": 2.2728,
"step": 1824
},
{
"epoch": 1.8356698308920603,
"grad_norm": 17.47207904072863,
"learning_rate": 7.033310976972949e-06,
"loss": 2.2823,
"step": 1825
},
{
"epoch": 1.8366756773747408,
"grad_norm": 18.9806791066407,
"learning_rate": 7.031075340934497e-06,
"loss": 2.3478,
"step": 1826
},
{
"epoch": 1.8376815238574213,
"grad_norm": 16.279698640644096,
"learning_rate": 7.0288397048960435e-06,
"loss": 2.2949,
"step": 1827
},
{
"epoch": 1.8386873703401019,
"grad_norm": 17.345829236953485,
"learning_rate": 7.026604068857591e-06,
"loss": 2.3412,
"step": 1828
},
{
"epoch": 1.8396932168227824,
"grad_norm": 19.52245272171043,
"learning_rate": 7.0243684328191375e-06,
"loss": 2.3069,
"step": 1829
},
{
"epoch": 1.840699063305463,
"grad_norm": 20.41333556532244,
"learning_rate": 7.022132796780685e-06,
"loss": 2.2643,
"step": 1830
},
{
"epoch": 1.8417049097881435,
"grad_norm": 16.19207248247733,
"learning_rate": 7.0198971607422314e-06,
"loss": 2.2304,
"step": 1831
},
{
"epoch": 1.842710756270824,
"grad_norm": 17.371107507447974,
"learning_rate": 7.017661524703778e-06,
"loss": 2.3066,
"step": 1832
},
{
"epoch": 1.8437166027535048,
"grad_norm": 15.865840975698038,
"learning_rate": 7.015425888665326e-06,
"loss": 2.2728,
"step": 1833
},
{
"epoch": 1.8447224492361853,
"grad_norm": 18.433634282678582,
"learning_rate": 7.013190252626873e-06,
"loss": 2.2587,
"step": 1834
},
{
"epoch": 1.8457282957188659,
"grad_norm": 19.746009444985077,
"learning_rate": 7.01095461658842e-06,
"loss": 2.3259,
"step": 1835
},
{
"epoch": 1.8467341422015466,
"grad_norm": 19.45097779972063,
"learning_rate": 7.008718980549967e-06,
"loss": 2.3208,
"step": 1836
},
{
"epoch": 1.8477399886842272,
"grad_norm": 17.122946543898664,
"learning_rate": 7.006483344511514e-06,
"loss": 2.2804,
"step": 1837
},
{
"epoch": 1.8487458351669077,
"grad_norm": 17.990726363404274,
"learning_rate": 7.004247708473061e-06,
"loss": 2.2692,
"step": 1838
},
{
"epoch": 1.8497516816495883,
"grad_norm": 17.06390467461542,
"learning_rate": 7.002012072434608e-06,
"loss": 2.2984,
"step": 1839
},
{
"epoch": 1.8507575281322688,
"grad_norm": 20.004890911595176,
"learning_rate": 6.999776436396155e-06,
"loss": 2.3004,
"step": 1840
},
{
"epoch": 1.8517633746149493,
"grad_norm": 19.17768686986296,
"learning_rate": 6.997540800357703e-06,
"loss": 2.2812,
"step": 1841
},
{
"epoch": 1.8527692210976299,
"grad_norm": 15.70661341588764,
"learning_rate": 6.9953051643192495e-06,
"loss": 2.2868,
"step": 1842
},
{
"epoch": 1.8537750675803104,
"grad_norm": 16.556570727346433,
"learning_rate": 6.993069528280797e-06,
"loss": 2.2866,
"step": 1843
},
{
"epoch": 1.8547809140629912,
"grad_norm": 15.967921511409507,
"learning_rate": 6.9908338922423435e-06,
"loss": 2.249,
"step": 1844
},
{
"epoch": 1.8557867605456717,
"grad_norm": 16.70715263496682,
"learning_rate": 6.98859825620389e-06,
"loss": 2.3087,
"step": 1845
},
{
"epoch": 1.8567926070283522,
"grad_norm": 20.495584603518225,
"learning_rate": 6.9863626201654375e-06,
"loss": 2.2572,
"step": 1846
},
{
"epoch": 1.857798453511033,
"grad_norm": 17.47988908055069,
"learning_rate": 6.984126984126984e-06,
"loss": 2.2883,
"step": 1847
},
{
"epoch": 1.8588042999937135,
"grad_norm": 17.017278697671724,
"learning_rate": 6.981891348088532e-06,
"loss": 2.2793,
"step": 1848
},
{
"epoch": 1.859810146476394,
"grad_norm": 18.59038874515862,
"learning_rate": 6.979655712050079e-06,
"loss": 2.2574,
"step": 1849
},
{
"epoch": 1.8608159929590746,
"grad_norm": 19.53865063389471,
"learning_rate": 6.977420076011626e-06,
"loss": 2.3442,
"step": 1850
},
{
"epoch": 1.8618218394417552,
"grad_norm": 16.338863083503824,
"learning_rate": 6.975184439973173e-06,
"loss": 2.2877,
"step": 1851
},
{
"epoch": 1.8628276859244357,
"grad_norm": 17.241535588150537,
"learning_rate": 6.97294880393472e-06,
"loss": 2.296,
"step": 1852
},
{
"epoch": 1.8638335324071162,
"grad_norm": 16.442352629632072,
"learning_rate": 6.970713167896267e-06,
"loss": 2.3199,
"step": 1853
},
{
"epoch": 1.8648393788897968,
"grad_norm": 18.764930586467397,
"learning_rate": 6.968477531857815e-06,
"loss": 2.2887,
"step": 1854
},
{
"epoch": 1.8658452253724775,
"grad_norm": 16.0002800988153,
"learning_rate": 6.966241895819362e-06,
"loss": 2.267,
"step": 1855
},
{
"epoch": 1.866851071855158,
"grad_norm": 16.1416440519751,
"learning_rate": 6.964006259780907e-06,
"loss": 2.2854,
"step": 1856
},
{
"epoch": 1.8678569183378388,
"grad_norm": 18.03816884235956,
"learning_rate": 6.9617706237424556e-06,
"loss": 2.2578,
"step": 1857
},
{
"epoch": 1.8688627648205194,
"grad_norm": 15.317109741476976,
"learning_rate": 6.959534987704002e-06,
"loss": 2.263,
"step": 1858
},
{
"epoch": 1.8698686113032,
"grad_norm": 14.828646134819525,
"learning_rate": 6.9572993516655495e-06,
"loss": 2.2415,
"step": 1859
},
{
"epoch": 1.8708744577858805,
"grad_norm": 14.97271984899026,
"learning_rate": 6.955063715627096e-06,
"loss": 2.2852,
"step": 1860
},
{
"epoch": 1.871880304268561,
"grad_norm": 16.65094367595834,
"learning_rate": 6.9528280795886435e-06,
"loss": 2.3065,
"step": 1861
},
{
"epoch": 1.8728861507512415,
"grad_norm": 15.882338299411797,
"learning_rate": 6.95059244355019e-06,
"loss": 2.2813,
"step": 1862
},
{
"epoch": 1.873891997233922,
"grad_norm": 15.05351225160577,
"learning_rate": 6.948356807511738e-06,
"loss": 2.2995,
"step": 1863
},
{
"epoch": 1.8748978437166026,
"grad_norm": 17.677697128378234,
"learning_rate": 6.946121171473285e-06,
"loss": 2.2247,
"step": 1864
},
{
"epoch": 1.8759036901992834,
"grad_norm": 14.832365800788716,
"learning_rate": 6.943885535434832e-06,
"loss": 2.3025,
"step": 1865
},
{
"epoch": 1.876909536681964,
"grad_norm": 15.956879223213827,
"learning_rate": 6.941649899396379e-06,
"loss": 2.3567,
"step": 1866
},
{
"epoch": 1.8779153831646445,
"grad_norm": 17.36600813497739,
"learning_rate": 6.939414263357926e-06,
"loss": 2.32,
"step": 1867
},
{
"epoch": 1.8789212296473252,
"grad_norm": 16.959525128835914,
"learning_rate": 6.937178627319473e-06,
"loss": 2.3037,
"step": 1868
},
{
"epoch": 1.8799270761300058,
"grad_norm": 16.21373348481673,
"learning_rate": 6.934942991281019e-06,
"loss": 2.3315,
"step": 1869
},
{
"epoch": 1.8809329226126863,
"grad_norm": 16.510470523686852,
"learning_rate": 6.932707355242568e-06,
"loss": 2.2608,
"step": 1870
},
{
"epoch": 1.8819387690953668,
"grad_norm": 17.030510814810018,
"learning_rate": 6.930471719204114e-06,
"loss": 2.2677,
"step": 1871
},
{
"epoch": 1.8829446155780474,
"grad_norm": 18.22176597603188,
"learning_rate": 6.9282360831656616e-06,
"loss": 2.3149,
"step": 1872
},
{
"epoch": 1.883950462060728,
"grad_norm": 16.792187637096355,
"learning_rate": 6.926000447127208e-06,
"loss": 2.2751,
"step": 1873
},
{
"epoch": 1.8849563085434085,
"grad_norm": 16.183554558324985,
"learning_rate": 6.9237648110887555e-06,
"loss": 2.2863,
"step": 1874
},
{
"epoch": 1.885962155026089,
"grad_norm": 17.300579827350667,
"learning_rate": 6.921529175050302e-06,
"loss": 2.2912,
"step": 1875
},
{
"epoch": 1.8869680015087698,
"grad_norm": 18.153546483178403,
"learning_rate": 6.9192935390118495e-06,
"loss": 2.2814,
"step": 1876
},
{
"epoch": 1.8879738479914503,
"grad_norm": 19.594562516452527,
"learning_rate": 6.917057902973396e-06,
"loss": 2.3179,
"step": 1877
},
{
"epoch": 1.8889796944741308,
"grad_norm": 17.096198547393634,
"learning_rate": 6.914822266934944e-06,
"loss": 2.3012,
"step": 1878
},
{
"epoch": 1.8899855409568116,
"grad_norm": 18.147615561617283,
"learning_rate": 6.912586630896491e-06,
"loss": 2.2432,
"step": 1879
},
{
"epoch": 1.8909913874394921,
"grad_norm": 16.826685763955773,
"learning_rate": 6.9103509948580374e-06,
"loss": 2.3103,
"step": 1880
},
{
"epoch": 1.8919972339221727,
"grad_norm": 16.737903754646183,
"learning_rate": 6.908115358819585e-06,
"loss": 2.3051,
"step": 1881
},
{
"epoch": 1.8930030804048532,
"grad_norm": 16.529744230872915,
"learning_rate": 6.905879722781131e-06,
"loss": 2.2755,
"step": 1882
},
{
"epoch": 1.8940089268875338,
"grad_norm": 19.09942006258242,
"learning_rate": 6.903644086742679e-06,
"loss": 2.3269,
"step": 1883
},
{
"epoch": 1.8950147733702143,
"grad_norm": 19.531445891717162,
"learning_rate": 6.901408450704225e-06,
"loss": 2.2548,
"step": 1884
},
{
"epoch": 1.8960206198528948,
"grad_norm": 18.997923333530903,
"learning_rate": 6.899172814665774e-06,
"loss": 2.3261,
"step": 1885
},
{
"epoch": 1.8970264663355754,
"grad_norm": 18.461677933705914,
"learning_rate": 6.89693717862732e-06,
"loss": 2.2897,
"step": 1886
},
{
"epoch": 1.8980323128182561,
"grad_norm": 18.61814161726463,
"learning_rate": 6.894701542588868e-06,
"loss": 2.3123,
"step": 1887
},
{
"epoch": 1.8990381593009367,
"grad_norm": 18.092851713644524,
"learning_rate": 6.892465906550414e-06,
"loss": 2.2905,
"step": 1888
},
{
"epoch": 1.9000440057836174,
"grad_norm": 17.044430585074025,
"learning_rate": 6.8902302705119616e-06,
"loss": 2.2941,
"step": 1889
},
{
"epoch": 1.901049852266298,
"grad_norm": 17.0026458608325,
"learning_rate": 6.887994634473508e-06,
"loss": 2.2681,
"step": 1890
},
{
"epoch": 1.9020556987489785,
"grad_norm": 15.129874898168884,
"learning_rate": 6.885758998435055e-06,
"loss": 2.3103,
"step": 1891
},
{
"epoch": 1.903061545231659,
"grad_norm": 18.273974027399,
"learning_rate": 6.883523362396602e-06,
"loss": 2.3008,
"step": 1892
},
{
"epoch": 1.9040673917143396,
"grad_norm": 19.742323390440422,
"learning_rate": 6.881287726358149e-06,
"loss": 2.2705,
"step": 1893
},
{
"epoch": 1.9050732381970201,
"grad_norm": 17.714002726675265,
"learning_rate": 6.879052090319697e-06,
"loss": 2.263,
"step": 1894
},
{
"epoch": 1.9060790846797007,
"grad_norm": 18.144590858661886,
"learning_rate": 6.8768164542812435e-06,
"loss": 2.2798,
"step": 1895
},
{
"epoch": 1.9070849311623812,
"grad_norm": 19.275985168763043,
"learning_rate": 6.874580818242791e-06,
"loss": 2.2764,
"step": 1896
},
{
"epoch": 1.908090777645062,
"grad_norm": 16.984422505971622,
"learning_rate": 6.872345182204337e-06,
"loss": 2.2937,
"step": 1897
},
{
"epoch": 1.9090966241277425,
"grad_norm": 17.473288710210316,
"learning_rate": 6.870109546165885e-06,
"loss": 2.3078,
"step": 1898
},
{
"epoch": 1.910102470610423,
"grad_norm": 17.934114969579213,
"learning_rate": 6.867873910127431e-06,
"loss": 2.2774,
"step": 1899
},
{
"epoch": 1.9111083170931038,
"grad_norm": 16.511999530124953,
"learning_rate": 6.86563827408898e-06,
"loss": 2.2898,
"step": 1900
},
{
"epoch": 1.9121141635757843,
"grad_norm": 17.374441373909356,
"learning_rate": 6.863402638050526e-06,
"loss": 2.289,
"step": 1901
},
{
"epoch": 1.9131200100584649,
"grad_norm": 17.336050692279382,
"learning_rate": 6.861167002012074e-06,
"loss": 2.2862,
"step": 1902
},
{
"epoch": 1.9141258565411454,
"grad_norm": 17.63830681128583,
"learning_rate": 6.85893136597362e-06,
"loss": 2.3034,
"step": 1903
},
{
"epoch": 1.915131703023826,
"grad_norm": 15.818875289328755,
"learning_rate": 6.856695729935167e-06,
"loss": 2.2395,
"step": 1904
},
{
"epoch": 1.9161375495065065,
"grad_norm": 21.700760974979666,
"learning_rate": 6.854460093896714e-06,
"loss": 2.2894,
"step": 1905
},
{
"epoch": 1.917143395989187,
"grad_norm": 18.11623312546713,
"learning_rate": 6.852224457858261e-06,
"loss": 2.2581,
"step": 1906
},
{
"epoch": 1.9181492424718676,
"grad_norm": 17.35648839707561,
"learning_rate": 6.849988821819808e-06,
"loss": 2.2989,
"step": 1907
},
{
"epoch": 1.9191550889545483,
"grad_norm": 15.924160561726598,
"learning_rate": 6.847753185781355e-06,
"loss": 2.306,
"step": 1908
},
{
"epoch": 1.9201609354372289,
"grad_norm": 16.623839964461403,
"learning_rate": 6.845517549742903e-06,
"loss": 2.332,
"step": 1909
},
{
"epoch": 1.9211667819199096,
"grad_norm": 19.17988497622896,
"learning_rate": 6.8432819137044495e-06,
"loss": 2.3114,
"step": 1910
},
{
"epoch": 1.9221726284025902,
"grad_norm": 16.50083651306228,
"learning_rate": 6.841046277665997e-06,
"loss": 2.3017,
"step": 1911
},
{
"epoch": 1.9231784748852707,
"grad_norm": 17.153877632238867,
"learning_rate": 6.8388106416275434e-06,
"loss": 2.2815,
"step": 1912
},
{
"epoch": 1.9241843213679513,
"grad_norm": 17.99505447335445,
"learning_rate": 6.836575005589091e-06,
"loss": 2.256,
"step": 1913
},
{
"epoch": 1.9251901678506318,
"grad_norm": 16.618949942022258,
"learning_rate": 6.834339369550637e-06,
"loss": 2.2815,
"step": 1914
},
{
"epoch": 1.9261960143333123,
"grad_norm": 17.30100921512153,
"learning_rate": 6.832103733512184e-06,
"loss": 2.2528,
"step": 1915
},
{
"epoch": 1.9272018608159929,
"grad_norm": 16.833482665869667,
"learning_rate": 6.829868097473732e-06,
"loss": 2.2769,
"step": 1916
},
{
"epoch": 1.9282077072986734,
"grad_norm": 17.327557752569497,
"learning_rate": 6.827632461435279e-06,
"loss": 2.2319,
"step": 1917
},
{
"epoch": 1.9292135537813542,
"grad_norm": 16.753107835060206,
"learning_rate": 6.825396825396826e-06,
"loss": 2.3046,
"step": 1918
},
{
"epoch": 1.9302194002640347,
"grad_norm": 16.23041816840295,
"learning_rate": 6.823161189358373e-06,
"loss": 2.3226,
"step": 1919
},
{
"epoch": 1.9312252467467153,
"grad_norm": 15.740962564667383,
"learning_rate": 6.82092555331992e-06,
"loss": 2.3109,
"step": 1920
},
{
"epoch": 1.932231093229396,
"grad_norm": 17.19442784322785,
"learning_rate": 6.818689917281467e-06,
"loss": 2.2663,
"step": 1921
},
{
"epoch": 1.9332369397120766,
"grad_norm": 16.598926970052784,
"learning_rate": 6.816454281243015e-06,
"loss": 2.2847,
"step": 1922
},
{
"epoch": 1.934242786194757,
"grad_norm": 16.403215664959518,
"learning_rate": 6.8142186452045615e-06,
"loss": 2.2704,
"step": 1923
},
{
"epoch": 1.9352486326774376,
"grad_norm": 16.847958638967047,
"learning_rate": 6.811983009166109e-06,
"loss": 2.3015,
"step": 1924
},
{
"epoch": 1.9362544791601182,
"grad_norm": 16.187628346715247,
"learning_rate": 6.8097473731276555e-06,
"loss": 2.2388,
"step": 1925
},
{
"epoch": 1.9372603256427987,
"grad_norm": 17.867635463374597,
"learning_rate": 6.807511737089203e-06,
"loss": 2.3398,
"step": 1926
},
{
"epoch": 1.9382661721254792,
"grad_norm": 16.06214084830298,
"learning_rate": 6.8052761010507495e-06,
"loss": 2.2553,
"step": 1927
},
{
"epoch": 1.9392720186081598,
"grad_norm": 15.857951038791413,
"learning_rate": 6.803040465012296e-06,
"loss": 2.3089,
"step": 1928
},
{
"epoch": 1.9402778650908405,
"grad_norm": 16.151187192882507,
"learning_rate": 6.800804828973843e-06,
"loss": 2.2297,
"step": 1929
},
{
"epoch": 1.941283711573521,
"grad_norm": 20.836594910648667,
"learning_rate": 6.79856919293539e-06,
"loss": 2.285,
"step": 1930
},
{
"epoch": 1.9422895580562016,
"grad_norm": 14.215712608950872,
"learning_rate": 6.796333556896938e-06,
"loss": 2.3463,
"step": 1931
},
{
"epoch": 1.9432954045388824,
"grad_norm": 16.87665782003356,
"learning_rate": 6.794097920858485e-06,
"loss": 2.3153,
"step": 1932
},
{
"epoch": 1.944301251021563,
"grad_norm": 17.73419471480312,
"learning_rate": 6.791862284820032e-06,
"loss": 2.3007,
"step": 1933
},
{
"epoch": 1.9453070975042435,
"grad_norm": 18.054020098005882,
"learning_rate": 6.789626648781579e-06,
"loss": 2.2164,
"step": 1934
},
{
"epoch": 1.946312943986924,
"grad_norm": 20.446090011166692,
"learning_rate": 6.787391012743126e-06,
"loss": 2.2585,
"step": 1935
},
{
"epoch": 1.9473187904696045,
"grad_norm": 15.36859727143387,
"learning_rate": 6.785155376704673e-06,
"loss": 2.3062,
"step": 1936
},
{
"epoch": 1.948324636952285,
"grad_norm": 18.920666147270758,
"learning_rate": 6.782919740666221e-06,
"loss": 2.2437,
"step": 1937
},
{
"epoch": 1.9493304834349656,
"grad_norm": 20.03365854926881,
"learning_rate": 6.7806841046277675e-06,
"loss": 2.3588,
"step": 1938
},
{
"epoch": 1.9503363299176462,
"grad_norm": 17.9178072777866,
"learning_rate": 6.778448468589314e-06,
"loss": 2.3013,
"step": 1939
},
{
"epoch": 1.951342176400327,
"grad_norm": 15.531749211477312,
"learning_rate": 6.7762128325508615e-06,
"loss": 2.2827,
"step": 1940
},
{
"epoch": 1.9523480228830075,
"grad_norm": 17.621182445872154,
"learning_rate": 6.773977196512408e-06,
"loss": 2.3119,
"step": 1941
},
{
"epoch": 1.9533538693656882,
"grad_norm": 14.977545170236327,
"learning_rate": 6.7717415604739555e-06,
"loss": 2.2564,
"step": 1942
},
{
"epoch": 1.9543597158483688,
"grad_norm": 20.097789651883318,
"learning_rate": 6.769505924435502e-06,
"loss": 2.2334,
"step": 1943
},
{
"epoch": 1.9553655623310493,
"grad_norm": 16.81766797993441,
"learning_rate": 6.7672702883970494e-06,
"loss": 2.3052,
"step": 1944
},
{
"epoch": 1.9563714088137298,
"grad_norm": 15.694137260145133,
"learning_rate": 6.765034652358596e-06,
"loss": 2.2931,
"step": 1945
},
{
"epoch": 1.9573772552964104,
"grad_norm": 18.560231807833976,
"learning_rate": 6.762799016320144e-06,
"loss": 2.2836,
"step": 1946
},
{
"epoch": 1.958383101779091,
"grad_norm": 14.368601263956439,
"learning_rate": 6.760563380281691e-06,
"loss": 2.2922,
"step": 1947
},
{
"epoch": 1.9593889482617715,
"grad_norm": 15.779224845368637,
"learning_rate": 6.758327744243238e-06,
"loss": 2.3322,
"step": 1948
},
{
"epoch": 1.960394794744452,
"grad_norm": 17.026069508766124,
"learning_rate": 6.756092108204785e-06,
"loss": 2.3136,
"step": 1949
},
{
"epoch": 1.9614006412271328,
"grad_norm": 13.895810317269211,
"learning_rate": 6.753856472166331e-06,
"loss": 2.3046,
"step": 1950
},
{
"epoch": 1.9624064877098133,
"grad_norm": 15.082309738104678,
"learning_rate": 6.751620836127879e-06,
"loss": 2.2669,
"step": 1951
},
{
"epoch": 1.9634123341924938,
"grad_norm": 16.69092843841514,
"learning_rate": 6.749385200089425e-06,
"loss": 2.2927,
"step": 1952
},
{
"epoch": 1.9644181806751746,
"grad_norm": 17.305936411163774,
"learning_rate": 6.7471495640509736e-06,
"loss": 2.2891,
"step": 1953
},
{
"epoch": 1.9654240271578551,
"grad_norm": 17.003792051735022,
"learning_rate": 6.74491392801252e-06,
"loss": 2.2635,
"step": 1954
},
{
"epoch": 1.9664298736405357,
"grad_norm": 16.2653781486701,
"learning_rate": 6.7426782919740675e-06,
"loss": 2.2911,
"step": 1955
},
{
"epoch": 1.9674357201232162,
"grad_norm": 17.781504567455677,
"learning_rate": 6.740442655935614e-06,
"loss": 2.3029,
"step": 1956
},
{
"epoch": 1.9684415666058968,
"grad_norm": 19.581271792774587,
"learning_rate": 6.7382070198971615e-06,
"loss": 2.3463,
"step": 1957
},
{
"epoch": 1.9694474130885773,
"grad_norm": 16.588824708722353,
"learning_rate": 6.735971383858708e-06,
"loss": 2.3208,
"step": 1958
},
{
"epoch": 1.9704532595712578,
"grad_norm": 20.341400759529318,
"learning_rate": 6.7337357478202555e-06,
"loss": 2.2971,
"step": 1959
},
{
"epoch": 1.9714591060539384,
"grad_norm": 23.39973002026867,
"learning_rate": 6.731500111781802e-06,
"loss": 2.2627,
"step": 1960
},
{
"epoch": 1.9724649525366191,
"grad_norm": 17.628348506173374,
"learning_rate": 6.72926447574335e-06,
"loss": 2.2762,
"step": 1961
},
{
"epoch": 1.9734707990192997,
"grad_norm": 22.430240616517345,
"learning_rate": 6.727028839704897e-06,
"loss": 2.3183,
"step": 1962
},
{
"epoch": 1.9744766455019802,
"grad_norm": 22.262878514597467,
"learning_rate": 6.724793203666443e-06,
"loss": 2.2925,
"step": 1963
},
{
"epoch": 1.975482491984661,
"grad_norm": 16.358599953165104,
"learning_rate": 6.722557567627991e-06,
"loss": 2.3594,
"step": 1964
},
{
"epoch": 1.9764883384673415,
"grad_norm": 17.417739772341868,
"learning_rate": 6.720321931589537e-06,
"loss": 2.3343,
"step": 1965
},
{
"epoch": 1.977494184950022,
"grad_norm": 19.997471037391072,
"learning_rate": 6.718086295551085e-06,
"loss": 2.2964,
"step": 1966
},
{
"epoch": 1.9785000314327026,
"grad_norm": 18.199676913970578,
"learning_rate": 6.715850659512631e-06,
"loss": 2.3029,
"step": 1967
},
{
"epoch": 1.9795058779153831,
"grad_norm": 20.29139543009488,
"learning_rate": 6.71361502347418e-06,
"loss": 2.3143,
"step": 1968
},
{
"epoch": 1.9805117243980637,
"grad_norm": 19.02342024523025,
"learning_rate": 6.711379387435726e-06,
"loss": 2.3403,
"step": 1969
},
{
"epoch": 1.9815175708807442,
"grad_norm": 18.105251981323825,
"learning_rate": 6.7091437513972735e-06,
"loss": 2.237,
"step": 1970
},
{
"epoch": 1.9825234173634247,
"grad_norm": 18.408253755204424,
"learning_rate": 6.70690811535882e-06,
"loss": 2.307,
"step": 1971
},
{
"epoch": 1.9835292638461055,
"grad_norm": 18.72993515240626,
"learning_rate": 6.7046724793203675e-06,
"loss": 2.3008,
"step": 1972
},
{
"epoch": 1.984535110328786,
"grad_norm": 17.73016244923382,
"learning_rate": 6.702436843281914e-06,
"loss": 2.2776,
"step": 1973
},
{
"epoch": 1.9855409568114668,
"grad_norm": 17.602188647925164,
"learning_rate": 6.700201207243461e-06,
"loss": 2.2459,
"step": 1974
},
{
"epoch": 1.9865468032941473,
"grad_norm": 20.771872926086225,
"learning_rate": 6.697965571205008e-06,
"loss": 2.284,
"step": 1975
},
{
"epoch": 1.9875526497768279,
"grad_norm": 16.413973765717518,
"learning_rate": 6.695729935166555e-06,
"loss": 2.3418,
"step": 1976
},
{
"epoch": 1.9885584962595084,
"grad_norm": 21.346503688042784,
"learning_rate": 6.693494299128103e-06,
"loss": 2.3004,
"step": 1977
},
{
"epoch": 1.989564342742189,
"grad_norm": 20.864761239960973,
"learning_rate": 6.691258663089649e-06,
"loss": 2.3018,
"step": 1978
},
{
"epoch": 1.9905701892248695,
"grad_norm": 15.97287960488781,
"learning_rate": 6.689023027051197e-06,
"loss": 2.2562,
"step": 1979
},
{
"epoch": 1.99157603570755,
"grad_norm": 17.090597044595953,
"learning_rate": 6.686787391012743e-06,
"loss": 2.2424,
"step": 1980
},
{
"epoch": 1.9925818821902306,
"grad_norm": 20.64289156909885,
"learning_rate": 6.684551754974291e-06,
"loss": 2.3371,
"step": 1981
},
{
"epoch": 1.9935877286729113,
"grad_norm": 17.23349871929263,
"learning_rate": 6.682316118935837e-06,
"loss": 2.2772,
"step": 1982
},
{
"epoch": 1.9945935751555919,
"grad_norm": 18.728810327340735,
"learning_rate": 6.680080482897386e-06,
"loss": 2.3038,
"step": 1983
},
{
"epoch": 1.9955994216382724,
"grad_norm": 20.432953519649597,
"learning_rate": 6.677844846858932e-06,
"loss": 2.3487,
"step": 1984
},
{
"epoch": 1.9966052681209532,
"grad_norm": 17.389497689218956,
"learning_rate": 6.6756092108204796e-06,
"loss": 2.2715,
"step": 1985
},
{
"epoch": 1.9976111146036337,
"grad_norm": 15.683209711373696,
"learning_rate": 6.673373574782026e-06,
"loss": 2.2951,
"step": 1986
},
{
"epoch": 1.9986169610863143,
"grad_norm": 19.70295022855433,
"learning_rate": 6.671137938743573e-06,
"loss": 2.3248,
"step": 1987
},
{
"epoch": 1.9996228075689948,
"grad_norm": 17.3150299364818,
"learning_rate": 6.66890230270512e-06,
"loss": 2.2834,
"step": 1988
},
{
"epoch": 2.0006286540516753,
"grad_norm": 16.761655273211517,
"learning_rate": 6.666666666666667e-06,
"loss": 2.1124,
"step": 1989
},
{
"epoch": 2.001634500534356,
"grad_norm": 19.19275918856666,
"learning_rate": 6.664431030628215e-06,
"loss": 2.0861,
"step": 1990
},
{
"epoch": 2.0026403470170364,
"grad_norm": 18.3665234076943,
"learning_rate": 6.6621953945897615e-06,
"loss": 2.0598,
"step": 1991
},
{
"epoch": 2.003646193499717,
"grad_norm": 16.111865050848618,
"learning_rate": 6.659959758551309e-06,
"loss": 2.0566,
"step": 1992
},
{
"epoch": 2.0046520399823975,
"grad_norm": 18.032869066711967,
"learning_rate": 6.6577241225128554e-06,
"loss": 2.045,
"step": 1993
},
{
"epoch": 2.0056578864650785,
"grad_norm": 16.739105171425805,
"learning_rate": 6.655488486474403e-06,
"loss": 2.0277,
"step": 1994
},
{
"epoch": 2.006663732947759,
"grad_norm": 15.937915097580044,
"learning_rate": 6.653252850435949e-06,
"loss": 2.0072,
"step": 1995
},
{
"epoch": 2.0076695794304396,
"grad_norm": 18.04853977752509,
"learning_rate": 6.651017214397497e-06,
"loss": 1.984,
"step": 1996
},
{
"epoch": 2.00867542591312,
"grad_norm": 17.54238079716516,
"learning_rate": 6.648781578359043e-06,
"loss": 2.041,
"step": 1997
},
{
"epoch": 2.0096812723958006,
"grad_norm": 16.72820920478064,
"learning_rate": 6.64654594232059e-06,
"loss": 2.0229,
"step": 1998
},
{
"epoch": 2.010687118878481,
"grad_norm": 18.224151388488885,
"learning_rate": 6.644310306282138e-06,
"loss": 2.0096,
"step": 1999
},
{
"epoch": 2.0116929653611617,
"grad_norm": 17.796645791493617,
"learning_rate": 6.642074670243685e-06,
"loss": 2.0127,
"step": 2000
}
],
"logging_steps": 1.0,
"max_steps": 4970,
"num_input_tokens_seen": 0,
"num_train_epochs": 5,
"save_steps": 1000,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 0.0,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}