Qwen7B-Full-ADEP2 / trainer_state.json
ZMC2019's picture
Model save
f1fe5a8 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.9967396367023755,
"eval_steps": 500,
"global_step": 4292,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.004657661853749418,
"grad_norm": 6.92752168275594,
"learning_rate": 5.813953488372093e-07,
"loss": 0.8029,
"step": 5
},
{
"epoch": 0.009315323707498836,
"grad_norm": 3.1206366140040105,
"learning_rate": 1.1627906976744186e-06,
"loss": 0.7421,
"step": 10
},
{
"epoch": 0.013972985561248253,
"grad_norm": 1.9977796339339233,
"learning_rate": 1.744186046511628e-06,
"loss": 0.6853,
"step": 15
},
{
"epoch": 0.018630647414997672,
"grad_norm": 2.4072474593437447,
"learning_rate": 2.325581395348837e-06,
"loss": 0.6596,
"step": 20
},
{
"epoch": 0.02328830926874709,
"grad_norm": 1.1116914304374472,
"learning_rate": 2.9069767441860468e-06,
"loss": 0.6228,
"step": 25
},
{
"epoch": 0.027945971122496506,
"grad_norm": 0.9277881373299325,
"learning_rate": 3.488372093023256e-06,
"loss": 0.6003,
"step": 30
},
{
"epoch": 0.032603632976245925,
"grad_norm": 0.7408949966084368,
"learning_rate": 4.0697674418604655e-06,
"loss": 0.5837,
"step": 35
},
{
"epoch": 0.037261294829995344,
"grad_norm": 0.7107225566304457,
"learning_rate": 4.651162790697674e-06,
"loss": 0.5644,
"step": 40
},
{
"epoch": 0.04191895668374476,
"grad_norm": 0.5848145784293148,
"learning_rate": 5.232558139534884e-06,
"loss": 0.5269,
"step": 45
},
{
"epoch": 0.04657661853749418,
"grad_norm": 0.5140529676345665,
"learning_rate": 5.8139534883720935e-06,
"loss": 0.5285,
"step": 50
},
{
"epoch": 0.05123428039124359,
"grad_norm": 0.4961933729423096,
"learning_rate": 6.395348837209303e-06,
"loss": 0.5164,
"step": 55
},
{
"epoch": 0.05589194224499301,
"grad_norm": 0.6027754913232912,
"learning_rate": 6.976744186046512e-06,
"loss": 0.5123,
"step": 60
},
{
"epoch": 0.06054960409874243,
"grad_norm": 0.5112041376492991,
"learning_rate": 7.558139534883721e-06,
"loss": 0.5051,
"step": 65
},
{
"epoch": 0.06520726595249185,
"grad_norm": 0.5749336250262522,
"learning_rate": 8.139534883720931e-06,
"loss": 0.4849,
"step": 70
},
{
"epoch": 0.06986492780624126,
"grad_norm": 0.6201245684634927,
"learning_rate": 8.72093023255814e-06,
"loss": 0.4873,
"step": 75
},
{
"epoch": 0.07452258965999069,
"grad_norm": 0.5673174922712911,
"learning_rate": 9.302325581395349e-06,
"loss": 0.4987,
"step": 80
},
{
"epoch": 0.0791802515137401,
"grad_norm": 0.5262178696818207,
"learning_rate": 9.883720930232558e-06,
"loss": 0.4779,
"step": 85
},
{
"epoch": 0.08383791336748952,
"grad_norm": 0.5695886644689018,
"learning_rate": 1.0465116279069768e-05,
"loss": 0.4848,
"step": 90
},
{
"epoch": 0.08849557522123894,
"grad_norm": 0.565733598911647,
"learning_rate": 1.1046511627906977e-05,
"loss": 0.4899,
"step": 95
},
{
"epoch": 0.09315323707498836,
"grad_norm": 0.5859263883507883,
"learning_rate": 1.1627906976744187e-05,
"loss": 0.4808,
"step": 100
},
{
"epoch": 0.09781089892873777,
"grad_norm": 0.7262514612828005,
"learning_rate": 1.2209302325581395e-05,
"loss": 0.4838,
"step": 105
},
{
"epoch": 0.10246856078248719,
"grad_norm": 0.614430086368894,
"learning_rate": 1.2790697674418606e-05,
"loss": 0.4728,
"step": 110
},
{
"epoch": 0.10712622263623661,
"grad_norm": 0.6018869957383723,
"learning_rate": 1.3372093023255814e-05,
"loss": 0.473,
"step": 115
},
{
"epoch": 0.11178388448998602,
"grad_norm": 0.6977395666291641,
"learning_rate": 1.3953488372093024e-05,
"loss": 0.4844,
"step": 120
},
{
"epoch": 0.11644154634373545,
"grad_norm": 0.6403151057589088,
"learning_rate": 1.4534883720930233e-05,
"loss": 0.4652,
"step": 125
},
{
"epoch": 0.12109920819748486,
"grad_norm": 0.6871710421028616,
"learning_rate": 1.5116279069767441e-05,
"loss": 0.468,
"step": 130
},
{
"epoch": 0.1257568700512343,
"grad_norm": 0.6015122476741315,
"learning_rate": 1.569767441860465e-05,
"loss": 0.4752,
"step": 135
},
{
"epoch": 0.1304145319049837,
"grad_norm": 0.7033847892042502,
"learning_rate": 1.6279069767441862e-05,
"loss": 0.4692,
"step": 140
},
{
"epoch": 0.1350721937587331,
"grad_norm": 0.8170324351792959,
"learning_rate": 1.686046511627907e-05,
"loss": 0.4792,
"step": 145
},
{
"epoch": 0.13972985561248252,
"grad_norm": 0.7719634478496668,
"learning_rate": 1.744186046511628e-05,
"loss": 0.4739,
"step": 150
},
{
"epoch": 0.14438751746623196,
"grad_norm": 0.8600962139069631,
"learning_rate": 1.802325581395349e-05,
"loss": 0.4566,
"step": 155
},
{
"epoch": 0.14904517931998137,
"grad_norm": 0.8190039066561372,
"learning_rate": 1.8604651162790697e-05,
"loss": 0.4636,
"step": 160
},
{
"epoch": 0.1537028411737308,
"grad_norm": 0.7921799218082787,
"learning_rate": 1.918604651162791e-05,
"loss": 0.4648,
"step": 165
},
{
"epoch": 0.1583605030274802,
"grad_norm": 0.9330951643065268,
"learning_rate": 1.9767441860465116e-05,
"loss": 0.4631,
"step": 170
},
{
"epoch": 0.1630181648812296,
"grad_norm": 0.6469151250901515,
"learning_rate": 2.0348837209302328e-05,
"loss": 0.4742,
"step": 175
},
{
"epoch": 0.16767582673497905,
"grad_norm": 0.7444083364826932,
"learning_rate": 2.0930232558139536e-05,
"loss": 0.4694,
"step": 180
},
{
"epoch": 0.17233348858872846,
"grad_norm": 0.9053753262064802,
"learning_rate": 2.1511627906976744e-05,
"loss": 0.4679,
"step": 185
},
{
"epoch": 0.17699115044247787,
"grad_norm": 0.696915456370512,
"learning_rate": 2.2093023255813955e-05,
"loss": 0.4617,
"step": 190
},
{
"epoch": 0.18164881229622729,
"grad_norm": 0.7765718674340899,
"learning_rate": 2.2674418604651163e-05,
"loss": 0.4677,
"step": 195
},
{
"epoch": 0.18630647414997673,
"grad_norm": 0.5706209115733197,
"learning_rate": 2.3255813953488374e-05,
"loss": 0.4422,
"step": 200
},
{
"epoch": 0.19096413600372614,
"grad_norm": 0.8866665224464126,
"learning_rate": 2.3837209302325582e-05,
"loss": 0.4634,
"step": 205
},
{
"epoch": 0.19562179785747555,
"grad_norm": 0.6314129248025072,
"learning_rate": 2.441860465116279e-05,
"loss": 0.453,
"step": 210
},
{
"epoch": 0.20027945971122496,
"grad_norm": 0.7431495556273111,
"learning_rate": 2.5e-05,
"loss": 0.4641,
"step": 215
},
{
"epoch": 0.20493712156497437,
"grad_norm": 0.5892331057656331,
"learning_rate": 2.5581395348837212e-05,
"loss": 0.4604,
"step": 220
},
{
"epoch": 0.2095947834187238,
"grad_norm": 0.6573016213315429,
"learning_rate": 2.616279069767442e-05,
"loss": 0.4634,
"step": 225
},
{
"epoch": 0.21425244527247322,
"grad_norm": 0.705234238344836,
"learning_rate": 2.674418604651163e-05,
"loss": 0.4618,
"step": 230
},
{
"epoch": 0.21891010712622264,
"grad_norm": 0.7161035621507849,
"learning_rate": 2.7325581395348836e-05,
"loss": 0.4573,
"step": 235
},
{
"epoch": 0.22356776897997205,
"grad_norm": 0.7220402302111609,
"learning_rate": 2.7906976744186048e-05,
"loss": 0.4566,
"step": 240
},
{
"epoch": 0.22822543083372146,
"grad_norm": 0.8317272534751984,
"learning_rate": 2.848837209302326e-05,
"loss": 0.448,
"step": 245
},
{
"epoch": 0.2328830926874709,
"grad_norm": 1.0116718049789115,
"learning_rate": 2.9069767441860467e-05,
"loss": 0.4463,
"step": 250
},
{
"epoch": 0.2375407545412203,
"grad_norm": 1.2520264875182285,
"learning_rate": 2.9651162790697678e-05,
"loss": 0.4572,
"step": 255
},
{
"epoch": 0.24219841639496972,
"grad_norm": 0.6273556563905193,
"learning_rate": 3.0232558139534883e-05,
"loss": 0.4611,
"step": 260
},
{
"epoch": 0.24685607824871914,
"grad_norm": 0.7803028876004741,
"learning_rate": 3.081395348837209e-05,
"loss": 0.4557,
"step": 265
},
{
"epoch": 0.2515137401024686,
"grad_norm": 0.7348185111841634,
"learning_rate": 3.13953488372093e-05,
"loss": 0.4459,
"step": 270
},
{
"epoch": 0.25617140195621796,
"grad_norm": 0.5907461262762532,
"learning_rate": 3.197674418604651e-05,
"loss": 0.4619,
"step": 275
},
{
"epoch": 0.2608290638099674,
"grad_norm": 0.966120981626802,
"learning_rate": 3.2558139534883724e-05,
"loss": 0.4505,
"step": 280
},
{
"epoch": 0.26548672566371684,
"grad_norm": 1.0403496213648828,
"learning_rate": 3.313953488372093e-05,
"loss": 0.4513,
"step": 285
},
{
"epoch": 0.2701443875174662,
"grad_norm": 1.2436057586513931,
"learning_rate": 3.372093023255814e-05,
"loss": 0.452,
"step": 290
},
{
"epoch": 0.27480204937121566,
"grad_norm": 2.3331539482866757,
"learning_rate": 3.430232558139535e-05,
"loss": 0.4451,
"step": 295
},
{
"epoch": 0.27945971122496505,
"grad_norm": 1.0379481464446643,
"learning_rate": 3.488372093023256e-05,
"loss": 0.4573,
"step": 300
},
{
"epoch": 0.2841173730787145,
"grad_norm": 3.6031638667280483,
"learning_rate": 3.5465116279069774e-05,
"loss": 0.4601,
"step": 305
},
{
"epoch": 0.2887750349324639,
"grad_norm": 0.6833415227247001,
"learning_rate": 3.604651162790698e-05,
"loss": 0.4558,
"step": 310
},
{
"epoch": 0.2934326967862133,
"grad_norm": 0.8639270782587997,
"learning_rate": 3.662790697674418e-05,
"loss": 0.4546,
"step": 315
},
{
"epoch": 0.29809035863996275,
"grad_norm": 1.8215303829903469,
"learning_rate": 3.7209302325581394e-05,
"loss": 0.4634,
"step": 320
},
{
"epoch": 0.30274802049371213,
"grad_norm": 1.2592473661230983,
"learning_rate": 3.7790697674418606e-05,
"loss": 0.4672,
"step": 325
},
{
"epoch": 0.3074056823474616,
"grad_norm": 1.1786650053013983,
"learning_rate": 3.837209302325582e-05,
"loss": 0.4573,
"step": 330
},
{
"epoch": 0.312063344201211,
"grad_norm": 0.9018705956781983,
"learning_rate": 3.895348837209303e-05,
"loss": 0.4428,
"step": 335
},
{
"epoch": 0.3167210060549604,
"grad_norm": 0.8940848456165655,
"learning_rate": 3.953488372093023e-05,
"loss": 0.4548,
"step": 340
},
{
"epoch": 0.32137866790870984,
"grad_norm": 0.8345182625802969,
"learning_rate": 4.0116279069767444e-05,
"loss": 0.4453,
"step": 345
},
{
"epoch": 0.3260363297624592,
"grad_norm": 0.7788858852128394,
"learning_rate": 4.0697674418604655e-05,
"loss": 0.4497,
"step": 350
},
{
"epoch": 0.33069399161620866,
"grad_norm": 2.835448275262579,
"learning_rate": 4.127906976744187e-05,
"loss": 0.469,
"step": 355
},
{
"epoch": 0.3353516534699581,
"grad_norm": 1.921513390569175,
"learning_rate": 4.186046511627907e-05,
"loss": 0.4716,
"step": 360
},
{
"epoch": 0.3400093153237075,
"grad_norm": 1.1947987859585383,
"learning_rate": 4.2441860465116276e-05,
"loss": 0.4624,
"step": 365
},
{
"epoch": 0.3446669771774569,
"grad_norm": 0.9700341651993428,
"learning_rate": 4.302325581395349e-05,
"loss": 0.4661,
"step": 370
},
{
"epoch": 0.3493246390312063,
"grad_norm": 0.8969571742098159,
"learning_rate": 4.36046511627907e-05,
"loss": 0.4627,
"step": 375
},
{
"epoch": 0.35398230088495575,
"grad_norm": 0.6881782636935027,
"learning_rate": 4.418604651162791e-05,
"loss": 0.4468,
"step": 380
},
{
"epoch": 0.3586399627387052,
"grad_norm": 0.8164164954054413,
"learning_rate": 4.476744186046512e-05,
"loss": 0.4422,
"step": 385
},
{
"epoch": 0.36329762459245457,
"grad_norm": 0.7087648580913801,
"learning_rate": 4.5348837209302326e-05,
"loss": 0.4474,
"step": 390
},
{
"epoch": 0.367955286446204,
"grad_norm": 0.7023296376912398,
"learning_rate": 4.593023255813954e-05,
"loss": 0.4424,
"step": 395
},
{
"epoch": 0.37261294829995345,
"grad_norm": 0.8333076016425238,
"learning_rate": 4.651162790697675e-05,
"loss": 0.4414,
"step": 400
},
{
"epoch": 0.37727061015370283,
"grad_norm": 3.1226077583229617,
"learning_rate": 4.709302325581396e-05,
"loss": 0.4464,
"step": 405
},
{
"epoch": 0.3819282720074523,
"grad_norm": 3.1795716272396377,
"learning_rate": 4.7674418604651164e-05,
"loss": 0.4659,
"step": 410
},
{
"epoch": 0.38658593386120166,
"grad_norm": 1.2388415204725036,
"learning_rate": 4.8255813953488375e-05,
"loss": 0.4707,
"step": 415
},
{
"epoch": 0.3912435957149511,
"grad_norm": 11.6644368029688,
"learning_rate": 4.883720930232558e-05,
"loss": 0.4515,
"step": 420
},
{
"epoch": 0.39590125756870054,
"grad_norm": 113.48820494896151,
"learning_rate": 4.941860465116279e-05,
"loss": 0.4821,
"step": 425
},
{
"epoch": 0.4005589194224499,
"grad_norm": 1.0083170302325644,
"learning_rate": 5e-05,
"loss": 0.4793,
"step": 430
},
{
"epoch": 0.40521658127619936,
"grad_norm": 0.8343252729292476,
"learning_rate": 4.9935266701191095e-05,
"loss": 0.4626,
"step": 435
},
{
"epoch": 0.40987424312994875,
"grad_norm": 1.3705818840350565,
"learning_rate": 4.987053340238219e-05,
"loss": 0.4755,
"step": 440
},
{
"epoch": 0.4145319049836982,
"grad_norm": 0.9134655651221893,
"learning_rate": 4.980580010357328e-05,
"loss": 0.4525,
"step": 445
},
{
"epoch": 0.4191895668374476,
"grad_norm": 0.8116030121985857,
"learning_rate": 4.9741066804764374e-05,
"loss": 0.4514,
"step": 450
},
{
"epoch": 0.423847228691197,
"grad_norm": 0.9800010870367113,
"learning_rate": 4.967633350595546e-05,
"loss": 0.4684,
"step": 455
},
{
"epoch": 0.42850489054494645,
"grad_norm": 0.6033357918612593,
"learning_rate": 4.961160020714656e-05,
"loss": 0.448,
"step": 460
},
{
"epoch": 0.43316255239869583,
"grad_norm": 0.7024480175816733,
"learning_rate": 4.954686690833765e-05,
"loss": 0.4561,
"step": 465
},
{
"epoch": 0.43782021425244527,
"grad_norm": 0.9064312821375221,
"learning_rate": 4.948213360952874e-05,
"loss": 0.4492,
"step": 470
},
{
"epoch": 0.4424778761061947,
"grad_norm": 0.7172332562259943,
"learning_rate": 4.941740031071983e-05,
"loss": 0.4579,
"step": 475
},
{
"epoch": 0.4471355379599441,
"grad_norm": 0.5153145090485904,
"learning_rate": 4.935266701191093e-05,
"loss": 0.443,
"step": 480
},
{
"epoch": 0.45179319981369354,
"grad_norm": 0.649992420910321,
"learning_rate": 4.9287933713102025e-05,
"loss": 0.4449,
"step": 485
},
{
"epoch": 0.4564508616674429,
"grad_norm": 0.7128785726809049,
"learning_rate": 4.922320041429311e-05,
"loss": 0.449,
"step": 490
},
{
"epoch": 0.46110852352119236,
"grad_norm": 0.518743601600328,
"learning_rate": 4.915846711548421e-05,
"loss": 0.4413,
"step": 495
},
{
"epoch": 0.4657661853749418,
"grad_norm": 0.7520128926024633,
"learning_rate": 4.9093733816675304e-05,
"loss": 0.4483,
"step": 500
},
{
"epoch": 0.4704238472286912,
"grad_norm": 1.0686968926245148,
"learning_rate": 4.902900051786639e-05,
"loss": 0.4598,
"step": 505
},
{
"epoch": 0.4750815090824406,
"grad_norm": 1.0997940553078784,
"learning_rate": 4.8964267219057483e-05,
"loss": 0.4433,
"step": 510
},
{
"epoch": 0.47973917093619,
"grad_norm": 0.9644379037478946,
"learning_rate": 4.889953392024858e-05,
"loss": 0.4508,
"step": 515
},
{
"epoch": 0.48439683278993945,
"grad_norm": 0.7240877417772227,
"learning_rate": 4.883480062143967e-05,
"loss": 0.4465,
"step": 520
},
{
"epoch": 0.4890544946436889,
"grad_norm": 0.6057475216698976,
"learning_rate": 4.877006732263076e-05,
"loss": 0.4538,
"step": 525
},
{
"epoch": 0.49371215649743827,
"grad_norm": 1.0805852332397097,
"learning_rate": 4.8705334023821855e-05,
"loss": 0.441,
"step": 530
},
{
"epoch": 0.4983698183511877,
"grad_norm": 0.7101163390203344,
"learning_rate": 4.864060072501295e-05,
"loss": 0.4474,
"step": 535
},
{
"epoch": 0.5030274802049371,
"grad_norm": 0.5607989501155178,
"learning_rate": 4.857586742620404e-05,
"loss": 0.4498,
"step": 540
},
{
"epoch": 0.5076851420586865,
"grad_norm": 0.7141168081675577,
"learning_rate": 4.8511134127395134e-05,
"loss": 0.4398,
"step": 545
},
{
"epoch": 0.5123428039124359,
"grad_norm": 0.6293431988298551,
"learning_rate": 4.844640082858623e-05,
"loss": 0.4536,
"step": 550
},
{
"epoch": 0.5170004657661854,
"grad_norm": 0.6257261340859561,
"learning_rate": 4.838166752977732e-05,
"loss": 0.4418,
"step": 555
},
{
"epoch": 0.5216581276199348,
"grad_norm": 0.6232117352293421,
"learning_rate": 4.831693423096841e-05,
"loss": 0.4466,
"step": 560
},
{
"epoch": 0.5263157894736842,
"grad_norm": 0.5876801848079608,
"learning_rate": 4.82522009321595e-05,
"loss": 0.4449,
"step": 565
},
{
"epoch": 0.5309734513274337,
"grad_norm": 0.6227366864183466,
"learning_rate": 4.81874676333506e-05,
"loss": 0.4442,
"step": 570
},
{
"epoch": 0.5356311131811831,
"grad_norm": 0.5475728503047741,
"learning_rate": 4.812273433454169e-05,
"loss": 0.4412,
"step": 575
},
{
"epoch": 0.5402887750349324,
"grad_norm": 0.5077825170970056,
"learning_rate": 4.8058001035732785e-05,
"loss": 0.4411,
"step": 580
},
{
"epoch": 0.5449464368886818,
"grad_norm": 0.7313103597196117,
"learning_rate": 4.799326773692387e-05,
"loss": 0.438,
"step": 585
},
{
"epoch": 0.5496040987424313,
"grad_norm": 0.6058579101984163,
"learning_rate": 4.792853443811497e-05,
"loss": 0.4528,
"step": 590
},
{
"epoch": 0.5542617605961807,
"grad_norm": 0.7064967055994522,
"learning_rate": 4.7863801139306064e-05,
"loss": 0.4435,
"step": 595
},
{
"epoch": 0.5589194224499301,
"grad_norm": 0.6236154362787276,
"learning_rate": 4.779906784049715e-05,
"loss": 0.4488,
"step": 600
},
{
"epoch": 0.5635770843036796,
"grad_norm": 0.7252402093905195,
"learning_rate": 4.773433454168825e-05,
"loss": 0.4248,
"step": 605
},
{
"epoch": 0.568234746157429,
"grad_norm": 0.668295888260378,
"learning_rate": 4.766960124287934e-05,
"loss": 0.4422,
"step": 610
},
{
"epoch": 0.5728924080111784,
"grad_norm": 0.6155123020308337,
"learning_rate": 4.760486794407043e-05,
"loss": 0.4348,
"step": 615
},
{
"epoch": 0.5775500698649279,
"grad_norm": 0.594045014249402,
"learning_rate": 4.754013464526152e-05,
"loss": 0.4377,
"step": 620
},
{
"epoch": 0.5822077317186772,
"grad_norm": 0.9764084282981423,
"learning_rate": 4.747540134645262e-05,
"loss": 0.448,
"step": 625
},
{
"epoch": 0.5868653935724266,
"grad_norm": 0.7417553163037858,
"learning_rate": 4.741066804764371e-05,
"loss": 0.4353,
"step": 630
},
{
"epoch": 0.5915230554261761,
"grad_norm": 0.6844689847415286,
"learning_rate": 4.73459347488348e-05,
"loss": 0.4343,
"step": 635
},
{
"epoch": 0.5961807172799255,
"grad_norm": 0.5636372991521233,
"learning_rate": 4.7281201450025894e-05,
"loss": 0.439,
"step": 640
},
{
"epoch": 0.6008383791336749,
"grad_norm": 0.7487968219317778,
"learning_rate": 4.721646815121699e-05,
"loss": 0.4384,
"step": 645
},
{
"epoch": 0.6054960409874243,
"grad_norm": 0.6545338432685719,
"learning_rate": 4.715173485240808e-05,
"loss": 0.4365,
"step": 650
},
{
"epoch": 0.6101537028411738,
"grad_norm": 0.5457324276584636,
"learning_rate": 4.708700155359917e-05,
"loss": 0.4423,
"step": 655
},
{
"epoch": 0.6148113646949231,
"grad_norm": 0.6673827973910538,
"learning_rate": 4.7022268254790266e-05,
"loss": 0.443,
"step": 660
},
{
"epoch": 0.6194690265486725,
"grad_norm": 0.9349454934006193,
"learning_rate": 4.695753495598136e-05,
"loss": 0.4383,
"step": 665
},
{
"epoch": 0.624126688402422,
"grad_norm": 0.63544553454262,
"learning_rate": 4.689280165717245e-05,
"loss": 0.441,
"step": 670
},
{
"epoch": 0.6287843502561714,
"grad_norm": 0.5837923413074558,
"learning_rate": 4.6828068358363545e-05,
"loss": 0.4328,
"step": 675
},
{
"epoch": 0.6334420121099208,
"grad_norm": 0.5320619946457391,
"learning_rate": 4.676333505955464e-05,
"loss": 0.4366,
"step": 680
},
{
"epoch": 0.6380996739636703,
"grad_norm": 0.605917494319971,
"learning_rate": 4.669860176074573e-05,
"loss": 0.4474,
"step": 685
},
{
"epoch": 0.6427573358174197,
"grad_norm": 0.6653824724316433,
"learning_rate": 4.6633868461936824e-05,
"loss": 0.4389,
"step": 690
},
{
"epoch": 0.6474149976711691,
"grad_norm": 0.5860908364136305,
"learning_rate": 4.656913516312791e-05,
"loss": 0.4369,
"step": 695
},
{
"epoch": 0.6520726595249184,
"grad_norm": 0.5966356195762597,
"learning_rate": 4.650440186431901e-05,
"loss": 0.4341,
"step": 700
},
{
"epoch": 0.6567303213786679,
"grad_norm": 0.564451568252952,
"learning_rate": 4.64396685655101e-05,
"loss": 0.4314,
"step": 705
},
{
"epoch": 0.6613879832324173,
"grad_norm": 0.7035115996231885,
"learning_rate": 4.637493526670119e-05,
"loss": 0.4429,
"step": 710
},
{
"epoch": 0.6660456450861667,
"grad_norm": 0.5778739558938046,
"learning_rate": 4.631020196789229e-05,
"loss": 0.4369,
"step": 715
},
{
"epoch": 0.6707033069399162,
"grad_norm": 0.8934302769291517,
"learning_rate": 4.624546866908338e-05,
"loss": 0.4288,
"step": 720
},
{
"epoch": 0.6753609687936656,
"grad_norm": 0.734087332564624,
"learning_rate": 4.618073537027447e-05,
"loss": 0.4337,
"step": 725
},
{
"epoch": 0.680018630647415,
"grad_norm": 0.7048613981585907,
"learning_rate": 4.611600207146556e-05,
"loss": 0.4399,
"step": 730
},
{
"epoch": 0.6846762925011645,
"grad_norm": 0.5604335558697083,
"learning_rate": 4.605126877265666e-05,
"loss": 0.4287,
"step": 735
},
{
"epoch": 0.6893339543549138,
"grad_norm": 0.4872919097245967,
"learning_rate": 4.598653547384775e-05,
"loss": 0.4333,
"step": 740
},
{
"epoch": 0.6939916162086632,
"grad_norm": 0.6669275046240741,
"learning_rate": 4.592180217503884e-05,
"loss": 0.4304,
"step": 745
},
{
"epoch": 0.6986492780624126,
"grad_norm": 0.630044864392535,
"learning_rate": 4.585706887622993e-05,
"loss": 0.4364,
"step": 750
},
{
"epoch": 0.7033069399161621,
"grad_norm": 0.6450622968857808,
"learning_rate": 4.5792335577421026e-05,
"loss": 0.4264,
"step": 755
},
{
"epoch": 0.7079646017699115,
"grad_norm": 0.6200426709320188,
"learning_rate": 4.572760227861212e-05,
"loss": 0.4292,
"step": 760
},
{
"epoch": 0.7126222636236609,
"grad_norm": 0.5777601231289204,
"learning_rate": 4.566286897980321e-05,
"loss": 0.437,
"step": 765
},
{
"epoch": 0.7172799254774104,
"grad_norm": 0.6569186197655689,
"learning_rate": 4.5598135680994305e-05,
"loss": 0.4362,
"step": 770
},
{
"epoch": 0.7219375873311598,
"grad_norm": 0.5598715684009743,
"learning_rate": 4.55334023821854e-05,
"loss": 0.4252,
"step": 775
},
{
"epoch": 0.7265952491849091,
"grad_norm": 0.6869231997726726,
"learning_rate": 4.546866908337649e-05,
"loss": 0.4408,
"step": 780
},
{
"epoch": 0.7312529110386586,
"grad_norm": 0.5837022632387678,
"learning_rate": 4.5403935784567584e-05,
"loss": 0.4342,
"step": 785
},
{
"epoch": 0.735910572892408,
"grad_norm": 0.6124526257152769,
"learning_rate": 4.533920248575868e-05,
"loss": 0.4262,
"step": 790
},
{
"epoch": 0.7405682347461574,
"grad_norm": 0.6724737198480268,
"learning_rate": 4.527446918694977e-05,
"loss": 0.4328,
"step": 795
},
{
"epoch": 0.7452258965999069,
"grad_norm": 0.5255076953476321,
"learning_rate": 4.520973588814086e-05,
"loss": 0.4286,
"step": 800
},
{
"epoch": 0.7498835584536563,
"grad_norm": 0.5944855149791919,
"learning_rate": 4.5145002589331956e-05,
"loss": 0.4435,
"step": 805
},
{
"epoch": 0.7545412203074057,
"grad_norm": 0.8048901594295268,
"learning_rate": 4.508026929052305e-05,
"loss": 0.4407,
"step": 810
},
{
"epoch": 0.759198882161155,
"grad_norm": 0.6454599524510284,
"learning_rate": 4.501553599171414e-05,
"loss": 0.4327,
"step": 815
},
{
"epoch": 0.7638565440149045,
"grad_norm": 0.5072677303046597,
"learning_rate": 4.495080269290523e-05,
"loss": 0.4319,
"step": 820
},
{
"epoch": 0.7685142058686539,
"grad_norm": 0.5559797354764922,
"learning_rate": 4.488606939409633e-05,
"loss": 0.4206,
"step": 825
},
{
"epoch": 0.7731718677224033,
"grad_norm": 0.5345595601661192,
"learning_rate": 4.482133609528742e-05,
"loss": 0.4274,
"step": 830
},
{
"epoch": 0.7778295295761528,
"grad_norm": 0.7547024968931794,
"learning_rate": 4.475660279647851e-05,
"loss": 0.4358,
"step": 835
},
{
"epoch": 0.7824871914299022,
"grad_norm": 0.6227865373803352,
"learning_rate": 4.46918694976696e-05,
"loss": 0.4212,
"step": 840
},
{
"epoch": 0.7871448532836516,
"grad_norm": 0.7744646863428748,
"learning_rate": 4.46271361988607e-05,
"loss": 0.4266,
"step": 845
},
{
"epoch": 0.7918025151374011,
"grad_norm": 0.7178370613933127,
"learning_rate": 4.4562402900051786e-05,
"loss": 0.4286,
"step": 850
},
{
"epoch": 0.7964601769911505,
"grad_norm": 0.6045256278401768,
"learning_rate": 4.449766960124288e-05,
"loss": 0.4209,
"step": 855
},
{
"epoch": 0.8011178388448998,
"grad_norm": 0.6908919040235558,
"learning_rate": 4.443293630243397e-05,
"loss": 0.4232,
"step": 860
},
{
"epoch": 0.8057755006986492,
"grad_norm": 0.607340379361927,
"learning_rate": 4.436820300362507e-05,
"loss": 0.4258,
"step": 865
},
{
"epoch": 0.8104331625523987,
"grad_norm": 0.5338338967801528,
"learning_rate": 4.430346970481616e-05,
"loss": 0.4204,
"step": 870
},
{
"epoch": 0.8150908244061481,
"grad_norm": 0.4097730121893453,
"learning_rate": 4.423873640600725e-05,
"loss": 0.4295,
"step": 875
},
{
"epoch": 0.8197484862598975,
"grad_norm": 0.41680877490809976,
"learning_rate": 4.4174003107198344e-05,
"loss": 0.4141,
"step": 880
},
{
"epoch": 0.824406148113647,
"grad_norm": 0.45846922466853185,
"learning_rate": 4.410926980838944e-05,
"loss": 0.4293,
"step": 885
},
{
"epoch": 0.8290638099673964,
"grad_norm": 0.624287985275869,
"learning_rate": 4.404453650958053e-05,
"loss": 0.4276,
"step": 890
},
{
"epoch": 0.8337214718211458,
"grad_norm": 0.6427305945113978,
"learning_rate": 4.397980321077162e-05,
"loss": 0.4285,
"step": 895
},
{
"epoch": 0.8383791336748952,
"grad_norm": 0.656722812459809,
"learning_rate": 4.3915069911962716e-05,
"loss": 0.4235,
"step": 900
},
{
"epoch": 0.8430367955286446,
"grad_norm": 0.5762150457354466,
"learning_rate": 4.385033661315381e-05,
"loss": 0.4262,
"step": 905
},
{
"epoch": 0.847694457382394,
"grad_norm": 0.5465031708835169,
"learning_rate": 4.37856033143449e-05,
"loss": 0.4324,
"step": 910
},
{
"epoch": 0.8523521192361434,
"grad_norm": 0.5937693113190048,
"learning_rate": 4.3720870015535995e-05,
"loss": 0.424,
"step": 915
},
{
"epoch": 0.8570097810898929,
"grad_norm": 0.5935366518742683,
"learning_rate": 4.365613671672709e-05,
"loss": 0.4313,
"step": 920
},
{
"epoch": 0.8616674429436423,
"grad_norm": 0.6991248808211339,
"learning_rate": 4.359140341791818e-05,
"loss": 0.4267,
"step": 925
},
{
"epoch": 0.8663251047973917,
"grad_norm": 0.7346783269595882,
"learning_rate": 4.352667011910927e-05,
"loss": 0.4277,
"step": 930
},
{
"epoch": 0.8709827666511412,
"grad_norm": 0.5339138640718899,
"learning_rate": 4.346193682030037e-05,
"loss": 0.4156,
"step": 935
},
{
"epoch": 0.8756404285048905,
"grad_norm": 0.5509815538831199,
"learning_rate": 4.339720352149146e-05,
"loss": 0.416,
"step": 940
},
{
"epoch": 0.8802980903586399,
"grad_norm": 0.5245746762008456,
"learning_rate": 4.3332470222682546e-05,
"loss": 0.4197,
"step": 945
},
{
"epoch": 0.8849557522123894,
"grad_norm": 0.49232142300598086,
"learning_rate": 4.326773692387364e-05,
"loss": 0.4094,
"step": 950
},
{
"epoch": 0.8896134140661388,
"grad_norm": 0.5051062616308978,
"learning_rate": 4.320300362506474e-05,
"loss": 0.422,
"step": 955
},
{
"epoch": 0.8942710759198882,
"grad_norm": 0.7687269777401123,
"learning_rate": 4.313827032625583e-05,
"loss": 0.4276,
"step": 960
},
{
"epoch": 0.8989287377736377,
"grad_norm": 0.5572032044424499,
"learning_rate": 4.307353702744692e-05,
"loss": 0.4223,
"step": 965
},
{
"epoch": 0.9035863996273871,
"grad_norm": 0.5643322714638882,
"learning_rate": 4.300880372863801e-05,
"loss": 0.4259,
"step": 970
},
{
"epoch": 0.9082440614811365,
"grad_norm": 0.5250152051480017,
"learning_rate": 4.294407042982911e-05,
"loss": 0.4203,
"step": 975
},
{
"epoch": 0.9129017233348858,
"grad_norm": 1.775669542950946,
"learning_rate": 4.28793371310202e-05,
"loss": 0.4235,
"step": 980
},
{
"epoch": 0.9175593851886353,
"grad_norm": 0.6989407506091581,
"learning_rate": 4.281460383221129e-05,
"loss": 0.4292,
"step": 985
},
{
"epoch": 0.9222170470423847,
"grad_norm": 1.0108097597477825,
"learning_rate": 4.274987053340238e-05,
"loss": 0.4022,
"step": 990
},
{
"epoch": 0.9268747088961341,
"grad_norm": 0.6870512928826218,
"learning_rate": 4.2685137234593476e-05,
"loss": 0.4175,
"step": 995
},
{
"epoch": 0.9315323707498836,
"grad_norm": 0.5676290103181388,
"learning_rate": 4.262040393578457e-05,
"loss": 0.4197,
"step": 1000
},
{
"epoch": 0.936190032603633,
"grad_norm": 0.6052775213447519,
"learning_rate": 4.255567063697566e-05,
"loss": 0.4285,
"step": 1005
},
{
"epoch": 0.9408476944573824,
"grad_norm": 0.48531586665515636,
"learning_rate": 4.2490937338166755e-05,
"loss": 0.417,
"step": 1010
},
{
"epoch": 0.9455053563111319,
"grad_norm": 0.5661010529095274,
"learning_rate": 4.242620403935785e-05,
"loss": 0.4199,
"step": 1015
},
{
"epoch": 0.9501630181648812,
"grad_norm": 0.6079831482890822,
"learning_rate": 4.236147074054894e-05,
"loss": 0.4278,
"step": 1020
},
{
"epoch": 0.9548206800186306,
"grad_norm": 0.5746246985979553,
"learning_rate": 4.2296737441740034e-05,
"loss": 0.4201,
"step": 1025
},
{
"epoch": 0.95947834187238,
"grad_norm": 0.5884834814710899,
"learning_rate": 4.223200414293113e-05,
"loss": 0.412,
"step": 1030
},
{
"epoch": 0.9641360037261295,
"grad_norm": 0.5304972411175706,
"learning_rate": 4.216727084412222e-05,
"loss": 0.4157,
"step": 1035
},
{
"epoch": 0.9687936655798789,
"grad_norm": 0.7966837800227516,
"learning_rate": 4.2102537545313306e-05,
"loss": 0.4154,
"step": 1040
},
{
"epoch": 0.9734513274336283,
"grad_norm": 0.8281870140369813,
"learning_rate": 4.2037804246504406e-05,
"loss": 0.4179,
"step": 1045
},
{
"epoch": 0.9781089892873778,
"grad_norm": 0.5860620536296465,
"learning_rate": 4.19730709476955e-05,
"loss": 0.4177,
"step": 1050
},
{
"epoch": 0.9827666511411272,
"grad_norm": 0.5890427048399142,
"learning_rate": 4.190833764888659e-05,
"loss": 0.4192,
"step": 1055
},
{
"epoch": 0.9874243129948765,
"grad_norm": 0.5213412179370963,
"learning_rate": 4.184360435007768e-05,
"loss": 0.4222,
"step": 1060
},
{
"epoch": 0.992081974848626,
"grad_norm": 0.7453282325289396,
"learning_rate": 4.177887105126878e-05,
"loss": 0.4189,
"step": 1065
},
{
"epoch": 0.9967396367023754,
"grad_norm": 0.6133847478397344,
"learning_rate": 4.171413775245987e-05,
"loss": 0.4104,
"step": 1070
},
{
"epoch": 1.00093153237075,
"grad_norm": 1.090678146779406,
"learning_rate": 4.164940445365096e-05,
"loss": 0.4209,
"step": 1075
},
{
"epoch": 1.0055891942244992,
"grad_norm": 0.6095624445469302,
"learning_rate": 4.158467115484205e-05,
"loss": 0.3619,
"step": 1080
},
{
"epoch": 1.0102468560782487,
"grad_norm": 0.5834527079795324,
"learning_rate": 4.151993785603315e-05,
"loss": 0.3498,
"step": 1085
},
{
"epoch": 1.0149045179319982,
"grad_norm": 0.7818406149318912,
"learning_rate": 4.1455204557224236e-05,
"loss": 0.3651,
"step": 1090
},
{
"epoch": 1.0195621797857475,
"grad_norm": 0.5600327410720667,
"learning_rate": 4.139047125841533e-05,
"loss": 0.3608,
"step": 1095
},
{
"epoch": 1.024219841639497,
"grad_norm": 0.5373438690472865,
"learning_rate": 4.132573795960642e-05,
"loss": 0.3657,
"step": 1100
},
{
"epoch": 1.0288775034932465,
"grad_norm": 0.4862220275338719,
"learning_rate": 4.1261004660797515e-05,
"loss": 0.3661,
"step": 1105
},
{
"epoch": 1.0335351653469957,
"grad_norm": 0.5361122770568882,
"learning_rate": 4.119627136198861e-05,
"loss": 0.3635,
"step": 1110
},
{
"epoch": 1.0381928272007452,
"grad_norm": 0.4625151982803019,
"learning_rate": 4.11315380631797e-05,
"loss": 0.3499,
"step": 1115
},
{
"epoch": 1.0428504890544947,
"grad_norm": 0.6531186567479866,
"learning_rate": 4.1066804764370794e-05,
"loss": 0.3618,
"step": 1120
},
{
"epoch": 1.047508150908244,
"grad_norm": 0.6272666912027586,
"learning_rate": 4.100207146556189e-05,
"loss": 0.3652,
"step": 1125
},
{
"epoch": 1.0521658127619935,
"grad_norm": 0.5373624121001779,
"learning_rate": 4.093733816675298e-05,
"loss": 0.3494,
"step": 1130
},
{
"epoch": 1.056823474615743,
"grad_norm": 0.4965004694353307,
"learning_rate": 4.087260486794407e-05,
"loss": 0.3596,
"step": 1135
},
{
"epoch": 1.0614811364694923,
"grad_norm": 0.5285596102050132,
"learning_rate": 4.0807871569135166e-05,
"loss": 0.3606,
"step": 1140
},
{
"epoch": 1.0661387983232418,
"grad_norm": 0.45339136110843775,
"learning_rate": 4.074313827032626e-05,
"loss": 0.3595,
"step": 1145
},
{
"epoch": 1.0707964601769913,
"grad_norm": 0.47470548262310025,
"learning_rate": 4.067840497151735e-05,
"loss": 0.3552,
"step": 1150
},
{
"epoch": 1.0754541220307405,
"grad_norm": 0.5222608731509295,
"learning_rate": 4.0613671672708445e-05,
"loss": 0.3581,
"step": 1155
},
{
"epoch": 1.08011178388449,
"grad_norm": 0.5951129546363467,
"learning_rate": 4.054893837389954e-05,
"loss": 0.3653,
"step": 1160
},
{
"epoch": 1.0847694457382393,
"grad_norm": 0.47330368757580166,
"learning_rate": 4.048420507509063e-05,
"loss": 0.3682,
"step": 1165
},
{
"epoch": 1.0894271075919888,
"grad_norm": 0.6630467479912063,
"learning_rate": 4.041947177628172e-05,
"loss": 0.3641,
"step": 1170
},
{
"epoch": 1.0940847694457383,
"grad_norm": 0.479092390372685,
"learning_rate": 4.035473847747282e-05,
"loss": 0.3636,
"step": 1175
},
{
"epoch": 1.0987424312994876,
"grad_norm": 0.5347502580113855,
"learning_rate": 4.029000517866391e-05,
"loss": 0.3559,
"step": 1180
},
{
"epoch": 1.103400093153237,
"grad_norm": 0.5580827812560017,
"learning_rate": 4.0225271879854996e-05,
"loss": 0.3572,
"step": 1185
},
{
"epoch": 1.1080577550069866,
"grad_norm": 0.40826648699641077,
"learning_rate": 4.016053858104609e-05,
"loss": 0.3396,
"step": 1190
},
{
"epoch": 1.1127154168607358,
"grad_norm": 0.4868993696437974,
"learning_rate": 4.009580528223719e-05,
"loss": 0.3481,
"step": 1195
},
{
"epoch": 1.1173730787144853,
"grad_norm": 0.44958393221221327,
"learning_rate": 4.0031071983428275e-05,
"loss": 0.3574,
"step": 1200
},
{
"epoch": 1.1220307405682348,
"grad_norm": 0.4529924774506899,
"learning_rate": 3.996633868461937e-05,
"loss": 0.366,
"step": 1205
},
{
"epoch": 1.126688402421984,
"grad_norm": 0.45382106058655186,
"learning_rate": 3.990160538581046e-05,
"loss": 0.3611,
"step": 1210
},
{
"epoch": 1.1313460642757336,
"grad_norm": 0.774011131965164,
"learning_rate": 3.9836872087001554e-05,
"loss": 0.3574,
"step": 1215
},
{
"epoch": 1.136003726129483,
"grad_norm": 0.5708209085490166,
"learning_rate": 3.977213878819265e-05,
"loss": 0.3679,
"step": 1220
},
{
"epoch": 1.1406613879832324,
"grad_norm": 0.5407084277053351,
"learning_rate": 3.970740548938374e-05,
"loss": 0.3602,
"step": 1225
},
{
"epoch": 1.1453190498369819,
"grad_norm": 0.5607895201020598,
"learning_rate": 3.964267219057483e-05,
"loss": 0.3569,
"step": 1230
},
{
"epoch": 1.1499767116907313,
"grad_norm": 0.5762919296213727,
"learning_rate": 3.9577938891765926e-05,
"loss": 0.3579,
"step": 1235
},
{
"epoch": 1.1546343735444806,
"grad_norm": 0.4915060554356313,
"learning_rate": 3.951320559295702e-05,
"loss": 0.3631,
"step": 1240
},
{
"epoch": 1.1592920353982301,
"grad_norm": 0.47937400758343035,
"learning_rate": 3.944847229414811e-05,
"loss": 0.3661,
"step": 1245
},
{
"epoch": 1.1639496972519794,
"grad_norm": 0.5488280878505598,
"learning_rate": 3.9383738995339205e-05,
"loss": 0.369,
"step": 1250
},
{
"epoch": 1.1686073591057289,
"grad_norm": 0.6567448003318919,
"learning_rate": 3.93190056965303e-05,
"loss": 0.3574,
"step": 1255
},
{
"epoch": 1.1732650209594784,
"grad_norm": 0.5486815256330948,
"learning_rate": 3.925427239772139e-05,
"loss": 0.3679,
"step": 1260
},
{
"epoch": 1.1779226828132279,
"grad_norm": 0.4616890041926306,
"learning_rate": 3.9189539098912484e-05,
"loss": 0.3579,
"step": 1265
},
{
"epoch": 1.1825803446669771,
"grad_norm": 0.5392889454010211,
"learning_rate": 3.912480580010358e-05,
"loss": 0.3483,
"step": 1270
},
{
"epoch": 1.1872380065207266,
"grad_norm": 0.4479725279421751,
"learning_rate": 3.906007250129467e-05,
"loss": 0.3587,
"step": 1275
},
{
"epoch": 1.191895668374476,
"grad_norm": 0.7590029961366441,
"learning_rate": 3.8995339202485756e-05,
"loss": 0.3637,
"step": 1280
},
{
"epoch": 1.1965533302282254,
"grad_norm": 0.6157692881026866,
"learning_rate": 3.8930605903676856e-05,
"loss": 0.3668,
"step": 1285
},
{
"epoch": 1.201210992081975,
"grad_norm": 0.5380563110707145,
"learning_rate": 3.886587260486795e-05,
"loss": 0.358,
"step": 1290
},
{
"epoch": 1.2058686539357242,
"grad_norm": 0.9956632223371663,
"learning_rate": 3.8801139306059035e-05,
"loss": 0.3593,
"step": 1295
},
{
"epoch": 1.2105263157894737,
"grad_norm": 0.6675088366219036,
"learning_rate": 3.873640600725013e-05,
"loss": 0.358,
"step": 1300
},
{
"epoch": 1.2151839776432232,
"grad_norm": 0.6151654966839248,
"learning_rate": 3.867167270844123e-05,
"loss": 0.3669,
"step": 1305
},
{
"epoch": 1.2198416394969724,
"grad_norm": 0.46562609640922176,
"learning_rate": 3.8606939409632314e-05,
"loss": 0.3684,
"step": 1310
},
{
"epoch": 1.224499301350722,
"grad_norm": 0.45769392273694903,
"learning_rate": 3.854220611082341e-05,
"loss": 0.3583,
"step": 1315
},
{
"epoch": 1.2291569632044714,
"grad_norm": 0.4734611105934788,
"learning_rate": 3.84774728120145e-05,
"loss": 0.3673,
"step": 1320
},
{
"epoch": 1.2338146250582207,
"grad_norm": 0.48854960034703904,
"learning_rate": 3.841273951320559e-05,
"loss": 0.3602,
"step": 1325
},
{
"epoch": 1.2384722869119702,
"grad_norm": 0.491431877053754,
"learning_rate": 3.8348006214396686e-05,
"loss": 0.3659,
"step": 1330
},
{
"epoch": 1.2431299487657197,
"grad_norm": 0.5080088344027641,
"learning_rate": 3.828327291558778e-05,
"loss": 0.3569,
"step": 1335
},
{
"epoch": 1.247787610619469,
"grad_norm": 0.8013937265298156,
"learning_rate": 3.821853961677888e-05,
"loss": 0.3607,
"step": 1340
},
{
"epoch": 1.2524452724732185,
"grad_norm": 0.4152250898013425,
"learning_rate": 3.8153806317969965e-05,
"loss": 0.353,
"step": 1345
},
{
"epoch": 1.257102934326968,
"grad_norm": 0.5085970118735269,
"learning_rate": 3.808907301916106e-05,
"loss": 0.3589,
"step": 1350
},
{
"epoch": 1.2617605961807172,
"grad_norm": 0.5551559795100467,
"learning_rate": 3.802433972035215e-05,
"loss": 0.356,
"step": 1355
},
{
"epoch": 1.2664182580344667,
"grad_norm": 0.6050619124995583,
"learning_rate": 3.7959606421543244e-05,
"loss": 0.3495,
"step": 1360
},
{
"epoch": 1.271075919888216,
"grad_norm": 0.3819757584386992,
"learning_rate": 3.789487312273434e-05,
"loss": 0.3612,
"step": 1365
},
{
"epoch": 1.2757335817419655,
"grad_norm": 0.518505991982242,
"learning_rate": 3.783013982392543e-05,
"loss": 0.3606,
"step": 1370
},
{
"epoch": 1.280391243595715,
"grad_norm": 0.489703544385954,
"learning_rate": 3.776540652511652e-05,
"loss": 0.3603,
"step": 1375
},
{
"epoch": 1.2850489054494645,
"grad_norm": 0.4189177736950793,
"learning_rate": 3.7700673226307616e-05,
"loss": 0.3586,
"step": 1380
},
{
"epoch": 1.2897065673032138,
"grad_norm": 0.49544734897634235,
"learning_rate": 3.763593992749871e-05,
"loss": 0.3584,
"step": 1385
},
{
"epoch": 1.2943642291569633,
"grad_norm": 0.7339020852038594,
"learning_rate": 3.7571206628689795e-05,
"loss": 0.3671,
"step": 1390
},
{
"epoch": 1.2990218910107125,
"grad_norm": 0.47812102647655946,
"learning_rate": 3.7506473329880895e-05,
"loss": 0.3709,
"step": 1395
},
{
"epoch": 1.303679552864462,
"grad_norm": 0.4842166642543789,
"learning_rate": 3.744174003107199e-05,
"loss": 0.3645,
"step": 1400
},
{
"epoch": 1.3083372147182115,
"grad_norm": 0.5461761326327322,
"learning_rate": 3.7377006732263074e-05,
"loss": 0.3582,
"step": 1405
},
{
"epoch": 1.312994876571961,
"grad_norm": 0.3782413550156958,
"learning_rate": 3.731227343345417e-05,
"loss": 0.357,
"step": 1410
},
{
"epoch": 1.3176525384257103,
"grad_norm": 0.5206824458146679,
"learning_rate": 3.724754013464527e-05,
"loss": 0.376,
"step": 1415
},
{
"epoch": 1.3223102002794598,
"grad_norm": 0.46613034806590914,
"learning_rate": 3.718280683583635e-05,
"loss": 0.3621,
"step": 1420
},
{
"epoch": 1.326967862133209,
"grad_norm": 0.4076821120515726,
"learning_rate": 3.7118073537027446e-05,
"loss": 0.3603,
"step": 1425
},
{
"epoch": 1.3316255239869585,
"grad_norm": 0.574827732477907,
"learning_rate": 3.705334023821854e-05,
"loss": 0.3577,
"step": 1430
},
{
"epoch": 1.336283185840708,
"grad_norm": 0.4990608353362272,
"learning_rate": 3.698860693940964e-05,
"loss": 0.3621,
"step": 1435
},
{
"epoch": 1.3409408476944573,
"grad_norm": 0.4258017041885792,
"learning_rate": 3.6923873640600725e-05,
"loss": 0.3612,
"step": 1440
},
{
"epoch": 1.3455985095482068,
"grad_norm": 0.5518768048000128,
"learning_rate": 3.685914034179182e-05,
"loss": 0.3679,
"step": 1445
},
{
"epoch": 1.350256171401956,
"grad_norm": 0.4451614301869116,
"learning_rate": 3.679440704298292e-05,
"loss": 0.358,
"step": 1450
},
{
"epoch": 1.3549138332557056,
"grad_norm": 0.512257774552283,
"learning_rate": 3.6729673744174004e-05,
"loss": 0.3676,
"step": 1455
},
{
"epoch": 1.359571495109455,
"grad_norm": 0.5162438863691285,
"learning_rate": 3.66649404453651e-05,
"loss": 0.3519,
"step": 1460
},
{
"epoch": 1.3642291569632046,
"grad_norm": 0.471777012702105,
"learning_rate": 3.660020714655619e-05,
"loss": 0.356,
"step": 1465
},
{
"epoch": 1.3688868188169538,
"grad_norm": 0.48449733457712785,
"learning_rate": 3.653547384774728e-05,
"loss": 0.3623,
"step": 1470
},
{
"epoch": 1.3735444806707033,
"grad_norm": 0.4576323028052415,
"learning_rate": 3.6470740548938376e-05,
"loss": 0.3535,
"step": 1475
},
{
"epoch": 1.3782021425244526,
"grad_norm": 0.5006209765273532,
"learning_rate": 3.640600725012947e-05,
"loss": 0.3558,
"step": 1480
},
{
"epoch": 1.382859804378202,
"grad_norm": 0.4418094810372605,
"learning_rate": 3.634127395132056e-05,
"loss": 0.3647,
"step": 1485
},
{
"epoch": 1.3875174662319516,
"grad_norm": 0.5278545308450533,
"learning_rate": 3.6276540652511655e-05,
"loss": 0.3663,
"step": 1490
},
{
"epoch": 1.392175128085701,
"grad_norm": 0.8122089772240753,
"learning_rate": 3.621180735370275e-05,
"loss": 0.365,
"step": 1495
},
{
"epoch": 1.3968327899394504,
"grad_norm": 0.4991551907714853,
"learning_rate": 3.6147074054893834e-05,
"loss": 0.3558,
"step": 1500
},
{
"epoch": 1.4014904517931999,
"grad_norm": 0.45673339355259496,
"learning_rate": 3.6082340756084934e-05,
"loss": 0.3601,
"step": 1505
},
{
"epoch": 1.4061481136469491,
"grad_norm": 0.45172383447115166,
"learning_rate": 3.601760745727603e-05,
"loss": 0.3577,
"step": 1510
},
{
"epoch": 1.4108057755006986,
"grad_norm": 0.5433134841533829,
"learning_rate": 3.595287415846711e-05,
"loss": 0.3676,
"step": 1515
},
{
"epoch": 1.4154634373544481,
"grad_norm": 0.48998833758767707,
"learning_rate": 3.5888140859658206e-05,
"loss": 0.3606,
"step": 1520
},
{
"epoch": 1.4201210992081974,
"grad_norm": 0.6104280836329665,
"learning_rate": 3.5823407560849306e-05,
"loss": 0.3617,
"step": 1525
},
{
"epoch": 1.424778761061947,
"grad_norm": 1.3807202434190187,
"learning_rate": 3.57586742620404e-05,
"loss": 0.3511,
"step": 1530
},
{
"epoch": 1.4294364229156964,
"grad_norm": 0.5813011933063881,
"learning_rate": 3.5693940963231485e-05,
"loss": 0.3668,
"step": 1535
},
{
"epoch": 1.4340940847694457,
"grad_norm": 0.4086292338502313,
"learning_rate": 3.5629207664422585e-05,
"loss": 0.3574,
"step": 1540
},
{
"epoch": 1.4387517466231952,
"grad_norm": 0.45242586909239896,
"learning_rate": 3.556447436561368e-05,
"loss": 0.3579,
"step": 1545
},
{
"epoch": 1.4434094084769447,
"grad_norm": 0.42955196340267765,
"learning_rate": 3.5499741066804764e-05,
"loss": 0.3551,
"step": 1550
},
{
"epoch": 1.448067070330694,
"grad_norm": 0.4355035350563131,
"learning_rate": 3.543500776799586e-05,
"loss": 0.3501,
"step": 1555
},
{
"epoch": 1.4527247321844434,
"grad_norm": 0.42304027877574124,
"learning_rate": 3.537027446918696e-05,
"loss": 0.3561,
"step": 1560
},
{
"epoch": 1.4573823940381927,
"grad_norm": 0.3943828294567786,
"learning_rate": 3.530554117037804e-05,
"loss": 0.3631,
"step": 1565
},
{
"epoch": 1.4620400558919422,
"grad_norm": 0.4544956891841179,
"learning_rate": 3.5240807871569136e-05,
"loss": 0.376,
"step": 1570
},
{
"epoch": 1.4666977177456917,
"grad_norm": 0.555764704348166,
"learning_rate": 3.517607457276023e-05,
"loss": 0.3584,
"step": 1575
},
{
"epoch": 1.4713553795994412,
"grad_norm": 0.6196585992504084,
"learning_rate": 3.511134127395132e-05,
"loss": 0.3514,
"step": 1580
},
{
"epoch": 1.4760130414531905,
"grad_norm": 0.5330586804823,
"learning_rate": 3.5046607975142415e-05,
"loss": 0.3576,
"step": 1585
},
{
"epoch": 1.48067070330694,
"grad_norm": 0.5553874975958871,
"learning_rate": 3.498187467633351e-05,
"loss": 0.3596,
"step": 1590
},
{
"epoch": 1.4853283651606892,
"grad_norm": 0.45769356056194815,
"learning_rate": 3.49171413775246e-05,
"loss": 0.3557,
"step": 1595
},
{
"epoch": 1.4899860270144387,
"grad_norm": 0.5584387615516803,
"learning_rate": 3.4852408078715694e-05,
"loss": 0.3613,
"step": 1600
},
{
"epoch": 1.4946436888681882,
"grad_norm": 0.43343277775624545,
"learning_rate": 3.478767477990679e-05,
"loss": 0.3524,
"step": 1605
},
{
"epoch": 1.4993013507219377,
"grad_norm": 0.4377140537194227,
"learning_rate": 3.472294148109787e-05,
"loss": 0.3607,
"step": 1610
},
{
"epoch": 1.503959012575687,
"grad_norm": 0.5053071933612086,
"learning_rate": 3.465820818228897e-05,
"loss": 0.3582,
"step": 1615
},
{
"epoch": 1.5086166744294365,
"grad_norm": 0.47265497816294294,
"learning_rate": 3.4593474883480066e-05,
"loss": 0.359,
"step": 1620
},
{
"epoch": 1.5132743362831858,
"grad_norm": 0.4617635092436261,
"learning_rate": 3.452874158467116e-05,
"loss": 0.3626,
"step": 1625
},
{
"epoch": 1.5179319981369352,
"grad_norm": 0.5305726325688056,
"learning_rate": 3.4464008285862245e-05,
"loss": 0.3596,
"step": 1630
},
{
"epoch": 1.5225896599906847,
"grad_norm": 0.4457785432446353,
"learning_rate": 3.4399274987053345e-05,
"loss": 0.3722,
"step": 1635
},
{
"epoch": 1.5272473218444342,
"grad_norm": 0.4196162887632958,
"learning_rate": 3.433454168824444e-05,
"loss": 0.365,
"step": 1640
},
{
"epoch": 1.5319049836981835,
"grad_norm": 0.4499181015182438,
"learning_rate": 3.4269808389435524e-05,
"loss": 0.3665,
"step": 1645
},
{
"epoch": 1.5365626455519328,
"grad_norm": 0.41777827066281287,
"learning_rate": 3.4205075090626624e-05,
"loss": 0.3511,
"step": 1650
},
{
"epoch": 1.5412203074056823,
"grad_norm": 0.483686609320562,
"learning_rate": 3.414034179181772e-05,
"loss": 0.3513,
"step": 1655
},
{
"epoch": 1.5458779692594318,
"grad_norm": 0.42024302687692083,
"learning_rate": 3.40756084930088e-05,
"loss": 0.3502,
"step": 1660
},
{
"epoch": 1.5505356311131813,
"grad_norm": 0.6580744072852233,
"learning_rate": 3.4010875194199896e-05,
"loss": 0.3629,
"step": 1665
},
{
"epoch": 1.5551932929669308,
"grad_norm": 0.4143100503219979,
"learning_rate": 3.3946141895390996e-05,
"loss": 0.3583,
"step": 1670
},
{
"epoch": 1.55985095482068,
"grad_norm": 0.45443895548063523,
"learning_rate": 3.388140859658208e-05,
"loss": 0.3569,
"step": 1675
},
{
"epoch": 1.5645086166744293,
"grad_norm": 0.42295760226587914,
"learning_rate": 3.3816675297773175e-05,
"loss": 0.3635,
"step": 1680
},
{
"epoch": 1.5691662785281788,
"grad_norm": 0.5110004227388002,
"learning_rate": 3.375194199896427e-05,
"loss": 0.3533,
"step": 1685
},
{
"epoch": 1.5738239403819283,
"grad_norm": 0.42320283733825365,
"learning_rate": 3.368720870015536e-05,
"loss": 0.3597,
"step": 1690
},
{
"epoch": 1.5784816022356778,
"grad_norm": 0.41482962057863415,
"learning_rate": 3.3622475401346454e-05,
"loss": 0.358,
"step": 1695
},
{
"epoch": 1.583139264089427,
"grad_norm": 0.4150667378311711,
"learning_rate": 3.355774210253755e-05,
"loss": 0.3632,
"step": 1700
},
{
"epoch": 1.5877969259431766,
"grad_norm": 0.3666610812445483,
"learning_rate": 3.349300880372864e-05,
"loss": 0.3552,
"step": 1705
},
{
"epoch": 1.5924545877969258,
"grad_norm": 0.4530536221435045,
"learning_rate": 3.342827550491973e-05,
"loss": 0.363,
"step": 1710
},
{
"epoch": 1.5971122496506753,
"grad_norm": 0.4941170831154173,
"learning_rate": 3.3363542206110826e-05,
"loss": 0.3631,
"step": 1715
},
{
"epoch": 1.6017699115044248,
"grad_norm": 0.4707664848927128,
"learning_rate": 3.329880890730191e-05,
"loss": 0.35,
"step": 1720
},
{
"epoch": 1.6064275733581743,
"grad_norm": 0.46158780862776105,
"learning_rate": 3.323407560849301e-05,
"loss": 0.3562,
"step": 1725
},
{
"epoch": 1.6110852352119236,
"grad_norm": 0.41962064552472267,
"learning_rate": 3.3169342309684105e-05,
"loss": 0.3674,
"step": 1730
},
{
"epoch": 1.6157428970656729,
"grad_norm": 0.45419850851419985,
"learning_rate": 3.31046090108752e-05,
"loss": 0.3683,
"step": 1735
},
{
"epoch": 1.6204005589194224,
"grad_norm": 0.46971165862694936,
"learning_rate": 3.3039875712066284e-05,
"loss": 0.3478,
"step": 1740
},
{
"epoch": 1.6250582207731719,
"grad_norm": 0.47443096230607346,
"learning_rate": 3.2975142413257384e-05,
"loss": 0.3772,
"step": 1745
},
{
"epoch": 1.6297158826269214,
"grad_norm": 0.37835298699147984,
"learning_rate": 3.291040911444848e-05,
"loss": 0.3578,
"step": 1750
},
{
"epoch": 1.6343735444806708,
"grad_norm": 0.4582780296219518,
"learning_rate": 3.284567581563956e-05,
"loss": 0.3664,
"step": 1755
},
{
"epoch": 1.6390312063344201,
"grad_norm": 0.4354429562476338,
"learning_rate": 3.278094251683066e-05,
"loss": 0.3534,
"step": 1760
},
{
"epoch": 1.6436888681881694,
"grad_norm": 0.5789259252606435,
"learning_rate": 3.2716209218021756e-05,
"loss": 0.356,
"step": 1765
},
{
"epoch": 1.648346530041919,
"grad_norm": 0.464453568996508,
"learning_rate": 3.265147591921284e-05,
"loss": 0.3503,
"step": 1770
},
{
"epoch": 1.6530041918956684,
"grad_norm": 0.4568793337848766,
"learning_rate": 3.2586742620403935e-05,
"loss": 0.3654,
"step": 1775
},
{
"epoch": 1.6576618537494179,
"grad_norm": 0.4717572823402282,
"learning_rate": 3.2522009321595035e-05,
"loss": 0.3542,
"step": 1780
},
{
"epoch": 1.6623195156031674,
"grad_norm": 0.4567156289592614,
"learning_rate": 3.245727602278612e-05,
"loss": 0.3629,
"step": 1785
},
{
"epoch": 1.6669771774569166,
"grad_norm": 0.4208675681151171,
"learning_rate": 3.2392542723977214e-05,
"loss": 0.3527,
"step": 1790
},
{
"epoch": 1.671634839310666,
"grad_norm": 0.5044454939254893,
"learning_rate": 3.232780942516831e-05,
"loss": 0.3525,
"step": 1795
},
{
"epoch": 1.6762925011644154,
"grad_norm": 0.42387939427190896,
"learning_rate": 3.22630761263594e-05,
"loss": 0.3506,
"step": 1800
},
{
"epoch": 1.680950163018165,
"grad_norm": 0.5023473389380669,
"learning_rate": 3.219834282755049e-05,
"loss": 0.3579,
"step": 1805
},
{
"epoch": 1.6856078248719144,
"grad_norm": 0.47206054679476617,
"learning_rate": 3.2133609528741586e-05,
"loss": 0.3564,
"step": 1810
},
{
"epoch": 1.6902654867256637,
"grad_norm": 0.4607885248484138,
"learning_rate": 3.206887622993268e-05,
"loss": 0.3589,
"step": 1815
},
{
"epoch": 1.6949231485794132,
"grad_norm": 0.35782128979620814,
"learning_rate": 3.200414293112377e-05,
"loss": 0.3454,
"step": 1820
},
{
"epoch": 1.6995808104331624,
"grad_norm": 0.44305913674826386,
"learning_rate": 3.1939409632314865e-05,
"loss": 0.3488,
"step": 1825
},
{
"epoch": 1.704238472286912,
"grad_norm": 0.38935978775442764,
"learning_rate": 3.187467633350596e-05,
"loss": 0.3532,
"step": 1830
},
{
"epoch": 1.7088961341406614,
"grad_norm": 0.40971733347800393,
"learning_rate": 3.180994303469705e-05,
"loss": 0.3606,
"step": 1835
},
{
"epoch": 1.713553795994411,
"grad_norm": 0.44109214451792117,
"learning_rate": 3.1745209735888144e-05,
"loss": 0.3547,
"step": 1840
},
{
"epoch": 1.7182114578481602,
"grad_norm": 0.40169684579135284,
"learning_rate": 3.168047643707924e-05,
"loss": 0.3596,
"step": 1845
},
{
"epoch": 1.7228691197019095,
"grad_norm": 0.3962129300962754,
"learning_rate": 3.161574313827032e-05,
"loss": 0.3668,
"step": 1850
},
{
"epoch": 1.727526781555659,
"grad_norm": 0.42734260387858874,
"learning_rate": 3.155100983946142e-05,
"loss": 0.348,
"step": 1855
},
{
"epoch": 1.7321844434094085,
"grad_norm": 0.6659934996226169,
"learning_rate": 3.1486276540652516e-05,
"loss": 0.3716,
"step": 1860
},
{
"epoch": 1.736842105263158,
"grad_norm": 0.5555095008887161,
"learning_rate": 3.14215432418436e-05,
"loss": 0.3524,
"step": 1865
},
{
"epoch": 1.7414997671169075,
"grad_norm": 0.3709679427144339,
"learning_rate": 3.13568099430347e-05,
"loss": 0.3547,
"step": 1870
},
{
"epoch": 1.7461574289706567,
"grad_norm": 0.4145921631194794,
"learning_rate": 3.1292076644225795e-05,
"loss": 0.3712,
"step": 1875
},
{
"epoch": 1.750815090824406,
"grad_norm": 0.4814783579328554,
"learning_rate": 3.122734334541688e-05,
"loss": 0.3613,
"step": 1880
},
{
"epoch": 1.7554727526781555,
"grad_norm": 0.49050194254390594,
"learning_rate": 3.1162610046607974e-05,
"loss": 0.3635,
"step": 1885
},
{
"epoch": 1.760130414531905,
"grad_norm": 0.4333631796725332,
"learning_rate": 3.1097876747799074e-05,
"loss": 0.3605,
"step": 1890
},
{
"epoch": 1.7647880763856545,
"grad_norm": 0.4868988400346533,
"learning_rate": 3.103314344899016e-05,
"loss": 0.3689,
"step": 1895
},
{
"epoch": 1.7694457382394038,
"grad_norm": 0.4001143060111157,
"learning_rate": 3.096841015018125e-05,
"loss": 0.3592,
"step": 1900
},
{
"epoch": 1.7741034000931533,
"grad_norm": 0.45824721223774795,
"learning_rate": 3.0903676851372346e-05,
"loss": 0.3582,
"step": 1905
},
{
"epoch": 1.7787610619469025,
"grad_norm": 0.5424585883204478,
"learning_rate": 3.083894355256344e-05,
"loss": 0.3534,
"step": 1910
},
{
"epoch": 1.783418723800652,
"grad_norm": 0.5014177386049438,
"learning_rate": 3.077421025375453e-05,
"loss": 0.3482,
"step": 1915
},
{
"epoch": 1.7880763856544015,
"grad_norm": 0.4851314989650092,
"learning_rate": 3.0709476954945625e-05,
"loss": 0.3514,
"step": 1920
},
{
"epoch": 1.792734047508151,
"grad_norm": 0.42176578290218586,
"learning_rate": 3.064474365613672e-05,
"loss": 0.3623,
"step": 1925
},
{
"epoch": 1.7973917093619003,
"grad_norm": 0.49197555336823995,
"learning_rate": 3.058001035732781e-05,
"loss": 0.3517,
"step": 1930
},
{
"epoch": 1.8020493712156498,
"grad_norm": 0.5114907876049536,
"learning_rate": 3.0515277058518904e-05,
"loss": 0.3471,
"step": 1935
},
{
"epoch": 1.806707033069399,
"grad_norm": 0.4579020379986667,
"learning_rate": 3.045054375971e-05,
"loss": 0.3499,
"step": 1940
},
{
"epoch": 1.8113646949231486,
"grad_norm": 0.4162796744779169,
"learning_rate": 3.0385810460901086e-05,
"loss": 0.3602,
"step": 1945
},
{
"epoch": 1.816022356776898,
"grad_norm": 0.43293484925432135,
"learning_rate": 3.0321077162092183e-05,
"loss": 0.3516,
"step": 1950
},
{
"epoch": 1.8206800186306475,
"grad_norm": 0.38712551527265004,
"learning_rate": 3.0256343863283276e-05,
"loss": 0.3553,
"step": 1955
},
{
"epoch": 1.8253376804843968,
"grad_norm": 0.5392482471776439,
"learning_rate": 3.0191610564474365e-05,
"loss": 0.3565,
"step": 1960
},
{
"epoch": 1.829995342338146,
"grad_norm": 0.3943059237946394,
"learning_rate": 3.0126877265665458e-05,
"loss": 0.3593,
"step": 1965
},
{
"epoch": 1.8346530041918956,
"grad_norm": 0.4256479628934368,
"learning_rate": 3.0062143966856555e-05,
"loss": 0.3536,
"step": 1970
},
{
"epoch": 1.839310666045645,
"grad_norm": 0.5122761549068479,
"learning_rate": 2.9997410668047644e-05,
"loss": 0.3589,
"step": 1975
},
{
"epoch": 1.8439683278993946,
"grad_norm": 0.370961688185249,
"learning_rate": 2.9932677369238737e-05,
"loss": 0.3503,
"step": 1980
},
{
"epoch": 1.848625989753144,
"grad_norm": 0.4063678003465277,
"learning_rate": 2.9867944070429834e-05,
"loss": 0.3649,
"step": 1985
},
{
"epoch": 1.8532836516068933,
"grad_norm": 0.5043143808118292,
"learning_rate": 2.980321077162092e-05,
"loss": 0.3597,
"step": 1990
},
{
"epoch": 1.8579413134606426,
"grad_norm": 0.3888766687196224,
"learning_rate": 2.9738477472812016e-05,
"loss": 0.3478,
"step": 1995
},
{
"epoch": 1.8625989753143921,
"grad_norm": 0.3653540581805115,
"learning_rate": 2.967374417400311e-05,
"loss": 0.3563,
"step": 2000
},
{
"epoch": 1.8672566371681416,
"grad_norm": 0.42346683909033195,
"learning_rate": 2.96090108751942e-05,
"loss": 0.3626,
"step": 2005
},
{
"epoch": 1.871914299021891,
"grad_norm": 0.37053462856669767,
"learning_rate": 2.9544277576385292e-05,
"loss": 0.3561,
"step": 2010
},
{
"epoch": 1.8765719608756404,
"grad_norm": 0.5423835623045288,
"learning_rate": 2.9479544277576388e-05,
"loss": 0.368,
"step": 2015
},
{
"epoch": 1.8812296227293899,
"grad_norm": 0.4430785063248816,
"learning_rate": 2.941481097876748e-05,
"loss": 0.3532,
"step": 2020
},
{
"epoch": 1.8858872845831391,
"grad_norm": 0.5390136431233741,
"learning_rate": 2.935007767995857e-05,
"loss": 0.3568,
"step": 2025
},
{
"epoch": 1.8905449464368886,
"grad_norm": 0.42451261658331807,
"learning_rate": 2.9285344381149664e-05,
"loss": 0.3546,
"step": 2030
},
{
"epoch": 1.8952026082906381,
"grad_norm": 0.4704880484210268,
"learning_rate": 2.922061108234076e-05,
"loss": 0.3528,
"step": 2035
},
{
"epoch": 1.8998602701443876,
"grad_norm": 10.0374479730041,
"learning_rate": 2.915587778353185e-05,
"loss": 0.3662,
"step": 2040
},
{
"epoch": 1.904517931998137,
"grad_norm": 0.44064361841438693,
"learning_rate": 2.9091144484722943e-05,
"loss": 0.3639,
"step": 2045
},
{
"epoch": 1.9091755938518864,
"grad_norm": 0.6176856815800884,
"learning_rate": 2.902641118591404e-05,
"loss": 0.3606,
"step": 2050
},
{
"epoch": 1.9138332557056357,
"grad_norm": 0.5399754968576305,
"learning_rate": 2.8961677887105125e-05,
"loss": 0.3578,
"step": 2055
},
{
"epoch": 1.9184909175593852,
"grad_norm": 0.4645100850311685,
"learning_rate": 2.889694458829622e-05,
"loss": 0.3533,
"step": 2060
},
{
"epoch": 1.9231485794131347,
"grad_norm": 0.5042470599105712,
"learning_rate": 2.8832211289487315e-05,
"loss": 0.356,
"step": 2065
},
{
"epoch": 1.9278062412668842,
"grad_norm": 0.4839587513961929,
"learning_rate": 2.8767477990678404e-05,
"loss": 0.3584,
"step": 2070
},
{
"epoch": 1.9324639031206334,
"grad_norm": 0.5205667060014961,
"learning_rate": 2.8702744691869497e-05,
"loss": 0.3568,
"step": 2075
},
{
"epoch": 1.9371215649743827,
"grad_norm": 0.5177813037309902,
"learning_rate": 2.8638011393060594e-05,
"loss": 0.3505,
"step": 2080
},
{
"epoch": 1.9417792268281322,
"grad_norm": 0.4545545367610249,
"learning_rate": 2.8573278094251683e-05,
"loss": 0.3608,
"step": 2085
},
{
"epoch": 1.9464368886818817,
"grad_norm": 0.41189147778170226,
"learning_rate": 2.8508544795442776e-05,
"loss": 0.3452,
"step": 2090
},
{
"epoch": 1.9510945505356312,
"grad_norm": 0.4202290799914547,
"learning_rate": 2.8443811496633873e-05,
"loss": 0.3514,
"step": 2095
},
{
"epoch": 1.9557522123893807,
"grad_norm": 0.4377924322321175,
"learning_rate": 2.837907819782496e-05,
"loss": 0.3542,
"step": 2100
},
{
"epoch": 1.96040987424313,
"grad_norm": 0.5132414811078034,
"learning_rate": 2.8314344899016055e-05,
"loss": 0.3611,
"step": 2105
},
{
"epoch": 1.9650675360968792,
"grad_norm": 0.4929000956532224,
"learning_rate": 2.8249611600207148e-05,
"loss": 0.3487,
"step": 2110
},
{
"epoch": 1.9697251979506287,
"grad_norm": 0.49429590172686955,
"learning_rate": 2.8184878301398244e-05,
"loss": 0.3548,
"step": 2115
},
{
"epoch": 1.9743828598043782,
"grad_norm": 0.45457503823460055,
"learning_rate": 2.812014500258933e-05,
"loss": 0.3566,
"step": 2120
},
{
"epoch": 1.9790405216581277,
"grad_norm": 0.3816140696343873,
"learning_rate": 2.8055411703780427e-05,
"loss": 0.3517,
"step": 2125
},
{
"epoch": 1.983698183511877,
"grad_norm": 0.6242736771442058,
"learning_rate": 2.799067840497152e-05,
"loss": 0.3543,
"step": 2130
},
{
"epoch": 1.9883558453656265,
"grad_norm": 0.4277411480118539,
"learning_rate": 2.792594510616261e-05,
"loss": 0.3592,
"step": 2135
},
{
"epoch": 1.9930135072193758,
"grad_norm": 0.5159703918534981,
"learning_rate": 2.7861211807353703e-05,
"loss": 0.3579,
"step": 2140
},
{
"epoch": 1.9976711690731253,
"grad_norm": 0.4202566030642709,
"learning_rate": 2.77964785085448e-05,
"loss": 0.3585,
"step": 2145
},
{
"epoch": 2.0018630647415,
"grad_norm": 0.5609402586843426,
"learning_rate": 2.773174520973589e-05,
"loss": 0.3298,
"step": 2150
},
{
"epoch": 2.0065207265952494,
"grad_norm": 0.425941696873025,
"learning_rate": 2.766701191092698e-05,
"loss": 0.2798,
"step": 2155
},
{
"epoch": 2.0111783884489984,
"grad_norm": 0.3916657744615872,
"learning_rate": 2.7602278612118078e-05,
"loss": 0.2704,
"step": 2160
},
{
"epoch": 2.015836050302748,
"grad_norm": 0.43810490702164956,
"learning_rate": 2.7537545313309164e-05,
"loss": 0.2683,
"step": 2165
},
{
"epoch": 2.0204937121564974,
"grad_norm": 0.42179753955288946,
"learning_rate": 2.747281201450026e-05,
"loss": 0.2647,
"step": 2170
},
{
"epoch": 2.025151374010247,
"grad_norm": 0.41065922439891167,
"learning_rate": 2.7408078715691354e-05,
"loss": 0.2722,
"step": 2175
},
{
"epoch": 2.0298090358639964,
"grad_norm": 1.1811479608004933,
"learning_rate": 2.7343345416882443e-05,
"loss": 0.2658,
"step": 2180
},
{
"epoch": 2.034466697717746,
"grad_norm": 0.3840477047643718,
"learning_rate": 2.7278612118073536e-05,
"loss": 0.2684,
"step": 2185
},
{
"epoch": 2.039124359571495,
"grad_norm": 0.45197057438747157,
"learning_rate": 2.7213878819264633e-05,
"loss": 0.2725,
"step": 2190
},
{
"epoch": 2.0437820214252445,
"grad_norm": 0.36416091004272566,
"learning_rate": 2.7149145520455722e-05,
"loss": 0.268,
"step": 2195
},
{
"epoch": 2.048439683278994,
"grad_norm": 0.43811123788548967,
"learning_rate": 2.7084412221646815e-05,
"loss": 0.269,
"step": 2200
},
{
"epoch": 2.0530973451327434,
"grad_norm": 0.45910013565579505,
"learning_rate": 2.701967892283791e-05,
"loss": 0.2736,
"step": 2205
},
{
"epoch": 2.057755006986493,
"grad_norm": 0.4013422169686014,
"learning_rate": 2.6954945624029004e-05,
"loss": 0.2642,
"step": 2210
},
{
"epoch": 2.062412668840242,
"grad_norm": 0.4295693677688835,
"learning_rate": 2.6890212325220094e-05,
"loss": 0.2705,
"step": 2215
},
{
"epoch": 2.0670703306939915,
"grad_norm": 0.3370168623171349,
"learning_rate": 2.6825479026411187e-05,
"loss": 0.2677,
"step": 2220
},
{
"epoch": 2.071727992547741,
"grad_norm": 0.40197526047576093,
"learning_rate": 2.6760745727602283e-05,
"loss": 0.2668,
"step": 2225
},
{
"epoch": 2.0763856544014905,
"grad_norm": 0.45672730531903405,
"learning_rate": 2.669601242879337e-05,
"loss": 0.2701,
"step": 2230
},
{
"epoch": 2.08104331625524,
"grad_norm": 0.39868426985110295,
"learning_rate": 2.6631279129984466e-05,
"loss": 0.2648,
"step": 2235
},
{
"epoch": 2.0857009781089895,
"grad_norm": 0.36305570920930336,
"learning_rate": 2.656654583117556e-05,
"loss": 0.2761,
"step": 2240
},
{
"epoch": 2.0903586399627385,
"grad_norm": 0.4023887365515303,
"learning_rate": 2.650181253236665e-05,
"loss": 0.271,
"step": 2245
},
{
"epoch": 2.095016301816488,
"grad_norm": 0.3716267176543075,
"learning_rate": 2.643707923355774e-05,
"loss": 0.2667,
"step": 2250
},
{
"epoch": 2.0996739636702375,
"grad_norm": 0.4200410072696504,
"learning_rate": 2.6372345934748838e-05,
"loss": 0.2762,
"step": 2255
},
{
"epoch": 2.104331625523987,
"grad_norm": 0.3687755529593597,
"learning_rate": 2.6307612635939928e-05,
"loss": 0.2655,
"step": 2260
},
{
"epoch": 2.1089892873777365,
"grad_norm": 0.32809338851619163,
"learning_rate": 2.624287933713102e-05,
"loss": 0.2713,
"step": 2265
},
{
"epoch": 2.113646949231486,
"grad_norm": 0.3409360403110876,
"learning_rate": 2.6178146038322117e-05,
"loss": 0.2685,
"step": 2270
},
{
"epoch": 2.118304611085235,
"grad_norm": 0.7982011287578665,
"learning_rate": 2.6113412739513203e-05,
"loss": 0.2746,
"step": 2275
},
{
"epoch": 2.1229622729389845,
"grad_norm": 0.3617141903798149,
"learning_rate": 2.60486794407043e-05,
"loss": 0.2715,
"step": 2280
},
{
"epoch": 2.127619934792734,
"grad_norm": 0.3727969530757095,
"learning_rate": 2.5983946141895393e-05,
"loss": 0.2738,
"step": 2285
},
{
"epoch": 2.1322775966464835,
"grad_norm": 0.40537658839373925,
"learning_rate": 2.5919212843086482e-05,
"loss": 0.2713,
"step": 2290
},
{
"epoch": 2.136935258500233,
"grad_norm": 0.4592856921511742,
"learning_rate": 2.5854479544277575e-05,
"loss": 0.2786,
"step": 2295
},
{
"epoch": 2.1415929203539825,
"grad_norm": 0.3602958767186378,
"learning_rate": 2.578974624546867e-05,
"loss": 0.2763,
"step": 2300
},
{
"epoch": 2.1462505822077316,
"grad_norm": 0.36531905883050036,
"learning_rate": 2.5725012946659764e-05,
"loss": 0.2732,
"step": 2305
},
{
"epoch": 2.150908244061481,
"grad_norm": 0.3623277574575608,
"learning_rate": 2.5660279647850854e-05,
"loss": 0.2643,
"step": 2310
},
{
"epoch": 2.1555659059152306,
"grad_norm": 0.3884139128194696,
"learning_rate": 2.559554634904195e-05,
"loss": 0.2693,
"step": 2315
},
{
"epoch": 2.16022356776898,
"grad_norm": 0.3919757281405461,
"learning_rate": 2.5530813050233043e-05,
"loss": 0.2656,
"step": 2320
},
{
"epoch": 2.1648812296227296,
"grad_norm": 0.4016863934977521,
"learning_rate": 2.5466079751424133e-05,
"loss": 0.2742,
"step": 2325
},
{
"epoch": 2.1695388914764786,
"grad_norm": 0.31887836235970607,
"learning_rate": 2.5401346452615226e-05,
"loss": 0.2737,
"step": 2330
},
{
"epoch": 2.174196553330228,
"grad_norm": 0.40350840902874313,
"learning_rate": 2.5336613153806322e-05,
"loss": 0.2662,
"step": 2335
},
{
"epoch": 2.1788542151839776,
"grad_norm": 0.4135565972212336,
"learning_rate": 2.527187985499741e-05,
"loss": 0.2717,
"step": 2340
},
{
"epoch": 2.183511877037727,
"grad_norm": 0.31536031277914545,
"learning_rate": 2.5207146556188505e-05,
"loss": 0.2772,
"step": 2345
},
{
"epoch": 2.1881695388914766,
"grad_norm": 0.34236501501838795,
"learning_rate": 2.5142413257379598e-05,
"loss": 0.271,
"step": 2350
},
{
"epoch": 2.192827200745226,
"grad_norm": 0.317656594570138,
"learning_rate": 2.5077679958570688e-05,
"loss": 0.2682,
"step": 2355
},
{
"epoch": 2.197484862598975,
"grad_norm": 0.3661371683010254,
"learning_rate": 2.5012946659761784e-05,
"loss": 0.2755,
"step": 2360
},
{
"epoch": 2.2021425244527246,
"grad_norm": 0.41181534702112593,
"learning_rate": 2.4948213360952874e-05,
"loss": 0.2738,
"step": 2365
},
{
"epoch": 2.206800186306474,
"grad_norm": 0.3653865073116719,
"learning_rate": 2.488348006214397e-05,
"loss": 0.2736,
"step": 2370
},
{
"epoch": 2.2114578481602236,
"grad_norm": 0.39203577537195466,
"learning_rate": 2.481874676333506e-05,
"loss": 0.2742,
"step": 2375
},
{
"epoch": 2.216115510013973,
"grad_norm": 0.3720934862670963,
"learning_rate": 2.4754013464526156e-05,
"loss": 0.2724,
"step": 2380
},
{
"epoch": 2.2207731718677226,
"grad_norm": 0.34171876985926397,
"learning_rate": 2.4689280165717246e-05,
"loss": 0.2831,
"step": 2385
},
{
"epoch": 2.2254308337214717,
"grad_norm": 0.39944570037740157,
"learning_rate": 2.462454686690834e-05,
"loss": 0.2794,
"step": 2390
},
{
"epoch": 2.230088495575221,
"grad_norm": 0.3550895446151112,
"learning_rate": 2.455981356809943e-05,
"loss": 0.2725,
"step": 2395
},
{
"epoch": 2.2347461574289706,
"grad_norm": 0.4970447794860612,
"learning_rate": 2.4495080269290524e-05,
"loss": 0.2644,
"step": 2400
},
{
"epoch": 2.23940381928272,
"grad_norm": 0.4050207598981995,
"learning_rate": 2.4430346970481614e-05,
"loss": 0.2732,
"step": 2405
},
{
"epoch": 2.2440614811364696,
"grad_norm": 0.3707865091431949,
"learning_rate": 2.436561367167271e-05,
"loss": 0.2764,
"step": 2410
},
{
"epoch": 2.248719142990219,
"grad_norm": 0.4063161340916692,
"learning_rate": 2.4300880372863803e-05,
"loss": 0.2708,
"step": 2415
},
{
"epoch": 2.253376804843968,
"grad_norm": 0.42036563058341386,
"learning_rate": 2.4236147074054896e-05,
"loss": 0.265,
"step": 2420
},
{
"epoch": 2.2580344666977177,
"grad_norm": 0.3543048649400106,
"learning_rate": 2.417141377524599e-05,
"loss": 0.2751,
"step": 2425
},
{
"epoch": 2.262692128551467,
"grad_norm": 0.3983712556686238,
"learning_rate": 2.410668047643708e-05,
"loss": 0.2675,
"step": 2430
},
{
"epoch": 2.2673497904052167,
"grad_norm": 0.3530424561412936,
"learning_rate": 2.4041947177628175e-05,
"loss": 0.2775,
"step": 2435
},
{
"epoch": 2.272007452258966,
"grad_norm": 0.3831906451800883,
"learning_rate": 2.3977213878819265e-05,
"loss": 0.2718,
"step": 2440
},
{
"epoch": 2.276665114112715,
"grad_norm": 0.3629826382220657,
"learning_rate": 2.3912480580010358e-05,
"loss": 0.2714,
"step": 2445
},
{
"epoch": 2.2813227759664647,
"grad_norm": 0.3723371127396783,
"learning_rate": 2.384774728120145e-05,
"loss": 0.2713,
"step": 2450
},
{
"epoch": 2.285980437820214,
"grad_norm": 0.33116565655664815,
"learning_rate": 2.3783013982392544e-05,
"loss": 0.2714,
"step": 2455
},
{
"epoch": 2.2906380996739637,
"grad_norm": 0.4241473682280153,
"learning_rate": 2.3718280683583634e-05,
"loss": 0.2697,
"step": 2460
},
{
"epoch": 2.295295761527713,
"grad_norm": 0.39705043895634146,
"learning_rate": 2.365354738477473e-05,
"loss": 0.2822,
"step": 2465
},
{
"epoch": 2.2999534233814627,
"grad_norm": 0.33163639660401323,
"learning_rate": 2.3588814085965823e-05,
"loss": 0.2718,
"step": 2470
},
{
"epoch": 2.3046110852352117,
"grad_norm": 0.33512581809854164,
"learning_rate": 2.3524080787156916e-05,
"loss": 0.2784,
"step": 2475
},
{
"epoch": 2.3092687470889612,
"grad_norm": 0.3821131258454766,
"learning_rate": 2.345934748834801e-05,
"loss": 0.2718,
"step": 2480
},
{
"epoch": 2.3139264089427107,
"grad_norm": 0.39074319401185925,
"learning_rate": 2.33946141895391e-05,
"loss": 0.2783,
"step": 2485
},
{
"epoch": 2.3185840707964602,
"grad_norm": 0.33902978499643044,
"learning_rate": 2.3329880890730195e-05,
"loss": 0.2783,
"step": 2490
},
{
"epoch": 2.3232417326502097,
"grad_norm": 0.3827711970500847,
"learning_rate": 2.3265147591921284e-05,
"loss": 0.2703,
"step": 2495
},
{
"epoch": 2.3278993945039588,
"grad_norm": 0.355781107832294,
"learning_rate": 2.3200414293112377e-05,
"loss": 0.2652,
"step": 2500
},
{
"epoch": 2.3325570563577083,
"grad_norm": 0.33162879789458566,
"learning_rate": 2.313568099430347e-05,
"loss": 0.2717,
"step": 2505
},
{
"epoch": 2.3372147182114578,
"grad_norm": 0.4338763189528674,
"learning_rate": 2.3070947695494563e-05,
"loss": 0.275,
"step": 2510
},
{
"epoch": 2.3418723800652073,
"grad_norm": 0.383583192216027,
"learning_rate": 2.3006214396685656e-05,
"loss": 0.2729,
"step": 2515
},
{
"epoch": 2.3465300419189568,
"grad_norm": 0.32609506468335425,
"learning_rate": 2.294148109787675e-05,
"loss": 0.275,
"step": 2520
},
{
"epoch": 2.3511877037727063,
"grad_norm": 0.3335278023149569,
"learning_rate": 2.2876747799067842e-05,
"loss": 0.2755,
"step": 2525
},
{
"epoch": 2.3558453656264557,
"grad_norm": 0.3350637802852781,
"learning_rate": 2.2812014500258935e-05,
"loss": 0.2732,
"step": 2530
},
{
"epoch": 2.360503027480205,
"grad_norm": 0.37866460267486196,
"learning_rate": 2.274728120145003e-05,
"loss": 0.268,
"step": 2535
},
{
"epoch": 2.3651606893339543,
"grad_norm": 0.34555155639658347,
"learning_rate": 2.2682547902641118e-05,
"loss": 0.2676,
"step": 2540
},
{
"epoch": 2.369818351187704,
"grad_norm": 0.3258434102060813,
"learning_rate": 2.2617814603832214e-05,
"loss": 0.2773,
"step": 2545
},
{
"epoch": 2.3744760130414533,
"grad_norm": 0.3658190762241647,
"learning_rate": 2.2553081305023304e-05,
"loss": 0.2778,
"step": 2550
},
{
"epoch": 2.3791336748952028,
"grad_norm": 0.4542579563180514,
"learning_rate": 2.2488348006214397e-05,
"loss": 0.2732,
"step": 2555
},
{
"epoch": 2.383791336748952,
"grad_norm": 0.3464667680960837,
"learning_rate": 2.242361470740549e-05,
"loss": 0.2727,
"step": 2560
},
{
"epoch": 2.3884489986027013,
"grad_norm": 0.38709489671449077,
"learning_rate": 2.2358881408596583e-05,
"loss": 0.2751,
"step": 2565
},
{
"epoch": 2.393106660456451,
"grad_norm": 0.35458794378570924,
"learning_rate": 2.2294148109787676e-05,
"loss": 0.2766,
"step": 2570
},
{
"epoch": 2.3977643223102003,
"grad_norm": 0.31586159359748556,
"learning_rate": 2.222941481097877e-05,
"loss": 0.2747,
"step": 2575
},
{
"epoch": 2.40242198416395,
"grad_norm": 0.47950668564649523,
"learning_rate": 2.2164681512169862e-05,
"loss": 0.2678,
"step": 2580
},
{
"epoch": 2.4070796460176993,
"grad_norm": 0.3547033989457256,
"learning_rate": 2.2099948213360955e-05,
"loss": 0.2714,
"step": 2585
},
{
"epoch": 2.4117373078714484,
"grad_norm": 0.3417274119177762,
"learning_rate": 2.2035214914552048e-05,
"loss": 0.2656,
"step": 2590
},
{
"epoch": 2.416394969725198,
"grad_norm": 0.41003338636699,
"learning_rate": 2.1970481615743137e-05,
"loss": 0.2814,
"step": 2595
},
{
"epoch": 2.4210526315789473,
"grad_norm": 0.3344591203299114,
"learning_rate": 2.1905748316934234e-05,
"loss": 0.2773,
"step": 2600
},
{
"epoch": 2.425710293432697,
"grad_norm": 0.3688753394872757,
"learning_rate": 2.1841015018125323e-05,
"loss": 0.2714,
"step": 2605
},
{
"epoch": 2.4303679552864463,
"grad_norm": 0.4134992480901522,
"learning_rate": 2.177628171931642e-05,
"loss": 0.2688,
"step": 2610
},
{
"epoch": 2.4350256171401954,
"grad_norm": 0.354197807356868,
"learning_rate": 2.171154842050751e-05,
"loss": 0.2695,
"step": 2615
},
{
"epoch": 2.439683278993945,
"grad_norm": 0.384555222770943,
"learning_rate": 2.1646815121698602e-05,
"loss": 0.2724,
"step": 2620
},
{
"epoch": 2.4443409408476944,
"grad_norm": 0.33945672656475084,
"learning_rate": 2.1582081822889695e-05,
"loss": 0.2672,
"step": 2625
},
{
"epoch": 2.448998602701444,
"grad_norm": 0.35223398070317613,
"learning_rate": 2.151734852408079e-05,
"loss": 0.266,
"step": 2630
},
{
"epoch": 2.4536562645551934,
"grad_norm": 0.33538981080275043,
"learning_rate": 2.145261522527188e-05,
"loss": 0.2752,
"step": 2635
},
{
"epoch": 2.458313926408943,
"grad_norm": 1.7424507890291698,
"learning_rate": 2.1387881926462974e-05,
"loss": 0.2756,
"step": 2640
},
{
"epoch": 2.4629715882626924,
"grad_norm": 0.36857578732146257,
"learning_rate": 2.1323148627654067e-05,
"loss": 0.2753,
"step": 2645
},
{
"epoch": 2.4676292501164414,
"grad_norm": 0.37901471619034555,
"learning_rate": 2.1258415328845157e-05,
"loss": 0.2708,
"step": 2650
},
{
"epoch": 2.472286911970191,
"grad_norm": 0.3728918334905263,
"learning_rate": 2.1193682030036253e-05,
"loss": 0.2721,
"step": 2655
},
{
"epoch": 2.4769445738239404,
"grad_norm": 0.3349864522210704,
"learning_rate": 2.1128948731227343e-05,
"loss": 0.2701,
"step": 2660
},
{
"epoch": 2.48160223567769,
"grad_norm": 0.3221018339130229,
"learning_rate": 2.106421543241844e-05,
"loss": 0.2714,
"step": 2665
},
{
"epoch": 2.4862598975314394,
"grad_norm": 0.372009777674902,
"learning_rate": 2.099948213360953e-05,
"loss": 0.2712,
"step": 2670
},
{
"epoch": 2.4909175593851884,
"grad_norm": 0.31144393603273973,
"learning_rate": 2.0934748834800622e-05,
"loss": 0.2686,
"step": 2675
},
{
"epoch": 2.495575221238938,
"grad_norm": 0.37310898554154925,
"learning_rate": 2.0870015535991715e-05,
"loss": 0.268,
"step": 2680
},
{
"epoch": 2.5002328830926874,
"grad_norm": 0.32429912612731604,
"learning_rate": 2.0805282237182808e-05,
"loss": 0.2722,
"step": 2685
},
{
"epoch": 2.504890544946437,
"grad_norm": 0.3645155853567906,
"learning_rate": 2.07405489383739e-05,
"loss": 0.2625,
"step": 2690
},
{
"epoch": 2.5095482068001864,
"grad_norm": 0.37285481357452527,
"learning_rate": 2.0675815639564994e-05,
"loss": 0.269,
"step": 2695
},
{
"epoch": 2.514205868653936,
"grad_norm": 0.4059330000248755,
"learning_rate": 2.0611082340756087e-05,
"loss": 0.2647,
"step": 2700
},
{
"epoch": 2.5188635305076854,
"grad_norm": 0.3436088518040214,
"learning_rate": 2.054634904194718e-05,
"loss": 0.2742,
"step": 2705
},
{
"epoch": 2.5235211923614345,
"grad_norm": 0.3972933011802129,
"learning_rate": 2.0481615743138273e-05,
"loss": 0.269,
"step": 2710
},
{
"epoch": 2.528178854215184,
"grad_norm": 0.37360697308076835,
"learning_rate": 2.0416882444329362e-05,
"loss": 0.2742,
"step": 2715
},
{
"epoch": 2.5328365160689335,
"grad_norm": 0.33376962508013164,
"learning_rate": 2.035214914552046e-05,
"loss": 0.2788,
"step": 2720
},
{
"epoch": 2.537494177922683,
"grad_norm": 0.3907774958559364,
"learning_rate": 2.028741584671155e-05,
"loss": 0.2749,
"step": 2725
},
{
"epoch": 2.542151839776432,
"grad_norm": 0.36884223475006717,
"learning_rate": 2.022268254790264e-05,
"loss": 0.2826,
"step": 2730
},
{
"epoch": 2.5468095016301815,
"grad_norm": 0.37074560443330334,
"learning_rate": 2.0157949249093734e-05,
"loss": 0.2695,
"step": 2735
},
{
"epoch": 2.551467163483931,
"grad_norm": 0.3160412629320368,
"learning_rate": 2.0093215950284827e-05,
"loss": 0.2679,
"step": 2740
},
{
"epoch": 2.5561248253376805,
"grad_norm": 0.3034071644528673,
"learning_rate": 2.002848265147592e-05,
"loss": 0.2698,
"step": 2745
},
{
"epoch": 2.56078248719143,
"grad_norm": 0.31258186729345505,
"learning_rate": 1.9963749352667013e-05,
"loss": 0.2679,
"step": 2750
},
{
"epoch": 2.5654401490451795,
"grad_norm": 0.292440405711894,
"learning_rate": 1.9899016053858106e-05,
"loss": 0.2684,
"step": 2755
},
{
"epoch": 2.570097810898929,
"grad_norm": 0.3363867700320085,
"learning_rate": 1.98342827550492e-05,
"loss": 0.2711,
"step": 2760
},
{
"epoch": 2.574755472752678,
"grad_norm": 0.33149638022316674,
"learning_rate": 1.9769549456240292e-05,
"loss": 0.2681,
"step": 2765
},
{
"epoch": 2.5794131346064275,
"grad_norm": 0.34029720041256184,
"learning_rate": 1.9704816157431382e-05,
"loss": 0.2694,
"step": 2770
},
{
"epoch": 2.584070796460177,
"grad_norm": 0.3279656882775838,
"learning_rate": 1.9640082858622478e-05,
"loss": 0.2671,
"step": 2775
},
{
"epoch": 2.5887284583139265,
"grad_norm": 0.3447709556024024,
"learning_rate": 1.9575349559813568e-05,
"loss": 0.2757,
"step": 2780
},
{
"epoch": 2.5933861201676756,
"grad_norm": 0.3106673292073178,
"learning_rate": 1.951061626100466e-05,
"loss": 0.2752,
"step": 2785
},
{
"epoch": 2.598043782021425,
"grad_norm": 0.3105299991093135,
"learning_rate": 1.9445882962195754e-05,
"loss": 0.2771,
"step": 2790
},
{
"epoch": 2.6027014438751745,
"grad_norm": 0.2982018872389286,
"learning_rate": 1.9381149663386847e-05,
"loss": 0.2711,
"step": 2795
},
{
"epoch": 2.607359105728924,
"grad_norm": 0.31038185590696776,
"learning_rate": 1.931641636457794e-05,
"loss": 0.2719,
"step": 2800
},
{
"epoch": 2.6120167675826735,
"grad_norm": 0.4576009994202879,
"learning_rate": 1.9251683065769033e-05,
"loss": 0.272,
"step": 2805
},
{
"epoch": 2.616674429436423,
"grad_norm": 0.38717499530156124,
"learning_rate": 1.9186949766960126e-05,
"loss": 0.272,
"step": 2810
},
{
"epoch": 2.6213320912901725,
"grad_norm": 0.3134007638188876,
"learning_rate": 1.912221646815122e-05,
"loss": 0.2679,
"step": 2815
},
{
"epoch": 2.625989753143922,
"grad_norm": 0.3662008822975538,
"learning_rate": 1.9057483169342312e-05,
"loss": 0.275,
"step": 2820
},
{
"epoch": 2.630647414997671,
"grad_norm": 0.338465603582536,
"learning_rate": 1.89927498705334e-05,
"loss": 0.275,
"step": 2825
},
{
"epoch": 2.6353050768514206,
"grad_norm": 0.39664923263948576,
"learning_rate": 1.8928016571724498e-05,
"loss": 0.2696,
"step": 2830
},
{
"epoch": 2.63996273870517,
"grad_norm": 0.3358218724089584,
"learning_rate": 1.8863283272915587e-05,
"loss": 0.277,
"step": 2835
},
{
"epoch": 2.6446204005589196,
"grad_norm": 0.39806694330638576,
"learning_rate": 1.879854997410668e-05,
"loss": 0.2801,
"step": 2840
},
{
"epoch": 2.6492780624126686,
"grad_norm": 0.3426052650712918,
"learning_rate": 1.8733816675297773e-05,
"loss": 0.2702,
"step": 2845
},
{
"epoch": 2.653935724266418,
"grad_norm": 0.321019545867809,
"learning_rate": 1.8669083376488866e-05,
"loss": 0.2769,
"step": 2850
},
{
"epoch": 2.6585933861201676,
"grad_norm": 0.32384371615715174,
"learning_rate": 1.860435007767996e-05,
"loss": 0.2678,
"step": 2855
},
{
"epoch": 2.663251047973917,
"grad_norm": 0.3290749039774749,
"learning_rate": 1.8539616778871052e-05,
"loss": 0.273,
"step": 2860
},
{
"epoch": 2.6679087098276666,
"grad_norm": 0.33856866467686797,
"learning_rate": 1.8474883480062145e-05,
"loss": 0.2725,
"step": 2865
},
{
"epoch": 2.672566371681416,
"grad_norm": 0.3500432840827306,
"learning_rate": 1.8410150181253238e-05,
"loss": 0.2731,
"step": 2870
},
{
"epoch": 2.6772240335351656,
"grad_norm": 0.38586231997017223,
"learning_rate": 1.834541688244433e-05,
"loss": 0.2719,
"step": 2875
},
{
"epoch": 2.6818816953889146,
"grad_norm": 0.31604905067500316,
"learning_rate": 1.828068358363542e-05,
"loss": 0.2713,
"step": 2880
},
{
"epoch": 2.686539357242664,
"grad_norm": 0.34688343054005744,
"learning_rate": 1.8215950284826517e-05,
"loss": 0.2793,
"step": 2885
},
{
"epoch": 2.6911970190964136,
"grad_norm": 0.30974497872868545,
"learning_rate": 1.8151216986017607e-05,
"loss": 0.2691,
"step": 2890
},
{
"epoch": 2.695854680950163,
"grad_norm": 0.38316570331805355,
"learning_rate": 1.8086483687208703e-05,
"loss": 0.2678,
"step": 2895
},
{
"epoch": 2.700512342803912,
"grad_norm": 0.33852769233141367,
"learning_rate": 1.8021750388399793e-05,
"loss": 0.2726,
"step": 2900
},
{
"epoch": 2.7051700046576617,
"grad_norm": 0.3593989177077844,
"learning_rate": 1.7957017089590886e-05,
"loss": 0.2641,
"step": 2905
},
{
"epoch": 2.709827666511411,
"grad_norm": 0.30875552983332166,
"learning_rate": 1.789228379078198e-05,
"loss": 0.2673,
"step": 2910
},
{
"epoch": 2.7144853283651607,
"grad_norm": 0.35727081049388676,
"learning_rate": 1.7827550491973072e-05,
"loss": 0.277,
"step": 2915
},
{
"epoch": 2.71914299021891,
"grad_norm": 0.3339865640541083,
"learning_rate": 1.7762817193164165e-05,
"loss": 0.2685,
"step": 2920
},
{
"epoch": 2.7238006520726596,
"grad_norm": 0.33169965373972965,
"learning_rate": 1.7698083894355258e-05,
"loss": 0.2756,
"step": 2925
},
{
"epoch": 2.728458313926409,
"grad_norm": 0.3157634299132718,
"learning_rate": 1.763335059554635e-05,
"loss": 0.2774,
"step": 2930
},
{
"epoch": 2.7331159757801586,
"grad_norm": 0.3396944635203724,
"learning_rate": 1.756861729673744e-05,
"loss": 0.2728,
"step": 2935
},
{
"epoch": 2.7377736376339077,
"grad_norm": 0.35200393420014703,
"learning_rate": 1.7503883997928537e-05,
"loss": 0.2725,
"step": 2940
},
{
"epoch": 2.742431299487657,
"grad_norm": 0.37998854310284846,
"learning_rate": 1.7439150699119626e-05,
"loss": 0.2776,
"step": 2945
},
{
"epoch": 2.7470889613414067,
"grad_norm": 0.3576281147469936,
"learning_rate": 1.7374417400310723e-05,
"loss": 0.2668,
"step": 2950
},
{
"epoch": 2.751746623195156,
"grad_norm": 0.3635900744428691,
"learning_rate": 1.7309684101501812e-05,
"loss": 0.272,
"step": 2955
},
{
"epoch": 2.7564042850489052,
"grad_norm": 0.33511685467814833,
"learning_rate": 1.7244950802692905e-05,
"loss": 0.2717,
"step": 2960
},
{
"epoch": 2.7610619469026547,
"grad_norm": 0.3497985322649408,
"learning_rate": 1.7180217503883998e-05,
"loss": 0.2734,
"step": 2965
},
{
"epoch": 2.765719608756404,
"grad_norm": 0.353483359237389,
"learning_rate": 1.711548420507509e-05,
"loss": 0.2753,
"step": 2970
},
{
"epoch": 2.7703772706101537,
"grad_norm": 0.3721193002897029,
"learning_rate": 1.7050750906266184e-05,
"loss": 0.2712,
"step": 2975
},
{
"epoch": 2.775034932463903,
"grad_norm": 0.3312140027398143,
"learning_rate": 1.6986017607457277e-05,
"loss": 0.2692,
"step": 2980
},
{
"epoch": 2.7796925943176527,
"grad_norm": 0.39640844681767,
"learning_rate": 1.692128430864837e-05,
"loss": 0.2768,
"step": 2985
},
{
"epoch": 2.784350256171402,
"grad_norm": 0.3175338400111014,
"learning_rate": 1.6856551009839463e-05,
"loss": 0.2716,
"step": 2990
},
{
"epoch": 2.7890079180251512,
"grad_norm": 0.31652286586007294,
"learning_rate": 1.6791817711030556e-05,
"loss": 0.2685,
"step": 2995
},
{
"epoch": 2.7936655798789007,
"grad_norm": 0.28761607644050446,
"learning_rate": 1.6727084412221646e-05,
"loss": 0.2755,
"step": 3000
},
{
"epoch": 2.7983232417326502,
"grad_norm": 0.3099040855305692,
"learning_rate": 1.6662351113412742e-05,
"loss": 0.2693,
"step": 3005
},
{
"epoch": 2.8029809035863997,
"grad_norm": 0.3553610602893777,
"learning_rate": 1.6597617814603832e-05,
"loss": 0.2765,
"step": 3010
},
{
"epoch": 2.807638565440149,
"grad_norm": 0.31104188100335034,
"learning_rate": 1.6532884515794925e-05,
"loss": 0.2596,
"step": 3015
},
{
"epoch": 2.8122962272938983,
"grad_norm": 1.220512975098523,
"learning_rate": 1.6468151216986018e-05,
"loss": 0.2753,
"step": 3020
},
{
"epoch": 2.8169538891476478,
"grad_norm": 0.3631142231781718,
"learning_rate": 1.640341791817711e-05,
"loss": 0.2712,
"step": 3025
},
{
"epoch": 2.8216115510013973,
"grad_norm": 0.30471054985328155,
"learning_rate": 1.6338684619368204e-05,
"loss": 0.2713,
"step": 3030
},
{
"epoch": 2.8262692128551468,
"grad_norm": 0.3372934949628952,
"learning_rate": 1.6273951320559297e-05,
"loss": 0.2745,
"step": 3035
},
{
"epoch": 2.8309268747088963,
"grad_norm": 0.3077738977129297,
"learning_rate": 1.620921802175039e-05,
"loss": 0.2794,
"step": 3040
},
{
"epoch": 2.8355845365626458,
"grad_norm": 0.3009609913128321,
"learning_rate": 1.6144484722941483e-05,
"loss": 0.2706,
"step": 3045
},
{
"epoch": 2.840242198416395,
"grad_norm": 0.30480428019972094,
"learning_rate": 1.6079751424132576e-05,
"loss": 0.2674,
"step": 3050
},
{
"epoch": 2.8448998602701443,
"grad_norm": 0.38749267527788284,
"learning_rate": 1.6015018125323665e-05,
"loss": 0.2763,
"step": 3055
},
{
"epoch": 2.849557522123894,
"grad_norm": 0.2810429589384943,
"learning_rate": 1.595028482651476e-05,
"loss": 0.2789,
"step": 3060
},
{
"epoch": 2.8542151839776433,
"grad_norm": 0.30523685729564104,
"learning_rate": 1.588555152770585e-05,
"loss": 0.2737,
"step": 3065
},
{
"epoch": 2.858872845831393,
"grad_norm": 0.35926120588650146,
"learning_rate": 1.5820818228896944e-05,
"loss": 0.2643,
"step": 3070
},
{
"epoch": 2.863530507685142,
"grad_norm": 0.3313225013919279,
"learning_rate": 1.5756084930088037e-05,
"loss": 0.2724,
"step": 3075
},
{
"epoch": 2.8681881695388913,
"grad_norm": 0.3244478248567013,
"learning_rate": 1.569135163127913e-05,
"loss": 0.2626,
"step": 3080
},
{
"epoch": 2.872845831392641,
"grad_norm": 0.28666274499426153,
"learning_rate": 1.5626618332470223e-05,
"loss": 0.2696,
"step": 3085
},
{
"epoch": 2.8775034932463903,
"grad_norm": 0.35856467204541537,
"learning_rate": 1.5561885033661316e-05,
"loss": 0.2767,
"step": 3090
},
{
"epoch": 2.88216115510014,
"grad_norm": 0.3245668012785978,
"learning_rate": 1.549715173485241e-05,
"loss": 0.2677,
"step": 3095
},
{
"epoch": 2.8868188169538893,
"grad_norm": 0.35666892295965125,
"learning_rate": 1.5432418436043502e-05,
"loss": 0.2733,
"step": 3100
},
{
"epoch": 2.891476478807639,
"grad_norm": 0.3356911176206139,
"learning_rate": 1.5367685137234595e-05,
"loss": 0.2749,
"step": 3105
},
{
"epoch": 2.896134140661388,
"grad_norm": 0.29808214193997173,
"learning_rate": 1.5302951838425685e-05,
"loss": 0.2658,
"step": 3110
},
{
"epoch": 2.9007918025151374,
"grad_norm": 0.3365027074012271,
"learning_rate": 1.5238218539616781e-05,
"loss": 0.2638,
"step": 3115
},
{
"epoch": 2.905449464368887,
"grad_norm": 0.31649157528469446,
"learning_rate": 1.5173485240807872e-05,
"loss": 0.2713,
"step": 3120
},
{
"epoch": 2.9101071262226363,
"grad_norm": 0.39941610710037784,
"learning_rate": 1.5108751941998964e-05,
"loss": 0.2728,
"step": 3125
},
{
"epoch": 2.9147647880763854,
"grad_norm": 0.3350365625256805,
"learning_rate": 1.5044018643190058e-05,
"loss": 0.2636,
"step": 3130
},
{
"epoch": 2.919422449930135,
"grad_norm": 0.3121317275501377,
"learning_rate": 1.497928534438115e-05,
"loss": 0.2683,
"step": 3135
},
{
"epoch": 2.9240801117838844,
"grad_norm": 0.32992397885946134,
"learning_rate": 1.4914552045572244e-05,
"loss": 0.2771,
"step": 3140
},
{
"epoch": 2.928737773637634,
"grad_norm": 0.3104785612542341,
"learning_rate": 1.4849818746763336e-05,
"loss": 0.2802,
"step": 3145
},
{
"epoch": 2.9333954354913834,
"grad_norm": 0.3284169515856172,
"learning_rate": 1.4785085447954427e-05,
"loss": 0.2717,
"step": 3150
},
{
"epoch": 2.938053097345133,
"grad_norm": 0.5149080066378856,
"learning_rate": 1.4720352149145522e-05,
"loss": 0.2696,
"step": 3155
},
{
"epoch": 2.9427107591988824,
"grad_norm": 0.33286678134806846,
"learning_rate": 1.4655618850336613e-05,
"loss": 0.2715,
"step": 3160
},
{
"epoch": 2.9473684210526314,
"grad_norm": 0.29991185736571746,
"learning_rate": 1.4590885551527706e-05,
"loss": 0.2706,
"step": 3165
},
{
"epoch": 2.952026082906381,
"grad_norm": 0.2981828082571127,
"learning_rate": 1.45261522527188e-05,
"loss": 0.2692,
"step": 3170
},
{
"epoch": 2.9566837447601304,
"grad_norm": 0.30858159031175714,
"learning_rate": 1.4461418953909892e-05,
"loss": 0.2723,
"step": 3175
},
{
"epoch": 2.96134140661388,
"grad_norm": 0.34811263846999796,
"learning_rate": 1.4396685655100987e-05,
"loss": 0.2701,
"step": 3180
},
{
"epoch": 2.9659990684676294,
"grad_norm": 0.3873651952945082,
"learning_rate": 1.4331952356292078e-05,
"loss": 0.2688,
"step": 3185
},
{
"epoch": 2.9706567303213784,
"grad_norm": 0.3522941439771578,
"learning_rate": 1.4267219057483169e-05,
"loss": 0.2634,
"step": 3190
},
{
"epoch": 2.975314392175128,
"grad_norm": 0.34341967428607423,
"learning_rate": 1.4202485758674264e-05,
"loss": 0.2753,
"step": 3195
},
{
"epoch": 2.9799720540288774,
"grad_norm": 0.3284918143676433,
"learning_rate": 1.4137752459865355e-05,
"loss": 0.2725,
"step": 3200
},
{
"epoch": 2.984629715882627,
"grad_norm": 0.3150251773196763,
"learning_rate": 1.4073019161056446e-05,
"loss": 0.2729,
"step": 3205
},
{
"epoch": 2.9892873777363764,
"grad_norm": 0.3115665963455036,
"learning_rate": 1.4008285862247541e-05,
"loss": 0.2587,
"step": 3210
},
{
"epoch": 2.993945039590126,
"grad_norm": 0.2930991072668908,
"learning_rate": 1.3943552563438634e-05,
"loss": 0.2585,
"step": 3215
},
{
"epoch": 2.9986027014438754,
"grad_norm": 0.3122070563393351,
"learning_rate": 1.3878819264629725e-05,
"loss": 0.2753,
"step": 3220
},
{
"epoch": 3.0027945971122496,
"grad_norm": 0.3304819813611403,
"learning_rate": 1.381408596582082e-05,
"loss": 0.2101,
"step": 3225
},
{
"epoch": 3.007452258965999,
"grad_norm": 0.37549362163456124,
"learning_rate": 1.3749352667011911e-05,
"loss": 0.183,
"step": 3230
},
{
"epoch": 3.0121099208197486,
"grad_norm": 0.3748665980288928,
"learning_rate": 1.3684619368203006e-05,
"loss": 0.179,
"step": 3235
},
{
"epoch": 3.016767582673498,
"grad_norm": 0.3134994608529235,
"learning_rate": 1.3619886069394097e-05,
"loss": 0.1742,
"step": 3240
},
{
"epoch": 3.021425244527247,
"grad_norm": 0.3056502084704337,
"learning_rate": 1.3555152770585189e-05,
"loss": 0.1806,
"step": 3245
},
{
"epoch": 3.0260829063809966,
"grad_norm": 0.31098559463773384,
"learning_rate": 1.3490419471776283e-05,
"loss": 0.1759,
"step": 3250
},
{
"epoch": 3.030740568234746,
"grad_norm": 0.3156350772890806,
"learning_rate": 1.3425686172967375e-05,
"loss": 0.1796,
"step": 3255
},
{
"epoch": 3.0353982300884956,
"grad_norm": 0.34419970542768846,
"learning_rate": 1.3360952874158466e-05,
"loss": 0.1804,
"step": 3260
},
{
"epoch": 3.040055891942245,
"grad_norm": 0.3199306178559233,
"learning_rate": 1.329621957534956e-05,
"loss": 0.1796,
"step": 3265
},
{
"epoch": 3.0447135537959946,
"grad_norm": 0.3153751181374002,
"learning_rate": 1.3231486276540654e-05,
"loss": 0.1782,
"step": 3270
},
{
"epoch": 3.0493712156497437,
"grad_norm": 0.3079364441771498,
"learning_rate": 1.3166752977731747e-05,
"loss": 0.1805,
"step": 3275
},
{
"epoch": 3.054028877503493,
"grad_norm": 0.28739311203736595,
"learning_rate": 1.310201967892284e-05,
"loss": 0.174,
"step": 3280
},
{
"epoch": 3.0586865393572427,
"grad_norm": 0.2984939778181826,
"learning_rate": 1.303728638011393e-05,
"loss": 0.1823,
"step": 3285
},
{
"epoch": 3.063344201210992,
"grad_norm": 0.2929108597306291,
"learning_rate": 1.2972553081305026e-05,
"loss": 0.1772,
"step": 3290
},
{
"epoch": 3.0680018630647417,
"grad_norm": 0.28458008501674803,
"learning_rate": 1.2907819782496117e-05,
"loss": 0.1759,
"step": 3295
},
{
"epoch": 3.0726595249184907,
"grad_norm": 0.292581882013171,
"learning_rate": 1.2843086483687208e-05,
"loss": 0.1721,
"step": 3300
},
{
"epoch": 3.07731718677224,
"grad_norm": 0.2943365822596842,
"learning_rate": 1.2778353184878303e-05,
"loss": 0.1751,
"step": 3305
},
{
"epoch": 3.0819748486259897,
"grad_norm": 0.29377045603091445,
"learning_rate": 1.2713619886069394e-05,
"loss": 0.1723,
"step": 3310
},
{
"epoch": 3.086632510479739,
"grad_norm": 0.3851561974798118,
"learning_rate": 1.2648886587260485e-05,
"loss": 0.1774,
"step": 3315
},
{
"epoch": 3.0912901723334887,
"grad_norm": 0.27567094909537965,
"learning_rate": 1.258415328845158e-05,
"loss": 0.1725,
"step": 3320
},
{
"epoch": 3.095947834187238,
"grad_norm": 0.29697155443114787,
"learning_rate": 1.2519419989642673e-05,
"loss": 0.1817,
"step": 3325
},
{
"epoch": 3.1006054960409872,
"grad_norm": 0.6608481087066058,
"learning_rate": 1.2454686690833766e-05,
"loss": 0.1775,
"step": 3330
},
{
"epoch": 3.1052631578947367,
"grad_norm": 0.29223487514964575,
"learning_rate": 1.2389953392024859e-05,
"loss": 0.176,
"step": 3335
},
{
"epoch": 3.109920819748486,
"grad_norm": 0.29753580371777344,
"learning_rate": 1.2325220093215952e-05,
"loss": 0.1757,
"step": 3340
},
{
"epoch": 3.1145784816022357,
"grad_norm": 0.28709197068860887,
"learning_rate": 1.2260486794407043e-05,
"loss": 0.1784,
"step": 3345
},
{
"epoch": 3.119236143455985,
"grad_norm": 0.29533810100485525,
"learning_rate": 1.2195753495598136e-05,
"loss": 0.1853,
"step": 3350
},
{
"epoch": 3.1238938053097347,
"grad_norm": 0.3018378671525719,
"learning_rate": 1.213102019678923e-05,
"loss": 0.1755,
"step": 3355
},
{
"epoch": 3.1285514671634838,
"grad_norm": 0.2891240704748271,
"learning_rate": 1.2066286897980322e-05,
"loss": 0.177,
"step": 3360
},
{
"epoch": 3.1332091290172333,
"grad_norm": 0.309067605426475,
"learning_rate": 1.2001553599171414e-05,
"loss": 0.1824,
"step": 3365
},
{
"epoch": 3.1378667908709827,
"grad_norm": 0.2766369819985356,
"learning_rate": 1.1936820300362507e-05,
"loss": 0.1784,
"step": 3370
},
{
"epoch": 3.1425244527247322,
"grad_norm": 0.2979597544234227,
"learning_rate": 1.18720870015536e-05,
"loss": 0.1784,
"step": 3375
},
{
"epoch": 3.1471821145784817,
"grad_norm": 0.33228924530041026,
"learning_rate": 1.1807353702744693e-05,
"loss": 0.1714,
"step": 3380
},
{
"epoch": 3.1518397764322312,
"grad_norm": 0.28703822230981474,
"learning_rate": 1.1742620403935786e-05,
"loss": 0.1767,
"step": 3385
},
{
"epoch": 3.1564974382859803,
"grad_norm": 0.3121502606416562,
"learning_rate": 1.1677887105126878e-05,
"loss": 0.1767,
"step": 3390
},
{
"epoch": 3.16115510013973,
"grad_norm": 0.28866525138955385,
"learning_rate": 1.1613153806317971e-05,
"loss": 0.1733,
"step": 3395
},
{
"epoch": 3.1658127619934793,
"grad_norm": 0.292655197435037,
"learning_rate": 1.1548420507509063e-05,
"loss": 0.1733,
"step": 3400
},
{
"epoch": 3.1704704238472288,
"grad_norm": 0.2959791687636977,
"learning_rate": 1.1483687208700156e-05,
"loss": 0.1758,
"step": 3405
},
{
"epoch": 3.1751280857009783,
"grad_norm": 0.2982918136803687,
"learning_rate": 1.1418953909891249e-05,
"loss": 0.1836,
"step": 3410
},
{
"epoch": 3.1797857475547273,
"grad_norm": 0.30298460591386067,
"learning_rate": 1.1354220611082342e-05,
"loss": 0.1751,
"step": 3415
},
{
"epoch": 3.184443409408477,
"grad_norm": 0.31276656779659445,
"learning_rate": 1.1289487312273433e-05,
"loss": 0.1811,
"step": 3420
},
{
"epoch": 3.1891010712622263,
"grad_norm": 0.29832050835385665,
"learning_rate": 1.1224754013464526e-05,
"loss": 0.1795,
"step": 3425
},
{
"epoch": 3.193758733115976,
"grad_norm": 0.30514056117416016,
"learning_rate": 1.1160020714655619e-05,
"loss": 0.1805,
"step": 3430
},
{
"epoch": 3.1984163949697253,
"grad_norm": 0.3029193167142536,
"learning_rate": 1.1095287415846712e-05,
"loss": 0.181,
"step": 3435
},
{
"epoch": 3.203074056823475,
"grad_norm": 0.30746488264340455,
"learning_rate": 1.1030554117037805e-05,
"loss": 0.1811,
"step": 3440
},
{
"epoch": 3.207731718677224,
"grad_norm": 0.2914819757567237,
"learning_rate": 1.0965820818228898e-05,
"loss": 0.1726,
"step": 3445
},
{
"epoch": 3.2123893805309733,
"grad_norm": 0.2886147432095147,
"learning_rate": 1.0901087519419991e-05,
"loss": 0.1775,
"step": 3450
},
{
"epoch": 3.217047042384723,
"grad_norm": 0.29564777298103045,
"learning_rate": 1.0836354220611084e-05,
"loss": 0.1789,
"step": 3455
},
{
"epoch": 3.2217047042384723,
"grad_norm": 0.2892508127991348,
"learning_rate": 1.0771620921802175e-05,
"loss": 0.1778,
"step": 3460
},
{
"epoch": 3.226362366092222,
"grad_norm": 0.28774906064692896,
"learning_rate": 1.0706887622993268e-05,
"loss": 0.176,
"step": 3465
},
{
"epoch": 3.2310200279459713,
"grad_norm": 0.30800900764013556,
"learning_rate": 1.0642154324184361e-05,
"loss": 0.1783,
"step": 3470
},
{
"epoch": 3.2356776897997204,
"grad_norm": 0.2793667780450472,
"learning_rate": 1.0577421025375453e-05,
"loss": 0.1776,
"step": 3475
},
{
"epoch": 3.24033535165347,
"grad_norm": 0.2960876247684058,
"learning_rate": 1.0512687726566546e-05,
"loss": 0.1765,
"step": 3480
},
{
"epoch": 3.2449930135072194,
"grad_norm": 0.32930610906590374,
"learning_rate": 1.0447954427757638e-05,
"loss": 0.1816,
"step": 3485
},
{
"epoch": 3.249650675360969,
"grad_norm": 0.3159028312568502,
"learning_rate": 1.0383221128948731e-05,
"loss": 0.1814,
"step": 3490
},
{
"epoch": 3.2543083372147183,
"grad_norm": 0.3037903738896479,
"learning_rate": 1.0318487830139824e-05,
"loss": 0.1873,
"step": 3495
},
{
"epoch": 3.258965999068468,
"grad_norm": 0.30690210747220553,
"learning_rate": 1.0253754531330917e-05,
"loss": 0.1792,
"step": 3500
},
{
"epoch": 3.263623660922217,
"grad_norm": 0.2971216838179025,
"learning_rate": 1.018902123252201e-05,
"loss": 0.182,
"step": 3505
},
{
"epoch": 3.2682813227759664,
"grad_norm": 0.3047994804606214,
"learning_rate": 1.0124287933713103e-05,
"loss": 0.1802,
"step": 3510
},
{
"epoch": 3.272938984629716,
"grad_norm": 0.29190460822979236,
"learning_rate": 1.0059554634904195e-05,
"loss": 0.1788,
"step": 3515
},
{
"epoch": 3.2775966464834654,
"grad_norm": 0.3046401879655743,
"learning_rate": 9.994821336095288e-06,
"loss": 0.1788,
"step": 3520
},
{
"epoch": 3.282254308337215,
"grad_norm": 0.31991453702020917,
"learning_rate": 9.93008803728638e-06,
"loss": 0.1816,
"step": 3525
},
{
"epoch": 3.286911970190964,
"grad_norm": 0.28860232259659196,
"learning_rate": 9.865354738477474e-06,
"loss": 0.1734,
"step": 3530
},
{
"epoch": 3.2915696320447134,
"grad_norm": 0.28646435016895677,
"learning_rate": 9.800621439668565e-06,
"loss": 0.1729,
"step": 3535
},
{
"epoch": 3.296227293898463,
"grad_norm": 0.2878206575379541,
"learning_rate": 9.735888140859658e-06,
"loss": 0.1802,
"step": 3540
},
{
"epoch": 3.3008849557522124,
"grad_norm": 0.278596793180263,
"learning_rate": 9.671154842050751e-06,
"loss": 0.1783,
"step": 3545
},
{
"epoch": 3.305542617605962,
"grad_norm": 0.29157821066173406,
"learning_rate": 9.606421543241844e-06,
"loss": 0.1753,
"step": 3550
},
{
"epoch": 3.3102002794597114,
"grad_norm": 0.28868610667936917,
"learning_rate": 9.541688244432937e-06,
"loss": 0.1734,
"step": 3555
},
{
"epoch": 3.3148579413134605,
"grad_norm": 0.2829748445043861,
"learning_rate": 9.47695494562403e-06,
"loss": 0.1791,
"step": 3560
},
{
"epoch": 3.31951560316721,
"grad_norm": 0.31936811621546346,
"learning_rate": 9.412221646815123e-06,
"loss": 0.1801,
"step": 3565
},
{
"epoch": 3.3241732650209594,
"grad_norm": 0.28502956960820763,
"learning_rate": 9.347488348006214e-06,
"loss": 0.1771,
"step": 3570
},
{
"epoch": 3.328830926874709,
"grad_norm": 0.2990286469354999,
"learning_rate": 9.282755049197307e-06,
"loss": 0.1783,
"step": 3575
},
{
"epoch": 3.3334885887284584,
"grad_norm": 0.3076827669029228,
"learning_rate": 9.2180217503884e-06,
"loss": 0.1762,
"step": 3580
},
{
"epoch": 3.3381462505822075,
"grad_norm": 0.30032433256579627,
"learning_rate": 9.153288451579493e-06,
"loss": 0.1812,
"step": 3585
},
{
"epoch": 3.342803912435957,
"grad_norm": 0.2763729371529282,
"learning_rate": 9.088555152770584e-06,
"loss": 0.1756,
"step": 3590
},
{
"epoch": 3.3474615742897065,
"grad_norm": 0.28665241347886994,
"learning_rate": 9.023821853961677e-06,
"loss": 0.1781,
"step": 3595
},
{
"epoch": 3.352119236143456,
"grad_norm": 0.2852578172845054,
"learning_rate": 8.95908855515277e-06,
"loss": 0.1766,
"step": 3600
},
{
"epoch": 3.3567768979972055,
"grad_norm": 0.2999218538137478,
"learning_rate": 8.894355256343863e-06,
"loss": 0.1742,
"step": 3605
},
{
"epoch": 3.361434559850955,
"grad_norm": 0.2778982809107516,
"learning_rate": 8.829621957534956e-06,
"loss": 0.1766,
"step": 3610
},
{
"epoch": 3.3660922217047045,
"grad_norm": 0.2870631712767666,
"learning_rate": 8.76488865872605e-06,
"loss": 0.1735,
"step": 3615
},
{
"epoch": 3.3707498835584535,
"grad_norm": 0.28630778046736993,
"learning_rate": 8.700155359917142e-06,
"loss": 0.1758,
"step": 3620
},
{
"epoch": 3.375407545412203,
"grad_norm": 0.28571554107591934,
"learning_rate": 8.635422061108235e-06,
"loss": 0.1802,
"step": 3625
},
{
"epoch": 3.3800652072659525,
"grad_norm": 0.2994504977989836,
"learning_rate": 8.570688762299327e-06,
"loss": 0.1808,
"step": 3630
},
{
"epoch": 3.384722869119702,
"grad_norm": 0.28410097054733224,
"learning_rate": 8.50595546349042e-06,
"loss": 0.1808,
"step": 3635
},
{
"epoch": 3.3893805309734515,
"grad_norm": 0.28236954251213214,
"learning_rate": 8.441222164681513e-06,
"loss": 0.1741,
"step": 3640
},
{
"epoch": 3.3940381928272005,
"grad_norm": 0.3084957965343309,
"learning_rate": 8.376488865872606e-06,
"loss": 0.1907,
"step": 3645
},
{
"epoch": 3.39869585468095,
"grad_norm": 0.29554611127145863,
"learning_rate": 8.311755567063697e-06,
"loss": 0.1788,
"step": 3650
},
{
"epoch": 3.4033535165346995,
"grad_norm": 0.3144424173115396,
"learning_rate": 8.24702226825479e-06,
"loss": 0.1809,
"step": 3655
},
{
"epoch": 3.408011178388449,
"grad_norm": 0.2918858454691064,
"learning_rate": 8.182288969445883e-06,
"loss": 0.174,
"step": 3660
},
{
"epoch": 3.4126688402421985,
"grad_norm": 0.2892152457631711,
"learning_rate": 8.117555670636976e-06,
"loss": 0.1763,
"step": 3665
},
{
"epoch": 3.417326502095948,
"grad_norm": 0.28373511140587954,
"learning_rate": 8.052822371828069e-06,
"loss": 0.1782,
"step": 3670
},
{
"epoch": 3.421984163949697,
"grad_norm": 0.272525018346227,
"learning_rate": 7.988089073019162e-06,
"loss": 0.1728,
"step": 3675
},
{
"epoch": 3.4266418258034466,
"grad_norm": 0.27049351592069754,
"learning_rate": 7.923355774210255e-06,
"loss": 0.1748,
"step": 3680
},
{
"epoch": 3.431299487657196,
"grad_norm": 0.2908050514006356,
"learning_rate": 7.858622475401346e-06,
"loss": 0.1816,
"step": 3685
},
{
"epoch": 3.4359571495109456,
"grad_norm": 0.2771715804654571,
"learning_rate": 7.793889176592439e-06,
"loss": 0.1775,
"step": 3690
},
{
"epoch": 3.440614811364695,
"grad_norm": 0.307408789296849,
"learning_rate": 7.729155877783532e-06,
"loss": 0.1803,
"step": 3695
},
{
"epoch": 3.445272473218444,
"grad_norm": 0.2896861472425251,
"learning_rate": 7.664422578974625e-06,
"loss": 0.1751,
"step": 3700
},
{
"epoch": 3.4499301350721936,
"grad_norm": 0.279195615233317,
"learning_rate": 7.599689280165717e-06,
"loss": 0.1798,
"step": 3705
},
{
"epoch": 3.454587796925943,
"grad_norm": 0.29432038056377485,
"learning_rate": 7.53495598135681e-06,
"loss": 0.1828,
"step": 3710
},
{
"epoch": 3.4592454587796926,
"grad_norm": 0.29060489107041276,
"learning_rate": 7.470222682547903e-06,
"loss": 0.1745,
"step": 3715
},
{
"epoch": 3.463903120633442,
"grad_norm": 0.28785175023856524,
"learning_rate": 7.405489383738996e-06,
"loss": 0.1774,
"step": 3720
},
{
"epoch": 3.4685607824871916,
"grad_norm": 0.2824942037705507,
"learning_rate": 7.340756084930088e-06,
"loss": 0.1798,
"step": 3725
},
{
"epoch": 3.473218444340941,
"grad_norm": 0.2885091784401793,
"learning_rate": 7.276022786121181e-06,
"loss": 0.1791,
"step": 3730
},
{
"epoch": 3.47787610619469,
"grad_norm": 0.2838658738954116,
"learning_rate": 7.211289487312274e-06,
"loss": 0.1746,
"step": 3735
},
{
"epoch": 3.4825337680484396,
"grad_norm": 0.28887842909118483,
"learning_rate": 7.146556188503367e-06,
"loss": 0.175,
"step": 3740
},
{
"epoch": 3.487191429902189,
"grad_norm": 0.2843809341741585,
"learning_rate": 7.081822889694459e-06,
"loss": 0.177,
"step": 3745
},
{
"epoch": 3.4918490917559386,
"grad_norm": 0.28621203602989626,
"learning_rate": 7.017089590885552e-06,
"loss": 0.178,
"step": 3750
},
{
"epoch": 3.496506753609688,
"grad_norm": 0.28912146796078997,
"learning_rate": 6.952356292076645e-06,
"loss": 0.1786,
"step": 3755
},
{
"epoch": 3.501164415463437,
"grad_norm": 0.28458889587226993,
"learning_rate": 6.887622993267737e-06,
"loss": 0.1755,
"step": 3760
},
{
"epoch": 3.5058220773171866,
"grad_norm": 0.29550375520040006,
"learning_rate": 6.82288969445883e-06,
"loss": 0.173,
"step": 3765
},
{
"epoch": 3.510479739170936,
"grad_norm": 0.31286759510416146,
"learning_rate": 6.758156395649923e-06,
"loss": 0.174,
"step": 3770
},
{
"epoch": 3.5151374010246856,
"grad_norm": 0.28829238017584236,
"learning_rate": 6.693423096841016e-06,
"loss": 0.1718,
"step": 3775
},
{
"epoch": 3.519795062878435,
"grad_norm": 0.28945905000643063,
"learning_rate": 6.628689798032108e-06,
"loss": 0.1717,
"step": 3780
},
{
"epoch": 3.5244527247321846,
"grad_norm": 0.29706213456606245,
"learning_rate": 6.563956499223201e-06,
"loss": 0.1767,
"step": 3785
},
{
"epoch": 3.529110386585934,
"grad_norm": 0.2697220597100272,
"learning_rate": 6.499223200414294e-06,
"loss": 0.1759,
"step": 3790
},
{
"epoch": 3.533768048439683,
"grad_norm": 0.28625641383280975,
"learning_rate": 6.434489901605387e-06,
"loss": 0.1848,
"step": 3795
},
{
"epoch": 3.5384257102934327,
"grad_norm": 0.27959609788340817,
"learning_rate": 6.369756602796478e-06,
"loss": 0.1759,
"step": 3800
},
{
"epoch": 3.543083372147182,
"grad_norm": 0.28228154116471305,
"learning_rate": 6.305023303987571e-06,
"loss": 0.1769,
"step": 3805
},
{
"epoch": 3.5477410340009317,
"grad_norm": 0.26985107405345043,
"learning_rate": 6.240290005178664e-06,
"loss": 0.1778,
"step": 3810
},
{
"epoch": 3.5523986958546807,
"grad_norm": 0.27337149644653264,
"learning_rate": 6.175556706369757e-06,
"loss": 0.1772,
"step": 3815
},
{
"epoch": 3.55705635770843,
"grad_norm": 0.27104309189982073,
"learning_rate": 6.11082340756085e-06,
"loss": 0.1754,
"step": 3820
},
{
"epoch": 3.5617140195621797,
"grad_norm": 0.30932575066266116,
"learning_rate": 6.046090108751942e-06,
"loss": 0.1776,
"step": 3825
},
{
"epoch": 3.566371681415929,
"grad_norm": 0.29740580773652014,
"learning_rate": 5.981356809943035e-06,
"loss": 0.176,
"step": 3830
},
{
"epoch": 3.5710293432696787,
"grad_norm": 0.276523784239008,
"learning_rate": 5.916623511134127e-06,
"loss": 0.1799,
"step": 3835
},
{
"epoch": 3.575687005123428,
"grad_norm": 0.29013154798571805,
"learning_rate": 5.85189021232522e-06,
"loss": 0.1768,
"step": 3840
},
{
"epoch": 3.5803446669771777,
"grad_norm": 0.29252747153466957,
"learning_rate": 5.787156913516313e-06,
"loss": 0.1738,
"step": 3845
},
{
"epoch": 3.5850023288309267,
"grad_norm": 0.28051730805375724,
"learning_rate": 5.7224236147074054e-06,
"loss": 0.1756,
"step": 3850
},
{
"epoch": 3.5896599906846762,
"grad_norm": 0.29837466182008804,
"learning_rate": 5.6576903158984984e-06,
"loss": 0.1806,
"step": 3855
},
{
"epoch": 3.5943176525384257,
"grad_norm": 0.2902254049349514,
"learning_rate": 5.5929570170895906e-06,
"loss": 0.1758,
"step": 3860
},
{
"epoch": 3.598975314392175,
"grad_norm": 0.29690468932528485,
"learning_rate": 5.5282237182806836e-06,
"loss": 0.1812,
"step": 3865
},
{
"epoch": 3.6036329762459243,
"grad_norm": 0.41903251792225776,
"learning_rate": 5.4634904194717765e-06,
"loss": 0.1761,
"step": 3870
},
{
"epoch": 3.6082906380996738,
"grad_norm": 0.2757754067100353,
"learning_rate": 5.3987571206628695e-06,
"loss": 0.1759,
"step": 3875
},
{
"epoch": 3.6129482999534233,
"grad_norm": 0.28245805312769273,
"learning_rate": 5.334023821853962e-06,
"loss": 0.1736,
"step": 3880
},
{
"epoch": 3.6176059618071728,
"grad_norm": 0.2664074642960957,
"learning_rate": 5.269290523045055e-06,
"loss": 0.1732,
"step": 3885
},
{
"epoch": 3.6222636236609222,
"grad_norm": 0.2736243477212168,
"learning_rate": 5.204557224236147e-06,
"loss": 0.1743,
"step": 3890
},
{
"epoch": 3.6269212855146717,
"grad_norm": 0.2775685835752688,
"learning_rate": 5.139823925427241e-06,
"loss": 0.1754,
"step": 3895
},
{
"epoch": 3.6315789473684212,
"grad_norm": 0.26500655531124734,
"learning_rate": 5.075090626618333e-06,
"loss": 0.1719,
"step": 3900
},
{
"epoch": 3.6362366092221707,
"grad_norm": 0.27561475493184445,
"learning_rate": 5.010357327809426e-06,
"loss": 0.1784,
"step": 3905
},
{
"epoch": 3.64089427107592,
"grad_norm": 0.28363025571766304,
"learning_rate": 4.945624029000518e-06,
"loss": 0.1711,
"step": 3910
},
{
"epoch": 3.6455519329296693,
"grad_norm": 0.27578950261131024,
"learning_rate": 4.880890730191611e-06,
"loss": 0.1745,
"step": 3915
},
{
"epoch": 3.6502095947834188,
"grad_norm": 0.2689287311474901,
"learning_rate": 4.816157431382704e-06,
"loss": 0.1673,
"step": 3920
},
{
"epoch": 3.6548672566371683,
"grad_norm": 0.2768899738432259,
"learning_rate": 4.751424132573796e-06,
"loss": 0.1668,
"step": 3925
},
{
"epoch": 3.6595249184909173,
"grad_norm": 0.2762059518655408,
"learning_rate": 4.686690833764889e-06,
"loss": 0.1746,
"step": 3930
},
{
"epoch": 3.664182580344667,
"grad_norm": 0.2854409744347062,
"learning_rate": 4.621957534955981e-06,
"loss": 0.1771,
"step": 3935
},
{
"epoch": 3.6688402421984163,
"grad_norm": 0.27071826247278635,
"learning_rate": 4.557224236147074e-06,
"loss": 0.1734,
"step": 3940
},
{
"epoch": 3.673497904052166,
"grad_norm": 0.31409279940314927,
"learning_rate": 4.492490937338166e-06,
"loss": 0.173,
"step": 3945
},
{
"epoch": 3.6781555659059153,
"grad_norm": 0.2712537638578198,
"learning_rate": 4.42775763852926e-06,
"loss": 0.1729,
"step": 3950
},
{
"epoch": 3.682813227759665,
"grad_norm": 0.2659242739993353,
"learning_rate": 4.363024339720352e-06,
"loss": 0.1781,
"step": 3955
},
{
"epoch": 3.6874708896134143,
"grad_norm": 0.2878431083722301,
"learning_rate": 4.298291040911445e-06,
"loss": 0.1755,
"step": 3960
},
{
"epoch": 3.6921285514671633,
"grad_norm": 0.27004564279309073,
"learning_rate": 4.233557742102537e-06,
"loss": 0.1722,
"step": 3965
},
{
"epoch": 3.696786213320913,
"grad_norm": 0.2726539565276091,
"learning_rate": 4.16882444329363e-06,
"loss": 0.1778,
"step": 3970
},
{
"epoch": 3.7014438751746623,
"grad_norm": 0.26268681246368897,
"learning_rate": 4.104091144484723e-06,
"loss": 0.1695,
"step": 3975
},
{
"epoch": 3.706101537028412,
"grad_norm": 0.2760735546995598,
"learning_rate": 4.039357845675816e-06,
"loss": 0.1726,
"step": 3980
},
{
"epoch": 3.710759198882161,
"grad_norm": 0.28424803977449115,
"learning_rate": 3.9746245468669085e-06,
"loss": 0.1799,
"step": 3985
},
{
"epoch": 3.7154168607359104,
"grad_norm": 0.2941948305095177,
"learning_rate": 3.9098912480580015e-06,
"loss": 0.1706,
"step": 3990
},
{
"epoch": 3.72007452258966,
"grad_norm": 0.2804830682382998,
"learning_rate": 3.845157949249094e-06,
"loss": 0.175,
"step": 3995
},
{
"epoch": 3.7247321844434094,
"grad_norm": 0.25894069673660725,
"learning_rate": 3.780424650440187e-06,
"loss": 0.179,
"step": 4000
},
{
"epoch": 3.729389846297159,
"grad_norm": 0.2770495343014349,
"learning_rate": 3.715691351631279e-06,
"loss": 0.1718,
"step": 4005
},
{
"epoch": 3.7340475081509084,
"grad_norm": 0.27245259036566644,
"learning_rate": 3.650958052822372e-06,
"loss": 0.1745,
"step": 4010
},
{
"epoch": 3.738705170004658,
"grad_norm": 0.3178518519661775,
"learning_rate": 3.5862247540134647e-06,
"loss": 0.1718,
"step": 4015
},
{
"epoch": 3.7433628318584073,
"grad_norm": 0.2725475491559217,
"learning_rate": 3.521491455204557e-06,
"loss": 0.1787,
"step": 4020
},
{
"epoch": 3.7480204937121564,
"grad_norm": 0.2815887851859046,
"learning_rate": 3.4567581563956503e-06,
"loss": 0.1804,
"step": 4025
},
{
"epoch": 3.752678155565906,
"grad_norm": 0.2671335476290762,
"learning_rate": 3.3920248575867424e-06,
"loss": 0.1725,
"step": 4030
},
{
"epoch": 3.7573358174196554,
"grad_norm": 0.2884011738724322,
"learning_rate": 3.3272915587778354e-06,
"loss": 0.172,
"step": 4035
},
{
"epoch": 3.761993479273405,
"grad_norm": 0.272777124868328,
"learning_rate": 3.262558259968928e-06,
"loss": 0.1719,
"step": 4040
},
{
"epoch": 3.766651141127154,
"grad_norm": 0.27639814202614327,
"learning_rate": 3.197824961160021e-06,
"loss": 0.1755,
"step": 4045
},
{
"epoch": 3.7713088029809034,
"grad_norm": 0.26949037354351396,
"learning_rate": 3.1330916623511135e-06,
"loss": 0.1759,
"step": 4050
},
{
"epoch": 3.775966464834653,
"grad_norm": 0.2722470274160135,
"learning_rate": 3.0683583635422065e-06,
"loss": 0.1817,
"step": 4055
},
{
"epoch": 3.7806241266884024,
"grad_norm": 0.2602766776236453,
"learning_rate": 3.003625064733299e-06,
"loss": 0.1715,
"step": 4060
},
{
"epoch": 3.785281788542152,
"grad_norm": 0.28711780692476924,
"learning_rate": 2.9388917659243916e-06,
"loss": 0.1711,
"step": 4065
},
{
"epoch": 3.7899394503959014,
"grad_norm": 0.27051801081640436,
"learning_rate": 2.874158467115484e-06,
"loss": 0.1728,
"step": 4070
},
{
"epoch": 3.794597112249651,
"grad_norm": 0.2828893708081034,
"learning_rate": 2.8094251683065768e-06,
"loss": 0.171,
"step": 4075
},
{
"epoch": 3.7992547741034,
"grad_norm": 0.2723968547135529,
"learning_rate": 2.7446918694976698e-06,
"loss": 0.1723,
"step": 4080
},
{
"epoch": 3.8039124359571495,
"grad_norm": 0.26655651659598223,
"learning_rate": 2.6799585706887623e-06,
"loss": 0.1729,
"step": 4085
},
{
"epoch": 3.808570097810899,
"grad_norm": 0.26770981285481243,
"learning_rate": 2.615225271879855e-06,
"loss": 0.1749,
"step": 4090
},
{
"epoch": 3.8132277596646484,
"grad_norm": 0.2758365036666948,
"learning_rate": 2.550491973070948e-06,
"loss": 0.1716,
"step": 4095
},
{
"epoch": 3.8178854215183975,
"grad_norm": 0.2610019392205348,
"learning_rate": 2.4857586742620404e-06,
"loss": 0.1718,
"step": 4100
},
{
"epoch": 3.822543083372147,
"grad_norm": 0.2685711043716591,
"learning_rate": 2.421025375453133e-06,
"loss": 0.1735,
"step": 4105
},
{
"epoch": 3.8272007452258965,
"grad_norm": 0.27773939842189077,
"learning_rate": 2.356292076644226e-06,
"loss": 0.1729,
"step": 4110
},
{
"epoch": 3.831858407079646,
"grad_norm": 0.27818722043603183,
"learning_rate": 2.2915587778353186e-06,
"loss": 0.18,
"step": 4115
},
{
"epoch": 3.8365160689333955,
"grad_norm": 0.29022869128773615,
"learning_rate": 2.2268254790264115e-06,
"loss": 0.1701,
"step": 4120
},
{
"epoch": 3.841173730787145,
"grad_norm": 0.26892060504547427,
"learning_rate": 2.162092180217504e-06,
"loss": 0.1748,
"step": 4125
},
{
"epoch": 3.8458313926408945,
"grad_norm": 0.2742194085553554,
"learning_rate": 2.0973588814085967e-06,
"loss": 0.1736,
"step": 4130
},
{
"epoch": 3.850489054494644,
"grad_norm": 0.280664869183797,
"learning_rate": 2.0326255825996897e-06,
"loss": 0.1742,
"step": 4135
},
{
"epoch": 3.855146716348393,
"grad_norm": 0.2663084540169044,
"learning_rate": 1.9678922837907822e-06,
"loss": 0.1755,
"step": 4140
},
{
"epoch": 3.8598043782021425,
"grad_norm": 0.2683512464822063,
"learning_rate": 1.903158984981875e-06,
"loss": 0.1765,
"step": 4145
},
{
"epoch": 3.864462040055892,
"grad_norm": 0.28077450679398896,
"learning_rate": 1.8384256861729676e-06,
"loss": 0.1713,
"step": 4150
},
{
"epoch": 3.8691197019096415,
"grad_norm": 0.2629212409957542,
"learning_rate": 1.77369238736406e-06,
"loss": 0.1746,
"step": 4155
},
{
"epoch": 3.8737773637633905,
"grad_norm": 0.2862486384413756,
"learning_rate": 1.7089590885551527e-06,
"loss": 0.1816,
"step": 4160
},
{
"epoch": 3.87843502561714,
"grad_norm": 0.2767392538386716,
"learning_rate": 1.6442257897462455e-06,
"loss": 0.1754,
"step": 4165
},
{
"epoch": 3.8830926874708895,
"grad_norm": 0.26967392829121106,
"learning_rate": 1.579492490937338e-06,
"loss": 0.1699,
"step": 4170
},
{
"epoch": 3.887750349324639,
"grad_norm": 0.26837525839027015,
"learning_rate": 1.514759192128431e-06,
"loss": 0.177,
"step": 4175
},
{
"epoch": 3.8924080111783885,
"grad_norm": 0.2627808696747621,
"learning_rate": 1.4500258933195236e-06,
"loss": 0.1719,
"step": 4180
},
{
"epoch": 3.897065673032138,
"grad_norm": 0.2639957998761549,
"learning_rate": 1.3852925945106164e-06,
"loss": 0.1708,
"step": 4185
},
{
"epoch": 3.9017233348858875,
"grad_norm": 0.27125749404584937,
"learning_rate": 1.320559295701709e-06,
"loss": 0.174,
"step": 4190
},
{
"epoch": 3.9063809967396366,
"grad_norm": 0.2769883473331165,
"learning_rate": 1.2558259968928017e-06,
"loss": 0.1719,
"step": 4195
},
{
"epoch": 3.911038658593386,
"grad_norm": 0.8395951166898823,
"learning_rate": 1.1910926980838945e-06,
"loss": 0.1696,
"step": 4200
},
{
"epoch": 3.9156963204471356,
"grad_norm": 0.2571958789773536,
"learning_rate": 1.126359399274987e-06,
"loss": 0.173,
"step": 4205
},
{
"epoch": 3.920353982300885,
"grad_norm": 0.3193999846164976,
"learning_rate": 1.0616261004660798e-06,
"loss": 0.1719,
"step": 4210
},
{
"epoch": 3.925011644154634,
"grad_norm": 0.27512694081708594,
"learning_rate": 9.968928016571726e-07,
"loss": 0.1703,
"step": 4215
},
{
"epoch": 3.9296693060083836,
"grad_norm": 0.27073480012745654,
"learning_rate": 9.321595028482653e-07,
"loss": 0.1701,
"step": 4220
},
{
"epoch": 3.934326967862133,
"grad_norm": 0.26389705940762015,
"learning_rate": 8.674262040393578e-07,
"loss": 0.1708,
"step": 4225
},
{
"epoch": 3.9389846297158826,
"grad_norm": 0.2575701459397594,
"learning_rate": 8.026929052304505e-07,
"loss": 0.1747,
"step": 4230
},
{
"epoch": 3.943642291569632,
"grad_norm": 0.27561192360509723,
"learning_rate": 7.379596064215433e-07,
"loss": 0.1705,
"step": 4235
},
{
"epoch": 3.9482999534233816,
"grad_norm": 0.256804333426344,
"learning_rate": 6.73226307612636e-07,
"loss": 0.1719,
"step": 4240
},
{
"epoch": 3.952957615277131,
"grad_norm": 0.25955876150091883,
"learning_rate": 6.084930088037286e-07,
"loss": 0.1714,
"step": 4245
},
{
"epoch": 3.9576152771308806,
"grad_norm": 0.26478526146557524,
"learning_rate": 5.437597099948214e-07,
"loss": 0.1713,
"step": 4250
},
{
"epoch": 3.9622729389846296,
"grad_norm": 0.27991810492833324,
"learning_rate": 4.790264111859141e-07,
"loss": 0.176,
"step": 4255
},
{
"epoch": 3.966930600838379,
"grad_norm": 0.2665648771975109,
"learning_rate": 4.1429311237700674e-07,
"loss": 0.1718,
"step": 4260
},
{
"epoch": 3.9715882626921286,
"grad_norm": 0.2644112154207203,
"learning_rate": 3.495598135680994e-07,
"loss": 0.1756,
"step": 4265
},
{
"epoch": 3.976245924545878,
"grad_norm": 0.26058860902480596,
"learning_rate": 2.8482651475919213e-07,
"loss": 0.1725,
"step": 4270
},
{
"epoch": 3.980903586399627,
"grad_norm": 0.27155646802982203,
"learning_rate": 2.200932159502848e-07,
"loss": 0.1751,
"step": 4275
},
{
"epoch": 3.9855612482533767,
"grad_norm": 0.279201457240123,
"learning_rate": 1.5535991714137753e-07,
"loss": 0.1732,
"step": 4280
},
{
"epoch": 3.990218910107126,
"grad_norm": 0.25945923118786696,
"learning_rate": 9.062661833247022e-08,
"loss": 0.1738,
"step": 4285
},
{
"epoch": 3.9948765719608756,
"grad_norm": 0.2669644915648517,
"learning_rate": 2.589331952356292e-08,
"loss": 0.1721,
"step": 4290
},
{
"epoch": 3.9967396367023755,
"step": 4292,
"total_flos": 3.675734279595229e+18,
"train_loss": 0.31530814574126986,
"train_runtime": 166213.7776,
"train_samples_per_second": 0.413,
"train_steps_per_second": 0.026
}
],
"logging_steps": 5,
"max_steps": 4292,
"num_input_tokens_seen": 0,
"num_train_epochs": 4,
"save_steps": 500,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 3.675734279595229e+18,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}