{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.22994080894442617,
  "eval_steps": 500,
  "global_step": 2797,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008220979940808945,
      "grad_norm": 0.7510645985603333,
      "learning_rate": 1.1061797752808989e-06,
      "loss": 0.5669,
      "step": 10
    },
    {
      "epoch": 0.001644195988161789,
      "grad_norm": 0.6885573267936707,
      "learning_rate": 2.2241573033707866e-06,
      "loss": 0.5563,
      "step": 20
    },
    {
      "epoch": 0.0024662939822426835,
      "grad_norm": 0.6587523221969604,
      "learning_rate": 3.342134831460675e-06,
      "loss": 0.556,
      "step": 30
    },
    {
      "epoch": 0.003288391976323578,
      "grad_norm": 0.6111008524894714,
      "learning_rate": 4.4601123595505624e-06,
      "loss": 0.5686,
      "step": 40
    },
    {
      "epoch": 0.004110489970404472,
      "grad_norm": 0.6500401496887207,
      "learning_rate": 5.5780898876404495e-06,
      "loss": 0.5576,
      "step": 50
    },
    {
      "epoch": 0.004932587964485367,
      "grad_norm": 0.6981553435325623,
      "learning_rate": 6.6960674157303374e-06,
      "loss": 0.5667,
      "step": 60
    },
    {
      "epoch": 0.005754685958566261,
      "grad_norm": 0.5952242612838745,
      "learning_rate": 7.814044943820226e-06,
      "loss": 0.5455,
      "step": 70
    },
    {
      "epoch": 0.006576783952647156,
      "grad_norm": 0.5906009674072266,
      "learning_rate": 8.932022471910112e-06,
      "loss": 0.5634,
      "step": 80
    },
    {
      "epoch": 0.00739888194672805,
      "grad_norm": 0.5970085859298706,
      "learning_rate": 1.005e-05,
      "loss": 0.5666,
      "step": 90
    },
    {
      "epoch": 0.008220979940808944,
      "grad_norm": 0.598000168800354,
      "learning_rate": 1.1167977528089888e-05,
      "loss": 0.5561,
      "step": 100
    },
    {
      "epoch": 0.00904307793488984,
      "grad_norm": 0.6920904517173767,
      "learning_rate": 1.2285955056179776e-05,
      "loss": 0.5571,
      "step": 110
    },
    {
      "epoch": 0.009865175928970734,
      "grad_norm": 0.5882319211959839,
      "learning_rate": 1.3403932584269664e-05,
      "loss": 0.5831,
      "step": 120
    },
    {
      "epoch": 0.010687273923051627,
      "grad_norm": 0.5964992642402649,
      "learning_rate": 1.4521910112359552e-05,
      "loss": 0.5511,
      "step": 130
    },
    {
      "epoch": 0.011509371917132522,
      "grad_norm": 0.6679021120071411,
      "learning_rate": 1.563988764044944e-05,
      "loss": 0.5583,
      "step": 140
    },
    {
      "epoch": 0.012331469911213417,
      "grad_norm": 0.5619528889656067,
      "learning_rate": 1.6757865168539328e-05,
      "loss": 0.5377,
      "step": 150
    },
    {
      "epoch": 0.013153567905294311,
      "grad_norm": 0.5494176149368286,
      "learning_rate": 1.7875842696629214e-05,
      "loss": 0.5539,
      "step": 160
    },
    {
      "epoch": 0.013975665899375206,
      "grad_norm": 0.5273856520652771,
      "learning_rate": 1.89938202247191e-05,
      "loss": 0.5493,
      "step": 170
    },
    {
      "epoch": 0.0147977638934561,
      "grad_norm": 0.5497198104858398,
      "learning_rate": 2e-05,
      "loss": 0.5684,
      "step": 180
    },
    {
      "epoch": 0.015619861887536994,
      "grad_norm": 0.5500702261924744,
      "learning_rate": 2e-05,
      "loss": 0.5553,
      "step": 190
    },
    {
      "epoch": 0.01644195988161789,
      "grad_norm": 0.5789015889167786,
      "learning_rate": 2e-05,
      "loss": 0.5661,
      "step": 200
    },
    {
      "epoch": 0.017264057875698784,
      "grad_norm": 0.44678372144699097,
      "learning_rate": 2e-05,
      "loss": 0.5739,
      "step": 210
    },
    {
      "epoch": 0.01808615586977968,
      "grad_norm": 0.4713243246078491,
      "learning_rate": 2e-05,
      "loss": 0.546,
      "step": 220
    },
    {
      "epoch": 0.018908253863860573,
      "grad_norm": 0.5556458234786987,
      "learning_rate": 2e-05,
      "loss": 0.5751,
      "step": 230
    },
    {
      "epoch": 0.019730351857941468,
      "grad_norm": 0.5001806616783142,
      "learning_rate": 2e-05,
      "loss": 0.5627,
      "step": 240
    },
    {
      "epoch": 0.020552449852022363,
      "grad_norm": 0.4647273123264313,
      "learning_rate": 2e-05,
      "loss": 0.5765,
      "step": 250
    },
    {
      "epoch": 0.021374547846103254,
      "grad_norm": 0.4200420379638672,
      "learning_rate": 2e-05,
      "loss": 0.5885,
      "step": 260
    },
    {
      "epoch": 0.02219664584018415,
      "grad_norm": 0.5070334076881409,
      "learning_rate": 2e-05,
      "loss": 0.5428,
      "step": 270
    },
    {
      "epoch": 0.023018743834265044,
      "grad_norm": 0.41507670283317566,
      "learning_rate": 2e-05,
      "loss": 0.5836,
      "step": 280
    },
    {
      "epoch": 0.02384084182834594,
      "grad_norm": 0.4536267817020416,
      "learning_rate": 2e-05,
      "loss": 0.5687,
      "step": 290
    },
    {
      "epoch": 0.024662939822426833,
      "grad_norm": 0.4120713472366333,
      "learning_rate": 2e-05,
      "loss": 0.5512,
      "step": 300
    },
    {
      "epoch": 0.025485037816507728,
      "grad_norm": 0.44302648305892944,
      "learning_rate": 2e-05,
      "loss": 0.553,
      "step": 310
    },
    {
      "epoch": 0.026307135810588623,
      "grad_norm": 0.41790035367012024,
      "learning_rate": 2e-05,
      "loss": 0.5793,
      "step": 320
    },
    {
      "epoch": 0.027129233804669518,
      "grad_norm": 0.4572747051715851,
      "learning_rate": 2e-05,
      "loss": 0.5838,
      "step": 330
    },
    {
      "epoch": 0.027951331798750412,
      "grad_norm": 0.4104056656360626,
      "learning_rate": 2e-05,
      "loss": 0.5716,
      "step": 340
    },
    {
      "epoch": 0.028773429792831307,
      "grad_norm": 0.41988691687583923,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 350
    },
    {
      "epoch": 0.0295955277869122,
      "grad_norm": 0.3765335977077484,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 360
    },
    {
      "epoch": 0.030417625780993093,
      "grad_norm": 0.42008137702941895,
      "learning_rate": 2e-05,
      "loss": 0.544,
      "step": 370
    },
    {
      "epoch": 0.031239723775073988,
      "grad_norm": 0.3445565402507782,
      "learning_rate": 2e-05,
      "loss": 0.5757,
      "step": 380
    },
    {
      "epoch": 0.03206182176915488,
      "grad_norm": 0.4134063124656677,
      "learning_rate": 2e-05,
      "loss": 0.5484,
      "step": 390
    },
    {
      "epoch": 0.03288391976323578,
      "grad_norm": 0.400969535112381,
      "learning_rate": 2e-05,
      "loss": 0.5582,
      "step": 400
    },
    {
      "epoch": 0.03370601775731667,
      "grad_norm": 0.5791268348693848,
      "learning_rate": 2e-05,
      "loss": 0.5586,
      "step": 410
    },
    {
      "epoch": 0.03452811575139757,
      "grad_norm": 0.3688918650150299,
      "learning_rate": 2e-05,
      "loss": 0.566,
      "step": 420
    },
    {
      "epoch": 0.03535021374547846,
      "grad_norm": 0.41482818126678467,
      "learning_rate": 2e-05,
      "loss": 0.5501,
      "step": 430
    },
    {
      "epoch": 0.03617231173955936,
      "grad_norm": 0.4072619676589966,
      "learning_rate": 2e-05,
      "loss": 0.5756,
      "step": 440
    },
    {
      "epoch": 0.03699440973364025,
      "grad_norm": 0.389475554227829,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 450
    },
    {
      "epoch": 0.037816507727721146,
      "grad_norm": 0.3854443430900574,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 460
    },
    {
      "epoch": 0.03863860572180204,
      "grad_norm": 0.35742488503456116,
      "learning_rate": 2e-05,
      "loss": 0.5758,
      "step": 470
    },
    {
      "epoch": 0.039460703715882936,
      "grad_norm": 0.3568844795227051,
      "learning_rate": 2e-05,
      "loss": 0.5463,
      "step": 480
    },
    {
      "epoch": 0.04028280170996383,
      "grad_norm": 0.3795979917049408,
      "learning_rate": 2e-05,
      "loss": 0.5436,
      "step": 490
    },
    {
      "epoch": 0.041104899704044726,
      "grad_norm": 0.37757909297943115,
      "learning_rate": 2e-05,
      "loss": 0.5663,
      "step": 500
    },
    {
      "epoch": 0.04192699769812561,
      "grad_norm": 0.35400980710983276,
      "learning_rate": 2e-05,
      "loss": 0.5598,
      "step": 510
    },
    {
      "epoch": 0.04274909569220651,
      "grad_norm": 0.3492357134819031,
      "learning_rate": 2e-05,
      "loss": 0.595,
      "step": 520
    },
    {
      "epoch": 0.0435711936862874,
      "grad_norm": 0.3667033910751343,
      "learning_rate": 2e-05,
      "loss": 0.5476,
      "step": 530
    },
    {
      "epoch": 0.0443932916803683,
      "grad_norm": 0.42593854665756226,
      "learning_rate": 2e-05,
      "loss": 0.5971,
      "step": 540
    },
    {
      "epoch": 0.04521538967444919,
      "grad_norm": 0.37045472860336304,
      "learning_rate": 2e-05,
      "loss": 0.5591,
      "step": 550
    },
    {
      "epoch": 0.04603748766853009,
      "grad_norm": 0.38231995701789856,
      "learning_rate": 2e-05,
      "loss": 0.5601,
      "step": 560
    },
    {
      "epoch": 0.04685958566261098,
      "grad_norm": 0.37826621532440186,
      "learning_rate": 2e-05,
      "loss": 0.543,
      "step": 570
    },
    {
      "epoch": 0.04768168365669188,
      "grad_norm": 0.3782760202884674,
      "learning_rate": 2e-05,
      "loss": 0.5883,
      "step": 580
    },
    {
      "epoch": 0.04850378165077277,
      "grad_norm": 0.3855355679988861,
      "learning_rate": 2e-05,
      "loss": 0.561,
      "step": 590
    },
    {
      "epoch": 0.04932587964485367,
      "grad_norm": 0.40973329544067383,
      "learning_rate": 2e-05,
      "loss": 0.5722,
      "step": 600
    },
    {
      "epoch": 0.05014797763893456,
      "grad_norm": 0.3679392337799072,
      "learning_rate": 2e-05,
      "loss": 0.566,
      "step": 610
    },
    {
      "epoch": 0.050970075633015456,
      "grad_norm": 0.36318325996398926,
      "learning_rate": 2e-05,
      "loss": 0.5539,
      "step": 620
    },
    {
      "epoch": 0.05179217362709635,
      "grad_norm": 0.3448237478733063,
      "learning_rate": 2e-05,
      "loss": 0.5757,
      "step": 630
    },
    {
      "epoch": 0.052614271621177246,
      "grad_norm": 0.48463839292526245,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 640
    },
    {
      "epoch": 0.05343636961525814,
      "grad_norm": 0.368011474609375,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 650
    },
    {
      "epoch": 0.054258467609339035,
      "grad_norm": 0.3750072419643402,
      "learning_rate": 2e-05,
      "loss": 0.5709,
      "step": 660
    },
    {
      "epoch": 0.05508056560341993,
      "grad_norm": 0.39397040009498596,
      "learning_rate": 2e-05,
      "loss": 0.5442,
      "step": 670
    },
    {
      "epoch": 0.055902663597500825,
      "grad_norm": 0.38440752029418945,
      "learning_rate": 2e-05,
      "loss": 0.5653,
      "step": 680
    },
    {
      "epoch": 0.05672476159158172,
      "grad_norm": 0.44925448298454285,
      "learning_rate": 2e-05,
      "loss": 0.5776,
      "step": 690
    },
    {
      "epoch": 0.057546859585662614,
      "grad_norm": 0.4260270297527313,
      "learning_rate": 2e-05,
      "loss": 0.5737,
      "step": 700
    },
    {
      "epoch": 0.0583689575797435,
      "grad_norm": 0.39539211988449097,
      "learning_rate": 2e-05,
      "loss": 0.5613,
      "step": 710
    },
    {
      "epoch": 0.0591910555738244,
      "grad_norm": 0.35687392950057983,
      "learning_rate": 2e-05,
      "loss": 0.5827,
      "step": 720
    },
    {
      "epoch": 0.06001315356790529,
      "grad_norm": 0.6145081520080566,
      "learning_rate": 2e-05,
      "loss": 0.5687,
      "step": 730
    },
    {
      "epoch": 0.06083525156198619,
      "grad_norm": 0.4325614273548126,
      "learning_rate": 2e-05,
      "loss": 0.5366,
      "step": 740
    },
    {
      "epoch": 0.06165734955606708,
      "grad_norm": 0.35582131147384644,
      "learning_rate": 2e-05,
      "loss": 0.5812,
      "step": 750
    },
    {
      "epoch": 0.062479447550147976,
      "grad_norm": 0.38189223408699036,
      "learning_rate": 2e-05,
      "loss": 0.5593,
      "step": 760
    },
    {
      "epoch": 0.06330154554422887,
      "grad_norm": 0.4139549434185028,
      "learning_rate": 2e-05,
      "loss": 0.5186,
      "step": 770
    },
    {
      "epoch": 0.06412364353830977,
      "grad_norm": 0.39057818055152893,
      "learning_rate": 2e-05,
      "loss": 0.5161,
      "step": 780
    },
    {
      "epoch": 0.06494574153239066,
      "grad_norm": 0.46800991892814636,
      "learning_rate": 2e-05,
      "loss": 0.5564,
      "step": 790
    },
    {
      "epoch": 0.06576783952647156,
      "grad_norm": 0.3704215884208679,
      "learning_rate": 2e-05,
      "loss": 0.5404,
      "step": 800
    },
    {
      "epoch": 0.06658993752055245,
      "grad_norm": 0.3691960871219635,
      "learning_rate": 2e-05,
      "loss": 0.5429,
      "step": 810
    },
    {
      "epoch": 0.06741203551463334,
      "grad_norm": 0.35629382729530334,
      "learning_rate": 2e-05,
      "loss": 0.5053,
      "step": 820
    },
    {
      "epoch": 0.06823413350871424,
      "grad_norm": 0.45477181673049927,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 830
    },
    {
      "epoch": 0.06905623150279513,
      "grad_norm": 0.36613255739212036,
      "learning_rate": 2e-05,
      "loss": 0.5621,
      "step": 840
    },
    {
      "epoch": 0.06987832949687603,
      "grad_norm": 0.38469594717025757,
      "learning_rate": 2e-05,
      "loss": 0.5734,
      "step": 850
    },
    {
      "epoch": 0.07070042749095692,
      "grad_norm": 0.38618406653404236,
      "learning_rate": 2e-05,
      "loss": 0.5549,
      "step": 860
    },
    {
      "epoch": 0.07152252548503782,
      "grad_norm": 0.3400881588459015,
      "learning_rate": 2e-05,
      "loss": 0.5348,
      "step": 870
    },
    {
      "epoch": 0.07234462347911871,
      "grad_norm": 0.36379608511924744,
      "learning_rate": 2e-05,
      "loss": 0.5436,
      "step": 880
    },
    {
      "epoch": 0.07316672147319961,
      "grad_norm": 0.4529801309108734,
      "learning_rate": 2e-05,
      "loss": 0.5772,
      "step": 890
    },
    {
      "epoch": 0.0739888194672805,
      "grad_norm": 0.37483784556388855,
      "learning_rate": 2e-05,
      "loss": 0.5383,
      "step": 900
    },
    {
      "epoch": 0.0748109174613614,
      "grad_norm": 0.42458927631378174,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 910
    },
    {
      "epoch": 0.07563301545544229,
      "grad_norm": 0.3337687849998474,
      "learning_rate": 2e-05,
      "loss": 0.5676,
      "step": 920
    },
    {
      "epoch": 0.07645511344952319,
      "grad_norm": 0.40329763293266296,
      "learning_rate": 2e-05,
      "loss": 0.5502,
      "step": 930
    },
    {
      "epoch": 0.07727721144360408,
      "grad_norm": 0.37778130173683167,
      "learning_rate": 2e-05,
      "loss": 0.5855,
      "step": 940
    },
    {
      "epoch": 0.07809930943768498,
      "grad_norm": 0.4349541962146759,
      "learning_rate": 2e-05,
      "loss": 0.5457,
      "step": 950
    },
    {
      "epoch": 0.07892140743176587,
      "grad_norm": 0.33621495962142944,
      "learning_rate": 2e-05,
      "loss": 0.5314,
      "step": 960
    },
    {
      "epoch": 0.07974350542584677,
      "grad_norm": 0.3857935965061188,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 970
    },
    {
      "epoch": 0.08056560341992766,
      "grad_norm": 0.3538459539413452,
      "learning_rate": 2e-05,
      "loss": 0.5577,
      "step": 980
    },
    {
      "epoch": 0.08138770141400856,
      "grad_norm": 0.3277754783630371,
      "learning_rate": 2e-05,
      "loss": 0.5736,
      "step": 990
    },
    {
      "epoch": 0.08220979940808945,
      "grad_norm": 0.33061307668685913,
      "learning_rate": 2e-05,
      "loss": 0.5463,
      "step": 1000
    },
    {
      "epoch": 0.08303189740217033,
      "grad_norm": 0.5522162914276123,
      "learning_rate": 2e-05,
      "loss": 0.5415,
      "step": 1010
    },
    {
      "epoch": 0.08385399539625123,
      "grad_norm": 0.337895929813385,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 1020
    },
    {
      "epoch": 0.08467609339033212,
      "grad_norm": 0.4023716449737549,
      "learning_rate": 2e-05,
      "loss": 0.5925,
      "step": 1030
    },
    {
      "epoch": 0.08549819138441302,
      "grad_norm": 0.35799404978752136,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 1040
    },
    {
      "epoch": 0.08632028937849391,
      "grad_norm": 0.3547460734844208,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 1050
    },
    {
      "epoch": 0.0871423873725748,
      "grad_norm": 0.44388723373413086,
      "learning_rate": 2e-05,
      "loss": 0.5746,
      "step": 1060
    },
    {
      "epoch": 0.0879644853666557,
      "grad_norm": 0.3479222059249878,
      "learning_rate": 2e-05,
      "loss": 0.5638,
      "step": 1070
    },
    {
      "epoch": 0.0887865833607366,
      "grad_norm": 0.37281978130340576,
      "learning_rate": 2e-05,
      "loss": 0.5595,
      "step": 1080
    },
    {
      "epoch": 0.08960868135481749,
      "grad_norm": 0.3608080744743347,
      "learning_rate": 2e-05,
      "loss": 0.5655,
      "step": 1090
    },
    {
      "epoch": 0.09043077934889839,
      "grad_norm": 0.3156391680240631,
      "learning_rate": 2e-05,
      "loss": 0.5364,
      "step": 1100
    },
    {
      "epoch": 0.09125287734297928,
      "grad_norm": 0.3983137905597687,
      "learning_rate": 2e-05,
      "loss": 0.5511,
      "step": 1110
    },
    {
      "epoch": 0.09207497533706017,
      "grad_norm": 0.3987083435058594,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 1120
    },
    {
      "epoch": 0.09289707333114107,
      "grad_norm": 0.3358446955680847,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 1130
    },
    {
      "epoch": 0.09371917132522196,
      "grad_norm": 0.3355688154697418,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 1140
    },
    {
      "epoch": 0.09454126931930286,
      "grad_norm": 0.39273035526275635,
      "learning_rate": 2e-05,
      "loss": 0.5551,
      "step": 1150
    },
    {
      "epoch": 0.09536336731338375,
      "grad_norm": 0.37203627824783325,
      "learning_rate": 2e-05,
      "loss": 0.5894,
      "step": 1160
    },
    {
      "epoch": 0.09618546530746465,
      "grad_norm": 0.34250086545944214,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 1170
    },
    {
      "epoch": 0.09700756330154554,
      "grad_norm": 0.3525017499923706,
      "learning_rate": 2e-05,
      "loss": 0.5527,
      "step": 1180
    },
    {
      "epoch": 0.09782966129562644,
      "grad_norm": 0.37996408343315125,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 1190
    },
    {
      "epoch": 0.09865175928970733,
      "grad_norm": 0.46024736762046814,
      "learning_rate": 2e-05,
      "loss": 0.5672,
      "step": 1200
    },
    {
      "epoch": 0.09947385728378823,
      "grad_norm": 0.3837468922138214,
      "learning_rate": 2e-05,
      "loss": 0.5551,
      "step": 1210
    },
    {
      "epoch": 0.10029595527786912,
      "grad_norm": 0.40620625019073486,
      "learning_rate": 2e-05,
      "loss": 0.5468,
      "step": 1220
    },
    {
      "epoch": 0.10111805327195002,
      "grad_norm": 0.35995301604270935,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 1230
    },
    {
      "epoch": 0.10194015126603091,
      "grad_norm": 0.396985799074173,
      "learning_rate": 2e-05,
      "loss": 0.5298,
      "step": 1240
    },
    {
      "epoch": 0.10276224926011181,
      "grad_norm": 0.38154250383377075,
      "learning_rate": 2e-05,
      "loss": 0.5574,
      "step": 1250
    },
    {
      "epoch": 0.1035843472541927,
      "grad_norm": 0.3399718403816223,
      "learning_rate": 2e-05,
      "loss": 0.5873,
      "step": 1260
    },
    {
      "epoch": 0.1044064452482736,
      "grad_norm": 0.39190253615379333,
      "learning_rate": 2e-05,
      "loss": 0.523,
      "step": 1270
    },
    {
      "epoch": 0.10522854324235449,
      "grad_norm": 0.34379804134368896,
      "learning_rate": 2e-05,
      "loss": 0.5429,
      "step": 1280
    },
    {
      "epoch": 0.10605064123643539,
      "grad_norm": 0.34028005599975586,
      "learning_rate": 2e-05,
      "loss": 0.5527,
      "step": 1290
    },
    {
      "epoch": 0.10687273923051628,
      "grad_norm": 0.3677566349506378,
      "learning_rate": 2e-05,
      "loss": 0.544,
      "step": 1300
    },
    {
      "epoch": 0.10769483722459718,
      "grad_norm": 0.3977677524089813,
      "learning_rate": 2e-05,
      "loss": 0.53,
      "step": 1310
    },
    {
      "epoch": 0.10851693521867807,
      "grad_norm": 0.33426037430763245,
      "learning_rate": 2e-05,
      "loss": 0.54,
      "step": 1320
    },
    {
      "epoch": 0.10933903321275897,
      "grad_norm": 0.3396483361721039,
      "learning_rate": 2e-05,
      "loss": 0.5681,
      "step": 1330
    },
    {
      "epoch": 0.11016113120683986,
      "grad_norm": 0.3713797926902771,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 1340
    },
    {
      "epoch": 0.11098322920092075,
      "grad_norm": 0.3870110511779785,
      "learning_rate": 2e-05,
      "loss": 0.5489,
      "step": 1350
    },
    {
      "epoch": 0.11180532719500165,
      "grad_norm": 0.4049339294433594,
      "learning_rate": 2e-05,
      "loss": 0.5656,
      "step": 1360
    },
    {
      "epoch": 0.11262742518908254,
      "grad_norm": 0.3298855423927307,
      "learning_rate": 2e-05,
      "loss": 0.5605,
      "step": 1370
    },
    {
      "epoch": 0.11344952318316344,
      "grad_norm": 0.3444337546825409,
      "learning_rate": 2e-05,
      "loss": 0.5446,
      "step": 1380
    },
    {
      "epoch": 0.11427162117724433,
      "grad_norm": 0.329982191324234,
      "learning_rate": 2e-05,
      "loss": 0.5602,
      "step": 1390
    },
    {
      "epoch": 0.11509371917132523,
      "grad_norm": 0.3550669252872467,
      "learning_rate": 2e-05,
      "loss": 0.5293,
      "step": 1400
    },
    {
      "epoch": 0.11591581716540611,
      "grad_norm": 0.3423064053058624,
      "learning_rate": 2e-05,
      "loss": 0.5502,
      "step": 1410
    },
    {
      "epoch": 0.116737915159487,
      "grad_norm": 0.4455831050872803,
      "learning_rate": 2e-05,
      "loss": 0.5537,
      "step": 1420
    },
    {
      "epoch": 0.1175600131535679,
      "grad_norm": 0.3592797517776489,
      "learning_rate": 2e-05,
      "loss": 0.5289,
      "step": 1430
    },
    {
      "epoch": 0.1183821111476488,
      "grad_norm": 0.3526698648929596,
      "learning_rate": 2e-05,
      "loss": 0.5458,
      "step": 1440
    },
    {
      "epoch": 0.11920420914172969,
      "grad_norm": 0.3346903324127197,
      "learning_rate": 2e-05,
      "loss": 0.5394,
      "step": 1450
    },
    {
      "epoch": 0.12002630713581058,
      "grad_norm": 0.36272871494293213,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 1460
    },
    {
      "epoch": 0.12084840512989148,
      "grad_norm": 0.38370686769485474,
      "learning_rate": 2e-05,
      "loss": 0.5616,
      "step": 1470
    },
    {
      "epoch": 0.12167050312397237,
      "grad_norm": 0.43002524971961975,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 1480
    },
    {
      "epoch": 0.12249260111805327,
      "grad_norm": 0.3865845501422882,
      "learning_rate": 2e-05,
      "loss": 0.5561,
      "step": 1490
    },
    {
      "epoch": 0.12331469911213416,
      "grad_norm": 0.3935903012752533,
      "learning_rate": 2e-05,
      "loss": 0.5519,
      "step": 1500
    },
    {
      "epoch": 0.12413679710621506,
      "grad_norm": 0.3224612772464752,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 1510
    },
    {
      "epoch": 0.12495889510029595,
      "grad_norm": 0.39488735795021057,
      "learning_rate": 2e-05,
      "loss": 0.5589,
      "step": 1520
    },
    {
      "epoch": 0.12578099309437685,
      "grad_norm": 0.366058349609375,
      "learning_rate": 2e-05,
      "loss": 0.5673,
      "step": 1530
    },
    {
      "epoch": 0.12660309108845774,
      "grad_norm": 0.40682297945022583,
      "learning_rate": 2e-05,
      "loss": 0.5565,
      "step": 1540
    },
    {
      "epoch": 0.12742518908253864,
      "grad_norm": 0.3384777903556824,
      "learning_rate": 2e-05,
      "loss": 0.5319,
      "step": 1550
    },
    {
      "epoch": 0.12824728707661953,
      "grad_norm": 0.3520301282405853,
      "learning_rate": 2e-05,
      "loss": 0.5555,
      "step": 1560
    },
    {
      "epoch": 0.12906938507070043,
      "grad_norm": 0.37539342045783997,
      "learning_rate": 2e-05,
      "loss": 0.5639,
      "step": 1570
    },
    {
      "epoch": 0.12989148306478132,
      "grad_norm": 0.418217271566391,
      "learning_rate": 2e-05,
      "loss": 0.5483,
      "step": 1580
    },
    {
      "epoch": 0.13071358105886222,
      "grad_norm": 0.37258896231651306,
      "learning_rate": 2e-05,
      "loss": 0.6018,
      "step": 1590
    },
    {
      "epoch": 0.1315356790529431,
      "grad_norm": 0.35443881154060364,
      "learning_rate": 2e-05,
      "loss": 0.5485,
      "step": 1600
    },
    {
      "epoch": 0.132357777047024,
      "grad_norm": 0.3421747386455536,
      "learning_rate": 2e-05,
      "loss": 0.5602,
      "step": 1610
    },
    {
      "epoch": 0.1331798750411049,
      "grad_norm": 0.3465687036514282,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 1620
    },
    {
      "epoch": 0.1340019730351858,
      "grad_norm": 0.7691576480865479,
      "learning_rate": 2e-05,
      "loss": 0.5456,
      "step": 1630
    },
    {
      "epoch": 0.1348240710292667,
      "grad_norm": 0.6803935766220093,
      "learning_rate": 2e-05,
      "loss": 0.58,
      "step": 1640
    },
    {
      "epoch": 0.13564616902334758,
      "grad_norm": 0.3570682108402252,
      "learning_rate": 2e-05,
      "loss": 0.5721,
      "step": 1650
    },
    {
      "epoch": 0.13646826701742848,
      "grad_norm": 0.5207669734954834,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 1660
    },
    {
      "epoch": 0.13729036501150937,
      "grad_norm": 0.32501304149627686,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 1670
    },
    {
      "epoch": 0.13811246300559027,
      "grad_norm": 0.39292895793914795,
      "learning_rate": 2e-05,
      "loss": 0.5437,
      "step": 1680
    },
    {
      "epoch": 0.13893456099967116,
      "grad_norm": 0.3334016501903534,
      "learning_rate": 2e-05,
      "loss": 0.5599,
      "step": 1690
    },
    {
      "epoch": 0.13975665899375206,
      "grad_norm": 0.31479495763778687,
      "learning_rate": 2e-05,
      "loss": 0.5658,
      "step": 1700
    },
    {
      "epoch": 0.14057875698783295,
      "grad_norm": 0.38708022236824036,
      "learning_rate": 2e-05,
      "loss": 0.5554,
      "step": 1710
    },
    {
      "epoch": 0.14140085498191385,
      "grad_norm": 0.29850685596466064,
      "learning_rate": 2e-05,
      "loss": 0.5217,
      "step": 1720
    },
    {
      "epoch": 0.14222295297599474,
      "grad_norm": 0.3497243821620941,
      "learning_rate": 2e-05,
      "loss": 0.5571,
      "step": 1730
    },
    {
      "epoch": 0.14304505097007564,
      "grad_norm": 0.34561386704444885,
      "learning_rate": 2e-05,
      "loss": 0.5815,
      "step": 1740
    },
    {
      "epoch": 0.14386714896415653,
      "grad_norm": 0.3378225862979889,
      "learning_rate": 2e-05,
      "loss": 0.5522,
      "step": 1750
    },
    {
      "epoch": 0.14468924695823743,
      "grad_norm": 0.3681405782699585,
      "learning_rate": 2e-05,
      "loss": 0.5257,
      "step": 1760
    },
    {
      "epoch": 0.14551134495231832,
      "grad_norm": 0.3730759024620056,
      "learning_rate": 2e-05,
      "loss": 0.5182,
      "step": 1770
    },
    {
      "epoch": 0.14633344294639922,
      "grad_norm": 0.41251009702682495,
      "learning_rate": 2e-05,
      "loss": 0.5824,
      "step": 1780
    },
    {
      "epoch": 0.1471555409404801,
      "grad_norm": 0.34546494483947754,
      "learning_rate": 2e-05,
      "loss": 0.5333,
      "step": 1790
    },
    {
      "epoch": 0.147977638934561,
      "grad_norm": 0.3387209177017212,
      "learning_rate": 2e-05,
      "loss": 0.5125,
      "step": 1800
    },
    {
      "epoch": 0.1487997369286419,
      "grad_norm": 0.3762285113334656,
      "learning_rate": 2e-05,
      "loss": 0.5307,
      "step": 1810
    },
    {
      "epoch": 0.1496218349227228,
      "grad_norm": 0.34216463565826416,
      "learning_rate": 2e-05,
      "loss": 0.5536,
      "step": 1820
    },
    {
      "epoch": 0.1504439329168037,
      "grad_norm": 0.3812594711780548,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 1830
    },
    {
      "epoch": 0.15126603091088459,
      "grad_norm": 0.34243831038475037,
      "learning_rate": 2e-05,
      "loss": 0.5408,
      "step": 1840
    },
    {
      "epoch": 0.15208812890496548,
      "grad_norm": 0.3886333107948303,
      "learning_rate": 2e-05,
      "loss": 0.5536,
      "step": 1850
    },
    {
      "epoch": 0.15291022689904638,
      "grad_norm": 0.37051960825920105,
      "learning_rate": 2e-05,
      "loss": 0.5584,
      "step": 1860
    },
    {
      "epoch": 0.15373232489312727,
      "grad_norm": 0.3963000178337097,
      "learning_rate": 2e-05,
      "loss": 0.5415,
      "step": 1870
    },
    {
      "epoch": 0.15455442288720816,
      "grad_norm": 0.36416611075401306,
      "learning_rate": 2e-05,
      "loss": 0.5514,
      "step": 1880
    },
    {
      "epoch": 0.15537652088128906,
      "grad_norm": 0.765783965587616,
      "learning_rate": 2e-05,
      "loss": 0.5476,
      "step": 1890
    },
    {
      "epoch": 0.15619861887536995,
      "grad_norm": 0.3947313129901886,
      "learning_rate": 2e-05,
      "loss": 0.5351,
      "step": 1900
    },
    {
      "epoch": 0.15702071686945085,
      "grad_norm": 0.42167967557907104,
      "learning_rate": 2e-05,
      "loss": 0.584,
      "step": 1910
    },
    {
      "epoch": 0.15784281486353174,
      "grad_norm": 0.3771551251411438,
      "learning_rate": 2e-05,
      "loss": 0.5411,
      "step": 1920
    },
    {
      "epoch": 0.15866491285761264,
      "grad_norm": 0.34432297945022583,
      "learning_rate": 2e-05,
      "loss": 0.5238,
      "step": 1930
    },
    {
      "epoch": 0.15948701085169353,
      "grad_norm": 0.3422872722148895,
      "learning_rate": 2e-05,
      "loss": 0.5669,
      "step": 1940
    },
    {
      "epoch": 0.16030910884577443,
      "grad_norm": 0.35941794514656067,
      "learning_rate": 2e-05,
      "loss": 0.5323,
      "step": 1950
    },
    {
      "epoch": 0.16113120683985532,
      "grad_norm": 0.4349963068962097,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 1960
    },
    {
      "epoch": 0.16195330483393622,
      "grad_norm": 0.49831265211105347,
      "learning_rate": 2e-05,
      "loss": 0.5511,
      "step": 1970
    },
    {
      "epoch": 0.1627754028280171,
      "grad_norm": 0.3435792624950409,
      "learning_rate": 2e-05,
      "loss": 0.5323,
      "step": 1980
    },
    {
      "epoch": 0.163597500822098,
      "grad_norm": 0.5475247502326965,
      "learning_rate": 2e-05,
      "loss": 0.5605,
      "step": 1990
    },
    {
      "epoch": 0.1644195988161789,
      "grad_norm": 0.3261669874191284,
      "learning_rate": 2e-05,
      "loss": 0.5466,
      "step": 2000
    },
    {
      "epoch": 0.16524169681025977,
      "grad_norm": 0.3482590615749359,
      "learning_rate": 2e-05,
      "loss": 0.5449,
      "step": 2010
    },
    {
      "epoch": 0.16606379480434066,
      "grad_norm": 0.41366517543792725,
      "learning_rate": 2e-05,
      "loss": 0.5535,
      "step": 2020
    },
    {
      "epoch": 0.16688589279842156,
      "grad_norm": 0.36402422189712524,
      "learning_rate": 2e-05,
      "loss": 0.5613,
      "step": 2030
    },
    {
      "epoch": 0.16770799079250245,
      "grad_norm": 0.347074419260025,
      "learning_rate": 2e-05,
      "loss": 0.5455,
      "step": 2040
    },
    {
      "epoch": 0.16853008878658335,
      "grad_norm": 0.35407862067222595,
      "learning_rate": 2e-05,
      "loss": 0.5549,
      "step": 2050
    },
    {
      "epoch": 0.16935218678066424,
      "grad_norm": 0.3733304738998413,
      "learning_rate": 2e-05,
      "loss": 0.5534,
      "step": 2060
    },
    {
      "epoch": 0.17017428477474514,
      "grad_norm": 0.36466488242149353,
      "learning_rate": 2e-05,
      "loss": 0.5256,
      "step": 2070
    },
    {
      "epoch": 0.17099638276882603,
      "grad_norm": 0.3579888641834259,
      "learning_rate": 2e-05,
      "loss": 0.5472,
      "step": 2080
    },
    {
      "epoch": 0.17181848076290693,
      "grad_norm": 0.3732098639011383,
      "learning_rate": 2e-05,
      "loss": 0.5302,
      "step": 2090
    },
    {
      "epoch": 0.17264057875698782,
      "grad_norm": 0.3745439946651459,
      "learning_rate": 2e-05,
      "loss": 0.5729,
      "step": 2100
    },
    {
      "epoch": 0.17346267675106872,
      "grad_norm": 0.33778682351112366,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 2110
    },
    {
      "epoch": 0.1742847747451496,
      "grad_norm": 0.4035145044326782,
      "learning_rate": 2e-05,
      "loss": 0.5604,
      "step": 2120
    },
    {
      "epoch": 0.1751068727392305,
      "grad_norm": 0.41023483872413635,
      "learning_rate": 2e-05,
      "loss": 0.5526,
      "step": 2130
    },
    {
      "epoch": 0.1759289707333114,
      "grad_norm": 0.3626518249511719,
      "learning_rate": 2e-05,
      "loss": 0.5384,
      "step": 2140
    },
    {
      "epoch": 0.1767510687273923,
      "grad_norm": 0.3237241506576538,
      "learning_rate": 2e-05,
      "loss": 0.5577,
      "step": 2150
    },
    {
      "epoch": 0.1775731667214732,
      "grad_norm": 0.37217530608177185,
      "learning_rate": 2e-05,
      "loss": 0.5607,
      "step": 2160
    },
    {
      "epoch": 0.17839526471555409,
      "grad_norm": 0.3166368305683136,
      "learning_rate": 2e-05,
      "loss": 0.5184,
      "step": 2170
    },
    {
      "epoch": 0.17921736270963498,
      "grad_norm": 4.73347282409668,
      "learning_rate": 2e-05,
      "loss": 0.5469,
      "step": 2180
    },
    {
      "epoch": 0.18003946070371588,
      "grad_norm": 0.3166395127773285,
      "learning_rate": 2e-05,
      "loss": 0.548,
      "step": 2190
    },
    {
      "epoch": 0.18086155869779677,
      "grad_norm": 0.37149322032928467,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 2200
    },
    {
      "epoch": 0.18168365669187767,
      "grad_norm": 0.3564207851886749,
      "learning_rate": 2e-05,
      "loss": 0.5534,
      "step": 2210
    },
    {
      "epoch": 0.18250575468595856,
      "grad_norm": 0.35720905661582947,
      "learning_rate": 2e-05,
      "loss": 0.5442,
      "step": 2220
    },
    {
      "epoch": 0.18332785268003945,
      "grad_norm": 0.3910946249961853,
      "learning_rate": 2e-05,
      "loss": 0.5389,
      "step": 2230
    },
    {
      "epoch": 0.18414995067412035,
      "grad_norm": 0.35082006454467773,
      "learning_rate": 2e-05,
      "loss": 0.5641,
      "step": 2240
    },
    {
      "epoch": 0.18497204866820124,
      "grad_norm": 0.352914035320282,
      "learning_rate": 2e-05,
      "loss": 0.572,
      "step": 2250
    },
    {
      "epoch": 0.18579414666228214,
      "grad_norm": 0.3582814335823059,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 2260
    },
    {
      "epoch": 0.18661624465636303,
      "grad_norm": 0.4184710681438446,
      "learning_rate": 2e-05,
      "loss": 0.5692,
      "step": 2270
    },
    {
      "epoch": 0.18743834265044393,
      "grad_norm": 0.5157349109649658,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 2280
    },
    {
      "epoch": 0.18826044064452482,
      "grad_norm": 0.438728928565979,
      "learning_rate": 2e-05,
      "loss": 0.5388,
      "step": 2290
    },
    {
      "epoch": 0.18908253863860572,
      "grad_norm": 0.31124523282051086,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 2300
    },
    {
      "epoch": 0.1899046366326866,
      "grad_norm": 0.35619640350341797,
      "learning_rate": 2e-05,
      "loss": 0.5229,
      "step": 2310
    },
    {
      "epoch": 0.1907267346267675,
      "grad_norm": 0.3216730058193207,
      "learning_rate": 2e-05,
      "loss": 0.5358,
      "step": 2320
    },
    {
      "epoch": 0.1915488326208484,
      "grad_norm": 0.38024744391441345,
      "learning_rate": 2e-05,
      "loss": 0.562,
      "step": 2330
    },
    {
      "epoch": 0.1923709306149293,
      "grad_norm": 0.3398594856262207,
      "learning_rate": 2e-05,
      "loss": 0.5448,
      "step": 2340
    },
    {
      "epoch": 0.1931930286090102,
      "grad_norm": 0.3224997818470001,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 2350
    },
    {
      "epoch": 0.1940151266030911,
      "grad_norm": 0.43699830770492554,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 2360
    },
    {
      "epoch": 0.19483722459717198,
      "grad_norm": 0.7130771279335022,
      "learning_rate": 2e-05,
      "loss": 0.5407,
      "step": 2370
    },
    {
      "epoch": 0.19565932259125288,
      "grad_norm": 0.345954567193985,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 2380
    },
    {
      "epoch": 0.19648142058533377,
      "grad_norm": 0.3121144473552704,
      "learning_rate": 2e-05,
      "loss": 0.5435,
      "step": 2390
    },
    {
      "epoch": 0.19730351857941467,
      "grad_norm": 0.47650909423828125,
      "learning_rate": 2e-05,
      "loss": 0.542,
      "step": 2400
    },
    {
      "epoch": 0.19812561657349556,
      "grad_norm": 0.3822428286075592,
      "learning_rate": 2e-05,
      "loss": 0.5577,
      "step": 2410
    },
    {
      "epoch": 0.19894771456757646,
      "grad_norm": 0.3568475842475891,
      "learning_rate": 2e-05,
      "loss": 0.5418,
      "step": 2420
    },
    {
      "epoch": 0.19976981256165735,
      "grad_norm": 0.3287544846534729,
      "learning_rate": 2e-05,
      "loss": 0.5381,
      "step": 2430
    },
    {
      "epoch": 0.20059191055573825,
      "grad_norm": 0.3641667664051056,
      "learning_rate": 2e-05,
      "loss": 0.5694,
      "step": 2440
    },
    {
      "epoch": 0.20141400854981914,
      "grad_norm": 0.3377435803413391,
      "learning_rate": 2e-05,
      "loss": 0.5552,
      "step": 2450
    },
    {
      "epoch": 0.20223610654390003,
      "grad_norm": 0.41057026386260986,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 2460
    },
    {
      "epoch": 0.20305820453798093,
      "grad_norm": 0.3275510370731354,
      "learning_rate": 2e-05,
      "loss": 0.5646,
      "step": 2470
    },
    {
      "epoch": 0.20388030253206182,
      "grad_norm": 0.31972360610961914,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 2480
    },
    {
      "epoch": 0.20470240052614272,
      "grad_norm": 0.37606051564216614,
      "learning_rate": 2e-05,
      "loss": 0.5597,
      "step": 2490
    },
    {
      "epoch": 0.20552449852022361,
      "grad_norm": 0.3712795674800873,
      "learning_rate": 2e-05,
      "loss": 0.5722,
      "step": 2500
    },
    {
      "epoch": 0.2063465965143045,
      "grad_norm": 0.4065472483634949,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 2510
    },
    {
      "epoch": 0.2071686945083854,
      "grad_norm": 0.34219712018966675,
      "learning_rate": 2e-05,
      "loss": 0.5525,
      "step": 2520
    },
    {
      "epoch": 0.2079907925024663,
      "grad_norm": 0.3456852436065674,
      "learning_rate": 2e-05,
      "loss": 0.5529,
      "step": 2530
    },
    {
      "epoch": 0.2088128904965472,
      "grad_norm": 0.7063271999359131,
      "learning_rate": 2e-05,
      "loss": 0.5781,
      "step": 2540
    },
    {
      "epoch": 0.2096349884906281,
      "grad_norm": 0.3391132950782776,
      "learning_rate": 1.997261400182901e-05,
      "loss": 0.5553,
      "step": 2550
    },
    {
      "epoch": 0.21045708648470898,
      "grad_norm": 0.39100420475006104,
      "learning_rate": 1.9805800746810423e-05,
      "loss": 0.5239,
      "step": 2560
    },
    {
      "epoch": 0.21127918447878988,
      "grad_norm": 0.34427765011787415,
      "learning_rate": 1.9489934220110274e-05,
      "loss": 0.5361,
      "step": 2570
    },
    {
      "epoch": 0.21210128247287077,
      "grad_norm": 0.345601350069046,
      "learning_rate": 1.9029840364820392e-05,
      "loss": 0.536,
      "step": 2580
    },
    {
      "epoch": 0.21292338046695167,
      "grad_norm": 0.36759981513023376,
      "learning_rate": 1.8432548690479908e-05,
      "loss": 0.5126,
      "step": 2590
    },
    {
      "epoch": 0.21374547846103256,
      "grad_norm": 0.41115498542785645,
      "learning_rate": 1.7707184873239772e-05,
      "loss": 0.5633,
      "step": 2600
    },
    {
      "epoch": 0.21456757645511346,
      "grad_norm": 0.33417969942092896,
      "learning_rate": 1.6864831329902565e-05,
      "loss": 0.5164,
      "step": 2610
    },
    {
      "epoch": 0.21538967444919435,
      "grad_norm": 0.34763196110725403,
      "learning_rate": 1.591835789604667e-05,
      "loss": 0.549,
      "step": 2620
    },
    {
      "epoch": 0.21621177244327525,
      "grad_norm": 0.3870541751384735,
      "learning_rate": 1.4882225195206459e-05,
      "loss": 0.581,
      "step": 2630
    },
    {
      "epoch": 0.21703387043735614,
      "grad_norm": 0.33340081572532654,
      "learning_rate": 1.3772263703317695e-05,
      "loss": 0.5486,
      "step": 2640
    },
    {
      "epoch": 0.21785596843143704,
      "grad_norm": 0.34669554233551025,
      "learning_rate": 1.2605431883975547e-05,
      "loss": 0.5124,
      "step": 2650
    },
    {
      "epoch": 0.21867806642551793,
      "grad_norm": 0.35023757815361023,
      "learning_rate": 1.1399557089817574e-05,
      "loss": 0.525,
      "step": 2660
    },
    {
      "epoch": 0.21950016441959883,
      "grad_norm": 0.34071967005729675,
      "learning_rate": 1.0173063188650463e-05,
      "loss": 0.5688,
      "step": 2670
    },
    {
      "epoch": 0.22032226241367972,
      "grad_norm": 0.34128445386886597,
      "learning_rate": 8.944689075764396e-06,
      "loss": 0.5328,
      "step": 2680
    },
    {
      "epoch": 0.22114436040776062,
      "grad_norm": 0.3424649238586426,
      "learning_rate": 7.73320237312362e-06,
      "loss": 0.5367,
      "step": 2690
    },
    {
      "epoch": 0.2219664584018415,
      "grad_norm": 0.3262079954147339,
      "learning_rate": 6.557112689658629e-06,
      "loss": 0.5535,
      "step": 2700
    },
    {
      "epoch": 0.2227885563959224,
      "grad_norm": 0.3555135726928711,
      "learning_rate": 5.434388823591213e-06,
      "loss": 0.5433,
      "step": 2710
    },
    {
      "epoch": 0.2236106543900033,
      "grad_norm": 0.3241407573223114,
      "learning_rate": 4.382184227495424e-06,
      "loss": 0.546,
      "step": 2720
    },
    {
      "epoch": 0.2244327523840842,
      "grad_norm": 0.35963329672813416,
      "learning_rate": 3.4165749305560203e-06,
      "loss": 0.527,
      "step": 2730
    },
    {
      "epoch": 0.2252548503781651,
      "grad_norm": 0.3601798117160797,
      "learning_rate": 2.5523139221595556e-06,
      "loss": 0.563,
      "step": 2740
    },
    {
      "epoch": 0.22607694837224598,
      "grad_norm": 0.397764652967453,
      "learning_rate": 1.8026057494499821e-06,
      "loss": 0.5403,
      "step": 2750
    },
    {
      "epoch": 0.22689904636632688,
      "grad_norm": 0.5001698732376099,
      "learning_rate": 1.178904772643223e-06,
      "loss": 0.544,
      "step": 2760
    },
    {
      "epoch": 0.22772114436040777,
      "grad_norm": 0.3501325249671936,
      "learning_rate": 6.90740160441754e-07,
      "loss": 0.5313,
      "step": 2770
    },
    {
      "epoch": 0.22854324235448867,
      "grad_norm": 0.34042757749557495,
      "learning_rate": 3.45570299343632e-07,
      "loss": 0.5273,
      "step": 2780
    },
    {
      "epoch": 0.22936534034856956,
      "grad_norm": 0.3410598933696747,
      "learning_rate": 1.486688412425152e-07,
      "loss": 0.5213,
      "step": 2790
    },
    {
      "epoch": 0.22994080894442617,
      "step": 2797,
      "total_flos": 1.2485883762455871e+19,
      "train_loss": 0.04882683965364524,
      "train_runtime": 2858.7108,
      "train_samples_per_second": 11.741,
      "train_steps_per_second": 0.978
    }
  ],
  "logging_steps": 10,
  "max_steps": 2797,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2543,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2485883762455871e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}