{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9201742847747452,
  "eval_steps": 500,
  "global_step": 11193,
  "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": 0.3616321086883545,
      "learning_rate": 2e-05,
      "loss": 0.5553,
      "step": 2550
    },
    {
      "epoch": 0.21045708648470898,
      "grad_norm": 0.34892845153808594,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 2560
    },
    {
      "epoch": 0.21127918447878988,
      "grad_norm": 0.33791273832321167,
      "learning_rate": 2e-05,
      "loss": 0.5362,
      "step": 2570
    },
    {
      "epoch": 0.21210128247287077,
      "grad_norm": 0.34417805075645447,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 2580
    },
    {
      "epoch": 0.21292338046695167,
      "grad_norm": 0.37011924386024475,
      "learning_rate": 2e-05,
      "loss": 0.5126,
      "step": 2590
    },
    {
      "epoch": 0.21374547846103256,
      "grad_norm": 0.39199256896972656,
      "learning_rate": 2e-05,
      "loss": 0.5637,
      "step": 2600
    },
    {
      "epoch": 0.21456757645511346,
      "grad_norm": 0.326111763715744,
      "learning_rate": 2e-05,
      "loss": 0.5169,
      "step": 2610
    },
    {
      "epoch": 0.21538967444919435,
      "grad_norm": 0.33117803931236267,
      "learning_rate": 2e-05,
      "loss": 0.5498,
      "step": 2620
    },
    {
      "epoch": 0.21621177244327525,
      "grad_norm": 0.39731863141059875,
      "learning_rate": 2e-05,
      "loss": 0.5821,
      "step": 2630
    },
    {
      "epoch": 0.21703387043735614,
      "grad_norm": 0.33802589774131775,
      "learning_rate": 2e-05,
      "loss": 0.5496,
      "step": 2640
    },
    {
      "epoch": 0.21785596843143704,
      "grad_norm": 0.36961087584495544,
      "learning_rate": 2e-05,
      "loss": 0.5138,
      "step": 2650
    },
    {
      "epoch": 0.21867806642551793,
      "grad_norm": 0.35770997405052185,
      "learning_rate": 2e-05,
      "loss": 0.5267,
      "step": 2660
    },
    {
      "epoch": 0.21950016441959883,
      "grad_norm": 0.337510347366333,
      "learning_rate": 2e-05,
      "loss": 0.5712,
      "step": 2670
    },
    {
      "epoch": 0.22032226241367972,
      "grad_norm": 0.3274226486682892,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 2680
    },
    {
      "epoch": 0.22114436040776062,
      "grad_norm": 0.41182374954223633,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 2690
    },
    {
      "epoch": 0.2219664584018415,
      "grad_norm": 0.3141595721244812,
      "learning_rate": 2e-05,
      "loss": 0.5567,
      "step": 2700
    },
    {
      "epoch": 0.2227885563959224,
      "grad_norm": 0.35677820444107056,
      "learning_rate": 2e-05,
      "loss": 0.5467,
      "step": 2710
    },
    {
      "epoch": 0.2236106543900033,
      "grad_norm": 0.45076557993888855,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 2720
    },
    {
      "epoch": 0.2244327523840842,
      "grad_norm": 0.3377240002155304,
      "learning_rate": 2e-05,
      "loss": 0.5301,
      "step": 2730
    },
    {
      "epoch": 0.2252548503781651,
      "grad_norm": 0.36233970522880554,
      "learning_rate": 2e-05,
      "loss": 0.5674,
      "step": 2740
    },
    {
      "epoch": 0.22607694837224598,
      "grad_norm": 0.38470005989074707,
      "learning_rate": 2e-05,
      "loss": 0.5442,
      "step": 2750
    },
    {
      "epoch": 0.22689904636632688,
      "grad_norm": 0.4208364486694336,
      "learning_rate": 2e-05,
      "loss": 0.5489,
      "step": 2760
    },
    {
      "epoch": 0.22772114436040777,
      "grad_norm": 0.38245609402656555,
      "learning_rate": 2e-05,
      "loss": 0.5351,
      "step": 2770
    },
    {
      "epoch": 0.22854324235448867,
      "grad_norm": 0.3385438919067383,
      "learning_rate": 2e-05,
      "loss": 0.5315,
      "step": 2780
    },
    {
      "epoch": 0.22936534034856956,
      "grad_norm": 0.36279070377349854,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 2790
    },
    {
      "epoch": 0.23018743834265046,
      "grad_norm": 0.3103238344192505,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 2800
    },
    {
      "epoch": 0.23100953633673132,
      "grad_norm": 0.37252289056777954,
      "learning_rate": 2e-05,
      "loss": 0.511,
      "step": 2810
    },
    {
      "epoch": 0.23183163433081222,
      "grad_norm": 0.38126927614212036,
      "learning_rate": 2e-05,
      "loss": 0.5819,
      "step": 2820
    },
    {
      "epoch": 0.23265373232489311,
      "grad_norm": 0.34690383076667786,
      "learning_rate": 2e-05,
      "loss": 0.5587,
      "step": 2830
    },
    {
      "epoch": 0.233475830318974,
      "grad_norm": 0.43841126561164856,
      "learning_rate": 2e-05,
      "loss": 0.5591,
      "step": 2840
    },
    {
      "epoch": 0.2342979283130549,
      "grad_norm": 0.3482837975025177,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 2850
    },
    {
      "epoch": 0.2351200263071358,
      "grad_norm": 0.34099289774894714,
      "learning_rate": 2e-05,
      "loss": 0.5371,
      "step": 2860
    },
    {
      "epoch": 0.2359421243012167,
      "grad_norm": 0.3340098559856415,
      "learning_rate": 2e-05,
      "loss": 0.5801,
      "step": 2870
    },
    {
      "epoch": 0.2367642222952976,
      "grad_norm": 0.3647443354129791,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 2880
    },
    {
      "epoch": 0.23758632028937848,
      "grad_norm": 0.34515708684921265,
      "learning_rate": 2e-05,
      "loss": 0.5279,
      "step": 2890
    },
    {
      "epoch": 0.23840841828345938,
      "grad_norm": 0.31294670701026917,
      "learning_rate": 2e-05,
      "loss": 0.5585,
      "step": 2900
    },
    {
      "epoch": 0.23923051627754027,
      "grad_norm": 0.32953768968582153,
      "learning_rate": 2e-05,
      "loss": 0.5323,
      "step": 2910
    },
    {
      "epoch": 0.24005261427162117,
      "grad_norm": 0.35687941312789917,
      "learning_rate": 2e-05,
      "loss": 0.5396,
      "step": 2920
    },
    {
      "epoch": 0.24087471226570206,
      "grad_norm": 0.35154175758361816,
      "learning_rate": 2e-05,
      "loss": 0.5532,
      "step": 2930
    },
    {
      "epoch": 0.24169681025978296,
      "grad_norm": 0.3024180829524994,
      "learning_rate": 2e-05,
      "loss": 0.5115,
      "step": 2940
    },
    {
      "epoch": 0.24251890825386385,
      "grad_norm": 0.37728214263916016,
      "learning_rate": 2e-05,
      "loss": 0.5792,
      "step": 2950
    },
    {
      "epoch": 0.24334100624794475,
      "grad_norm": 0.2942720949649811,
      "learning_rate": 2e-05,
      "loss": 0.5171,
      "step": 2960
    },
    {
      "epoch": 0.24416310424202564,
      "grad_norm": 0.29982566833496094,
      "learning_rate": 2e-05,
      "loss": 0.5597,
      "step": 2970
    },
    {
      "epoch": 0.24498520223610654,
      "grad_norm": 0.506549596786499,
      "learning_rate": 2e-05,
      "loss": 0.5301,
      "step": 2980
    },
    {
      "epoch": 0.24580730023018743,
      "grad_norm": 0.3588046431541443,
      "learning_rate": 2e-05,
      "loss": 0.5574,
      "step": 2990
    },
    {
      "epoch": 0.24662939822426833,
      "grad_norm": 0.32193976640701294,
      "learning_rate": 2e-05,
      "loss": 0.5524,
      "step": 3000
    },
    {
      "epoch": 0.24745149621834922,
      "grad_norm": 0.386301726102829,
      "learning_rate": 2e-05,
      "loss": 0.5558,
      "step": 3010
    },
    {
      "epoch": 0.24827359421243012,
      "grad_norm": 0.33212605118751526,
      "learning_rate": 2e-05,
      "loss": 0.5639,
      "step": 3020
    },
    {
      "epoch": 0.249095692206511,
      "grad_norm": 0.3520735204219818,
      "learning_rate": 2e-05,
      "loss": 0.5247,
      "step": 3030
    },
    {
      "epoch": 0.2499177902005919,
      "grad_norm": 0.4220942258834839,
      "learning_rate": 2e-05,
      "loss": 0.5788,
      "step": 3040
    },
    {
      "epoch": 0.2507398881946728,
      "grad_norm": 0.3613930344581604,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 3050
    },
    {
      "epoch": 0.2515619861887537,
      "grad_norm": 0.35459303855895996,
      "learning_rate": 2e-05,
      "loss": 0.5566,
      "step": 3060
    },
    {
      "epoch": 0.2523840841828346,
      "grad_norm": 0.36353611946105957,
      "learning_rate": 2e-05,
      "loss": 0.5542,
      "step": 3070
    },
    {
      "epoch": 0.2532061821769155,
      "grad_norm": 0.35317590832710266,
      "learning_rate": 2e-05,
      "loss": 0.5784,
      "step": 3080
    },
    {
      "epoch": 0.2540282801709964,
      "grad_norm": 0.4215488135814667,
      "learning_rate": 2e-05,
      "loss": 0.5443,
      "step": 3090
    },
    {
      "epoch": 0.2548503781650773,
      "grad_norm": 0.32683974504470825,
      "learning_rate": 2e-05,
      "loss": 0.543,
      "step": 3100
    },
    {
      "epoch": 0.25567247615915817,
      "grad_norm": 0.3655347228050232,
      "learning_rate": 2e-05,
      "loss": 0.5361,
      "step": 3110
    },
    {
      "epoch": 0.25649457415323906,
      "grad_norm": 0.37827545404434204,
      "learning_rate": 2e-05,
      "loss": 0.5533,
      "step": 3120
    },
    {
      "epoch": 0.25731667214731996,
      "grad_norm": 0.34333717823028564,
      "learning_rate": 2e-05,
      "loss": 0.5558,
      "step": 3130
    },
    {
      "epoch": 0.25813877014140085,
      "grad_norm": 0.3396783471107483,
      "learning_rate": 2e-05,
      "loss": 0.5436,
      "step": 3140
    },
    {
      "epoch": 0.25896086813548175,
      "grad_norm": 0.37508583068847656,
      "learning_rate": 2e-05,
      "loss": 0.5169,
      "step": 3150
    },
    {
      "epoch": 0.25978296612956264,
      "grad_norm": 0.3585703372955322,
      "learning_rate": 2e-05,
      "loss": 0.5422,
      "step": 3160
    },
    {
      "epoch": 0.26060506412364354,
      "grad_norm": 0.3634900450706482,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 3170
    },
    {
      "epoch": 0.26142716211772443,
      "grad_norm": 0.34407326579093933,
      "learning_rate": 2e-05,
      "loss": 0.5208,
      "step": 3180
    },
    {
      "epoch": 0.2622492601118053,
      "grad_norm": 0.41253870725631714,
      "learning_rate": 2e-05,
      "loss": 0.5141,
      "step": 3190
    },
    {
      "epoch": 0.2630713581058862,
      "grad_norm": 0.3736776113510132,
      "learning_rate": 2e-05,
      "loss": 0.5594,
      "step": 3200
    },
    {
      "epoch": 0.2638934560999671,
      "grad_norm": 0.3139011859893799,
      "learning_rate": 2e-05,
      "loss": 0.5385,
      "step": 3210
    },
    {
      "epoch": 0.264715554094048,
      "grad_norm": 0.3514135777950287,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 3220
    },
    {
      "epoch": 0.2655376520881289,
      "grad_norm": 0.38396161794662476,
      "learning_rate": 2e-05,
      "loss": 0.521,
      "step": 3230
    },
    {
      "epoch": 0.2663597500822098,
      "grad_norm": 0.32594597339630127,
      "learning_rate": 2e-05,
      "loss": 0.5404,
      "step": 3240
    },
    {
      "epoch": 0.2671818480762907,
      "grad_norm": 0.3474389612674713,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 3250
    },
    {
      "epoch": 0.2680039460703716,
      "grad_norm": 0.4189068675041199,
      "learning_rate": 2e-05,
      "loss": 0.549,
      "step": 3260
    },
    {
      "epoch": 0.2688260440644525,
      "grad_norm": 0.5115881562232971,
      "learning_rate": 2e-05,
      "loss": 0.538,
      "step": 3270
    },
    {
      "epoch": 0.2696481420585334,
      "grad_norm": 0.31304091215133667,
      "learning_rate": 2e-05,
      "loss": 0.5662,
      "step": 3280
    },
    {
      "epoch": 0.2704702400526143,
      "grad_norm": 0.2987343370914459,
      "learning_rate": 2e-05,
      "loss": 0.5443,
      "step": 3290
    },
    {
      "epoch": 0.27129233804669517,
      "grad_norm": 0.33361944556236267,
      "learning_rate": 2e-05,
      "loss": 0.5654,
      "step": 3300
    },
    {
      "epoch": 0.27211443604077606,
      "grad_norm": 0.330786794424057,
      "learning_rate": 2e-05,
      "loss": 0.5063,
      "step": 3310
    },
    {
      "epoch": 0.27293653403485696,
      "grad_norm": 0.33479249477386475,
      "learning_rate": 2e-05,
      "loss": 0.5536,
      "step": 3320
    },
    {
      "epoch": 0.27375863202893785,
      "grad_norm": 0.3566909432411194,
      "learning_rate": 2e-05,
      "loss": 0.5505,
      "step": 3330
    },
    {
      "epoch": 0.27458073002301875,
      "grad_norm": 0.3387537896633148,
      "learning_rate": 2e-05,
      "loss": 0.5342,
      "step": 3340
    },
    {
      "epoch": 0.27540282801709964,
      "grad_norm": 0.36577555537223816,
      "learning_rate": 2e-05,
      "loss": 0.5352,
      "step": 3350
    },
    {
      "epoch": 0.27622492601118054,
      "grad_norm": 0.35958272218704224,
      "learning_rate": 2e-05,
      "loss": 0.5746,
      "step": 3360
    },
    {
      "epoch": 0.27704702400526143,
      "grad_norm": 0.391726553440094,
      "learning_rate": 2e-05,
      "loss": 0.5379,
      "step": 3370
    },
    {
      "epoch": 0.27786912199934233,
      "grad_norm": 0.3955790400505066,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 3380
    },
    {
      "epoch": 0.2786912199934232,
      "grad_norm": 0.36854982376098633,
      "learning_rate": 2e-05,
      "loss": 0.5418,
      "step": 3390
    },
    {
      "epoch": 0.2795133179875041,
      "grad_norm": 0.366266131401062,
      "learning_rate": 2e-05,
      "loss": 0.5826,
      "step": 3400
    },
    {
      "epoch": 0.280335415981585,
      "grad_norm": 0.4051331579685211,
      "learning_rate": 2e-05,
      "loss": 0.5352,
      "step": 3410
    },
    {
      "epoch": 0.2811575139756659,
      "grad_norm": 0.3959217369556427,
      "learning_rate": 2e-05,
      "loss": 0.5248,
      "step": 3420
    },
    {
      "epoch": 0.2819796119697468,
      "grad_norm": 0.3575945496559143,
      "learning_rate": 2e-05,
      "loss": 0.5547,
      "step": 3430
    },
    {
      "epoch": 0.2828017099638277,
      "grad_norm": 0.35974130034446716,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 3440
    },
    {
      "epoch": 0.2836238079579086,
      "grad_norm": 0.31665098667144775,
      "learning_rate": 2e-05,
      "loss": 0.5389,
      "step": 3450
    },
    {
      "epoch": 0.2844459059519895,
      "grad_norm": 0.3459745943546295,
      "learning_rate": 2e-05,
      "loss": 0.5666,
      "step": 3460
    },
    {
      "epoch": 0.2852680039460704,
      "grad_norm": 0.47385168075561523,
      "learning_rate": 2e-05,
      "loss": 0.5415,
      "step": 3470
    },
    {
      "epoch": 0.2860901019401513,
      "grad_norm": 0.29525819420814514,
      "learning_rate": 2e-05,
      "loss": 0.5201,
      "step": 3480
    },
    {
      "epoch": 0.28691219993423217,
      "grad_norm": 0.32592856884002686,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 3490
    },
    {
      "epoch": 0.28773429792831307,
      "grad_norm": 0.3753909766674042,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 3500
    },
    {
      "epoch": 0.28855639592239396,
      "grad_norm": 0.35471951961517334,
      "learning_rate": 2e-05,
      "loss": 0.5548,
      "step": 3510
    },
    {
      "epoch": 0.28937849391647485,
      "grad_norm": 0.31236377358436584,
      "learning_rate": 2e-05,
      "loss": 0.5608,
      "step": 3520
    },
    {
      "epoch": 0.29020059191055575,
      "grad_norm": 0.3410734534263611,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 3530
    },
    {
      "epoch": 0.29102268990463664,
      "grad_norm": 0.362771213054657,
      "learning_rate": 2e-05,
      "loss": 0.5247,
      "step": 3540
    },
    {
      "epoch": 0.29184478789871754,
      "grad_norm": 0.36238428950309753,
      "learning_rate": 2e-05,
      "loss": 0.5464,
      "step": 3550
    },
    {
      "epoch": 0.29266688589279843,
      "grad_norm": 0.38895243406295776,
      "learning_rate": 2e-05,
      "loss": 0.519,
      "step": 3560
    },
    {
      "epoch": 0.29348898388687933,
      "grad_norm": 0.32610785961151123,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 3570
    },
    {
      "epoch": 0.2943110818809602,
      "grad_norm": 0.4277496337890625,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 3580
    },
    {
      "epoch": 0.2951331798750411,
      "grad_norm": 0.35077282786369324,
      "learning_rate": 2e-05,
      "loss": 0.5421,
      "step": 3590
    },
    {
      "epoch": 0.295955277869122,
      "grad_norm": 0.3360082507133484,
      "learning_rate": 2e-05,
      "loss": 0.5236,
      "step": 3600
    },
    {
      "epoch": 0.2967773758632029,
      "grad_norm": 0.41645798087120056,
      "learning_rate": 2e-05,
      "loss": 0.5403,
      "step": 3610
    },
    {
      "epoch": 0.2975994738572838,
      "grad_norm": 0.3737342953681946,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 3620
    },
    {
      "epoch": 0.2984215718513647,
      "grad_norm": 0.45222207903862,
      "learning_rate": 2e-05,
      "loss": 0.5524,
      "step": 3630
    },
    {
      "epoch": 0.2992436698454456,
      "grad_norm": 0.3541415333747864,
      "learning_rate": 2e-05,
      "loss": 0.5506,
      "step": 3640
    },
    {
      "epoch": 0.3000657678395265,
      "grad_norm": 0.3760773837566376,
      "learning_rate": 2e-05,
      "loss": 0.5374,
      "step": 3650
    },
    {
      "epoch": 0.3008878658336074,
      "grad_norm": 0.374584436416626,
      "learning_rate": 2e-05,
      "loss": 0.5814,
      "step": 3660
    },
    {
      "epoch": 0.3017099638276883,
      "grad_norm": 0.35466334223747253,
      "learning_rate": 2e-05,
      "loss": 0.5587,
      "step": 3670
    },
    {
      "epoch": 0.30253206182176917,
      "grad_norm": 0.3851785659790039,
      "learning_rate": 2e-05,
      "loss": 0.5636,
      "step": 3680
    },
    {
      "epoch": 0.30335415981585007,
      "grad_norm": 0.3262382447719574,
      "learning_rate": 2e-05,
      "loss": 0.5509,
      "step": 3690
    },
    {
      "epoch": 0.30417625780993096,
      "grad_norm": 0.35221946239471436,
      "learning_rate": 2e-05,
      "loss": 0.5395,
      "step": 3700
    },
    {
      "epoch": 0.30499835580401186,
      "grad_norm": 0.33840522170066833,
      "learning_rate": 2e-05,
      "loss": 0.4874,
      "step": 3710
    },
    {
      "epoch": 0.30582045379809275,
      "grad_norm": 0.3443030118942261,
      "learning_rate": 2e-05,
      "loss": 0.5604,
      "step": 3720
    },
    {
      "epoch": 0.30664255179217365,
      "grad_norm": 0.34090104699134827,
      "learning_rate": 2e-05,
      "loss": 0.5368,
      "step": 3730
    },
    {
      "epoch": 0.30746464978625454,
      "grad_norm": 0.4112664461135864,
      "learning_rate": 2e-05,
      "loss": 0.5413,
      "step": 3740
    },
    {
      "epoch": 0.30828674778033544,
      "grad_norm": 0.36865583062171936,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 3750
    },
    {
      "epoch": 0.30910884577441633,
      "grad_norm": 0.2920708954334259,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 3760
    },
    {
      "epoch": 0.3099309437684972,
      "grad_norm": 0.3259790539741516,
      "learning_rate": 2e-05,
      "loss": 0.5138,
      "step": 3770
    },
    {
      "epoch": 0.3107530417625781,
      "grad_norm": 0.3496236801147461,
      "learning_rate": 2e-05,
      "loss": 0.5521,
      "step": 3780
    },
    {
      "epoch": 0.311575139756659,
      "grad_norm": 0.36148640513420105,
      "learning_rate": 2e-05,
      "loss": 0.5167,
      "step": 3790
    },
    {
      "epoch": 0.3123972377507399,
      "grad_norm": 0.3326259255409241,
      "learning_rate": 2e-05,
      "loss": 0.512,
      "step": 3800
    },
    {
      "epoch": 0.3132193357448208,
      "grad_norm": 0.3559156656265259,
      "learning_rate": 2e-05,
      "loss": 0.5413,
      "step": 3810
    },
    {
      "epoch": 0.3140414337389017,
      "grad_norm": 0.3168695867061615,
      "learning_rate": 2e-05,
      "loss": 0.5592,
      "step": 3820
    },
    {
      "epoch": 0.3148635317329826,
      "grad_norm": 0.3359394967556,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 3830
    },
    {
      "epoch": 0.3156856297270635,
      "grad_norm": 0.3557104766368866,
      "learning_rate": 2e-05,
      "loss": 0.5495,
      "step": 3840
    },
    {
      "epoch": 0.3165077277211444,
      "grad_norm": 0.3635749816894531,
      "learning_rate": 2e-05,
      "loss": 0.543,
      "step": 3850
    },
    {
      "epoch": 0.3173298257152253,
      "grad_norm": 0.35071372985839844,
      "learning_rate": 2e-05,
      "loss": 0.5198,
      "step": 3860
    },
    {
      "epoch": 0.3181519237093062,
      "grad_norm": 0.38483187556266785,
      "learning_rate": 2e-05,
      "loss": 0.572,
      "step": 3870
    },
    {
      "epoch": 0.31897402170338707,
      "grad_norm": 0.33359649777412415,
      "learning_rate": 2e-05,
      "loss": 0.5426,
      "step": 3880
    },
    {
      "epoch": 0.31979611969746796,
      "grad_norm": 0.31382572650909424,
      "learning_rate": 2e-05,
      "loss": 0.5469,
      "step": 3890
    },
    {
      "epoch": 0.32061821769154886,
      "grad_norm": 0.35558342933654785,
      "learning_rate": 2e-05,
      "loss": 0.5293,
      "step": 3900
    },
    {
      "epoch": 0.32144031568562975,
      "grad_norm": 0.35809141397476196,
      "learning_rate": 2e-05,
      "loss": 0.5275,
      "step": 3910
    },
    {
      "epoch": 0.32226241367971065,
      "grad_norm": 0.34361934661865234,
      "learning_rate": 2e-05,
      "loss": 0.5106,
      "step": 3920
    },
    {
      "epoch": 0.32308451167379154,
      "grad_norm": 0.338867723941803,
      "learning_rate": 2e-05,
      "loss": 0.5193,
      "step": 3930
    },
    {
      "epoch": 0.32390660966787244,
      "grad_norm": 0.34359851479530334,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 3940
    },
    {
      "epoch": 0.32472870766195333,
      "grad_norm": 0.3797745406627655,
      "learning_rate": 2e-05,
      "loss": 0.5256,
      "step": 3950
    },
    {
      "epoch": 0.3255508056560342,
      "grad_norm": 0.3485299348831177,
      "learning_rate": 2e-05,
      "loss": 0.5275,
      "step": 3960
    },
    {
      "epoch": 0.3263729036501151,
      "grad_norm": 0.33082035183906555,
      "learning_rate": 2e-05,
      "loss": 0.5398,
      "step": 3970
    },
    {
      "epoch": 0.327195001644196,
      "grad_norm": 0.3318018913269043,
      "learning_rate": 2e-05,
      "loss": 0.5226,
      "step": 3980
    },
    {
      "epoch": 0.3280170996382769,
      "grad_norm": 0.3452273905277252,
      "learning_rate": 2e-05,
      "loss": 0.5526,
      "step": 3990
    },
    {
      "epoch": 0.3288391976323578,
      "grad_norm": 0.32671037316322327,
      "learning_rate": 2e-05,
      "loss": 0.539,
      "step": 4000
    },
    {
      "epoch": 0.32966129562643864,
      "grad_norm": 0.44799134135246277,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 4010
    },
    {
      "epoch": 0.33048339362051954,
      "grad_norm": 0.3295125663280487,
      "learning_rate": 2e-05,
      "loss": 0.5346,
      "step": 4020
    },
    {
      "epoch": 0.33130549161460043,
      "grad_norm": 0.32488885521888733,
      "learning_rate": 2e-05,
      "loss": 0.531,
      "step": 4030
    },
    {
      "epoch": 0.33212758960868133,
      "grad_norm": 0.3428230881690979,
      "learning_rate": 2e-05,
      "loss": 0.5293,
      "step": 4040
    },
    {
      "epoch": 0.3329496876027622,
      "grad_norm": 0.34698739647865295,
      "learning_rate": 2e-05,
      "loss": 0.5167,
      "step": 4050
    },
    {
      "epoch": 0.3337717855968431,
      "grad_norm": 0.47139301896095276,
      "learning_rate": 2e-05,
      "loss": 0.5732,
      "step": 4060
    },
    {
      "epoch": 0.334593883590924,
      "grad_norm": 0.3707517683506012,
      "learning_rate": 2e-05,
      "loss": 0.5125,
      "step": 4070
    },
    {
      "epoch": 0.3354159815850049,
      "grad_norm": 0.3662269115447998,
      "learning_rate": 2e-05,
      "loss": 0.5531,
      "step": 4080
    },
    {
      "epoch": 0.3362380795790858,
      "grad_norm": 0.34573638439178467,
      "learning_rate": 2e-05,
      "loss": 0.4987,
      "step": 4090
    },
    {
      "epoch": 0.3370601775731667,
      "grad_norm": 0.35743385553359985,
      "learning_rate": 2e-05,
      "loss": 0.544,
      "step": 4100
    },
    {
      "epoch": 0.3378822755672476,
      "grad_norm": 0.3641563057899475,
      "learning_rate": 2e-05,
      "loss": 0.5418,
      "step": 4110
    },
    {
      "epoch": 0.3387043735613285,
      "grad_norm": 0.3505076467990875,
      "learning_rate": 2e-05,
      "loss": 0.5492,
      "step": 4120
    },
    {
      "epoch": 0.3395264715554094,
      "grad_norm": 0.35745880007743835,
      "learning_rate": 2e-05,
      "loss": 0.5515,
      "step": 4130
    },
    {
      "epoch": 0.3403485695494903,
      "grad_norm": 0.41468533873558044,
      "learning_rate": 2e-05,
      "loss": 0.52,
      "step": 4140
    },
    {
      "epoch": 0.34117066754357117,
      "grad_norm": 0.3421972692012787,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 4150
    },
    {
      "epoch": 0.34199276553765207,
      "grad_norm": 0.3675749897956848,
      "learning_rate": 2e-05,
      "loss": 0.5501,
      "step": 4160
    },
    {
      "epoch": 0.34281486353173296,
      "grad_norm": 0.39348721504211426,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 4170
    },
    {
      "epoch": 0.34363696152581386,
      "grad_norm": 0.34833019971847534,
      "learning_rate": 2e-05,
      "loss": 0.53,
      "step": 4180
    },
    {
      "epoch": 0.34445905951989475,
      "grad_norm": 0.36116981506347656,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 4190
    },
    {
      "epoch": 0.34528115751397565,
      "grad_norm": 0.3307689130306244,
      "learning_rate": 2e-05,
      "loss": 0.516,
      "step": 4200
    },
    {
      "epoch": 0.34610325550805654,
      "grad_norm": 0.3135853409767151,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 4210
    },
    {
      "epoch": 0.34692535350213743,
      "grad_norm": 0.3765367269515991,
      "learning_rate": 2e-05,
      "loss": 0.5044,
      "step": 4220
    },
    {
      "epoch": 0.34774745149621833,
      "grad_norm": 0.3693373501300812,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 4230
    },
    {
      "epoch": 0.3485695494902992,
      "grad_norm": 0.37614792585372925,
      "learning_rate": 2e-05,
      "loss": 0.5231,
      "step": 4240
    },
    {
      "epoch": 0.3493916474843801,
      "grad_norm": 0.32215288281440735,
      "learning_rate": 2e-05,
      "loss": 0.5654,
      "step": 4250
    },
    {
      "epoch": 0.350213745478461,
      "grad_norm": 0.4110385775566101,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 4260
    },
    {
      "epoch": 0.3510358434725419,
      "grad_norm": 0.34171709418296814,
      "learning_rate": 2e-05,
      "loss": 0.5149,
      "step": 4270
    },
    {
      "epoch": 0.3518579414666228,
      "grad_norm": 0.35466277599334717,
      "learning_rate": 2e-05,
      "loss": 0.5064,
      "step": 4280
    },
    {
      "epoch": 0.3526800394607037,
      "grad_norm": 0.3796570897102356,
      "learning_rate": 2e-05,
      "loss": 0.5486,
      "step": 4290
    },
    {
      "epoch": 0.3535021374547846,
      "grad_norm": 0.32320407032966614,
      "learning_rate": 2e-05,
      "loss": 0.5647,
      "step": 4300
    },
    {
      "epoch": 0.3543242354488655,
      "grad_norm": 0.3509526550769806,
      "learning_rate": 2e-05,
      "loss": 0.5017,
      "step": 4310
    },
    {
      "epoch": 0.3551463334429464,
      "grad_norm": 0.34390702843666077,
      "learning_rate": 2e-05,
      "loss": 0.5591,
      "step": 4320
    },
    {
      "epoch": 0.3559684314370273,
      "grad_norm": 0.3451758623123169,
      "learning_rate": 2e-05,
      "loss": 0.5305,
      "step": 4330
    },
    {
      "epoch": 0.35679052943110817,
      "grad_norm": 0.40349626541137695,
      "learning_rate": 2e-05,
      "loss": 0.5279,
      "step": 4340
    },
    {
      "epoch": 0.35761262742518907,
      "grad_norm": 0.32908904552459717,
      "learning_rate": 2e-05,
      "loss": 0.5327,
      "step": 4350
    },
    {
      "epoch": 0.35843472541926996,
      "grad_norm": 0.3217417597770691,
      "learning_rate": 2e-05,
      "loss": 0.5414,
      "step": 4360
    },
    {
      "epoch": 0.35925682341335086,
      "grad_norm": 0.37074434757232666,
      "learning_rate": 2e-05,
      "loss": 0.5376,
      "step": 4370
    },
    {
      "epoch": 0.36007892140743175,
      "grad_norm": 0.5618411302566528,
      "learning_rate": 2e-05,
      "loss": 0.5039,
      "step": 4380
    },
    {
      "epoch": 0.36090101940151265,
      "grad_norm": 0.3570129871368408,
      "learning_rate": 2e-05,
      "loss": 0.5209,
      "step": 4390
    },
    {
      "epoch": 0.36172311739559354,
      "grad_norm": 0.34965282678604126,
      "learning_rate": 2e-05,
      "loss": 0.5073,
      "step": 4400
    },
    {
      "epoch": 0.36254521538967444,
      "grad_norm": 0.3438706696033478,
      "learning_rate": 2e-05,
      "loss": 0.512,
      "step": 4410
    },
    {
      "epoch": 0.36336731338375533,
      "grad_norm": 0.3440127670764923,
      "learning_rate": 2e-05,
      "loss": 0.5234,
      "step": 4420
    },
    {
      "epoch": 0.3641894113778362,
      "grad_norm": 0.3571926951408386,
      "learning_rate": 2e-05,
      "loss": 0.5249,
      "step": 4430
    },
    {
      "epoch": 0.3650115093719171,
      "grad_norm": 0.4331763684749603,
      "learning_rate": 2e-05,
      "loss": 0.5334,
      "step": 4440
    },
    {
      "epoch": 0.365833607365998,
      "grad_norm": 0.36819732189178467,
      "learning_rate": 2e-05,
      "loss": 0.5538,
      "step": 4450
    },
    {
      "epoch": 0.3666557053600789,
      "grad_norm": 0.36837688088417053,
      "learning_rate": 2e-05,
      "loss": 0.5046,
      "step": 4460
    },
    {
      "epoch": 0.3674778033541598,
      "grad_norm": 0.35506975650787354,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 4470
    },
    {
      "epoch": 0.3682999013482407,
      "grad_norm": 0.3211697041988373,
      "learning_rate": 2e-05,
      "loss": 0.5151,
      "step": 4480
    },
    {
      "epoch": 0.3691219993423216,
      "grad_norm": 0.32097750902175903,
      "learning_rate": 2e-05,
      "loss": 0.5478,
      "step": 4490
    },
    {
      "epoch": 0.3699440973364025,
      "grad_norm": 0.35200196504592896,
      "learning_rate": 2e-05,
      "loss": 0.5249,
      "step": 4500
    },
    {
      "epoch": 0.3707661953304834,
      "grad_norm": 0.3686334192752838,
      "learning_rate": 2e-05,
      "loss": 0.5445,
      "step": 4510
    },
    {
      "epoch": 0.3715882933245643,
      "grad_norm": 0.3223293423652649,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 4520
    },
    {
      "epoch": 0.3724103913186452,
      "grad_norm": 0.3505318760871887,
      "learning_rate": 2e-05,
      "loss": 0.5244,
      "step": 4530
    },
    {
      "epoch": 0.37323248931272607,
      "grad_norm": 0.37446868419647217,
      "learning_rate": 2e-05,
      "loss": 0.5585,
      "step": 4540
    },
    {
      "epoch": 0.37405458730680696,
      "grad_norm": 0.3326321542263031,
      "learning_rate": 2e-05,
      "loss": 0.541,
      "step": 4550
    },
    {
      "epoch": 0.37487668530088786,
      "grad_norm": 0.3477605879306793,
      "learning_rate": 2e-05,
      "loss": 0.5146,
      "step": 4560
    },
    {
      "epoch": 0.37569878329496875,
      "grad_norm": 0.36050575971603394,
      "learning_rate": 2e-05,
      "loss": 0.5205,
      "step": 4570
    },
    {
      "epoch": 0.37652088128904965,
      "grad_norm": 0.33603909611701965,
      "learning_rate": 2e-05,
      "loss": 0.5254,
      "step": 4580
    },
    {
      "epoch": 0.37734297928313054,
      "grad_norm": 0.3777899742126465,
      "learning_rate": 2e-05,
      "loss": 0.5236,
      "step": 4590
    },
    {
      "epoch": 0.37816507727721144,
      "grad_norm": 0.3817446827888489,
      "learning_rate": 2e-05,
      "loss": 0.5453,
      "step": 4600
    },
    {
      "epoch": 0.37898717527129233,
      "grad_norm": 0.42952242493629456,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 4610
    },
    {
      "epoch": 0.3798092732653732,
      "grad_norm": 0.3556438386440277,
      "learning_rate": 2e-05,
      "loss": 0.5129,
      "step": 4620
    },
    {
      "epoch": 0.3806313712594541,
      "grad_norm": 0.3687664866447449,
      "learning_rate": 2e-05,
      "loss": 0.5171,
      "step": 4630
    },
    {
      "epoch": 0.381453469253535,
      "grad_norm": 0.33991408348083496,
      "learning_rate": 2e-05,
      "loss": 0.5303,
      "step": 4640
    },
    {
      "epoch": 0.3822755672476159,
      "grad_norm": 0.3397250175476074,
      "learning_rate": 2e-05,
      "loss": 0.5286,
      "step": 4650
    },
    {
      "epoch": 0.3830976652416968,
      "grad_norm": 0.34458306431770325,
      "learning_rate": 2e-05,
      "loss": 0.5626,
      "step": 4660
    },
    {
      "epoch": 0.3839197632357777,
      "grad_norm": 0.32431814074516296,
      "learning_rate": 2e-05,
      "loss": 0.5157,
      "step": 4670
    },
    {
      "epoch": 0.3847418612298586,
      "grad_norm": 0.3668116331100464,
      "learning_rate": 2e-05,
      "loss": 0.5457,
      "step": 4680
    },
    {
      "epoch": 0.3855639592239395,
      "grad_norm": 0.36735913157463074,
      "learning_rate": 2e-05,
      "loss": 0.535,
      "step": 4690
    },
    {
      "epoch": 0.3863860572180204,
      "grad_norm": 0.35322412848472595,
      "learning_rate": 2e-05,
      "loss": 0.5197,
      "step": 4700
    },
    {
      "epoch": 0.3872081552121013,
      "grad_norm": 0.3773740828037262,
      "learning_rate": 2e-05,
      "loss": 0.5211,
      "step": 4710
    },
    {
      "epoch": 0.3880302532061822,
      "grad_norm": 0.335239052772522,
      "learning_rate": 2e-05,
      "loss": 0.4863,
      "step": 4720
    },
    {
      "epoch": 0.38885235120026307,
      "grad_norm": 0.43967121839523315,
      "learning_rate": 2e-05,
      "loss": 0.4973,
      "step": 4730
    },
    {
      "epoch": 0.38967444919434396,
      "grad_norm": 0.33292362093925476,
      "learning_rate": 2e-05,
      "loss": 0.5264,
      "step": 4740
    },
    {
      "epoch": 0.39049654718842486,
      "grad_norm": 0.3542301058769226,
      "learning_rate": 2e-05,
      "loss": 0.543,
      "step": 4750
    },
    {
      "epoch": 0.39131864518250575,
      "grad_norm": 0.3485114872455597,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 4760
    },
    {
      "epoch": 0.39214074317658665,
      "grad_norm": 0.35498690605163574,
      "learning_rate": 2e-05,
      "loss": 0.5293,
      "step": 4770
    },
    {
      "epoch": 0.39296284117066754,
      "grad_norm": 0.3507038950920105,
      "learning_rate": 2e-05,
      "loss": 0.5295,
      "step": 4780
    },
    {
      "epoch": 0.39378493916474844,
      "grad_norm": 0.32928594946861267,
      "learning_rate": 2e-05,
      "loss": 0.5537,
      "step": 4790
    },
    {
      "epoch": 0.39460703715882933,
      "grad_norm": 0.3561805486679077,
      "learning_rate": 2e-05,
      "loss": 0.521,
      "step": 4800
    },
    {
      "epoch": 0.3954291351529102,
      "grad_norm": 0.3253518342971802,
      "learning_rate": 2e-05,
      "loss": 0.5484,
      "step": 4810
    },
    {
      "epoch": 0.3962512331469911,
      "grad_norm": 0.3671138882637024,
      "learning_rate": 2e-05,
      "loss": 0.5726,
      "step": 4820
    },
    {
      "epoch": 0.397073331141072,
      "grad_norm": 0.3315047025680542,
      "learning_rate": 2e-05,
      "loss": 0.5463,
      "step": 4830
    },
    {
      "epoch": 0.3978954291351529,
      "grad_norm": 0.33220726251602173,
      "learning_rate": 2e-05,
      "loss": 0.485,
      "step": 4840
    },
    {
      "epoch": 0.3987175271292338,
      "grad_norm": 0.3390117287635803,
      "learning_rate": 2e-05,
      "loss": 0.5407,
      "step": 4850
    },
    {
      "epoch": 0.3995396251233147,
      "grad_norm": 0.34227901697158813,
      "learning_rate": 2e-05,
      "loss": 0.5372,
      "step": 4860
    },
    {
      "epoch": 0.4003617231173956,
      "grad_norm": 0.36416247487068176,
      "learning_rate": 2e-05,
      "loss": 0.5212,
      "step": 4870
    },
    {
      "epoch": 0.4011838211114765,
      "grad_norm": 0.31804022192955017,
      "learning_rate": 2e-05,
      "loss": 0.5,
      "step": 4880
    },
    {
      "epoch": 0.4020059191055574,
      "grad_norm": 0.350694864988327,
      "learning_rate": 2e-05,
      "loss": 0.5213,
      "step": 4890
    },
    {
      "epoch": 0.4028280170996383,
      "grad_norm": 0.3451733887195587,
      "learning_rate": 2e-05,
      "loss": 0.5365,
      "step": 4900
    },
    {
      "epoch": 0.4036501150937192,
      "grad_norm": 0.3070090711116791,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 4910
    },
    {
      "epoch": 0.40447221308780007,
      "grad_norm": 0.33814436197280884,
      "learning_rate": 2e-05,
      "loss": 0.5272,
      "step": 4920
    },
    {
      "epoch": 0.40529431108188096,
      "grad_norm": 0.3543502986431122,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 4930
    },
    {
      "epoch": 0.40611640907596186,
      "grad_norm": 0.33732470870018005,
      "learning_rate": 2e-05,
      "loss": 0.5044,
      "step": 4940
    },
    {
      "epoch": 0.40693850707004275,
      "grad_norm": 0.34642818570137024,
      "learning_rate": 2e-05,
      "loss": 0.5282,
      "step": 4950
    },
    {
      "epoch": 0.40776060506412365,
      "grad_norm": 0.34865498542785645,
      "learning_rate": 2e-05,
      "loss": 0.5013,
      "step": 4960
    },
    {
      "epoch": 0.40858270305820454,
      "grad_norm": 0.34857216477394104,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 4970
    },
    {
      "epoch": 0.40940480105228544,
      "grad_norm": 0.3917066752910614,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 4980
    },
    {
      "epoch": 0.41022689904636633,
      "grad_norm": 0.4312105178833008,
      "learning_rate": 2e-05,
      "loss": 0.5726,
      "step": 4990
    },
    {
      "epoch": 0.41104899704044723,
      "grad_norm": 0.40886908769607544,
      "learning_rate": 2e-05,
      "loss": 0.4993,
      "step": 5000
    },
    {
      "epoch": 0.4118710950345281,
      "grad_norm": 0.7453745603561401,
      "learning_rate": 2e-05,
      "loss": 0.5222,
      "step": 5010
    },
    {
      "epoch": 0.412693193028609,
      "grad_norm": 0.32858678698539734,
      "learning_rate": 2e-05,
      "loss": 0.5316,
      "step": 5020
    },
    {
      "epoch": 0.4135152910226899,
      "grad_norm": 0.3431444764137268,
      "learning_rate": 2e-05,
      "loss": 0.5494,
      "step": 5030
    },
    {
      "epoch": 0.4143373890167708,
      "grad_norm": 0.3860490024089813,
      "learning_rate": 2e-05,
      "loss": 0.5413,
      "step": 5040
    },
    {
      "epoch": 0.4151594870108517,
      "grad_norm": 0.34055987000465393,
      "learning_rate": 2e-05,
      "loss": 0.5274,
      "step": 5050
    },
    {
      "epoch": 0.4159815850049326,
      "grad_norm": 0.3648587763309479,
      "learning_rate": 2e-05,
      "loss": 0.5389,
      "step": 5060
    },
    {
      "epoch": 0.4168036829990135,
      "grad_norm": 0.4195209741592407,
      "learning_rate": 2e-05,
      "loss": 0.5467,
      "step": 5070
    },
    {
      "epoch": 0.4176257809930944,
      "grad_norm": 0.36065131425857544,
      "learning_rate": 2e-05,
      "loss": 0.5558,
      "step": 5080
    },
    {
      "epoch": 0.4184478789871753,
      "grad_norm": 0.3311932384967804,
      "learning_rate": 2e-05,
      "loss": 0.5286,
      "step": 5090
    },
    {
      "epoch": 0.4192699769812562,
      "grad_norm": 0.37731996178627014,
      "learning_rate": 2e-05,
      "loss": 0.5856,
      "step": 5100
    },
    {
      "epoch": 0.42009207497533707,
      "grad_norm": 0.4098602533340454,
      "learning_rate": 2e-05,
      "loss": 0.5198,
      "step": 5110
    },
    {
      "epoch": 0.42091417296941797,
      "grad_norm": 0.4512919485569,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 5120
    },
    {
      "epoch": 0.42173627096349886,
      "grad_norm": 0.3168100416660309,
      "learning_rate": 2e-05,
      "loss": 0.5616,
      "step": 5130
    },
    {
      "epoch": 0.42255836895757976,
      "grad_norm": 0.303036630153656,
      "learning_rate": 2e-05,
      "loss": 0.5064,
      "step": 5140
    },
    {
      "epoch": 0.42338046695166065,
      "grad_norm": 0.35976848006248474,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 5150
    },
    {
      "epoch": 0.42420256494574154,
      "grad_norm": 0.347128301858902,
      "learning_rate": 2e-05,
      "loss": 0.5357,
      "step": 5160
    },
    {
      "epoch": 0.42502466293982244,
      "grad_norm": 0.35909566283226013,
      "learning_rate": 2e-05,
      "loss": 0.5468,
      "step": 5170
    },
    {
      "epoch": 0.42584676093390333,
      "grad_norm": 0.3777230978012085,
      "learning_rate": 2e-05,
      "loss": 0.5426,
      "step": 5180
    },
    {
      "epoch": 0.42666885892798423,
      "grad_norm": 0.31758278608322144,
      "learning_rate": 2e-05,
      "loss": 0.5275,
      "step": 5190
    },
    {
      "epoch": 0.4274909569220651,
      "grad_norm": 0.31867358088493347,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 5200
    },
    {
      "epoch": 0.428313054916146,
      "grad_norm": 0.38007357716560364,
      "learning_rate": 2e-05,
      "loss": 0.506,
      "step": 5210
    },
    {
      "epoch": 0.4291351529102269,
      "grad_norm": 0.3436846435070038,
      "learning_rate": 2e-05,
      "loss": 0.5587,
      "step": 5220
    },
    {
      "epoch": 0.4299572509043078,
      "grad_norm": 0.37001490592956543,
      "learning_rate": 2e-05,
      "loss": 0.5335,
      "step": 5230
    },
    {
      "epoch": 0.4307793488983887,
      "grad_norm": 0.38903769850730896,
      "learning_rate": 2e-05,
      "loss": 0.5337,
      "step": 5240
    },
    {
      "epoch": 0.4316014468924696,
      "grad_norm": 0.385413259267807,
      "learning_rate": 2e-05,
      "loss": 0.5354,
      "step": 5250
    },
    {
      "epoch": 0.4324235448865505,
      "grad_norm": 0.3405441343784332,
      "learning_rate": 2e-05,
      "loss": 0.5634,
      "step": 5260
    },
    {
      "epoch": 0.4332456428806314,
      "grad_norm": 0.34991002082824707,
      "learning_rate": 2e-05,
      "loss": 0.5181,
      "step": 5270
    },
    {
      "epoch": 0.4340677408747123,
      "grad_norm": 0.31718170642852783,
      "learning_rate": 2e-05,
      "loss": 0.4945,
      "step": 5280
    },
    {
      "epoch": 0.4348898388687932,
      "grad_norm": 0.32966628670692444,
      "learning_rate": 2e-05,
      "loss": 0.5284,
      "step": 5290
    },
    {
      "epoch": 0.43571193686287407,
      "grad_norm": 0.400497704744339,
      "learning_rate": 2e-05,
      "loss": 0.5353,
      "step": 5300
    },
    {
      "epoch": 0.43653403485695497,
      "grad_norm": 0.3342435956001282,
      "learning_rate": 2e-05,
      "loss": 0.5056,
      "step": 5310
    },
    {
      "epoch": 0.43735613285103586,
      "grad_norm": 0.33096522092819214,
      "learning_rate": 2e-05,
      "loss": 0.504,
      "step": 5320
    },
    {
      "epoch": 0.43817823084511676,
      "grad_norm": 0.3500446379184723,
      "learning_rate": 2e-05,
      "loss": 0.5035,
      "step": 5330
    },
    {
      "epoch": 0.43900032883919765,
      "grad_norm": 0.35873088240623474,
      "learning_rate": 2e-05,
      "loss": 0.5338,
      "step": 5340
    },
    {
      "epoch": 0.43982242683327855,
      "grad_norm": 0.36536699533462524,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 5350
    },
    {
      "epoch": 0.44064452482735944,
      "grad_norm": 0.34878480434417725,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 5360
    },
    {
      "epoch": 0.44146662282144034,
      "grad_norm": 0.36636587977409363,
      "learning_rate": 2e-05,
      "loss": 0.5667,
      "step": 5370
    },
    {
      "epoch": 0.44228872081552123,
      "grad_norm": 0.309857040643692,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 5380
    },
    {
      "epoch": 0.4431108188096021,
      "grad_norm": 0.3562362790107727,
      "learning_rate": 2e-05,
      "loss": 0.5242,
      "step": 5390
    },
    {
      "epoch": 0.443932916803683,
      "grad_norm": 0.33130940794944763,
      "learning_rate": 2e-05,
      "loss": 0.5785,
      "step": 5400
    },
    {
      "epoch": 0.4447550147977639,
      "grad_norm": 0.3700507879257202,
      "learning_rate": 2e-05,
      "loss": 0.5127,
      "step": 5410
    },
    {
      "epoch": 0.4455771127918448,
      "grad_norm": 0.39629971981048584,
      "learning_rate": 2e-05,
      "loss": 0.5603,
      "step": 5420
    },
    {
      "epoch": 0.4463992107859257,
      "grad_norm": 0.39275819063186646,
      "learning_rate": 2e-05,
      "loss": 0.5221,
      "step": 5430
    },
    {
      "epoch": 0.4472213087800066,
      "grad_norm": 0.3409693241119385,
      "learning_rate": 2e-05,
      "loss": 0.5681,
      "step": 5440
    },
    {
      "epoch": 0.4480434067740875,
      "grad_norm": 0.323932409286499,
      "learning_rate": 2e-05,
      "loss": 0.5391,
      "step": 5450
    },
    {
      "epoch": 0.4488655047681684,
      "grad_norm": 0.3399604856967926,
      "learning_rate": 2e-05,
      "loss": 0.4894,
      "step": 5460
    },
    {
      "epoch": 0.4496876027622493,
      "grad_norm": 0.3412494659423828,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 5470
    },
    {
      "epoch": 0.4505097007563302,
      "grad_norm": 0.34528857469558716,
      "learning_rate": 2e-05,
      "loss": 0.5363,
      "step": 5480
    },
    {
      "epoch": 0.4513317987504111,
      "grad_norm": 0.5859479904174805,
      "learning_rate": 2e-05,
      "loss": 0.5219,
      "step": 5490
    },
    {
      "epoch": 0.45215389674449197,
      "grad_norm": 0.3615967631340027,
      "learning_rate": 2e-05,
      "loss": 0.5419,
      "step": 5500
    },
    {
      "epoch": 0.45297599473857286,
      "grad_norm": 0.3209433853626251,
      "learning_rate": 2e-05,
      "loss": 0.5627,
      "step": 5510
    },
    {
      "epoch": 0.45379809273265376,
      "grad_norm": 0.3443490266799927,
      "learning_rate": 2e-05,
      "loss": 0.5169,
      "step": 5520
    },
    {
      "epoch": 0.45462019072673465,
      "grad_norm": 0.32966867089271545,
      "learning_rate": 2e-05,
      "loss": 0.5301,
      "step": 5530
    },
    {
      "epoch": 0.45544228872081555,
      "grad_norm": 0.3491547703742981,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 5540
    },
    {
      "epoch": 0.45626438671489644,
      "grad_norm": 0.32784348726272583,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 5550
    },
    {
      "epoch": 0.45708648470897734,
      "grad_norm": 0.5522540807723999,
      "learning_rate": 2e-05,
      "loss": 0.5401,
      "step": 5560
    },
    {
      "epoch": 0.45790858270305823,
      "grad_norm": 0.35558345913887024,
      "learning_rate": 2e-05,
      "loss": 0.5154,
      "step": 5570
    },
    {
      "epoch": 0.4587306806971391,
      "grad_norm": 0.3482027053833008,
      "learning_rate": 2e-05,
      "loss": 0.5258,
      "step": 5580
    },
    {
      "epoch": 0.45955277869122,
      "grad_norm": 0.3334774076938629,
      "learning_rate": 2e-05,
      "loss": 0.5182,
      "step": 5590
    },
    {
      "epoch": 0.4603748766853009,
      "grad_norm": 0.40814051032066345,
      "learning_rate": 2e-05,
      "loss": 0.4762,
      "step": 5600
    },
    {
      "epoch": 0.46119697467938175,
      "grad_norm": 0.3433951139450073,
      "learning_rate": 2e-05,
      "loss": 0.5228,
      "step": 5610
    },
    {
      "epoch": 0.46201907267346265,
      "grad_norm": 0.35481464862823486,
      "learning_rate": 2e-05,
      "loss": 0.5292,
      "step": 5620
    },
    {
      "epoch": 0.46284117066754354,
      "grad_norm": 0.34293460845947266,
      "learning_rate": 2e-05,
      "loss": 0.5257,
      "step": 5630
    },
    {
      "epoch": 0.46366326866162444,
      "grad_norm": 0.3450120687484741,
      "learning_rate": 2e-05,
      "loss": 0.5404,
      "step": 5640
    },
    {
      "epoch": 0.46448536665570533,
      "grad_norm": 0.33979350328445435,
      "learning_rate": 2e-05,
      "loss": 0.5046,
      "step": 5650
    },
    {
      "epoch": 0.46530746464978623,
      "grad_norm": 0.3271808624267578,
      "learning_rate": 2e-05,
      "loss": 0.5049,
      "step": 5660
    },
    {
      "epoch": 0.4661295626438671,
      "grad_norm": 0.42547348141670227,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 5670
    },
    {
      "epoch": 0.466951660637948,
      "grad_norm": 0.3803088665008545,
      "learning_rate": 2e-05,
      "loss": 0.5102,
      "step": 5680
    },
    {
      "epoch": 0.4677737586320289,
      "grad_norm": 0.35167670249938965,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 5690
    },
    {
      "epoch": 0.4685958566261098,
      "grad_norm": 0.39087656140327454,
      "learning_rate": 2e-05,
      "loss": 0.5278,
      "step": 5700
    },
    {
      "epoch": 0.4694179546201907,
      "grad_norm": 0.4157888889312744,
      "learning_rate": 2e-05,
      "loss": 0.5272,
      "step": 5710
    },
    {
      "epoch": 0.4702400526142716,
      "grad_norm": 0.37098145484924316,
      "learning_rate": 2e-05,
      "loss": 0.5462,
      "step": 5720
    },
    {
      "epoch": 0.4710621506083525,
      "grad_norm": 0.35868436098098755,
      "learning_rate": 2e-05,
      "loss": 0.5239,
      "step": 5730
    },
    {
      "epoch": 0.4718842486024334,
      "grad_norm": 0.3059246838092804,
      "learning_rate": 2e-05,
      "loss": 0.521,
      "step": 5740
    },
    {
      "epoch": 0.4727063465965143,
      "grad_norm": 0.3306145668029785,
      "learning_rate": 2e-05,
      "loss": 0.5244,
      "step": 5750
    },
    {
      "epoch": 0.4735284445905952,
      "grad_norm": 0.4216914176940918,
      "learning_rate": 2e-05,
      "loss": 0.5182,
      "step": 5760
    },
    {
      "epoch": 0.47435054258467607,
      "grad_norm": 0.34667181968688965,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 5770
    },
    {
      "epoch": 0.47517264057875697,
      "grad_norm": 0.36054208874702454,
      "learning_rate": 2e-05,
      "loss": 0.5266,
      "step": 5780
    },
    {
      "epoch": 0.47599473857283786,
      "grad_norm": 0.3428150713443756,
      "learning_rate": 2e-05,
      "loss": 0.5121,
      "step": 5790
    },
    {
      "epoch": 0.47681683656691876,
      "grad_norm": 0.30918461084365845,
      "learning_rate": 2e-05,
      "loss": 0.5091,
      "step": 5800
    },
    {
      "epoch": 0.47763893456099965,
      "grad_norm": 0.3570272922515869,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 5810
    },
    {
      "epoch": 0.47846103255508055,
      "grad_norm": 0.3141920268535614,
      "learning_rate": 2e-05,
      "loss": 0.5393,
      "step": 5820
    },
    {
      "epoch": 0.47928313054916144,
      "grad_norm": 0.36845454573631287,
      "learning_rate": 2e-05,
      "loss": 0.4958,
      "step": 5830
    },
    {
      "epoch": 0.48010522854324233,
      "grad_norm": 0.43613797426223755,
      "learning_rate": 2e-05,
      "loss": 0.5342,
      "step": 5840
    },
    {
      "epoch": 0.48092732653732323,
      "grad_norm": 0.3770115375518799,
      "learning_rate": 2e-05,
      "loss": 0.5136,
      "step": 5850
    },
    {
      "epoch": 0.4817494245314041,
      "grad_norm": 0.3940106928348541,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 5860
    },
    {
      "epoch": 0.482571522525485,
      "grad_norm": 0.339157372713089,
      "learning_rate": 2e-05,
      "loss": 0.4965,
      "step": 5870
    },
    {
      "epoch": 0.4833936205195659,
      "grad_norm": 0.33463016152381897,
      "learning_rate": 2e-05,
      "loss": 0.5228,
      "step": 5880
    },
    {
      "epoch": 0.4842157185136468,
      "grad_norm": 0.3496641516685486,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 5890
    },
    {
      "epoch": 0.4850378165077277,
      "grad_norm": 0.39122918248176575,
      "learning_rate": 2e-05,
      "loss": 0.5291,
      "step": 5900
    },
    {
      "epoch": 0.4858599145018086,
      "grad_norm": 0.4266890585422516,
      "learning_rate": 2e-05,
      "loss": 0.5344,
      "step": 5910
    },
    {
      "epoch": 0.4866820124958895,
      "grad_norm": 0.36491334438323975,
      "learning_rate": 2e-05,
      "loss": 0.5151,
      "step": 5920
    },
    {
      "epoch": 0.4875041104899704,
      "grad_norm": 0.3526418209075928,
      "learning_rate": 2e-05,
      "loss": 0.5253,
      "step": 5930
    },
    {
      "epoch": 0.4883262084840513,
      "grad_norm": 0.45887047052383423,
      "learning_rate": 2e-05,
      "loss": 0.4971,
      "step": 5940
    },
    {
      "epoch": 0.4891483064781322,
      "grad_norm": 0.3713309168815613,
      "learning_rate": 2e-05,
      "loss": 0.4947,
      "step": 5950
    },
    {
      "epoch": 0.4899704044722131,
      "grad_norm": 0.3308202624320984,
      "learning_rate": 2e-05,
      "loss": 0.517,
      "step": 5960
    },
    {
      "epoch": 0.49079250246629397,
      "grad_norm": 0.36872732639312744,
      "learning_rate": 2e-05,
      "loss": 0.5121,
      "step": 5970
    },
    {
      "epoch": 0.49161460046037486,
      "grad_norm": 0.3550807535648346,
      "learning_rate": 2e-05,
      "loss": 0.5143,
      "step": 5980
    },
    {
      "epoch": 0.49243669845445576,
      "grad_norm": 0.525942862033844,
      "learning_rate": 2e-05,
      "loss": 0.5459,
      "step": 5990
    },
    {
      "epoch": 0.49325879644853665,
      "grad_norm": 0.343132883310318,
      "learning_rate": 2e-05,
      "loss": 0.5265,
      "step": 6000
    },
    {
      "epoch": 0.49408089444261755,
      "grad_norm": 0.3554967939853668,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 6010
    },
    {
      "epoch": 0.49490299243669844,
      "grad_norm": 0.4032709002494812,
      "learning_rate": 2e-05,
      "loss": 0.5275,
      "step": 6020
    },
    {
      "epoch": 0.49572509043077934,
      "grad_norm": 0.3412620723247528,
      "learning_rate": 2e-05,
      "loss": 0.5438,
      "step": 6030
    },
    {
      "epoch": 0.49654718842486023,
      "grad_norm": 0.38648051023483276,
      "learning_rate": 2e-05,
      "loss": 0.5212,
      "step": 6040
    },
    {
      "epoch": 0.4973692864189411,
      "grad_norm": 0.3341023921966553,
      "learning_rate": 2e-05,
      "loss": 0.5399,
      "step": 6050
    },
    {
      "epoch": 0.498191384413022,
      "grad_norm": 0.361113041639328,
      "learning_rate": 2e-05,
      "loss": 0.5138,
      "step": 6060
    },
    {
      "epoch": 0.4990134824071029,
      "grad_norm": 0.3458334803581238,
      "learning_rate": 2e-05,
      "loss": 0.5236,
      "step": 6070
    },
    {
      "epoch": 0.4998355804011838,
      "grad_norm": 0.3666258454322815,
      "learning_rate": 2e-05,
      "loss": 0.5084,
      "step": 6080
    },
    {
      "epoch": 0.5006576783952648,
      "grad_norm": 0.3324543237686157,
      "learning_rate": 2e-05,
      "loss": 0.52,
      "step": 6090
    },
    {
      "epoch": 0.5014797763893456,
      "grad_norm": 0.33686965703964233,
      "learning_rate": 2e-05,
      "loss": 0.5386,
      "step": 6100
    },
    {
      "epoch": 0.5023018743834265,
      "grad_norm": 0.33279576897621155,
      "learning_rate": 2e-05,
      "loss": 0.5427,
      "step": 6110
    },
    {
      "epoch": 0.5031239723775074,
      "grad_norm": 0.3806231915950775,
      "learning_rate": 2e-05,
      "loss": 0.5258,
      "step": 6120
    },
    {
      "epoch": 0.5039460703715883,
      "grad_norm": 0.336805522441864,
      "learning_rate": 2e-05,
      "loss": 0.5053,
      "step": 6130
    },
    {
      "epoch": 0.5047681683656692,
      "grad_norm": 0.35874515771865845,
      "learning_rate": 2e-05,
      "loss": 0.5204,
      "step": 6140
    },
    {
      "epoch": 0.5055902663597501,
      "grad_norm": 0.3264509439468384,
      "learning_rate": 2e-05,
      "loss": 0.4792,
      "step": 6150
    },
    {
      "epoch": 0.506412364353831,
      "grad_norm": 0.34842997789382935,
      "learning_rate": 2e-05,
      "loss": 0.5168,
      "step": 6160
    },
    {
      "epoch": 0.5072344623479119,
      "grad_norm": 0.3089509904384613,
      "learning_rate": 2e-05,
      "loss": 0.4946,
      "step": 6170
    },
    {
      "epoch": 0.5080565603419928,
      "grad_norm": 0.37455838918685913,
      "learning_rate": 2e-05,
      "loss": 0.5176,
      "step": 6180
    },
    {
      "epoch": 0.5088786583360737,
      "grad_norm": 0.38695502281188965,
      "learning_rate": 2e-05,
      "loss": 0.5377,
      "step": 6190
    },
    {
      "epoch": 0.5097007563301545,
      "grad_norm": 0.39155563712120056,
      "learning_rate": 2e-05,
      "loss": 0.5405,
      "step": 6200
    },
    {
      "epoch": 0.5105228543242355,
      "grad_norm": 0.37171003222465515,
      "learning_rate": 2e-05,
      "loss": 0.5134,
      "step": 6210
    },
    {
      "epoch": 0.5113449523183163,
      "grad_norm": 0.38868147134780884,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 6220
    },
    {
      "epoch": 0.5121670503123973,
      "grad_norm": 0.35871919989585876,
      "learning_rate": 2e-05,
      "loss": 0.5299,
      "step": 6230
    },
    {
      "epoch": 0.5129891483064781,
      "grad_norm": 0.339733749628067,
      "learning_rate": 2e-05,
      "loss": 0.5252,
      "step": 6240
    },
    {
      "epoch": 0.5138112463005591,
      "grad_norm": 0.3725637197494507,
      "learning_rate": 2e-05,
      "loss": 0.5454,
      "step": 6250
    },
    {
      "epoch": 0.5146333442946399,
      "grad_norm": 0.3703186511993408,
      "learning_rate": 2e-05,
      "loss": 0.5263,
      "step": 6260
    },
    {
      "epoch": 0.5154554422887209,
      "grad_norm": 0.3278525173664093,
      "learning_rate": 2e-05,
      "loss": 0.5368,
      "step": 6270
    },
    {
      "epoch": 0.5162775402828017,
      "grad_norm": 0.3236202299594879,
      "learning_rate": 2e-05,
      "loss": 0.5167,
      "step": 6280
    },
    {
      "epoch": 0.5170996382768827,
      "grad_norm": 0.3427184820175171,
      "learning_rate": 2e-05,
      "loss": 0.5101,
      "step": 6290
    },
    {
      "epoch": 0.5179217362709635,
      "grad_norm": 0.42214933037757874,
      "learning_rate": 2e-05,
      "loss": 0.5417,
      "step": 6300
    },
    {
      "epoch": 0.5187438342650444,
      "grad_norm": 0.34158939123153687,
      "learning_rate": 2e-05,
      "loss": 0.4989,
      "step": 6310
    },
    {
      "epoch": 0.5195659322591253,
      "grad_norm": 0.33726999163627625,
      "learning_rate": 2e-05,
      "loss": 0.496,
      "step": 6320
    },
    {
      "epoch": 0.5203880302532062,
      "grad_norm": 0.34772059321403503,
      "learning_rate": 2e-05,
      "loss": 0.5173,
      "step": 6330
    },
    {
      "epoch": 0.5212101282472871,
      "grad_norm": 0.31484803557395935,
      "learning_rate": 2e-05,
      "loss": 0.5217,
      "step": 6340
    },
    {
      "epoch": 0.522032226241368,
      "grad_norm": 0.358344167470932,
      "learning_rate": 2e-05,
      "loss": 0.5102,
      "step": 6350
    },
    {
      "epoch": 0.5228543242354489,
      "grad_norm": 0.32679757475852966,
      "learning_rate": 2e-05,
      "loss": 0.499,
      "step": 6360
    },
    {
      "epoch": 0.5236764222295298,
      "grad_norm": 0.3723878562450409,
      "learning_rate": 2e-05,
      "loss": 0.5315,
      "step": 6370
    },
    {
      "epoch": 0.5244985202236107,
      "grad_norm": 0.3221898674964905,
      "learning_rate": 2e-05,
      "loss": 0.5156,
      "step": 6380
    },
    {
      "epoch": 0.5253206182176916,
      "grad_norm": 0.31092241406440735,
      "learning_rate": 2e-05,
      "loss": 0.5131,
      "step": 6390
    },
    {
      "epoch": 0.5261427162117724,
      "grad_norm": 0.37889015674591064,
      "learning_rate": 2e-05,
      "loss": 0.5288,
      "step": 6400
    },
    {
      "epoch": 0.5269648142058533,
      "grad_norm": 0.3712537884712219,
      "learning_rate": 2e-05,
      "loss": 0.5401,
      "step": 6410
    },
    {
      "epoch": 0.5277869121999342,
      "grad_norm": 0.4032345414161682,
      "learning_rate": 2e-05,
      "loss": 0.5326,
      "step": 6420
    },
    {
      "epoch": 0.5286090101940151,
      "grad_norm": 0.3339247703552246,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 6430
    },
    {
      "epoch": 0.529431108188096,
      "grad_norm": 0.3389577567577362,
      "learning_rate": 2e-05,
      "loss": 0.5143,
      "step": 6440
    },
    {
      "epoch": 0.5302532061821769,
      "grad_norm": 0.3528316617012024,
      "learning_rate": 2e-05,
      "loss": 0.5023,
      "step": 6450
    },
    {
      "epoch": 0.5310753041762578,
      "grad_norm": 0.3738712668418884,
      "learning_rate": 2e-05,
      "loss": 0.5168,
      "step": 6460
    },
    {
      "epoch": 0.5318974021703387,
      "grad_norm": 0.3632975220680237,
      "learning_rate": 2e-05,
      "loss": 0.5298,
      "step": 6470
    },
    {
      "epoch": 0.5327195001644196,
      "grad_norm": 0.38012567162513733,
      "learning_rate": 2e-05,
      "loss": 0.527,
      "step": 6480
    },
    {
      "epoch": 0.5335415981585004,
      "grad_norm": 0.3689791262149811,
      "learning_rate": 2e-05,
      "loss": 0.5216,
      "step": 6490
    },
    {
      "epoch": 0.5343636961525814,
      "grad_norm": 0.3481782078742981,
      "learning_rate": 2e-05,
      "loss": 0.5438,
      "step": 6500
    },
    {
      "epoch": 0.5351857941466622,
      "grad_norm": 0.36696410179138184,
      "learning_rate": 2e-05,
      "loss": 0.5348,
      "step": 6510
    },
    {
      "epoch": 0.5360078921407432,
      "grad_norm": 0.3264205753803253,
      "learning_rate": 2e-05,
      "loss": 0.5325,
      "step": 6520
    },
    {
      "epoch": 0.536829990134824,
      "grad_norm": 0.319574236869812,
      "learning_rate": 2e-05,
      "loss": 0.4717,
      "step": 6530
    },
    {
      "epoch": 0.537652088128905,
      "grad_norm": 0.32346606254577637,
      "learning_rate": 2e-05,
      "loss": 0.5333,
      "step": 6540
    },
    {
      "epoch": 0.5384741861229858,
      "grad_norm": 0.3300493657588959,
      "learning_rate": 2e-05,
      "loss": 0.5291,
      "step": 6550
    },
    {
      "epoch": 0.5392962841170668,
      "grad_norm": 0.3276021182537079,
      "learning_rate": 2e-05,
      "loss": 0.4659,
      "step": 6560
    },
    {
      "epoch": 0.5401183821111476,
      "grad_norm": 0.3467969596385956,
      "learning_rate": 2e-05,
      "loss": 0.503,
      "step": 6570
    },
    {
      "epoch": 0.5409404801052285,
      "grad_norm": 0.35062652826309204,
      "learning_rate": 2e-05,
      "loss": 0.52,
      "step": 6580
    },
    {
      "epoch": 0.5417625780993094,
      "grad_norm": 0.3541053235530853,
      "learning_rate": 2e-05,
      "loss": 0.5457,
      "step": 6590
    },
    {
      "epoch": 0.5425846760933903,
      "grad_norm": 0.3582180440425873,
      "learning_rate": 2e-05,
      "loss": 0.4948,
      "step": 6600
    },
    {
      "epoch": 0.5434067740874712,
      "grad_norm": 0.34203892946243286,
      "learning_rate": 2e-05,
      "loss": 0.5176,
      "step": 6610
    },
    {
      "epoch": 0.5442288720815521,
      "grad_norm": 0.3508211076259613,
      "learning_rate": 2e-05,
      "loss": 0.5219,
      "step": 6620
    },
    {
      "epoch": 0.545050970075633,
      "grad_norm": 0.31727051734924316,
      "learning_rate": 2e-05,
      "loss": 0.5076,
      "step": 6630
    },
    {
      "epoch": 0.5458730680697139,
      "grad_norm": 0.3239530920982361,
      "learning_rate": 2e-05,
      "loss": 0.5226,
      "step": 6640
    },
    {
      "epoch": 0.5466951660637948,
      "grad_norm": 0.36701706051826477,
      "learning_rate": 2e-05,
      "loss": 0.5108,
      "step": 6650
    },
    {
      "epoch": 0.5475172640578757,
      "grad_norm": 0.33933332562446594,
      "learning_rate": 2e-05,
      "loss": 0.5154,
      "step": 6660
    },
    {
      "epoch": 0.5483393620519565,
      "grad_norm": 0.3401816785335541,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 6670
    },
    {
      "epoch": 0.5491614600460375,
      "grad_norm": 0.40791773796081543,
      "learning_rate": 2e-05,
      "loss": 0.5147,
      "step": 6680
    },
    {
      "epoch": 0.5499835580401183,
      "grad_norm": 0.4325167238712311,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 6690
    },
    {
      "epoch": 0.5508056560341993,
      "grad_norm": 0.3388225734233856,
      "learning_rate": 2e-05,
      "loss": 0.5175,
      "step": 6700
    },
    {
      "epoch": 0.5516277540282801,
      "grad_norm": 0.39247000217437744,
      "learning_rate": 2e-05,
      "loss": 0.5096,
      "step": 6710
    },
    {
      "epoch": 0.5524498520223611,
      "grad_norm": 0.345599502325058,
      "learning_rate": 2e-05,
      "loss": 0.5297,
      "step": 6720
    },
    {
      "epoch": 0.5532719500164419,
      "grad_norm": 0.33695724606513977,
      "learning_rate": 2e-05,
      "loss": 0.4973,
      "step": 6730
    },
    {
      "epoch": 0.5540940480105229,
      "grad_norm": 0.33853599429130554,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 6740
    },
    {
      "epoch": 0.5549161460046037,
      "grad_norm": 0.3400750458240509,
      "learning_rate": 2e-05,
      "loss": 0.5013,
      "step": 6750
    },
    {
      "epoch": 0.5557382439986847,
      "grad_norm": 0.36855965852737427,
      "learning_rate": 2e-05,
      "loss": 0.5198,
      "step": 6760
    },
    {
      "epoch": 0.5565603419927655,
      "grad_norm": 0.3576034605503082,
      "learning_rate": 2e-05,
      "loss": 0.5044,
      "step": 6770
    },
    {
      "epoch": 0.5573824399868464,
      "grad_norm": 0.4167182147502899,
      "learning_rate": 2e-05,
      "loss": 0.5005,
      "step": 6780
    },
    {
      "epoch": 0.5582045379809273,
      "grad_norm": 0.36729925870895386,
      "learning_rate": 2e-05,
      "loss": 0.5199,
      "step": 6790
    },
    {
      "epoch": 0.5590266359750082,
      "grad_norm": 0.34270405769348145,
      "learning_rate": 2e-05,
      "loss": 0.4932,
      "step": 6800
    },
    {
      "epoch": 0.5598487339690891,
      "grad_norm": 0.4501667320728302,
      "learning_rate": 2e-05,
      "loss": 0.5018,
      "step": 6810
    },
    {
      "epoch": 0.56067083196317,
      "grad_norm": 0.3375786244869232,
      "learning_rate": 2e-05,
      "loss": 0.5218,
      "step": 6820
    },
    {
      "epoch": 0.5614929299572509,
      "grad_norm": 0.38053855299949646,
      "learning_rate": 2e-05,
      "loss": 0.526,
      "step": 6830
    },
    {
      "epoch": 0.5623150279513318,
      "grad_norm": 0.31846582889556885,
      "learning_rate": 2e-05,
      "loss": 0.4945,
      "step": 6840
    },
    {
      "epoch": 0.5631371259454127,
      "grad_norm": 0.3517959415912628,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 6850
    },
    {
      "epoch": 0.5639592239394936,
      "grad_norm": 0.3572523891925812,
      "learning_rate": 2e-05,
      "loss": 0.4907,
      "step": 6860
    },
    {
      "epoch": 0.5647813219335744,
      "grad_norm": 0.36367860436439514,
      "learning_rate": 2e-05,
      "loss": 0.5354,
      "step": 6870
    },
    {
      "epoch": 0.5656034199276554,
      "grad_norm": 0.33389022946357727,
      "learning_rate": 2e-05,
      "loss": 0.5056,
      "step": 6880
    },
    {
      "epoch": 0.5664255179217362,
      "grad_norm": 0.33777540922164917,
      "learning_rate": 2e-05,
      "loss": 0.5115,
      "step": 6890
    },
    {
      "epoch": 0.5672476159158172,
      "grad_norm": 0.3262439966201782,
      "learning_rate": 2e-05,
      "loss": 0.4952,
      "step": 6900
    },
    {
      "epoch": 0.568069713909898,
      "grad_norm": 0.32361310720443726,
      "learning_rate": 2e-05,
      "loss": 0.5122,
      "step": 6910
    },
    {
      "epoch": 0.568891811903979,
      "grad_norm": 0.3598339557647705,
      "learning_rate": 2e-05,
      "loss": 0.5122,
      "step": 6920
    },
    {
      "epoch": 0.5697139098980598,
      "grad_norm": 0.3919849395751953,
      "learning_rate": 2e-05,
      "loss": 0.5044,
      "step": 6930
    },
    {
      "epoch": 0.5705360078921408,
      "grad_norm": 0.38242006301879883,
      "learning_rate": 2e-05,
      "loss": 0.516,
      "step": 6940
    },
    {
      "epoch": 0.5713581058862216,
      "grad_norm": 0.357115238904953,
      "learning_rate": 2e-05,
      "loss": 0.5195,
      "step": 6950
    },
    {
      "epoch": 0.5721802038803026,
      "grad_norm": 0.3921072483062744,
      "learning_rate": 2e-05,
      "loss": 0.5062,
      "step": 6960
    },
    {
      "epoch": 0.5730023018743834,
      "grad_norm": 0.4039548337459564,
      "learning_rate": 2e-05,
      "loss": 0.5043,
      "step": 6970
    },
    {
      "epoch": 0.5738243998684643,
      "grad_norm": 0.36437490582466125,
      "learning_rate": 2e-05,
      "loss": 0.5242,
      "step": 6980
    },
    {
      "epoch": 0.5746464978625452,
      "grad_norm": 0.32480210065841675,
      "learning_rate": 2e-05,
      "loss": 0.5106,
      "step": 6990
    },
    {
      "epoch": 0.5754685958566261,
      "grad_norm": 0.3924250602722168,
      "learning_rate": 2e-05,
      "loss": 0.4737,
      "step": 7000
    },
    {
      "epoch": 0.576290693850707,
      "grad_norm": 0.39116451144218445,
      "learning_rate": 2e-05,
      "loss": 0.525,
      "step": 7010
    },
    {
      "epoch": 0.5771127918447879,
      "grad_norm": 0.3894917070865631,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 7020
    },
    {
      "epoch": 0.5779348898388688,
      "grad_norm": 0.3350021541118622,
      "learning_rate": 2e-05,
      "loss": 0.483,
      "step": 7030
    },
    {
      "epoch": 0.5787569878329497,
      "grad_norm": 0.3306668698787689,
      "learning_rate": 2e-05,
      "loss": 0.516,
      "step": 7040
    },
    {
      "epoch": 0.5795790858270305,
      "grad_norm": 0.32512906193733215,
      "learning_rate": 2e-05,
      "loss": 0.4899,
      "step": 7050
    },
    {
      "epoch": 0.5804011838211115,
      "grad_norm": 0.38045719265937805,
      "learning_rate": 2e-05,
      "loss": 0.5022,
      "step": 7060
    },
    {
      "epoch": 0.5812232818151923,
      "grad_norm": 0.37048399448394775,
      "learning_rate": 2e-05,
      "loss": 0.536,
      "step": 7070
    },
    {
      "epoch": 0.5820453798092733,
      "grad_norm": 0.3797077238559723,
      "learning_rate": 2e-05,
      "loss": 0.5048,
      "step": 7080
    },
    {
      "epoch": 0.5828674778033541,
      "grad_norm": 0.3983193635940552,
      "learning_rate": 2e-05,
      "loss": 0.5016,
      "step": 7090
    },
    {
      "epoch": 0.5836895757974351,
      "grad_norm": 0.3394424319267273,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 7100
    },
    {
      "epoch": 0.5845116737915159,
      "grad_norm": 0.4480521082878113,
      "learning_rate": 2e-05,
      "loss": 0.5292,
      "step": 7110
    },
    {
      "epoch": 0.5853337717855969,
      "grad_norm": 0.5891214609146118,
      "learning_rate": 2e-05,
      "loss": 0.5497,
      "step": 7120
    },
    {
      "epoch": 0.5861558697796777,
      "grad_norm": 0.3389278054237366,
      "learning_rate": 2e-05,
      "loss": 0.4939,
      "step": 7130
    },
    {
      "epoch": 0.5869779677737587,
      "grad_norm": 0.32833293080329895,
      "learning_rate": 2e-05,
      "loss": 0.517,
      "step": 7140
    },
    {
      "epoch": 0.5878000657678395,
      "grad_norm": 0.36266469955444336,
      "learning_rate": 2e-05,
      "loss": 0.5313,
      "step": 7150
    },
    {
      "epoch": 0.5886221637619204,
      "grad_norm": 0.33986470103263855,
      "learning_rate": 2e-05,
      "loss": 0.5293,
      "step": 7160
    },
    {
      "epoch": 0.5894442617560013,
      "grad_norm": 0.3569768965244293,
      "learning_rate": 2e-05,
      "loss": 0.5213,
      "step": 7170
    },
    {
      "epoch": 0.5902663597500822,
      "grad_norm": 0.4939633011817932,
      "learning_rate": 2e-05,
      "loss": 0.5359,
      "step": 7180
    },
    {
      "epoch": 0.5910884577441631,
      "grad_norm": 0.35048767924308777,
      "learning_rate": 2e-05,
      "loss": 0.517,
      "step": 7190
    },
    {
      "epoch": 0.591910555738244,
      "grad_norm": 0.3796350061893463,
      "learning_rate": 2e-05,
      "loss": 0.5267,
      "step": 7200
    },
    {
      "epoch": 0.5927326537323249,
      "grad_norm": 0.45387715101242065,
      "learning_rate": 2e-05,
      "loss": 0.5139,
      "step": 7210
    },
    {
      "epoch": 0.5935547517264058,
      "grad_norm": 0.37782368063926697,
      "learning_rate": 2e-05,
      "loss": 0.5091,
      "step": 7220
    },
    {
      "epoch": 0.5943768497204867,
      "grad_norm": 0.374018132686615,
      "learning_rate": 2e-05,
      "loss": 0.5087,
      "step": 7230
    },
    {
      "epoch": 0.5951989477145676,
      "grad_norm": 0.35478952527046204,
      "learning_rate": 2e-05,
      "loss": 0.5349,
      "step": 7240
    },
    {
      "epoch": 0.5960210457086484,
      "grad_norm": 0.33503904938697815,
      "learning_rate": 2e-05,
      "loss": 0.5005,
      "step": 7250
    },
    {
      "epoch": 0.5968431437027294,
      "grad_norm": 0.3862592577934265,
      "learning_rate": 2e-05,
      "loss": 0.5,
      "step": 7260
    },
    {
      "epoch": 0.5976652416968102,
      "grad_norm": 0.36067014932632446,
      "learning_rate": 2e-05,
      "loss": 0.506,
      "step": 7270
    },
    {
      "epoch": 0.5984873396908912,
      "grad_norm": 0.3634778559207916,
      "learning_rate": 2e-05,
      "loss": 0.5246,
      "step": 7280
    },
    {
      "epoch": 0.599309437684972,
      "grad_norm": 0.36188751459121704,
      "learning_rate": 2e-05,
      "loss": 0.5132,
      "step": 7290
    },
    {
      "epoch": 0.600131535679053,
      "grad_norm": 0.34325480461120605,
      "learning_rate": 2e-05,
      "loss": 0.5047,
      "step": 7300
    },
    {
      "epoch": 0.6009536336731338,
      "grad_norm": 0.3932059705257416,
      "learning_rate": 2e-05,
      "loss": 0.5296,
      "step": 7310
    },
    {
      "epoch": 0.6017757316672148,
      "grad_norm": 0.33536094427108765,
      "learning_rate": 2e-05,
      "loss": 0.5098,
      "step": 7320
    },
    {
      "epoch": 0.6025978296612956,
      "grad_norm": 0.3565771281719208,
      "learning_rate": 2e-05,
      "loss": 0.5157,
      "step": 7330
    },
    {
      "epoch": 0.6034199276553766,
      "grad_norm": 0.3434689939022064,
      "learning_rate": 2e-05,
      "loss": 0.5144,
      "step": 7340
    },
    {
      "epoch": 0.6042420256494574,
      "grad_norm": 0.4176717698574066,
      "learning_rate": 2e-05,
      "loss": 0.5089,
      "step": 7350
    },
    {
      "epoch": 0.6050641236435383,
      "grad_norm": 0.3383750319480896,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 7360
    },
    {
      "epoch": 0.6058862216376192,
      "grad_norm": 0.3415583372116089,
      "learning_rate": 2e-05,
      "loss": 0.5572,
      "step": 7370
    },
    {
      "epoch": 0.6067083196317001,
      "grad_norm": 0.3192376494407654,
      "learning_rate": 2e-05,
      "loss": 0.5305,
      "step": 7380
    },
    {
      "epoch": 0.607530417625781,
      "grad_norm": 0.31960010528564453,
      "learning_rate": 2e-05,
      "loss": 0.4889,
      "step": 7390
    },
    {
      "epoch": 0.6083525156198619,
      "grad_norm": 0.39826664328575134,
      "learning_rate": 2e-05,
      "loss": 0.5466,
      "step": 7400
    },
    {
      "epoch": 0.6091746136139428,
      "grad_norm": 0.336336612701416,
      "learning_rate": 2e-05,
      "loss": 0.5126,
      "step": 7410
    },
    {
      "epoch": 0.6099967116080237,
      "grad_norm": 0.36106076836586,
      "learning_rate": 2e-05,
      "loss": 0.5206,
      "step": 7420
    },
    {
      "epoch": 0.6108188096021046,
      "grad_norm": 0.3358343541622162,
      "learning_rate": 2e-05,
      "loss": 0.4965,
      "step": 7430
    },
    {
      "epoch": 0.6116409075961855,
      "grad_norm": 0.3448723554611206,
      "learning_rate": 2e-05,
      "loss": 0.5434,
      "step": 7440
    },
    {
      "epoch": 0.6124630055902663,
      "grad_norm": 0.3757982552051544,
      "learning_rate": 2e-05,
      "loss": 0.5106,
      "step": 7450
    },
    {
      "epoch": 0.6132851035843473,
      "grad_norm": 0.4052806496620178,
      "learning_rate": 2e-05,
      "loss": 0.547,
      "step": 7460
    },
    {
      "epoch": 0.6141072015784281,
      "grad_norm": 0.3721083700656891,
      "learning_rate": 2e-05,
      "loss": 0.5352,
      "step": 7470
    },
    {
      "epoch": 0.6149292995725091,
      "grad_norm": 0.35850921273231506,
      "learning_rate": 2e-05,
      "loss": 0.5125,
      "step": 7480
    },
    {
      "epoch": 0.6157513975665899,
      "grad_norm": 0.3884437680244446,
      "learning_rate": 2e-05,
      "loss": 0.4988,
      "step": 7490
    },
    {
      "epoch": 0.6165734955606709,
      "grad_norm": 0.3600706160068512,
      "learning_rate": 2e-05,
      "loss": 0.5163,
      "step": 7500
    },
    {
      "epoch": 0.6173955935547517,
      "grad_norm": 0.3290514349937439,
      "learning_rate": 2e-05,
      "loss": 0.5399,
      "step": 7510
    },
    {
      "epoch": 0.6182176915488327,
      "grad_norm": 0.3773733377456665,
      "learning_rate": 2e-05,
      "loss": 0.5044,
      "step": 7520
    },
    {
      "epoch": 0.6190397895429135,
      "grad_norm": 0.4137864112854004,
      "learning_rate": 2e-05,
      "loss": 0.5233,
      "step": 7530
    },
    {
      "epoch": 0.6198618875369944,
      "grad_norm": 0.35345545411109924,
      "learning_rate": 2e-05,
      "loss": 0.504,
      "step": 7540
    },
    {
      "epoch": 0.6206839855310753,
      "grad_norm": 0.3582826554775238,
      "learning_rate": 2e-05,
      "loss": 0.5122,
      "step": 7550
    },
    {
      "epoch": 0.6215060835251562,
      "grad_norm": 0.39076730608940125,
      "learning_rate": 2e-05,
      "loss": 0.4979,
      "step": 7560
    },
    {
      "epoch": 0.6223281815192371,
      "grad_norm": 0.3597421944141388,
      "learning_rate": 2e-05,
      "loss": 0.4994,
      "step": 7570
    },
    {
      "epoch": 0.623150279513318,
      "grad_norm": 0.35309913754463196,
      "learning_rate": 2e-05,
      "loss": 0.5294,
      "step": 7580
    },
    {
      "epoch": 0.6239723775073989,
      "grad_norm": 0.3912990391254425,
      "learning_rate": 2e-05,
      "loss": 0.5423,
      "step": 7590
    },
    {
      "epoch": 0.6247944755014798,
      "grad_norm": 0.3692512810230255,
      "learning_rate": 2e-05,
      "loss": 0.5031,
      "step": 7600
    },
    {
      "epoch": 0.6256165734955607,
      "grad_norm": 0.32391855120658875,
      "learning_rate": 2e-05,
      "loss": 0.5225,
      "step": 7610
    },
    {
      "epoch": 0.6264386714896416,
      "grad_norm": 0.37337398529052734,
      "learning_rate": 2e-05,
      "loss": 0.5262,
      "step": 7620
    },
    {
      "epoch": 0.6272607694837224,
      "grad_norm": 0.3550015091896057,
      "learning_rate": 2e-05,
      "loss": 0.5339,
      "step": 7630
    },
    {
      "epoch": 0.6280828674778034,
      "grad_norm": 0.3937951624393463,
      "learning_rate": 2e-05,
      "loss": 0.5271,
      "step": 7640
    },
    {
      "epoch": 0.6289049654718842,
      "grad_norm": 0.3871081471443176,
      "learning_rate": 2e-05,
      "loss": 0.5112,
      "step": 7650
    },
    {
      "epoch": 0.6297270634659652,
      "grad_norm": 0.3938255310058594,
      "learning_rate": 2e-05,
      "loss": 0.5185,
      "step": 7660
    },
    {
      "epoch": 0.630549161460046,
      "grad_norm": 0.33170223236083984,
      "learning_rate": 2e-05,
      "loss": 0.5192,
      "step": 7670
    },
    {
      "epoch": 0.631371259454127,
      "grad_norm": 0.35777175426483154,
      "learning_rate": 2e-05,
      "loss": 0.5016,
      "step": 7680
    },
    {
      "epoch": 0.6321933574482078,
      "grad_norm": 0.3342248201370239,
      "learning_rate": 2e-05,
      "loss": 0.5196,
      "step": 7690
    },
    {
      "epoch": 0.6330154554422888,
      "grad_norm": 0.3633832335472107,
      "learning_rate": 2e-05,
      "loss": 0.5153,
      "step": 7700
    },
    {
      "epoch": 0.6338375534363696,
      "grad_norm": 0.35452553629875183,
      "learning_rate": 2e-05,
      "loss": 0.5007,
      "step": 7710
    },
    {
      "epoch": 0.6346596514304506,
      "grad_norm": 0.37376514077186584,
      "learning_rate": 2e-05,
      "loss": 0.499,
      "step": 7720
    },
    {
      "epoch": 0.6354817494245314,
      "grad_norm": 0.38084471225738525,
      "learning_rate": 2e-05,
      "loss": 0.5293,
      "step": 7730
    },
    {
      "epoch": 0.6363038474186123,
      "grad_norm": 0.3683341145515442,
      "learning_rate": 2e-05,
      "loss": 0.4898,
      "step": 7740
    },
    {
      "epoch": 0.6371259454126932,
      "grad_norm": 0.33271658420562744,
      "learning_rate": 2e-05,
      "loss": 0.4953,
      "step": 7750
    },
    {
      "epoch": 0.6379480434067741,
      "grad_norm": 0.3332103192806244,
      "learning_rate": 2e-05,
      "loss": 0.5063,
      "step": 7760
    },
    {
      "epoch": 0.638770141400855,
      "grad_norm": 0.39206597208976746,
      "learning_rate": 2e-05,
      "loss": 0.5191,
      "step": 7770
    },
    {
      "epoch": 0.6395922393949359,
      "grad_norm": 0.36504417657852173,
      "learning_rate": 2e-05,
      "loss": 0.5297,
      "step": 7780
    },
    {
      "epoch": 0.6404143373890168,
      "grad_norm": 0.34102264046669006,
      "learning_rate": 2e-05,
      "loss": 0.5065,
      "step": 7790
    },
    {
      "epoch": 0.6412364353830977,
      "grad_norm": 0.3812370300292969,
      "learning_rate": 2e-05,
      "loss": 0.5237,
      "step": 7800
    },
    {
      "epoch": 0.6420585333771786,
      "grad_norm": 0.3697664439678192,
      "learning_rate": 2e-05,
      "loss": 0.4925,
      "step": 7810
    },
    {
      "epoch": 0.6428806313712595,
      "grad_norm": 0.3846948742866516,
      "learning_rate": 2e-05,
      "loss": 0.5387,
      "step": 7820
    },
    {
      "epoch": 0.6437027293653403,
      "grad_norm": 0.37630146741867065,
      "learning_rate": 2e-05,
      "loss": 0.515,
      "step": 7830
    },
    {
      "epoch": 0.6445248273594213,
      "grad_norm": 0.38242560625076294,
      "learning_rate": 2e-05,
      "loss": 0.5049,
      "step": 7840
    },
    {
      "epoch": 0.6453469253535021,
      "grad_norm": 0.3737841844558716,
      "learning_rate": 2e-05,
      "loss": 0.5162,
      "step": 7850
    },
    {
      "epoch": 0.6461690233475831,
      "grad_norm": 0.37755560874938965,
      "learning_rate": 2e-05,
      "loss": 0.5201,
      "step": 7860
    },
    {
      "epoch": 0.6469911213416639,
      "grad_norm": 0.37589287757873535,
      "learning_rate": 2e-05,
      "loss": 0.5309,
      "step": 7870
    },
    {
      "epoch": 0.6478132193357449,
      "grad_norm": 0.3560100197792053,
      "learning_rate": 2e-05,
      "loss": 0.505,
      "step": 7880
    },
    {
      "epoch": 0.6486353173298257,
      "grad_norm": 0.3803607225418091,
      "learning_rate": 2e-05,
      "loss": 0.5268,
      "step": 7890
    },
    {
      "epoch": 0.6494574153239067,
      "grad_norm": 0.48573926091194153,
      "learning_rate": 2e-05,
      "loss": 0.5066,
      "step": 7900
    },
    {
      "epoch": 0.6502795133179875,
      "grad_norm": 0.3593197464942932,
      "learning_rate": 2e-05,
      "loss": 0.4861,
      "step": 7910
    },
    {
      "epoch": 0.6511016113120685,
      "grad_norm": 0.3650665581226349,
      "learning_rate": 2e-05,
      "loss": 0.5159,
      "step": 7920
    },
    {
      "epoch": 0.6519237093061493,
      "grad_norm": 0.3694579303264618,
      "learning_rate": 2e-05,
      "loss": 0.5081,
      "step": 7930
    },
    {
      "epoch": 0.6527458073002302,
      "grad_norm": 0.3657929599285126,
      "learning_rate": 2e-05,
      "loss": 0.5244,
      "step": 7940
    },
    {
      "epoch": 0.6535679052943111,
      "grad_norm": 0.436565101146698,
      "learning_rate": 2e-05,
      "loss": 0.4739,
      "step": 7950
    },
    {
      "epoch": 0.654390003288392,
      "grad_norm": 0.38532212376594543,
      "learning_rate": 2e-05,
      "loss": 0.5215,
      "step": 7960
    },
    {
      "epoch": 0.6552121012824729,
      "grad_norm": 0.423753947019577,
      "learning_rate": 2e-05,
      "loss": 0.4934,
      "step": 7970
    },
    {
      "epoch": 0.6560341992765538,
      "grad_norm": 0.31683191657066345,
      "learning_rate": 2e-05,
      "loss": 0.514,
      "step": 7980
    },
    {
      "epoch": 0.6568562972706347,
      "grad_norm": 0.34565532207489014,
      "learning_rate": 2e-05,
      "loss": 0.4985,
      "step": 7990
    },
    {
      "epoch": 0.6576783952647156,
      "grad_norm": 0.3449639081954956,
      "learning_rate": 2e-05,
      "loss": 0.5045,
      "step": 8000
    },
    {
      "epoch": 0.6585004932587964,
      "grad_norm": 0.35967814922332764,
      "learning_rate": 2e-05,
      "loss": 0.5264,
      "step": 8010
    },
    {
      "epoch": 0.6593225912528773,
      "grad_norm": 0.35341161489486694,
      "learning_rate": 2e-05,
      "loss": 0.5075,
      "step": 8020
    },
    {
      "epoch": 0.6601446892469582,
      "grad_norm": 0.36577513813972473,
      "learning_rate": 2e-05,
      "loss": 0.5114,
      "step": 8030
    },
    {
      "epoch": 0.6609667872410391,
      "grad_norm": 0.4042232930660248,
      "learning_rate": 2e-05,
      "loss": 0.52,
      "step": 8040
    },
    {
      "epoch": 0.66178888523512,
      "grad_norm": 0.36703282594680786,
      "learning_rate": 2e-05,
      "loss": 0.4994,
      "step": 8050
    },
    {
      "epoch": 0.6626109832292009,
      "grad_norm": 0.33281126618385315,
      "learning_rate": 2e-05,
      "loss": 0.53,
      "step": 8060
    },
    {
      "epoch": 0.6634330812232818,
      "grad_norm": 0.43817874789237976,
      "learning_rate": 2e-05,
      "loss": 0.5218,
      "step": 8070
    },
    {
      "epoch": 0.6642551792173627,
      "grad_norm": 0.3248680531978607,
      "learning_rate": 2e-05,
      "loss": 0.546,
      "step": 8080
    },
    {
      "epoch": 0.6650772772114436,
      "grad_norm": 0.35722923278808594,
      "learning_rate": 2e-05,
      "loss": 0.5356,
      "step": 8090
    },
    {
      "epoch": 0.6658993752055244,
      "grad_norm": 0.3979380428791046,
      "learning_rate": 2e-05,
      "loss": 0.4944,
      "step": 8100
    },
    {
      "epoch": 0.6667214731996054,
      "grad_norm": 0.33463358879089355,
      "learning_rate": 2e-05,
      "loss": 0.5042,
      "step": 8110
    },
    {
      "epoch": 0.6675435711936862,
      "grad_norm": 0.3657388985157013,
      "learning_rate": 2e-05,
      "loss": 0.5138,
      "step": 8120
    },
    {
      "epoch": 0.6683656691877672,
      "grad_norm": 0.3372856080532074,
      "learning_rate": 2e-05,
      "loss": 0.5294,
      "step": 8130
    },
    {
      "epoch": 0.669187767181848,
      "grad_norm": 0.3818710446357727,
      "learning_rate": 2e-05,
      "loss": 0.5145,
      "step": 8140
    },
    {
      "epoch": 0.670009865175929,
      "grad_norm": 0.3899519741535187,
      "learning_rate": 2e-05,
      "loss": 0.5184,
      "step": 8150
    },
    {
      "epoch": 0.6708319631700098,
      "grad_norm": 0.5708285570144653,
      "learning_rate": 2e-05,
      "loss": 0.4873,
      "step": 8160
    },
    {
      "epoch": 0.6716540611640908,
      "grad_norm": 0.39678239822387695,
      "learning_rate": 2e-05,
      "loss": 0.4744,
      "step": 8170
    },
    {
      "epoch": 0.6724761591581716,
      "grad_norm": 0.38429194688796997,
      "learning_rate": 2e-05,
      "loss": 0.5196,
      "step": 8180
    },
    {
      "epoch": 0.6732982571522526,
      "grad_norm": 0.37952762842178345,
      "learning_rate": 2e-05,
      "loss": 0.4905,
      "step": 8190
    },
    {
      "epoch": 0.6741203551463334,
      "grad_norm": 0.32266339659690857,
      "learning_rate": 2e-05,
      "loss": 0.5257,
      "step": 8200
    },
    {
      "epoch": 0.6749424531404143,
      "grad_norm": 0.36519643664360046,
      "learning_rate": 2e-05,
      "loss": 0.5054,
      "step": 8210
    },
    {
      "epoch": 0.6757645511344952,
      "grad_norm": 0.3923785090446472,
      "learning_rate": 2e-05,
      "loss": 0.5029,
      "step": 8220
    },
    {
      "epoch": 0.6765866491285761,
      "grad_norm": 0.3507627248764038,
      "learning_rate": 2e-05,
      "loss": 0.4914,
      "step": 8230
    },
    {
      "epoch": 0.677408747122657,
      "grad_norm": 0.33368611335754395,
      "learning_rate": 2e-05,
      "loss": 0.533,
      "step": 8240
    },
    {
      "epoch": 0.6782308451167379,
      "grad_norm": 0.33496859669685364,
      "learning_rate": 2e-05,
      "loss": 0.5263,
      "step": 8250
    },
    {
      "epoch": 0.6790529431108188,
      "grad_norm": 0.38564130663871765,
      "learning_rate": 2e-05,
      "loss": 0.4809,
      "step": 8260
    },
    {
      "epoch": 0.6798750411048997,
      "grad_norm": 0.41036850214004517,
      "learning_rate": 2e-05,
      "loss": 0.5919,
      "step": 8270
    },
    {
      "epoch": 0.6806971390989806,
      "grad_norm": 0.43069151043891907,
      "learning_rate": 2e-05,
      "loss": 0.4873,
      "step": 8280
    },
    {
      "epoch": 0.6815192370930615,
      "grad_norm": 0.37820202112197876,
      "learning_rate": 2e-05,
      "loss": 0.5172,
      "step": 8290
    },
    {
      "epoch": 0.6823413350871423,
      "grad_norm": 0.3682824969291687,
      "learning_rate": 2e-05,
      "loss": 0.4875,
      "step": 8300
    },
    {
      "epoch": 0.6831634330812233,
      "grad_norm": 0.3456193506717682,
      "learning_rate": 2e-05,
      "loss": 0.4976,
      "step": 8310
    },
    {
      "epoch": 0.6839855310753041,
      "grad_norm": 0.3795250654220581,
      "learning_rate": 2e-05,
      "loss": 0.4917,
      "step": 8320
    },
    {
      "epoch": 0.6848076290693851,
      "grad_norm": 0.40203621983528137,
      "learning_rate": 2e-05,
      "loss": 0.5312,
      "step": 8330
    },
    {
      "epoch": 0.6856297270634659,
      "grad_norm": 0.35154861211776733,
      "learning_rate": 2e-05,
      "loss": 0.501,
      "step": 8340
    },
    {
      "epoch": 0.6864518250575469,
      "grad_norm": 0.4113715887069702,
      "learning_rate": 2e-05,
      "loss": 0.4898,
      "step": 8350
    },
    {
      "epoch": 0.6872739230516277,
      "grad_norm": 0.33968862891197205,
      "learning_rate": 2e-05,
      "loss": 0.5201,
      "step": 8360
    },
    {
      "epoch": 0.6880960210457087,
      "grad_norm": 0.37374597787857056,
      "learning_rate": 2e-05,
      "loss": 0.5087,
      "step": 8370
    },
    {
      "epoch": 0.6889181190397895,
      "grad_norm": 0.3904866874217987,
      "learning_rate": 2e-05,
      "loss": 0.5033,
      "step": 8380
    },
    {
      "epoch": 0.6897402170338705,
      "grad_norm": 0.3659636676311493,
      "learning_rate": 2e-05,
      "loss": 0.4989,
      "step": 8390
    },
    {
      "epoch": 0.6905623150279513,
      "grad_norm": 0.3810175657272339,
      "learning_rate": 2e-05,
      "loss": 0.5295,
      "step": 8400
    },
    {
      "epoch": 0.6913844130220322,
      "grad_norm": 0.40234172344207764,
      "learning_rate": 2e-05,
      "loss": 0.5174,
      "step": 8410
    },
    {
      "epoch": 0.6922065110161131,
      "grad_norm": 0.3226257264614105,
      "learning_rate": 2e-05,
      "loss": 0.5014,
      "step": 8420
    },
    {
      "epoch": 0.693028609010194,
      "grad_norm": 0.34616246819496155,
      "learning_rate": 2e-05,
      "loss": 0.5076,
      "step": 8430
    },
    {
      "epoch": 0.6938507070042749,
      "grad_norm": 0.3630603849887848,
      "learning_rate": 2e-05,
      "loss": 0.5207,
      "step": 8440
    },
    {
      "epoch": 0.6946728049983558,
      "grad_norm": 0.3440687358379364,
      "learning_rate": 2e-05,
      "loss": 0.5451,
      "step": 8450
    },
    {
      "epoch": 0.6954949029924367,
      "grad_norm": 0.43776196241378784,
      "learning_rate": 2e-05,
      "loss": 0.5257,
      "step": 8460
    },
    {
      "epoch": 0.6963170009865176,
      "grad_norm": 0.4499776363372803,
      "learning_rate": 2e-05,
      "loss": 0.4877,
      "step": 8470
    },
    {
      "epoch": 0.6971390989805984,
      "grad_norm": 0.3535178303718567,
      "learning_rate": 2e-05,
      "loss": 0.542,
      "step": 8480
    },
    {
      "epoch": 0.6979611969746794,
      "grad_norm": 0.39740246534347534,
      "learning_rate": 2e-05,
      "loss": 0.518,
      "step": 8490
    },
    {
      "epoch": 0.6987832949687602,
      "grad_norm": 0.35805901885032654,
      "learning_rate": 2e-05,
      "loss": 0.5118,
      "step": 8500
    },
    {
      "epoch": 0.6996053929628412,
      "grad_norm": 0.3331608772277832,
      "learning_rate": 2e-05,
      "loss": 0.4918,
      "step": 8510
    },
    {
      "epoch": 0.700427490956922,
      "grad_norm": 0.38446810841560364,
      "learning_rate": 2e-05,
      "loss": 0.54,
      "step": 8520
    },
    {
      "epoch": 0.701249588951003,
      "grad_norm": 0.3549647629261017,
      "learning_rate": 2e-05,
      "loss": 0.4938,
      "step": 8530
    },
    {
      "epoch": 0.7020716869450838,
      "grad_norm": 0.3349737524986267,
      "learning_rate": 2e-05,
      "loss": 0.5165,
      "step": 8540
    },
    {
      "epoch": 0.7028937849391648,
      "grad_norm": 0.37007954716682434,
      "learning_rate": 2e-05,
      "loss": 0.5355,
      "step": 8550
    },
    {
      "epoch": 0.7037158829332456,
      "grad_norm": 0.445494145154953,
      "learning_rate": 2e-05,
      "loss": 0.5155,
      "step": 8560
    },
    {
      "epoch": 0.7045379809273266,
      "grad_norm": 0.36776721477508545,
      "learning_rate": 2e-05,
      "loss": 0.501,
      "step": 8570
    },
    {
      "epoch": 0.7053600789214074,
      "grad_norm": 0.32858461141586304,
      "learning_rate": 2e-05,
      "loss": 0.4978,
      "step": 8580
    },
    {
      "epoch": 0.7061821769154883,
      "grad_norm": 0.3639601469039917,
      "learning_rate": 2e-05,
      "loss": 0.5048,
      "step": 8590
    },
    {
      "epoch": 0.7070042749095692,
      "grad_norm": 0.5855494141578674,
      "learning_rate": 2e-05,
      "loss": 0.5149,
      "step": 8600
    },
    {
      "epoch": 0.7078263729036501,
      "grad_norm": 0.36862504482269287,
      "learning_rate": 2e-05,
      "loss": 0.5094,
      "step": 8610
    },
    {
      "epoch": 0.708648470897731,
      "grad_norm": 0.36281338334083557,
      "learning_rate": 2e-05,
      "loss": 0.5154,
      "step": 8620
    },
    {
      "epoch": 0.7094705688918119,
      "grad_norm": 0.34635740518569946,
      "learning_rate": 2e-05,
      "loss": 0.51,
      "step": 8630
    },
    {
      "epoch": 0.7102926668858928,
      "grad_norm": 0.39140135049819946,
      "learning_rate": 2e-05,
      "loss": 0.4956,
      "step": 8640
    },
    {
      "epoch": 0.7111147648799737,
      "grad_norm": 0.4316105842590332,
      "learning_rate": 2e-05,
      "loss": 0.5174,
      "step": 8650
    },
    {
      "epoch": 0.7119368628740546,
      "grad_norm": 0.333999902009964,
      "learning_rate": 2e-05,
      "loss": 0.5069,
      "step": 8660
    },
    {
      "epoch": 0.7127589608681355,
      "grad_norm": 0.33733758330345154,
      "learning_rate": 2e-05,
      "loss": 0.488,
      "step": 8670
    },
    {
      "epoch": 0.7135810588622163,
      "grad_norm": 0.3597734868526459,
      "learning_rate": 2e-05,
      "loss": 0.5254,
      "step": 8680
    },
    {
      "epoch": 0.7144031568562973,
      "grad_norm": 0.40565571188926697,
      "learning_rate": 2e-05,
      "loss": 0.5001,
      "step": 8690
    },
    {
      "epoch": 0.7152252548503781,
      "grad_norm": 0.39464282989501953,
      "learning_rate": 2e-05,
      "loss": 0.5038,
      "step": 8700
    },
    {
      "epoch": 0.7160473528444591,
      "grad_norm": 0.3704948127269745,
      "learning_rate": 2e-05,
      "loss": 0.4997,
      "step": 8710
    },
    {
      "epoch": 0.7168694508385399,
      "grad_norm": 0.3553788959980011,
      "learning_rate": 2e-05,
      "loss": 0.4885,
      "step": 8720
    },
    {
      "epoch": 0.7176915488326209,
      "grad_norm": 0.3571178913116455,
      "learning_rate": 2e-05,
      "loss": 0.5151,
      "step": 8730
    },
    {
      "epoch": 0.7185136468267017,
      "grad_norm": 0.3622753620147705,
      "learning_rate": 2e-05,
      "loss": 0.5087,
      "step": 8740
    },
    {
      "epoch": 0.7193357448207827,
      "grad_norm": 0.31488320231437683,
      "learning_rate": 2e-05,
      "loss": 0.4718,
      "step": 8750
    },
    {
      "epoch": 0.7201578428148635,
      "grad_norm": 0.39543792605400085,
      "learning_rate": 2e-05,
      "loss": 0.5125,
      "step": 8760
    },
    {
      "epoch": 0.7209799408089445,
      "grad_norm": 0.36313340067863464,
      "learning_rate": 2e-05,
      "loss": 0.4671,
      "step": 8770
    },
    {
      "epoch": 0.7218020388030253,
      "grad_norm": 0.3390205502510071,
      "learning_rate": 2e-05,
      "loss": 0.4998,
      "step": 8780
    },
    {
      "epoch": 0.7226241367971062,
      "grad_norm": 0.36342447996139526,
      "learning_rate": 2e-05,
      "loss": 0.5148,
      "step": 8790
    },
    {
      "epoch": 0.7234462347911871,
      "grad_norm": 0.38755878806114197,
      "learning_rate": 2e-05,
      "loss": 0.5402,
      "step": 8800
    },
    {
      "epoch": 0.724268332785268,
      "grad_norm": 0.3711983263492584,
      "learning_rate": 2e-05,
      "loss": 0.5106,
      "step": 8810
    },
    {
      "epoch": 0.7250904307793489,
      "grad_norm": 0.3663436472415924,
      "learning_rate": 2e-05,
      "loss": 0.5131,
      "step": 8820
    },
    {
      "epoch": 0.7259125287734298,
      "grad_norm": 0.3488328754901886,
      "learning_rate": 2e-05,
      "loss": 0.5007,
      "step": 8830
    },
    {
      "epoch": 0.7267346267675107,
      "grad_norm": 0.3441716730594635,
      "learning_rate": 2e-05,
      "loss": 0.485,
      "step": 8840
    },
    {
      "epoch": 0.7275567247615916,
      "grad_norm": 0.3426259160041809,
      "learning_rate": 2e-05,
      "loss": 0.5023,
      "step": 8850
    },
    {
      "epoch": 0.7283788227556725,
      "grad_norm": 0.3334996998310089,
      "learning_rate": 2e-05,
      "loss": 0.5091,
      "step": 8860
    },
    {
      "epoch": 0.7292009207497534,
      "grad_norm": 0.39734187722206116,
      "learning_rate": 2e-05,
      "loss": 0.5241,
      "step": 8870
    },
    {
      "epoch": 0.7300230187438342,
      "grad_norm": 0.3918741047382355,
      "learning_rate": 2e-05,
      "loss": 0.4874,
      "step": 8880
    },
    {
      "epoch": 0.7308451167379152,
      "grad_norm": 0.37887081503868103,
      "learning_rate": 2e-05,
      "loss": 0.4935,
      "step": 8890
    },
    {
      "epoch": 0.731667214731996,
      "grad_norm": 0.33841803669929504,
      "learning_rate": 2e-05,
      "loss": 0.4733,
      "step": 8900
    },
    {
      "epoch": 0.732489312726077,
      "grad_norm": 0.3494199216365814,
      "learning_rate": 2e-05,
      "loss": 0.4979,
      "step": 8910
    },
    {
      "epoch": 0.7333114107201578,
      "grad_norm": 0.3692823350429535,
      "learning_rate": 2e-05,
      "loss": 0.5028,
      "step": 8920
    },
    {
      "epoch": 0.7341335087142388,
      "grad_norm": 0.39580994844436646,
      "learning_rate": 2e-05,
      "loss": 0.4979,
      "step": 8930
    },
    {
      "epoch": 0.7349556067083196,
      "grad_norm": 0.386745423078537,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 8940
    },
    {
      "epoch": 0.7357777047024006,
      "grad_norm": 0.35717809200286865,
      "learning_rate": 2e-05,
      "loss": 0.4884,
      "step": 8950
    },
    {
      "epoch": 0.7365998026964814,
      "grad_norm": 0.37287530303001404,
      "learning_rate": 2e-05,
      "loss": 0.5247,
      "step": 8960
    },
    {
      "epoch": 0.7374219006905623,
      "grad_norm": 0.45678460597991943,
      "learning_rate": 2e-05,
      "loss": 0.504,
      "step": 8970
    },
    {
      "epoch": 0.7382439986846432,
      "grad_norm": 0.33139747381210327,
      "learning_rate": 2e-05,
      "loss": 0.5159,
      "step": 8980
    },
    {
      "epoch": 0.7390660966787241,
      "grad_norm": 0.3368031680583954,
      "learning_rate": 2e-05,
      "loss": 0.524,
      "step": 8990
    },
    {
      "epoch": 0.739888194672805,
      "grad_norm": 0.4049912989139557,
      "learning_rate": 2e-05,
      "loss": 0.4921,
      "step": 9000
    },
    {
      "epoch": 0.7407102926668859,
      "grad_norm": 0.3987921476364136,
      "learning_rate": 2e-05,
      "loss": 0.5119,
      "step": 9010
    },
    {
      "epoch": 0.7415323906609668,
      "grad_norm": 0.35913628339767456,
      "learning_rate": 2e-05,
      "loss": 0.4984,
      "step": 9020
    },
    {
      "epoch": 0.7423544886550477,
      "grad_norm": 0.37858593463897705,
      "learning_rate": 2e-05,
      "loss": 0.4944,
      "step": 9030
    },
    {
      "epoch": 0.7431765866491286,
      "grad_norm": 0.5864289402961731,
      "learning_rate": 2e-05,
      "loss": 0.4928,
      "step": 9040
    },
    {
      "epoch": 0.7439986846432095,
      "grad_norm": 0.342980295419693,
      "learning_rate": 2e-05,
      "loss": 0.4839,
      "step": 9050
    },
    {
      "epoch": 0.7448207826372903,
      "grad_norm": 0.39347073435783386,
      "learning_rate": 2e-05,
      "loss": 0.5071,
      "step": 9060
    },
    {
      "epoch": 0.7456428806313713,
      "grad_norm": 0.3552831709384918,
      "learning_rate": 2e-05,
      "loss": 0.5141,
      "step": 9070
    },
    {
      "epoch": 0.7464649786254521,
      "grad_norm": 0.36579394340515137,
      "learning_rate": 2e-05,
      "loss": 0.5085,
      "step": 9080
    },
    {
      "epoch": 0.7472870766195331,
      "grad_norm": 0.3493228554725647,
      "learning_rate": 2e-05,
      "loss": 0.5007,
      "step": 9090
    },
    {
      "epoch": 0.7481091746136139,
      "grad_norm": 0.3832164704799652,
      "learning_rate": 2e-05,
      "loss": 0.5038,
      "step": 9100
    },
    {
      "epoch": 0.7489312726076949,
      "grad_norm": 0.3576708137989044,
      "learning_rate": 2e-05,
      "loss": 0.4902,
      "step": 9110
    },
    {
      "epoch": 0.7497533706017757,
      "grad_norm": 0.4196131229400635,
      "learning_rate": 2e-05,
      "loss": 0.4963,
      "step": 9120
    },
    {
      "epoch": 0.7505754685958567,
      "grad_norm": 0.39867278933525085,
      "learning_rate": 2e-05,
      "loss": 0.5124,
      "step": 9130
    },
    {
      "epoch": 0.7513975665899375,
      "grad_norm": 0.4016173481941223,
      "learning_rate": 2e-05,
      "loss": 0.5048,
      "step": 9140
    },
    {
      "epoch": 0.7522196645840185,
      "grad_norm": 0.3744872808456421,
      "learning_rate": 2e-05,
      "loss": 0.4967,
      "step": 9150
    },
    {
      "epoch": 0.7530417625780993,
      "grad_norm": 0.3431645333766937,
      "learning_rate": 2e-05,
      "loss": 0.5194,
      "step": 9160
    },
    {
      "epoch": 0.7538638605721802,
      "grad_norm": 0.41768670082092285,
      "learning_rate": 2e-05,
      "loss": 0.483,
      "step": 9170
    },
    {
      "epoch": 0.7546859585662611,
      "grad_norm": 0.4457896947860718,
      "learning_rate": 2e-05,
      "loss": 0.5131,
      "step": 9180
    },
    {
      "epoch": 0.755508056560342,
      "grad_norm": 0.3251597285270691,
      "learning_rate": 2e-05,
      "loss": 0.4695,
      "step": 9190
    },
    {
      "epoch": 0.7563301545544229,
      "grad_norm": 0.3738825023174286,
      "learning_rate": 2e-05,
      "loss": 0.5078,
      "step": 9200
    },
    {
      "epoch": 0.7571522525485038,
      "grad_norm": 0.3414279818534851,
      "learning_rate": 2e-05,
      "loss": 0.5069,
      "step": 9210
    },
    {
      "epoch": 0.7579743505425847,
      "grad_norm": 0.3814016878604889,
      "learning_rate": 2e-05,
      "loss": 0.4935,
      "step": 9220
    },
    {
      "epoch": 0.7587964485366656,
      "grad_norm": 0.38303613662719727,
      "learning_rate": 2e-05,
      "loss": 0.4908,
      "step": 9230
    },
    {
      "epoch": 0.7596185465307465,
      "grad_norm": 0.3854353129863739,
      "learning_rate": 2e-05,
      "loss": 0.4985,
      "step": 9240
    },
    {
      "epoch": 0.7604406445248274,
      "grad_norm": 0.37634292244911194,
      "learning_rate": 2e-05,
      "loss": 0.5238,
      "step": 9250
    },
    {
      "epoch": 0.7612627425189082,
      "grad_norm": 0.3755147457122803,
      "learning_rate": 2e-05,
      "loss": 0.5051,
      "step": 9260
    },
    {
      "epoch": 0.7620848405129892,
      "grad_norm": 0.4416508674621582,
      "learning_rate": 2e-05,
      "loss": 0.491,
      "step": 9270
    },
    {
      "epoch": 0.76290693850707,
      "grad_norm": 0.3706270456314087,
      "learning_rate": 2e-05,
      "loss": 0.5108,
      "step": 9280
    },
    {
      "epoch": 0.763729036501151,
      "grad_norm": 0.41300439834594727,
      "learning_rate": 2e-05,
      "loss": 0.495,
      "step": 9290
    },
    {
      "epoch": 0.7645511344952318,
      "grad_norm": 0.30869555473327637,
      "learning_rate": 2e-05,
      "loss": 0.4836,
      "step": 9300
    },
    {
      "epoch": 0.7653732324893128,
      "grad_norm": 0.4632079005241394,
      "learning_rate": 2e-05,
      "loss": 0.499,
      "step": 9310
    },
    {
      "epoch": 0.7661953304833936,
      "grad_norm": 0.3383975625038147,
      "learning_rate": 2e-05,
      "loss": 0.4705,
      "step": 9320
    },
    {
      "epoch": 0.7670174284774746,
      "grad_norm": 0.3273196518421173,
      "learning_rate": 2e-05,
      "loss": 0.5128,
      "step": 9330
    },
    {
      "epoch": 0.7678395264715554,
      "grad_norm": 0.34336844086647034,
      "learning_rate": 2e-05,
      "loss": 0.4835,
      "step": 9340
    },
    {
      "epoch": 0.7686616244656364,
      "grad_norm": 0.4296633005142212,
      "learning_rate": 2e-05,
      "loss": 0.503,
      "step": 9350
    },
    {
      "epoch": 0.7694837224597172,
      "grad_norm": 0.3802550137042999,
      "learning_rate": 2e-05,
      "loss": 0.5189,
      "step": 9360
    },
    {
      "epoch": 0.7703058204537981,
      "grad_norm": 0.36910149455070496,
      "learning_rate": 2e-05,
      "loss": 0.4948,
      "step": 9370
    },
    {
      "epoch": 0.771127918447879,
      "grad_norm": 0.3573198616504669,
      "learning_rate": 2e-05,
      "loss": 0.4885,
      "step": 9380
    },
    {
      "epoch": 0.7719500164419599,
      "grad_norm": 0.34607523679733276,
      "learning_rate": 2e-05,
      "loss": 0.4926,
      "step": 9390
    },
    {
      "epoch": 0.7727721144360408,
      "grad_norm": 0.3863622546195984,
      "learning_rate": 2e-05,
      "loss": 0.4973,
      "step": 9400
    },
    {
      "epoch": 0.7735942124301217,
      "grad_norm": 0.42345476150512695,
      "learning_rate": 2e-05,
      "loss": 0.5223,
      "step": 9410
    },
    {
      "epoch": 0.7744163104242026,
      "grad_norm": 0.41988059878349304,
      "learning_rate": 2e-05,
      "loss": 0.5202,
      "step": 9420
    },
    {
      "epoch": 0.7752384084182835,
      "grad_norm": 0.3765261173248291,
      "learning_rate": 2e-05,
      "loss": 0.5259,
      "step": 9430
    },
    {
      "epoch": 0.7760605064123643,
      "grad_norm": 0.3944740295410156,
      "learning_rate": 2e-05,
      "loss": 0.5103,
      "step": 9440
    },
    {
      "epoch": 0.7768826044064453,
      "grad_norm": 0.35089200735092163,
      "learning_rate": 2e-05,
      "loss": 0.5001,
      "step": 9450
    },
    {
      "epoch": 0.7777047024005261,
      "grad_norm": 0.3804813027381897,
      "learning_rate": 2e-05,
      "loss": 0.5052,
      "step": 9460
    },
    {
      "epoch": 0.7785268003946071,
      "grad_norm": 0.35043665766716003,
      "learning_rate": 2e-05,
      "loss": 0.5123,
      "step": 9470
    },
    {
      "epoch": 0.7793488983886879,
      "grad_norm": 0.3926442861557007,
      "learning_rate": 2e-05,
      "loss": 0.4918,
      "step": 9480
    },
    {
      "epoch": 0.7801709963827689,
      "grad_norm": 0.3696306347846985,
      "learning_rate": 2e-05,
      "loss": 0.5218,
      "step": 9490
    },
    {
      "epoch": 0.7809930943768497,
      "grad_norm": 0.3850564658641815,
      "learning_rate": 2e-05,
      "loss": 0.5108,
      "step": 9500
    },
    {
      "epoch": 0.7818151923709307,
      "grad_norm": 0.3178589344024658,
      "learning_rate": 2e-05,
      "loss": 0.4651,
      "step": 9510
    },
    {
      "epoch": 0.7826372903650115,
      "grad_norm": 0.37762412428855896,
      "learning_rate": 2e-05,
      "loss": 0.5112,
      "step": 9520
    },
    {
      "epoch": 0.7834593883590925,
      "grad_norm": 0.3568861186504364,
      "learning_rate": 2e-05,
      "loss": 0.5242,
      "step": 9530
    },
    {
      "epoch": 0.7842814863531733,
      "grad_norm": 0.3443784713745117,
      "learning_rate": 2e-05,
      "loss": 0.5151,
      "step": 9540
    },
    {
      "epoch": 0.7851035843472542,
      "grad_norm": 0.4007415771484375,
      "learning_rate": 2e-05,
      "loss": 0.5052,
      "step": 9550
    },
    {
      "epoch": 0.7859256823413351,
      "grad_norm": 0.3236842155456543,
      "learning_rate": 2e-05,
      "loss": 0.4916,
      "step": 9560
    },
    {
      "epoch": 0.786747780335416,
      "grad_norm": 0.34756213426589966,
      "learning_rate": 2e-05,
      "loss": 0.4942,
      "step": 9570
    },
    {
      "epoch": 0.7875698783294969,
      "grad_norm": 0.33382245898246765,
      "learning_rate": 2e-05,
      "loss": 0.5103,
      "step": 9580
    },
    {
      "epoch": 0.7883919763235778,
      "grad_norm": 0.3905704915523529,
      "learning_rate": 2e-05,
      "loss": 0.5178,
      "step": 9590
    },
    {
      "epoch": 0.7892140743176587,
      "grad_norm": 0.33488261699676514,
      "learning_rate": 2e-05,
      "loss": 0.5035,
      "step": 9600
    },
    {
      "epoch": 0.7900361723117395,
      "grad_norm": 0.38006189465522766,
      "learning_rate": 2e-05,
      "loss": 0.4984,
      "step": 9610
    },
    {
      "epoch": 0.7908582703058205,
      "grad_norm": 0.36203011870384216,
      "learning_rate": 2e-05,
      "loss": 0.4933,
      "step": 9620
    },
    {
      "epoch": 0.7916803682999013,
      "grad_norm": 0.335334837436676,
      "learning_rate": 2e-05,
      "loss": 0.5143,
      "step": 9630
    },
    {
      "epoch": 0.7925024662939822,
      "grad_norm": 0.3651992678642273,
      "learning_rate": 2e-05,
      "loss": 0.5214,
      "step": 9640
    },
    {
      "epoch": 0.7933245642880631,
      "grad_norm": 0.35433998703956604,
      "learning_rate": 2e-05,
      "loss": 0.5152,
      "step": 9650
    },
    {
      "epoch": 0.794146662282144,
      "grad_norm": 0.37448662519454956,
      "learning_rate": 2e-05,
      "loss": 0.5407,
      "step": 9660
    },
    {
      "epoch": 0.7949687602762249,
      "grad_norm": 0.4155105650424957,
      "learning_rate": 2e-05,
      "loss": 0.5261,
      "step": 9670
    },
    {
      "epoch": 0.7957908582703058,
      "grad_norm": 0.35761407017707825,
      "learning_rate": 2e-05,
      "loss": 0.4952,
      "step": 9680
    },
    {
      "epoch": 0.7966129562643867,
      "grad_norm": 0.37693333625793457,
      "learning_rate": 2e-05,
      "loss": 0.5377,
      "step": 9690
    },
    {
      "epoch": 0.7974350542584676,
      "grad_norm": 0.37470677495002747,
      "learning_rate": 2e-05,
      "loss": 0.4983,
      "step": 9700
    },
    {
      "epoch": 0.7982571522525485,
      "grad_norm": 0.36826571822166443,
      "learning_rate": 2e-05,
      "loss": 0.5197,
      "step": 9710
    },
    {
      "epoch": 0.7990792502466294,
      "grad_norm": 0.3359779417514801,
      "learning_rate": 2e-05,
      "loss": 0.4686,
      "step": 9720
    },
    {
      "epoch": 0.7999013482407102,
      "grad_norm": 0.3511999249458313,
      "learning_rate": 2e-05,
      "loss": 0.5093,
      "step": 9730
    },
    {
      "epoch": 0.8007234462347912,
      "grad_norm": 0.3566172420978546,
      "learning_rate": 2e-05,
      "loss": 0.5475,
      "step": 9740
    },
    {
      "epoch": 0.801545544228872,
      "grad_norm": 0.4056861102581024,
      "learning_rate": 2e-05,
      "loss": 0.5156,
      "step": 9750
    },
    {
      "epoch": 0.802367642222953,
      "grad_norm": 0.4533374011516571,
      "learning_rate": 2e-05,
      "loss": 0.5175,
      "step": 9760
    },
    {
      "epoch": 0.8031897402170338,
      "grad_norm": 0.3790939748287201,
      "learning_rate": 2e-05,
      "loss": 0.4827,
      "step": 9770
    },
    {
      "epoch": 0.8040118382111148,
      "grad_norm": 0.4070254862308502,
      "learning_rate": 2e-05,
      "loss": 0.4929,
      "step": 9780
    },
    {
      "epoch": 0.8048339362051956,
      "grad_norm": 0.3422532081604004,
      "learning_rate": 2e-05,
      "loss": 0.4922,
      "step": 9790
    },
    {
      "epoch": 0.8056560341992766,
      "grad_norm": 0.34460949897766113,
      "learning_rate": 2e-05,
      "loss": 0.5269,
      "step": 9800
    },
    {
      "epoch": 0.8064781321933574,
      "grad_norm": 0.35433098673820496,
      "learning_rate": 2e-05,
      "loss": 0.5054,
      "step": 9810
    },
    {
      "epoch": 0.8073002301874384,
      "grad_norm": 0.3727871775627136,
      "learning_rate": 2e-05,
      "loss": 0.492,
      "step": 9820
    },
    {
      "epoch": 0.8081223281815192,
      "grad_norm": 0.3413172662258148,
      "learning_rate": 2e-05,
      "loss": 0.4864,
      "step": 9830
    },
    {
      "epoch": 0.8089444261756001,
      "grad_norm": 0.35182371735572815,
      "learning_rate": 2e-05,
      "loss": 0.4955,
      "step": 9840
    },
    {
      "epoch": 0.809766524169681,
      "grad_norm": 0.3433065414428711,
      "learning_rate": 2e-05,
      "loss": 0.4569,
      "step": 9850
    },
    {
      "epoch": 0.8105886221637619,
      "grad_norm": 0.40070515871047974,
      "learning_rate": 2e-05,
      "loss": 0.4886,
      "step": 9860
    },
    {
      "epoch": 0.8114107201578428,
      "grad_norm": 0.4108491539955139,
      "learning_rate": 2e-05,
      "loss": 0.4909,
      "step": 9870
    },
    {
      "epoch": 0.8122328181519237,
      "grad_norm": 0.38729992508888245,
      "learning_rate": 2e-05,
      "loss": 0.5166,
      "step": 9880
    },
    {
      "epoch": 0.8130549161460046,
      "grad_norm": 0.36251741647720337,
      "learning_rate": 2e-05,
      "loss": 0.5084,
      "step": 9890
    },
    {
      "epoch": 0.8138770141400855,
      "grad_norm": 0.34179598093032837,
      "learning_rate": 2e-05,
      "loss": 0.4977,
      "step": 9900
    },
    {
      "epoch": 0.8146991121341663,
      "grad_norm": 0.38883981108665466,
      "learning_rate": 2e-05,
      "loss": 0.5223,
      "step": 9910
    },
    {
      "epoch": 0.8155212101282473,
      "grad_norm": 0.44009190797805786,
      "learning_rate": 2e-05,
      "loss": 0.5189,
      "step": 9920
    },
    {
      "epoch": 0.8163433081223281,
      "grad_norm": 0.37525853514671326,
      "learning_rate": 2e-05,
      "loss": 0.5115,
      "step": 9930
    },
    {
      "epoch": 0.8171654061164091,
      "grad_norm": 0.40035396814346313,
      "learning_rate": 2e-05,
      "loss": 0.516,
      "step": 9940
    },
    {
      "epoch": 0.8179875041104899,
      "grad_norm": 0.3889773190021515,
      "learning_rate": 2e-05,
      "loss": 0.5,
      "step": 9950
    },
    {
      "epoch": 0.8188096021045709,
      "grad_norm": 0.3463926613330841,
      "learning_rate": 2e-05,
      "loss": 0.5139,
      "step": 9960
    },
    {
      "epoch": 0.8196317000986517,
      "grad_norm": 0.3888430595397949,
      "learning_rate": 2e-05,
      "loss": 0.4992,
      "step": 9970
    },
    {
      "epoch": 0.8204537980927327,
      "grad_norm": 0.3606644570827484,
      "learning_rate": 2e-05,
      "loss": 0.499,
      "step": 9980
    },
    {
      "epoch": 0.8212758960868135,
      "grad_norm": 0.4134714603424072,
      "learning_rate": 2e-05,
      "loss": 0.482,
      "step": 9990
    },
    {
      "epoch": 0.8220979940808945,
      "grad_norm": 0.35072022676467896,
      "learning_rate": 2e-05,
      "loss": 0.501,
      "step": 10000
    },
    {
      "epoch": 0.8229200920749753,
      "grad_norm": 0.39270690083503723,
      "learning_rate": 2e-05,
      "loss": 0.4829,
      "step": 10010
    },
    {
      "epoch": 0.8237421900690562,
      "grad_norm": 0.3883296847343445,
      "learning_rate": 2e-05,
      "loss": 0.515,
      "step": 10020
    },
    {
      "epoch": 0.8245642880631371,
      "grad_norm": 0.40071243047714233,
      "learning_rate": 2e-05,
      "loss": 0.5151,
      "step": 10030
    },
    {
      "epoch": 0.825386386057218,
      "grad_norm": 0.39978209137916565,
      "learning_rate": 2e-05,
      "loss": 0.4621,
      "step": 10040
    },
    {
      "epoch": 0.8262084840512989,
      "grad_norm": 0.3997681438922882,
      "learning_rate": 2e-05,
      "loss": 0.4945,
      "step": 10050
    },
    {
      "epoch": 0.8270305820453798,
      "grad_norm": 0.4024221897125244,
      "learning_rate": 2e-05,
      "loss": 0.4847,
      "step": 10060
    },
    {
      "epoch": 0.8278526800394607,
      "grad_norm": 0.38962456583976746,
      "learning_rate": 2e-05,
      "loss": 0.5183,
      "step": 10070
    },
    {
      "epoch": 0.8286747780335416,
      "grad_norm": 0.39697688817977905,
      "learning_rate": 2e-05,
      "loss": 0.4806,
      "step": 10080
    },
    {
      "epoch": 0.8294968760276225,
      "grad_norm": 0.41418227553367615,
      "learning_rate": 2e-05,
      "loss": 0.4961,
      "step": 10090
    },
    {
      "epoch": 0.8303189740217034,
      "grad_norm": 0.3687451481819153,
      "learning_rate": 2e-05,
      "loss": 0.5016,
      "step": 10100
    },
    {
      "epoch": 0.8311410720157842,
      "grad_norm": 0.3828966021537781,
      "learning_rate": 2e-05,
      "loss": 0.4788,
      "step": 10110
    },
    {
      "epoch": 0.8319631700098652,
      "grad_norm": 0.36220890283584595,
      "learning_rate": 2e-05,
      "loss": 0.5094,
      "step": 10120
    },
    {
      "epoch": 0.832785268003946,
      "grad_norm": 0.36459845304489136,
      "learning_rate": 2e-05,
      "loss": 0.5091,
      "step": 10130
    },
    {
      "epoch": 0.833607365998027,
      "grad_norm": 0.43089812994003296,
      "learning_rate": 2e-05,
      "loss": 0.4652,
      "step": 10140
    },
    {
      "epoch": 0.8344294639921078,
      "grad_norm": 0.3603317141532898,
      "learning_rate": 2e-05,
      "loss": 0.4917,
      "step": 10150
    },
    {
      "epoch": 0.8352515619861888,
      "grad_norm": 0.4343354105949402,
      "learning_rate": 2e-05,
      "loss": 0.4729,
      "step": 10160
    },
    {
      "epoch": 0.8360736599802696,
      "grad_norm": 0.42944860458374023,
      "learning_rate": 2e-05,
      "loss": 0.5077,
      "step": 10170
    },
    {
      "epoch": 0.8368957579743506,
      "grad_norm": 0.3678496479988098,
      "learning_rate": 1.999957274187931e-05,
      "loss": 0.5026,
      "step": 10180
    },
    {
      "epoch": 0.8377178559684314,
      "grad_norm": 0.4069693684577942,
      "learning_rate": 1.9991978062682906e-05,
      "loss": 0.4882,
      "step": 10190
    },
    {
      "epoch": 0.8385399539625124,
      "grad_norm": 0.4372115135192871,
      "learning_rate": 1.997489709987378e-05,
      "loss": 0.5085,
      "step": 10200
    },
    {
      "epoch": 0.8393620519565932,
      "grad_norm": 0.39724889397621155,
      "learning_rate": 1.9948346151502135e-05,
      "loss": 0.5352,
      "step": 10210
    },
    {
      "epoch": 0.8401841499506741,
      "grad_norm": 0.3725707530975342,
      "learning_rate": 1.991235055154325e-05,
      "loss": 0.498,
      "step": 10220
    },
    {
      "epoch": 0.841006247944755,
      "grad_norm": 0.38640648126602173,
      "learning_rate": 1.9866944645724693e-05,
      "loss": 0.5018,
      "step": 10230
    },
    {
      "epoch": 0.8418283459388359,
      "grad_norm": 0.40359997749328613,
      "learning_rate": 1.9812171758754842e-05,
      "loss": 0.4924,
      "step": 10240
    },
    {
      "epoch": 0.8426504439329168,
      "grad_norm": 0.3489453196525574,
      "learning_rate": 1.9748084152983987e-05,
      "loss": 0.4729,
      "step": 10250
    },
    {
      "epoch": 0.8434725419269977,
      "grad_norm": 0.4023876190185547,
      "learning_rate": 1.967474297853744e-05,
      "loss": 0.4988,
      "step": 10260
    },
    {
      "epoch": 0.8442946399210786,
      "grad_norm": 0.3495236337184906,
      "learning_rate": 1.9592218214968256e-05,
      "loss": 0.5056,
      "step": 10270
    },
    {
      "epoch": 0.8451167379151595,
      "grad_norm": 0.3577699363231659,
      "learning_rate": 1.9500588604485213e-05,
      "loss": 0.5312,
      "step": 10280
    },
    {
      "epoch": 0.8459388359092403,
      "grad_norm": 0.41998547315597534,
      "learning_rate": 1.939994157681979e-05,
      "loss": 0.4994,
      "step": 10290
    },
    {
      "epoch": 0.8467609339033213,
      "grad_norm": 0.3391339182853699,
      "learning_rate": 1.9290373165803803e-05,
      "loss": 0.5153,
      "step": 10300
    },
    {
      "epoch": 0.8475830318974021,
      "grad_norm": 0.34198668599128723,
      "learning_rate": 1.9171987917737328e-05,
      "loss": 0.5095,
      "step": 10310
    },
    {
      "epoch": 0.8484051298914831,
      "grad_norm": 0.43322861194610596,
      "learning_rate": 1.9044898791634322e-05,
      "loss": 0.4983,
      "step": 10320
    },
    {
      "epoch": 0.8492272278855639,
      "grad_norm": 0.3970557451248169,
      "learning_rate": 1.8909227051441126e-05,
      "loss": 0.4937,
      "step": 10330
    },
    {
      "epoch": 0.8500493258796449,
      "grad_norm": 0.38077935576438904,
      "learning_rate": 1.8765102150330692e-05,
      "loss": 0.4876,
      "step": 10340
    },
    {
      "epoch": 0.8508714238737257,
      "grad_norm": 0.3480050563812256,
      "learning_rate": 1.861266160718297e-05,
      "loss": 0.507,
      "step": 10350
    },
    {
      "epoch": 0.8516935218678067,
      "grad_norm": 0.4220784306526184,
      "learning_rate": 1.845205087536923e-05,
      "loss": 0.5377,
      "step": 10360
    },
    {
      "epoch": 0.8525156198618875,
      "grad_norm": 0.3862803876399994,
      "learning_rate": 1.8283423203965633e-05,
      "loss": 0.4924,
      "step": 10370
    },
    {
      "epoch": 0.8533377178559685,
      "grad_norm": 0.46266141533851624,
      "learning_rate": 1.810693949152835e-05,
      "loss": 0.4753,
      "step": 10380
    },
    {
      "epoch": 0.8541598158500493,
      "grad_norm": 0.4024084508419037,
      "learning_rate": 1.7922768132569902e-05,
      "loss": 0.5299,
      "step": 10390
    },
    {
      "epoch": 0.8549819138441302,
      "grad_norm": 0.3871736526489258,
      "learning_rate": 1.773108485688302e-05,
      "loss": 0.4923,
      "step": 10400
    },
    {
      "epoch": 0.8558040118382111,
      "grad_norm": 0.3565307855606079,
      "learning_rate": 1.7532072561865537e-05,
      "loss": 0.4867,
      "step": 10410
    },
    {
      "epoch": 0.856626109832292,
      "grad_norm": 0.3408243656158447,
      "learning_rate": 1.732592113800616e-05,
      "loss": 0.5045,
      "step": 10420
    },
    {
      "epoch": 0.8574482078263729,
      "grad_norm": 0.3976184129714966,
      "learning_rate": 1.7112827287697668e-05,
      "loss": 0.492,
      "step": 10430
    },
    {
      "epoch": 0.8582703058204538,
      "grad_norm": 0.3392387926578522,
      "learning_rate": 1.6892994337550452e-05,
      "loss": 0.4724,
      "step": 10440
    },
    {
      "epoch": 0.8590924038145347,
      "grad_norm": 0.412242591381073,
      "learning_rate": 1.6666632044385487e-05,
      "loss": 0.5202,
      "step": 10450
    },
    {
      "epoch": 0.8599145018086156,
      "grad_norm": 0.32808980345726013,
      "learning_rate": 1.6433956395091757e-05,
      "loss": 0.4974,
      "step": 10460
    },
    {
      "epoch": 0.8607365998026965,
      "grad_norm": 0.3927038609981537,
      "learning_rate": 1.6195189400539238e-05,
      "loss": 0.5116,
      "step": 10470
    },
    {
      "epoch": 0.8615586977967774,
      "grad_norm": 0.4286595284938812,
      "learning_rate": 1.595055888374398e-05,
      "loss": 0.4981,
      "step": 10480
    },
    {
      "epoch": 0.8623807957908582,
      "grad_norm": 0.3861750364303589,
      "learning_rate": 1.5700298262487434e-05,
      "loss": 0.5167,
      "step": 10490
    },
    {
      "epoch": 0.8632028937849392,
      "grad_norm": 0.38458073139190674,
      "learning_rate": 1.5444646326597467e-05,
      "loss": 0.5213,
      "step": 10500
    },
    {
      "epoch": 0.86402499177902,
      "grad_norm": 0.3661424219608307,
      "learning_rate": 1.5183847010103586e-05,
      "loss": 0.5231,
      "step": 10510
    },
    {
      "epoch": 0.864847089773101,
      "grad_norm": 0.35300156474113464,
      "learning_rate": 1.4918149158483686e-05,
      "loss": 0.4915,
      "step": 10520
    },
    {
      "epoch": 0.8656691877671818,
      "grad_norm": 0.357980340719223,
      "learning_rate": 1.464780629122452e-05,
      "loss": 0.5199,
      "step": 10530
    },
    {
      "epoch": 0.8664912857612628,
      "grad_norm": 0.37720081210136414,
      "learning_rate": 1.4373076359922374e-05,
      "loss": 0.5058,
      "step": 10540
    },
    {
      "epoch": 0.8673133837553436,
      "grad_norm": 0.35986006259918213,
      "learning_rate": 1.4094221502154774e-05,
      "loss": 0.5016,
      "step": 10550
    },
    {
      "epoch": 0.8681354817494246,
      "grad_norm": 0.40953224897384644,
      "learning_rate": 1.3811507791358113e-05,
      "loss": 0.4828,
      "step": 10560
    },
    {
      "epoch": 0.8689575797435054,
      "grad_norm": 0.45023179054260254,
      "learning_rate": 1.3525204982949753e-05,
      "loss": 0.4963,
      "step": 10570
    },
    {
      "epoch": 0.8697796777375864,
      "grad_norm": 0.4509366750717163,
      "learning_rate": 1.3235586256936996e-05,
      "loss": 0.4885,
      "step": 10580
    },
    {
      "epoch": 0.8706017757316672,
      "grad_norm": 0.34612149000167847,
      "learning_rate": 1.294292795725836e-05,
      "loss": 0.5087,
      "step": 10590
    },
    {
      "epoch": 0.8714238737257481,
      "grad_norm": 0.3661915957927704,
      "learning_rate": 1.2647509328105978e-05,
      "loss": 0.4889,
      "step": 10600
    },
    {
      "epoch": 0.872245971719829,
      "grad_norm": 0.3972892165184021,
      "learning_rate": 1.2349612247480686e-05,
      "loss": 0.4742,
      "step": 10610
    },
    {
      "epoch": 0.8730680697139099,
      "grad_norm": 0.3790651261806488,
      "learning_rate": 1.2049520958234e-05,
      "loss": 0.491,
      "step": 10620
    },
    {
      "epoch": 0.8738901677079908,
      "grad_norm": 0.36708495020866394,
      "learning_rate": 1.1747521796853638e-05,
      "loss": 0.4843,
      "step": 10630
    },
    {
      "epoch": 0.8747122657020717,
      "grad_norm": 0.34987008571624756,
      "learning_rate": 1.1443902920251395e-05,
      "loss": 0.499,
      "step": 10640
    },
    {
      "epoch": 0.8755343636961526,
      "grad_norm": 0.3802943229675293,
      "learning_rate": 1.1138954030814009e-05,
      "loss": 0.4821,
      "step": 10650
    },
    {
      "epoch": 0.8763564616902335,
      "grad_norm": 0.3647462725639343,
      "learning_rate": 1.0832966099979388e-05,
      "loss": 0.5257,
      "step": 10660
    },
    {
      "epoch": 0.8771785596843144,
      "grad_norm": 0.35518011450767517,
      "learning_rate": 1.0526231090601969e-05,
      "loss": 0.4773,
      "step": 10670
    },
    {
      "epoch": 0.8780006576783953,
      "grad_norm": 0.35710275173187256,
      "learning_rate": 1.0219041678372106e-05,
      "loss": 0.4942,
      "step": 10680
    },
    {
      "epoch": 0.8788227556724761,
      "grad_norm": 0.36514362692832947,
      "learning_rate": 9.911690972555268e-06,
      "loss": 0.4967,
      "step": 10690
    },
    {
      "epoch": 0.8796448536665571,
      "grad_norm": 0.32222339510917664,
      "learning_rate": 9.604472236317541e-06,
      "loss": 0.4721,
      "step": 10700
    },
    {
      "epoch": 0.8804669516606379,
      "grad_norm": 0.3266163468360901,
      "learning_rate": 9.297678606904335e-06,
      "loss": 0.4956,
      "step": 10710
    },
    {
      "epoch": 0.8812890496547189,
      "grad_norm": 0.38403791189193726,
      "learning_rate": 8.991602815939155e-06,
      "loss": 0.4957,
      "step": 10720
    },
    {
      "epoch": 0.8821111476487997,
      "grad_norm": 0.44961464405059814,
      "learning_rate": 8.686536910109464e-06,
      "loss": 0.4822,
      "step": 10730
    },
    {
      "epoch": 0.8829332456428807,
      "grad_norm": 0.3640117049217224,
      "learning_rate": 8.382771972506084e-06,
      "loss": 0.4984,
      "step": 10740
    },
    {
      "epoch": 0.8837553436369615,
      "grad_norm": 0.34173232316970825,
      "learning_rate": 8.08059784488198e-06,
      "loss": 0.4845,
      "step": 10750
    },
    {
      "epoch": 0.8845774416310425,
      "grad_norm": 0.4196718633174896,
      "learning_rate": 7.780302851095544e-06,
      "loss": 0.5055,
      "step": 10760
    },
    {
      "epoch": 0.8853995396251233,
      "grad_norm": 0.37520596385002136,
      "learning_rate": 7.482173522002166e-06,
      "loss": 0.4911,
      "step": 10770
    },
    {
      "epoch": 0.8862216376192042,
      "grad_norm": 0.38429462909698486,
      "learning_rate": 7.186494322056624e-06,
      "loss": 0.4848,
      "step": 10780
    },
    {
      "epoch": 0.8870437356132851,
      "grad_norm": 0.32563698291778564,
      "learning_rate": 6.893547377887183e-06,
      "loss": 0.4774,
      "step": 10790
    },
    {
      "epoch": 0.887865833607366,
      "grad_norm": 0.3838485777378082,
      "learning_rate": 6.6036122091003365e-06,
      "loss": 0.4976,
      "step": 10800
    },
    {
      "epoch": 0.8886879316014469,
      "grad_norm": 0.3423444628715515,
      "learning_rate": 6.316965461573097e-06,
      "loss": 0.4918,
      "step": 10810
    },
    {
      "epoch": 0.8895100295955278,
      "grad_norm": 0.3551524877548218,
      "learning_rate": 6.0338806434872965e-06,
      "loss": 0.467,
      "step": 10820
    },
    {
      "epoch": 0.8903321275896087,
      "grad_norm": 0.3420228958129883,
      "learning_rate": 5.754627864357749e-06,
      "loss": 0.4652,
      "step": 10830
    },
    {
      "epoch": 0.8911542255836896,
      "grad_norm": 0.4045047461986542,
      "learning_rate": 5.479473577303319e-06,
      "loss": 0.4692,
      "step": 10840
    },
    {
      "epoch": 0.8919763235777705,
      "grad_norm": 0.3642389476299286,
      "learning_rate": 5.208680324806778e-06,
      "loss": 0.4667,
      "step": 10850
    },
    {
      "epoch": 0.8927984215718514,
      "grad_norm": 0.3561105728149414,
      "learning_rate": 4.94250648820606e-06,
      "loss": 0.4823,
      "step": 10860
    },
    {
      "epoch": 0.8936205195659322,
      "grad_norm": 0.3457697927951813,
      "learning_rate": 4.681206041155951e-06,
      "loss": 0.4819,
      "step": 10870
    },
    {
      "epoch": 0.8944426175600132,
      "grad_norm": 0.3958311676979065,
      "learning_rate": 4.42502830729538e-06,
      "loss": 0.4923,
      "step": 10880
    },
    {
      "epoch": 0.895264715554094,
      "grad_norm": 0.35813573002815247,
      "learning_rate": 4.174217722351669e-06,
      "loss": 0.5102,
      "step": 10890
    },
    {
      "epoch": 0.896086813548175,
      "grad_norm": 0.3853065073490143,
      "learning_rate": 3.929013600908575e-06,
      "loss": 0.5052,
      "step": 10900
    },
    {
      "epoch": 0.8969089115422558,
      "grad_norm": 0.36961492896080017,
      "learning_rate": 3.689649908060807e-06,
      "loss": 0.5013,
      "step": 10910
    },
    {
      "epoch": 0.8977310095363368,
      "grad_norm": 0.3522484600543976,
      "learning_rate": 3.4563550361728028e-06,
      "loss": 0.4642,
      "step": 10920
    },
    {
      "epoch": 0.8985531075304176,
      "grad_norm": 0.3807116746902466,
      "learning_rate": 3.2293515869548216e-06,
      "loss": 0.4849,
      "step": 10930
    },
    {
      "epoch": 0.8993752055244986,
      "grad_norm": 0.3610656261444092,
      "learning_rate": 3.0088561590642746e-06,
      "loss": 0.4599,
      "step": 10940
    },
    {
      "epoch": 0.9001973035185794,
      "grad_norm": 0.3911413848400116,
      "learning_rate": 2.795079141434941e-06,
      "loss": 0.4826,
      "step": 10950
    },
    {
      "epoch": 0.9010194015126604,
      "grad_norm": 0.304314523935318,
      "learning_rate": 2.588224512531299e-06,
      "loss": 0.4722,
      "step": 10960
    },
    {
      "epoch": 0.9018414995067412,
      "grad_norm": 0.35844188928604126,
      "learning_rate": 2.388489645719495e-06,
      "loss": 0.4951,
      "step": 10970
    },
    {
      "epoch": 0.9026635975008221,
      "grad_norm": 0.36753714084625244,
      "learning_rate": 2.196065120940663e-06,
      "loss": 0.5232,
      "step": 10980
    },
    {
      "epoch": 0.903485695494903,
      "grad_norm": 0.35546278953552246,
      "learning_rate": 2.0111345428662808e-06,
      "loss": 0.4908,
      "step": 10990
    },
    {
      "epoch": 0.9043077934889839,
      "grad_norm": 0.3371787667274475,
      "learning_rate": 1.8338743657091102e-06,
      "loss": 0.4759,
      "step": 11000
    },
    {
      "epoch": 0.9051298914830648,
      "grad_norm": 0.36303243041038513,
      "learning_rate": 1.6644537248568113e-06,
      "loss": 0.4835,
      "step": 11010
    },
    {
      "epoch": 0.9059519894771457,
      "grad_norm": 0.3602112829685211,
      "learning_rate": 1.503034275488951e-06,
      "loss": 0.5041,
      "step": 11020
    },
    {
      "epoch": 0.9067740874712266,
      "grad_norm": 0.3440713882446289,
      "learning_rate": 1.3497700383313495e-06,
      "loss": 0.4791,
      "step": 11030
    },
    {
      "epoch": 0.9075961854653075,
      "grad_norm": 0.34699884057044983,
      "learning_rate": 1.2048072526949562e-06,
      "loss": 0.4805,
      "step": 11040
    },
    {
      "epoch": 0.9084182834593884,
      "grad_norm": 0.41581302881240845,
      "learning_rate": 1.0682842369394756e-06,
      "loss": 0.4717,
      "step": 11050
    },
    {
      "epoch": 0.9092403814534693,
      "grad_norm": 0.3425777852535248,
      "learning_rate": 9.403312564948792e-07,
      "loss": 0.4739,
      "step": 11060
    },
    {
      "epoch": 0.9100624794475501,
      "grad_norm": 0.33630093932151794,
      "learning_rate": 8.210703995667471e-07,
      "loss": 0.5151,
      "step": 11070
    },
    {
      "epoch": 0.9108845774416311,
      "grad_norm": 0.3680729568004608,
      "learning_rate": 7.106154606440245e-07,
      "loss": 0.506,
      "step": 11080
    },
    {
      "epoch": 0.9117066754357119,
      "grad_norm": 0.36699941754341125,
      "learning_rate": 6.090718319203393e-07,
      "loss": 0.4659,
      "step": 11090
    },
    {
      "epoch": 0.9125287734297929,
      "grad_norm": 0.3412867784500122,
      "learning_rate": 5.165364027325089e-07,
      "loss": 0.4984,
      "step": 11100
    },
    {
      "epoch": 0.9133508714238737,
      "grad_norm": 0.37780678272247314,
      "learning_rate": 4.330974671121585e-07,
      "loss": 0.4926,
      "step": 11110
    },
    {
      "epoch": 0.9141729694179547,
      "grad_norm": 0.3810194432735443,
      "learning_rate": 3.588346395386846e-07,
      "loss": 0.4883,
      "step": 11120
    },
    {
      "epoch": 0.9149950674120355,
      "grad_norm": 0.3762509226799011,
      "learning_rate": 2.938187789739395e-07,
      "loss": 0.4812,
      "step": 11130
    },
    {
      "epoch": 0.9158171654061165,
      "grad_norm": 0.4267049729824066,
      "learning_rate": 2.3811192125110724e-07,
      "loss": 0.4897,
      "step": 11140
    },
    {
      "epoch": 0.9166392634001973,
      "grad_norm": 0.37740764021873474,
      "learning_rate": 1.917672198823177e-07,
      "loss": 0.4817,
      "step": 11150
    },
    {
      "epoch": 0.9174613613942783,
      "grad_norm": 0.3235510289669037,
      "learning_rate": 1.5482889534143795e-07,
      "loss": 0.4791,
      "step": 11160
    },
    {
      "epoch": 0.9182834593883591,
      "grad_norm": 0.4926914870738983,
      "learning_rate": 1.273321928704694e-07,
      "loss": 0.468,
      "step": 11170
    },
    {
      "epoch": 0.91910555738244,
      "grad_norm": 0.3373006582260132,
      "learning_rate": 1.0930334884977866e-07,
      "loss": 0.4804,
      "step": 11180
    },
    {
      "epoch": 0.9199276553765209,
      "grad_norm": 1.1631683111190796,
      "learning_rate": 1.0075956576427487e-07,
      "loss": 0.4888,
      "step": 11190
    },
    {
      "epoch": 0.9201742847747452,
      "step": 11193,
      "total_flos": 4.9965855215861105e+19,
      "train_loss": 0.0448528764331606,
      "train_runtime": 11360.4359,
      "train_samples_per_second": 11.823,
      "train_steps_per_second": 0.985
    }
  ],
  "logging_steps": 10,
  "max_steps": 11193,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 636,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.9965855215861105e+19,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}