{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2424,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00041254125412541255,
      "grad_norm": 25.364087477258124,
      "learning_rate": 4.1152263374485605e-08,
      "loss": 1.4505,
      "step": 1
    },
    {
      "epoch": 0.0020627062706270625,
      "grad_norm": 23.584120739330842,
      "learning_rate": 2.05761316872428e-07,
      "loss": 1.4234,
      "step": 5
    },
    {
      "epoch": 0.004125412541254125,
      "grad_norm": 14.928598653810912,
      "learning_rate": 4.11522633744856e-07,
      "loss": 1.3906,
      "step": 10
    },
    {
      "epoch": 0.006188118811881188,
      "grad_norm": 8.160909185975878,
      "learning_rate": 6.17283950617284e-07,
      "loss": 1.2594,
      "step": 15
    },
    {
      "epoch": 0.00825082508250825,
      "grad_norm": 10.23630031823055,
      "learning_rate": 8.23045267489712e-07,
      "loss": 1.1692,
      "step": 20
    },
    {
      "epoch": 0.010313531353135313,
      "grad_norm": 4.891097073750768,
      "learning_rate": 1.02880658436214e-06,
      "loss": 1.0515,
      "step": 25
    },
    {
      "epoch": 0.012376237623762377,
      "grad_norm": 3.5104107671564515,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.9833,
      "step": 30
    },
    {
      "epoch": 0.014438943894389438,
      "grad_norm": 3.531054710795857,
      "learning_rate": 1.440329218106996e-06,
      "loss": 0.9681,
      "step": 35
    },
    {
      "epoch": 0.0165016501650165,
      "grad_norm": 3.0121852103996622,
      "learning_rate": 1.646090534979424e-06,
      "loss": 0.9485,
      "step": 40
    },
    {
      "epoch": 0.018564356435643563,
      "grad_norm": 3.069656613093857,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.9428,
      "step": 45
    },
    {
      "epoch": 0.020627062706270627,
      "grad_norm": 3.017501156282318,
      "learning_rate": 2.05761316872428e-06,
      "loss": 0.9155,
      "step": 50
    },
    {
      "epoch": 0.02268976897689769,
      "grad_norm": 2.9523272195417474,
      "learning_rate": 2.263374485596708e-06,
      "loss": 0.9189,
      "step": 55
    },
    {
      "epoch": 0.024752475247524754,
      "grad_norm": 3.115591318120338,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.8972,
      "step": 60
    },
    {
      "epoch": 0.026815181518151814,
      "grad_norm": 3.315252968948637,
      "learning_rate": 2.674897119341564e-06,
      "loss": 0.9052,
      "step": 65
    },
    {
      "epoch": 0.028877887788778877,
      "grad_norm": 3.0856931589962424,
      "learning_rate": 2.880658436213992e-06,
      "loss": 0.8861,
      "step": 70
    },
    {
      "epoch": 0.03094059405940594,
      "grad_norm": 3.0575333985043316,
      "learning_rate": 3.08641975308642e-06,
      "loss": 0.881,
      "step": 75
    },
    {
      "epoch": 0.033003300330033,
      "grad_norm": 3.4245071366441606,
      "learning_rate": 3.292181069958848e-06,
      "loss": 0.8804,
      "step": 80
    },
    {
      "epoch": 0.03506600660066007,
      "grad_norm": 3.1231813544872074,
      "learning_rate": 3.4979423868312762e-06,
      "loss": 0.8866,
      "step": 85
    },
    {
      "epoch": 0.03712871287128713,
      "grad_norm": 3.095931330170451,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.8754,
      "step": 90
    },
    {
      "epoch": 0.039191419141914194,
      "grad_norm": 3.129665723198324,
      "learning_rate": 3.909465020576132e-06,
      "loss": 0.8535,
      "step": 95
    },
    {
      "epoch": 0.041254125412541254,
      "grad_norm": 3.201387303972087,
      "learning_rate": 4.11522633744856e-06,
      "loss": 0.8763,
      "step": 100
    },
    {
      "epoch": 0.043316831683168314,
      "grad_norm": 3.198510757288822,
      "learning_rate": 4.3209876543209875e-06,
      "loss": 0.8806,
      "step": 105
    },
    {
      "epoch": 0.04537953795379538,
      "grad_norm": 3.0677019856853622,
      "learning_rate": 4.526748971193416e-06,
      "loss": 0.853,
      "step": 110
    },
    {
      "epoch": 0.04744224422442244,
      "grad_norm": 3.063859140817453,
      "learning_rate": 4.732510288065844e-06,
      "loss": 0.8482,
      "step": 115
    },
    {
      "epoch": 0.04950495049504951,
      "grad_norm": 3.0863428454414743,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.8677,
      "step": 120
    },
    {
      "epoch": 0.05156765676567657,
      "grad_norm": 3.3814586153531554,
      "learning_rate": 5.1440329218107e-06,
      "loss": 0.8636,
      "step": 125
    },
    {
      "epoch": 0.05363036303630363,
      "grad_norm": 3.142802292993217,
      "learning_rate": 5.349794238683128e-06,
      "loss": 0.8614,
      "step": 130
    },
    {
      "epoch": 0.055693069306930694,
      "grad_norm": 3.1420236113465694,
      "learning_rate": 5.555555555555557e-06,
      "loss": 0.8516,
      "step": 135
    },
    {
      "epoch": 0.057755775577557754,
      "grad_norm": 2.994047093745851,
      "learning_rate": 5.761316872427984e-06,
      "loss": 0.8516,
      "step": 140
    },
    {
      "epoch": 0.05981848184818482,
      "grad_norm": 3.1021850083230937,
      "learning_rate": 5.967078189300412e-06,
      "loss": 0.8449,
      "step": 145
    },
    {
      "epoch": 0.06188118811881188,
      "grad_norm": 3.2584063367905562,
      "learning_rate": 6.17283950617284e-06,
      "loss": 0.8274,
      "step": 150
    },
    {
      "epoch": 0.06394389438943894,
      "grad_norm": 3.170049017525715,
      "learning_rate": 6.3786008230452675e-06,
      "loss": 0.8385,
      "step": 155
    },
    {
      "epoch": 0.066006600660066,
      "grad_norm": 3.4425094609983353,
      "learning_rate": 6.584362139917696e-06,
      "loss": 0.8563,
      "step": 160
    },
    {
      "epoch": 0.06806930693069307,
      "grad_norm": 2.927681759000597,
      "learning_rate": 6.790123456790124e-06,
      "loss": 0.8301,
      "step": 165
    },
    {
      "epoch": 0.07013201320132013,
      "grad_norm": 3.1321713688594235,
      "learning_rate": 6.9958847736625525e-06,
      "loss": 0.8301,
      "step": 170
    },
    {
      "epoch": 0.0721947194719472,
      "grad_norm": 3.320556237455856,
      "learning_rate": 7.201646090534981e-06,
      "loss": 0.8419,
      "step": 175
    },
    {
      "epoch": 0.07425742574257425,
      "grad_norm": 3.441590967273474,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.8311,
      "step": 180
    },
    {
      "epoch": 0.07632013201320131,
      "grad_norm": 3.081130693491232,
      "learning_rate": 7.613168724279836e-06,
      "loss": 0.8371,
      "step": 185
    },
    {
      "epoch": 0.07838283828382839,
      "grad_norm": 2.929265287656251,
      "learning_rate": 7.818930041152263e-06,
      "loss": 0.8286,
      "step": 190
    },
    {
      "epoch": 0.08044554455445545,
      "grad_norm": 3.369317740510869,
      "learning_rate": 8.024691358024692e-06,
      "loss": 0.8141,
      "step": 195
    },
    {
      "epoch": 0.08250825082508251,
      "grad_norm": 3.143333891707379,
      "learning_rate": 8.23045267489712e-06,
      "loss": 0.8365,
      "step": 200
    },
    {
      "epoch": 0.08457095709570957,
      "grad_norm": 3.085954619911263,
      "learning_rate": 8.43621399176955e-06,
      "loss": 0.8186,
      "step": 205
    },
    {
      "epoch": 0.08663366336633663,
      "grad_norm": 2.9994310068455183,
      "learning_rate": 8.641975308641975e-06,
      "loss": 0.8318,
      "step": 210
    },
    {
      "epoch": 0.0886963696369637,
      "grad_norm": 2.992649765505128,
      "learning_rate": 8.847736625514404e-06,
      "loss": 0.8174,
      "step": 215
    },
    {
      "epoch": 0.09075907590759076,
      "grad_norm": 2.931292838310062,
      "learning_rate": 9.053497942386832e-06,
      "loss": 0.8192,
      "step": 220
    },
    {
      "epoch": 0.09282178217821782,
      "grad_norm": 3.1717689547907684,
      "learning_rate": 9.25925925925926e-06,
      "loss": 0.8162,
      "step": 225
    },
    {
      "epoch": 0.09488448844884488,
      "grad_norm": 3.015985778332564,
      "learning_rate": 9.465020576131688e-06,
      "loss": 0.8253,
      "step": 230
    },
    {
      "epoch": 0.09694719471947194,
      "grad_norm": 2.8717973335844986,
      "learning_rate": 9.670781893004116e-06,
      "loss": 0.8314,
      "step": 235
    },
    {
      "epoch": 0.09900990099009901,
      "grad_norm": 3.1365360963455227,
      "learning_rate": 9.876543209876543e-06,
      "loss": 0.8291,
      "step": 240
    },
    {
      "epoch": 0.10107260726072607,
      "grad_norm": 3.121729098249164,
      "learning_rate": 9.9999792514327e-06,
      "loss": 0.8185,
      "step": 245
    },
    {
      "epoch": 0.10313531353135313,
      "grad_norm": 2.899940654946387,
      "learning_rate": 9.999745832028163e-06,
      "loss": 0.815,
      "step": 250
    },
    {
      "epoch": 0.1051980198019802,
      "grad_norm": 2.979183990670769,
      "learning_rate": 9.999253069658074e-06,
      "loss": 0.7941,
      "step": 255
    },
    {
      "epoch": 0.10726072607260725,
      "grad_norm": 2.915884290090879,
      "learning_rate": 9.998500989882627e-06,
      "loss": 0.7928,
      "step": 260
    },
    {
      "epoch": 0.10932343234323433,
      "grad_norm": 3.0656511848073045,
      "learning_rate": 9.997489631713117e-06,
      "loss": 0.8052,
      "step": 265
    },
    {
      "epoch": 0.11138613861386139,
      "grad_norm": 2.779236934182846,
      "learning_rate": 9.996219047609943e-06,
      "loss": 0.8042,
      "step": 270
    },
    {
      "epoch": 0.11344884488448845,
      "grad_norm": 3.1406451139744553,
      "learning_rate": 9.99468930347986e-06,
      "loss": 0.7934,
      "step": 275
    },
    {
      "epoch": 0.11551155115511551,
      "grad_norm": 2.908579072433511,
      "learning_rate": 9.99290047867258e-06,
      "loss": 0.8036,
      "step": 280
    },
    {
      "epoch": 0.11757425742574257,
      "grad_norm": 2.885133657343927,
      "learning_rate": 9.990852665976648e-06,
      "loss": 0.7966,
      "step": 285
    },
    {
      "epoch": 0.11963696369636964,
      "grad_norm": 2.8220528871968464,
      "learning_rate": 9.98854597161462e-06,
      "loss": 0.7872,
      "step": 290
    },
    {
      "epoch": 0.1216996699669967,
      "grad_norm": 3.0815967855392343,
      "learning_rate": 9.98598051523758e-06,
      "loss": 0.8156,
      "step": 295
    },
    {
      "epoch": 0.12376237623762376,
      "grad_norm": 2.9269865697026427,
      "learning_rate": 9.983156429918895e-06,
      "loss": 0.797,
      "step": 300
    },
    {
      "epoch": 0.12582508250825084,
      "grad_norm": 3.094829019950906,
      "learning_rate": 9.98007386214735e-06,
      "loss": 0.7914,
      "step": 305
    },
    {
      "epoch": 0.12788778877887788,
      "grad_norm": 2.9433436175076633,
      "learning_rate": 9.976732971819526e-06,
      "loss": 0.7811,
      "step": 310
    },
    {
      "epoch": 0.12995049504950495,
      "grad_norm": 2.763941972916272,
      "learning_rate": 9.973133932231514e-06,
      "loss": 0.7783,
      "step": 315
    },
    {
      "epoch": 0.132013201320132,
      "grad_norm": 2.876279072787343,
      "learning_rate": 9.96927693006992e-06,
      "loss": 0.7948,
      "step": 320
    },
    {
      "epoch": 0.13407590759075907,
      "grad_norm": 2.951672309789002,
      "learning_rate": 9.965162165402194e-06,
      "loss": 0.7815,
      "step": 325
    },
    {
      "epoch": 0.13613861386138615,
      "grad_norm": 2.9838776306124726,
      "learning_rate": 9.960789851666237e-06,
      "loss": 0.7745,
      "step": 330
    },
    {
      "epoch": 0.1382013201320132,
      "grad_norm": 2.822284810816944,
      "learning_rate": 9.956160215659342e-06,
      "loss": 0.777,
      "step": 335
    },
    {
      "epoch": 0.14026402640264027,
      "grad_norm": 2.9059843659239637,
      "learning_rate": 9.951273497526423e-06,
      "loss": 0.7697,
      "step": 340
    },
    {
      "epoch": 0.14232673267326731,
      "grad_norm": 2.870130006561793,
      "learning_rate": 9.94612995074756e-06,
      "loss": 0.7516,
      "step": 345
    },
    {
      "epoch": 0.1443894389438944,
      "grad_norm": 2.772459191587491,
      "learning_rate": 9.94072984212485e-06,
      "loss": 0.7643,
      "step": 350
    },
    {
      "epoch": 0.14645214521452146,
      "grad_norm": 2.826949039936636,
      "learning_rate": 9.935073451768567e-06,
      "loss": 0.7648,
      "step": 355
    },
    {
      "epoch": 0.1485148514851485,
      "grad_norm": 3.0414673762126996,
      "learning_rate": 9.929161073082636e-06,
      "loss": 0.7781,
      "step": 360
    },
    {
      "epoch": 0.15057755775577558,
      "grad_norm": 2.7722425713109855,
      "learning_rate": 9.922993012749413e-06,
      "loss": 0.7614,
      "step": 365
    },
    {
      "epoch": 0.15264026402640263,
      "grad_norm": 2.748341237442781,
      "learning_rate": 9.916569590713775e-06,
      "loss": 0.7762,
      "step": 370
    },
    {
      "epoch": 0.1547029702970297,
      "grad_norm": 3.0368108721146045,
      "learning_rate": 9.90989114016652e-06,
      "loss": 0.7641,
      "step": 375
    },
    {
      "epoch": 0.15676567656765678,
      "grad_norm": 2.764173640131257,
      "learning_rate": 9.902958007527092e-06,
      "loss": 0.7237,
      "step": 380
    },
    {
      "epoch": 0.15882838283828382,
      "grad_norm": 2.7786342966630158,
      "learning_rate": 9.89577055242561e-06,
      "loss": 0.7513,
      "step": 385
    },
    {
      "epoch": 0.1608910891089109,
      "grad_norm": 2.921540805258061,
      "learning_rate": 9.88832914768421e-06,
      "loss": 0.7575,
      "step": 390
    },
    {
      "epoch": 0.16295379537953794,
      "grad_norm": 2.7315756627341443,
      "learning_rate": 9.880634179297706e-06,
      "loss": 0.754,
      "step": 395
    },
    {
      "epoch": 0.16501650165016502,
      "grad_norm": 2.7136927749412734,
      "learning_rate": 9.872686046413575e-06,
      "loss": 0.7215,
      "step": 400
    },
    {
      "epoch": 0.1670792079207921,
      "grad_norm": 2.783986178315758,
      "learning_rate": 9.864485161311242e-06,
      "loss": 0.7651,
      "step": 405
    },
    {
      "epoch": 0.16914191419141913,
      "grad_norm": 2.8016707367343887,
      "learning_rate": 9.856031949380707e-06,
      "loss": 0.7256,
      "step": 410
    },
    {
      "epoch": 0.1712046204620462,
      "grad_norm": 2.9437239528614216,
      "learning_rate": 9.847326849100467e-06,
      "loss": 0.7228,
      "step": 415
    },
    {
      "epoch": 0.17326732673267325,
      "grad_norm": 2.739229372104028,
      "learning_rate": 9.838370312014783e-06,
      "loss": 0.7411,
      "step": 420
    },
    {
      "epoch": 0.17533003300330033,
      "grad_norm": 2.747689406028468,
      "learning_rate": 9.829162802710246e-06,
      "loss": 0.7373,
      "step": 425
    },
    {
      "epoch": 0.1773927392739274,
      "grad_norm": 2.879870332673276,
      "learning_rate": 9.819704798791691e-06,
      "loss": 0.7376,
      "step": 430
    },
    {
      "epoch": 0.17945544554455445,
      "grad_norm": 3.057579380919091,
      "learning_rate": 9.80999679085741e-06,
      "loss": 0.7233,
      "step": 435
    },
    {
      "epoch": 0.18151815181518152,
      "grad_norm": 3.100440102401277,
      "learning_rate": 9.800039282473719e-06,
      "loss": 0.7001,
      "step": 440
    },
    {
      "epoch": 0.18358085808580857,
      "grad_norm": 2.7070231210764844,
      "learning_rate": 9.789832790148822e-06,
      "loss": 0.7085,
      "step": 445
    },
    {
      "epoch": 0.18564356435643564,
      "grad_norm": 2.733916165732209,
      "learning_rate": 9.77937784330603e-06,
      "loss": 0.7012,
      "step": 450
    },
    {
      "epoch": 0.18770627062706272,
      "grad_norm": 2.563129838542594,
      "learning_rate": 9.768674984256292e-06,
      "loss": 0.7123,
      "step": 455
    },
    {
      "epoch": 0.18976897689768976,
      "grad_norm": 2.74260970646788,
      "learning_rate": 9.757724768170074e-06,
      "loss": 0.7061,
      "step": 460
    },
    {
      "epoch": 0.19183168316831684,
      "grad_norm": 2.9775386795661607,
      "learning_rate": 9.74652776304855e-06,
      "loss": 0.7035,
      "step": 465
    },
    {
      "epoch": 0.19389438943894388,
      "grad_norm": 11.7972903732034,
      "learning_rate": 9.73508454969415e-06,
      "loss": 0.7031,
      "step": 470
    },
    {
      "epoch": 0.19595709570957096,
      "grad_norm": 3.0823242258683203,
      "learning_rate": 9.723395721680418e-06,
      "loss": 0.7208,
      "step": 475
    },
    {
      "epoch": 0.19801980198019803,
      "grad_norm": 2.984181238427559,
      "learning_rate": 9.711461885321247e-06,
      "loss": 0.7082,
      "step": 480
    },
    {
      "epoch": 0.20008250825082508,
      "grad_norm": 2.7986743538102505,
      "learning_rate": 9.699283659639402e-06,
      "loss": 0.6877,
      "step": 485
    },
    {
      "epoch": 0.20214521452145215,
      "grad_norm": 2.8619815735669403,
      "learning_rate": 9.68686167633443e-06,
      "loss": 0.764,
      "step": 490
    },
    {
      "epoch": 0.2042079207920792,
      "grad_norm": 4.637672201020113,
      "learning_rate": 9.67419657974988e-06,
      "loss": 0.6984,
      "step": 495
    },
    {
      "epoch": 0.20627062706270627,
      "grad_norm": 2.7511677692038217,
      "learning_rate": 9.661289026839889e-06,
      "loss": 0.698,
      "step": 500
    },
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 3.1629886058739594,
      "learning_rate": 9.648139687135106e-06,
      "loss": 0.7025,
      "step": 505
    },
    {
      "epoch": 0.2103960396039604,
      "grad_norm": 2.835404473287634,
      "learning_rate": 9.634749242707948e-06,
      "loss": 0.6829,
      "step": 510
    },
    {
      "epoch": 0.21245874587458746,
      "grad_norm": 2.6239704035151687,
      "learning_rate": 9.62111838813724e-06,
      "loss": 0.6772,
      "step": 515
    },
    {
      "epoch": 0.2145214521452145,
      "grad_norm": 2.740976973488732,
      "learning_rate": 9.607247830472174e-06,
      "loss": 0.6731,
      "step": 520
    },
    {
      "epoch": 0.21658415841584158,
      "grad_norm": 2.735519864410512,
      "learning_rate": 9.593138289195634e-06,
      "loss": 0.6806,
      "step": 525
    },
    {
      "epoch": 0.21864686468646866,
      "grad_norm": 2.684453002911694,
      "learning_rate": 9.578790496186879e-06,
      "loss": 0.6962,
      "step": 530
    },
    {
      "epoch": 0.2207095709570957,
      "grad_norm": 2.7970751535674037,
      "learning_rate": 9.56420519568358e-06,
      "loss": 0.6854,
      "step": 535
    },
    {
      "epoch": 0.22277227722772278,
      "grad_norm": 2.893389666992377,
      "learning_rate": 9.549383144243213e-06,
      "loss": 0.6713,
      "step": 540
    },
    {
      "epoch": 0.22483498349834982,
      "grad_norm": 2.7169375882464926,
      "learning_rate": 9.534325110703814e-06,
      "loss": 0.7436,
      "step": 545
    },
    {
      "epoch": 0.2268976897689769,
      "grad_norm": 2.7286325422508537,
      "learning_rate": 9.519031876144106e-06,
      "loss": 0.6758,
      "step": 550
    },
    {
      "epoch": 0.22896039603960397,
      "grad_norm": 2.8735702079028775,
      "learning_rate": 9.503504233842973e-06,
      "loss": 0.6657,
      "step": 555
    },
    {
      "epoch": 0.23102310231023102,
      "grad_norm": 3.0387173259825735,
      "learning_rate": 9.487742989238318e-06,
      "loss": 0.6782,
      "step": 560
    },
    {
      "epoch": 0.2330858085808581,
      "grad_norm": 2.538376376963473,
      "learning_rate": 9.471748959885284e-06,
      "loss": 0.6776,
      "step": 565
    },
    {
      "epoch": 0.23514851485148514,
      "grad_norm": 2.6593328363095816,
      "learning_rate": 9.455522975413846e-06,
      "loss": 0.674,
      "step": 570
    },
    {
      "epoch": 0.2372112211221122,
      "grad_norm": 2.7283552755825715,
      "learning_rate": 9.439065877485774e-06,
      "loss": 0.6533,
      "step": 575
    },
    {
      "epoch": 0.23927392739273928,
      "grad_norm": 2.7031827262094708,
      "learning_rate": 9.422378519750978e-06,
      "loss": 0.6738,
      "step": 580
    },
    {
      "epoch": 0.24133663366336633,
      "grad_norm": 2.9208550727888554,
      "learning_rate": 9.40546176780323e-06,
      "loss": 0.6792,
      "step": 585
    },
    {
      "epoch": 0.2433993399339934,
      "grad_norm": 2.8493025404273555,
      "learning_rate": 9.388316499135259e-06,
      "loss": 0.6728,
      "step": 590
    },
    {
      "epoch": 0.24546204620462045,
      "grad_norm": 2.6354127779313985,
      "learning_rate": 9.370943603093235e-06,
      "loss": 0.6745,
      "step": 595
    },
    {
      "epoch": 0.24752475247524752,
      "grad_norm": 3.598421971885995,
      "learning_rate": 9.353343980830644e-06,
      "loss": 0.6581,
      "step": 600
    },
    {
      "epoch": 0.2495874587458746,
      "grad_norm": 2.5853090904417497,
      "learning_rate": 9.33551854526154e-06,
      "loss": 0.6673,
      "step": 605
    },
    {
      "epoch": 0.25165016501650167,
      "grad_norm": 2.793105077025725,
      "learning_rate": 9.31746822101319e-06,
      "loss": 0.6637,
      "step": 610
    },
    {
      "epoch": 0.2537128712871287,
      "grad_norm": 2.7270681052691765,
      "learning_rate": 9.299193944378112e-06,
      "loss": 0.6356,
      "step": 615
    },
    {
      "epoch": 0.25577557755775576,
      "grad_norm": 2.7008720069429115,
      "learning_rate": 9.280696663265512e-06,
      "loss": 0.6464,
      "step": 620
    },
    {
      "epoch": 0.25783828382838286,
      "grad_norm": 2.671865487166939,
      "learning_rate": 9.261977337152107e-06,
      "loss": 0.6527,
      "step": 625
    },
    {
      "epoch": 0.2599009900990099,
      "grad_norm": 2.653313117664496,
      "learning_rate": 9.243036937032373e-06,
      "loss": 0.638,
      "step": 630
    },
    {
      "epoch": 0.26196369636963696,
      "grad_norm": 2.5458051542477502,
      "learning_rate": 9.223876445368153e-06,
      "loss": 0.6545,
      "step": 635
    },
    {
      "epoch": 0.264026402640264,
      "grad_norm": 2.6965322964778027,
      "learning_rate": 9.204496856037718e-06,
      "loss": 0.6616,
      "step": 640
    },
    {
      "epoch": 0.2660891089108911,
      "grad_norm": 2.830059150957898,
      "learning_rate": 9.184899174284201e-06,
      "loss": 0.6492,
      "step": 645
    },
    {
      "epoch": 0.26815181518151815,
      "grad_norm": 2.796697680039645,
      "learning_rate": 9.16508441666346e-06,
      "loss": 0.6324,
      "step": 650
    },
    {
      "epoch": 0.2702145214521452,
      "grad_norm": 3.091827022287661,
      "learning_rate": 9.14505361099134e-06,
      "loss": 0.6603,
      "step": 655
    },
    {
      "epoch": 0.2722772277227723,
      "grad_norm": 2.77180371501166,
      "learning_rate": 9.124807796290366e-06,
      "loss": 0.6265,
      "step": 660
    },
    {
      "epoch": 0.27433993399339934,
      "grad_norm": 2.6247932532421894,
      "learning_rate": 9.104348022735853e-06,
      "loss": 0.623,
      "step": 665
    },
    {
      "epoch": 0.2764026402640264,
      "grad_norm": 2.909646454868501,
      "learning_rate": 9.083675351601417e-06,
      "loss": 0.6554,
      "step": 670
    },
    {
      "epoch": 0.2784653465346535,
      "grad_norm": 2.8355016443543795,
      "learning_rate": 9.062790855203932e-06,
      "loss": 0.6248,
      "step": 675
    },
    {
      "epoch": 0.28052805280528054,
      "grad_norm": 2.845588283250197,
      "learning_rate": 9.041695616847915e-06,
      "loss": 0.636,
      "step": 680
    },
    {
      "epoch": 0.2825907590759076,
      "grad_norm": 2.5668723810217657,
      "learning_rate": 9.020390730769324e-06,
      "loss": 0.647,
      "step": 685
    },
    {
      "epoch": 0.28465346534653463,
      "grad_norm": 2.7870591118074515,
      "learning_rate": 8.998877302078803e-06,
      "loss": 0.6163,
      "step": 690
    },
    {
      "epoch": 0.28671617161716173,
      "grad_norm": 2.8993841087894867,
      "learning_rate": 8.97715644670436e-06,
      "loss": 0.6308,
      "step": 695
    },
    {
      "epoch": 0.2887788778877888,
      "grad_norm": 2.669040121356427,
      "learning_rate": 8.955229291333473e-06,
      "loss": 0.6293,
      "step": 700
    },
    {
      "epoch": 0.2908415841584158,
      "grad_norm": 2.6715972640429095,
      "learning_rate": 8.933096973354665e-06,
      "loss": 0.6414,
      "step": 705
    },
    {
      "epoch": 0.2929042904290429,
      "grad_norm": 2.721741671637072,
      "learning_rate": 8.910760640798487e-06,
      "loss": 0.6048,
      "step": 710
    },
    {
      "epoch": 0.29496699669966997,
      "grad_norm": 2.694393411500226,
      "learning_rate": 8.88822145227798e-06,
      "loss": 0.6253,
      "step": 715
    },
    {
      "epoch": 0.297029702970297,
      "grad_norm": 2.7216607887245834,
      "learning_rate": 8.865480576928578e-06,
      "loss": 0.5975,
      "step": 720
    },
    {
      "epoch": 0.2990924092409241,
      "grad_norm": 2.56544703422047,
      "learning_rate": 8.842539194347448e-06,
      "loss": 0.6166,
      "step": 725
    },
    {
      "epoch": 0.30115511551155116,
      "grad_norm": 2.687926593730155,
      "learning_rate": 8.819398494532328e-06,
      "loss": 0.6162,
      "step": 730
    },
    {
      "epoch": 0.3032178217821782,
      "grad_norm": 2.715855361797108,
      "learning_rate": 8.796059677819773e-06,
      "loss": 0.6159,
      "step": 735
    },
    {
      "epoch": 0.30528052805280526,
      "grad_norm": 2.7177077570753574,
      "learning_rate": 8.77252395482291e-06,
      "loss": 0.6199,
      "step": 740
    },
    {
      "epoch": 0.30734323432343236,
      "grad_norm": 2.7537980998977596,
      "learning_rate": 8.748792546368641e-06,
      "loss": 0.5942,
      "step": 745
    },
    {
      "epoch": 0.3094059405940594,
      "grad_norm": 2.7754021446494894,
      "learning_rate": 8.72486668343431e-06,
      "loss": 0.6204,
      "step": 750
    },
    {
      "epoch": 0.31146864686468645,
      "grad_norm": 2.6145595604114744,
      "learning_rate": 8.700747607083851e-06,
      "loss": 0.606,
      "step": 755
    },
    {
      "epoch": 0.31353135313531355,
      "grad_norm": 2.5551572752880785,
      "learning_rate": 8.676436568403422e-06,
      "loss": 0.615,
      "step": 760
    },
    {
      "epoch": 0.3155940594059406,
      "grad_norm": 2.584531027457074,
      "learning_rate": 8.651934828436497e-06,
      "loss": 0.6094,
      "step": 765
    },
    {
      "epoch": 0.31765676567656764,
      "grad_norm": 2.7585647106017865,
      "learning_rate": 8.627243658118466e-06,
      "loss": 0.6082,
      "step": 770
    },
    {
      "epoch": 0.31971947194719474,
      "grad_norm": 2.654651285995417,
      "learning_rate": 8.602364338210699e-06,
      "loss": 0.5939,
      "step": 775
    },
    {
      "epoch": 0.3217821782178218,
      "grad_norm": 2.6787685912910204,
      "learning_rate": 8.57729815923412e-06,
      "loss": 0.5876,
      "step": 780
    },
    {
      "epoch": 0.32384488448844884,
      "grad_norm": 2.6267227792982357,
      "learning_rate": 8.55204642140226e-06,
      "loss": 0.599,
      "step": 785
    },
    {
      "epoch": 0.3259075907590759,
      "grad_norm": 2.6802025518663015,
      "learning_rate": 8.52661043455382e-06,
      "loss": 0.594,
      "step": 790
    },
    {
      "epoch": 0.327970297029703,
      "grad_norm": 2.57110758651426,
      "learning_rate": 8.50099151808472e-06,
      "loss": 0.5884,
      "step": 795
    },
    {
      "epoch": 0.33003300330033003,
      "grad_norm": 2.99724938552448,
      "learning_rate": 8.47519100087967e-06,
      "loss": 0.5821,
      "step": 800
    },
    {
      "epoch": 0.3320957095709571,
      "grad_norm": 2.7370257356141616,
      "learning_rate": 8.449210221243225e-06,
      "loss": 0.5761,
      "step": 805
    },
    {
      "epoch": 0.3341584158415842,
      "grad_norm": 3.0576827360675214,
      "learning_rate": 8.42305052683038e-06,
      "loss": 0.5824,
      "step": 810
    },
    {
      "epoch": 0.3362211221122112,
      "grad_norm": 2.562013124794496,
      "learning_rate": 8.39671327457666e-06,
      "loss": 0.5837,
      "step": 815
    },
    {
      "epoch": 0.33828382838283827,
      "grad_norm": 2.7787602704146916,
      "learning_rate": 8.370199830627732e-06,
      "loss": 0.5924,
      "step": 820
    },
    {
      "epoch": 0.34034653465346537,
      "grad_norm": 2.6149487374961136,
      "learning_rate": 8.343511570268541e-06,
      "loss": 0.594,
      "step": 825
    },
    {
      "epoch": 0.3424092409240924,
      "grad_norm": 2.7424901857047925,
      "learning_rate": 8.316649877851977e-06,
      "loss": 0.5696,
      "step": 830
    },
    {
      "epoch": 0.34447194719471946,
      "grad_norm": 2.654634886694336,
      "learning_rate": 8.289616146727062e-06,
      "loss": 0.5821,
      "step": 835
    },
    {
      "epoch": 0.3465346534653465,
      "grad_norm": 2.5386950387254514,
      "learning_rate": 8.262411779166681e-06,
      "loss": 0.5722,
      "step": 840
    },
    {
      "epoch": 0.3485973597359736,
      "grad_norm": 2.5882867695406353,
      "learning_rate": 8.235038186294836e-06,
      "loss": 0.5896,
      "step": 845
    },
    {
      "epoch": 0.35066006600660066,
      "grad_norm": 2.6558091815806226,
      "learning_rate": 8.207496788013456e-06,
      "loss": 0.5843,
      "step": 850
    },
    {
      "epoch": 0.3527227722772277,
      "grad_norm": 2.766934233648842,
      "learning_rate": 8.179789012928747e-06,
      "loss": 0.5785,
      "step": 855
    },
    {
      "epoch": 0.3547854785478548,
      "grad_norm": 2.6862925251030427,
      "learning_rate": 8.151916298277078e-06,
      "loss": 0.5641,
      "step": 860
    },
    {
      "epoch": 0.35684818481848185,
      "grad_norm": 2.6591167593333256,
      "learning_rate": 8.123880089850438e-06,
      "loss": 0.5696,
      "step": 865
    },
    {
      "epoch": 0.3589108910891089,
      "grad_norm": 2.6326244768332745,
      "learning_rate": 8.095681841921441e-06,
      "loss": 0.5791,
      "step": 870
    },
    {
      "epoch": 0.360973597359736,
      "grad_norm": 2.65447071424182,
      "learning_rate": 8.06732301716789e-06,
      "loss": 0.5663,
      "step": 875
    },
    {
      "epoch": 0.36303630363036304,
      "grad_norm": 2.7561863484341207,
      "learning_rate": 8.038805086596903e-06,
      "loss": 0.572,
      "step": 880
    },
    {
      "epoch": 0.3650990099009901,
      "grad_norm": 2.6239959610961665,
      "learning_rate": 8.010129529468614e-06,
      "loss": 0.5703,
      "step": 885
    },
    {
      "epoch": 0.36716171617161714,
      "grad_norm": 2.8054930191127627,
      "learning_rate": 7.981297833219435e-06,
      "loss": 0.5697,
      "step": 890
    },
    {
      "epoch": 0.36922442244224424,
      "grad_norm": 2.556672661006346,
      "learning_rate": 7.952311493384916e-06,
      "loss": 0.5454,
      "step": 895
    },
    {
      "epoch": 0.3712871287128713,
      "grad_norm": 2.711104263166826,
      "learning_rate": 7.923172013522153e-06,
      "loss": 0.5477,
      "step": 900
    },
    {
      "epoch": 0.37334983498349833,
      "grad_norm": 2.6047389045795137,
      "learning_rate": 7.893880905131807e-06,
      "loss": 0.5548,
      "step": 905
    },
    {
      "epoch": 0.37541254125412543,
      "grad_norm": 2.6009558864259077,
      "learning_rate": 7.864439687579695e-06,
      "loss": 0.5487,
      "step": 910
    },
    {
      "epoch": 0.3774752475247525,
      "grad_norm": 2.776081930223416,
      "learning_rate": 7.834849888017979e-06,
      "loss": 0.5535,
      "step": 915
    },
    {
      "epoch": 0.3795379537953795,
      "grad_norm": 2.7439984836733227,
      "learning_rate": 7.805113041305958e-06,
      "loss": 0.5647,
      "step": 920
    },
    {
      "epoch": 0.3816006600660066,
      "grad_norm": 2.621923636099921,
      "learning_rate": 7.775230689930445e-06,
      "loss": 0.5364,
      "step": 925
    },
    {
      "epoch": 0.38366336633663367,
      "grad_norm": 2.5440712763463296,
      "learning_rate": 7.745204383925753e-06,
      "loss": 0.5392,
      "step": 930
    },
    {
      "epoch": 0.3857260726072607,
      "grad_norm": 2.6239950992845134,
      "learning_rate": 7.715035680793311e-06,
      "loss": 0.548,
      "step": 935
    },
    {
      "epoch": 0.38778877887788776,
      "grad_norm": 2.558970533428198,
      "learning_rate": 7.684726145420853e-06,
      "loss": 0.5481,
      "step": 940
    },
    {
      "epoch": 0.38985148514851486,
      "grad_norm": 2.733683917435511,
      "learning_rate": 7.654277350001255e-06,
      "loss": 0.5735,
      "step": 945
    },
    {
      "epoch": 0.3919141914191419,
      "grad_norm": 2.4440362195756,
      "learning_rate": 7.623690873950988e-06,
      "loss": 0.5361,
      "step": 950
    },
    {
      "epoch": 0.39397689768976896,
      "grad_norm": 2.7948133819884347,
      "learning_rate": 7.592968303828181e-06,
      "loss": 0.532,
      "step": 955
    },
    {
      "epoch": 0.39603960396039606,
      "grad_norm": 2.712027777278135,
      "learning_rate": 7.5621112332503325e-06,
      "loss": 0.5323,
      "step": 960
    },
    {
      "epoch": 0.3981023102310231,
      "grad_norm": 2.5614258873150724,
      "learning_rate": 7.531121262811645e-06,
      "loss": 0.532,
      "step": 965
    },
    {
      "epoch": 0.40016501650165015,
      "grad_norm": 2.5258391365444557,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.5271,
      "step": 970
    },
    {
      "epoch": 0.40222772277227725,
      "grad_norm": 2.543456979300113,
      "learning_rate": 7.468749059113578e-06,
      "loss": 0.5168,
      "step": 975
    },
    {
      "epoch": 0.4042904290429043,
      "grad_norm": 2.5683535191631623,
      "learning_rate": 7.437370061177116e-06,
      "loss": 0.5341,
      "step": 980
    },
    {
      "epoch": 0.40635313531353134,
      "grad_norm": 2.598132530083866,
      "learning_rate": 7.40586463385783e-06,
      "loss": 0.5278,
      "step": 985
    },
    {
      "epoch": 0.4084158415841584,
      "grad_norm": 2.6824293845426666,
      "learning_rate": 7.374234411380987e-06,
      "loss": 0.5388,
      "step": 990
    },
    {
      "epoch": 0.4104785478547855,
      "grad_norm": 2.6715849470717576,
      "learning_rate": 7.342481034445127e-06,
      "loss": 0.5476,
      "step": 995
    },
    {
      "epoch": 0.41254125412541254,
      "grad_norm": 2.676941475984213,
      "learning_rate": 7.310606150136965e-06,
      "loss": 0.5291,
      "step": 1000
    },
    {
      "epoch": 0.4146039603960396,
      "grad_norm": 2.7331658451074627,
      "learning_rate": 7.2786114118459564e-06,
      "loss": 0.5277,
      "step": 1005
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 2.6472987943602813,
      "learning_rate": 7.246498479178523e-06,
      "loss": 0.5221,
      "step": 1010
    },
    {
      "epoch": 0.41872937293729373,
      "grad_norm": 2.586821568507126,
      "learning_rate": 7.214269017871981e-06,
      "loss": 0.5163,
      "step": 1015
    },
    {
      "epoch": 0.4207920792079208,
      "grad_norm": 2.4588164960210834,
      "learning_rate": 7.181924699708127e-06,
      "loss": 0.5151,
      "step": 1020
    },
    {
      "epoch": 0.4228547854785479,
      "grad_norm": 2.5130611281830006,
      "learning_rate": 7.149467202426525e-06,
      "loss": 0.5326,
      "step": 1025
    },
    {
      "epoch": 0.4249174917491749,
      "grad_norm": 2.6743950877544265,
      "learning_rate": 7.116898209637478e-06,
      "loss": 0.5324,
      "step": 1030
    },
    {
      "epoch": 0.42698019801980197,
      "grad_norm": 2.5588441225148735,
      "learning_rate": 7.084219410734701e-06,
      "loss": 0.5277,
      "step": 1035
    },
    {
      "epoch": 0.429042904290429,
      "grad_norm": 2.6168950669777935,
      "learning_rate": 7.051432500807682e-06,
      "loss": 0.4983,
      "step": 1040
    },
    {
      "epoch": 0.4311056105610561,
      "grad_norm": 2.5513680083363677,
      "learning_rate": 7.018539180553768e-06,
      "loss": 0.5032,
      "step": 1045
    },
    {
      "epoch": 0.43316831683168316,
      "grad_norm": 2.590369266646011,
      "learning_rate": 6.985541156189932e-06,
      "loss": 0.5151,
      "step": 1050
    },
    {
      "epoch": 0.4352310231023102,
      "grad_norm": 2.5434958160268986,
      "learning_rate": 6.952440139364286e-06,
      "loss": 0.5231,
      "step": 1055
    },
    {
      "epoch": 0.4372937293729373,
      "grad_norm": 2.630962973641557,
      "learning_rate": 6.919237847067282e-06,
      "loss": 0.5248,
      "step": 1060
    },
    {
      "epoch": 0.43935643564356436,
      "grad_norm": 2.4427367015076737,
      "learning_rate": 6.885936001542658e-06,
      "loss": 0.5074,
      "step": 1065
    },
    {
      "epoch": 0.4414191419141914,
      "grad_norm": 2.529834433774531,
      "learning_rate": 6.852536330198099e-06,
      "loss": 0.52,
      "step": 1070
    },
    {
      "epoch": 0.4434818481848185,
      "grad_norm": 2.5714905566940622,
      "learning_rate": 6.819040565515636e-06,
      "loss": 0.5273,
      "step": 1075
    },
    {
      "epoch": 0.44554455445544555,
      "grad_norm": 2.4185173202106585,
      "learning_rate": 6.785450444961783e-06,
      "loss": 0.5058,
      "step": 1080
    },
    {
      "epoch": 0.4476072607260726,
      "grad_norm": 2.6463305618629884,
      "learning_rate": 6.751767710897404e-06,
      "loss": 0.5186,
      "step": 1085
    },
    {
      "epoch": 0.44966996699669964,
      "grad_norm": 2.4417765175517374,
      "learning_rate": 6.71799411048734e-06,
      "loss": 0.5119,
      "step": 1090
    },
    {
      "epoch": 0.45173267326732675,
      "grad_norm": 2.4466915489716246,
      "learning_rate": 6.684131395609784e-06,
      "loss": 0.5091,
      "step": 1095
    },
    {
      "epoch": 0.4537953795379538,
      "grad_norm": 2.5469460375657045,
      "learning_rate": 6.650181322765407e-06,
      "loss": 0.4972,
      "step": 1100
    },
    {
      "epoch": 0.45585808580858084,
      "grad_norm": 2.55425172030808,
      "learning_rate": 6.61614565298624e-06,
      "loss": 0.5163,
      "step": 1105
    },
    {
      "epoch": 0.45792079207920794,
      "grad_norm": 2.6047811461540635,
      "learning_rate": 6.5820261517443365e-06,
      "loss": 0.5068,
      "step": 1110
    },
    {
      "epoch": 0.459983498349835,
      "grad_norm": 2.4314690079287344,
      "learning_rate": 6.54782458886019e-06,
      "loss": 0.4936,
      "step": 1115
    },
    {
      "epoch": 0.46204620462046203,
      "grad_norm": 2.4496252297971512,
      "learning_rate": 6.5135427384109315e-06,
      "loss": 0.5055,
      "step": 1120
    },
    {
      "epoch": 0.46410891089108913,
      "grad_norm": 2.4983465170752996,
      "learning_rate": 6.479182378638308e-06,
      "loss": 0.508,
      "step": 1125
    },
    {
      "epoch": 0.4661716171617162,
      "grad_norm": 2.493954436696378,
      "learning_rate": 6.444745291856442e-06,
      "loss": 0.5074,
      "step": 1130
    },
    {
      "epoch": 0.4682343234323432,
      "grad_norm": 2.4743870926317104,
      "learning_rate": 6.410233264359379e-06,
      "loss": 0.4817,
      "step": 1135
    },
    {
      "epoch": 0.47029702970297027,
      "grad_norm": 2.6015106871562717,
      "learning_rate": 6.375648086328431e-06,
      "loss": 0.503,
      "step": 1140
    },
    {
      "epoch": 0.47235973597359737,
      "grad_norm": 2.4748853313920876,
      "learning_rate": 6.340991551739319e-06,
      "loss": 0.4898,
      "step": 1145
    },
    {
      "epoch": 0.4744224422442244,
      "grad_norm": 2.571563250995017,
      "learning_rate": 6.3062654582691175e-06,
      "loss": 0.5038,
      "step": 1150
    },
    {
      "epoch": 0.47648514851485146,
      "grad_norm": 2.56828630674544,
      "learning_rate": 6.271471607203006e-06,
      "loss": 0.4964,
      "step": 1155
    },
    {
      "epoch": 0.47854785478547857,
      "grad_norm": 2.4472686378385013,
      "learning_rate": 6.236611803340829e-06,
      "loss": 0.4837,
      "step": 1160
    },
    {
      "epoch": 0.4806105610561056,
      "grad_norm": 2.4926046687862864,
      "learning_rate": 6.201687854903492e-06,
      "loss": 0.4783,
      "step": 1165
    },
    {
      "epoch": 0.48267326732673266,
      "grad_norm": 2.4605803697540893,
      "learning_rate": 6.16670157343915e-06,
      "loss": 0.4855,
      "step": 1170
    },
    {
      "epoch": 0.48473597359735976,
      "grad_norm": 2.4830994265099906,
      "learning_rate": 6.131654773729255e-06,
      "loss": 0.492,
      "step": 1175
    },
    {
      "epoch": 0.4867986798679868,
      "grad_norm": 2.4315421147193184,
      "learning_rate": 6.096549273694411e-06,
      "loss": 0.4906,
      "step": 1180
    },
    {
      "epoch": 0.48886138613861385,
      "grad_norm": 2.599517694694856,
      "learning_rate": 6.061386894300082e-06,
      "loss": 0.4818,
      "step": 1185
    },
    {
      "epoch": 0.4909240924092409,
      "grad_norm": 2.5808864192189835,
      "learning_rate": 6.026169459462132e-06,
      "loss": 0.4882,
      "step": 1190
    },
    {
      "epoch": 0.492986798679868,
      "grad_norm": 2.378628769357962,
      "learning_rate": 5.990898795952225e-06,
      "loss": 0.4645,
      "step": 1195
    },
    {
      "epoch": 0.49504950495049505,
      "grad_norm": 2.5396332086908906,
      "learning_rate": 5.955576733303053e-06,
      "loss": 0.4864,
      "step": 1200
    },
    {
      "epoch": 0.4971122112211221,
      "grad_norm": 2.597727268666024,
      "learning_rate": 5.920205103713449e-06,
      "loss": 0.4801,
      "step": 1205
    },
    {
      "epoch": 0.4991749174917492,
      "grad_norm": 2.5073049896152604,
      "learning_rate": 5.884785741953345e-06,
      "loss": 0.493,
      "step": 1210
    },
    {
      "epoch": 0.5012376237623762,
      "grad_norm": 2.5498231092866352,
      "learning_rate": 5.849320485268597e-06,
      "loss": 0.4847,
      "step": 1215
    },
    {
      "epoch": 0.5033003300330033,
      "grad_norm": 2.4434059859729267,
      "learning_rate": 5.8138111732856906e-06,
      "loss": 0.4816,
      "step": 1220
    },
    {
      "epoch": 0.5053630363036303,
      "grad_norm": 2.427941557179141,
      "learning_rate": 5.778259647916309e-06,
      "loss": 0.4752,
      "step": 1225
    },
    {
      "epoch": 0.5074257425742574,
      "grad_norm": 2.5066438897899523,
      "learning_rate": 5.7426677532618e-06,
      "loss": 0.4751,
      "step": 1230
    },
    {
      "epoch": 0.5094884488448845,
      "grad_norm": 2.442917336254435,
      "learning_rate": 5.707037335517514e-06,
      "loss": 0.4516,
      "step": 1235
    },
    {
      "epoch": 0.5115511551155115,
      "grad_norm": 2.4869384945022066,
      "learning_rate": 5.67137024287704e-06,
      "loss": 0.4697,
      "step": 1240
    },
    {
      "epoch": 0.5136138613861386,
      "grad_norm": 2.4916982755437016,
      "learning_rate": 5.635668325436343e-06,
      "loss": 0.4593,
      "step": 1245
    },
    {
      "epoch": 0.5156765676567657,
      "grad_norm": 2.400223999637477,
      "learning_rate": 5.599933435097791e-06,
      "loss": 0.4627,
      "step": 1250
    },
    {
      "epoch": 0.5177392739273927,
      "grad_norm": 2.5691310342129006,
      "learning_rate": 5.564167425474093e-06,
      "loss": 0.4773,
      "step": 1255
    },
    {
      "epoch": 0.5198019801980198,
      "grad_norm": 2.5534505692218135,
      "learning_rate": 5.528372151792161e-06,
      "loss": 0.4757,
      "step": 1260
    },
    {
      "epoch": 0.5218646864686468,
      "grad_norm": 2.3595984093788465,
      "learning_rate": 5.492549470796865e-06,
      "loss": 0.4598,
      "step": 1265
    },
    {
      "epoch": 0.5239273927392739,
      "grad_norm": 2.484377807128577,
      "learning_rate": 5.456701240654726e-06,
      "loss": 0.4591,
      "step": 1270
    },
    {
      "epoch": 0.525990099009901,
      "grad_norm": 2.4760013171249966,
      "learning_rate": 5.420829320857532e-06,
      "loss": 0.4707,
      "step": 1275
    },
    {
      "epoch": 0.528052805280528,
      "grad_norm": 2.462101837921461,
      "learning_rate": 5.384935572125882e-06,
      "loss": 0.454,
      "step": 1280
    },
    {
      "epoch": 0.5301155115511551,
      "grad_norm": 2.588115117622585,
      "learning_rate": 5.349021856312669e-06,
      "loss": 0.4678,
      "step": 1285
    },
    {
      "epoch": 0.5321782178217822,
      "grad_norm": 2.5754165982294626,
      "learning_rate": 5.3130900363065055e-06,
      "loss": 0.4573,
      "step": 1290
    },
    {
      "epoch": 0.5342409240924092,
      "grad_norm": 2.4762653761884064,
      "learning_rate": 5.277141975935083e-06,
      "loss": 0.4733,
      "step": 1295
    },
    {
      "epoch": 0.5363036303630363,
      "grad_norm": 2.527567420790079,
      "learning_rate": 5.24117953986851e-06,
      "loss": 0.4678,
      "step": 1300
    },
    {
      "epoch": 0.5383663366336634,
      "grad_norm": 2.55248617117022,
      "learning_rate": 5.2052045935225725e-06,
      "loss": 0.4676,
      "step": 1305
    },
    {
      "epoch": 0.5404290429042904,
      "grad_norm": 2.569839654328644,
      "learning_rate": 5.169219002961987e-06,
      "loss": 0.4516,
      "step": 1310
    },
    {
      "epoch": 0.5424917491749175,
      "grad_norm": 2.4340359025464458,
      "learning_rate": 5.133224634803594e-06,
      "loss": 0.4487,
      "step": 1315
    },
    {
      "epoch": 0.5445544554455446,
      "grad_norm": 2.5185591909324985,
      "learning_rate": 5.097223356119538e-06,
      "loss": 0.4508,
      "step": 1320
    },
    {
      "epoch": 0.5466171617161716,
      "grad_norm": 2.473389867364878,
      "learning_rate": 5.061217034340426e-06,
      "loss": 0.4641,
      "step": 1325
    },
    {
      "epoch": 0.5486798679867987,
      "grad_norm": 2.4259356317621585,
      "learning_rate": 5.02520753715845e-06,
      "loss": 0.4489,
      "step": 1330
    },
    {
      "epoch": 0.5507425742574258,
      "grad_norm": 2.419683877837546,
      "learning_rate": 4.989196732430518e-06,
      "loss": 0.4523,
      "step": 1335
    },
    {
      "epoch": 0.5528052805280528,
      "grad_norm": 2.5368183953560077,
      "learning_rate": 4.953186488081362e-06,
      "loss": 0.4575,
      "step": 1340
    },
    {
      "epoch": 0.5548679867986799,
      "grad_norm": 2.6640720825787727,
      "learning_rate": 4.9171786720066465e-06,
      "loss": 0.4613,
      "step": 1345
    },
    {
      "epoch": 0.556930693069307,
      "grad_norm": 2.3570386936334025,
      "learning_rate": 4.881175151976075e-06,
      "loss": 0.4533,
      "step": 1350
    },
    {
      "epoch": 0.558993399339934,
      "grad_norm": 2.92205965965078,
      "learning_rate": 4.845177795536516e-06,
      "loss": 0.4315,
      "step": 1355
    },
    {
      "epoch": 0.5610561056105611,
      "grad_norm": 2.350803495083776,
      "learning_rate": 4.809188469915121e-06,
      "loss": 0.4611,
      "step": 1360
    },
    {
      "epoch": 0.5631188118811881,
      "grad_norm": 2.4307822932274084,
      "learning_rate": 4.773209041922472e-06,
      "loss": 0.4502,
      "step": 1365
    },
    {
      "epoch": 0.5651815181518152,
      "grad_norm": 2.471289928751954,
      "learning_rate": 4.737241377855751e-06,
      "loss": 0.4505,
      "step": 1370
    },
    {
      "epoch": 0.5672442244224423,
      "grad_norm": 2.493867736661699,
      "learning_rate": 4.7012873434019296e-06,
      "loss": 0.4518,
      "step": 1375
    },
    {
      "epoch": 0.5693069306930693,
      "grad_norm": 2.3665679103622974,
      "learning_rate": 4.6653488035409975e-06,
      "loss": 0.4303,
      "step": 1380
    },
    {
      "epoch": 0.5713696369636964,
      "grad_norm": 2.642220991131861,
      "learning_rate": 4.629427622449217e-06,
      "loss": 0.4458,
      "step": 1385
    },
    {
      "epoch": 0.5734323432343235,
      "grad_norm": 2.534010674982419,
      "learning_rate": 4.59352566340243e-06,
      "loss": 0.4407,
      "step": 1390
    },
    {
      "epoch": 0.5754950495049505,
      "grad_norm": 2.36153551659038,
      "learning_rate": 4.557644788679413e-06,
      "loss": 0.4343,
      "step": 1395
    },
    {
      "epoch": 0.5775577557755776,
      "grad_norm": 2.535141668952679,
      "learning_rate": 4.521786859465263e-06,
      "loss": 0.4412,
      "step": 1400
    },
    {
      "epoch": 0.5796204620462047,
      "grad_norm": 2.3990701439475446,
      "learning_rate": 4.485953735754872e-06,
      "loss": 0.4302,
      "step": 1405
    },
    {
      "epoch": 0.5816831683168316,
      "grad_norm": 2.463873380082242,
      "learning_rate": 4.450147276256439e-06,
      "loss": 0.4422,
      "step": 1410
    },
    {
      "epoch": 0.5837458745874587,
      "grad_norm": 2.3323366385949953,
      "learning_rate": 4.414369338295056e-06,
      "loss": 0.4358,
      "step": 1415
    },
    {
      "epoch": 0.5858085808580858,
      "grad_norm": 2.3533697334953874,
      "learning_rate": 4.37862177771637e-06,
      "loss": 0.4475,
      "step": 1420
    },
    {
      "epoch": 0.5878712871287128,
      "grad_norm": 2.4813703004487655,
      "learning_rate": 4.342906448790315e-06,
      "loss": 0.4244,
      "step": 1425
    },
    {
      "epoch": 0.5899339933993399,
      "grad_norm": 2.4263959721663686,
      "learning_rate": 4.307225204114927e-06,
      "loss": 0.4249,
      "step": 1430
    },
    {
      "epoch": 0.591996699669967,
      "grad_norm": 2.3890090123031675,
      "learning_rate": 4.271579894520254e-06,
      "loss": 0.4374,
      "step": 1435
    },
    {
      "epoch": 0.594059405940594,
      "grad_norm": 2.4162520655229316,
      "learning_rate": 4.235972368972343e-06,
      "loss": 0.42,
      "step": 1440
    },
    {
      "epoch": 0.5961221122112211,
      "grad_norm": 2.4663434097906416,
      "learning_rate": 4.200404474477341e-06,
      "loss": 0.4333,
      "step": 1445
    },
    {
      "epoch": 0.5981848184818482,
      "grad_norm": 2.643133094563938,
      "learning_rate": 4.16487805598568e-06,
      "loss": 0.4383,
      "step": 1450
    },
    {
      "epoch": 0.6002475247524752,
      "grad_norm": 2.369553776540664,
      "learning_rate": 4.12939495629638e-06,
      "loss": 0.421,
      "step": 1455
    },
    {
      "epoch": 0.6023102310231023,
      "grad_norm": 2.3739230876100277,
      "learning_rate": 4.093957015961465e-06,
      "loss": 0.4205,
      "step": 1460
    },
    {
      "epoch": 0.6043729372937293,
      "grad_norm": 2.7065139401959857,
      "learning_rate": 4.0585660731904855e-06,
      "loss": 0.4281,
      "step": 1465
    },
    {
      "epoch": 0.6064356435643564,
      "grad_norm": 2.303889526435788,
      "learning_rate": 4.023223963755168e-06,
      "loss": 0.4284,
      "step": 1470
    },
    {
      "epoch": 0.6084983498349835,
      "grad_norm": 2.4782043827381974,
      "learning_rate": 3.987932520894201e-06,
      "loss": 0.4347,
      "step": 1475
    },
    {
      "epoch": 0.6105610561056105,
      "grad_norm": 2.385221087724949,
      "learning_rate": 3.9526935752181275e-06,
      "loss": 0.4189,
      "step": 1480
    },
    {
      "epoch": 0.6126237623762376,
      "grad_norm": 2.4606523274098397,
      "learning_rate": 3.917508954614401e-06,
      "loss": 0.431,
      "step": 1485
    },
    {
      "epoch": 0.6146864686468647,
      "grad_norm": 2.3670999832545405,
      "learning_rate": 3.882380484152567e-06,
      "loss": 0.4284,
      "step": 1490
    },
    {
      "epoch": 0.6167491749174917,
      "grad_norm": 2.476043722834365,
      "learning_rate": 3.847309985989593e-06,
      "loss": 0.4241,
      "step": 1495
    },
    {
      "epoch": 0.6188118811881188,
      "grad_norm": 2.37313210444145,
      "learning_rate": 3.8122992792753534e-06,
      "loss": 0.4314,
      "step": 1500
    },
    {
      "epoch": 0.6208745874587459,
      "grad_norm": 2.382925261546362,
      "learning_rate": 3.777350180058264e-06,
      "loss": 0.4257,
      "step": 1505
    },
    {
      "epoch": 0.6229372937293729,
      "grad_norm": 2.2462536557274766,
      "learning_rate": 3.7424645011910847e-06,
      "loss": 0.4076,
      "step": 1510
    },
    {
      "epoch": 0.625,
      "grad_norm": 2.3505024535992733,
      "learning_rate": 3.707644052236887e-06,
      "loss": 0.3984,
      "step": 1515
    },
    {
      "epoch": 0.6270627062706271,
      "grad_norm": 2.4771391837897516,
      "learning_rate": 3.672890639375184e-06,
      "loss": 0.4172,
      "step": 1520
    },
    {
      "epoch": 0.6291254125412541,
      "grad_norm": 2.3721712805814197,
      "learning_rate": 3.6382060653082434e-06,
      "loss": 0.4097,
      "step": 1525
    },
    {
      "epoch": 0.6311881188118812,
      "grad_norm": 2.4306878354961454,
      "learning_rate": 3.6035921291675815e-06,
      "loss": 0.4076,
      "step": 1530
    },
    {
      "epoch": 0.6332508250825083,
      "grad_norm": 2.492688891911326,
      "learning_rate": 3.569050626420636e-06,
      "loss": 0.4253,
      "step": 1535
    },
    {
      "epoch": 0.6353135313531353,
      "grad_norm": 2.5661742951165585,
      "learning_rate": 3.5345833487776404e-06,
      "loss": 0.4238,
      "step": 1540
    },
    {
      "epoch": 0.6373762376237624,
      "grad_norm": 2.331662853575241,
      "learning_rate": 3.500192084098677e-06,
      "loss": 0.4183,
      "step": 1545
    },
    {
      "epoch": 0.6394389438943895,
      "grad_norm": 2.497643635749368,
      "learning_rate": 3.4658786163009416e-06,
      "loss": 0.4246,
      "step": 1550
    },
    {
      "epoch": 0.6415016501650165,
      "grad_norm": 2.5479933583941663,
      "learning_rate": 3.4316447252662142e-06,
      "loss": 0.4153,
      "step": 1555
    },
    {
      "epoch": 0.6435643564356436,
      "grad_norm": 2.455261221854015,
      "learning_rate": 3.3974921867485238e-06,
      "loss": 0.4206,
      "step": 1560
    },
    {
      "epoch": 0.6456270627062707,
      "grad_norm": 2.6087745317564712,
      "learning_rate": 3.3634227722820496e-06,
      "loss": 0.4078,
      "step": 1565
    },
    {
      "epoch": 0.6476897689768977,
      "grad_norm": 2.3621897623815302,
      "learning_rate": 3.3294382490892226e-06,
      "loss": 0.4306,
      "step": 1570
    },
    {
      "epoch": 0.6497524752475248,
      "grad_norm": 2.3806385163770067,
      "learning_rate": 3.2955403799890567e-06,
      "loss": 0.3937,
      "step": 1575
    },
    {
      "epoch": 0.6518151815181518,
      "grad_norm": 2.3247110625837863,
      "learning_rate": 3.261730923305717e-06,
      "loss": 0.4038,
      "step": 1580
    },
    {
      "epoch": 0.6538778877887789,
      "grad_norm": 2.375164955670525,
      "learning_rate": 3.2280116327773028e-06,
      "loss": 0.3958,
      "step": 1585
    },
    {
      "epoch": 0.655940594059406,
      "grad_norm": 2.4926818374164674,
      "learning_rate": 3.194384257464884e-06,
      "loss": 0.3996,
      "step": 1590
    },
    {
      "epoch": 0.658003300330033,
      "grad_norm": 2.495497820744883,
      "learning_rate": 3.160850541661779e-06,
      "loss": 0.3989,
      "step": 1595
    },
    {
      "epoch": 0.6600660066006601,
      "grad_norm": 2.4045930607358392,
      "learning_rate": 3.127412224803068e-06,
      "loss": 0.4063,
      "step": 1600
    },
    {
      "epoch": 0.6621287128712872,
      "grad_norm": 2.454505000870822,
      "learning_rate": 3.094071041375375e-06,
      "loss": 0.3978,
      "step": 1605
    },
    {
      "epoch": 0.6641914191419142,
      "grad_norm": 2.415988215095147,
      "learning_rate": 3.060828720826889e-06,
      "loss": 0.401,
      "step": 1610
    },
    {
      "epoch": 0.6662541254125413,
      "grad_norm": 2.4233639484301945,
      "learning_rate": 3.0276869874776632e-06,
      "loss": 0.4037,
      "step": 1615
    },
    {
      "epoch": 0.6683168316831684,
      "grad_norm": 2.4871754101710923,
      "learning_rate": 2.994647560430167e-06,
      "loss": 0.3957,
      "step": 1620
    },
    {
      "epoch": 0.6703795379537953,
      "grad_norm": 2.3870434326246475,
      "learning_rate": 2.961712153480118e-06,
      "loss": 0.406,
      "step": 1625
    },
    {
      "epoch": 0.6724422442244224,
      "grad_norm": 2.291964987725719,
      "learning_rate": 2.9288824750275803e-06,
      "loss": 0.4039,
      "step": 1630
    },
    {
      "epoch": 0.6745049504950495,
      "grad_norm": 2.3781197855311613,
      "learning_rate": 2.896160227988357e-06,
      "loss": 0.392,
      "step": 1635
    },
    {
      "epoch": 0.6765676567656765,
      "grad_norm": 2.419352678270268,
      "learning_rate": 2.8635471097056423e-06,
      "loss": 0.3948,
      "step": 1640
    },
    {
      "epoch": 0.6786303630363036,
      "grad_norm": 2.350196159609065,
      "learning_rate": 2.8310448118619967e-06,
      "loss": 0.3962,
      "step": 1645
    },
    {
      "epoch": 0.6806930693069307,
      "grad_norm": 2.359226335753887,
      "learning_rate": 2.7986550203915807e-06,
      "loss": 0.4032,
      "step": 1650
    },
    {
      "epoch": 0.6827557755775577,
      "grad_norm": 2.4118762465962558,
      "learning_rate": 2.7663794153927165e-06,
      "loss": 0.4063,
      "step": 1655
    },
    {
      "epoch": 0.6848184818481848,
      "grad_norm": 2.409985641673926,
      "learning_rate": 2.7342196710407337e-06,
      "loss": 0.3873,
      "step": 1660
    },
    {
      "epoch": 0.6868811881188119,
      "grad_norm": 2.3453850159224916,
      "learning_rate": 2.7021774555011214e-06,
      "loss": 0.3935,
      "step": 1665
    },
    {
      "epoch": 0.6889438943894389,
      "grad_norm": 2.38810093387651,
      "learning_rate": 2.6702544308430122e-06,
      "loss": 0.3948,
      "step": 1670
    },
    {
      "epoch": 0.691006600660066,
      "grad_norm": 2.4973217878271763,
      "learning_rate": 2.6384522529529542e-06,
      "loss": 0.3922,
      "step": 1675
    },
    {
      "epoch": 0.693069306930693,
      "grad_norm": 2.4765249088902044,
      "learning_rate": 2.6067725714490307e-06,
      "loss": 0.3971,
      "step": 1680
    },
    {
      "epoch": 0.6951320132013201,
      "grad_norm": 2.3774272170618076,
      "learning_rate": 2.5752170295952856e-06,
      "loss": 0.3845,
      "step": 1685
    },
    {
      "epoch": 0.6971947194719472,
      "grad_norm": 2.321619743547469,
      "learning_rate": 2.5437872642164818e-06,
      "loss": 0.3976,
      "step": 1690
    },
    {
      "epoch": 0.6992574257425742,
      "grad_norm": 2.4409520235076516,
      "learning_rate": 2.5124849056132094e-06,
      "loss": 0.4011,
      "step": 1695
    },
    {
      "epoch": 0.7013201320132013,
      "grad_norm": 2.3383182088114416,
      "learning_rate": 2.4813115774773046e-06,
      "loss": 0.3937,
      "step": 1700
    },
    {
      "epoch": 0.7033828382838284,
      "grad_norm": 2.297945871228005,
      "learning_rate": 2.4502688968076416e-06,
      "loss": 0.3734,
      "step": 1705
    },
    {
      "epoch": 0.7054455445544554,
      "grad_norm": 2.326552654875149,
      "learning_rate": 2.4193584738262426e-06,
      "loss": 0.378,
      "step": 1710
    },
    {
      "epoch": 0.7075082508250825,
      "grad_norm": 2.584655230215262,
      "learning_rate": 2.388581911894767e-06,
      "loss": 0.3768,
      "step": 1715
    },
    {
      "epoch": 0.7095709570957096,
      "grad_norm": 2.3738045585977177,
      "learning_rate": 2.357940807431339e-06,
      "loss": 0.3969,
      "step": 1720
    },
    {
      "epoch": 0.7116336633663366,
      "grad_norm": 2.393762942918226,
      "learning_rate": 2.3274367498277246e-06,
      "loss": 0.3981,
      "step": 1725
    },
    {
      "epoch": 0.7136963696369637,
      "grad_norm": 2.5701720220513935,
      "learning_rate": 2.2970713213669127e-06,
      "loss": 0.3977,
      "step": 1730
    },
    {
      "epoch": 0.7157590759075908,
      "grad_norm": 2.3432875976560337,
      "learning_rate": 2.266846097141026e-06,
      "loss": 0.3853,
      "step": 1735
    },
    {
      "epoch": 0.7178217821782178,
      "grad_norm": 2.266576155932609,
      "learning_rate": 2.2367626449696168e-06,
      "loss": 0.3873,
      "step": 1740
    },
    {
      "epoch": 0.7198844884488449,
      "grad_norm": 2.4100813326697113,
      "learning_rate": 2.206822525318352e-06,
      "loss": 0.3796,
      "step": 1745
    },
    {
      "epoch": 0.721947194719472,
      "grad_norm": 2.3097612430692824,
      "learning_rate": 2.1770272912180577e-06,
      "loss": 0.3828,
      "step": 1750
    },
    {
      "epoch": 0.724009900990099,
      "grad_norm": 2.381872158371206,
      "learning_rate": 2.1473784881841753e-06,
      "loss": 0.3877,
      "step": 1755
    },
    {
      "epoch": 0.7260726072607261,
      "grad_norm": 2.3045933376924728,
      "learning_rate": 2.117877654136584e-06,
      "loss": 0.3808,
      "step": 1760
    },
    {
      "epoch": 0.7281353135313532,
      "grad_norm": 2.326638441980535,
      "learning_rate": 2.088526319319827e-06,
      "loss": 0.3921,
      "step": 1765
    },
    {
      "epoch": 0.7301980198019802,
      "grad_norm": 2.4833712672454147,
      "learning_rate": 2.059326006223743e-06,
      "loss": 0.3787,
      "step": 1770
    },
    {
      "epoch": 0.7322607260726073,
      "grad_norm": 2.3365364725152933,
      "learning_rate": 2.030278229504484e-06,
      "loss": 0.3849,
      "step": 1775
    },
    {
      "epoch": 0.7343234323432343,
      "grad_norm": 2.4948287704930365,
      "learning_rate": 2.001384495905954e-06,
      "loss": 0.3774,
      "step": 1780
    },
    {
      "epoch": 0.7363861386138614,
      "grad_norm": 2.562327216184289,
      "learning_rate": 1.972646304181656e-06,
      "loss": 0.3958,
      "step": 1785
    },
    {
      "epoch": 0.7384488448844885,
      "grad_norm": 2.4385787815481432,
      "learning_rate": 1.944065145016935e-06,
      "loss": 0.3796,
      "step": 1790
    },
    {
      "epoch": 0.7405115511551155,
      "grad_norm": 2.2522811433040837,
      "learning_rate": 1.9156425009516736e-06,
      "loss": 0.3877,
      "step": 1795
    },
    {
      "epoch": 0.7425742574257426,
      "grad_norm": 2.5063965305246416,
      "learning_rate": 1.8873798463033742e-06,
      "loss": 0.3803,
      "step": 1800
    },
    {
      "epoch": 0.7446369636963697,
      "grad_norm": 2.6814884076074628,
      "learning_rate": 1.8592786470906932e-06,
      "loss": 0.3836,
      "step": 1805
    },
    {
      "epoch": 0.7466996699669967,
      "grad_norm": 2.2538094160236324,
      "learning_rate": 1.8313403609573976e-06,
      "loss": 0.3723,
      "step": 1810
    },
    {
      "epoch": 0.7487623762376238,
      "grad_norm": 2.4236325470496274,
      "learning_rate": 1.8035664370967493e-06,
      "loss": 0.3831,
      "step": 1815
    },
    {
      "epoch": 0.7508250825082509,
      "grad_norm": 2.4653728464781537,
      "learning_rate": 1.775958316176339e-06,
      "loss": 0.3809,
      "step": 1820
    },
    {
      "epoch": 0.7528877887788779,
      "grad_norm": 2.473806197305363,
      "learning_rate": 1.7485174302633557e-06,
      "loss": 0.3798,
      "step": 1825
    },
    {
      "epoch": 0.754950495049505,
      "grad_norm": 2.530281290392794,
      "learning_rate": 1.721245202750299e-06,
      "loss": 0.3837,
      "step": 1830
    },
    {
      "epoch": 0.7570132013201321,
      "grad_norm": 2.307304666583442,
      "learning_rate": 1.694143048281156e-06,
      "loss": 0.378,
      "step": 1835
    },
    {
      "epoch": 0.759075907590759,
      "grad_norm": 2.498849418663323,
      "learning_rate": 1.6672123726780083e-06,
      "loss": 0.376,
      "step": 1840
    },
    {
      "epoch": 0.7611386138613861,
      "grad_norm": 2.4025100016563186,
      "learning_rate": 1.6404545728681232e-06,
      "loss": 0.373,
      "step": 1845
    },
    {
      "epoch": 0.7632013201320133,
      "grad_norm": 2.2171879939222894,
      "learning_rate": 1.613871036811489e-06,
      "loss": 0.3648,
      "step": 1850
    },
    {
      "epoch": 0.7652640264026402,
      "grad_norm": 2.3047287367214837,
      "learning_rate": 1.5874631434288128e-06,
      "loss": 0.3655,
      "step": 1855
    },
    {
      "epoch": 0.7673267326732673,
      "grad_norm": 2.4662734075880848,
      "learning_rate": 1.5612322625300064e-06,
      "loss": 0.3714,
      "step": 1860
    },
    {
      "epoch": 0.7693894389438944,
      "grad_norm": 2.331832423235198,
      "learning_rate": 1.5351797547431212e-06,
      "loss": 0.3675,
      "step": 1865
    },
    {
      "epoch": 0.7714521452145214,
      "grad_norm": 2.4200832008390196,
      "learning_rate": 1.5093069714437803e-06,
      "loss": 0.3793,
      "step": 1870
    },
    {
      "epoch": 0.7735148514851485,
      "grad_norm": 2.4710921538024717,
      "learning_rate": 1.483615254685075e-06,
      "loss": 0.3767,
      "step": 1875
    },
    {
      "epoch": 0.7755775577557755,
      "grad_norm": 2.3547080899947543,
      "learning_rate": 1.4581059371279516e-06,
      "loss": 0.3487,
      "step": 1880
    },
    {
      "epoch": 0.7776402640264026,
      "grad_norm": 2.3257391997447927,
      "learning_rate": 1.4327803419720836e-06,
      "loss": 0.35,
      "step": 1885
    },
    {
      "epoch": 0.7797029702970297,
      "grad_norm": 2.313586644368121,
      "learning_rate": 1.4076397828872441e-06,
      "loss": 0.3606,
      "step": 1890
    },
    {
      "epoch": 0.7817656765676567,
      "grad_norm": 2.3428094787987574,
      "learning_rate": 1.3826855639451492e-06,
      "loss": 0.3653,
      "step": 1895
    },
    {
      "epoch": 0.7838283828382838,
      "grad_norm": 2.45005288836788,
      "learning_rate": 1.357918979551831e-06,
      "loss": 0.3754,
      "step": 1900
    },
    {
      "epoch": 0.7858910891089109,
      "grad_norm": 2.3976786255699754,
      "learning_rate": 1.333341314380479e-06,
      "loss": 0.3667,
      "step": 1905
    },
    {
      "epoch": 0.7879537953795379,
      "grad_norm": 2.5250135396691604,
      "learning_rate": 1.308953843304816e-06,
      "loss": 0.3614,
      "step": 1910
    },
    {
      "epoch": 0.790016501650165,
      "grad_norm": 2.348295501687973,
      "learning_rate": 1.2847578313329623e-06,
      "loss": 0.3608,
      "step": 1915
    },
    {
      "epoch": 0.7920792079207921,
      "grad_norm": 2.367799355908408,
      "learning_rate": 1.2607545335418154e-06,
      "loss": 0.3633,
      "step": 1920
    },
    {
      "epoch": 0.7941419141914191,
      "grad_norm": 2.45330068107534,
      "learning_rate": 1.2369451950119553e-06,
      "loss": 0.3692,
      "step": 1925
    },
    {
      "epoch": 0.7962046204620462,
      "grad_norm": 2.249645433984358,
      "learning_rate": 1.2133310507630535e-06,
      "loss": 0.3586,
      "step": 1930
    },
    {
      "epoch": 0.7982673267326733,
      "grad_norm": 2.4175222905557012,
      "learning_rate": 1.189913325689816e-06,
      "loss": 0.3622,
      "step": 1935
    },
    {
      "epoch": 0.8003300330033003,
      "grad_norm": 2.3871779225388616,
      "learning_rate": 1.166693234498446e-06,
      "loss": 0.3536,
      "step": 1940
    },
    {
      "epoch": 0.8023927392739274,
      "grad_norm": 2.4452723810844743,
      "learning_rate": 1.1436719816436293e-06,
      "loss": 0.3613,
      "step": 1945
    },
    {
      "epoch": 0.8044554455445545,
      "grad_norm": 2.48712631644359,
      "learning_rate": 1.120850761266068e-06,
      "loss": 0.3669,
      "step": 1950
    },
    {
      "epoch": 0.8065181518151815,
      "grad_norm": 2.5253662509329393,
      "learning_rate": 1.098230757130529e-06,
      "loss": 0.3727,
      "step": 1955
    },
    {
      "epoch": 0.8085808580858086,
      "grad_norm": 2.380725750843834,
      "learning_rate": 1.075813142564448e-06,
      "loss": 0.3513,
      "step": 1960
    },
    {
      "epoch": 0.8106435643564357,
      "grad_norm": 2.358274468399545,
      "learning_rate": 1.053599080397068e-06,
      "loss": 0.3702,
      "step": 1965
    },
    {
      "epoch": 0.8127062706270627,
      "grad_norm": 2.2547142529874806,
      "learning_rate": 1.031589722899109e-06,
      "loss": 0.3612,
      "step": 1970
    },
    {
      "epoch": 0.8147689768976898,
      "grad_norm": 2.3254385156527664,
      "learning_rate": 1.0097862117230162e-06,
      "loss": 0.3468,
      "step": 1975
    },
    {
      "epoch": 0.8168316831683168,
      "grad_norm": 2.3457610696785425,
      "learning_rate": 9.881896778437328e-07,
      "loss": 0.3565,
      "step": 1980
    },
    {
      "epoch": 0.8188943894389439,
      "grad_norm": 2.3354355561954314,
      "learning_rate": 9.6680124150003e-07,
      "loss": 0.3545,
      "step": 1985
    },
    {
      "epoch": 0.820957095709571,
      "grad_norm": 2.3649448904745904,
      "learning_rate": 9.456220121364091e-07,
      "loss": 0.3544,
      "step": 1990
    },
    {
      "epoch": 0.823019801980198,
      "grad_norm": 2.562831507635733,
      "learning_rate": 9.24653088345544e-07,
      "loss": 0.3487,
      "step": 1995
    },
    {
      "epoch": 0.8250825082508251,
      "grad_norm": 2.5452000028587145,
      "learning_rate": 9.038955578113018e-07,
      "loss": 0.3511,
      "step": 2000
    },
    {
      "epoch": 0.8271452145214522,
      "grad_norm": 2.279868733040485,
      "learning_rate": 8.833504972523238e-07,
      "loss": 0.3507,
      "step": 2005
    },
    {
      "epoch": 0.8292079207920792,
      "grad_norm": 2.447577859196897,
      "learning_rate": 8.630189723661663e-07,
      "loss": 0.3505,
      "step": 2010
    },
    {
      "epoch": 0.8312706270627063,
      "grad_norm": 2.3258834832843855,
      "learning_rate": 8.429020377740338e-07,
      "loss": 0.3436,
      "step": 2015
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 2.419682969212795,
      "learning_rate": 8.230007369660636e-07,
      "loss": 0.3488,
      "step": 2020
    },
    {
      "epoch": 0.8353960396039604,
      "grad_norm": 2.386888258570233,
      "learning_rate": 8.033161022472063e-07,
      "loss": 0.3535,
      "step": 2025
    },
    {
      "epoch": 0.8374587458745875,
      "grad_norm": 2.3745894252664916,
      "learning_rate": 7.838491546836763e-07,
      "loss": 0.3508,
      "step": 2030
    },
    {
      "epoch": 0.8395214521452146,
      "grad_norm": 2.384176050336693,
      "learning_rate": 7.646009040499846e-07,
      "loss": 0.3431,
      "step": 2035
    },
    {
      "epoch": 0.8415841584158416,
      "grad_norm": 2.4506033032717274,
      "learning_rate": 7.455723487765664e-07,
      "loss": 0.349,
      "step": 2040
    },
    {
      "epoch": 0.8436468646864687,
      "grad_norm": 2.4755241753598956,
      "learning_rate": 7.267644758979869e-07,
      "loss": 0.3465,
      "step": 2045
    },
    {
      "epoch": 0.8457095709570958,
      "grad_norm": 2.3912218867869472,
      "learning_rate": 7.08178261001743e-07,
      "loss": 0.3606,
      "step": 2050
    },
    {
      "epoch": 0.8477722772277227,
      "grad_norm": 2.4752358064551925,
      "learning_rate": 6.898146681776629e-07,
      "loss": 0.3451,
      "step": 2055
    },
    {
      "epoch": 0.8498349834983498,
      "grad_norm": 2.4469151560024365,
      "learning_rate": 6.7167464996789e-07,
      "loss": 0.3515,
      "step": 2060
    },
    {
      "epoch": 0.851897689768977,
      "grad_norm": 2.4398936114143024,
      "learning_rate": 6.537591473174814e-07,
      "loss": 0.3602,
      "step": 2065
    },
    {
      "epoch": 0.8539603960396039,
      "grad_norm": 2.499314654767742,
      "learning_rate": 6.360690895255916e-07,
      "loss": 0.3624,
      "step": 2070
    },
    {
      "epoch": 0.856023102310231,
      "grad_norm": 2.386870027944476,
      "learning_rate": 6.186053941972775e-07,
      "loss": 0.3449,
      "step": 2075
    },
    {
      "epoch": 0.858085808580858,
      "grad_norm": 2.4376398779640347,
      "learning_rate": 6.013689671958944e-07,
      "loss": 0.3612,
      "step": 2080
    },
    {
      "epoch": 0.8601485148514851,
      "grad_norm": 2.530488463039945,
      "learning_rate": 5.84360702596109e-07,
      "loss": 0.3635,
      "step": 2085
    },
    {
      "epoch": 0.8622112211221122,
      "grad_norm": 2.393526534346384,
      "learning_rate": 5.67581482637527e-07,
      "loss": 0.3484,
      "step": 2090
    },
    {
      "epoch": 0.8642739273927392,
      "grad_norm": 2.439985682818376,
      "learning_rate": 5.510321776789213e-07,
      "loss": 0.3562,
      "step": 2095
    },
    {
      "epoch": 0.8663366336633663,
      "grad_norm": 2.404561027314301,
      "learning_rate": 5.347136461530966e-07,
      "loss": 0.3515,
      "step": 2100
    },
    {
      "epoch": 0.8683993399339934,
      "grad_norm": 2.529134431042306,
      "learning_rate": 5.186267345223539e-07,
      "loss": 0.336,
      "step": 2105
    },
    {
      "epoch": 0.8704620462046204,
      "grad_norm": 2.4549178066211796,
      "learning_rate": 5.027722772345828e-07,
      "loss": 0.3483,
      "step": 2110
    },
    {
      "epoch": 0.8725247524752475,
      "grad_norm": 2.404643832863399,
      "learning_rate": 4.871510966799847e-07,
      "loss": 0.3406,
      "step": 2115
    },
    {
      "epoch": 0.8745874587458746,
      "grad_norm": 2.503938904115535,
      "learning_rate": 4.717640031484055e-07,
      "loss": 0.3629,
      "step": 2120
    },
    {
      "epoch": 0.8766501650165016,
      "grad_norm": 2.3230215699664183,
      "learning_rate": 4.566117947873139e-07,
      "loss": 0.3542,
      "step": 2125
    },
    {
      "epoch": 0.8787128712871287,
      "grad_norm": 2.510839585616837,
      "learning_rate": 4.4169525756039164e-07,
      "loss": 0.3504,
      "step": 2130
    },
    {
      "epoch": 0.8807755775577558,
      "grad_norm": 2.4483049384414035,
      "learning_rate": 4.2701516520677054e-07,
      "loss": 0.3426,
      "step": 2135
    },
    {
      "epoch": 0.8828382838283828,
      "grad_norm": 2.3212952681789445,
      "learning_rate": 4.1257227920089684e-07,
      "loss": 0.3467,
      "step": 2140
    },
    {
      "epoch": 0.8849009900990099,
      "grad_norm": 2.306034813512439,
      "learning_rate": 3.983673487130313e-07,
      "loss": 0.3496,
      "step": 2145
    },
    {
      "epoch": 0.886963696369637,
      "grad_norm": 2.351525266795836,
      "learning_rate": 3.8440111057038874e-07,
      "loss": 0.3446,
      "step": 2150
    },
    {
      "epoch": 0.889026402640264,
      "grad_norm": 2.33661307568645,
      "learning_rate": 3.706742892189197e-07,
      "loss": 0.3385,
      "step": 2155
    },
    {
      "epoch": 0.8910891089108911,
      "grad_norm": 2.479408467495357,
      "learning_rate": 3.5718759668572913e-07,
      "loss": 0.3471,
      "step": 2160
    },
    {
      "epoch": 0.8931518151815182,
      "grad_norm": 2.4454103391168687,
      "learning_rate": 3.439417325421468e-07,
      "loss": 0.347,
      "step": 2165
    },
    {
      "epoch": 0.8952145214521452,
      "grad_norm": 2.4210315871292174,
      "learning_rate": 3.3093738386743734e-07,
      "loss": 0.3387,
      "step": 2170
    },
    {
      "epoch": 0.8972772277227723,
      "grad_norm": 2.415104298972758,
      "learning_rate": 3.1817522521316034e-07,
      "loss": 0.3328,
      "step": 2175
    },
    {
      "epoch": 0.8993399339933993,
      "grad_norm": 2.6057415546092955,
      "learning_rate": 3.0565591856818236e-07,
      "loss": 0.3384,
      "step": 2180
    },
    {
      "epoch": 0.9014026402640264,
      "grad_norm": 2.5470334526474083,
      "learning_rate": 2.9338011332433525e-07,
      "loss": 0.3456,
      "step": 2185
    },
    {
      "epoch": 0.9034653465346535,
      "grad_norm": 2.3087270354481495,
      "learning_rate": 2.813484462427357e-07,
      "loss": 0.3416,
      "step": 2190
    },
    {
      "epoch": 0.9055280528052805,
      "grad_norm": 2.3652461290680615,
      "learning_rate": 2.695615414207542e-07,
      "loss": 0.3419,
      "step": 2195
    },
    {
      "epoch": 0.9075907590759076,
      "grad_norm": 2.3153894157922084,
      "learning_rate": 2.5802001025963917e-07,
      "loss": 0.3457,
      "step": 2200
    },
    {
      "epoch": 0.9096534653465347,
      "grad_norm": 2.437439490370531,
      "learning_rate": 2.467244514328082e-07,
      "loss": 0.3336,
      "step": 2205
    },
    {
      "epoch": 0.9117161716171617,
      "grad_norm": 2.48712819745237,
      "learning_rate": 2.3567545085478983e-07,
      "loss": 0.3473,
      "step": 2210
    },
    {
      "epoch": 0.9137788778877888,
      "grad_norm": 2.467836274329266,
      "learning_rate": 2.248735816508324e-07,
      "loss": 0.3342,
      "step": 2215
    },
    {
      "epoch": 0.9158415841584159,
      "grad_norm": 2.43610224609289,
      "learning_rate": 2.1431940412717843e-07,
      "loss": 0.349,
      "step": 2220
    },
    {
      "epoch": 0.9179042904290429,
      "grad_norm": 2.4583486850305265,
      "learning_rate": 2.040134657419951e-07,
      "loss": 0.3577,
      "step": 2225
    },
    {
      "epoch": 0.91996699669967,
      "grad_norm": 2.307477095808075,
      "learning_rate": 1.9395630107698293e-07,
      "loss": 0.3375,
      "step": 2230
    },
    {
      "epoch": 0.9220297029702971,
      "grad_norm": 2.6904928261878207,
      "learning_rate": 1.8414843180964316e-07,
      "loss": 0.3381,
      "step": 2235
    },
    {
      "epoch": 0.9240924092409241,
      "grad_norm": 2.395973262946888,
      "learning_rate": 1.7459036668621586e-07,
      "loss": 0.3487,
      "step": 2240
    },
    {
      "epoch": 0.9261551155115512,
      "grad_norm": 2.5974442906075654,
      "learning_rate": 1.6528260149529573e-07,
      "loss": 0.3462,
      "step": 2245
    },
    {
      "epoch": 0.9282178217821783,
      "grad_norm": 2.4163928757500734,
      "learning_rate": 1.562256190421102e-07,
      "loss": 0.3338,
      "step": 2250
    },
    {
      "epoch": 0.9302805280528053,
      "grad_norm": 2.3929529097227595,
      "learning_rate": 1.4741988912347848e-07,
      "loss": 0.3327,
      "step": 2255
    },
    {
      "epoch": 0.9323432343234324,
      "grad_norm": 2.333345645365385,
      "learning_rate": 1.3886586850344276e-07,
      "loss": 0.3601,
      "step": 2260
    },
    {
      "epoch": 0.9344059405940595,
      "grad_norm": 2.302538628872532,
      "learning_rate": 1.30564000889572e-07,
      "loss": 0.3449,
      "step": 2265
    },
    {
      "epoch": 0.9364686468646864,
      "grad_norm": 2.453853466702325,
      "learning_rate": 1.225147169099511e-07,
      "loss": 0.3318,
      "step": 2270
    },
    {
      "epoch": 0.9385313531353136,
      "grad_norm": 2.490214641545018,
      "learning_rate": 1.147184340908386e-07,
      "loss": 0.3558,
      "step": 2275
    },
    {
      "epoch": 0.9405940594059405,
      "grad_norm": 2.4840208163150432,
      "learning_rate": 1.0717555683501413e-07,
      "loss": 0.3486,
      "step": 2280
    },
    {
      "epoch": 0.9426567656765676,
      "grad_norm": 2.378818883269412,
      "learning_rate": 9.988647640079785e-08,
      "loss": 0.3305,
      "step": 2285
    },
    {
      "epoch": 0.9447194719471947,
      "grad_norm": 2.254918829640657,
      "learning_rate": 9.285157088175678e-08,
      "loss": 0.3363,
      "step": 2290
    },
    {
      "epoch": 0.9467821782178217,
      "grad_norm": 2.432890954515314,
      "learning_rate": 8.607120518709156e-08,
      "loss": 0.3417,
      "step": 2295
    },
    {
      "epoch": 0.9488448844884488,
      "grad_norm": 2.4920984458242503,
      "learning_rate": 7.954573102271157e-08,
      "loss": 0.3476,
      "step": 2300
    },
    {
      "epoch": 0.9509075907590759,
      "grad_norm": 2.47335546582837,
      "learning_rate": 7.327548687298625e-08,
      "loss": 0.3504,
      "step": 2305
    },
    {
      "epoch": 0.9529702970297029,
      "grad_norm": 2.4140844536841572,
      "learning_rate": 6.726079798319185e-08,
      "loss": 0.337,
      "step": 2310
    },
    {
      "epoch": 0.95503300330033,
      "grad_norm": 2.5771622061139046,
      "learning_rate": 6.150197634263888e-08,
      "loss": 0.3285,
      "step": 2315
    },
    {
      "epoch": 0.9570957095709571,
      "grad_norm": 2.3569938271313795,
      "learning_rate": 5.599932066848834e-08,
      "loss": 0.3539,
      "step": 2320
    },
    {
      "epoch": 0.9591584158415841,
      "grad_norm": 2.4499385246168344,
      "learning_rate": 5.0753116390258594e-08,
      "loss": 0.3459,
      "step": 2325
    },
    {
      "epoch": 0.9612211221122112,
      "grad_norm": 2.4764510130658386,
      "learning_rate": 4.576363563501718e-08,
      "loss": 0.346,
      "step": 2330
    },
    {
      "epoch": 0.9632838283828383,
      "grad_norm": 2.596731808326133,
      "learning_rate": 4.103113721326768e-08,
      "loss": 0.3362,
      "step": 2335
    },
    {
      "epoch": 0.9653465346534653,
      "grad_norm": 2.5900213567140176,
      "learning_rate": 3.655586660552324e-08,
      "loss": 0.3363,
      "step": 2340
    },
    {
      "epoch": 0.9674092409240924,
      "grad_norm": 2.313646904697264,
      "learning_rate": 3.233805594957506e-08,
      "loss": 0.3447,
      "step": 2345
    },
    {
      "epoch": 0.9694719471947195,
      "grad_norm": 2.321374397889606,
      "learning_rate": 2.8377924028449855e-08,
      "loss": 0.3389,
      "step": 2350
    },
    {
      "epoch": 0.9715346534653465,
      "grad_norm": 2.809082658776519,
      "learning_rate": 2.4675676259059976e-08,
      "loss": 0.3488,
      "step": 2355
    },
    {
      "epoch": 0.9735973597359736,
      "grad_norm": 2.373910272093463,
      "learning_rate": 2.123150468155144e-08,
      "loss": 0.3475,
      "step": 2360
    },
    {
      "epoch": 0.9756600660066007,
      "grad_norm": 2.5332366911693596,
      "learning_rate": 1.8045587949339637e-08,
      "loss": 0.3328,
      "step": 2365
    },
    {
      "epoch": 0.9777227722772277,
      "grad_norm": 2.313810796400828,
      "learning_rate": 1.5118091319843985e-08,
      "loss": 0.3407,
      "step": 2370
    },
    {
      "epoch": 0.9797854785478548,
      "grad_norm": 2.2683681366110777,
      "learning_rate": 1.2449166645915333e-08,
      "loss": 0.3341,
      "step": 2375
    },
    {
      "epoch": 0.9818481848184818,
      "grad_norm": 2.5404146692025638,
      "learning_rate": 1.0038952367958376e-08,
      "loss": 0.3287,
      "step": 2380
    },
    {
      "epoch": 0.9839108910891089,
      "grad_norm": 2.433802925714093,
      "learning_rate": 7.887573506752954e-09,
      "loss": 0.3484,
      "step": 2385
    },
    {
      "epoch": 0.985973597359736,
      "grad_norm": 2.3794895761915953,
      "learning_rate": 5.9951416569659085e-09,
      "loss": 0.3433,
      "step": 2390
    },
    {
      "epoch": 0.988036303630363,
      "grad_norm": 2.476927805960015,
      "learning_rate": 4.361754981365152e-09,
      "loss": 0.3393,
      "step": 2395
    },
    {
      "epoch": 0.9900990099009901,
      "grad_norm": 2.438512650786407,
      "learning_rate": 2.98749820572708e-09,
      "loss": 0.3389,
      "step": 2400
    },
    {
      "epoch": 0.9921617161716172,
      "grad_norm": 2.3365090706536735,
      "learning_rate": 1.8724426144395293e-09,
      "loss": 0.3354,
      "step": 2405
    },
    {
      "epoch": 0.9942244224422442,
      "grad_norm": 2.4286642972043313,
      "learning_rate": 1.0166460468080674e-09,
      "loss": 0.3527,
      "step": 2410
    },
    {
      "epoch": 0.9962871287128713,
      "grad_norm": 2.4446712259075682,
      "learning_rate": 4.2015289405339386e-10,
      "loss": 0.3393,
      "step": 2415
    },
    {
      "epoch": 0.9983498349834984,
      "grad_norm": 2.395544519671724,
      "learning_rate": 8.29940970092924e-11,
      "loss": 0.3433,
      "step": 2420
    },
    {
      "epoch": 1.0,
      "eval_runtime": 3.4199,
      "eval_samples_per_second": 2.924,
      "eval_steps_per_second": 0.877,
      "step": 2424
    },
    {
      "epoch": 1.0,
      "step": 2424,
      "total_flos": 253768142684160.0,
      "train_loss": 0.0,
      "train_runtime": 0.0092,
      "train_samples_per_second": 4199719.219,
      "train_steps_per_second": 262529.834
    }
  ],
  "logging_steps": 5,
  "max_steps": 2424,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 253768142684160.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}