{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9991254154276719,
  "eval_steps": 500,
  "global_step": 1071,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004664451052416769,
      "grad_norm": 0.0,
      "learning_rate": 0.0,
      "loss": 0.8476,
      "step": 5
    },
    {
      "epoch": 0.009328902104833538,
      "grad_norm": 0.0,
      "learning_rate": 0.0,
      "loss": 0.861,
      "step": 10
    },
    {
      "epoch": 0.013993353157250307,
      "grad_norm": 0.0,
      "learning_rate": 0.0,
      "loss": 0.8574,
      "step": 15
    },
    {
      "epoch": 0.018657804209667075,
      "grad_norm": 1.7217303169710205,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.8586,
      "step": 20
    },
    {
      "epoch": 0.023322255262083844,
      "grad_norm": 1.0811304401552388,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.8027,
      "step": 25
    },
    {
      "epoch": 0.027986706314500613,
      "grad_norm": 0.6678730431826253,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 0.7386,
      "step": 30
    },
    {
      "epoch": 0.03265115736691738,
      "grad_norm": 0.49414971424324633,
      "learning_rate": 1.7592592592592595e-05,
      "loss": 0.6911,
      "step": 35
    },
    {
      "epoch": 0.03731560841933415,
      "grad_norm": 0.4956290038506038,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.6738,
      "step": 40
    },
    {
      "epoch": 0.041980059471750916,
      "grad_norm": 0.45113968047451636,
      "learning_rate": 2.6851851851851855e-05,
      "loss": 0.6615,
      "step": 45
    },
    {
      "epoch": 0.04664451052416769,
      "grad_norm": 0.4151578102626258,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.637,
      "step": 50
    },
    {
      "epoch": 0.051308961576584454,
      "grad_norm": 0.41598749293660114,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.6521,
      "step": 55
    },
    {
      "epoch": 0.055973412629001226,
      "grad_norm": 0.44364992697109346,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.6416,
      "step": 60
    },
    {
      "epoch": 0.06063786368141799,
      "grad_norm": 0.4334859115222668,
      "learning_rate": 4.5370370370370374e-05,
      "loss": 0.6179,
      "step": 65
    },
    {
      "epoch": 0.06530231473383476,
      "grad_norm": 0.4168421218480663,
      "learning_rate": 5e-05,
      "loss": 0.6455,
      "step": 70
    },
    {
      "epoch": 0.06996676578625154,
      "grad_norm": 0.5598377478471296,
      "learning_rate": 4.999731625198103e-05,
      "loss": 0.6218,
      "step": 75
    },
    {
      "epoch": 0.0746312168386683,
      "grad_norm": 0.5562791478235322,
      "learning_rate": 4.998926564814665e-05,
      "loss": 0.6164,
      "step": 80
    },
    {
      "epoch": 0.07929566789108507,
      "grad_norm": 0.494836009809662,
      "learning_rate": 4.997585010901171e-05,
      "loss": 0.6083,
      "step": 85
    },
    {
      "epoch": 0.08396011894350183,
      "grad_norm": 0.47761536307054875,
      "learning_rate": 4.995707283492525e-05,
      "loss": 0.597,
      "step": 90
    },
    {
      "epoch": 0.08862456999591861,
      "grad_norm": 0.5012276873640602,
      "learning_rate": 4.9932938305306997e-05,
      "loss": 0.5986,
      "step": 95
    },
    {
      "epoch": 0.09328902104833538,
      "grad_norm": 0.558780335812584,
      "learning_rate": 4.990345227757883e-05,
      "loss": 0.6108,
      "step": 100
    },
    {
      "epoch": 0.09795347210075214,
      "grad_norm": 0.5027530843851625,
      "learning_rate": 4.986862178579129e-05,
      "loss": 0.616,
      "step": 105
    },
    {
      "epoch": 0.10261792315316891,
      "grad_norm": 0.5276174777450547,
      "learning_rate": 4.982845513894555e-05,
      "loss": 0.6028,
      "step": 110
    },
    {
      "epoch": 0.10728237420558567,
      "grad_norm": 0.48105561141886855,
      "learning_rate": 4.9782961919011294e-05,
      "loss": 0.5926,
      "step": 115
    },
    {
      "epoch": 0.11194682525800245,
      "grad_norm": 0.5884754531701646,
      "learning_rate": 4.973215297864088e-05,
      "loss": 0.5997,
      "step": 120
    },
    {
      "epoch": 0.11661127631041922,
      "grad_norm": 0.5038947888532336,
      "learning_rate": 4.967604043858034e-05,
      "loss": 0.5966,
      "step": 125
    },
    {
      "epoch": 0.12127572736283598,
      "grad_norm": 0.4628854625972218,
      "learning_rate": 4.9614637684777964e-05,
      "loss": 0.5889,
      "step": 130
    },
    {
      "epoch": 0.12594017841525276,
      "grad_norm": 0.369569834468325,
      "learning_rate": 4.954795936519099e-05,
      "loss": 0.5886,
      "step": 135
    },
    {
      "epoch": 0.13060462946766951,
      "grad_norm": 0.510224312267307,
      "learning_rate": 4.9476021386291255e-05,
      "loss": 0.5904,
      "step": 140
    },
    {
      "epoch": 0.1352690805200863,
      "grad_norm": 0.5165955737252187,
      "learning_rate": 4.93988409092706e-05,
      "loss": 0.5931,
      "step": 145
    },
    {
      "epoch": 0.13993353157250307,
      "grad_norm": 0.45509182070489973,
      "learning_rate": 4.931643634594701e-05,
      "loss": 0.5764,
      "step": 150
    },
    {
      "epoch": 0.14459798262491982,
      "grad_norm": 0.41968982538759186,
      "learning_rate": 4.9228827354372374e-05,
      "loss": 0.5809,
      "step": 155
    },
    {
      "epoch": 0.1492624336773366,
      "grad_norm": 0.41229365733365814,
      "learning_rate": 4.9136034834142906e-05,
      "loss": 0.5886,
      "step": 160
    },
    {
      "epoch": 0.15392688472975335,
      "grad_norm": 0.3826932795909075,
      "learning_rate": 4.9038080921413485e-05,
      "loss": 0.5842,
      "step": 165
    },
    {
      "epoch": 0.15859133578217013,
      "grad_norm": 0.41134892045101074,
      "learning_rate": 4.893498898361693e-05,
      "loss": 0.5946,
      "step": 170
    },
    {
      "epoch": 0.1632557868345869,
      "grad_norm": 0.38992267173979334,
      "learning_rate": 4.882678361388958e-05,
      "loss": 0.5885,
      "step": 175
    },
    {
      "epoch": 0.16792023788700366,
      "grad_norm": 0.40073952235248766,
      "learning_rate": 4.871349062520448e-05,
      "loss": 0.5841,
      "step": 180
    },
    {
      "epoch": 0.17258468893942044,
      "grad_norm": 0.4200184545679776,
      "learning_rate": 4.859513704421353e-05,
      "loss": 0.5771,
      "step": 185
    },
    {
      "epoch": 0.17724913999183722,
      "grad_norm": 0.4292182704616412,
      "learning_rate": 4.847175110480015e-05,
      "loss": 0.5708,
      "step": 190
    },
    {
      "epoch": 0.18191359104425397,
      "grad_norm": 0.450661348267434,
      "learning_rate": 4.8343362241343944e-05,
      "loss": 0.5708,
      "step": 195
    },
    {
      "epoch": 0.18657804209667075,
      "grad_norm": 0.3869287419256418,
      "learning_rate": 4.8210001081698954e-05,
      "loss": 0.5708,
      "step": 200
    },
    {
      "epoch": 0.1912424931490875,
      "grad_norm": 0.42553333511533853,
      "learning_rate": 4.8071699439887215e-05,
      "loss": 0.5845,
      "step": 205
    },
    {
      "epoch": 0.19590694420150429,
      "grad_norm": 0.4146813048215204,
      "learning_rate": 4.792849030850938e-05,
      "loss": 0.5762,
      "step": 210
    },
    {
      "epoch": 0.20057139525392106,
      "grad_norm": 0.5810537942195573,
      "learning_rate": 4.778040785087412e-05,
      "loss": 0.5736,
      "step": 215
    },
    {
      "epoch": 0.20523584630633782,
      "grad_norm": 0.4540901088901895,
      "learning_rate": 4.762748739284831e-05,
      "loss": 0.5695,
      "step": 220
    },
    {
      "epoch": 0.2099002973587546,
      "grad_norm": 0.39778098326448963,
      "learning_rate": 4.746976541442986e-05,
      "loss": 0.562,
      "step": 225
    },
    {
      "epoch": 0.21456474841117135,
      "grad_norm": 0.45060914716273687,
      "learning_rate": 4.730727954104515e-05,
      "loss": 0.5708,
      "step": 230
    },
    {
      "epoch": 0.21922919946358813,
      "grad_norm": 0.40075705306410814,
      "learning_rate": 4.714006853457339e-05,
      "loss": 0.5779,
      "step": 235
    },
    {
      "epoch": 0.2238936505160049,
      "grad_norm": 0.3988838422959827,
      "learning_rate": 4.6968172284099654e-05,
      "loss": 0.5769,
      "step": 240
    },
    {
      "epoch": 0.22855810156842166,
      "grad_norm": 0.40730912286641724,
      "learning_rate": 4.679163179639923e-05,
      "loss": 0.5622,
      "step": 245
    },
    {
      "epoch": 0.23322255262083844,
      "grad_norm": 0.41570250249981516,
      "learning_rate": 4.661048918615513e-05,
      "loss": 0.5626,
      "step": 250
    },
    {
      "epoch": 0.23788700367325522,
      "grad_norm": 0.37266061168544923,
      "learning_rate": 4.642478766591148e-05,
      "loss": 0.5603,
      "step": 255
    },
    {
      "epoch": 0.24255145472567197,
      "grad_norm": 0.41204417027814066,
      "learning_rate": 4.623457153576493e-05,
      "loss": 0.5692,
      "step": 260
    },
    {
      "epoch": 0.24721590577808875,
      "grad_norm": 0.3517500791021223,
      "learning_rate": 4.603988617279655e-05,
      "loss": 0.5575,
      "step": 265
    },
    {
      "epoch": 0.2518803568305055,
      "grad_norm": 0.3626203299511348,
      "learning_rate": 4.5840778020247025e-05,
      "loss": 0.5561,
      "step": 270
    },
    {
      "epoch": 0.2565448078829223,
      "grad_norm": 0.4323865340964901,
      "learning_rate": 4.56372945764372e-05,
      "loss": 0.5605,
      "step": 275
    },
    {
      "epoch": 0.26120925893533903,
      "grad_norm": 0.4558769125407336,
      "learning_rate": 4.542948438343725e-05,
      "loss": 0.564,
      "step": 280
    },
    {
      "epoch": 0.26587370998775584,
      "grad_norm": 0.3788810919577419,
      "learning_rate": 4.521739701548662e-05,
      "loss": 0.5689,
      "step": 285
    },
    {
      "epoch": 0.2705381610401726,
      "grad_norm": 0.3494299182914199,
      "learning_rate": 4.500108306716784e-05,
      "loss": 0.5532,
      "step": 290
    },
    {
      "epoch": 0.27520261209258934,
      "grad_norm": 0.3243682374762204,
      "learning_rate": 4.478059414133695e-05,
      "loss": 0.5472,
      "step": 295
    },
    {
      "epoch": 0.27986706314500615,
      "grad_norm": 0.3774629263697534,
      "learning_rate": 4.455598283681331e-05,
      "loss": 0.5631,
      "step": 300
    },
    {
      "epoch": 0.2845315141974229,
      "grad_norm": 0.3590842486979046,
      "learning_rate": 4.43273027358319e-05,
      "loss": 0.5473,
      "step": 305
    },
    {
      "epoch": 0.28919596524983965,
      "grad_norm": 0.33869526982866255,
      "learning_rate": 4.4094608391260996e-05,
      "loss": 0.5556,
      "step": 310
    },
    {
      "epoch": 0.29386041630225646,
      "grad_norm": 0.3740834437948184,
      "learning_rate": 4.3857955313588256e-05,
      "loss": 0.5586,
      "step": 315
    },
    {
      "epoch": 0.2985248673546732,
      "grad_norm": 0.4069053631763161,
      "learning_rate": 4.3617399957678426e-05,
      "loss": 0.5699,
      "step": 320
    },
    {
      "epoch": 0.30318931840708996,
      "grad_norm": 0.42775595832918634,
      "learning_rate": 4.3372999709305726e-05,
      "loss": 0.5541,
      "step": 325
    },
    {
      "epoch": 0.3078537694595067,
      "grad_norm": 0.4225047260621452,
      "learning_rate": 4.3124812871464146e-05,
      "loss": 0.5612,
      "step": 330
    },
    {
      "epoch": 0.3125182205119235,
      "grad_norm": 0.4064532119864525,
      "learning_rate": 4.287289865045895e-05,
      "loss": 0.5566,
      "step": 335
    },
    {
      "epoch": 0.31718267156434027,
      "grad_norm": 0.40813784397443764,
      "learning_rate": 4.261731714178274e-05,
      "loss": 0.5427,
      "step": 340
    },
    {
      "epoch": 0.321847122616757,
      "grad_norm": 0.37505518830201817,
      "learning_rate": 4.23581293157793e-05,
      "loss": 0.5573,
      "step": 345
    },
    {
      "epoch": 0.3265115736691738,
      "grad_norm": 0.40656917887904453,
      "learning_rate": 4.20953970030988e-05,
      "loss": 0.5531,
      "step": 350
    },
    {
      "epoch": 0.3311760247215906,
      "grad_norm": 0.39600365804673093,
      "learning_rate": 4.182918287994781e-05,
      "loss": 0.5475,
      "step": 355
    },
    {
      "epoch": 0.33584047577400733,
      "grad_norm": 0.3703296359847103,
      "learning_rate": 4.155955045313748e-05,
      "loss": 0.5651,
      "step": 360
    },
    {
      "epoch": 0.34050492682642414,
      "grad_norm": 0.3666872032980013,
      "learning_rate": 4.128656404493371e-05,
      "loss": 0.559,
      "step": 365
    },
    {
      "epoch": 0.3451693778788409,
      "grad_norm": 0.33550827998787563,
      "learning_rate": 4.10102887777127e-05,
      "loss": 0.5609,
      "step": 370
    },
    {
      "epoch": 0.34983382893125764,
      "grad_norm": 0.3730387291098042,
      "learning_rate": 4.073079055842566e-05,
      "loss": 0.5485,
      "step": 375
    },
    {
      "epoch": 0.35449827998367445,
      "grad_norm": 0.3970093889085142,
      "learning_rate": 4.044813606287634e-05,
      "loss": 0.5475,
      "step": 380
    },
    {
      "epoch": 0.3591627310360912,
      "grad_norm": 0.32485469708579506,
      "learning_rate": 4.016239271981519e-05,
      "loss": 0.5461,
      "step": 385
    },
    {
      "epoch": 0.36382718208850795,
      "grad_norm": 0.3382837402395568,
      "learning_rate": 3.987362869485384e-05,
      "loss": 0.5496,
      "step": 390
    },
    {
      "epoch": 0.3684916331409247,
      "grad_norm": 0.3609893182037058,
      "learning_rate": 3.9581912874203854e-05,
      "loss": 0.5561,
      "step": 395
    },
    {
      "epoch": 0.3731560841933415,
      "grad_norm": 0.35168703173333654,
      "learning_rate": 3.928731484824351e-05,
      "loss": 0.5479,
      "step": 400
    },
    {
      "epoch": 0.37782053524575826,
      "grad_norm": 0.3629018488370036,
      "learning_rate": 3.898990489491668e-05,
      "loss": 0.5501,
      "step": 405
    },
    {
      "epoch": 0.382484986298175,
      "grad_norm": 0.37390402781019416,
      "learning_rate": 3.8689753962967636e-05,
      "loss": 0.5354,
      "step": 410
    },
    {
      "epoch": 0.3871494373505918,
      "grad_norm": 0.335380363177113,
      "learning_rate": 3.838693365501586e-05,
      "loss": 0.5469,
      "step": 415
    },
    {
      "epoch": 0.39181388840300857,
      "grad_norm": 0.3424832891358978,
      "learning_rate": 3.80815162104748e-05,
      "loss": 0.552,
      "step": 420
    },
    {
      "epoch": 0.3964783394554253,
      "grad_norm": 0.37221668230659216,
      "learning_rate": 3.7773574488318854e-05,
      "loss": 0.5553,
      "step": 425
    },
    {
      "epoch": 0.40114279050784213,
      "grad_norm": 0.37074728239538524,
      "learning_rate": 3.746318194970239e-05,
      "loss": 0.5523,
      "step": 430
    },
    {
      "epoch": 0.4058072415602589,
      "grad_norm": 0.38839870040744245,
      "learning_rate": 3.715041264043525e-05,
      "loss": 0.556,
      "step": 435
    },
    {
      "epoch": 0.41047169261267563,
      "grad_norm": 0.3575823882886992,
      "learning_rate": 3.683534117331869e-05,
      "loss": 0.5485,
      "step": 440
    },
    {
      "epoch": 0.41513614366509244,
      "grad_norm": 0.3161641377206891,
      "learning_rate": 3.65180427103461e-05,
      "loss": 0.5467,
      "step": 445
    },
    {
      "epoch": 0.4198005947175092,
      "grad_norm": 0.3703781300095459,
      "learning_rate": 3.619859294477273e-05,
      "loss": 0.5492,
      "step": 450
    },
    {
      "epoch": 0.42446504576992594,
      "grad_norm": 0.3492760299630012,
      "learning_rate": 3.587706808305861e-05,
      "loss": 0.5421,
      "step": 455
    },
    {
      "epoch": 0.4291294968223427,
      "grad_norm": 0.3267136941264168,
      "learning_rate": 3.5553544826689145e-05,
      "loss": 0.538,
      "step": 460
    },
    {
      "epoch": 0.4337939478747595,
      "grad_norm": 0.29422503748430795,
      "learning_rate": 3.522810035387752e-05,
      "loss": 0.5288,
      "step": 465
    },
    {
      "epoch": 0.43845839892717625,
      "grad_norm": 0.34737447664291465,
      "learning_rate": 3.490081230115343e-05,
      "loss": 0.5525,
      "step": 470
    },
    {
      "epoch": 0.443122849979593,
      "grad_norm": 0.3404046261795078,
      "learning_rate": 3.4571758744842507e-05,
      "loss": 0.5522,
      "step": 475
    },
    {
      "epoch": 0.4477873010320098,
      "grad_norm": 0.36447616317715154,
      "learning_rate": 3.4241018182440735e-05,
      "loss": 0.5441,
      "step": 480
    },
    {
      "epoch": 0.45245175208442656,
      "grad_norm": 0.31157047733964094,
      "learning_rate": 3.390866951388847e-05,
      "loss": 0.5391,
      "step": 485
    },
    {
      "epoch": 0.4571162031368433,
      "grad_norm": 0.3116582428424639,
      "learning_rate": 3.3574792022748466e-05,
      "loss": 0.5535,
      "step": 490
    },
    {
      "epoch": 0.4617806541892601,
      "grad_norm": 0.3279027907069977,
      "learning_rate": 3.3239465357292304e-05,
      "loss": 0.5376,
      "step": 495
    },
    {
      "epoch": 0.46644510524167687,
      "grad_norm": 0.3484015010403029,
      "learning_rate": 3.290276951149992e-05,
      "loss": 0.5336,
      "step": 500
    },
    {
      "epoch": 0.4711095562940936,
      "grad_norm": 0.33469705746842926,
      "learning_rate": 3.256478480597656e-05,
      "loss": 0.546,
      "step": 505
    },
    {
      "epoch": 0.47577400734651043,
      "grad_norm": 0.30392678850247695,
      "learning_rate": 3.222559186879191e-05,
      "loss": 0.5407,
      "step": 510
    },
    {
      "epoch": 0.4804384583989272,
      "grad_norm": 0.29705201467826636,
      "learning_rate": 3.18852716162458e-05,
      "loss": 0.5395,
      "step": 515
    },
    {
      "epoch": 0.48510290945134393,
      "grad_norm": 0.31309471816872475,
      "learning_rate": 3.154390523356523e-05,
      "loss": 0.5309,
      "step": 520
    },
    {
      "epoch": 0.48976736050376074,
      "grad_norm": 0.3532094930450357,
      "learning_rate": 3.1201574155537155e-05,
      "loss": 0.5341,
      "step": 525
    },
    {
      "epoch": 0.4944318115561775,
      "grad_norm": 0.3130231914781531,
      "learning_rate": 3.085836004708179e-05,
      "loss": 0.5243,
      "step": 530
    },
    {
      "epoch": 0.49909626260859424,
      "grad_norm": 0.32746390804605524,
      "learning_rate": 3.0514344783771015e-05,
      "loss": 0.5458,
      "step": 535
    },
    {
      "epoch": 0.503760713661011,
      "grad_norm": 0.3140774516982654,
      "learning_rate": 3.0169610432296513e-05,
      "loss": 0.542,
      "step": 540
    },
    {
      "epoch": 0.5084251647134278,
      "grad_norm": 0.33623222677088005,
      "learning_rate": 2.9824239230892316e-05,
      "loss": 0.5449,
      "step": 545
    },
    {
      "epoch": 0.5130896157658446,
      "grad_norm": 0.34034229987032566,
      "learning_rate": 2.9478313569716425e-05,
      "loss": 0.5399,
      "step": 550
    },
    {
      "epoch": 0.5177540668182613,
      "grad_norm": 0.31316724089217596,
      "learning_rate": 2.9131915971196216e-05,
      "loss": 0.538,
      "step": 555
    },
    {
      "epoch": 0.5224185178706781,
      "grad_norm": 0.3088718959081153,
      "learning_rate": 2.8785129070342247e-05,
      "loss": 0.5415,
      "step": 560
    },
    {
      "epoch": 0.5270829689230948,
      "grad_norm": 0.34309616279067395,
      "learning_rate": 2.8438035595035235e-05,
      "loss": 0.5323,
      "step": 565
    },
    {
      "epoch": 0.5317474199755117,
      "grad_norm": 0.3160243240149518,
      "learning_rate": 2.8090718346290902e-05,
      "loss": 0.5227,
      "step": 570
    },
    {
      "epoch": 0.5364118710279284,
      "grad_norm": 0.30318761428213864,
      "learning_rate": 2.77432601785073e-05,
      "loss": 0.538,
      "step": 575
    },
    {
      "epoch": 0.5410763220803452,
      "grad_norm": 0.33554969124573514,
      "learning_rate": 2.7395743979699527e-05,
      "loss": 0.5423,
      "step": 580
    },
    {
      "epoch": 0.5457407731327619,
      "grad_norm": 0.29358246176347985,
      "learning_rate": 2.7048252651726237e-05,
      "loss": 0.5302,
      "step": 585
    },
    {
      "epoch": 0.5504052241851787,
      "grad_norm": 0.27227963891831986,
      "learning_rate": 2.6700869090513025e-05,
      "loss": 0.5276,
      "step": 590
    },
    {
      "epoch": 0.5550696752375954,
      "grad_norm": 0.3315788476523275,
      "learning_rate": 2.6353676166277175e-05,
      "loss": 0.5375,
      "step": 595
    },
    {
      "epoch": 0.5597341262900123,
      "grad_norm": 0.29167775719810907,
      "learning_rate": 2.6006756703758462e-05,
      "loss": 0.5287,
      "step": 600
    },
    {
      "epoch": 0.564398577342429,
      "grad_norm": 0.2836903485895201,
      "learning_rate": 2.5660193462460914e-05,
      "loss": 0.5209,
      "step": 605
    },
    {
      "epoch": 0.5690630283948458,
      "grad_norm": 0.312902944542489,
      "learning_rate": 2.5314069116910073e-05,
      "loss": 0.5352,
      "step": 610
    },
    {
      "epoch": 0.5737274794472625,
      "grad_norm": 0.3061911987604843,
      "learning_rate": 2.496846623693052e-05,
      "loss": 0.5418,
      "step": 615
    },
    {
      "epoch": 0.5783919304996793,
      "grad_norm": 0.30918289084780787,
      "learning_rate": 2.4623467267948453e-05,
      "loss": 0.5264,
      "step": 620
    },
    {
      "epoch": 0.583056381552096,
      "grad_norm": 0.29432202760005055,
      "learning_rate": 2.427915451132382e-05,
      "loss": 0.5303,
      "step": 625
    },
    {
      "epoch": 0.5877208326045129,
      "grad_norm": 0.30248260299992763,
      "learning_rate": 2.3935610104716934e-05,
      "loss": 0.5318,
      "step": 630
    },
    {
      "epoch": 0.5923852836569297,
      "grad_norm": 0.2921703613190239,
      "learning_rate": 2.359291600249407e-05,
      "loss": 0.5351,
      "step": 635
    },
    {
      "epoch": 0.5970497347093464,
      "grad_norm": 0.2861561331468405,
      "learning_rate": 2.325115395617683e-05,
      "loss": 0.5191,
      "step": 640
    },
    {
      "epoch": 0.6017141857617632,
      "grad_norm": 0.29199262406988685,
      "learning_rate": 2.291040549493985e-05,
      "loss": 0.5314,
      "step": 645
    },
    {
      "epoch": 0.6063786368141799,
      "grad_norm": 0.28022740610521435,
      "learning_rate": 2.2570751906161624e-05,
      "loss": 0.5189,
      "step": 650
    },
    {
      "epoch": 0.6110430878665967,
      "grad_norm": 0.32229685451288487,
      "learning_rate": 2.223227421603289e-05,
      "loss": 0.5277,
      "step": 655
    },
    {
      "epoch": 0.6157075389190134,
      "grad_norm": 0.2812720351153964,
      "learning_rate": 2.1895053170227464e-05,
      "loss": 0.5315,
      "step": 660
    },
    {
      "epoch": 0.6203719899714303,
      "grad_norm": 0.2888622086540738,
      "learning_rate": 2.1559169214639884e-05,
      "loss": 0.5277,
      "step": 665
    },
    {
      "epoch": 0.625036441023847,
      "grad_norm": 0.3120048352568711,
      "learning_rate": 2.122470247619464e-05,
      "loss": 0.525,
      "step": 670
    },
    {
      "epoch": 0.6297008920762638,
      "grad_norm": 0.2818693243090786,
      "learning_rate": 2.0891732743731434e-05,
      "loss": 0.5322,
      "step": 675
    },
    {
      "epoch": 0.6343653431286805,
      "grad_norm": 0.2926059214892839,
      "learning_rate": 2.0560339448971146e-05,
      "loss": 0.525,
      "step": 680
    },
    {
      "epoch": 0.6390297941810973,
      "grad_norm": 0.29268503655464984,
      "learning_rate": 2.0230601647566966e-05,
      "loss": 0.5247,
      "step": 685
    },
    {
      "epoch": 0.643694245233514,
      "grad_norm": 0.2728644227058383,
      "learning_rate": 1.9902598000245222e-05,
      "loss": 0.5185,
      "step": 690
    },
    {
      "epoch": 0.6483586962859309,
      "grad_norm": 0.25962809295866146,
      "learning_rate": 1.9576406754040467e-05,
      "loss": 0.5228,
      "step": 695
    },
    {
      "epoch": 0.6530231473383477,
      "grad_norm": 0.2939794591562482,
      "learning_rate": 1.925210572362922e-05,
      "loss": 0.5402,
      "step": 700
    },
    {
      "epoch": 0.6576875983907644,
      "grad_norm": 0.2675031379318846,
      "learning_rate": 1.892977227276685e-05,
      "loss": 0.5224,
      "step": 705
    },
    {
      "epoch": 0.6623520494431812,
      "grad_norm": 0.33077413707192227,
      "learning_rate": 1.8609483295832036e-05,
      "loss": 0.5145,
      "step": 710
    },
    {
      "epoch": 0.6670165004955979,
      "grad_norm": 0.28316621823190463,
      "learning_rate": 1.829131519948323e-05,
      "loss": 0.518,
      "step": 715
    },
    {
      "epoch": 0.6716809515480147,
      "grad_norm": 0.28091675149417483,
      "learning_rate": 1.7975343884431357e-05,
      "loss": 0.5204,
      "step": 720
    },
    {
      "epoch": 0.6763454026004314,
      "grad_norm": 0.30963978843403667,
      "learning_rate": 1.7661644727333403e-05,
      "loss": 0.524,
      "step": 725
    },
    {
      "epoch": 0.6810098536528483,
      "grad_norm": 0.27818187692784135,
      "learning_rate": 1.7350292562810832e-05,
      "loss": 0.5202,
      "step": 730
    },
    {
      "epoch": 0.685674304705265,
      "grad_norm": 0.28024391562302353,
      "learning_rate": 1.704136166559737e-05,
      "loss": 0.5216,
      "step": 735
    },
    {
      "epoch": 0.6903387557576818,
      "grad_norm": 0.25776425960390403,
      "learning_rate": 1.6734925732820454e-05,
      "loss": 0.5201,
      "step": 740
    },
    {
      "epoch": 0.6950032068100985,
      "grad_norm": 0.25912854344445496,
      "learning_rate": 1.6431057866420313e-05,
      "loss": 0.5154,
      "step": 745
    },
    {
      "epoch": 0.6996676578625153,
      "grad_norm": 0.288843246635576,
      "learning_rate": 1.61298305557111e-05,
      "loss": 0.5319,
      "step": 750
    },
    {
      "epoch": 0.704332108914932,
      "grad_norm": 0.27843803013272117,
      "learning_rate": 1.583131566008825e-05,
      "loss": 0.5322,
      "step": 755
    },
    {
      "epoch": 0.7089965599673489,
      "grad_norm": 0.2724850636300469,
      "learning_rate": 1.553558439188594e-05,
      "loss": 0.5188,
      "step": 760
    },
    {
      "epoch": 0.7136610110197656,
      "grad_norm": 0.27405031166570964,
      "learning_rate": 1.5242707299389086e-05,
      "loss": 0.5264,
      "step": 765
    },
    {
      "epoch": 0.7183254620721824,
      "grad_norm": 0.28279666997673636,
      "learning_rate": 1.4952754250003637e-05,
      "loss": 0.5296,
      "step": 770
    },
    {
      "epoch": 0.7229899131245991,
      "grad_norm": 0.276837438024271,
      "learning_rate": 1.4665794413589298e-05,
      "loss": 0.5364,
      "step": 775
    },
    {
      "epoch": 0.7276543641770159,
      "grad_norm": 0.2724777598286785,
      "learning_rate": 1.4381896245958752e-05,
      "loss": 0.5158,
      "step": 780
    },
    {
      "epoch": 0.7323188152294327,
      "grad_norm": 0.2649273346430747,
      "learning_rate": 1.4101127472547084e-05,
      "loss": 0.519,
      "step": 785
    },
    {
      "epoch": 0.7369832662818494,
      "grad_norm": 0.2699794671741458,
      "learning_rate": 1.3823555072255606e-05,
      "loss": 0.5276,
      "step": 790
    },
    {
      "epoch": 0.7416477173342663,
      "grad_norm": 0.2597781223574146,
      "learning_rate": 1.354924526147357e-05,
      "loss": 0.5158,
      "step": 795
    },
    {
      "epoch": 0.746312168386683,
      "grad_norm": 0.28723301584203204,
      "learning_rate": 1.3278263478281994e-05,
      "loss": 0.5224,
      "step": 800
    },
    {
      "epoch": 0.7509766194390998,
      "grad_norm": 0.2535306464938602,
      "learning_rate": 1.3010674366843001e-05,
      "loss": 0.5361,
      "step": 805
    },
    {
      "epoch": 0.7556410704915165,
      "grad_norm": 0.27848610430564497,
      "learning_rate": 1.2746541761978592e-05,
      "loss": 0.5148,
      "step": 810
    },
    {
      "epoch": 0.7603055215439333,
      "grad_norm": 0.26508085752140736,
      "learning_rate": 1.2485928673942568e-05,
      "loss": 0.5197,
      "step": 815
    },
    {
      "epoch": 0.76496997259635,
      "grad_norm": 0.2765576189082103,
      "learning_rate": 1.2228897273389022e-05,
      "loss": 0.5043,
      "step": 820
    },
    {
      "epoch": 0.7696344236487669,
      "grad_norm": 0.25179624599466677,
      "learning_rate": 1.1975508876541262e-05,
      "loss": 0.5252,
      "step": 825
    },
    {
      "epoch": 0.7742988747011836,
      "grad_norm": 0.2597455495271655,
      "learning_rate": 1.1725823930564436e-05,
      "loss": 0.5292,
      "step": 830
    },
    {
      "epoch": 0.7789633257536004,
      "grad_norm": 0.2525298953468885,
      "learning_rate": 1.1479901999145583e-05,
      "loss": 0.5194,
      "step": 835
    },
    {
      "epoch": 0.7836277768060171,
      "grad_norm": 0.26668419176346814,
      "learning_rate": 1.1237801748284375e-05,
      "loss": 0.5183,
      "step": 840
    },
    {
      "epoch": 0.7882922278584339,
      "grad_norm": 0.2576707752493972,
      "learning_rate": 1.099958093229802e-05,
      "loss": 0.5145,
      "step": 845
    },
    {
      "epoch": 0.7929566789108506,
      "grad_norm": 0.2519900873666899,
      "learning_rate": 1.0765296380043684e-05,
      "loss": 0.52,
      "step": 850
    },
    {
      "epoch": 0.7976211299632674,
      "grad_norm": 0.25029147728099826,
      "learning_rate": 1.0535003981361613e-05,
      "loss": 0.5085,
      "step": 855
    },
    {
      "epoch": 0.8022855810156843,
      "grad_norm": 0.25974689320051614,
      "learning_rate": 1.030875867374238e-05,
      "loss": 0.5276,
      "step": 860
    },
    {
      "epoch": 0.806950032068101,
      "grad_norm": 0.2750183412160561,
      "learning_rate": 1.008661442922118e-05,
      "loss": 0.5288,
      "step": 865
    },
    {
      "epoch": 0.8116144831205178,
      "grad_norm": 0.24684508884873066,
      "learning_rate": 9.868624241502573e-06,
      "loss": 0.5108,
      "step": 870
    },
    {
      "epoch": 0.8162789341729345,
      "grad_norm": 0.24917506535926326,
      "learning_rate": 9.654840113318506e-06,
      "loss": 0.5153,
      "step": 875
    },
    {
      "epoch": 0.8209433852253513,
      "grad_norm": 0.26588122762475325,
      "learning_rate": 9.445313044022797e-06,
      "loss": 0.5134,
      "step": 880
    },
    {
      "epoch": 0.825607836277768,
      "grad_norm": 0.25205075163043134,
      "learning_rate": 9.240093017424978e-06,
      "loss": 0.5234,
      "step": 885
    },
    {
      "epoch": 0.8302722873301849,
      "grad_norm": 0.24453056572204862,
      "learning_rate": 9.039228989866358e-06,
      "loss": 0.5256,
      "step": 890
    },
    {
      "epoch": 0.8349367383826016,
      "grad_norm": 0.25964646851376255,
      "learning_rate": 8.84276887854126e-06,
      "loss": 0.5197,
      "step": 895
    },
    {
      "epoch": 0.8396011894350184,
      "grad_norm": 0.2380789074605979,
      "learning_rate": 8.650759550066084e-06,
      "loss": 0.5226,
      "step": 900
    },
    {
      "epoch": 0.8442656404874351,
      "grad_norm": 0.25537829061619166,
      "learning_rate": 8.46324680929905e-06,
      "loss": 0.5109,
      "step": 905
    },
    {
      "epoch": 0.8489300915398519,
      "grad_norm": 0.24326625950869546,
      "learning_rate": 8.280275388413186e-06,
      "loss": 0.5171,
      "step": 910
    },
    {
      "epoch": 0.8535945425922686,
      "grad_norm": 0.283956175304338,
      "learning_rate": 8.10188893622523e-06,
      "loss": 0.5138,
      "step": 915
    },
    {
      "epoch": 0.8582589936446854,
      "grad_norm": 0.26712492993115533,
      "learning_rate": 7.928130007782977e-06,
      "loss": 0.5347,
      "step": 920
    },
    {
      "epoch": 0.8629234446971022,
      "grad_norm": 0.2770179693009444,
      "learning_rate": 7.759040054213531e-06,
      "loss": 0.5332,
      "step": 925
    },
    {
      "epoch": 0.867587895749519,
      "grad_norm": 0.25471359044347464,
      "learning_rate": 7.59465941283494e-06,
      "loss": 0.5245,
      "step": 930
    },
    {
      "epoch": 0.8722523468019358,
      "grad_norm": 0.2474269162576819,
      "learning_rate": 7.435027297533474e-06,
      "loss": 0.5122,
      "step": 935
    },
    {
      "epoch": 0.8769167978543525,
      "grad_norm": 0.25130162397546496,
      "learning_rate": 7.2801817894089756e-06,
      "loss": 0.5063,
      "step": 940
    },
    {
      "epoch": 0.8815812489067693,
      "grad_norm": 0.24881956227908078,
      "learning_rate": 7.130159827690404e-06,
      "loss": 0.5177,
      "step": 945
    },
    {
      "epoch": 0.886245699959186,
      "grad_norm": 0.25545295509666294,
      "learning_rate": 6.98499720092375e-06,
      "loss": 0.508,
      "step": 950
    },
    {
      "epoch": 0.8909101510116029,
      "grad_norm": 0.24814116495940802,
      "learning_rate": 6.844728538434536e-06,
      "loss": 0.5156,
      "step": 955
    },
    {
      "epoch": 0.8955746020640196,
      "grad_norm": 0.24345586049308032,
      "learning_rate": 6.709387302066758e-06,
      "loss": 0.5192,
      "step": 960
    },
    {
      "epoch": 0.9002390531164364,
      "grad_norm": 0.25035521575621933,
      "learning_rate": 6.579005778200434e-06,
      "loss": 0.5253,
      "step": 965
    },
    {
      "epoch": 0.9049035041688531,
      "grad_norm": 0.2580394858408089,
      "learning_rate": 6.453615070049447e-06,
      "loss": 0.5251,
      "step": 970
    },
    {
      "epoch": 0.9095679552212699,
      "grad_norm": 0.23956216998636262,
      "learning_rate": 6.3332450902417666e-06,
      "loss": 0.5107,
      "step": 975
    },
    {
      "epoch": 0.9142324062736866,
      "grad_norm": 0.23764348877193034,
      "learning_rate": 6.21792455368361e-06,
      "loss": 0.5129,
      "step": 980
    },
    {
      "epoch": 0.9188968573261035,
      "grad_norm": 0.2520198832874535,
      "learning_rate": 6.1076809707093225e-06,
      "loss": 0.5208,
      "step": 985
    },
    {
      "epoch": 0.9235613083785202,
      "grad_norm": 0.23664963056881916,
      "learning_rate": 6.002540640518684e-06,
      "loss": 0.5066,
      "step": 990
    },
    {
      "epoch": 0.928225759430937,
      "grad_norm": 0.3088138680489082,
      "learning_rate": 5.902528644903055e-06,
      "loss": 0.5153,
      "step": 995
    },
    {
      "epoch": 0.9328902104833537,
      "grad_norm": 0.24722065847341207,
      "learning_rate": 5.807668842262004e-06,
      "loss": 0.5169,
      "step": 1000
    },
    {
      "epoch": 0.9375546615357705,
      "grad_norm": 0.2351634623111495,
      "learning_rate": 5.7179838619117394e-06,
      "loss": 0.5097,
      "step": 1005
    },
    {
      "epoch": 0.9422191125881872,
      "grad_norm": 0.2422992129098815,
      "learning_rate": 5.633495098686789e-06,
      "loss": 0.5119,
      "step": 1010
    },
    {
      "epoch": 0.946883563640604,
      "grad_norm": 0.2550904076289393,
      "learning_rate": 5.554222707836121e-06,
      "loss": 0.5147,
      "step": 1015
    },
    {
      "epoch": 0.9515480146930209,
      "grad_norm": 0.25885529407119634,
      "learning_rate": 5.480185600215012e-06,
      "loss": 0.5079,
      "step": 1020
    },
    {
      "epoch": 0.9562124657454376,
      "grad_norm": 0.23792385142576641,
      "learning_rate": 5.411401437773773e-06,
      "loss": 0.5143,
      "step": 1025
    },
    {
      "epoch": 0.9608769167978544,
      "grad_norm": 0.24312465281652115,
      "learning_rate": 5.347886629344369e-06,
      "loss": 0.5107,
      "step": 1030
    },
    {
      "epoch": 0.9655413678502711,
      "grad_norm": 0.23595847074173298,
      "learning_rate": 5.28965632672603e-06,
      "loss": 0.5187,
      "step": 1035
    },
    {
      "epoch": 0.9702058189026879,
      "grad_norm": 0.2479378393066615,
      "learning_rate": 5.236724421070693e-06,
      "loss": 0.5144,
      "step": 1040
    },
    {
      "epoch": 0.9748702699551046,
      "grad_norm": 0.24722093987308796,
      "learning_rate": 5.189103539569195e-06,
      "loss": 0.5215,
      "step": 1045
    },
    {
      "epoch": 0.9795347210075215,
      "grad_norm": 0.23704918121696106,
      "learning_rate": 5.146805042438997e-06,
      "loss": 0.516,
      "step": 1050
    },
    {
      "epoch": 0.9841991720599382,
      "grad_norm": 0.2499713377733996,
      "learning_rate": 5.10983902021413e-06,
      "loss": 0.5086,
      "step": 1055
    },
    {
      "epoch": 0.988863623112355,
      "grad_norm": 0.24126794921726105,
      "learning_rate": 5.078214291338054e-06,
      "loss": 0.5263,
      "step": 1060
    },
    {
      "epoch": 0.9935280741647717,
      "grad_norm": 0.2575581210389819,
      "learning_rate": 5.051938400059965e-06,
      "loss": 0.5173,
      "step": 1065
    },
    {
      "epoch": 0.9981925252171885,
      "grad_norm": 0.24967221419990557,
      "learning_rate": 5.031017614635082e-06,
      "loss": 0.5177,
      "step": 1070
    },
    {
      "epoch": 0.9991254154276719,
      "step": 1071,
      "total_flos": 486995857768448.0,
      "train_loss": 0.5541811234532507,
      "train_runtime": 156721.9484,
      "train_samples_per_second": 0.219,
      "train_steps_per_second": 0.007
    }
  ],
  "logging_steps": 5,
  "max_steps": 1071,
  "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": 486995857768448.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}