{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1044,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0009578544061302681,
      "grad_norm": 0.6597593171619324,
      "learning_rate": 9.523809523809525e-08,
      "loss": 1.1525,
      "step": 1
    },
    {
      "epoch": 0.004789272030651341,
      "grad_norm": 0.6607550915738474,
      "learning_rate": 4.7619047619047623e-07,
      "loss": 1.1632,
      "step": 5
    },
    {
      "epoch": 0.009578544061302681,
      "grad_norm": 0.6985874969645404,
      "learning_rate": 9.523809523809525e-07,
      "loss": 1.1753,
      "step": 10
    },
    {
      "epoch": 0.014367816091954023,
      "grad_norm": 0.6553768135986224,
      "learning_rate": 1.4285714285714286e-06,
      "loss": 1.1488,
      "step": 15
    },
    {
      "epoch": 0.019157088122605363,
      "grad_norm": 0.6087816631352209,
      "learning_rate": 1.904761904761905e-06,
      "loss": 1.1384,
      "step": 20
    },
    {
      "epoch": 0.023946360153256706,
      "grad_norm": 0.5225466109798246,
      "learning_rate": 2.380952380952381e-06,
      "loss": 1.122,
      "step": 25
    },
    {
      "epoch": 0.028735632183908046,
      "grad_norm": 0.5080052683286845,
      "learning_rate": 2.8571428571428573e-06,
      "loss": 1.1505,
      "step": 30
    },
    {
      "epoch": 0.033524904214559385,
      "grad_norm": 0.4341207873557241,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 1.1339,
      "step": 35
    },
    {
      "epoch": 0.038314176245210725,
      "grad_norm": 0.3985137882747194,
      "learning_rate": 3.80952380952381e-06,
      "loss": 1.1282,
      "step": 40
    },
    {
      "epoch": 0.04310344827586207,
      "grad_norm": 0.36938748261120696,
      "learning_rate": 4.2857142857142855e-06,
      "loss": 1.1141,
      "step": 45
    },
    {
      "epoch": 0.04789272030651341,
      "grad_norm": 0.31186688358132403,
      "learning_rate": 4.761904761904762e-06,
      "loss": 1.0914,
      "step": 50
    },
    {
      "epoch": 0.05268199233716475,
      "grad_norm": 0.3279958963323018,
      "learning_rate": 5.2380952380952384e-06,
      "loss": 1.1115,
      "step": 55
    },
    {
      "epoch": 0.05747126436781609,
      "grad_norm": 0.3107640896000187,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 1.0897,
      "step": 60
    },
    {
      "epoch": 0.06226053639846743,
      "grad_norm": 0.3071723661825106,
      "learning_rate": 6.1904761904761914e-06,
      "loss": 1.1045,
      "step": 65
    },
    {
      "epoch": 0.06704980842911877,
      "grad_norm": 0.2936092764163838,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.0661,
      "step": 70
    },
    {
      "epoch": 0.07183908045977011,
      "grad_norm": 0.28053978486957076,
      "learning_rate": 7.1428571428571436e-06,
      "loss": 1.0887,
      "step": 75
    },
    {
      "epoch": 0.07662835249042145,
      "grad_norm": 0.2891456438800518,
      "learning_rate": 7.61904761904762e-06,
      "loss": 1.0756,
      "step": 80
    },
    {
      "epoch": 0.08141762452107279,
      "grad_norm": 0.24950135731624706,
      "learning_rate": 8.095238095238097e-06,
      "loss": 1.0478,
      "step": 85
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 0.23670461068088686,
      "learning_rate": 8.571428571428571e-06,
      "loss": 1.0389,
      "step": 90
    },
    {
      "epoch": 0.09099616858237548,
      "grad_norm": 0.2507461508986584,
      "learning_rate": 9.047619047619049e-06,
      "loss": 1.0509,
      "step": 95
    },
    {
      "epoch": 0.09578544061302682,
      "grad_norm": 0.22930855800514297,
      "learning_rate": 9.523809523809525e-06,
      "loss": 1.0377,
      "step": 100
    },
    {
      "epoch": 0.10057471264367816,
      "grad_norm": 0.24445927836597864,
      "learning_rate": 1e-05,
      "loss": 1.0506,
      "step": 105
    },
    {
      "epoch": 0.1053639846743295,
      "grad_norm": 0.22388714842997454,
      "learning_rate": 9.999300418283908e-06,
      "loss": 1.0377,
      "step": 110
    },
    {
      "epoch": 0.11015325670498084,
      "grad_norm": 0.2406057364217601,
      "learning_rate": 9.997201868901463e-06,
      "loss": 1.0374,
      "step": 115
    },
    {
      "epoch": 0.11494252873563218,
      "grad_norm": 0.2275701466092952,
      "learning_rate": 9.993704939095376e-06,
      "loss": 1.0663,
      "step": 120
    },
    {
      "epoch": 0.11973180076628352,
      "grad_norm": 0.22200651495287974,
      "learning_rate": 9.988810607420912e-06,
      "loss": 1.0448,
      "step": 125
    },
    {
      "epoch": 0.12452107279693486,
      "grad_norm": 0.22051254365376602,
      "learning_rate": 9.982520243472044e-06,
      "loss": 1.0099,
      "step": 130
    },
    {
      "epoch": 0.12931034482758622,
      "grad_norm": 0.23279527845715264,
      "learning_rate": 9.974835607498224e-06,
      "loss": 1.0212,
      "step": 135
    },
    {
      "epoch": 0.13409961685823754,
      "grad_norm": 0.22813702069156522,
      "learning_rate": 9.965758849911774e-06,
      "loss": 1.023,
      "step": 140
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.23555996785615885,
      "learning_rate": 9.955292510686156e-06,
      "loss": 0.9997,
      "step": 145
    },
    {
      "epoch": 0.14367816091954022,
      "grad_norm": 0.2560040654609538,
      "learning_rate": 9.943439518645193e-06,
      "loss": 1.0172,
      "step": 150
    },
    {
      "epoch": 0.14846743295019157,
      "grad_norm": 0.24096780952464245,
      "learning_rate": 9.930203190643491e-06,
      "loss": 0.9876,
      "step": 155
    },
    {
      "epoch": 0.1532567049808429,
      "grad_norm": 0.24193989881098193,
      "learning_rate": 9.915587230638269e-06,
      "loss": 1.0417,
      "step": 160
    },
    {
      "epoch": 0.15804597701149425,
      "grad_norm": 0.24788380685714667,
      "learning_rate": 9.899595728652883e-06,
      "loss": 1.0332,
      "step": 165
    },
    {
      "epoch": 0.16283524904214558,
      "grad_norm": 0.2570877531631883,
      "learning_rate": 9.882233159632297e-06,
      "loss": 1.0129,
      "step": 170
    },
    {
      "epoch": 0.16762452107279693,
      "grad_norm": 0.26735645578353723,
      "learning_rate": 9.863504382190838e-06,
      "loss": 1.0255,
      "step": 175
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.24596191852493296,
      "learning_rate": 9.843414637252615e-06,
      "loss": 1.0125,
      "step": 180
    },
    {
      "epoch": 0.1772030651340996,
      "grad_norm": 0.2510254446995529,
      "learning_rate": 9.821969546584922e-06,
      "loss": 1.022,
      "step": 185
    },
    {
      "epoch": 0.18199233716475097,
      "grad_norm": 0.2626288820781652,
      "learning_rate": 9.79917511122509e-06,
      "loss": 1.0016,
      "step": 190
    },
    {
      "epoch": 0.1867816091954023,
      "grad_norm": 0.2588872211996587,
      "learning_rate": 9.775037709801206e-06,
      "loss": 1.0292,
      "step": 195
    },
    {
      "epoch": 0.19157088122605365,
      "grad_norm": 0.28272889728543066,
      "learning_rate": 9.749564096747148e-06,
      "loss": 1.0255,
      "step": 200
    },
    {
      "epoch": 0.19636015325670497,
      "grad_norm": 0.25691560926098406,
      "learning_rate": 9.722761400412496e-06,
      "loss": 1.0205,
      "step": 205
    },
    {
      "epoch": 0.20114942528735633,
      "grad_norm": 0.31206698507075104,
      "learning_rate": 9.694637121067764e-06,
      "loss": 1.0018,
      "step": 210
    },
    {
      "epoch": 0.20593869731800765,
      "grad_norm": 0.24725309050346184,
      "learning_rate": 9.6651991288056e-06,
      "loss": 1.013,
      "step": 215
    },
    {
      "epoch": 0.210727969348659,
      "grad_norm": 0.2530356632425841,
      "learning_rate": 9.63445566133846e-06,
      "loss": 0.9921,
      "step": 220
    },
    {
      "epoch": 0.21551724137931033,
      "grad_norm": 0.2784122469884546,
      "learning_rate": 9.602415321693434e-06,
      "loss": 1.0066,
      "step": 225
    },
    {
      "epoch": 0.22030651340996169,
      "grad_norm": 0.3098632552557301,
      "learning_rate": 9.569087075804842e-06,
      "loss": 1.0062,
      "step": 230
    },
    {
      "epoch": 0.22509578544061304,
      "grad_norm": 0.26451308039501314,
      "learning_rate": 9.534480250005263e-06,
      "loss": 0.9951,
      "step": 235
    },
    {
      "epoch": 0.22988505747126436,
      "grad_norm": 0.27301629134605937,
      "learning_rate": 9.498604528415731e-06,
      "loss": 1.0347,
      "step": 240
    },
    {
      "epoch": 0.23467432950191572,
      "grad_norm": 0.2891326880180606,
      "learning_rate": 9.461469950235795e-06,
      "loss": 1.0114,
      "step": 245
    },
    {
      "epoch": 0.23946360153256704,
      "grad_norm": 0.26574747916476,
      "learning_rate": 9.423086906934228e-06,
      "loss": 1.0272,
      "step": 250
    },
    {
      "epoch": 0.2442528735632184,
      "grad_norm": 0.26114773270931335,
      "learning_rate": 9.38346613934115e-06,
      "loss": 1.0039,
      "step": 255
    },
    {
      "epoch": 0.24904214559386972,
      "grad_norm": 0.28383740993290196,
      "learning_rate": 9.342618734642395e-06,
      "loss": 1.0077,
      "step": 260
    },
    {
      "epoch": 0.25383141762452105,
      "grad_norm": 0.28127484887890697,
      "learning_rate": 9.300556123276955e-06,
      "loss": 1.0306,
      "step": 265
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.2702807344360773,
      "learning_rate": 9.257290075738365e-06,
      "loss": 0.9924,
      "step": 270
    },
    {
      "epoch": 0.26340996168582376,
      "grad_norm": 0.2912366855364206,
      "learning_rate": 9.212832699280942e-06,
      "loss": 1.026,
      "step": 275
    },
    {
      "epoch": 0.2681992337164751,
      "grad_norm": 0.30617702200806085,
      "learning_rate": 9.16719643453177e-06,
      "loss": 1.0247,
      "step": 280
    },
    {
      "epoch": 0.27298850574712646,
      "grad_norm": 0.2590934020046758,
      "learning_rate": 9.120394052009412e-06,
      "loss": 1.0211,
      "step": 285
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.27434400868197933,
      "learning_rate": 9.072438648550304e-06,
      "loss": 1.0243,
      "step": 290
    },
    {
      "epoch": 0.2825670498084291,
      "grad_norm": 0.2813695079919259,
      "learning_rate": 9.023343643643821e-06,
      "loss": 1.0008,
      "step": 295
    },
    {
      "epoch": 0.28735632183908044,
      "grad_norm": 0.2764043745947579,
      "learning_rate": 8.973122775677078e-06,
      "loss": 1.0066,
      "step": 300
    },
    {
      "epoch": 0.2921455938697318,
      "grad_norm": 0.2993769190948342,
      "learning_rate": 8.921790098090477e-06,
      "loss": 1.015,
      "step": 305
    },
    {
      "epoch": 0.29693486590038315,
      "grad_norm": 0.28718834695385625,
      "learning_rate": 8.869359975445085e-06,
      "loss": 1.0212,
      "step": 310
    },
    {
      "epoch": 0.3017241379310345,
      "grad_norm": 0.3787907086651271,
      "learning_rate": 8.815847079402972e-06,
      "loss": 1.0079,
      "step": 315
    },
    {
      "epoch": 0.3065134099616858,
      "grad_norm": 0.3012426621088511,
      "learning_rate": 8.761266384621599e-06,
      "loss": 1.0245,
      "step": 320
    },
    {
      "epoch": 0.3113026819923372,
      "grad_norm": 0.545229044151294,
      "learning_rate": 8.705633164563413e-06,
      "loss": 1.0014,
      "step": 325
    },
    {
      "epoch": 0.3160919540229885,
      "grad_norm": 0.3018913245413703,
      "learning_rate": 8.648962987221837e-06,
      "loss": 1.0035,
      "step": 330
    },
    {
      "epoch": 0.32088122605363983,
      "grad_norm": 0.2937330943277994,
      "learning_rate": 8.591271710764839e-06,
      "loss": 0.9932,
      "step": 335
    },
    {
      "epoch": 0.32567049808429116,
      "grad_norm": 0.30415781861083496,
      "learning_rate": 8.532575479097294e-06,
      "loss": 0.982,
      "step": 340
    },
    {
      "epoch": 0.33045977011494254,
      "grad_norm": 0.2726446228184506,
      "learning_rate": 8.472890717343391e-06,
      "loss": 0.9992,
      "step": 345
    },
    {
      "epoch": 0.33524904214559387,
      "grad_norm": 0.2841460417699516,
      "learning_rate": 8.412234127250353e-06,
      "loss": 1.0007,
      "step": 350
    },
    {
      "epoch": 0.3400383141762452,
      "grad_norm": 0.313253280455998,
      "learning_rate": 8.350622682514735e-06,
      "loss": 0.9951,
      "step": 355
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.33772538052784323,
      "learning_rate": 8.288073624032634e-06,
      "loss": 1.0169,
      "step": 360
    },
    {
      "epoch": 0.3496168582375479,
      "grad_norm": 0.2810004404857808,
      "learning_rate": 8.224604455075115e-06,
      "loss": 1.0086,
      "step": 365
    },
    {
      "epoch": 0.3544061302681992,
      "grad_norm": 0.2817546952474833,
      "learning_rate": 8.160232936390239e-06,
      "loss": 0.9888,
      "step": 370
    },
    {
      "epoch": 0.35919540229885055,
      "grad_norm": 0.32316453097441,
      "learning_rate": 8.094977081233006e-06,
      "loss": 0.997,
      "step": 375
    },
    {
      "epoch": 0.36398467432950193,
      "grad_norm": 0.2920029553424864,
      "learning_rate": 8.02885515032467e-06,
      "loss": 1.0172,
      "step": 380
    },
    {
      "epoch": 0.36877394636015326,
      "grad_norm": 0.28736749227532504,
      "learning_rate": 7.961885646742793e-06,
      "loss": 1.0092,
      "step": 385
    },
    {
      "epoch": 0.3735632183908046,
      "grad_norm": 0.29552008976856375,
      "learning_rate": 7.894087310743468e-06,
      "loss": 0.9952,
      "step": 390
    },
    {
      "epoch": 0.3783524904214559,
      "grad_norm": 0.28037261402408636,
      "learning_rate": 7.825479114517197e-06,
      "loss": 1.0148,
      "step": 395
    },
    {
      "epoch": 0.3831417624521073,
      "grad_norm": 0.3259544387907705,
      "learning_rate": 7.756080256879837e-06,
      "loss": 1.0172,
      "step": 400
    },
    {
      "epoch": 0.3879310344827586,
      "grad_norm": 0.2866306787997861,
      "learning_rate": 7.685910157900158e-06,
      "loss": 0.9969,
      "step": 405
    },
    {
      "epoch": 0.39272030651340994,
      "grad_norm": 0.27836252924957877,
      "learning_rate": 7.614988453465469e-06,
      "loss": 0.9981,
      "step": 410
    },
    {
      "epoch": 0.3975095785440613,
      "grad_norm": 0.2881265537652179,
      "learning_rate": 7.5433349897868445e-06,
      "loss": 1.0075,
      "step": 415
    },
    {
      "epoch": 0.40229885057471265,
      "grad_norm": 0.286210264814618,
      "learning_rate": 7.470969817845518e-06,
      "loss": 1.0025,
      "step": 420
    },
    {
      "epoch": 0.407088122605364,
      "grad_norm": 0.28784515054514276,
      "learning_rate": 7.397913187781962e-06,
      "loss": 0.9918,
      "step": 425
    },
    {
      "epoch": 0.4118773946360153,
      "grad_norm": 0.28225636678596183,
      "learning_rate": 7.324185543229226e-06,
      "loss": 1.0164,
      "step": 430
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.30273570716493303,
      "learning_rate": 7.249807515592149e-06,
      "loss": 0.991,
      "step": 435
    },
    {
      "epoch": 0.421455938697318,
      "grad_norm": 0.29638416230646264,
      "learning_rate": 7.174799918274018e-06,
      "loss": 1.0103,
      "step": 440
    },
    {
      "epoch": 0.42624521072796934,
      "grad_norm": 0.27796285045314706,
      "learning_rate": 7.099183740852296e-06,
      "loss": 0.9929,
      "step": 445
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 0.3024800017009751,
      "learning_rate": 7.022980143205046e-06,
      "loss": 0.9945,
      "step": 450
    },
    {
      "epoch": 0.43582375478927204,
      "grad_norm": 0.308729839004599,
      "learning_rate": 6.946210449589714e-06,
      "loss": 1.0131,
      "step": 455
    },
    {
      "epoch": 0.44061302681992337,
      "grad_norm": 0.29787252415350113,
      "learning_rate": 6.868896142675903e-06,
      "loss": 1.0053,
      "step": 460
    },
    {
      "epoch": 0.4454022988505747,
      "grad_norm": 0.28555796101935293,
      "learning_rate": 6.791058857533814e-06,
      "loss": 1.0106,
      "step": 465
    },
    {
      "epoch": 0.4501915708812261,
      "grad_norm": 0.27754653040108523,
      "learning_rate": 6.712720375580057e-06,
      "loss": 1.0127,
      "step": 470
    },
    {
      "epoch": 0.4549808429118774,
      "grad_norm": 0.3064883256256046,
      "learning_rate": 6.633902618482484e-06,
      "loss": 1.0137,
      "step": 475
    },
    {
      "epoch": 0.45977011494252873,
      "grad_norm": 0.2871757827962217,
      "learning_rate": 6.554627642025807e-06,
      "loss": 0.9808,
      "step": 480
    },
    {
      "epoch": 0.46455938697318006,
      "grad_norm": 0.32016740057261645,
      "learning_rate": 6.474917629939652e-06,
      "loss": 1.0154,
      "step": 485
    },
    {
      "epoch": 0.46934865900383144,
      "grad_norm": 0.28945157272232624,
      "learning_rate": 6.394794887690838e-06,
      "loss": 0.987,
      "step": 490
    },
    {
      "epoch": 0.47413793103448276,
      "grad_norm": 0.45234637623400176,
      "learning_rate": 6.314281836241573e-06,
      "loss": 1.0072,
      "step": 495
    },
    {
      "epoch": 0.4789272030651341,
      "grad_norm": 0.3043561526142259,
      "learning_rate": 6.233401005775339e-06,
      "loss": 0.9947,
      "step": 500
    },
    {
      "epoch": 0.4837164750957854,
      "grad_norm": 0.3021339898048032,
      "learning_rate": 6.1521750293922035e-06,
      "loss": 1.0168,
      "step": 505
    },
    {
      "epoch": 0.4885057471264368,
      "grad_norm": 0.262840671697266,
      "learning_rate": 6.070626636775349e-06,
      "loss": 0.9854,
      "step": 510
    },
    {
      "epoch": 0.4932950191570881,
      "grad_norm": 0.2717384304425024,
      "learning_rate": 5.988778647830554e-06,
      "loss": 0.9847,
      "step": 515
    },
    {
      "epoch": 0.49808429118773945,
      "grad_norm": 0.3005959006799932,
      "learning_rate": 5.906653966300444e-06,
      "loss": 1.0007,
      "step": 520
    },
    {
      "epoch": 0.5028735632183908,
      "grad_norm": 0.2810709946318669,
      "learning_rate": 5.824275573355278e-06,
      "loss": 0.9891,
      "step": 525
    },
    {
      "epoch": 0.5076628352490421,
      "grad_norm": 0.31324328313914135,
      "learning_rate": 5.741666521162055e-06,
      "loss": 1.0049,
      "step": 530
    },
    {
      "epoch": 0.5124521072796935,
      "grad_norm": 0.29241234641844166,
      "learning_rate": 5.658849926433774e-06,
      "loss": 1.0019,
      "step": 535
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.29226941844525284,
      "learning_rate": 5.575848963960621e-06,
      "loss": 0.9964,
      "step": 540
    },
    {
      "epoch": 0.5220306513409961,
      "grad_norm": 0.2869562590824223,
      "learning_rate": 5.4926868601249e-06,
      "loss": 1.003,
      "step": 545
    },
    {
      "epoch": 0.5268199233716475,
      "grad_norm": 0.2887086296052449,
      "learning_rate": 5.4093868864015405e-06,
      "loss": 0.9911,
      "step": 550
    },
    {
      "epoch": 0.5316091954022989,
      "grad_norm": 0.2990975733324071,
      "learning_rate": 5.325972352845965e-06,
      "loss": 0.9961,
      "step": 555
    },
    {
      "epoch": 0.5363984674329502,
      "grad_norm": 0.2964344269886325,
      "learning_rate": 5.24246660157119e-06,
      "loss": 1.0045,
      "step": 560
    },
    {
      "epoch": 0.5411877394636015,
      "grad_norm": 0.3080697315686108,
      "learning_rate": 5.1588930002159255e-06,
      "loss": 0.9897,
      "step": 565
    },
    {
      "epoch": 0.5459770114942529,
      "grad_norm": 0.3217970889035437,
      "learning_rate": 5.075274935405554e-06,
      "loss": 1.0022,
      "step": 570
    },
    {
      "epoch": 0.5507662835249042,
      "grad_norm": 0.32311237689799727,
      "learning_rate": 4.991635806207788e-06,
      "loss": 0.9918,
      "step": 575
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 0.3076150767243147,
      "learning_rate": 4.90799901758484e-06,
      "loss": 1.0156,
      "step": 580
    },
    {
      "epoch": 0.5603448275862069,
      "grad_norm": 0.29579117268352634,
      "learning_rate": 4.824387973843957e-06,
      "loss": 0.9859,
      "step": 585
    },
    {
      "epoch": 0.5651340996168582,
      "grad_norm": 0.27808702972070926,
      "learning_rate": 4.74082607208812e-06,
      "loss": 0.988,
      "step": 590
    },
    {
      "epoch": 0.5699233716475096,
      "grad_norm": 0.2772865889207572,
      "learning_rate": 4.6573366956687885e-06,
      "loss": 1.0042,
      "step": 595
    },
    {
      "epoch": 0.5747126436781609,
      "grad_norm": 0.2741988412251713,
      "learning_rate": 4.573943207642452e-06,
      "loss": 1.018,
      "step": 600
    },
    {
      "epoch": 0.5795019157088123,
      "grad_norm": 0.3041843000888798,
      "learning_rate": 4.4906689442328935e-06,
      "loss": 1.0095,
      "step": 605
    },
    {
      "epoch": 0.5842911877394636,
      "grad_norm": 0.3234627396285633,
      "learning_rate": 4.407537208300957e-06,
      "loss": 0.9981,
      "step": 610
    },
    {
      "epoch": 0.5890804597701149,
      "grad_norm": 0.30361837699199173,
      "learning_rate": 4.3245712628236356e-06,
      "loss": 0.9945,
      "step": 615
    },
    {
      "epoch": 0.5938697318007663,
      "grad_norm": 0.3068819956047309,
      "learning_rate": 4.241794324384334e-06,
      "loss": 0.9829,
      "step": 620
    },
    {
      "epoch": 0.5986590038314177,
      "grad_norm": 0.3010857723276443,
      "learning_rate": 4.159229556676111e-06,
      "loss": 0.9778,
      "step": 625
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 0.3212249651416007,
      "learning_rate": 4.076900064019721e-06,
      "loss": 1.007,
      "step": 630
    },
    {
      "epoch": 0.6082375478927203,
      "grad_norm": 0.2842660737481042,
      "learning_rate": 3.994828884898267e-06,
      "loss": 1.0056,
      "step": 635
    },
    {
      "epoch": 0.6130268199233716,
      "grad_norm": 0.2887728882458368,
      "learning_rate": 3.91303898551028e-06,
      "loss": 1.0131,
      "step": 640
    },
    {
      "epoch": 0.617816091954023,
      "grad_norm": 0.31374850764975326,
      "learning_rate": 3.8315532533430285e-06,
      "loss": 0.9979,
      "step": 645
    },
    {
      "epoch": 0.6226053639846744,
      "grad_norm": 0.2990812859204113,
      "learning_rate": 3.7503944907678543e-06,
      "loss": 0.9979,
      "step": 650
    },
    {
      "epoch": 0.6273946360153256,
      "grad_norm": 0.31174935012060845,
      "learning_rate": 3.6695854086593126e-06,
      "loss": 0.9907,
      "step": 655
    },
    {
      "epoch": 0.632183908045977,
      "grad_norm": 0.30150077718156976,
      "learning_rate": 3.5891486200399413e-06,
      "loss": 0.9937,
      "step": 660
    },
    {
      "epoch": 0.6369731800766284,
      "grad_norm": 0.304753149818871,
      "learning_rate": 3.509106633752387e-06,
      "loss": 1.0164,
      "step": 665
    },
    {
      "epoch": 0.6417624521072797,
      "grad_norm": 0.2993950641947178,
      "learning_rate": 3.429481848160702e-06,
      "loss": 1.0093,
      "step": 670
    },
    {
      "epoch": 0.646551724137931,
      "grad_norm": 0.3102107207132909,
      "learning_rate": 3.350296544882543e-06,
      "loss": 0.969,
      "step": 675
    },
    {
      "epoch": 0.6513409961685823,
      "grad_norm": 0.27647028696280174,
      "learning_rate": 3.2715728825540525e-06,
      "loss": 1.0102,
      "step": 680
    },
    {
      "epoch": 0.6561302681992337,
      "grad_norm": 0.30329159137172645,
      "learning_rate": 3.19333289062915e-06,
      "loss": 0.9992,
      "step": 685
    },
    {
      "epoch": 0.6609195402298851,
      "grad_norm": 0.3233989260192753,
      "learning_rate": 3.1155984632149565e-06,
      "loss": 0.9984,
      "step": 690
    },
    {
      "epoch": 0.6657088122605364,
      "grad_norm": 0.284592333087901,
      "learning_rate": 3.0383913529451286e-06,
      "loss": 1.0097,
      "step": 695
    },
    {
      "epoch": 0.6704980842911877,
      "grad_norm": 0.284014773245373,
      "learning_rate": 2.961733164892744e-06,
      "loss": 1.0048,
      "step": 700
    },
    {
      "epoch": 0.6752873563218391,
      "grad_norm": 0.29864923395210635,
      "learning_rate": 2.8856453505245018e-06,
      "loss": 1.008,
      "step": 705
    },
    {
      "epoch": 0.6800766283524904,
      "grad_norm": 0.28455877244795186,
      "learning_rate": 2.8101492016979027e-06,
      "loss": 1.0082,
      "step": 710
    },
    {
      "epoch": 0.6848659003831418,
      "grad_norm": 0.29297293441472344,
      "learning_rate": 2.7352658447030882e-06,
      "loss": 1.0137,
      "step": 715
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.29774107321754356,
      "learning_rate": 2.6610162343510183e-06,
      "loss": 0.9878,
      "step": 720
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.2886426546973218,
      "learning_rate": 2.587421148109619e-06,
      "loss": 0.9855,
      "step": 725
    },
    {
      "epoch": 0.6992337164750958,
      "grad_norm": 0.30340545161406907,
      "learning_rate": 2.5145011802895835e-06,
      "loss": 1.004,
      "step": 730
    },
    {
      "epoch": 0.7040229885057471,
      "grad_norm": 0.28629744556011416,
      "learning_rate": 2.4422767362814045e-06,
      "loss": 0.9935,
      "step": 735
    },
    {
      "epoch": 0.7088122605363985,
      "grad_norm": 0.29480150488982737,
      "learning_rate": 2.370768026845276e-06,
      "loss": 1.0013,
      "step": 740
    },
    {
      "epoch": 0.7136015325670498,
      "grad_norm": 0.30424158130358797,
      "learning_rate": 2.299995062455459e-06,
      "loss": 0.9932,
      "step": 745
    },
    {
      "epoch": 0.7183908045977011,
      "grad_norm": 0.3179663498265742,
      "learning_rate": 2.2299776477007073e-06,
      "loss": 1.007,
      "step": 750
    },
    {
      "epoch": 0.7231800766283525,
      "grad_norm": 0.303515618658323,
      "learning_rate": 2.16073537574229e-06,
      "loss": 0.9963,
      "step": 755
    },
    {
      "epoch": 0.7279693486590039,
      "grad_norm": 0.30307569076630475,
      "learning_rate": 2.0922876228311833e-06,
      "loss": 0.9772,
      "step": 760
    },
    {
      "epoch": 0.7327586206896551,
      "grad_norm": 0.3143565291039063,
      "learning_rate": 2.0246535428859652e-06,
      "loss": 0.9899,
      "step": 765
    },
    {
      "epoch": 0.7375478927203065,
      "grad_norm": 0.28738666111079514,
      "learning_rate": 1.957852062132924e-06,
      "loss": 0.9848,
      "step": 770
    },
    {
      "epoch": 0.7423371647509579,
      "grad_norm": 0.2850271965375348,
      "learning_rate": 1.8919018738098704e-06,
      "loss": 1.0076,
      "step": 775
    },
    {
      "epoch": 0.7471264367816092,
      "grad_norm": 0.29940367717031596,
      "learning_rate": 1.8268214329351797e-06,
      "loss": 0.9864,
      "step": 780
    },
    {
      "epoch": 0.7519157088122606,
      "grad_norm": 0.2957695153675884,
      "learning_rate": 1.762628951143454e-06,
      "loss": 0.9972,
      "step": 785
    },
    {
      "epoch": 0.7567049808429118,
      "grad_norm": 0.2884019240313734,
      "learning_rate": 1.6993423915893241e-06,
      "loss": 0.9969,
      "step": 790
    },
    {
      "epoch": 0.7614942528735632,
      "grad_norm": 0.30882836660635715,
      "learning_rate": 1.6369794639207626e-06,
      "loss": 1.0005,
      "step": 795
    },
    {
      "epoch": 0.7662835249042146,
      "grad_norm": 0.2880996272538283,
      "learning_rate": 1.575557619323353e-06,
      "loss": 0.9853,
      "step": 800
    },
    {
      "epoch": 0.7710727969348659,
      "grad_norm": 0.28083999464104503,
      "learning_rate": 1.5150940456368784e-06,
      "loss": 0.9579,
      "step": 805
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.2977795145319976,
      "learning_rate": 1.4556056625455922e-06,
      "loss": 0.9944,
      "step": 810
    },
    {
      "epoch": 0.7806513409961686,
      "grad_norm": 0.3044689990672244,
      "learning_rate": 1.3971091168435463e-06,
      "loss": 0.997,
      "step": 815
    },
    {
      "epoch": 0.7854406130268199,
      "grad_norm": 0.2951506289424605,
      "learning_rate": 1.3396207777762732e-06,
      "loss": 1.0116,
      "step": 820
    },
    {
      "epoch": 0.7902298850574713,
      "grad_norm": 0.284297818009739,
      "learning_rate": 1.2831567324601325e-06,
      "loss": 0.9792,
      "step": 825
    },
    {
      "epoch": 0.7950191570881227,
      "grad_norm": 0.31450502170794314,
      "learning_rate": 1.2277327813806123e-06,
      "loss": 0.9927,
      "step": 830
    },
    {
      "epoch": 0.7998084291187739,
      "grad_norm": 0.2896076523698672,
      "learning_rate": 1.173364433970835e-06,
      "loss": 0.9795,
      "step": 835
    },
    {
      "epoch": 0.8045977011494253,
      "grad_norm": 0.2834770010415917,
      "learning_rate": 1.1200669042715163e-06,
      "loss": 0.9966,
      "step": 840
    },
    {
      "epoch": 0.8093869731800766,
      "grad_norm": 0.34009482243032485,
      "learning_rate": 1.0678551066735671e-06,
      "loss": 0.9767,
      "step": 845
    },
    {
      "epoch": 0.814176245210728,
      "grad_norm": 0.29936688994813515,
      "learning_rate": 1.0167436517445777e-06,
      "loss": 1.003,
      "step": 850
    },
    {
      "epoch": 0.8189655172413793,
      "grad_norm": 0.29389391810900556,
      "learning_rate": 9.66746842140287e-07,
      "loss": 0.9888,
      "step": 855
    },
    {
      "epoch": 0.8237547892720306,
      "grad_norm": 0.29840441886793445,
      "learning_rate": 9.178786686022417e-07,
      "loss": 1.0011,
      "step": 860
    },
    {
      "epoch": 0.828544061302682,
      "grad_norm": 0.3050673773124508,
      "learning_rate": 8.701528060427194e-07,
      "loss": 0.9867,
      "step": 865
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 0.2809885669064277,
      "learning_rate": 8.235826097180566e-07,
      "loss": 0.9802,
      "step": 870
    },
    {
      "epoch": 0.8381226053639846,
      "grad_norm": 0.29288746432276136,
      "learning_rate": 7.781811114913995e-07,
      "loss": 0.9965,
      "step": 875
    },
    {
      "epoch": 0.842911877394636,
      "grad_norm": 0.29110203494522685,
      "learning_rate": 7.339610161859618e-07,
      "loss": 0.9809,
      "step": 880
    },
    {
      "epoch": 0.8477011494252874,
      "grad_norm": 0.30532295542101273,
      "learning_rate": 6.909346980298093e-07,
      "loss": 1.0039,
      "step": 885
    },
    {
      "epoch": 0.8524904214559387,
      "grad_norm": 0.3178843529727934,
      "learning_rate": 6.49114197193137e-07,
      "loss": 0.9992,
      "step": 890
    },
    {
      "epoch": 0.85727969348659,
      "grad_norm": 0.30030807864509074,
      "learning_rate": 6.085112164190466e-07,
      "loss": 0.9967,
      "step": 895
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.283276426877559,
      "learning_rate": 5.691371177487215e-07,
      "loss": 0.9951,
      "step": 900
    },
    {
      "epoch": 0.8668582375478927,
      "grad_norm": 0.2771047728402987,
      "learning_rate": 5.310029193419697e-07,
      "loss": 0.9823,
      "step": 905
    },
    {
      "epoch": 0.8716475095785441,
      "grad_norm": 0.30200927801260424,
      "learning_rate": 4.941192923939769e-07,
      "loss": 0.9944,
      "step": 910
    },
    {
      "epoch": 0.8764367816091954,
      "grad_norm": 0.29243200143497705,
      "learning_rate": 4.5849655814915683e-07,
      "loss": 0.9923,
      "step": 915
    },
    {
      "epoch": 0.8812260536398467,
      "grad_norm": 0.38073762499381975,
      "learning_rate": 4.2414468501293217e-07,
      "loss": 0.9931,
      "step": 920
    },
    {
      "epoch": 0.8860153256704981,
      "grad_norm": 0.28457589633356245,
      "learning_rate": 3.9107328576224736e-07,
      "loss": 0.9879,
      "step": 925
    },
    {
      "epoch": 0.8908045977011494,
      "grad_norm": 0.29913915061920765,
      "learning_rate": 3.5929161485559694e-07,
      "loss": 1.0269,
      "step": 930
    },
    {
      "epoch": 0.8955938697318008,
      "grad_norm": 0.28795408054019833,
      "learning_rate": 3.2880856584333043e-07,
      "loss": 0.984,
      "step": 935
    },
    {
      "epoch": 0.9003831417624522,
      "grad_norm": 0.28456219117461123,
      "learning_rate": 2.9963266887894526e-07,
      "loss": 1.0007,
      "step": 940
    },
    {
      "epoch": 0.9051724137931034,
      "grad_norm": 0.30655155945282825,
      "learning_rate": 2.717720883320685e-07,
      "loss": 1.0093,
      "step": 945
    },
    {
      "epoch": 0.9099616858237548,
      "grad_norm": 0.28653578966574017,
      "learning_rate": 2.4523462050379864e-07,
      "loss": 0.9861,
      "step": 950
    },
    {
      "epoch": 0.9147509578544061,
      "grad_norm": 0.30075605446410747,
      "learning_rate": 2.2002769144504943e-07,
      "loss": 0.997,
      "step": 955
    },
    {
      "epoch": 0.9195402298850575,
      "grad_norm": 0.28984184857418177,
      "learning_rate": 1.9615835487849677e-07,
      "loss": 0.9772,
      "step": 960
    },
    {
      "epoch": 0.9243295019157088,
      "grad_norm": 0.2971882152661699,
      "learning_rate": 1.7363329022471564e-07,
      "loss": 1.0125,
      "step": 965
    },
    {
      "epoch": 0.9291187739463601,
      "grad_norm": 0.30026576782404996,
      "learning_rate": 1.5245880073305963e-07,
      "loss": 1.0128,
      "step": 970
    },
    {
      "epoch": 0.9339080459770115,
      "grad_norm": 0.2971224564720333,
      "learning_rate": 1.3264081171780797e-07,
      "loss": 1.0114,
      "step": 975
    },
    {
      "epoch": 0.9386973180076629,
      "grad_norm": 0.278422035094591,
      "learning_rate": 1.1418486890006574e-07,
      "loss": 0.982,
      "step": 980
    },
    {
      "epoch": 0.9434865900383141,
      "grad_norm": 0.29059990973676236,
      "learning_rate": 9.709613685589314e-08,
      "loss": 0.998,
      "step": 985
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 0.29092073403074437,
      "learning_rate": 8.137939757108526e-08,
      "loss": 1.011,
      "step": 990
    },
    {
      "epoch": 0.9530651340996169,
      "grad_norm": 0.2923297366913393,
      "learning_rate": 6.703904910301929e-08,
      "loss": 0.9656,
      "step": 995
    },
    {
      "epoch": 0.9578544061302682,
      "grad_norm": 0.2931478943843858,
      "learning_rate": 5.4079104349929465e-08,
      "loss": 1.0036,
      "step": 1000
    },
    {
      "epoch": 0.9626436781609196,
      "grad_norm": 0.2923376776086664,
      "learning_rate": 4.250318992797375e-08,
      "loss": 1.0083,
      "step": 1005
    },
    {
      "epoch": 0.9674329501915708,
      "grad_norm": 0.2863912260830555,
      "learning_rate": 3.231454515638221e-08,
      "loss": 0.9955,
      "step": 1010
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.2974177186982834,
      "learning_rate": 2.351602115099272e-08,
      "loss": 0.9865,
      "step": 1015
    },
    {
      "epoch": 0.9770114942528736,
      "grad_norm": 0.29561321400349233,
      "learning_rate": 1.6110080026414123e-08,
      "loss": 1.0083,
      "step": 1020
    },
    {
      "epoch": 0.9818007662835249,
      "grad_norm": 0.29932685669861786,
      "learning_rate": 1.0098794207047402e-08,
      "loss": 1.0118,
      "step": 1025
    },
    {
      "epoch": 0.9865900383141762,
      "grad_norm": 0.29155813115628504,
      "learning_rate": 5.483845847151226e-09,
      "loss": 0.9846,
      "step": 1030
    },
    {
      "epoch": 0.9913793103448276,
      "grad_norm": 0.2910040393119768,
      "learning_rate": 2.2665263601240328e-09,
      "loss": 0.9812,
      "step": 1035
    },
    {
      "epoch": 0.9961685823754789,
      "grad_norm": 0.28919827046140006,
      "learning_rate": 4.4773605712089554e-10,
      "loss": 1.0115,
      "step": 1040
    },
    {
      "epoch": 1.0,
      "eval_runtime": 6595.9355,
      "eval_samples_per_second": 3.504,
      "eval_steps_per_second": 0.876,
      "step": 1044
    },
    {
      "epoch": 1.0,
      "step": 1044,
      "total_flos": 1940427569627136.0,
      "train_loss": 1.0122476654034465,
      "train_runtime": 20247.7008,
      "train_samples_per_second": 3.298,
      "train_steps_per_second": 0.052
    }
  ],
  "logging_steps": 5,
  "max_steps": 1044,
  "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": 1940427569627136.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}