{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1073,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0046641791044776115,
      "grad_norm": 1.9791202613768635,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.8462,
      "step": 5
    },
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.4305500674160407,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8236,
      "step": 10
    },
    {
      "epoch": 0.013992537313432836,
      "grad_norm": 1.1206530311176968,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.7682,
      "step": 15
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 0.6720846360747776,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.7273,
      "step": 20
    },
    {
      "epoch": 0.02332089552238806,
      "grad_norm": 0.5048504624958363,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.6963,
      "step": 25
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.459675485239456,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6737,
      "step": 30
    },
    {
      "epoch": 0.03264925373134328,
      "grad_norm": 0.4155808328686281,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.6448,
      "step": 35
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.42045736753620744,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6378,
      "step": 40
    },
    {
      "epoch": 0.04197761194029851,
      "grad_norm": 0.40302027504447946,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6333,
      "step": 45
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.3893242949214179,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.6148,
      "step": 50
    },
    {
      "epoch": 0.051305970149253734,
      "grad_norm": 0.399280608624726,
      "learning_rate": 4.999989285883431e-05,
      "loss": 0.6341,
      "step": 55
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.42810480822035357,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6272,
      "step": 60
    },
    {
      "epoch": 0.06063432835820896,
      "grad_norm": 0.39853587993512773,
      "learning_rate": 4.99870371535606e-05,
      "loss": 0.6059,
      "step": 65
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.3862224380592072,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.635,
      "step": 70
    },
    {
      "epoch": 0.06996268656716417,
      "grad_norm": 0.4763664268940608,
      "learning_rate": 4.9952767243121146e-05,
      "loss": 0.6122,
      "step": 75
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.4964064315332045,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.6076,
      "step": 80
    },
    {
      "epoch": 0.07929104477611941,
      "grad_norm": 0.5168599776716694,
      "learning_rate": 4.989711576244639e-05,
      "loss": 0.6,
      "step": 85
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.41578311172819316,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.5894,
      "step": 90
    },
    {
      "epoch": 0.08861940298507463,
      "grad_norm": 0.4718528588737532,
      "learning_rate": 4.9820135707945634e-05,
      "loss": 0.5914,
      "step": 95
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.49638008772305653,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.6045,
      "step": 100
    },
    {
      "epoch": 0.09794776119402986,
      "grad_norm": 0.44140653927951257,
      "learning_rate": 4.972190038703905e-05,
      "loss": 0.6098,
      "step": 105
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.42118133939437635,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.5969,
      "step": 110
    },
    {
      "epoch": 0.10727611940298508,
      "grad_norm": 0.45053532317430645,
      "learning_rate": 4.9602503348347625e-05,
      "loss": 0.5874,
      "step": 115
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.4793106564516387,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.5948,
      "step": 120
    },
    {
      "epoch": 0.1166044776119403,
      "grad_norm": 0.4880433070764506,
      "learning_rate": 4.9462058292607735e-05,
      "loss": 0.5921,
      "step": 125
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.4235522421922073,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.5849,
      "step": 130
    },
    {
      "epoch": 0.1259328358208955,
      "grad_norm": 0.399275985838528,
      "learning_rate": 4.930069896439485e-05,
      "loss": 0.5846,
      "step": 135
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.5344724244547288,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.5867,
      "step": 140
    },
    {
      "epoch": 0.13526119402985073,
      "grad_norm": 0.4368309623159553,
      "learning_rate": 4.9118579024759854e-05,
      "loss": 0.5891,
      "step": 145
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.471415466306611,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.5729,
      "step": 150
    },
    {
      "epoch": 0.14458955223880596,
      "grad_norm": 0.42978553492277366,
      "learning_rate": 4.891587190489891e-05,
      "loss": 0.5775,
      "step": 155
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.42206017886782643,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.5853,
      "step": 160
    },
    {
      "epoch": 0.15391791044776118,
      "grad_norm": 0.3845543469543206,
      "learning_rate": 4.869277064099654e-05,
      "loss": 0.5811,
      "step": 165
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.41035761845429175,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.5916,
      "step": 170
    },
    {
      "epoch": 0.16324626865671643,
      "grad_norm": 0.4026284324321276,
      "learning_rate": 4.844948769039896e-05,
      "loss": 0.5857,
      "step": 175
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.3918826778611494,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.5812,
      "step": 180
    },
    {
      "epoch": 0.17257462686567165,
      "grad_norm": 0.39449144634351974,
      "learning_rate": 4.818625472929286e-05,
      "loss": 0.5741,
      "step": 185
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.3478654496926465,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.5677,
      "step": 190
    },
    {
      "epoch": 0.18190298507462688,
      "grad_norm": 0.42972864545531936,
      "learning_rate": 4.7903322432082185e-05,
      "loss": 0.568,
      "step": 195
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.36762920021108547,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5678,
      "step": 200
    },
    {
      "epoch": 0.1912313432835821,
      "grad_norm": 0.41776776027606394,
      "learning_rate": 4.760096023267322e-05,
      "loss": 0.5815,
      "step": 205
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.3651361955796744,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5732,
      "step": 210
    },
    {
      "epoch": 0.20055970149253732,
      "grad_norm": 0.507956038275049,
      "learning_rate": 4.7279456067895e-05,
      "loss": 0.5708,
      "step": 215
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.38678146760762727,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.5667,
      "step": 220
    },
    {
      "epoch": 0.20988805970149255,
      "grad_norm": 0.3746088635815425,
      "learning_rate": 4.6939116103299655e-05,
      "loss": 0.5594,
      "step": 225
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.4210888636990193,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5684,
      "step": 230
    },
    {
      "epoch": 0.21921641791044777,
      "grad_norm": 0.37863146424275174,
      "learning_rate": 4.6580264441603724e-05,
      "loss": 0.5754,
      "step": 235
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.3812578385049777,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.5745,
      "step": 240
    },
    {
      "epoch": 0.228544776119403,
      "grad_norm": 0.3755525320196136,
      "learning_rate": 4.6203242814047946e-05,
      "loss": 0.5594,
      "step": 245
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.35892802480588903,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5598,
      "step": 250
    },
    {
      "epoch": 0.23787313432835822,
      "grad_norm": 0.34690780233039625,
      "learning_rate": 4.580841025496974e-05,
      "loss": 0.5574,
      "step": 255
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.34148169877209095,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.5667,
      "step": 260
    },
    {
      "epoch": 0.24720149253731344,
      "grad_norm": 0.32369902312164034,
      "learning_rate": 4.539614275989793e-05,
      "loss": 0.5551,
      "step": 265
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.33514891723560625,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.5539,
      "step": 270
    },
    {
      "epoch": 0.25652985074626866,
      "grad_norm": 0.4073054681680018,
      "learning_rate": 4.496683292749555e-05,
      "loss": 0.5583,
      "step": 275
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.39594817481956696,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5616,
      "step": 280
    },
    {
      "epoch": 0.2658582089552239,
      "grad_norm": 0.3005600705660207,
      "learning_rate": 4.4520889585691705e-05,
      "loss": 0.5669,
      "step": 285
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.32063856721430456,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.551,
      "step": 290
    },
    {
      "epoch": 0.2751865671641791,
      "grad_norm": 0.29805045488753207,
      "learning_rate": 4.4058737402358295e-05,
      "loss": 0.5452,
      "step": 295
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.38405299320875,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5612,
      "step": 300
    },
    {
      "epoch": 0.28451492537313433,
      "grad_norm": 0.33664016303790095,
      "learning_rate": 4.3580816480902656e-05,
      "loss": 0.5454,
      "step": 305
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.3342664338536646,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5538,
      "step": 310
    },
    {
      "epoch": 0.29384328358208955,
      "grad_norm": 0.35017572756661175,
      "learning_rate": 4.308758194116094e-05,
      "loss": 0.5565,
      "step": 315
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.40426731913580016,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5679,
      "step": 320
    },
    {
      "epoch": 0.3031716417910448,
      "grad_norm": 0.36961111936209573,
      "learning_rate": 4.2579503485991567e-05,
      "loss": 0.5523,
      "step": 325
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.37687261195329114,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5591,
      "step": 330
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.4162442205962276,
      "learning_rate": 4.205706495398143e-05,
      "loss": 0.5547,
      "step": 335
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.34051096681409126,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.5406,
      "step": 340
    },
    {
      "epoch": 0.3218283582089552,
      "grad_norm": 0.3420797027133927,
      "learning_rate": 4.1520763858690644e-05,
      "loss": 0.5553,
      "step": 345
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.3352653830273245,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.5517,
      "step": 350
    },
    {
      "epoch": 0.33115671641791045,
      "grad_norm": 0.3248618140478854,
      "learning_rate": 4.097111091487486e-05,
      "loss": 0.5455,
      "step": 355
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.38398590733518817,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.5631,
      "step": 360
    },
    {
      "epoch": 0.34048507462686567,
      "grad_norm": 0.328922480165823,
      "learning_rate": 4.040862955213615e-05,
      "loss": 0.5571,
      "step": 365
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.3202207126152803,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5592,
      "step": 370
    },
    {
      "epoch": 0.3498134328358209,
      "grad_norm": 0.3498933076665203,
      "learning_rate": 3.9833855416465624e-05,
      "loss": 0.5468,
      "step": 375
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.3457169453229565,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5458,
      "step": 380
    },
    {
      "epoch": 0.3591417910447761,
      "grad_norm": 0.3049103435082426,
      "learning_rate": 3.924733586015257e-05,
      "loss": 0.5443,
      "step": 385
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.30856535885379005,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.5477,
      "step": 390
    },
    {
      "epoch": 0.36847014925373134,
      "grad_norm": 0.3344038208963223,
      "learning_rate": 3.864962942054572e-05,
      "loss": 0.5543,
      "step": 395
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.32351896377380496,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.546,
      "step": 400
    },
    {
      "epoch": 0.37779850746268656,
      "grad_norm": 0.32892338756590944,
      "learning_rate": 3.804130528816312e-05,
      "loss": 0.5486,
      "step": 405
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.3199167315472929,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.534,
      "step": 410
    },
    {
      "epoch": 0.3871268656716418,
      "grad_norm": 0.2986495023865257,
      "learning_rate": 3.7422942764657054e-05,
      "loss": 0.5454,
      "step": 415
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.33847840678383695,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.5505,
      "step": 420
    },
    {
      "epoch": 0.396455223880597,
      "grad_norm": 0.32240507890071096,
      "learning_rate": 3.679513071115025e-05,
      "loss": 0.5537,
      "step": 425
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.31943339464931536,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5508,
      "step": 430
    },
    {
      "epoch": 0.40578358208955223,
      "grad_norm": 0.3297548396557319,
      "learning_rate": 3.615846698746869e-05,
      "loss": 0.5544,
      "step": 435
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.32095008956501014,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5468,
      "step": 440
    },
    {
      "epoch": 0.41511194029850745,
      "grad_norm": 0.2954412311074952,
      "learning_rate": 3.5513557882805e-05,
      "loss": 0.5451,
      "step": 445
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.35657724937539037,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5477,
      "step": 450
    },
    {
      "epoch": 0.4244402985074627,
      "grad_norm": 0.32588100741949194,
      "learning_rate": 3.486101753835468e-05,
      "loss": 0.5405,
      "step": 455
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.31836225652258193,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5364,
      "step": 460
    },
    {
      "epoch": 0.4337686567164179,
      "grad_norm": 0.2667312497141442,
      "learning_rate": 3.420146736247487e-05,
      "loss": 0.5273,
      "step": 465
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.32043847430305067,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.5513,
      "step": 470
    },
    {
      "epoch": 0.4430970149253731,
      "grad_norm": 0.32874167757362127,
      "learning_rate": 3.353553543892277e-05,
      "loss": 0.5507,
      "step": 475
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.33258043796881387,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.5427,
      "step": 480
    },
    {
      "epoch": 0.45242537313432835,
      "grad_norm": 0.3020542500158025,
      "learning_rate": 3.2863855928737026e-05,
      "loss": 0.5375,
      "step": 485
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.2999989976107512,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.552,
      "step": 490
    },
    {
      "epoch": 0.46175373134328357,
      "grad_norm": 0.3135899658936416,
      "learning_rate": 3.218706846633183e-05,
      "loss": 0.5361,
      "step": 495
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.31783999641909394,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5321,
      "step": 500
    },
    {
      "epoch": 0.4710820895522388,
      "grad_norm": 0.3140003079271038,
      "learning_rate": 3.150581755037877e-05,
      "loss": 0.5444,
      "step": 505
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.2851153733261266,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.5391,
      "step": 510
    },
    {
      "epoch": 0.480410447761194,
      "grad_norm": 0.28512732601606067,
      "learning_rate": 3.08207519300565e-05,
      "loss": 0.5379,
      "step": 515
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.2823127624665095,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.5294,
      "step": 520
    },
    {
      "epoch": 0.48973880597014924,
      "grad_norm": 0.32198571530210485,
      "learning_rate": 3.0132523987252658e-05,
      "loss": 0.5327,
      "step": 525
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.287820704906656,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5229,
      "step": 530
    },
    {
      "epoch": 0.49906716417910446,
      "grad_norm": 0.2972454504515111,
      "learning_rate": 2.9441789115306402e-05,
      "loss": 0.5444,
      "step": 535
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.29514050211964044,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5405,
      "step": 540
    },
    {
      "epoch": 0.5083955223880597,
      "grad_norm": 0.3223030195906274,
      "learning_rate": 2.874920509488319e-05,
      "loss": 0.5436,
      "step": 545
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.331237041377847,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.5386,
      "step": 550
    },
    {
      "epoch": 0.5177238805970149,
      "grad_norm": 0.28692038439269246,
      "learning_rate": 2.8055431467576106e-05,
      "loss": 0.5366,
      "step": 555
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.30635958523804346,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.54,
      "step": 560
    },
    {
      "epoch": 0.5270522388059702,
      "grad_norm": 0.31895459933403253,
      "learning_rate": 2.7361128907830253e-05,
      "loss": 0.5309,
      "step": 565
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.28189697192197444,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5214,
      "step": 570
    },
    {
      "epoch": 0.5363805970149254,
      "grad_norm": 0.28815634636872006,
      "learning_rate": 2.6666958593788405e-05,
      "loss": 0.5364,
      "step": 575
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.29779272601870393,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.5409,
      "step": 580
    },
    {
      "epoch": 0.5457089552238806,
      "grad_norm": 0.2736771548372716,
      "learning_rate": 2.597358157765692e-05,
      "loss": 0.5287,
      "step": 585
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.2610666235288935,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.5261,
      "step": 590
    },
    {
      "epoch": 0.5550373134328358,
      "grad_norm": 0.30003927928430624,
      "learning_rate": 2.528165815619162e-05,
      "loss": 0.5361,
      "step": 595
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.27184633082761,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5272,
      "step": 600
    },
    {
      "epoch": 0.5643656716417911,
      "grad_norm": 0.2673745453598303,
      "learning_rate": 2.459184724190308e-05,
      "loss": 0.5195,
      "step": 605
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.30502816322055604,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5339,
      "step": 610
    },
    {
      "epoch": 0.5736940298507462,
      "grad_norm": 0.2915347636882713,
      "learning_rate": 2.390480573558012e-05,
      "loss": 0.5404,
      "step": 615
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.3009151953810016,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5251,
      "step": 620
    },
    {
      "epoch": 0.5830223880597015,
      "grad_norm": 0.273216079030479,
      "learning_rate": 2.3221187900729003e-05,
      "loss": 0.5289,
      "step": 625
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.2814635050429347,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.5304,
      "step": 630
    },
    {
      "epoch": 0.5923507462686567,
      "grad_norm": 0.2671887715565127,
      "learning_rate": 2.254164474052416e-05,
      "loss": 0.5336,
      "step": 635
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.2813730919512758,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5179,
      "step": 640
    },
    {
      "epoch": 0.601679104477612,
      "grad_norm": 0.27596121024924675,
      "learning_rate": 2.186682337786365e-05,
      "loss": 0.53,
      "step": 645
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.2799192543432165,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5177,
      "step": 650
    },
    {
      "epoch": 0.6110074626865671,
      "grad_norm": 0.2885695286696909,
      "learning_rate": 2.119736643911979e-05,
      "loss": 0.5266,
      "step": 655
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.2744442740400871,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.5304,
      "step": 660
    },
    {
      "epoch": 0.6203358208955224,
      "grad_norm": 0.26479139181117806,
      "learning_rate": 2.0533911442171805e-05,
      "loss": 0.5265,
      "step": 665
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.29307727200532646,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.5237,
      "step": 670
    },
    {
      "epoch": 0.6296641791044776,
      "grad_norm": 0.26780149637777595,
      "learning_rate": 1.9877090189303182e-05,
      "loss": 0.531,
      "step": 675
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.2724139697039071,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.5237,
      "step": 680
    },
    {
      "epoch": 0.6389925373134329,
      "grad_norm": 0.2603052625446588,
      "learning_rate": 1.922752816554204e-05,
      "loss": 0.5234,
      "step": 685
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.26133005941810766,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.5174,
      "step": 690
    },
    {
      "epoch": 0.648320895522388,
      "grad_norm": 0.24984442702607385,
      "learning_rate": 1.858584394301728e-05,
      "loss": 0.5215,
      "step": 695
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.28057878918319706,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.5386,
      "step": 700
    },
    {
      "epoch": 0.6576492537313433,
      "grad_norm": 0.2481915917558959,
      "learning_rate": 1.7952648591897858e-05,
      "loss": 0.5212,
      "step": 705
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.2908975085954642,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.5132,
      "step": 710
    },
    {
      "epoch": 0.6669776119402985,
      "grad_norm": 0.24930505384787335,
      "learning_rate": 1.7328545098476106e-05,
      "loss": 0.5167,
      "step": 715
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.26823214875899426,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.5191,
      "step": 720
    },
    {
      "epoch": 0.6763059701492538,
      "grad_norm": 0.2613619981319557,
      "learning_rate": 1.671412779094926e-05,
      "loss": 0.5229,
      "step": 725
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.2524866364696518,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.5191,
      "step": 730
    },
    {
      "epoch": 0.6856343283582089,
      "grad_norm": 0.2754948559566439,
      "learning_rate": 1.6109981773446036e-05,
      "loss": 0.5204,
      "step": 735
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.2473197406930695,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.5188,
      "step": 740
    },
    {
      "epoch": 0.6949626865671642,
      "grad_norm": 0.24213175335701867,
      "learning_rate": 1.5516682368837133e-05,
      "loss": 0.5142,
      "step": 745
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.2760399820439403,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.5307,
      "step": 750
    },
    {
      "epoch": 0.7042910447761194,
      "grad_norm": 0.265171148488907,
      "learning_rate": 1.4934794570860416e-05,
      "loss": 0.5314,
      "step": 755
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.2636748634412273,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5178,
      "step": 760
    },
    {
      "epoch": 0.7136194029850746,
      "grad_norm": 0.26114188558324736,
      "learning_rate": 1.4364872506082425e-05,
      "loss": 0.5255,
      "step": 765
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.27463450094575076,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.5286,
      "step": 770
    },
    {
      "epoch": 0.7229477611940298,
      "grad_norm": 0.2627758464521469,
      "learning_rate": 1.3807458906208546e-05,
      "loss": 0.5355,
      "step": 775
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.2535387674795409,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5148,
      "step": 780
    },
    {
      "epoch": 0.7322761194029851,
      "grad_norm": 0.25143264464640164,
      "learning_rate": 1.326308459124447e-05,
      "loss": 0.518,
      "step": 785
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.24678458633586137,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.5265,
      "step": 790
    },
    {
      "epoch": 0.7416044776119403,
      "grad_norm": 0.25397199889705413,
      "learning_rate": 1.2732267964001033e-05,
      "loss": 0.5147,
      "step": 795
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.2612625440498218,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5212,
      "step": 800
    },
    {
      "epoch": 0.7509328358208955,
      "grad_norm": 0.24164108913385374,
      "learning_rate": 1.2215514516423813e-05,
      "loss": 0.5352,
      "step": 805
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.2560594323046853,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.5138,
      "step": 810
    },
    {
      "epoch": 0.7602611940298507,
      "grad_norm": 0.24857174780101957,
      "learning_rate": 1.1713316348217673e-05,
      "loss": 0.5187,
      "step": 815
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.2522325967641256,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.5033,
      "step": 820
    },
    {
      "epoch": 0.769589552238806,
      "grad_norm": 0.23930223016635088,
      "learning_rate": 1.1226151698224597e-05,
      "loss": 0.5243,
      "step": 825
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.24668012695988362,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.5283,
      "step": 830
    },
    {
      "epoch": 0.7789179104477612,
      "grad_norm": 0.2440582920115209,
      "learning_rate": 1.0754484489001085e-05,
      "loss": 0.5185,
      "step": 835
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.2518319679837789,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.5174,
      "step": 840
    },
    {
      "epoch": 0.7882462686567164,
      "grad_norm": 0.24479017106886222,
      "learning_rate": 1.0298763885028839e-05,
      "loss": 0.5135,
      "step": 845
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.23990431396616274,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5192,
      "step": 850
    },
    {
      "epoch": 0.7975746268656716,
      "grad_norm": 0.22863144922943815,
      "learning_rate": 9.859423864979441e-06,
      "loss": 0.5076,
      "step": 855
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.2526688437058162,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.5266,
      "step": 860
    },
    {
      "epoch": 0.8069029850746269,
      "grad_norm": 0.2571652401412218,
      "learning_rate": 9.436882808440334e-06,
      "loss": 0.5279,
      "step": 865
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.23297383232884672,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.51,
      "step": 870
    },
    {
      "epoch": 0.816231343283582,
      "grad_norm": 0.23792626827153376,
      "learning_rate": 9.031543097495638e-06,
      "loss": 0.5145,
      "step": 875
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.259491381893308,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5125,
      "step": 880
    },
    {
      "epoch": 0.8255597014925373,
      "grad_norm": 0.2544129442607262,
      "learning_rate": 8.6437907335413e-06,
      "loss": 0.5225,
      "step": 885
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.23802013309346642,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5248,
      "step": 890
    },
    {
      "epoch": 0.8348880597014925,
      "grad_norm": 0.24098732939917877,
      "learning_rate": 8.273994969699394e-06,
      "loss": 0.5188,
      "step": 895
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.22509509146201456,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.5218,
      "step": 900
    },
    {
      "epoch": 0.8434296365330848,
      "grad_norm": 0.25503644802838177,
      "learning_rate": 7.95115281652163e-06,
      "loss": 0.4928,
      "step": 905
    },
    {
      "epoch": 0.848089468779124,
      "grad_norm": 0.25554829568809356,
      "learning_rate": 7.781763120363474e-06,
      "loss": 0.4939,
      "step": 910
    },
    {
      "epoch": 0.8527493010251631,
      "grad_norm": 0.24785150988406393,
      "learning_rate": 7.6170588694362915e-06,
      "loss": 0.4962,
      "step": 915
    },
    {
      "epoch": 0.8574091332712023,
      "grad_norm": 0.2345884436747603,
      "learning_rate": 7.4570792007456745e-06,
      "loss": 0.5042,
      "step": 920
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.26221430699973647,
      "learning_rate": 7.3018621286425035e-06,
      "loss": 0.4985,
      "step": 925
    },
    {
      "epoch": 0.8667287977632805,
      "grad_norm": 0.23632019810695779,
      "learning_rate": 7.151444535790017e-06,
      "loss": 0.4879,
      "step": 930
    },
    {
      "epoch": 0.8713886300093197,
      "grad_norm": 0.23685269779407658,
      "learning_rate": 7.005862164399716e-06,
      "loss": 0.4952,
      "step": 935
    },
    {
      "epoch": 0.8760484622553588,
      "grad_norm": 0.2263847108294601,
      "learning_rate": 6.865149607738324e-06,
      "loss": 0.4842,
      "step": 940
    },
    {
      "epoch": 0.880708294501398,
      "grad_norm": 0.2493209183511488,
      "learning_rate": 6.7293403019077394e-06,
      "loss": 0.4875,
      "step": 945
    },
    {
      "epoch": 0.8853681267474371,
      "grad_norm": 0.24952354183015024,
      "learning_rate": 6.598466517899961e-06,
      "loss": 0.5012,
      "step": 950
    },
    {
      "epoch": 0.8900279589934762,
      "grad_norm": 0.24015418129527374,
      "learning_rate": 6.472559353928814e-06,
      "loss": 0.497,
      "step": 955
    },
    {
      "epoch": 0.8946877912395154,
      "grad_norm": 0.2406838118135352,
      "learning_rate": 6.35164872804046e-06,
      "loss": 0.4904,
      "step": 960
    },
    {
      "epoch": 0.8993476234855545,
      "grad_norm": 0.23781577622705932,
      "learning_rate": 6.235763371004234e-06,
      "loss": 0.4853,
      "step": 965
    },
    {
      "epoch": 0.9040074557315937,
      "grad_norm": 0.24384534616887826,
      "learning_rate": 6.124930819485644e-06,
      "loss": 0.5014,
      "step": 970
    },
    {
      "epoch": 0.9086672879776329,
      "grad_norm": 0.24824064166520088,
      "learning_rate": 6.0191774095031244e-06,
      "loss": 0.4794,
      "step": 975
    },
    {
      "epoch": 0.9133271202236719,
      "grad_norm": 0.2397496796949414,
      "learning_rate": 5.9185282701700745e-06,
      "loss": 0.4995,
      "step": 980
    },
    {
      "epoch": 0.9179869524697111,
      "grad_norm": 0.23793615513302432,
      "learning_rate": 5.823007317723664e-06,
      "loss": 0.4909,
      "step": 985
    },
    {
      "epoch": 0.9226467847157502,
      "grad_norm": 0.2417670178260445,
      "learning_rate": 5.732637249841873e-06,
      "loss": 0.4824,
      "step": 990
    },
    {
      "epoch": 0.9273066169617894,
      "grad_norm": 0.23712042301378178,
      "learning_rate": 5.647439540250082e-06,
      "loss": 0.4876,
      "step": 995
    },
    {
      "epoch": 0.9319664492078286,
      "grad_norm": 0.23577490401974563,
      "learning_rate": 5.567434433618465e-06,
      "loss": 0.4848,
      "step": 1000
    },
    {
      "epoch": 0.9366262814538676,
      "grad_norm": 0.239146845460939,
      "learning_rate": 5.492640940751462e-06,
      "loss": 0.4797,
      "step": 1005
    },
    {
      "epoch": 0.9412861136999068,
      "grad_norm": 0.2542109123530308,
      "learning_rate": 5.423076834070447e-06,
      "loss": 0.4787,
      "step": 1010
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 0.23090507379059155,
      "learning_rate": 5.358758643390628e-06,
      "loss": 0.5016,
      "step": 1015
    },
    {
      "epoch": 0.9506057781919851,
      "grad_norm": 0.23532827969201903,
      "learning_rate": 5.299701651993246e-06,
      "loss": 0.5013,
      "step": 1020
    },
    {
      "epoch": 0.9552656104380243,
      "grad_norm": 0.23516814839669387,
      "learning_rate": 5.245919892993957e-06,
      "loss": 0.4826,
      "step": 1025
    },
    {
      "epoch": 0.9599254426840633,
      "grad_norm": 0.23105280417402752,
      "learning_rate": 5.197426146008291e-06,
      "loss": 0.4931,
      "step": 1030
    },
    {
      "epoch": 0.9645852749301025,
      "grad_norm": 0.2663559537808514,
      "learning_rate": 5.1542319341149565e-06,
      "loss": 0.4899,
      "step": 1035
    },
    {
      "epoch": 0.9692451071761417,
      "grad_norm": 0.23949167751897144,
      "learning_rate": 5.1163475211177235e-06,
      "loss": 0.4988,
      "step": 1040
    },
    {
      "epoch": 0.9739049394221808,
      "grad_norm": 0.24000618926829137,
      "learning_rate": 5.083781909106557e-06,
      "loss": 0.4828,
      "step": 1045
    },
    {
      "epoch": 0.97856477166822,
      "grad_norm": 0.24152911669491808,
      "learning_rate": 5.056542836318518e-06,
      "loss": 0.498,
      "step": 1050
    },
    {
      "epoch": 0.983224603914259,
      "grad_norm": 0.22134585139948434,
      "learning_rate": 5.034636775299023e-06,
      "loss": 0.4894,
      "step": 1055
    },
    {
      "epoch": 0.9878844361602982,
      "grad_norm": 0.23386321836027193,
      "learning_rate": 5.018068931363828e-06,
      "loss": 0.4874,
      "step": 1060
    },
    {
      "epoch": 0.9925442684063374,
      "grad_norm": 0.23532118958350487,
      "learning_rate": 5.006843241362149e-06,
      "loss": 0.4972,
      "step": 1065
    },
    {
      "epoch": 0.9972041006523765,
      "grad_norm": 0.22265388724532062,
      "learning_rate": 5.000962372741178e-06,
      "loss": 0.498,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "step": 1073,
      "total_flos": 489077052801024.0,
      "train_loss": 0.07929676676729741,
      "train_runtime": 3102.6198,
      "train_samples_per_second": 11.059,
      "train_steps_per_second": 0.346
    }
  ],
  "logging_steps": 5,
  "max_steps": 1073,
  "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": 489077052801024.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}