{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1072,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0046641791044776115,
      "grad_norm": 1.9737565699550477,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.8462,
      "step": 5
    },
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.4313045221260292,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8235,
      "step": 10
    },
    {
      "epoch": 0.013992537313432836,
      "grad_norm": 1.1346996876854334,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.7682,
      "step": 15
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 0.6828641731714449,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.7273,
      "step": 20
    },
    {
      "epoch": 0.02332089552238806,
      "grad_norm": 0.5132486218762813,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.6962,
      "step": 25
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.48646435519095227,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6736,
      "step": 30
    },
    {
      "epoch": 0.03264925373134328,
      "grad_norm": 0.4320653730399992,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.6448,
      "step": 35
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.46506998468975,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.638,
      "step": 40
    },
    {
      "epoch": 0.04197761194029851,
      "grad_norm": 0.4190129866016933,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6334,
      "step": 45
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.4197147311398817,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.615,
      "step": 50
    },
    {
      "epoch": 0.051305970149253734,
      "grad_norm": 0.454150411866044,
      "learning_rate": 4.999989285883431e-05,
      "loss": 0.6343,
      "step": 55
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.4976545347994734,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6274,
      "step": 60
    },
    {
      "epoch": 0.06063432835820896,
      "grad_norm": 0.45598284917949133,
      "learning_rate": 4.99870371535606e-05,
      "loss": 0.6063,
      "step": 65
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.5227086559500899,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.6353,
      "step": 70
    },
    {
      "epoch": 0.06996268656716417,
      "grad_norm": 0.6174850179632522,
      "learning_rate": 4.9952767243121146e-05,
      "loss": 0.6128,
      "step": 75
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.5880731170032467,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.6087,
      "step": 80
    },
    {
      "epoch": 0.07929104477611941,
      "grad_norm": 0.6375946180801447,
      "learning_rate": 4.989711576244639e-05,
      "loss": 0.6006,
      "step": 85
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.48790239748391595,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.5899,
      "step": 90
    },
    {
      "epoch": 0.08861940298507463,
      "grad_norm": 0.5318335956060131,
      "learning_rate": 4.9820135707945634e-05,
      "loss": 0.5917,
      "step": 95
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.5656690396811953,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.6048,
      "step": 100
    },
    {
      "epoch": 0.09794776119402986,
      "grad_norm": 0.4691363598516851,
      "learning_rate": 4.972190038703905e-05,
      "loss": 0.6103,
      "step": 105
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.49198235678257524,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.5973,
      "step": 110
    },
    {
      "epoch": 0.10727611940298508,
      "grad_norm": 0.4828283304659039,
      "learning_rate": 4.9602503348347625e-05,
      "loss": 0.5879,
      "step": 115
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.5222495128976433,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.5953,
      "step": 120
    },
    {
      "epoch": 0.1166044776119403,
      "grad_norm": 0.5173262535486767,
      "learning_rate": 4.9462058292607735e-05,
      "loss": 0.5924,
      "step": 125
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.49449726128671867,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.5853,
      "step": 130
    },
    {
      "epoch": 0.1259328358208955,
      "grad_norm": 0.3981488723042445,
      "learning_rate": 4.930069896439485e-05,
      "loss": 0.5848,
      "step": 135
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.5616720737443401,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.5869,
      "step": 140
    },
    {
      "epoch": 0.13526119402985073,
      "grad_norm": 0.5645188109079204,
      "learning_rate": 4.9118579024759854e-05,
      "loss": 0.5896,
      "step": 145
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.4528930971557964,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.5733,
      "step": 150
    },
    {
      "epoch": 0.14458955223880596,
      "grad_norm": 0.5119372178393736,
      "learning_rate": 4.891587190489891e-05,
      "loss": 0.5779,
      "step": 155
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.545404890594189,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.5855,
      "step": 160
    },
    {
      "epoch": 0.15391791044776118,
      "grad_norm": 0.5516112369637339,
      "learning_rate": 4.869277064099654e-05,
      "loss": 0.5814,
      "step": 165
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.42360505235042945,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.5918,
      "step": 170
    },
    {
      "epoch": 0.16324626865671643,
      "grad_norm": 0.41597973453963927,
      "learning_rate": 4.844948769039896e-05,
      "loss": 0.5859,
      "step": 175
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.44936882842852305,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.5817,
      "step": 180
    },
    {
      "epoch": 0.17257462686567165,
      "grad_norm": 0.46299205203501975,
      "learning_rate": 4.818625472929286e-05,
      "loss": 0.5746,
      "step": 185
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.4673688881878735,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.5684,
      "step": 190
    },
    {
      "epoch": 0.18190298507462688,
      "grad_norm": 0.4435776856144207,
      "learning_rate": 4.7903322432082185e-05,
      "loss": 0.5686,
      "step": 195
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.4058371589175082,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5684,
      "step": 200
    },
    {
      "epoch": 0.1912313432835821,
      "grad_norm": 0.49977241209761997,
      "learning_rate": 4.760096023267322e-05,
      "loss": 0.5821,
      "step": 205
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.43523914588954843,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5739,
      "step": 210
    },
    {
      "epoch": 0.20055970149253732,
      "grad_norm": 0.7085255215907048,
      "learning_rate": 4.7279456067895e-05,
      "loss": 0.5716,
      "step": 215
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.5653424685417476,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.5674,
      "step": 220
    },
    {
      "epoch": 0.20988805970149255,
      "grad_norm": 0.5244121990272786,
      "learning_rate": 4.6939116103299655e-05,
      "loss": 0.56,
      "step": 225
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.5077725577171148,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5688,
      "step": 230
    },
    {
      "epoch": 0.21921641791044777,
      "grad_norm": 0.4425026395421933,
      "learning_rate": 4.6580264441603724e-05,
      "loss": 0.576,
      "step": 235
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.4341800111451734,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.5752,
      "step": 240
    },
    {
      "epoch": 0.228544776119403,
      "grad_norm": 0.4996634399185324,
      "learning_rate": 4.6203242814047946e-05,
      "loss": 0.5603,
      "step": 245
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.5912347448686226,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5607,
      "step": 250
    },
    {
      "epoch": 0.23787313432835822,
      "grad_norm": 0.43002434255127664,
      "learning_rate": 4.580841025496974e-05,
      "loss": 0.5581,
      "step": 255
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.4485234966770534,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.5677,
      "step": 260
    },
    {
      "epoch": 0.24720149253731344,
      "grad_norm": 0.37138561390050095,
      "learning_rate": 4.539614275989793e-05,
      "loss": 0.5565,
      "step": 265
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.4133626295112158,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.5548,
      "step": 270
    },
    {
      "epoch": 0.25652985074626866,
      "grad_norm": 0.45815619256972684,
      "learning_rate": 4.496683292749555e-05,
      "loss": 0.5592,
      "step": 275
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.5382642082030428,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5627,
      "step": 280
    },
    {
      "epoch": 0.2658582089552239,
      "grad_norm": 0.46388012094196474,
      "learning_rate": 4.4520889585691705e-05,
      "loss": 0.5677,
      "step": 285
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.4651562768167999,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.5518,
      "step": 290
    },
    {
      "epoch": 0.2751865671641791,
      "grad_norm": 0.3962720245968155,
      "learning_rate": 4.4058737402358295e-05,
      "loss": 0.5458,
      "step": 295
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.4883867917348728,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5616,
      "step": 300
    },
    {
      "epoch": 0.28451492537313433,
      "grad_norm": 0.4225077326378658,
      "learning_rate": 4.3580816480902656e-05,
      "loss": 0.5459,
      "step": 305
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.3662378172000391,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5542,
      "step": 310
    },
    {
      "epoch": 0.29384328358208955,
      "grad_norm": 0.4446516792918299,
      "learning_rate": 4.308758194116094e-05,
      "loss": 0.5574,
      "step": 315
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.4359436549059787,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5686,
      "step": 320
    },
    {
      "epoch": 0.3031716417910448,
      "grad_norm": 0.562147315461778,
      "learning_rate": 4.2579503485991567e-05,
      "loss": 0.5529,
      "step": 325
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.5107829000076599,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5597,
      "step": 330
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.4735116577263397,
      "learning_rate": 4.205706495398143e-05,
      "loss": 0.5553,
      "step": 335
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.5036979789350788,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.5414,
      "step": 340
    },
    {
      "epoch": 0.3218283582089552,
      "grad_norm": 0.3917546906129853,
      "learning_rate": 4.1520763858690644e-05,
      "loss": 0.5558,
      "step": 345
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.5111185433617897,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.552,
      "step": 350
    },
    {
      "epoch": 0.33115671641791045,
      "grad_norm": 0.45005650174171213,
      "learning_rate": 4.097111091487486e-05,
      "loss": 0.5461,
      "step": 355
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.4185868959063299,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.5638,
      "step": 360
    },
    {
      "epoch": 0.34048507462686567,
      "grad_norm": 0.4385018433239723,
      "learning_rate": 4.040862955213615e-05,
      "loss": 0.5579,
      "step": 365
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.3658683971284041,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5598,
      "step": 370
    },
    {
      "epoch": 0.3498134328358209,
      "grad_norm": 0.44442062333546467,
      "learning_rate": 3.9833855416465624e-05,
      "loss": 0.5473,
      "step": 375
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.45116064271828604,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5463,
      "step": 380
    },
    {
      "epoch": 0.3591417910447761,
      "grad_norm": 0.36498533110429815,
      "learning_rate": 3.924733586015257e-05,
      "loss": 0.545,
      "step": 385
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.39587844840520986,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.5484,
      "step": 390
    },
    {
      "epoch": 0.36847014925373134,
      "grad_norm": 0.4001846361793331,
      "learning_rate": 3.864962942054572e-05,
      "loss": 0.5549,
      "step": 395
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.38715828447829004,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.5466,
      "step": 400
    },
    {
      "epoch": 0.37779850746268656,
      "grad_norm": 0.4275094946094768,
      "learning_rate": 3.804130528816312e-05,
      "loss": 0.549,
      "step": 405
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.5465207240404546,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.5347,
      "step": 410
    },
    {
      "epoch": 0.3871268656716418,
      "grad_norm": 0.3810626185760908,
      "learning_rate": 3.7422942764657054e-05,
      "loss": 0.5462,
      "step": 415
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.42546821730605433,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.5513,
      "step": 420
    },
    {
      "epoch": 0.396455223880597,
      "grad_norm": 0.40660329464977896,
      "learning_rate": 3.679513071115025e-05,
      "loss": 0.5545,
      "step": 425
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.4901995684742552,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5514,
      "step": 430
    },
    {
      "epoch": 0.40578358208955223,
      "grad_norm": 0.4561473177324483,
      "learning_rate": 3.615846698746869e-05,
      "loss": 0.5551,
      "step": 435
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.4545066058248777,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5475,
      "step": 440
    },
    {
      "epoch": 0.41511194029850745,
      "grad_norm": 0.3567683329232947,
      "learning_rate": 3.5513557882805e-05,
      "loss": 0.5457,
      "step": 445
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.4935516493631999,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5483,
      "step": 450
    },
    {
      "epoch": 0.4244402985074627,
      "grad_norm": 0.4278828841184784,
      "learning_rate": 3.486101753835468e-05,
      "loss": 0.5411,
      "step": 455
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.3831980820409706,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5369,
      "step": 460
    },
    {
      "epoch": 0.4337686567164179,
      "grad_norm": 0.35346791316580967,
      "learning_rate": 3.420146736247487e-05,
      "loss": 0.5278,
      "step": 465
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.4341647119894046,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.552,
      "step": 470
    },
    {
      "epoch": 0.4430970149253731,
      "grad_norm": 0.38104723235081495,
      "learning_rate": 3.353553543892277e-05,
      "loss": 0.5513,
      "step": 475
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.42615290307324083,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.5432,
      "step": 480
    },
    {
      "epoch": 0.45242537313432835,
      "grad_norm": 0.3462873785234133,
      "learning_rate": 3.2863855928737026e-05,
      "loss": 0.5381,
      "step": 485
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.3586109539340555,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.5527,
      "step": 490
    },
    {
      "epoch": 0.46175373134328357,
      "grad_norm": 0.3606896599871606,
      "learning_rate": 3.218706846633183e-05,
      "loss": 0.5367,
      "step": 495
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.4086083805613451,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5327,
      "step": 500
    },
    {
      "epoch": 0.4710820895522388,
      "grad_norm": 0.39895267005686896,
      "learning_rate": 3.150581755037877e-05,
      "loss": 0.5449,
      "step": 505
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.32662220814932813,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.5397,
      "step": 510
    },
    {
      "epoch": 0.480410447761194,
      "grad_norm": 0.3375934491741763,
      "learning_rate": 3.08207519300565e-05,
      "loss": 0.5385,
      "step": 515
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.34787210971968974,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.53,
      "step": 520
    },
    {
      "epoch": 0.48973880597014924,
      "grad_norm": 0.39620710906533535,
      "learning_rate": 3.0132523987252658e-05,
      "loss": 0.5335,
      "step": 525
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.3784201727069867,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5234,
      "step": 530
    },
    {
      "epoch": 0.49906716417910446,
      "grad_norm": 0.36372530063574576,
      "learning_rate": 2.9441789115306402e-05,
      "loss": 0.5449,
      "step": 535
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.33385903740532513,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5411,
      "step": 540
    },
    {
      "epoch": 0.5083955223880597,
      "grad_norm": 0.3634795150059451,
      "learning_rate": 2.874920509488319e-05,
      "loss": 0.544,
      "step": 545
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.3740217231457619,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.5392,
      "step": 550
    },
    {
      "epoch": 0.5177238805970149,
      "grad_norm": 0.32673424561583636,
      "learning_rate": 2.8055431467576106e-05,
      "loss": 0.5372,
      "step": 555
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.32989889996277566,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.5406,
      "step": 560
    },
    {
      "epoch": 0.5270522388059702,
      "grad_norm": 0.34578592805977293,
      "learning_rate": 2.7361128907830253e-05,
      "loss": 0.5315,
      "step": 565
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.3355523828461531,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5219,
      "step": 570
    },
    {
      "epoch": 0.5363805970149254,
      "grad_norm": 0.32374626691721103,
      "learning_rate": 2.6666958593788405e-05,
      "loss": 0.5373,
      "step": 575
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.3498631050042572,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.5415,
      "step": 580
    },
    {
      "epoch": 0.5457089552238806,
      "grad_norm": 0.31152996247408016,
      "learning_rate": 2.597358157765692e-05,
      "loss": 0.5293,
      "step": 585
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.3008488970775489,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.5269,
      "step": 590
    },
    {
      "epoch": 0.5550373134328358,
      "grad_norm": 0.3679099326165536,
      "learning_rate": 2.528165815619162e-05,
      "loss": 0.5367,
      "step": 595
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.3216377349194825,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5278,
      "step": 600
    },
    {
      "epoch": 0.5643656716417911,
      "grad_norm": 0.3087685158708632,
      "learning_rate": 2.459184724190308e-05,
      "loss": 0.5202,
      "step": 605
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.3391573089678486,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5345,
      "step": 610
    },
    {
      "epoch": 0.5736940298507462,
      "grad_norm": 0.3255460099079437,
      "learning_rate": 2.390480573558012e-05,
      "loss": 0.5411,
      "step": 615
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.35983471442574044,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5257,
      "step": 620
    },
    {
      "epoch": 0.5830223880597015,
      "grad_norm": 0.32544671257325375,
      "learning_rate": 2.3221187900729003e-05,
      "loss": 0.5293,
      "step": 625
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.34956657987333445,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.531,
      "step": 630
    },
    {
      "epoch": 0.5923507462686567,
      "grad_norm": 0.32518866529909685,
      "learning_rate": 2.254164474052416e-05,
      "loss": 0.5342,
      "step": 635
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.33351983201189994,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5183,
      "step": 640
    },
    {
      "epoch": 0.601679104477612,
      "grad_norm": 0.3254337838136604,
      "learning_rate": 2.186682337786365e-05,
      "loss": 0.5306,
      "step": 645
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.3010923556830084,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5183,
      "step": 650
    },
    {
      "epoch": 0.6110074626865671,
      "grad_norm": 0.3728436902441161,
      "learning_rate": 2.119736643911979e-05,
      "loss": 0.5271,
      "step": 655
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.32934721950944257,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.531,
      "step": 660
    },
    {
      "epoch": 0.6203358208955224,
      "grad_norm": 0.32241891261617145,
      "learning_rate": 2.0533911442171805e-05,
      "loss": 0.527,
      "step": 665
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.34418976349661284,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.5243,
      "step": 670
    },
    {
      "epoch": 0.6296641791044776,
      "grad_norm": 0.3108019920881547,
      "learning_rate": 1.9877090189303182e-05,
      "loss": 0.5316,
      "step": 675
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.31926070186961547,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.5243,
      "step": 680
    },
    {
      "epoch": 0.6389925373134329,
      "grad_norm": 0.3155811744495747,
      "learning_rate": 1.922752816554204e-05,
      "loss": 0.5241,
      "step": 685
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.2937615368459596,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.518,
      "step": 690
    },
    {
      "epoch": 0.648320895522388,
      "grad_norm": 0.29046163409682196,
      "learning_rate": 1.858584394301728e-05,
      "loss": 0.5221,
      "step": 695
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.3145402617953687,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.5394,
      "step": 700
    },
    {
      "epoch": 0.6576492537313433,
      "grad_norm": 0.29543998229149865,
      "learning_rate": 1.7952648591897858e-05,
      "loss": 0.522,
      "step": 705
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.36052987636160755,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.5138,
      "step": 710
    },
    {
      "epoch": 0.6669776119402985,
      "grad_norm": 0.3209855069869533,
      "learning_rate": 1.7328545098476106e-05,
      "loss": 0.5173,
      "step": 715
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.308357685515476,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.5197,
      "step": 720
    },
    {
      "epoch": 0.6763059701492538,
      "grad_norm": 0.3329034887769105,
      "learning_rate": 1.671412779094926e-05,
      "loss": 0.5235,
      "step": 725
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.3047861045310467,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.5197,
      "step": 730
    },
    {
      "epoch": 0.6856343283582089,
      "grad_norm": 0.31416211733218163,
      "learning_rate": 1.6109981773446036e-05,
      "loss": 0.521,
      "step": 735
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.28132655200766216,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.5195,
      "step": 740
    },
    {
      "epoch": 0.6949626865671642,
      "grad_norm": 0.2814236111750379,
      "learning_rate": 1.5516682368837133e-05,
      "loss": 0.5149,
      "step": 745
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.30912591121226046,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.5313,
      "step": 750
    },
    {
      "epoch": 0.7042910447761194,
      "grad_norm": 0.30995429354155263,
      "learning_rate": 1.4934794570860416e-05,
      "loss": 0.5322,
      "step": 755
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.3087346682651283,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5183,
      "step": 760
    },
    {
      "epoch": 0.7136194029850746,
      "grad_norm": 0.3167565478672606,
      "learning_rate": 1.4364872506082425e-05,
      "loss": 0.526,
      "step": 765
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.3133737836701582,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.5292,
      "step": 770
    },
    {
      "epoch": 0.7229477611940298,
      "grad_norm": 0.30698622747486176,
      "learning_rate": 1.3807458906208546e-05,
      "loss": 0.536,
      "step": 775
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.292344961823975,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5153,
      "step": 780
    },
    {
      "epoch": 0.7322761194029851,
      "grad_norm": 0.2958271040668866,
      "learning_rate": 1.326308459124447e-05,
      "loss": 0.5185,
      "step": 785
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.29371243569106203,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.5272,
      "step": 790
    },
    {
      "epoch": 0.7416044776119403,
      "grad_norm": 0.2917942436195178,
      "learning_rate": 1.2732267964001033e-05,
      "loss": 0.5153,
      "step": 795
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.30518780549956975,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5217,
      "step": 800
    },
    {
      "epoch": 0.7509328358208955,
      "grad_norm": 0.2723694367867143,
      "learning_rate": 1.2215514516423813e-05,
      "loss": 0.5355,
      "step": 805
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.30521862254139703,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.5144,
      "step": 810
    },
    {
      "epoch": 0.7602611940298507,
      "grad_norm": 0.2820686752314443,
      "learning_rate": 1.1713316348217673e-05,
      "loss": 0.5193,
      "step": 815
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.3048405563450722,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.5039,
      "step": 820
    },
    {
      "epoch": 0.769589552238806,
      "grad_norm": 0.27703352458545016,
      "learning_rate": 1.1226151698224597e-05,
      "loss": 0.5249,
      "step": 825
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.2806197704847487,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.5289,
      "step": 830
    },
    {
      "epoch": 0.7789179104477612,
      "grad_norm": 0.2748109945673585,
      "learning_rate": 1.0754484489001085e-05,
      "loss": 0.5192,
      "step": 835
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.29771643664474096,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.518,
      "step": 840
    },
    {
      "epoch": 0.7882462686567164,
      "grad_norm": 0.27387569498834374,
      "learning_rate": 1.0298763885028839e-05,
      "loss": 0.5141,
      "step": 845
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.2670369542440758,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5196,
      "step": 850
    },
    {
      "epoch": 0.7975746268656716,
      "grad_norm": 0.262789307862455,
      "learning_rate": 9.859423864979441e-06,
      "loss": 0.5081,
      "step": 855
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.28509078304386815,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.5272,
      "step": 860
    },
    {
      "epoch": 0.8069029850746269,
      "grad_norm": 0.30907587460324537,
      "learning_rate": 9.436882808440334e-06,
      "loss": 0.5285,
      "step": 865
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.26048833998062026,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.5105,
      "step": 870
    },
    {
      "epoch": 0.816231343283582,
      "grad_norm": 0.26919840136919315,
      "learning_rate": 9.031543097495638e-06,
      "loss": 0.515,
      "step": 875
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.2992648125729761,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5131,
      "step": 880
    },
    {
      "epoch": 0.8255597014925373,
      "grad_norm": 0.2723483896878325,
      "learning_rate": 8.6437907335413e-06,
      "loss": 0.5231,
      "step": 885
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.262308992285445,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5253,
      "step": 890
    },
    {
      "epoch": 0.8348880597014925,
      "grad_norm": 0.27940346598038684,
      "learning_rate": 8.273994969699394e-06,
      "loss": 0.5195,
      "step": 895
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.2573250994947132,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.5224,
      "step": 900
    },
    {
      "epoch": 0.8442164179104478,
      "grad_norm": 0.2746489902348466,
      "learning_rate": 7.922507959181673e-06,
      "loss": 0.5107,
      "step": 905
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 0.2546718927089244,
      "learning_rate": 7.753735525674059e-06,
      "loss": 0.5169,
      "step": 910
    },
    {
      "epoch": 0.8535447761194029,
      "grad_norm": 0.3128597706256909,
      "learning_rate": 7.58966441993719e-06,
      "loss": 0.5135,
      "step": 915
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.29147238348317533,
      "learning_rate": 7.430333705159286e-06,
      "loss": 0.5346,
      "step": 920
    },
    {
      "epoch": 0.8628731343283582,
      "grad_norm": 0.3039841418661845,
      "learning_rate": 7.275781315903374e-06,
      "loss": 0.533,
      "step": 925
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 0.26748082202802587,
      "learning_rate": 7.126044049075548e-06,
      "loss": 0.5244,
      "step": 930
    },
    {
      "epoch": 0.8722014925373134,
      "grad_norm": 0.2574220292531603,
      "learning_rate": 6.9811575551641224e-06,
      "loss": 0.512,
      "step": 935
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.2697051023022897,
      "learning_rate": 6.8411563297516995e-06,
      "loss": 0.506,
      "step": 940
    },
    {
      "epoch": 0.8815298507462687,
      "grad_norm": 0.2727834127820519,
      "learning_rate": 6.706073705302254e-06,
      "loss": 0.5175,
      "step": 945
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 0.2726393047900149,
      "learning_rate": 6.575941843225068e-06,
      "loss": 0.5078,
      "step": 950
    },
    {
      "epoch": 0.8908582089552238,
      "grad_norm": 0.25953360021866434,
      "learning_rate": 6.450791726217538e-06,
      "loss": 0.5153,
      "step": 955
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.25309315963307893,
      "learning_rate": 6.330653150888617e-06,
      "loss": 0.519,
      "step": 960
    },
    {
      "epoch": 0.9001865671641791,
      "grad_norm": 0.25585840582334424,
      "learning_rate": 6.215554720664598e-06,
      "loss": 0.5251,
      "step": 965
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 0.2710531901317892,
      "learning_rate": 6.105523838979022e-06,
      "loss": 0.525,
      "step": 970
    },
    {
      "epoch": 0.9095149253731343,
      "grad_norm": 0.2602937977745067,
      "learning_rate": 6.000586702748301e-06,
      "loss": 0.5108,
      "step": 975
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.254290516488359,
      "learning_rate": 5.900768296134551e-06,
      "loss": 0.5125,
      "step": 980
    },
    {
      "epoch": 0.9188432835820896,
      "grad_norm": 0.2734352638891298,
      "learning_rate": 5.8060923845971825e-06,
      "loss": 0.5206,
      "step": 985
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 0.24782218153928054,
      "learning_rate": 5.7165815092346825e-06,
      "loss": 0.5065,
      "step": 990
    },
    {
      "epoch": 0.9281716417910447,
      "grad_norm": 0.32481746469443235,
      "learning_rate": 5.632256981417845e-06,
      "loss": 0.5148,
      "step": 995
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.26237184302764716,
      "learning_rate": 5.553138877715833e-06,
      "loss": 0.5168,
      "step": 1000
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.24993877270733902,
      "learning_rate": 5.479246035116201e-06,
      "loss": 0.5096,
      "step": 1005
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 0.2577951338789917,
      "learning_rate": 5.410596046540051e-06,
      "loss": 0.5119,
      "step": 1010
    },
    {
      "epoch": 0.9468283582089553,
      "grad_norm": 0.27362141957349273,
      "learning_rate": 5.347205256653387e-06,
      "loss": 0.5146,
      "step": 1015
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.284901166454586,
      "learning_rate": 5.28908875797568e-06,
      "loss": 0.5078,
      "step": 1020
    },
    {
      "epoch": 0.9561567164179104,
      "grad_norm": 0.2547318958730913,
      "learning_rate": 5.236260387286509e-06,
      "loss": 0.514,
      "step": 1025
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 0.25616709172231145,
      "learning_rate": 5.1887327223312296e-06,
      "loss": 0.5108,
      "step": 1030
    },
    {
      "epoch": 0.9654850746268657,
      "grad_norm": 0.24877070884034994,
      "learning_rate": 5.1465170788263595e-06,
      "loss": 0.5187,
      "step": 1035
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.26291431270775706,
      "learning_rate": 5.109623507765466e-06,
      "loss": 0.5143,
      "step": 1040
    },
    {
      "epoch": 0.9748134328358209,
      "grad_norm": 0.26168106195147633,
      "learning_rate": 5.07806079302615e-06,
      "loss": 0.5214,
      "step": 1045
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 0.2525330885754495,
      "learning_rate": 5.051836449278715e-06,
      "loss": 0.5161,
      "step": 1050
    },
    {
      "epoch": 0.9841417910447762,
      "grad_norm": 0.26797294104450187,
      "learning_rate": 5.030956720197035e-06,
      "loss": 0.5086,
      "step": 1055
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.2522869759491293,
      "learning_rate": 5.015426576972003e-06,
      "loss": 0.5263,
      "step": 1060
    },
    {
      "epoch": 0.9934701492537313,
      "grad_norm": 0.27283752347105966,
      "learning_rate": 5.005249717127964e-06,
      "loss": 0.5173,
      "step": 1065
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 0.26831346058816563,
      "learning_rate": 5.000428563642382e-06,
      "loss": 0.5177,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "step": 1072,
      "total_flos": 488507298545664.0,
      "train_loss": 0.5501598766935405,
      "train_runtime": 34536.877,
      "train_samples_per_second": 0.993,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 5,
  "max_steps": 1072,
  "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": 488507298545664.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}