{
  "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": 2.0296592712402344,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.8606,
      "step": 5
    },
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.3905454874038696,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8093,
      "step": 10
    },
    {
      "epoch": 0.013992537313432836,
      "grad_norm": 1.113548755645752,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.77,
      "step": 15
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 0.6648120880126953,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.7113,
      "step": 20
    },
    {
      "epoch": 0.02332089552238806,
      "grad_norm": 0.5477202534675598,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.692,
      "step": 25
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.4421328008174896,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.6771,
      "step": 30
    },
    {
      "epoch": 0.03264925373134328,
      "grad_norm": 0.41744956374168396,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.6643,
      "step": 35
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.42775049805641174,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6361,
      "step": 40
    },
    {
      "epoch": 0.04197761194029851,
      "grad_norm": 0.4391756057739258,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.6508,
      "step": 45
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.4183003604412079,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.6313,
      "step": 50
    },
    {
      "epoch": 0.051305970149253734,
      "grad_norm": 0.37660548090934753,
      "learning_rate": 4.999989285883431e-05,
      "loss": 0.6322,
      "step": 55
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.4807550013065338,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6349,
      "step": 60
    },
    {
      "epoch": 0.06063432835820896,
      "grad_norm": 0.5458044409751892,
      "learning_rate": 4.99870371535606e-05,
      "loss": 0.6033,
      "step": 65
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.45444127917289734,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.6164,
      "step": 70
    },
    {
      "epoch": 0.06996268656716417,
      "grad_norm": 0.5141887068748474,
      "learning_rate": 4.9952767243121146e-05,
      "loss": 0.5996,
      "step": 75
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.4654217064380646,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.6155,
      "step": 80
    },
    {
      "epoch": 0.07929104477611941,
      "grad_norm": 0.5317748785018921,
      "learning_rate": 4.989711576244639e-05,
      "loss": 0.5991,
      "step": 85
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.49785587191581726,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.6001,
      "step": 90
    },
    {
      "epoch": 0.08861940298507463,
      "grad_norm": 0.4556206464767456,
      "learning_rate": 4.9820135707945634e-05,
      "loss": 0.6039,
      "step": 95
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.4123242199420929,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.5986,
      "step": 100
    },
    {
      "epoch": 0.09794776119402986,
      "grad_norm": 0.5058124661445618,
      "learning_rate": 4.972190038703905e-05,
      "loss": 0.5939,
      "step": 105
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.4481053948402405,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.5974,
      "step": 110
    },
    {
      "epoch": 0.10727611940298508,
      "grad_norm": 0.4899747967720032,
      "learning_rate": 4.9602503348347625e-05,
      "loss": 0.6127,
      "step": 115
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.44825080037117004,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.5927,
      "step": 120
    },
    {
      "epoch": 0.1166044776119403,
      "grad_norm": 0.3983883857727051,
      "learning_rate": 4.9462058292607735e-05,
      "loss": 0.5907,
      "step": 125
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.39468082785606384,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.5798,
      "step": 130
    },
    {
      "epoch": 0.1259328358208955,
      "grad_norm": 0.4177343547344208,
      "learning_rate": 4.930069896439485e-05,
      "loss": 0.5832,
      "step": 135
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.3727002739906311,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.5839,
      "step": 140
    },
    {
      "epoch": 0.13526119402985073,
      "grad_norm": 0.4837956726551056,
      "learning_rate": 4.9118579024759854e-05,
      "loss": 0.5715,
      "step": 145
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.41040709614753723,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.6,
      "step": 150
    },
    {
      "epoch": 0.14458955223880596,
      "grad_norm": 0.40860089659690857,
      "learning_rate": 4.891587190489891e-05,
      "loss": 0.5902,
      "step": 155
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.45564886927604675,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.5737,
      "step": 160
    },
    {
      "epoch": 0.15391791044776118,
      "grad_norm": 0.349163293838501,
      "learning_rate": 4.869277064099654e-05,
      "loss": 0.5882,
      "step": 165
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.3984043002128601,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.5899,
      "step": 170
    },
    {
      "epoch": 0.16324626865671643,
      "grad_norm": 0.4434589147567749,
      "learning_rate": 4.844948769039896e-05,
      "loss": 0.592,
      "step": 175
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.4797697365283966,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.5755,
      "step": 180
    },
    {
      "epoch": 0.17257462686567165,
      "grad_norm": 0.4126896262168884,
      "learning_rate": 4.818625472929286e-05,
      "loss": 0.5794,
      "step": 185
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.3637039363384247,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.5907,
      "step": 190
    },
    {
      "epoch": 0.18190298507462688,
      "grad_norm": 0.4578976631164551,
      "learning_rate": 4.7903322432082185e-05,
      "loss": 0.5794,
      "step": 195
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.37413689494132996,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5717,
      "step": 200
    },
    {
      "epoch": 0.1912313432835821,
      "grad_norm": 0.35682445764541626,
      "learning_rate": 4.760096023267322e-05,
      "loss": 0.5606,
      "step": 205
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.37927520275115967,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5622,
      "step": 210
    },
    {
      "epoch": 0.20055970149253732,
      "grad_norm": 0.37045177817344666,
      "learning_rate": 4.7279456067895e-05,
      "loss": 0.5713,
      "step": 215
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.4183138608932495,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.5654,
      "step": 220
    },
    {
      "epoch": 0.20988805970149255,
      "grad_norm": 0.3651997447013855,
      "learning_rate": 4.6939116103299655e-05,
      "loss": 0.5666,
      "step": 225
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.4040168821811676,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5673,
      "step": 230
    },
    {
      "epoch": 0.21921641791044777,
      "grad_norm": 0.43385499715805054,
      "learning_rate": 4.6580264441603724e-05,
      "loss": 0.5687,
      "step": 235
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.4357352554798126,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.5645,
      "step": 240
    },
    {
      "epoch": 0.228544776119403,
      "grad_norm": 0.38218751549720764,
      "learning_rate": 4.6203242814047946e-05,
      "loss": 0.5556,
      "step": 245
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.3532406687736511,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5796,
      "step": 250
    },
    {
      "epoch": 0.23787313432835822,
      "grad_norm": 0.36546623706817627,
      "learning_rate": 4.580841025496974e-05,
      "loss": 0.563,
      "step": 255
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.3052341639995575,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.561,
      "step": 260
    },
    {
      "epoch": 0.24720149253731344,
      "grad_norm": 0.34654292464256287,
      "learning_rate": 4.539614275989793e-05,
      "loss": 0.5704,
      "step": 265
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.41322067379951477,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.5633,
      "step": 270
    },
    {
      "epoch": 0.25652985074626866,
      "grad_norm": 0.39780569076538086,
      "learning_rate": 4.496683292749555e-05,
      "loss": 0.5739,
      "step": 275
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.38549038767814636,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5583,
      "step": 280
    },
    {
      "epoch": 0.2658582089552239,
      "grad_norm": 0.39039453864097595,
      "learning_rate": 4.4520889585691705e-05,
      "loss": 0.565,
      "step": 285
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.3481265604496002,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.5558,
      "step": 290
    },
    {
      "epoch": 0.2751865671641791,
      "grad_norm": 0.34713584184646606,
      "learning_rate": 4.4058737402358295e-05,
      "loss": 0.5573,
      "step": 295
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.32922685146331787,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5578,
      "step": 300
    },
    {
      "epoch": 0.28451492537313433,
      "grad_norm": 0.333051472902298,
      "learning_rate": 4.3580816480902656e-05,
      "loss": 0.5604,
      "step": 305
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.2980387508869171,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5515,
      "step": 310
    },
    {
      "epoch": 0.29384328358208955,
      "grad_norm": 0.31041234731674194,
      "learning_rate": 4.308758194116094e-05,
      "loss": 0.5559,
      "step": 315
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.3864513635635376,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5625,
      "step": 320
    },
    {
      "epoch": 0.3031716417910448,
      "grad_norm": 0.3100121319293976,
      "learning_rate": 4.2579503485991567e-05,
      "loss": 0.5433,
      "step": 325
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.3470633327960968,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5619,
      "step": 330
    },
    {
      "epoch": 0.3125,
      "grad_norm": 0.3305334448814392,
      "learning_rate": 4.205706495398143e-05,
      "loss": 0.555,
      "step": 335
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.32060301303863525,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.5642,
      "step": 340
    },
    {
      "epoch": 0.3218283582089552,
      "grad_norm": 0.2954115569591522,
      "learning_rate": 4.1520763858690644e-05,
      "loss": 0.5454,
      "step": 345
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.30578792095184326,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.5503,
      "step": 350
    },
    {
      "epoch": 0.33115671641791045,
      "grad_norm": 0.3552355468273163,
      "learning_rate": 4.097111091487486e-05,
      "loss": 0.5441,
      "step": 355
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.3189811706542969,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.5338,
      "step": 360
    },
    {
      "epoch": 0.34048507462686567,
      "grad_norm": 0.34771403670310974,
      "learning_rate": 4.040862955213615e-05,
      "loss": 0.5515,
      "step": 365
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.3125436007976532,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5639,
      "step": 370
    },
    {
      "epoch": 0.3498134328358209,
      "grad_norm": 0.3404907286167145,
      "learning_rate": 3.9833855416465624e-05,
      "loss": 0.5442,
      "step": 375
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.3328476548194885,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5501,
      "step": 380
    },
    {
      "epoch": 0.3591417910447761,
      "grad_norm": 0.3313455581665039,
      "learning_rate": 3.924733586015257e-05,
      "loss": 0.5389,
      "step": 385
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.30594876408576965,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.5541,
      "step": 390
    },
    {
      "epoch": 0.36847014925373134,
      "grad_norm": 0.309085875749588,
      "learning_rate": 3.864962942054572e-05,
      "loss": 0.5479,
      "step": 395
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.3359319269657135,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.5558,
      "step": 400
    },
    {
      "epoch": 0.37779850746268656,
      "grad_norm": 0.3179395794868469,
      "learning_rate": 3.804130528816312e-05,
      "loss": 0.5611,
      "step": 405
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.338645875453949,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.5427,
      "step": 410
    },
    {
      "epoch": 0.3871268656716418,
      "grad_norm": 0.35845276713371277,
      "learning_rate": 3.7422942764657054e-05,
      "loss": 0.5453,
      "step": 415
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.2944061756134033,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.5395,
      "step": 420
    },
    {
      "epoch": 0.396455223880597,
      "grad_norm": 0.36072415113449097,
      "learning_rate": 3.679513071115025e-05,
      "loss": 0.5355,
      "step": 425
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.3697458803653717,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5407,
      "step": 430
    },
    {
      "epoch": 0.40578358208955223,
      "grad_norm": 0.3033182919025421,
      "learning_rate": 3.615846698746869e-05,
      "loss": 0.5456,
      "step": 435
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.2976251244544983,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5473,
      "step": 440
    },
    {
      "epoch": 0.41511194029850745,
      "grad_norm": 0.32828009128570557,
      "learning_rate": 3.5513557882805e-05,
      "loss": 0.5488,
      "step": 445
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.29798197746276855,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5257,
      "step": 450
    },
    {
      "epoch": 0.4244402985074627,
      "grad_norm": 0.3206273317337036,
      "learning_rate": 3.486101753835468e-05,
      "loss": 0.5481,
      "step": 455
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.30574238300323486,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5437,
      "step": 460
    },
    {
      "epoch": 0.4337686567164179,
      "grad_norm": 0.3001498878002167,
      "learning_rate": 3.420146736247487e-05,
      "loss": 0.5539,
      "step": 465
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.33013975620269775,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.5405,
      "step": 470
    },
    {
      "epoch": 0.4430970149253731,
      "grad_norm": 0.2987500727176666,
      "learning_rate": 3.353553543892277e-05,
      "loss": 0.5267,
      "step": 475
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.292449951171875,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.5483,
      "step": 480
    },
    {
      "epoch": 0.45242537313432835,
      "grad_norm": 0.3454206585884094,
      "learning_rate": 3.2863855928737026e-05,
      "loss": 0.5369,
      "step": 485
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.3288365304470062,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.5547,
      "step": 490
    },
    {
      "epoch": 0.46175373134328357,
      "grad_norm": 0.29698145389556885,
      "learning_rate": 3.218706846633183e-05,
      "loss": 0.5277,
      "step": 495
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.2867559790611267,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5431,
      "step": 500
    },
    {
      "epoch": 0.4710820895522388,
      "grad_norm": 0.2875625789165497,
      "learning_rate": 3.150581755037877e-05,
      "loss": 0.5293,
      "step": 505
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.3133365511894226,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.544,
      "step": 510
    },
    {
      "epoch": 0.480410447761194,
      "grad_norm": 0.31600460410118103,
      "learning_rate": 3.08207519300565e-05,
      "loss": 0.5527,
      "step": 515
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.3297214210033417,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.5476,
      "step": 520
    },
    {
      "epoch": 0.48973880597014924,
      "grad_norm": 0.3018137812614441,
      "learning_rate": 3.0132523987252658e-05,
      "loss": 0.5303,
      "step": 525
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.32776540517807007,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5346,
      "step": 530
    },
    {
      "epoch": 0.49906716417910446,
      "grad_norm": 0.3238219618797302,
      "learning_rate": 2.9441789115306402e-05,
      "loss": 0.5257,
      "step": 535
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.38174256682395935,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5345,
      "step": 540
    },
    {
      "epoch": 0.5083955223880597,
      "grad_norm": 0.2929813265800476,
      "learning_rate": 2.874920509488319e-05,
      "loss": 0.5337,
      "step": 545
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.2884758710861206,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.5286,
      "step": 550
    },
    {
      "epoch": 0.5177238805970149,
      "grad_norm": 0.3295556902885437,
      "learning_rate": 2.8055431467576106e-05,
      "loss": 0.5405,
      "step": 555
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.28779491782188416,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.5321,
      "step": 560
    },
    {
      "epoch": 0.5270522388059702,
      "grad_norm": 0.2856288254261017,
      "learning_rate": 2.7361128907830253e-05,
      "loss": 0.5345,
      "step": 565
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.2876598536968231,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5247,
      "step": 570
    },
    {
      "epoch": 0.5363805970149254,
      "grad_norm": 0.30823373794555664,
      "learning_rate": 2.6666958593788405e-05,
      "loss": 0.5263,
      "step": 575
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.2999833822250366,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.525,
      "step": 580
    },
    {
      "epoch": 0.5457089552238806,
      "grad_norm": 0.2924771010875702,
      "learning_rate": 2.597358157765692e-05,
      "loss": 0.5275,
      "step": 585
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.2781645357608795,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.5307,
      "step": 590
    },
    {
      "epoch": 0.5550373134328358,
      "grad_norm": 0.31060436367988586,
      "learning_rate": 2.528165815619162e-05,
      "loss": 0.525,
      "step": 595
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.2785758674144745,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5414,
      "step": 600
    },
    {
      "epoch": 0.5643656716417911,
      "grad_norm": 0.2810611426830292,
      "learning_rate": 2.459184724190308e-05,
      "loss": 0.5305,
      "step": 605
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.310566246509552,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5308,
      "step": 610
    },
    {
      "epoch": 0.5736940298507462,
      "grad_norm": 0.2734331786632538,
      "learning_rate": 2.390480573558012e-05,
      "loss": 0.5191,
      "step": 615
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.28134024143218994,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5218,
      "step": 620
    },
    {
      "epoch": 0.5830223880597015,
      "grad_norm": 0.28448307514190674,
      "learning_rate": 2.3221187900729003e-05,
      "loss": 0.5245,
      "step": 625
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.2660384476184845,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.5387,
      "step": 630
    },
    {
      "epoch": 0.5923507462686567,
      "grad_norm": 0.28395891189575195,
      "learning_rate": 2.254164474052416e-05,
      "loss": 0.5342,
      "step": 635
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.2656669616699219,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5337,
      "step": 640
    },
    {
      "epoch": 0.601679104477612,
      "grad_norm": 0.2691911458969116,
      "learning_rate": 2.186682337786365e-05,
      "loss": 0.5273,
      "step": 645
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.2647033929824829,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5311,
      "step": 650
    },
    {
      "epoch": 0.6110074626865671,
      "grad_norm": 0.2803230881690979,
      "learning_rate": 2.119736643911979e-05,
      "loss": 0.5196,
      "step": 655
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.2749485373497009,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.5412,
      "step": 660
    },
    {
      "epoch": 0.6203358208955224,
      "grad_norm": 0.27285492420196533,
      "learning_rate": 2.0533911442171805e-05,
      "loss": 0.523,
      "step": 665
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.2703399956226349,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.5156,
      "step": 670
    },
    {
      "epoch": 0.6296641791044776,
      "grad_norm": 0.3498228192329407,
      "learning_rate": 1.9877090189303182e-05,
      "loss": 0.5282,
      "step": 675
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.3206793963909149,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.5193,
      "step": 680
    },
    {
      "epoch": 0.6389925373134329,
      "grad_norm": 0.28730154037475586,
      "learning_rate": 1.922752816554204e-05,
      "loss": 0.5345,
      "step": 685
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.25482338666915894,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.5324,
      "step": 690
    },
    {
      "epoch": 0.648320895522388,
      "grad_norm": 0.2529654800891876,
      "learning_rate": 1.858584394301728e-05,
      "loss": 0.5175,
      "step": 695
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.2732594907283783,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.5206,
      "step": 700
    },
    {
      "epoch": 0.6576492537313433,
      "grad_norm": 0.26254940032958984,
      "learning_rate": 1.7952648591897858e-05,
      "loss": 0.5156,
      "step": 705
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.2806949019432068,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.5325,
      "step": 710
    },
    {
      "epoch": 0.6669776119402985,
      "grad_norm": 0.2589339315891266,
      "learning_rate": 1.7328545098476106e-05,
      "loss": 0.5264,
      "step": 715
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.2677186131477356,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.5286,
      "step": 720
    },
    {
      "epoch": 0.6763059701492538,
      "grad_norm": 0.25420376658439636,
      "learning_rate": 1.671412779094926e-05,
      "loss": 0.5164,
      "step": 725
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.2698172330856323,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.5332,
      "step": 730
    },
    {
      "epoch": 0.6856343283582089,
      "grad_norm": 0.26080092787742615,
      "learning_rate": 1.6109981773446036e-05,
      "loss": 0.5315,
      "step": 735
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.24189084768295288,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.5102,
      "step": 740
    },
    {
      "epoch": 0.6949626865671642,
      "grad_norm": 0.24392467737197876,
      "learning_rate": 1.5516682368837133e-05,
      "loss": 0.5104,
      "step": 745
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.23985299468040466,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.5193,
      "step": 750
    },
    {
      "epoch": 0.7042910447761194,
      "grad_norm": 0.2562742829322815,
      "learning_rate": 1.4934794570860416e-05,
      "loss": 0.5218,
      "step": 755
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.2763282060623169,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5296,
      "step": 760
    },
    {
      "epoch": 0.7136194029850746,
      "grad_norm": 0.24829770624637604,
      "learning_rate": 1.4364872506082425e-05,
      "loss": 0.5325,
      "step": 765
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.256150484085083,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.5161,
      "step": 770
    },
    {
      "epoch": 0.7229477611940298,
      "grad_norm": 0.2693561017513275,
      "learning_rate": 1.3807458906208546e-05,
      "loss": 0.5315,
      "step": 775
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.2647959887981415,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5149,
      "step": 780
    },
    {
      "epoch": 0.7322761194029851,
      "grad_norm": 0.24539123475551605,
      "learning_rate": 1.326308459124447e-05,
      "loss": 0.5156,
      "step": 785
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.2749805152416229,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.5323,
      "step": 790
    },
    {
      "epoch": 0.7416044776119403,
      "grad_norm": 0.2615833580493927,
      "learning_rate": 1.2732267964001033e-05,
      "loss": 0.5284,
      "step": 795
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.25320667028427124,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5185,
      "step": 800
    },
    {
      "epoch": 0.7509328358208955,
      "grad_norm": 0.261538028717041,
      "learning_rate": 1.2215514516423813e-05,
      "loss": 0.5222,
      "step": 805
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.25271981954574585,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.5202,
      "step": 810
    },
    {
      "epoch": 0.7602611940298507,
      "grad_norm": 0.24026913940906525,
      "learning_rate": 1.1713316348217673e-05,
      "loss": 0.5095,
      "step": 815
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.24491222202777863,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.5137,
      "step": 820
    },
    {
      "epoch": 0.769589552238806,
      "grad_norm": 0.2599160969257355,
      "learning_rate": 1.1226151698224597e-05,
      "loss": 0.5271,
      "step": 825
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.2577415108680725,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.526,
      "step": 830
    },
    {
      "epoch": 0.7789179104477612,
      "grad_norm": 0.23763109743595123,
      "learning_rate": 1.0754484489001085e-05,
      "loss": 0.5025,
      "step": 835
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.24150507152080536,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.518,
      "step": 840
    },
    {
      "epoch": 0.7882462686567164,
      "grad_norm": 0.2546503245830536,
      "learning_rate": 1.0298763885028839e-05,
      "loss": 0.521,
      "step": 845
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.2274450659751892,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5169,
      "step": 850
    },
    {
      "epoch": 0.7975746268656716,
      "grad_norm": 0.23724405467510223,
      "learning_rate": 9.859423864979441e-06,
      "loss": 0.5298,
      "step": 855
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.24516107141971588,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.5117,
      "step": 860
    },
    {
      "epoch": 0.8069029850746269,
      "grad_norm": 0.2544306516647339,
      "learning_rate": 9.436882808440334e-06,
      "loss": 0.5131,
      "step": 865
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.25421103835105896,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.5168,
      "step": 870
    },
    {
      "epoch": 0.816231343283582,
      "grad_norm": 0.24098147451877594,
      "learning_rate": 9.031543097495638e-06,
      "loss": 0.5127,
      "step": 875
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.23913885653018951,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5214,
      "step": 880
    },
    {
      "epoch": 0.8255597014925373,
      "grad_norm": 0.24211934208869934,
      "learning_rate": 8.6437907335413e-06,
      "loss": 0.5254,
      "step": 885
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.23977182805538177,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5241,
      "step": 890
    },
    {
      "epoch": 0.8348880597014925,
      "grad_norm": 0.23448112607002258,
      "learning_rate": 8.273994969699394e-06,
      "loss": 0.5127,
      "step": 895
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.2343115657567978,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.5135,
      "step": 900
    },
    {
      "epoch": 0.8442164179104478,
      "grad_norm": 0.23262952268123627,
      "learning_rate": 7.922507959181673e-06,
      "loss": 0.5342,
      "step": 905
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 0.24626874923706055,
      "learning_rate": 7.753735525674059e-06,
      "loss": 0.5305,
      "step": 910
    },
    {
      "epoch": 0.8535447761194029,
      "grad_norm": 0.24683545529842377,
      "learning_rate": 7.58966441993719e-06,
      "loss": 0.5297,
      "step": 915
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.23083548247814178,
      "learning_rate": 7.430333705159286e-06,
      "loss": 0.5243,
      "step": 920
    },
    {
      "epoch": 0.8628731343283582,
      "grad_norm": 0.24129389226436615,
      "learning_rate": 7.275781315903374e-06,
      "loss": 0.5109,
      "step": 925
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 0.25144246220588684,
      "learning_rate": 7.126044049075548e-06,
      "loss": 0.5152,
      "step": 930
    },
    {
      "epoch": 0.8722014925373134,
      "grad_norm": 0.24448096752166748,
      "learning_rate": 6.9811575551641224e-06,
      "loss": 0.5037,
      "step": 935
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.23240606486797333,
      "learning_rate": 6.8411563297516995e-06,
      "loss": 0.5062,
      "step": 940
    },
    {
      "epoch": 0.8815298507462687,
      "grad_norm": 0.2471693903207779,
      "learning_rate": 6.706073705302254e-06,
      "loss": 0.516,
      "step": 945
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 0.25190070271492004,
      "learning_rate": 6.575941843225068e-06,
      "loss": 0.522,
      "step": 950
    },
    {
      "epoch": 0.8908582089552238,
      "grad_norm": 0.25332111120224,
      "learning_rate": 6.450791726217538e-06,
      "loss": 0.5173,
      "step": 955
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.23512767255306244,
      "learning_rate": 6.330653150888617e-06,
      "loss": 0.5172,
      "step": 960
    },
    {
      "epoch": 0.9001865671641791,
      "grad_norm": 0.2349744737148285,
      "learning_rate": 6.215554720664598e-06,
      "loss": 0.5134,
      "step": 965
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 0.2327372282743454,
      "learning_rate": 6.105523838979022e-06,
      "loss": 0.5214,
      "step": 970
    },
    {
      "epoch": 0.9095149253731343,
      "grad_norm": 0.23093824088573456,
      "learning_rate": 6.000586702748301e-06,
      "loss": 0.5011,
      "step": 975
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.2209603190422058,
      "learning_rate": 5.900768296134551e-06,
      "loss": 0.5048,
      "step": 980
    },
    {
      "epoch": 0.9188432835820896,
      "grad_norm": 0.2304598093032837,
      "learning_rate": 5.8060923845971825e-06,
      "loss": 0.5192,
      "step": 985
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 0.23641374707221985,
      "learning_rate": 5.7165815092346825e-06,
      "loss": 0.5014,
      "step": 990
    },
    {
      "epoch": 0.9281716417910447,
      "grad_norm": 0.25322774052619934,
      "learning_rate": 5.632256981417845e-06,
      "loss": 0.5247,
      "step": 995
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.226620152592659,
      "learning_rate": 5.553138877715833e-06,
      "loss": 0.5086,
      "step": 1000
    },
    {
      "epoch": 0.9375,
      "grad_norm": 0.23227237164974213,
      "learning_rate": 5.479246035116201e-06,
      "loss": 0.5078,
      "step": 1005
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 0.23838132619857788,
      "learning_rate": 5.410596046540051e-06,
      "loss": 0.5263,
      "step": 1010
    },
    {
      "epoch": 0.9468283582089553,
      "grad_norm": 0.2337532490491867,
      "learning_rate": 5.347205256653387e-06,
      "loss": 0.5168,
      "step": 1015
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.24512432515621185,
      "learning_rate": 5.28908875797568e-06,
      "loss": 0.5092,
      "step": 1020
    },
    {
      "epoch": 0.9561567164179104,
      "grad_norm": 0.24788177013397217,
      "learning_rate": 5.236260387286509e-06,
      "loss": 0.5251,
      "step": 1025
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 0.24356013536453247,
      "learning_rate": 5.1887327223312296e-06,
      "loss": 0.5109,
      "step": 1030
    },
    {
      "epoch": 0.9654850746268657,
      "grad_norm": 0.2298063486814499,
      "learning_rate": 5.1465170788263595e-06,
      "loss": 0.5288,
      "step": 1035
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.21959343552589417,
      "learning_rate": 5.109623507765466e-06,
      "loss": 0.506,
      "step": 1040
    },
    {
      "epoch": 0.9748134328358209,
      "grad_norm": 0.22912734746932983,
      "learning_rate": 5.07806079302615e-06,
      "loss": 0.508,
      "step": 1045
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 0.2325100600719452,
      "learning_rate": 5.051836449278715e-06,
      "loss": 0.5001,
      "step": 1050
    },
    {
      "epoch": 0.9841417910447762,
      "grad_norm": 0.2450973242521286,
      "learning_rate": 5.030956720197035e-06,
      "loss": 0.5107,
      "step": 1055
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.22343529760837555,
      "learning_rate": 5.015426576972003e-06,
      "loss": 0.509,
      "step": 1060
    },
    {
      "epoch": 0.9934701492537313,
      "grad_norm": 0.23454374074935913,
      "learning_rate": 5.005249717127964e-06,
      "loss": 0.5137,
      "step": 1065
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 0.23174691200256348,
      "learning_rate": 5.000428563642382e-06,
      "loss": 0.5,
      "step": 1070
    },
    {
      "epoch": 1.0,
      "step": 1072,
      "total_flos": 4.418757235321078e+18,
      "train_loss": 0.5497396646150902,
      "train_runtime": 11730.1118,
      "train_samples_per_second": 2.924,
      "train_steps_per_second": 0.091
    }
  ],
  "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": 4.418757235321078e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}