{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 536,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.009328358208955223,
      "grad_norm": 1.4529272838242249,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8491,
      "step": 5
    },
    {
      "epoch": 0.018656716417910446,
      "grad_norm": 1.1689508861038835,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.8005,
      "step": 10
    },
    {
      "epoch": 0.027985074626865673,
      "grad_norm": 0.7491351005527522,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.7315,
      "step": 15
    },
    {
      "epoch": 0.03731343283582089,
      "grad_norm": 0.46441624130037606,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.6803,
      "step": 20
    },
    {
      "epoch": 0.04664179104477612,
      "grad_norm": 0.4634226186946367,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.6577,
      "step": 25
    },
    {
      "epoch": 0.055970149253731345,
      "grad_norm": 0.4030963463525461,
      "learning_rate": 4.999614302517356e-05,
      "loss": 0.6581,
      "step": 30
    },
    {
      "epoch": 0.06529850746268656,
      "grad_norm": 0.39406929868265494,
      "learning_rate": 4.997257741198456e-05,
      "loss": 0.6419,
      "step": 35
    },
    {
      "epoch": 0.07462686567164178,
      "grad_norm": 0.31318784447485654,
      "learning_rate": 4.992761136351291e-05,
      "loss": 0.627,
      "step": 40
    },
    {
      "epoch": 0.08395522388059702,
      "grad_norm": 0.3353753917281023,
      "learning_rate": 4.986128770052603e-05,
      "loss": 0.6084,
      "step": 45
    },
    {
      "epoch": 0.09328358208955224,
      "grad_norm": 0.2954075007188722,
      "learning_rate": 4.9773669582457364e-05,
      "loss": 0.61,
      "step": 50
    },
    {
      "epoch": 0.10261194029850747,
      "grad_norm": 0.3692314525331676,
      "learning_rate": 4.966484044726024e-05,
      "loss": 0.6137,
      "step": 55
    },
    {
      "epoch": 0.11194029850746269,
      "grad_norm": 0.3600438394484195,
      "learning_rate": 4.953490393195063e-05,
      "loss": 0.6004,
      "step": 60
    },
    {
      "epoch": 0.12126865671641791,
      "grad_norm": 0.39009463063899535,
      "learning_rate": 4.938398377391461e-05,
      "loss": 0.5973,
      "step": 65
    },
    {
      "epoch": 0.13059701492537312,
      "grad_norm": 0.5329218771045787,
      "learning_rate": 4.921222369307427e-05,
      "loss": 0.594,
      "step": 70
    },
    {
      "epoch": 0.13992537313432835,
      "grad_norm": 0.46575182507115465,
      "learning_rate": 4.901978725502454e-05,
      "loss": 0.5895,
      "step": 75
    },
    {
      "epoch": 0.14925373134328357,
      "grad_norm": 0.39800373886063,
      "learning_rate": 4.880685771527114e-05,
      "loss": 0.5895,
      "step": 80
    },
    {
      "epoch": 0.15858208955223882,
      "grad_norm": 0.42200587253551747,
      "learning_rate": 4.8573637844718e-05,
      "loss": 0.5944,
      "step": 85
    },
    {
      "epoch": 0.16791044776119404,
      "grad_norm": 0.4495424363631174,
      "learning_rate": 4.83203497365703e-05,
      "loss": 0.5911,
      "step": 90
    },
    {
      "epoch": 0.17723880597014927,
      "grad_norm": 0.46522791184194295,
      "learning_rate": 4.8047234594837143e-05,
      "loss": 0.5782,
      "step": 95
    },
    {
      "epoch": 0.1865671641791045,
      "grad_norm": 0.4425302293891964,
      "learning_rate": 4.775455250463507e-05,
      "loss": 0.5749,
      "step": 100
    },
    {
      "epoch": 0.1958955223880597,
      "grad_norm": 0.40683117962693205,
      "learning_rate": 4.744258218451135e-05,
      "loss": 0.5848,
      "step": 105
    },
    {
      "epoch": 0.20522388059701493,
      "grad_norm": 0.28500611318023544,
      "learning_rate": 4.71116207210228e-05,
      "loss": 0.576,
      "step": 110
    },
    {
      "epoch": 0.21455223880597016,
      "grad_norm": 0.35259387788620616,
      "learning_rate": 4.676198328582288e-05,
      "loss": 0.5712,
      "step": 115
    },
    {
      "epoch": 0.22388059701492538,
      "grad_norm": 0.3411126136195717,
      "learning_rate": 4.6394002835526535e-05,
      "loss": 0.5822,
      "step": 120
    },
    {
      "epoch": 0.2332089552238806,
      "grad_norm": 0.41571669987803755,
      "learning_rate": 4.6008029794638596e-05,
      "loss": 0.5668,
      "step": 125
    },
    {
      "epoch": 0.24253731343283583,
      "grad_norm": 0.3783814665428634,
      "learning_rate": 4.560443172184763e-05,
      "loss": 0.5695,
      "step": 130
    },
    {
      "epoch": 0.251865671641791,
      "grad_norm": 0.36003096332997486,
      "learning_rate": 4.5183592960003104e-05,
      "loss": 0.562,
      "step": 135
    },
    {
      "epoch": 0.26119402985074625,
      "grad_norm": 0.29625254021632597,
      "learning_rate": 4.4745914270109055e-05,
      "loss": 0.5672,
      "step": 140
    },
    {
      "epoch": 0.27052238805970147,
      "grad_norm": 0.4041052584419407,
      "learning_rate": 4.429181244968301e-05,
      "loss": 0.5665,
      "step": 145
    },
    {
      "epoch": 0.2798507462686567,
      "grad_norm": 0.28581434778217296,
      "learning_rate": 4.38217199358434e-05,
      "loss": 0.5604,
      "step": 150
    },
    {
      "epoch": 0.2891791044776119,
      "grad_norm": 0.4031760526528498,
      "learning_rate": 4.3336084393503545e-05,
      "loss": 0.5573,
      "step": 155
    },
    {
      "epoch": 0.29850746268656714,
      "grad_norm": 0.3787628774367758,
      "learning_rate": 4.283536828906436e-05,
      "loss": 0.5698,
      "step": 160
    },
    {
      "epoch": 0.30783582089552236,
      "grad_norm": 0.4316425611729101,
      "learning_rate": 4.2320048450011684e-05,
      "loss": 0.5631,
      "step": 165
    },
    {
      "epoch": 0.31716417910447764,
      "grad_norm": 0.32889164452115366,
      "learning_rate": 4.179061561083777e-05,
      "loss": 0.555,
      "step": 170
    },
    {
      "epoch": 0.32649253731343286,
      "grad_norm": 0.3304034607225048,
      "learning_rate": 4.124757394571914e-05,
      "loss": 0.5612,
      "step": 175
    },
    {
      "epoch": 0.3358208955223881,
      "grad_norm": 0.28777523778739394,
      "learning_rate": 4.069144058839605e-05,
      "loss": 0.562,
      "step": 180
    },
    {
      "epoch": 0.3451492537313433,
      "grad_norm": 0.3339855615199367,
      "learning_rate": 4.012274513971061e-05,
      "loss": 0.5662,
      "step": 185
    },
    {
      "epoch": 0.35447761194029853,
      "grad_norm": 0.3197899362876164,
      "learning_rate": 3.954202916327264e-05,
      "loss": 0.5544,
      "step": 190
    },
    {
      "epoch": 0.36380597014925375,
      "grad_norm": 0.3250020382984713,
      "learning_rate": 3.894984566973346e-05,
      "loss": 0.5544,
      "step": 195
    },
    {
      "epoch": 0.373134328358209,
      "grad_norm": 0.3329747882916182,
      "learning_rate": 3.834675859015876e-05,
      "loss": 0.5584,
      "step": 200
    },
    {
      "epoch": 0.3824626865671642,
      "grad_norm": 0.2931796751529486,
      "learning_rate": 3.77333422390021e-05,
      "loss": 0.5492,
      "step": 205
    },
    {
      "epoch": 0.3917910447761194,
      "grad_norm": 0.309605999568671,
      "learning_rate": 3.711018076719034e-05,
      "loss": 0.5563,
      "step": 210
    },
    {
      "epoch": 0.40111940298507465,
      "grad_norm": 0.29067978843774794,
      "learning_rate": 3.647786760584194e-05,
      "loss": 0.5606,
      "step": 215
    },
    {
      "epoch": 0.41044776119402987,
      "grad_norm": 0.2720420372815392,
      "learning_rate": 3.583700490114776e-05,
      "loss": 0.5587,
      "step": 220
    },
    {
      "epoch": 0.4197761194029851,
      "grad_norm": 0.3190566242172115,
      "learning_rate": 3.518820294095267e-05,
      "loss": 0.5546,
      "step": 225
    },
    {
      "epoch": 0.4291044776119403,
      "grad_norm": 0.2972056748474866,
      "learning_rate": 3.453207957358377e-05,
      "loss": 0.5465,
      "step": 230
    },
    {
      "epoch": 0.43843283582089554,
      "grad_norm": 0.31648213924949936,
      "learning_rate": 3.386925961947906e-05,
      "loss": 0.5478,
      "step": 235
    },
    {
      "epoch": 0.44776119402985076,
      "grad_norm": 0.3483684766363109,
      "learning_rate": 3.320037427617639e-05,
      "loss": 0.5552,
      "step": 240
    },
    {
      "epoch": 0.457089552238806,
      "grad_norm": 0.2870111399875252,
      "learning_rate": 3.252606051722972e-05,
      "loss": 0.5536,
      "step": 245
    },
    {
      "epoch": 0.4664179104477612,
      "grad_norm": 0.3290811706730747,
      "learning_rate": 3.1846960485624886e-05,
      "loss": 0.5429,
      "step": 250
    },
    {
      "epoch": 0.47574626865671643,
      "grad_norm": 0.29255944971499237,
      "learning_rate": 3.1163720882272516e-05,
      "loss": 0.5506,
      "step": 255
    },
    {
      "epoch": 0.48507462686567165,
      "grad_norm": 0.31439453371777304,
      "learning_rate": 3.047699235016056e-05,
      "loss": 0.5429,
      "step": 260
    },
    {
      "epoch": 0.4944029850746269,
      "grad_norm": 0.3425522743911677,
      "learning_rate": 2.9787428854752736e-05,
      "loss": 0.5368,
      "step": 265
    },
    {
      "epoch": 0.503731343283582,
      "grad_norm": 0.27793747214527936,
      "learning_rate": 2.9095687061223058e-05,
      "loss": 0.5514,
      "step": 270
    },
    {
      "epoch": 0.5130597014925373,
      "grad_norm": 0.3383812023651381,
      "learning_rate": 2.8402425709119435e-05,
      "loss": 0.5506,
      "step": 275
    },
    {
      "epoch": 0.5223880597014925,
      "grad_norm": 0.27903329251231174,
      "learning_rate": 2.7708304985051868e-05,
      "loss": 0.5476,
      "step": 280
    },
    {
      "epoch": 0.5317164179104478,
      "grad_norm": 0.26659760763870816,
      "learning_rate": 2.7013985894002623e-05,
      "loss": 0.5355,
      "step": 285
    },
    {
      "epoch": 0.5410447761194029,
      "grad_norm": 0.27151810632711876,
      "learning_rate": 2.6320129629857093e-05,
      "loss": 0.5483,
      "step": 290
    },
    {
      "epoch": 0.5503731343283582,
      "grad_norm": 0.26109145689554275,
      "learning_rate": 2.56273969457547e-05,
      "loss": 0.5372,
      "step": 295
    },
    {
      "epoch": 0.5597014925373134,
      "grad_norm": 0.25106284404522056,
      "learning_rate": 2.4936447524859625e-05,
      "loss": 0.5415,
      "step": 300
    },
    {
      "epoch": 0.5690298507462687,
      "grad_norm": 0.3274744617644554,
      "learning_rate": 2.4247939352150386e-05,
      "loss": 0.5366,
      "step": 305
    },
    {
      "epoch": 0.5783582089552238,
      "grad_norm": 0.26437036147373966,
      "learning_rate": 2.3562528087826573e-05,
      "loss": 0.5428,
      "step": 310
    },
    {
      "epoch": 0.5876865671641791,
      "grad_norm": 0.2446519338891704,
      "learning_rate": 2.2880866442929544e-05,
      "loss": 0.5398,
      "step": 315
    },
    {
      "epoch": 0.5970149253731343,
      "grad_norm": 0.24911911219120392,
      "learning_rate": 2.2203603557771447e-05,
      "loss": 0.5358,
      "step": 320
    },
    {
      "epoch": 0.6063432835820896,
      "grad_norm": 0.2516922089571222,
      "learning_rate": 2.153138438376473e-05,
      "loss": 0.5341,
      "step": 325
    },
    {
      "epoch": 0.6156716417910447,
      "grad_norm": 0.2412997869982118,
      "learning_rate": 2.0864849069240645e-05,
      "loss": 0.5388,
      "step": 330
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.2565617737795945,
      "learning_rate": 2.0204632349841667e-05,
      "loss": 0.5356,
      "step": 335
    },
    {
      "epoch": 0.6343283582089553,
      "grad_norm": 0.23064002794765903,
      "learning_rate": 1.9551362944068462e-05,
      "loss": 0.5379,
      "step": 340
    },
    {
      "epoch": 0.6436567164179104,
      "grad_norm": 0.24633590530445001,
      "learning_rate": 1.890566295455678e-05,
      "loss": 0.5312,
      "step": 345
    },
    {
      "epoch": 0.6529850746268657,
      "grad_norm": 0.2387974629546731,
      "learning_rate": 1.8268147275654707e-05,
      "loss": 0.5412,
      "step": 350
    },
    {
      "epoch": 0.6623134328358209,
      "grad_norm": 0.23072321595406606,
      "learning_rate": 1.7639423007864252e-05,
      "loss": 0.528,
      "step": 355
    },
    {
      "epoch": 0.6716417910447762,
      "grad_norm": 0.22136194889078148,
      "learning_rate": 1.702008887970491e-05,
      "loss": 0.5288,
      "step": 360
    },
    {
      "epoch": 0.6809701492537313,
      "grad_norm": 0.22646201154385875,
      "learning_rate": 1.6410734677549872e-05,
      "loss": 0.5321,
      "step": 365
    },
    {
      "epoch": 0.6902985074626866,
      "grad_norm": 0.22901452524086416,
      "learning_rate": 1.58119406839777e-05,
      "loss": 0.531,
      "step": 370
    },
    {
      "epoch": 0.6996268656716418,
      "grad_norm": 0.25990616664969085,
      "learning_rate": 1.5224277125174388e-05,
      "loss": 0.534,
      "step": 375
    },
    {
      "epoch": 0.7089552238805971,
      "grad_norm": 0.2294421044982824,
      "learning_rate": 1.464830362791204e-05,
      "loss": 0.5361,
      "step": 380
    },
    {
      "epoch": 0.7182835820895522,
      "grad_norm": 0.2217290229674116,
      "learning_rate": 1.4084568686621314e-05,
      "loss": 0.5384,
      "step": 385
    },
    {
      "epoch": 0.7276119402985075,
      "grad_norm": 0.2162028885642014,
      "learning_rate": 1.3533609141065008e-05,
      "loss": 0.5368,
      "step": 390
    },
    {
      "epoch": 0.7369402985074627,
      "grad_norm": 0.23748115013736273,
      "learning_rate": 1.299594966511038e-05,
      "loss": 0.534,
      "step": 395
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 0.22565894301779696,
      "learning_rate": 1.2472102267086904e-05,
      "loss": 0.5297,
      "step": 400
    },
    {
      "epoch": 0.7555970149253731,
      "grad_norm": 0.21905287170194962,
      "learning_rate": 1.1962565802205255e-05,
      "loss": 0.5364,
      "step": 405
    },
    {
      "epoch": 0.7649253731343284,
      "grad_norm": 0.20940519149092718,
      "learning_rate": 1.1467825497501954e-05,
      "loss": 0.5227,
      "step": 410
    },
    {
      "epoch": 0.7742537313432836,
      "grad_norm": 0.23494185563834197,
      "learning_rate": 1.0988352489762006e-05,
      "loss": 0.5386,
      "step": 415
    },
    {
      "epoch": 0.7835820895522388,
      "grad_norm": 0.21057179681993918,
      "learning_rate": 1.052460337685951e-05,
      "loss": 0.5301,
      "step": 420
    },
    {
      "epoch": 0.792910447761194,
      "grad_norm": 0.2131633170898678,
      "learning_rate": 1.0077019782943584e-05,
      "loss": 0.5283,
      "step": 425
    },
    {
      "epoch": 0.8022388059701493,
      "grad_norm": 0.219677996534737,
      "learning_rate": 9.646027937883622e-06,
      "loss": 0.5292,
      "step": 430
    },
    {
      "epoch": 0.8115671641791045,
      "grad_norm": 0.20532036673430643,
      "learning_rate": 9.232038271374377e-06,
      "loss": 0.5312,
      "step": 435
    },
    {
      "epoch": 0.8208955223880597,
      "grad_norm": 0.2274426668213202,
      "learning_rate": 8.835445022087426e-06,
      "loss": 0.5258,
      "step": 440
    },
    {
      "epoch": 0.8302238805970149,
      "grad_norm": 0.21658031336344372,
      "learning_rate": 8.456625862241193e-06,
      "loss": 0.5361,
      "step": 445
    },
    {
      "epoch": 0.8395522388059702,
      "grad_norm": 0.1969825219644592,
      "learning_rate": 8.095941537947057e-06,
      "loss": 0.533,
      "step": 450
    },
    {
      "epoch": 0.8488805970149254,
      "grad_norm": 0.23076534818899264,
      "learning_rate": 7.753735525674059e-06,
      "loss": 0.5257,
      "step": 455
    },
    {
      "epoch": 0.8582089552238806,
      "grad_norm": 0.20880898976821344,
      "learning_rate": 7.430333705159286e-06,
      "loss": 0.5362,
      "step": 460
    },
    {
      "epoch": 0.8675373134328358,
      "grad_norm": 0.20216937595208878,
      "learning_rate": 7.126044049075548e-06,
      "loss": 0.541,
      "step": 465
    },
    {
      "epoch": 0.8768656716417911,
      "grad_norm": 0.1944238128936159,
      "learning_rate": 6.8411563297516995e-06,
      "loss": 0.521,
      "step": 470
    },
    {
      "epoch": 0.8861940298507462,
      "grad_norm": 0.18302437521460974,
      "learning_rate": 6.575941843225068e-06,
      "loss": 0.5248,
      "step": 475
    },
    {
      "epoch": 0.8955223880597015,
      "grad_norm": 0.2194502098879558,
      "learning_rate": 6.330653150888617e-06,
      "loss": 0.5294,
      "step": 480
    },
    {
      "epoch": 0.9048507462686567,
      "grad_norm": 0.21482900804537108,
      "learning_rate": 6.105523838979022e-06,
      "loss": 0.5375,
      "step": 485
    },
    {
      "epoch": 0.914179104477612,
      "grad_norm": 0.19312573942582062,
      "learning_rate": 5.900768296134551e-06,
      "loss": 0.524,
      "step": 490
    },
    {
      "epoch": 0.9235074626865671,
      "grad_norm": 0.18903344539396477,
      "learning_rate": 5.7165815092346825e-06,
      "loss": 0.5262,
      "step": 495
    },
    {
      "epoch": 0.9328358208955224,
      "grad_norm": 0.1911384927096407,
      "learning_rate": 5.553138877715833e-06,
      "loss": 0.5281,
      "step": 500
    },
    {
      "epoch": 0.9421641791044776,
      "grad_norm": 0.19367838422770262,
      "learning_rate": 5.410596046540051e-06,
      "loss": 0.5231,
      "step": 505
    },
    {
      "epoch": 0.9514925373134329,
      "grad_norm": 0.20829703892741963,
      "learning_rate": 5.28908875797568e-06,
      "loss": 0.5237,
      "step": 510
    },
    {
      "epoch": 0.960820895522388,
      "grad_norm": 0.19163012582670946,
      "learning_rate": 5.1887327223312296e-06,
      "loss": 0.5248,
      "step": 515
    },
    {
      "epoch": 0.9701492537313433,
      "grad_norm": 0.20150691323789874,
      "learning_rate": 5.109623507765466e-06,
      "loss": 0.5291,
      "step": 520
    },
    {
      "epoch": 0.9794776119402985,
      "grad_norm": 0.20723359673688968,
      "learning_rate": 5.051836449278715e-06,
      "loss": 0.5313,
      "step": 525
    },
    {
      "epoch": 0.9888059701492538,
      "grad_norm": 0.20425578544955647,
      "learning_rate": 5.015426576972003e-06,
      "loss": 0.53,
      "step": 530
    },
    {
      "epoch": 0.9981343283582089,
      "grad_norm": 0.18922418280122189,
      "learning_rate": 5.000428563642382e-06,
      "loss": 0.5301,
      "step": 535
    },
    {
      "epoch": 1.0,
      "step": 536,
      "total_flos": 488621249396736.0,
      "train_loss": 0.5612209483758727,
      "train_runtime": 11742.8979,
      "train_samples_per_second": 2.921,
      "train_steps_per_second": 0.046
    }
  ],
  "logging_steps": 5,
  "max_steps": 536,
  "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": 488621249396736.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}