{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 3202,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.006246096189881324,
      "grad_norm": 0.47147321701049805,
      "learning_rate": 4.968769519050593e-05,
      "loss": 4.3946,
      "step": 20
    },
    {
      "epoch": 0.012492192379762648,
      "grad_norm": 1.0567554235458374,
      "learning_rate": 4.937539038101187e-05,
      "loss": 4.4724,
      "step": 40
    },
    {
      "epoch": 0.018738288569643973,
      "grad_norm": 1.647952675819397,
      "learning_rate": 4.90630855715178e-05,
      "loss": 4.162,
      "step": 60
    },
    {
      "epoch": 0.024984384759525295,
      "grad_norm": 1.286374568939209,
      "learning_rate": 4.875078076202374e-05,
      "loss": 3.7767,
      "step": 80
    },
    {
      "epoch": 0.03123048094940662,
      "grad_norm": 1.465066909790039,
      "learning_rate": 4.843847595252967e-05,
      "loss": 3.9672,
      "step": 100
    },
    {
      "epoch": 0.037476577139287946,
      "grad_norm": 1.7631100416183472,
      "learning_rate": 4.81261711430356e-05,
      "loss": 3.6736,
      "step": 120
    },
    {
      "epoch": 0.04372267332916927,
      "grad_norm": 1.5231457948684692,
      "learning_rate": 4.781386633354154e-05,
      "loss": 3.4314,
      "step": 140
    },
    {
      "epoch": 0.04996876951905059,
      "grad_norm": 3.9318079948425293,
      "learning_rate": 4.750156152404747e-05,
      "loss": 3.501,
      "step": 160
    },
    {
      "epoch": 0.056214865708931916,
      "grad_norm": 2.3082265853881836,
      "learning_rate": 4.71892567145534e-05,
      "loss": 3.3423,
      "step": 180
    },
    {
      "epoch": 0.06246096189881324,
      "grad_norm": 2.0189499855041504,
      "learning_rate": 4.6876951905059344e-05,
      "loss": 3.4936,
      "step": 200
    },
    {
      "epoch": 0.06870705808869457,
      "grad_norm": 2.6035380363464355,
      "learning_rate": 4.656464709556527e-05,
      "loss": 2.9782,
      "step": 220
    },
    {
      "epoch": 0.07495315427857589,
      "grad_norm": 1.5224251747131348,
      "learning_rate": 4.625234228607121e-05,
      "loss": 3.0411,
      "step": 240
    },
    {
      "epoch": 0.08119925046845722,
      "grad_norm": 3.759772777557373,
      "learning_rate": 4.5940037476577143e-05,
      "loss": 3.0212,
      "step": 260
    },
    {
      "epoch": 0.08744534665833854,
      "grad_norm": 3.752288341522217,
      "learning_rate": 4.562773266708307e-05,
      "loss": 3.0461,
      "step": 280
    },
    {
      "epoch": 0.09369144284821987,
      "grad_norm": 4.104545593261719,
      "learning_rate": 4.531542785758901e-05,
      "loss": 3.0462,
      "step": 300
    },
    {
      "epoch": 0.09993753903810118,
      "grad_norm": 4.520786762237549,
      "learning_rate": 4.500312304809494e-05,
      "loss": 2.8036,
      "step": 320
    },
    {
      "epoch": 0.1061836352279825,
      "grad_norm": 2.0589022636413574,
      "learning_rate": 4.469081823860088e-05,
      "loss": 2.9551,
      "step": 340
    },
    {
      "epoch": 0.11242973141786383,
      "grad_norm": 3.8759829998016357,
      "learning_rate": 4.437851342910681e-05,
      "loss": 2.5833,
      "step": 360
    },
    {
      "epoch": 0.11867582760774516,
      "grad_norm": 3.9548556804656982,
      "learning_rate": 4.406620861961274e-05,
      "loss": 3.0385,
      "step": 380
    },
    {
      "epoch": 0.12492192379762648,
      "grad_norm": 2.6232974529266357,
      "learning_rate": 4.375390381011868e-05,
      "loss": 2.7124,
      "step": 400
    },
    {
      "epoch": 0.1311680199875078,
      "grad_norm": 2.080956220626831,
      "learning_rate": 4.3441599000624614e-05,
      "loss": 2.8504,
      "step": 420
    },
    {
      "epoch": 0.13741411617738913,
      "grad_norm": 2.893707036972046,
      "learning_rate": 4.312929419113055e-05,
      "loss": 2.8075,
      "step": 440
    },
    {
      "epoch": 0.14366021236727045,
      "grad_norm": 3.174916982650757,
      "learning_rate": 4.281698938163648e-05,
      "loss": 2.6756,
      "step": 460
    },
    {
      "epoch": 0.14990630855715179,
      "grad_norm": 2.7715773582458496,
      "learning_rate": 4.2504684572142414e-05,
      "loss": 2.6994,
      "step": 480
    },
    {
      "epoch": 0.1561524047470331,
      "grad_norm": 3.1229937076568604,
      "learning_rate": 4.219237976264835e-05,
      "loss": 2.5,
      "step": 500
    },
    {
      "epoch": 0.16239850093691444,
      "grad_norm": 3.7433629035949707,
      "learning_rate": 4.188007495315428e-05,
      "loss": 2.7145,
      "step": 520
    },
    {
      "epoch": 0.16864459712679575,
      "grad_norm": 4.687792778015137,
      "learning_rate": 4.156777014366021e-05,
      "loss": 2.7268,
      "step": 540
    },
    {
      "epoch": 0.1748906933166771,
      "grad_norm": 4.73487663269043,
      "learning_rate": 4.125546533416615e-05,
      "loss": 3.1233,
      "step": 560
    },
    {
      "epoch": 0.1811367895065584,
      "grad_norm": 3.7153172492980957,
      "learning_rate": 4.094316052467208e-05,
      "loss": 2.7226,
      "step": 580
    },
    {
      "epoch": 0.18738288569643974,
      "grad_norm": 4.160110950469971,
      "learning_rate": 4.063085571517802e-05,
      "loss": 2.6628,
      "step": 600
    },
    {
      "epoch": 0.19362898188632105,
      "grad_norm": 3.126046657562256,
      "learning_rate": 4.031855090568395e-05,
      "loss": 3.0178,
      "step": 620
    },
    {
      "epoch": 0.19987507807620236,
      "grad_norm": 2.5289721488952637,
      "learning_rate": 4.0006246096189884e-05,
      "loss": 2.6422,
      "step": 640
    },
    {
      "epoch": 0.2061211742660837,
      "grad_norm": 3.9148857593536377,
      "learning_rate": 3.969394128669582e-05,
      "loss": 2.6393,
      "step": 660
    },
    {
      "epoch": 0.212367270455965,
      "grad_norm": 2.8504717350006104,
      "learning_rate": 3.938163647720175e-05,
      "loss": 2.9792,
      "step": 680
    },
    {
      "epoch": 0.21861336664584635,
      "grad_norm": 3.4905455112457275,
      "learning_rate": 3.906933166770769e-05,
      "loss": 2.8159,
      "step": 700
    },
    {
      "epoch": 0.22485946283572766,
      "grad_norm": 3.08007550239563,
      "learning_rate": 3.875702685821362e-05,
      "loss": 2.6937,
      "step": 720
    },
    {
      "epoch": 0.231105559025609,
      "grad_norm": 2.3973348140716553,
      "learning_rate": 3.844472204871955e-05,
      "loss": 2.9008,
      "step": 740
    },
    {
      "epoch": 0.23735165521549031,
      "grad_norm": 2.920398473739624,
      "learning_rate": 3.813241723922549e-05,
      "loss": 2.7197,
      "step": 760
    },
    {
      "epoch": 0.24359775140537165,
      "grad_norm": 4.2394280433654785,
      "learning_rate": 3.782011242973142e-05,
      "loss": 2.5608,
      "step": 780
    },
    {
      "epoch": 0.24984384759525297,
      "grad_norm": 4.320163249969482,
      "learning_rate": 3.750780762023735e-05,
      "loss": 2.7325,
      "step": 800
    },
    {
      "epoch": 0.2560899437851343,
      "grad_norm": 7.346144199371338,
      "learning_rate": 3.719550281074329e-05,
      "loss": 2.7655,
      "step": 820
    },
    {
      "epoch": 0.2623360399750156,
      "grad_norm": 3.444692850112915,
      "learning_rate": 3.688319800124922e-05,
      "loss": 2.6974,
      "step": 840
    },
    {
      "epoch": 0.26858213616489696,
      "grad_norm": 2.8311755657196045,
      "learning_rate": 3.6570893191755154e-05,
      "loss": 2.571,
      "step": 860
    },
    {
      "epoch": 0.27482823235477827,
      "grad_norm": 3.064018726348877,
      "learning_rate": 3.625858838226109e-05,
      "loss": 2.8051,
      "step": 880
    },
    {
      "epoch": 0.2810743285446596,
      "grad_norm": 2.4530911445617676,
      "learning_rate": 3.594628357276702e-05,
      "loss": 2.4469,
      "step": 900
    },
    {
      "epoch": 0.2873204247345409,
      "grad_norm": 3.4664344787597656,
      "learning_rate": 3.563397876327296e-05,
      "loss": 2.7295,
      "step": 920
    },
    {
      "epoch": 0.29356652092442226,
      "grad_norm": 4.130941867828369,
      "learning_rate": 3.532167395377889e-05,
      "loss": 2.8405,
      "step": 940
    },
    {
      "epoch": 0.29981261711430357,
      "grad_norm": 2.76470685005188,
      "learning_rate": 3.5009369144284825e-05,
      "loss": 2.7108,
      "step": 960
    },
    {
      "epoch": 0.3060587133041849,
      "grad_norm": 4.205817222595215,
      "learning_rate": 3.469706433479076e-05,
      "loss": 2.659,
      "step": 980
    },
    {
      "epoch": 0.3123048094940662,
      "grad_norm": 3.306403160095215,
      "learning_rate": 3.438475952529669e-05,
      "loss": 2.7492,
      "step": 1000
    },
    {
      "epoch": 0.3185509056839475,
      "grad_norm": 2.0320382118225098,
      "learning_rate": 3.4072454715802624e-05,
      "loss": 2.5899,
      "step": 1020
    },
    {
      "epoch": 0.32479700187382887,
      "grad_norm": 4.874257564544678,
      "learning_rate": 3.376014990630856e-05,
      "loss": 2.8844,
      "step": 1040
    },
    {
      "epoch": 0.3310430980637102,
      "grad_norm": 3.102703809738159,
      "learning_rate": 3.3447845096814495e-05,
      "loss": 2.6927,
      "step": 1060
    },
    {
      "epoch": 0.3372891942535915,
      "grad_norm": 3.8374264240264893,
      "learning_rate": 3.3135540287320424e-05,
      "loss": 2.7193,
      "step": 1080
    },
    {
      "epoch": 0.3435352904434728,
      "grad_norm": 4.128782749176025,
      "learning_rate": 3.282323547782636e-05,
      "loss": 2.7898,
      "step": 1100
    },
    {
      "epoch": 0.3497813866333542,
      "grad_norm": 2.540234088897705,
      "learning_rate": 3.2510930668332295e-05,
      "loss": 2.492,
      "step": 1120
    },
    {
      "epoch": 0.3560274828232355,
      "grad_norm": 5.775226593017578,
      "learning_rate": 3.219862585883823e-05,
      "loss": 2.7778,
      "step": 1140
    },
    {
      "epoch": 0.3622735790131168,
      "grad_norm": 3.100594997406006,
      "learning_rate": 3.188632104934416e-05,
      "loss": 2.2936,
      "step": 1160
    },
    {
      "epoch": 0.3685196752029981,
      "grad_norm": 8.619044303894043,
      "learning_rate": 3.1574016239850095e-05,
      "loss": 2.5594,
      "step": 1180
    },
    {
      "epoch": 0.3747657713928795,
      "grad_norm": 3.1239614486694336,
      "learning_rate": 3.126171143035603e-05,
      "loss": 2.8581,
      "step": 1200
    },
    {
      "epoch": 0.3810118675827608,
      "grad_norm": 3.2376351356506348,
      "learning_rate": 3.0949406620861966e-05,
      "loss": 2.4649,
      "step": 1220
    },
    {
      "epoch": 0.3872579637726421,
      "grad_norm": 2.9151909351348877,
      "learning_rate": 3.0637101811367894e-05,
      "loss": 2.6663,
      "step": 1240
    },
    {
      "epoch": 0.3935040599625234,
      "grad_norm": 4.819116592407227,
      "learning_rate": 3.0324797001873826e-05,
      "loss": 2.5575,
      "step": 1260
    },
    {
      "epoch": 0.3997501561524047,
      "grad_norm": 5.1033430099487305,
      "learning_rate": 3.0012492192379765e-05,
      "loss": 2.5924,
      "step": 1280
    },
    {
      "epoch": 0.4059962523422861,
      "grad_norm": 3.506040096282959,
      "learning_rate": 2.9700187382885697e-05,
      "loss": 2.6526,
      "step": 1300
    },
    {
      "epoch": 0.4122423485321674,
      "grad_norm": 5.795025825500488,
      "learning_rate": 2.9387882573391633e-05,
      "loss": 2.8599,
      "step": 1320
    },
    {
      "epoch": 0.4184884447220487,
      "grad_norm": 3.329371452331543,
      "learning_rate": 2.9075577763897565e-05,
      "loss": 2.5105,
      "step": 1340
    },
    {
      "epoch": 0.42473454091193,
      "grad_norm": 2.9050350189208984,
      "learning_rate": 2.8763272954403497e-05,
      "loss": 2.4808,
      "step": 1360
    },
    {
      "epoch": 0.4309806371018114,
      "grad_norm": 5.538816928863525,
      "learning_rate": 2.8450968144909436e-05,
      "loss": 2.6096,
      "step": 1380
    },
    {
      "epoch": 0.4372267332916927,
      "grad_norm": 4.255838394165039,
      "learning_rate": 2.8138663335415365e-05,
      "loss": 2.6359,
      "step": 1400
    },
    {
      "epoch": 0.443472829481574,
      "grad_norm": 6.425328731536865,
      "learning_rate": 2.7826358525921297e-05,
      "loss": 2.3929,
      "step": 1420
    },
    {
      "epoch": 0.4497189256714553,
      "grad_norm": 4.074633598327637,
      "learning_rate": 2.7514053716427236e-05,
      "loss": 2.3325,
      "step": 1440
    },
    {
      "epoch": 0.45596502186133664,
      "grad_norm": 4.7358245849609375,
      "learning_rate": 2.7201748906933168e-05,
      "loss": 2.4945,
      "step": 1460
    },
    {
      "epoch": 0.462211118051218,
      "grad_norm": 4.505492210388184,
      "learning_rate": 2.6889444097439103e-05,
      "loss": 2.5168,
      "step": 1480
    },
    {
      "epoch": 0.4684572142410993,
      "grad_norm": 6.606976509094238,
      "learning_rate": 2.6577139287945035e-05,
      "loss": 2.5106,
      "step": 1500
    },
    {
      "epoch": 0.47470331043098063,
      "grad_norm": 4.111032962799072,
      "learning_rate": 2.6264834478450968e-05,
      "loss": 2.7365,
      "step": 1520
    },
    {
      "epoch": 0.48094940662086194,
      "grad_norm": 6.724081993103027,
      "learning_rate": 2.5952529668956903e-05,
      "loss": 2.2786,
      "step": 1540
    },
    {
      "epoch": 0.4871955028107433,
      "grad_norm": 4.5777268409729,
      "learning_rate": 2.5640224859462835e-05,
      "loss": 2.1842,
      "step": 1560
    },
    {
      "epoch": 0.4934415990006246,
      "grad_norm": 5.641876697540283,
      "learning_rate": 2.5327920049968774e-05,
      "loss": 2.2779,
      "step": 1580
    },
    {
      "epoch": 0.49968769519050593,
      "grad_norm": 7.818845748901367,
      "learning_rate": 2.5015615240474706e-05,
      "loss": 2.1692,
      "step": 1600
    },
    {
      "epoch": 0.5059337913803873,
      "grad_norm": 5.147799968719482,
      "learning_rate": 2.4703310430980638e-05,
      "loss": 2.2715,
      "step": 1620
    },
    {
      "epoch": 0.5121798875702686,
      "grad_norm": 10.44902229309082,
      "learning_rate": 2.4391005621486574e-05,
      "loss": 2.4603,
      "step": 1640
    },
    {
      "epoch": 0.5184259837601499,
      "grad_norm": 3.7985732555389404,
      "learning_rate": 2.4078700811992506e-05,
      "loss": 2.3849,
      "step": 1660
    },
    {
      "epoch": 0.5246720799500312,
      "grad_norm": 5.865043640136719,
      "learning_rate": 2.3766396002498438e-05,
      "loss": 2.4521,
      "step": 1680
    },
    {
      "epoch": 0.5309181761399125,
      "grad_norm": 6.91799783706665,
      "learning_rate": 2.3454091193004373e-05,
      "loss": 2.159,
      "step": 1700
    },
    {
      "epoch": 0.5371642723297939,
      "grad_norm": 4.18131685256958,
      "learning_rate": 2.314178638351031e-05,
      "loss": 2.1075,
      "step": 1720
    },
    {
      "epoch": 0.5434103685196752,
      "grad_norm": 5.681297779083252,
      "learning_rate": 2.282948157401624e-05,
      "loss": 2.3221,
      "step": 1740
    },
    {
      "epoch": 0.5496564647095565,
      "grad_norm": 5.769021034240723,
      "learning_rate": 2.2517176764522173e-05,
      "loss": 2.3497,
      "step": 1760
    },
    {
      "epoch": 0.5559025608994379,
      "grad_norm": 9.655105590820312,
      "learning_rate": 2.220487195502811e-05,
      "loss": 2.5534,
      "step": 1780
    },
    {
      "epoch": 0.5621486570893192,
      "grad_norm": 8.151744842529297,
      "learning_rate": 2.1892567145534044e-05,
      "loss": 2.1035,
      "step": 1800
    },
    {
      "epoch": 0.5683947532792005,
      "grad_norm": 5.74683141708374,
      "learning_rate": 2.1580262336039976e-05,
      "loss": 2.154,
      "step": 1820
    },
    {
      "epoch": 0.5746408494690818,
      "grad_norm": 3.5920357704162598,
      "learning_rate": 2.1267957526545908e-05,
      "loss": 2.0646,
      "step": 1840
    },
    {
      "epoch": 0.5808869456589631,
      "grad_norm": 10.239977836608887,
      "learning_rate": 2.0955652717051844e-05,
      "loss": 2.198,
      "step": 1860
    },
    {
      "epoch": 0.5871330418488445,
      "grad_norm": 3.631121873855591,
      "learning_rate": 2.064334790755778e-05,
      "loss": 2.1863,
      "step": 1880
    },
    {
      "epoch": 0.5933791380387258,
      "grad_norm": 9.566389083862305,
      "learning_rate": 2.033104309806371e-05,
      "loss": 1.9127,
      "step": 1900
    },
    {
      "epoch": 0.5996252342286071,
      "grad_norm": 7.540945529937744,
      "learning_rate": 2.0018738288569643e-05,
      "loss": 2.2545,
      "step": 1920
    },
    {
      "epoch": 0.6058713304184884,
      "grad_norm": 5.457769393920898,
      "learning_rate": 1.970643347907558e-05,
      "loss": 1.7829,
      "step": 1940
    },
    {
      "epoch": 0.6121174266083698,
      "grad_norm": 4.6898088455200195,
      "learning_rate": 1.939412866958151e-05,
      "loss": 2.092,
      "step": 1960
    },
    {
      "epoch": 0.6183635227982511,
      "grad_norm": 4.707203388214111,
      "learning_rate": 1.9081823860087447e-05,
      "loss": 2.0059,
      "step": 1980
    },
    {
      "epoch": 0.6246096189881324,
      "grad_norm": 5.7073469161987305,
      "learning_rate": 1.876951905059338e-05,
      "loss": 2.0865,
      "step": 2000
    },
    {
      "epoch": 0.6308557151780138,
      "grad_norm": 6.401088237762451,
      "learning_rate": 1.8457214241099314e-05,
      "loss": 1.7292,
      "step": 2020
    },
    {
      "epoch": 0.637101811367895,
      "grad_norm": 8.511351585388184,
      "learning_rate": 1.8144909431605246e-05,
      "loss": 2.2739,
      "step": 2040
    },
    {
      "epoch": 0.6433479075577764,
      "grad_norm": 6.70497465133667,
      "learning_rate": 1.7832604622111182e-05,
      "loss": 2.042,
      "step": 2060
    },
    {
      "epoch": 0.6495940037476577,
      "grad_norm": 5.246578693389893,
      "learning_rate": 1.7520299812617117e-05,
      "loss": 2.0974,
      "step": 2080
    },
    {
      "epoch": 0.655840099937539,
      "grad_norm": 11.206995964050293,
      "learning_rate": 1.720799500312305e-05,
      "loss": 2.0204,
      "step": 2100
    },
    {
      "epoch": 0.6620861961274204,
      "grad_norm": 6.476535320281982,
      "learning_rate": 1.689569019362898e-05,
      "loss": 2.3546,
      "step": 2120
    },
    {
      "epoch": 0.6683322923173017,
      "grad_norm": 2.9628567695617676,
      "learning_rate": 1.6583385384134917e-05,
      "loss": 2.0813,
      "step": 2140
    },
    {
      "epoch": 0.674578388507183,
      "grad_norm": 5.9349188804626465,
      "learning_rate": 1.6271080574640852e-05,
      "loss": 2.2729,
      "step": 2160
    },
    {
      "epoch": 0.6808244846970644,
      "grad_norm": 6.260815143585205,
      "learning_rate": 1.5958775765146785e-05,
      "loss": 1.9583,
      "step": 2180
    },
    {
      "epoch": 0.6870705808869456,
      "grad_norm": 5.544328689575195,
      "learning_rate": 1.5646470955652717e-05,
      "loss": 2.0548,
      "step": 2200
    },
    {
      "epoch": 0.693316677076827,
      "grad_norm": 11.215269088745117,
      "learning_rate": 1.5334166146158652e-05,
      "loss": 1.7925,
      "step": 2220
    },
    {
      "epoch": 0.6995627732667083,
      "grad_norm": 7.354444980621338,
      "learning_rate": 1.5021861336664586e-05,
      "loss": 2.2688,
      "step": 2240
    },
    {
      "epoch": 0.7058088694565896,
      "grad_norm": 6.186635971069336,
      "learning_rate": 1.470955652717052e-05,
      "loss": 1.8723,
      "step": 2260
    },
    {
      "epoch": 0.712054965646471,
      "grad_norm": 6.673027992248535,
      "learning_rate": 1.4397251717676452e-05,
      "loss": 2.2833,
      "step": 2280
    },
    {
      "epoch": 0.7183010618363522,
      "grad_norm": 6.882096290588379,
      "learning_rate": 1.4084946908182387e-05,
      "loss": 1.8425,
      "step": 2300
    },
    {
      "epoch": 0.7245471580262336,
      "grad_norm": 5.77081823348999,
      "learning_rate": 1.3772642098688321e-05,
      "loss": 2.2437,
      "step": 2320
    },
    {
      "epoch": 0.730793254216115,
      "grad_norm": 4.107306957244873,
      "learning_rate": 1.3460337289194255e-05,
      "loss": 1.7821,
      "step": 2340
    },
    {
      "epoch": 0.7370393504059962,
      "grad_norm": 6.597945690155029,
      "learning_rate": 1.3148032479700187e-05,
      "loss": 2.0115,
      "step": 2360
    },
    {
      "epoch": 0.7432854465958776,
      "grad_norm": 13.545961380004883,
      "learning_rate": 1.283572767020612e-05,
      "loss": 2.325,
      "step": 2380
    },
    {
      "epoch": 0.749531542785759,
      "grad_norm": 5.746728420257568,
      "learning_rate": 1.2523422860712056e-05,
      "loss": 2.0989,
      "step": 2400
    },
    {
      "epoch": 0.7557776389756402,
      "grad_norm": 7.114035606384277,
      "learning_rate": 1.2211118051217988e-05,
      "loss": 1.8291,
      "step": 2420
    },
    {
      "epoch": 0.7620237351655216,
      "grad_norm": 6.319833278656006,
      "learning_rate": 1.1898813241723924e-05,
      "loss": 2.0398,
      "step": 2440
    },
    {
      "epoch": 0.7682698313554028,
      "grad_norm": 8.076285362243652,
      "learning_rate": 1.1586508432229858e-05,
      "loss": 2.365,
      "step": 2460
    },
    {
      "epoch": 0.7745159275452842,
      "grad_norm": 5.4164628982543945,
      "learning_rate": 1.127420362273579e-05,
      "loss": 2.2452,
      "step": 2480
    },
    {
      "epoch": 0.7807620237351656,
      "grad_norm": 8.143169403076172,
      "learning_rate": 1.0961898813241725e-05,
      "loss": 2.3809,
      "step": 2500
    },
    {
      "epoch": 0.7870081199250468,
      "grad_norm": 8.056733131408691,
      "learning_rate": 1.0649594003747657e-05,
      "loss": 2.2368,
      "step": 2520
    },
    {
      "epoch": 0.7932542161149282,
      "grad_norm": 7.611058712005615,
      "learning_rate": 1.0337289194253593e-05,
      "loss": 2.0644,
      "step": 2540
    },
    {
      "epoch": 0.7995003123048094,
      "grad_norm": 7.84691858291626,
      "learning_rate": 1.0024984384759525e-05,
      "loss": 1.8021,
      "step": 2560
    },
    {
      "epoch": 0.8057464084946908,
      "grad_norm": 5.361764430999756,
      "learning_rate": 9.71267957526546e-06,
      "loss": 1.9417,
      "step": 2580
    },
    {
      "epoch": 0.8119925046845722,
      "grad_norm": 6.095186233520508,
      "learning_rate": 9.400374765771393e-06,
      "loss": 2.3096,
      "step": 2600
    },
    {
      "epoch": 0.8182386008744534,
      "grad_norm": 7.828145980834961,
      "learning_rate": 9.088069956277328e-06,
      "loss": 2.2737,
      "step": 2620
    },
    {
      "epoch": 0.8244846970643348,
      "grad_norm": 5.065859794616699,
      "learning_rate": 8.77576514678326e-06,
      "loss": 2.182,
      "step": 2640
    },
    {
      "epoch": 0.8307307932542161,
      "grad_norm": 5.14509391784668,
      "learning_rate": 8.463460337289194e-06,
      "loss": 2.1187,
      "step": 2660
    },
    {
      "epoch": 0.8369768894440974,
      "grad_norm": 8.84304428100586,
      "learning_rate": 8.15115552779513e-06,
      "loss": 1.8987,
      "step": 2680
    },
    {
      "epoch": 0.8432229856339788,
      "grad_norm": 6.434870719909668,
      "learning_rate": 7.838850718301062e-06,
      "loss": 1.9413,
      "step": 2700
    },
    {
      "epoch": 0.84946908182386,
      "grad_norm": 4.600275039672852,
      "learning_rate": 7.526545908806996e-06,
      "loss": 2.1107,
      "step": 2720
    },
    {
      "epoch": 0.8557151780137414,
      "grad_norm": 9.691720962524414,
      "learning_rate": 7.214241099312929e-06,
      "loss": 1.8678,
      "step": 2740
    },
    {
      "epoch": 0.8619612742036228,
      "grad_norm": 5.024956226348877,
      "learning_rate": 6.901936289818864e-06,
      "loss": 2.1413,
      "step": 2760
    },
    {
      "epoch": 0.868207370393504,
      "grad_norm": 4.373764991760254,
      "learning_rate": 6.589631480324797e-06,
      "loss": 2.2157,
      "step": 2780
    },
    {
      "epoch": 0.8744534665833854,
      "grad_norm": 5.964173316955566,
      "learning_rate": 6.277326670830731e-06,
      "loss": 1.9203,
      "step": 2800
    },
    {
      "epoch": 0.8806995627732667,
      "grad_norm": 4.583627223968506,
      "learning_rate": 5.965021861336665e-06,
      "loss": 1.7377,
      "step": 2820
    },
    {
      "epoch": 0.886945658963148,
      "grad_norm": 5.218972206115723,
      "learning_rate": 5.652717051842599e-06,
      "loss": 2.0339,
      "step": 2840
    },
    {
      "epoch": 0.8931917551530294,
      "grad_norm": 4.644482135772705,
      "learning_rate": 5.340412242348533e-06,
      "loss": 1.9294,
      "step": 2860
    },
    {
      "epoch": 0.8994378513429107,
      "grad_norm": 3.316526412963867,
      "learning_rate": 5.028107432854466e-06,
      "loss": 1.9216,
      "step": 2880
    },
    {
      "epoch": 0.905683947532792,
      "grad_norm": 3.9178998470306396,
      "learning_rate": 4.7158026233603995e-06,
      "loss": 2.1137,
      "step": 2900
    },
    {
      "epoch": 0.9119300437226733,
      "grad_norm": 5.773944854736328,
      "learning_rate": 4.403497813866333e-06,
      "loss": 2.3483,
      "step": 2920
    },
    {
      "epoch": 0.9181761399125546,
      "grad_norm": 6.133700847625732,
      "learning_rate": 4.091193004372267e-06,
      "loss": 2.0656,
      "step": 2940
    },
    {
      "epoch": 0.924422236102436,
      "grad_norm": 3.0790977478027344,
      "learning_rate": 3.7788881948782017e-06,
      "loss": 1.9562,
      "step": 2960
    },
    {
      "epoch": 0.9306683322923173,
      "grad_norm": 4.9188103675842285,
      "learning_rate": 3.4665833853841355e-06,
      "loss": 2.4327,
      "step": 2980
    },
    {
      "epoch": 0.9369144284821986,
      "grad_norm": 4.5354509353637695,
      "learning_rate": 3.154278575890069e-06,
      "loss": 2.0456,
      "step": 3000
    },
    {
      "epoch": 0.94316052467208,
      "grad_norm": 4.375840187072754,
      "learning_rate": 2.8419737663960027e-06,
      "loss": 1.9895,
      "step": 3020
    },
    {
      "epoch": 0.9494066208619613,
      "grad_norm": 5.6817307472229,
      "learning_rate": 2.5296689569019365e-06,
      "loss": 1.9656,
      "step": 3040
    },
    {
      "epoch": 0.9556527170518426,
      "grad_norm": 6.242130756378174,
      "learning_rate": 2.2173641474078703e-06,
      "loss": 2.0195,
      "step": 3060
    },
    {
      "epoch": 0.9618988132417239,
      "grad_norm": 6.264340877532959,
      "learning_rate": 1.9050593379138039e-06,
      "loss": 1.826,
      "step": 3080
    },
    {
      "epoch": 0.9681449094316052,
      "grad_norm": 6.317758083343506,
      "learning_rate": 1.5927545284197377e-06,
      "loss": 2.2237,
      "step": 3100
    },
    {
      "epoch": 0.9743910056214866,
      "grad_norm": 7.347826957702637,
      "learning_rate": 1.2804497189256715e-06,
      "loss": 2.0525,
      "step": 3120
    },
    {
      "epoch": 0.9806371018113679,
      "grad_norm": 4.710007667541504,
      "learning_rate": 9.681449094316053e-07,
      "loss": 2.1825,
      "step": 3140
    },
    {
      "epoch": 0.9868831980012492,
      "grad_norm": 3.5454931259155273,
      "learning_rate": 6.558400999375391e-07,
      "loss": 2.1784,
      "step": 3160
    },
    {
      "epoch": 0.9931292941911305,
      "grad_norm": 4.577638149261475,
      "learning_rate": 3.435352904434728e-07,
      "loss": 1.703,
      "step": 3180
    },
    {
      "epoch": 0.9993753903810119,
      "grad_norm": 8.630655288696289,
      "learning_rate": 3.123048094940662e-08,
      "loss": 2.0709,
      "step": 3200
    }
  ],
  "logging_steps": 20,
  "max_steps": 3202,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2326032083828736.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}