{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.042687531465010906,
  "eval_steps": 500,
  "global_step": 2035,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00020976673938580297,
      "grad_norm": 236.7567785411443,
      "learning_rate": 2.910112359550562e-06,
      "loss": 1.5920657157897948,
      "step": 10
    },
    {
      "epoch": 0.00041953347877160595,
      "grad_norm": 152.92314363522394,
      "learning_rate": 3.921348314606742e-06,
      "loss": 0.9421598434448242,
      "step": 20
    },
    {
      "epoch": 0.000629300218157409,
      "grad_norm": 97.19978632147773,
      "learning_rate": 4.932584269662922e-06,
      "loss": 0.7996736526489258,
      "step": 30
    },
    {
      "epoch": 0.0008390669575432119,
      "grad_norm": 157.7423675487546,
      "learning_rate": 5.943820224719102e-06,
      "loss": 0.760670280456543,
      "step": 40
    },
    {
      "epoch": 0.001048833696929015,
      "grad_norm": 49.50204924369943,
      "learning_rate": 6.955056179775282e-06,
      "loss": 0.7635975837707519,
      "step": 50
    },
    {
      "epoch": 0.001258600436314818,
      "grad_norm": 16.48624285559217,
      "learning_rate": 7.966292134831462e-06,
      "loss": 0.7791356563568115,
      "step": 60
    },
    {
      "epoch": 0.0014683671757006209,
      "grad_norm": 6.858907596379825,
      "learning_rate": 8.977528089887641e-06,
      "loss": 0.7495433330535889,
      "step": 70
    },
    {
      "epoch": 0.0016781339150864238,
      "grad_norm": 26.435928023185784,
      "learning_rate": 9.988764044943822e-06,
      "loss": 0.7318498134613037,
      "step": 80
    },
    {
      "epoch": 0.001887900654472227,
      "grad_norm": 214.57004030820568,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.7413080215454102,
      "step": 90
    },
    {
      "epoch": 0.00209766739385803,
      "grad_norm": 7.939148544393301,
      "learning_rate": 1.2011235955056182e-05,
      "loss": 0.8061410903930664,
      "step": 100
    },
    {
      "epoch": 0.0023074341332438328,
      "grad_norm": 14.183647299272815,
      "learning_rate": 1.3022471910112362e-05,
      "loss": 0.7331802368164062,
      "step": 110
    },
    {
      "epoch": 0.002517200872629636,
      "grad_norm": 22.423640491145502,
      "learning_rate": 1.403370786516854e-05,
      "loss": 0.7498276710510254,
      "step": 120
    },
    {
      "epoch": 0.002726967612015439,
      "grad_norm": 3.136265232048693,
      "learning_rate": 1.504494382022472e-05,
      "loss": 0.724825668334961,
      "step": 130
    },
    {
      "epoch": 0.0029367343514012417,
      "grad_norm": 8.214245546263463,
      "learning_rate": 1.60561797752809e-05,
      "loss": 0.6886782646179199,
      "step": 140
    },
    {
      "epoch": 0.003146501090787045,
      "grad_norm": 5.619530923664304,
      "learning_rate": 1.706741573033708e-05,
      "loss": 0.7081296443939209,
      "step": 150
    },
    {
      "epoch": 0.0033562678301728476,
      "grad_norm": 10.033723007819054,
      "learning_rate": 1.8078651685393256e-05,
      "loss": 0.7299324035644531,
      "step": 160
    },
    {
      "epoch": 0.0035660345695586507,
      "grad_norm": 11.134808180954863,
      "learning_rate": 1.908988764044944e-05,
      "loss": 0.745941162109375,
      "step": 170
    },
    {
      "epoch": 0.003775801308944454,
      "grad_norm": 2.4304907872424795,
      "learning_rate": 2e-05,
      "loss": 0.7129830360412598,
      "step": 180
    },
    {
      "epoch": 0.0039855680483302566,
      "grad_norm": 4.018584134102583,
      "learning_rate": 2e-05,
      "loss": 0.7092458724975585,
      "step": 190
    },
    {
      "epoch": 0.00419533478771606,
      "grad_norm": 5.006464475899677,
      "learning_rate": 2e-05,
      "loss": 0.7565587520599365,
      "step": 200
    },
    {
      "epoch": 0.004405101527101863,
      "grad_norm": 2.800599112711121,
      "learning_rate": 2e-05,
      "loss": 0.6959824562072754,
      "step": 210
    },
    {
      "epoch": 0.0046148682664876655,
      "grad_norm": 6.231621527897506,
      "learning_rate": 2e-05,
      "loss": 0.6992167472839356,
      "step": 220
    },
    {
      "epoch": 0.004824635005873469,
      "grad_norm": 10.811727778354081,
      "learning_rate": 2e-05,
      "loss": 0.6952213287353516,
      "step": 230
    },
    {
      "epoch": 0.005034401745259272,
      "grad_norm": 11.261067852316236,
      "learning_rate": 2e-05,
      "loss": 0.7460278987884521,
      "step": 240
    },
    {
      "epoch": 0.0052441684846450745,
      "grad_norm": 7.649125680617738,
      "learning_rate": 2e-05,
      "loss": 0.7858521461486816,
      "step": 250
    },
    {
      "epoch": 0.005453935224030878,
      "grad_norm": 4.490247060675485,
      "learning_rate": 2e-05,
      "loss": 0.7642570018768311,
      "step": 260
    },
    {
      "epoch": 0.005663701963416681,
      "grad_norm": 5.320163840820431,
      "learning_rate": 2e-05,
      "loss": 0.7289691925048828,
      "step": 270
    },
    {
      "epoch": 0.0058734687028024835,
      "grad_norm": 3.4799647445683717,
      "learning_rate": 2e-05,
      "loss": 0.6995810508728028,
      "step": 280
    },
    {
      "epoch": 0.006083235442188286,
      "grad_norm": 34.442893954128465,
      "learning_rate": 2e-05,
      "loss": 0.7102492809295654,
      "step": 290
    },
    {
      "epoch": 0.00629300218157409,
      "grad_norm": 4.179341348142214,
      "learning_rate": 2e-05,
      "loss": 0.7110370635986328,
      "step": 300
    },
    {
      "epoch": 0.0065027689209598925,
      "grad_norm": 4.270936726323635,
      "learning_rate": 2e-05,
      "loss": 0.7388000011444091,
      "step": 310
    },
    {
      "epoch": 0.006712535660345695,
      "grad_norm": 3.52144124715383,
      "learning_rate": 2e-05,
      "loss": 0.7715152740478516,
      "step": 320
    },
    {
      "epoch": 0.006922302399731499,
      "grad_norm": 2.1760255438109892,
      "learning_rate": 2e-05,
      "loss": 0.7474425315856934,
      "step": 330
    },
    {
      "epoch": 0.0071320691391173014,
      "grad_norm": 1.8391052662069023,
      "learning_rate": 2e-05,
      "loss": 0.7086240291595459,
      "step": 340
    },
    {
      "epoch": 0.007341835878503104,
      "grad_norm": 9.85499125185226,
      "learning_rate": 2e-05,
      "loss": 0.736334228515625,
      "step": 350
    },
    {
      "epoch": 0.007551602617888908,
      "grad_norm": 7.061981460856605,
      "learning_rate": 2e-05,
      "loss": 0.7058637619018555,
      "step": 360
    },
    {
      "epoch": 0.00776136935727471,
      "grad_norm": 2.399773463771947,
      "learning_rate": 2e-05,
      "loss": 0.7035298824310303,
      "step": 370
    },
    {
      "epoch": 0.007971136096660513,
      "grad_norm": 1.8715337008198967,
      "learning_rate": 2e-05,
      "loss": 0.7319884300231934,
      "step": 380
    },
    {
      "epoch": 0.008180902836046316,
      "grad_norm": 4.402380567892121,
      "learning_rate": 2e-05,
      "loss": 0.6931228637695312,
      "step": 390
    },
    {
      "epoch": 0.00839066957543212,
      "grad_norm": 1.997385308499001,
      "learning_rate": 2e-05,
      "loss": 0.6841664791107178,
      "step": 400
    },
    {
      "epoch": 0.008600436314817923,
      "grad_norm": 8.77520106541026,
      "learning_rate": 2e-05,
      "loss": 0.6674545288085938,
      "step": 410
    },
    {
      "epoch": 0.008810203054203726,
      "grad_norm": 227.00469733968774,
      "learning_rate": 2e-05,
      "loss": 0.7310820579528808,
      "step": 420
    },
    {
      "epoch": 0.009019969793589528,
      "grad_norm": 3.283083883742064,
      "learning_rate": 2e-05,
      "loss": 0.7967383384704589,
      "step": 430
    },
    {
      "epoch": 0.009229736532975331,
      "grad_norm": 12.336376030486466,
      "learning_rate": 2e-05,
      "loss": 0.7576329231262207,
      "step": 440
    },
    {
      "epoch": 0.009439503272361134,
      "grad_norm": 1.8242777860324726,
      "learning_rate": 2e-05,
      "loss": 0.706461238861084,
      "step": 450
    },
    {
      "epoch": 0.009649270011746938,
      "grad_norm": 1.8090965567853239,
      "learning_rate": 2e-05,
      "loss": 0.7032674312591553,
      "step": 460
    },
    {
      "epoch": 0.009859036751132741,
      "grad_norm": 2.5475991670097953,
      "learning_rate": 2e-05,
      "loss": 0.7029439449310303,
      "step": 470
    },
    {
      "epoch": 0.010068803490518544,
      "grad_norm": 1.4692934900457209,
      "learning_rate": 2e-05,
      "loss": 0.6630709648132325,
      "step": 480
    },
    {
      "epoch": 0.010278570229904346,
      "grad_norm": 2.1331924790853534,
      "learning_rate": 2e-05,
      "loss": 0.7047128677368164,
      "step": 490
    },
    {
      "epoch": 0.010488336969290149,
      "grad_norm": 2.4224465959516164,
      "learning_rate": 2e-05,
      "loss": 0.7062530517578125,
      "step": 500
    },
    {
      "epoch": 0.010698103708675952,
      "grad_norm": 1.9672184607046945,
      "learning_rate": 2e-05,
      "loss": 0.6654829978942871,
      "step": 510
    },
    {
      "epoch": 0.010907870448061756,
      "grad_norm": 1.8632555104139144,
      "learning_rate": 2e-05,
      "loss": 0.6872918128967285,
      "step": 520
    },
    {
      "epoch": 0.011117637187447559,
      "grad_norm": 2.696765708839419,
      "learning_rate": 2e-05,
      "loss": 0.712721586227417,
      "step": 530
    },
    {
      "epoch": 0.011327403926833362,
      "grad_norm": 3.3114397630142145,
      "learning_rate": 2e-05,
      "loss": 0.665414571762085,
      "step": 540
    },
    {
      "epoch": 0.011537170666219164,
      "grad_norm": 1.7287592058057928,
      "learning_rate": 2e-05,
      "loss": 0.7328654289245605,
      "step": 550
    },
    {
      "epoch": 0.011746937405604967,
      "grad_norm": 1.7622716775547056,
      "learning_rate": 2e-05,
      "loss": 0.692367696762085,
      "step": 560
    },
    {
      "epoch": 0.01195670414499077,
      "grad_norm": 2.3708233226098807,
      "learning_rate": 2e-05,
      "loss": 0.6588814735412598,
      "step": 570
    },
    {
      "epoch": 0.012166470884376572,
      "grad_norm": 4.863667938523831,
      "learning_rate": 2e-05,
      "loss": 0.7014576911926269,
      "step": 580
    },
    {
      "epoch": 0.012376237623762377,
      "grad_norm": 2.1632337576331073,
      "learning_rate": 2e-05,
      "loss": 0.7060187816619873,
      "step": 590
    },
    {
      "epoch": 0.01258600436314818,
      "grad_norm": 2.1446479504758584,
      "learning_rate": 2e-05,
      "loss": 0.7127746105194092,
      "step": 600
    },
    {
      "epoch": 0.012795771102533982,
      "grad_norm": 2.103648073321234,
      "learning_rate": 2e-05,
      "loss": 0.7010409355163574,
      "step": 610
    },
    {
      "epoch": 0.013005537841919785,
      "grad_norm": 2.300281953023738,
      "learning_rate": 2e-05,
      "loss": 0.7145485877990723,
      "step": 620
    },
    {
      "epoch": 0.013215304581305588,
      "grad_norm": 2.1745603402017855,
      "learning_rate": 2e-05,
      "loss": 0.6661243438720703,
      "step": 630
    },
    {
      "epoch": 0.01342507132069139,
      "grad_norm": 3.3911386633188085,
      "learning_rate": 2e-05,
      "loss": 0.7290327072143554,
      "step": 640
    },
    {
      "epoch": 0.013634838060077195,
      "grad_norm": 2.838234667045874,
      "learning_rate": 2e-05,
      "loss": 0.7416918754577637,
      "step": 650
    },
    {
      "epoch": 0.013844604799462997,
      "grad_norm": 2.6030763255310494,
      "learning_rate": 2e-05,
      "loss": 0.6962085247039795,
      "step": 660
    },
    {
      "epoch": 0.0140543715388488,
      "grad_norm": 3.9879154447494263,
      "learning_rate": 2e-05,
      "loss": 0.7320832729339599,
      "step": 670
    },
    {
      "epoch": 0.014264138278234603,
      "grad_norm": 5.147234411879652,
      "learning_rate": 2e-05,
      "loss": 0.690255880355835,
      "step": 680
    },
    {
      "epoch": 0.014473905017620406,
      "grad_norm": 4.41792994494731,
      "learning_rate": 2e-05,
      "loss": 0.7107495784759521,
      "step": 690
    },
    {
      "epoch": 0.014683671757006208,
      "grad_norm": 4.0092748239949465,
      "learning_rate": 2e-05,
      "loss": 0.7144205093383789,
      "step": 700
    },
    {
      "epoch": 0.014893438496392013,
      "grad_norm": 3.663722199190995,
      "learning_rate": 2e-05,
      "loss": 0.678815221786499,
      "step": 710
    },
    {
      "epoch": 0.015103205235777815,
      "grad_norm": 374.6324936776676,
      "learning_rate": 2e-05,
      "loss": 0.7061020851135253,
      "step": 720
    },
    {
      "epoch": 0.015312971975163618,
      "grad_norm": 2.830711027072347,
      "learning_rate": 2e-05,
      "loss": 0.716461181640625,
      "step": 730
    },
    {
      "epoch": 0.01552273871454942,
      "grad_norm": 17.687806831387658,
      "learning_rate": 2e-05,
      "loss": 0.6958938598632812,
      "step": 740
    },
    {
      "epoch": 0.015732505453935224,
      "grad_norm": 67.22665881659304,
      "learning_rate": 2e-05,
      "loss": 0.6571521759033203,
      "step": 750
    },
    {
      "epoch": 0.015942272193321026,
      "grad_norm": 3.2787186518009888,
      "learning_rate": 2e-05,
      "loss": 0.737470006942749,
      "step": 760
    },
    {
      "epoch": 0.01615203893270683,
      "grad_norm": 5.764665619355419,
      "learning_rate": 2e-05,
      "loss": 0.6784905433654785,
      "step": 770
    },
    {
      "epoch": 0.01636180567209263,
      "grad_norm": 2.8187442313274156,
      "learning_rate": 2e-05,
      "loss": 0.698319387435913,
      "step": 780
    },
    {
      "epoch": 0.016571572411478434,
      "grad_norm": 11.052681670850987,
      "learning_rate": 2e-05,
      "loss": 0.7136563301086426,
      "step": 790
    },
    {
      "epoch": 0.01678133915086424,
      "grad_norm": 2.738088534726021,
      "learning_rate": 2e-05,
      "loss": 0.7217898368835449,
      "step": 800
    },
    {
      "epoch": 0.016991105890250043,
      "grad_norm": 22.99130603074463,
      "learning_rate": 2e-05,
      "loss": 0.6949929237365723,
      "step": 810
    },
    {
      "epoch": 0.017200872629635846,
      "grad_norm": 2.2673846343594293,
      "learning_rate": 2e-05,
      "loss": 0.6527021408081055,
      "step": 820
    },
    {
      "epoch": 0.01741063936902165,
      "grad_norm": 4.040669613703935,
      "learning_rate": 2e-05,
      "loss": 0.6478307723999024,
      "step": 830
    },
    {
      "epoch": 0.01762040610840745,
      "grad_norm": 1.7450269280537714,
      "learning_rate": 2e-05,
      "loss": 0.7058829307556153,
      "step": 840
    },
    {
      "epoch": 0.017830172847793254,
      "grad_norm": 1.3888768876108615,
      "learning_rate": 2e-05,
      "loss": 0.6917304039001465,
      "step": 850
    },
    {
      "epoch": 0.018039939587179057,
      "grad_norm": 2.497192781134032,
      "learning_rate": 2e-05,
      "loss": 0.6497130393981934,
      "step": 860
    },
    {
      "epoch": 0.01824970632656486,
      "grad_norm": 2.096909255558357,
      "learning_rate": 2e-05,
      "loss": 0.6941553115844726,
      "step": 870
    },
    {
      "epoch": 0.018459473065950662,
      "grad_norm": 1.7631454307114556,
      "learning_rate": 2e-05,
      "loss": 0.7313367366790772,
      "step": 880
    },
    {
      "epoch": 0.018669239805336465,
      "grad_norm": 1.2137215799580678,
      "learning_rate": 2e-05,
      "loss": 0.734074592590332,
      "step": 890
    },
    {
      "epoch": 0.018879006544722268,
      "grad_norm": 8.527470156498259,
      "learning_rate": 2e-05,
      "loss": 0.7139863967895508,
      "step": 900
    },
    {
      "epoch": 0.01908877328410807,
      "grad_norm": 1.150419868331103,
      "learning_rate": 2e-05,
      "loss": 0.7071406364440918,
      "step": 910
    },
    {
      "epoch": 0.019298540023493876,
      "grad_norm": 2.4598251273735596,
      "learning_rate": 2e-05,
      "loss": 0.7090801239013672,
      "step": 920
    },
    {
      "epoch": 0.01950830676287968,
      "grad_norm": 2.0085089193495707,
      "learning_rate": 2e-05,
      "loss": 0.6835993766784668,
      "step": 930
    },
    {
      "epoch": 0.019718073502265482,
      "grad_norm": 3.0778601053919536,
      "learning_rate": 2e-05,
      "loss": 0.7152371883392334,
      "step": 940
    },
    {
      "epoch": 0.019927840241651285,
      "grad_norm": 6.504260026804486,
      "learning_rate": 2e-05,
      "loss": 0.6743127346038819,
      "step": 950
    },
    {
      "epoch": 0.020137606981037087,
      "grad_norm": 1.7969857643634526,
      "learning_rate": 2e-05,
      "loss": 0.6778482437133789,
      "step": 960
    },
    {
      "epoch": 0.02034737372042289,
      "grad_norm": 29.433979140864412,
      "learning_rate": 2e-05,
      "loss": 0.6873385429382324,
      "step": 970
    },
    {
      "epoch": 0.020557140459808693,
      "grad_norm": 2.035543156408046,
      "learning_rate": 2e-05,
      "loss": 0.6699578762054443,
      "step": 980
    },
    {
      "epoch": 0.020766907199194495,
      "grad_norm": 1.1357536583710182,
      "learning_rate": 2e-05,
      "loss": 0.695745849609375,
      "step": 990
    },
    {
      "epoch": 0.020976673938580298,
      "grad_norm": 3.7046794359871447,
      "learning_rate": 2e-05,
      "loss": 0.6789576530456543,
      "step": 1000
    },
    {
      "epoch": 0.0211864406779661,
      "grad_norm": 3.130957938855976,
      "learning_rate": 2e-05,
      "loss": 0.6930006980895996,
      "step": 1010
    },
    {
      "epoch": 0.021396207417351903,
      "grad_norm": 11.276914781871405,
      "learning_rate": 2e-05,
      "loss": 0.6952384948730469,
      "step": 1020
    },
    {
      "epoch": 0.021605974156737706,
      "grad_norm": 6.337861678743805,
      "learning_rate": 2e-05,
      "loss": 0.6785932064056397,
      "step": 1030
    },
    {
      "epoch": 0.021815740896123512,
      "grad_norm": 4.784777304261159,
      "learning_rate": 2e-05,
      "loss": 0.7035085678100585,
      "step": 1040
    },
    {
      "epoch": 0.022025507635509315,
      "grad_norm": 5.679658714316126,
      "learning_rate": 2e-05,
      "loss": 0.6616763114929199,
      "step": 1050
    },
    {
      "epoch": 0.022235274374895118,
      "grad_norm": 3.478649303767567,
      "learning_rate": 2e-05,
      "loss": 0.6685798645019532,
      "step": 1060
    },
    {
      "epoch": 0.02244504111428092,
      "grad_norm": 5.634898233949741,
      "learning_rate": 2e-05,
      "loss": 0.7499291896820068,
      "step": 1070
    },
    {
      "epoch": 0.022654807853666723,
      "grad_norm": 5.773543600338411,
      "learning_rate": 2e-05,
      "loss": 0.6558557033538819,
      "step": 1080
    },
    {
      "epoch": 0.022864574593052526,
      "grad_norm": 39.700232901357666,
      "learning_rate": 2e-05,
      "loss": 0.6942555904388428,
      "step": 1090
    },
    {
      "epoch": 0.02307434133243833,
      "grad_norm": 6.2820558818505265,
      "learning_rate": 2e-05,
      "loss": 0.6771759033203125,
      "step": 1100
    },
    {
      "epoch": 0.02328410807182413,
      "grad_norm": 2.9316401080763823,
      "learning_rate": 2e-05,
      "loss": 0.6969615936279296,
      "step": 1110
    },
    {
      "epoch": 0.023493874811209934,
      "grad_norm": 2.8974904226392466,
      "learning_rate": 2e-05,
      "loss": 0.6931879997253418,
      "step": 1120
    },
    {
      "epoch": 0.023703641550595737,
      "grad_norm": 4.189761025368283,
      "learning_rate": 2e-05,
      "loss": 0.6718255043029785,
      "step": 1130
    },
    {
      "epoch": 0.02391340828998154,
      "grad_norm": 2.417352201543177,
      "learning_rate": 2e-05,
      "loss": 0.7471983909606934,
      "step": 1140
    },
    {
      "epoch": 0.024123175029367342,
      "grad_norm": 2.839365825197699,
      "learning_rate": 2e-05,
      "loss": 0.6793410301208496,
      "step": 1150
    },
    {
      "epoch": 0.024332941768753145,
      "grad_norm": 3.141367299942584,
      "learning_rate": 2e-05,
      "loss": 0.7086249351501465,
      "step": 1160
    },
    {
      "epoch": 0.02454270850813895,
      "grad_norm": 2.060642183486379,
      "learning_rate": 2e-05,
      "loss": 0.6504176139831543,
      "step": 1170
    },
    {
      "epoch": 0.024752475247524754,
      "grad_norm": 23.95719687060173,
      "learning_rate": 2e-05,
      "loss": 0.6888862609863281,
      "step": 1180
    },
    {
      "epoch": 0.024962241986910556,
      "grad_norm": 5.209642741753617,
      "learning_rate": 2e-05,
      "loss": 0.6663257598876953,
      "step": 1190
    },
    {
      "epoch": 0.02517200872629636,
      "grad_norm": 4.668122000612013,
      "learning_rate": 2e-05,
      "loss": 0.6789710521697998,
      "step": 1200
    },
    {
      "epoch": 0.025381775465682162,
      "grad_norm": 2.21317699920472,
      "learning_rate": 2e-05,
      "loss": 0.6676199913024903,
      "step": 1210
    },
    {
      "epoch": 0.025591542205067964,
      "grad_norm": 4.368293636030707,
      "learning_rate": 2e-05,
      "loss": 0.690043592453003,
      "step": 1220
    },
    {
      "epoch": 0.025801308944453767,
      "grad_norm": 3.8199119576336176,
      "learning_rate": 2e-05,
      "loss": 0.7009891033172607,
      "step": 1230
    },
    {
      "epoch": 0.02601107568383957,
      "grad_norm": 2.3977516666695142,
      "learning_rate": 2e-05,
      "loss": 0.6778505325317383,
      "step": 1240
    },
    {
      "epoch": 0.026220842423225373,
      "grad_norm": 12.32747378491102,
      "learning_rate": 2e-05,
      "loss": 0.7028230190277099,
      "step": 1250
    },
    {
      "epoch": 0.026430609162611175,
      "grad_norm": 5.6463852018519844,
      "learning_rate": 2e-05,
      "loss": 0.6816394329071045,
      "step": 1260
    },
    {
      "epoch": 0.026640375901996978,
      "grad_norm": 3.0036641576788443,
      "learning_rate": 2e-05,
      "loss": 0.6653374671936035,
      "step": 1270
    },
    {
      "epoch": 0.02685014264138278,
      "grad_norm": 3.963136105803816,
      "learning_rate": 2e-05,
      "loss": 0.70212721824646,
      "step": 1280
    },
    {
      "epoch": 0.027059909380768587,
      "grad_norm": 2.3922579167108338,
      "learning_rate": 2e-05,
      "loss": 0.6652139186859131,
      "step": 1290
    },
    {
      "epoch": 0.02726967612015439,
      "grad_norm": 1.430307897978201,
      "learning_rate": 2e-05,
      "loss": 0.649704122543335,
      "step": 1300
    },
    {
      "epoch": 0.027479442859540192,
      "grad_norm": 12.579409728745457,
      "learning_rate": 2e-05,
      "loss": 0.7093433380126953,
      "step": 1310
    },
    {
      "epoch": 0.027689209598925995,
      "grad_norm": 1.8768014730990727,
      "learning_rate": 2e-05,
      "loss": 0.6947019100189209,
      "step": 1320
    },
    {
      "epoch": 0.027898976338311798,
      "grad_norm": 4.087020619174566,
      "learning_rate": 2e-05,
      "loss": 0.6580727577209473,
      "step": 1330
    },
    {
      "epoch": 0.0281087430776976,
      "grad_norm": 2.2104247401520505,
      "learning_rate": 2e-05,
      "loss": 0.6761454582214356,
      "step": 1340
    },
    {
      "epoch": 0.028318509817083403,
      "grad_norm": 1.54359381136085,
      "learning_rate": 2e-05,
      "loss": 0.7064968109130859,
      "step": 1350
    },
    {
      "epoch": 0.028528276556469206,
      "grad_norm": 9.445161276396444,
      "learning_rate": 2e-05,
      "loss": 0.669188404083252,
      "step": 1360
    },
    {
      "epoch": 0.02873804329585501,
      "grad_norm": 2.7373325593304836,
      "learning_rate": 2e-05,
      "loss": 0.6705168724060059,
      "step": 1370
    },
    {
      "epoch": 0.02894781003524081,
      "grad_norm": 2.276476538972258,
      "learning_rate": 2e-05,
      "loss": 0.7053794860839844,
      "step": 1380
    },
    {
      "epoch": 0.029157576774626614,
      "grad_norm": 9.277889149036925,
      "learning_rate": 2e-05,
      "loss": 0.6553175926208497,
      "step": 1390
    },
    {
      "epoch": 0.029367343514012417,
      "grad_norm": 3.368267525079364,
      "learning_rate": 2e-05,
      "loss": 0.6869246959686279,
      "step": 1400
    },
    {
      "epoch": 0.029577110253398223,
      "grad_norm": 3.2226378523733645,
      "learning_rate": 2e-05,
      "loss": 0.6814833641052246,
      "step": 1410
    },
    {
      "epoch": 0.029786876992784025,
      "grad_norm": 1.930571790233891,
      "learning_rate": 2e-05,
      "loss": 0.6552035331726074,
      "step": 1420
    },
    {
      "epoch": 0.029996643732169828,
      "grad_norm": 3.5201808130997003,
      "learning_rate": 2e-05,
      "loss": 0.6717945098876953,
      "step": 1430
    },
    {
      "epoch": 0.03020641047155563,
      "grad_norm": 660.3628631506493,
      "learning_rate": 2e-05,
      "loss": 0.6539801597595215,
      "step": 1440
    },
    {
      "epoch": 0.030416177210941434,
      "grad_norm": 5.251919957664849,
      "learning_rate": 2e-05,
      "loss": 0.6830947875976563,
      "step": 1450
    },
    {
      "epoch": 0.030625943950327236,
      "grad_norm": 5.249421083313879,
      "learning_rate": 2e-05,
      "loss": 0.681186580657959,
      "step": 1460
    },
    {
      "epoch": 0.03083571068971304,
      "grad_norm": 2.0509657202306752,
      "learning_rate": 2e-05,
      "loss": 0.6720983028411865,
      "step": 1470
    },
    {
      "epoch": 0.03104547742909884,
      "grad_norm": 2.2734683613081064,
      "learning_rate": 2e-05,
      "loss": 0.67911696434021,
      "step": 1480
    },
    {
      "epoch": 0.031255244168484644,
      "grad_norm": 1.4795523059161764,
      "learning_rate": 2e-05,
      "loss": 0.6568414211273194,
      "step": 1490
    },
    {
      "epoch": 0.03146501090787045,
      "grad_norm": 1.2453772205405704,
      "learning_rate": 2e-05,
      "loss": 0.6762202262878418,
      "step": 1500
    },
    {
      "epoch": 0.03167477764725625,
      "grad_norm": 2.0010667543841123,
      "learning_rate": 2e-05,
      "loss": 0.6813722133636475,
      "step": 1510
    },
    {
      "epoch": 0.03188454438664205,
      "grad_norm": 39.56264214911426,
      "learning_rate": 2e-05,
      "loss": 0.7298669815063477,
      "step": 1520
    },
    {
      "epoch": 0.032094311126027855,
      "grad_norm": 1.9507720129351338,
      "learning_rate": 2e-05,
      "loss": 0.6247622489929199,
      "step": 1530
    },
    {
      "epoch": 0.03230407786541366,
      "grad_norm": 1.484907363192865,
      "learning_rate": 2e-05,
      "loss": 0.6759968757629394,
      "step": 1540
    },
    {
      "epoch": 0.03251384460479946,
      "grad_norm": 1.9147431404888071,
      "learning_rate": 2e-05,
      "loss": 0.6800020217895508,
      "step": 1550
    },
    {
      "epoch": 0.03272361134418526,
      "grad_norm": 3.5428313404092,
      "learning_rate": 2e-05,
      "loss": 0.6901631355285645,
      "step": 1560
    },
    {
      "epoch": 0.032933378083571066,
      "grad_norm": 4.7250773473958105,
      "learning_rate": 2e-05,
      "loss": 0.6709762573242187,
      "step": 1570
    },
    {
      "epoch": 0.03314314482295687,
      "grad_norm": 5.898810367933252,
      "learning_rate": 2e-05,
      "loss": 0.66232271194458,
      "step": 1580
    },
    {
      "epoch": 0.03335291156234268,
      "grad_norm": 1.9369336440014253,
      "learning_rate": 2e-05,
      "loss": 0.7334044933319092,
      "step": 1590
    },
    {
      "epoch": 0.03356267830172848,
      "grad_norm": 8.737915524161552,
      "learning_rate": 2e-05,
      "loss": 0.6792474269866944,
      "step": 1600
    },
    {
      "epoch": 0.033772445041114284,
      "grad_norm": 1.9836799556383249,
      "learning_rate": 2e-05,
      "loss": 0.7033929824829102,
      "step": 1610
    },
    {
      "epoch": 0.033982211780500086,
      "grad_norm": 1.1854978830042706,
      "learning_rate": 2e-05,
      "loss": 0.6635163307189942,
      "step": 1620
    },
    {
      "epoch": 0.03419197851988589,
      "grad_norm": 1.2197328294566572,
      "learning_rate": 2e-05,
      "loss": 0.6634726524353027,
      "step": 1630
    },
    {
      "epoch": 0.03440174525927169,
      "grad_norm": 1.4709834426885486,
      "learning_rate": 2e-05,
      "loss": 0.6794222831726074,
      "step": 1640
    },
    {
      "epoch": 0.034611511998657495,
      "grad_norm": 1.5366826867181045,
      "learning_rate": 2e-05,
      "loss": 0.6605133056640625,
      "step": 1650
    },
    {
      "epoch": 0.0348212787380433,
      "grad_norm": 2.9147800494231335,
      "learning_rate": 2e-05,
      "loss": 0.6418501853942871,
      "step": 1660
    },
    {
      "epoch": 0.0350310454774291,
      "grad_norm": 1.202686476162776,
      "learning_rate": 2e-05,
      "loss": 0.633223295211792,
      "step": 1670
    },
    {
      "epoch": 0.0352408122168149,
      "grad_norm": 0.9631875057504979,
      "learning_rate": 2e-05,
      "loss": 0.666954517364502,
      "step": 1680
    },
    {
      "epoch": 0.035450578956200705,
      "grad_norm": 1.4192365422706907,
      "learning_rate": 2e-05,
      "loss": 0.6560389518737793,
      "step": 1690
    },
    {
      "epoch": 0.03566034569558651,
      "grad_norm": 1.1713801817552898,
      "learning_rate": 2e-05,
      "loss": 0.641888427734375,
      "step": 1700
    },
    {
      "epoch": 0.03587011243497231,
      "grad_norm": 1.4022563568499706,
      "learning_rate": 2e-05,
      "loss": 0.7102468013763428,
      "step": 1710
    },
    {
      "epoch": 0.036079879174358113,
      "grad_norm": 2.0986091486389826,
      "learning_rate": 2e-05,
      "loss": 0.6136519908905029,
      "step": 1720
    },
    {
      "epoch": 0.036289645913743916,
      "grad_norm": 1.1634525834102774,
      "learning_rate": 2e-05,
      "loss": 0.6924193382263184,
      "step": 1730
    },
    {
      "epoch": 0.03649941265312972,
      "grad_norm": 2.081748039936143,
      "learning_rate": 2e-05,
      "loss": 0.6236453533172608,
      "step": 1740
    },
    {
      "epoch": 0.03670917939251552,
      "grad_norm": 2.484079612019191,
      "learning_rate": 2e-05,
      "loss": 0.6446088314056396,
      "step": 1750
    },
    {
      "epoch": 0.036918946131901324,
      "grad_norm": 2.142979943466157,
      "learning_rate": 2e-05,
      "loss": 0.6566426753997803,
      "step": 1760
    },
    {
      "epoch": 0.03712871287128713,
      "grad_norm": 2.727971395956995,
      "learning_rate": 2e-05,
      "loss": 0.6372243404388428,
      "step": 1770
    },
    {
      "epoch": 0.03733847961067293,
      "grad_norm": 74.34931845644519,
      "learning_rate": 2e-05,
      "loss": 0.6198902130126953,
      "step": 1780
    },
    {
      "epoch": 0.03754824635005873,
      "grad_norm": 2.066092794611252,
      "learning_rate": 2e-05,
      "loss": 0.6555921554565429,
      "step": 1790
    },
    {
      "epoch": 0.037758013089444535,
      "grad_norm": 3.0945696621385244,
      "learning_rate": 2e-05,
      "loss": 0.6903239250183105,
      "step": 1800
    },
    {
      "epoch": 0.03796777982883034,
      "grad_norm": 2.0553070965106786,
      "learning_rate": 2e-05,
      "loss": 0.6789304733276367,
      "step": 1810
    },
    {
      "epoch": 0.03817754656821614,
      "grad_norm": 1.9236355838089434,
      "learning_rate": 2e-05,
      "loss": 0.6835772514343261,
      "step": 1820
    },
    {
      "epoch": 0.03838731330760194,
      "grad_norm": 1.1647199736335656,
      "learning_rate": 2e-05,
      "loss": 0.6774552822113037,
      "step": 1830
    },
    {
      "epoch": 0.03859708004698775,
      "grad_norm": 5.471968133112629,
      "learning_rate": 2e-05,
      "loss": 0.6373446464538575,
      "step": 1840
    },
    {
      "epoch": 0.038806846786373556,
      "grad_norm": 1.4710591021325636,
      "learning_rate": 2e-05,
      "loss": 0.6392849922180176,
      "step": 1850
    },
    {
      "epoch": 0.03901661352575936,
      "grad_norm": 2.6999041311336756,
      "learning_rate": 2e-05,
      "loss": 0.6585689067840577,
      "step": 1860
    },
    {
      "epoch": 0.03922638026514516,
      "grad_norm": 1.1333104342309857,
      "learning_rate": 2e-05,
      "loss": 0.6750880241394043,
      "step": 1870
    },
    {
      "epoch": 0.039436147004530964,
      "grad_norm": 1.8072514966793143,
      "learning_rate": 2e-05,
      "loss": 0.6438388824462891,
      "step": 1880
    },
    {
      "epoch": 0.039645913743916766,
      "grad_norm": 1.0781542511331554,
      "learning_rate": 2e-05,
      "loss": 0.6553819179534912,
      "step": 1890
    },
    {
      "epoch": 0.03985568048330257,
      "grad_norm": 3.0719828705316616,
      "learning_rate": 2e-05,
      "loss": 0.6419807434082031,
      "step": 1900
    },
    {
      "epoch": 0.04006544722268837,
      "grad_norm": 1.5436345848872006,
      "learning_rate": 2e-05,
      "loss": 0.6960256099700928,
      "step": 1910
    },
    {
      "epoch": 0.040275213962074174,
      "grad_norm": 1.0011408997937439,
      "learning_rate": 2e-05,
      "loss": 0.6509086608886718,
      "step": 1920
    },
    {
      "epoch": 0.04048498070145998,
      "grad_norm": 1.294803313028635,
      "learning_rate": 2e-05,
      "loss": 0.6446637153625489,
      "step": 1930
    },
    {
      "epoch": 0.04069474744084578,
      "grad_norm": 1.4561867936361599,
      "learning_rate": 2e-05,
      "loss": 0.6471066474914551,
      "step": 1940
    },
    {
      "epoch": 0.04090451418023158,
      "grad_norm": 1.420758879294619,
      "learning_rate": 2e-05,
      "loss": 0.6668454647064209,
      "step": 1950
    },
    {
      "epoch": 0.041114280919617385,
      "grad_norm": 1.339731233422897,
      "learning_rate": 2e-05,
      "loss": 0.6452267646789551,
      "step": 1960
    },
    {
      "epoch": 0.04132404765900319,
      "grad_norm": 3.1201962896167976,
      "learning_rate": 2e-05,
      "loss": 0.643893814086914,
      "step": 1970
    },
    {
      "epoch": 0.04153381439838899,
      "grad_norm": 1.724067302763631,
      "learning_rate": 2e-05,
      "loss": 0.6555222988128662,
      "step": 1980
    },
    {
      "epoch": 0.04174358113777479,
      "grad_norm": 2.166362359535913,
      "learning_rate": 2e-05,
      "loss": 0.6665771484375,
      "step": 1990
    },
    {
      "epoch": 0.041953347877160596,
      "grad_norm": 2.744806451285816,
      "learning_rate": 2e-05,
      "loss": 0.6609567165374756,
      "step": 2000
    },
    {
      "epoch": 0.0421631146165464,
      "grad_norm": 1.1718498443702758,
      "learning_rate": 2e-05,
      "loss": 0.6552879333496093,
      "step": 2010
    },
    {
      "epoch": 0.0423728813559322,
      "grad_norm": 20.838352988266458,
      "learning_rate": 2e-05,
      "loss": 0.6685161590576172,
      "step": 2020
    },
    {
      "epoch": 0.042582648095318004,
      "grad_norm": 1.0046844547391474,
      "learning_rate": 2e-05,
      "loss": 0.6807273864746094,
      "step": 2030
    }
  ],
  "logging_steps": 10,
  "max_steps": 5935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 2035,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 963613493821440.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}