{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12449655982547407,
  "eval_steps": 500,
  "global_step": 5935,
  "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
    },
    {
      "epoch": 0.04279241483470381,
      "grad_norm": 1.0422462861274395,
      "learning_rate": 2e-05,
      "loss": 0.6735074996948243,
      "step": 2040
    },
    {
      "epoch": 0.04300218157408961,
      "grad_norm": 2.3271030192407123,
      "learning_rate": 2e-05,
      "loss": 0.7008447170257568,
      "step": 2050
    },
    {
      "epoch": 0.04321194831347541,
      "grad_norm": 1.0466437770509067,
      "learning_rate": 2e-05,
      "loss": 0.6856519222259522,
      "step": 2060
    },
    {
      "epoch": 0.043421715052861215,
      "grad_norm": 1.8010026654547595,
      "learning_rate": 2e-05,
      "loss": 0.6528799057006835,
      "step": 2070
    },
    {
      "epoch": 0.043631481792247025,
      "grad_norm": 1.3505658857593894,
      "learning_rate": 2e-05,
      "loss": 0.6574287414550781,
      "step": 2080
    },
    {
      "epoch": 0.04384124853163283,
      "grad_norm": 1.4621149054171736,
      "learning_rate": 2e-05,
      "loss": 0.6371768951416016,
      "step": 2090
    },
    {
      "epoch": 0.04405101527101863,
      "grad_norm": 1.063808823913143,
      "learning_rate": 2e-05,
      "loss": 0.6493932247161865,
      "step": 2100
    },
    {
      "epoch": 0.04426078201040443,
      "grad_norm": 1.0941475684679507,
      "learning_rate": 2e-05,
      "loss": 0.6469996452331543,
      "step": 2110
    },
    {
      "epoch": 0.044470548749790235,
      "grad_norm": 2.9013958178163155,
      "learning_rate": 2e-05,
      "loss": 0.6639802932739258,
      "step": 2120
    },
    {
      "epoch": 0.04468031548917604,
      "grad_norm": 1.3274950977190392,
      "learning_rate": 2e-05,
      "loss": 0.6418691158294678,
      "step": 2130
    },
    {
      "epoch": 0.04489008222856184,
      "grad_norm": 0.9289121119554267,
      "learning_rate": 2e-05,
      "loss": 0.6633244514465332,
      "step": 2140
    },
    {
      "epoch": 0.045099848967947644,
      "grad_norm": 1.3315812946093735,
      "learning_rate": 2e-05,
      "loss": 0.6491587638854981,
      "step": 2150
    },
    {
      "epoch": 0.045309615707333446,
      "grad_norm": 1.4176253250948148,
      "learning_rate": 2e-05,
      "loss": 0.6659342765808105,
      "step": 2160
    },
    {
      "epoch": 0.04551938244671925,
      "grad_norm": 1.417591137202183,
      "learning_rate": 2e-05,
      "loss": 0.6725269317626953,
      "step": 2170
    },
    {
      "epoch": 0.04572914918610505,
      "grad_norm": 1.13378390472357,
      "learning_rate": 2e-05,
      "loss": 0.6531981468200684,
      "step": 2180
    },
    {
      "epoch": 0.045938915925490854,
      "grad_norm": 1.4942397260353406,
      "learning_rate": 2e-05,
      "loss": 0.6550272941589356,
      "step": 2190
    },
    {
      "epoch": 0.04614868266487666,
      "grad_norm": 7.502507584526351,
      "learning_rate": 2e-05,
      "loss": 0.6380607604980468,
      "step": 2200
    },
    {
      "epoch": 0.04635844940426246,
      "grad_norm": 2.0666055878713925,
      "learning_rate": 2e-05,
      "loss": 0.6765716552734375,
      "step": 2210
    },
    {
      "epoch": 0.04656821614364826,
      "grad_norm": 3.7526650968480615,
      "learning_rate": 2e-05,
      "loss": 0.6970747947692871,
      "step": 2220
    },
    {
      "epoch": 0.046777982883034065,
      "grad_norm": 1.0839625187073827,
      "learning_rate": 2e-05,
      "loss": 0.6326667785644531,
      "step": 2230
    },
    {
      "epoch": 0.04698774962241987,
      "grad_norm": 1.4444230099278503,
      "learning_rate": 2e-05,
      "loss": 0.6718984127044678,
      "step": 2240
    },
    {
      "epoch": 0.04719751636180567,
      "grad_norm": 5.2812918783484015,
      "learning_rate": 2e-05,
      "loss": 0.6714650630950928,
      "step": 2250
    },
    {
      "epoch": 0.04740728310119147,
      "grad_norm": 1.3513967222907057,
      "learning_rate": 2e-05,
      "loss": 0.623006534576416,
      "step": 2260
    },
    {
      "epoch": 0.047617049840577276,
      "grad_norm": 1.2616632620184862,
      "learning_rate": 2e-05,
      "loss": 0.6513717174530029,
      "step": 2270
    },
    {
      "epoch": 0.04782681657996308,
      "grad_norm": 0.9944118591830172,
      "learning_rate": 2e-05,
      "loss": 0.6423852443695068,
      "step": 2280
    },
    {
      "epoch": 0.04803658331934888,
      "grad_norm": 2.5602239897084993,
      "learning_rate": 2e-05,
      "loss": 0.6237055778503418,
      "step": 2290
    },
    {
      "epoch": 0.048246350058734684,
      "grad_norm": 5.098047996535586,
      "learning_rate": 2e-05,
      "loss": 0.6563852310180665,
      "step": 2300
    },
    {
      "epoch": 0.04845611679812049,
      "grad_norm": 2.4908638567535495,
      "learning_rate": 2e-05,
      "loss": 0.6325714111328125,
      "step": 2310
    },
    {
      "epoch": 0.04866588353750629,
      "grad_norm": 7.314280111917294,
      "learning_rate": 2e-05,
      "loss": 0.6511282920837402,
      "step": 2320
    },
    {
      "epoch": 0.0488756502768921,
      "grad_norm": 3.9514336521890736,
      "learning_rate": 2e-05,
      "loss": 0.6881688594818115,
      "step": 2330
    },
    {
      "epoch": 0.0490854170162779,
      "grad_norm": 5.593476626022951,
      "learning_rate": 2e-05,
      "loss": 0.678310775756836,
      "step": 2340
    },
    {
      "epoch": 0.049295183755663705,
      "grad_norm": 1.09495038750096,
      "learning_rate": 2e-05,
      "loss": 0.6575408935546875,
      "step": 2350
    },
    {
      "epoch": 0.04950495049504951,
      "grad_norm": 1.7774709237343953,
      "learning_rate": 2e-05,
      "loss": 0.631895637512207,
      "step": 2360
    },
    {
      "epoch": 0.04971471723443531,
      "grad_norm": 6.054243872558155,
      "learning_rate": 2e-05,
      "loss": 0.6792542934417725,
      "step": 2370
    },
    {
      "epoch": 0.04992448397382111,
      "grad_norm": 10.448245720311721,
      "learning_rate": 2e-05,
      "loss": 0.6161588668823242,
      "step": 2380
    },
    {
      "epoch": 0.050134250713206915,
      "grad_norm": 4.349668514735628,
      "learning_rate": 2e-05,
      "loss": 0.6573549270629883,
      "step": 2390
    },
    {
      "epoch": 0.05034401745259272,
      "grad_norm": 1.0492871670169195,
      "learning_rate": 2e-05,
      "loss": 0.6595412254333496,
      "step": 2400
    },
    {
      "epoch": 0.05055378419197852,
      "grad_norm": 4.208402496962076,
      "learning_rate": 2e-05,
      "loss": 0.6440282821655273,
      "step": 2410
    },
    {
      "epoch": 0.050763550931364324,
      "grad_norm": 1.6377864001891818,
      "learning_rate": 2e-05,
      "loss": 0.6259332180023194,
      "step": 2420
    },
    {
      "epoch": 0.050973317670750126,
      "grad_norm": 3.5201211960751166,
      "learning_rate": 2e-05,
      "loss": 0.6841580867767334,
      "step": 2430
    },
    {
      "epoch": 0.05118308441013593,
      "grad_norm": 3.1099904753941923,
      "learning_rate": 2e-05,
      "loss": 0.6848432064056397,
      "step": 2440
    },
    {
      "epoch": 0.05139285114952173,
      "grad_norm": 5.180282203706244,
      "learning_rate": 2e-05,
      "loss": 0.6627151489257812,
      "step": 2450
    },
    {
      "epoch": 0.051602617888907534,
      "grad_norm": 4.740946754901601,
      "learning_rate": 2e-05,
      "loss": 0.6182768821716309,
      "step": 2460
    },
    {
      "epoch": 0.05181238462829334,
      "grad_norm": 4.622580865957454,
      "learning_rate": 2e-05,
      "loss": 0.6499597549438476,
      "step": 2470
    },
    {
      "epoch": 0.05202215136767914,
      "grad_norm": 5.26590300113533,
      "learning_rate": 2e-05,
      "loss": 0.6509732723236084,
      "step": 2480
    },
    {
      "epoch": 0.05223191810706494,
      "grad_norm": 19.25978934862213,
      "learning_rate": 2e-05,
      "loss": 0.6599420070648193,
      "step": 2490
    },
    {
      "epoch": 0.052441684846450745,
      "grad_norm": 6.960936314568276,
      "learning_rate": 2e-05,
      "loss": 0.6727987289428711,
      "step": 2500
    },
    {
      "epoch": 0.05265145158583655,
      "grad_norm": 1.414075477389571,
      "learning_rate": 2e-05,
      "loss": 0.635031795501709,
      "step": 2510
    },
    {
      "epoch": 0.05286121832522235,
      "grad_norm": 4.066649172039289,
      "learning_rate": 2e-05,
      "loss": 0.6597941398620606,
      "step": 2520
    },
    {
      "epoch": 0.05307098506460815,
      "grad_norm": 2.364634340031721,
      "learning_rate": 2e-05,
      "loss": 0.6823457241058349,
      "step": 2530
    },
    {
      "epoch": 0.053280751803993956,
      "grad_norm": 4.716601083750949,
      "learning_rate": 2e-05,
      "loss": 0.6498630523681641,
      "step": 2540
    },
    {
      "epoch": 0.05349051854337976,
      "grad_norm": 5.285387148880005,
      "learning_rate": 2e-05,
      "loss": 0.639859962463379,
      "step": 2550
    },
    {
      "epoch": 0.05370028528276556,
      "grad_norm": 1.5808950827421306,
      "learning_rate": 2e-05,
      "loss": 0.6604787349700928,
      "step": 2560
    },
    {
      "epoch": 0.05391005202215137,
      "grad_norm": 1.5822303036177494,
      "learning_rate": 2e-05,
      "loss": 0.6555892944335937,
      "step": 2570
    },
    {
      "epoch": 0.054119818761537174,
      "grad_norm": 2.9431810050217284,
      "learning_rate": 2e-05,
      "loss": 0.6334521293640136,
      "step": 2580
    },
    {
      "epoch": 0.054329585500922976,
      "grad_norm": 1.7788715048102133,
      "learning_rate": 2e-05,
      "loss": 0.6453732967376709,
      "step": 2590
    },
    {
      "epoch": 0.05453935224030878,
      "grad_norm": 2.50270616666147,
      "learning_rate": 2e-05,
      "loss": 0.6774460792541503,
      "step": 2600
    },
    {
      "epoch": 0.05474911897969458,
      "grad_norm": 10.707562798836856,
      "learning_rate": 2e-05,
      "loss": 0.6564057350158692,
      "step": 2610
    },
    {
      "epoch": 0.054958885719080385,
      "grad_norm": 7.655005229658258,
      "learning_rate": 2e-05,
      "loss": 0.6521840572357178,
      "step": 2620
    },
    {
      "epoch": 0.05516865245846619,
      "grad_norm": 3.2841435118348516,
      "learning_rate": 2e-05,
      "loss": 0.656426191329956,
      "step": 2630
    },
    {
      "epoch": 0.05537841919785199,
      "grad_norm": 1.4775967192316672,
      "learning_rate": 2e-05,
      "loss": 0.6167282104492188,
      "step": 2640
    },
    {
      "epoch": 0.05558818593723779,
      "grad_norm": 2.1146359434093567,
      "learning_rate": 2e-05,
      "loss": 0.6516013145446777,
      "step": 2650
    },
    {
      "epoch": 0.055797952676623595,
      "grad_norm": 3.922483673727723,
      "learning_rate": 2e-05,
      "loss": 0.6396924018859863,
      "step": 2660
    },
    {
      "epoch": 0.0560077194160094,
      "grad_norm": 3.5454952032437412,
      "learning_rate": 2e-05,
      "loss": 0.6680498123168945,
      "step": 2670
    },
    {
      "epoch": 0.0562174861553952,
      "grad_norm": 1.3429385611798919,
      "learning_rate": 2e-05,
      "loss": 0.6435057640075683,
      "step": 2680
    },
    {
      "epoch": 0.056427252894781,
      "grad_norm": 3.0538281935853266,
      "learning_rate": 2e-05,
      "loss": 0.6330676078796387,
      "step": 2690
    },
    {
      "epoch": 0.056637019634166806,
      "grad_norm": 16.829414730202203,
      "learning_rate": 2e-05,
      "loss": 0.6667710781097412,
      "step": 2700
    },
    {
      "epoch": 0.05684678637355261,
      "grad_norm": 4.39087328540998,
      "learning_rate": 2e-05,
      "loss": 0.633282995223999,
      "step": 2710
    },
    {
      "epoch": 0.05705655311293841,
      "grad_norm": 5.287462256642327,
      "learning_rate": 2e-05,
      "loss": 0.6073061943054199,
      "step": 2720
    },
    {
      "epoch": 0.057266319852324214,
      "grad_norm": 1.0838576240977882,
      "learning_rate": 2e-05,
      "loss": 0.6258288383483886,
      "step": 2730
    },
    {
      "epoch": 0.05747608659171002,
      "grad_norm": 0.9900910697266448,
      "learning_rate": 2e-05,
      "loss": 0.6326725006103515,
      "step": 2740
    },
    {
      "epoch": 0.05768585333109582,
      "grad_norm": 3.1076601428219357,
      "learning_rate": 2e-05,
      "loss": 0.6563446044921875,
      "step": 2750
    },
    {
      "epoch": 0.05789562007048162,
      "grad_norm": 1.0038588549026581,
      "learning_rate": 2e-05,
      "loss": 0.6112082481384278,
      "step": 2760
    },
    {
      "epoch": 0.058105386809867425,
      "grad_norm": 1.3442378974062874,
      "learning_rate": 2e-05,
      "loss": 0.6177424907684326,
      "step": 2770
    },
    {
      "epoch": 0.05831515354925323,
      "grad_norm": 2.8108346453547948,
      "learning_rate": 2e-05,
      "loss": 0.6193189144134521,
      "step": 2780
    },
    {
      "epoch": 0.05852492028863903,
      "grad_norm": 1.905205374029777,
      "learning_rate": 2e-05,
      "loss": 0.648875379562378,
      "step": 2790
    },
    {
      "epoch": 0.05873468702802483,
      "grad_norm": 3.74823423744243,
      "learning_rate": 2e-05,
      "loss": 0.6043904781341553,
      "step": 2800
    },
    {
      "epoch": 0.05894445376741064,
      "grad_norm": 18.168630622234243,
      "learning_rate": 2e-05,
      "loss": 0.6348827362060547,
      "step": 2810
    },
    {
      "epoch": 0.059154220506796446,
      "grad_norm": 4.41375082760251,
      "learning_rate": 2e-05,
      "loss": 0.6650748252868652,
      "step": 2820
    },
    {
      "epoch": 0.05936398724618225,
      "grad_norm": 4.666265939789063,
      "learning_rate": 2e-05,
      "loss": 0.646610164642334,
      "step": 2830
    },
    {
      "epoch": 0.05957375398556805,
      "grad_norm": 7.434065397758597,
      "learning_rate": 2e-05,
      "loss": 0.6720587730407714,
      "step": 2840
    },
    {
      "epoch": 0.059783520724953854,
      "grad_norm": 14.577190757849174,
      "learning_rate": 2e-05,
      "loss": 0.6914210319519043,
      "step": 2850
    },
    {
      "epoch": 0.059993287464339656,
      "grad_norm": 7.690469349544798,
      "learning_rate": 2e-05,
      "loss": 0.6587874889373779,
      "step": 2860
    },
    {
      "epoch": 0.06020305420372546,
      "grad_norm": 4.5317526555397265,
      "learning_rate": 2e-05,
      "loss": 0.648529052734375,
      "step": 2870
    },
    {
      "epoch": 0.06041282094311126,
      "grad_norm": 2.4738472787325754,
      "learning_rate": 2e-05,
      "loss": 0.6423982143402099,
      "step": 2880
    },
    {
      "epoch": 0.060622587682497064,
      "grad_norm": 9.389047282481764,
      "learning_rate": 2e-05,
      "loss": 0.6508796215057373,
      "step": 2890
    },
    {
      "epoch": 0.06083235442188287,
      "grad_norm": 1.319738088613341,
      "learning_rate": 2e-05,
      "loss": 0.6483406066894531,
      "step": 2900
    },
    {
      "epoch": 0.06104212116126867,
      "grad_norm": 1.3986494540414636,
      "learning_rate": 2e-05,
      "loss": 0.6722330093383789,
      "step": 2910
    },
    {
      "epoch": 0.06125188790065447,
      "grad_norm": 1.3248596727395392,
      "learning_rate": 2e-05,
      "loss": 0.6521361827850342,
      "step": 2920
    },
    {
      "epoch": 0.061461654640040275,
      "grad_norm": 8.01789018839688,
      "learning_rate": 2e-05,
      "loss": 0.6545214653015137,
      "step": 2930
    },
    {
      "epoch": 0.06167142137942608,
      "grad_norm": 1.867876365359386,
      "learning_rate": 2e-05,
      "loss": 0.6807507514953614,
      "step": 2940
    },
    {
      "epoch": 0.06188118811881188,
      "grad_norm": 1.9579123185059468,
      "learning_rate": 2e-05,
      "loss": 0.6590943813323975,
      "step": 2950
    },
    {
      "epoch": 0.06209095485819768,
      "grad_norm": 4.666099434423608,
      "learning_rate": 2e-05,
      "loss": 0.6239662170410156,
      "step": 2960
    },
    {
      "epoch": 0.062300721597583486,
      "grad_norm": 6.629107363464002,
      "learning_rate": 2e-05,
      "loss": 0.654464864730835,
      "step": 2970
    },
    {
      "epoch": 0.06251048833696929,
      "grad_norm": 1.2941207879585257,
      "learning_rate": 2e-05,
      "loss": 0.6336467742919922,
      "step": 2980
    },
    {
      "epoch": 0.06272025507635509,
      "grad_norm": 2.7567451703280867,
      "learning_rate": 2e-05,
      "loss": 0.6776866912841797,
      "step": 2990
    },
    {
      "epoch": 0.0629300218157409,
      "grad_norm": 1.8713254083285533,
      "learning_rate": 2e-05,
      "loss": 0.6494972229003906,
      "step": 3000
    },
    {
      "epoch": 0.0631397885551267,
      "grad_norm": 4.968645545352801,
      "learning_rate": 2e-05,
      "loss": 0.6557379722595215,
      "step": 3010
    },
    {
      "epoch": 0.0633495552945125,
      "grad_norm": 5.29522254711333,
      "learning_rate": 2e-05,
      "loss": 0.6359572887420655,
      "step": 3020
    },
    {
      "epoch": 0.0635593220338983,
      "grad_norm": 2.273924269870783,
      "learning_rate": 2e-05,
      "loss": 0.6379573822021485,
      "step": 3030
    },
    {
      "epoch": 0.0637690887732841,
      "grad_norm": 1.1328251395273925,
      "learning_rate": 2e-05,
      "loss": 0.6374843120574951,
      "step": 3040
    },
    {
      "epoch": 0.06397885551266991,
      "grad_norm": 15.368295822345088,
      "learning_rate": 2e-05,
      "loss": 0.6400310516357421,
      "step": 3050
    },
    {
      "epoch": 0.06418862225205571,
      "grad_norm": 1.4607094526795188,
      "learning_rate": 2e-05,
      "loss": 0.6543275356292725,
      "step": 3060
    },
    {
      "epoch": 0.06439838899144151,
      "grad_norm": 0.961287901503162,
      "learning_rate": 2e-05,
      "loss": 0.6664879798889161,
      "step": 3070
    },
    {
      "epoch": 0.06460815573082732,
      "grad_norm": 2.191551250716189,
      "learning_rate": 2e-05,
      "loss": 0.6127400398254395,
      "step": 3080
    },
    {
      "epoch": 0.06481792247021312,
      "grad_norm": 1.2296826571837989,
      "learning_rate": 2e-05,
      "loss": 0.6447425365447998,
      "step": 3090
    },
    {
      "epoch": 0.06502768920959892,
      "grad_norm": 1.4143474486735077,
      "learning_rate": 2e-05,
      "loss": 0.6479999542236328,
      "step": 3100
    },
    {
      "epoch": 0.06523745594898472,
      "grad_norm": 2.5009608440595854,
      "learning_rate": 2e-05,
      "loss": 0.6653982162475586,
      "step": 3110
    },
    {
      "epoch": 0.06544722268837053,
      "grad_norm": 1.2857554395308046,
      "learning_rate": 2e-05,
      "loss": 0.6534938812255859,
      "step": 3120
    },
    {
      "epoch": 0.06565698942775633,
      "grad_norm": 1.083653801140928,
      "learning_rate": 2e-05,
      "loss": 0.6390992641448975,
      "step": 3130
    },
    {
      "epoch": 0.06586675616714213,
      "grad_norm": 1.8441477694886703,
      "learning_rate": 2e-05,
      "loss": 0.6085839748382569,
      "step": 3140
    },
    {
      "epoch": 0.06607652290652793,
      "grad_norm": 2.533533734029522,
      "learning_rate": 2e-05,
      "loss": 0.6358766555786133,
      "step": 3150
    },
    {
      "epoch": 0.06628628964591374,
      "grad_norm": 1.141802239759352,
      "learning_rate": 2e-05,
      "loss": 0.6275746822357178,
      "step": 3160
    },
    {
      "epoch": 0.06649605638529954,
      "grad_norm": 1.4525638657170827,
      "learning_rate": 2e-05,
      "loss": 0.6507892608642578,
      "step": 3170
    },
    {
      "epoch": 0.06670582312468536,
      "grad_norm": 1.1588031097230553,
      "learning_rate": 2e-05,
      "loss": 0.6174406051635742,
      "step": 3180
    },
    {
      "epoch": 0.06691558986407116,
      "grad_norm": 2.5106093744433733,
      "learning_rate": 2e-05,
      "loss": 0.6564050674438476,
      "step": 3190
    },
    {
      "epoch": 0.06712535660345696,
      "grad_norm": 1.5797241142582132,
      "learning_rate": 2e-05,
      "loss": 0.6210603713989258,
      "step": 3200
    },
    {
      "epoch": 0.06733512334284276,
      "grad_norm": 1.1249954656400734,
      "learning_rate": 2e-05,
      "loss": 0.649308729171753,
      "step": 3210
    },
    {
      "epoch": 0.06754489008222857,
      "grad_norm": 2.20992490602421,
      "learning_rate": 2e-05,
      "loss": 0.6323455810546875,
      "step": 3220
    },
    {
      "epoch": 0.06775465682161437,
      "grad_norm": 4.469544409281363,
      "learning_rate": 2e-05,
      "loss": 0.6276360988616944,
      "step": 3230
    },
    {
      "epoch": 0.06796442356100017,
      "grad_norm": 1.5415451708593162,
      "learning_rate": 2e-05,
      "loss": 0.6344407558441162,
      "step": 3240
    },
    {
      "epoch": 0.06817419030038598,
      "grad_norm": 1.5949984481189907,
      "learning_rate": 2e-05,
      "loss": 0.6543967247009277,
      "step": 3250
    },
    {
      "epoch": 0.06838395703977178,
      "grad_norm": 4.337420652054659,
      "learning_rate": 2e-05,
      "loss": 0.6937810897827148,
      "step": 3260
    },
    {
      "epoch": 0.06859372377915758,
      "grad_norm": 5.830490852479801,
      "learning_rate": 2e-05,
      "loss": 0.6038520812988282,
      "step": 3270
    },
    {
      "epoch": 0.06880349051854338,
      "grad_norm": 4.631265483483824,
      "learning_rate": 2e-05,
      "loss": 0.5950284004211426,
      "step": 3280
    },
    {
      "epoch": 0.06901325725792919,
      "grad_norm": 1.3176387751671785,
      "learning_rate": 2e-05,
      "loss": 0.6093755245208741,
      "step": 3290
    },
    {
      "epoch": 0.06922302399731499,
      "grad_norm": 1.4717375829615742,
      "learning_rate": 2e-05,
      "loss": 0.6016129970550537,
      "step": 3300
    },
    {
      "epoch": 0.06943279073670079,
      "grad_norm": 1.030735076162183,
      "learning_rate": 2e-05,
      "loss": 0.6150879859924316,
      "step": 3310
    },
    {
      "epoch": 0.0696425574760866,
      "grad_norm": 1.0121046050263844,
      "learning_rate": 2e-05,
      "loss": 0.6329584121704102,
      "step": 3320
    },
    {
      "epoch": 0.0698523242154724,
      "grad_norm": 1.0842693289830008,
      "learning_rate": 2e-05,
      "loss": 0.6599737167358398,
      "step": 3330
    },
    {
      "epoch": 0.0700620909548582,
      "grad_norm": 0.9212902872267731,
      "learning_rate": 2e-05,
      "loss": 0.5934088706970215,
      "step": 3340
    },
    {
      "epoch": 0.070271857694244,
      "grad_norm": 1.4550776446428906,
      "learning_rate": 2e-05,
      "loss": 0.6075980186462402,
      "step": 3350
    },
    {
      "epoch": 0.0704816244336298,
      "grad_norm": 1.5913770936750142,
      "learning_rate": 2e-05,
      "loss": 0.6325076580047607,
      "step": 3360
    },
    {
      "epoch": 0.07069139117301561,
      "grad_norm": 1.7644895084736516,
      "learning_rate": 2e-05,
      "loss": 0.6227485656738281,
      "step": 3370
    },
    {
      "epoch": 0.07090115791240141,
      "grad_norm": 1.6486442155005259,
      "learning_rate": 2e-05,
      "loss": 0.6122469902038574,
      "step": 3380
    },
    {
      "epoch": 0.07111092465178721,
      "grad_norm": 2.1016110075472705,
      "learning_rate": 2e-05,
      "loss": 0.6439970970153809,
      "step": 3390
    },
    {
      "epoch": 0.07132069139117302,
      "grad_norm": 1.1474703819153493,
      "learning_rate": 2e-05,
      "loss": 0.6407495021820069,
      "step": 3400
    },
    {
      "epoch": 0.07153045813055882,
      "grad_norm": 1.6594212119031326,
      "learning_rate": 2e-05,
      "loss": 0.6275229454040527,
      "step": 3410
    },
    {
      "epoch": 0.07174022486994462,
      "grad_norm": 1.5566863992273559,
      "learning_rate": 2e-05,
      "loss": 0.6257375717163086,
      "step": 3420
    },
    {
      "epoch": 0.07194999160933042,
      "grad_norm": 1.3482677008573798,
      "learning_rate": 2e-05,
      "loss": 0.600681209564209,
      "step": 3430
    },
    {
      "epoch": 0.07215975834871623,
      "grad_norm": 1.1085116434458875,
      "learning_rate": 2e-05,
      "loss": 0.6113192558288574,
      "step": 3440
    },
    {
      "epoch": 0.07236952508810203,
      "grad_norm": 1.0470200412519313,
      "learning_rate": 2e-05,
      "loss": 0.6392256736755371,
      "step": 3450
    },
    {
      "epoch": 0.07257929182748783,
      "grad_norm": 1.208091709460663,
      "learning_rate": 2e-05,
      "loss": 0.6172694206237793,
      "step": 3460
    },
    {
      "epoch": 0.07278905856687364,
      "grad_norm": 1.391808307290741,
      "learning_rate": 2e-05,
      "loss": 0.6202544212341309,
      "step": 3470
    },
    {
      "epoch": 0.07299882530625944,
      "grad_norm": 1.3324061305312844,
      "learning_rate": 2e-05,
      "loss": 0.6044974327087402,
      "step": 3480
    },
    {
      "epoch": 0.07320859204564524,
      "grad_norm": 1.603768952138755,
      "learning_rate": 2e-05,
      "loss": 0.6273996353149414,
      "step": 3490
    },
    {
      "epoch": 0.07341835878503104,
      "grad_norm": 1.1515236435171623,
      "learning_rate": 2e-05,
      "loss": 0.6415897369384765,
      "step": 3500
    },
    {
      "epoch": 0.07362812552441685,
      "grad_norm": 1.2465427409602832,
      "learning_rate": 2e-05,
      "loss": 0.6239260673522949,
      "step": 3510
    },
    {
      "epoch": 0.07383789226380265,
      "grad_norm": 3.195547073882619,
      "learning_rate": 2e-05,
      "loss": 0.6058053493499755,
      "step": 3520
    },
    {
      "epoch": 0.07404765900318845,
      "grad_norm": 1.2612335270585324,
      "learning_rate": 2e-05,
      "loss": 0.6198372840881348,
      "step": 3530
    },
    {
      "epoch": 0.07425742574257425,
      "grad_norm": 0.9419472600007744,
      "learning_rate": 2e-05,
      "loss": 0.6054319381713867,
      "step": 3540
    },
    {
      "epoch": 0.07446719248196006,
      "grad_norm": 1.089531938815818,
      "learning_rate": 2e-05,
      "loss": 0.6167111873626709,
      "step": 3550
    },
    {
      "epoch": 0.07467695922134586,
      "grad_norm": 1.3477436055009677,
      "learning_rate": 2e-05,
      "loss": 0.6194707870483398,
      "step": 3560
    },
    {
      "epoch": 0.07488672596073166,
      "grad_norm": 1.4774128762036365,
      "learning_rate": 2e-05,
      "loss": 0.6233315944671631,
      "step": 3570
    },
    {
      "epoch": 0.07509649270011746,
      "grad_norm": 2.180364295402137,
      "learning_rate": 2e-05,
      "loss": 0.6441976070404053,
      "step": 3580
    },
    {
      "epoch": 0.07530625943950327,
      "grad_norm": 2.348014561727009,
      "learning_rate": 2e-05,
      "loss": 0.622609281539917,
      "step": 3590
    },
    {
      "epoch": 0.07551602617888907,
      "grad_norm": 2.0546033771693244,
      "learning_rate": 2e-05,
      "loss": 0.6218820571899414,
      "step": 3600
    },
    {
      "epoch": 0.07572579291827487,
      "grad_norm": 3.948630055653689,
      "learning_rate": 2e-05,
      "loss": 0.6077123165130616,
      "step": 3610
    },
    {
      "epoch": 0.07593555965766068,
      "grad_norm": 1.425460243175804,
      "learning_rate": 2e-05,
      "loss": 0.6164697170257568,
      "step": 3620
    },
    {
      "epoch": 0.07614532639704648,
      "grad_norm": 2.8560957332391355,
      "learning_rate": 2e-05,
      "loss": 0.6065490245819092,
      "step": 3630
    },
    {
      "epoch": 0.07635509313643228,
      "grad_norm": 1.4382360748720746,
      "learning_rate": 2e-05,
      "loss": 0.631071949005127,
      "step": 3640
    },
    {
      "epoch": 0.07656485987581808,
      "grad_norm": 2.4612191373604095,
      "learning_rate": 2e-05,
      "loss": 0.6297952175140381,
      "step": 3650
    },
    {
      "epoch": 0.07677462661520389,
      "grad_norm": 0.9581431067857364,
      "learning_rate": 2e-05,
      "loss": 0.6396968841552735,
      "step": 3660
    },
    {
      "epoch": 0.0769843933545897,
      "grad_norm": 0.84258061503981,
      "learning_rate": 2e-05,
      "loss": 0.6055449962615966,
      "step": 3670
    },
    {
      "epoch": 0.0771941600939755,
      "grad_norm": 3.58877286495232,
      "learning_rate": 2e-05,
      "loss": 0.6315743923187256,
      "step": 3680
    },
    {
      "epoch": 0.07740392683336131,
      "grad_norm": 1.293421173281621,
      "learning_rate": 2e-05,
      "loss": 0.6112199783325195,
      "step": 3690
    },
    {
      "epoch": 0.07761369357274711,
      "grad_norm": 1.5684550738154965,
      "learning_rate": 2e-05,
      "loss": 0.6496015071868897,
      "step": 3700
    },
    {
      "epoch": 0.07782346031213291,
      "grad_norm": 1.2200393519933037,
      "learning_rate": 2e-05,
      "loss": 0.5959799766540528,
      "step": 3710
    },
    {
      "epoch": 0.07803322705151872,
      "grad_norm": 1.0163116397855736,
      "learning_rate": 2e-05,
      "loss": 0.6234962463378906,
      "step": 3720
    },
    {
      "epoch": 0.07824299379090452,
      "grad_norm": 1.717517892283433,
      "learning_rate": 2e-05,
      "loss": 0.616788673400879,
      "step": 3730
    },
    {
      "epoch": 0.07845276053029032,
      "grad_norm": 1.2080101695844014,
      "learning_rate": 2e-05,
      "loss": 0.6270909786224366,
      "step": 3740
    },
    {
      "epoch": 0.07866252726967612,
      "grad_norm": 1.2642428576235951,
      "learning_rate": 2e-05,
      "loss": 0.634865665435791,
      "step": 3750
    },
    {
      "epoch": 0.07887229400906193,
      "grad_norm": 1.7755472921599644,
      "learning_rate": 2e-05,
      "loss": 0.6161201477050782,
      "step": 3760
    },
    {
      "epoch": 0.07908206074844773,
      "grad_norm": 1.1611024568844615,
      "learning_rate": 2e-05,
      "loss": 0.6064630508422851,
      "step": 3770
    },
    {
      "epoch": 0.07929182748783353,
      "grad_norm": 1.3367012479877545,
      "learning_rate": 2e-05,
      "loss": 0.6257253170013428,
      "step": 3780
    },
    {
      "epoch": 0.07950159422721934,
      "grad_norm": 1.0410919406167345,
      "learning_rate": 2e-05,
      "loss": 0.6141583442687988,
      "step": 3790
    },
    {
      "epoch": 0.07971136096660514,
      "grad_norm": 0.9751600085611505,
      "learning_rate": 2e-05,
      "loss": 0.6254030227661133,
      "step": 3800
    },
    {
      "epoch": 0.07992112770599094,
      "grad_norm": 1.4908385976253524,
      "learning_rate": 2e-05,
      "loss": 0.6212417602539062,
      "step": 3810
    },
    {
      "epoch": 0.08013089444537674,
      "grad_norm": 1.3485953686507417,
      "learning_rate": 2e-05,
      "loss": 0.6176995754241943,
      "step": 3820
    },
    {
      "epoch": 0.08034066118476255,
      "grad_norm": 1.2206130924099254,
      "learning_rate": 2e-05,
      "loss": 0.5810882568359375,
      "step": 3830
    },
    {
      "epoch": 0.08055042792414835,
      "grad_norm": 1.6334570441244893,
      "learning_rate": 2e-05,
      "loss": 0.6379716873168946,
      "step": 3840
    },
    {
      "epoch": 0.08076019466353415,
      "grad_norm": 1.3369545468363697,
      "learning_rate": 2e-05,
      "loss": 0.6377719879150391,
      "step": 3850
    },
    {
      "epoch": 0.08096996140291995,
      "grad_norm": 1.8818167536690795,
      "learning_rate": 2e-05,
      "loss": 0.6293519496917724,
      "step": 3860
    },
    {
      "epoch": 0.08117972814230576,
      "grad_norm": 1.4023282025205475,
      "learning_rate": 2e-05,
      "loss": 0.6685831069946289,
      "step": 3870
    },
    {
      "epoch": 0.08138949488169156,
      "grad_norm": 1.2005685243156285,
      "learning_rate": 2e-05,
      "loss": 0.6329050064086914,
      "step": 3880
    },
    {
      "epoch": 0.08159926162107736,
      "grad_norm": 1.0700487697578784,
      "learning_rate": 2e-05,
      "loss": 0.6079745292663574,
      "step": 3890
    },
    {
      "epoch": 0.08180902836046317,
      "grad_norm": 0.9160942375676819,
      "learning_rate": 2e-05,
      "loss": 0.6010311126708985,
      "step": 3900
    },
    {
      "epoch": 0.08201879509984897,
      "grad_norm": 0.9783151957554443,
      "learning_rate": 2e-05,
      "loss": 0.5906445503234863,
      "step": 3910
    },
    {
      "epoch": 0.08222856183923477,
      "grad_norm": 2.1241581007326023,
      "learning_rate": 2e-05,
      "loss": 0.6245768547058106,
      "step": 3920
    },
    {
      "epoch": 0.08243832857862057,
      "grad_norm": 1.1142880534951982,
      "learning_rate": 2e-05,
      "loss": 0.6391409873962403,
      "step": 3930
    },
    {
      "epoch": 0.08264809531800638,
      "grad_norm": 5.801643529836836,
      "learning_rate": 2e-05,
      "loss": 0.5939390182495117,
      "step": 3940
    },
    {
      "epoch": 0.08285786205739218,
      "grad_norm": 1.1381215245542802,
      "learning_rate": 2e-05,
      "loss": 0.6113842964172364,
      "step": 3950
    },
    {
      "epoch": 0.08306762879677798,
      "grad_norm": 1.0450250941065826,
      "learning_rate": 2e-05,
      "loss": 0.640468978881836,
      "step": 3960
    },
    {
      "epoch": 0.08327739553616378,
      "grad_norm": 0.9218406597772941,
      "learning_rate": 2e-05,
      "loss": 0.5912976264953613,
      "step": 3970
    },
    {
      "epoch": 0.08348716227554959,
      "grad_norm": 0.9837293096882525,
      "learning_rate": 2e-05,
      "loss": 0.6223003387451171,
      "step": 3980
    },
    {
      "epoch": 0.08369692901493539,
      "grad_norm": 0.990643501319215,
      "learning_rate": 2e-05,
      "loss": 0.6180656433105469,
      "step": 3990
    },
    {
      "epoch": 0.08390669575432119,
      "grad_norm": 0.9212243888272387,
      "learning_rate": 2e-05,
      "loss": 0.6109961986541748,
      "step": 4000
    },
    {
      "epoch": 0.084116462493707,
      "grad_norm": 1.0851483978977774,
      "learning_rate": 2e-05,
      "loss": 0.6065635681152344,
      "step": 4010
    },
    {
      "epoch": 0.0843262292330928,
      "grad_norm": 1.1576097032428736,
      "learning_rate": 2e-05,
      "loss": 0.5832537651062012,
      "step": 4020
    },
    {
      "epoch": 0.0845359959724786,
      "grad_norm": 1.8856590779704239,
      "learning_rate": 2e-05,
      "loss": 0.6252013683319092,
      "step": 4030
    },
    {
      "epoch": 0.0847457627118644,
      "grad_norm": 0.9200345861822318,
      "learning_rate": 2e-05,
      "loss": 0.5908494472503663,
      "step": 4040
    },
    {
      "epoch": 0.0849555294512502,
      "grad_norm": 0.9332338359441046,
      "learning_rate": 2e-05,
      "loss": 0.5820940971374512,
      "step": 4050
    },
    {
      "epoch": 0.08516529619063601,
      "grad_norm": 2.7597827875708423,
      "learning_rate": 2e-05,
      "loss": 0.6051089286804199,
      "step": 4060
    },
    {
      "epoch": 0.08537506293002181,
      "grad_norm": 1.2095388396803786,
      "learning_rate": 2e-05,
      "loss": 0.6083034038543701,
      "step": 4070
    },
    {
      "epoch": 0.08558482966940761,
      "grad_norm": 0.880204726346596,
      "learning_rate": 2e-05,
      "loss": 0.5899758815765381,
      "step": 4080
    },
    {
      "epoch": 0.08579459640879342,
      "grad_norm": 13.99737327303483,
      "learning_rate": 2e-05,
      "loss": 0.6198636531829834,
      "step": 4090
    },
    {
      "epoch": 0.08600436314817922,
      "grad_norm": 1.9394187728290382,
      "learning_rate": 2e-05,
      "loss": 0.5917937278747558,
      "step": 4100
    },
    {
      "epoch": 0.08621412988756502,
      "grad_norm": 2.1910617512998343,
      "learning_rate": 2e-05,
      "loss": 0.5839457511901855,
      "step": 4110
    },
    {
      "epoch": 0.08642389662695082,
      "grad_norm": 2.458567419264388,
      "learning_rate": 2e-05,
      "loss": 0.6433148860931397,
      "step": 4120
    },
    {
      "epoch": 0.08663366336633663,
      "grad_norm": 1.2519100126455327,
      "learning_rate": 2e-05,
      "loss": 0.5895092010498046,
      "step": 4130
    },
    {
      "epoch": 0.08684343010572243,
      "grad_norm": 1.5648241470149864,
      "learning_rate": 2e-05,
      "loss": 0.6126773834228516,
      "step": 4140
    },
    {
      "epoch": 0.08705319684510823,
      "grad_norm": 1.4911748241638616,
      "learning_rate": 2e-05,
      "loss": 0.6426522731781006,
      "step": 4150
    },
    {
      "epoch": 0.08726296358449405,
      "grad_norm": 1.1804016879572379,
      "learning_rate": 2e-05,
      "loss": 0.6006235122680664,
      "step": 4160
    },
    {
      "epoch": 0.08747273032387985,
      "grad_norm": 1.1725890052832575,
      "learning_rate": 2e-05,
      "loss": 0.6514087200164795,
      "step": 4170
    },
    {
      "epoch": 0.08768249706326565,
      "grad_norm": 15.673057708262041,
      "learning_rate": 2e-05,
      "loss": 0.6518727779388428,
      "step": 4180
    },
    {
      "epoch": 0.08789226380265146,
      "grad_norm": 1.6006242501053345,
      "learning_rate": 2e-05,
      "loss": 0.6194744110107422,
      "step": 4190
    },
    {
      "epoch": 0.08810203054203726,
      "grad_norm": 1.1276936228169603,
      "learning_rate": 2e-05,
      "loss": 0.6109624385833741,
      "step": 4200
    },
    {
      "epoch": 0.08831179728142306,
      "grad_norm": 1.584446038250469,
      "learning_rate": 2e-05,
      "loss": 0.6263072490692139,
      "step": 4210
    },
    {
      "epoch": 0.08852156402080887,
      "grad_norm": 1.2221015879302377,
      "learning_rate": 2e-05,
      "loss": 0.6244864463806152,
      "step": 4220
    },
    {
      "epoch": 0.08873133076019467,
      "grad_norm": 2.9036287459628856,
      "learning_rate": 2e-05,
      "loss": 0.586827564239502,
      "step": 4230
    },
    {
      "epoch": 0.08894109749958047,
      "grad_norm": 2.188502843775794,
      "learning_rate": 2e-05,
      "loss": 0.6110862255096435,
      "step": 4240
    },
    {
      "epoch": 0.08915086423896627,
      "grad_norm": 1.4652217633255118,
      "learning_rate": 2e-05,
      "loss": 0.6276896476745606,
      "step": 4250
    },
    {
      "epoch": 0.08936063097835208,
      "grad_norm": 0.9342440880753969,
      "learning_rate": 2e-05,
      "loss": 0.5945404052734375,
      "step": 4260
    },
    {
      "epoch": 0.08957039771773788,
      "grad_norm": 1.54669126712079,
      "learning_rate": 2e-05,
      "loss": 0.6257088661193848,
      "step": 4270
    },
    {
      "epoch": 0.08978016445712368,
      "grad_norm": 1.2526285060658828,
      "learning_rate": 2e-05,
      "loss": 0.5948711395263672,
      "step": 4280
    },
    {
      "epoch": 0.08998993119650948,
      "grad_norm": 2.2998188765928975,
      "learning_rate": 2e-05,
      "loss": 0.587238073348999,
      "step": 4290
    },
    {
      "epoch": 0.09019969793589529,
      "grad_norm": 0.9476221451941326,
      "learning_rate": 2e-05,
      "loss": 0.6262912750244141,
      "step": 4300
    },
    {
      "epoch": 0.09040946467528109,
      "grad_norm": 1.9858074065406148,
      "learning_rate": 2e-05,
      "loss": 0.6163738250732422,
      "step": 4310
    },
    {
      "epoch": 0.09061923141466689,
      "grad_norm": 1.4574692039905501,
      "learning_rate": 2e-05,
      "loss": 0.5749715328216553,
      "step": 4320
    },
    {
      "epoch": 0.0908289981540527,
      "grad_norm": 1.1822905265235701,
      "learning_rate": 2e-05,
      "loss": 0.6017665386199951,
      "step": 4330
    },
    {
      "epoch": 0.0910387648934385,
      "grad_norm": 2.6178294462719007,
      "learning_rate": 2e-05,
      "loss": 0.598384952545166,
      "step": 4340
    },
    {
      "epoch": 0.0912485316328243,
      "grad_norm": 1.3915518872207258,
      "learning_rate": 2e-05,
      "loss": 0.6022363662719726,
      "step": 4350
    },
    {
      "epoch": 0.0914582983722101,
      "grad_norm": 1.1267311162756137,
      "learning_rate": 2e-05,
      "loss": 0.6115863800048829,
      "step": 4360
    },
    {
      "epoch": 0.0916680651115959,
      "grad_norm": 1.3452371147848081,
      "learning_rate": 2e-05,
      "loss": 0.5858843803405762,
      "step": 4370
    },
    {
      "epoch": 0.09187783185098171,
      "grad_norm": 2.090534029998743,
      "learning_rate": 2e-05,
      "loss": 0.6155561447143555,
      "step": 4380
    },
    {
      "epoch": 0.09208759859036751,
      "grad_norm": 1.0662968091332854,
      "learning_rate": 2e-05,
      "loss": 0.5968133449554444,
      "step": 4390
    },
    {
      "epoch": 0.09229736532975331,
      "grad_norm": 1.012479952015296,
      "learning_rate": 2e-05,
      "loss": 0.6149543762207031,
      "step": 4400
    },
    {
      "epoch": 0.09250713206913912,
      "grad_norm": 2.0908100806982555,
      "learning_rate": 2e-05,
      "loss": 0.585263442993164,
      "step": 4410
    },
    {
      "epoch": 0.09271689880852492,
      "grad_norm": 24.463013245437715,
      "learning_rate": 2e-05,
      "loss": 0.5892629623413086,
      "step": 4420
    },
    {
      "epoch": 0.09292666554791072,
      "grad_norm": 1.3925974703363353,
      "learning_rate": 2e-05,
      "loss": 0.6281230926513672,
      "step": 4430
    },
    {
      "epoch": 0.09313643228729653,
      "grad_norm": 2.374116636980276,
      "learning_rate": 2e-05,
      "loss": 0.6405997276306152,
      "step": 4440
    },
    {
      "epoch": 0.09334619902668233,
      "grad_norm": 2.9038231624553967,
      "learning_rate": 2e-05,
      "loss": 0.5980388641357421,
      "step": 4450
    },
    {
      "epoch": 0.09355596576606813,
      "grad_norm": 1.6107356306209257,
      "learning_rate": 2e-05,
      "loss": 0.6009420394897461,
      "step": 4460
    },
    {
      "epoch": 0.09376573250545393,
      "grad_norm": 3.1236022163918205,
      "learning_rate": 2e-05,
      "loss": 0.5700675010681152,
      "step": 4470
    },
    {
      "epoch": 0.09397549924483974,
      "grad_norm": 4.687484342274098,
      "learning_rate": 2e-05,
      "loss": 0.6176850318908691,
      "step": 4480
    },
    {
      "epoch": 0.09418526598422554,
      "grad_norm": 1.6823532790223308,
      "learning_rate": 2e-05,
      "loss": 0.6400545120239258,
      "step": 4490
    },
    {
      "epoch": 0.09439503272361134,
      "grad_norm": 2.384482481608329,
      "learning_rate": 2e-05,
      "loss": 0.6063694000244141,
      "step": 4500
    },
    {
      "epoch": 0.09460479946299714,
      "grad_norm": 2.2944313464564225,
      "learning_rate": 2e-05,
      "loss": 0.619162654876709,
      "step": 4510
    },
    {
      "epoch": 0.09481456620238295,
      "grad_norm": 1.3800436816178512,
      "learning_rate": 2e-05,
      "loss": 0.6205848693847656,
      "step": 4520
    },
    {
      "epoch": 0.09502433294176875,
      "grad_norm": 1.121425313214809,
      "learning_rate": 2e-05,
      "loss": 0.5997221946716309,
      "step": 4530
    },
    {
      "epoch": 0.09523409968115455,
      "grad_norm": 1.4874578729489585,
      "learning_rate": 2e-05,
      "loss": 0.650821590423584,
      "step": 4540
    },
    {
      "epoch": 0.09544386642054035,
      "grad_norm": 3.4404647545988687,
      "learning_rate": 2e-05,
      "loss": 0.5590650081634522,
      "step": 4550
    },
    {
      "epoch": 0.09565363315992616,
      "grad_norm": 0.7928852424235432,
      "learning_rate": 2e-05,
      "loss": 0.6436666011810303,
      "step": 4560
    },
    {
      "epoch": 0.09586339989931196,
      "grad_norm": 1.0186786828865984,
      "learning_rate": 2e-05,
      "loss": 0.5799013137817383,
      "step": 4570
    },
    {
      "epoch": 0.09607316663869776,
      "grad_norm": 0.9674204825763407,
      "learning_rate": 2e-05,
      "loss": 0.5709643363952637,
      "step": 4580
    },
    {
      "epoch": 0.09628293337808357,
      "grad_norm": 0.9123899428954066,
      "learning_rate": 2e-05,
      "loss": 0.5932353019714356,
      "step": 4590
    },
    {
      "epoch": 0.09649270011746937,
      "grad_norm": 1.0977780539857185,
      "learning_rate": 2e-05,
      "loss": 0.6155364990234375,
      "step": 4600
    },
    {
      "epoch": 0.09670246685685517,
      "grad_norm": 0.8784644285863972,
      "learning_rate": 2e-05,
      "loss": 0.5827277660369873,
      "step": 4610
    },
    {
      "epoch": 0.09691223359624097,
      "grad_norm": 0.9479104974332059,
      "learning_rate": 2e-05,
      "loss": 0.6517523765563965,
      "step": 4620
    },
    {
      "epoch": 0.09712200033562678,
      "grad_norm": 1.4989206042086718,
      "learning_rate": 2e-05,
      "loss": 0.627266263961792,
      "step": 4630
    },
    {
      "epoch": 0.09733176707501258,
      "grad_norm": 1.4414272285427632,
      "learning_rate": 2e-05,
      "loss": 0.6052295684814453,
      "step": 4640
    },
    {
      "epoch": 0.0975415338143984,
      "grad_norm": 1.2936914842609621,
      "learning_rate": 2e-05,
      "loss": 0.6232361793518066,
      "step": 4650
    },
    {
      "epoch": 0.0977513005537842,
      "grad_norm": 2.6763593778706896,
      "learning_rate": 2e-05,
      "loss": 0.5736616134643555,
      "step": 4660
    },
    {
      "epoch": 0.09796106729317,
      "grad_norm": 4.9323630958755915,
      "learning_rate": 2e-05,
      "loss": 0.5895534515380859,
      "step": 4670
    },
    {
      "epoch": 0.0981708340325558,
      "grad_norm": 5.390349224539633,
      "learning_rate": 2e-05,
      "loss": 0.5882411479949952,
      "step": 4680
    },
    {
      "epoch": 0.0983806007719416,
      "grad_norm": 3.6167575999336226,
      "learning_rate": 2e-05,
      "loss": 0.5855002403259277,
      "step": 4690
    },
    {
      "epoch": 0.09859036751132741,
      "grad_norm": 5.473090445718808,
      "learning_rate": 2e-05,
      "loss": 0.6358242034912109,
      "step": 4700
    },
    {
      "epoch": 0.09880013425071321,
      "grad_norm": 1.2271635714244669,
      "learning_rate": 2e-05,
      "loss": 0.5759749412536621,
      "step": 4710
    },
    {
      "epoch": 0.09900990099009901,
      "grad_norm": 1.9854527274279405,
      "learning_rate": 2e-05,
      "loss": 0.6502627372741699,
      "step": 4720
    },
    {
      "epoch": 0.09921966772948482,
      "grad_norm": 1.3908654896898205,
      "learning_rate": 2e-05,
      "loss": 0.6431674003601074,
      "step": 4730
    },
    {
      "epoch": 0.09942943446887062,
      "grad_norm": 2.05555311175049,
      "learning_rate": 2e-05,
      "loss": 0.5834898948669434,
      "step": 4740
    },
    {
      "epoch": 0.09963920120825642,
      "grad_norm": 1.0867268082115882,
      "learning_rate": 2e-05,
      "loss": 0.5954546928405762,
      "step": 4750
    },
    {
      "epoch": 0.09984896794764223,
      "grad_norm": 1.5199611963172561,
      "learning_rate": 2e-05,
      "loss": 0.6293166637420654,
      "step": 4760
    },
    {
      "epoch": 0.10005873468702803,
      "grad_norm": 1.8011963951072958,
      "learning_rate": 2e-05,
      "loss": 0.6096612453460694,
      "step": 4770
    },
    {
      "epoch": 0.10026850142641383,
      "grad_norm": 1.0463461910078327,
      "learning_rate": 2e-05,
      "loss": 0.6111156463623046,
      "step": 4780
    },
    {
      "epoch": 0.10047826816579963,
      "grad_norm": 0.9586640118425774,
      "learning_rate": 2e-05,
      "loss": 0.5941665649414063,
      "step": 4790
    },
    {
      "epoch": 0.10068803490518544,
      "grad_norm": 1.4987846100151212,
      "learning_rate": 2e-05,
      "loss": 0.5871945858001709,
      "step": 4800
    },
    {
      "epoch": 0.10089780164457124,
      "grad_norm": 1.2560346834896248,
      "learning_rate": 2e-05,
      "loss": 0.5902459621429443,
      "step": 4810
    },
    {
      "epoch": 0.10110756838395704,
      "grad_norm": 1.5169412620900642,
      "learning_rate": 2e-05,
      "loss": 0.6164841651916504,
      "step": 4820
    },
    {
      "epoch": 0.10131733512334284,
      "grad_norm": 1.0148108831813412,
      "learning_rate": 2e-05,
      "loss": 0.5755373001098633,
      "step": 4830
    },
    {
      "epoch": 0.10152710186272865,
      "grad_norm": 1.4086602722751511,
      "learning_rate": 2e-05,
      "loss": 0.6044353485107422,
      "step": 4840
    },
    {
      "epoch": 0.10173686860211445,
      "grad_norm": 1.1169745292020286,
      "learning_rate": 2e-05,
      "loss": 0.5700086116790771,
      "step": 4850
    },
    {
      "epoch": 0.10194663534150025,
      "grad_norm": 3.1811349693514455,
      "learning_rate": 2e-05,
      "loss": 0.6365950584411622,
      "step": 4860
    },
    {
      "epoch": 0.10215640208088606,
      "grad_norm": 3.07768864058665,
      "learning_rate": 2e-05,
      "loss": 0.6404936790466309,
      "step": 4870
    },
    {
      "epoch": 0.10236616882027186,
      "grad_norm": 1.3371706439487367,
      "learning_rate": 2e-05,
      "loss": 0.5982741355895996,
      "step": 4880
    },
    {
      "epoch": 0.10257593555965766,
      "grad_norm": 3.9578498104127275,
      "learning_rate": 2e-05,
      "loss": 0.6071772575378418,
      "step": 4890
    },
    {
      "epoch": 0.10278570229904346,
      "grad_norm": 1.456772676589832,
      "learning_rate": 2e-05,
      "loss": 0.5611026287078857,
      "step": 4900
    },
    {
      "epoch": 0.10299546903842927,
      "grad_norm": 2.08248864632305,
      "learning_rate": 2e-05,
      "loss": 0.5639925956726074,
      "step": 4910
    },
    {
      "epoch": 0.10320523577781507,
      "grad_norm": 1.2070126625656041,
      "learning_rate": 2e-05,
      "loss": 0.6080421447753906,
      "step": 4920
    },
    {
      "epoch": 0.10341500251720087,
      "grad_norm": 1.68288239255359,
      "learning_rate": 2e-05,
      "loss": 0.5914509773254395,
      "step": 4930
    },
    {
      "epoch": 0.10362476925658667,
      "grad_norm": 1.0321862993849868,
      "learning_rate": 2e-05,
      "loss": 0.6066112518310547,
      "step": 4940
    },
    {
      "epoch": 0.10383453599597248,
      "grad_norm": 1.0358142285880108,
      "learning_rate": 2e-05,
      "loss": 0.5691138744354248,
      "step": 4950
    },
    {
      "epoch": 0.10404430273535828,
      "grad_norm": 2.04547804175922,
      "learning_rate": 2e-05,
      "loss": 0.5506986618041992,
      "step": 4960
    },
    {
      "epoch": 0.10425406947474408,
      "grad_norm": 98.9202567272448,
      "learning_rate": 2e-05,
      "loss": 0.5544290542602539,
      "step": 4970
    },
    {
      "epoch": 0.10446383621412988,
      "grad_norm": 1.9359890131951925,
      "learning_rate": 2e-05,
      "loss": 0.613715124130249,
      "step": 4980
    },
    {
      "epoch": 0.10467360295351569,
      "grad_norm": 1.1406714637642643,
      "learning_rate": 2e-05,
      "loss": 0.5773385047912598,
      "step": 4990
    },
    {
      "epoch": 0.10488336969290149,
      "grad_norm": 4.981458134760419,
      "learning_rate": 2e-05,
      "loss": 0.5920042991638184,
      "step": 5000
    },
    {
      "epoch": 0.10509313643228729,
      "grad_norm": 1.169798427779728,
      "learning_rate": 2e-05,
      "loss": 0.6323759078979492,
      "step": 5010
    },
    {
      "epoch": 0.1053029031716731,
      "grad_norm": 1.579152953063113,
      "learning_rate": 2e-05,
      "loss": 0.6000487327575683,
      "step": 5020
    },
    {
      "epoch": 0.1055126699110589,
      "grad_norm": 1.6369923106332778,
      "learning_rate": 2e-05,
      "loss": 0.6221375465393066,
      "step": 5030
    },
    {
      "epoch": 0.1057224366504447,
      "grad_norm": 1.4142838696907318,
      "learning_rate": 2e-05,
      "loss": 0.6185449123382568,
      "step": 5040
    },
    {
      "epoch": 0.1059322033898305,
      "grad_norm": 1.3703125022723703,
      "learning_rate": 2e-05,
      "loss": 0.5900285720825196,
      "step": 5050
    },
    {
      "epoch": 0.1061419701292163,
      "grad_norm": 1.304764476675948,
      "learning_rate": 2e-05,
      "loss": 0.5960735321044922,
      "step": 5060
    },
    {
      "epoch": 0.10635173686860211,
      "grad_norm": 1.3807964375545567,
      "learning_rate": 2e-05,
      "loss": 0.6175256729125976,
      "step": 5070
    },
    {
      "epoch": 0.10656150360798791,
      "grad_norm": 0.9840300146029017,
      "learning_rate": 2e-05,
      "loss": 0.5989031791687012,
      "step": 5080
    },
    {
      "epoch": 0.10677127034737371,
      "grad_norm": 1.4408251580913778,
      "learning_rate": 2e-05,
      "loss": 0.5669477462768555,
      "step": 5090
    },
    {
      "epoch": 0.10698103708675952,
      "grad_norm": 1.325874687900724,
      "learning_rate": 2e-05,
      "loss": 0.5900368690490723,
      "step": 5100
    },
    {
      "epoch": 0.10719080382614532,
      "grad_norm": 2.693162506836015,
      "learning_rate": 2e-05,
      "loss": 0.5921677589416504,
      "step": 5110
    },
    {
      "epoch": 0.10740057056553112,
      "grad_norm": 1.1703238617712772,
      "learning_rate": 2e-05,
      "loss": 0.5660878658294678,
      "step": 5120
    },
    {
      "epoch": 0.10761033730491694,
      "grad_norm": 4.401054105717419,
      "learning_rate": 2e-05,
      "loss": 0.5696834087371826,
      "step": 5130
    },
    {
      "epoch": 0.10782010404430274,
      "grad_norm": 2.4847522662668204,
      "learning_rate": 2e-05,
      "loss": 0.5897496700286865,
      "step": 5140
    },
    {
      "epoch": 0.10802987078368854,
      "grad_norm": 1.1775643898345467,
      "learning_rate": 2e-05,
      "loss": 0.5921326637268066,
      "step": 5150
    },
    {
      "epoch": 0.10823963752307435,
      "grad_norm": 0.9088004576635755,
      "learning_rate": 2e-05,
      "loss": 0.6369561195373535,
      "step": 5160
    },
    {
      "epoch": 0.10844940426246015,
      "grad_norm": 1.0115107317913363,
      "learning_rate": 2e-05,
      "loss": 0.6052901744842529,
      "step": 5170
    },
    {
      "epoch": 0.10865917100184595,
      "grad_norm": 1.497599015977304,
      "learning_rate": 2e-05,
      "loss": 0.5822335243225097,
      "step": 5180
    },
    {
      "epoch": 0.10886893774123176,
      "grad_norm": 2.139327351776706,
      "learning_rate": 2e-05,
      "loss": 0.6052592277526856,
      "step": 5190
    },
    {
      "epoch": 0.10907870448061756,
      "grad_norm": 0.9882567592278001,
      "learning_rate": 2e-05,
      "loss": 0.5900141716003418,
      "step": 5200
    },
    {
      "epoch": 0.10928847122000336,
      "grad_norm": 1.1343671316050472,
      "learning_rate": 2e-05,
      "loss": 0.5901280879974365,
      "step": 5210
    },
    {
      "epoch": 0.10949823795938916,
      "grad_norm": 1.0767577115372522,
      "learning_rate": 2e-05,
      "loss": 0.6030196189880371,
      "step": 5220
    },
    {
      "epoch": 0.10970800469877497,
      "grad_norm": 0.8948989818967813,
      "learning_rate": 2e-05,
      "loss": 0.5701768875122071,
      "step": 5230
    },
    {
      "epoch": 0.10991777143816077,
      "grad_norm": 2.7335532106661815,
      "learning_rate": 2e-05,
      "loss": 0.5429616451263428,
      "step": 5240
    },
    {
      "epoch": 0.11012753817754657,
      "grad_norm": 1.2309401479613467,
      "learning_rate": 2e-05,
      "loss": 0.5788998603820801,
      "step": 5250
    },
    {
      "epoch": 0.11033730491693237,
      "grad_norm": 0.8089735320831154,
      "learning_rate": 2e-05,
      "loss": 0.5558887004852295,
      "step": 5260
    },
    {
      "epoch": 0.11054707165631818,
      "grad_norm": 0.7663344983915933,
      "learning_rate": 2e-05,
      "loss": 0.5819116592407226,
      "step": 5270
    },
    {
      "epoch": 0.11075683839570398,
      "grad_norm": 0.9903153722948607,
      "learning_rate": 2e-05,
      "loss": 0.585750961303711,
      "step": 5280
    },
    {
      "epoch": 0.11096660513508978,
      "grad_norm": 4.8208287010493995,
      "learning_rate": 2e-05,
      "loss": 0.5778745651245117,
      "step": 5290
    },
    {
      "epoch": 0.11117637187447559,
      "grad_norm": 1.1451638920555645,
      "learning_rate": 2e-05,
      "loss": 0.5597260951995849,
      "step": 5300
    },
    {
      "epoch": 0.11138613861386139,
      "grad_norm": 1.055701613028047,
      "learning_rate": 2e-05,
      "loss": 0.5642464637756348,
      "step": 5310
    },
    {
      "epoch": 0.11159590535324719,
      "grad_norm": 1.1312994156250342,
      "learning_rate": 2e-05,
      "loss": 0.6001087188720703,
      "step": 5320
    },
    {
      "epoch": 0.111805672092633,
      "grad_norm": 0.8752863720788787,
      "learning_rate": 2e-05,
      "loss": 0.6106997966766358,
      "step": 5330
    },
    {
      "epoch": 0.1120154388320188,
      "grad_norm": 0.9348132788295025,
      "learning_rate": 2e-05,
      "loss": 0.6121906280517578,
      "step": 5340
    },
    {
      "epoch": 0.1122252055714046,
      "grad_norm": 2.383424365131304,
      "learning_rate": 1.999194520388485e-05,
      "loss": 0.6219453811645508,
      "step": 5350
    },
    {
      "epoch": 0.1124349723107904,
      "grad_norm": 0.7610279930468801,
      "learning_rate": 1.9959247303157763e-05,
      "loss": 0.5907040596008301,
      "step": 5360
    },
    {
      "epoch": 0.1126447390501762,
      "grad_norm": 1.0145482661998457,
      "learning_rate": 1.990149424459817e-05,
      "loss": 0.5997787475585937,
      "step": 5370
    },
    {
      "epoch": 0.112854505789562,
      "grad_norm": 0.924539741605453,
      "learning_rate": 1.9818847538648688e-05,
      "loss": 0.5890125274658203,
      "step": 5380
    },
    {
      "epoch": 0.11306427252894781,
      "grad_norm": 1.1045725380977878,
      "learning_rate": 1.971153831256721e-05,
      "loss": 0.5800723075866699,
      "step": 5390
    },
    {
      "epoch": 0.11327403926833361,
      "grad_norm": 1.1670127903880712,
      "learning_rate": 1.9579866664063435e-05,
      "loss": 0.6123351097106934,
      "step": 5400
    },
    {
      "epoch": 0.11348380600771941,
      "grad_norm": 1.6290991602977665,
      "learning_rate": 1.9424200822054782e-05,
      "loss": 0.601279354095459,
      "step": 5410
    },
    {
      "epoch": 0.11369357274710522,
      "grad_norm": 0.8865092070142566,
      "learning_rate": 1.9244976116888626e-05,
      "loss": 0.6134068965911865,
      "step": 5420
    },
    {
      "epoch": 0.11390333948649102,
      "grad_norm": 1.0106891293364595,
      "learning_rate": 1.904269376291071e-05,
      "loss": 0.5925784587860108,
      "step": 5430
    },
    {
      "epoch": 0.11411310622587682,
      "grad_norm": 1.2550004844543956,
      "learning_rate": 1.88179194567844e-05,
      "loss": 0.6154834747314453,
      "step": 5440
    },
    {
      "epoch": 0.11432287296526263,
      "grad_norm": 0.8663903834409756,
      "learning_rate": 1.8571281795480632e-05,
      "loss": 0.5740708827972412,
      "step": 5450
    },
    {
      "epoch": 0.11453263970464843,
      "grad_norm": 0.8757909567668538,
      "learning_rate": 1.8303470518362808e-05,
      "loss": 0.5779492378234863,
      "step": 5460
    },
    {
      "epoch": 0.11474240644403423,
      "grad_norm": 0.8251232745658752,
      "learning_rate": 1.801523457828271e-05,
      "loss": 0.62754225730896,
      "step": 5470
    },
    {
      "epoch": 0.11495217318342003,
      "grad_norm": 1.0996938286110867,
      "learning_rate": 1.7707380047081796e-05,
      "loss": 0.6147668361663818,
      "step": 5480
    },
    {
      "epoch": 0.11516193992280584,
      "grad_norm": 1.0781120330099085,
      "learning_rate": 1.7380767861355252e-05,
      "loss": 0.5535259246826172,
      "step": 5490
    },
    {
      "epoch": 0.11537170666219164,
      "grad_norm": 0.906305376984634,
      "learning_rate": 1.703631141478299e-05,
      "loss": 0.6055543422698975,
      "step": 5500
    },
    {
      "epoch": 0.11558147340157744,
      "grad_norm": 1.2492203615805852,
      "learning_rate": 1.6674974003760705e-05,
      "loss": 0.5707786560058594,
      "step": 5510
    },
    {
      "epoch": 0.11579124014096324,
      "grad_norm": 0.8680069069328755,
      "learning_rate": 1.6297766133474556e-05,
      "loss": 0.5344223022460938,
      "step": 5520
    },
    {
      "epoch": 0.11600100688034905,
      "grad_norm": 1.3797078496824349,
      "learning_rate": 1.5905742691953186e-05,
      "loss": 0.5875358581542969,
      "step": 5530
    },
    {
      "epoch": 0.11621077361973485,
      "grad_norm": 0.8002536048829909,
      "learning_rate": 1.55e-05,
      "loss": 0.5772517681121826,
      "step": 5540
    },
    {
      "epoch": 0.11642054035912065,
      "grad_norm": 1.6687264519213083,
      "learning_rate": 1.5081672745255862e-05,
      "loss": 0.5724442958831787,
      "step": 5550
    },
    {
      "epoch": 0.11663030709850646,
      "grad_norm": 0.8572764844626785,
      "learning_rate": 1.4651930808966206e-05,
      "loss": 0.5642001152038574,
      "step": 5560
    },
    {
      "epoch": 0.11684007383789226,
      "grad_norm": 0.8711913905310741,
      "learning_rate": 1.4211975994326848e-05,
      "loss": 0.5909633159637451,
      "step": 5570
    },
    {
      "epoch": 0.11704984057727806,
      "grad_norm": 1.0259849198987272,
      "learning_rate": 1.376303866555779e-05,
      "loss": 0.5776866912841797,
      "step": 5580
    },
    {
      "epoch": 0.11725960731666386,
      "grad_norm": 0.9578797703119694,
      "learning_rate": 1.3306374307104164e-05,
      "loss": 0.560568380355835,
      "step": 5590
    },
    {
      "epoch": 0.11746937405604967,
      "grad_norm": 0.8829821140590774,
      "learning_rate": 1.2843260012586718e-05,
      "loss": 0.5382630348205566,
      "step": 5600
    },
    {
      "epoch": 0.11767914079543547,
      "grad_norm": 1.2266656760931693,
      "learning_rate": 1.2374990913320726e-05,
      "loss": 0.5363844871520996,
      "step": 5610
    },
    {
      "epoch": 0.11788890753482129,
      "grad_norm": 0.7853552747954475,
      "learning_rate": 1.1902876556391247e-05,
      "loss": 0.553427267074585,
      "step": 5620
    },
    {
      "epoch": 0.11809867427420709,
      "grad_norm": 0.7175625831319892,
      "learning_rate": 1.1428237242413685e-05,
      "loss": 0.5885550498962402,
      "step": 5630
    },
    {
      "epoch": 0.11830844101359289,
      "grad_norm": 2.6592699455321704,
      "learning_rate": 1.095240033322132e-05,
      "loss": 0.5815681457519531,
      "step": 5640
    },
    {
      "epoch": 0.1185182077529787,
      "grad_norm": 0.9017977305325245,
      "learning_rate": 1.047669653980572e-05,
      "loss": 0.5832801818847656,
      "step": 5650
    },
    {
      "epoch": 0.1187279744923645,
      "grad_norm": 1.4002225586484862,
      "learning_rate": 1.0002456200890902e-05,
      "loss": 0.55551118850708,
      "step": 5660
    },
    {
      "epoch": 0.1189377412317503,
      "grad_norm": 2.8152432045880196,
      "learning_rate": 9.531005562548668e-06,
      "loss": 0.5489104270935059,
      "step": 5670
    },
    {
      "epoch": 0.1191475079711361,
      "grad_norm": 0.7537869065799321,
      "learning_rate": 9.06366306925925e-06,
      "loss": 0.563069486618042,
      "step": 5680
    },
    {
      "epoch": 0.1193572747105219,
      "grad_norm": 0.8333914557096338,
      "learning_rate": 8.601735676789686e-06,
      "loss": 0.5415831565856933,
      "step": 5690
    },
    {
      "epoch": 0.11956704144990771,
      "grad_norm": 1.0676281247770207,
      "learning_rate": 8.14651519720118e-06,
      "loss": 0.512755012512207,
      "step": 5700
    },
    {
      "epoch": 0.11977680818929351,
      "grad_norm": 0.7582486717894659,
      "learning_rate": 7.699274686206866e-06,
      "loss": 0.5345975399017334,
      "step": 5710
    },
    {
      "epoch": 0.11998657492867931,
      "grad_norm": 1.199189385155412,
      "learning_rate": 7.261264882983024e-06,
      "loss": 0.5936285018920898,
      "step": 5720
    },
    {
      "epoch": 0.12019634166806512,
      "grad_norm": 0.7805313274838336,
      "learning_rate": 6.8337107123900386e-06,
      "loss": 0.5691483974456787,
      "step": 5730
    },
    {
      "epoch": 0.12040610840745092,
      "grad_norm": 0.8011712185240046,
      "learning_rate": 6.417807859384881e-06,
      "loss": 0.5548898220062256,
      "step": 5740
    },
    {
      "epoch": 0.12061587514683672,
      "grad_norm": 0.8678294445663222,
      "learning_rate": 6.014719425205009e-06,
      "loss": 0.5622466087341309,
      "step": 5750
    },
    {
      "epoch": 0.12082564188622252,
      "grad_norm": 0.7725690149400918,
      "learning_rate": 5.6255726746749275e-06,
      "loss": 0.5937882900238037,
      "step": 5760
    },
    {
      "epoch": 0.12103540862560833,
      "grad_norm": 0.8596328292180833,
      "learning_rate": 5.251455883731728e-06,
      "loss": 0.5242228507995605,
      "step": 5770
    },
    {
      "epoch": 0.12124517536499413,
      "grad_norm": 0.9536416138058138,
      "learning_rate": 4.893415295985812e-06,
      "loss": 0.5733167171478272,
      "step": 5780
    },
    {
      "epoch": 0.12145494210437993,
      "grad_norm": 0.9835130802666492,
      "learning_rate": 4.552452196827885e-06,
      "loss": 0.584380054473877,
      "step": 5790
    },
    {
      "epoch": 0.12166470884376573,
      "grad_norm": 0.9348104194657836,
      "learning_rate": 4.229520113264785e-06,
      "loss": 0.5617354869842529,
      "step": 5800
    },
    {
      "epoch": 0.12187447558315154,
      "grad_norm": 0.8703438766658813,
      "learning_rate": 3.925522147314915e-06,
      "loss": 0.5718338966369629,
      "step": 5810
    },
    {
      "epoch": 0.12208424232253734,
      "grad_norm": 0.9716207991267166,
      "learning_rate": 3.641308450420675e-06,
      "loss": 0.5608897686004639,
      "step": 5820
    },
    {
      "epoch": 0.12229400906192314,
      "grad_norm": 0.8629618250765935,
      "learning_rate": 3.37767384594087e-06,
      "loss": 0.5632754802703858,
      "step": 5830
    },
    {
      "epoch": 0.12250377580130895,
      "grad_norm": 1.2935840329203043,
      "learning_rate": 3.1353556063719357e-06,
      "loss": 0.527442455291748,
      "step": 5840
    },
    {
      "epoch": 0.12271354254069475,
      "grad_norm": 1.1540952728987506,
      "learning_rate": 2.9150313915141474e-06,
      "loss": 0.569704818725586,
      "step": 5850
    },
    {
      "epoch": 0.12292330928008055,
      "grad_norm": 1.0988001335036446,
      "learning_rate": 2.717317353348891e-06,
      "loss": 0.5773658752441406,
      "step": 5860
    },
    {
      "epoch": 0.12313307601946635,
      "grad_norm": 0.8451781751659095,
      "learning_rate": 2.5427664129268253e-06,
      "loss": 0.51842041015625,
      "step": 5870
    },
    {
      "epoch": 0.12334284275885216,
      "grad_norm": 0.8900096983097106,
      "learning_rate": 2.391866714085717e-06,
      "loss": 0.5635234355926514,
      "step": 5880
    },
    {
      "epoch": 0.12355260949823796,
      "grad_norm": 0.818907072961598,
      "learning_rate": 2.2650402583222753e-06,
      "loss": 0.5465801239013672,
      "step": 5890
    },
    {
      "epoch": 0.12376237623762376,
      "grad_norm": 0.8458346610901077,
      "learning_rate": 2.1626417246356398e-06,
      "loss": 0.551777458190918,
      "step": 5900
    },
    {
      "epoch": 0.12397214297700956,
      "grad_norm": 1.1752967528704192,
      "learning_rate": 2.084957477642894e-06,
      "loss": 0.5042602062225342,
      "step": 5910
    },
    {
      "epoch": 0.12418190971639537,
      "grad_norm": 0.8224999422202722,
      "learning_rate": 2.0322047667405147e-06,
      "loss": 0.5769207000732421,
      "step": 5920
    },
    {
      "epoch": 0.12439167645578117,
      "grad_norm": 0.9458165701271866,
      "learning_rate": 2.0045311185513342e-06,
      "loss": 0.5461589336395264,
      "step": 5930
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2810342056919040.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}