{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 3125,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 11.83144098371552,
      "learning_rate": 1.437699680511182e-07,
      "loss": 0.6488185882568359,
      "step": 10
    },
    {
      "epoch": 0.0064,
      "grad_norm": 9.59315923624666,
      "learning_rate": 3.0351437699680514e-07,
      "loss": 0.6357447624206543,
      "step": 20
    },
    {
      "epoch": 0.0096,
      "grad_norm": 8.611135612577282,
      "learning_rate": 4.6325878594249205e-07,
      "loss": 0.5702734470367432,
      "step": 30
    },
    {
      "epoch": 0.0128,
      "grad_norm": 3.380574205681449,
      "learning_rate": 6.230031948881789e-07,
      "loss": 0.4336705207824707,
      "step": 40
    },
    {
      "epoch": 0.016,
      "grad_norm": 1.3215372913591454,
      "learning_rate": 7.82747603833866e-07,
      "loss": 0.3196877956390381,
      "step": 50
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.49205152753117876,
      "learning_rate": 9.424920127795528e-07,
      "loss": 0.26679015159606934,
      "step": 60
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.45222798092529587,
      "learning_rate": 1.1022364217252397e-06,
      "loss": 0.267962121963501,
      "step": 70
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.38880908025022265,
      "learning_rate": 1.2619808306709266e-06,
      "loss": 0.2507643222808838,
      "step": 80
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.3466305556727367,
      "learning_rate": 1.4217252396166134e-06,
      "loss": 0.24545693397521973,
      "step": 90
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.3545745065230796,
      "learning_rate": 1.5814696485623005e-06,
      "loss": 0.24865365028381348,
      "step": 100
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.36611277528700825,
      "learning_rate": 1.7412140575079875e-06,
      "loss": 0.24806594848632812,
      "step": 110
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.4168127418004872,
      "learning_rate": 1.9009584664536742e-06,
      "loss": 0.23844523429870607,
      "step": 120
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.3478506619909293,
      "learning_rate": 2.060702875399361e-06,
      "loss": 0.24028465747833253,
      "step": 130
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.37897531613175517,
      "learning_rate": 2.220447284345048e-06,
      "loss": 0.23141345977783204,
      "step": 140
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.44434967369549566,
      "learning_rate": 2.380191693290735e-06,
      "loss": 0.227997350692749,
      "step": 150
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.3846551259722682,
      "learning_rate": 2.539936102236422e-06,
      "loss": 0.22319293022155762,
      "step": 160
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.42029854714577086,
      "learning_rate": 2.699680511182109e-06,
      "loss": 0.23121466636657714,
      "step": 170
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.5531278369713543,
      "learning_rate": 2.8594249201277955e-06,
      "loss": 0.22436304092407228,
      "step": 180
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.41335994292883266,
      "learning_rate": 3.0191693290734825e-06,
      "loss": 0.2279865026473999,
      "step": 190
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.41777451150300726,
      "learning_rate": 3.17891373801917e-06,
      "loss": 0.22936224937438965,
      "step": 200
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.4429107133096889,
      "learning_rate": 3.3386581469648564e-06,
      "loss": 0.22612979412078857,
      "step": 210
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.37955292136954716,
      "learning_rate": 3.4984025559105434e-06,
      "loss": 0.21961536407470703,
      "step": 220
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.3916244375225685,
      "learning_rate": 3.6581469648562303e-06,
      "loss": 0.22160472869873046,
      "step": 230
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.4190841730835042,
      "learning_rate": 3.817891373801918e-06,
      "loss": 0.21761112213134765,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.39560583196094196,
      "learning_rate": 3.977635782747604e-06,
      "loss": 0.21524910926818847,
      "step": 250
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.4078924637485393,
      "learning_rate": 4.137380191693291e-06,
      "loss": 0.21542978286743164,
      "step": 260
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.4106351573034691,
      "learning_rate": 4.297124600638978e-06,
      "loss": 0.2175312042236328,
      "step": 270
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.350392701468648,
      "learning_rate": 4.456869009584665e-06,
      "loss": 0.21332988739013672,
      "step": 280
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.33990023576201683,
      "learning_rate": 4.616613418530352e-06,
      "loss": 0.21945428848266602,
      "step": 290
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.3930734349391618,
      "learning_rate": 4.776357827476039e-06,
      "loss": 0.21993803977966309,
      "step": 300
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.3988914165504776,
      "learning_rate": 4.936102236421725e-06,
      "loss": 0.20795869827270508,
      "step": 310
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.35731910595219907,
      "learning_rate": 4.999943833158769e-06,
      "loss": 0.20925359725952147,
      "step": 320
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.3059149411920343,
      "learning_rate": 4.999600600490783e-06,
      "loss": 0.2151254415512085,
      "step": 330
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.38132024364024864,
      "learning_rate": 4.9989453817439345e-06,
      "loss": 0.20901703834533691,
      "step": 340
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.3345902579307589,
      "learning_rate": 4.997978258698942e-06,
      "loss": 0.20827054977416992,
      "step": 350
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.3654908632001385,
      "learning_rate": 4.996699352066659e-06,
      "loss": 0.21013379096984863,
      "step": 360
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.39442024987420615,
      "learning_rate": 4.995108821473014e-06,
      "loss": 0.20931053161621094,
      "step": 370
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.4182616523944993,
      "learning_rate": 4.993206865439084e-06,
      "loss": 0.20571417808532716,
      "step": 380
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.341180560016227,
      "learning_rate": 4.990993721356317e-06,
      "loss": 0.2043307065963745,
      "step": 390
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.3752267388801998,
      "learning_rate": 4.988469665456901e-06,
      "loss": 0.20494937896728516,
      "step": 400
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.3245176299142898,
      "learning_rate": 4.985635012779288e-06,
      "loss": 0.21275849342346193,
      "step": 410
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.30464125237598766,
      "learning_rate": 4.98249011712887e-06,
      "loss": 0.20537421703338624,
      "step": 420
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.32422803949501344,
      "learning_rate": 4.979035371033824e-06,
      "loss": 0.20476508140563965,
      "step": 430
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.3539302093186753,
      "learning_rate": 4.975271205696115e-06,
      "loss": 0.2044001340866089,
      "step": 440
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.3405614059989652,
      "learning_rate": 4.971198090937671e-06,
      "loss": 0.198509681224823,
      "step": 450
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.38543797495647464,
      "learning_rate": 4.966816535141756e-06,
      "loss": 0.19933369159698486,
      "step": 460
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.3243047107311659,
      "learning_rate": 4.9621270851895035e-06,
      "loss": 0.20511775016784667,
      "step": 470
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.3856847456921783,
      "learning_rate": 4.957130326391662e-06,
      "loss": 0.2059837818145752,
      "step": 480
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.39057481445802306,
      "learning_rate": 4.951826882415544e-06,
      "loss": 0.20367493629455566,
      "step": 490
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.3868568198929062,
      "learning_rate": 4.946217415207177e-06,
      "loss": 0.19742422103881835,
      "step": 500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.3509586614208196,
      "learning_rate": 4.940302624908689e-06,
      "loss": 0.20141959190368652,
      "step": 510
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.39255560366454684,
      "learning_rate": 4.934083249770912e-06,
      "loss": 0.19597206115722657,
      "step": 520
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.3753941656538999,
      "learning_rate": 4.927560066061251e-06,
      "loss": 0.19957098960876465,
      "step": 530
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.33505547223235554,
      "learning_rate": 4.920733887966783e-06,
      "loss": 0.2104717254638672,
      "step": 540
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.3397525078929753,
      "learning_rate": 4.913605567492636e-06,
      "loss": 0.19930799007415773,
      "step": 550
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.41307748920624604,
      "learning_rate": 4.906175994355656e-06,
      "loss": 0.20131049156188965,
      "step": 560
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.46727990153877497,
      "learning_rate": 4.898446095873345e-06,
      "loss": 0.20283927917480468,
      "step": 570
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.3544063329370901,
      "learning_rate": 4.890416836848128e-06,
      "loss": 0.200484037399292,
      "step": 580
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.3342685909168471,
      "learning_rate": 4.882089219446925e-06,
      "loss": 0.19547154903411865,
      "step": 590
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.37096928349741987,
      "learning_rate": 4.873464283076074e-06,
      "loss": 0.20015311241149902,
      "step": 600
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.3027007752332859,
      "learning_rate": 4.864543104251587e-06,
      "loss": 0.19957488775253296,
      "step": 610
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.33671892410389453,
      "learning_rate": 4.855326796464798e-06,
      "loss": 0.20140202045440675,
      "step": 620
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.3380893311921056,
      "learning_rate": 4.8458165100433725e-06,
      "loss": 0.1972325325012207,
      "step": 630
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.33781332412651466,
      "learning_rate": 4.836013432007738e-06,
      "loss": 0.19717284440994262,
      "step": 640
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.34031999571424276,
      "learning_rate": 4.825918785922921e-06,
      "loss": 0.20164141654968262,
      "step": 650
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.3761741276870676,
      "learning_rate": 4.8155338317458315e-06,
      "loss": 0.19852631092071532,
      "step": 660
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.33766360947275226,
      "learning_rate": 4.804859865668002e-06,
      "loss": 0.1955265998840332,
      "step": 670
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.3263649532195094,
      "learning_rate": 4.793898219953804e-06,
      "loss": 0.1965425729751587,
      "step": 680
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.32263376893110113,
      "learning_rate": 4.782650262774164e-06,
      "loss": 0.19828790426254272,
      "step": 690
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.3517548043501075,
      "learning_rate": 4.7711173980357886e-06,
      "loss": 0.19447412490844726,
      "step": 700
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.338792238112512,
      "learning_rate": 4.759301065205947e-06,
      "loss": 0.1949925899505615,
      "step": 710
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.37194857535123965,
      "learning_rate": 4.7472027391328e-06,
      "loss": 0.19610352516174318,
      "step": 720
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.3915181276276985,
      "learning_rate": 4.734823929861317e-06,
      "loss": 0.1986656069755554,
      "step": 730
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.38957297489256615,
      "learning_rate": 4.722166182444801e-06,
      "loss": 0.19531142711639404,
      "step": 740
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.35280725988589035,
      "learning_rate": 4.709231076752045e-06,
      "loss": 0.1945881724357605,
      "step": 750
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.328859584018874,
      "learning_rate": 4.696020227270142e-06,
      "loss": 0.1986994504928589,
      "step": 760
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.42079847846495083,
      "learning_rate": 4.6825352829029705e-06,
      "loss": 0.1951799511909485,
      "step": 770
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.3564546225835861,
      "learning_rate": 4.668777926765392e-06,
      "loss": 0.19262158870697021,
      "step": 780
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.30910181888519894,
      "learning_rate": 4.6547498759731725e-06,
      "loss": 0.1888942837715149,
      "step": 790
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.3800555896836864,
      "learning_rate": 4.6404528814286575e-06,
      "loss": 0.191619348526001,
      "step": 800
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.3214924880946795,
      "learning_rate": 4.6258887276022425e-06,
      "loss": 0.1980231523513794,
      "step": 810
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.3380732634719034,
      "learning_rate": 4.611059232309639e-06,
      "loss": 0.19314407110214232,
      "step": 820
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.389068899000227,
      "learning_rate": 4.595966246484986e-06,
      "loss": 0.19506924152374266,
      "step": 830
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.3863125802433344,
      "learning_rate": 4.580611653949829e-06,
      "loss": 0.19590425491333008,
      "step": 840
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.3386867491077868,
      "learning_rate": 4.564997371177992e-06,
      "loss": 0.19343934059143067,
      "step": 850
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.32498475663304227,
      "learning_rate": 4.54912534705637e-06,
      "loss": 0.19271985292434693,
      "step": 860
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.37076633469721365,
      "learning_rate": 4.532997562641683e-06,
      "loss": 0.19003691673278808,
      "step": 870
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.34537647808173055,
      "learning_rate": 4.516616030913214e-06,
      "loss": 0.19376157522201537,
      "step": 880
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.3396204448169827,
      "learning_rate": 4.499982796521556e-06,
      "loss": 0.1885458469390869,
      "step": 890
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.3278931050248095,
      "learning_rate": 4.48309993553341e-06,
      "loss": 0.1910359740257263,
      "step": 900
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.31081499529851564,
      "learning_rate": 4.465969555172468e-06,
      "loss": 0.1909637451171875,
      "step": 910
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.36721402125547814,
      "learning_rate": 4.448593793556391e-06,
      "loss": 0.19662458896636964,
      "step": 920
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.3535696512302074,
      "learning_rate": 4.430974819429954e-06,
      "loss": 0.18980509042739868,
      "step": 930
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.38926250103548876,
      "learning_rate": 4.413114831894344e-06,
      "loss": 0.18494791984558107,
      "step": 940
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.34453334571881833,
      "learning_rate": 4.3950160601326865e-06,
      "loss": 0.184212589263916,
      "step": 950
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.3762340897100377,
      "learning_rate": 4.376680763131811e-06,
      "loss": 0.18981791734695436,
      "step": 960
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.3530324206457789,
      "learning_rate": 4.358111229400296e-06,
      "loss": 0.19288237094879152,
      "step": 970
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.34771997443076413,
      "learning_rate": 4.33930977668283e-06,
      "loss": 0.19517345428466798,
      "step": 980
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.3259036212205655,
      "learning_rate": 4.320278751670922e-06,
      "loss": 0.18931124210357667,
      "step": 990
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3860199237155075,
      "learning_rate": 4.301020529710009e-06,
      "loss": 0.19042216539382933,
      "step": 1000
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.376632222430791,
      "learning_rate": 4.281537514502962e-06,
      "loss": 0.19394458532333375,
      "step": 1010
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.3295954306779813,
      "learning_rate": 4.261832137810093e-06,
      "loss": 0.1905941605567932,
      "step": 1020
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.36350379944390687,
      "learning_rate": 4.241906859145611e-06,
      "loss": 0.1911671996116638,
      "step": 1030
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.3515414089838784,
      "learning_rate": 4.221764165470661e-06,
      "loss": 0.1879589319229126,
      "step": 1040
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.36056971430032736,
      "learning_rate": 4.201406570882898e-06,
      "loss": 0.1876499891281128,
      "step": 1050
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.38550285088851427,
      "learning_rate": 4.180836616302704e-06,
      "loss": 0.18704679012298583,
      "step": 1060
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.3907983397501218,
      "learning_rate": 4.160056869156041e-06,
      "loss": 0.1898512363433838,
      "step": 1070
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.3185324338246927,
      "learning_rate": 4.139069923053995e-06,
      "loss": 0.18859152793884276,
      "step": 1080
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.33476796180481766,
      "learning_rate": 4.117878397469062e-06,
      "loss": 0.19204466342926024,
      "step": 1090
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.36080516187755207,
      "learning_rate": 4.096484937408195e-06,
      "loss": 0.18328707218170165,
      "step": 1100
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.3643905053944042,
      "learning_rate": 4.074892213082676e-06,
      "loss": 0.18360459804534912,
      "step": 1110
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.3451314428178685,
      "learning_rate": 4.0531029195748265e-06,
      "loss": 0.19115616083145143,
      "step": 1120
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.3503646826533139,
      "learning_rate": 4.03111977650163e-06,
      "loss": 0.1892825484275818,
      "step": 1130
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.38647304833372,
      "learning_rate": 4.008945527675281e-06,
      "loss": 0.18974583148956298,
      "step": 1140
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.3923573548206856,
      "learning_rate": 3.986582940760717e-06,
      "loss": 0.17893118858337403,
      "step": 1150
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.3381639516581911,
      "learning_rate": 3.9640348069301785e-06,
      "loss": 0.18569667339324952,
      "step": 1160
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.34840921452583046,
      "learning_rate": 3.941303940514826e-06,
      "loss": 0.18692519664764404,
      "step": 1170
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.3621025779320011,
      "learning_rate": 3.918393178653472e-06,
      "loss": 0.18706123828887938,
      "step": 1180
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.3409994515842346,
      "learning_rate": 3.895305380938468e-06,
      "loss": 0.1850137233734131,
      "step": 1190
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.32134108140002676,
      "learning_rate": 3.872043429058783e-06,
      "loss": 0.18305637836456298,
      "step": 1200
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.34976165687661737,
      "learning_rate": 3.84861022644033e-06,
      "loss": 0.18524043560028075,
      "step": 1210
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.35397661097444644,
      "learning_rate": 3.825008697883574e-06,
      "loss": 0.19263406991958618,
      "step": 1220
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.3047568820190958,
      "learning_rate": 3.8012417891984776e-06,
      "loss": 0.18630678653717042,
      "step": 1230
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.3509346490656729,
      "learning_rate": 3.777312466836819e-06,
      "loss": 0.18509007692337037,
      "step": 1240
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.35442414080318474,
      "learning_rate": 3.7532237175219378e-06,
      "loss": 0.183595871925354,
      "step": 1250
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.3763561451452722,
      "learning_rate": 3.728978547875948e-06,
      "loss": 0.18734192848205566,
      "step": 1260
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.3806442700695017,
      "learning_rate": 3.7045799840444712e-06,
      "loss": 0.18231661319732667,
      "step": 1270
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.3693294758862003,
      "learning_rate": 3.6800310713189258e-06,
      "loss": 0.18366684913635253,
      "step": 1280
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.4679876778528945,
      "learning_rate": 3.6553348737564328e-06,
      "loss": 0.1844363331794739,
      "step": 1290
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.3188887295686398,
      "learning_rate": 3.6304944737973794e-06,
      "loss": 0.189242947101593,
      "step": 1300
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.38141292954197614,
      "learning_rate": 3.6055129718806836e-06,
      "loss": 0.18387432098388673,
      "step": 1310
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.42807961078436774,
      "learning_rate": 3.5803934860568134e-06,
      "loss": 0.1800466775894165,
      "step": 1320
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.407419776715016,
      "learning_rate": 3.5551391515986163e-06,
      "loss": 0.1842915654182434,
      "step": 1330
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.3457735305030836,
      "learning_rate": 3.529753120609982e-06,
      "loss": 0.18095966577529907,
      "step": 1340
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.40768516482856293,
      "learning_rate": 3.5042385616324243e-06,
      "loss": 0.1833919644355774,
      "step": 1350
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.3218920190656802,
      "learning_rate": 3.4785986592495934e-06,
      "loss": 0.18204569816589355,
      "step": 1360
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.3296264433157612,
      "learning_rate": 3.452836613689803e-06,
      "loss": 0.1804790735244751,
      "step": 1370
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.3252281867414127,
      "learning_rate": 3.426955640426584e-06,
      "loss": 0.18308441638946532,
      "step": 1380
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.3453938909284546,
      "learning_rate": 3.4009589697773605e-06,
      "loss": 0.1829667568206787,
      "step": 1390
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.3705007173736256,
      "learning_rate": 3.3748498465002475e-06,
      "loss": 0.17947460412979127,
      "step": 1400
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.3546369769618238,
      "learning_rate": 3.3486315293890693e-06,
      "loss": 0.18765747547149658,
      "step": 1410
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.3781284245896334,
      "learning_rate": 3.3223072908666053e-06,
      "loss": 0.17857763767242432,
      "step": 1420
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.3588497462943525,
      "learning_rate": 3.295880416576153e-06,
      "loss": 0.18716816902160643,
      "step": 1430
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.4018300783669189,
      "learning_rate": 3.269354204971427e-06,
      "loss": 0.1817032814025879,
      "step": 1440
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.43569079636878416,
      "learning_rate": 3.242731966904865e-06,
      "loss": 0.17936263084411622,
      "step": 1450
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.4112327697697191,
      "learning_rate": 3.2160170252143913e-06,
      "loss": 0.17932071685791015,
      "step": 1460
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.32882512900562777,
      "learning_rate": 3.1892127143086716e-06,
      "loss": 0.1854783296585083,
      "step": 1470
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.34718382356494926,
      "learning_rate": 3.1623223797509347e-06,
      "loss": 0.17551472187042236,
      "step": 1480
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.40014884763121933,
      "learning_rate": 3.135349377841396e-06,
      "loss": 0.17985255718231202,
      "step": 1490
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.37002552374237196,
      "learning_rate": 3.1082970751983497e-06,
      "loss": 0.1858470320701599,
      "step": 1500
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.3189241194339908,
      "learning_rate": 3.0811688483379546e-06,
      "loss": 0.1844353437423706,
      "step": 1510
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.3292240464587196,
      "learning_rate": 3.0539680832528074e-06,
      "loss": 0.183160138130188,
      "step": 1520
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.3489032040504687,
      "learning_rate": 3.026698174989316e-06,
      "loss": 0.18127987384796143,
      "step": 1530
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.33504467669019317,
      "learning_rate": 2.999362527223952e-06,
      "loss": 0.1818302035331726,
      "step": 1540
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.3278367328955766,
      "learning_rate": 2.9719645518384194e-06,
      "loss": 0.18281204700469972,
      "step": 1550
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.3932354078093087,
      "learning_rate": 2.944507668493807e-06,
      "loss": 0.18620465993881224,
      "step": 1560
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.3669969573194585,
      "learning_rate": 2.9169953042037623e-06,
      "loss": 0.18195055723190307,
      "step": 1570
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.3798793826233043,
      "learning_rate": 2.889430892906754e-06,
      "loss": 0.17828736305236817,
      "step": 1580
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.3169054303302339,
      "learning_rate": 2.861817875037462e-06,
      "loss": 0.1758497476577759,
      "step": 1590
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.3427628658073453,
      "learning_rate": 2.8341596970973683e-06,
      "loss": 0.18466551303863527,
      "step": 1600
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.3313402627200071,
      "learning_rate": 2.80645981122458e-06,
      "loss": 0.17985165119171143,
      "step": 1610
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.3299040944947936,
      "learning_rate": 2.7787216747629508e-06,
      "loss": 0.186691951751709,
      "step": 1620
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.34753331909668905,
      "learning_rate": 2.7509487498305615e-06,
      "loss": 0.1754052758216858,
      "step": 1630
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.3813115280828973,
      "learning_rate": 2.7231445028875924e-06,
      "loss": 0.18044731616973878,
      "step": 1640
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.34093546280867537,
      "learning_rate": 2.6953124043036604e-06,
      "loss": 0.18008266687393187,
      "step": 1650
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.37253670534687966,
      "learning_rate": 2.667455927924667e-06,
      "loss": 0.17874090671539306,
      "step": 1660
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.3605836169669652,
      "learning_rate": 2.6395785506392164e-06,
      "loss": 0.17416150569915773,
      "step": 1670
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.3528045717062405,
      "learning_rate": 2.6116837519446407e-06,
      "loss": 0.17853121757507323,
      "step": 1680
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.35663168003844714,
      "learning_rate": 2.5837750135127192e-06,
      "loss": 0.1759209394454956,
      "step": 1690
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.35462371645127194,
      "learning_rate": 2.555855818755108e-06,
      "loss": 0.1807964563369751,
      "step": 1700
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.40893120307942266,
      "learning_rate": 2.5279296523885636e-06,
      "loss": 0.1787680983543396,
      "step": 1710
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.35660931340798585,
      "learning_rate": 2.5e-06,
      "loss": 0.17910590171813964,
      "step": 1720
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.36395042718977416,
      "learning_rate": 2.472070347611437e-06,
      "loss": 0.17353482246398927,
      "step": 1730
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.38459148561610207,
      "learning_rate": 2.444144181244893e-06,
      "loss": 0.17830569744110109,
      "step": 1740
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3483704729147774,
      "learning_rate": 2.416224986487282e-06,
      "loss": 0.17659709453582764,
      "step": 1750
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.37643842809936706,
      "learning_rate": 2.3883162480553605e-06,
      "loss": 0.17362364530563354,
      "step": 1760
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.3616237003343465,
      "learning_rate": 2.3604214493607844e-06,
      "loss": 0.17707206010818483,
      "step": 1770
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.36273369687906065,
      "learning_rate": 2.332544072075333e-06,
      "loss": 0.1810509443283081,
      "step": 1780
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.3538045704417264,
      "learning_rate": 2.30468759569634e-06,
      "loss": 0.1798029899597168,
      "step": 1790
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.3665300341057018,
      "learning_rate": 2.276855497112408e-06,
      "loss": 0.17657459974288942,
      "step": 1800
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.3888504913735948,
      "learning_rate": 2.2490512501694394e-06,
      "loss": 0.1715422749519348,
      "step": 1810
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.3299991726946758,
      "learning_rate": 2.2212783252370496e-06,
      "loss": 0.17716653347015382,
      "step": 1820
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.33385446323519324,
      "learning_rate": 2.1935401887754213e-06,
      "loss": 0.18214735984802247,
      "step": 1830
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.38150050429665633,
      "learning_rate": 2.165840302902632e-06,
      "loss": 0.17498456239700316,
      "step": 1840
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.3772340757946472,
      "learning_rate": 2.1381821249625383e-06,
      "loss": 0.1807033061981201,
      "step": 1850
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.3995250415291945,
      "learning_rate": 2.1105691070932465e-06,
      "loss": 0.17290885448455812,
      "step": 1860
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.3722805742071356,
      "learning_rate": 2.083004695796238e-06,
      "loss": 0.17120599746704102,
      "step": 1870
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.4127408400570583,
      "learning_rate": 2.055492331506194e-06,
      "loss": 0.17853354215621947,
      "step": 1880
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.38023468325708004,
      "learning_rate": 2.0280354481615814e-06,
      "loss": 0.17626445293426513,
      "step": 1890
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.39569320233551497,
      "learning_rate": 2.000637472776049e-06,
      "loss": 0.17760238647460938,
      "step": 1900
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.3395104645209661,
      "learning_rate": 1.973301825010685e-06,
      "loss": 0.17229535579681396,
      "step": 1910
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.40844052109667645,
      "learning_rate": 1.9460319167471934e-06,
      "loss": 0.17733073234558105,
      "step": 1920
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.37810430448153487,
      "learning_rate": 1.9188311516620466e-06,
      "loss": 0.17335065603256225,
      "step": 1930
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.40814146676733887,
      "learning_rate": 1.891702924801651e-06,
      "loss": 0.17898330688476563,
      "step": 1940
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.3602487349877917,
      "learning_rate": 1.864650622158604e-06,
      "loss": 0.18250794410705568,
      "step": 1950
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.3393538607720307,
      "learning_rate": 1.8376776202490666e-06,
      "loss": 0.1782406210899353,
      "step": 1960
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.3915429710722421,
      "learning_rate": 1.8107872856913293e-06,
      "loss": 0.1735108494758606,
      "step": 1970
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.37985437272954925,
      "learning_rate": 1.7839829747856096e-06,
      "loss": 0.1750550627708435,
      "step": 1980
    },
    {
      "epoch": 0.6368,
      "grad_norm": 0.38539358407326346,
      "learning_rate": 1.7572680330951359e-06,
      "loss": 0.1733097553253174,
      "step": 1990
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3738259217786014,
      "learning_rate": 1.7306457950285747e-06,
      "loss": 0.17515769004821777,
      "step": 2000
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.3819943763249579,
      "learning_rate": 1.704119583423848e-06,
      "loss": 0.16884570121765136,
      "step": 2010
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.389801745248532,
      "learning_rate": 1.677692709133396e-06,
      "loss": 0.17741096019744873,
      "step": 2020
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.3653120667345927,
      "learning_rate": 1.6513684706109311e-06,
      "loss": 0.17773522138595582,
      "step": 2030
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.37933212087707896,
      "learning_rate": 1.6251501534997529e-06,
      "loss": 0.17607669830322265,
      "step": 2040
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.3321821531044163,
      "learning_rate": 1.5990410302226405e-06,
      "loss": 0.16934506893157958,
      "step": 2050
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.4060257595183473,
      "learning_rate": 1.5730443595734162e-06,
      "loss": 0.17470494508743287,
      "step": 2060
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.4081327316011303,
      "learning_rate": 1.5471633863101982e-06,
      "loss": 0.1757789969444275,
      "step": 2070
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.38926392491151846,
      "learning_rate": 1.521401340750407e-06,
      "loss": 0.17842782735824586,
      "step": 2080
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.3688818898128435,
      "learning_rate": 1.495761438367577e-06,
      "loss": 0.17657792568206787,
      "step": 2090
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.3996797165946016,
      "learning_rate": 1.4702468793900187e-06,
      "loss": 0.17507438659667968,
      "step": 2100
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.3660129183813682,
      "learning_rate": 1.444860848401384e-06,
      "loss": 0.17568498849868774,
      "step": 2110
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.40919180930559174,
      "learning_rate": 1.4196065139431866e-06,
      "loss": 0.17516255378723145,
      "step": 2120
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.3883216412697579,
      "learning_rate": 1.3944870281193178e-06,
      "loss": 0.17390661239624022,
      "step": 2130
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.36101851282241604,
      "learning_rate": 1.3695055262026208e-06,
      "loss": 0.1793433666229248,
      "step": 2140
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.351474002817543,
      "learning_rate": 1.3446651262435679e-06,
      "loss": 0.17248008251190186,
      "step": 2150
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.3584227522515397,
      "learning_rate": 1.3199689286810746e-06,
      "loss": 0.17412493228912354,
      "step": 2160
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.3698435441682052,
      "learning_rate": 1.2954200159555294e-06,
      "loss": 0.16935389041900634,
      "step": 2170
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.3875871103452335,
      "learning_rate": 1.2710214521240527e-06,
      "loss": 0.17429906129837036,
      "step": 2180
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.38818689314212534,
      "learning_rate": 1.246776282478063e-06,
      "loss": 0.175516939163208,
      "step": 2190
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.35684099766648697,
      "learning_rate": 1.222687533163181e-06,
      "loss": 0.1735387325286865,
      "step": 2200
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.42314786020598566,
      "learning_rate": 1.1987582108015228e-06,
      "loss": 0.1752473831176758,
      "step": 2210
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.3896139405923189,
      "learning_rate": 1.1749913021164255e-06,
      "loss": 0.1759577751159668,
      "step": 2220
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.4351615457717856,
      "learning_rate": 1.1513897735596702e-06,
      "loss": 0.17103490829467774,
      "step": 2230
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.3834559121154836,
      "learning_rate": 1.127956570941218e-06,
      "loss": 0.16780270338058473,
      "step": 2240
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.39841337538803123,
      "learning_rate": 1.104694619061533e-06,
      "loss": 0.1768646717071533,
      "step": 2250
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.37995888893004054,
      "learning_rate": 1.0816068213465295e-06,
      "loss": 0.17878193855285646,
      "step": 2260
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.3883854717990393,
      "learning_rate": 1.0586960594851762e-06,
      "loss": 0.17407444715499878,
      "step": 2270
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.41133967353441564,
      "learning_rate": 1.0359651930698217e-06,
      "loss": 0.17313239574432374,
      "step": 2280
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.42036949314037747,
      "learning_rate": 1.0134170592392837e-06,
      "loss": 0.17422957420349122,
      "step": 2290
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.4147236855875292,
      "learning_rate": 9.910544723247204e-07,
      "loss": 0.17318263053894042,
      "step": 2300
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.43489255396692184,
      "learning_rate": 9.688802234983706e-07,
      "loss": 0.17206907272338867,
      "step": 2310
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.4244354677425118,
      "learning_rate": 9.468970804251742e-07,
      "loss": 0.17518973350524902,
      "step": 2320
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.40784179462155445,
      "learning_rate": 9.251077869173244e-07,
      "loss": 0.1695675015449524,
      "step": 2330
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.3987641796361722,
      "learning_rate": 9.035150625918054e-07,
      "loss": 0.17051217555999756,
      "step": 2340
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.4187928539614081,
      "learning_rate": 8.821216025309395e-07,
      "loss": 0.17391297817230225,
      "step": 2350
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.3690037869142762,
      "learning_rate": 8.609300769460055e-07,
      "loss": 0.16910953521728517,
      "step": 2360
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.3825088904788983,
      "learning_rate": 8.399431308439592e-07,
      "loss": 0.17401565313339235,
      "step": 2370
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.42264696630595905,
      "learning_rate": 8.191633836972962e-07,
      "loss": 0.17544336318969728,
      "step": 2380
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.37637912026899967,
      "learning_rate": 7.985934291171024e-07,
      "loss": 0.17184804677963256,
      "step": 2390
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.41579932435136646,
      "learning_rate": 7.7823583452934e-07,
      "loss": 0.17463115453720093,
      "step": 2400
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.3686323403281178,
      "learning_rate": 7.58093140854389e-07,
      "loss": 0.17356055974960327,
      "step": 2410
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.38576880068059977,
      "learning_rate": 7.381678621899077e-07,
      "loss": 0.17460007667541505,
      "step": 2420
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.42703129906735826,
      "learning_rate": 7.184624854970379e-07,
      "loss": 0.17067699432373046,
      "step": 2430
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.34858414304666985,
      "learning_rate": 6.989794702899932e-07,
      "loss": 0.17033281326293945,
      "step": 2440
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.4152200143790169,
      "learning_rate": 6.797212483290777e-07,
      "loss": 0.17171205282211305,
      "step": 2450
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.38517462156418525,
      "learning_rate": 6.60690223317171e-07,
      "loss": 0.16951452493667601,
      "step": 2460
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.3434889573405646,
      "learning_rate": 6.418887705997046e-07,
      "loss": 0.17201969623565674,
      "step": 2470
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.44971621368685466,
      "learning_rate": 6.23319236868189e-07,
      "loss": 0.1749892234802246,
      "step": 2480
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.36196201020004853,
      "learning_rate": 6.049839398673141e-07,
      "loss": 0.17593114376068114,
      "step": 2490
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.43060409309666475,
      "learning_rate": 5.868851681056567e-07,
      "loss": 0.17676409482955932,
      "step": 2500
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.3991002538591168,
      "learning_rate": 5.690251805700467e-07,
      "loss": 0.17424919605255126,
      "step": 2510
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.40310381220451585,
      "learning_rate": 5.514062064436096e-07,
      "loss": 0.17589831352233887,
      "step": 2520
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.39399011188949445,
      "learning_rate": 5.34030444827533e-07,
      "loss": 0.17281873226165773,
      "step": 2530
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.4083878901260424,
      "learning_rate": 5.169000644665895e-07,
      "loss": 0.16947317123413086,
      "step": 2540
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.3851361432363648,
      "learning_rate": 5.000172034784442e-07,
      "loss": 0.1711735486984253,
      "step": 2550
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.39337090058301416,
      "learning_rate": 4.833839690867853e-07,
      "loss": 0.17315969467163086,
      "step": 2560
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.40037871082344,
      "learning_rate": 4.6700243735831705e-07,
      "loss": 0.16745550632476808,
      "step": 2570
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.37587338743617393,
      "learning_rate": 4.508746529436311e-07,
      "loss": 0.16563820838928223,
      "step": 2580
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.4123037543150221,
      "learning_rate": 4.350026288220083e-07,
      "loss": 0.17215006351470946,
      "step": 2590
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.38068425945027606,
      "learning_rate": 4.1938834605017133e-07,
      "loss": 0.1699705719947815,
      "step": 2600
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.3857308249335943,
      "learning_rate": 4.0403375351501515e-07,
      "loss": 0.16985273361206055,
      "step": 2610
    },
    {
      "epoch": 0.8384,
      "grad_norm": 0.3771439406828394,
      "learning_rate": 3.88940767690362e-07,
      "loss": 0.16914334297180175,
      "step": 2620
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.42087726200191267,
      "learning_rate": 3.7411127239775774e-07,
      "loss": 0.1676808476448059,
      "step": 2630
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.4027649556029035,
      "learning_rate": 3.595471185713431e-07,
      "loss": 0.16704294681549073,
      "step": 2640
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.41244639471418854,
      "learning_rate": 3.4525012402682826e-07,
      "loss": 0.16766990423202516,
      "step": 2650
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.4188553008501556,
      "learning_rate": 3.3122207323460804e-07,
      "loss": 0.17007026672363282,
      "step": 2660
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.37308514287025385,
      "learning_rate": 3.1746471709702963e-07,
      "loss": 0.16677470207214357,
      "step": 2670
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.3866325428538323,
      "learning_rate": 3.039797727298585e-07,
      "loss": 0.17320499420166016,
      "step": 2680
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.44679886990626905,
      "learning_rate": 2.9076892324795546e-07,
      "loss": 0.17161571979522705,
      "step": 2690
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.39358272297855784,
      "learning_rate": 2.778338175551995e-07,
      "loss": 0.16702463626861572,
      "step": 2700
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.40122291418570893,
      "learning_rate": 2.6517607013868326e-07,
      "loss": 0.17485493421554565,
      "step": 2710
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.41570716353907156,
      "learning_rate": 2.527972608672002e-07,
      "loss": 0.17134263515472412,
      "step": 2720
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.4519701719039968,
      "learning_rate": 2.40698934794053e-07,
      "loss": 0.17035775184631347,
      "step": 2730
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.4118651442289031,
      "learning_rate": 2.2888260196421237e-07,
      "loss": 0.1699577808380127,
      "step": 2740
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3941928006067303,
      "learning_rate": 2.1734973722583735e-07,
      "loss": 0.17237366437911988,
      "step": 2750
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.40832217608875215,
      "learning_rate": 2.0610178004619564e-07,
      "loss": 0.16440961360931397,
      "step": 2760
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.3971180240809628,
      "learning_rate": 1.9514013433199834e-07,
      "loss": 0.17652008533477784,
      "step": 2770
    },
    {
      "epoch": 0.8896,
      "grad_norm": 0.407761577951783,
      "learning_rate": 1.8446616825416958e-07,
      "loss": 0.174829363822937,
      "step": 2780
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.43193458090135123,
      "learning_rate": 1.7408121407708007e-07,
      "loss": 0.1740306258201599,
      "step": 2790
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.42658943363969254,
      "learning_rate": 1.6398656799226253e-07,
      "loss": 0.16446399688720703,
      "step": 2800
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.3958915075423234,
      "learning_rate": 1.5418348995662773e-07,
      "loss": 0.170265531539917,
      "step": 2810
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.42859321093856184,
      "learning_rate": 1.4467320353520275e-07,
      "loss": 0.16803120374679564,
      "step": 2820
    },
    {
      "epoch": 0.9056,
      "grad_norm": 0.38305141869358017,
      "learning_rate": 1.3545689574841341e-07,
      "loss": 0.16951898336410523,
      "step": 2830
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.3798038400686433,
      "learning_rate": 1.26535716923927e-07,
      "loss": 0.17529479265213013,
      "step": 2840
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.4110364408684957,
      "learning_rate": 1.1791078055307493e-07,
      "loss": 0.1713550090789795,
      "step": 2850
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.42376391140315484,
      "learning_rate": 1.0958316315187289e-07,
      "loss": 0.17042005062103271,
      "step": 2860
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.3797604889728542,
      "learning_rate": 1.0155390412665528e-07,
      "loss": 0.1684396505355835,
      "step": 2870
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.3914880075566837,
      "learning_rate": 9.38240056443443e-08,
      "loss": 0.16648833751678466,
      "step": 2880
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.4028554721018558,
      "learning_rate": 8.639443250736402e-08,
      "loss": 0.16861410140991212,
      "step": 2890
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.4114822506067345,
      "learning_rate": 7.926611203321777e-08,
      "loss": 0.1701727867126465,
      "step": 2900
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.4152545647271041,
      "learning_rate": 7.243993393874882e-08,
      "loss": 0.17078380584716796,
      "step": 2910
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.38756806680774125,
      "learning_rate": 6.591675022908805e-08,
      "loss": 0.16773648262023927,
      "step": 2920
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.42784086073295813,
      "learning_rate": 5.969737509131241e-08,
      "loss": 0.16553585529327391,
      "step": 2930
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.4042746869722402,
      "learning_rate": 5.3782584792823334e-08,
      "loss": 0.17082666158676146,
      "step": 2940
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.4061618058805029,
      "learning_rate": 4.817311758445686e-08,
      "loss": 0.17140612602233887,
      "step": 2950
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.391081181391559,
      "learning_rate": 4.286967360833866e-08,
      "loss": 0.17159396409988403,
      "step": 2960
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.41735466488493406,
      "learning_rate": 3.787291481049754e-08,
      "loss": 0.17192131280899048,
      "step": 2970
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.4185419841597216,
      "learning_rate": 3.3183464858244364e-08,
      "loss": 0.17720068693161012,
      "step": 2980
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.39722761835716225,
      "learning_rate": 2.8801909062328992e-08,
      "loss": 0.16387460231781006,
      "step": 2990
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3849231647626283,
      "learning_rate": 2.4728794303886248e-08,
      "loss": 0.16302640438079835,
      "step": 3000
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.41574175274160263,
      "learning_rate": 2.0964628966175794e-08,
      "loss": 0.1683335542678833,
      "step": 3010
    },
    {
      "epoch": 0.9664,
      "grad_norm": 0.39544014121768384,
      "learning_rate": 1.750988287113009e-08,
      "loss": 0.16783595085144043,
      "step": 3020
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.37372955897537213,
      "learning_rate": 1.4364987220713278e-08,
      "loss": 0.17508283853530884,
      "step": 3030
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.40018108643956146,
      "learning_rate": 1.1530334543099763e-08,
      "loss": 0.17274413108825684,
      "step": 3040
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.41874909868220905,
      "learning_rate": 9.006278643683697e-09,
      "loss": 0.17360517978668213,
      "step": 3050
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.43201391828339053,
      "learning_rate": 6.793134560916514e-09,
      "loss": 0.17629475593566896,
      "step": 3060
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.35261808222980445,
      "learning_rate": 4.891178526986451e-09,
      "loss": 0.1717047333717346,
      "step": 3070
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.4309234115032205,
      "learning_rate": 3.3006479333413943e-09,
      "loss": 0.17459846735000611,
      "step": 3080
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.4212724049580533,
      "learning_rate": 2.021741301058422e-09,
      "loss": 0.1706598997116089,
      "step": 3090
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.45412833541759956,
      "learning_rate": 1.0546182560652872e-09,
      "loss": 0.17277355194091798,
      "step": 3100
    },
    {
      "epoch": 0.9952,
      "grad_norm": 0.3904315097822292,
      "learning_rate": 3.9939950921774607e-10,
      "loss": 0.1682264804840088,
      "step": 3110
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.3934819932959913,
      "learning_rate": 5.616684123160854e-11,
      "loss": 0.1676420211791992,
      "step": 3120
    },
    {
      "epoch": 1.0,
      "step": 3125,
      "total_flos": 1865119859998720.0,
      "train_loss": 0.1913537948036194,
      "train_runtime": 24530.7619,
      "train_samples_per_second": 8.153,
      "train_steps_per_second": 0.127
    }
  ],
  "logging_steps": 10,
  "max_steps": 3125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1865119859998720.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}