{
  "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": 20.093808181688754,
      "learning_rate": 1.437699680511182e-07,
      "loss": 0.7523126602172852,
      "step": 10
    },
    {
      "epoch": 0.0064,
      "grad_norm": 16.520568445399164,
      "learning_rate": 3.0351437699680514e-07,
      "loss": 0.684361743927002,
      "step": 20
    },
    {
      "epoch": 0.0096,
      "grad_norm": 7.062991511064744,
      "learning_rate": 4.6325878594249205e-07,
      "loss": 0.46736898422241213,
      "step": 30
    },
    {
      "epoch": 0.0128,
      "grad_norm": 1.0572338350229438,
      "learning_rate": 6.230031948881789e-07,
      "loss": 0.3222517013549805,
      "step": 40
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.768970780796944,
      "learning_rate": 7.82747603833866e-07,
      "loss": 0.29146518707275393,
      "step": 50
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.8158618748659492,
      "learning_rate": 9.424920127795528e-07,
      "loss": 0.28341834545135497,
      "step": 60
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.7218086220464439,
      "learning_rate": 1.1022364217252397e-06,
      "loss": 0.2903137683868408,
      "step": 70
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.7459109221323802,
      "learning_rate": 1.2619808306709266e-06,
      "loss": 0.2718811988830566,
      "step": 80
    },
    {
      "epoch": 0.0288,
      "grad_norm": 0.7186860317140319,
      "learning_rate": 1.4217252396166134e-06,
      "loss": 0.2660067558288574,
      "step": 90
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.765918500231858,
      "learning_rate": 1.5814696485623005e-06,
      "loss": 0.26980152130126955,
      "step": 100
    },
    {
      "epoch": 0.0352,
      "grad_norm": 0.7344200083929374,
      "learning_rate": 1.7412140575079875e-06,
      "loss": 0.2695180416107178,
      "step": 110
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.7057487416602337,
      "learning_rate": 1.9009584664536742e-06,
      "loss": 0.2582674264907837,
      "step": 120
    },
    {
      "epoch": 0.0416,
      "grad_norm": 0.6996888798419932,
      "learning_rate": 2.060702875399361e-06,
      "loss": 0.2612154960632324,
      "step": 130
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.7150606291134206,
      "learning_rate": 2.220447284345048e-06,
      "loss": 0.2520437717437744,
      "step": 140
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.7697242977250355,
      "learning_rate": 2.380191693290735e-06,
      "loss": 0.2501786470413208,
      "step": 150
    },
    {
      "epoch": 0.0512,
      "grad_norm": 0.6327215717833664,
      "learning_rate": 2.539936102236422e-06,
      "loss": 0.24434318542480468,
      "step": 160
    },
    {
      "epoch": 0.0544,
      "grad_norm": 0.7947096523807732,
      "learning_rate": 2.699680511182109e-06,
      "loss": 0.25281600952148436,
      "step": 170
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.6717890611061146,
      "learning_rate": 2.8594249201277955e-06,
      "loss": 0.2454531669616699,
      "step": 180
    },
    {
      "epoch": 0.0608,
      "grad_norm": 0.7151585341922304,
      "learning_rate": 3.0191693290734825e-06,
      "loss": 0.2505363464355469,
      "step": 190
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.8601334705182279,
      "learning_rate": 3.17891373801917e-06,
      "loss": 0.2505714178085327,
      "step": 200
    },
    {
      "epoch": 0.0672,
      "grad_norm": 0.6106680426063227,
      "learning_rate": 3.3386581469648564e-06,
      "loss": 0.24775364398956298,
      "step": 210
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.6262984320818072,
      "learning_rate": 3.4984025559105434e-06,
      "loss": 0.24066565036773682,
      "step": 220
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.6078537303186395,
      "learning_rate": 3.6581469648562303e-06,
      "loss": 0.24378209114074706,
      "step": 230
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.5889510426869463,
      "learning_rate": 3.817891373801918e-06,
      "loss": 0.23820171356201172,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5658292689427505,
      "learning_rate": 3.977635782747604e-06,
      "loss": 0.23654117584228515,
      "step": 250
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.5757166706348428,
      "learning_rate": 4.137380191693291e-06,
      "loss": 0.23743386268615724,
      "step": 260
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.5807034355359694,
      "learning_rate": 4.297124600638978e-06,
      "loss": 0.23970918655395507,
      "step": 270
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.5634022487351626,
      "learning_rate": 4.456869009584665e-06,
      "loss": 0.23490209579467775,
      "step": 280
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.5520223075835592,
      "learning_rate": 4.616613418530352e-06,
      "loss": 0.2404552936553955,
      "step": 290
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.5587222430473198,
      "learning_rate": 4.776357827476039e-06,
      "loss": 0.24298410415649413,
      "step": 300
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.542281258937415,
      "learning_rate": 4.936102236421725e-06,
      "loss": 0.22964231967926024,
      "step": 310
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.6339707011249724,
      "learning_rate": 4.999943833158769e-06,
      "loss": 0.22938170433044433,
      "step": 320
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.5290859105179109,
      "learning_rate": 4.999600600490783e-06,
      "loss": 0.23717782497406006,
      "step": 330
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.574404257271199,
      "learning_rate": 4.9989453817439345e-06,
      "loss": 0.23035426139831544,
      "step": 340
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.5887719210155044,
      "learning_rate": 4.997978258698942e-06,
      "loss": 0.230421781539917,
      "step": 350
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.5618660264892863,
      "learning_rate": 4.996699352066659e-06,
      "loss": 0.23192777633666992,
      "step": 360
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.589113954603133,
      "learning_rate": 4.995108821473014e-06,
      "loss": 0.23194873332977295,
      "step": 370
    },
    {
      "epoch": 0.1216,
      "grad_norm": 0.552581223712263,
      "learning_rate": 4.993206865439084e-06,
      "loss": 0.22629022598266602,
      "step": 380
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.5506631212695152,
      "learning_rate": 4.990993721356317e-06,
      "loss": 0.22567858695983886,
      "step": 390
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.5210832665844604,
      "learning_rate": 4.988469665456901e-06,
      "loss": 0.22596418857574463,
      "step": 400
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.5132503738005023,
      "learning_rate": 4.985635012779288e-06,
      "loss": 0.23435051441192628,
      "step": 410
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.5264119522984109,
      "learning_rate": 4.98249011712887e-06,
      "loss": 0.2258882999420166,
      "step": 420
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.5122311697688684,
      "learning_rate": 4.979035371033824e-06,
      "loss": 0.22527906894683838,
      "step": 430
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.5105227090020142,
      "learning_rate": 4.975271205696115e-06,
      "loss": 0.2246992588043213,
      "step": 440
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.5307268054645026,
      "learning_rate": 4.971198090937671e-06,
      "loss": 0.2193459987640381,
      "step": 450
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.46923570087876276,
      "learning_rate": 4.966816535141756e-06,
      "loss": 0.21553544998168944,
      "step": 460
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.4881836025298746,
      "learning_rate": 4.9621270851895035e-06,
      "loss": 0.22505784034729004,
      "step": 470
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.50506411723612,
      "learning_rate": 4.957130326391662e-06,
      "loss": 0.22673957347869872,
      "step": 480
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.5086993434891525,
      "learning_rate": 4.951826882415544e-06,
      "loss": 0.22294471263885499,
      "step": 490
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5280465251135189,
      "learning_rate": 4.946217415207177e-06,
      "loss": 0.21789300441741943,
      "step": 500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.5337843871964275,
      "learning_rate": 4.940302624908689e-06,
      "loss": 0.22192811965942383,
      "step": 510
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.4884343559217744,
      "learning_rate": 4.934083249770912e-06,
      "loss": 0.21614904403686525,
      "step": 520
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.5316592538281818,
      "learning_rate": 4.927560066061251e-06,
      "loss": 0.21973915100097657,
      "step": 530
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.518761429695226,
      "learning_rate": 4.920733887966783e-06,
      "loss": 0.23207192420959472,
      "step": 540
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.511452747175852,
      "learning_rate": 4.913605567492636e-06,
      "loss": 0.21878607273101808,
      "step": 550
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.49924599926539726,
      "learning_rate": 4.906175994355656e-06,
      "loss": 0.22075920104980468,
      "step": 560
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.5259698850641532,
      "learning_rate": 4.898446095873345e-06,
      "loss": 0.22276382446289061,
      "step": 570
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.501751014152873,
      "learning_rate": 4.890416836848128e-06,
      "loss": 0.21954989433288574,
      "step": 580
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.5167201593356286,
      "learning_rate": 4.882089219446925e-06,
      "loss": 0.2145029067993164,
      "step": 590
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.5006060240232905,
      "learning_rate": 4.873464283076074e-06,
      "loss": 0.22003324031829835,
      "step": 600
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.4477538874438277,
      "learning_rate": 4.864543104251587e-06,
      "loss": 0.21916275024414061,
      "step": 610
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.4832933241270485,
      "learning_rate": 4.855326796464798e-06,
      "loss": 0.2203526973724365,
      "step": 620
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.5359361967005408,
      "learning_rate": 4.8458165100433725e-06,
      "loss": 0.21596732139587402,
      "step": 630
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.5708003689943741,
      "learning_rate": 4.836013432007738e-06,
      "loss": 0.2171140193939209,
      "step": 640
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.4831169531465719,
      "learning_rate": 4.825918785922921e-06,
      "loss": 0.22040581703186035,
      "step": 650
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.4982382400104379,
      "learning_rate": 4.8155338317458315e-06,
      "loss": 0.21841506958007811,
      "step": 660
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.4741071764041748,
      "learning_rate": 4.804859865668002e-06,
      "loss": 0.2143453598022461,
      "step": 670
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.47853550451884025,
      "learning_rate": 4.793898219953804e-06,
      "loss": 0.21545085906982422,
      "step": 680
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.4902247743421047,
      "learning_rate": 4.782650262774164e-06,
      "loss": 0.2166231393814087,
      "step": 690
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.4611717059287351,
      "learning_rate": 4.7711173980357886e-06,
      "loss": 0.21284222602844238,
      "step": 700
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.4815654128340087,
      "learning_rate": 4.759301065205947e-06,
      "loss": 0.21358721256256102,
      "step": 710
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.5049245613626656,
      "learning_rate": 4.7472027391328e-06,
      "loss": 0.21447527408599854,
      "step": 720
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.4758997167389971,
      "learning_rate": 4.734823929861317e-06,
      "loss": 0.21809780597686768,
      "step": 730
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.5423173365143716,
      "learning_rate": 4.722166182444801e-06,
      "loss": 0.21390962600708008,
      "step": 740
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.44572231492476455,
      "learning_rate": 4.709231076752045e-06,
      "loss": 0.21404554843902587,
      "step": 750
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.4848421373802031,
      "learning_rate": 4.696020227270142e-06,
      "loss": 0.21710457801818847,
      "step": 760
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.518532765750562,
      "learning_rate": 4.6825352829029705e-06,
      "loss": 0.21285481452941896,
      "step": 770
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.5008678397970389,
      "learning_rate": 4.668777926765392e-06,
      "loss": 0.21155524253845215,
      "step": 780
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.48720974823345864,
      "learning_rate": 4.6547498759731725e-06,
      "loss": 0.20655455589294433,
      "step": 790
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.49528977499161353,
      "learning_rate": 4.6404528814286575e-06,
      "loss": 0.2101435422897339,
      "step": 800
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.4532686250809506,
      "learning_rate": 4.6258887276022425e-06,
      "loss": 0.21684365272521972,
      "step": 810
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.49803115837380546,
      "learning_rate": 4.611059232309639e-06,
      "loss": 0.21193151473999022,
      "step": 820
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.5153783225404047,
      "learning_rate": 4.595966246484986e-06,
      "loss": 0.21344296932220458,
      "step": 830
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.4765272009238815,
      "learning_rate": 4.580611653949829e-06,
      "loss": 0.21319386959075928,
      "step": 840
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.5228745905777464,
      "learning_rate": 4.564997371177992e-06,
      "loss": 0.21112470626831054,
      "step": 850
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.4583805155148445,
      "learning_rate": 4.54912534705637e-06,
      "loss": 0.2108391284942627,
      "step": 860
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.4920259584441244,
      "learning_rate": 4.532997562641683e-06,
      "loss": 0.20768051147460936,
      "step": 870
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.5200095181799963,
      "learning_rate": 4.516616030913214e-06,
      "loss": 0.21211957931518555,
      "step": 880
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.4788503683270311,
      "learning_rate": 4.499982796521556e-06,
      "loss": 0.20693025588989258,
      "step": 890
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.4666456137071941,
      "learning_rate": 4.48309993553341e-06,
      "loss": 0.20890872478485106,
      "step": 900
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.4794527139448749,
      "learning_rate": 4.465969555172468e-06,
      "loss": 0.20777955055236816,
      "step": 910
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.4616610840587355,
      "learning_rate": 4.448593793556391e-06,
      "loss": 0.21416122913360597,
      "step": 920
    },
    {
      "epoch": 0.2976,
      "grad_norm": 0.47725407011391663,
      "learning_rate": 4.430974819429954e-06,
      "loss": 0.20783448219299316,
      "step": 930
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.4596350013424985,
      "learning_rate": 4.413114831894344e-06,
      "loss": 0.20199823379516602,
      "step": 940
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.4940149958405755,
      "learning_rate": 4.3950160601326865e-06,
      "loss": 0.20049993991851806,
      "step": 950
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.4891958940488766,
      "learning_rate": 4.376680763131811e-06,
      "loss": 0.20765538215637208,
      "step": 960
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.5373640149223949,
      "learning_rate": 4.358111229400296e-06,
      "loss": 0.2103745460510254,
      "step": 970
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.5035919946088194,
      "learning_rate": 4.33930977668283e-06,
      "loss": 0.2148181438446045,
      "step": 980
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.498832420199319,
      "learning_rate": 4.320278751670922e-06,
      "loss": 0.20667800903320313,
      "step": 990
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5016480811009209,
      "learning_rate": 4.301020529710009e-06,
      "loss": 0.20847175121307374,
      "step": 1000
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.5355131410598809,
      "learning_rate": 4.281537514502962e-06,
      "loss": 0.21192097663879395,
      "step": 1010
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.49710771531514497,
      "learning_rate": 4.261832137810093e-06,
      "loss": 0.20849306583404542,
      "step": 1020
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.4702938633516668,
      "learning_rate": 4.241906859145611e-06,
      "loss": 0.20947628021240233,
      "step": 1030
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.47328762785100176,
      "learning_rate": 4.221764165470661e-06,
      "loss": 0.20568199157714845,
      "step": 1040
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.48090607151875236,
      "learning_rate": 4.201406570882898e-06,
      "loss": 0.20522446632385255,
      "step": 1050
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.46870182419574746,
      "learning_rate": 4.180836616302704e-06,
      "loss": 0.2044762134552002,
      "step": 1060
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.49284234006242156,
      "learning_rate": 4.160056869156041e-06,
      "loss": 0.20835609436035157,
      "step": 1070
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.425482663225026,
      "learning_rate": 4.139069923053995e-06,
      "loss": 0.20575876235961915,
      "step": 1080
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.46647669293000804,
      "learning_rate": 4.117878397469062e-06,
      "loss": 0.20992250442504884,
      "step": 1090
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.4464343988416538,
      "learning_rate": 4.096484937408195e-06,
      "loss": 0.20092244148254396,
      "step": 1100
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.5116088744854695,
      "learning_rate": 4.074892213082676e-06,
      "loss": 0.20036702156066893,
      "step": 1110
    },
    {
      "epoch": 0.3584,
      "grad_norm": 4.940314739525779,
      "learning_rate": 4.0531029195748265e-06,
      "loss": 0.21338913440704346,
      "step": 1120
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.4721397920115156,
      "learning_rate": 4.03111977650163e-06,
      "loss": 0.20792775154113768,
      "step": 1130
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.5105519348301445,
      "learning_rate": 4.008945527675281e-06,
      "loss": 0.2061443328857422,
      "step": 1140
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.523180958068929,
      "learning_rate": 3.986582940760717e-06,
      "loss": 0.1962942123413086,
      "step": 1150
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.5027335828799008,
      "learning_rate": 3.9640348069301785e-06,
      "loss": 0.2031947612762451,
      "step": 1160
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.48735270934050073,
      "learning_rate": 3.941303940514826e-06,
      "loss": 0.20448057651519774,
      "step": 1170
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.5075332440871839,
      "learning_rate": 3.918393178653472e-06,
      "loss": 0.20594587326049804,
      "step": 1180
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.4485083644552742,
      "learning_rate": 3.895305380938468e-06,
      "loss": 0.20264167785644532,
      "step": 1190
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.4568492727427137,
      "learning_rate": 3.872043429058783e-06,
      "loss": 0.20010733604431152,
      "step": 1200
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.46103501808297814,
      "learning_rate": 3.84861022644033e-06,
      "loss": 0.2026883602142334,
      "step": 1210
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.46609834517793386,
      "learning_rate": 3.825008697883574e-06,
      "loss": 0.21079249382019044,
      "step": 1220
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.49992288047242467,
      "learning_rate": 3.8012417891984776e-06,
      "loss": 0.2031094551086426,
      "step": 1230
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.4746264528155682,
      "learning_rate": 3.777312466836819e-06,
      "loss": 0.20238199234008789,
      "step": 1240
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.45243385346205817,
      "learning_rate": 3.7532237175219378e-06,
      "loss": 0.20085253715515136,
      "step": 1250
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.48931316379420287,
      "learning_rate": 3.728978547875948e-06,
      "loss": 0.20520598888397218,
      "step": 1260
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.5229456414008956,
      "learning_rate": 3.7045799840444712e-06,
      "loss": 0.19984333515167235,
      "step": 1270
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.4773055647919508,
      "learning_rate": 3.6800310713189258e-06,
      "loss": 0.20064287185668944,
      "step": 1280
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.4824962267097886,
      "learning_rate": 3.6553348737564328e-06,
      "loss": 0.20138092041015626,
      "step": 1290
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.47245858486532044,
      "learning_rate": 3.6304944737973794e-06,
      "loss": 0.20704314708709717,
      "step": 1300
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.47670774891547607,
      "learning_rate": 3.6055129718806836e-06,
      "loss": 0.20015296936035157,
      "step": 1310
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.4553061754046557,
      "learning_rate": 3.5803934860568134e-06,
      "loss": 0.19692450761795044,
      "step": 1320
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.5124220374815842,
      "learning_rate": 3.5551391515986163e-06,
      "loss": 0.2016448497772217,
      "step": 1330
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.4809826187082155,
      "learning_rate": 3.529753120609982e-06,
      "loss": 0.19793987274169922,
      "step": 1340
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.48798480914379067,
      "learning_rate": 3.5042385616324243e-06,
      "loss": 0.20041651725769044,
      "step": 1350
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.4589600174491072,
      "learning_rate": 3.4785986592495934e-06,
      "loss": 0.19874777793884277,
      "step": 1360
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.44810416886840765,
      "learning_rate": 3.452836613689803e-06,
      "loss": 0.19696075916290284,
      "step": 1370
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.4584133576368786,
      "learning_rate": 3.426955640426584e-06,
      "loss": 0.20014967918395996,
      "step": 1380
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.46474214573205574,
      "learning_rate": 3.4009589697773605e-06,
      "loss": 0.19937365055084227,
      "step": 1390
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.4671452045462699,
      "learning_rate": 3.3748498465002475e-06,
      "loss": 0.19703936576843262,
      "step": 1400
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.48450994567172556,
      "learning_rate": 3.3486315293890693e-06,
      "loss": 0.20506525039672852,
      "step": 1410
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.48940983460177095,
      "learning_rate": 3.3223072908666053e-06,
      "loss": 0.19508613348007203,
      "step": 1420
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.5510507698314822,
      "learning_rate": 3.295880416576153e-06,
      "loss": 0.20555310249328612,
      "step": 1430
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.45473195837081576,
      "learning_rate": 3.269354204971427e-06,
      "loss": 0.19813575744628906,
      "step": 1440
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.4854091562037593,
      "learning_rate": 3.242731966904865e-06,
      "loss": 0.19694712162017822,
      "step": 1450
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.4637441174996577,
      "learning_rate": 3.2160170252143913e-06,
      "loss": 0.1959088087081909,
      "step": 1460
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.4460606032902631,
      "learning_rate": 3.1892127143086716e-06,
      "loss": 0.20340628623962403,
      "step": 1470
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.4768689558424143,
      "learning_rate": 3.1623223797509347e-06,
      "loss": 0.19146734476089478,
      "step": 1480
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.46631038217283505,
      "learning_rate": 3.135349377841396e-06,
      "loss": 0.19588179588317872,
      "step": 1490
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.48197350793708515,
      "learning_rate": 3.1082970751983497e-06,
      "loss": 0.20245718955993652,
      "step": 1500
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.44408940491911375,
      "learning_rate": 3.0811688483379546e-06,
      "loss": 0.19959219694137573,
      "step": 1510
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.47255519902507054,
      "learning_rate": 3.0539680832528074e-06,
      "loss": 0.1994904398918152,
      "step": 1520
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.48800627171777977,
      "learning_rate": 3.026698174989316e-06,
      "loss": 0.19807126522064208,
      "step": 1530
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.4748737132528679,
      "learning_rate": 2.999362527223952e-06,
      "loss": 0.19806113243103027,
      "step": 1540
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.47637730688550123,
      "learning_rate": 2.9719645518384194e-06,
      "loss": 0.19955278635025026,
      "step": 1550
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.5411554495039922,
      "learning_rate": 2.944507668493807e-06,
      "loss": 0.202299165725708,
      "step": 1560
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.48642193804707995,
      "learning_rate": 2.9169953042037623e-06,
      "loss": 0.19863581657409668,
      "step": 1570
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.5363553346933208,
      "learning_rate": 2.889430892906754e-06,
      "loss": 0.19409118890762328,
      "step": 1580
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.47187050499878397,
      "learning_rate": 2.861817875037462e-06,
      "loss": 0.1912764310836792,
      "step": 1590
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.5163595948637988,
      "learning_rate": 2.8341596970973683e-06,
      "loss": 0.20115599632263184,
      "step": 1600
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.5033907485073755,
      "learning_rate": 2.80645981122458e-06,
      "loss": 0.19687057733535768,
      "step": 1610
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.4753722793172304,
      "learning_rate": 2.7787216747629508e-06,
      "loss": 0.20292258262634277,
      "step": 1620
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.46781165760957,
      "learning_rate": 2.7509487498305615e-06,
      "loss": 0.18959319591522217,
      "step": 1630
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.4803554793777817,
      "learning_rate": 2.7231445028875924e-06,
      "loss": 0.19619333744049072,
      "step": 1640
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.43719126287209875,
      "learning_rate": 2.6953124043036604e-06,
      "loss": 0.19511375427246094,
      "step": 1650
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.4689037514921924,
      "learning_rate": 2.667455927924667e-06,
      "loss": 0.19399585723876953,
      "step": 1660
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.48479905355532704,
      "learning_rate": 2.6395785506392164e-06,
      "loss": 0.1896076202392578,
      "step": 1670
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.516453973005613,
      "learning_rate": 2.6116837519446407e-06,
      "loss": 0.1939442992210388,
      "step": 1680
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.47710575683228795,
      "learning_rate": 2.5837750135127192e-06,
      "loss": 0.19078316688537597,
      "step": 1690
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.47654319681013313,
      "learning_rate": 2.555855818755108e-06,
      "loss": 0.19690483808517456,
      "step": 1700
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.5030326386548561,
      "learning_rate": 2.5279296523885636e-06,
      "loss": 0.19325432777404786,
      "step": 1710
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.49452423153374125,
      "learning_rate": 2.5e-06,
      "loss": 0.19436432123184205,
      "step": 1720
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.5135088244704792,
      "learning_rate": 2.472070347611437e-06,
      "loss": 0.1878933072090149,
      "step": 1730
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.5160118206798595,
      "learning_rate": 2.444144181244893e-06,
      "loss": 0.19355961084365844,
      "step": 1740
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.5069308846787346,
      "learning_rate": 2.416224986487282e-06,
      "loss": 0.19122695922851562,
      "step": 1750
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.5385800538703149,
      "learning_rate": 2.3883162480553605e-06,
      "loss": 0.18820159435272216,
      "step": 1760
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.49129457413116234,
      "learning_rate": 2.3604214493607844e-06,
      "loss": 0.19197521209716797,
      "step": 1770
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.4908165776123557,
      "learning_rate": 2.332544072075333e-06,
      "loss": 0.19534649848937988,
      "step": 1780
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.49497656453552125,
      "learning_rate": 2.30468759569634e-06,
      "loss": 0.19484236240386962,
      "step": 1790
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.466973816624908,
      "learning_rate": 2.276855497112408e-06,
      "loss": 0.191474986076355,
      "step": 1800
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.498294237386886,
      "learning_rate": 2.2490512501694394e-06,
      "loss": 0.18636202812194824,
      "step": 1810
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.5110432771457695,
      "learning_rate": 2.2212783252370496e-06,
      "loss": 0.19112749099731446,
      "step": 1820
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.4923044532988948,
      "learning_rate": 2.1935401887754213e-06,
      "loss": 0.19590845108032226,
      "step": 1830
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.49881036242858373,
      "learning_rate": 2.165840302902632e-06,
      "loss": 0.18917866945266723,
      "step": 1840
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.5070848566140863,
      "learning_rate": 2.1381821249625383e-06,
      "loss": 0.1955878973007202,
      "step": 1850
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.5245919327161893,
      "learning_rate": 2.1105691070932465e-06,
      "loss": 0.18681724071502687,
      "step": 1860
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.5043139368489675,
      "learning_rate": 2.083004695796238e-06,
      "loss": 0.185194993019104,
      "step": 1870
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.5180452275250914,
      "learning_rate": 2.055492331506194e-06,
      "loss": 0.1928567886352539,
      "step": 1880
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.5320215436686966,
      "learning_rate": 2.0280354481615814e-06,
      "loss": 0.19074957370758056,
      "step": 1890
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.4725862343819939,
      "learning_rate": 2.000637472776049e-06,
      "loss": 0.19257795810699463,
      "step": 1900
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.46908638481055026,
      "learning_rate": 1.973301825010685e-06,
      "loss": 0.18594731092453004,
      "step": 1910
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.5595713557618127,
      "learning_rate": 1.9460319167471934e-06,
      "loss": 0.19121139049530028,
      "step": 1920
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.507704360185881,
      "learning_rate": 1.9188311516620466e-06,
      "loss": 0.18624544143676758,
      "step": 1930
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.4860192603301521,
      "learning_rate": 1.891702924801651e-06,
      "loss": 0.19231630563735963,
      "step": 1940
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.5275367662218493,
      "learning_rate": 1.864650622158604e-06,
      "loss": 0.19608126878738402,
      "step": 1950
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.49282562967431837,
      "learning_rate": 1.8376776202490666e-06,
      "loss": 0.19235665798187257,
      "step": 1960
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.5182260002744055,
      "learning_rate": 1.8107872856913293e-06,
      "loss": 0.18613014221191407,
      "step": 1970
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.5103313601861706,
      "learning_rate": 1.7839829747856096e-06,
      "loss": 0.1881113052368164,
      "step": 1980
    },
    {
      "epoch": 0.6368,
      "grad_norm": 0.5451499180289584,
      "learning_rate": 1.7572680330951359e-06,
      "loss": 0.18735458850860595,
      "step": 1990
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5090636315844644,
      "learning_rate": 1.7306457950285747e-06,
      "loss": 0.1885282278060913,
      "step": 2000
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.4758742975901025,
      "learning_rate": 1.704119583423848e-06,
      "loss": 0.18241598606109619,
      "step": 2010
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.49602490022248863,
      "learning_rate": 1.677692709133396e-06,
      "loss": 0.19074147939682007,
      "step": 2020
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.520455285125112,
      "learning_rate": 1.6513684706109311e-06,
      "loss": 0.19024887084960937,
      "step": 2030
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.5234524283247538,
      "learning_rate": 1.6251501534997529e-06,
      "loss": 0.18900917768478392,
      "step": 2040
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.4762667999370438,
      "learning_rate": 1.5990410302226405e-06,
      "loss": 0.18147594928741456,
      "step": 2050
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.4931916769975977,
      "learning_rate": 1.5730443595734162e-06,
      "loss": 0.18815698623657226,
      "step": 2060
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.5595459804684163,
      "learning_rate": 1.5471633863101982e-06,
      "loss": 0.18958520889282227,
      "step": 2070
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.551381176131532,
      "learning_rate": 1.521401340750407e-06,
      "loss": 0.1908926248550415,
      "step": 2080
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.5155022860725758,
      "learning_rate": 1.495761438367577e-06,
      "loss": 0.18872777223587037,
      "step": 2090
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.6037433446756716,
      "learning_rate": 1.4702468793900187e-06,
      "loss": 0.18800405263900757,
      "step": 2100
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.5613773833705744,
      "learning_rate": 1.444860848401384e-06,
      "loss": 0.18743778467178346,
      "step": 2110
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.5277286435676816,
      "learning_rate": 1.4196065139431866e-06,
      "loss": 0.18769149780273436,
      "step": 2120
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.5487755330646784,
      "learning_rate": 1.3944870281193178e-06,
      "loss": 0.1866753101348877,
      "step": 2130
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.5319334450957595,
      "learning_rate": 1.3695055262026208e-06,
      "loss": 0.19193503856658936,
      "step": 2140
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.5061777243502238,
      "learning_rate": 1.3446651262435679e-06,
      "loss": 0.18499069213867186,
      "step": 2150
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.5063080834031065,
      "learning_rate": 1.3199689286810746e-06,
      "loss": 0.18700281381607056,
      "step": 2160
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.5014045449596041,
      "learning_rate": 1.2954200159555294e-06,
      "loss": 0.18185386657714844,
      "step": 2170
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.5417896517828541,
      "learning_rate": 1.2710214521240527e-06,
      "loss": 0.18632771968841552,
      "step": 2180
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.5710908799443121,
      "learning_rate": 1.246776282478063e-06,
      "loss": 0.18732945919036864,
      "step": 2190
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.5180508096448415,
      "learning_rate": 1.222687533163181e-06,
      "loss": 0.18602204322814941,
      "step": 2200
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.5480758918229119,
      "learning_rate": 1.1987582108015228e-06,
      "loss": 0.18710973262786865,
      "step": 2210
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.5631818126474104,
      "learning_rate": 1.1749913021164255e-06,
      "loss": 0.18828771114349366,
      "step": 2220
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.4833634541431531,
      "learning_rate": 1.1513897735596702e-06,
      "loss": 0.18257718086242675,
      "step": 2230
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.5051522117897481,
      "learning_rate": 1.127956570941218e-06,
      "loss": 0.17966469526290893,
      "step": 2240
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5404271805851407,
      "learning_rate": 1.104694619061533e-06,
      "loss": 0.18814800977706908,
      "step": 2250
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.5147342090287059,
      "learning_rate": 1.0816068213465295e-06,
      "loss": 0.1908186197280884,
      "step": 2260
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.5558495401174878,
      "learning_rate": 1.0586960594851762e-06,
      "loss": 0.1859324097633362,
      "step": 2270
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.6185737554957568,
      "learning_rate": 1.0359651930698217e-06,
      "loss": 0.18477405309677125,
      "step": 2280
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.5398647348951853,
      "learning_rate": 1.0134170592392837e-06,
      "loss": 0.1857767939567566,
      "step": 2290
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.5450678028060058,
      "learning_rate": 9.910544723247204e-07,
      "loss": 0.184822678565979,
      "step": 2300
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.5999082382312588,
      "learning_rate": 9.688802234983706e-07,
      "loss": 0.18381783962249756,
      "step": 2310
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.5175099712487172,
      "learning_rate": 9.468970804251742e-07,
      "loss": 0.18641353845596315,
      "step": 2320
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.5367638040398911,
      "learning_rate": 9.251077869173244e-07,
      "loss": 0.18090612888336183,
      "step": 2330
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.563594153188617,
      "learning_rate": 9.035150625918054e-07,
      "loss": 0.18149322271347046,
      "step": 2340
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.5304713442318342,
      "learning_rate": 8.821216025309395e-07,
      "loss": 0.18464915752410888,
      "step": 2350
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.535119183480021,
      "learning_rate": 8.609300769460055e-07,
      "loss": 0.1792607307434082,
      "step": 2360
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.5724539486438234,
      "learning_rate": 8.399431308439592e-07,
      "loss": 0.183684778213501,
      "step": 2370
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.5589161632397335,
      "learning_rate": 8.191633836972962e-07,
      "loss": 0.18650429248809813,
      "step": 2380
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.5386156132762686,
      "learning_rate": 7.985934291171024e-07,
      "loss": 0.1821720838546753,
      "step": 2390
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.5321288466713382,
      "learning_rate": 7.7823583452934e-07,
      "loss": 0.18489625453948974,
      "step": 2400
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.5670301824645666,
      "learning_rate": 7.58093140854389e-07,
      "loss": 0.18495336771011353,
      "step": 2410
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.6058756306995335,
      "learning_rate": 7.381678621899077e-07,
      "loss": 0.1848145008087158,
      "step": 2420
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.5477002870283818,
      "learning_rate": 7.184624854970379e-07,
      "loss": 0.1817490816116333,
      "step": 2430
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.5458027173632266,
      "learning_rate": 6.989794702899932e-07,
      "loss": 0.18078404664993286,
      "step": 2440
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.5772130708628379,
      "learning_rate": 6.797212483290777e-07,
      "loss": 0.18299766778945922,
      "step": 2450
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.5674146932938366,
      "learning_rate": 6.60690223317171e-07,
      "loss": 0.1799448013305664,
      "step": 2460
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.5238538237059384,
      "learning_rate": 6.418887705997046e-07,
      "loss": 0.1826066255569458,
      "step": 2470
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.5857270779434125,
      "learning_rate": 6.23319236868189e-07,
      "loss": 0.18549437522888185,
      "step": 2480
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.5274424793724192,
      "learning_rate": 6.049839398673141e-07,
      "loss": 0.1865037798881531,
      "step": 2490
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5820741885019232,
      "learning_rate": 5.868851681056567e-07,
      "loss": 0.18739759922027588,
      "step": 2500
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.559971376703767,
      "learning_rate": 5.690251805700467e-07,
      "loss": 0.1853170394897461,
      "step": 2510
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.5456407872897143,
      "learning_rate": 5.514062064436096e-07,
      "loss": 0.18589026927948,
      "step": 2520
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.5866178273652722,
      "learning_rate": 5.34030444827533e-07,
      "loss": 0.1827709197998047,
      "step": 2530
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.588749656654477,
      "learning_rate": 5.169000644665895e-07,
      "loss": 0.1794450044631958,
      "step": 2540
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.5778176841150756,
      "learning_rate": 5.000172034784442e-07,
      "loss": 0.18060548305511476,
      "step": 2550
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.566426267196354,
      "learning_rate": 4.833839690867853e-07,
      "loss": 0.18326361179351808,
      "step": 2560
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.5763812670051818,
      "learning_rate": 4.6700243735831705e-07,
      "loss": 0.17798151969909667,
      "step": 2570
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.5465254160649792,
      "learning_rate": 4.508746529436311e-07,
      "loss": 0.1761394739151001,
      "step": 2580
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.5717164779412172,
      "learning_rate": 4.350026288220083e-07,
      "loss": 0.18241602182388306,
      "step": 2590
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.5532919690194787,
      "learning_rate": 4.1938834605017133e-07,
      "loss": 0.1799800157546997,
      "step": 2600
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.5485503614596886,
      "learning_rate": 4.0403375351501515e-07,
      "loss": 0.18037915229797363,
      "step": 2610
    },
    {
      "epoch": 0.8384,
      "grad_norm": 0.5921392059955939,
      "learning_rate": 3.88940767690362e-07,
      "loss": 0.17850807905197144,
      "step": 2620
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.6173777417506611,
      "learning_rate": 3.7411127239775774e-07,
      "loss": 0.17773046493530273,
      "step": 2630
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.5704461135916385,
      "learning_rate": 3.595471185713431e-07,
      "loss": 0.17534157037734985,
      "step": 2640
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.6016600022490033,
      "learning_rate": 3.4525012402682826e-07,
      "loss": 0.17784465551376344,
      "step": 2650
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.5793357844007763,
      "learning_rate": 3.3122207323460804e-07,
      "loss": 0.17941689491271973,
      "step": 2660
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.5402101980665998,
      "learning_rate": 3.1746471709702963e-07,
      "loss": 0.17694177627563476,
      "step": 2670
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.5764717205309013,
      "learning_rate": 3.039797727298585e-07,
      "loss": 0.18307201862335204,
      "step": 2680
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.6021889152147203,
      "learning_rate": 2.9076892324795546e-07,
      "loss": 0.18175405263900757,
      "step": 2690
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.5783244972157141,
      "learning_rate": 2.778338175551995e-07,
      "loss": 0.17646790742874147,
      "step": 2700
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.573282650162234,
      "learning_rate": 2.6517607013868326e-07,
      "loss": 0.18459818363189698,
      "step": 2710
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.6039696058732922,
      "learning_rate": 2.527972608672002e-07,
      "loss": 0.18084490299224854,
      "step": 2720
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.5916439702722857,
      "learning_rate": 2.40698934794053e-07,
      "loss": 0.18053301572799682,
      "step": 2730
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.5703451942226244,
      "learning_rate": 2.2888260196421237e-07,
      "loss": 0.1792958378791809,
      "step": 2740
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.5672304805383847,
      "learning_rate": 2.1734973722583735e-07,
      "loss": 0.1819172501564026,
      "step": 2750
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.5784570642525821,
      "learning_rate": 2.0610178004619564e-07,
      "loss": 0.17332799434661866,
      "step": 2760
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.575451427907292,
      "learning_rate": 1.9514013433199834e-07,
      "loss": 0.18558990955352783,
      "step": 2770
    },
    {
      "epoch": 0.8896,
      "grad_norm": 0.5133461724908028,
      "learning_rate": 1.8446616825416958e-07,
      "loss": 0.18399085998535156,
      "step": 2780
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.6123280023323261,
      "learning_rate": 1.7408121407708007e-07,
      "loss": 0.1844745397567749,
      "step": 2790
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.5761361465385083,
      "learning_rate": 1.6398656799226253e-07,
      "loss": 0.17304511070251466,
      "step": 2800
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.6034414454227958,
      "learning_rate": 1.5418348995662773e-07,
      "loss": 0.17871806621551514,
      "step": 2810
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.5923974971972374,
      "learning_rate": 1.4467320353520275e-07,
      "loss": 0.17667040824890137,
      "step": 2820
    },
    {
      "epoch": 0.9056,
      "grad_norm": 0.603734748014922,
      "learning_rate": 1.3545689574841341e-07,
      "loss": 0.1787508487701416,
      "step": 2830
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.5750783540393263,
      "learning_rate": 1.26535716923927e-07,
      "loss": 0.18438329696655273,
      "step": 2840
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.5716942434142535,
      "learning_rate": 1.1791078055307493e-07,
      "loss": 0.1802410364151001,
      "step": 2850
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.6031535401501658,
      "learning_rate": 1.0958316315187289e-07,
      "loss": 0.17950894832611083,
      "step": 2860
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.5724651470732645,
      "learning_rate": 1.0155390412665528e-07,
      "loss": 0.17800890207290648,
      "step": 2870
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.5920847136083833,
      "learning_rate": 9.38240056443443e-08,
      "loss": 0.17559461593627929,
      "step": 2880
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.5600845233888927,
      "learning_rate": 8.639443250736402e-08,
      "loss": 0.17780338525772094,
      "step": 2890
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.5760602589693042,
      "learning_rate": 7.926611203321777e-08,
      "loss": 0.1794909954071045,
      "step": 2900
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.59057677772977,
      "learning_rate": 7.243993393874882e-08,
      "loss": 0.1795297384262085,
      "step": 2910
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.5693422129621047,
      "learning_rate": 6.591675022908805e-08,
      "loss": 0.17676992416381837,
      "step": 2920
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.5656532345210596,
      "learning_rate": 5.969737509131241e-08,
      "loss": 0.17433459758758546,
      "step": 2930
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.5865348817236666,
      "learning_rate": 5.3782584792823334e-08,
      "loss": 0.1795581579208374,
      "step": 2940
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.6034375830769324,
      "learning_rate": 4.817311758445686e-08,
      "loss": 0.18066773414611817,
      "step": 2950
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.598761782830776,
      "learning_rate": 4.286967360833866e-08,
      "loss": 0.1803189516067505,
      "step": 2960
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.5410244646488507,
      "learning_rate": 3.787291481049754e-08,
      "loss": 0.18075671195983886,
      "step": 2970
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.6102805369465131,
      "learning_rate": 3.3183464858244364e-08,
      "loss": 0.18705531358718872,
      "step": 2980
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.5798299084498433,
      "learning_rate": 2.8801909062328992e-08,
      "loss": 0.17331962585449218,
      "step": 2990
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.5999449762716584,
      "learning_rate": 2.4728794303886248e-08,
      "loss": 0.17158935070037842,
      "step": 3000
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.6212882795186798,
      "learning_rate": 2.0964628966175794e-08,
      "loss": 0.17738908529281616,
      "step": 3010
    },
    {
      "epoch": 0.9664,
      "grad_norm": 0.564746561855876,
      "learning_rate": 1.750988287113009e-08,
      "loss": 0.17667733430862426,
      "step": 3020
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.5852806549215316,
      "learning_rate": 1.4364987220713278e-08,
      "loss": 0.18457986116409303,
      "step": 3030
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.5991233203919278,
      "learning_rate": 1.1530334543099763e-08,
      "loss": 0.18215363025665282,
      "step": 3040
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.6041102228390866,
      "learning_rate": 9.006278643683697e-09,
      "loss": 0.18243587017059326,
      "step": 3050
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.5869697890802611,
      "learning_rate": 6.793134560916514e-09,
      "loss": 0.18486570119857787,
      "step": 3060
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.5595978682216465,
      "learning_rate": 4.891178526986451e-09,
      "loss": 0.18047856092453002,
      "step": 3070
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.5638404572903396,
      "learning_rate": 3.3006479333413943e-09,
      "loss": 0.18349089622497558,
      "step": 3080
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.5582534730189623,
      "learning_rate": 2.021741301058422e-09,
      "loss": 0.18032891750335694,
      "step": 3090
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.5757824692806152,
      "learning_rate": 1.0546182560652872e-09,
      "loss": 0.1812995433807373,
      "step": 3100
    },
    {
      "epoch": 0.9952,
      "grad_norm": 0.5718406851297113,
      "learning_rate": 3.9939950921774607e-10,
      "loss": 0.17747504711151124,
      "step": 3110
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.549457935685087,
      "learning_rate": 5.616684123160854e-11,
      "loss": 0.17633507251739503,
      "step": 3120
    },
    {
      "epoch": 1.0,
      "step": 3125,
      "total_flos": 2477163648385024.0,
      "train_loss": 0.20598802658081056,
      "train_runtime": 35266.4791,
      "train_samples_per_second": 5.671,
      "train_steps_per_second": 0.089
    }
  ],
  "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": 2477163648385024.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}