{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 2414,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00041425020712510354,
      "grad_norm": 24.199668764830136,
      "learning_rate": 4.132231404958678e-08,
      "loss": 1.3991,
      "step": 1
    },
    {
      "epoch": 0.002071251035625518,
      "grad_norm": 23.414571187298456,
      "learning_rate": 2.066115702479339e-07,
      "loss": 1.4112,
      "step": 5
    },
    {
      "epoch": 0.004142502071251036,
      "grad_norm": 14.988102359744419,
      "learning_rate": 4.132231404958678e-07,
      "loss": 1.3662,
      "step": 10
    },
    {
      "epoch": 0.006213753106876553,
      "grad_norm": 8.653198188938534,
      "learning_rate": 6.198347107438018e-07,
      "loss": 1.2641,
      "step": 15
    },
    {
      "epoch": 0.008285004142502071,
      "grad_norm": 10.616310133905728,
      "learning_rate": 8.264462809917356e-07,
      "loss": 1.1368,
      "step": 20
    },
    {
      "epoch": 0.010356255178127589,
      "grad_norm": 4.668322782816374,
      "learning_rate": 1.0330578512396695e-06,
      "loss": 1.0379,
      "step": 25
    },
    {
      "epoch": 0.012427506213753107,
      "grad_norm": 3.5182741608551535,
      "learning_rate": 1.2396694214876035e-06,
      "loss": 0.9937,
      "step": 30
    },
    {
      "epoch": 0.014498757249378625,
      "grad_norm": 3.3411660569699504,
      "learning_rate": 1.4462809917355372e-06,
      "loss": 0.9572,
      "step": 35
    },
    {
      "epoch": 0.016570008285004142,
      "grad_norm": 3.037596563067607,
      "learning_rate": 1.6528925619834712e-06,
      "loss": 0.9289,
      "step": 40
    },
    {
      "epoch": 0.018641259320629662,
      "grad_norm": 3.2528069002854374,
      "learning_rate": 1.859504132231405e-06,
      "loss": 0.9214,
      "step": 45
    },
    {
      "epoch": 0.020712510356255178,
      "grad_norm": 3.2107014690984803,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.9086,
      "step": 50
    },
    {
      "epoch": 0.022783761391880698,
      "grad_norm": 2.9859094433454554,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.8991,
      "step": 55
    },
    {
      "epoch": 0.024855012427506214,
      "grad_norm": 3.0590584148615627,
      "learning_rate": 2.479338842975207e-06,
      "loss": 0.8916,
      "step": 60
    },
    {
      "epoch": 0.026926263463131733,
      "grad_norm": 2.9914820848818446,
      "learning_rate": 2.6859504132231405e-06,
      "loss": 0.8785,
      "step": 65
    },
    {
      "epoch": 0.02899751449875725,
      "grad_norm": 2.9775144313765445,
      "learning_rate": 2.8925619834710743e-06,
      "loss": 0.8897,
      "step": 70
    },
    {
      "epoch": 0.03106876553438277,
      "grad_norm": 3.128569537433572,
      "learning_rate": 3.0991735537190086e-06,
      "loss": 0.8834,
      "step": 75
    },
    {
      "epoch": 0.033140016570008285,
      "grad_norm": 3.1281090024503757,
      "learning_rate": 3.3057851239669424e-06,
      "loss": 0.8651,
      "step": 80
    },
    {
      "epoch": 0.035211267605633804,
      "grad_norm": 3.0548507580726625,
      "learning_rate": 3.5123966942148763e-06,
      "loss": 0.8737,
      "step": 85
    },
    {
      "epoch": 0.037282518641259324,
      "grad_norm": 3.088551853774246,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.8623,
      "step": 90
    },
    {
      "epoch": 0.03935376967688484,
      "grad_norm": 3.117570527250618,
      "learning_rate": 3.925619834710744e-06,
      "loss": 0.838,
      "step": 95
    },
    {
      "epoch": 0.041425020712510356,
      "grad_norm": 3.248544738727649,
      "learning_rate": 4.132231404958678e-06,
      "loss": 0.8675,
      "step": 100
    },
    {
      "epoch": 0.043496271748135876,
      "grad_norm": 3.096171447698008,
      "learning_rate": 4.338842975206612e-06,
      "loss": 0.8394,
      "step": 105
    },
    {
      "epoch": 0.045567522783761395,
      "grad_norm": 3.233248587377124,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.857,
      "step": 110
    },
    {
      "epoch": 0.04763877381938691,
      "grad_norm": 3.6001636024321293,
      "learning_rate": 4.75206611570248e-06,
      "loss": 0.8321,
      "step": 115
    },
    {
      "epoch": 0.04971002485501243,
      "grad_norm": 2.9956895468112084,
      "learning_rate": 4.958677685950414e-06,
      "loss": 0.833,
      "step": 120
    },
    {
      "epoch": 0.05178127589063795,
      "grad_norm": 3.2710428266169953,
      "learning_rate": 5.165289256198347e-06,
      "loss": 0.847,
      "step": 125
    },
    {
      "epoch": 0.053852526926263466,
      "grad_norm": 3.249727950554207,
      "learning_rate": 5.371900826446281e-06,
      "loss": 0.8491,
      "step": 130
    },
    {
      "epoch": 0.05592377796188898,
      "grad_norm": 3.638097729325915,
      "learning_rate": 5.578512396694216e-06,
      "loss": 0.8218,
      "step": 135
    },
    {
      "epoch": 0.0579950289975145,
      "grad_norm": 3.136515122305162,
      "learning_rate": 5.785123966942149e-06,
      "loss": 0.8367,
      "step": 140
    },
    {
      "epoch": 0.06006628003314002,
      "grad_norm": 2.992975297054502,
      "learning_rate": 5.991735537190083e-06,
      "loss": 0.8331,
      "step": 145
    },
    {
      "epoch": 0.06213753106876554,
      "grad_norm": 3.1773736798479653,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.8159,
      "step": 150
    },
    {
      "epoch": 0.06420878210439106,
      "grad_norm": 3.2512013569633,
      "learning_rate": 6.404958677685951e-06,
      "loss": 0.8261,
      "step": 155
    },
    {
      "epoch": 0.06628003314001657,
      "grad_norm": 3.0823463887606852,
      "learning_rate": 6.611570247933885e-06,
      "loss": 0.8376,
      "step": 160
    },
    {
      "epoch": 0.06835128417564208,
      "grad_norm": 3.1208323742016897,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.8275,
      "step": 165
    },
    {
      "epoch": 0.07042253521126761,
      "grad_norm": 3.0745780302802332,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.8249,
      "step": 170
    },
    {
      "epoch": 0.07249378624689312,
      "grad_norm": 3.119480709134969,
      "learning_rate": 7.231404958677687e-06,
      "loss": 0.8042,
      "step": 175
    },
    {
      "epoch": 0.07456503728251865,
      "grad_norm": 3.3248978246402663,
      "learning_rate": 7.43801652892562e-06,
      "loss": 0.8284,
      "step": 180
    },
    {
      "epoch": 0.07663628831814416,
      "grad_norm": 3.101088437755532,
      "learning_rate": 7.644628099173555e-06,
      "loss": 0.8158,
      "step": 185
    },
    {
      "epoch": 0.07870753935376967,
      "grad_norm": 3.3761252638628707,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.8082,
      "step": 190
    },
    {
      "epoch": 0.0807787903893952,
      "grad_norm": 3.211138010347251,
      "learning_rate": 8.057851239669421e-06,
      "loss": 0.8066,
      "step": 195
    },
    {
      "epoch": 0.08285004142502071,
      "grad_norm": 3.390864723290171,
      "learning_rate": 8.264462809917356e-06,
      "loss": 0.8071,
      "step": 200
    },
    {
      "epoch": 0.08492129246064622,
      "grad_norm": 2.9884296341537775,
      "learning_rate": 8.47107438016529e-06,
      "loss": 0.8173,
      "step": 205
    },
    {
      "epoch": 0.08699254349627175,
      "grad_norm": 2.9929265463289774,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.8139,
      "step": 210
    },
    {
      "epoch": 0.08906379453189726,
      "grad_norm": 3.178365772310986,
      "learning_rate": 8.884297520661158e-06,
      "loss": 0.8077,
      "step": 215
    },
    {
      "epoch": 0.09113504556752279,
      "grad_norm": 3.0990381736081556,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.7962,
      "step": 220
    },
    {
      "epoch": 0.0932062966031483,
      "grad_norm": 3.3389255178168322,
      "learning_rate": 9.297520661157025e-06,
      "loss": 0.7985,
      "step": 225
    },
    {
      "epoch": 0.09527754763877382,
      "grad_norm": 3.3752236468584456,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.796,
      "step": 230
    },
    {
      "epoch": 0.09734879867439934,
      "grad_norm": 3.0456999521377877,
      "learning_rate": 9.710743801652894e-06,
      "loss": 0.7983,
      "step": 235
    },
    {
      "epoch": 0.09942004971002485,
      "grad_norm": 3.2120044737399187,
      "learning_rate": 9.917355371900828e-06,
      "loss": 0.8055,
      "step": 240
    },
    {
      "epoch": 0.10149130074565037,
      "grad_norm": 3.132342089203479,
      "learning_rate": 9.999952928077044e-06,
      "loss": 0.8021,
      "step": 245
    },
    {
      "epoch": 0.1035625517812759,
      "grad_norm": 2.9894402985924766,
      "learning_rate": 9.999665269535307e-06,
      "loss": 0.8013,
      "step": 250
    },
    {
      "epoch": 0.1056338028169014,
      "grad_norm": 3.046903074957559,
      "learning_rate": 9.99911611854702e-06,
      "loss": 0.7979,
      "step": 255
    },
    {
      "epoch": 0.10770505385252693,
      "grad_norm": 2.8665185067493626,
      "learning_rate": 9.998305503833872e-06,
      "loss": 0.8073,
      "step": 260
    },
    {
      "epoch": 0.10977630488815245,
      "grad_norm": 3.2409413642928278,
      "learning_rate": 9.997233467792626e-06,
      "loss": 0.7976,
      "step": 265
    },
    {
      "epoch": 0.11184755592377796,
      "grad_norm": 2.944077687828054,
      "learning_rate": 9.995900066492902e-06,
      "loss": 0.8016,
      "step": 270
    },
    {
      "epoch": 0.11391880695940348,
      "grad_norm": 2.9912116980156145,
      "learning_rate": 9.994305369674242e-06,
      "loss": 0.7853,
      "step": 275
    },
    {
      "epoch": 0.115990057995029,
      "grad_norm": 2.892932100424063,
      "learning_rate": 9.992449460742464e-06,
      "loss": 0.8046,
      "step": 280
    },
    {
      "epoch": 0.11806130903065451,
      "grad_norm": 3.055501054823764,
      "learning_rate": 9.9903324367653e-06,
      "loss": 0.7794,
      "step": 285
    },
    {
      "epoch": 0.12013256006628004,
      "grad_norm": 2.7659846238569052,
      "learning_rate": 9.98795440846732e-06,
      "loss": 0.7814,
      "step": 290
    },
    {
      "epoch": 0.12220381110190555,
      "grad_norm": 3.0121040906813126,
      "learning_rate": 9.985315500224135e-06,
      "loss": 0.7809,
      "step": 295
    },
    {
      "epoch": 0.12427506213753108,
      "grad_norm": 3.01982583757692,
      "learning_rate": 9.982415850055902e-06,
      "loss": 0.781,
      "step": 300
    },
    {
      "epoch": 0.1263463131731566,
      "grad_norm": 2.716394804747616,
      "learning_rate": 9.979255609620095e-06,
      "loss": 0.7734,
      "step": 305
    },
    {
      "epoch": 0.12841756420878211,
      "grad_norm": 2.690802383393494,
      "learning_rate": 9.975834944203581e-06,
      "loss": 0.7667,
      "step": 310
    },
    {
      "epoch": 0.1304888152444076,
      "grad_norm": 2.9488089043416124,
      "learning_rate": 9.972154032713973e-06,
      "loss": 0.7805,
      "step": 315
    },
    {
      "epoch": 0.13256006628003314,
      "grad_norm": 2.719266979190075,
      "learning_rate": 9.968213067670265e-06,
      "loss": 0.7763,
      "step": 320
    },
    {
      "epoch": 0.13463131731565867,
      "grad_norm": 2.812329679333236,
      "learning_rate": 9.964012255192776e-06,
      "loss": 0.7533,
      "step": 325
    },
    {
      "epoch": 0.13670256835128416,
      "grad_norm": 2.939179450726379,
      "learning_rate": 9.959551814992364e-06,
      "loss": 0.7538,
      "step": 330
    },
    {
      "epoch": 0.1387738193869097,
      "grad_norm": 3.0343927565408566,
      "learning_rate": 9.954831980358928e-06,
      "loss": 0.7761,
      "step": 335
    },
    {
      "epoch": 0.14084507042253522,
      "grad_norm": 3.051369274408743,
      "learning_rate": 9.949852998149217e-06,
      "loss": 0.7623,
      "step": 340
    },
    {
      "epoch": 0.14291632145816072,
      "grad_norm": 2.8926023239726666,
      "learning_rate": 9.944615128773911e-06,
      "loss": 0.7449,
      "step": 345
    },
    {
      "epoch": 0.14498757249378624,
      "grad_norm": 3.0274876780182725,
      "learning_rate": 9.939118646184007e-06,
      "loss": 0.7564,
      "step": 350
    },
    {
      "epoch": 0.14705882352941177,
      "grad_norm": 3.030086237395326,
      "learning_rate": 9.933363837856485e-06,
      "loss": 0.7636,
      "step": 355
    },
    {
      "epoch": 0.1491300745650373,
      "grad_norm": 2.7216718925144443,
      "learning_rate": 9.927351004779275e-06,
      "loss": 0.7523,
      "step": 360
    },
    {
      "epoch": 0.1512013256006628,
      "grad_norm": 2.8963011471510423,
      "learning_rate": 9.921080461435522e-06,
      "loss": 0.747,
      "step": 365
    },
    {
      "epoch": 0.15327257663628832,
      "grad_norm": 2.908273407418608,
      "learning_rate": 9.914552535787122e-06,
      "loss": 0.7325,
      "step": 370
    },
    {
      "epoch": 0.15534382767191385,
      "grad_norm": 2.8676437077981722,
      "learning_rate": 9.90776756925758e-06,
      "loss": 0.7349,
      "step": 375
    },
    {
      "epoch": 0.15741507870753935,
      "grad_norm": 2.8556939586506687,
      "learning_rate": 9.900725916714157e-06,
      "loss": 0.7524,
      "step": 380
    },
    {
      "epoch": 0.15948632974316487,
      "grad_norm": 2.87186116366176,
      "learning_rate": 9.893427946449297e-06,
      "loss": 0.7214,
      "step": 385
    },
    {
      "epoch": 0.1615575807787904,
      "grad_norm": 2.8757200745024485,
      "learning_rate": 9.885874040161373e-06,
      "loss": 0.7326,
      "step": 390
    },
    {
      "epoch": 0.1636288318144159,
      "grad_norm": 3.215961320189587,
      "learning_rate": 9.878064592934723e-06,
      "loss": 0.7227,
      "step": 395
    },
    {
      "epoch": 0.16570008285004142,
      "grad_norm": 2.9059437980483693,
      "learning_rate": 9.87000001321898e-06,
      "loss": 0.7509,
      "step": 400
    },
    {
      "epoch": 0.16777133388566695,
      "grad_norm": 2.685626883478994,
      "learning_rate": 9.86168072280772e-06,
      "loss": 0.7223,
      "step": 405
    },
    {
      "epoch": 0.16984258492129245,
      "grad_norm": 3.033273475512452,
      "learning_rate": 9.853107156816393e-06,
      "loss": 0.7385,
      "step": 410
    },
    {
      "epoch": 0.17191383595691798,
      "grad_norm": 2.7784947677053955,
      "learning_rate": 9.844279763659566e-06,
      "loss": 0.7237,
      "step": 415
    },
    {
      "epoch": 0.1739850869925435,
      "grad_norm": 2.9117670580445543,
      "learning_rate": 9.835199005027477e-06,
      "loss": 0.7161,
      "step": 420
    },
    {
      "epoch": 0.176056338028169,
      "grad_norm": 2.8843325957862764,
      "learning_rate": 9.825865355861878e-06,
      "loss": 0.7256,
      "step": 425
    },
    {
      "epoch": 0.17812758906379453,
      "grad_norm": 2.856399843872642,
      "learning_rate": 9.816279304331202e-06,
      "loss": 0.7142,
      "step": 430
    },
    {
      "epoch": 0.18019884009942005,
      "grad_norm": 2.9556721265181163,
      "learning_rate": 9.806441351805025e-06,
      "loss": 0.7306,
      "step": 435
    },
    {
      "epoch": 0.18227009113504558,
      "grad_norm": 2.9810648069102172,
      "learning_rate": 9.79635201282785e-06,
      "loss": 0.705,
      "step": 440
    },
    {
      "epoch": 0.18434134217067108,
      "grad_norm": 2.650650945302262,
      "learning_rate": 9.786011815092193e-06,
      "loss": 0.7026,
      "step": 445
    },
    {
      "epoch": 0.1864125932062966,
      "grad_norm": 2.7614731795144727,
      "learning_rate": 9.775421299410977e-06,
      "loss": 0.6993,
      "step": 450
    },
    {
      "epoch": 0.18848384424192213,
      "grad_norm": 2.8508808311574487,
      "learning_rate": 9.764581019689255e-06,
      "loss": 0.707,
      "step": 455
    },
    {
      "epoch": 0.19055509527754763,
      "grad_norm": 3.1150961713118646,
      "learning_rate": 9.753491542895237e-06,
      "loss": 0.7186,
      "step": 460
    },
    {
      "epoch": 0.19262634631317316,
      "grad_norm": 3.0323106493041325,
      "learning_rate": 9.742153449030639e-06,
      "loss": 0.6893,
      "step": 465
    },
    {
      "epoch": 0.19469759734879868,
      "grad_norm": 3.0499193466413654,
      "learning_rate": 9.730567331100333e-06,
      "loss": 0.7106,
      "step": 470
    },
    {
      "epoch": 0.19676884838442418,
      "grad_norm": 2.8549221082922,
      "learning_rate": 9.71873379508136e-06,
      "loss": 0.6801,
      "step": 475
    },
    {
      "epoch": 0.1988400994200497,
      "grad_norm": 2.8978651489251703,
      "learning_rate": 9.706653459891207e-06,
      "loss": 0.6906,
      "step": 480
    },
    {
      "epoch": 0.20091135045567524,
      "grad_norm": 2.7131289500779237,
      "learning_rate": 9.694326957355452e-06,
      "loss": 0.6793,
      "step": 485
    },
    {
      "epoch": 0.20298260149130073,
      "grad_norm": 2.7450048929619033,
      "learning_rate": 9.681754932174719e-06,
      "loss": 0.6987,
      "step": 490
    },
    {
      "epoch": 0.20505385252692626,
      "grad_norm": 2.752895531078258,
      "learning_rate": 9.668938041890952e-06,
      "loss": 0.693,
      "step": 495
    },
    {
      "epoch": 0.2071251035625518,
      "grad_norm": 2.808137119286013,
      "learning_rate": 9.655876956853025e-06,
      "loss": 0.681,
      "step": 500
    },
    {
      "epoch": 0.2091963545981773,
      "grad_norm": 2.724476444787799,
      "learning_rate": 9.64257236018169e-06,
      "loss": 0.6716,
      "step": 505
    },
    {
      "epoch": 0.2112676056338028,
      "grad_norm": 2.748867424602178,
      "learning_rate": 9.629024947733836e-06,
      "loss": 0.6944,
      "step": 510
    },
    {
      "epoch": 0.21333885666942834,
      "grad_norm": 2.9490226550436676,
      "learning_rate": 9.615235428066106e-06,
      "loss": 0.6844,
      "step": 515
    },
    {
      "epoch": 0.21541010770505387,
      "grad_norm": 2.6625828983237634,
      "learning_rate": 9.601204522397826e-06,
      "loss": 0.6876,
      "step": 520
    },
    {
      "epoch": 0.21748135874067936,
      "grad_norm": 2.8580713323817672,
      "learning_rate": 9.586932964573298e-06,
      "loss": 0.6675,
      "step": 525
    },
    {
      "epoch": 0.2195526097763049,
      "grad_norm": 2.65085890829761,
      "learning_rate": 9.572421501023403e-06,
      "loss": 0.6871,
      "step": 530
    },
    {
      "epoch": 0.22162386081193042,
      "grad_norm": 2.8073591225727066,
      "learning_rate": 9.557670890726576e-06,
      "loss": 0.6806,
      "step": 535
    },
    {
      "epoch": 0.22369511184755592,
      "grad_norm": 2.890518621456693,
      "learning_rate": 9.5426819051691e-06,
      "loss": 0.6842,
      "step": 540
    },
    {
      "epoch": 0.22576636288318144,
      "grad_norm": 3.037365740542274,
      "learning_rate": 9.527455328304756e-06,
      "loss": 0.6706,
      "step": 545
    },
    {
      "epoch": 0.22783761391880697,
      "grad_norm": 3.007466890807781,
      "learning_rate": 9.511991956513828e-06,
      "loss": 0.6768,
      "step": 550
    },
    {
      "epoch": 0.22990886495443247,
      "grad_norm": 2.991724717996849,
      "learning_rate": 9.496292598561445e-06,
      "loss": 0.6399,
      "step": 555
    },
    {
      "epoch": 0.231980115990058,
      "grad_norm": 2.6643291004806144,
      "learning_rate": 9.480358075555278e-06,
      "loss": 0.6745,
      "step": 560
    },
    {
      "epoch": 0.23405136702568352,
      "grad_norm": 2.7494300595409076,
      "learning_rate": 9.464189220902603e-06,
      "loss": 0.6659,
      "step": 565
    },
    {
      "epoch": 0.23612261806130902,
      "grad_norm": 2.778050795482066,
      "learning_rate": 9.447786880266706e-06,
      "loss": 0.6682,
      "step": 570
    },
    {
      "epoch": 0.23819386909693455,
      "grad_norm": 2.8808954872434107,
      "learning_rate": 9.431151911522656e-06,
      "loss": 0.6603,
      "step": 575
    },
    {
      "epoch": 0.24026512013256007,
      "grad_norm": 2.7631310193416008,
      "learning_rate": 9.414285184712432e-06,
      "loss": 0.6479,
      "step": 580
    },
    {
      "epoch": 0.24233637116818557,
      "grad_norm": 2.7504091304269505,
      "learning_rate": 9.397187581999424e-06,
      "loss": 0.656,
      "step": 585
    },
    {
      "epoch": 0.2444076222038111,
      "grad_norm": 2.7704909990609345,
      "learning_rate": 9.37985999762229e-06,
      "loss": 0.6494,
      "step": 590
    },
    {
      "epoch": 0.24647887323943662,
      "grad_norm": 2.7880341771333086,
      "learning_rate": 9.362303337848188e-06,
      "loss": 0.6498,
      "step": 595
    },
    {
      "epoch": 0.24855012427506215,
      "grad_norm": 2.7372277502845064,
      "learning_rate": 9.344518520925377e-06,
      "loss": 0.635,
      "step": 600
    },
    {
      "epoch": 0.2506213753106877,
      "grad_norm": 2.700220180234652,
      "learning_rate": 9.326506477035179e-06,
      "loss": 0.6205,
      "step": 605
    },
    {
      "epoch": 0.2526926263463132,
      "grad_norm": 2.9446333047323714,
      "learning_rate": 9.308268148243355e-06,
      "loss": 0.6377,
      "step": 610
    },
    {
      "epoch": 0.2547638773819387,
      "grad_norm": 2.851475449493436,
      "learning_rate": 9.289804488450805e-06,
      "loss": 0.6395,
      "step": 615
    },
    {
      "epoch": 0.25683512841756423,
      "grad_norm": 2.868033711564291,
      "learning_rate": 9.271116463343692e-06,
      "loss": 0.6421,
      "step": 620
    },
    {
      "epoch": 0.25890637945318973,
      "grad_norm": 2.6514436144580307,
      "learning_rate": 9.25220505034293e-06,
      "loss": 0.6372,
      "step": 625
    },
    {
      "epoch": 0.2609776304888152,
      "grad_norm": 2.9338066764643598,
      "learning_rate": 9.23307123855307e-06,
      "loss": 0.636,
      "step": 630
    },
    {
      "epoch": 0.2630488815244408,
      "grad_norm": 3.023298105516956,
      "learning_rate": 9.213716028710558e-06,
      "loss": 0.6103,
      "step": 635
    },
    {
      "epoch": 0.2651201325600663,
      "grad_norm": 2.871190920425405,
      "learning_rate": 9.194140433131397e-06,
      "loss": 0.6159,
      "step": 640
    },
    {
      "epoch": 0.2671913835956918,
      "grad_norm": 2.7631786024911187,
      "learning_rate": 9.174345475658208e-06,
      "loss": 0.6228,
      "step": 645
    },
    {
      "epoch": 0.26926263463131733,
      "grad_norm": 2.998454912972155,
      "learning_rate": 9.154332191606671e-06,
      "loss": 0.6335,
      "step": 650
    },
    {
      "epoch": 0.27133388566694283,
      "grad_norm": 2.803386942098811,
      "learning_rate": 9.134101627711384e-06,
      "loss": 0.6119,
      "step": 655
    },
    {
      "epoch": 0.27340513670256833,
      "grad_norm": 2.6219329149002277,
      "learning_rate": 9.113654842071114e-06,
      "loss": 0.6171,
      "step": 660
    },
    {
      "epoch": 0.2754763877381939,
      "grad_norm": 2.61515633927928,
      "learning_rate": 9.092992904093451e-06,
      "loss": 0.6165,
      "step": 665
    },
    {
      "epoch": 0.2775476387738194,
      "grad_norm": 2.8142275849395473,
      "learning_rate": 9.072116894438885e-06,
      "loss": 0.6091,
      "step": 670
    },
    {
      "epoch": 0.2796188898094449,
      "grad_norm": 2.8036261300008856,
      "learning_rate": 9.051027904964279e-06,
      "loss": 0.6266,
      "step": 675
    },
    {
      "epoch": 0.28169014084507044,
      "grad_norm": 2.5728545238650122,
      "learning_rate": 9.029727038665765e-06,
      "loss": 0.625,
      "step": 680
    },
    {
      "epoch": 0.28376139188069593,
      "grad_norm": 2.911382813259161,
      "learning_rate": 9.008215409621053e-06,
      "loss": 0.6023,
      "step": 685
    },
    {
      "epoch": 0.28583264291632143,
      "grad_norm": 2.8794319732821956,
      "learning_rate": 8.986494142931168e-06,
      "loss": 0.6113,
      "step": 690
    },
    {
      "epoch": 0.287903893951947,
      "grad_norm": 2.6994578026382956,
      "learning_rate": 8.964564374661597e-06,
      "loss": 0.6083,
      "step": 695
    },
    {
      "epoch": 0.2899751449875725,
      "grad_norm": 2.901823826406794,
      "learning_rate": 8.94242725178288e-06,
      "loss": 0.616,
      "step": 700
    },
    {
      "epoch": 0.292046396023198,
      "grad_norm": 2.653196617958314,
      "learning_rate": 8.920083932110608e-06,
      "loss": 0.6196,
      "step": 705
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 2.6749585402475264,
      "learning_rate": 8.89753558424488e-06,
      "loss": 0.597,
      "step": 710
    },
    {
      "epoch": 0.29618889809444904,
      "grad_norm": 2.7369834176016723,
      "learning_rate": 8.874783387509181e-06,
      "loss": 0.6012,
      "step": 715
    },
    {
      "epoch": 0.2982601491300746,
      "grad_norm": 2.845681780523314,
      "learning_rate": 8.851828531888692e-06,
      "loss": 0.6223,
      "step": 720
    },
    {
      "epoch": 0.3003314001657001,
      "grad_norm": 2.7849231768312883,
      "learning_rate": 8.828672217968055e-06,
      "loss": 0.6192,
      "step": 725
    },
    {
      "epoch": 0.3024026512013256,
      "grad_norm": 2.7476355382279385,
      "learning_rate": 8.805315656868587e-06,
      "loss": 0.6053,
      "step": 730
    },
    {
      "epoch": 0.30447390223695114,
      "grad_norm": 2.686604793998387,
      "learning_rate": 8.781760070184933e-06,
      "loss": 0.5934,
      "step": 735
    },
    {
      "epoch": 0.30654515327257664,
      "grad_norm": 2.997703713470155,
      "learning_rate": 8.75800668992117e-06,
      "loss": 0.5956,
      "step": 740
    },
    {
      "epoch": 0.30861640430820214,
      "grad_norm": 2.621004060213254,
      "learning_rate": 8.734056758426367e-06,
      "loss": 0.5974,
      "step": 745
    },
    {
      "epoch": 0.3106876553438277,
      "grad_norm": 2.6549364932267303,
      "learning_rate": 8.709911528329623e-06,
      "loss": 0.5948,
      "step": 750
    },
    {
      "epoch": 0.3127589063794532,
      "grad_norm": 2.7965881278503435,
      "learning_rate": 8.685572262474538e-06,
      "loss": 0.5967,
      "step": 755
    },
    {
      "epoch": 0.3148301574150787,
      "grad_norm": 2.5906530552932185,
      "learning_rate": 8.661040233853166e-06,
      "loss": 0.5824,
      "step": 760
    },
    {
      "epoch": 0.31690140845070425,
      "grad_norm": 2.96203173026875,
      "learning_rate": 8.636316725539445e-06,
      "loss": 0.5833,
      "step": 765
    },
    {
      "epoch": 0.31897265948632975,
      "grad_norm": 2.6456839069829967,
      "learning_rate": 8.611403030622074e-06,
      "loss": 0.5913,
      "step": 770
    },
    {
      "epoch": 0.32104391052195524,
      "grad_norm": 2.552077771616609,
      "learning_rate": 8.586300452136895e-06,
      "loss": 0.5604,
      "step": 775
    },
    {
      "epoch": 0.3231151615575808,
      "grad_norm": 2.681514794026766,
      "learning_rate": 8.561010302998734e-06,
      "loss": 0.5878,
      "step": 780
    },
    {
      "epoch": 0.3251864125932063,
      "grad_norm": 2.810277058352099,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.5904,
      "step": 785
    },
    {
      "epoch": 0.3272576636288318,
      "grad_norm": 2.823366500468815,
      "learning_rate": 8.509872593405189e-06,
      "loss": 0.5785,
      "step": 790
    },
    {
      "epoch": 0.32932891466445735,
      "grad_norm": 2.537302794266039,
      "learning_rate": 8.484027707553818e-06,
      "loss": 0.5749,
      "step": 795
    },
    {
      "epoch": 0.33140016570008285,
      "grad_norm": 2.679914960811432,
      "learning_rate": 8.458000600117604e-06,
      "loss": 0.5758,
      "step": 800
    },
    {
      "epoch": 0.33347141673570835,
      "grad_norm": 2.714722412588362,
      "learning_rate": 8.43179263236608e-06,
      "loss": 0.5703,
      "step": 805
    },
    {
      "epoch": 0.3355426677713339,
      "grad_norm": 2.8825027456443593,
      "learning_rate": 8.40540517502813e-06,
      "loss": 0.5741,
      "step": 810
    },
    {
      "epoch": 0.3376139188069594,
      "grad_norm": 2.69779454090861,
      "learning_rate": 8.378839608220304e-06,
      "loss": 0.5758,
      "step": 815
    },
    {
      "epoch": 0.3396851698425849,
      "grad_norm": 2.591195744363521,
      "learning_rate": 8.35209732137463e-06,
      "loss": 0.5703,
      "step": 820
    },
    {
      "epoch": 0.34175642087821045,
      "grad_norm": 2.5928484463176056,
      "learning_rate": 8.32517971316595e-06,
      "loss": 0.5789,
      "step": 825
    },
    {
      "epoch": 0.34382767191383595,
      "grad_norm": 2.8392376730929962,
      "learning_rate": 8.298088191438753e-06,
      "loss": 0.5573,
      "step": 830
    },
    {
      "epoch": 0.34589892294946145,
      "grad_norm": 2.717204690264741,
      "learning_rate": 8.270824173133563e-06,
      "loss": 0.5553,
      "step": 835
    },
    {
      "epoch": 0.347970173985087,
      "grad_norm": 2.6426626328945124,
      "learning_rate": 8.243389084212808e-06,
      "loss": 0.5503,
      "step": 840
    },
    {
      "epoch": 0.3500414250207125,
      "grad_norm": 2.6361988946032806,
      "learning_rate": 8.215784359586257e-06,
      "loss": 0.5498,
      "step": 845
    },
    {
      "epoch": 0.352112676056338,
      "grad_norm": 2.452756091881905,
      "learning_rate": 8.188011443035962e-06,
      "loss": 0.538,
      "step": 850
    },
    {
      "epoch": 0.35418392709196356,
      "grad_norm": 2.740977671213533,
      "learning_rate": 8.160071787140742e-06,
      "loss": 0.5452,
      "step": 855
    },
    {
      "epoch": 0.35625517812758906,
      "grad_norm": 2.650236162364244,
      "learning_rate": 8.131966853200226e-06,
      "loss": 0.5485,
      "step": 860
    },
    {
      "epoch": 0.35832642916321455,
      "grad_norm": 2.701846208365303,
      "learning_rate": 8.103698111158405e-06,
      "loss": 0.5495,
      "step": 865
    },
    {
      "epoch": 0.3603976801988401,
      "grad_norm": 2.7179474541169104,
      "learning_rate": 8.075267039526764e-06,
      "loss": 0.548,
      "step": 870
    },
    {
      "epoch": 0.3624689312344656,
      "grad_norm": 2.833260463644198,
      "learning_rate": 8.046675125306948e-06,
      "loss": 0.5632,
      "step": 875
    },
    {
      "epoch": 0.36454018227009116,
      "grad_norm": 2.7066133568619724,
      "learning_rate": 8.017923863912989e-06,
      "loss": 0.5475,
      "step": 880
    },
    {
      "epoch": 0.36661143330571666,
      "grad_norm": 2.554105810407388,
      "learning_rate": 7.989014759093095e-06,
      "loss": 0.5455,
      "step": 885
    },
    {
      "epoch": 0.36868268434134216,
      "grad_norm": 2.530083794168148,
      "learning_rate": 7.959949322850994e-06,
      "loss": 0.5397,
      "step": 890
    },
    {
      "epoch": 0.3707539353769677,
      "grad_norm": 2.6016133268460937,
      "learning_rate": 7.930729075366867e-06,
      "loss": 0.542,
      "step": 895
    },
    {
      "epoch": 0.3728251864125932,
      "grad_norm": 2.5754320630229754,
      "learning_rate": 7.901355544917827e-06,
      "loss": 0.5362,
      "step": 900
    },
    {
      "epoch": 0.3748964374482187,
      "grad_norm": 2.576938951435503,
      "learning_rate": 7.87183026779799e-06,
      "loss": 0.5414,
      "step": 905
    },
    {
      "epoch": 0.37696768848384427,
      "grad_norm": 2.6785725784997796,
      "learning_rate": 7.842154788238124e-06,
      "loss": 0.5512,
      "step": 910
    },
    {
      "epoch": 0.37903893951946976,
      "grad_norm": 2.6795394480092414,
      "learning_rate": 7.812330658324884e-06,
      "loss": 0.5375,
      "step": 915
    },
    {
      "epoch": 0.38111019055509526,
      "grad_norm": 2.672382602041246,
      "learning_rate": 7.782359437919644e-06,
      "loss": 0.5457,
      "step": 920
    },
    {
      "epoch": 0.3831814415907208,
      "grad_norm": 2.884729781500387,
      "learning_rate": 7.75224269457689e-06,
      "loss": 0.5239,
      "step": 925
    },
    {
      "epoch": 0.3852526926263463,
      "grad_norm": 2.56507336027056,
      "learning_rate": 7.721982003462255e-06,
      "loss": 0.5405,
      "step": 930
    },
    {
      "epoch": 0.3873239436619718,
      "grad_norm": 2.5473419393915,
      "learning_rate": 7.691578947270122e-06,
      "loss": 0.5438,
      "step": 935
    },
    {
      "epoch": 0.38939519469759737,
      "grad_norm": 2.783657224970263,
      "learning_rate": 7.661035116140856e-06,
      "loss": 0.5198,
      "step": 940
    },
    {
      "epoch": 0.39146644573322287,
      "grad_norm": 2.6274430429540927,
      "learning_rate": 7.63035210757763e-06,
      "loss": 0.5535,
      "step": 945
    },
    {
      "epoch": 0.39353769676884837,
      "grad_norm": 2.7442552412406958,
      "learning_rate": 7.599531526362873e-06,
      "loss": 0.5304,
      "step": 950
    },
    {
      "epoch": 0.3956089478044739,
      "grad_norm": 2.7785345881187786,
      "learning_rate": 7.568574984474335e-06,
      "loss": 0.5225,
      "step": 955
    },
    {
      "epoch": 0.3976801988400994,
      "grad_norm": 2.720728749217881,
      "learning_rate": 7.537484101000787e-06,
      "loss": 0.5217,
      "step": 960
    },
    {
      "epoch": 0.3997514498757249,
      "grad_norm": 2.637160358823368,
      "learning_rate": 7.506260502057325e-06,
      "loss": 0.5297,
      "step": 965
    },
    {
      "epoch": 0.40182270091135047,
      "grad_norm": 2.6335954311355216,
      "learning_rate": 7.474905820700334e-06,
      "loss": 0.5291,
      "step": 970
    },
    {
      "epoch": 0.40389395194697597,
      "grad_norm": 2.656261545637842,
      "learning_rate": 7.443421696842066e-06,
      "loss": 0.5076,
      "step": 975
    },
    {
      "epoch": 0.40596520298260147,
      "grad_norm": 2.6508585415601362,
      "learning_rate": 7.411809777164873e-06,
      "loss": 0.5183,
      "step": 980
    },
    {
      "epoch": 0.408036454018227,
      "grad_norm": 2.665237164150542,
      "learning_rate": 7.380071715035089e-06,
      "loss": 0.5241,
      "step": 985
    },
    {
      "epoch": 0.4101077050538525,
      "grad_norm": 2.5770691509075565,
      "learning_rate": 7.3482091704165405e-06,
      "loss": 0.5164,
      "step": 990
    },
    {
      "epoch": 0.412178956089478,
      "grad_norm": 2.5586140708653584,
      "learning_rate": 7.316223809783745e-06,
      "loss": 0.4982,
      "step": 995
    },
    {
      "epoch": 0.4142502071251036,
      "grad_norm": 2.6304400469333697,
      "learning_rate": 7.284117306034733e-06,
      "loss": 0.514,
      "step": 1000
    },
    {
      "epoch": 0.4163214581607291,
      "grad_norm": 2.5369318681193667,
      "learning_rate": 7.2518913384035685e-06,
      "loss": 0.5119,
      "step": 1005
    },
    {
      "epoch": 0.4183927091963546,
      "grad_norm": 2.7061992404592683,
      "learning_rate": 7.219547592372512e-06,
      "loss": 0.5175,
      "step": 1010
    },
    {
      "epoch": 0.4204639602319801,
      "grad_norm": 2.420325437701678,
      "learning_rate": 7.187087759583869e-06,
      "loss": 0.5063,
      "step": 1015
    },
    {
      "epoch": 0.4225352112676056,
      "grad_norm": 2.7148650041029514,
      "learning_rate": 7.15451353775151e-06,
      "loss": 0.5237,
      "step": 1020
    },
    {
      "epoch": 0.4246064623032311,
      "grad_norm": 2.654478692235913,
      "learning_rate": 7.121826630572084e-06,
      "loss": 0.4966,
      "step": 1025
    },
    {
      "epoch": 0.4266777133388567,
      "grad_norm": 2.6379021813183354,
      "learning_rate": 7.089028747635908e-06,
      "loss": 0.4938,
      "step": 1030
    },
    {
      "epoch": 0.4287489643744822,
      "grad_norm": 2.713654107850754,
      "learning_rate": 7.056121604337554e-06,
      "loss": 0.5074,
      "step": 1035
    },
    {
      "epoch": 0.43082021541010773,
      "grad_norm": 2.540085709590772,
      "learning_rate": 7.023106921786118e-06,
      "loss": 0.4995,
      "step": 1040
    },
    {
      "epoch": 0.43289146644573323,
      "grad_norm": 2.5985604745407374,
      "learning_rate": 6.9899864267152275e-06,
      "loss": 0.5318,
      "step": 1045
    },
    {
      "epoch": 0.43496271748135873,
      "grad_norm": 2.4411457867844417,
      "learning_rate": 6.956761851392706e-06,
      "loss": 0.4858,
      "step": 1050
    },
    {
      "epoch": 0.4370339685169843,
      "grad_norm": 2.4714643955598072,
      "learning_rate": 6.9234349335299835e-06,
      "loss": 0.5171,
      "step": 1055
    },
    {
      "epoch": 0.4391052195526098,
      "grad_norm": 2.4510153141350868,
      "learning_rate": 6.890007416191209e-06,
      "loss": 0.4875,
      "step": 1060
    },
    {
      "epoch": 0.4411764705882353,
      "grad_norm": 2.502867427304289,
      "learning_rate": 6.8564810477020835e-06,
      "loss": 0.4982,
      "step": 1065
    },
    {
      "epoch": 0.44324772162386084,
      "grad_norm": 2.5224619340497183,
      "learning_rate": 6.822857581558423e-06,
      "loss": 0.4971,
      "step": 1070
    },
    {
      "epoch": 0.44531897265948633,
      "grad_norm": 2.622904137029192,
      "learning_rate": 6.789138776334441e-06,
      "loss": 0.5063,
      "step": 1075
    },
    {
      "epoch": 0.44739022369511183,
      "grad_norm": 2.4481597241493174,
      "learning_rate": 6.7553263955907755e-06,
      "loss": 0.4841,
      "step": 1080
    },
    {
      "epoch": 0.4494614747307374,
      "grad_norm": 2.505241922300399,
      "learning_rate": 6.721422207782249e-06,
      "loss": 0.486,
      "step": 1085
    },
    {
      "epoch": 0.4515327257663629,
      "grad_norm": 2.51038092005137,
      "learning_rate": 6.687427986165379e-06,
      "loss": 0.4866,
      "step": 1090
    },
    {
      "epoch": 0.4536039768019884,
      "grad_norm": 2.618228992641671,
      "learning_rate": 6.653345508705629e-06,
      "loss": 0.4889,
      "step": 1095
    },
    {
      "epoch": 0.45567522783761394,
      "grad_norm": 2.6041268512577873,
      "learning_rate": 6.6191765579844205e-06,
      "loss": 0.4957,
      "step": 1100
    },
    {
      "epoch": 0.45774647887323944,
      "grad_norm": 2.5703382132992414,
      "learning_rate": 6.584922921105894e-06,
      "loss": 0.4805,
      "step": 1105
    },
    {
      "epoch": 0.45981772990886494,
      "grad_norm": 2.4810077168179894,
      "learning_rate": 6.550586389603451e-06,
      "loss": 0.4843,
      "step": 1110
    },
    {
      "epoch": 0.4618889809444905,
      "grad_norm": 2.7146694842288492,
      "learning_rate": 6.5161687593460395e-06,
      "loss": 0.4805,
      "step": 1115
    },
    {
      "epoch": 0.463960231980116,
      "grad_norm": 2.661840488355652,
      "learning_rate": 6.481671830444243e-06,
      "loss": 0.4766,
      "step": 1120
    },
    {
      "epoch": 0.4660314830157415,
      "grad_norm": 2.542982024246636,
      "learning_rate": 6.447097407156114e-06,
      "loss": 0.4714,
      "step": 1125
    },
    {
      "epoch": 0.46810273405136704,
      "grad_norm": 2.545352453639695,
      "learning_rate": 6.412447297792818e-06,
      "loss": 0.4627,
      "step": 1130
    },
    {
      "epoch": 0.47017398508699254,
      "grad_norm": 2.601762512884624,
      "learning_rate": 6.377723314624057e-06,
      "loss": 0.4617,
      "step": 1135
    },
    {
      "epoch": 0.47224523612261804,
      "grad_norm": 2.5271935902650746,
      "learning_rate": 6.3429272737832726e-06,
      "loss": 0.482,
      "step": 1140
    },
    {
      "epoch": 0.4743164871582436,
      "grad_norm": 2.8257956277427265,
      "learning_rate": 6.308060995172673e-06,
      "loss": 0.4875,
      "step": 1145
    },
    {
      "epoch": 0.4763877381938691,
      "grad_norm": 2.6203637102255843,
      "learning_rate": 6.273126302368037e-06,
      "loss": 0.4725,
      "step": 1150
    },
    {
      "epoch": 0.4784589892294946,
      "grad_norm": 2.4615255657443855,
      "learning_rate": 6.238125022523343e-06,
      "loss": 0.462,
      "step": 1155
    },
    {
      "epoch": 0.48053024026512015,
      "grad_norm": 2.399337323008823,
      "learning_rate": 6.203058986275207e-06,
      "loss": 0.4706,
      "step": 1160
    },
    {
      "epoch": 0.48260149130074564,
      "grad_norm": 2.568546498877998,
      "learning_rate": 6.1679300276471285e-06,
      "loss": 0.46,
      "step": 1165
    },
    {
      "epoch": 0.48467274233637114,
      "grad_norm": 2.5774193338530886,
      "learning_rate": 6.132739983953579e-06,
      "loss": 0.4639,
      "step": 1170
    },
    {
      "epoch": 0.4867439933719967,
      "grad_norm": 2.433169814085788,
      "learning_rate": 6.097490695703896e-06,
      "loss": 0.4706,
      "step": 1175
    },
    {
      "epoch": 0.4888152444076222,
      "grad_norm": 2.5672509798739975,
      "learning_rate": 6.062184006506027e-06,
      "loss": 0.469,
      "step": 1180
    },
    {
      "epoch": 0.4908864954432477,
      "grad_norm": 2.566412335613826,
      "learning_rate": 6.026821762970102e-06,
      "loss": 0.4667,
      "step": 1185
    },
    {
      "epoch": 0.49295774647887325,
      "grad_norm": 2.4468040851684885,
      "learning_rate": 5.991405814611855e-06,
      "loss": 0.4547,
      "step": 1190
    },
    {
      "epoch": 0.49502899751449875,
      "grad_norm": 2.490658198568155,
      "learning_rate": 5.955938013755888e-06,
      "loss": 0.4628,
      "step": 1195
    },
    {
      "epoch": 0.4971002485501243,
      "grad_norm": 2.463293970038353,
      "learning_rate": 5.920420215438794e-06,
      "loss": 0.4804,
      "step": 1200
    },
    {
      "epoch": 0.4991714995857498,
      "grad_norm": 2.6047443491478393,
      "learning_rate": 5.8848542773121285e-06,
      "loss": 0.468,
      "step": 1205
    },
    {
      "epoch": 0.5012427506213754,
      "grad_norm": 2.4826419630150407,
      "learning_rate": 5.849242059545259e-06,
      "loss": 0.4565,
      "step": 1210
    },
    {
      "epoch": 0.5033140016570008,
      "grad_norm": 2.757538020740051,
      "learning_rate": 5.81358542472807e-06,
      "loss": 0.4598,
      "step": 1215
    },
    {
      "epoch": 0.5053852526926264,
      "grad_norm": 2.522678699637508,
      "learning_rate": 5.777886237773542e-06,
      "loss": 0.4451,
      "step": 1220
    },
    {
      "epoch": 0.5074565037282519,
      "grad_norm": 2.697025952581437,
      "learning_rate": 5.742146365820223e-06,
      "loss": 0.4606,
      "step": 1225
    },
    {
      "epoch": 0.5095277547638773,
      "grad_norm": 2.5255536162735095,
      "learning_rate": 5.706367678134562e-06,
      "loss": 0.4587,
      "step": 1230
    },
    {
      "epoch": 0.5115990057995029,
      "grad_norm": 2.5706493150820866,
      "learning_rate": 5.670552046013151e-06,
      "loss": 0.4626,
      "step": 1235
    },
    {
      "epoch": 0.5136702568351285,
      "grad_norm": 2.478635300584232,
      "learning_rate": 5.634701342684852e-06,
      "loss": 0.4495,
      "step": 1240
    },
    {
      "epoch": 0.5157415078707539,
      "grad_norm": 2.413081651555011,
      "learning_rate": 5.598817443212813e-06,
      "loss": 0.4545,
      "step": 1245
    },
    {
      "epoch": 0.5178127589063795,
      "grad_norm": 2.4514043208682614,
      "learning_rate": 5.562902224396416e-06,
      "loss": 0.4487,
      "step": 1250
    },
    {
      "epoch": 0.519884009942005,
      "grad_norm": 2.8553325471084894,
      "learning_rate": 5.526957564673098e-06,
      "loss": 0.4491,
      "step": 1255
    },
    {
      "epoch": 0.5219552609776305,
      "grad_norm": 2.3499605624954043,
      "learning_rate": 5.49098534402012e-06,
      "loss": 0.455,
      "step": 1260
    },
    {
      "epoch": 0.524026512013256,
      "grad_norm": 2.511645835881994,
      "learning_rate": 5.454987443856235e-06,
      "loss": 0.4405,
      "step": 1265
    },
    {
      "epoch": 0.5260977630488816,
      "grad_norm": 2.5573246473567663,
      "learning_rate": 5.418965746943281e-06,
      "loss": 0.4573,
      "step": 1270
    },
    {
      "epoch": 0.528169014084507,
      "grad_norm": 2.6069496932706446,
      "learning_rate": 5.3829221372877175e-06,
      "loss": 0.4385,
      "step": 1275
    },
    {
      "epoch": 0.5302402651201326,
      "grad_norm": 2.3669759764626286,
      "learning_rate": 5.34685850004208e-06,
      "loss": 0.4474,
      "step": 1280
    },
    {
      "epoch": 0.5323115161557581,
      "grad_norm": 2.5262162943745428,
      "learning_rate": 5.310776721406392e-06,
      "loss": 0.4465,
      "step": 1285
    },
    {
      "epoch": 0.5343827671913836,
      "grad_norm": 2.367388706229585,
      "learning_rate": 5.2746786885295034e-06,
      "loss": 0.4477,
      "step": 1290
    },
    {
      "epoch": 0.5364540182270091,
      "grad_norm": 2.4424670349956834,
      "learning_rate": 5.238566289410396e-06,
      "loss": 0.4347,
      "step": 1295
    },
    {
      "epoch": 0.5385252692626347,
      "grad_norm": 2.374597209278358,
      "learning_rate": 5.2024414127994325e-06,
      "loss": 0.4414,
      "step": 1300
    },
    {
      "epoch": 0.5405965202982601,
      "grad_norm": 2.363317895955179,
      "learning_rate": 5.166305948099574e-06,
      "loss": 0.4211,
      "step": 1305
    },
    {
      "epoch": 0.5426677713338857,
      "grad_norm": 2.445116140490052,
      "learning_rate": 5.13016178526756e-06,
      "loss": 0.4341,
      "step": 1310
    },
    {
      "epoch": 0.5447390223695112,
      "grad_norm": 2.451842751824045,
      "learning_rate": 5.094010814715062e-06,
      "loss": 0.4323,
      "step": 1315
    },
    {
      "epoch": 0.5468102734051367,
      "grad_norm": 2.4108281284288116,
      "learning_rate": 5.057854927209804e-06,
      "loss": 0.4363,
      "step": 1320
    },
    {
      "epoch": 0.5488815244407622,
      "grad_norm": 2.47478290810414,
      "learning_rate": 5.0216960137766805e-06,
      "loss": 0.429,
      "step": 1325
    },
    {
      "epoch": 0.5509527754763878,
      "grad_norm": 2.462991177532584,
      "learning_rate": 4.985535965598843e-06,
      "loss": 0.4504,
      "step": 1330
    },
    {
      "epoch": 0.5530240265120132,
      "grad_norm": 2.437965270822768,
      "learning_rate": 4.949376673918802e-06,
      "loss": 0.432,
      "step": 1335
    },
    {
      "epoch": 0.5550952775476388,
      "grad_norm": 2.588594999778493,
      "learning_rate": 4.913220029939491e-06,
      "loss": 0.4335,
      "step": 1340
    },
    {
      "epoch": 0.5571665285832643,
      "grad_norm": 2.624887614176281,
      "learning_rate": 4.877067924725368e-06,
      "loss": 0.4268,
      "step": 1345
    },
    {
      "epoch": 0.5592377796188898,
      "grad_norm": 2.59768554910079,
      "learning_rate": 4.840922249103506e-06,
      "loss": 0.4304,
      "step": 1350
    },
    {
      "epoch": 0.5613090306545153,
      "grad_norm": 2.3988418599389063,
      "learning_rate": 4.804784893564697e-06,
      "loss": 0.4411,
      "step": 1355
    },
    {
      "epoch": 0.5633802816901409,
      "grad_norm": 2.580765462139435,
      "learning_rate": 4.7686577481645745e-06,
      "loss": 0.4082,
      "step": 1360
    },
    {
      "epoch": 0.5654515327257663,
      "grad_norm": 2.4467281357535855,
      "learning_rate": 4.732542702424759e-06,
      "loss": 0.4382,
      "step": 1365
    },
    {
      "epoch": 0.5675227837613919,
      "grad_norm": 2.6370177774043073,
      "learning_rate": 4.696441645234042e-06,
      "loss": 0.4271,
      "step": 1370
    },
    {
      "epoch": 0.5695940347970174,
      "grad_norm": 2.475111070321159,
      "learning_rate": 4.660356464749578e-06,
      "loss": 0.4185,
      "step": 1375
    },
    {
      "epoch": 0.5716652858326429,
      "grad_norm": 2.6994305643163776,
      "learning_rate": 4.624289048298147e-06,
      "loss": 0.4259,
      "step": 1380
    },
    {
      "epoch": 0.5737365368682684,
      "grad_norm": 2.4520066328265977,
      "learning_rate": 4.588241282277428e-06,
      "loss": 0.4179,
      "step": 1385
    },
    {
      "epoch": 0.575807787903894,
      "grad_norm": 2.3795955584858386,
      "learning_rate": 4.55221505205734e-06,
      "loss": 0.4287,
      "step": 1390
    },
    {
      "epoch": 0.5778790389395194,
      "grad_norm": 2.381541096732435,
      "learning_rate": 4.516212241881448e-06,
      "loss": 0.4157,
      "step": 1395
    },
    {
      "epoch": 0.579950289975145,
      "grad_norm": 2.566306679289895,
      "learning_rate": 4.480234734768393e-06,
      "loss": 0.4213,
      "step": 1400
    },
    {
      "epoch": 0.5820215410107705,
      "grad_norm": 2.396097079278926,
      "learning_rate": 4.444284412413418e-06,
      "loss": 0.4172,
      "step": 1405
    },
    {
      "epoch": 0.584092792046396,
      "grad_norm": 2.3514162631646274,
      "learning_rate": 4.408363155089952e-06,
      "loss": 0.4211,
      "step": 1410
    },
    {
      "epoch": 0.5861640430820215,
      "grad_norm": 2.3172982824966906,
      "learning_rate": 4.3724728415512585e-06,
      "loss": 0.4193,
      "step": 1415
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 2.472751185964712,
      "learning_rate": 4.3366153489321855e-06,
      "loss": 0.4264,
      "step": 1420
    },
    {
      "epoch": 0.5903065451532725,
      "grad_norm": 2.4154833816633365,
      "learning_rate": 4.30079255265098e-06,
      "loss": 0.4048,
      "step": 1425
    },
    {
      "epoch": 0.5923777961888981,
      "grad_norm": 2.5211717793789803,
      "learning_rate": 4.265006326311199e-06,
      "loss": 0.4089,
      "step": 1430
    },
    {
      "epoch": 0.5944490472245236,
      "grad_norm": 2.3701199367093504,
      "learning_rate": 4.229258541603723e-06,
      "loss": 0.4156,
      "step": 1435
    },
    {
      "epoch": 0.5965202982601492,
      "grad_norm": 2.3994950566503994,
      "learning_rate": 4.1935510682088545e-06,
      "loss": 0.4065,
      "step": 1440
    },
    {
      "epoch": 0.5985915492957746,
      "grad_norm": 2.4968713927832273,
      "learning_rate": 4.157885773698535e-06,
      "loss": 0.4057,
      "step": 1445
    },
    {
      "epoch": 0.6006628003314002,
      "grad_norm": 2.416064472998044,
      "learning_rate": 4.122264523438668e-06,
      "loss": 0.3931,
      "step": 1450
    },
    {
      "epoch": 0.6027340513670257,
      "grad_norm": 2.2816034379976355,
      "learning_rate": 4.086689180491554e-06,
      "loss": 0.4099,
      "step": 1455
    },
    {
      "epoch": 0.6048053024026512,
      "grad_norm": 2.40826781123055,
      "learning_rate": 4.051161605518453e-06,
      "loss": 0.4149,
      "step": 1460
    },
    {
      "epoch": 0.6068765534382767,
      "grad_norm": 2.538687087204467,
      "learning_rate": 4.015683656682255e-06,
      "loss": 0.4175,
      "step": 1465
    },
    {
      "epoch": 0.6089478044739023,
      "grad_norm": 2.430104423449275,
      "learning_rate": 3.980257189550316e-06,
      "loss": 0.4053,
      "step": 1470
    },
    {
      "epoch": 0.6110190555095277,
      "grad_norm": 2.3721045844407143,
      "learning_rate": 3.94488405699739e-06,
      "loss": 0.4031,
      "step": 1475
    },
    {
      "epoch": 0.6130903065451533,
      "grad_norm": 2.338789854463612,
      "learning_rate": 3.909566109108727e-06,
      "loss": 0.3972,
      "step": 1480
    },
    {
      "epoch": 0.6151615575807788,
      "grad_norm": 2.2787527608868943,
      "learning_rate": 3.874305193083313e-06,
      "loss": 0.411,
      "step": 1485
    },
    {
      "epoch": 0.6172328086164043,
      "grad_norm": 2.454450620709443,
      "learning_rate": 3.839103153137247e-06,
      "loss": 0.4008,
      "step": 1490
    },
    {
      "epoch": 0.6193040596520298,
      "grad_norm": 2.4225686330793823,
      "learning_rate": 3.803961830407297e-06,
      "loss": 0.3974,
      "step": 1495
    },
    {
      "epoch": 0.6213753106876554,
      "grad_norm": 2.564739658366535,
      "learning_rate": 3.768883062854598e-06,
      "loss": 0.4152,
      "step": 1500
    },
    {
      "epoch": 0.6234465617232808,
      "grad_norm": 2.4055933426253837,
      "learning_rate": 3.7338686851685267e-06,
      "loss": 0.4021,
      "step": 1505
    },
    {
      "epoch": 0.6255178127589064,
      "grad_norm": 2.3297777756780986,
      "learning_rate": 3.6989205286707398e-06,
      "loss": 0.4087,
      "step": 1510
    },
    {
      "epoch": 0.6275890637945319,
      "grad_norm": 2.276174828544462,
      "learning_rate": 3.664040421219393e-06,
      "loss": 0.4,
      "step": 1515
    },
    {
      "epoch": 0.6296603148301574,
      "grad_norm": 2.558523087497906,
      "learning_rate": 3.6292301871135425e-06,
      "loss": 0.3919,
      "step": 1520
    },
    {
      "epoch": 0.6317315658657829,
      "grad_norm": 2.599732843239929,
      "learning_rate": 3.59449164699773e-06,
      "loss": 0.4016,
      "step": 1525
    },
    {
      "epoch": 0.6338028169014085,
      "grad_norm": 2.4726991401898952,
      "learning_rate": 3.55982661776676e-06,
      "loss": 0.3986,
      "step": 1530
    },
    {
      "epoch": 0.6358740679370339,
      "grad_norm": 2.321690978258901,
      "learning_rate": 3.5252369124706697e-06,
      "loss": 0.3978,
      "step": 1535
    },
    {
      "epoch": 0.6379453189726595,
      "grad_norm": 2.6754380236168664,
      "learning_rate": 3.4907243402199013e-06,
      "loss": 0.4113,
      "step": 1540
    },
    {
      "epoch": 0.640016570008285,
      "grad_norm": 2.4052468560081475,
      "learning_rate": 3.4562907060906908e-06,
      "loss": 0.3813,
      "step": 1545
    },
    {
      "epoch": 0.6420878210439105,
      "grad_norm": 2.4570458861707696,
      "learning_rate": 3.4219378110306523e-06,
      "loss": 0.392,
      "step": 1550
    },
    {
      "epoch": 0.644159072079536,
      "grad_norm": 2.3101859891504866,
      "learning_rate": 3.3876674517645815e-06,
      "loss": 0.3996,
      "step": 1555
    },
    {
      "epoch": 0.6462303231151616,
      "grad_norm": 2.4539409882344483,
      "learning_rate": 3.353481420700495e-06,
      "loss": 0.3979,
      "step": 1560
    },
    {
      "epoch": 0.648301574150787,
      "grad_norm": 2.2847760045122087,
      "learning_rate": 3.319381505835868e-06,
      "loss": 0.3749,
      "step": 1565
    },
    {
      "epoch": 0.6503728251864126,
      "grad_norm": 2.2918085600506717,
      "learning_rate": 3.285369490664133e-06,
      "loss": 0.3895,
      "step": 1570
    },
    {
      "epoch": 0.6524440762220381,
      "grad_norm": 2.4126038265192538,
      "learning_rate": 3.251447154081394e-06,
      "loss": 0.3885,
      "step": 1575
    },
    {
      "epoch": 0.6545153272576636,
      "grad_norm": 2.4831233203755603,
      "learning_rate": 3.2176162702933816e-06,
      "loss": 0.413,
      "step": 1580
    },
    {
      "epoch": 0.6565865782932891,
      "grad_norm": 2.385955156565656,
      "learning_rate": 3.183878608722669e-06,
      "loss": 0.3838,
      "step": 1585
    },
    {
      "epoch": 0.6586578293289147,
      "grad_norm": 2.5426984744026124,
      "learning_rate": 3.150235933916115e-06,
      "loss": 0.3755,
      "step": 1590
    },
    {
      "epoch": 0.6607290803645401,
      "grad_norm": 2.388904551962294,
      "learning_rate": 3.1166900054525873e-06,
      "loss": 0.3868,
      "step": 1595
    },
    {
      "epoch": 0.6628003314001657,
      "grad_norm": 2.5869530818499147,
      "learning_rate": 3.0832425778509235e-06,
      "loss": 0.3821,
      "step": 1600
    },
    {
      "epoch": 0.6648715824357913,
      "grad_norm": 2.47920960360778,
      "learning_rate": 3.049895400478174e-06,
      "loss": 0.3935,
      "step": 1605
    },
    {
      "epoch": 0.6669428334714167,
      "grad_norm": 2.4320164826856234,
      "learning_rate": 3.0166502174581012e-06,
      "loss": 0.3732,
      "step": 1610
    },
    {
      "epoch": 0.6690140845070423,
      "grad_norm": 2.3447158283526783,
      "learning_rate": 2.983508767579956e-06,
      "loss": 0.3805,
      "step": 1615
    },
    {
      "epoch": 0.6710853355426678,
      "grad_norm": 2.336702612441604,
      "learning_rate": 2.950472784207544e-06,
      "loss": 0.3814,
      "step": 1620
    },
    {
      "epoch": 0.6731565865782932,
      "grad_norm": 2.403683822019676,
      "learning_rate": 2.917543995188562e-06,
      "loss": 0.3799,
      "step": 1625
    },
    {
      "epoch": 0.6752278376139188,
      "grad_norm": 2.32245444127231,
      "learning_rate": 2.8847241227642255e-06,
      "loss": 0.3878,
      "step": 1630
    },
    {
      "epoch": 0.6772990886495444,
      "grad_norm": 2.2336601101273894,
      "learning_rate": 2.852014883479198e-06,
      "loss": 0.3668,
      "step": 1635
    },
    {
      "epoch": 0.6793703396851698,
      "grad_norm": 2.4572013362431666,
      "learning_rate": 2.819417988091814e-06,
      "loss": 0.3789,
      "step": 1640
    },
    {
      "epoch": 0.6814415907207954,
      "grad_norm": 2.374979599510779,
      "learning_rate": 2.786935141484586e-06,
      "loss": 0.3776,
      "step": 1645
    },
    {
      "epoch": 0.6835128417564209,
      "grad_norm": 2.387695271521185,
      "learning_rate": 2.754568042575061e-06,
      "loss": 0.3785,
      "step": 1650
    },
    {
      "epoch": 0.6855840927920464,
      "grad_norm": 2.4380470816357573,
      "learning_rate": 2.7223183842269442e-06,
      "loss": 0.3762,
      "step": 1655
    },
    {
      "epoch": 0.6876553438276719,
      "grad_norm": 2.2802764216409996,
      "learning_rate": 2.6901878531615677e-06,
      "loss": 0.377,
      "step": 1660
    },
    {
      "epoch": 0.6897265948632975,
      "grad_norm": 2.3338582614060748,
      "learning_rate": 2.658178129869672e-06,
      "loss": 0.382,
      "step": 1665
    },
    {
      "epoch": 0.6917978458989229,
      "grad_norm": 2.4939738702851932,
      "learning_rate": 2.6262908885235046e-06,
      "loss": 0.3789,
      "step": 1670
    },
    {
      "epoch": 0.6938690969345485,
      "grad_norm": 2.409154689853176,
      "learning_rate": 2.594527796889265e-06,
      "loss": 0.3859,
      "step": 1675
    },
    {
      "epoch": 0.695940347970174,
      "grad_norm": 2.4232651472750724,
      "learning_rate": 2.5628905162398797e-06,
      "loss": 0.3759,
      "step": 1680
    },
    {
      "epoch": 0.6980115990057995,
      "grad_norm": 2.438606094251203,
      "learning_rate": 2.531380701268108e-06,
      "loss": 0.3707,
      "step": 1685
    },
    {
      "epoch": 0.700082850041425,
      "grad_norm": 2.2068267980004164,
      "learning_rate": 2.5000000000000015e-06,
      "loss": 0.3635,
      "step": 1690
    },
    {
      "epoch": 0.7021541010770506,
      "grad_norm": 2.4849851469693487,
      "learning_rate": 2.4687500537087027e-06,
      "loss": 0.3765,
      "step": 1695
    },
    {
      "epoch": 0.704225352112676,
      "grad_norm": 2.3676710626755604,
      "learning_rate": 2.4376324968286154e-06,
      "loss": 0.3706,
      "step": 1700
    },
    {
      "epoch": 0.7062966031483016,
      "grad_norm": 2.4665247552413194,
      "learning_rate": 2.40664895686991e-06,
      "loss": 0.359,
      "step": 1705
    },
    {
      "epoch": 0.7083678541839271,
      "grad_norm": 2.4228989801005882,
      "learning_rate": 2.375801054333409e-06,
      "loss": 0.3574,
      "step": 1710
    },
    {
      "epoch": 0.7104391052195526,
      "grad_norm": 2.447240340812193,
      "learning_rate": 2.345090402625822e-06,
      "loss": 0.3629,
      "step": 1715
    },
    {
      "epoch": 0.7125103562551781,
      "grad_norm": 2.494665611756937,
      "learning_rate": 2.3145186079753685e-06,
      "loss": 0.3607,
      "step": 1720
    },
    {
      "epoch": 0.7145816072908037,
      "grad_norm": 2.3466738040218433,
      "learning_rate": 2.2840872693477694e-06,
      "loss": 0.3582,
      "step": 1725
    },
    {
      "epoch": 0.7166528583264291,
      "grad_norm": 2.4579034785541274,
      "learning_rate": 2.253797978362617e-06,
      "loss": 0.37,
      "step": 1730
    },
    {
      "epoch": 0.7187241093620547,
      "grad_norm": 2.294277539263346,
      "learning_rate": 2.2236523192101264e-06,
      "loss": 0.371,
      "step": 1735
    },
    {
      "epoch": 0.7207953603976802,
      "grad_norm": 2.3165105524366676,
      "learning_rate": 2.193651868568285e-06,
      "loss": 0.3533,
      "step": 1740
    },
    {
      "epoch": 0.7228666114333057,
      "grad_norm": 2.3412049707939766,
      "learning_rate": 2.16379819552038e-06,
      "loss": 0.3598,
      "step": 1745
    },
    {
      "epoch": 0.7249378624689312,
      "grad_norm": 2.37970184238566,
      "learning_rate": 2.1340928614729445e-06,
      "loss": 0.3553,
      "step": 1750
    },
    {
      "epoch": 0.7270091135045568,
      "grad_norm": 2.5101331147915733,
      "learning_rate": 2.1045374200740863e-06,
      "loss": 0.3589,
      "step": 1755
    },
    {
      "epoch": 0.7290803645401823,
      "grad_norm": 2.381929011616353,
      "learning_rate": 2.075133417132223e-06,
      "loss": 0.3544,
      "step": 1760
    },
    {
      "epoch": 0.7311516155758078,
      "grad_norm": 2.64277485599053,
      "learning_rate": 2.045882390535248e-06,
      "loss": 0.354,
      "step": 1765
    },
    {
      "epoch": 0.7332228666114333,
      "grad_norm": 2.2983110375410756,
      "learning_rate": 2.016785870170079e-06,
      "loss": 0.359,
      "step": 1770
    },
    {
      "epoch": 0.7352941176470589,
      "grad_norm": 2.338818197474944,
      "learning_rate": 1.987845377842656e-06,
      "loss": 0.3515,
      "step": 1775
    },
    {
      "epoch": 0.7373653686826843,
      "grad_norm": 2.412156205295859,
      "learning_rate": 1.9590624271983406e-06,
      "loss": 0.3707,
      "step": 1780
    },
    {
      "epoch": 0.7394366197183099,
      "grad_norm": 2.3221544897617394,
      "learning_rate": 1.9304385236427505e-06,
      "loss": 0.3522,
      "step": 1785
    },
    {
      "epoch": 0.7415078707539354,
      "grad_norm": 2.4316222609963427,
      "learning_rate": 1.9019751642630252e-06,
      "loss": 0.3574,
      "step": 1790
    },
    {
      "epoch": 0.7435791217895609,
      "grad_norm": 2.2641479623423635,
      "learning_rate": 1.8736738377495196e-06,
      "loss": 0.3592,
      "step": 1795
    },
    {
      "epoch": 0.7456503728251864,
      "grad_norm": 2.3420647549229265,
      "learning_rate": 1.8455360243179537e-06,
      "loss": 0.3542,
      "step": 1800
    },
    {
      "epoch": 0.747721623860812,
      "grad_norm": 2.456638840774933,
      "learning_rate": 1.8175631956319823e-06,
      "loss": 0.3542,
      "step": 1805
    },
    {
      "epoch": 0.7497928748964374,
      "grad_norm": 2.315911960148082,
      "learning_rate": 1.7897568147262323e-06,
      "loss": 0.3651,
      "step": 1810
    },
    {
      "epoch": 0.751864125932063,
      "grad_norm": 2.430456263244821,
      "learning_rate": 1.7621183359297817e-06,
      "loss": 0.3513,
      "step": 1815
    },
    {
      "epoch": 0.7539353769676885,
      "grad_norm": 2.2820850020652137,
      "learning_rate": 1.7346492047900897e-06,
      "loss": 0.344,
      "step": 1820
    },
    {
      "epoch": 0.756006628003314,
      "grad_norm": 2.4006568265757755,
      "learning_rate": 1.7073508579973996e-06,
      "loss": 0.3526,
      "step": 1825
    },
    {
      "epoch": 0.7580778790389395,
      "grad_norm": 2.2901041487912774,
      "learning_rate": 1.6802247233095914e-06,
      "loss": 0.349,
      "step": 1830
    },
    {
      "epoch": 0.7601491300745651,
      "grad_norm": 2.4009693108682204,
      "learning_rate": 1.6532722194775108e-06,
      "loss": 0.3537,
      "step": 1835
    },
    {
      "epoch": 0.7622203811101905,
      "grad_norm": 2.343710211021654,
      "learning_rate": 1.626494756170765e-06,
      "loss": 0.3536,
      "step": 1840
    },
    {
      "epoch": 0.7642916321458161,
      "grad_norm": 2.3742949969171363,
      "learning_rate": 1.5998937339039889e-06,
      "loss": 0.3542,
      "step": 1845
    },
    {
      "epoch": 0.7663628831814416,
      "grad_norm": 2.3672793937996093,
      "learning_rate": 1.5734705439636017e-06,
      "loss": 0.3428,
      "step": 1850
    },
    {
      "epoch": 0.7684341342170671,
      "grad_norm": 2.2889634008759803,
      "learning_rate": 1.5472265683350397e-06,
      "loss": 0.3535,
      "step": 1855
    },
    {
      "epoch": 0.7705053852526926,
      "grad_norm": 2.2816526862864164,
      "learning_rate": 1.5211631796304721e-06,
      "loss": 0.3452,
      "step": 1860
    },
    {
      "epoch": 0.7725766362883182,
      "grad_norm": 2.256494769623704,
      "learning_rate": 1.495281741017016e-06,
      "loss": 0.3569,
      "step": 1865
    },
    {
      "epoch": 0.7746478873239436,
      "grad_norm": 2.2248161459647333,
      "learning_rate": 1.46958360614543e-06,
      "loss": 0.3439,
      "step": 1870
    },
    {
      "epoch": 0.7767191383595692,
      "grad_norm": 2.3902740970389065,
      "learning_rate": 1.4440701190793278e-06,
      "loss": 0.3301,
      "step": 1875
    },
    {
      "epoch": 0.7787903893951947,
      "grad_norm": 2.439067358960243,
      "learning_rate": 1.4187426142248723e-06,
      "loss": 0.3462,
      "step": 1880
    },
    {
      "epoch": 0.7808616404308202,
      "grad_norm": 2.3392746632984585,
      "learning_rate": 1.3936024162609897e-06,
      "loss": 0.3408,
      "step": 1885
    },
    {
      "epoch": 0.7829328914664457,
      "grad_norm": 2.377498700024158,
      "learning_rate": 1.3686508400700787e-06,
      "loss": 0.3549,
      "step": 1890
    },
    {
      "epoch": 0.7850041425020713,
      "grad_norm": 2.3496620916951767,
      "learning_rate": 1.3438891906692447e-06,
      "loss": 0.3472,
      "step": 1895
    },
    {
      "epoch": 0.7870753935376967,
      "grad_norm": 2.289494721325453,
      "learning_rate": 1.3193187631420462e-06,
      "loss": 0.3355,
      "step": 1900
    },
    {
      "epoch": 0.7891466445733223,
      "grad_norm": 2.2947285370456396,
      "learning_rate": 1.2949408425707566e-06,
      "loss": 0.3394,
      "step": 1905
    },
    {
      "epoch": 0.7912178956089478,
      "grad_norm": 2.3617970794893126,
      "learning_rate": 1.2707567039691505e-06,
      "loss": 0.3497,
      "step": 1910
    },
    {
      "epoch": 0.7932891466445733,
      "grad_norm": 2.321811184391228,
      "learning_rate": 1.2467676122158224e-06,
      "loss": 0.3412,
      "step": 1915
    },
    {
      "epoch": 0.7953603976801988,
      "grad_norm": 2.3211063248548967,
      "learning_rate": 1.222974821988024e-06,
      "loss": 0.3436,
      "step": 1920
    },
    {
      "epoch": 0.7974316487158244,
      "grad_norm": 2.367048239376449,
      "learning_rate": 1.1993795776960498e-06,
      "loss": 0.3387,
      "step": 1925
    },
    {
      "epoch": 0.7995028997514498,
      "grad_norm": 2.415017409117694,
      "learning_rate": 1.1759831134181504e-06,
      "loss": 0.3477,
      "step": 1930
    },
    {
      "epoch": 0.8015741507870754,
      "grad_norm": 2.2878955632108737,
      "learning_rate": 1.1527866528359805e-06,
      "loss": 0.3424,
      "step": 1935
    },
    {
      "epoch": 0.8036454018227009,
      "grad_norm": 2.4168712983822407,
      "learning_rate": 1.1297914091706086e-06,
      "loss": 0.3395,
      "step": 1940
    },
    {
      "epoch": 0.8057166528583264,
      "grad_norm": 2.332403241899359,
      "learning_rate": 1.1069985851190524e-06,
      "loss": 0.3413,
      "step": 1945
    },
    {
      "epoch": 0.8077879038939519,
      "grad_norm": 2.245314869032349,
      "learning_rate": 1.0844093727913868e-06,
      "loss": 0.3302,
      "step": 1950
    },
    {
      "epoch": 0.8098591549295775,
      "grad_norm": 2.3766452170038814,
      "learning_rate": 1.062024953648384e-06,
      "loss": 0.3363,
      "step": 1955
    },
    {
      "epoch": 0.8119304059652029,
      "grad_norm": 2.3725445423461076,
      "learning_rate": 1.039846498439727e-06,
      "loss": 0.3312,
      "step": 1960
    },
    {
      "epoch": 0.8140016570008285,
      "grad_norm": 2.400378472961016,
      "learning_rate": 1.0178751671427755e-06,
      "loss": 0.3406,
      "step": 1965
    },
    {
      "epoch": 0.816072908036454,
      "grad_norm": 2.600161757489673,
      "learning_rate": 9.961121089018933e-07,
      "loss": 0.3286,
      "step": 1970
    },
    {
      "epoch": 0.8181441590720795,
      "grad_norm": 2.4280753174812597,
      "learning_rate": 9.745584619683524e-07,
      "loss": 0.3439,
      "step": 1975
    },
    {
      "epoch": 0.820215410107705,
      "grad_norm": 2.260545698321657,
      "learning_rate": 9.532153536407923e-07,
      "loss": 0.3317,
      "step": 1980
    },
    {
      "epoch": 0.8222866611433306,
      "grad_norm": 2.4227916089483204,
      "learning_rate": 9.320839002062682e-07,
      "loss": 0.3312,
      "step": 1985
    },
    {
      "epoch": 0.824357912178956,
      "grad_norm": 2.8477868846172067,
      "learning_rate": 9.111652068818621e-07,
      "loss": 0.3334,
      "step": 1990
    },
    {
      "epoch": 0.8264291632145816,
      "grad_norm": 2.370728174821394,
      "learning_rate": 8.904603677568785e-07,
      "loss": 0.3278,
      "step": 1995
    },
    {
      "epoch": 0.8285004142502072,
      "grad_norm": 2.3115016796994183,
      "learning_rate": 8.699704657356195e-07,
      "loss": 0.3358,
      "step": 2000
    },
    {
      "epoch": 0.8305716652858326,
      "grad_norm": 2.253238724272844,
      "learning_rate": 8.496965724807516e-07,
      "loss": 0.3348,
      "step": 2005
    },
    {
      "epoch": 0.8326429163214581,
      "grad_norm": 2.3696595779036573,
      "learning_rate": 8.296397483572515e-07,
      "loss": 0.3322,
      "step": 2010
    },
    {
      "epoch": 0.8347141673570837,
      "grad_norm": 2.3994710782454143,
      "learning_rate": 8.098010423769503e-07,
      "loss": 0.3159,
      "step": 2015
    },
    {
      "epoch": 0.8367854183927091,
      "grad_norm": 2.1347268704516638,
      "learning_rate": 7.901814921436624e-07,
      "loss": 0.3318,
      "step": 2020
    },
    {
      "epoch": 0.8388566694283347,
      "grad_norm": 2.2628490853732397,
      "learning_rate": 7.70782123798921e-07,
      "loss": 0.3318,
      "step": 2025
    },
    {
      "epoch": 0.8409279204639603,
      "grad_norm": 2.2582184248706683,
      "learning_rate": 7.516039519683105e-07,
      "loss": 0.3234,
      "step": 2030
    },
    {
      "epoch": 0.8429991714995857,
      "grad_norm": 2.476131958227405,
      "learning_rate": 7.326479797083963e-07,
      "loss": 0.3278,
      "step": 2035
    },
    {
      "epoch": 0.8450704225352113,
      "grad_norm": 2.3728222901700677,
      "learning_rate": 7.139151984542636e-07,
      "loss": 0.3306,
      "step": 2040
    },
    {
      "epoch": 0.8471416735708368,
      "grad_norm": 2.3665157862844994,
      "learning_rate": 6.954065879676653e-07,
      "loss": 0.3309,
      "step": 2045
    },
    {
      "epoch": 0.8492129246064622,
      "grad_norm": 2.5008625602985304,
      "learning_rate": 6.771231162857722e-07,
      "loss": 0.3362,
      "step": 2050
    },
    {
      "epoch": 0.8512841756420878,
      "grad_norm": 2.2712774856509688,
      "learning_rate": 6.590657396705525e-07,
      "loss": 0.3215,
      "step": 2055
    },
    {
      "epoch": 0.8533554266777134,
      "grad_norm": 2.535681331577295,
      "learning_rate": 6.412354025587509e-07,
      "loss": 0.3229,
      "step": 2060
    },
    {
      "epoch": 0.8554266777133389,
      "grad_norm": 2.547194182489716,
      "learning_rate": 6.236330375124921e-07,
      "loss": 0.3461,
      "step": 2065
    },
    {
      "epoch": 0.8574979287489644,
      "grad_norm": 2.371373910066343,
      "learning_rate": 6.062595651705111e-07,
      "loss": 0.3389,
      "step": 2070
    },
    {
      "epoch": 0.8595691797845899,
      "grad_norm": 2.439136346587766,
      "learning_rate": 5.891158941999959e-07,
      "loss": 0.3299,
      "step": 2075
    },
    {
      "epoch": 0.8616404308202155,
      "grad_norm": 2.466656635870232,
      "learning_rate": 5.722029212490666e-07,
      "loss": 0.3298,
      "step": 2080
    },
    {
      "epoch": 0.8637116818558409,
      "grad_norm": 2.288892577992347,
      "learning_rate": 5.555215308998779e-07,
      "loss": 0.3223,
      "step": 2085
    },
    {
      "epoch": 0.8657829328914665,
      "grad_norm": 2.401858716049078,
      "learning_rate": 5.390725956223531e-07,
      "loss": 0.3218,
      "step": 2090
    },
    {
      "epoch": 0.867854183927092,
      "grad_norm": 2.33005782306982,
      "learning_rate": 5.22856975728554e-07,
      "loss": 0.3216,
      "step": 2095
    },
    {
      "epoch": 0.8699254349627175,
      "grad_norm": 2.4143761257261303,
      "learning_rate": 5.068755193276798e-07,
      "loss": 0.3307,
      "step": 2100
    },
    {
      "epoch": 0.871996685998343,
      "grad_norm": 2.352096096095919,
      "learning_rate": 4.911290622817161e-07,
      "loss": 0.3204,
      "step": 2105
    },
    {
      "epoch": 0.8740679370339686,
      "grad_norm": 2.3702490110719303,
      "learning_rate": 4.756184281617121e-07,
      "loss": 0.3245,
      "step": 2110
    },
    {
      "epoch": 0.876139188069594,
      "grad_norm": 2.2456862499766306,
      "learning_rate": 4.6034442820471037e-07,
      "loss": 0.3182,
      "step": 2115
    },
    {
      "epoch": 0.8782104391052196,
      "grad_norm": 2.2534896280263816,
      "learning_rate": 4.4530786127131575e-07,
      "loss": 0.3312,
      "step": 2120
    },
    {
      "epoch": 0.8802816901408451,
      "grad_norm": 2.309108572623016,
      "learning_rate": 4.305095138039106e-07,
      "loss": 0.326,
      "step": 2125
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 2.5387736296808563,
      "learning_rate": 4.159501597855287e-07,
      "loss": 0.3327,
      "step": 2130
    },
    {
      "epoch": 0.8844241922120961,
      "grad_norm": 2.4068093993598962,
      "learning_rate": 4.0163056069936757e-07,
      "loss": 0.3265,
      "step": 2135
    },
    {
      "epoch": 0.8864954432477217,
      "grad_norm": 2.410912519062069,
      "learning_rate": 3.8755146548896784e-07,
      "loss": 0.3286,
      "step": 2140
    },
    {
      "epoch": 0.8885666942833471,
      "grad_norm": 2.3441979713103476,
      "learning_rate": 3.737136105190337e-07,
      "loss": 0.3215,
      "step": 2145
    },
    {
      "epoch": 0.8906379453189727,
      "grad_norm": 2.4124574145028843,
      "learning_rate": 3.6011771953693044e-07,
      "loss": 0.3222,
      "step": 2150
    },
    {
      "epoch": 0.8927091963545982,
      "grad_norm": 2.3579671682789467,
      "learning_rate": 3.4676450363481937e-07,
      "loss": 0.3227,
      "step": 2155
    },
    {
      "epoch": 0.8947804473902237,
      "grad_norm": 2.3464323541089214,
      "learning_rate": 3.336546612124758e-07,
      "loss": 0.3197,
      "step": 2160
    },
    {
      "epoch": 0.8968516984258492,
      "grad_norm": 2.4445949626198242,
      "learning_rate": 3.20788877940757e-07,
      "loss": 0.3164,
      "step": 2165
    },
    {
      "epoch": 0.8989229494614748,
      "grad_norm": 2.668457673876882,
      "learning_rate": 3.081678267257404e-07,
      "loss": 0.3233,
      "step": 2170
    },
    {
      "epoch": 0.9009942004971002,
      "grad_norm": 2.2794632035182865,
      "learning_rate": 2.9579216767352815e-07,
      "loss": 0.314,
      "step": 2175
    },
    {
      "epoch": 0.9030654515327258,
      "grad_norm": 2.511600762449477,
      "learning_rate": 2.836625480557265e-07,
      "loss": 0.3195,
      "step": 2180
    },
    {
      "epoch": 0.9051367025683513,
      "grad_norm": 2.290095274854777,
      "learning_rate": 2.7177960227558863e-07,
      "loss": 0.323,
      "step": 2185
    },
    {
      "epoch": 0.9072079536039768,
      "grad_norm": 2.490832241300732,
      "learning_rate": 2.601439518348331e-07,
      "loss": 0.3229,
      "step": 2190
    },
    {
      "epoch": 0.9092792046396023,
      "grad_norm": 2.3369810868473833,
      "learning_rate": 2.487562053011422e-07,
      "loss": 0.3298,
      "step": 2195
    },
    {
      "epoch": 0.9113504556752279,
      "grad_norm": 2.511875787240971,
      "learning_rate": 2.376169582763288e-07,
      "loss": 0.3273,
      "step": 2200
    },
    {
      "epoch": 0.9134217067108533,
      "grad_norm": 2.4194421741312078,
      "learning_rate": 2.2672679336518789e-07,
      "loss": 0.3187,
      "step": 2205
    },
    {
      "epoch": 0.9154929577464789,
      "grad_norm": 2.4729733542961516,
      "learning_rate": 2.1608628014502364e-07,
      "loss": 0.3209,
      "step": 2210
    },
    {
      "epoch": 0.9175642087821044,
      "grad_norm": 2.379186052198249,
      "learning_rate": 2.0569597513586004e-07,
      "loss": 0.3229,
      "step": 2215
    },
    {
      "epoch": 0.9196354598177299,
      "grad_norm": 2.3319074686966674,
      "learning_rate": 1.955564217713335e-07,
      "loss": 0.3181,
      "step": 2220
    },
    {
      "epoch": 0.9217067108533554,
      "grad_norm": 2.6517343047945334,
      "learning_rate": 1.8566815037026897e-07,
      "loss": 0.323,
      "step": 2225
    },
    {
      "epoch": 0.923777961888981,
      "grad_norm": 2.232882122217696,
      "learning_rate": 1.7603167810894662e-07,
      "loss": 0.3196,
      "step": 2230
    },
    {
      "epoch": 0.9258492129246064,
      "grad_norm": 2.3988532707568186,
      "learning_rate": 1.6664750899404892e-07,
      "loss": 0.3183,
      "step": 2235
    },
    {
      "epoch": 0.927920463960232,
      "grad_norm": 2.3744099648372403,
      "learning_rate": 1.5751613383630128e-07,
      "loss": 0.3181,
      "step": 2240
    },
    {
      "epoch": 0.9299917149958575,
      "grad_norm": 2.419787212184274,
      "learning_rate": 1.4863803022480362e-07,
      "loss": 0.3244,
      "step": 2245
    },
    {
      "epoch": 0.932062966031483,
      "grad_norm": 2.3333109862483856,
      "learning_rate": 1.4001366250204762e-07,
      "loss": 0.325,
      "step": 2250
    },
    {
      "epoch": 0.9341342170671085,
      "grad_norm": 2.5299591194957505,
      "learning_rate": 1.3164348173963392e-07,
      "loss": 0.3251,
      "step": 2255
    },
    {
      "epoch": 0.9362054681027341,
      "grad_norm": 2.415328780928961,
      "learning_rate": 1.235279257146804e-07,
      "loss": 0.3317,
      "step": 2260
    },
    {
      "epoch": 0.9382767191383595,
      "grad_norm": 2.39855625868432,
      "learning_rate": 1.1566741888692168e-07,
      "loss": 0.3106,
      "step": 2265
    },
    {
      "epoch": 0.9403479701739851,
      "grad_norm": 2.51301146564224,
      "learning_rate": 1.080623723765134e-07,
      "loss": 0.3207,
      "step": 2270
    },
    {
      "epoch": 0.9424192212096106,
      "grad_norm": 2.5451397411628425,
      "learning_rate": 1.0071318394252849e-07,
      "loss": 0.321,
      "step": 2275
    },
    {
      "epoch": 0.9444904722452361,
      "grad_norm": 2.3238922238777753,
      "learning_rate": 9.362023796215036e-08,
      "loss": 0.3115,
      "step": 2280
    },
    {
      "epoch": 0.9465617232808616,
      "grad_norm": 2.3960161103125595,
      "learning_rate": 8.678390541057512e-08,
      "loss": 0.3334,
      "step": 2285
    },
    {
      "epoch": 0.9486329743164872,
      "grad_norm": 2.455291465574181,
      "learning_rate": 8.020454384160437e-08,
      "loss": 0.321,
      "step": 2290
    },
    {
      "epoch": 0.9507042253521126,
      "grad_norm": 2.313389085031741,
      "learning_rate": 7.388249736894615e-08,
      "loss": 0.3155,
      "step": 2295
    },
    {
      "epoch": 0.9527754763877382,
      "grad_norm": 2.4859967083644507,
      "learning_rate": 6.78180966482156e-08,
      "loss": 0.3189,
      "step": 2300
    },
    {
      "epoch": 0.9548467274233637,
      "grad_norm": 2.5150728286365878,
      "learning_rate": 6.201165885964311e-08,
      "loss": 0.3196,
      "step": 2305
    },
    {
      "epoch": 0.9569179784589892,
      "grad_norm": 2.5646449395309134,
      "learning_rate": 5.646348769148491e-08,
      "loss": 0.32,
      "step": 2310
    },
    {
      "epoch": 0.9589892294946147,
      "grad_norm": 2.4648895794572145,
      "learning_rate": 5.117387332413737e-08,
      "loss": 0.3142,
      "step": 2315
    },
    {
      "epoch": 0.9610604805302403,
      "grad_norm": 2.4075943066752292,
      "learning_rate": 4.6143092414961396e-08,
      "loss": 0.3275,
      "step": 2320
    },
    {
      "epoch": 0.9631317315658657,
      "grad_norm": 2.5717189064452604,
      "learning_rate": 4.1371408083815126e-08,
      "loss": 0.3326,
      "step": 2325
    },
    {
      "epoch": 0.9652029826014913,
      "grad_norm": 2.2405081152048982,
      "learning_rate": 3.685906989928656e-08,
      "loss": 0.3198,
      "step": 2330
    },
    {
      "epoch": 0.9672742336371168,
      "grad_norm": 2.493323605860437,
      "learning_rate": 3.2606313865646276e-08,
      "loss": 0.3158,
      "step": 2335
    },
    {
      "epoch": 0.9693454846727423,
      "grad_norm": 2.3221489971336307,
      "learning_rate": 2.861336241050061e-08,
      "loss": 0.3192,
      "step": 2340
    },
    {
      "epoch": 0.9714167357083678,
      "grad_norm": 2.224686624765676,
      "learning_rate": 2.488042437315985e-08,
      "loss": 0.3144,
      "step": 2345
    },
    {
      "epoch": 0.9734879867439934,
      "grad_norm": 2.517150826241084,
      "learning_rate": 2.1407694993714755e-08,
      "loss": 0.3201,
      "step": 2350
    },
    {
      "epoch": 0.9755592377796188,
      "grad_norm": 2.4950963422344126,
      "learning_rate": 1.8195355902824174e-08,
      "loss": 0.3094,
      "step": 2355
    },
    {
      "epoch": 0.9776304888152444,
      "grad_norm": 2.3719000652949376,
      "learning_rate": 1.5243575112218744e-08,
      "loss": 0.3069,
      "step": 2360
    },
    {
      "epoch": 0.97970173985087,
      "grad_norm": 2.532095428796904,
      "learning_rate": 1.2552507005909042e-08,
      "loss": 0.317,
      "step": 2365
    },
    {
      "epoch": 0.9817729908864954,
      "grad_norm": 2.508218831541988,
      "learning_rate": 1.0122292332114814e-08,
      "loss": 0.3214,
      "step": 2370
    },
    {
      "epoch": 0.9838442419221209,
      "grad_norm": 2.3230365500226156,
      "learning_rate": 7.953058195900864e-09,
      "loss": 0.3214,
      "step": 2375
    },
    {
      "epoch": 0.9859154929577465,
      "grad_norm": 2.3842329300121565,
      "learning_rate": 6.044918052531268e-09,
      "loss": 0.3267,
      "step": 2380
    },
    {
      "epoch": 0.987986743993372,
      "grad_norm": 2.450379118891722,
      "learning_rate": 4.397971701533554e-09,
      "loss": 0.3231,
      "step": 2385
    },
    {
      "epoch": 0.9900579950289975,
      "grad_norm": 2.3274963425198423,
      "learning_rate": 3.0123052814812203e-09,
      "loss": 0.3284,
      "step": 2390
    },
    {
      "epoch": 0.992129246064623,
      "grad_norm": 2.4027725490076075,
      "learning_rate": 1.887991265486222e-09,
      "loss": 0.309,
      "step": 2395
    },
    {
      "epoch": 0.9942004971002486,
      "grad_norm": 2.5888360240441015,
      "learning_rate": 1.025088457409229e-09,
      "loss": 0.3177,
      "step": 2400
    },
    {
      "epoch": 0.996271748135874,
      "grad_norm": 2.5018345143420917,
      "learning_rate": 4.2364198878597216e-10,
      "loss": 0.3297,
      "step": 2405
    },
    {
      "epoch": 0.9983429991714996,
      "grad_norm": 2.22608195027372,
      "learning_rate": 8.368331646302353e-11,
      "loss": 0.3161,
      "step": 2410
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.28806865215301514,
      "eval_runtime": 1.191,
      "eval_samples_per_second": 2.519,
      "eval_steps_per_second": 0.84,
      "step": 2414
    },
    {
      "epoch": 1.0,
      "step": 2414,
      "total_flos": 252721244405760.0,
      "train_loss": 0.5156875643339054,
      "train_runtime": 24515.4963,
      "train_samples_per_second": 1.575,
      "train_steps_per_second": 0.098
    }
  ],
  "logging_steps": 5,
  "max_steps": 2414,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 252721244405760.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}