{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.23539970133662894,
  "eval_steps": 500,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 30.342304229736328,
      "learning_rate": 5e-06,
      "loss": 16.1136,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 24.945159912109375,
      "learning_rate": 1e-05,
      "loss": 15.6196,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 16.91960906982422,
      "learning_rate": 1.5e-05,
      "loss": 14.2988,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 13.145479202270508,
      "learning_rate": 2e-05,
      "loss": 12.62,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.406893730163574,
      "learning_rate": 2.5e-05,
      "loss": 11.1458,
      "step": 25
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.820833206176758,
      "learning_rate": 3e-05,
      "loss": 10.3118,
      "step": 30
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.9837632179260254,
      "learning_rate": 3.5e-05,
      "loss": 9.7074,
      "step": 35
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.2150299549102783,
      "learning_rate": 4e-05,
      "loss": 9.3075,
      "step": 40
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.4824776649475098,
      "learning_rate": 4.5e-05,
      "loss": 9.009,
      "step": 45
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.035980463027954,
      "learning_rate": 5e-05,
      "loss": 8.8959,
      "step": 50
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.1182140111923218,
      "learning_rate": 5.500000000000001e-05,
      "loss": 8.8498,
      "step": 55
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.9272787570953369,
      "learning_rate": 6e-05,
      "loss": 8.7864,
      "step": 60
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7738469839096069,
      "learning_rate": 6.500000000000001e-05,
      "loss": 8.746,
      "step": 65
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.07902991771698,
      "learning_rate": 7e-05,
      "loss": 8.6857,
      "step": 70
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.960619330406189,
      "learning_rate": 7.500000000000001e-05,
      "loss": 8.6806,
      "step": 75
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.2761393785476685,
      "learning_rate": 8e-05,
      "loss": 8.5453,
      "step": 80
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.8235496878623962,
      "learning_rate": 8.5e-05,
      "loss": 8.4478,
      "step": 85
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.1294881105422974,
      "learning_rate": 9e-05,
      "loss": 8.4099,
      "step": 90
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.8258363008499146,
      "learning_rate": 9.5e-05,
      "loss": 8.3111,
      "step": 95
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.0850114822387695,
      "learning_rate": 0.0001,
      "loss": 8.1401,
      "step": 100
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.1717969179153442,
      "learning_rate": 9.999999462732701e-05,
      "loss": 8.0193,
      "step": 105
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.4094077348709106,
      "learning_rate": 9.999997850930916e-05,
      "loss": 7.9089,
      "step": 110
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.7466275691986084,
      "learning_rate": 9.999995164594991e-05,
      "loss": 7.8535,
      "step": 115
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.6117960214614868,
      "learning_rate": 9.999991403725507e-05,
      "loss": 7.6416,
      "step": 120
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.4604828357696533,
      "learning_rate": 9.99998656832327e-05,
      "loss": 7.548,
      "step": 125
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.610573410987854,
      "learning_rate": 9.999980658389317e-05,
      "loss": 7.4099,
      "step": 130
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.906869649887085,
      "learning_rate": 9.999973673924921e-05,
      "loss": 7.3182,
      "step": 135
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.7330366373062134,
      "learning_rate": 9.999965614931583e-05,
      "loss": 7.3278,
      "step": 140
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.3011655807495117,
      "learning_rate": 9.999956481411035e-05,
      "loss": 7.2329,
      "step": 145
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.4379724264144897,
      "learning_rate": 9.999946273365238e-05,
      "loss": 7.0154,
      "step": 150
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.5722861289978027,
      "learning_rate": 9.999934990796387e-05,
      "loss": 7.004,
      "step": 155
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.751556634902954,
      "learning_rate": 9.999922633706906e-05,
      "loss": 6.8289,
      "step": 160
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.822228193283081,
      "learning_rate": 9.999909202099453e-05,
      "loss": 6.8306,
      "step": 165
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.300720453262329,
      "learning_rate": 9.999894695976912e-05,
      "loss": 6.6871,
      "step": 170
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.469855546951294,
      "learning_rate": 9.9998791153424e-05,
      "loss": 6.7028,
      "step": 175
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4493306875228882,
      "learning_rate": 9.999862460199267e-05,
      "loss": 6.6762,
      "step": 180
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3964134454727173,
      "learning_rate": 9.999844730551092e-05,
      "loss": 6.5773,
      "step": 185
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.482433557510376,
      "learning_rate": 9.999825926401686e-05,
      "loss": 6.5673,
      "step": 190
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.41461980342865,
      "learning_rate": 9.999806047755088e-05,
      "loss": 6.4353,
      "step": 195
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5991195440292358,
      "learning_rate": 9.999785094615573e-05,
      "loss": 6.436,
      "step": 200
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.873603105545044,
      "learning_rate": 9.99976306698764e-05,
      "loss": 6.3738,
      "step": 205
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.9190030097961426,
      "learning_rate": 9.999739964876026e-05,
      "loss": 6.3186,
      "step": 210
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9261058568954468,
      "learning_rate": 9.999715788285695e-05,
      "loss": 6.3737,
      "step": 215
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.460803747177124,
      "learning_rate": 9.999690537221844e-05,
      "loss": 6.2825,
      "step": 220
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.8361531496047974,
      "learning_rate": 9.999664211689897e-05,
      "loss": 6.3246,
      "step": 225
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3946270942687988,
      "learning_rate": 9.999636811695513e-05,
      "loss": 6.2914,
      "step": 230
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.054090738296509,
      "learning_rate": 9.99960833724458e-05,
      "loss": 6.1841,
      "step": 235
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6103662252426147,
      "learning_rate": 9.999578788343218e-05,
      "loss": 6.1829,
      "step": 240
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3430672883987427,
      "learning_rate": 9.999548164997778e-05,
      "loss": 6.1341,
      "step": 245
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4969396591186523,
      "learning_rate": 9.999516467214839e-05,
      "loss": 5.9567,
      "step": 250
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.8334124088287354,
      "learning_rate": 9.999483695001214e-05,
      "loss": 6.0918,
      "step": 255
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6682161092758179,
      "learning_rate": 9.999449848363946e-05,
      "loss": 6.0409,
      "step": 260
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9023274183273315,
      "learning_rate": 9.999414927310309e-05,
      "loss": 6.0427,
      "step": 265
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5176289081573486,
      "learning_rate": 9.99937893184781e-05,
      "loss": 5.8567,
      "step": 270
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1307928562164307,
      "learning_rate": 9.99934186198418e-05,
      "loss": 5.8957,
      "step": 275
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.072376251220703,
      "learning_rate": 9.999303717727389e-05,
      "loss": 6.0866,
      "step": 280
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.852491021156311,
      "learning_rate": 9.999264499085633e-05,
      "loss": 6.0052,
      "step": 285
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9469953775405884,
      "learning_rate": 9.999224206067341e-05,
      "loss": 5.9109,
      "step": 290
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7127127647399902,
      "learning_rate": 9.999182838681171e-05,
      "loss": 5.9224,
      "step": 295
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4940752983093262,
      "learning_rate": 9.999140396936016e-05,
      "loss": 5.9318,
      "step": 300
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5373684167861938,
      "learning_rate": 9.999096880840994e-05,
      "loss": 5.7734,
      "step": 305
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2774112224578857,
      "learning_rate": 9.999052290405459e-05,
      "loss": 5.8201,
      "step": 310
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.0223042964935303,
      "learning_rate": 9.999006625638994e-05,
      "loss": 5.8437,
      "step": 315
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.9257631301879883,
      "learning_rate": 9.998959886551411e-05,
      "loss": 5.7747,
      "step": 320
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4398373365402222,
      "learning_rate": 9.998912073152755e-05,
      "loss": 5.8213,
      "step": 325
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7308235168457031,
      "learning_rate": 9.998863185453302e-05,
      "loss": 5.6854,
      "step": 330
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.008397340774536,
      "learning_rate": 9.998813223463557e-05,
      "loss": 5.7672,
      "step": 335
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6252214908599854,
      "learning_rate": 9.998762187194258e-05,
      "loss": 5.7493,
      "step": 340
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.831465244293213,
      "learning_rate": 9.998710076656373e-05,
      "loss": 5.7602,
      "step": 345
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6400951147079468,
      "learning_rate": 9.998656891861102e-05,
      "loss": 5.8053,
      "step": 350
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4044883251190186,
      "learning_rate": 9.998602632819874e-05,
      "loss": 5.7024,
      "step": 355
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5869430303573608,
      "learning_rate": 9.998547299544348e-05,
      "loss": 5.6819,
      "step": 360
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4006961584091187,
      "learning_rate": 9.99849089204642e-05,
      "loss": 5.7766,
      "step": 365
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4306929111480713,
      "learning_rate": 9.998433410338206e-05,
      "loss": 5.7059,
      "step": 370
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7405836582183838,
      "learning_rate": 9.998374854432064e-05,
      "loss": 5.6134,
      "step": 375
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2912834882736206,
      "learning_rate": 9.998315224340577e-05,
      "loss": 5.6073,
      "step": 380
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2825273275375366,
      "learning_rate": 9.99825452007656e-05,
      "loss": 5.649,
      "step": 385
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4239997863769531,
      "learning_rate": 9.998192741653057e-05,
      "loss": 5.6027,
      "step": 390
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7178236246109009,
      "learning_rate": 9.998129889083346e-05,
      "loss": 5.4732,
      "step": 395
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7059869766235352,
      "learning_rate": 9.998065962380936e-05,
      "loss": 5.545,
      "step": 400
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4597516059875488,
      "learning_rate": 9.998000961559562e-05,
      "loss": 5.5552,
      "step": 405
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7380889654159546,
      "learning_rate": 9.997934886633195e-05,
      "loss": 5.5077,
      "step": 410
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7295619249343872,
      "learning_rate": 9.997867737616035e-05,
      "loss": 5.5829,
      "step": 415
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6434414386749268,
      "learning_rate": 9.997799514522513e-05,
      "loss": 5.4583,
      "step": 420
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.8824145793914795,
      "learning_rate": 9.99773021736729e-05,
      "loss": 5.481,
      "step": 425
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5333040952682495,
      "learning_rate": 9.99765984616526e-05,
      "loss": 5.505,
      "step": 430
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3971092700958252,
      "learning_rate": 9.997588400931542e-05,
      "loss": 5.4592,
      "step": 435
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.2320480346679688,
      "learning_rate": 9.997515881681496e-05,
      "loss": 5.4835,
      "step": 440
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.052046060562134,
      "learning_rate": 9.9974422884307e-05,
      "loss": 5.3817,
      "step": 445
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.453912615776062,
      "learning_rate": 9.997367621194977e-05,
      "loss": 5.5673,
      "step": 450
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3289484977722168,
      "learning_rate": 9.99729187999037e-05,
      "loss": 5.4446,
      "step": 455
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7679824829101562,
      "learning_rate": 9.997215064833155e-05,
      "loss": 5.4021,
      "step": 460
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.6922376155853271,
      "learning_rate": 9.99713717573984e-05,
      "loss": 5.4794,
      "step": 465
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4465811252593994,
      "learning_rate": 9.997058212727169e-05,
      "loss": 5.4322,
      "step": 470
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2094805240631104,
      "learning_rate": 9.996978175812105e-05,
      "loss": 5.393,
      "step": 475
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1705032587051392,
      "learning_rate": 9.996897065011852e-05,
      "loss": 5.4092,
      "step": 480
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.0050983428955078,
      "learning_rate": 9.996814880343842e-05,
      "loss": 5.3538,
      "step": 485
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.7262133359909058,
      "learning_rate": 9.996731621825736e-05,
      "loss": 5.4662,
      "step": 490
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5162723064422607,
      "learning_rate": 9.996647289475426e-05,
      "loss": 5.407,
      "step": 495
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4311221837997437,
      "learning_rate": 9.996561883311036e-05,
      "loss": 5.3465,
      "step": 500
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.5752923488616943,
      "learning_rate": 9.996475403350921e-05,
      "loss": 5.3002,
      "step": 505
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4409325122833252,
      "learning_rate": 9.996387849613667e-05,
      "loss": 5.3934,
      "step": 510
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5187398195266724,
      "learning_rate": 9.996299222118086e-05,
      "loss": 5.3454,
      "step": 515
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5572065114974976,
      "learning_rate": 9.99620952088323e-05,
      "loss": 5.2132,
      "step": 520
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5349783897399902,
      "learning_rate": 9.996118745928373e-05,
      "loss": 5.3789,
      "step": 525
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.250296711921692,
      "learning_rate": 9.996026897273024e-05,
      "loss": 5.212,
      "step": 530
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4555943012237549,
      "learning_rate": 9.995933974936922e-05,
      "loss": 5.2334,
      "step": 535
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3377940654754639,
      "learning_rate": 9.995839978940036e-05,
      "loss": 5.2216,
      "step": 540
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4281011819839478,
      "learning_rate": 9.995744909302567e-05,
      "loss": 5.3381,
      "step": 545
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2415715456008911,
      "learning_rate": 9.995648766044947e-05,
      "loss": 5.2049,
      "step": 550
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.664896845817566,
      "learning_rate": 9.995551549187836e-05,
      "loss": 5.1661,
      "step": 555
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6059741973876953,
      "learning_rate": 9.995453258752127e-05,
      "loss": 5.2048,
      "step": 560
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.733182668685913,
      "learning_rate": 9.995353894758945e-05,
      "loss": 5.1547,
      "step": 565
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.7354639768600464,
      "learning_rate": 9.995253457229643e-05,
      "loss": 5.1943,
      "step": 570
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.958396315574646,
      "learning_rate": 9.995151946185804e-05,
      "loss": 5.2572,
      "step": 575
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3888930082321167,
      "learning_rate": 9.995049361649247e-05,
      "loss": 5.1437,
      "step": 580
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4347337484359741,
      "learning_rate": 9.994945703642013e-05,
      "loss": 5.1534,
      "step": 585
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2853283882141113,
      "learning_rate": 9.994840972186386e-05,
      "loss": 5.2076,
      "step": 590
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6410499811172485,
      "learning_rate": 9.994735167304866e-05,
      "loss": 5.082,
      "step": 595
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.413332223892212,
      "learning_rate": 9.994628289020195e-05,
      "loss": 5.08,
      "step": 600
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2463523149490356,
      "learning_rate": 9.994520337355344e-05,
      "loss": 5.0535,
      "step": 605
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3588193655014038,
      "learning_rate": 9.994411312333507e-05,
      "loss": 5.1552,
      "step": 610
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1142165660858154,
      "learning_rate": 9.99430121397812e-05,
      "loss": 5.1473,
      "step": 615
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2580281496047974,
      "learning_rate": 9.994190042312839e-05,
      "loss": 5.0683,
      "step": 620
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2235815525054932,
      "learning_rate": 9.994077797361559e-05,
      "loss": 5.0866,
      "step": 625
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.139379858970642,
      "learning_rate": 9.9939644791484e-05,
      "loss": 5.1228,
      "step": 630
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.6412758827209473,
      "learning_rate": 9.993850087697716e-05,
      "loss": 5.0153,
      "step": 635
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4560657739639282,
      "learning_rate": 9.99373462303409e-05,
      "loss": 5.0367,
      "step": 640
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.8065565824508667,
      "learning_rate": 9.993618085182337e-05,
      "loss": 5.134,
      "step": 645
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.8256549835205078,
      "learning_rate": 9.993500474167501e-05,
      "loss": 5.1075,
      "step": 650
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3655732870101929,
      "learning_rate": 9.99338179001486e-05,
      "loss": 5.1308,
      "step": 655
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2854615449905396,
      "learning_rate": 9.993262032749914e-05,
      "loss": 5.1831,
      "step": 660
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1920758485794067,
      "learning_rate": 9.993141202398407e-05,
      "loss": 5.1133,
      "step": 665
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.388599157333374,
      "learning_rate": 9.993019298986301e-05,
      "loss": 5.0677,
      "step": 670
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5200777053833008,
      "learning_rate": 9.992896322539796e-05,
      "loss": 5.0587,
      "step": 675
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1357358694076538,
      "learning_rate": 9.992772273085319e-05,
      "loss": 5.2216,
      "step": 680
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3757842779159546,
      "learning_rate": 9.99264715064953e-05,
      "loss": 5.0554,
      "step": 685
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3100160360336304,
      "learning_rate": 9.992520955259319e-05,
      "loss": 5.1055,
      "step": 690
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0510196685791016,
      "learning_rate": 9.992393686941808e-05,
      "loss": 5.079,
      "step": 695
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.5555288791656494,
      "learning_rate": 9.992265345724344e-05,
      "loss": 5.0962,
      "step": 700
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.9240226745605469,
      "learning_rate": 9.99213593163451e-05,
      "loss": 5.0331,
      "step": 705
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.22459077835083,
      "learning_rate": 9.992005444700121e-05,
      "loss": 5.0045,
      "step": 710
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.268713116645813,
      "learning_rate": 9.991873884949214e-05,
      "loss": 4.9937,
      "step": 715
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2946181297302246,
      "learning_rate": 9.991741252410067e-05,
      "loss": 4.961,
      "step": 720
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1600682735443115,
      "learning_rate": 9.99160754711118e-05,
      "loss": 4.8899,
      "step": 725
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1104838848114014,
      "learning_rate": 9.99147276908129e-05,
      "loss": 4.9027,
      "step": 730
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2879794836044312,
      "learning_rate": 9.991336918349361e-05,
      "loss": 4.952,
      "step": 735
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3048343658447266,
      "learning_rate": 9.991199994944586e-05,
      "loss": 5.0505,
      "step": 740
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0726879835128784,
      "learning_rate": 9.991061998896393e-05,
      "loss": 4.9215,
      "step": 745
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4467452764511108,
      "learning_rate": 9.99092293023444e-05,
      "loss": 4.993,
      "step": 750
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2162690162658691,
      "learning_rate": 9.990782788988611e-05,
      "loss": 4.9725,
      "step": 755
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.302495002746582,
      "learning_rate": 9.990641575189022e-05,
      "loss": 5.0366,
      "step": 760
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1818315982818604,
      "learning_rate": 9.990499288866024e-05,
      "loss": 5.0003,
      "step": 765
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4780274629592896,
      "learning_rate": 9.990355930050194e-05,
      "loss": 4.9482,
      "step": 770
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1718212366104126,
      "learning_rate": 9.990211498772341e-05,
      "loss": 4.9948,
      "step": 775
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3258137702941895,
      "learning_rate": 9.990065995063506e-05,
      "loss": 4.9081,
      "step": 780
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.2919809818267822,
      "learning_rate": 9.989919418954956e-05,
      "loss": 4.8432,
      "step": 785
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4691888093948364,
      "learning_rate": 9.989771770478191e-05,
      "loss": 4.7886,
      "step": 790
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3927934169769287,
      "learning_rate": 9.989623049664944e-05,
      "loss": 4.9309,
      "step": 795
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3581048250198364,
      "learning_rate": 9.989473256547175e-05,
      "loss": 4.9269,
      "step": 800
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1874853372573853,
      "learning_rate": 9.989322391157076e-05,
      "loss": 4.8723,
      "step": 805
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.1008710861206055,
      "learning_rate": 9.989170453527068e-05,
      "loss": 5.0152,
      "step": 810
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0841079950332642,
      "learning_rate": 9.989017443689805e-05,
      "loss": 4.8612,
      "step": 815
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0537878274917603,
      "learning_rate": 9.988863361678169e-05,
      "loss": 4.8228,
      "step": 820
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3282699584960938,
      "learning_rate": 9.988708207525272e-05,
      "loss": 4.9457,
      "step": 825
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.3590075969696045,
      "learning_rate": 9.98855198126446e-05,
      "loss": 4.8463,
      "step": 830
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.1755837202072144,
      "learning_rate": 9.988394682929305e-05,
      "loss": 4.865,
      "step": 835
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.4010895490646362,
      "learning_rate": 9.988236312553614e-05,
      "loss": 4.9516,
      "step": 840
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0931832790374756,
      "learning_rate": 9.988076870171419e-05,
      "loss": 4.9097,
      "step": 845
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1834778785705566,
      "learning_rate": 9.987916355816988e-05,
      "loss": 4.8992,
      "step": 850
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0853551626205444,
      "learning_rate": 9.987754769524814e-05,
      "loss": 4.8215,
      "step": 855
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1411994695663452,
      "learning_rate": 9.987592111329626e-05,
      "loss": 4.8857,
      "step": 860
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.2159109115600586,
      "learning_rate": 9.987428381266377e-05,
      "loss": 4.9341,
      "step": 865
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5336378812789917,
      "learning_rate": 9.987263579370257e-05,
      "loss": 4.9056,
      "step": 870
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4192923307418823,
      "learning_rate": 9.987097705676681e-05,
      "loss": 4.8045,
      "step": 875
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.271614909172058,
      "learning_rate": 9.986930760221296e-05,
      "loss": 4.913,
      "step": 880
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.2139419317245483,
      "learning_rate": 9.986762743039983e-05,
      "loss": 4.8847,
      "step": 885
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.261183500289917,
      "learning_rate": 9.986593654168846e-05,
      "loss": 4.7993,
      "step": 890
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0047696828842163,
      "learning_rate": 9.986423493644225e-05,
      "loss": 4.8312,
      "step": 895
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4076634645462036,
      "learning_rate": 9.98625226150269e-05,
      "loss": 4.7532,
      "step": 900
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0993361473083496,
      "learning_rate": 9.986079957781038e-05,
      "loss": 4.7618,
      "step": 905
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0709277391433716,
      "learning_rate": 9.9859065825163e-05,
      "loss": 4.7713,
      "step": 910
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1873985528945923,
      "learning_rate": 9.985732135745733e-05,
      "loss": 4.7592,
      "step": 915
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.012856364250183,
      "learning_rate": 9.985556617506828e-05,
      "loss": 4.7884,
      "step": 920
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1248955726623535,
      "learning_rate": 9.985380027837308e-05,
      "loss": 4.8139,
      "step": 925
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3157622814178467,
      "learning_rate": 9.98520236677512e-05,
      "loss": 4.7854,
      "step": 930
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.096008539199829,
      "learning_rate": 9.985023634358443e-05,
      "loss": 4.7017,
      "step": 935
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4006218910217285,
      "learning_rate": 9.984843830625691e-05,
      "loss": 4.8483,
      "step": 940
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.4066104888916016,
      "learning_rate": 9.984662955615507e-05,
      "loss": 4.7217,
      "step": 945
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1498020887374878,
      "learning_rate": 9.984481009366758e-05,
      "loss": 4.9245,
      "step": 950
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.09044349193573,
      "learning_rate": 9.984297991918548e-05,
      "loss": 4.8242,
      "step": 955
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0431567430496216,
      "learning_rate": 9.984113903310206e-05,
      "loss": 4.8223,
      "step": 960
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9916888475418091,
      "learning_rate": 9.983928743581296e-05,
      "loss": 4.6107,
      "step": 965
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0403705835342407,
      "learning_rate": 9.983742512771611e-05,
      "loss": 4.7688,
      "step": 970
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1474281549453735,
      "learning_rate": 9.983555210921171e-05,
      "loss": 4.7933,
      "step": 975
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1478239297866821,
      "learning_rate": 9.983366838070231e-05,
      "loss": 4.7062,
      "step": 980
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.103904128074646,
      "learning_rate": 9.983177394259271e-05,
      "loss": 4.6975,
      "step": 985
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.066270112991333,
      "learning_rate": 9.982986879529006e-05,
      "loss": 4.7508,
      "step": 990
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1204413175582886,
      "learning_rate": 9.982795293920377e-05,
      "loss": 4.7611,
      "step": 995
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1344144344329834,
      "learning_rate": 9.982602637474558e-05,
      "loss": 4.7102,
      "step": 1000
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3472899198532104,
      "learning_rate": 9.982408910232952e-05,
      "loss": 4.6872,
      "step": 1005
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0646827220916748,
      "learning_rate": 9.982214112237193e-05,
      "loss": 4.5932,
      "step": 1010
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1303422451019287,
      "learning_rate": 9.982018243529146e-05,
      "loss": 4.8744,
      "step": 1015
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3574433326721191,
      "learning_rate": 9.981821304150901e-05,
      "loss": 4.7773,
      "step": 1020
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.232305884361267,
      "learning_rate": 9.981623294144782e-05,
      "loss": 4.699,
      "step": 1025
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.256428599357605,
      "learning_rate": 9.981424213553344e-05,
      "loss": 4.7425,
      "step": 1030
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0305962562561035,
      "learning_rate": 9.981224062419372e-05,
      "loss": 4.7519,
      "step": 1035
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9948828220367432,
      "learning_rate": 9.981022840785877e-05,
      "loss": 4.6761,
      "step": 1040
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.095750331878662,
      "learning_rate": 9.980820548696106e-05,
      "loss": 4.7341,
      "step": 1045
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0619468688964844,
      "learning_rate": 9.98061718619353e-05,
      "loss": 4.7778,
      "step": 1050
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9546065330505371,
      "learning_rate": 9.980412753321856e-05,
      "loss": 4.6604,
      "step": 1055
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9763450622558594,
      "learning_rate": 9.980207250125015e-05,
      "loss": 4.7935,
      "step": 1060
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9340956807136536,
      "learning_rate": 9.980000676647174e-05,
      "loss": 4.556,
      "step": 1065
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.092559576034546,
      "learning_rate": 9.979793032932725e-05,
      "loss": 4.7046,
      "step": 1070
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.2036818265914917,
      "learning_rate": 9.979584319026292e-05,
      "loss": 4.6607,
      "step": 1075
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0511127710342407,
      "learning_rate": 9.97937453497273e-05,
      "loss": 4.6717,
      "step": 1080
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1158552169799805,
      "learning_rate": 9.979163680817124e-05,
      "loss": 4.6354,
      "step": 1085
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1708827018737793,
      "learning_rate": 9.978951756604786e-05,
      "loss": 4.591,
      "step": 1090
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0847101211547852,
      "learning_rate": 9.978738762381262e-05,
      "loss": 4.674,
      "step": 1095
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1602849960327148,
      "learning_rate": 9.978524698192322e-05,
      "loss": 4.7425,
      "step": 1100
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9692180752754211,
      "learning_rate": 9.978309564083975e-05,
      "loss": 4.698,
      "step": 1105
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1480087041854858,
      "learning_rate": 9.97809336010245e-05,
      "loss": 4.8214,
      "step": 1110
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0540591478347778,
      "learning_rate": 9.977876086294216e-05,
      "loss": 4.6241,
      "step": 1115
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9855077862739563,
      "learning_rate": 9.977657742705962e-05,
      "loss": 4.6663,
      "step": 1120
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0006612539291382,
      "learning_rate": 9.977438329384614e-05,
      "loss": 4.7089,
      "step": 1125
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9681260585784912,
      "learning_rate": 9.977217846377324e-05,
      "loss": 4.661,
      "step": 1130
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0346499681472778,
      "learning_rate": 9.976996293731476e-05,
      "loss": 4.6153,
      "step": 1135
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1809090375900269,
      "learning_rate": 9.976773671494684e-05,
      "loss": 4.6053,
      "step": 1140
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9507768154144287,
      "learning_rate": 9.976549979714791e-05,
      "loss": 4.7405,
      "step": 1145
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0678818225860596,
      "learning_rate": 9.976325218439868e-05,
      "loss": 4.6906,
      "step": 1150
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0061429738998413,
      "learning_rate": 9.976099387718218e-05,
      "loss": 4.5844,
      "step": 1155
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9220618605613708,
      "learning_rate": 9.975872487598377e-05,
      "loss": 4.7097,
      "step": 1160
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.068864107131958,
      "learning_rate": 9.975644518129104e-05,
      "loss": 4.5737,
      "step": 1165
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.010648250579834,
      "learning_rate": 9.975415479359394e-05,
      "loss": 4.746,
      "step": 1170
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.037503719329834,
      "learning_rate": 9.975185371338464e-05,
      "loss": 4.6889,
      "step": 1175
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.030385136604309,
      "learning_rate": 9.97495419411577e-05,
      "loss": 4.7419,
      "step": 1180
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0794814825057983,
      "learning_rate": 9.974721947740995e-05,
      "loss": 4.6252,
      "step": 1185
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1710423231124878,
      "learning_rate": 9.974488632264046e-05,
      "loss": 4.717,
      "step": 1190
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1813603639602661,
      "learning_rate": 9.974254247735066e-05,
      "loss": 4.6835,
      "step": 1195
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9003679752349854,
      "learning_rate": 9.974018794204426e-05,
      "loss": 4.5403,
      "step": 1200
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0389988422393799,
      "learning_rate": 9.973782271722727e-05,
      "loss": 4.472,
      "step": 1205
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.157646894454956,
      "learning_rate": 9.973544680340799e-05,
      "loss": 4.5818,
      "step": 1210
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9575796127319336,
      "learning_rate": 9.973306020109701e-05,
      "loss": 4.6471,
      "step": 1215
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9345293641090393,
      "learning_rate": 9.973066291080724e-05,
      "loss": 4.5219,
      "step": 1220
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9667677879333496,
      "learning_rate": 9.972825493305387e-05,
      "loss": 4.5209,
      "step": 1225
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9334149956703186,
      "learning_rate": 9.97258362683544e-05,
      "loss": 4.5731,
      "step": 1230
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.981364905834198,
      "learning_rate": 9.972340691722858e-05,
      "loss": 4.6526,
      "step": 1235
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9291808605194092,
      "learning_rate": 9.972096688019856e-05,
      "loss": 4.5769,
      "step": 1240
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2989933490753174,
      "learning_rate": 9.971851615778868e-05,
      "loss": 4.5498,
      "step": 1245
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0652649402618408,
      "learning_rate": 9.971605475052561e-05,
      "loss": 4.5796,
      "step": 1250
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0848689079284668,
      "learning_rate": 9.971358265893834e-05,
      "loss": 4.6682,
      "step": 1255
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0221582651138306,
      "learning_rate": 9.971109988355814e-05,
      "loss": 4.6457,
      "step": 1260
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9533518552780151,
      "learning_rate": 9.970860642491854e-05,
      "loss": 4.5238,
      "step": 1265
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0245946645736694,
      "learning_rate": 9.970610228355548e-05,
      "loss": 4.6149,
      "step": 1270
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9755035042762756,
      "learning_rate": 9.970358746000704e-05,
      "loss": 4.5613,
      "step": 1275
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.917687177658081,
      "learning_rate": 9.97010619548137e-05,
      "loss": 4.5225,
      "step": 1280
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9751790761947632,
      "learning_rate": 9.969852576851822e-05,
      "loss": 4.5566,
      "step": 1285
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0446383953094482,
      "learning_rate": 9.969597890166565e-05,
      "loss": 4.6131,
      "step": 1290
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9377387762069702,
      "learning_rate": 9.969342135480328e-05,
      "loss": 4.7582,
      "step": 1295
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.826242208480835,
      "learning_rate": 9.969085312848081e-05,
      "loss": 4.503,
      "step": 1300
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1156399250030518,
      "learning_rate": 9.968827422325012e-05,
      "loss": 4.5403,
      "step": 1305
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9961814880371094,
      "learning_rate": 9.968568463966545e-05,
      "loss": 4.6479,
      "step": 1310
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9816222786903381,
      "learning_rate": 9.968308437828333e-05,
      "loss": 4.5221,
      "step": 1315
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9534478783607483,
      "learning_rate": 9.968047343966256e-05,
      "loss": 4.4231,
      "step": 1320
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9924973249435425,
      "learning_rate": 9.967785182436426e-05,
      "loss": 4.6264,
      "step": 1325
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.121100902557373,
      "learning_rate": 9.967521953295185e-05,
      "loss": 4.4787,
      "step": 1330
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0099316835403442,
      "learning_rate": 9.967257656599099e-05,
      "loss": 4.5235,
      "step": 1335
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8959918022155762,
      "learning_rate": 9.966992292404967e-05,
      "loss": 4.519,
      "step": 1340
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.043068289756775,
      "learning_rate": 9.966725860769821e-05,
      "loss": 4.5237,
      "step": 1345
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9585849046707153,
      "learning_rate": 9.966458361750916e-05,
      "loss": 4.452,
      "step": 1350
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8837548494338989,
      "learning_rate": 9.966189795405743e-05,
      "loss": 4.4559,
      "step": 1355
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9389224648475647,
      "learning_rate": 9.965920161792016e-05,
      "loss": 4.5855,
      "step": 1360
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.914502739906311,
      "learning_rate": 9.96564946096768e-05,
      "loss": 4.5115,
      "step": 1365
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9476872682571411,
      "learning_rate": 9.965377692990914e-05,
      "loss": 4.5574,
      "step": 1370
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.907654345035553,
      "learning_rate": 9.965104857920121e-05,
      "loss": 4.499,
      "step": 1375
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2423409223556519,
      "learning_rate": 9.964830955813935e-05,
      "loss": 4.5499,
      "step": 1380
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0977503061294556,
      "learning_rate": 9.964555986731219e-05,
      "loss": 4.5591,
      "step": 1385
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0124132633209229,
      "learning_rate": 9.964279950731066e-05,
      "loss": 4.449,
      "step": 1390
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1032521724700928,
      "learning_rate": 9.9640028478728e-05,
      "loss": 4.4399,
      "step": 1395
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9310466647148132,
      "learning_rate": 9.963724678215968e-05,
      "loss": 4.5857,
      "step": 1400
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8802070021629333,
      "learning_rate": 9.963445441820356e-05,
      "loss": 4.5798,
      "step": 1405
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8232974410057068,
      "learning_rate": 9.963165138745969e-05,
      "loss": 4.4769,
      "step": 1410
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8830395340919495,
      "learning_rate": 9.962883769053051e-05,
      "loss": 4.4153,
      "step": 1415
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8238136172294617,
      "learning_rate": 9.962601332802065e-05,
      "loss": 4.498,
      "step": 1420
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8898188471794128,
      "learning_rate": 9.962317830053711e-05,
      "loss": 4.5213,
      "step": 1425
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9314691424369812,
      "learning_rate": 9.962033260868916e-05,
      "loss": 4.4017,
      "step": 1430
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9342561364173889,
      "learning_rate": 9.961747625308838e-05,
      "loss": 4.5264,
      "step": 1435
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9425392746925354,
      "learning_rate": 9.961460923434858e-05,
      "loss": 4.4474,
      "step": 1440
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.036442756652832,
      "learning_rate": 9.961173155308591e-05,
      "loss": 4.4184,
      "step": 1445
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9934285879135132,
      "learning_rate": 9.960884320991881e-05,
      "loss": 4.5285,
      "step": 1450
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2279188632965088,
      "learning_rate": 9.960594420546802e-05,
      "loss": 4.3292,
      "step": 1455
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1246490478515625,
      "learning_rate": 9.960303454035654e-05,
      "loss": 4.5094,
      "step": 1460
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1780115365982056,
      "learning_rate": 9.960011421520969e-05,
      "loss": 4.4499,
      "step": 1465
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.2334381341934204,
      "learning_rate": 9.959718323065503e-05,
      "loss": 4.3353,
      "step": 1470
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.060360074043274,
      "learning_rate": 9.959424158732252e-05,
      "loss": 4.6173,
      "step": 1475
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8194041848182678,
      "learning_rate": 9.959128928584426e-05,
      "loss": 4.4082,
      "step": 1480
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0778460502624512,
      "learning_rate": 9.958832632685478e-05,
      "loss": 4.3102,
      "step": 1485
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9119295477867126,
      "learning_rate": 9.95853527109908e-05,
      "loss": 4.4194,
      "step": 1490
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9570429921150208,
      "learning_rate": 9.95823684388914e-05,
      "loss": 4.4514,
      "step": 1495
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9421054124832153,
      "learning_rate": 9.95793735111979e-05,
      "loss": 4.3578,
      "step": 1500
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.89156174659729,
      "learning_rate": 9.957636792855394e-05,
      "loss": 4.4693,
      "step": 1505
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8912834525108337,
      "learning_rate": 9.957335169160545e-05,
      "loss": 4.4528,
      "step": 1510
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9975301027297974,
      "learning_rate": 9.957032480100061e-05,
      "loss": 4.3657,
      "step": 1515
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9152092933654785,
      "learning_rate": 9.956728725738996e-05,
      "loss": 4.3876,
      "step": 1520
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8747090697288513,
      "learning_rate": 9.956423906142626e-05,
      "loss": 4.4496,
      "step": 1525
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8948420286178589,
      "learning_rate": 9.956118021376461e-05,
      "loss": 4.3097,
      "step": 1530
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8913704752922058,
      "learning_rate": 9.955811071506236e-05,
      "loss": 4.4389,
      "step": 1535
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9597818851470947,
      "learning_rate": 9.955503056597917e-05,
      "loss": 4.4291,
      "step": 1540
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9380824565887451,
      "learning_rate": 9.955193976717698e-05,
      "loss": 4.5208,
      "step": 1545
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9058547616004944,
      "learning_rate": 9.954883831932003e-05,
      "loss": 4.428,
      "step": 1550
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9552749991416931,
      "learning_rate": 9.954572622307485e-05,
      "loss": 4.4769,
      "step": 1555
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9138779044151306,
      "learning_rate": 9.954260347911025e-05,
      "loss": 4.4486,
      "step": 1560
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8599411249160767,
      "learning_rate": 9.953947008809731e-05,
      "loss": 4.4214,
      "step": 1565
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.799152135848999,
      "learning_rate": 9.953632605070944e-05,
      "loss": 4.3476,
      "step": 1570
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7961298823356628,
      "learning_rate": 9.95331713676223e-05,
      "loss": 4.4594,
      "step": 1575
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.972852885723114,
      "learning_rate": 9.953000603951386e-05,
      "loss": 4.4933,
      "step": 1580
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9430463910102844,
      "learning_rate": 9.952683006706438e-05,
      "loss": 4.4243,
      "step": 1585
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9489074945449829,
      "learning_rate": 9.952364345095638e-05,
      "loss": 4.3052,
      "step": 1590
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7771168351173401,
      "learning_rate": 9.95204461918747e-05,
      "loss": 4.4476,
      "step": 1595
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9220024943351746,
      "learning_rate": 9.951723829050646e-05,
      "loss": 4.407,
      "step": 1600
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8158936500549316,
      "learning_rate": 9.951401974754101e-05,
      "loss": 4.3259,
      "step": 1605
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.948128342628479,
      "learning_rate": 9.951079056367013e-05,
      "loss": 4.4767,
      "step": 1610
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9961679577827454,
      "learning_rate": 9.95075507395877e-05,
      "loss": 4.406,
      "step": 1615
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.885222852230072,
      "learning_rate": 9.950430027599003e-05,
      "loss": 4.353,
      "step": 1620
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8581100106239319,
      "learning_rate": 9.950103917357566e-05,
      "loss": 4.3167,
      "step": 1625
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9207460284233093,
      "learning_rate": 9.949776743304542e-05,
      "loss": 4.4399,
      "step": 1630
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9526963233947754,
      "learning_rate": 9.949448505510243e-05,
      "loss": 4.4291,
      "step": 1635
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7979224324226379,
      "learning_rate": 9.94911920404521e-05,
      "loss": 4.4032,
      "step": 1640
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8377354741096497,
      "learning_rate": 9.94878883898021e-05,
      "loss": 4.2944,
      "step": 1645
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8064927458763123,
      "learning_rate": 9.948457410386244e-05,
      "loss": 4.3732,
      "step": 1650
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9221789240837097,
      "learning_rate": 9.948124918334538e-05,
      "loss": 4.3199,
      "step": 1655
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8794273138046265,
      "learning_rate": 9.947791362896543e-05,
      "loss": 4.262,
      "step": 1660
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.826052188873291,
      "learning_rate": 9.947456744143947e-05,
      "loss": 4.2841,
      "step": 1665
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9316311478614807,
      "learning_rate": 9.947121062148658e-05,
      "loss": 4.4079,
      "step": 1670
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8867166638374329,
      "learning_rate": 9.94678431698282e-05,
      "loss": 4.4114,
      "step": 1675
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8845940828323364,
      "learning_rate": 9.946446508718801e-05,
      "loss": 4.3226,
      "step": 1680
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9194801449775696,
      "learning_rate": 9.946107637429196e-05,
      "loss": 4.3153,
      "step": 1685
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9783342480659485,
      "learning_rate": 9.945767703186832e-05,
      "loss": 4.4386,
      "step": 1690
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8650766611099243,
      "learning_rate": 9.945426706064764e-05,
      "loss": 4.2955,
      "step": 1695
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.843528151512146,
      "learning_rate": 9.945084646136275e-05,
      "loss": 4.3029,
      "step": 1700
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8105334639549255,
      "learning_rate": 9.944741523474874e-05,
      "loss": 4.4107,
      "step": 1705
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8164493441581726,
      "learning_rate": 9.944397338154301e-05,
      "loss": 4.3618,
      "step": 1710
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0284205675125122,
      "learning_rate": 9.944052090248526e-05,
      "loss": 4.3867,
      "step": 1715
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8768856525421143,
      "learning_rate": 9.943705779831742e-05,
      "loss": 4.4049,
      "step": 1720
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8829424381256104,
      "learning_rate": 9.943358406978377e-05,
      "loss": 4.3811,
      "step": 1725
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0121821165084839,
      "learning_rate": 9.943009971763082e-05,
      "loss": 4.3873,
      "step": 1730
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.86043381690979,
      "learning_rate": 9.942660474260737e-05,
      "loss": 4.2826,
      "step": 1735
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9295215606689453,
      "learning_rate": 9.942309914546453e-05,
      "loss": 4.2991,
      "step": 1740
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9926019906997681,
      "learning_rate": 9.941958292695565e-05,
      "loss": 4.2777,
      "step": 1745
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9335598349571228,
      "learning_rate": 9.941605608783644e-05,
      "loss": 4.3978,
      "step": 1750
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8112255334854126,
      "learning_rate": 9.94125186288648e-05,
      "loss": 4.3488,
      "step": 1755
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8819780945777893,
      "learning_rate": 9.940897055080098e-05,
      "loss": 4.362,
      "step": 1760
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7420908808708191,
      "learning_rate": 9.940541185440746e-05,
      "loss": 4.2821,
      "step": 1765
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.885562539100647,
      "learning_rate": 9.940184254044904e-05,
      "loss": 4.3117,
      "step": 1770
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9075621366500854,
      "learning_rate": 9.939826260969282e-05,
      "loss": 4.4021,
      "step": 1775
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9036880731582642,
      "learning_rate": 9.93946720629081e-05,
      "loss": 4.4039,
      "step": 1780
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9084277749061584,
      "learning_rate": 9.939107090086654e-05,
      "loss": 4.3635,
      "step": 1785
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9255689382553101,
      "learning_rate": 9.938745912434208e-05,
      "loss": 4.4293,
      "step": 1790
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9867833852767944,
      "learning_rate": 9.938383673411087e-05,
      "loss": 4.3411,
      "step": 1795
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.903582751750946,
      "learning_rate": 9.938020373095141e-05,
      "loss": 4.3567,
      "step": 1800
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8810187578201294,
      "learning_rate": 9.937656011564445e-05,
      "loss": 4.3881,
      "step": 1805
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8416256308555603,
      "learning_rate": 9.937290588897302e-05,
      "loss": 4.2842,
      "step": 1810
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7998962998390198,
      "learning_rate": 9.936924105172248e-05,
      "loss": 4.3901,
      "step": 1815
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8087441921234131,
      "learning_rate": 9.936556560468037e-05,
      "loss": 4.3258,
      "step": 1820
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.4025551080703735,
      "learning_rate": 9.93618795486366e-05,
      "loss": 4.265,
      "step": 1825
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9645698070526123,
      "learning_rate": 9.935818288438334e-05,
      "loss": 4.2497,
      "step": 1830
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8740422129631042,
      "learning_rate": 9.935447561271501e-05,
      "loss": 4.302,
      "step": 1835
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8668932914733887,
      "learning_rate": 9.935075773442835e-05,
      "loss": 4.3326,
      "step": 1840
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8833674192428589,
      "learning_rate": 9.934702925032232e-05,
      "loss": 4.2073,
      "step": 1845
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8068246245384216,
      "learning_rate": 9.934329016119824e-05,
      "loss": 4.2498,
      "step": 1850
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8597808480262756,
      "learning_rate": 9.933954046785963e-05,
      "loss": 4.3479,
      "step": 1855
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8278979063034058,
      "learning_rate": 9.933578017111233e-05,
      "loss": 4.2928,
      "step": 1860
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8309856653213501,
      "learning_rate": 9.933200927176449e-05,
      "loss": 4.2951,
      "step": 1865
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9148632884025574,
      "learning_rate": 9.932822777062646e-05,
      "loss": 4.4142,
      "step": 1870
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8819959759712219,
      "learning_rate": 9.932443566851092e-05,
      "loss": 4.2107,
      "step": 1875
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8736444115638733,
      "learning_rate": 9.932063296623284e-05,
      "loss": 4.2355,
      "step": 1880
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9972934722900391,
      "learning_rate": 9.931681966460944e-05,
      "loss": 4.2829,
      "step": 1885
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.2609264850616455,
      "learning_rate": 9.93129957644602e-05,
      "loss": 4.3043,
      "step": 1890
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8282685875892639,
      "learning_rate": 9.930916126660691e-05,
      "loss": 4.3125,
      "step": 1895
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8143281936645508,
      "learning_rate": 9.930531617187367e-05,
      "loss": 4.3372,
      "step": 1900
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8509781360626221,
      "learning_rate": 9.930146048108678e-05,
      "loss": 4.2629,
      "step": 1905
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.942048192024231,
      "learning_rate": 9.929759419507487e-05,
      "loss": 4.2383,
      "step": 1910
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9580615758895874,
      "learning_rate": 9.92937173146688e-05,
      "loss": 4.2488,
      "step": 1915
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.98505699634552,
      "learning_rate": 9.928982984070179e-05,
      "loss": 4.3019,
      "step": 1920
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9058492183685303,
      "learning_rate": 9.928593177400925e-05,
      "loss": 4.3298,
      "step": 1925
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.023591160774231,
      "learning_rate": 9.928202311542889e-05,
      "loss": 4.3555,
      "step": 1930
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7966316938400269,
      "learning_rate": 9.927810386580075e-05,
      "loss": 4.3547,
      "step": 1935
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7503045797348022,
      "learning_rate": 9.927417402596707e-05,
      "loss": 4.2031,
      "step": 1940
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7294631004333496,
      "learning_rate": 9.927023359677241e-05,
      "loss": 4.2089,
      "step": 1945
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8023434281349182,
      "learning_rate": 9.92662825790636e-05,
      "loss": 4.2852,
      "step": 1950
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8379714488983154,
      "learning_rate": 9.926232097368974e-05,
      "loss": 4.2487,
      "step": 1955
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9109610915184021,
      "learning_rate": 9.925834878150219e-05,
      "loss": 4.2328,
      "step": 1960
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7228615880012512,
      "learning_rate": 9.925436600335464e-05,
      "loss": 4.1495,
      "step": 1965
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.823662281036377,
      "learning_rate": 9.925037264010298e-05,
      "loss": 4.2498,
      "step": 1970
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8762710690498352,
      "learning_rate": 9.924636869260542e-05,
      "loss": 4.2575,
      "step": 1975
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9420381784439087,
      "learning_rate": 9.924235416172243e-05,
      "loss": 4.2382,
      "step": 1980
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8953169584274292,
      "learning_rate": 9.923832904831679e-05,
      "loss": 4.216,
      "step": 1985
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8370233178138733,
      "learning_rate": 9.923429335325349e-05,
      "loss": 4.0623,
      "step": 1990
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8436635136604309,
      "learning_rate": 9.923024707739985e-05,
      "loss": 4.1317,
      "step": 1995
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8946409821510315,
      "learning_rate": 9.922619022162545e-05,
      "loss": 4.2259,
      "step": 2000
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8554151654243469,
      "learning_rate": 9.92221227868021e-05,
      "loss": 4.3721,
      "step": 2005
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7739545106887817,
      "learning_rate": 9.921804477380394e-05,
      "loss": 4.2641,
      "step": 2010
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.93932044506073,
      "learning_rate": 9.921395618350739e-05,
      "loss": 4.2049,
      "step": 2015
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.752130925655365,
      "learning_rate": 9.920985701679106e-05,
      "loss": 4.3533,
      "step": 2020
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8123421669006348,
      "learning_rate": 9.920574727453595e-05,
      "loss": 4.3507,
      "step": 2025
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8544089198112488,
      "learning_rate": 9.920162695762522e-05,
      "loss": 4.246,
      "step": 2030
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0053143501281738,
      "learning_rate": 9.919749606694439e-05,
      "loss": 4.1843,
      "step": 2035
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8623014092445374,
      "learning_rate": 9.919335460338122e-05,
      "loss": 4.0687,
      "step": 2040
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.627628207206726,
      "learning_rate": 9.918920256782572e-05,
      "loss": 4.1736,
      "step": 2045
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8666718006134033,
      "learning_rate": 9.918503996117016e-05,
      "loss": 4.1948,
      "step": 2050
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9772979021072388,
      "learning_rate": 9.918086678430917e-05,
      "loss": 4.1805,
      "step": 2055
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8658679127693176,
      "learning_rate": 9.917668303813959e-05,
      "loss": 4.2882,
      "step": 2060
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8397815227508545,
      "learning_rate": 9.917248872356051e-05,
      "loss": 4.3114,
      "step": 2065
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8537928462028503,
      "learning_rate": 9.916828384147331e-05,
      "loss": 4.1891,
      "step": 2070
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7911549210548401,
      "learning_rate": 9.916406839278167e-05,
      "loss": 4.1786,
      "step": 2075
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0153640508651733,
      "learning_rate": 9.915984237839152e-05,
      "loss": 4.2403,
      "step": 2080
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8811365962028503,
      "learning_rate": 9.915560579921106e-05,
      "loss": 4.1511,
      "step": 2085
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8019070029258728,
      "learning_rate": 9.915135865615074e-05,
      "loss": 4.2289,
      "step": 2090
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8007054924964905,
      "learning_rate": 9.914710095012332e-05,
      "loss": 4.209,
      "step": 2095
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.903588593006134,
      "learning_rate": 9.91428326820438e-05,
      "loss": 4.2377,
      "step": 2100
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7710729837417603,
      "learning_rate": 9.913855385282947e-05,
      "loss": 4.1974,
      "step": 2105
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7649155855178833,
      "learning_rate": 9.913426446339986e-05,
      "loss": 4.1877,
      "step": 2110
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8055683970451355,
      "learning_rate": 9.912996451467682e-05,
      "loss": 4.1663,
      "step": 2115
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7791784405708313,
      "learning_rate": 9.912565400758443e-05,
      "loss": 4.2177,
      "step": 2120
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7775405049324036,
      "learning_rate": 9.912133294304901e-05,
      "loss": 4.2863,
      "step": 2125
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9319812655448914,
      "learning_rate": 9.911700132199924e-05,
      "loss": 4.2167,
      "step": 2130
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8496304154396057,
      "learning_rate": 9.911265914536598e-05,
      "loss": 4.2538,
      "step": 2135
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7550316452980042,
      "learning_rate": 9.910830641408242e-05,
      "loss": 4.3086,
      "step": 2140
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9063945412635803,
      "learning_rate": 9.910394312908397e-05,
      "loss": 4.2478,
      "step": 2145
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7556061148643494,
      "learning_rate": 9.909956929130834e-05,
      "loss": 4.3032,
      "step": 2150
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7757620215415955,
      "learning_rate": 9.90951849016955e-05,
      "loss": 4.1497,
      "step": 2155
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.897804319858551,
      "learning_rate": 9.909078996118768e-05,
      "loss": 4.2201,
      "step": 2160
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7761451005935669,
      "learning_rate": 9.908638447072939e-05,
      "loss": 4.1276,
      "step": 2165
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.744706392288208,
      "learning_rate": 9.90819684312674e-05,
      "loss": 4.0948,
      "step": 2170
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8259772658348083,
      "learning_rate": 9.907754184375074e-05,
      "loss": 4.1926,
      "step": 2175
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7987555265426636,
      "learning_rate": 9.90731047091307e-05,
      "loss": 4.1669,
      "step": 2180
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.725737452507019,
      "learning_rate": 9.906865702836088e-05,
      "loss": 4.2674,
      "step": 2185
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.7878186106681824,
      "learning_rate": 9.90641988023971e-05,
      "loss": 4.1683,
      "step": 2190
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.8156335949897766,
      "learning_rate": 9.905973003219748e-05,
      "loss": 4.1548,
      "step": 2195
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.100070834159851,
      "learning_rate": 9.905525071872237e-05,
      "loss": 4.1883,
      "step": 2200
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.830297589302063,
      "learning_rate": 9.905076086293443e-05,
      "loss": 4.1837,
      "step": 2205
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8370123505592346,
      "learning_rate": 9.904626046579852e-05,
      "loss": 4.1875,
      "step": 2210
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7700148820877075,
      "learning_rate": 9.904174952828186e-05,
      "loss": 4.2746,
      "step": 2215
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7468922138214111,
      "learning_rate": 9.903722805135385e-05,
      "loss": 4.096,
      "step": 2220
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8280553817749023,
      "learning_rate": 9.903269603598617e-05,
      "loss": 4.1166,
      "step": 2225
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7329906821250916,
      "learning_rate": 9.902815348315282e-05,
      "loss": 4.1924,
      "step": 2230
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7978401184082031,
      "learning_rate": 9.902360039383e-05,
      "loss": 4.1268,
      "step": 2235
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8253445029258728,
      "learning_rate": 9.901903676899621e-05,
      "loss": 4.2038,
      "step": 2240
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7740457653999329,
      "learning_rate": 9.901446260963221e-05,
      "loss": 4.2709,
      "step": 2245
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7907018661499023,
      "learning_rate": 9.9009877916721e-05,
      "loss": 4.137,
      "step": 2250
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8358339071273804,
      "learning_rate": 9.900528269124787e-05,
      "loss": 4.1409,
      "step": 2255
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8193193078041077,
      "learning_rate": 9.900067693420039e-05,
      "loss": 4.2023,
      "step": 2260
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.736617922782898,
      "learning_rate": 9.899606064656832e-05,
      "loss": 4.1218,
      "step": 2265
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.88621985912323,
      "learning_rate": 9.899143382934379e-05,
      "loss": 4.1579,
      "step": 2270
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.956885576248169,
      "learning_rate": 9.898679648352108e-05,
      "loss": 4.2118,
      "step": 2275
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8070094585418701,
      "learning_rate": 9.898214861009683e-05,
      "loss": 4.188,
      "step": 2280
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7357026934623718,
      "learning_rate": 9.897749021006988e-05,
      "loss": 4.1517,
      "step": 2285
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8290369510650635,
      "learning_rate": 9.897282128444135e-05,
      "loss": 4.1974,
      "step": 2290
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7857857942581177,
      "learning_rate": 9.896814183421464e-05,
      "loss": 4.1493,
      "step": 2295
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.754823625087738,
      "learning_rate": 9.896345186039538e-05,
      "loss": 4.1327,
      "step": 2300
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8335931897163391,
      "learning_rate": 9.895875136399149e-05,
      "loss": 4.0114,
      "step": 2305
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.914398729801178,
      "learning_rate": 9.895404034601313e-05,
      "loss": 4.1531,
      "step": 2310
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7224910259246826,
      "learning_rate": 9.894931880747273e-05,
      "loss": 4.1535,
      "step": 2315
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8029696941375732,
      "learning_rate": 9.894458674938499e-05,
      "loss": 4.1922,
      "step": 2320
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7707552909851074,
      "learning_rate": 9.893984417276686e-05,
      "loss": 4.1648,
      "step": 2325
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7824752926826477,
      "learning_rate": 9.893509107863754e-05,
      "loss": 4.052,
      "step": 2330
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9048404097557068,
      "learning_rate": 9.893032746801854e-05,
      "loss": 4.2044,
      "step": 2335
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7872731685638428,
      "learning_rate": 9.892555334193353e-05,
      "loss": 4.2525,
      "step": 2340
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7767019867897034,
      "learning_rate": 9.892076870140857e-05,
      "loss": 4.0323,
      "step": 2345
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8217889666557312,
      "learning_rate": 9.891597354747186e-05,
      "loss": 4.1726,
      "step": 2350
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7746922373771667,
      "learning_rate": 9.891116788115393e-05,
      "loss": 4.2547,
      "step": 2355
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7832623720169067,
      "learning_rate": 9.890635170348756e-05,
      "loss": 4.1057,
      "step": 2360
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8045166730880737,
      "learning_rate": 9.890152501550777e-05,
      "loss": 4.161,
      "step": 2365
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7111038565635681,
      "learning_rate": 9.889668781825185e-05,
      "loss": 4.1347,
      "step": 2370
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8069304823875427,
      "learning_rate": 9.889184011275936e-05,
      "loss": 4.2351,
      "step": 2375
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8643245100975037,
      "learning_rate": 9.888698190007209e-05,
      "loss": 4.1146,
      "step": 2380
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9556590914726257,
      "learning_rate": 9.88821131812341e-05,
      "loss": 4.2547,
      "step": 2385
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8787074685096741,
      "learning_rate": 9.887723395729174e-05,
      "loss": 4.2104,
      "step": 2390
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8544966578483582,
      "learning_rate": 9.887234422929355e-05,
      "loss": 4.1943,
      "step": 2395
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7704616189002991,
      "learning_rate": 9.886744399829039e-05,
      "loss": 4.1598,
      "step": 2400
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8291452527046204,
      "learning_rate": 9.886253326533535e-05,
      "loss": 4.1079,
      "step": 2405
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9139868021011353,
      "learning_rate": 9.885761203148378e-05,
      "loss": 4.1134,
      "step": 2410
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8632022738456726,
      "learning_rate": 9.885268029779327e-05,
      "loss": 4.1528,
      "step": 2415
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.775946319103241,
      "learning_rate": 9.884773806532373e-05,
      "loss": 4.1804,
      "step": 2420
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.803955614566803,
      "learning_rate": 9.884278533513723e-05,
      "loss": 4.1509,
      "step": 2425
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7114711403846741,
      "learning_rate": 9.883782210829816e-05,
      "loss": 4.0719,
      "step": 2430
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.850292980670929,
      "learning_rate": 9.883284838587317e-05,
      "loss": 4.172,
      "step": 2435
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9531928896903992,
      "learning_rate": 9.882786416893112e-05,
      "loss": 4.0284,
      "step": 2440
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.898213803768158,
      "learning_rate": 9.882286945854319e-05,
      "loss": 4.1218,
      "step": 2445
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8438432812690735,
      "learning_rate": 9.881786425578274e-05,
      "loss": 4.1996,
      "step": 2450
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7946485280990601,
      "learning_rate": 9.881284856172545e-05,
      "loss": 4.1655,
      "step": 2455
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7994732856750488,
      "learning_rate": 9.880782237744921e-05,
      "loss": 4.1008,
      "step": 2460
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7926275134086609,
      "learning_rate": 9.880278570403419e-05,
      "loss": 4.0829,
      "step": 2465
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7710028886795044,
      "learning_rate": 9.87977385425628e-05,
      "loss": 4.1255,
      "step": 2470
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9420976638793945,
      "learning_rate": 9.87926808941197e-05,
      "loss": 4.2486,
      "step": 2475
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.841746985912323,
      "learning_rate": 9.878761275979184e-05,
      "loss": 4.0446,
      "step": 2480
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.79327791929245,
      "learning_rate": 9.878253414066839e-05,
      "loss": 4.1392,
      "step": 2485
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7650776505470276,
      "learning_rate": 9.877744503784076e-05,
      "loss": 4.098,
      "step": 2490
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7901017665863037,
      "learning_rate": 9.877234545240265e-05,
      "loss": 4.222,
      "step": 2495
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7802971005439758,
      "learning_rate": 9.876723538545001e-05,
      "loss": 4.2193,
      "step": 2500
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7363675832748413,
      "learning_rate": 9.8762114838081e-05,
      "loss": 4.0839,
      "step": 2505
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8111417889595032,
      "learning_rate": 9.875698381139607e-05,
      "loss": 4.061,
      "step": 2510
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8356532454490662,
      "learning_rate": 9.875184230649792e-05,
      "loss": 4.1394,
      "step": 2515
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.9372687339782715,
      "learning_rate": 9.874669032449149e-05,
      "loss": 4.0415,
      "step": 2520
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7597430944442749,
      "learning_rate": 9.874152786648399e-05,
      "loss": 4.0942,
      "step": 2525
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.7658441066741943,
      "learning_rate": 9.873635493358484e-05,
      "loss": 4.1127,
      "step": 2530
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8129209876060486,
      "learning_rate": 9.873117152690576e-05,
      "loss": 4.1872,
      "step": 2535
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8080174922943115,
      "learning_rate": 9.872597764756071e-05,
      "loss": 4.0864,
      "step": 2540
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8307592272758484,
      "learning_rate": 9.872077329666587e-05,
      "loss": 4.1109,
      "step": 2545
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7644115686416626,
      "learning_rate": 9.871555847533968e-05,
      "loss": 4.1058,
      "step": 2550
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7532459497451782,
      "learning_rate": 9.871033318470289e-05,
      "loss": 4.1253,
      "step": 2555
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.06892991065979,
      "learning_rate": 9.870509742587838e-05,
      "loss": 4.0732,
      "step": 2560
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8135694265365601,
      "learning_rate": 9.869985119999143e-05,
      "loss": 4.0016,
      "step": 2565
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7824971079826355,
      "learning_rate": 9.869459450816942e-05,
      "loss": 4.1079,
      "step": 2570
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9020003080368042,
      "learning_rate": 9.86893273515421e-05,
      "loss": 4.224,
      "step": 2575
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7438197731971741,
      "learning_rate": 9.868404973124139e-05,
      "loss": 4.028,
      "step": 2580
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8091161847114563,
      "learning_rate": 9.86787616484015e-05,
      "loss": 3.9873,
      "step": 2585
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8219226598739624,
      "learning_rate": 9.867346310415886e-05,
      "loss": 4.0072,
      "step": 2590
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8258355259895325,
      "learning_rate": 9.866815409965219e-05,
      "loss": 4.0855,
      "step": 2595
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8613013029098511,
      "learning_rate": 9.86628346360224e-05,
      "loss": 4.0931,
      "step": 2600
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8585067987442017,
      "learning_rate": 9.865750471441271e-05,
      "loss": 4.1578,
      "step": 2605
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9461768269538879,
      "learning_rate": 9.865216433596853e-05,
      "loss": 4.0245,
      "step": 2610
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8886264562606812,
      "learning_rate": 9.864681350183758e-05,
      "loss": 4.132,
      "step": 2615
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.743223249912262,
      "learning_rate": 9.864145221316975e-05,
      "loss": 4.0742,
      "step": 2620
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7838162779808044,
      "learning_rate": 9.863608047111725e-05,
      "loss": 4.1214,
      "step": 2625
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.734805703163147,
      "learning_rate": 9.863069827683446e-05,
      "loss": 4.1283,
      "step": 2630
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6818572878837585,
      "learning_rate": 9.862530563147812e-05,
      "loss": 4.0488,
      "step": 2635
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7708154916763306,
      "learning_rate": 9.861990253620709e-05,
      "loss": 4.0592,
      "step": 2640
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7283186316490173,
      "learning_rate": 9.861448899218255e-05,
      "loss": 4.0855,
      "step": 2645
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7363342046737671,
      "learning_rate": 9.860906500056792e-05,
      "loss": 4.1319,
      "step": 2650
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7398841977119446,
      "learning_rate": 9.860363056252883e-05,
      "loss": 4.1345,
      "step": 2655
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8120643496513367,
      "learning_rate": 9.859818567923321e-05,
      "loss": 4.0916,
      "step": 2660
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8443947434425354,
      "learning_rate": 9.859273035185118e-05,
      "loss": 4.129,
      "step": 2665
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7855377197265625,
      "learning_rate": 9.858726458155512e-05,
      "loss": 4.1368,
      "step": 2670
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7563038468360901,
      "learning_rate": 9.858178836951967e-05,
      "loss": 4.101,
      "step": 2675
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8669852018356323,
      "learning_rate": 9.857630171692174e-05,
      "loss": 3.9216,
      "step": 2680
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.73479163646698,
      "learning_rate": 9.857080462494039e-05,
      "loss": 4.0213,
      "step": 2685
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7990299463272095,
      "learning_rate": 9.856529709475703e-05,
      "loss": 4.0142,
      "step": 2690
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.706426739692688,
      "learning_rate": 9.855977912755524e-05,
      "loss": 4.0286,
      "step": 2695
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7951425313949585,
      "learning_rate": 9.855425072452086e-05,
      "loss": 4.0383,
      "step": 2700
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7398020625114441,
      "learning_rate": 9.854871188684204e-05,
      "loss": 4.0851,
      "step": 2705
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7143636345863342,
      "learning_rate": 9.854316261570905e-05,
      "loss": 4.0571,
      "step": 2710
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6799008250236511,
      "learning_rate": 9.85376029123145e-05,
      "loss": 4.0486,
      "step": 2715
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7127188444137573,
      "learning_rate": 9.853203277785318e-05,
      "loss": 3.94,
      "step": 2720
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7105900645256042,
      "learning_rate": 9.85264522135222e-05,
      "loss": 3.9955,
      "step": 2725
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7587236166000366,
      "learning_rate": 9.852086122052083e-05,
      "loss": 4.1314,
      "step": 2730
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7126836776733398,
      "learning_rate": 9.85152598000506e-05,
      "loss": 4.0848,
      "step": 2735
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7512907981872559,
      "learning_rate": 9.850964795331531e-05,
      "loss": 4.0144,
      "step": 2740
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7720962166786194,
      "learning_rate": 9.8504025681521e-05,
      "loss": 4.077,
      "step": 2745
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8098607659339905,
      "learning_rate": 9.84983929858759e-05,
      "loss": 4.0165,
      "step": 2750
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7108052968978882,
      "learning_rate": 9.849274986759054e-05,
      "loss": 4.0131,
      "step": 2755
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7802889943122864,
      "learning_rate": 9.848709632787768e-05,
      "loss": 4.0673,
      "step": 2760
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7834372520446777,
      "learning_rate": 9.848143236795225e-05,
      "loss": 4.0162,
      "step": 2765
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7018157243728638,
      "learning_rate": 9.847575798903152e-05,
      "loss": 4.0526,
      "step": 2770
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7318530082702637,
      "learning_rate": 9.847007319233496e-05,
      "loss": 4.0448,
      "step": 2775
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7086035013198853,
      "learning_rate": 9.846437797908424e-05,
      "loss": 3.9831,
      "step": 2780
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7520937919616699,
      "learning_rate": 9.845867235050333e-05,
      "loss": 3.9638,
      "step": 2785
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8007222414016724,
      "learning_rate": 9.845295630781837e-05,
      "loss": 4.0585,
      "step": 2790
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7911286950111389,
      "learning_rate": 9.844722985225783e-05,
      "loss": 3.9913,
      "step": 2795
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8257914781570435,
      "learning_rate": 9.844149298505232e-05,
      "loss": 4.0944,
      "step": 2800
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6999296545982361,
      "learning_rate": 9.843574570743474e-05,
      "loss": 4.0192,
      "step": 2805
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7232578992843628,
      "learning_rate": 9.842998802064025e-05,
      "loss": 4.013,
      "step": 2810
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7358812093734741,
      "learning_rate": 9.842421992590619e-05,
      "loss": 4.0258,
      "step": 2815
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7485278248786926,
      "learning_rate": 9.841844142447215e-05,
      "loss": 4.0108,
      "step": 2820
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1929200887680054,
      "learning_rate": 9.841265251758e-05,
      "loss": 3.9899,
      "step": 2825
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8642017841339111,
      "learning_rate": 9.840685320647381e-05,
      "loss": 4.0631,
      "step": 2830
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7665454745292664,
      "learning_rate": 9.840104349239988e-05,
      "loss": 4.0785,
      "step": 2835
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7468354105949402,
      "learning_rate": 9.839522337660677e-05,
      "loss": 3.9164,
      "step": 2840
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7683619856834412,
      "learning_rate": 9.838939286034526e-05,
      "loss": 3.9471,
      "step": 2845
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7758325338363647,
      "learning_rate": 9.838355194486836e-05,
      "loss": 4.054,
      "step": 2850
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7746393084526062,
      "learning_rate": 9.837770063143133e-05,
      "loss": 4.0433,
      "step": 2855
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7358125448226929,
      "learning_rate": 9.837183892129164e-05,
      "loss": 3.9496,
      "step": 2860
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7970110177993774,
      "learning_rate": 9.836596681570906e-05,
      "loss": 4.0682,
      "step": 2865
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7705901861190796,
      "learning_rate": 9.83600843159455e-05,
      "loss": 4.0357,
      "step": 2870
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7316420674324036,
      "learning_rate": 9.835419142326517e-05,
      "loss": 3.9176,
      "step": 2875
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7848609089851379,
      "learning_rate": 9.834828813893448e-05,
      "loss": 3.927,
      "step": 2880
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7528284192085266,
      "learning_rate": 9.83423744642221e-05,
      "loss": 4.0622,
      "step": 2885
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.760507345199585,
      "learning_rate": 9.833645040039892e-05,
      "loss": 4.063,
      "step": 2890
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.745658814907074,
      "learning_rate": 9.833051594873805e-05,
      "loss": 4.0438,
      "step": 2895
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.722758948802948,
      "learning_rate": 9.832457111051485e-05,
      "loss": 3.9339,
      "step": 2900
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7566544413566589,
      "learning_rate": 9.831861588700692e-05,
      "loss": 4.0537,
      "step": 2905
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.710358202457428,
      "learning_rate": 9.831265027949406e-05,
      "loss": 4.0193,
      "step": 2910
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7195625901222229,
      "learning_rate": 9.830667428925833e-05,
      "loss": 4.0685,
      "step": 2915
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7128478288650513,
      "learning_rate": 9.8300687917584e-05,
      "loss": 3.962,
      "step": 2920
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7533696293830872,
      "learning_rate": 9.82946911657576e-05,
      "loss": 3.9918,
      "step": 2925
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.883273184299469,
      "learning_rate": 9.828868403506788e-05,
      "loss": 4.0376,
      "step": 2930
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7158867716789246,
      "learning_rate": 9.828266652680577e-05,
      "loss": 3.9555,
      "step": 2935
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.734056293964386,
      "learning_rate": 9.827663864226453e-05,
      "loss": 3.8866,
      "step": 2940
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.891761064529419,
      "learning_rate": 9.827060038273956e-05,
      "loss": 4.0411,
      "step": 2945
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7480966448783875,
      "learning_rate": 9.826455174952852e-05,
      "loss": 4.1173,
      "step": 2950
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7737939357757568,
      "learning_rate": 9.825849274393133e-05,
      "loss": 4.0405,
      "step": 2955
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7104798555374146,
      "learning_rate": 9.825242336725009e-05,
      "loss": 4.034,
      "step": 2960
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7378020882606506,
      "learning_rate": 9.824634362078915e-05,
      "loss": 3.9363,
      "step": 2965
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6872981190681458,
      "learning_rate": 9.824025350585511e-05,
      "loss": 4.0676,
      "step": 2970
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8083837628364563,
      "learning_rate": 9.823415302375676e-05,
      "loss": 4.0098,
      "step": 2975
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6961967945098877,
      "learning_rate": 9.822804217580515e-05,
      "loss": 4.1249,
      "step": 2980
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7386844754219055,
      "learning_rate": 9.822192096331351e-05,
      "loss": 4.0251,
      "step": 2985
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6958508491516113,
      "learning_rate": 9.82157893875974e-05,
      "loss": 4.0632,
      "step": 2990
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6928688883781433,
      "learning_rate": 9.820964744997445e-05,
      "loss": 3.9394,
      "step": 2995
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7715601921081543,
      "learning_rate": 9.820349515176466e-05,
      "loss": 3.9267,
      "step": 3000
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7515993714332581,
      "learning_rate": 9.819733249429019e-05,
      "loss": 4.018,
      "step": 3005
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7219489216804504,
      "learning_rate": 9.819115947887545e-05,
      "loss": 3.9966,
      "step": 3010
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7461873888969421,
      "learning_rate": 9.818497610684705e-05,
      "loss": 4.001,
      "step": 3015
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8548391461372375,
      "learning_rate": 9.817878237953383e-05,
      "loss": 3.9714,
      "step": 3020
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7424699664115906,
      "learning_rate": 9.817257829826689e-05,
      "loss": 4.0579,
      "step": 3025
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.719058096408844,
      "learning_rate": 9.81663638643795e-05,
      "loss": 4.044,
      "step": 3030
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6738602519035339,
      "learning_rate": 9.81601390792072e-05,
      "loss": 3.9057,
      "step": 3035
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7466130256652832,
      "learning_rate": 9.815390394408776e-05,
      "loss": 3.8893,
      "step": 3040
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.848037838935852,
      "learning_rate": 9.814765846036111e-05,
      "loss": 3.9566,
      "step": 3045
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.768531084060669,
      "learning_rate": 9.81414026293695e-05,
      "loss": 4.0175,
      "step": 3050
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7544642686843872,
      "learning_rate": 9.813513645245729e-05,
      "loss": 4.0611,
      "step": 3055
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.710389256477356,
      "learning_rate": 9.812885993097118e-05,
      "loss": 4.0544,
      "step": 3060
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7147810459136963,
      "learning_rate": 9.812257306626002e-05,
      "loss": 4.0297,
      "step": 3065
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7230045795440674,
      "learning_rate": 9.811627585967487e-05,
      "loss": 3.9483,
      "step": 3070
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.696534276008606,
      "learning_rate": 9.810996831256909e-05,
      "loss": 3.9449,
      "step": 3075
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7990797162055969,
      "learning_rate": 9.81036504262982e-05,
      "loss": 4.0815,
      "step": 3080
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.809975266456604,
      "learning_rate": 9.809732220221995e-05,
      "loss": 3.9233,
      "step": 3085
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8132238984107971,
      "learning_rate": 9.80909836416943e-05,
      "loss": 4.1062,
      "step": 3090
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7771387100219727,
      "learning_rate": 9.80846347460835e-05,
      "loss": 3.965,
      "step": 3095
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6894800662994385,
      "learning_rate": 9.807827551675192e-05,
      "loss": 3.9332,
      "step": 3100
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0841281414031982,
      "learning_rate": 9.807190595506623e-05,
      "loss": 3.9767,
      "step": 3105
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.127347707748413,
      "learning_rate": 9.80655260623953e-05,
      "loss": 4.004,
      "step": 3110
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.8833109140396118,
      "learning_rate": 9.805913584011019e-05,
      "loss": 3.986,
      "step": 3115
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7759318947792053,
      "learning_rate": 9.805273528958423e-05,
      "loss": 3.9719,
      "step": 3120
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.78520667552948,
      "learning_rate": 9.804632441219293e-05,
      "loss": 4.0629,
      "step": 3125
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7258091568946838,
      "learning_rate": 9.803990320931402e-05,
      "loss": 3.8967,
      "step": 3130
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7688604593276978,
      "learning_rate": 9.803347168232746e-05,
      "loss": 3.9809,
      "step": 3135
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7502864599227905,
      "learning_rate": 9.802702983261545e-05,
      "loss": 3.9897,
      "step": 3140
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.705725908279419,
      "learning_rate": 9.802057766156238e-05,
      "loss": 4.0531,
      "step": 3145
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7458087801933289,
      "learning_rate": 9.801411517055488e-05,
      "loss": 3.9821,
      "step": 3150
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7353714108467102,
      "learning_rate": 9.800764236098175e-05,
      "loss": 3.9987,
      "step": 3155
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7288916110992432,
      "learning_rate": 9.800115923423405e-05,
      "loss": 3.8548,
      "step": 3160
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7385783195495605,
      "learning_rate": 9.799466579170508e-05,
      "loss": 3.908,
      "step": 3165
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7545756101608276,
      "learning_rate": 9.798816203479029e-05,
      "loss": 3.8721,
      "step": 3170
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0583322048187256,
      "learning_rate": 9.79816479648874e-05,
      "loss": 3.9334,
      "step": 3175
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7293004989624023,
      "learning_rate": 9.797512358339633e-05,
      "loss": 3.9651,
      "step": 3180
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7027021646499634,
      "learning_rate": 9.796858889171921e-05,
      "loss": 4.097,
      "step": 3185
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7191122174263,
      "learning_rate": 9.79620438912604e-05,
      "loss": 3.9319,
      "step": 3190
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6871917247772217,
      "learning_rate": 9.795548858342645e-05,
      "loss": 3.8479,
      "step": 3195
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7160236835479736,
      "learning_rate": 9.794892296962615e-05,
      "loss": 4.0371,
      "step": 3200
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7150247693061829,
      "learning_rate": 9.794234705127047e-05,
      "loss": 3.9839,
      "step": 3205
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7422568798065186,
      "learning_rate": 9.793576082977267e-05,
      "loss": 3.9326,
      "step": 3210
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7878756523132324,
      "learning_rate": 9.792916430654815e-05,
      "loss": 3.9372,
      "step": 3215
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.7709748148918152,
      "learning_rate": 9.792255748301453e-05,
      "loss": 3.8904,
      "step": 3220
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.6990944743156433,
      "learning_rate": 9.791594036059168e-05,
      "loss": 3.8485,
      "step": 3225
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7754520177841187,
      "learning_rate": 9.790931294070169e-05,
      "loss": 3.9795,
      "step": 3230
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.746785581111908,
      "learning_rate": 9.79026752247688e-05,
      "loss": 3.891,
      "step": 3235
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.9178284406661987,
      "learning_rate": 9.78960272142195e-05,
      "loss": 4.0349,
      "step": 3240
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7732784748077393,
      "learning_rate": 9.788936891048251e-05,
      "loss": 3.794,
      "step": 3245
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.9610726833343506,
      "learning_rate": 9.788270031498876e-05,
      "loss": 3.9541,
      "step": 3250
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0118210315704346,
      "learning_rate": 9.787602142917135e-05,
      "loss": 4.0247,
      "step": 3255
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7599555850028992,
      "learning_rate": 9.786933225446563e-05,
      "loss": 3.9734,
      "step": 3260
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7824902534484863,
      "learning_rate": 9.786263279230915e-05,
      "loss": 3.838,
      "step": 3265
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7005599141120911,
      "learning_rate": 9.785592304414167e-05,
      "loss": 3.9649,
      "step": 3270
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7226083874702454,
      "learning_rate": 9.784920301140516e-05,
      "loss": 3.912,
      "step": 3275
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7067767977714539,
      "learning_rate": 9.78424726955438e-05,
      "loss": 3.9454,
      "step": 3280
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8311893939971924,
      "learning_rate": 9.783573209800398e-05,
      "loss": 3.9107,
      "step": 3285
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.714250385761261,
      "learning_rate": 9.782898122023432e-05,
      "loss": 4.0142,
      "step": 3290
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7652574181556702,
      "learning_rate": 9.78222200636856e-05,
      "loss": 3.9509,
      "step": 3295
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7526282668113708,
      "learning_rate": 9.781544862981087e-05,
      "loss": 3.9731,
      "step": 3300
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7795079350471497,
      "learning_rate": 9.780866692006534e-05,
      "loss": 3.9641,
      "step": 3305
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6870121359825134,
      "learning_rate": 9.780187493590645e-05,
      "loss": 4.0577,
      "step": 3310
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6691737771034241,
      "learning_rate": 9.779507267879385e-05,
      "loss": 3.9093,
      "step": 3315
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7428581118583679,
      "learning_rate": 9.778826015018936e-05,
      "loss": 3.9984,
      "step": 3320
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6960693597793579,
      "learning_rate": 9.778143735155709e-05,
      "loss": 3.8689,
      "step": 3325
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8860224485397339,
      "learning_rate": 9.777460428436327e-05,
      "loss": 3.8314,
      "step": 3330
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.673819363117218,
      "learning_rate": 9.776776095007638e-05,
      "loss": 3.9711,
      "step": 3335
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.733471691608429,
      "learning_rate": 9.776090735016711e-05,
      "loss": 4.0542,
      "step": 3340
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8403627872467041,
      "learning_rate": 9.775404348610835e-05,
      "loss": 3.9245,
      "step": 3345
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8133549094200134,
      "learning_rate": 9.774716935937518e-05,
      "loss": 3.9721,
      "step": 3350
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7566184401512146,
      "learning_rate": 9.77402849714449e-05,
      "loss": 3.9774,
      "step": 3355
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.0229859352111816,
      "learning_rate": 9.773339032379701e-05,
      "loss": 3.8129,
      "step": 3360
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7952930331230164,
      "learning_rate": 9.772648541791323e-05,
      "loss": 3.8309,
      "step": 3365
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6955248117446899,
      "learning_rate": 9.771957025527745e-05,
      "loss": 3.993,
      "step": 3370
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8217131495475769,
      "learning_rate": 9.771264483737581e-05,
      "loss": 3.9863,
      "step": 3375
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7798985838890076,
      "learning_rate": 9.770570916569661e-05,
      "loss": 4.0512,
      "step": 3380
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8685747981071472,
      "learning_rate": 9.76987632417304e-05,
      "loss": 3.8787,
      "step": 3385
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6639370918273926,
      "learning_rate": 9.769180706696988e-05,
      "loss": 3.9231,
      "step": 3390
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7095580697059631,
      "learning_rate": 9.768484064290997e-05,
      "loss": 3.9405,
      "step": 3395
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7198642492294312,
      "learning_rate": 9.767786397104786e-05,
      "loss": 3.8722,
      "step": 3400
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7233620285987854,
      "learning_rate": 9.767087705288283e-05,
      "loss": 3.7939,
      "step": 3405
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8588070869445801,
      "learning_rate": 9.766387988991642e-05,
      "loss": 4.0136,
      "step": 3410
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.8157956004142761,
      "learning_rate": 9.765687248365241e-05,
      "loss": 3.8411,
      "step": 3415
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6806758642196655,
      "learning_rate": 9.76498548355967e-05,
      "loss": 3.9107,
      "step": 3420
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6882597208023071,
      "learning_rate": 9.764282694725744e-05,
      "loss": 4.0274,
      "step": 3425
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7129148244857788,
      "learning_rate": 9.7635788820145e-05,
      "loss": 3.9048,
      "step": 3430
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6672071218490601,
      "learning_rate": 9.762874045577189e-05,
      "loss": 3.957,
      "step": 3435
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6573688387870789,
      "learning_rate": 9.762168185565284e-05,
      "loss": 3.8637,
      "step": 3440
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7076075673103333,
      "learning_rate": 9.761461302130484e-05,
      "loss": 4.0383,
      "step": 3445
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7405285239219666,
      "learning_rate": 9.760753395424699e-05,
      "loss": 3.9946,
      "step": 3450
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7034798264503479,
      "learning_rate": 9.760044465600065e-05,
      "loss": 3.9384,
      "step": 3455
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7300727963447571,
      "learning_rate": 9.759334512808935e-05,
      "loss": 3.9314,
      "step": 3460
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7228839993476868,
      "learning_rate": 9.758623537203883e-05,
      "loss": 4.0292,
      "step": 3465
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6920320391654968,
      "learning_rate": 9.757911538937706e-05,
      "loss": 3.9242,
      "step": 3470
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7290514707565308,
      "learning_rate": 9.757198518163411e-05,
      "loss": 3.9039,
      "step": 3475
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6719415187835693,
      "learning_rate": 9.756484475034235e-05,
      "loss": 3.9086,
      "step": 3480
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6778905987739563,
      "learning_rate": 9.75576940970363e-05,
      "loss": 4.0269,
      "step": 3485
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7163796424865723,
      "learning_rate": 9.755053322325269e-05,
      "loss": 3.8789,
      "step": 3490
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7407890558242798,
      "learning_rate": 9.754336213053043e-05,
      "loss": 3.9847,
      "step": 3495
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7084596157073975,
      "learning_rate": 9.753618082041066e-05,
      "loss": 3.9628,
      "step": 3500
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6744286417961121,
      "learning_rate": 9.752898929443666e-05,
      "loss": 4.0142,
      "step": 3505
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7069187760353088,
      "learning_rate": 9.752178755415396e-05,
      "loss": 3.925,
      "step": 3510
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7191082239151001,
      "learning_rate": 9.751457560111026e-05,
      "loss": 3.8514,
      "step": 3515
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6958099603652954,
      "learning_rate": 9.750735343685547e-05,
      "loss": 3.8532,
      "step": 3520
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7785604000091553,
      "learning_rate": 9.750012106294167e-05,
      "loss": 3.9123,
      "step": 3525
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7711930274963379,
      "learning_rate": 9.749287848092315e-05,
      "loss": 3.8649,
      "step": 3530
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7367296814918518,
      "learning_rate": 9.748562569235637e-05,
      "loss": 3.9417,
      "step": 3535
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7231921553611755,
      "learning_rate": 9.747836269880003e-05,
      "loss": 3.923,
      "step": 3540
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7479574680328369,
      "learning_rate": 9.7471089501815e-05,
      "loss": 3.8587,
      "step": 3545
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7554340362548828,
      "learning_rate": 9.746380610296435e-05,
      "loss": 3.9534,
      "step": 3550
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0150195360183716,
      "learning_rate": 9.74565125038133e-05,
      "loss": 3.8286,
      "step": 3555
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7512678503990173,
      "learning_rate": 9.744920870592931e-05,
      "loss": 3.9535,
      "step": 3560
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.7426272034645081,
      "learning_rate": 9.7441894710882e-05,
      "loss": 3.9938,
      "step": 3565
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8531988263130188,
      "learning_rate": 9.743457052024323e-05,
      "loss": 3.8793,
      "step": 3570
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7055252194404602,
      "learning_rate": 9.742723613558702e-05,
      "loss": 3.9246,
      "step": 3575
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7694268822669983,
      "learning_rate": 9.741989155848953e-05,
      "loss": 3.8223,
      "step": 3580
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7166606187820435,
      "learning_rate": 9.741253679052923e-05,
      "loss": 4.0092,
      "step": 3585
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7191266417503357,
      "learning_rate": 9.740517183328666e-05,
      "loss": 3.8883,
      "step": 3590
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7761995792388916,
      "learning_rate": 9.739779668834461e-05,
      "loss": 3.7922,
      "step": 3595
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7263098955154419,
      "learning_rate": 9.739041135728806e-05,
      "loss": 3.8842,
      "step": 3600
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6826450824737549,
      "learning_rate": 9.738301584170417e-05,
      "loss": 3.8217,
      "step": 3605
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6834452152252197,
      "learning_rate": 9.737561014318228e-05,
      "loss": 3.9195,
      "step": 3610
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7843515872955322,
      "learning_rate": 9.736819426331392e-05,
      "loss": 3.8591,
      "step": 3615
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6866840124130249,
      "learning_rate": 9.736076820369283e-05,
      "loss": 3.8311,
      "step": 3620
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6714634895324707,
      "learning_rate": 9.735333196591492e-05,
      "loss": 3.8794,
      "step": 3625
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6876804232597351,
      "learning_rate": 9.734588555157826e-05,
      "loss": 3.9511,
      "step": 3630
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6832573413848877,
      "learning_rate": 9.733842896228318e-05,
      "loss": 3.8994,
      "step": 3635
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7441936731338501,
      "learning_rate": 9.733096219963211e-05,
      "loss": 3.8149,
      "step": 3640
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7112553715705872,
      "learning_rate": 9.732348526522974e-05,
      "loss": 3.9951,
      "step": 3645
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7188517451286316,
      "learning_rate": 9.73159981606829e-05,
      "loss": 3.9197,
      "step": 3650
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6519314050674438,
      "learning_rate": 9.730850088760064e-05,
      "loss": 3.8815,
      "step": 3655
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7158154845237732,
      "learning_rate": 9.730099344759413e-05,
      "loss": 3.8634,
      "step": 3660
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6900362968444824,
      "learning_rate": 9.729347584227684e-05,
      "loss": 3.8966,
      "step": 3665
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.671099841594696,
      "learning_rate": 9.72859480732643e-05,
      "loss": 3.8888,
      "step": 3670
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6918987035751343,
      "learning_rate": 9.727841014217429e-05,
      "loss": 3.8126,
      "step": 3675
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6621039509773254,
      "learning_rate": 9.727086205062676e-05,
      "loss": 3.8732,
      "step": 3680
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.675125002861023,
      "learning_rate": 9.72633038002439e-05,
      "loss": 3.8875,
      "step": 3685
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6571565270423889,
      "learning_rate": 9.725573539264994e-05,
      "loss": 3.8438,
      "step": 3690
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8042651414871216,
      "learning_rate": 9.724815682947146e-05,
      "loss": 3.7984,
      "step": 3695
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6985417008399963,
      "learning_rate": 9.72405681123371e-05,
      "loss": 3.9439,
      "step": 3700
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6829345226287842,
      "learning_rate": 9.723296924287775e-05,
      "loss": 3.7782,
      "step": 3705
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7554473876953125,
      "learning_rate": 9.722536022272646e-05,
      "loss": 3.8013,
      "step": 3710
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7119689583778381,
      "learning_rate": 9.721774105351845e-05,
      "loss": 3.8787,
      "step": 3715
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6767385601997375,
      "learning_rate": 9.721011173689114e-05,
      "loss": 3.8514,
      "step": 3720
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7423023581504822,
      "learning_rate": 9.720247227448411e-05,
      "loss": 3.8765,
      "step": 3725
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7341432571411133,
      "learning_rate": 9.719482266793916e-05,
      "loss": 3.8507,
      "step": 3730
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6785480380058289,
      "learning_rate": 9.718716291890022e-05,
      "loss": 3.8608,
      "step": 3735
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7413129210472107,
      "learning_rate": 9.717949302901343e-05,
      "loss": 3.7917,
      "step": 3740
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7120551466941833,
      "learning_rate": 9.71718129999271e-05,
      "loss": 3.9548,
      "step": 3745
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7153645157814026,
      "learning_rate": 9.716412283329172e-05,
      "loss": 3.9423,
      "step": 3750
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7615500092506409,
      "learning_rate": 9.715642253075998e-05,
      "loss": 3.8956,
      "step": 3755
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.744097888469696,
      "learning_rate": 9.714871209398671e-05,
      "loss": 4.0276,
      "step": 3760
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6623550057411194,
      "learning_rate": 9.714099152462892e-05,
      "loss": 3.9646,
      "step": 3765
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7312632203102112,
      "learning_rate": 9.713326082434584e-05,
      "loss": 3.8148,
      "step": 3770
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.0714316368103027,
      "learning_rate": 9.712551999479884e-05,
      "loss": 3.8356,
      "step": 3775
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7808007001876831,
      "learning_rate": 9.71177690376515e-05,
      "loss": 3.8913,
      "step": 3780
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7386226654052734,
      "learning_rate": 9.711000795456951e-05,
      "loss": 3.9646,
      "step": 3785
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6828973889350891,
      "learning_rate": 9.710223674722083e-05,
      "loss": 3.8761,
      "step": 3790
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7019652128219604,
      "learning_rate": 9.709445541727551e-05,
      "loss": 3.8537,
      "step": 3795
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7127235531806946,
      "learning_rate": 9.708666396640582e-05,
      "loss": 3.8199,
      "step": 3800
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6864252686500549,
      "learning_rate": 9.707886239628621e-05,
      "loss": 3.8968,
      "step": 3805
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7741761207580566,
      "learning_rate": 9.707105070859326e-05,
      "loss": 3.8223,
      "step": 3810
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6799846887588501,
      "learning_rate": 9.706322890500579e-05,
      "loss": 3.8571,
      "step": 3815
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.767926812171936,
      "learning_rate": 9.705539698720476e-05,
      "loss": 3.9378,
      "step": 3820
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7681795954704285,
      "learning_rate": 9.704755495687328e-05,
      "loss": 3.8918,
      "step": 3825
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7288364768028259,
      "learning_rate": 9.703970281569668e-05,
      "loss": 3.7672,
      "step": 3830
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6848310232162476,
      "learning_rate": 9.70318405653624e-05,
      "loss": 3.8422,
      "step": 3835
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.9907037019729614,
      "learning_rate": 9.702396820756014e-05,
      "loss": 3.8872,
      "step": 3840
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6901751160621643,
      "learning_rate": 9.701608574398169e-05,
      "loss": 3.8675,
      "step": 3845
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6914875507354736,
      "learning_rate": 9.700819317632107e-05,
      "loss": 3.8182,
      "step": 3850
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7442486882209778,
      "learning_rate": 9.700029050627443e-05,
      "loss": 3.8043,
      "step": 3855
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6181342005729675,
      "learning_rate": 9.699237773554012e-05,
      "loss": 3.8554,
      "step": 3860
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7173374891281128,
      "learning_rate": 9.698445486581864e-05,
      "loss": 3.9489,
      "step": 3865
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6859890818595886,
      "learning_rate": 9.697652189881266e-05,
      "loss": 3.9066,
      "step": 3870
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.643197774887085,
      "learning_rate": 9.696857883622706e-05,
      "loss": 3.8238,
      "step": 3875
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7117764949798584,
      "learning_rate": 9.696062567976885e-05,
      "loss": 3.9098,
      "step": 3880
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7453239560127258,
      "learning_rate": 9.695266243114718e-05,
      "loss": 3.8484,
      "step": 3885
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6525713801383972,
      "learning_rate": 9.694468909207345e-05,
      "loss": 3.9237,
      "step": 3890
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6787622570991516,
      "learning_rate": 9.693670566426119e-05,
      "loss": 3.7985,
      "step": 3895
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7121519446372986,
      "learning_rate": 9.692871214942608e-05,
      "loss": 3.8035,
      "step": 3900
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6890889406204224,
      "learning_rate": 9.692070854928596e-05,
      "loss": 3.7433,
      "step": 3905
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6926892995834351,
      "learning_rate": 9.691269486556088e-05,
      "loss": 3.7421,
      "step": 3910
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6847178339958191,
      "learning_rate": 9.690467109997303e-05,
      "loss": 3.8563,
      "step": 3915
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7610287666320801,
      "learning_rate": 9.68966372542468e-05,
      "loss": 3.889,
      "step": 3920
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7582743763923645,
      "learning_rate": 9.688859333010866e-05,
      "loss": 3.9254,
      "step": 3925
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6970685124397278,
      "learning_rate": 9.688053932928737e-05,
      "loss": 3.6337,
      "step": 3930
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6435913443565369,
      "learning_rate": 9.687247525351375e-05,
      "loss": 3.7458,
      "step": 3935
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6628790497779846,
      "learning_rate": 9.686440110452084e-05,
      "loss": 3.8691,
      "step": 3940
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6733722686767578,
      "learning_rate": 9.685631688404384e-05,
      "loss": 3.8358,
      "step": 3945
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7022401094436646,
      "learning_rate": 9.684822259382008e-05,
      "loss": 3.9326,
      "step": 3950
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7360368967056274,
      "learning_rate": 9.684011823558909e-05,
      "loss": 3.8302,
      "step": 3955
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7467257380485535,
      "learning_rate": 9.683200381109254e-05,
      "loss": 3.8588,
      "step": 3960
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6862025856971741,
      "learning_rate": 9.68238793220743e-05,
      "loss": 3.8104,
      "step": 3965
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7375925779342651,
      "learning_rate": 9.681574477028039e-05,
      "loss": 3.8592,
      "step": 3970
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6502492427825928,
      "learning_rate": 9.680760015745894e-05,
      "loss": 3.8981,
      "step": 3975
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6648892760276794,
      "learning_rate": 9.67994454853603e-05,
      "loss": 3.8021,
      "step": 3980
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6658400297164917,
      "learning_rate": 9.679128075573698e-05,
      "loss": 3.9363,
      "step": 3985
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7797056436538696,
      "learning_rate": 9.678310597034364e-05,
      "loss": 3.7606,
      "step": 3990
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8219534754753113,
      "learning_rate": 9.677492113093706e-05,
      "loss": 3.927,
      "step": 3995
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8211590647697449,
      "learning_rate": 9.676672623927625e-05,
      "loss": 3.8555,
      "step": 4000
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6888617873191833,
      "learning_rate": 9.675852129712234e-05,
      "loss": 3.7574,
      "step": 4005
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7110888957977295,
      "learning_rate": 9.675030630623864e-05,
      "loss": 3.8277,
      "step": 4010
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6909633874893188,
      "learning_rate": 9.67420812683906e-05,
      "loss": 3.8568,
      "step": 4015
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6736234426498413,
      "learning_rate": 9.673384618534582e-05,
      "loss": 3.8707,
      "step": 4020
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8059706687927246,
      "learning_rate": 9.672560105887409e-05,
      "loss": 3.8764,
      "step": 4025
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7178688645362854,
      "learning_rate": 9.671734589074736e-05,
      "loss": 3.7723,
      "step": 4030
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.731417715549469,
      "learning_rate": 9.670908068273972e-05,
      "loss": 3.8978,
      "step": 4035
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6916252970695496,
      "learning_rate": 9.67008054366274e-05,
      "loss": 3.6844,
      "step": 4040
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7116053700447083,
      "learning_rate": 9.669252015418882e-05,
      "loss": 3.8742,
      "step": 4045
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7012292146682739,
      "learning_rate": 9.668422483720456e-05,
      "loss": 3.7863,
      "step": 4050
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7366868257522583,
      "learning_rate": 9.66759194874573e-05,
      "loss": 3.6905,
      "step": 4055
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7654955983161926,
      "learning_rate": 9.666760410673195e-05,
      "loss": 3.879,
      "step": 4060
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6694311499595642,
      "learning_rate": 9.665927869681553e-05,
      "loss": 3.9847,
      "step": 4065
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7594870924949646,
      "learning_rate": 9.665094325949723e-05,
      "loss": 3.7149,
      "step": 4070
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.753724217414856,
      "learning_rate": 9.664259779656842e-05,
      "loss": 3.8137,
      "step": 4075
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7392582297325134,
      "learning_rate": 9.663424230982254e-05,
      "loss": 3.8971,
      "step": 4080
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7506909370422363,
      "learning_rate": 9.66258768010553e-05,
      "loss": 3.8533,
      "step": 4085
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.669894278049469,
      "learning_rate": 9.661750127206445e-05,
      "loss": 3.8371,
      "step": 4090
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.671461284160614,
      "learning_rate": 9.660911572465e-05,
      "loss": 3.7493,
      "step": 4095
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7372993230819702,
      "learning_rate": 9.660072016061406e-05,
      "loss": 3.8309,
      "step": 4100
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7047749757766724,
      "learning_rate": 9.659231458176085e-05,
      "loss": 3.7541,
      "step": 4105
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6656426191329956,
      "learning_rate": 9.658389898989684e-05,
      "loss": 3.8268,
      "step": 4110
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6715406179428101,
      "learning_rate": 9.657547338683055e-05,
      "loss": 3.825,
      "step": 4115
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6625576019287109,
      "learning_rate": 9.656703777437273e-05,
      "loss": 3.6975,
      "step": 4120
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7435094714164734,
      "learning_rate": 9.655859215433625e-05,
      "loss": 3.9107,
      "step": 4125
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7117746472358704,
      "learning_rate": 9.655013652853612e-05,
      "loss": 3.8793,
      "step": 4130
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6769823431968689,
      "learning_rate": 9.654167089878952e-05,
      "loss": 3.8291,
      "step": 4135
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.662699818611145,
      "learning_rate": 9.653319526691577e-05,
      "loss": 3.8443,
      "step": 4140
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6602285504341125,
      "learning_rate": 9.652470963473635e-05,
      "loss": 3.8831,
      "step": 4145
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6535061001777649,
      "learning_rate": 9.651621400407487e-05,
      "loss": 3.9066,
      "step": 4150
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6752974987030029,
      "learning_rate": 9.650770837675712e-05,
      "loss": 3.8004,
      "step": 4155
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6730858087539673,
      "learning_rate": 9.649919275461099e-05,
      "loss": 3.8749,
      "step": 4160
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7307189106941223,
      "learning_rate": 9.649066713946655e-05,
      "loss": 3.8207,
      "step": 4165
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6608849763870239,
      "learning_rate": 9.648213153315603e-05,
      "loss": 3.8392,
      "step": 4170
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7603358030319214,
      "learning_rate": 9.647358593751378e-05,
      "loss": 3.794,
      "step": 4175
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7025288939476013,
      "learning_rate": 9.646503035437633e-05,
      "loss": 3.8138,
      "step": 4180
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6737129092216492,
      "learning_rate": 9.64564647855823e-05,
      "loss": 3.7189,
      "step": 4185
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7193328738212585,
      "learning_rate": 9.644788923297253e-05,
      "loss": 3.7526,
      "step": 4190
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7209916114807129,
      "learning_rate": 9.64393036983899e-05,
      "loss": 3.7228,
      "step": 4195
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7370156049728394,
      "learning_rate": 9.643070818367957e-05,
      "loss": 3.8044,
      "step": 4200
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7239187955856323,
      "learning_rate": 9.642210269068874e-05,
      "loss": 3.867,
      "step": 4205
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6631564497947693,
      "learning_rate": 9.641348722126679e-05,
      "loss": 3.7893,
      "step": 4210
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7152836918830872,
      "learning_rate": 9.640486177726527e-05,
      "loss": 3.7757,
      "step": 4215
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8161007165908813,
      "learning_rate": 9.639622636053781e-05,
      "loss": 3.8219,
      "step": 4220
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7268388867378235,
      "learning_rate": 9.638758097294024e-05,
      "loss": 3.893,
      "step": 4225
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7126551866531372,
      "learning_rate": 9.637892561633054e-05,
      "loss": 3.811,
      "step": 4230
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7863583564758301,
      "learning_rate": 9.637026029256874e-05,
      "loss": 3.6538,
      "step": 4235
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.6742908358573914,
      "learning_rate": 9.636158500351713e-05,
      "loss": 3.8133,
      "step": 4240
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.672213077545166,
      "learning_rate": 9.635289975104009e-05,
      "loss": 3.8122,
      "step": 4245
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6957303881645203,
      "learning_rate": 9.634420453700413e-05,
      "loss": 3.9459,
      "step": 4250
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7409310340881348,
      "learning_rate": 9.633549936327789e-05,
      "loss": 3.9024,
      "step": 4255
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6747995615005493,
      "learning_rate": 9.63267842317322e-05,
      "loss": 3.8916,
      "step": 4260
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.8072021007537842,
      "learning_rate": 9.631805914424e-05,
      "loss": 3.8789,
      "step": 4265
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6807817220687866,
      "learning_rate": 9.630932410267637e-05,
      "loss": 3.812,
      "step": 4270
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7510599493980408,
      "learning_rate": 9.630057910891852e-05,
      "loss": 3.7439,
      "step": 4275
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6785033345222473,
      "learning_rate": 9.629182416484582e-05,
      "loss": 3.8062,
      "step": 4280
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6535221338272095,
      "learning_rate": 9.628305927233975e-05,
      "loss": 3.8264,
      "step": 4285
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.659803032875061,
      "learning_rate": 9.627428443328397e-05,
      "loss": 3.7252,
      "step": 4290
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7842913269996643,
      "learning_rate": 9.626549964956424e-05,
      "loss": 3.8096,
      "step": 4295
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7154276371002197,
      "learning_rate": 9.625670492306848e-05,
      "loss": 3.8639,
      "step": 4300
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7371067404747009,
      "learning_rate": 9.624790025568673e-05,
      "loss": 3.8137,
      "step": 4305
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6887011528015137,
      "learning_rate": 9.623908564931117e-05,
      "loss": 3.7101,
      "step": 4310
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6945094466209412,
      "learning_rate": 9.623026110583612e-05,
      "loss": 3.814,
      "step": 4315
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7022725939750671,
      "learning_rate": 9.622142662715808e-05,
      "loss": 3.8278,
      "step": 4320
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7546177506446838,
      "learning_rate": 9.621258221517556e-05,
      "loss": 3.814,
      "step": 4325
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6581098437309265,
      "learning_rate": 9.620372787178934e-05,
      "loss": 3.7013,
      "step": 4330
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.65794837474823,
      "learning_rate": 9.619486359890228e-05,
      "loss": 3.7558,
      "step": 4335
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6979213953018188,
      "learning_rate": 9.618598939841934e-05,
      "loss": 3.7498,
      "step": 4340
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6551359295845032,
      "learning_rate": 9.617710527224768e-05,
      "loss": 3.8007,
      "step": 4345
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6268786787986755,
      "learning_rate": 9.616821122229656e-05,
      "loss": 3.7119,
      "step": 4350
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6691540479660034,
      "learning_rate": 9.615930725047732e-05,
      "loss": 3.7953,
      "step": 4355
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6890245079994202,
      "learning_rate": 9.615039335870357e-05,
      "loss": 3.8028,
      "step": 4360
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6479136347770691,
      "learning_rate": 9.614146954889091e-05,
      "loss": 3.8816,
      "step": 4365
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7748972773551941,
      "learning_rate": 9.613253582295714e-05,
      "loss": 3.7866,
      "step": 4370
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7267974615097046,
      "learning_rate": 9.612359218282216e-05,
      "loss": 3.6501,
      "step": 4375
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.699131965637207,
      "learning_rate": 9.611463863040807e-05,
      "loss": 3.7923,
      "step": 4380
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7759175300598145,
      "learning_rate": 9.610567516763902e-05,
      "loss": 3.7385,
      "step": 4385
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6975424289703369,
      "learning_rate": 9.60967017964413e-05,
      "loss": 3.7876,
      "step": 4390
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6473740935325623,
      "learning_rate": 9.608771851874338e-05,
      "loss": 3.7894,
      "step": 4395
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6984918117523193,
      "learning_rate": 9.607872533647584e-05,
      "loss": 3.6996,
      "step": 4400
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6749013662338257,
      "learning_rate": 9.606972225157134e-05,
      "loss": 3.8394,
      "step": 4405
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.661348819732666,
      "learning_rate": 9.606070926596471e-05,
      "loss": 3.7939,
      "step": 4410
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6563800573348999,
      "learning_rate": 9.605168638159294e-05,
      "loss": 3.892,
      "step": 4415
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6903115510940552,
      "learning_rate": 9.604265360039507e-05,
      "loss": 3.7467,
      "step": 4420
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6968739628791809,
      "learning_rate": 9.603361092431234e-05,
      "loss": 3.8499,
      "step": 4425
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7806746363639832,
      "learning_rate": 9.602455835528804e-05,
      "loss": 3.7042,
      "step": 4430
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6835498213768005,
      "learning_rate": 9.601549589526767e-05,
      "loss": 3.8456,
      "step": 4435
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6580449342727661,
      "learning_rate": 9.60064235461988e-05,
      "loss": 3.7861,
      "step": 4440
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6747674942016602,
      "learning_rate": 9.599734131003114e-05,
      "loss": 3.8373,
      "step": 4445
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7133957743644714,
      "learning_rate": 9.598824918871654e-05,
      "loss": 3.8257,
      "step": 4450
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6850093007087708,
      "learning_rate": 9.597914718420892e-05,
      "loss": 3.8961,
      "step": 4455
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6808018088340759,
      "learning_rate": 9.597003529846442e-05,
      "loss": 3.7672,
      "step": 4460
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6689706444740295,
      "learning_rate": 9.59609135334412e-05,
      "loss": 3.9405,
      "step": 4465
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.66801917552948,
      "learning_rate": 9.595178189109963e-05,
      "loss": 3.8017,
      "step": 4470
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7193124294281006,
      "learning_rate": 9.594264037340212e-05,
      "loss": 3.7868,
      "step": 4475
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7194704413414001,
      "learning_rate": 9.593348898231326e-05,
      "loss": 3.8373,
      "step": 4480
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.7037748694419861,
      "learning_rate": 9.592432771979978e-05,
      "loss": 3.7439,
      "step": 4485
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6788994669914246,
      "learning_rate": 9.591515658783046e-05,
      "loss": 3.7141,
      "step": 4490
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6741676330566406,
      "learning_rate": 9.590597558837626e-05,
      "loss": 3.6127,
      "step": 4495
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6605724096298218,
      "learning_rate": 9.589678472341023e-05,
      "loss": 3.7314,
      "step": 4500
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.8783450722694397,
      "learning_rate": 9.588758399490755e-05,
      "loss": 3.7755,
      "step": 4505
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6835188865661621,
      "learning_rate": 9.587837340484553e-05,
      "loss": 3.8178,
      "step": 4510
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.677359402179718,
      "learning_rate": 9.586915295520358e-05,
      "loss": 3.7829,
      "step": 4515
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6628224849700928,
      "learning_rate": 9.585992264796324e-05,
      "loss": 3.7673,
      "step": 4520
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6344485878944397,
      "learning_rate": 9.585068248510817e-05,
      "loss": 3.7322,
      "step": 4525
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6477779746055603,
      "learning_rate": 9.584143246862414e-05,
      "loss": 3.7501,
      "step": 4530
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.635899007320404,
      "learning_rate": 9.583217260049907e-05,
      "loss": 3.8804,
      "step": 4535
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6486161947250366,
      "learning_rate": 9.582290288272291e-05,
      "loss": 3.7213,
      "step": 4540
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6272599697113037,
      "learning_rate": 9.581362331728784e-05,
      "loss": 3.7009,
      "step": 4545
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6917307376861572,
      "learning_rate": 9.580433390618808e-05,
      "loss": 3.7676,
      "step": 4550
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6916787624359131,
      "learning_rate": 9.579503465142e-05,
      "loss": 3.8048,
      "step": 4555
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6864062547683716,
      "learning_rate": 9.578572555498207e-05,
      "loss": 3.7873,
      "step": 4560
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6955726146697998,
      "learning_rate": 9.577640661887488e-05,
      "loss": 3.7229,
      "step": 4565
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6768925189971924,
      "learning_rate": 9.576707784510113e-05,
      "loss": 3.7331,
      "step": 4570
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6550329327583313,
      "learning_rate": 9.575773923566563e-05,
      "loss": 3.7877,
      "step": 4575
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6363929510116577,
      "learning_rate": 9.574839079257532e-05,
      "loss": 3.7705,
      "step": 4580
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.6772913932800293,
      "learning_rate": 9.573903251783925e-05,
      "loss": 3.8064,
      "step": 4585
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7799834609031677,
      "learning_rate": 9.572966441346858e-05,
      "loss": 3.714,
      "step": 4590
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.685915470123291,
      "learning_rate": 9.572028648147657e-05,
      "loss": 3.831,
      "step": 4595
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6995981335639954,
      "learning_rate": 9.571089872387862e-05,
      "loss": 3.6812,
      "step": 4600
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7328318357467651,
      "learning_rate": 9.57015011426922e-05,
      "loss": 3.7356,
      "step": 4605
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6539381146430969,
      "learning_rate": 9.569209373993693e-05,
      "loss": 3.8145,
      "step": 4610
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6630015969276428,
      "learning_rate": 9.568267651763451e-05,
      "loss": 3.8633,
      "step": 4615
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7794010639190674,
      "learning_rate": 9.56732494778088e-05,
      "loss": 3.7724,
      "step": 4620
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7095695734024048,
      "learning_rate": 9.56638126224857e-05,
      "loss": 3.7096,
      "step": 4625
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7243617177009583,
      "learning_rate": 9.565436595369328e-05,
      "loss": 3.7648,
      "step": 4630
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.678071916103363,
      "learning_rate": 9.564490947346168e-05,
      "loss": 3.8175,
      "step": 4635
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6386250853538513,
      "learning_rate": 9.563544318382316e-05,
      "loss": 3.6974,
      "step": 4640
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7747935056686401,
      "learning_rate": 9.56259670868121e-05,
      "loss": 3.7325,
      "step": 4645
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6498805284500122,
      "learning_rate": 9.5616481184465e-05,
      "loss": 3.7781,
      "step": 4650
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6850088238716125,
      "learning_rate": 9.560698547882041e-05,
      "loss": 3.8033,
      "step": 4655
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6730029582977295,
      "learning_rate": 9.559747997191903e-05,
      "loss": 3.7273,
      "step": 4660
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6881535053253174,
      "learning_rate": 9.558796466580366e-05,
      "loss": 3.8005,
      "step": 4665
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6282068490982056,
      "learning_rate": 9.557843956251924e-05,
      "loss": 3.7958,
      "step": 4670
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6996836066246033,
      "learning_rate": 9.556890466411272e-05,
      "loss": 3.7962,
      "step": 4675
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.733283519744873,
      "learning_rate": 9.555935997263328e-05,
      "loss": 3.8089,
      "step": 4680
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6561275720596313,
      "learning_rate": 9.554980549013209e-05,
      "loss": 3.7965,
      "step": 4685
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7193552255630493,
      "learning_rate": 9.554024121866249e-05,
      "loss": 3.651,
      "step": 4690
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7125024199485779,
      "learning_rate": 9.55306671602799e-05,
      "loss": 3.7269,
      "step": 4695
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8615260124206543,
      "learning_rate": 9.552108331704189e-05,
      "loss": 3.6698,
      "step": 4700
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6783967018127441,
      "learning_rate": 9.551148969100804e-05,
      "loss": 3.8896,
      "step": 4705
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8152195811271667,
      "learning_rate": 9.550188628424012e-05,
      "loss": 3.728,
      "step": 4710
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.663604199886322,
      "learning_rate": 9.549227309880197e-05,
      "loss": 3.7658,
      "step": 4715
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.8130972385406494,
      "learning_rate": 9.548265013675951e-05,
      "loss": 3.6762,
      "step": 4720
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7327086925506592,
      "learning_rate": 9.54730174001808e-05,
      "loss": 3.7147,
      "step": 4725
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6797347664833069,
      "learning_rate": 9.546337489113597e-05,
      "loss": 3.8123,
      "step": 4730
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7810825705528259,
      "learning_rate": 9.545372261169725e-05,
      "loss": 3.6678,
      "step": 4735
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7133763432502747,
      "learning_rate": 9.544406056393903e-05,
      "loss": 3.8079,
      "step": 4740
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7472773790359497,
      "learning_rate": 9.54343887499377e-05,
      "loss": 3.8487,
      "step": 4745
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7636592984199524,
      "learning_rate": 9.542470717177182e-05,
      "loss": 3.6375,
      "step": 4750
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7216145992279053,
      "learning_rate": 9.541501583152204e-05,
      "loss": 3.8123,
      "step": 4755
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6921109557151794,
      "learning_rate": 9.540531473127106e-05,
      "loss": 3.7531,
      "step": 4760
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6648208498954773,
      "learning_rate": 9.539560387310375e-05,
      "loss": 3.7358,
      "step": 4765
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6873924136161804,
      "learning_rate": 9.538588325910702e-05,
      "loss": 3.7759,
      "step": 4770
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6596155762672424,
      "learning_rate": 9.53761528913699e-05,
      "loss": 3.7702,
      "step": 4775
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7167657017707825,
      "learning_rate": 9.536641277198353e-05,
      "loss": 3.8388,
      "step": 4780
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7047612071037292,
      "learning_rate": 9.535666290304111e-05,
      "loss": 3.7945,
      "step": 4785
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.659663200378418,
      "learning_rate": 9.534690328663796e-05,
      "loss": 3.7356,
      "step": 4790
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8161595463752747,
      "learning_rate": 9.533713392487149e-05,
      "loss": 3.6676,
      "step": 4795
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6260289549827576,
      "learning_rate": 9.532735481984121e-05,
      "loss": 3.8547,
      "step": 4800
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6579049825668335,
      "learning_rate": 9.53175659736487e-05,
      "loss": 3.7357,
      "step": 4805
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6839591860771179,
      "learning_rate": 9.530776738839766e-05,
      "loss": 3.8128,
      "step": 4810
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6603372693061829,
      "learning_rate": 9.529795906619387e-05,
      "loss": 3.7549,
      "step": 4815
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6463632583618164,
      "learning_rate": 9.528814100914522e-05,
      "loss": 3.8035,
      "step": 4820
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6729907393455505,
      "learning_rate": 9.52783132193617e-05,
      "loss": 3.7683,
      "step": 4825
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6462814211845398,
      "learning_rate": 9.52684756989553e-05,
      "loss": 3.8915,
      "step": 4830
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6726704239845276,
      "learning_rate": 9.525862845004022e-05,
      "loss": 3.7551,
      "step": 4835
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6637750864028931,
      "learning_rate": 9.52487714747327e-05,
      "loss": 3.6835,
      "step": 4840
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6965750455856323,
      "learning_rate": 9.523890477515105e-05,
      "loss": 3.8619,
      "step": 4845
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7050938010215759,
      "learning_rate": 9.522902835341573e-05,
      "loss": 3.7381,
      "step": 4850
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.564842700958252,
      "learning_rate": 9.521914221164922e-05,
      "loss": 3.7634,
      "step": 4855
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6671020984649658,
      "learning_rate": 9.520924635197612e-05,
      "loss": 3.7613,
      "step": 4860
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6357429027557373,
      "learning_rate": 9.519934077652312e-05,
      "loss": 3.7063,
      "step": 4865
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6546706557273865,
      "learning_rate": 9.518942548741902e-05,
      "loss": 3.7353,
      "step": 4870
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6571557521820068,
      "learning_rate": 9.517950048679467e-05,
      "loss": 3.736,
      "step": 4875
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6127691268920898,
      "learning_rate": 9.516956577678302e-05,
      "loss": 3.7657,
      "step": 4880
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6815075278282166,
      "learning_rate": 9.51596213595191e-05,
      "loss": 3.8573,
      "step": 4885
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.8378604650497437,
      "learning_rate": 9.514966723714004e-05,
      "loss": 3.7592,
      "step": 4890
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6306765675544739,
      "learning_rate": 9.513970341178506e-05,
      "loss": 3.772,
      "step": 4895
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.642233669757843,
      "learning_rate": 9.512972988559545e-05,
      "loss": 3.6963,
      "step": 4900
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.7159611582756042,
      "learning_rate": 9.511974666071458e-05,
      "loss": 3.7957,
      "step": 4905
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6733197569847107,
      "learning_rate": 9.510975373928792e-05,
      "loss": 3.5881,
      "step": 4910
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6531829833984375,
      "learning_rate": 9.509975112346303e-05,
      "loss": 3.5557,
      "step": 4915
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6851134896278381,
      "learning_rate": 9.508973881538952e-05,
      "loss": 3.7943,
      "step": 4920
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6288054585456848,
      "learning_rate": 9.507971681721911e-05,
      "loss": 3.7582,
      "step": 4925
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6526462435722351,
      "learning_rate": 9.50696851311056e-05,
      "loss": 3.9354,
      "step": 4930
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6528685688972473,
      "learning_rate": 9.505964375920489e-05,
      "loss": 3.8026,
      "step": 4935
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6465728878974915,
      "learning_rate": 9.50495927036749e-05,
      "loss": 3.702,
      "step": 4940
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.723405122756958,
      "learning_rate": 9.503953196667569e-05,
      "loss": 3.7258,
      "step": 4945
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6770041584968567,
      "learning_rate": 9.502946155036938e-05,
      "loss": 3.5705,
      "step": 4950
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6670681834220886,
      "learning_rate": 9.50193814569202e-05,
      "loss": 3.7954,
      "step": 4955
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6247360706329346,
      "learning_rate": 9.500929168849439e-05,
      "loss": 3.6959,
      "step": 4960
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7034873962402344,
      "learning_rate": 9.499919224726032e-05,
      "loss": 3.646,
      "step": 4965
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6678271889686584,
      "learning_rate": 9.498908313538846e-05,
      "loss": 3.7168,
      "step": 4970
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6549263000488281,
      "learning_rate": 9.49789643550513e-05,
      "loss": 3.8374,
      "step": 4975
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6765576601028442,
      "learning_rate": 9.496883590842343e-05,
      "loss": 3.799,
      "step": 4980
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6775615215301514,
      "learning_rate": 9.495869779768154e-05,
      "loss": 3.7856,
      "step": 4985
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7025602459907532,
      "learning_rate": 9.494855002500438e-05,
      "loss": 3.7661,
      "step": 4990
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6586291790008545,
      "learning_rate": 9.493839259257278e-05,
      "loss": 3.7444,
      "step": 4995
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.701615035533905,
      "learning_rate": 9.492822550256962e-05,
      "loss": 3.611,
      "step": 5000
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6458895206451416,
      "learning_rate": 9.49180487571799e-05,
      "loss": 3.8019,
      "step": 5005
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6729027032852173,
      "learning_rate": 9.490786235859066e-05,
      "loss": 3.7056,
      "step": 5010
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6592373251914978,
      "learning_rate": 9.489766630899104e-05,
      "loss": 3.7932,
      "step": 5015
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6353498101234436,
      "learning_rate": 9.488746061057223e-05,
      "loss": 3.7134,
      "step": 5020
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6946417689323425,
      "learning_rate": 9.487724526552753e-05,
      "loss": 3.7368,
      "step": 5025
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6795774102210999,
      "learning_rate": 9.486702027605224e-05,
      "loss": 3.657,
      "step": 5030
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.739536702632904,
      "learning_rate": 9.485678564434381e-05,
      "loss": 3.7658,
      "step": 5035
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8021363615989685,
      "learning_rate": 9.484654137260176e-05,
      "loss": 3.7651,
      "step": 5040
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.674990713596344,
      "learning_rate": 9.483628746302761e-05,
      "loss": 3.8117,
      "step": 5045
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6410651803016663,
      "learning_rate": 9.482602391782501e-05,
      "loss": 3.6741,
      "step": 5050
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7386654615402222,
      "learning_rate": 9.481575073919968e-05,
      "loss": 3.7987,
      "step": 5055
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6734520792961121,
      "learning_rate": 9.480546792935939e-05,
      "loss": 3.6728,
      "step": 5060
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.723974347114563,
      "learning_rate": 9.479517549051398e-05,
      "loss": 3.7382,
      "step": 5065
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6471766829490662,
      "learning_rate": 9.478487342487536e-05,
      "loss": 3.6281,
      "step": 5070
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6364583373069763,
      "learning_rate": 9.477456173465754e-05,
      "loss": 3.6388,
      "step": 5075
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.0277632474899292,
      "learning_rate": 9.476424042207655e-05,
      "loss": 3.7042,
      "step": 5080
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7114141583442688,
      "learning_rate": 9.475390948935054e-05,
      "loss": 3.9085,
      "step": 5085
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6893407702445984,
      "learning_rate": 9.474356893869967e-05,
      "loss": 3.6869,
      "step": 5090
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6581876277923584,
      "learning_rate": 9.47332187723462e-05,
      "loss": 3.6621,
      "step": 5095
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7156861424446106,
      "learning_rate": 9.472285899251445e-05,
      "loss": 3.5443,
      "step": 5100
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7322091460227966,
      "learning_rate": 9.471248960143085e-05,
      "loss": 3.6498,
      "step": 5105
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.9005959630012512,
      "learning_rate": 9.47021106013238e-05,
      "loss": 3.6537,
      "step": 5110
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7185865044593811,
      "learning_rate": 9.469172199442383e-05,
      "loss": 3.7121,
      "step": 5115
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6447934508323669,
      "learning_rate": 9.468132378296355e-05,
      "loss": 3.6444,
      "step": 5120
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6421835422515869,
      "learning_rate": 9.467091596917758e-05,
      "loss": 3.7709,
      "step": 5125
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6536871194839478,
      "learning_rate": 9.466049855530265e-05,
      "loss": 3.7238,
      "step": 5130
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.712164044380188,
      "learning_rate": 9.465007154357752e-05,
      "loss": 3.7529,
      "step": 5135
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6509156823158264,
      "learning_rate": 9.463963493624304e-05,
      "loss": 3.6418,
      "step": 5140
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8072320818901062,
      "learning_rate": 9.462918873554209e-05,
      "loss": 3.7931,
      "step": 5145
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6603420376777649,
      "learning_rate": 9.461873294371965e-05,
      "loss": 3.8258,
      "step": 5150
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7887171506881714,
      "learning_rate": 9.460826756302275e-05,
      "loss": 3.7943,
      "step": 5155
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.8414411544799805,
      "learning_rate": 9.459779259570045e-05,
      "loss": 3.7072,
      "step": 5160
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6599427461624146,
      "learning_rate": 9.458730804400389e-05,
      "loss": 3.7216,
      "step": 5165
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7204203605651855,
      "learning_rate": 9.457681391018629e-05,
      "loss": 3.6586,
      "step": 5170
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.662329375743866,
      "learning_rate": 9.45663101965029e-05,
      "loss": 3.7691,
      "step": 5175
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7182522416114807,
      "learning_rate": 9.455579690521106e-05,
      "loss": 3.6721,
      "step": 5180
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7199786305427551,
      "learning_rate": 9.454527403857013e-05,
      "loss": 3.7708,
      "step": 5185
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6762434244155884,
      "learning_rate": 9.453474159884155e-05,
      "loss": 3.7196,
      "step": 5190
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6246855854988098,
      "learning_rate": 9.452419958828883e-05,
      "loss": 3.669,
      "step": 5195
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6661684513092041,
      "learning_rate": 9.45136480091775e-05,
      "loss": 3.6504,
      "step": 5200
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6106119751930237,
      "learning_rate": 9.450308686377516e-05,
      "loss": 3.5993,
      "step": 5205
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.639525294303894,
      "learning_rate": 9.449251615435151e-05,
      "loss": 3.7018,
      "step": 5210
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.659537136554718,
      "learning_rate": 9.448193588317822e-05,
      "loss": 3.694,
      "step": 5215
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6245831847190857,
      "learning_rate": 9.447134605252912e-05,
      "loss": 3.7093,
      "step": 5220
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6542061567306519,
      "learning_rate": 9.446074666468e-05,
      "loss": 3.683,
      "step": 5225
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6588606238365173,
      "learning_rate": 9.445013772190874e-05,
      "loss": 3.6529,
      "step": 5230
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7639594078063965,
      "learning_rate": 9.443951922649529e-05,
      "loss": 3.6337,
      "step": 5235
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6120252013206482,
      "learning_rate": 9.442889118072162e-05,
      "loss": 3.7133,
      "step": 5240
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.678114116191864,
      "learning_rate": 9.441825358687181e-05,
      "loss": 3.6693,
      "step": 5245
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6213470697402954,
      "learning_rate": 9.440760644723192e-05,
      "loss": 3.6048,
      "step": 5250
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6561489701271057,
      "learning_rate": 9.439694976409008e-05,
      "loss": 3.7111,
      "step": 5255
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.7540062665939331,
      "learning_rate": 9.438628353973653e-05,
      "loss": 3.7596,
      "step": 5260
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.6423439979553223,
      "learning_rate": 9.437560777646348e-05,
      "loss": 3.7498,
      "step": 5265
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6140279173851013,
      "learning_rate": 9.436492247656523e-05,
      "loss": 3.8221,
      "step": 5270
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6377617716789246,
      "learning_rate": 9.435422764233814e-05,
      "loss": 3.673,
      "step": 5275
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6614524722099304,
      "learning_rate": 9.43435232760806e-05,
      "loss": 3.8153,
      "step": 5280
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6483317017555237,
      "learning_rate": 9.433280938009304e-05,
      "loss": 3.6476,
      "step": 5285
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.68013995885849,
      "learning_rate": 9.432208595667797e-05,
      "loss": 3.7058,
      "step": 5290
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6414756774902344,
      "learning_rate": 9.431135300813989e-05,
      "loss": 3.7271,
      "step": 5295
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6088926196098328,
      "learning_rate": 9.430061053678542e-05,
      "loss": 3.623,
      "step": 5300
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6551048755645752,
      "learning_rate": 9.42898585449232e-05,
      "loss": 3.7571,
      "step": 5305
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6506572365760803,
      "learning_rate": 9.427909703486388e-05,
      "loss": 3.729,
      "step": 5310
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6532192230224609,
      "learning_rate": 9.426832600892019e-05,
      "loss": 3.7489,
      "step": 5315
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6376852989196777,
      "learning_rate": 9.425754546940691e-05,
      "loss": 3.7032,
      "step": 5320
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6398579478263855,
      "learning_rate": 9.424675541864081e-05,
      "loss": 3.6222,
      "step": 5325
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6369754076004028,
      "learning_rate": 9.423595585894082e-05,
      "loss": 3.7356,
      "step": 5330
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8388092517852783,
      "learning_rate": 9.422514679262778e-05,
      "loss": 3.6958,
      "step": 5335
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6580761075019836,
      "learning_rate": 9.421432822202465e-05,
      "loss": 3.713,
      "step": 5340
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6614803075790405,
      "learning_rate": 9.420350014945643e-05,
      "loss": 3.7516,
      "step": 5345
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6593483686447144,
      "learning_rate": 9.419266257725013e-05,
      "loss": 3.7292,
      "step": 5350
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6902868747711182,
      "learning_rate": 9.418181550773483e-05,
      "loss": 3.6948,
      "step": 5355
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6123176217079163,
      "learning_rate": 9.417095894324163e-05,
      "loss": 3.6234,
      "step": 5360
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6449345946311951,
      "learning_rate": 9.41600928861037e-05,
      "loss": 3.6446,
      "step": 5365
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.638054370880127,
      "learning_rate": 9.41492173386562e-05,
      "loss": 3.6357,
      "step": 5370
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6688047647476196,
      "learning_rate": 9.413833230323639e-05,
      "loss": 3.7164,
      "step": 5375
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6318192481994629,
      "learning_rate": 9.412743778218351e-05,
      "loss": 3.672,
      "step": 5380
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6329336166381836,
      "learning_rate": 9.41165337778389e-05,
      "loss": 3.7187,
      "step": 5385
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.658408522605896,
      "learning_rate": 9.410562029254589e-05,
      "loss": 3.5515,
      "step": 5390
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.66224604845047,
      "learning_rate": 9.409469732864986e-05,
      "loss": 3.673,
      "step": 5395
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6777649521827698,
      "learning_rate": 9.408376488849824e-05,
      "loss": 3.7212,
      "step": 5400
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7027958035469055,
      "learning_rate": 9.407282297444048e-05,
      "loss": 3.7322,
      "step": 5405
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6730536818504333,
      "learning_rate": 9.406187158882808e-05,
      "loss": 3.7315,
      "step": 5410
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6176627278327942,
      "learning_rate": 9.405091073401458e-05,
      "loss": 3.6071,
      "step": 5415
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6213515996932983,
      "learning_rate": 9.40399404123555e-05,
      "loss": 3.6307,
      "step": 5420
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6307554841041565,
      "learning_rate": 9.402896062620848e-05,
      "loss": 3.6878,
      "step": 5425
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6286924481391907,
      "learning_rate": 9.401797137793313e-05,
      "loss": 3.6388,
      "step": 5430
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6225752234458923,
      "learning_rate": 9.400697266989112e-05,
      "loss": 3.7247,
      "step": 5435
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6127519011497498,
      "learning_rate": 9.399596450444617e-05,
      "loss": 3.4657,
      "step": 5440
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6851731538772583,
      "learning_rate": 9.398494688396398e-05,
      "loss": 3.6639,
      "step": 5445
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7170841693878174,
      "learning_rate": 9.397391981081232e-05,
      "loss": 3.6518,
      "step": 5450
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6797364354133606,
      "learning_rate": 9.3962883287361e-05,
      "loss": 3.6479,
      "step": 5455
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7039297223091125,
      "learning_rate": 9.395183731598183e-05,
      "loss": 3.5929,
      "step": 5460
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6286032199859619,
      "learning_rate": 9.394078189904867e-05,
      "loss": 3.6493,
      "step": 5465
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.7035037875175476,
      "learning_rate": 9.392971703893741e-05,
      "loss": 3.6224,
      "step": 5470
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6568278074264526,
      "learning_rate": 9.391864273802595e-05,
      "loss": 3.6451,
      "step": 5475
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6430177092552185,
      "learning_rate": 9.390755899869425e-05,
      "loss": 3.7398,
      "step": 5480
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6650875210762024,
      "learning_rate": 9.38964658233243e-05,
      "loss": 3.6274,
      "step": 5485
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6013226509094238,
      "learning_rate": 9.388536321430005e-05,
      "loss": 3.6124,
      "step": 5490
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6057133674621582,
      "learning_rate": 9.387425117400757e-05,
      "loss": 3.707,
      "step": 5495
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6174566745758057,
      "learning_rate": 9.386312970483489e-05,
      "loss": 3.6808,
      "step": 5500
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6537914872169495,
      "learning_rate": 9.385199880917211e-05,
      "loss": 3.7363,
      "step": 5505
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6281108260154724,
      "learning_rate": 9.384085848941132e-05,
      "loss": 3.6029,
      "step": 5510
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6802313327789307,
      "learning_rate": 9.382970874794666e-05,
      "loss": 3.6571,
      "step": 5515
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6608865857124329,
      "learning_rate": 9.381854958717429e-05,
      "loss": 3.7788,
      "step": 5520
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6382231116294861,
      "learning_rate": 9.380738100949238e-05,
      "loss": 3.6338,
      "step": 5525
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6531025767326355,
      "learning_rate": 9.379620301730114e-05,
      "loss": 3.6649,
      "step": 5530
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6510147452354431,
      "learning_rate": 9.378501561300282e-05,
      "loss": 3.7148,
      "step": 5535
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6271425485610962,
      "learning_rate": 9.377381879900162e-05,
      "loss": 3.6267,
      "step": 5540
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6044381260871887,
      "learning_rate": 9.376261257770384e-05,
      "loss": 3.7015,
      "step": 5545
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6446499228477478,
      "learning_rate": 9.37513969515178e-05,
      "loss": 3.601,
      "step": 5550
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6631693840026855,
      "learning_rate": 9.374017192285378e-05,
      "loss": 3.7201,
      "step": 5555
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6370187997817993,
      "learning_rate": 9.372893749412413e-05,
      "loss": 3.778,
      "step": 5560
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6660693287849426,
      "learning_rate": 9.371769366774321e-05,
      "loss": 3.6891,
      "step": 5565
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6291117072105408,
      "learning_rate": 9.370644044612738e-05,
      "loss": 3.6522,
      "step": 5570
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.611329197883606,
      "learning_rate": 9.369517783169505e-05,
      "loss": 3.7306,
      "step": 5575
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.673376739025116,
      "learning_rate": 9.368390582686663e-05,
      "loss": 3.641,
      "step": 5580
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6396419405937195,
      "learning_rate": 9.367262443406456e-05,
      "loss": 3.71,
      "step": 5585
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6642711162567139,
      "learning_rate": 9.366133365571328e-05,
      "loss": 3.6017,
      "step": 5590
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6573745012283325,
      "learning_rate": 9.365003349423924e-05,
      "loss": 3.6857,
      "step": 5595
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6645369529724121,
      "learning_rate": 9.363872395207097e-05,
      "loss": 3.7149,
      "step": 5600
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6369231939315796,
      "learning_rate": 9.362740503163891e-05,
      "loss": 3.6096,
      "step": 5605
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6577723622322083,
      "learning_rate": 9.361607673537563e-05,
      "loss": 3.519,
      "step": 5610
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.660521388053894,
      "learning_rate": 9.36047390657156e-05,
      "loss": 3.7015,
      "step": 5615
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6788069009780884,
      "learning_rate": 9.359339202509543e-05,
      "loss": 3.6674,
      "step": 5620
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7483208179473877,
      "learning_rate": 9.358203561595363e-05,
      "loss": 3.6124,
      "step": 5625
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6983548402786255,
      "learning_rate": 9.357066984073079e-05,
      "loss": 3.6969,
      "step": 5630
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.656515896320343,
      "learning_rate": 9.355929470186948e-05,
      "loss": 3.5438,
      "step": 5635
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6570074558258057,
      "learning_rate": 9.354791020181431e-05,
      "loss": 3.6706,
      "step": 5640
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6258523464202881,
      "learning_rate": 9.353651634301188e-05,
      "loss": 3.7789,
      "step": 5645
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6447742581367493,
      "learning_rate": 9.352511312791083e-05,
      "loss": 3.7588,
      "step": 5650
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6349491477012634,
      "learning_rate": 9.351370055896175e-05,
      "loss": 3.7511,
      "step": 5655
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6817086338996887,
      "learning_rate": 9.350227863861732e-05,
      "loss": 3.5931,
      "step": 5660
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6746085286140442,
      "learning_rate": 9.349084736933215e-05,
      "loss": 3.6319,
      "step": 5665
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6272738575935364,
      "learning_rate": 9.347940675356292e-05,
      "loss": 3.7066,
      "step": 5670
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6712009310722351,
      "learning_rate": 9.346795679376832e-05,
      "loss": 3.7163,
      "step": 5675
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.9074142575263977,
      "learning_rate": 9.345649749240898e-05,
      "loss": 3.7303,
      "step": 5680
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6244702935218811,
      "learning_rate": 9.344502885194763e-05,
      "loss": 3.6351,
      "step": 5685
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6687377691268921,
      "learning_rate": 9.343355087484894e-05,
      "loss": 3.6668,
      "step": 5690
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6336571574211121,
      "learning_rate": 9.342206356357956e-05,
      "loss": 3.6673,
      "step": 5695
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.606383204460144,
      "learning_rate": 9.341056692060828e-05,
      "loss": 3.7283,
      "step": 5700
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.679284393787384,
      "learning_rate": 9.339906094840575e-05,
      "loss": 3.6364,
      "step": 5705
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7427834272384644,
      "learning_rate": 9.33875456494447e-05,
      "loss": 3.6058,
      "step": 5710
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6680914163589478,
      "learning_rate": 9.337602102619985e-05,
      "loss": 3.5716,
      "step": 5715
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6169113516807556,
      "learning_rate": 9.336448708114789e-05,
      "loss": 3.6582,
      "step": 5720
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6662060618400574,
      "learning_rate": 9.33529438167676e-05,
      "loss": 3.8104,
      "step": 5725
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6639431715011597,
      "learning_rate": 9.334139123553967e-05,
      "loss": 3.7137,
      "step": 5730
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6679248809814453,
      "learning_rate": 9.332982933994683e-05,
      "loss": 3.5704,
      "step": 5735
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6353369951248169,
      "learning_rate": 9.331825813247381e-05,
      "loss": 3.6295,
      "step": 5740
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6266745328903198,
      "learning_rate": 9.330667761560739e-05,
      "loss": 3.5429,
      "step": 5745
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6587311029434204,
      "learning_rate": 9.329508779183623e-05,
      "loss": 3.581,
      "step": 5750
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.61961829662323,
      "learning_rate": 9.328348866365111e-05,
      "loss": 3.6455,
      "step": 5755
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6712757349014282,
      "learning_rate": 9.327188023354474e-05,
      "loss": 3.5309,
      "step": 5760
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6361719369888306,
      "learning_rate": 9.326026250401188e-05,
      "loss": 3.6823,
      "step": 5765
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7460502982139587,
      "learning_rate": 9.324863547754924e-05,
      "loss": 3.6253,
      "step": 5770
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6312098503112793,
      "learning_rate": 9.323699915665555e-05,
      "loss": 3.7332,
      "step": 5775
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6962384581565857,
      "learning_rate": 9.322535354383153e-05,
      "loss": 3.6475,
      "step": 5780
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6418073177337646,
      "learning_rate": 9.321369864157992e-05,
      "loss": 3.5619,
      "step": 5785
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6265275478363037,
      "learning_rate": 9.320203445240543e-05,
      "loss": 3.5751,
      "step": 5790
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6218913197517395,
      "learning_rate": 9.319036097881477e-05,
      "loss": 3.6047,
      "step": 5795
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6409159898757935,
      "learning_rate": 9.317867822331667e-05,
      "loss": 3.6499,
      "step": 5800
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6067190766334534,
      "learning_rate": 9.31669861884218e-05,
      "loss": 3.7431,
      "step": 5805
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6058268547058105,
      "learning_rate": 9.315528487664289e-05,
      "loss": 3.7276,
      "step": 5810
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6447632908821106,
      "learning_rate": 9.314357429049462e-05,
      "loss": 3.6851,
      "step": 5815
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.610950767993927,
      "learning_rate": 9.313185443249367e-05,
      "loss": 3.5887,
      "step": 5820
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6097620129585266,
      "learning_rate": 9.312012530515875e-05,
      "loss": 3.6819,
      "step": 5825
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6040717363357544,
      "learning_rate": 9.310838691101048e-05,
      "loss": 3.6389,
      "step": 5830
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.628102719783783,
      "learning_rate": 9.309663925257156e-05,
      "loss": 3.4789,
      "step": 5835
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7199088335037231,
      "learning_rate": 9.308488233236663e-05,
      "loss": 3.6635,
      "step": 5840
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.665327787399292,
      "learning_rate": 9.307311615292234e-05,
      "loss": 3.6998,
      "step": 5845
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6038660407066345,
      "learning_rate": 9.306134071676734e-05,
      "loss": 3.5653,
      "step": 5850
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6003241539001465,
      "learning_rate": 9.30495560264322e-05,
      "loss": 3.5575,
      "step": 5855
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6340860724449158,
      "learning_rate": 9.303776208444958e-05,
      "loss": 3.6162,
      "step": 5860
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6892641186714172,
      "learning_rate": 9.302595889335405e-05,
      "loss": 3.6516,
      "step": 5865
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.592067301273346,
      "learning_rate": 9.301414645568223e-05,
      "loss": 3.6828,
      "step": 5870
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7419446110725403,
      "learning_rate": 9.300232477397266e-05,
      "loss": 3.708,
      "step": 5875
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6518017053604126,
      "learning_rate": 9.299049385076592e-05,
      "loss": 3.5758,
      "step": 5880
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6585884690284729,
      "learning_rate": 9.297865368860456e-05,
      "loss": 3.5911,
      "step": 5885
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5948323607444763,
      "learning_rate": 9.296680429003311e-05,
      "loss": 3.6147,
      "step": 5890
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6271572709083557,
      "learning_rate": 9.295494565759808e-05,
      "loss": 3.634,
      "step": 5895
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6015599370002747,
      "learning_rate": 9.294307779384796e-05,
      "loss": 3.6892,
      "step": 5900
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.7203601598739624,
      "learning_rate": 9.293120070133327e-05,
      "loss": 3.6682,
      "step": 5905
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.626017689704895,
      "learning_rate": 9.291931438260646e-05,
      "loss": 3.7067,
      "step": 5910
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6437251567840576,
      "learning_rate": 9.290741884022198e-05,
      "loss": 3.7392,
      "step": 5915
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6172105669975281,
      "learning_rate": 9.28955140767363e-05,
      "loss": 3.504,
      "step": 5920
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6975985169410706,
      "learning_rate": 9.288360009470776e-05,
      "loss": 3.8211,
      "step": 5925
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6059762239456177,
      "learning_rate": 9.287167689669683e-05,
      "loss": 3.6423,
      "step": 5930
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6702840328216553,
      "learning_rate": 9.285974448526585e-05,
      "loss": 3.5623,
      "step": 5935
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6284217238426208,
      "learning_rate": 9.284780286297918e-05,
      "loss": 3.6705,
      "step": 5940
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.6288384795188904,
      "learning_rate": 9.283585203240317e-05,
      "loss": 3.7277,
      "step": 5945
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6619918346405029,
      "learning_rate": 9.282389199610613e-05,
      "loss": 3.6503,
      "step": 5950
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.695856511592865,
      "learning_rate": 9.281192275665836e-05,
      "loss": 3.6841,
      "step": 5955
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5982354283332825,
      "learning_rate": 9.279994431663212e-05,
      "loss": 3.5915,
      "step": 5960
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6896786093711853,
      "learning_rate": 9.278795667860167e-05,
      "loss": 3.6684,
      "step": 5965
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5959322452545166,
      "learning_rate": 9.277595984514325e-05,
      "loss": 3.6434,
      "step": 5970
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6684930920600891,
      "learning_rate": 9.276395381883502e-05,
      "loss": 3.69,
      "step": 5975
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6274242997169495,
      "learning_rate": 9.275193860225722e-05,
      "loss": 3.6293,
      "step": 5980
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6387612223625183,
      "learning_rate": 9.273991419799193e-05,
      "loss": 3.6725,
      "step": 5985
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6110684275627136,
      "learning_rate": 9.272788060862334e-05,
      "loss": 3.6898,
      "step": 5990
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2275980710983276,
      "learning_rate": 9.271583783673751e-05,
      "loss": 3.6676,
      "step": 5995
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6144578456878662,
      "learning_rate": 9.270378588492255e-05,
      "loss": 3.6788,
      "step": 6000
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6123051047325134,
      "learning_rate": 9.269172475576849e-05,
      "loss": 3.6181,
      "step": 6005
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6386008262634277,
      "learning_rate": 9.267965445186733e-05,
      "loss": 3.7491,
      "step": 6010
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6545584201812744,
      "learning_rate": 9.266757497581308e-05,
      "loss": 3.6518,
      "step": 6015
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5987696647644043,
      "learning_rate": 9.265548633020174e-05,
      "loss": 3.4672,
      "step": 6020
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6671674251556396,
      "learning_rate": 9.264338851763117e-05,
      "loss": 3.6619,
      "step": 6025
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6036574244499207,
      "learning_rate": 9.263128154070133e-05,
      "loss": 3.5954,
      "step": 6030
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6052005290985107,
      "learning_rate": 9.261916540201407e-05,
      "loss": 3.6473,
      "step": 6035
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.0243521928787231,
      "learning_rate": 9.260704010417325e-05,
      "loss": 3.6587,
      "step": 6040
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.613764226436615,
      "learning_rate": 9.259490564978466e-05,
      "loss": 3.6221,
      "step": 6045
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6378306150436401,
      "learning_rate": 9.258276204145609e-05,
      "loss": 3.6011,
      "step": 6050
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.7736099362373352,
      "learning_rate": 9.257060928179726e-05,
      "loss": 3.7215,
      "step": 6055
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.646984338760376,
      "learning_rate": 9.255844737341992e-05,
      "loss": 3.6253,
      "step": 6060
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.7412344813346863,
      "learning_rate": 9.254627631893773e-05,
      "loss": 3.7016,
      "step": 6065
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.710990846157074,
      "learning_rate": 9.253409612096635e-05,
      "loss": 3.6102,
      "step": 6070
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6395878791809082,
      "learning_rate": 9.252190678212335e-05,
      "loss": 3.5317,
      "step": 6075
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.615039587020874,
      "learning_rate": 9.250970830502834e-05,
      "loss": 3.6583,
      "step": 6080
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.7352080941200256,
      "learning_rate": 9.249750069230285e-05,
      "loss": 3.7558,
      "step": 6085
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6084471940994263,
      "learning_rate": 9.248528394657036e-05,
      "loss": 3.6231,
      "step": 6090
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6312777996063232,
      "learning_rate": 9.247305807045637e-05,
      "loss": 3.615,
      "step": 6095
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.626972496509552,
      "learning_rate": 9.246082306658827e-05,
      "loss": 3.6639,
      "step": 6100
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6172739267349243,
      "learning_rate": 9.244857893759545e-05,
      "loss": 3.557,
      "step": 6105
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5875005722045898,
      "learning_rate": 9.243632568610929e-05,
      "loss": 3.7367,
      "step": 6110
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5923462510108948,
      "learning_rate": 9.242406331476307e-05,
      "loss": 3.5937,
      "step": 6115
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6698502898216248,
      "learning_rate": 9.241179182619206e-05,
      "loss": 3.6492,
      "step": 6120
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6432156562805176,
      "learning_rate": 9.239951122303349e-05,
      "loss": 3.6879,
      "step": 6125
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5815703272819519,
      "learning_rate": 9.238722150792658e-05,
      "loss": 3.5971,
      "step": 6130
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.634823739528656,
      "learning_rate": 9.237492268351242e-05,
      "loss": 3.7397,
      "step": 6135
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6127801537513733,
      "learning_rate": 9.236261475243415e-05,
      "loss": 3.7115,
      "step": 6140
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6250133514404297,
      "learning_rate": 9.23502977173368e-05,
      "loss": 3.6788,
      "step": 6145
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5987094044685364,
      "learning_rate": 9.233797158086742e-05,
      "loss": 3.6377,
      "step": 6150
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5959830284118652,
      "learning_rate": 9.232563634567495e-05,
      "loss": 3.6572,
      "step": 6155
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6371004581451416,
      "learning_rate": 9.231329201441034e-05,
      "loss": 3.702,
      "step": 6160
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6539026498794556,
      "learning_rate": 9.230093858972646e-05,
      "loss": 3.6312,
      "step": 6165
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6082867980003357,
      "learning_rate": 9.228857607427816e-05,
      "loss": 3.7105,
      "step": 6170
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5815205574035645,
      "learning_rate": 9.227620447072221e-05,
      "loss": 3.5989,
      "step": 6175
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.608966052532196,
      "learning_rate": 9.226382378171738e-05,
      "loss": 3.6316,
      "step": 6180
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6275756359100342,
      "learning_rate": 9.225143400992435e-05,
      "loss": 3.5835,
      "step": 6185
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6804114580154419,
      "learning_rate": 9.223903515800576e-05,
      "loss": 3.7168,
      "step": 6190
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6216789484024048,
      "learning_rate": 9.222662722862622e-05,
      "loss": 3.6218,
      "step": 6195
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6242992281913757,
      "learning_rate": 9.221421022445228e-05,
      "loss": 3.6211,
      "step": 6200
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6428170800209045,
      "learning_rate": 9.220178414815242e-05,
      "loss": 3.5709,
      "step": 6205
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6073488593101501,
      "learning_rate": 9.218934900239713e-05,
      "loss": 3.6119,
      "step": 6210
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6136678457260132,
      "learning_rate": 9.217690478985878e-05,
      "loss": 3.7387,
      "step": 6215
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5773927569389343,
      "learning_rate": 9.216445151321171e-05,
      "loss": 3.5675,
      "step": 6220
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6443408727645874,
      "learning_rate": 9.215198917513223e-05,
      "loss": 3.6316,
      "step": 6225
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6348914504051208,
      "learning_rate": 9.213951777829861e-05,
      "loss": 3.5198,
      "step": 6230
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6297686696052551,
      "learning_rate": 9.212703732539098e-05,
      "loss": 3.5835,
      "step": 6235
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6197727918624878,
      "learning_rate": 9.211454781909153e-05,
      "loss": 3.603,
      "step": 6240
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6774610877037048,
      "learning_rate": 9.21020492620843e-05,
      "loss": 3.7656,
      "step": 6245
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6623004078865051,
      "learning_rate": 9.208954165705534e-05,
      "loss": 3.577,
      "step": 6250
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6406750679016113,
      "learning_rate": 9.207702500669262e-05,
      "loss": 3.5956,
      "step": 6255
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5959295034408569,
      "learning_rate": 9.206449931368604e-05,
      "loss": 3.4743,
      "step": 6260
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6465792655944824,
      "learning_rate": 9.205196458072747e-05,
      "loss": 3.6595,
      "step": 6265
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6831611394882202,
      "learning_rate": 9.203942081051071e-05,
      "loss": 3.5805,
      "step": 6270
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6011339426040649,
      "learning_rate": 9.20268680057315e-05,
      "loss": 3.7398,
      "step": 6275
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6345436573028564,
      "learning_rate": 9.201430616908752e-05,
      "loss": 3.6244,
      "step": 6280
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.6435637474060059,
      "learning_rate": 9.200173530327842e-05,
      "loss": 3.4508,
      "step": 6285
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6558165550231934,
      "learning_rate": 9.198915541100575e-05,
      "loss": 3.678,
      "step": 6290
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6507555842399597,
      "learning_rate": 9.1976566494973e-05,
      "loss": 3.5086,
      "step": 6295
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.968017339706421,
      "learning_rate": 9.196396855788563e-05,
      "loss": 3.6104,
      "step": 6300
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6781919598579407,
      "learning_rate": 9.195136160245104e-05,
      "loss": 3.6742,
      "step": 6305
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6069244146347046,
      "learning_rate": 9.193874563137853e-05,
      "loss": 3.6004,
      "step": 6310
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.7005805373191833,
      "learning_rate": 9.192612064737934e-05,
      "loss": 3.6106,
      "step": 6315
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6405349969863892,
      "learning_rate": 9.19134866531667e-05,
      "loss": 3.5918,
      "step": 6320
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6165139079093933,
      "learning_rate": 9.190084365145575e-05,
      "loss": 3.6314,
      "step": 6325
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6977831721305847,
      "learning_rate": 9.188819164496354e-05,
      "loss": 3.6207,
      "step": 6330
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6094171404838562,
      "learning_rate": 9.187553063640906e-05,
      "loss": 3.614,
      "step": 6335
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5684142112731934,
      "learning_rate": 9.186286062851327e-05,
      "loss": 3.548,
      "step": 6340
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5863828063011169,
      "learning_rate": 9.185018162399905e-05,
      "loss": 3.5041,
      "step": 6345
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5964145660400391,
      "learning_rate": 9.183749362559118e-05,
      "loss": 3.646,
      "step": 6350
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5859694480895996,
      "learning_rate": 9.182479663601642e-05,
      "loss": 3.6313,
      "step": 6355
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6039595007896423,
      "learning_rate": 9.181209065800342e-05,
      "loss": 3.6152,
      "step": 6360
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5884960293769836,
      "learning_rate": 9.17993756942828e-05,
      "loss": 3.4615,
      "step": 6365
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5988009572029114,
      "learning_rate": 9.17866517475871e-05,
      "loss": 3.5826,
      "step": 6370
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6679307818412781,
      "learning_rate": 9.177391882065077e-05,
      "loss": 3.5082,
      "step": 6375
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6062582731246948,
      "learning_rate": 9.176117691621017e-05,
      "loss": 3.6778,
      "step": 6380
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6468236446380615,
      "learning_rate": 9.17484260370037e-05,
      "loss": 3.5851,
      "step": 6385
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5810919404029846,
      "learning_rate": 9.173566618577156e-05,
      "loss": 3.6776,
      "step": 6390
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6626738905906677,
      "learning_rate": 9.172289736525594e-05,
      "loss": 3.6528,
      "step": 6395
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.66367107629776,
      "learning_rate": 9.171011957820097e-05,
      "loss": 3.6644,
      "step": 6400
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6790648102760315,
      "learning_rate": 9.169733282735265e-05,
      "loss": 3.5701,
      "step": 6405
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6435784101486206,
      "learning_rate": 9.168453711545895e-05,
      "loss": 3.5698,
      "step": 6410
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6229742765426636,
      "learning_rate": 9.167173244526977e-05,
      "loss": 3.641,
      "step": 6415
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6029466986656189,
      "learning_rate": 9.165891881953692e-05,
      "loss": 3.5662,
      "step": 6420
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5852835774421692,
      "learning_rate": 9.164609624101414e-05,
      "loss": 3.5949,
      "step": 6425
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6254869103431702,
      "learning_rate": 9.163326471245707e-05,
      "loss": 3.5645,
      "step": 6430
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6965833902359009,
      "learning_rate": 9.162042423662331e-05,
      "loss": 3.6241,
      "step": 6435
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6012727618217468,
      "learning_rate": 9.160757481627237e-05,
      "loss": 3.5747,
      "step": 6440
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.679567277431488,
      "learning_rate": 9.159471645416567e-05,
      "loss": 3.5661,
      "step": 6445
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6437513828277588,
      "learning_rate": 9.158184915306655e-05,
      "loss": 3.6698,
      "step": 6450
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6113952994346619,
      "learning_rate": 9.156897291574034e-05,
      "loss": 3.5914,
      "step": 6455
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5789247751235962,
      "learning_rate": 9.155608774495415e-05,
      "loss": 3.6495,
      "step": 6460
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.59614497423172,
      "learning_rate": 9.154319364347715e-05,
      "loss": 3.6431,
      "step": 6465
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6178292632102966,
      "learning_rate": 9.153029061408037e-05,
      "loss": 3.6156,
      "step": 6470
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6082344651222229,
      "learning_rate": 9.151737865953671e-05,
      "loss": 3.6467,
      "step": 6475
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5963026285171509,
      "learning_rate": 9.150445778262109e-05,
      "loss": 3.5968,
      "step": 6480
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.678371012210846,
      "learning_rate": 9.149152798611027e-05,
      "loss": 3.6371,
      "step": 6485
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5853849649429321,
      "learning_rate": 9.147858927278296e-05,
      "loss": 3.5786,
      "step": 6490
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5865886211395264,
      "learning_rate": 9.14656416454198e-05,
      "loss": 3.6203,
      "step": 6495
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6075390577316284,
      "learning_rate": 9.14526851068033e-05,
      "loss": 3.7387,
      "step": 6500
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6003337502479553,
      "learning_rate": 9.143971965971789e-05,
      "loss": 3.5012,
      "step": 6505
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.664794921875,
      "learning_rate": 9.142674530694999e-05,
      "loss": 3.6539,
      "step": 6510
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6633325815200806,
      "learning_rate": 9.141376205128784e-05,
      "loss": 3.5997,
      "step": 6515
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6330863833427429,
      "learning_rate": 9.140076989552162e-05,
      "loss": 3.6448,
      "step": 6520
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6024140119552612,
      "learning_rate": 9.138776884244347e-05,
      "loss": 3.617,
      "step": 6525
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5953519344329834,
      "learning_rate": 9.137475889484739e-05,
      "loss": 3.5202,
      "step": 6530
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.611735999584198,
      "learning_rate": 9.13617400555293e-05,
      "loss": 3.5439,
      "step": 6535
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6424632668495178,
      "learning_rate": 9.134871232728706e-05,
      "loss": 3.6314,
      "step": 6540
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6273808479309082,
      "learning_rate": 9.133567571292038e-05,
      "loss": 3.5669,
      "step": 6545
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6691052317619324,
      "learning_rate": 9.132263021523096e-05,
      "loss": 3.6719,
      "step": 6550
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5857458710670471,
      "learning_rate": 9.130957583702236e-05,
      "loss": 3.5734,
      "step": 6555
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6955469846725464,
      "learning_rate": 9.129651258110003e-05,
      "loss": 3.5797,
      "step": 6560
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5966945886611938,
      "learning_rate": 9.128344045027137e-05,
      "loss": 3.5484,
      "step": 6565
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.648418664932251,
      "learning_rate": 9.127035944734568e-05,
      "loss": 3.6428,
      "step": 6570
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6223288178443909,
      "learning_rate": 9.125726957513414e-05,
      "loss": 3.6082,
      "step": 6575
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6146071553230286,
      "learning_rate": 9.124417083644988e-05,
      "loss": 3.5208,
      "step": 6580
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5741205811500549,
      "learning_rate": 9.12310632341079e-05,
      "loss": 3.6599,
      "step": 6585
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.648072361946106,
      "learning_rate": 9.121794677092509e-05,
      "loss": 3.5329,
      "step": 6590
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.597624659538269,
      "learning_rate": 9.120482144972029e-05,
      "loss": 3.6819,
      "step": 6595
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5847507119178772,
      "learning_rate": 9.119168727331424e-05,
      "loss": 3.5743,
      "step": 6600
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6404077410697937,
      "learning_rate": 9.117854424452952e-05,
      "loss": 3.5292,
      "step": 6605
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6022205352783203,
      "learning_rate": 9.116539236619071e-05,
      "loss": 3.519,
      "step": 6610
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6334814429283142,
      "learning_rate": 9.115223164112418e-05,
      "loss": 3.7497,
      "step": 6615
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.8721258640289307,
      "learning_rate": 9.113906207215832e-05,
      "loss": 3.4376,
      "step": 6620
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6144387125968933,
      "learning_rate": 9.112588366212333e-05,
      "loss": 3.5812,
      "step": 6625
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.616588830947876,
      "learning_rate": 9.111269641385134e-05,
      "loss": 3.6377,
      "step": 6630
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6814590692520142,
      "learning_rate": 9.10995003301764e-05,
      "loss": 3.571,
      "step": 6635
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.595249354839325,
      "learning_rate": 9.10862954139344e-05,
      "loss": 3.6674,
      "step": 6640
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6267328262329102,
      "learning_rate": 9.107308166796322e-05,
      "loss": 3.5308,
      "step": 6645
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5850332379341125,
      "learning_rate": 9.105985909510255e-05,
      "loss": 3.6179,
      "step": 6650
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6673199534416199,
      "learning_rate": 9.104662769819404e-05,
      "loss": 3.5432,
      "step": 6655
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6039637923240662,
      "learning_rate": 9.103338748008118e-05,
      "loss": 3.6104,
      "step": 6660
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6271899938583374,
      "learning_rate": 9.102013844360939e-05,
      "loss": 3.6062,
      "step": 6665
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.588601291179657,
      "learning_rate": 9.100688059162599e-05,
      "loss": 3.667,
      "step": 6670
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5945859551429749,
      "learning_rate": 9.099361392698018e-05,
      "loss": 3.5979,
      "step": 6675
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6507734060287476,
      "learning_rate": 9.098033845252307e-05,
      "loss": 3.5821,
      "step": 6680
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5882008075714111,
      "learning_rate": 9.096705417110762e-05,
      "loss": 3.5787,
      "step": 6685
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5724419355392456,
      "learning_rate": 9.095376108558875e-05,
      "loss": 3.6373,
      "step": 6690
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6695144772529602,
      "learning_rate": 9.094045919882322e-05,
      "loss": 3.7305,
      "step": 6695
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6830370426177979,
      "learning_rate": 9.092714851366968e-05,
      "loss": 3.5444,
      "step": 6700
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8158934712409973,
      "learning_rate": 9.091382903298872e-05,
      "loss": 3.6295,
      "step": 6705
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6257811784744263,
      "learning_rate": 9.090050075964276e-05,
      "loss": 3.7174,
      "step": 6710
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5905994176864624,
      "learning_rate": 9.088716369649616e-05,
      "loss": 3.612,
      "step": 6715
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5851759910583496,
      "learning_rate": 9.087381784641515e-05,
      "loss": 3.5161,
      "step": 6720
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6264608502388,
      "learning_rate": 9.086046321226783e-05,
      "loss": 3.6345,
      "step": 6725
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6070705056190491,
      "learning_rate": 9.08470997969242e-05,
      "loss": 3.6562,
      "step": 6730
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5795014500617981,
      "learning_rate": 9.083372760325617e-05,
      "loss": 3.6278,
      "step": 6735
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6471191048622131,
      "learning_rate": 9.08203466341375e-05,
      "loss": 3.5972,
      "step": 6740
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6771706342697144,
      "learning_rate": 9.080695689244385e-05,
      "loss": 3.5645,
      "step": 6745
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7141193747520447,
      "learning_rate": 9.079355838105279e-05,
      "loss": 3.5511,
      "step": 6750
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6002992987632751,
      "learning_rate": 9.078015110284373e-05,
      "loss": 3.5784,
      "step": 6755
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6891971826553345,
      "learning_rate": 9.076673506069801e-05,
      "loss": 3.684,
      "step": 6760
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6122922301292419,
      "learning_rate": 9.07533102574988e-05,
      "loss": 3.5214,
      "step": 6765
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6337205171585083,
      "learning_rate": 9.07398766961312e-05,
      "loss": 3.4895,
      "step": 6770
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.637915849685669,
      "learning_rate": 9.072643437948218e-05,
      "loss": 3.691,
      "step": 6775
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5818666815757751,
      "learning_rate": 9.07129833104406e-05,
      "loss": 3.6058,
      "step": 6780
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6148473620414734,
      "learning_rate": 9.069952349189714e-05,
      "loss": 3.5751,
      "step": 6785
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.8701967000961304,
      "learning_rate": 9.068605492674446e-05,
      "loss": 3.574,
      "step": 6790
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6925360560417175,
      "learning_rate": 9.0672577617877e-05,
      "loss": 3.5583,
      "step": 6795
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6491031646728516,
      "learning_rate": 9.065909156819118e-05,
      "loss": 3.5166,
      "step": 6800
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5695008039474487,
      "learning_rate": 9.06455967805852e-05,
      "loss": 3.5519,
      "step": 6805
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6708154082298279,
      "learning_rate": 9.063209325795921e-05,
      "loss": 3.54,
      "step": 6810
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6315829753875732,
      "learning_rate": 9.06185810032152e-05,
      "loss": 3.6595,
      "step": 6815
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6070363521575928,
      "learning_rate": 9.060506001925707e-05,
      "loss": 3.5702,
      "step": 6820
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6195346713066101,
      "learning_rate": 9.059153030899051e-05,
      "loss": 3.6997,
      "step": 6825
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6161662936210632,
      "learning_rate": 9.05779918753232e-05,
      "loss": 3.6189,
      "step": 6830
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5959815382957458,
      "learning_rate": 9.056444472116465e-05,
      "loss": 3.5487,
      "step": 6835
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6107397079467773,
      "learning_rate": 9.055088884942622e-05,
      "loss": 3.6382,
      "step": 6840
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.610625684261322,
      "learning_rate": 9.053732426302116e-05,
      "loss": 3.4782,
      "step": 6845
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6375144720077515,
      "learning_rate": 9.052375096486458e-05,
      "loss": 3.5106,
      "step": 6850
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.640093207359314,
      "learning_rate": 9.051016895787352e-05,
      "loss": 3.6933,
      "step": 6855
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6102331876754761,
      "learning_rate": 9.049657824496678e-05,
      "loss": 3.5619,
      "step": 6860
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5834786295890808,
      "learning_rate": 9.048297882906516e-05,
      "loss": 3.6086,
      "step": 6865
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7011825442314148,
      "learning_rate": 9.046937071309124e-05,
      "loss": 3.5739,
      "step": 6870
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6685060858726501,
      "learning_rate": 9.04557538999695e-05,
      "loss": 3.5737,
      "step": 6875
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6052543520927429,
      "learning_rate": 9.04421283926263e-05,
      "loss": 3.461,
      "step": 6880
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6505953669548035,
      "learning_rate": 9.042849419398982e-05,
      "loss": 3.533,
      "step": 6885
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5944388508796692,
      "learning_rate": 9.041485130699018e-05,
      "loss": 3.611,
      "step": 6890
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.63641357421875,
      "learning_rate": 9.040119973455934e-05,
      "loss": 3.4501,
      "step": 6895
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.700008749961853,
      "learning_rate": 9.038753947963108e-05,
      "loss": 3.4921,
      "step": 6900
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5949934124946594,
      "learning_rate": 9.03738705451411e-05,
      "loss": 3.5667,
      "step": 6905
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6110278367996216,
      "learning_rate": 9.036019293402697e-05,
      "loss": 3.5798,
      "step": 6910
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.7038879990577698,
      "learning_rate": 9.034650664922807e-05,
      "loss": 3.6259,
      "step": 6915
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6816919445991516,
      "learning_rate": 9.033281169368568e-05,
      "loss": 3.51,
      "step": 6920
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6436446309089661,
      "learning_rate": 9.031910807034296e-05,
      "loss": 3.5742,
      "step": 6925
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5824409127235413,
      "learning_rate": 9.03053957821449e-05,
      "loss": 3.5673,
      "step": 6930
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6068187952041626,
      "learning_rate": 9.029167483203837e-05,
      "loss": 3.6403,
      "step": 6935
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6095534563064575,
      "learning_rate": 9.027794522297211e-05,
      "loss": 3.5818,
      "step": 6940
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5963279604911804,
      "learning_rate": 9.026420695789668e-05,
      "loss": 3.5195,
      "step": 6945
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6488138437271118,
      "learning_rate": 9.025046003976457e-05,
      "loss": 3.5798,
      "step": 6950
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.6168668270111084,
      "learning_rate": 9.023670447153003e-05,
      "loss": 3.678,
      "step": 6955
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5797130465507507,
      "learning_rate": 9.022294025614926e-05,
      "loss": 3.551,
      "step": 6960
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5982871055603027,
      "learning_rate": 9.02091673965803e-05,
      "loss": 3.4867,
      "step": 6965
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5738404393196106,
      "learning_rate": 9.0195385895783e-05,
      "loss": 3.5454,
      "step": 6970
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6368952393531799,
      "learning_rate": 9.018159575671914e-05,
      "loss": 3.5836,
      "step": 6975
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6302484273910522,
      "learning_rate": 9.016779698235227e-05,
      "loss": 3.5608,
      "step": 6980
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6043098568916321,
      "learning_rate": 9.015398957564787e-05,
      "loss": 3.5575,
      "step": 6985
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5688521862030029,
      "learning_rate": 9.014017353957324e-05,
      "loss": 3.4991,
      "step": 6990
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5898637771606445,
      "learning_rate": 9.012634887709754e-05,
      "loss": 3.543,
      "step": 6995
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5719659924507141,
      "learning_rate": 9.01125155911918e-05,
      "loss": 3.5972,
      "step": 7000
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5865889191627502,
      "learning_rate": 9.009867368482887e-05,
      "loss": 3.5141,
      "step": 7005
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.7704794406890869,
      "learning_rate": 9.008482316098348e-05,
      "loss": 3.6008,
      "step": 7010
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6483211517333984,
      "learning_rate": 9.007096402263221e-05,
      "loss": 3.5989,
      "step": 7015
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.025435209274292,
      "learning_rate": 9.005709627275348e-05,
      "loss": 3.5176,
      "step": 7020
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6267207860946655,
      "learning_rate": 9.004321991432753e-05,
      "loss": 3.5604,
      "step": 7025
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6454509496688843,
      "learning_rate": 9.002933495033657e-05,
      "loss": 3.553,
      "step": 7030
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6539108753204346,
      "learning_rate": 9.001544138376447e-05,
      "loss": 3.5748,
      "step": 7035
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6201371550559998,
      "learning_rate": 9.000153921759714e-05,
      "loss": 3.5863,
      "step": 7040
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6270084977149963,
      "learning_rate": 8.998762845482223e-05,
      "loss": 3.5132,
      "step": 7045
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6105707883834839,
      "learning_rate": 8.997370909842921e-05,
      "loss": 3.4429,
      "step": 7050
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6434949636459351,
      "learning_rate": 8.99597811514095e-05,
      "loss": 3.4811,
      "step": 7055
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.7177697420120239,
      "learning_rate": 8.99458446167563e-05,
      "loss": 3.5908,
      "step": 7060
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5924736261367798,
      "learning_rate": 8.993189949746468e-05,
      "loss": 3.517,
      "step": 7065
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5866804718971252,
      "learning_rate": 8.99179457965315e-05,
      "loss": 3.5137,
      "step": 7070
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6293784976005554,
      "learning_rate": 8.990398351695555e-05,
      "loss": 3.5589,
      "step": 7075
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5955253839492798,
      "learning_rate": 8.989001266173738e-05,
      "loss": 3.4918,
      "step": 7080
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5830385088920593,
      "learning_rate": 8.987603323387948e-05,
      "loss": 3.5162,
      "step": 7085
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6224161386489868,
      "learning_rate": 8.986204523638606e-05,
      "loss": 3.584,
      "step": 7090
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5759114027023315,
      "learning_rate": 8.984804867226328e-05,
      "loss": 3.6106,
      "step": 7095
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5731534361839294,
      "learning_rate": 8.98340435445191e-05,
      "loss": 3.5851,
      "step": 7100
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6229199767112732,
      "learning_rate": 8.982002985616331e-05,
      "loss": 3.544,
      "step": 7105
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6196799278259277,
      "learning_rate": 8.980600761020754e-05,
      "loss": 3.467,
      "step": 7110
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5989375710487366,
      "learning_rate": 8.979197680966527e-05,
      "loss": 3.5417,
      "step": 7115
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6297645568847656,
      "learning_rate": 8.977793745755182e-05,
      "loss": 3.6671,
      "step": 7120
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6044521927833557,
      "learning_rate": 8.976388955688434e-05,
      "loss": 3.5335,
      "step": 7125
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5773036479949951,
      "learning_rate": 8.974983311068184e-05,
      "loss": 3.5591,
      "step": 7130
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5920732021331787,
      "learning_rate": 8.973576812196513e-05,
      "loss": 3.5548,
      "step": 7135
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6109089255332947,
      "learning_rate": 8.972169459375686e-05,
      "loss": 3.5663,
      "step": 7140
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6437410116195679,
      "learning_rate": 8.970761252908156e-05,
      "loss": 3.5592,
      "step": 7145
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5664073824882507,
      "learning_rate": 8.969352193096554e-05,
      "loss": 3.5121,
      "step": 7150
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6073634624481201,
      "learning_rate": 8.967942280243699e-05,
      "loss": 3.493,
      "step": 7155
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6586644649505615,
      "learning_rate": 8.966531514652588e-05,
      "loss": 3.5006,
      "step": 7160
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.62142413854599,
      "learning_rate": 8.965119896626406e-05,
      "loss": 3.544,
      "step": 7165
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6072076559066772,
      "learning_rate": 8.963707426468519e-05,
      "loss": 3.6376,
      "step": 7170
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6039524674415588,
      "learning_rate": 8.962294104482478e-05,
      "loss": 3.5711,
      "step": 7175
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5711743831634521,
      "learning_rate": 8.960879930972014e-05,
      "loss": 3.6351,
      "step": 7180
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5628252029418945,
      "learning_rate": 8.959464906241044e-05,
      "loss": 3.5012,
      "step": 7185
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6577252745628357,
      "learning_rate": 8.958049030593665e-05,
      "loss": 3.4604,
      "step": 7190
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6371538639068604,
      "learning_rate": 8.956632304334158e-05,
      "loss": 3.5749,
      "step": 7195
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5645363330841064,
      "learning_rate": 8.955214727766989e-05,
      "loss": 3.5743,
      "step": 7200
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6304428577423096,
      "learning_rate": 8.953796301196807e-05,
      "loss": 3.548,
      "step": 7205
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5620599985122681,
      "learning_rate": 8.952377024928437e-05,
      "loss": 3.6211,
      "step": 7210
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5865871906280518,
      "learning_rate": 8.950956899266895e-05,
      "loss": 3.5616,
      "step": 7215
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5945559740066528,
      "learning_rate": 8.949535924517374e-05,
      "loss": 3.5095,
      "step": 7220
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6027511358261108,
      "learning_rate": 8.948114100985252e-05,
      "loss": 3.4283,
      "step": 7225
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5858290791511536,
      "learning_rate": 8.946691428976088e-05,
      "loss": 3.5367,
      "step": 7230
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6115530133247375,
      "learning_rate": 8.945267908795625e-05,
      "loss": 3.5585,
      "step": 7235
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.57766193151474,
      "learning_rate": 8.943843540749785e-05,
      "loss": 3.5059,
      "step": 7240
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5847091674804688,
      "learning_rate": 8.942418325144679e-05,
      "loss": 3.5182,
      "step": 7245
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6164994239807129,
      "learning_rate": 8.940992262286592e-05,
      "loss": 3.6382,
      "step": 7250
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6507517695426941,
      "learning_rate": 8.939565352481996e-05,
      "loss": 3.5568,
      "step": 7255
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5813296437263489,
      "learning_rate": 8.938137596037544e-05,
      "loss": 3.5178,
      "step": 7260
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6194491982460022,
      "learning_rate": 8.93670899326007e-05,
      "loss": 3.4812,
      "step": 7265
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5926390886306763,
      "learning_rate": 8.93527954445659e-05,
      "loss": 3.5327,
      "step": 7270
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.677503764629364,
      "learning_rate": 8.933849249934306e-05,
      "loss": 3.5543,
      "step": 7275
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6172518730163574,
      "learning_rate": 8.932418110000594e-05,
      "loss": 3.5056,
      "step": 7280
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6466969847679138,
      "learning_rate": 8.930986124963016e-05,
      "loss": 3.5868,
      "step": 7285
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5908235907554626,
      "learning_rate": 8.92955329512932e-05,
      "loss": 3.4444,
      "step": 7290
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.672362208366394,
      "learning_rate": 8.928119620807426e-05,
      "loss": 3.5163,
      "step": 7295
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.66802579164505,
      "learning_rate": 8.926685102305443e-05,
      "loss": 3.4768,
      "step": 7300
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.6499452590942383,
      "learning_rate": 8.925249739931659e-05,
      "loss": 3.6281,
      "step": 7305
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5987485647201538,
      "learning_rate": 8.923813533994543e-05,
      "loss": 3.5545,
      "step": 7310
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.610182523727417,
      "learning_rate": 8.922376484802745e-05,
      "loss": 3.5992,
      "step": 7315
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8330290913581848,
      "learning_rate": 8.920938592665098e-05,
      "loss": 3.497,
      "step": 7320
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6577211022377014,
      "learning_rate": 8.919499857890613e-05,
      "loss": 3.5281,
      "step": 7325
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6530924439430237,
      "learning_rate": 8.918060280788486e-05,
      "loss": 3.4609,
      "step": 7330
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5722185373306274,
      "learning_rate": 8.916619861668092e-05,
      "loss": 3.5758,
      "step": 7335
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.0003070831298828,
      "learning_rate": 8.915178600838985e-05,
      "loss": 3.4847,
      "step": 7340
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.600658655166626,
      "learning_rate": 8.913736498610906e-05,
      "loss": 3.5447,
      "step": 7345
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6058965921401978,
      "learning_rate": 8.912293555293767e-05,
      "loss": 3.5013,
      "step": 7350
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6027334332466125,
      "learning_rate": 8.910849771197671e-05,
      "loss": 3.4534,
      "step": 7355
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7254871129989624,
      "learning_rate": 8.909405146632896e-05,
      "loss": 3.5719,
      "step": 7360
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6278379559516907,
      "learning_rate": 8.907959681909902e-05,
      "loss": 3.552,
      "step": 7365
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.592115581035614,
      "learning_rate": 8.906513377339327e-05,
      "loss": 3.5147,
      "step": 7370
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6572948098182678,
      "learning_rate": 8.905066233231995e-05,
      "loss": 3.5169,
      "step": 7375
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5802149176597595,
      "learning_rate": 8.903618249898906e-05,
      "loss": 3.5188,
      "step": 7380
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6030763387680054,
      "learning_rate": 8.902169427651242e-05,
      "loss": 3.5097,
      "step": 7385
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6219210028648376,
      "learning_rate": 8.900719766800365e-05,
      "loss": 3.5152,
      "step": 7390
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6297764778137207,
      "learning_rate": 8.899269267657817e-05,
      "loss": 3.5511,
      "step": 7395
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6542302370071411,
      "learning_rate": 8.89781793053532e-05,
      "loss": 3.4734,
      "step": 7400
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7868180871009827,
      "learning_rate": 8.896365755744775e-05,
      "loss": 3.5185,
      "step": 7405
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6043024659156799,
      "learning_rate": 8.894912743598268e-05,
      "loss": 3.4987,
      "step": 7410
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6141871809959412,
      "learning_rate": 8.893458894408058e-05,
      "loss": 3.5125,
      "step": 7415
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.1623255014419556,
      "learning_rate": 8.89200420848659e-05,
      "loss": 3.5328,
      "step": 7420
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5948125720024109,
      "learning_rate": 8.890548686146483e-05,
      "loss": 3.5266,
      "step": 7425
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6076386570930481,
      "learning_rate": 8.889092327700541e-05,
      "loss": 3.4681,
      "step": 7430
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6311365365982056,
      "learning_rate": 8.887635133461745e-05,
      "loss": 3.5524,
      "step": 7435
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6062614917755127,
      "learning_rate": 8.886177103743257e-05,
      "loss": 3.5978,
      "step": 7440
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5807971954345703,
      "learning_rate": 8.884718238858416e-05,
      "loss": 3.5638,
      "step": 7445
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5995590090751648,
      "learning_rate": 8.883258539120742e-05,
      "loss": 3.5552,
      "step": 7450
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.643273651599884,
      "learning_rate": 8.881798004843938e-05,
      "loss": 3.496,
      "step": 7455
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6321567893028259,
      "learning_rate": 8.880336636341878e-05,
      "loss": 3.5047,
      "step": 7460
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6418941020965576,
      "learning_rate": 8.878874433928623e-05,
      "loss": 3.5702,
      "step": 7465
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6537648439407349,
      "learning_rate": 8.877411397918411e-05,
      "loss": 3.5627,
      "step": 7470
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6156619787216187,
      "learning_rate": 8.875947528625658e-05,
      "loss": 3.4358,
      "step": 7475
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6063427925109863,
      "learning_rate": 8.874482826364958e-05,
      "loss": 3.5379,
      "step": 7480
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5953460335731506,
      "learning_rate": 8.873017291451086e-05,
      "loss": 3.5107,
      "step": 7485
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.667246401309967,
      "learning_rate": 8.871550924198999e-05,
      "loss": 3.5201,
      "step": 7490
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6311941742897034,
      "learning_rate": 8.870083724923826e-05,
      "loss": 3.4702,
      "step": 7495
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5834953784942627,
      "learning_rate": 8.868615693940879e-05,
      "loss": 3.6,
      "step": 7500
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.7096093893051147,
      "learning_rate": 8.86714683156565e-05,
      "loss": 3.4905,
      "step": 7505
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6487685441970825,
      "learning_rate": 8.865677138113803e-05,
      "loss": 3.4489,
      "step": 7510
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6087823510169983,
      "learning_rate": 8.864206613901189e-05,
      "loss": 3.5198,
      "step": 7515
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6358893513679504,
      "learning_rate": 8.862735259243832e-05,
      "loss": 3.511,
      "step": 7520
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.596430778503418,
      "learning_rate": 8.861263074457941e-05,
      "loss": 3.5597,
      "step": 7525
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5572250485420227,
      "learning_rate": 8.859790059859893e-05,
      "loss": 3.5909,
      "step": 7530
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.568970263004303,
      "learning_rate": 8.858316215766252e-05,
      "loss": 3.5541,
      "step": 7535
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6048108339309692,
      "learning_rate": 8.856841542493754e-05,
      "loss": 3.5652,
      "step": 7540
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5846671462059021,
      "learning_rate": 8.855366040359321e-05,
      "loss": 3.3629,
      "step": 7545
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5682613253593445,
      "learning_rate": 8.853889709680045e-05,
      "loss": 3.5622,
      "step": 7550
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6080896258354187,
      "learning_rate": 8.852412550773202e-05,
      "loss": 3.6509,
      "step": 7555
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6251019239425659,
      "learning_rate": 8.850934563956244e-05,
      "loss": 3.5056,
      "step": 7560
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.9435561299324036,
      "learning_rate": 8.849455749546797e-05,
      "loss": 3.5072,
      "step": 7565
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5656173229217529,
      "learning_rate": 8.847976107862673e-05,
      "loss": 3.6198,
      "step": 7570
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5905395150184631,
      "learning_rate": 8.846495639221854e-05,
      "loss": 3.5957,
      "step": 7575
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5923274159431458,
      "learning_rate": 8.845014343942505e-05,
      "loss": 3.4418,
      "step": 7580
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5674275159835815,
      "learning_rate": 8.843532222342965e-05,
      "loss": 3.5745,
      "step": 7585
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5745781660079956,
      "learning_rate": 8.842049274741753e-05,
      "loss": 3.494,
      "step": 7590
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5866777300834656,
      "learning_rate": 8.840565501457565e-05,
      "loss": 3.6826,
      "step": 7595
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5641412138938904,
      "learning_rate": 8.839080902809274e-05,
      "loss": 3.5453,
      "step": 7600
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5640811920166016,
      "learning_rate": 8.83759547911593e-05,
      "loss": 3.4015,
      "step": 7605
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5902920365333557,
      "learning_rate": 8.83610923069676e-05,
      "loss": 3.5226,
      "step": 7610
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5834364295005798,
      "learning_rate": 8.83462215787117e-05,
      "loss": 3.7474,
      "step": 7615
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5967845916748047,
      "learning_rate": 8.833134260958745e-05,
      "loss": 3.6181,
      "step": 7620
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5804824829101562,
      "learning_rate": 8.831645540279238e-05,
      "loss": 3.5134,
      "step": 7625
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5917492508888245,
      "learning_rate": 8.830155996152591e-05,
      "loss": 3.4907,
      "step": 7630
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5945274829864502,
      "learning_rate": 8.828665628898918e-05,
      "loss": 3.375,
      "step": 7635
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6050729751586914,
      "learning_rate": 8.827174438838503e-05,
      "loss": 3.4622,
      "step": 7640
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5743839740753174,
      "learning_rate": 8.825682426291819e-05,
      "loss": 3.4933,
      "step": 7645
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6049296259880066,
      "learning_rate": 8.824189591579505e-05,
      "loss": 3.4459,
      "step": 7650
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6671527028083801,
      "learning_rate": 8.822695935022387e-05,
      "loss": 3.5631,
      "step": 7655
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6094750165939331,
      "learning_rate": 8.821201456941458e-05,
      "loss": 3.4512,
      "step": 7660
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6009117960929871,
      "learning_rate": 8.819706157657893e-05,
      "loss": 3.6082,
      "step": 7665
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.591812252998352,
      "learning_rate": 8.818210037493042e-05,
      "loss": 3.5465,
      "step": 7670
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.579538881778717,
      "learning_rate": 8.816713096768431e-05,
      "loss": 3.5407,
      "step": 7675
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6216089725494385,
      "learning_rate": 8.815215335805765e-05,
      "loss": 3.5524,
      "step": 7680
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5829979181289673,
      "learning_rate": 8.81371675492692e-05,
      "loss": 3.4794,
      "step": 7685
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.593366265296936,
      "learning_rate": 8.812217354453956e-05,
      "loss": 3.4285,
      "step": 7690
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5687240362167358,
      "learning_rate": 8.810717134709099e-05,
      "loss": 3.3687,
      "step": 7695
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.58964604139328,
      "learning_rate": 8.809216096014759e-05,
      "loss": 3.5439,
      "step": 7700
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5834695100784302,
      "learning_rate": 8.807714238693521e-05,
      "loss": 3.492,
      "step": 7705
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.608650267124176,
      "learning_rate": 8.806211563068145e-05,
      "loss": 3.5707,
      "step": 7710
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5998702645301819,
      "learning_rate": 8.804708069461563e-05,
      "loss": 3.5524,
      "step": 7715
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6006673574447632,
      "learning_rate": 8.803203758196887e-05,
      "loss": 3.5236,
      "step": 7720
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6444456577301025,
      "learning_rate": 8.801698629597407e-05,
      "loss": 3.4808,
      "step": 7725
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6684592962265015,
      "learning_rate": 8.800192683986583e-05,
      "loss": 3.4458,
      "step": 7730
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5931656360626221,
      "learning_rate": 8.798685921688053e-05,
      "loss": 3.4858,
      "step": 7735
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6286415457725525,
      "learning_rate": 8.797178343025631e-05,
      "loss": 3.446,
      "step": 7740
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5778789520263672,
      "learning_rate": 8.795669948323308e-05,
      "loss": 3.4408,
      "step": 7745
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6765512824058533,
      "learning_rate": 8.794160737905245e-05,
      "loss": 3.5104,
      "step": 7750
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5898998379707336,
      "learning_rate": 8.792650712095784e-05,
      "loss": 3.4419,
      "step": 7755
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5976765155792236,
      "learning_rate": 8.791139871219441e-05,
      "loss": 3.5347,
      "step": 7760
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5881169438362122,
      "learning_rate": 8.789628215600902e-05,
      "loss": 3.4074,
      "step": 7765
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5646470785140991,
      "learning_rate": 8.788115745565036e-05,
      "loss": 3.4967,
      "step": 7770
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6034087538719177,
      "learning_rate": 8.786602461436884e-05,
      "loss": 3.544,
      "step": 7775
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5686057806015015,
      "learning_rate": 8.785088363541657e-05,
      "loss": 3.5843,
      "step": 7780
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6128042936325073,
      "learning_rate": 8.78357345220475e-05,
      "loss": 3.5679,
      "step": 7785
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6089234948158264,
      "learning_rate": 8.782057727751722e-05,
      "loss": 3.535,
      "step": 7790
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.588179886341095,
      "learning_rate": 8.78054119050832e-05,
      "loss": 3.5179,
      "step": 7795
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5967940092086792,
      "learning_rate": 8.779023840800452e-05,
      "loss": 3.456,
      "step": 7800
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5849707126617432,
      "learning_rate": 8.777505678954211e-05,
      "loss": 3.464,
      "step": 7805
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6934648156166077,
      "learning_rate": 8.775986705295858e-05,
      "loss": 3.5303,
      "step": 7810
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.924924910068512,
      "learning_rate": 8.774466920151833e-05,
      "loss": 3.5763,
      "step": 7815
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5875627994537354,
      "learning_rate": 8.772946323848746e-05,
      "loss": 3.4861,
      "step": 7820
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5833917856216431,
      "learning_rate": 8.771424916713386e-05,
      "loss": 3.5592,
      "step": 7825
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6373021006584167,
      "learning_rate": 8.769902699072712e-05,
      "loss": 3.5631,
      "step": 7830
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5927019119262695,
      "learning_rate": 8.768379671253859e-05,
      "loss": 3.6129,
      "step": 7835
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5892027616500854,
      "learning_rate": 8.76685583358414e-05,
      "loss": 3.5132,
      "step": 7840
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5818294882774353,
      "learning_rate": 8.765331186391033e-05,
      "loss": 3.5326,
      "step": 7845
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5821009874343872,
      "learning_rate": 8.763805730002199e-05,
      "loss": 3.4772,
      "step": 7850
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5985898375511169,
      "learning_rate": 8.762279464745466e-05,
      "loss": 3.5566,
      "step": 7855
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6195432543754578,
      "learning_rate": 8.760752390948843e-05,
      "loss": 3.4931,
      "step": 7860
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7456226348876953,
      "learning_rate": 8.759224508940505e-05,
      "loss": 3.5742,
      "step": 7865
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5818597078323364,
      "learning_rate": 8.757695819048808e-05,
      "loss": 3.5416,
      "step": 7870
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.76542067527771,
      "learning_rate": 8.756166321602273e-05,
      "loss": 3.6407,
      "step": 7875
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.583821713924408,
      "learning_rate": 8.754636016929604e-05,
      "loss": 3.5353,
      "step": 7880
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5945658683776855,
      "learning_rate": 8.753104905359674e-05,
      "loss": 3.4333,
      "step": 7885
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5563685297966003,
      "learning_rate": 8.751572987221526e-05,
      "loss": 3.4463,
      "step": 7890
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.215826392173767,
      "learning_rate": 8.750040262844384e-05,
      "loss": 3.4546,
      "step": 7895
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5955301523208618,
      "learning_rate": 8.748506732557637e-05,
      "loss": 3.5934,
      "step": 7900
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6070915460586548,
      "learning_rate": 8.746972396690853e-05,
      "loss": 3.5119,
      "step": 7905
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5966696739196777,
      "learning_rate": 8.745437255573773e-05,
      "loss": 3.5278,
      "step": 7910
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6042582392692566,
      "learning_rate": 8.743901309536307e-05,
      "loss": 3.5883,
      "step": 7915
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.668655514717102,
      "learning_rate": 8.742364558908543e-05,
      "loss": 3.487,
      "step": 7920
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5582161545753479,
      "learning_rate": 8.740827004020735e-05,
      "loss": 3.4782,
      "step": 7925
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6002620458602905,
      "learning_rate": 8.739288645203319e-05,
      "loss": 3.5881,
      "step": 7930
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6492757201194763,
      "learning_rate": 8.737749482786897e-05,
      "loss": 3.496,
      "step": 7935
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.61732017993927,
      "learning_rate": 8.736209517102247e-05,
      "loss": 3.5241,
      "step": 7940
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5936669111251831,
      "learning_rate": 8.734668748480315e-05,
      "loss": 3.4722,
      "step": 7945
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5613396763801575,
      "learning_rate": 8.733127177252225e-05,
      "loss": 3.5153,
      "step": 7950
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5918736457824707,
      "learning_rate": 8.731584803749271e-05,
      "loss": 3.5113,
      "step": 7955
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5804104208946228,
      "learning_rate": 8.73004162830292e-05,
      "loss": 3.6084,
      "step": 7960
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5645163059234619,
      "learning_rate": 8.728497651244812e-05,
      "loss": 3.4528,
      "step": 7965
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.6022102236747742,
      "learning_rate": 8.726952872906755e-05,
      "loss": 3.5959,
      "step": 7970
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5786009430885315,
      "learning_rate": 8.725407293620737e-05,
      "loss": 3.4871,
      "step": 7975
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5923308730125427,
      "learning_rate": 8.72386091371891e-05,
      "loss": 3.4014,
      "step": 7980
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.579049289226532,
      "learning_rate": 8.722313733533603e-05,
      "loss": 3.4881,
      "step": 7985
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6084596514701843,
      "learning_rate": 8.720765753397317e-05,
      "loss": 3.5408,
      "step": 7990
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5969850420951843,
      "learning_rate": 8.719216973642723e-05,
      "loss": 3.4607,
      "step": 7995
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.6145032644271851,
      "learning_rate": 8.717667394602664e-05,
      "loss": 3.4901,
      "step": 8000
    }
  ],
  "logging_steps": 5,
  "max_steps": 33984,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 2.208473472093389e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}