{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.8313847752663029,
  "eval_steps": 500,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 122.31757354736328,
      "learning_rate": 5e-06,
      "loss": 12.8799,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 52.391082763671875,
      "learning_rate": 1e-05,
      "loss": 12.6172,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 29.490076065063477,
      "learning_rate": 1.5e-05,
      "loss": 11.8088,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 31.53589630126953,
      "learning_rate": 2e-05,
      "loss": 10.9282,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 61.79519271850586,
      "learning_rate": 2.5e-05,
      "loss": 10.1323,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.654998779296875,
      "learning_rate": 3e-05,
      "loss": 9.4484,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 5.740698337554932,
      "learning_rate": 3.5e-05,
      "loss": 8.8873,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.908849716186523,
      "learning_rate": 4e-05,
      "loss": 8.3969,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.360279083251953,
      "learning_rate": 4.5e-05,
      "loss": 8.1293,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.529783248901367,
      "learning_rate": 5e-05,
      "loss": 7.9849,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.2711334228515625,
      "learning_rate": 5.500000000000001e-05,
      "loss": 7.9591,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.8138530254364014,
      "learning_rate": 6e-05,
      "loss": 7.8974,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 3.322045087814331,
      "learning_rate": 6.500000000000001e-05,
      "loss": 7.9111,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 2.2868151664733887,
      "learning_rate": 7e-05,
      "loss": 7.9229,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.699502468109131,
      "learning_rate": 7.500000000000001e-05,
      "loss": 7.9082,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.17010760307312,
      "learning_rate": 8e-05,
      "loss": 7.914,
      "step": 80
    },
    {
      "epoch": 0.02,
      "grad_norm": 6.845041751861572,
      "learning_rate": 8.5e-05,
      "loss": 7.8946,
      "step": 85
    },
    {
      "epoch": 0.02,
      "grad_norm": 2.585634708404541,
      "learning_rate": 9e-05,
      "loss": 7.8897,
      "step": 90
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.026464462280273,
      "learning_rate": 9.5e-05,
      "loss": 7.8652,
      "step": 95
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.074096202850342,
      "learning_rate": 0.0001,
      "loss": 7.8336,
      "step": 100
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.498573303222656,
      "learning_rate": 9.999972205865686e-05,
      "loss": 7.7889,
      "step": 105
    },
    {
      "epoch": 0.02,
      "grad_norm": 5.521847724914551,
      "learning_rate": 9.999888823771751e-05,
      "loss": 7.7641,
      "step": 110
    },
    {
      "epoch": 0.02,
      "grad_norm": 8.344034194946289,
      "learning_rate": 9.999749854645204e-05,
      "loss": 7.7462,
      "step": 115
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.066705703735352,
      "learning_rate": 9.99955530003106e-05,
      "loss": 7.6825,
      "step": 120
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.9992802143096924,
      "learning_rate": 9.99930516209231e-05,
      "loss": 7.6298,
      "step": 125
    },
    {
      "epoch": 0.03,
      "grad_norm": 19.802461624145508,
      "learning_rate": 9.998999443609897e-05,
      "loss": 7.6001,
      "step": 130
    },
    {
      "epoch": 0.03,
      "grad_norm": 4.808511734008789,
      "learning_rate": 9.998638147982696e-05,
      "loss": 7.505,
      "step": 135
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.9778096675872803,
      "learning_rate": 9.998221279227467e-05,
      "loss": 7.473,
      "step": 140
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.969865560531616,
      "learning_rate": 9.997748841978812e-05,
      "loss": 7.3771,
      "step": 145
    },
    {
      "epoch": 0.03,
      "grad_norm": 6.17680549621582,
      "learning_rate": 9.997220841489122e-05,
      "loss": 7.3354,
      "step": 150
    },
    {
      "epoch": 0.03,
      "grad_norm": 5.137447357177734,
      "learning_rate": 9.996637283628528e-05,
      "loss": 7.228,
      "step": 155
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.0998005867004395,
      "learning_rate": 9.995998174884821e-05,
      "loss": 7.183,
      "step": 160
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.1692376136779785,
      "learning_rate": 9.995303522363394e-05,
      "loss": 7.0672,
      "step": 165
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.8801262378692627,
      "learning_rate": 9.99455333378715e-05,
      "loss": 7.0342,
      "step": 170
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.4446799755096436,
      "learning_rate": 9.993747617496428e-05,
      "loss": 6.9368,
      "step": 175
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.6471118927001953,
      "learning_rate": 9.9928863824489e-05,
      "loss": 6.8468,
      "step": 180
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.5301289558410645,
      "learning_rate": 9.99196963821948e-05,
      "loss": 6.7799,
      "step": 185
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.7082278728485107,
      "learning_rate": 9.990997395000217e-05,
      "loss": 6.6638,
      "step": 190
    },
    {
      "epoch": 0.04,
      "grad_norm": 3.3497087955474854,
      "learning_rate": 9.989969663600169e-05,
      "loss": 6.5953,
      "step": 195
    },
    {
      "epoch": 0.04,
      "grad_norm": 4.278082370758057,
      "learning_rate": 9.9888864554453e-05,
      "loss": 6.606,
      "step": 200
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.66937518119812,
      "learning_rate": 9.987747782578342e-05,
      "loss": 6.5057,
      "step": 205
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.571780204772949,
      "learning_rate": 9.986553657658668e-05,
      "loss": 6.4385,
      "step": 210
    },
    {
      "epoch": 0.04,
      "grad_norm": 5.088606834411621,
      "learning_rate": 9.985304093962145e-05,
      "loss": 6.3923,
      "step": 215
    },
    {
      "epoch": 0.05,
      "grad_norm": 5.042826175689697,
      "learning_rate": 9.983999105380988e-05,
      "loss": 6.3688,
      "step": 220
    },
    {
      "epoch": 0.05,
      "grad_norm": 7.568378925323486,
      "learning_rate": 9.982638706423608e-05,
      "loss": 6.2207,
      "step": 225
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.896998405456543,
      "learning_rate": 9.98122291221445e-05,
      "loss": 6.1763,
      "step": 230
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.744915723800659,
      "learning_rate": 9.979751738493826e-05,
      "loss": 6.2032,
      "step": 235
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.402007102966309,
      "learning_rate": 9.978225201617732e-05,
      "loss": 6.1301,
      "step": 240
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.604529857635498,
      "learning_rate": 9.976643318557678e-05,
      "loss": 6.1229,
      "step": 245
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.7452666759490967,
      "learning_rate": 9.975006106900495e-05,
      "loss": 6.0317,
      "step": 250
    },
    {
      "epoch": 0.05,
      "grad_norm": 4.346206188201904,
      "learning_rate": 9.973313584848132e-05,
      "loss": 5.9815,
      "step": 255
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.731186628341675,
      "learning_rate": 9.971565771217464e-05,
      "loss": 5.9493,
      "step": 260
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.345651865005493,
      "learning_rate": 9.969762685440076e-05,
      "loss": 5.9823,
      "step": 265
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.3746602535247803,
      "learning_rate": 9.967904347562054e-05,
      "loss": 5.8495,
      "step": 270
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.518210411071777,
      "learning_rate": 9.965990778243755e-05,
      "loss": 5.9199,
      "step": 275
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.3639678955078125,
      "learning_rate": 9.964021998759577e-05,
      "loss": 5.8102,
      "step": 280
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.2886486053466797,
      "learning_rate": 9.961998030997733e-05,
      "loss": 5.816,
      "step": 285
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.6282382011413574,
      "learning_rate": 9.95991889745999e-05,
      "loss": 5.7934,
      "step": 290
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.633136749267578,
      "learning_rate": 9.957784621261441e-05,
      "loss": 5.7084,
      "step": 295
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.706786155700684,
      "learning_rate": 9.955595226130226e-05,
      "loss": 5.726,
      "step": 300
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.696900367736816,
      "learning_rate": 9.953350736407282e-05,
      "loss": 5.741,
      "step": 305
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.607839345932007,
      "learning_rate": 9.951051177046069e-05,
      "loss": 5.6393,
      "step": 310
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.6848416328430176,
      "learning_rate": 9.948696573612292e-05,
      "loss": 5.6151,
      "step": 315
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.3749618530273438,
      "learning_rate": 9.946286952283618e-05,
      "loss": 5.5774,
      "step": 320
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.431562662124634,
      "learning_rate": 9.943822339849381e-05,
      "loss": 5.6212,
      "step": 325
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.472083806991577,
      "learning_rate": 9.941302763710288e-05,
      "loss": 5.578,
      "step": 330
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.792559862136841,
      "learning_rate": 9.938728251878116e-05,
      "loss": 5.5623,
      "step": 335
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.234732627868652,
      "learning_rate": 9.936098832975393e-05,
      "loss": 5.5339,
      "step": 340
    },
    {
      "epoch": 0.07,
      "grad_norm": 4.907464981079102,
      "learning_rate": 9.933414536235091e-05,
      "loss": 5.4366,
      "step": 345
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.1314311027526855,
      "learning_rate": 9.93067539150029e-05,
      "loss": 5.4759,
      "step": 350
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.1746504306793213,
      "learning_rate": 9.927881429223853e-05,
      "loss": 5.4098,
      "step": 355
    },
    {
      "epoch": 0.07,
      "grad_norm": 3.957454204559326,
      "learning_rate": 9.925032680468085e-05,
      "loss": 5.4251,
      "step": 360
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.583669900894165,
      "learning_rate": 9.922129176904388e-05,
      "loss": 5.4224,
      "step": 365
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.548809051513672,
      "learning_rate": 9.919170950812911e-05,
      "loss": 5.3477,
      "step": 370
    },
    {
      "epoch": 0.08,
      "grad_norm": 5.560839653015137,
      "learning_rate": 9.916158035082184e-05,
      "loss": 5.3688,
      "step": 375
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.999515533447266,
      "learning_rate": 9.913090463208763e-05,
      "loss": 5.4106,
      "step": 380
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.346980571746826,
      "learning_rate": 9.90996826929685e-05,
      "loss": 5.3671,
      "step": 385
    },
    {
      "epoch": 0.08,
      "grad_norm": 4.390699863433838,
      "learning_rate": 9.906791488057916e-05,
      "loss": 5.3009,
      "step": 390
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.8730216026306152,
      "learning_rate": 9.903560154810313e-05,
      "loss": 5.2528,
      "step": 395
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.8211958408355713,
      "learning_rate": 9.900274305478887e-05,
      "loss": 5.278,
      "step": 400
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.9018919467926025,
      "learning_rate": 9.896933976594572e-05,
      "loss": 5.2639,
      "step": 405
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.902015209197998,
      "learning_rate": 9.893539205293989e-05,
      "loss": 5.3053,
      "step": 410
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.365139961242676,
      "learning_rate": 9.890090029319028e-05,
      "loss": 5.2269,
      "step": 415
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.671627521514893,
      "learning_rate": 9.886586487016433e-05,
      "loss": 5.2006,
      "step": 420
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.9795711040496826,
      "learning_rate": 9.883028617337378e-05,
      "loss": 5.1443,
      "step": 425
    },
    {
      "epoch": 0.09,
      "grad_norm": 4.8376240730285645,
      "learning_rate": 9.879416459837022e-05,
      "loss": 5.1789,
      "step": 430
    },
    {
      "epoch": 0.09,
      "grad_norm": 74.21269226074219,
      "learning_rate": 9.875750054674082e-05,
      "loss": 5.164,
      "step": 435
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.394662618637085,
      "learning_rate": 9.872029442610382e-05,
      "loss": 5.229,
      "step": 440
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.8581583499908447,
      "learning_rate": 9.8682546650104e-05,
      "loss": 5.1748,
      "step": 445
    },
    {
      "epoch": 0.09,
      "grad_norm": 2.873598098754883,
      "learning_rate": 9.864425763840802e-05,
      "loss": 5.0188,
      "step": 450
    },
    {
      "epoch": 0.09,
      "grad_norm": 3.0684592723846436,
      "learning_rate": 9.860542781669988e-05,
      "loss": 5.0698,
      "step": 455
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.1893513202667236,
      "learning_rate": 9.85660576166761e-05,
      "loss": 5.0876,
      "step": 460
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.5682992935180664,
      "learning_rate": 9.852614747604093e-05,
      "loss": 5.012,
      "step": 465
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.143974781036377,
      "learning_rate": 9.848569783850145e-05,
      "loss": 5.0468,
      "step": 470
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.436988115310669,
      "learning_rate": 9.844470915376278e-05,
      "loss": 5.0128,
      "step": 475
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.3220009803771973,
      "learning_rate": 9.840318187752292e-05,
      "loss": 5.0599,
      "step": 480
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.984196901321411,
      "learning_rate": 9.836111647146771e-05,
      "loss": 4.9826,
      "step": 485
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.470804452896118,
      "learning_rate": 9.831851340326577e-05,
      "loss": 4.9622,
      "step": 490
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.7063026428222656,
      "learning_rate": 9.82753731465633e-05,
      "loss": 4.9792,
      "step": 495
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.18554401397705,
      "learning_rate": 9.823169618097871e-05,
      "loss": 4.8994,
      "step": 500
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.004983901977539,
      "learning_rate": 9.81874829920974e-05,
      "loss": 4.8866,
      "step": 505
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.266251564025879,
      "learning_rate": 9.814273407146623e-05,
      "loss": 4.8267,
      "step": 510
    },
    {
      "epoch": 0.11,
      "grad_norm": 4.0071024894714355,
      "learning_rate": 9.809744991658829e-05,
      "loss": 4.8254,
      "step": 515
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.009719133377075,
      "learning_rate": 9.805163103091708e-05,
      "loss": 4.8355,
      "step": 520
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.7107043266296387,
      "learning_rate": 9.800527792385112e-05,
      "loss": 4.9058,
      "step": 525
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.9798905849456787,
      "learning_rate": 9.79583911107282e-05,
      "loss": 4.8755,
      "step": 530
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.1199405193328857,
      "learning_rate": 9.791097111281968e-05,
      "loss": 4.877,
      "step": 535
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.966737747192383,
      "learning_rate": 9.786301845732467e-05,
      "loss": 4.8143,
      "step": 540
    },
    {
      "epoch": 0.11,
      "grad_norm": 2.3689117431640625,
      "learning_rate": 9.781453367736418e-05,
      "loss": 4.7737,
      "step": 545
    },
    {
      "epoch": 0.11,
      "grad_norm": 3.9130005836486816,
      "learning_rate": 9.776551731197524e-05,
      "loss": 4.849,
      "step": 550
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.18276047706604,
      "learning_rate": 9.771596990610478e-05,
      "loss": 4.7412,
      "step": 555
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.2997982501983643,
      "learning_rate": 9.766589201060372e-05,
      "loss": 4.7399,
      "step": 560
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.285614490509033,
      "learning_rate": 9.761528418222077e-05,
      "loss": 4.7657,
      "step": 565
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.5080738067626953,
      "learning_rate": 9.756414698359624e-05,
      "loss": 4.764,
      "step": 570
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.743520736694336,
      "learning_rate": 9.75124809832558e-05,
      "loss": 4.642,
      "step": 575
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.485870838165283,
      "learning_rate": 9.746028675560413e-05,
      "loss": 4.7463,
      "step": 580
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.757502794265747,
      "learning_rate": 9.740756488091861e-05,
      "loss": 4.6997,
      "step": 585
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.4718565940856934,
      "learning_rate": 9.735431594534277e-05,
      "loss": 4.6539,
      "step": 590
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.387404680252075,
      "learning_rate": 9.730054054087983e-05,
      "loss": 4.5988,
      "step": 595
    },
    {
      "epoch": 0.12,
      "grad_norm": 3.5060672760009766,
      "learning_rate": 9.724623926538612e-05,
      "loss": 4.6711,
      "step": 600
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.424964189529419,
      "learning_rate": 9.719141272256443e-05,
      "loss": 4.5616,
      "step": 605
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.4947853088378906,
      "learning_rate": 9.713606152195726e-05,
      "loss": 4.5994,
      "step": 610
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.7293379306793213,
      "learning_rate": 9.708018627894011e-05,
      "loss": 4.6421,
      "step": 615
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.5259528160095215,
      "learning_rate": 9.702378761471456e-05,
      "loss": 4.6012,
      "step": 620
    },
    {
      "epoch": 0.13,
      "grad_norm": 5.21116828918457,
      "learning_rate": 9.696686615630146e-05,
      "loss": 4.6564,
      "step": 625
    },
    {
      "epoch": 0.13,
      "grad_norm": 3.3022258281707764,
      "learning_rate": 9.690942253653385e-05,
      "loss": 4.6354,
      "step": 630
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.6259093284606934,
      "learning_rate": 9.685145739405002e-05,
      "loss": 4.4983,
      "step": 635
    },
    {
      "epoch": 0.13,
      "grad_norm": 2.3440661430358887,
      "learning_rate": 9.679297137328634e-05,
      "loss": 4.5076,
      "step": 640
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.7991883754730225,
      "learning_rate": 9.673396512447013e-05,
      "loss": 4.5399,
      "step": 645
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.315129041671753,
      "learning_rate": 9.667443930361247e-05,
      "loss": 4.5895,
      "step": 650
    },
    {
      "epoch": 0.14,
      "grad_norm": 13.325892448425293,
      "learning_rate": 9.661439457250076e-05,
      "loss": 4.522,
      "step": 655
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.741748332977295,
      "learning_rate": 9.655383159869158e-05,
      "loss": 4.4874,
      "step": 660
    },
    {
      "epoch": 0.14,
      "grad_norm": 6.063040256500244,
      "learning_rate": 9.649275105550309e-05,
      "loss": 4.6068,
      "step": 665
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.7268357276916504,
      "learning_rate": 9.643115362200762e-05,
      "loss": 4.4608,
      "step": 670
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.668646812438965,
      "learning_rate": 9.636903998302409e-05,
      "loss": 4.4986,
      "step": 675
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.6662468910217285,
      "learning_rate": 9.630641082911045e-05,
      "loss": 4.4831,
      "step": 680
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.068556785583496,
      "learning_rate": 9.624326685655593e-05,
      "loss": 4.5347,
      "step": 685
    },
    {
      "epoch": 0.14,
      "grad_norm": 3.145664930343628,
      "learning_rate": 9.617960876737337e-05,
      "loss": 4.3987,
      "step": 690
    },
    {
      "epoch": 0.14,
      "grad_norm": 2.836777687072754,
      "learning_rate": 9.611543726929134e-05,
      "loss": 4.4232,
      "step": 695
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.4169726371765137,
      "learning_rate": 9.605075307574635e-05,
      "loss": 4.419,
      "step": 700
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.363471508026123,
      "learning_rate": 9.598555690587487e-05,
      "loss": 4.4505,
      "step": 705
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.95708429813385,
      "learning_rate": 9.591984948450532e-05,
      "loss": 4.4358,
      "step": 710
    },
    {
      "epoch": 0.15,
      "grad_norm": 17.262592315673828,
      "learning_rate": 9.585363154215008e-05,
      "loss": 4.4204,
      "step": 715
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.599637269973755,
      "learning_rate": 9.578690381499728e-05,
      "loss": 4.4039,
      "step": 720
    },
    {
      "epoch": 0.15,
      "grad_norm": 4.246088027954102,
      "learning_rate": 9.571966704490271e-05,
      "loss": 4.3817,
      "step": 725
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.036510944366455,
      "learning_rate": 9.565192197938148e-05,
      "loss": 4.3224,
      "step": 730
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.655491590499878,
      "learning_rate": 9.558366937159977e-05,
      "loss": 4.3792,
      "step": 735
    },
    {
      "epoch": 0.15,
      "grad_norm": 2.5865752696990967,
      "learning_rate": 9.551490998036646e-05,
      "loss": 4.3374,
      "step": 740
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5395338535308838,
      "learning_rate": 9.544564457012463e-05,
      "loss": 4.3361,
      "step": 745
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.922009825706482,
      "learning_rate": 9.537587391094314e-05,
      "loss": 4.3745,
      "step": 750
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.555393934249878,
      "learning_rate": 9.5305598778508e-05,
      "loss": 4.3643,
      "step": 755
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.469945192337036,
      "learning_rate": 9.52348199541138e-05,
      "loss": 4.3793,
      "step": 760
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8235751390457153,
      "learning_rate": 9.516353822465504e-05,
      "loss": 4.282,
      "step": 765
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.8429944515228271,
      "learning_rate": 9.509175438261726e-05,
      "loss": 4.3457,
      "step": 770
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.9383187294006348,
      "learning_rate": 9.501946922606838e-05,
      "loss": 4.3501,
      "step": 775
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.5326876640319824,
      "learning_rate": 9.494668355864973e-05,
      "loss": 4.3599,
      "step": 780
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.431361198425293,
      "learning_rate": 9.487339818956716e-05,
      "loss": 4.3081,
      "step": 785
    },
    {
      "epoch": 0.16,
      "grad_norm": 2.223879098892212,
      "learning_rate": 9.479961393358203e-05,
      "loss": 4.2371,
      "step": 790
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4740023612976074,
      "learning_rate": 9.472533161100215e-05,
      "loss": 4.3332,
      "step": 795
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.1612377166748047,
      "learning_rate": 9.465055204767265e-05,
      "loss": 4.2704,
      "step": 800
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9543275833129883,
      "learning_rate": 9.457527607496685e-05,
      "loss": 4.2932,
      "step": 805
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.7458062171936035,
      "learning_rate": 9.44995045297769e-05,
      "loss": 4.1922,
      "step": 810
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.9623520374298096,
      "learning_rate": 9.442323825450464e-05,
      "loss": 4.2933,
      "step": 815
    },
    {
      "epoch": 0.17,
      "grad_norm": 3.298856735229492,
      "learning_rate": 9.43464780970521e-05,
      "loss": 4.1951,
      "step": 820
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.488687753677368,
      "learning_rate": 9.426922491081212e-05,
      "loss": 4.1681,
      "step": 825
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.890091896057129,
      "learning_rate": 9.419147955465888e-05,
      "loss": 4.2496,
      "step": 830
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.8966683149337769,
      "learning_rate": 9.411324289293832e-05,
      "loss": 4.2597,
      "step": 835
    },
    {
      "epoch": 0.17,
      "grad_norm": 2.0964736938476562,
      "learning_rate": 9.403451579545859e-05,
      "loss": 4.2082,
      "step": 840
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.873693823814392,
      "learning_rate": 9.395529913748025e-05,
      "loss": 4.2124,
      "step": 845
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.7219654321670532,
      "learning_rate": 9.387559379970672e-05,
      "loss": 4.1612,
      "step": 850
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.8131452798843384,
      "learning_rate": 9.379540066827431e-05,
      "loss": 4.2244,
      "step": 855
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3947608470916748,
      "learning_rate": 9.371472063474248e-05,
      "loss": 4.1973,
      "step": 860
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.269024610519409,
      "learning_rate": 9.363355459608394e-05,
      "loss": 4.2322,
      "step": 865
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.323319435119629,
      "learning_rate": 9.355190345467457e-05,
      "loss": 4.2178,
      "step": 870
    },
    {
      "epoch": 0.18,
      "grad_norm": 3.0089669227600098,
      "learning_rate": 9.346976811828352e-05,
      "loss": 4.1682,
      "step": 875
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.6600987911224365,
      "learning_rate": 9.338714950006297e-05,
      "loss": 4.1985,
      "step": 880
    },
    {
      "epoch": 0.18,
      "grad_norm": 2.142987012863159,
      "learning_rate": 9.330404851853817e-05,
      "loss": 4.1924,
      "step": 885
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.8982244729995728,
      "learning_rate": 9.3220466097597e-05,
      "loss": 4.1807,
      "step": 890
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.6516717672348022,
      "learning_rate": 9.313640316647991e-05,
      "loss": 4.1747,
      "step": 895
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.066444158554077,
      "learning_rate": 9.305186065976945e-05,
      "loss": 4.0994,
      "step": 900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.427220106124878,
      "learning_rate": 9.296683951737993e-05,
      "loss": 4.0989,
      "step": 905
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4269256591796875,
      "learning_rate": 9.288134068454697e-05,
      "loss": 4.1823,
      "step": 910
    },
    {
      "epoch": 0.19,
      "grad_norm": 2.578458070755005,
      "learning_rate": 9.2795365111817e-05,
      "loss": 4.1248,
      "step": 915
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.8514119386672974,
      "learning_rate": 9.270891375503665e-05,
      "loss": 4.2063,
      "step": 920
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.622624158859253,
      "learning_rate": 9.262198757534218e-05,
      "loss": 4.1254,
      "step": 925
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.5997337102890015,
      "learning_rate": 9.253458753914874e-05,
      "loss": 4.0921,
      "step": 930
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.544672966003418,
      "learning_rate": 9.244671461813969e-05,
      "loss": 4.1679,
      "step": 935
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.013326406478882,
      "learning_rate": 9.235836978925572e-05,
      "loss": 4.0929,
      "step": 940
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.0170209407806396,
      "learning_rate": 9.226955403468406e-05,
      "loss": 4.2025,
      "step": 945
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.877813696861267,
      "learning_rate": 9.21802683418475e-05,
      "loss": 4.1673,
      "step": 950
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.38816499710083,
      "learning_rate": 9.209051370339347e-05,
      "loss": 4.0372,
      "step": 955
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.6840617656707764,
      "learning_rate": 9.200029111718295e-05,
      "loss": 4.0814,
      "step": 960
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.1929237842559814,
      "learning_rate": 9.190960158627941e-05,
      "loss": 4.1309,
      "step": 965
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.2283923625946045,
      "learning_rate": 9.181844611893766e-05,
      "loss": 4.1529,
      "step": 970
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4596413373947144,
      "learning_rate": 9.172682572859261e-05,
      "loss": 4.1775,
      "step": 975
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5317531824111938,
      "learning_rate": 9.163474143384806e-05,
      "loss": 3.9884,
      "step": 980
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.49446964263916,
      "learning_rate": 9.154219425846528e-05,
      "loss": 4.1084,
      "step": 985
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.7860472202301025,
      "learning_rate": 9.144918523135175e-05,
      "loss": 4.0325,
      "step": 990
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.552417516708374,
      "learning_rate": 9.13557153865496e-05,
      "loss": 4.0353,
      "step": 995
    },
    {
      "epoch": 0.21,
      "grad_norm": 2.3973443508148193,
      "learning_rate": 9.12617857632242e-05,
      "loss": 4.064,
      "step": 1000
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8256909847259521,
      "learning_rate": 9.116739740565259e-05,
      "loss": 4.0357,
      "step": 1005
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6800718307495117,
      "learning_rate": 9.107255136321184e-05,
      "loss": 4.0421,
      "step": 1010
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.8746205568313599,
      "learning_rate": 9.09772486903674e-05,
      "loss": 3.992,
      "step": 1015
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4675569534301758,
      "learning_rate": 9.08814904466614e-05,
      "loss": 4.0656,
      "step": 1020
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.790181040763855,
      "learning_rate": 9.078527769670085e-05,
      "loss": 3.9793,
      "step": 1025
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.5290518999099731,
      "learning_rate": 9.068861151014575e-05,
      "loss": 3.9731,
      "step": 1030
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6314709186553955,
      "learning_rate": 9.05914929616973e-05,
      "loss": 4.001,
      "step": 1035
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6790775060653687,
      "learning_rate": 9.04939231310859e-05,
      "loss": 3.9733,
      "step": 1040
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.706252336502075,
      "learning_rate": 9.039590310305914e-05,
      "loss": 3.9434,
      "step": 1045
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.976989507675171,
      "learning_rate": 9.029743396736974e-05,
      "loss": 4.0118,
      "step": 1050
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6745636463165283,
      "learning_rate": 9.019851681876348e-05,
      "loss": 3.9574,
      "step": 1055
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4342231750488281,
      "learning_rate": 9.009915275696693e-05,
      "loss": 3.977,
      "step": 1060
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3511041402816772,
      "learning_rate": 8.999934288667534e-05,
      "loss": 4.0192,
      "step": 1065
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.1652772426605225,
      "learning_rate": 8.989908831754028e-05,
      "loss": 3.996,
      "step": 1070
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3712204694747925,
      "learning_rate": 8.979839016415735e-05,
      "loss": 3.9809,
      "step": 1075
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.658094048500061,
      "learning_rate": 8.969724954605373e-05,
      "loss": 3.9934,
      "step": 1080
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5667743682861328,
      "learning_rate": 8.959566758767581e-05,
      "loss": 4.0291,
      "step": 1085
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2648104429244995,
      "learning_rate": 8.949364541837661e-05,
      "loss": 4.007,
      "step": 1090
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.144688606262207,
      "learning_rate": 8.939118417240329e-05,
      "loss": 3.9619,
      "step": 1095
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.567773461341858,
      "learning_rate": 8.92882849888845e-05,
      "loss": 3.9128,
      "step": 1100
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5329694747924805,
      "learning_rate": 8.918494901181773e-05,
      "loss": 3.8977,
      "step": 1105
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.1690361499786377,
      "learning_rate": 8.908117739005659e-05,
      "loss": 4.0269,
      "step": 1110
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.9766258001327515,
      "learning_rate": 8.897697127729805e-05,
      "loss": 4.0738,
      "step": 1115
    },
    {
      "epoch": 0.23,
      "grad_norm": 2.04693865776062,
      "learning_rate": 8.887233183206957e-05,
      "loss": 3.9546,
      "step": 1120
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5112673044204712,
      "learning_rate": 8.876726021771627e-05,
      "loss": 4.0082,
      "step": 1125
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.8021695613861084,
      "learning_rate": 8.866175760238798e-05,
      "loss": 3.9489,
      "step": 1130
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.141040563583374,
      "learning_rate": 8.855582515902625e-05,
      "loss": 3.9274,
      "step": 1135
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.5247093439102173,
      "learning_rate": 8.844946406535131e-05,
      "loss": 3.9396,
      "step": 1140
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6850353479385376,
      "learning_rate": 8.834267550384893e-05,
      "loss": 3.9595,
      "step": 1145
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.969480037689209,
      "learning_rate": 8.823546066175741e-05,
      "loss": 3.9362,
      "step": 1150
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3904099464416504,
      "learning_rate": 8.81278207310542e-05,
      "loss": 3.8443,
      "step": 1155
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.679237961769104,
      "learning_rate": 8.801975690844278e-05,
      "loss": 3.9694,
      "step": 1160
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.516084909439087,
      "learning_rate": 8.791127039533934e-05,
      "loss": 3.9411,
      "step": 1165
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3940407037734985,
      "learning_rate": 8.780236239785935e-05,
      "loss": 3.949,
      "step": 1170
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.5583488941192627,
      "learning_rate": 8.76930341268042e-05,
      "loss": 3.9738,
      "step": 1175
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3650261163711548,
      "learning_rate": 8.758328679764776e-05,
      "loss": 3.9532,
      "step": 1180
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3828051090240479,
      "learning_rate": 8.747312163052284e-05,
      "loss": 3.8283,
      "step": 1185
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6516921520233154,
      "learning_rate": 8.736253985020761e-05,
      "loss": 3.8717,
      "step": 1190
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.7323459386825562,
      "learning_rate": 8.725154268611203e-05,
      "loss": 3.8824,
      "step": 1195
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.829392910003662,
      "learning_rate": 8.714013137226411e-05,
      "loss": 3.9052,
      "step": 1200
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3618247509002686,
      "learning_rate": 8.702830714729628e-05,
      "loss": 3.8535,
      "step": 1205
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2153682708740234,
      "learning_rate": 8.691607125443153e-05,
      "loss": 3.88,
      "step": 1210
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5574041604995728,
      "learning_rate": 8.680342494146967e-05,
      "loss": 3.8169,
      "step": 1215
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6505224704742432,
      "learning_rate": 8.66903694607734e-05,
      "loss": 3.8834,
      "step": 1220
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.343554139137268,
      "learning_rate": 8.65769060692544e-05,
      "loss": 3.8068,
      "step": 1225
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5355747938156128,
      "learning_rate": 8.646303602835936e-05,
      "loss": 3.9175,
      "step": 1230
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2311254739761353,
      "learning_rate": 8.634876060405597e-05,
      "loss": 3.82,
      "step": 1235
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1889828443527222,
      "learning_rate": 8.623408106681884e-05,
      "loss": 3.8909,
      "step": 1240
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.758866548538208,
      "learning_rate": 8.611899869161535e-05,
      "loss": 3.9627,
      "step": 1245
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4952363967895508,
      "learning_rate": 8.600351475789147e-05,
      "loss": 3.8229,
      "step": 1250
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1656522750854492,
      "learning_rate": 8.588763054955764e-05,
      "loss": 3.8274,
      "step": 1255
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5959575176239014,
      "learning_rate": 8.57713473549743e-05,
      "loss": 3.7958,
      "step": 1260
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.9823359251022339,
      "learning_rate": 8.565466646693778e-05,
      "loss": 3.9781,
      "step": 1265
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4883431196212769,
      "learning_rate": 8.553758918266578e-05,
      "loss": 3.7693,
      "step": 1270
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.681728482246399,
      "learning_rate": 8.5420116803783e-05,
      "loss": 3.774,
      "step": 1275
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.252820611000061,
      "learning_rate": 8.530225063630668e-05,
      "loss": 3.7305,
      "step": 1280
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.5661219358444214,
      "learning_rate": 8.518399199063205e-05,
      "loss": 3.8674,
      "step": 1285
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3736493587493896,
      "learning_rate": 8.50653421815178e-05,
      "loss": 3.9066,
      "step": 1290
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.458081603050232,
      "learning_rate": 8.494630252807138e-05,
      "loss": 3.6953,
      "step": 1295
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.8416239023208618,
      "learning_rate": 8.482687435373449e-05,
      "loss": 3.8985,
      "step": 1300
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2914624214172363,
      "learning_rate": 8.470705898626817e-05,
      "loss": 3.9091,
      "step": 1305
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4390232563018799,
      "learning_rate": 8.458685775773822e-05,
      "loss": 3.7945,
      "step": 1310
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.513663411140442,
      "learning_rate": 8.446627200450025e-05,
      "loss": 3.8353,
      "step": 1315
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3679267168045044,
      "learning_rate": 8.434530306718493e-05,
      "loss": 3.8589,
      "step": 1320
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.641858458518982,
      "learning_rate": 8.4223952290683e-05,
      "loss": 3.8278,
      "step": 1325
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3191187381744385,
      "learning_rate": 8.41022210241304e-05,
      "loss": 3.7688,
      "step": 1330
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3148295879364014,
      "learning_rate": 8.398011062089316e-05,
      "loss": 3.7961,
      "step": 1335
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.6754995584487915,
      "learning_rate": 8.385762243855249e-05,
      "loss": 3.8242,
      "step": 1340
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.521572232246399,
      "learning_rate": 8.373475783888958e-05,
      "loss": 3.805,
      "step": 1345
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.6051355600357056,
      "learning_rate": 8.36115181878705e-05,
      "loss": 3.8183,
      "step": 1350
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.645491123199463,
      "learning_rate": 8.348790485563101e-05,
      "loss": 3.8065,
      "step": 1355
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2536430358886719,
      "learning_rate": 8.336391921646134e-05,
      "loss": 3.7664,
      "step": 1360
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.691973328590393,
      "learning_rate": 8.323956264879089e-05,
      "loss": 3.7289,
      "step": 1365
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.4711138010025024,
      "learning_rate": 8.311483653517294e-05,
      "loss": 3.8468,
      "step": 1370
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3313995599746704,
      "learning_rate": 8.298974226226919e-05,
      "loss": 3.8141,
      "step": 1375
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.632800579071045,
      "learning_rate": 8.28642812208345e-05,
      "loss": 3.7395,
      "step": 1380
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.1504483222961426,
      "learning_rate": 8.273845480570123e-05,
      "loss": 3.802,
      "step": 1385
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.0767090320587158,
      "learning_rate": 8.26122644157639e-05,
      "loss": 3.7864,
      "step": 1390
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.284472107887268,
      "learning_rate": 8.248571145396362e-05,
      "loss": 3.6997,
      "step": 1395
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.715078353881836,
      "learning_rate": 8.235879732727236e-05,
      "loss": 3.8206,
      "step": 1400
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.6855794191360474,
      "learning_rate": 8.223152344667745e-05,
      "loss": 3.7517,
      "step": 1405
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.481239914894104,
      "learning_rate": 8.21038912271658e-05,
      "loss": 3.7843,
      "step": 1410
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.2233805656433105,
      "learning_rate": 8.197590208770824e-05,
      "loss": 3.8117,
      "step": 1415
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.6338953971862793,
      "learning_rate": 8.184755745124371e-05,
      "loss": 3.7475,
      "step": 1420
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.2792235612869263,
      "learning_rate": 8.171885874466342e-05,
      "loss": 3.7921,
      "step": 1425
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.2128022909164429,
      "learning_rate": 8.158980739879507e-05,
      "loss": 3.7265,
      "step": 1430
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4311732053756714,
      "learning_rate": 8.146040484838677e-05,
      "loss": 3.7395,
      "step": 1435
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.2157390117645264,
      "learning_rate": 8.133065253209132e-05,
      "loss": 3.6868,
      "step": 1440
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.5604244470596313,
      "learning_rate": 8.120055189245e-05,
      "loss": 3.7265,
      "step": 1445
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1982959508895874,
      "learning_rate": 8.10701043758767e-05,
      "loss": 3.7419,
      "step": 1450
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.1536376476287842,
      "learning_rate": 8.093931143264174e-05,
      "loss": 3.7326,
      "step": 1455
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.265428900718689,
      "learning_rate": 8.080817451685576e-05,
      "loss": 3.7235,
      "step": 1460
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3499423265457153,
      "learning_rate": 8.067669508645356e-05,
      "loss": 3.7707,
      "step": 1465
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.43877112865448,
      "learning_rate": 8.054487460317797e-05,
      "loss": 3.7049,
      "step": 1470
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4240456819534302,
      "learning_rate": 8.041271453256345e-05,
      "loss": 3.6702,
      "step": 1475
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.505245566368103,
      "learning_rate": 8.02802163439199e-05,
      "loss": 3.7357,
      "step": 1480
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.620792269706726,
      "learning_rate": 8.01473815103163e-05,
      "loss": 3.8004,
      "step": 1485
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3375896215438843,
      "learning_rate": 8.001421150856434e-05,
      "loss": 3.7365,
      "step": 1490
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.2824058532714844,
      "learning_rate": 7.988070781920197e-05,
      "loss": 3.7214,
      "step": 1495
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.2873374223709106,
      "learning_rate": 7.9746871926477e-05,
      "loss": 3.7474,
      "step": 1500
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.2541723251342773,
      "learning_rate": 7.961270531833052e-05,
      "loss": 3.6825,
      "step": 1505
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3854354619979858,
      "learning_rate": 7.947820948638045e-05,
      "loss": 3.7333,
      "step": 1510
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.0735524892807007,
      "learning_rate": 7.934338592590486e-05,
      "loss": 3.7878,
      "step": 1515
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.1718729734420776,
      "learning_rate": 7.92082361358254e-05,
      "loss": 3.7286,
      "step": 1520
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.987315058708191,
      "learning_rate": 7.907276161869065e-05,
      "loss": 3.7756,
      "step": 1525
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5922112464904785,
      "learning_rate": 7.893696388065936e-05,
      "loss": 3.6469,
      "step": 1530
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.1863898038864136,
      "learning_rate": 7.88008444314838e-05,
      "loss": 3.72,
      "step": 1535
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.0921562910079956,
      "learning_rate": 7.866440478449283e-05,
      "loss": 3.824,
      "step": 1540
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.4743787050247192,
      "learning_rate": 7.852764645657522e-05,
      "loss": 3.71,
      "step": 1545
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.2745715379714966,
      "learning_rate": 7.839057096816271e-05,
      "loss": 3.7467,
      "step": 1550
    },
    {
      "epoch": 0.32,
      "grad_norm": 5.968575954437256,
      "learning_rate": 7.82531798432131e-05,
      "loss": 3.7285,
      "step": 1555
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.130025863647461,
      "learning_rate": 7.811547460919333e-05,
      "loss": 3.6629,
      "step": 1560
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5758984088897705,
      "learning_rate": 7.797745679706254e-05,
      "loss": 3.6272,
      "step": 1565
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.1425355672836304,
      "learning_rate": 7.783912794125496e-05,
      "loss": 3.6044,
      "step": 1570
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5422570705413818,
      "learning_rate": 7.770048957966291e-05,
      "loss": 3.7452,
      "step": 1575
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6101562976837158,
      "learning_rate": 7.756154325361967e-05,
      "loss": 3.7062,
      "step": 1580
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.1914243698120117,
      "learning_rate": 7.74222905078824e-05,
      "loss": 3.6931,
      "step": 1585
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.2087386846542358,
      "learning_rate": 7.728273289061489e-05,
      "loss": 3.7102,
      "step": 1590
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6411712169647217,
      "learning_rate": 7.714287195337044e-05,
      "loss": 3.6645,
      "step": 1595
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.0920937061309814,
      "learning_rate": 7.700270925107448e-05,
      "loss": 3.7041,
      "step": 1600
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4154627323150635,
      "learning_rate": 7.686224634200742e-05,
      "loss": 3.7628,
      "step": 1605
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.9885333180427551,
      "learning_rate": 7.672148478778722e-05,
      "loss": 3.6455,
      "step": 1610
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1912422180175781,
      "learning_rate": 7.658042615335212e-05,
      "loss": 3.6583,
      "step": 1615
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4246773719787598,
      "learning_rate": 7.643907200694318e-05,
      "loss": 3.7332,
      "step": 1620
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1347945928573608,
      "learning_rate": 7.629742392008684e-05,
      "loss": 3.7321,
      "step": 1625
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3160018920898438,
      "learning_rate": 7.615548346757749e-05,
      "loss": 3.6345,
      "step": 1630
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3370493650436401,
      "learning_rate": 7.60132522274599e-05,
      "loss": 3.6661,
      "step": 1635
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0385992527008057,
      "learning_rate": 7.587073178101178e-05,
      "loss": 3.6488,
      "step": 1640
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.0909576416015625,
      "learning_rate": 7.572792371272609e-05,
      "loss": 3.6406,
      "step": 1645
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4987009763717651,
      "learning_rate": 7.55848296102935e-05,
      "loss": 3.6228,
      "step": 1650
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.1608306169509888,
      "learning_rate": 7.544145106458465e-05,
      "loss": 3.7531,
      "step": 1655
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.263434886932373,
      "learning_rate": 7.529778966963259e-05,
      "loss": 3.6643,
      "step": 1660
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.130403995513916,
      "learning_rate": 7.515384702261496e-05,
      "loss": 3.6908,
      "step": 1665
    },
    {
      "epoch": 0.35,
      "grad_norm": 2.0340819358825684,
      "learning_rate": 7.500962472383627e-05,
      "loss": 3.7466,
      "step": 1670
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.113906979560852,
      "learning_rate": 7.486512437671011e-05,
      "loss": 3.6592,
      "step": 1675
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.6292065382003784,
      "learning_rate": 7.472034758774128e-05,
      "loss": 3.5562,
      "step": 1680
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4088952541351318,
      "learning_rate": 7.457529596650797e-05,
      "loss": 3.6639,
      "step": 1685
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.2031278610229492,
      "learning_rate": 7.442997112564392e-05,
      "loss": 3.6144,
      "step": 1690
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3227585554122925,
      "learning_rate": 7.428437468082037e-05,
      "loss": 3.6286,
      "step": 1695
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.252436876296997,
      "learning_rate": 7.413850825072817e-05,
      "loss": 3.646,
      "step": 1700
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.394742727279663,
      "learning_rate": 7.39923734570598e-05,
      "loss": 3.5767,
      "step": 1705
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.2847375869750977,
      "learning_rate": 7.384597192449126e-05,
      "loss": 3.5754,
      "step": 1710
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.2518513202667236,
      "learning_rate": 7.369930528066412e-05,
      "loss": 3.634,
      "step": 1715
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0519161224365234,
      "learning_rate": 7.355237515616732e-05,
      "loss": 3.6772,
      "step": 1720
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.226135015487671,
      "learning_rate": 7.340518318451914e-05,
      "loss": 3.552,
      "step": 1725
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1285977363586426,
      "learning_rate": 7.325773100214893e-05,
      "loss": 3.601,
      "step": 1730
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1608659029006958,
      "learning_rate": 7.311002024837899e-05,
      "loss": 3.5961,
      "step": 1735
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.2595032453536987,
      "learning_rate": 7.296205256540633e-05,
      "loss": 3.6213,
      "step": 1740
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.1732516288757324,
      "learning_rate": 7.281382959828443e-05,
      "loss": 3.6298,
      "step": 1745
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.0036036968231201,
      "learning_rate": 7.26653529949049e-05,
      "loss": 3.6044,
      "step": 1750
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.2911688089370728,
      "learning_rate": 7.25166244059792e-05,
      "loss": 3.6088,
      "step": 1755
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.1788573265075684,
      "learning_rate": 7.236764548502029e-05,
      "loss": 3.6249,
      "step": 1760
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.5241605043411255,
      "learning_rate": 7.221841788832421e-05,
      "loss": 3.5965,
      "step": 1765
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3550124168395996,
      "learning_rate": 7.206894327495173e-05,
      "loss": 3.5928,
      "step": 1770
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.125823736190796,
      "learning_rate": 7.191922330670982e-05,
      "loss": 3.6486,
      "step": 1775
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.9978755116462708,
      "learning_rate": 7.176925964813326e-05,
      "loss": 3.5958,
      "step": 1780
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.6410948038101196,
      "learning_rate": 7.161905396646607e-05,
      "loss": 3.5321,
      "step": 1785
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.1125872135162354,
      "learning_rate": 7.146860793164299e-05,
      "loss": 3.5681,
      "step": 1790
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.2789889574050903,
      "learning_rate": 7.131792321627098e-05,
      "loss": 3.6238,
      "step": 1795
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.1455119848251343,
      "learning_rate": 7.116700149561048e-05,
      "loss": 3.5738,
      "step": 1800
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.9821574091911316,
      "learning_rate": 7.101584444755696e-05,
      "loss": 3.5523,
      "step": 1805
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.1989892721176147,
      "learning_rate": 7.086445375262212e-05,
      "loss": 3.7067,
      "step": 1810
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.1733030080795288,
      "learning_rate": 7.071283109391528e-05,
      "loss": 3.5572,
      "step": 1815
    },
    {
      "epoch": 0.38,
      "grad_norm": 2.4888556003570557,
      "learning_rate": 7.056097815712466e-05,
      "loss": 3.6334,
      "step": 1820
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.2262009382247925,
      "learning_rate": 7.040889663049862e-05,
      "loss": 3.6345,
      "step": 1825
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.1262528896331787,
      "learning_rate": 7.025658820482693e-05,
      "loss": 3.673,
      "step": 1830
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.1414169073104858,
      "learning_rate": 7.010405457342192e-05,
      "loss": 3.5342,
      "step": 1835
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.1732423305511475,
      "learning_rate": 6.995129743209967e-05,
      "loss": 3.6113,
      "step": 1840
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.1435227394104004,
      "learning_rate": 6.97983184791612e-05,
      "loss": 3.5576,
      "step": 1845
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.1150481700897217,
      "learning_rate": 6.964511941537355e-05,
      "loss": 3.659,
      "step": 1850
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.2644362449645996,
      "learning_rate": 6.949170194395083e-05,
      "loss": 3.5798,
      "step": 1855
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.1610716581344604,
      "learning_rate": 6.933806777053536e-05,
      "loss": 3.5271,
      "step": 1860
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3042726516723633,
      "learning_rate": 6.918421860317872e-05,
      "loss": 3.5707,
      "step": 1865
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.9652230143547058,
      "learning_rate": 6.903015615232263e-05,
      "loss": 3.6355,
      "step": 1870
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.2273874282836914,
      "learning_rate": 6.887588213078012e-05,
      "loss": 3.542,
      "step": 1875
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.090998649597168,
      "learning_rate": 6.87213982537163e-05,
      "loss": 3.5175,
      "step": 1880
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.203755259513855,
      "learning_rate": 6.856670623862943e-05,
      "loss": 3.515,
      "step": 1885
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0707592964172363,
      "learning_rate": 6.841180780533179e-05,
      "loss": 3.5394,
      "step": 1890
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.0272190570831299,
      "learning_rate": 6.82567046759305e-05,
      "loss": 3.4902,
      "step": 1895
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.9815902709960938,
      "learning_rate": 6.810139857480844e-05,
      "loss": 3.5563,
      "step": 1900
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.1668179035186768,
      "learning_rate": 6.794589122860509e-05,
      "loss": 3.5696,
      "step": 1905
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.574946641921997,
      "learning_rate": 6.779018436619725e-05,
      "loss": 3.6011,
      "step": 1910
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.2445915937423706,
      "learning_rate": 6.763427971867992e-05,
      "loss": 3.5176,
      "step": 1915
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.286269187927246,
      "learning_rate": 6.747817901934699e-05,
      "loss": 3.5828,
      "step": 1920
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.2477304935455322,
      "learning_rate": 6.732188400367197e-05,
      "loss": 3.6096,
      "step": 1925
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.2229552268981934,
      "learning_rate": 6.716539640928871e-05,
      "loss": 3.6102,
      "step": 1930
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.356566309928894,
      "learning_rate": 6.70087179759721e-05,
      "loss": 3.5343,
      "step": 1935
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.0600124597549438,
      "learning_rate": 6.685185044561874e-05,
      "loss": 3.5066,
      "step": 1940
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4194650650024414,
      "learning_rate": 6.669479556222747e-05,
      "loss": 3.5649,
      "step": 1945
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.2837903499603271,
      "learning_rate": 6.653755507188013e-05,
      "loss": 3.5501,
      "step": 1950
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.2244608402252197,
      "learning_rate": 6.638013072272205e-05,
      "loss": 3.4894,
      "step": 1955
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3000928163528442,
      "learning_rate": 6.622252426494259e-05,
      "loss": 3.5159,
      "step": 1960
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.141238808631897,
      "learning_rate": 6.606473745075581e-05,
      "loss": 3.6376,
      "step": 1965
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.2801133394241333,
      "learning_rate": 6.590677203438084e-05,
      "loss": 3.5289,
      "step": 1970
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.1038589477539062,
      "learning_rate": 6.574862977202252e-05,
      "loss": 3.511,
      "step": 1975
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3234045505523682,
      "learning_rate": 6.559031242185174e-05,
      "loss": 3.5653,
      "step": 1980
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.2162070274353027,
      "learning_rate": 6.543182174398597e-05,
      "loss": 3.5762,
      "step": 1985
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.0354394912719727,
      "learning_rate": 6.52731595004697e-05,
      "loss": 3.606,
      "step": 1990
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.1612788438796997,
      "learning_rate": 6.51143274552548e-05,
      "loss": 3.5454,
      "step": 1995
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.1168513298034668,
      "learning_rate": 6.495532737418098e-05,
      "loss": 3.6233,
      "step": 2000
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.1265037059783936,
      "learning_rate": 6.479616102495605e-05,
      "loss": 3.4706,
      "step": 2005
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.1584540605545044,
      "learning_rate": 6.463683017713638e-05,
      "loss": 3.4945,
      "step": 2010
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9387543201446533,
      "learning_rate": 6.447733660210715e-05,
      "loss": 3.5447,
      "step": 2015
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6024271249771118,
      "learning_rate": 6.431768207306272e-05,
      "loss": 3.519,
      "step": 2020
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9260468482971191,
      "learning_rate": 6.415786836498684e-05,
      "loss": 3.5743,
      "step": 2025
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.0614194869995117,
      "learning_rate": 6.399789725463298e-05,
      "loss": 3.5214,
      "step": 2030
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.9857592582702637,
      "learning_rate": 6.383777052050458e-05,
      "loss": 3.5064,
      "step": 2035
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.1745067834854126,
      "learning_rate": 6.367748994283518e-05,
      "loss": 3.4946,
      "step": 2040
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0638951063156128,
      "learning_rate": 6.351705730356877e-05,
      "loss": 3.5524,
      "step": 2045
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.14460289478302,
      "learning_rate": 6.335647438633987e-05,
      "loss": 3.4517,
      "step": 2050
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.394174575805664,
      "learning_rate": 6.319574297645374e-05,
      "loss": 3.4724,
      "step": 2055
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0267730951309204,
      "learning_rate": 6.303486486086654e-05,
      "loss": 3.515,
      "step": 2060
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.169547438621521,
      "learning_rate": 6.287384182816546e-05,
      "loss": 3.4679,
      "step": 2065
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.9727531671524048,
      "learning_rate": 6.271267566854883e-05,
      "loss": 3.4599,
      "step": 2070
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0682404041290283,
      "learning_rate": 6.255136817380618e-05,
      "loss": 3.4934,
      "step": 2075
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0919233560562134,
      "learning_rate": 6.23899211372984e-05,
      "loss": 3.5548,
      "step": 2080
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0446982383728027,
      "learning_rate": 6.222833635393772e-05,
      "loss": 3.4958,
      "step": 2085
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.0501346588134766,
      "learning_rate": 6.206661562016782e-05,
      "loss": 3.5291,
      "step": 2090
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0181680917739868,
      "learning_rate": 6.190476073394382e-05,
      "loss": 3.435,
      "step": 2095
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.117964506149292,
      "learning_rate": 6.17427734947123e-05,
      "loss": 3.5824,
      "step": 2100
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0905768871307373,
      "learning_rate": 6.158065570339127e-05,
      "loss": 3.4931,
      "step": 2105
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.1306370496749878,
      "learning_rate": 6.141840916235021e-05,
      "loss": 3.503,
      "step": 2110
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5290015935897827,
      "learning_rate": 6.125603567539001e-05,
      "loss": 3.5664,
      "step": 2115
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.1953859329223633,
      "learning_rate": 6.109353704772284e-05,
      "loss": 3.5314,
      "step": 2120
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0872337818145752,
      "learning_rate": 6.0930915085952164e-05,
      "loss": 3.5423,
      "step": 2125
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.055855393409729,
      "learning_rate": 6.076817159805267e-05,
      "loss": 3.5259,
      "step": 2130
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.106011986732483,
      "learning_rate": 6.06053083933501e-05,
      "loss": 3.393,
      "step": 2135
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.0738811492919922,
      "learning_rate": 6.044232728250116e-05,
      "loss": 3.4602,
      "step": 2140
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.1911791563034058,
      "learning_rate": 6.027923007747339e-05,
      "loss": 3.44,
      "step": 2145
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.1135449409484863,
      "learning_rate": 6.011601859152506e-05,
      "loss": 3.4929,
      "step": 2150
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9973747134208679,
      "learning_rate": 5.995269463918495e-05,
      "loss": 3.6224,
      "step": 2155
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9805868268013,
      "learning_rate": 5.97892600362322e-05,
      "loss": 3.4072,
      "step": 2160
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9871159195899963,
      "learning_rate": 5.962571659967614e-05,
      "loss": 3.4855,
      "step": 2165
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9133853316307068,
      "learning_rate": 5.946206614773606e-05,
      "loss": 3.5552,
      "step": 2170
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.06731379032135,
      "learning_rate": 5.929831049982103e-05,
      "loss": 3.4467,
      "step": 2175
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.9730843901634216,
      "learning_rate": 5.9134451476509633e-05,
      "loss": 3.5087,
      "step": 2180
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.1040936708450317,
      "learning_rate": 5.897049089952974e-05,
      "loss": 3.5024,
      "step": 2185
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.1260329484939575,
      "learning_rate": 5.880643059173826e-05,
      "loss": 3.4753,
      "step": 2190
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.0565217733383179,
      "learning_rate": 5.864227237710093e-05,
      "loss": 3.5087,
      "step": 2195
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.034422755241394,
      "learning_rate": 5.847801808067189e-05,
      "loss": 3.4757,
      "step": 2200
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.0072028636932373,
      "learning_rate": 5.831366952857357e-05,
      "loss": 3.431,
      "step": 2205
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.0306864976882935,
      "learning_rate": 5.814922854797622e-05,
      "loss": 3.5302,
      "step": 2210
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.169833779335022,
      "learning_rate": 5.798469696707775e-05,
      "loss": 3.425,
      "step": 2215
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.9371284246444702,
      "learning_rate": 5.782007661508331e-05,
      "loss": 3.4402,
      "step": 2220
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.21083402633667,
      "learning_rate": 5.765536932218495e-05,
      "loss": 3.4139,
      "step": 2225
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.1241546869277954,
      "learning_rate": 5.7490576919541315e-05,
      "loss": 3.4487,
      "step": 2230
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.0151013135910034,
      "learning_rate": 5.732570123925729e-05,
      "loss": 3.5626,
      "step": 2235
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0404661893844604,
      "learning_rate": 5.7160744114363593e-05,
      "loss": 3.357,
      "step": 2240
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0313767194747925,
      "learning_rate": 5.699570737879641e-05,
      "loss": 3.4451,
      "step": 2245
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9385614395141602,
      "learning_rate": 5.683059286737702e-05,
      "loss": 3.4489,
      "step": 2250
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.2350726127624512,
      "learning_rate": 5.666540241579139e-05,
      "loss": 3.5222,
      "step": 2255
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.2488938570022583,
      "learning_rate": 5.6500137860569766e-05,
      "loss": 3.474,
      "step": 2260
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.9713031649589539,
      "learning_rate": 5.633480103906624e-05,
      "loss": 3.4048,
      "step": 2265
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0884007215499878,
      "learning_rate": 5.616939378943834e-05,
      "loss": 3.4551,
      "step": 2270
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0945968627929688,
      "learning_rate": 5.6003917950626595e-05,
      "loss": 3.4321,
      "step": 2275
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.3760701417922974,
      "learning_rate": 5.583837536233407e-05,
      "loss": 3.4909,
      "step": 2280
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.0719841718673706,
      "learning_rate": 5.567276786500596e-05,
      "loss": 3.4773,
      "step": 2285
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0727452039718628,
      "learning_rate": 5.5507097299809054e-05,
      "loss": 3.4759,
      "step": 2290
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.101324439048767,
      "learning_rate": 5.534136550861133e-05,
      "loss": 3.378,
      "step": 2295
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.2684340476989746,
      "learning_rate": 5.5175574333961465e-05,
      "loss": 3.3992,
      "step": 2300
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.119226098060608,
      "learning_rate": 5.500972561906832e-05,
      "loss": 3.5095,
      "step": 2305
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.1630808115005493,
      "learning_rate": 5.484382120778048e-05,
      "loss": 3.4643,
      "step": 2310
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0694406032562256,
      "learning_rate": 5.467786294456575e-05,
      "loss": 3.4609,
      "step": 2315
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.0865049362182617,
      "learning_rate": 5.451185267449061e-05,
      "loss": 3.462,
      "step": 2320
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.9963206052780151,
      "learning_rate": 5.43457922431998e-05,
      "loss": 3.4361,
      "step": 2325
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.2405842542648315,
      "learning_rate": 5.417968349689566e-05,
      "loss": 3.3828,
      "step": 2330
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.3211381435394287,
      "learning_rate": 5.401352828231772e-05,
      "loss": 3.4652,
      "step": 2335
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.0762677192687988,
      "learning_rate": 5.384732844672211e-05,
      "loss": 3.4598,
      "step": 2340
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.0378797054290771,
      "learning_rate": 5.368108583786107e-05,
      "loss": 3.4911,
      "step": 2345
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.2162408828735352,
      "learning_rate": 5.3514802303962344e-05,
      "loss": 3.457,
      "step": 2350
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.988157331943512,
      "learning_rate": 5.334847969370868e-05,
      "loss": 3.4645,
      "step": 2355
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.0698318481445312,
      "learning_rate": 5.3182119856217284e-05,
      "loss": 3.4249,
      "step": 2360
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9144718050956726,
      "learning_rate": 5.3015724641019214e-05,
      "loss": 3.4205,
      "step": 2365
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.0714385509490967,
      "learning_rate": 5.284929589803884e-05,
      "loss": 3.3969,
      "step": 2370
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9795346856117249,
      "learning_rate": 5.2682835477573336e-05,
      "loss": 3.3832,
      "step": 2375
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.9754644632339478,
      "learning_rate": 5.2516345230271965e-05,
      "loss": 3.3691,
      "step": 2380
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.000421404838562,
      "learning_rate": 5.234982700711569e-05,
      "loss": 3.4015,
      "step": 2385
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9788889288902283,
      "learning_rate": 5.218328265939643e-05,
      "loss": 3.4135,
      "step": 2390
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.037010669708252,
      "learning_rate": 5.201671403869657e-05,
      "loss": 3.4492,
      "step": 2395
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.8524897694587708,
      "learning_rate": 5.1850122996868366e-05,
      "loss": 3.5092,
      "step": 2400
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.112398624420166,
      "learning_rate": 5.168351138601334e-05,
      "loss": 3.3871,
      "step": 2405
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.1374262571334839,
      "learning_rate": 5.1516881058461675e-05,
      "loss": 3.5498,
      "step": 2410
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.0454888343811035,
      "learning_rate": 5.135023386675166e-05,
      "loss": 3.3942,
      "step": 2415
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.1183348894119263,
      "learning_rate": 5.118357166360906e-05,
      "loss": 3.3447,
      "step": 2420
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.9044198989868164,
      "learning_rate": 5.101689630192655e-05,
      "loss": 3.4698,
      "step": 2425
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.0566176176071167,
      "learning_rate": 5.085020963474307e-05,
      "loss": 3.397,
      "step": 2430
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.0596601963043213,
      "learning_rate": 5.068351351522329e-05,
      "loss": 3.4018,
      "step": 2435
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.033179521560669,
      "learning_rate": 5.0516809796636935e-05,
      "loss": 3.3199,
      "step": 2440
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9158328771591187,
      "learning_rate": 5.035010033233821e-05,
      "loss": 3.3591,
      "step": 2445
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9977751970291138,
      "learning_rate": 5.018338697574523e-05,
      "loss": 3.4262,
      "step": 2450
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.0123366117477417,
      "learning_rate": 5.0016671580319354e-05,
      "loss": 3.481,
      "step": 2455
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.0733108520507812,
      "learning_rate": 4.984995599954461e-05,
      "loss": 3.4421,
      "step": 2460
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9767444133758545,
      "learning_rate": 4.968324208690712e-05,
      "loss": 3.3568,
      "step": 2465
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.9529309272766113,
      "learning_rate": 4.951653169587441e-05,
      "loss": 3.3881,
      "step": 2470
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.082856297492981,
      "learning_rate": 4.93498266798749e-05,
      "loss": 3.4472,
      "step": 2475
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.0122766494750977,
      "learning_rate": 4.918312889227722e-05,
      "loss": 3.3907,
      "step": 2480
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.982729971408844,
      "learning_rate": 4.901644018636966e-05,
      "loss": 3.3404,
      "step": 2485
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9235652089118958,
      "learning_rate": 4.8849762415339526e-05,
      "loss": 3.4075,
      "step": 2490
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9878057241439819,
      "learning_rate": 4.868309743225256e-05,
      "loss": 3.3355,
      "step": 2495
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.1668199300765991,
      "learning_rate": 4.851644709003233e-05,
      "loss": 3.4238,
      "step": 2500
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.0045877695083618,
      "learning_rate": 4.834981324143964e-05,
      "loss": 3.352,
      "step": 2505
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9545599818229675,
      "learning_rate": 4.818319773905191e-05,
      "loss": 3.5209,
      "step": 2510
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8998648524284363,
      "learning_rate": 4.801660243524261e-05,
      "loss": 3.4146,
      "step": 2515
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.0703359842300415,
      "learning_rate": 4.7850029182160626e-05,
      "loss": 3.4293,
      "step": 2520
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.9540716409683228,
      "learning_rate": 4.768347983170973e-05,
      "loss": 3.4172,
      "step": 2525
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.002467393875122,
      "learning_rate": 4.7516956235527884e-05,
      "loss": 3.3264,
      "step": 2530
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9552346467971802,
      "learning_rate": 4.735046024496682e-05,
      "loss": 3.3276,
      "step": 2535
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9977598786354065,
      "learning_rate": 4.7183993711071286e-05,
      "loss": 3.3882,
      "step": 2540
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.0754117965698242,
      "learning_rate": 4.7017558484558554e-05,
      "loss": 3.3855,
      "step": 2545
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.1346033811569214,
      "learning_rate": 4.6851156415797844e-05,
      "loss": 3.336,
      "step": 2550
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9791164994239807,
      "learning_rate": 4.6684789354789746e-05,
      "loss": 3.423,
      "step": 2555
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.0167591571807861,
      "learning_rate": 4.651845915114563e-05,
      "loss": 3.3428,
      "step": 2560
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.0157413482666016,
      "learning_rate": 4.6352167654067095e-05,
      "loss": 3.4264,
      "step": 2565
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.9716314077377319,
      "learning_rate": 4.618591671232544e-05,
      "loss": 3.3867,
      "step": 2570
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.0501482486724854,
      "learning_rate": 4.601970817424106e-05,
      "loss": 3.3322,
      "step": 2575
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.1115131378173828,
      "learning_rate": 4.585354388766292e-05,
      "loss": 3.376,
      "step": 2580
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9755733609199524,
      "learning_rate": 4.568742569994802e-05,
      "loss": 3.3261,
      "step": 2585
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.0900994539260864,
      "learning_rate": 4.552135545794086e-05,
      "loss": 3.4289,
      "step": 2590
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.934795618057251,
      "learning_rate": 4.535533500795288e-05,
      "loss": 3.3068,
      "step": 2595
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4602357149124146,
      "learning_rate": 4.5189366195741953e-05,
      "loss": 3.2843,
      "step": 2600
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.00444495677948,
      "learning_rate": 4.502345086649186e-05,
      "loss": 3.4315,
      "step": 2605
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.0988986492156982,
      "learning_rate": 4.485759086479179e-05,
      "loss": 3.3491,
      "step": 2610
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.0917633771896362,
      "learning_rate": 4.469178803461579e-05,
      "loss": 3.414,
      "step": 2615
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.9472582340240479,
      "learning_rate": 4.4526044219302326e-05,
      "loss": 3.3974,
      "step": 2620
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.026991367340088,
      "learning_rate": 4.4360361261533745e-05,
      "loss": 3.4343,
      "step": 2625
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9805732369422913,
      "learning_rate": 4.419474100331579e-05,
      "loss": 3.3849,
      "step": 2630
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.0283095836639404,
      "learning_rate": 4.402918528595715e-05,
      "loss": 3.3216,
      "step": 2635
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.1360772848129272,
      "learning_rate": 4.386369595004896e-05,
      "loss": 3.3478,
      "step": 2640
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9674432873725891,
      "learning_rate": 4.3698274835444354e-05,
      "loss": 3.3886,
      "step": 2645
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9571623802185059,
      "learning_rate": 4.3532923781238e-05,
      "loss": 3.3687,
      "step": 2650
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9667463302612305,
      "learning_rate": 4.336764462574566e-05,
      "loss": 3.4309,
      "step": 2655
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.9910677075386047,
      "learning_rate": 4.320243920648376e-05,
      "loss": 3.4069,
      "step": 2660
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3989684581756592,
      "learning_rate": 4.303730936014894e-05,
      "loss": 3.3443,
      "step": 2665
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.076253056526184,
      "learning_rate": 4.287225692259765e-05,
      "loss": 3.2979,
      "step": 2670
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.2015970945358276,
      "learning_rate": 4.270728372882575e-05,
      "loss": 3.4052,
      "step": 2675
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.1010020971298218,
      "learning_rate": 4.254239161294804e-05,
      "loss": 3.4477,
      "step": 2680
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9749462604522705,
      "learning_rate": 4.237758240817802e-05,
      "loss": 3.3347,
      "step": 2685
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9020056128501892,
      "learning_rate": 4.2212857946807336e-05,
      "loss": 3.4074,
      "step": 2690
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.8927140831947327,
      "learning_rate": 4.2048220060185516e-05,
      "loss": 3.4199,
      "step": 2695
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9286185503005981,
      "learning_rate": 4.188367057869957e-05,
      "loss": 3.4153,
      "step": 2700
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9509190320968628,
      "learning_rate": 4.171921133175365e-05,
      "loss": 3.2441,
      "step": 2705
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9190889000892639,
      "learning_rate": 4.155484414774872e-05,
      "loss": 3.3528,
      "step": 2710
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.9416577816009521,
      "learning_rate": 4.139057085406221e-05,
      "loss": 3.3397,
      "step": 2715
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9352203607559204,
      "learning_rate": 4.1226393277027726e-05,
      "loss": 3.3165,
      "step": 2720
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.0076632499694824,
      "learning_rate": 4.106231324191471e-05,
      "loss": 3.3903,
      "step": 2725
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.970344603061676,
      "learning_rate": 4.089833257290817e-05,
      "loss": 3.3232,
      "step": 2730
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9177244305610657,
      "learning_rate": 4.073445309308842e-05,
      "loss": 3.3932,
      "step": 2735
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.0965250730514526,
      "learning_rate": 4.0570676624410756e-05,
      "loss": 3.4276,
      "step": 2740
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.0091582536697388,
      "learning_rate": 4.040700498768525e-05,
      "loss": 3.3576,
      "step": 2745
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.025673508644104,
      "learning_rate": 4.024344000255648e-05,
      "loss": 3.4126,
      "step": 2750
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.998465359210968,
      "learning_rate": 4.0079983487483313e-05,
      "loss": 3.3526,
      "step": 2755
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.9938681125640869,
      "learning_rate": 3.9916637259718683e-05,
      "loss": 3.3711,
      "step": 2760
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.008997917175293,
      "learning_rate": 3.9753403135289396e-05,
      "loss": 3.3428,
      "step": 2765
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9713385701179504,
      "learning_rate": 3.9590282928975914e-05,
      "loss": 3.3509,
      "step": 2770
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9508611559867859,
      "learning_rate": 3.942727845429221e-05,
      "loss": 3.4893,
      "step": 2775
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9679702520370483,
      "learning_rate": 3.926439152346558e-05,
      "loss": 3.324,
      "step": 2780
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9336875081062317,
      "learning_rate": 3.910162394741653e-05,
      "loss": 3.3954,
      "step": 2785
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9901192784309387,
      "learning_rate": 3.893897753573861e-05,
      "loss": 3.3056,
      "step": 2790
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.20081627368927,
      "learning_rate": 3.877645409667829e-05,
      "loss": 3.3136,
      "step": 2795
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9638660550117493,
      "learning_rate": 3.861405543711491e-05,
      "loss": 3.4024,
      "step": 2800
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.9280533194541931,
      "learning_rate": 3.8451783362540507e-05,
      "loss": 3.3105,
      "step": 2805
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.077518105506897,
      "learning_rate": 3.828963967703983e-05,
      "loss": 3.4049,
      "step": 2810
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8907383680343628,
      "learning_rate": 3.8127626183270223e-05,
      "loss": 3.4606,
      "step": 2815
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9571870565414429,
      "learning_rate": 3.796574468244161e-05,
      "loss": 3.3415,
      "step": 2820
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8469297885894775,
      "learning_rate": 3.7803996974296444e-05,
      "loss": 3.3511,
      "step": 2825
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.0232789516448975,
      "learning_rate": 3.7642384857089776e-05,
      "loss": 3.3848,
      "step": 2830
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8824405670166016,
      "learning_rate": 3.748091012756915e-05,
      "loss": 3.3029,
      "step": 2835
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9363697171211243,
      "learning_rate": 3.731957458095467e-05,
      "loss": 3.4415,
      "step": 2840
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.9917386174201965,
      "learning_rate": 3.71583800109191e-05,
      "loss": 3.3475,
      "step": 2845
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.919383704662323,
      "learning_rate": 3.699732820956784e-05,
      "loss": 3.3903,
      "step": 2850
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8919963240623474,
      "learning_rate": 3.6836420967419057e-05,
      "loss": 3.4202,
      "step": 2855
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.8788687586784363,
      "learning_rate": 3.6675660073383745e-05,
      "loss": 3.3453,
      "step": 2860
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.1343214511871338,
      "learning_rate": 3.6515047314745856e-05,
      "loss": 3.4452,
      "step": 2865
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.8609405755996704,
      "learning_rate": 3.6354584477142437e-05,
      "loss": 3.3719,
      "step": 2870
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.921087384223938,
      "learning_rate": 3.6194273344543736e-05,
      "loss": 3.3104,
      "step": 2875
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9409319162368774,
      "learning_rate": 3.6034115699233425e-05,
      "loss": 3.2578,
      "step": 2880
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.0189098119735718,
      "learning_rate": 3.5874113321788736e-05,
      "loss": 3.3444,
      "step": 2885
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9592056274414062,
      "learning_rate": 3.571426799106071e-05,
      "loss": 3.3167,
      "step": 2890
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.0163627862930298,
      "learning_rate": 3.555458148415437e-05,
      "loss": 3.3269,
      "step": 2895
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9534558057785034,
      "learning_rate": 3.539505557640901e-05,
      "loss": 3.3309,
      "step": 2900
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.9127333164215088,
      "learning_rate": 3.523569204137843e-05,
      "loss": 3.3732,
      "step": 2905
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.0162489414215088,
      "learning_rate": 3.5076492650811246e-05,
      "loss": 3.2918,
      "step": 2910
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9052246809005737,
      "learning_rate": 3.491745917463113e-05,
      "loss": 3.3155,
      "step": 2915
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.2741327285766602,
      "learning_rate": 3.475859338091721e-05,
      "loss": 3.3678,
      "step": 2920
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9226068258285522,
      "learning_rate": 3.4599897035884374e-05,
      "loss": 3.3418,
      "step": 2925
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9619899392127991,
      "learning_rate": 3.444137190386363e-05,
      "loss": 3.3496,
      "step": 2930
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.0131815671920776,
      "learning_rate": 3.4283019747282514e-05,
      "loss": 3.324,
      "step": 2935
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9577188491821289,
      "learning_rate": 3.412484232664545e-05,
      "loss": 3.2871,
      "step": 2940
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.0029125213623047,
      "learning_rate": 3.396684140051424e-05,
      "loss": 3.356,
      "step": 2945
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.9922611713409424,
      "learning_rate": 3.3809018725488466e-05,
      "loss": 3.2954,
      "step": 2950
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.943458080291748,
      "learning_rate": 3.365137605618598e-05,
      "loss": 3.3505,
      "step": 2955
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9286159873008728,
      "learning_rate": 3.3493915145223395e-05,
      "loss": 3.3266,
      "step": 2960
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.8530454039573669,
      "learning_rate": 3.3336637743196584e-05,
      "loss": 3.2675,
      "step": 2965
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9328925609588623,
      "learning_rate": 3.317954559866126e-05,
      "loss": 3.2409,
      "step": 2970
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.931462287902832,
      "learning_rate": 3.302264045811344e-05,
      "loss": 3.3505,
      "step": 2975
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0057929754257202,
      "learning_rate": 3.286592406597021e-05,
      "loss": 3.3282,
      "step": 2980
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9094107747077942,
      "learning_rate": 3.270939816455012e-05,
      "loss": 3.29,
      "step": 2985
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9364352822303772,
      "learning_rate": 3.255306449405395e-05,
      "loss": 3.3582,
      "step": 2990
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9047064185142517,
      "learning_rate": 3.2396924792545304e-05,
      "loss": 3.4079,
      "step": 2995
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.9266229271888733,
      "learning_rate": 3.224098079593132e-05,
      "loss": 3.2917,
      "step": 3000
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.0187420845031738,
      "learning_rate": 3.2085234237943354e-05,
      "loss": 3.3806,
      "step": 3005
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8054633140563965,
      "learning_rate": 3.19296868501177e-05,
      "loss": 3.2928,
      "step": 3010
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8668712377548218,
      "learning_rate": 3.177434036177636e-05,
      "loss": 3.2372,
      "step": 3015
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8695892095565796,
      "learning_rate": 3.1619196500007804e-05,
      "loss": 3.2712,
      "step": 3020
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.868233323097229,
      "learning_rate": 3.146425698964776e-05,
      "loss": 3.3161,
      "step": 3025
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9072175025939941,
      "learning_rate": 3.1309523553260046e-05,
      "loss": 3.3399,
      "step": 3030
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.8848998546600342,
      "learning_rate": 3.115499791111743e-05,
      "loss": 3.4425,
      "step": 3035
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9443853497505188,
      "learning_rate": 3.10006817811825e-05,
      "loss": 3.3577,
      "step": 3040
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.031072974205017,
      "learning_rate": 3.084657687908855e-05,
      "loss": 3.3258,
      "step": 3045
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9901952147483826,
      "learning_rate": 3.069268491812052e-05,
      "loss": 3.2956,
      "step": 3050
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.9264650940895081,
      "learning_rate": 3.0539007609195934e-05,
      "loss": 3.3224,
      "step": 3055
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9035649299621582,
      "learning_rate": 3.0385546660845908e-05,
      "loss": 3.2863,
      "step": 3060
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9281037449836731,
      "learning_rate": 3.0232303779196132e-05,
      "loss": 3.3212,
      "step": 3065
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.096510648727417,
      "learning_rate": 3.0079280667947885e-05,
      "loss": 3.2815,
      "step": 3070
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.8994583487510681,
      "learning_rate": 2.9926479028359132e-05,
      "loss": 3.408,
      "step": 3075
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9495896100997925,
      "learning_rate": 2.97739005592256e-05,
      "loss": 3.2415,
      "step": 3080
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.040583848953247,
      "learning_rate": 2.962154695686187e-05,
      "loss": 3.2955,
      "step": 3085
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9189764261245728,
      "learning_rate": 2.9469419915082536e-05,
      "loss": 3.2703,
      "step": 3090
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9766911864280701,
      "learning_rate": 2.9317521125183368e-05,
      "loss": 3.3358,
      "step": 3095
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.9531775712966919,
      "learning_rate": 2.9165852275922524e-05,
      "loss": 3.3381,
      "step": 3100
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.0340322256088257,
      "learning_rate": 2.901441505350174e-05,
      "loss": 3.2312,
      "step": 3105
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9244762063026428,
      "learning_rate": 2.886321114154762e-05,
      "loss": 3.3023,
      "step": 3110
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.0161535739898682,
      "learning_rate": 2.87122422210929e-05,
      "loss": 3.2473,
      "step": 3115
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9165362119674683,
      "learning_rate": 2.8561509970557736e-05,
      "loss": 3.3181,
      "step": 3120
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9139227867126465,
      "learning_rate": 2.8411016065731146e-05,
      "loss": 3.2694,
      "step": 3125
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9805368185043335,
      "learning_rate": 2.826076217975222e-05,
      "loss": 3.2982,
      "step": 3130
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9517117142677307,
      "learning_rate": 2.8110749983091632e-05,
      "loss": 3.3033,
      "step": 3135
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.053392767906189,
      "learning_rate": 2.7960981143533053e-05,
      "loss": 3.2751,
      "step": 3140
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9289665222167969,
      "learning_rate": 2.781145732615457e-05,
      "loss": 3.3134,
      "step": 3145
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9292173981666565,
      "learning_rate": 2.7662180193310218e-05,
      "loss": 3.3168,
      "step": 3150
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.1169716119766235,
      "learning_rate": 2.751315140461145e-05,
      "loss": 3.2929,
      "step": 3155
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.919864296913147,
      "learning_rate": 2.7364372616908744e-05,
      "loss": 3.3529,
      "step": 3160
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9498540163040161,
      "learning_rate": 2.7215845484273152e-05,
      "loss": 3.3366,
      "step": 3165
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.0325037240982056,
      "learning_rate": 2.7067571657977893e-05,
      "loss": 3.2316,
      "step": 3170
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8834294080734253,
      "learning_rate": 2.691955278648003e-05,
      "loss": 3.278,
      "step": 3175
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9540823698043823,
      "learning_rate": 2.6771790515402112e-05,
      "loss": 3.2605,
      "step": 3180
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8876092433929443,
      "learning_rate": 2.6624286487513916e-05,
      "loss": 3.2087,
      "step": 3185
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.9012912511825562,
      "learning_rate": 2.6477042342714137e-05,
      "loss": 3.2982,
      "step": 3190
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.903864324092865,
      "learning_rate": 2.633005971801219e-05,
      "loss": 3.3177,
      "step": 3195
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9462845921516418,
      "learning_rate": 2.6183340247510013e-05,
      "loss": 3.255,
      "step": 3200
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.859810471534729,
      "learning_rate": 2.6036885562383856e-05,
      "loss": 3.24,
      "step": 3205
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.921997606754303,
      "learning_rate": 2.5890697290866206e-05,
      "loss": 3.3059,
      "step": 3210
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.8770375847816467,
      "learning_rate": 2.5744777058227642e-05,
      "loss": 3.3065,
      "step": 3215
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9048674702644348,
      "learning_rate": 2.5599126486758777e-05,
      "loss": 3.2547,
      "step": 3220
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.899581253528595,
      "learning_rate": 2.5453747195752243e-05,
      "loss": 3.2962,
      "step": 3225
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9766543507575989,
      "learning_rate": 2.530864080148464e-05,
      "loss": 3.3386,
      "step": 3230
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.956592857837677,
      "learning_rate": 2.5163808917198615e-05,
      "loss": 3.3397,
      "step": 3235
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9843078255653381,
      "learning_rate": 2.501925315308492e-05,
      "loss": 3.2581,
      "step": 3240
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.9282741546630859,
      "learning_rate": 2.4874975116264477e-05,
      "loss": 3.2057,
      "step": 3245
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8186929821968079,
      "learning_rate": 2.4730976410770534e-05,
      "loss": 3.2798,
      "step": 3250
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9256529808044434,
      "learning_rate": 2.458725863753084e-05,
      "loss": 3.2866,
      "step": 3255
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.030348300933838,
      "learning_rate": 2.4443823394349834e-05,
      "loss": 3.3567,
      "step": 3260
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8589518666267395,
      "learning_rate": 2.430067227589088e-05,
      "loss": 3.2716,
      "step": 3265
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9088854789733887,
      "learning_rate": 2.4157806873658517e-05,
      "loss": 3.3572,
      "step": 3270
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8666104078292847,
      "learning_rate": 2.401522877598087e-05,
      "loss": 3.2997,
      "step": 3275
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.873067319393158,
      "learning_rate": 2.3872939567991827e-05,
      "loss": 3.3017,
      "step": 3280
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8777161240577698,
      "learning_rate": 2.373094083161353e-05,
      "loss": 3.2611,
      "step": 3285
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.9435715079307556,
      "learning_rate": 2.358923414553877e-05,
      "loss": 3.2632,
      "step": 3290
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.8779476284980774,
      "learning_rate": 2.3447821085213405e-05,
      "loss": 3.3449,
      "step": 3295
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9076412916183472,
      "learning_rate": 2.3306703222818878e-05,
      "loss": 3.3196,
      "step": 3300
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.2708210945129395,
      "learning_rate": 2.3165882127254705e-05,
      "loss": 3.2802,
      "step": 3305
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8561325073242188,
      "learning_rate": 2.302535936412108e-05,
      "loss": 3.2155,
      "step": 3310
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9210352301597595,
      "learning_rate": 2.2885136495701415e-05,
      "loss": 3.2985,
      "step": 3315
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9178744554519653,
      "learning_rate": 2.274521508094501e-05,
      "loss": 3.3691,
      "step": 3320
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.894907534122467,
      "learning_rate": 2.2605596675449698e-05,
      "loss": 3.2381,
      "step": 3325
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9445805549621582,
      "learning_rate": 2.246628283144457e-05,
      "loss": 3.3258,
      "step": 3330
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.9801211953163147,
      "learning_rate": 2.232727509777269e-05,
      "loss": 3.2626,
      "step": 3335
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.8563469052314758,
      "learning_rate": 2.2188575019873932e-05,
      "loss": 3.2606,
      "step": 3340
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9590172171592712,
      "learning_rate": 2.2050184139767704e-05,
      "loss": 3.259,
      "step": 3345
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8452922701835632,
      "learning_rate": 2.191210399603591e-05,
      "loss": 3.3287,
      "step": 3350
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8957239985466003,
      "learning_rate": 2.1774336123805772e-05,
      "loss": 3.204,
      "step": 3355
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.0064430236816406,
      "learning_rate": 2.1636882054732776e-05,
      "loss": 3.2428,
      "step": 3360
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8556531071662903,
      "learning_rate": 2.1499743316983684e-05,
      "loss": 3.2985,
      "step": 3365
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8558668494224548,
      "learning_rate": 2.1362921435219473e-05,
      "loss": 3.2492,
      "step": 3370
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9391397833824158,
      "learning_rate": 2.1226417930578464e-05,
      "loss": 3.3506,
      "step": 3375
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.999201238155365,
      "learning_rate": 2.109023432065935e-05,
      "loss": 3.312,
      "step": 3380
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9704054594039917,
      "learning_rate": 2.095437211950434e-05,
      "loss": 3.2287,
      "step": 3385
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.9226841330528259,
      "learning_rate": 2.0818832837582352e-05,
      "loss": 3.2592,
      "step": 3390
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8675829172134399,
      "learning_rate": 2.068361798177218e-05,
      "loss": 3.2965,
      "step": 3395
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8511921763420105,
      "learning_rate": 2.0548729055345778e-05,
      "loss": 3.2273,
      "step": 3400
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8496411442756653,
      "learning_rate": 2.0414167557951514e-05,
      "loss": 3.3067,
      "step": 3405
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8244772553443909,
      "learning_rate": 2.0279934985597527e-05,
      "loss": 3.416,
      "step": 3410
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9402651190757751,
      "learning_rate": 2.0146032830635054e-05,
      "loss": 3.2653,
      "step": 3415
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9962188005447388,
      "learning_rate": 2.001246258174192e-05,
      "loss": 3.2729,
      "step": 3420
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8802465796470642,
      "learning_rate": 1.9879225723905886e-05,
      "loss": 3.2681,
      "step": 3425
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.8524779677391052,
      "learning_rate": 1.9746323738408203e-05,
      "loss": 3.2497,
      "step": 3430
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.9403968453407288,
      "learning_rate": 1.9613758102807117e-05,
      "loss": 3.2378,
      "step": 3435
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.0809732675552368,
      "learning_rate": 1.9481530290921474e-05,
      "loss": 3.2091,
      "step": 3440
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8482281565666199,
      "learning_rate": 1.934964177281428e-05,
      "loss": 3.2957,
      "step": 3445
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8832562565803528,
      "learning_rate": 1.9218094014776434e-05,
      "loss": 3.3247,
      "step": 3450
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8952575922012329,
      "learning_rate": 1.9086888479310333e-05,
      "loss": 3.2158,
      "step": 3455
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6820100545883179,
      "learning_rate": 1.895602662511371e-05,
      "loss": 3.2396,
      "step": 3460
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.9064893126487732,
      "learning_rate": 1.8825509907063327e-05,
      "loss": 3.3192,
      "step": 3465
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8885228633880615,
      "learning_rate": 1.8695339776198872e-05,
      "loss": 3.2687,
      "step": 3470
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.8821113705635071,
      "learning_rate": 1.8565517679706783e-05,
      "loss": 3.2581,
      "step": 3475
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.907341480255127,
      "learning_rate": 1.8436045060904174e-05,
      "loss": 3.2212,
      "step": 3480
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.1254795789718628,
      "learning_rate": 1.830692335922279e-05,
      "loss": 3.2318,
      "step": 3485
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8602316379547119,
      "learning_rate": 1.8178154010192994e-05,
      "loss": 3.3136,
      "step": 3490
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8210873603820801,
      "learning_rate": 1.8049738445427822e-05,
      "loss": 3.369,
      "step": 3495
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8428252935409546,
      "learning_rate": 1.7921678092607052e-05,
      "loss": 3.2728,
      "step": 3500
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8923355340957642,
      "learning_rate": 1.7793974375461352e-05,
      "loss": 3.336,
      "step": 3505
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8625919818878174,
      "learning_rate": 1.7666628713756417e-05,
      "loss": 3.3083,
      "step": 3510
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8860301375389099,
      "learning_rate": 1.7539642523277228e-05,
      "loss": 3.2761,
      "step": 3515
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9147588014602661,
      "learning_rate": 1.7413017215812273e-05,
      "loss": 3.288,
      "step": 3520
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9153177738189697,
      "learning_rate": 1.728675419913788e-05,
      "loss": 3.3425,
      "step": 3525
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.8802617192268372,
      "learning_rate": 1.716085487700253e-05,
      "loss": 3.3332,
      "step": 3530
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.9598279595375061,
      "learning_rate": 1.703532064911131e-05,
      "loss": 3.2123,
      "step": 3535
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9063637852668762,
      "learning_rate": 1.6910152911110283e-05,
      "loss": 3.1825,
      "step": 3540
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9381482005119324,
      "learning_rate": 1.6785353054571024e-05,
      "loss": 3.2832,
      "step": 3545
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8743510842323303,
      "learning_rate": 1.666092246697512e-05,
      "loss": 3.34,
      "step": 3550
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8579509854316711,
      "learning_rate": 1.6536862531698766e-05,
      "loss": 3.2737,
      "step": 3555
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9608656167984009,
      "learning_rate": 1.6413174627997328e-05,
      "loss": 3.2559,
      "step": 3560
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.941361665725708,
      "learning_rate": 1.6289860130990147e-05,
      "loss": 3.2698,
      "step": 3565
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9168158173561096,
      "learning_rate": 1.6166920411645064e-05,
      "loss": 3.3504,
      "step": 3570
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.9752878546714783,
      "learning_rate": 1.6044356836763315e-05,
      "loss": 3.2641,
      "step": 3575
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.8661676645278931,
      "learning_rate": 1.5922170768964285e-05,
      "loss": 3.287,
      "step": 3580
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8774048089981079,
      "learning_rate": 1.5800363566670362e-05,
      "loss": 3.2631,
      "step": 3585
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.819927990436554,
      "learning_rate": 1.5678936584091852e-05,
      "loss": 3.3321,
      "step": 3590
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.835914671421051,
      "learning_rate": 1.5557891171211892e-05,
      "loss": 3.3336,
      "step": 3595
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8567005395889282,
      "learning_rate": 1.5437228673771465e-05,
      "loss": 3.2515,
      "step": 3600
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8540778160095215,
      "learning_rate": 1.5316950433254445e-05,
      "loss": 3.1985,
      "step": 3605
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8710616827011108,
      "learning_rate": 1.5197057786872649e-05,
      "loss": 3.3092,
      "step": 3610
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.9007919430732727,
      "learning_rate": 1.5077552067551015e-05,
      "loss": 3.2059,
      "step": 3615
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8696431517601013,
      "learning_rate": 1.4958434603912747e-05,
      "loss": 3.3528,
      "step": 3620
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8940817713737488,
      "learning_rate": 1.4839706720264546e-05,
      "loss": 3.2701,
      "step": 3625
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.8489177227020264,
      "learning_rate": 1.4721369736581924e-05,
      "loss": 3.2409,
      "step": 3630
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9342685341835022,
      "learning_rate": 1.4603424968494484e-05,
      "loss": 3.2308,
      "step": 3635
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.9768315553665161,
      "learning_rate": 1.448587372727132e-05,
      "loss": 3.2947,
      "step": 3640
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8428696393966675,
      "learning_rate": 1.4368717319806419e-05,
      "loss": 3.2968,
      "step": 3645
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8810005784034729,
      "learning_rate": 1.4251957048604152e-05,
      "loss": 3.2555,
      "step": 3650
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8546096682548523,
      "learning_rate": 1.413559421176479e-05,
      "loss": 3.2297,
      "step": 3655
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8835750222206116,
      "learning_rate": 1.4019630102970056e-05,
      "loss": 3.2423,
      "step": 3660
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8346020579338074,
      "learning_rate": 1.3904066011468753e-05,
      "loss": 3.2519,
      "step": 3665
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8157615661621094,
      "learning_rate": 1.3788903222062433e-05,
      "loss": 3.2553,
      "step": 3670
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8050662279129028,
      "learning_rate": 1.3674143015091118e-05,
      "loss": 3.2356,
      "step": 3675
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.8705363273620605,
      "learning_rate": 1.355978666641905e-05,
      "loss": 3.2543,
      "step": 3680
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8249686360359192,
      "learning_rate": 1.3445835447420507e-05,
      "loss": 3.2328,
      "step": 3685
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8292253017425537,
      "learning_rate": 1.3332290624965688e-05,
      "loss": 3.2763,
      "step": 3690
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8558942079544067,
      "learning_rate": 1.3219153461406609e-05,
      "loss": 3.2703,
      "step": 3695
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8376969695091248,
      "learning_rate": 1.3106425214563078e-05,
      "loss": 3.1923,
      "step": 3700
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8428564071655273,
      "learning_rate": 1.2994107137708716e-05,
      "loss": 3.2835,
      "step": 3705
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8432294726371765,
      "learning_rate": 1.2882200479556988e-05,
      "loss": 3.252,
      "step": 3710
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8771489858627319,
      "learning_rate": 1.2770706484247397e-05,
      "loss": 3.2667,
      "step": 3715
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.9042550921440125,
      "learning_rate": 1.2659626391331564e-05,
      "loss": 3.213,
      "step": 3720
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.8231926560401917,
      "learning_rate": 1.2548961435759493e-05,
      "loss": 3.2232,
      "step": 3725
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.833065927028656,
      "learning_rate": 1.2438712847865846e-05,
      "loss": 3.1761,
      "step": 3730
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9881918430328369,
      "learning_rate": 1.2328881853356244e-05,
      "loss": 3.2745,
      "step": 3735
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8577405214309692,
      "learning_rate": 1.221946967329365e-05,
      "loss": 3.225,
      "step": 3740
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8643710017204285,
      "learning_rate": 1.2110477524084796e-05,
      "loss": 3.2312,
      "step": 3745
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.9053153395652771,
      "learning_rate": 1.2001906617466657e-05,
      "loss": 3.3218,
      "step": 3750
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6423649787902832,
      "learning_rate": 1.1893758160492978e-05,
      "loss": 3.2066,
      "step": 3755
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8391192555427551,
      "learning_rate": 1.1786033355520859e-05,
      "loss": 3.3188,
      "step": 3760
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8088417649269104,
      "learning_rate": 1.1678733400197373e-05,
      "loss": 3.2506,
      "step": 3765
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.829933226108551,
      "learning_rate": 1.1571859487446263e-05,
      "loss": 3.2841,
      "step": 3770
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.8629627227783203,
      "learning_rate": 1.1465412805454695e-05,
      "loss": 3.2619,
      "step": 3775
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8171300888061523,
      "learning_rate": 1.1359394537660011e-05,
      "loss": 3.1963,
      "step": 3780
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8476045727729797,
      "learning_rate": 1.125380586273661e-05,
      "loss": 3.2401,
      "step": 3785
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8052193522453308,
      "learning_rate": 1.1148647954582808e-05,
      "loss": 3.2368,
      "step": 3790
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8845239877700806,
      "learning_rate": 1.1043921982307819e-05,
      "loss": 3.2525,
      "step": 3795
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.843393087387085,
      "learning_rate": 1.0939629110218735e-05,
      "loss": 3.2181,
      "step": 3800
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8983203768730164,
      "learning_rate": 1.0835770497807596e-05,
      "loss": 3.2595,
      "step": 3805
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.8705394268035889,
      "learning_rate": 1.0732347299738493e-05,
      "loss": 3.2674,
      "step": 3810
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.841873288154602,
      "learning_rate": 1.0629360665834732e-05,
      "loss": 3.2982,
      "step": 3815
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.81884765625,
      "learning_rate": 1.052681174106604e-05,
      "loss": 3.2157,
      "step": 3820
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9083743095397949,
      "learning_rate": 1.0424701665535852e-05,
      "loss": 3.2913,
      "step": 3825
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8582718968391418,
      "learning_rate": 1.0323031574468638e-05,
      "loss": 3.244,
      "step": 3830
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9574825167655945,
      "learning_rate": 1.0221802598197261e-05,
      "loss": 3.21,
      "step": 3835
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8792760372161865,
      "learning_rate": 1.0121015862150423e-05,
      "loss": 3.2714,
      "step": 3840
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8689337372779846,
      "learning_rate": 1.0020672486840154e-05,
      "loss": 3.276,
      "step": 3845
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8836303353309631,
      "learning_rate": 9.920773587849364e-06,
      "loss": 3.2494,
      "step": 3850
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.822163462638855,
      "learning_rate": 9.821320275819401e-06,
      "loss": 3.2138,
      "step": 3855
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.8732519149780273,
      "learning_rate": 9.72231365643777e-06,
      "loss": 3.2743,
      "step": 3860
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.9054784178733826,
      "learning_rate": 9.623754830425779e-06,
      "loss": 3.1878,
      "step": 3865
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.2136757373809814,
      "learning_rate": 9.52564489352632e-06,
      "loss": 3.2454,
      "step": 3870
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9085641503334045,
      "learning_rate": 9.427984936491702e-06,
      "loss": 3.2894,
      "step": 3875
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9027767777442932,
      "learning_rate": 9.330776045071509e-06,
      "loss": 3.1571,
      "step": 3880
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.1602563858032227,
      "learning_rate": 9.23401930000054e-06,
      "loss": 3.2342,
      "step": 3885
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.913773775100708,
      "learning_rate": 9.137715776986772e-06,
      "loss": 3.2798,
      "step": 3890
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9193867444992065,
      "learning_rate": 9.041866546699434e-06,
      "loss": 3.2185,
      "step": 3895
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.0573819875717163,
      "learning_rate": 8.946472674757078e-06,
      "loss": 3.2578,
      "step": 3900
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.9102919697761536,
      "learning_rate": 8.851535221715735e-06,
      "loss": 3.2902,
      "step": 3905
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.1361302137374878,
      "learning_rate": 8.757055243057132e-06,
      "loss": 3.2576,
      "step": 3910
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.115525722503662,
      "learning_rate": 8.663033789176967e-06,
      "loss": 3.2436,
      "step": 3915
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.8845899105072021,
      "learning_rate": 8.5694719053732e-06,
      "loss": 3.247,
      "step": 3920
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9481613039970398,
      "learning_rate": 8.476370631834458e-06,
      "loss": 3.2395,
      "step": 3925
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8820552825927734,
      "learning_rate": 8.383731003628452e-06,
      "loss": 3.2178,
      "step": 3930
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8499136567115784,
      "learning_rate": 8.291554050690508e-06,
      "loss": 3.2706,
      "step": 3935
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8514552116394043,
      "learning_rate": 8.199840797812058e-06,
      "loss": 3.2353,
      "step": 3940
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8272942900657654,
      "learning_rate": 8.108592264629295e-06,
      "loss": 3.2831,
      "step": 3945
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8572783470153809,
      "learning_rate": 8.017809465611803e-06,
      "loss": 3.2337,
      "step": 3950
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.8681054711341858,
      "learning_rate": 7.927493410051324e-06,
      "loss": 3.2473,
      "step": 3955
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.9493885040283203,
      "learning_rate": 7.837645102050473e-06,
      "loss": 3.2802,
      "step": 3960
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.881588876247406,
      "learning_rate": 7.748265540511635e-06,
      "loss": 3.3412,
      "step": 3965
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8257955312728882,
      "learning_rate": 7.65935571912582e-06,
      "loss": 3.2538,
      "step": 3970
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.848596453666687,
      "learning_rate": 7.5709166263616405e-06,
      "loss": 3.2224,
      "step": 3975
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9227803349494934,
      "learning_rate": 7.482949245454302e-06,
      "loss": 3.286,
      "step": 3980
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8736422061920166,
      "learning_rate": 7.3954545543946876e-06,
      "loss": 3.286,
      "step": 3985
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8847970366477966,
      "learning_rate": 7.308433525918468e-06,
      "loss": 3.3588,
      "step": 3990
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.8576563000679016,
      "learning_rate": 7.221887127495313e-06,
      "loss": 3.2342,
      "step": 3995
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.9085653424263,
      "learning_rate": 7.1358163213181114e-06,
      "loss": 3.1536,
      "step": 4000
    }
  ],
  "logging_steps": 5,
  "max_steps": 4811,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 1.0865380348133376e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}