{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.5626695604991863,
  "eval_steps": 500,
  "global_step": 4500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0017362995116657625,
      "grad_norm": 0.357421875,
      "learning_rate": 0.0001,
      "loss": 1.5607,
      "step": 5
    },
    {
      "epoch": 0.003472599023331525,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0001,
      "loss": 0.8611,
      "step": 10
    },
    {
      "epoch": 0.005208898534997287,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0001,
      "loss": 0.746,
      "step": 15
    },
    {
      "epoch": 0.00694519804666305,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0001,
      "loss": 0.7252,
      "step": 20
    },
    {
      "epoch": 0.008681497558328812,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.7153,
      "step": 25
    },
    {
      "epoch": 0.010417797069994574,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0001,
      "loss": 0.666,
      "step": 30
    },
    {
      "epoch": 0.012154096581660336,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0001,
      "loss": 0.6443,
      "step": 35
    },
    {
      "epoch": 0.0138903960933261,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.6404,
      "step": 40
    },
    {
      "epoch": 0.01562669560499186,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.6227,
      "step": 45
    },
    {
      "epoch": 0.017362995116657624,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0001,
      "loss": 0.6077,
      "step": 50
    },
    {
      "epoch": 0.019099294628323386,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0001,
      "loss": 0.6814,
      "step": 55
    },
    {
      "epoch": 0.020835594139989148,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0001,
      "loss": 0.6831,
      "step": 60
    },
    {
      "epoch": 0.02257189365165491,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.6659,
      "step": 65
    },
    {
      "epoch": 0.02430819316332067,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0001,
      "loss": 0.6521,
      "step": 70
    },
    {
      "epoch": 0.026044492674986434,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.6309,
      "step": 75
    },
    {
      "epoch": 0.0277807921866522,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.6263,
      "step": 80
    },
    {
      "epoch": 0.02951709169831796,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.6268,
      "step": 85
    },
    {
      "epoch": 0.03125339120998372,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.6309,
      "step": 90
    },
    {
      "epoch": 0.032989690721649485,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.6103,
      "step": 95
    },
    {
      "epoch": 0.03472599023331525,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0001,
      "loss": 0.6125,
      "step": 100
    },
    {
      "epoch": 0.03646228974498101,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0001,
      "loss": 0.6828,
      "step": 105
    },
    {
      "epoch": 0.03819858925664677,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0001,
      "loss": 0.6517,
      "step": 110
    },
    {
      "epoch": 0.03993488876831253,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0001,
      "loss": 0.6702,
      "step": 115
    },
    {
      "epoch": 0.041671188279978295,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0001,
      "loss": 0.6158,
      "step": 120
    },
    {
      "epoch": 0.04340748779164406,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.6157,
      "step": 125
    },
    {
      "epoch": 0.04514378730330982,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0001,
      "loss": 0.6444,
      "step": 130
    },
    {
      "epoch": 0.04688008681497558,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.6053,
      "step": 135
    },
    {
      "epoch": 0.04861638632664134,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5974,
      "step": 140
    },
    {
      "epoch": 0.050352685838307105,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0001,
      "loss": 0.5874,
      "step": 145
    },
    {
      "epoch": 0.05208898534997287,
      "grad_norm": 0.109375,
      "learning_rate": 0.0001,
      "loss": 0.5656,
      "step": 150
    },
    {
      "epoch": 0.05382528486163863,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0001,
      "loss": 0.7098,
      "step": 155
    },
    {
      "epoch": 0.0555615843733044,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0001,
      "loss": 0.6968,
      "step": 160
    },
    {
      "epoch": 0.05729788388497016,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.652,
      "step": 165
    },
    {
      "epoch": 0.05903418339663592,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0001,
      "loss": 0.6084,
      "step": 170
    },
    {
      "epoch": 0.060770482908301685,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0001,
      "loss": 0.6243,
      "step": 175
    },
    {
      "epoch": 0.06250678241996745,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.6087,
      "step": 180
    },
    {
      "epoch": 0.06424308193163321,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.6189,
      "step": 185
    },
    {
      "epoch": 0.06597938144329897,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5926,
      "step": 190
    },
    {
      "epoch": 0.06771568095496473,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5846,
      "step": 195
    },
    {
      "epoch": 0.0694519804666305,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0001,
      "loss": 0.5668,
      "step": 200
    },
    {
      "epoch": 0.07118827997829626,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.6842,
      "step": 205
    },
    {
      "epoch": 0.07292457948996202,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0001,
      "loss": 0.6476,
      "step": 210
    },
    {
      "epoch": 0.07466087900162778,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.6269,
      "step": 215
    },
    {
      "epoch": 0.07639717851329354,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.6189,
      "step": 220
    },
    {
      "epoch": 0.0781334780249593,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.617,
      "step": 225
    },
    {
      "epoch": 0.07986977753662507,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.649,
      "step": 230
    },
    {
      "epoch": 0.08160607704829083,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5985,
      "step": 235
    },
    {
      "epoch": 0.08334237655995659,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0001,
      "loss": 0.5909,
      "step": 240
    },
    {
      "epoch": 0.08507867607162235,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.6015,
      "step": 245
    },
    {
      "epoch": 0.08681497558328811,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0001,
      "loss": 0.5706,
      "step": 250
    },
    {
      "epoch": 0.08855127509495388,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.6721,
      "step": 255
    },
    {
      "epoch": 0.09028757460661964,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0001,
      "loss": 0.6523,
      "step": 260
    },
    {
      "epoch": 0.0920238741182854,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.6232,
      "step": 265
    },
    {
      "epoch": 0.09376017362995116,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.642,
      "step": 270
    },
    {
      "epoch": 0.09549647314161692,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.6214,
      "step": 275
    },
    {
      "epoch": 0.09723277265328269,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5981,
      "step": 280
    },
    {
      "epoch": 0.09896907216494845,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.6205,
      "step": 285
    },
    {
      "epoch": 0.10070537167661421,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.6008,
      "step": 290
    },
    {
      "epoch": 0.10244167118827997,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5768,
      "step": 295
    },
    {
      "epoch": 0.10417797069994574,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5863,
      "step": 300
    },
    {
      "epoch": 0.1059142702116115,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.6621,
      "step": 305
    },
    {
      "epoch": 0.10765056972327726,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.0001,
      "loss": 0.6285,
      "step": 310
    },
    {
      "epoch": 0.10938686923494302,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.6254,
      "step": 315
    },
    {
      "epoch": 0.1111231687466088,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.6164,
      "step": 320
    },
    {
      "epoch": 0.11285946825827456,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.623,
      "step": 325
    },
    {
      "epoch": 0.11459576776994032,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5782,
      "step": 330
    },
    {
      "epoch": 0.11633206728160608,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0001,
      "loss": 0.5962,
      "step": 335
    },
    {
      "epoch": 0.11806836679327185,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5987,
      "step": 340
    },
    {
      "epoch": 0.11980466630493761,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5885,
      "step": 345
    },
    {
      "epoch": 0.12154096581660337,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5678,
      "step": 350
    },
    {
      "epoch": 0.12327726532826913,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.674,
      "step": 355
    },
    {
      "epoch": 0.1250135648399349,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0001,
      "loss": 0.631,
      "step": 360
    },
    {
      "epoch": 0.12674986435160066,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.6447,
      "step": 365
    },
    {
      "epoch": 0.12848616386326642,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.6277,
      "step": 370
    },
    {
      "epoch": 0.13022246337493218,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5902,
      "step": 375
    },
    {
      "epoch": 0.13195876288659794,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.6211,
      "step": 380
    },
    {
      "epoch": 0.1336950623982637,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.6151,
      "step": 385
    },
    {
      "epoch": 0.13543136190992947,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.6097,
      "step": 390
    },
    {
      "epoch": 0.13716766142159523,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5842,
      "step": 395
    },
    {
      "epoch": 0.138903960933261,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5522,
      "step": 400
    },
    {
      "epoch": 0.14064026044492675,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.6664,
      "step": 405
    },
    {
      "epoch": 0.1423765599565925,
      "grad_norm": 0.111328125,
      "learning_rate": 0.0001,
      "loss": 0.647,
      "step": 410
    },
    {
      "epoch": 0.14411285946825828,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6205,
      "step": 415
    },
    {
      "epoch": 0.14584915897992404,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.615,
      "step": 420
    },
    {
      "epoch": 0.1475854584915898,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.605,
      "step": 425
    },
    {
      "epoch": 0.14932175800325556,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5884,
      "step": 430
    },
    {
      "epoch": 0.15105805751492132,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.6043,
      "step": 435
    },
    {
      "epoch": 0.15279435702658709,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5842,
      "step": 440
    },
    {
      "epoch": 0.15453065653825285,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5745,
      "step": 445
    },
    {
      "epoch": 0.1562669560499186,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0001,
      "loss": 0.5975,
      "step": 450
    },
    {
      "epoch": 0.15800325556158437,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.6604,
      "step": 455
    },
    {
      "epoch": 0.15973955507325013,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0001,
      "loss": 0.6376,
      "step": 460
    },
    {
      "epoch": 0.1614758545849159,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.6241,
      "step": 465
    },
    {
      "epoch": 0.16321215409658166,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.6238,
      "step": 470
    },
    {
      "epoch": 0.16494845360824742,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5941,
      "step": 475
    },
    {
      "epoch": 0.16668475311991318,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.5897,
      "step": 480
    },
    {
      "epoch": 0.16842105263157894,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.6106,
      "step": 485
    },
    {
      "epoch": 0.1701573521432447,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5948,
      "step": 490
    },
    {
      "epoch": 0.17189365165491047,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5811,
      "step": 495
    },
    {
      "epoch": 0.17362995116657623,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.5604,
      "step": 500
    },
    {
      "epoch": 0.175366250678242,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0001,
      "loss": 0.6414,
      "step": 505
    },
    {
      "epoch": 0.17710255018990775,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0001,
      "loss": 0.6296,
      "step": 510
    },
    {
      "epoch": 0.17883884970157352,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.6161,
      "step": 515
    },
    {
      "epoch": 0.18057514921323928,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0001,
      "loss": 0.5981,
      "step": 520
    },
    {
      "epoch": 0.18231144872490504,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.6258,
      "step": 525
    },
    {
      "epoch": 0.1840477482365708,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.591,
      "step": 530
    },
    {
      "epoch": 0.18578404774823656,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5829,
      "step": 535
    },
    {
      "epoch": 0.18752034725990233,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5749,
      "step": 540
    },
    {
      "epoch": 0.1892566467715681,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5908,
      "step": 545
    },
    {
      "epoch": 0.19099294628323385,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0001,
      "loss": 0.5533,
      "step": 550
    },
    {
      "epoch": 0.1927292457948996,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0001,
      "loss": 0.6469,
      "step": 555
    },
    {
      "epoch": 0.19446554530656537,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.61,
      "step": 560
    },
    {
      "epoch": 0.19620184481823114,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.6342,
      "step": 565
    },
    {
      "epoch": 0.1979381443298969,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.6017,
      "step": 570
    },
    {
      "epoch": 0.19967444384156266,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5873,
      "step": 575
    },
    {
      "epoch": 0.20141074335322842,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5898,
      "step": 580
    },
    {
      "epoch": 0.20314704286489418,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5927,
      "step": 585
    },
    {
      "epoch": 0.20488334237655995,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5641,
      "step": 590
    },
    {
      "epoch": 0.2066196418882257,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5742,
      "step": 595
    },
    {
      "epoch": 0.20835594139989147,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0001,
      "loss": 0.566,
      "step": 600
    },
    {
      "epoch": 0.21009224091155723,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.6651,
      "step": 605
    },
    {
      "epoch": 0.211828540423223,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.6181,
      "step": 610
    },
    {
      "epoch": 0.21356483993488876,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.6136,
      "step": 615
    },
    {
      "epoch": 0.21530113944655452,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.5951,
      "step": 620
    },
    {
      "epoch": 0.21703743895822028,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5696,
      "step": 625
    },
    {
      "epoch": 0.21877373846988604,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5912,
      "step": 630
    },
    {
      "epoch": 0.2205100379815518,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.587,
      "step": 635
    },
    {
      "epoch": 0.2222463374932176,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0001,
      "loss": 0.5574,
      "step": 640
    },
    {
      "epoch": 0.22398263700488336,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5815,
      "step": 645
    },
    {
      "epoch": 0.22571893651654912,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0001,
      "loss": 0.5718,
      "step": 650
    },
    {
      "epoch": 0.22745523602821488,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.6307,
      "step": 655
    },
    {
      "epoch": 0.22919153553988064,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.6099,
      "step": 660
    },
    {
      "epoch": 0.2309278350515464,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.6032,
      "step": 665
    },
    {
      "epoch": 0.23266413456321217,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5917,
      "step": 670
    },
    {
      "epoch": 0.23440043407487793,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5869,
      "step": 675
    },
    {
      "epoch": 0.2361367335865437,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5737,
      "step": 680
    },
    {
      "epoch": 0.23787303309820945,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.586,
      "step": 685
    },
    {
      "epoch": 0.23960933260987521,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.5844,
      "step": 690
    },
    {
      "epoch": 0.24134563212154098,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5819,
      "step": 695
    },
    {
      "epoch": 0.24308193163320674,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5453,
      "step": 700
    },
    {
      "epoch": 0.2448182311448725,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.6409,
      "step": 705
    },
    {
      "epoch": 0.24655453065653826,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6099,
      "step": 710
    },
    {
      "epoch": 0.24829083016820402,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.6267,
      "step": 715
    },
    {
      "epoch": 0.2500271296798698,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.6319,
      "step": 720
    },
    {
      "epoch": 0.25176342919153555,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0001,
      "loss": 0.5841,
      "step": 725
    },
    {
      "epoch": 0.2534997287032013,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.5665,
      "step": 730
    },
    {
      "epoch": 0.2552360282148671,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5881,
      "step": 735
    },
    {
      "epoch": 0.25697232772653283,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5599,
      "step": 740
    },
    {
      "epoch": 0.2587086272381986,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.5614,
      "step": 745
    },
    {
      "epoch": 0.26044492674986436,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.5683,
      "step": 750
    },
    {
      "epoch": 0.2621812262615301,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.6652,
      "step": 755
    },
    {
      "epoch": 0.2639175257731959,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.6104,
      "step": 760
    },
    {
      "epoch": 0.26565382528486164,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5999,
      "step": 765
    },
    {
      "epoch": 0.2673901247965274,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5882,
      "step": 770
    },
    {
      "epoch": 0.26912642430819317,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5826,
      "step": 775
    },
    {
      "epoch": 0.27086272381985893,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5648,
      "step": 780
    },
    {
      "epoch": 0.2725990233315247,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5883,
      "step": 785
    },
    {
      "epoch": 0.27433532284319045,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5872,
      "step": 790
    },
    {
      "epoch": 0.2760716223548562,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5411,
      "step": 795
    },
    {
      "epoch": 0.277807921866522,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0001,
      "loss": 0.5518,
      "step": 800
    },
    {
      "epoch": 0.27954422137818774,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6338,
      "step": 805
    },
    {
      "epoch": 0.2812805208898535,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.6061,
      "step": 810
    },
    {
      "epoch": 0.28301682040151926,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.6213,
      "step": 815
    },
    {
      "epoch": 0.284753119913185,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5815,
      "step": 820
    },
    {
      "epoch": 0.2864894194248508,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.6088,
      "step": 825
    },
    {
      "epoch": 0.28822571893651655,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5935,
      "step": 830
    },
    {
      "epoch": 0.2899620184481823,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5708,
      "step": 835
    },
    {
      "epoch": 0.2916983179598481,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5919,
      "step": 840
    },
    {
      "epoch": 0.29343461747151384,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5402,
      "step": 845
    },
    {
      "epoch": 0.2951709169831796,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0001,
      "loss": 0.5421,
      "step": 850
    },
    {
      "epoch": 0.29690721649484536,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.6586,
      "step": 855
    },
    {
      "epoch": 0.2986435160065111,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.6076,
      "step": 860
    },
    {
      "epoch": 0.3003798155181769,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.6283,
      "step": 865
    },
    {
      "epoch": 0.30211611502984265,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.6065,
      "step": 870
    },
    {
      "epoch": 0.3038524145415084,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5651,
      "step": 875
    },
    {
      "epoch": 0.30558871405317417,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0001,
      "loss": 0.6081,
      "step": 880
    },
    {
      "epoch": 0.30732501356483993,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.5631,
      "step": 885
    },
    {
      "epoch": 0.3090613130765057,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5562,
      "step": 890
    },
    {
      "epoch": 0.31079761258817146,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5865,
      "step": 895
    },
    {
      "epoch": 0.3125339120998372,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0001,
      "loss": 0.546,
      "step": 900
    },
    {
      "epoch": 0.314270211611503,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.6489,
      "step": 905
    },
    {
      "epoch": 0.31600651112316874,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.6021,
      "step": 910
    },
    {
      "epoch": 0.3177428106348345,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.6069,
      "step": 915
    },
    {
      "epoch": 0.31947911014650027,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5969,
      "step": 920
    },
    {
      "epoch": 0.32121540965816603,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5781,
      "step": 925
    },
    {
      "epoch": 0.3229517091698318,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5975,
      "step": 930
    },
    {
      "epoch": 0.32468800868149755,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5856,
      "step": 935
    },
    {
      "epoch": 0.3264243081931633,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5709,
      "step": 940
    },
    {
      "epoch": 0.3281606077048291,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5523,
      "step": 945
    },
    {
      "epoch": 0.32989690721649484,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0001,
      "loss": 0.5505,
      "step": 950
    },
    {
      "epoch": 0.3316332067281606,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.6636,
      "step": 955
    },
    {
      "epoch": 0.33336950623982636,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.6165,
      "step": 960
    },
    {
      "epoch": 0.3351058057514921,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.6182,
      "step": 965
    },
    {
      "epoch": 0.3368421052631579,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5859,
      "step": 970
    },
    {
      "epoch": 0.33857840477482365,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.6026,
      "step": 975
    },
    {
      "epoch": 0.3403147042864894,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.5777,
      "step": 980
    },
    {
      "epoch": 0.3420510037981552,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5612,
      "step": 985
    },
    {
      "epoch": 0.34378730330982094,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5838,
      "step": 990
    },
    {
      "epoch": 0.3455236028214867,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5567,
      "step": 995
    },
    {
      "epoch": 0.34725990233315246,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5563,
      "step": 1000
    },
    {
      "epoch": 0.3489962018448182,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.6221,
      "step": 1005
    },
    {
      "epoch": 0.350732501356484,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.613,
      "step": 1010
    },
    {
      "epoch": 0.35246880086814975,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.603,
      "step": 1015
    },
    {
      "epoch": 0.3542051003798155,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.6122,
      "step": 1020
    },
    {
      "epoch": 0.35594139989148127,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.6032,
      "step": 1025
    },
    {
      "epoch": 0.35767769940314703,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.6027,
      "step": 1030
    },
    {
      "epoch": 0.3594139989148128,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5592,
      "step": 1035
    },
    {
      "epoch": 0.36115029842647856,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5676,
      "step": 1040
    },
    {
      "epoch": 0.3628865979381443,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5578,
      "step": 1045
    },
    {
      "epoch": 0.3646228974498101,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5349,
      "step": 1050
    },
    {
      "epoch": 0.36635919696147584,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.0001,
      "loss": 0.6616,
      "step": 1055
    },
    {
      "epoch": 0.3680954964731416,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.6331,
      "step": 1060
    },
    {
      "epoch": 0.36983179598480737,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.6057,
      "step": 1065
    },
    {
      "epoch": 0.3715680954964731,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.5951,
      "step": 1070
    },
    {
      "epoch": 0.3733043950081389,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.5747,
      "step": 1075
    },
    {
      "epoch": 0.37504069451980465,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5827,
      "step": 1080
    },
    {
      "epoch": 0.3767769940314704,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5855,
      "step": 1085
    },
    {
      "epoch": 0.3785132935431362,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5862,
      "step": 1090
    },
    {
      "epoch": 0.38024959305480194,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5525,
      "step": 1095
    },
    {
      "epoch": 0.3819858925664677,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0001,
      "loss": 0.5491,
      "step": 1100
    },
    {
      "epoch": 0.38372219207813346,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.6284,
      "step": 1105
    },
    {
      "epoch": 0.3854584915897992,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5915,
      "step": 1110
    },
    {
      "epoch": 0.387194791101465,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.6242,
      "step": 1115
    },
    {
      "epoch": 0.38893109061313075,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0001,
      "loss": 0.5679,
      "step": 1120
    },
    {
      "epoch": 0.3906673901247965,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5795,
      "step": 1125
    },
    {
      "epoch": 0.39240368963646227,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5973,
      "step": 1130
    },
    {
      "epoch": 0.39413998914812803,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0001,
      "loss": 0.5729,
      "step": 1135
    },
    {
      "epoch": 0.3958762886597938,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5757,
      "step": 1140
    },
    {
      "epoch": 0.39761258817145956,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5745,
      "step": 1145
    },
    {
      "epoch": 0.3993488876831253,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5563,
      "step": 1150
    },
    {
      "epoch": 0.4010851871947911,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.637,
      "step": 1155
    },
    {
      "epoch": 0.40282148670645684,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.597,
      "step": 1160
    },
    {
      "epoch": 0.4045577862181226,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.598,
      "step": 1165
    },
    {
      "epoch": 0.40629408572978837,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5984,
      "step": 1170
    },
    {
      "epoch": 0.40803038524145413,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.5883,
      "step": 1175
    },
    {
      "epoch": 0.4097666847531199,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5725,
      "step": 1180
    },
    {
      "epoch": 0.41150298426478565,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5848,
      "step": 1185
    },
    {
      "epoch": 0.4132392837764514,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5757,
      "step": 1190
    },
    {
      "epoch": 0.4149755832881172,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5538,
      "step": 1195
    },
    {
      "epoch": 0.41671188279978294,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0001,
      "loss": 0.531,
      "step": 1200
    },
    {
      "epoch": 0.4184481823114487,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.6415,
      "step": 1205
    },
    {
      "epoch": 0.42018448182311446,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.6296,
      "step": 1210
    },
    {
      "epoch": 0.4219207813347802,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.6157,
      "step": 1215
    },
    {
      "epoch": 0.423657080846446,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0001,
      "loss": 0.5636,
      "step": 1220
    },
    {
      "epoch": 0.42539338035811175,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5655,
      "step": 1225
    },
    {
      "epoch": 0.4271296798697775,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5727,
      "step": 1230
    },
    {
      "epoch": 0.4288659793814433,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5615,
      "step": 1235
    },
    {
      "epoch": 0.43060227889310904,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5715,
      "step": 1240
    },
    {
      "epoch": 0.4323385784047748,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5566,
      "step": 1245
    },
    {
      "epoch": 0.43407487791644056,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0001,
      "loss": 0.5504,
      "step": 1250
    },
    {
      "epoch": 0.4358111774281063,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.6203,
      "step": 1255
    },
    {
      "epoch": 0.4375474769397721,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.6021,
      "step": 1260
    },
    {
      "epoch": 0.43928377645143785,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.6065,
      "step": 1265
    },
    {
      "epoch": 0.4410200759631036,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5844,
      "step": 1270
    },
    {
      "epoch": 0.44275637547476937,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5717,
      "step": 1275
    },
    {
      "epoch": 0.4444926749864352,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5893,
      "step": 1280
    },
    {
      "epoch": 0.44622897449810095,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.5707,
      "step": 1285
    },
    {
      "epoch": 0.4479652740097667,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5494,
      "step": 1290
    },
    {
      "epoch": 0.4497015735214325,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5564,
      "step": 1295
    },
    {
      "epoch": 0.45143787303309824,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5536,
      "step": 1300
    },
    {
      "epoch": 0.453174172544764,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0001,
      "loss": 0.6417,
      "step": 1305
    },
    {
      "epoch": 0.45491047205642976,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0001,
      "loss": 0.6055,
      "step": 1310
    },
    {
      "epoch": 0.4566467715680955,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5962,
      "step": 1315
    },
    {
      "epoch": 0.4583830710797613,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0001,
      "loss": 0.6115,
      "step": 1320
    },
    {
      "epoch": 0.46011937059142705,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5898,
      "step": 1325
    },
    {
      "epoch": 0.4618556701030928,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5834,
      "step": 1330
    },
    {
      "epoch": 0.46359196961475857,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5721,
      "step": 1335
    },
    {
      "epoch": 0.46532826912642433,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.561,
      "step": 1340
    },
    {
      "epoch": 0.4670645686380901,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5637,
      "step": 1345
    },
    {
      "epoch": 0.46880086814975586,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5528,
      "step": 1350
    },
    {
      "epoch": 0.4705371676614216,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.6466,
      "step": 1355
    },
    {
      "epoch": 0.4722734671730874,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0001,
      "loss": 0.6073,
      "step": 1360
    },
    {
      "epoch": 0.47400976668475314,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.6299,
      "step": 1365
    },
    {
      "epoch": 0.4757460661964189,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5787,
      "step": 1370
    },
    {
      "epoch": 0.47748236570808467,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5815,
      "step": 1375
    },
    {
      "epoch": 0.47921866521975043,
      "grad_norm": 0.2451171875,
      "learning_rate": 0.0001,
      "loss": 0.5751,
      "step": 1380
    },
    {
      "epoch": 0.4809549647314162,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5652,
      "step": 1385
    },
    {
      "epoch": 0.48269126424308195,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5488,
      "step": 1390
    },
    {
      "epoch": 0.4844275637547477,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5472,
      "step": 1395
    },
    {
      "epoch": 0.4861638632664135,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0001,
      "loss": 0.5394,
      "step": 1400
    },
    {
      "epoch": 0.48790016277807924,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6331,
      "step": 1405
    },
    {
      "epoch": 0.489636462289745,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.6019,
      "step": 1410
    },
    {
      "epoch": 0.49137276180141076,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.6056,
      "step": 1415
    },
    {
      "epoch": 0.4931090613130765,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0001,
      "loss": 0.6167,
      "step": 1420
    },
    {
      "epoch": 0.4948453608247423,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0001,
      "loss": 0.5708,
      "step": 1425
    },
    {
      "epoch": 0.49658166033640805,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.591,
      "step": 1430
    },
    {
      "epoch": 0.4983179598480738,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0001,
      "loss": 0.5364,
      "step": 1435
    },
    {
      "epoch": 0.5000542593597396,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5634,
      "step": 1440
    },
    {
      "epoch": 0.5017905588714053,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5374,
      "step": 1445
    },
    {
      "epoch": 0.5035268583830711,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0001,
      "loss": 0.5533,
      "step": 1450
    },
    {
      "epoch": 0.5052631578947369,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.6605,
      "step": 1455
    },
    {
      "epoch": 0.5069994574064026,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.6167,
      "step": 1460
    },
    {
      "epoch": 0.5087357569180684,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.6013,
      "step": 1465
    },
    {
      "epoch": 0.5104720564297341,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5774,
      "step": 1470
    },
    {
      "epoch": 0.5122083559413999,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5837,
      "step": 1475
    },
    {
      "epoch": 0.5139446554530657,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5885,
      "step": 1480
    },
    {
      "epoch": 0.5156809549647314,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5565,
      "step": 1485
    },
    {
      "epoch": 0.5174172544763972,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0001,
      "loss": 0.5793,
      "step": 1490
    },
    {
      "epoch": 0.519153553988063,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5552,
      "step": 1495
    },
    {
      "epoch": 0.5208898534997287,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.0001,
      "loss": 0.5376,
      "step": 1500
    },
    {
      "epoch": 0.5226261530113945,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.6368,
      "step": 1505
    },
    {
      "epoch": 0.5243624525230602,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.6176,
      "step": 1510
    },
    {
      "epoch": 0.526098752034726,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.5909,
      "step": 1515
    },
    {
      "epoch": 0.5278350515463918,
      "grad_norm": 0.076171875,
      "learning_rate": 0.0001,
      "loss": 0.5816,
      "step": 1520
    },
    {
      "epoch": 0.5295713510580575,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.6136,
      "step": 1525
    },
    {
      "epoch": 0.5313076505697233,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5724,
      "step": 1530
    },
    {
      "epoch": 0.533043950081389,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5657,
      "step": 1535
    },
    {
      "epoch": 0.5347802495930548,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5642,
      "step": 1540
    },
    {
      "epoch": 0.5365165491047206,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5711,
      "step": 1545
    },
    {
      "epoch": 0.5382528486163863,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0001,
      "loss": 0.514,
      "step": 1550
    },
    {
      "epoch": 0.5399891481280521,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.6508,
      "step": 1555
    },
    {
      "epoch": 0.5417254476397179,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.6109,
      "step": 1560
    },
    {
      "epoch": 0.5434617471513836,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.6041,
      "step": 1565
    },
    {
      "epoch": 0.5451980466630494,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0001,
      "loss": 0.5893,
      "step": 1570
    },
    {
      "epoch": 0.5469343461747151,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5753,
      "step": 1575
    },
    {
      "epoch": 0.5486706456863809,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5626,
      "step": 1580
    },
    {
      "epoch": 0.5504069451980467,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5912,
      "step": 1585
    },
    {
      "epoch": 0.5521432447097124,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5629,
      "step": 1590
    },
    {
      "epoch": 0.5538795442213782,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.581,
      "step": 1595
    },
    {
      "epoch": 0.555615843733044,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0001,
      "loss": 0.5413,
      "step": 1600
    },
    {
      "epoch": 0.5573521432447097,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.6606,
      "step": 1605
    },
    {
      "epoch": 0.5590884427563755,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.5923,
      "step": 1610
    },
    {
      "epoch": 0.5608247422680412,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5958,
      "step": 1615
    },
    {
      "epoch": 0.562561041779707,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0001,
      "loss": 0.5899,
      "step": 1620
    },
    {
      "epoch": 0.5642973412913728,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5817,
      "step": 1625
    },
    {
      "epoch": 0.5660336408030385,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5579,
      "step": 1630
    },
    {
      "epoch": 0.5677699403147043,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5677,
      "step": 1635
    },
    {
      "epoch": 0.56950623982637,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.566,
      "step": 1640
    },
    {
      "epoch": 0.5712425393380358,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5668,
      "step": 1645
    },
    {
      "epoch": 0.5729788388497016,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5334,
      "step": 1650
    },
    {
      "epoch": 0.5747151383613673,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.6322,
      "step": 1655
    },
    {
      "epoch": 0.5764514378730331,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.588,
      "step": 1660
    },
    {
      "epoch": 0.5781877373846989,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5929,
      "step": 1665
    },
    {
      "epoch": 0.5799240368963646,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.6142,
      "step": 1670
    },
    {
      "epoch": 0.5816603364080304,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5919,
      "step": 1675
    },
    {
      "epoch": 0.5833966359196961,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.5811,
      "step": 1680
    },
    {
      "epoch": 0.5851329354313619,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0001,
      "loss": 0.5564,
      "step": 1685
    },
    {
      "epoch": 0.5868692349430277,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5658,
      "step": 1690
    },
    {
      "epoch": 0.5886055344546934,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5677,
      "step": 1695
    },
    {
      "epoch": 0.5903418339663592,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0001,
      "loss": 0.5537,
      "step": 1700
    },
    {
      "epoch": 0.592078133478025,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.6258,
      "step": 1705
    },
    {
      "epoch": 0.5938144329896907,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5946,
      "step": 1710
    },
    {
      "epoch": 0.5955507325013565,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5991,
      "step": 1715
    },
    {
      "epoch": 0.5972870320130222,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001,
      "loss": 0.602,
      "step": 1720
    },
    {
      "epoch": 0.599023331524688,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5778,
      "step": 1725
    },
    {
      "epoch": 0.6007596310363538,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5531,
      "step": 1730
    },
    {
      "epoch": 0.6024959305480195,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.56,
      "step": 1735
    },
    {
      "epoch": 0.6042322300596853,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5679,
      "step": 1740
    },
    {
      "epoch": 0.6059685295713511,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.557,
      "step": 1745
    },
    {
      "epoch": 0.6077048290830168,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5345,
      "step": 1750
    },
    {
      "epoch": 0.6094411285946826,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.6289,
      "step": 1755
    },
    {
      "epoch": 0.6111774281063483,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.6079,
      "step": 1760
    },
    {
      "epoch": 0.6129137276180141,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5773,
      "step": 1765
    },
    {
      "epoch": 0.6146500271296799,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5802,
      "step": 1770
    },
    {
      "epoch": 0.6163863266413456,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5654,
      "step": 1775
    },
    {
      "epoch": 0.6181226261530114,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5679,
      "step": 1780
    },
    {
      "epoch": 0.6198589256646772,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5849,
      "step": 1785
    },
    {
      "epoch": 0.6215952251763429,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5674,
      "step": 1790
    },
    {
      "epoch": 0.6233315246880087,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.56,
      "step": 1795
    },
    {
      "epoch": 0.6250678241996744,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.527,
      "step": 1800
    },
    {
      "epoch": 0.6268041237113402,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.6374,
      "step": 1805
    },
    {
      "epoch": 0.628540423223006,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001,
      "loss": 0.5962,
      "step": 1810
    },
    {
      "epoch": 0.6302767227346717,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.6013,
      "step": 1815
    },
    {
      "epoch": 0.6320130222463375,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6017,
      "step": 1820
    },
    {
      "epoch": 0.6337493217580032,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5826,
      "step": 1825
    },
    {
      "epoch": 0.635485621269669,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5977,
      "step": 1830
    },
    {
      "epoch": 0.6372219207813348,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5981,
      "step": 1835
    },
    {
      "epoch": 0.6389582202930005,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5768,
      "step": 1840
    },
    {
      "epoch": 0.6406945198046663,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5415,
      "step": 1845
    },
    {
      "epoch": 0.6424308193163321,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5317,
      "step": 1850
    },
    {
      "epoch": 0.6441671188279978,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.6237,
      "step": 1855
    },
    {
      "epoch": 0.6459034183396636,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5886,
      "step": 1860
    },
    {
      "epoch": 0.6476397178513293,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5769,
      "step": 1865
    },
    {
      "epoch": 0.6493760173629951,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5793,
      "step": 1870
    },
    {
      "epoch": 0.6511123168746609,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.5704,
      "step": 1875
    },
    {
      "epoch": 0.6528486163863266,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5597,
      "step": 1880
    },
    {
      "epoch": 0.6545849158979924,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5358,
      "step": 1885
    },
    {
      "epoch": 0.6563212154096582,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5479,
      "step": 1890
    },
    {
      "epoch": 0.6580575149213239,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5472,
      "step": 1895
    },
    {
      "epoch": 0.6597938144329897,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5398,
      "step": 1900
    },
    {
      "epoch": 0.6615301139446554,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.651,
      "step": 1905
    },
    {
      "epoch": 0.6632664134563212,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.5915,
      "step": 1910
    },
    {
      "epoch": 0.665002712967987,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.6007,
      "step": 1915
    },
    {
      "epoch": 0.6667390124796527,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5788,
      "step": 1920
    },
    {
      "epoch": 0.6684753119913185,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.568,
      "step": 1925
    },
    {
      "epoch": 0.6702116115029843,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0001,
      "loss": 0.5711,
      "step": 1930
    },
    {
      "epoch": 0.67194791101465,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5682,
      "step": 1935
    },
    {
      "epoch": 0.6736842105263158,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5764,
      "step": 1940
    },
    {
      "epoch": 0.6754205100379815,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5491,
      "step": 1945
    },
    {
      "epoch": 0.6771568095496473,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0001,
      "loss": 0.5629,
      "step": 1950
    },
    {
      "epoch": 0.6788931090613131,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.6379,
      "step": 1955
    },
    {
      "epoch": 0.6806294085729788,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.5983,
      "step": 1960
    },
    {
      "epoch": 0.6823657080846446,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5929,
      "step": 1965
    },
    {
      "epoch": 0.6841020075963103,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.588,
      "step": 1970
    },
    {
      "epoch": 0.6858383071079761,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5619,
      "step": 1975
    },
    {
      "epoch": 0.6875746066196419,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0001,
      "loss": 0.5683,
      "step": 1980
    },
    {
      "epoch": 0.6893109061313076,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5808,
      "step": 1985
    },
    {
      "epoch": 0.6910472056429734,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5536,
      "step": 1990
    },
    {
      "epoch": 0.6927835051546392,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5608,
      "step": 1995
    },
    {
      "epoch": 0.6945198046663049,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0001,
      "loss": 0.534,
      "step": 2000
    },
    {
      "epoch": 0.6962561041779707,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.6431,
      "step": 2005
    },
    {
      "epoch": 0.6979924036896364,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5889,
      "step": 2010
    },
    {
      "epoch": 0.6997287032013022,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5804,
      "step": 2015
    },
    {
      "epoch": 0.701465002712968,
      "grad_norm": 0.0703125,
      "learning_rate": 0.0001,
      "loss": 0.5916,
      "step": 2020
    },
    {
      "epoch": 0.7032013022246337,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5759,
      "step": 2025
    },
    {
      "epoch": 0.7049376017362995,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5853,
      "step": 2030
    },
    {
      "epoch": 0.7066739012479653,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5733,
      "step": 2035
    },
    {
      "epoch": 0.708410200759631,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5776,
      "step": 2040
    },
    {
      "epoch": 0.7101465002712968,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5489,
      "step": 2045
    },
    {
      "epoch": 0.7118827997829625,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5241,
      "step": 2050
    },
    {
      "epoch": 0.7136190992946283,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.6735,
      "step": 2055
    },
    {
      "epoch": 0.7153553988062941,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.6125,
      "step": 2060
    },
    {
      "epoch": 0.7170916983179598,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5805,
      "step": 2065
    },
    {
      "epoch": 0.7188279978296256,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.6037,
      "step": 2070
    },
    {
      "epoch": 0.7205642973412913,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5873,
      "step": 2075
    },
    {
      "epoch": 0.7223005968529571,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5727,
      "step": 2080
    },
    {
      "epoch": 0.7240368963646229,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5709,
      "step": 2085
    },
    {
      "epoch": 0.7257731958762886,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.519,
      "step": 2090
    },
    {
      "epoch": 0.7275094953879544,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5576,
      "step": 2095
    },
    {
      "epoch": 0.7292457948996202,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5285,
      "step": 2100
    },
    {
      "epoch": 0.7309820944112859,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.6482,
      "step": 2105
    },
    {
      "epoch": 0.7327183939229517,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.595,
      "step": 2110
    },
    {
      "epoch": 0.7344546934346174,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5953,
      "step": 2115
    },
    {
      "epoch": 0.7361909929462832,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.5665,
      "step": 2120
    },
    {
      "epoch": 0.737927292457949,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5612,
      "step": 2125
    },
    {
      "epoch": 0.7396635919696147,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5645,
      "step": 2130
    },
    {
      "epoch": 0.7413998914812805,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5865,
      "step": 2135
    },
    {
      "epoch": 0.7431361909929463,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5658,
      "step": 2140
    },
    {
      "epoch": 0.744872490504612,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5523,
      "step": 2145
    },
    {
      "epoch": 0.7466087900162778,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5358,
      "step": 2150
    },
    {
      "epoch": 0.7483450895279435,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.6148,
      "step": 2155
    },
    {
      "epoch": 0.7500813890396093,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.6092,
      "step": 2160
    },
    {
      "epoch": 0.7518176885512751,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5939,
      "step": 2165
    },
    {
      "epoch": 0.7535539880629408,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5823,
      "step": 2170
    },
    {
      "epoch": 0.7552902875746066,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.5779,
      "step": 2175
    },
    {
      "epoch": 0.7570265870862724,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5697,
      "step": 2180
    },
    {
      "epoch": 0.7587628865979381,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.5716,
      "step": 2185
    },
    {
      "epoch": 0.7604991861096039,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.5569,
      "step": 2190
    },
    {
      "epoch": 0.7622354856212696,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001,
      "loss": 0.5544,
      "step": 2195
    },
    {
      "epoch": 0.7639717851329354,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5332,
      "step": 2200
    },
    {
      "epoch": 0.7657080846446012,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.6517,
      "step": 2205
    },
    {
      "epoch": 0.7674443841562669,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5991,
      "step": 2210
    },
    {
      "epoch": 0.7691806836679327,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5823,
      "step": 2215
    },
    {
      "epoch": 0.7709169831795984,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.6052,
      "step": 2220
    },
    {
      "epoch": 0.7726532826912642,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.5711,
      "step": 2225
    },
    {
      "epoch": 0.77438958220293,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001,
      "loss": 0.5369,
      "step": 2230
    },
    {
      "epoch": 0.7761258817145957,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001,
      "loss": 0.5406,
      "step": 2235
    },
    {
      "epoch": 0.7778621812262615,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.604,
      "step": 2240
    },
    {
      "epoch": 0.7795984807379273,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5587,
      "step": 2245
    },
    {
      "epoch": 0.781334780249593,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5347,
      "step": 2250
    },
    {
      "epoch": 0.7830710797612588,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.6413,
      "step": 2255
    },
    {
      "epoch": 0.7848073792729245,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5683,
      "step": 2260
    },
    {
      "epoch": 0.7865436787845903,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.6132,
      "step": 2265
    },
    {
      "epoch": 0.7882799782962561,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5839,
      "step": 2270
    },
    {
      "epoch": 0.7900162778079218,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5857,
      "step": 2275
    },
    {
      "epoch": 0.7917525773195876,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5711,
      "step": 2280
    },
    {
      "epoch": 0.7934888768312534,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5748,
      "step": 2285
    },
    {
      "epoch": 0.7952251763429191,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5431,
      "step": 2290
    },
    {
      "epoch": 0.7969614758545849,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.549,
      "step": 2295
    },
    {
      "epoch": 0.7986977753662506,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0001,
      "loss": 0.5471,
      "step": 2300
    },
    {
      "epoch": 0.8004340748779164,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.6297,
      "step": 2305
    },
    {
      "epoch": 0.8021703743895822,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0001,
      "loss": 0.5896,
      "step": 2310
    },
    {
      "epoch": 0.8039066739012479,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5866,
      "step": 2315
    },
    {
      "epoch": 0.8056429734129137,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5779,
      "step": 2320
    },
    {
      "epoch": 0.8073792729245794,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.58,
      "step": 2325
    },
    {
      "epoch": 0.8091155724362452,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5887,
      "step": 2330
    },
    {
      "epoch": 0.810851871947911,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5622,
      "step": 2335
    },
    {
      "epoch": 0.8125881714595767,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5402,
      "step": 2340
    },
    {
      "epoch": 0.8143244709712425,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5459,
      "step": 2345
    },
    {
      "epoch": 0.8160607704829083,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5197,
      "step": 2350
    },
    {
      "epoch": 0.817797069994574,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6183,
      "step": 2355
    },
    {
      "epoch": 0.8195333695062398,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.5875,
      "step": 2360
    },
    {
      "epoch": 0.8212696690179055,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.605,
      "step": 2365
    },
    {
      "epoch": 0.8230059685295713,
      "grad_norm": 0.07421875,
      "learning_rate": 0.0001,
      "loss": 0.583,
      "step": 2370
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5785,
      "step": 2375
    },
    {
      "epoch": 0.8264785675529028,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.5611,
      "step": 2380
    },
    {
      "epoch": 0.8282148670645686,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5796,
      "step": 2385
    },
    {
      "epoch": 0.8299511665762344,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.548,
      "step": 2390
    },
    {
      "epoch": 0.8316874660879001,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.5653,
      "step": 2395
    },
    {
      "epoch": 0.8334237655995659,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5371,
      "step": 2400
    },
    {
      "epoch": 0.8351600651112316,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6198,
      "step": 2405
    },
    {
      "epoch": 0.8368963646228974,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5979,
      "step": 2410
    },
    {
      "epoch": 0.8386326641345632,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5892,
      "step": 2415
    },
    {
      "epoch": 0.8403689636462289,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.5635,
      "step": 2420
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5798,
      "step": 2425
    },
    {
      "epoch": 0.8438415626695605,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5589,
      "step": 2430
    },
    {
      "epoch": 0.8455778621812262,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5365,
      "step": 2435
    },
    {
      "epoch": 0.847314161692892,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5417,
      "step": 2440
    },
    {
      "epoch": 0.8490504612045577,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5503,
      "step": 2445
    },
    {
      "epoch": 0.8507867607162235,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5355,
      "step": 2450
    },
    {
      "epoch": 0.8525230602278893,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.6341,
      "step": 2455
    },
    {
      "epoch": 0.854259359739555,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5729,
      "step": 2460
    },
    {
      "epoch": 0.8559956592512208,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.584,
      "step": 2465
    },
    {
      "epoch": 0.8577319587628865,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.5905,
      "step": 2470
    },
    {
      "epoch": 0.8594682582745523,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5782,
      "step": 2475
    },
    {
      "epoch": 0.8612045577862181,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.5476,
      "step": 2480
    },
    {
      "epoch": 0.8629408572978838,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5952,
      "step": 2485
    },
    {
      "epoch": 0.8646771568095496,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5347,
      "step": 2490
    },
    {
      "epoch": 0.8664134563212154,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.554,
      "step": 2495
    },
    {
      "epoch": 0.8681497558328811,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.54,
      "step": 2500
    },
    {
      "epoch": 0.8698860553445469,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.6187,
      "step": 2505
    },
    {
      "epoch": 0.8716223548562126,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5818,
      "step": 2510
    },
    {
      "epoch": 0.8733586543678784,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0001,
      "loss": 0.585,
      "step": 2515
    },
    {
      "epoch": 0.8750949538795442,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.5996,
      "step": 2520
    },
    {
      "epoch": 0.8768312533912099,
      "grad_norm": 0.07470703125,
      "learning_rate": 0.0001,
      "loss": 0.5727,
      "step": 2525
    },
    {
      "epoch": 0.8785675529028757,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5711,
      "step": 2530
    },
    {
      "epoch": 0.8803038524145415,
      "grad_norm": 0.08056640625,
      "learning_rate": 0.0001,
      "loss": 0.5614,
      "step": 2535
    },
    {
      "epoch": 0.8820401519262072,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5451,
      "step": 2540
    },
    {
      "epoch": 0.883776451437873,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5539,
      "step": 2545
    },
    {
      "epoch": 0.8855127509495387,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0001,
      "loss": 0.5379,
      "step": 2550
    },
    {
      "epoch": 0.8872490504612045,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.6218,
      "step": 2555
    },
    {
      "epoch": 0.8889853499728704,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001,
      "loss": 0.6054,
      "step": 2560
    },
    {
      "epoch": 0.8907216494845361,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5887,
      "step": 2565
    },
    {
      "epoch": 0.8924579489962019,
      "grad_norm": 0.07666015625,
      "learning_rate": 0.0001,
      "loss": 0.5683,
      "step": 2570
    },
    {
      "epoch": 0.8941942485078677,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5753,
      "step": 2575
    },
    {
      "epoch": 0.8959305480195334,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5803,
      "step": 2580
    },
    {
      "epoch": 0.8976668475311992,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.5634,
      "step": 2585
    },
    {
      "epoch": 0.899403147042865,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5674,
      "step": 2590
    },
    {
      "epoch": 0.9011394465545307,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5618,
      "step": 2595
    },
    {
      "epoch": 0.9028757460661965,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5553,
      "step": 2600
    },
    {
      "epoch": 0.9046120455778622,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.6392,
      "step": 2605
    },
    {
      "epoch": 0.906348345089528,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0001,
      "loss": 0.5888,
      "step": 2610
    },
    {
      "epoch": 0.9080846446011938,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5674,
      "step": 2615
    },
    {
      "epoch": 0.9098209441128595,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.5688,
      "step": 2620
    },
    {
      "epoch": 0.9115572436245253,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5894,
      "step": 2625
    },
    {
      "epoch": 0.913293543136191,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.5691,
      "step": 2630
    },
    {
      "epoch": 0.9150298426478568,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.5588,
      "step": 2635
    },
    {
      "epoch": 0.9167661421595226,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.5453,
      "step": 2640
    },
    {
      "epoch": 0.9185024416711883,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5394,
      "step": 2645
    },
    {
      "epoch": 0.9202387411828541,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5313,
      "step": 2650
    },
    {
      "epoch": 0.9219750406945199,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.6438,
      "step": 2655
    },
    {
      "epoch": 0.9237113402061856,
      "grad_norm": 0.0771484375,
      "learning_rate": 0.0001,
      "loss": 0.5677,
      "step": 2660
    },
    {
      "epoch": 0.9254476397178514,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5836,
      "step": 2665
    },
    {
      "epoch": 0.9271839392295171,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0001,
      "loss": 0.5679,
      "step": 2670
    },
    {
      "epoch": 0.9289202387411829,
      "grad_norm": 0.07568359375,
      "learning_rate": 0.0001,
      "loss": 0.5684,
      "step": 2675
    },
    {
      "epoch": 0.9306565382528487,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.54,
      "step": 2680
    },
    {
      "epoch": 0.9323928377645144,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5564,
      "step": 2685
    },
    {
      "epoch": 0.9341291372761802,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5526,
      "step": 2690
    },
    {
      "epoch": 0.935865436787846,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5372,
      "step": 2695
    },
    {
      "epoch": 0.9376017362995117,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5369,
      "step": 2700
    },
    {
      "epoch": 0.9393380358111775,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.6357,
      "step": 2705
    },
    {
      "epoch": 0.9410743353228432,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.5584,
      "step": 2710
    },
    {
      "epoch": 0.942810634834509,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.6022,
      "step": 2715
    },
    {
      "epoch": 0.9445469343461748,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.566,
      "step": 2720
    },
    {
      "epoch": 0.9462832338578405,
      "grad_norm": 0.07958984375,
      "learning_rate": 0.0001,
      "loss": 0.5742,
      "step": 2725
    },
    {
      "epoch": 0.9480195333695063,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5775,
      "step": 2730
    },
    {
      "epoch": 0.949755832881172,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5666,
      "step": 2735
    },
    {
      "epoch": 0.9514921323928378,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0001,
      "loss": 0.5505,
      "step": 2740
    },
    {
      "epoch": 0.9532284319045036,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5276,
      "step": 2745
    },
    {
      "epoch": 0.9549647314161693,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5265,
      "step": 2750
    },
    {
      "epoch": 0.9567010309278351,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.6327,
      "step": 2755
    },
    {
      "epoch": 0.9584373304395009,
      "grad_norm": 0.0732421875,
      "learning_rate": 0.0001,
      "loss": 0.5947,
      "step": 2760
    },
    {
      "epoch": 0.9601736299511666,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5922,
      "step": 2765
    },
    {
      "epoch": 0.9619099294628324,
      "grad_norm": 0.07373046875,
      "learning_rate": 0.0001,
      "loss": 0.585,
      "step": 2770
    },
    {
      "epoch": 0.9636462289744981,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5712,
      "step": 2775
    },
    {
      "epoch": 0.9653825284861639,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5843,
      "step": 2780
    },
    {
      "epoch": 0.9671188279978297,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.551,
      "step": 2785
    },
    {
      "epoch": 0.9688551275094954,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5638,
      "step": 2790
    },
    {
      "epoch": 0.9705914270211612,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5495,
      "step": 2795
    },
    {
      "epoch": 0.972327726532827,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5238,
      "step": 2800
    },
    {
      "epoch": 0.9740640260444927,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.6252,
      "step": 2805
    },
    {
      "epoch": 0.9758003255561585,
      "grad_norm": 0.0712890625,
      "learning_rate": 0.0001,
      "loss": 0.6002,
      "step": 2810
    },
    {
      "epoch": 0.9775366250678242,
      "grad_norm": 0.08154296875,
      "learning_rate": 0.0001,
      "loss": 0.5987,
      "step": 2815
    },
    {
      "epoch": 0.97927292457949,
      "grad_norm": 0.0751953125,
      "learning_rate": 0.0001,
      "loss": 0.5911,
      "step": 2820
    },
    {
      "epoch": 0.9810092240911558,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0001,
      "loss": 0.5721,
      "step": 2825
    },
    {
      "epoch": 0.9827455236028215,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.5606,
      "step": 2830
    },
    {
      "epoch": 0.9844818231144873,
      "grad_norm": 0.2177734375,
      "learning_rate": 0.0001,
      "loss": 0.5618,
      "step": 2835
    },
    {
      "epoch": 0.986218122626153,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5074,
      "step": 2840
    },
    {
      "epoch": 0.9879544221378188,
      "grad_norm": 0.078125,
      "learning_rate": 0.0001,
      "loss": 0.5442,
      "step": 2845
    },
    {
      "epoch": 0.9896907216494846,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.556,
      "step": 2850
    },
    {
      "epoch": 0.9914270211611503,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.6428,
      "step": 2855
    },
    {
      "epoch": 0.9931633206728161,
      "grad_norm": 0.072265625,
      "learning_rate": 0.0001,
      "loss": 0.6115,
      "step": 2860
    },
    {
      "epoch": 0.9948996201844819,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5626,
      "step": 2865
    },
    {
      "epoch": 0.9966359196961476,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5743,
      "step": 2870
    },
    {
      "epoch": 0.9983722192078134,
      "grad_norm": 0.08447265625,
      "learning_rate": 0.0001,
      "loss": 0.5582,
      "step": 2875
    },
    {
      "epoch": 1.0001085187194791,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.5435,
      "step": 2880
    },
    {
      "epoch": 1.0018448182311448,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5697,
      "step": 2885
    },
    {
      "epoch": 1.0035811177428107,
      "grad_norm": 0.07763671875,
      "learning_rate": 0.0001,
      "loss": 0.5622,
      "step": 2890
    },
    {
      "epoch": 1.0053174172544763,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5732,
      "step": 2895
    },
    {
      "epoch": 1.0070537167661422,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5266,
      "step": 2900
    },
    {
      "epoch": 1.0087900162778078,
      "grad_norm": 0.07861328125,
      "learning_rate": 0.0001,
      "loss": 0.5615,
      "step": 2905
    },
    {
      "epoch": 1.0105263157894737,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5519,
      "step": 2910
    },
    {
      "epoch": 1.0122626153011394,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5355,
      "step": 2915
    },
    {
      "epoch": 1.0139989148128052,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5235,
      "step": 2920
    },
    {
      "epoch": 1.015735214324471,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.4988,
      "step": 2925
    },
    {
      "epoch": 1.0174715138361368,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5261,
      "step": 2930
    },
    {
      "epoch": 1.0192078133478024,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.6225,
      "step": 2935
    },
    {
      "epoch": 1.0209441128594683,
      "grad_norm": 0.0791015625,
      "learning_rate": 0.0001,
      "loss": 0.555,
      "step": 2940
    },
    {
      "epoch": 1.022680412371134,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5522,
      "step": 2945
    },
    {
      "epoch": 1.0244167118827998,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5617,
      "step": 2950
    },
    {
      "epoch": 1.0261530113944655,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5503,
      "step": 2955
    },
    {
      "epoch": 1.0278893109061313,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5318,
      "step": 2960
    },
    {
      "epoch": 1.029625610417797,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5166,
      "step": 2965
    },
    {
      "epoch": 1.0313619099294629,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5373,
      "step": 2970
    },
    {
      "epoch": 1.0330982094411285,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.535,
      "step": 2975
    },
    {
      "epoch": 1.0348345089527944,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5341,
      "step": 2980
    },
    {
      "epoch": 1.03657080846446,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.5795,
      "step": 2985
    },
    {
      "epoch": 1.038307107976126,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5764,
      "step": 2990
    },
    {
      "epoch": 1.0400434074877916,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.6015,
      "step": 2995
    },
    {
      "epoch": 1.0417797069994574,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5517,
      "step": 3000
    },
    {
      "epoch": 1.043516006511123,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5322,
      "step": 3005
    },
    {
      "epoch": 1.045252306022789,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5317,
      "step": 3010
    },
    {
      "epoch": 1.0469886055344546,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5288,
      "step": 3015
    },
    {
      "epoch": 1.0487249050461205,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5166,
      "step": 3020
    },
    {
      "epoch": 1.0504612045577861,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5185,
      "step": 3025
    },
    {
      "epoch": 1.052197504069452,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5088,
      "step": 3030
    },
    {
      "epoch": 1.0539338035811177,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.6082,
      "step": 3035
    },
    {
      "epoch": 1.0556701030927835,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5536,
      "step": 3040
    },
    {
      "epoch": 1.0574064026044492,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5774,
      "step": 3045
    },
    {
      "epoch": 1.059142702116115,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5647,
      "step": 3050
    },
    {
      "epoch": 1.0608790016277807,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5461,
      "step": 3055
    },
    {
      "epoch": 1.0626153011394466,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5357,
      "step": 3060
    },
    {
      "epoch": 1.0643516006511122,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5344,
      "step": 3065
    },
    {
      "epoch": 1.066087900162778,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5371,
      "step": 3070
    },
    {
      "epoch": 1.0678241996744438,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5229,
      "step": 3075
    },
    {
      "epoch": 1.0695604991861096,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5132,
      "step": 3080
    },
    {
      "epoch": 1.0712967986977753,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.5932,
      "step": 3085
    },
    {
      "epoch": 1.0730330982094411,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5733,
      "step": 3090
    },
    {
      "epoch": 1.0747693977211068,
      "grad_norm": 0.109375,
      "learning_rate": 0.0001,
      "loss": 0.5709,
      "step": 3095
    },
    {
      "epoch": 1.0765056972327727,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.557,
      "step": 3100
    },
    {
      "epoch": 1.0782419967444383,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5577,
      "step": 3105
    },
    {
      "epoch": 1.0799782962561042,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.521,
      "step": 3110
    },
    {
      "epoch": 1.0817145957677698,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5096,
      "step": 3115
    },
    {
      "epoch": 1.0834508952794357,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5221,
      "step": 3120
    },
    {
      "epoch": 1.0851871947911014,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5144,
      "step": 3125
    },
    {
      "epoch": 1.0869234943027672,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5188,
      "step": 3130
    },
    {
      "epoch": 1.088659793814433,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.6153,
      "step": 3135
    },
    {
      "epoch": 1.0903960933260988,
      "grad_norm": 0.0810546875,
      "learning_rate": 0.0001,
      "loss": 0.5514,
      "step": 3140
    },
    {
      "epoch": 1.0921323928377644,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5448,
      "step": 3145
    },
    {
      "epoch": 1.0938686923494303,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5636,
      "step": 3150
    },
    {
      "epoch": 1.095604991861096,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5605,
      "step": 3155
    },
    {
      "epoch": 1.0973412913727618,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5305,
      "step": 3160
    },
    {
      "epoch": 1.0990775908844275,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5507,
      "step": 3165
    },
    {
      "epoch": 1.1008138903960933,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.4921,
      "step": 3170
    },
    {
      "epoch": 1.102550189907759,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0001,
      "loss": 0.5324,
      "step": 3175
    },
    {
      "epoch": 1.1042864894194249,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5052,
      "step": 3180
    },
    {
      "epoch": 1.1060227889310905,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5938,
      "step": 3185
    },
    {
      "epoch": 1.1077590884427564,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5594,
      "step": 3190
    },
    {
      "epoch": 1.109495387954422,
      "grad_norm": 0.08251953125,
      "learning_rate": 0.0001,
      "loss": 0.5513,
      "step": 3195
    },
    {
      "epoch": 1.111231687466088,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5478,
      "step": 3200
    },
    {
      "epoch": 1.1129679869777536,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5314,
      "step": 3205
    },
    {
      "epoch": 1.1147042864894194,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5351,
      "step": 3210
    },
    {
      "epoch": 1.116440586001085,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5434,
      "step": 3215
    },
    {
      "epoch": 1.118176885512751,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5408,
      "step": 3220
    },
    {
      "epoch": 1.1199131850244166,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5003,
      "step": 3225
    },
    {
      "epoch": 1.1216494845360825,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0001,
      "loss": 0.5179,
      "step": 3230
    },
    {
      "epoch": 1.1233857840477481,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5892,
      "step": 3235
    },
    {
      "epoch": 1.125122083559414,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.5529,
      "step": 3240
    },
    {
      "epoch": 1.1268583830710797,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0001,
      "loss": 0.5669,
      "step": 3245
    },
    {
      "epoch": 1.1285946825827455,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.565,
      "step": 3250
    },
    {
      "epoch": 1.1303309820944114,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5467,
      "step": 3255
    },
    {
      "epoch": 1.132067281606077,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5411,
      "step": 3260
    },
    {
      "epoch": 1.1338035811177427,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5108,
      "step": 3265
    },
    {
      "epoch": 1.1355398806294086,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0001,
      "loss": 0.5241,
      "step": 3270
    },
    {
      "epoch": 1.1372761801410745,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5386,
      "step": 3275
    },
    {
      "epoch": 1.13901247965274,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5035,
      "step": 3280
    },
    {
      "epoch": 1.1407487791644058,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.5958,
      "step": 3285
    },
    {
      "epoch": 1.1424850786760716,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5763,
      "step": 3290
    },
    {
      "epoch": 1.1442213781877375,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5484,
      "step": 3295
    },
    {
      "epoch": 1.1459576776994032,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5687,
      "step": 3300
    },
    {
      "epoch": 1.1476939772110688,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.55,
      "step": 3305
    },
    {
      "epoch": 1.1494302767227347,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5231,
      "step": 3310
    },
    {
      "epoch": 1.1511665762344006,
      "grad_norm": 0.10302734375,
      "learning_rate": 0.0001,
      "loss": 0.5541,
      "step": 3315
    },
    {
      "epoch": 1.1529028757460662,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.54,
      "step": 3320
    },
    {
      "epoch": 1.1546391752577319,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5213,
      "step": 3325
    },
    {
      "epoch": 1.1563754747693977,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5192,
      "step": 3330
    },
    {
      "epoch": 1.1581117742810636,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5946,
      "step": 3335
    },
    {
      "epoch": 1.1598480737927293,
      "grad_norm": 0.08203125,
      "learning_rate": 0.0001,
      "loss": 0.5822,
      "step": 3340
    },
    {
      "epoch": 1.161584373304395,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0001,
      "loss": 0.5655,
      "step": 3345
    },
    {
      "epoch": 1.1633206728160608,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5687,
      "step": 3350
    },
    {
      "epoch": 1.1650569723277266,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5448,
      "step": 3355
    },
    {
      "epoch": 1.1667932718393923,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5355,
      "step": 3360
    },
    {
      "epoch": 1.168529571351058,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5321,
      "step": 3365
    },
    {
      "epoch": 1.1702658708627238,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5252,
      "step": 3370
    },
    {
      "epoch": 1.1720021703743897,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.5082,
      "step": 3375
    },
    {
      "epoch": 1.1737384698860553,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0001,
      "loss": 0.4836,
      "step": 3380
    },
    {
      "epoch": 1.175474769397721,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.6026,
      "step": 3385
    },
    {
      "epoch": 1.1772110689093869,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5656,
      "step": 3390
    },
    {
      "epoch": 1.1789473684210527,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.563,
      "step": 3395
    },
    {
      "epoch": 1.1806836679327184,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.553,
      "step": 3400
    },
    {
      "epoch": 1.182419967444384,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5587,
      "step": 3405
    },
    {
      "epoch": 1.18415626695605,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5344,
      "step": 3410
    },
    {
      "epoch": 1.1858925664677158,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5537,
      "step": 3415
    },
    {
      "epoch": 1.1876288659793814,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.5234,
      "step": 3420
    },
    {
      "epoch": 1.189365165491047,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0001,
      "loss": 0.5108,
      "step": 3425
    },
    {
      "epoch": 1.191101465002713,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5186,
      "step": 3430
    },
    {
      "epoch": 1.1928377645143788,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.619,
      "step": 3435
    },
    {
      "epoch": 1.1945740640260445,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.568,
      "step": 3440
    },
    {
      "epoch": 1.1963103635377101,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5474,
      "step": 3445
    },
    {
      "epoch": 1.198046663049376,
      "grad_norm": 0.080078125,
      "learning_rate": 0.0001,
      "loss": 0.5505,
      "step": 3450
    },
    {
      "epoch": 1.1997829625610419,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.5279,
      "step": 3455
    },
    {
      "epoch": 1.2015192620727075,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5384,
      "step": 3460
    },
    {
      "epoch": 1.2032555615843732,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5143,
      "step": 3465
    },
    {
      "epoch": 1.204991861096039,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5401,
      "step": 3470
    },
    {
      "epoch": 1.206728160607705,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0001,
      "loss": 0.526,
      "step": 3475
    },
    {
      "epoch": 1.2084644601193706,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.5094,
      "step": 3480
    },
    {
      "epoch": 1.2102007596310362,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.6154,
      "step": 3485
    },
    {
      "epoch": 1.2119370591427021,
      "grad_norm": 0.0830078125,
      "learning_rate": 0.0001,
      "loss": 0.566,
      "step": 3490
    },
    {
      "epoch": 1.213673358654368,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5844,
      "step": 3495
    },
    {
      "epoch": 1.2154096581660336,
      "grad_norm": 0.10498046875,
      "learning_rate": 0.0001,
      "loss": 0.5497,
      "step": 3500
    },
    {
      "epoch": 1.2171459576776993,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.5543,
      "step": 3505
    },
    {
      "epoch": 1.2188822571893652,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.543,
      "step": 3510
    },
    {
      "epoch": 1.220618556701031,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5354,
      "step": 3515
    },
    {
      "epoch": 1.2223548562126967,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.4978,
      "step": 3520
    },
    {
      "epoch": 1.2240911557243623,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0001,
      "loss": 0.5197,
      "step": 3525
    },
    {
      "epoch": 1.2258274552360282,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5303,
      "step": 3530
    },
    {
      "epoch": 1.227563754747694,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.599,
      "step": 3535
    },
    {
      "epoch": 1.2293000542593597,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5683,
      "step": 3540
    },
    {
      "epoch": 1.2310363537710254,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.5718,
      "step": 3545
    },
    {
      "epoch": 1.2327726532826913,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5621,
      "step": 3550
    },
    {
      "epoch": 1.2345089527943571,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5643,
      "step": 3555
    },
    {
      "epoch": 1.2362452523060228,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.545,
      "step": 3560
    },
    {
      "epoch": 1.2379815518176884,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5316,
      "step": 3565
    },
    {
      "epoch": 1.2397178513293543,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0001,
      "loss": 0.5377,
      "step": 3570
    },
    {
      "epoch": 1.2414541508410202,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.5073,
      "step": 3575
    },
    {
      "epoch": 1.2431904503526858,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5154,
      "step": 3580
    },
    {
      "epoch": 1.2449267498643515,
      "grad_norm": 0.111328125,
      "learning_rate": 0.0001,
      "loss": 0.605,
      "step": 3585
    },
    {
      "epoch": 1.2466630493760174,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5731,
      "step": 3590
    },
    {
      "epoch": 1.2483993488876832,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5519,
      "step": 3595
    },
    {
      "epoch": 1.2501356483993489,
      "grad_norm": 0.08349609375,
      "learning_rate": 0.0001,
      "loss": 0.5507,
      "step": 3600
    },
    {
      "epoch": 1.2518719479110145,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.5362,
      "step": 3605
    },
    {
      "epoch": 1.2536082474226804,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5723,
      "step": 3610
    },
    {
      "epoch": 1.2553445469343463,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.538,
      "step": 3615
    },
    {
      "epoch": 1.257080846446012,
      "grad_norm": 0.0869140625,
      "learning_rate": 0.0001,
      "loss": 0.5143,
      "step": 3620
    },
    {
      "epoch": 1.2588171459576776,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0001,
      "loss": 0.5278,
      "step": 3625
    },
    {
      "epoch": 1.2605534454693434,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0001,
      "loss": 0.5244,
      "step": 3630
    },
    {
      "epoch": 1.2622897449810093,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.5943,
      "step": 3635
    },
    {
      "epoch": 1.264026044492675,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5651,
      "step": 3640
    },
    {
      "epoch": 1.2657623440043406,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5607,
      "step": 3645
    },
    {
      "epoch": 1.2674986435160065,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5403,
      "step": 3650
    },
    {
      "epoch": 1.2692349430276724,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5385,
      "step": 3655
    },
    {
      "epoch": 1.270971242539338,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.528,
      "step": 3660
    },
    {
      "epoch": 1.2727075420510037,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5318,
      "step": 3665
    },
    {
      "epoch": 1.2744438415626695,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5335,
      "step": 3670
    },
    {
      "epoch": 1.2761801410743354,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.5265,
      "step": 3675
    },
    {
      "epoch": 1.277916440586001,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0001,
      "loss": 0.4965,
      "step": 3680
    },
    {
      "epoch": 1.2796527400976667,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.583,
      "step": 3685
    },
    {
      "epoch": 1.2813890396093326,
      "grad_norm": 0.08544921875,
      "learning_rate": 0.0001,
      "loss": 0.5798,
      "step": 3690
    },
    {
      "epoch": 1.2831253391209985,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5799,
      "step": 3695
    },
    {
      "epoch": 1.2848616386326641,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5753,
      "step": 3700
    },
    {
      "epoch": 1.2865979381443298,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.555,
      "step": 3705
    },
    {
      "epoch": 1.2883342376559956,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5462,
      "step": 3710
    },
    {
      "epoch": 1.2900705371676615,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5345,
      "step": 3715
    },
    {
      "epoch": 1.2918068366793272,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5211,
      "step": 3720
    },
    {
      "epoch": 1.2935431361909928,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5173,
      "step": 3725
    },
    {
      "epoch": 1.2952794357026587,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.508,
      "step": 3730
    },
    {
      "epoch": 1.2970157352143246,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.6064,
      "step": 3735
    },
    {
      "epoch": 1.2987520347259902,
      "grad_norm": 0.083984375,
      "learning_rate": 0.0001,
      "loss": 0.5432,
      "step": 3740
    },
    {
      "epoch": 1.3004883342376559,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5654,
      "step": 3745
    },
    {
      "epoch": 1.3022246337493217,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.5506,
      "step": 3750
    },
    {
      "epoch": 1.3039609332609876,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5412,
      "step": 3755
    },
    {
      "epoch": 1.3056972327726533,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5289,
      "step": 3760
    },
    {
      "epoch": 1.307433532284319,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5344,
      "step": 3765
    },
    {
      "epoch": 1.3091698317959848,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5394,
      "step": 3770
    },
    {
      "epoch": 1.3109061313076507,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.516,
      "step": 3775
    },
    {
      "epoch": 1.3126424308193163,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5243,
      "step": 3780
    },
    {
      "epoch": 1.314378730330982,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.6128,
      "step": 3785
    },
    {
      "epoch": 1.3161150298426478,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5888,
      "step": 3790
    },
    {
      "epoch": 1.3178513293543137,
      "grad_norm": 0.08642578125,
      "learning_rate": 0.0001,
      "loss": 0.5755,
      "step": 3795
    },
    {
      "epoch": 1.3195876288659794,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5551,
      "step": 3800
    },
    {
      "epoch": 1.321323928377645,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.0001,
      "loss": 0.5296,
      "step": 3805
    },
    {
      "epoch": 1.3230602278893109,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5562,
      "step": 3810
    },
    {
      "epoch": 1.3247965274009768,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5339,
      "step": 3815
    },
    {
      "epoch": 1.3265328269126424,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.526,
      "step": 3820
    },
    {
      "epoch": 1.328269126424308,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5213,
      "step": 3825
    },
    {
      "epoch": 1.330005425935974,
      "grad_norm": 0.1064453125,
      "learning_rate": 0.0001,
      "loss": 0.4972,
      "step": 3830
    },
    {
      "epoch": 1.3317417254476398,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5998,
      "step": 3835
    },
    {
      "epoch": 1.3334780249593055,
      "grad_norm": 0.08740234375,
      "learning_rate": 0.0001,
      "loss": 0.5902,
      "step": 3840
    },
    {
      "epoch": 1.3352143244709713,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.585,
      "step": 3845
    },
    {
      "epoch": 1.336950623982637,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0001,
      "loss": 0.5746,
      "step": 3850
    },
    {
      "epoch": 1.3386869234943028,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.5543,
      "step": 3855
    },
    {
      "epoch": 1.3404232230059685,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5311,
      "step": 3860
    },
    {
      "epoch": 1.3421595225176344,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5246,
      "step": 3865
    },
    {
      "epoch": 1.3438958220293,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5152,
      "step": 3870
    },
    {
      "epoch": 1.345632121540966,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0001,
      "loss": 0.5267,
      "step": 3875
    },
    {
      "epoch": 1.3473684210526315,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0001,
      "loss": 0.5042,
      "step": 3880
    },
    {
      "epoch": 1.3491047205642974,
      "grad_norm": 0.1015625,
      "learning_rate": 0.0001,
      "loss": 0.6091,
      "step": 3885
    },
    {
      "epoch": 1.350841020075963,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5629,
      "step": 3890
    },
    {
      "epoch": 1.352577319587629,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5481,
      "step": 3895
    },
    {
      "epoch": 1.3543136190992946,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5349,
      "step": 3900
    },
    {
      "epoch": 1.3560499186109605,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5352,
      "step": 3905
    },
    {
      "epoch": 1.3577862181226261,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5529,
      "step": 3910
    },
    {
      "epoch": 1.359522517634292,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.55,
      "step": 3915
    },
    {
      "epoch": 1.3612588171459576,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5221,
      "step": 3920
    },
    {
      "epoch": 1.3629951166576235,
      "grad_norm": 0.291015625,
      "learning_rate": 0.0001,
      "loss": 0.5193,
      "step": 3925
    },
    {
      "epoch": 1.3647314161692892,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0001,
      "loss": 0.5119,
      "step": 3930
    },
    {
      "epoch": 1.366467715680955,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5839,
      "step": 3935
    },
    {
      "epoch": 1.3682040151926207,
      "grad_norm": 0.087890625,
      "learning_rate": 0.0001,
      "loss": 0.5602,
      "step": 3940
    },
    {
      "epoch": 1.3699403147042866,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5468,
      "step": 3945
    },
    {
      "epoch": 1.3716766142159522,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0001,
      "loss": 0.548,
      "step": 3950
    },
    {
      "epoch": 1.373412913727618,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5601,
      "step": 3955
    },
    {
      "epoch": 1.3751492132392837,
      "grad_norm": 0.09130859375,
      "learning_rate": 0.0001,
      "loss": 0.5371,
      "step": 3960
    },
    {
      "epoch": 1.3768855127509496,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5431,
      "step": 3965
    },
    {
      "epoch": 1.3786218122626153,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5277,
      "step": 3970
    },
    {
      "epoch": 1.3803581117742811,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.5221,
      "step": 3975
    },
    {
      "epoch": 1.3820944112859468,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0001,
      "loss": 0.5064,
      "step": 3980
    },
    {
      "epoch": 1.3838307107976127,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.6115,
      "step": 3985
    },
    {
      "epoch": 1.3855670103092783,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.573,
      "step": 3990
    },
    {
      "epoch": 1.3873033098209442,
      "grad_norm": 0.0849609375,
      "learning_rate": 0.0001,
      "loss": 0.585,
      "step": 3995
    },
    {
      "epoch": 1.3890396093326098,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5566,
      "step": 4000
    },
    {
      "epoch": 1.3907759088442757,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5365,
      "step": 4005
    },
    {
      "epoch": 1.3925122083559414,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5477,
      "step": 4010
    },
    {
      "epoch": 1.3942485078676072,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5232,
      "step": 4015
    },
    {
      "epoch": 1.3959848073792729,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5557,
      "step": 4020
    },
    {
      "epoch": 1.3977211068909388,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5138,
      "step": 4025
    },
    {
      "epoch": 1.3994574064026044,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.5254,
      "step": 4030
    },
    {
      "epoch": 1.4011937059142703,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.604,
      "step": 4035
    },
    {
      "epoch": 1.402930005425936,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5583,
      "step": 4040
    },
    {
      "epoch": 1.4046663049376018,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0001,
      "loss": 0.5735,
      "step": 4045
    },
    {
      "epoch": 1.4064026044492675,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5454,
      "step": 4050
    },
    {
      "epoch": 1.4081389039609333,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5569,
      "step": 4055
    },
    {
      "epoch": 1.409875203472599,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5459,
      "step": 4060
    },
    {
      "epoch": 1.4116115029842649,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5285,
      "step": 4065
    },
    {
      "epoch": 1.4133478024959305,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0001,
      "loss": 0.5373,
      "step": 4070
    },
    {
      "epoch": 1.4150841020075964,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5274,
      "step": 4075
    },
    {
      "epoch": 1.416820401519262,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.4974,
      "step": 4080
    },
    {
      "epoch": 1.418556701030928,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0001,
      "loss": 0.5757,
      "step": 4085
    },
    {
      "epoch": 1.4202930005425936,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5569,
      "step": 4090
    },
    {
      "epoch": 1.4220293000542594,
      "grad_norm": 0.0947265625,
      "learning_rate": 0.0001,
      "loss": 0.5585,
      "step": 4095
    },
    {
      "epoch": 1.423765599565925,
      "grad_norm": 0.4375,
      "learning_rate": 0.0001,
      "loss": 0.5621,
      "step": 4100
    },
    {
      "epoch": 1.425501899077591,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.5391,
      "step": 4105
    },
    {
      "epoch": 1.4272381985892566,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5445,
      "step": 4110
    },
    {
      "epoch": 1.4289744981009225,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5327,
      "step": 4115
    },
    {
      "epoch": 1.4307107976125881,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0001,
      "loss": 0.5027,
      "step": 4120
    },
    {
      "epoch": 1.432447097124254,
      "grad_norm": 0.10107421875,
      "learning_rate": 0.0001,
      "loss": 0.5147,
      "step": 4125
    },
    {
      "epoch": 1.4341833966359196,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0001,
      "loss": 0.5243,
      "step": 4130
    },
    {
      "epoch": 1.4359196961475855,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5844,
      "step": 4135
    },
    {
      "epoch": 1.4376559956592512,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5563,
      "step": 4140
    },
    {
      "epoch": 1.439392295170917,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5539,
      "step": 4145
    },
    {
      "epoch": 1.4411285946825827,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.539,
      "step": 4150
    },
    {
      "epoch": 1.4428648941942486,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.539,
      "step": 4155
    },
    {
      "epoch": 1.4446011937059142,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5496,
      "step": 4160
    },
    {
      "epoch": 1.44633749321758,
      "grad_norm": 0.099609375,
      "learning_rate": 0.0001,
      "loss": 0.5393,
      "step": 4165
    },
    {
      "epoch": 1.4480737927292457,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0001,
      "loss": 0.5233,
      "step": 4170
    },
    {
      "epoch": 1.4498100922409116,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5267,
      "step": 4175
    },
    {
      "epoch": 1.4515463917525773,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0001,
      "loss": 0.529,
      "step": 4180
    },
    {
      "epoch": 1.4532826912642431,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.6023,
      "step": 4185
    },
    {
      "epoch": 1.4550189907759088,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5811,
      "step": 4190
    },
    {
      "epoch": 1.4567552902875747,
      "grad_norm": 0.09814453125,
      "learning_rate": 0.0001,
      "loss": 0.5586,
      "step": 4195
    },
    {
      "epoch": 1.4584915897992403,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5704,
      "step": 4200
    },
    {
      "epoch": 1.4602278893109062,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5394,
      "step": 4205
    },
    {
      "epoch": 1.4619641888225718,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.5105,
      "step": 4210
    },
    {
      "epoch": 1.4637004883342377,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5327,
      "step": 4215
    },
    {
      "epoch": 1.4654367878459034,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5387,
      "step": 4220
    },
    {
      "epoch": 1.4671730873575692,
      "grad_norm": 0.09326171875,
      "learning_rate": 0.0001,
      "loss": 0.518,
      "step": 4225
    },
    {
      "epoch": 1.468909386869235,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5103,
      "step": 4230
    },
    {
      "epoch": 1.4706456863809008,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.588,
      "step": 4235
    },
    {
      "epoch": 1.4723819858925664,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0001,
      "loss": 0.5676,
      "step": 4240
    },
    {
      "epoch": 1.4741182854042323,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5664,
      "step": 4245
    },
    {
      "epoch": 1.475854584915898,
      "grad_norm": 0.0927734375,
      "learning_rate": 0.0001,
      "loss": 0.5609,
      "step": 4250
    },
    {
      "epoch": 1.4775908844275638,
      "grad_norm": 0.09716796875,
      "learning_rate": 0.0001,
      "loss": 0.5587,
      "step": 4255
    },
    {
      "epoch": 1.4793271839392295,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5505,
      "step": 4260
    },
    {
      "epoch": 1.4810634834508953,
      "grad_norm": 0.0966796875,
      "learning_rate": 0.0001,
      "loss": 0.5338,
      "step": 4265
    },
    {
      "epoch": 1.482799782962561,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.5416,
      "step": 4270
    },
    {
      "epoch": 1.4845360824742269,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5411,
      "step": 4275
    },
    {
      "epoch": 1.4862723819858925,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0001,
      "loss": 0.5019,
      "step": 4280
    },
    {
      "epoch": 1.4880086814975584,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0001,
      "loss": 0.5715,
      "step": 4285
    },
    {
      "epoch": 1.489744981009224,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5747,
      "step": 4290
    },
    {
      "epoch": 1.49148128052089,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5858,
      "step": 4295
    },
    {
      "epoch": 1.4932175800325556,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5528,
      "step": 4300
    },
    {
      "epoch": 1.4949538795442214,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5545,
      "step": 4305
    },
    {
      "epoch": 1.496690179055887,
      "grad_norm": 0.09033203125,
      "learning_rate": 0.0001,
      "loss": 0.5409,
      "step": 4310
    },
    {
      "epoch": 1.498426478567553,
      "grad_norm": 0.103515625,
      "learning_rate": 0.0001,
      "loss": 0.5582,
      "step": 4315
    },
    {
      "epoch": 1.5001627780792188,
      "grad_norm": 0.1044921875,
      "learning_rate": 0.0001,
      "loss": 0.5204,
      "step": 4320
    },
    {
      "epoch": 1.5018990775908845,
      "grad_norm": 0.1083984375,
      "learning_rate": 0.0001,
      "loss": 0.5273,
      "step": 4325
    },
    {
      "epoch": 1.5036353771025501,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0001,
      "loss": 0.5047,
      "step": 4330
    },
    {
      "epoch": 1.505371676614216,
      "grad_norm": 0.10009765625,
      "learning_rate": 0.0001,
      "loss": 0.5932,
      "step": 4335
    },
    {
      "epoch": 1.5071079761258819,
      "grad_norm": 0.08837890625,
      "learning_rate": 0.0001,
      "loss": 0.5626,
      "step": 4340
    },
    {
      "epoch": 1.5088442756375475,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5594,
      "step": 4345
    },
    {
      "epoch": 1.5105805751492132,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0001,
      "loss": 0.5529,
      "step": 4350
    },
    {
      "epoch": 1.512316874660879,
      "grad_norm": 0.095703125,
      "learning_rate": 0.0001,
      "loss": 0.5572,
      "step": 4355
    },
    {
      "epoch": 1.514053174172545,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5454,
      "step": 4360
    },
    {
      "epoch": 1.5157894736842106,
      "grad_norm": 0.09619140625,
      "learning_rate": 0.0001,
      "loss": 0.5368,
      "step": 4365
    },
    {
      "epoch": 1.5175257731958762,
      "grad_norm": 0.10205078125,
      "learning_rate": 0.0001,
      "loss": 0.5221,
      "step": 4370
    },
    {
      "epoch": 1.519262072707542,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.0001,
      "loss": 0.5348,
      "step": 4375
    },
    {
      "epoch": 1.520998372219208,
      "grad_norm": 0.10693359375,
      "learning_rate": 0.0001,
      "loss": 0.5132,
      "step": 4380
    },
    {
      "epoch": 1.5227346717308736,
      "grad_norm": 0.09765625,
      "learning_rate": 0.0001,
      "loss": 0.5832,
      "step": 4385
    },
    {
      "epoch": 1.5244709712425393,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5648,
      "step": 4390
    },
    {
      "epoch": 1.5262072707542051,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.568,
      "step": 4395
    },
    {
      "epoch": 1.527943570265871,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.539,
      "step": 4400
    },
    {
      "epoch": 1.5296798697775367,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5463,
      "step": 4405
    },
    {
      "epoch": 1.5314161692892023,
      "grad_norm": 0.0908203125,
      "learning_rate": 0.0001,
      "loss": 0.5238,
      "step": 4410
    },
    {
      "epoch": 1.5331524688008682,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0001,
      "loss": 0.5254,
      "step": 4415
    },
    {
      "epoch": 1.534888768312534,
      "grad_norm": 0.09912109375,
      "learning_rate": 0.0001,
      "loss": 0.535,
      "step": 4420
    },
    {
      "epoch": 1.5366250678241997,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0001,
      "loss": 0.5234,
      "step": 4425
    },
    {
      "epoch": 1.5383613673358654,
      "grad_norm": 0.1025390625,
      "learning_rate": 0.0001,
      "loss": 0.4957,
      "step": 4430
    },
    {
      "epoch": 1.5400976668475312,
      "grad_norm": 0.09375,
      "learning_rate": 0.0001,
      "loss": 0.5816,
      "step": 4435
    },
    {
      "epoch": 1.5418339663591971,
      "grad_norm": 0.0859375,
      "learning_rate": 0.0001,
      "loss": 0.5775,
      "step": 4440
    },
    {
      "epoch": 1.5435702658708628,
      "grad_norm": 0.09228515625,
      "learning_rate": 0.0001,
      "loss": 0.5528,
      "step": 4445
    },
    {
      "epoch": 1.5453065653825284,
      "grad_norm": 0.091796875,
      "learning_rate": 0.0001,
      "loss": 0.5616,
      "step": 4450
    },
    {
      "epoch": 1.5470428648941943,
      "grad_norm": 0.0888671875,
      "learning_rate": 0.0001,
      "loss": 0.5281,
      "step": 4455
    },
    {
      "epoch": 1.5487791644058602,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5411,
      "step": 4460
    },
    {
      "epoch": 1.5505154639175258,
      "grad_norm": 0.09423828125,
      "learning_rate": 0.0001,
      "loss": 0.4934,
      "step": 4465
    },
    {
      "epoch": 1.5522517634291915,
      "grad_norm": 0.0986328125,
      "learning_rate": 0.0001,
      "loss": 0.5168,
      "step": 4470
    },
    {
      "epoch": 1.5539880629408573,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.0001,
      "loss": 0.5091,
      "step": 4475
    },
    {
      "epoch": 1.5557243624525232,
      "grad_norm": 0.10400390625,
      "learning_rate": 0.0001,
      "loss": 0.5026,
      "step": 4480
    },
    {
      "epoch": 1.5574606619641889,
      "grad_norm": 0.1005859375,
      "learning_rate": 0.0001,
      "loss": 0.6009,
      "step": 4485
    },
    {
      "epoch": 1.5591969614758545,
      "grad_norm": 0.08935546875,
      "learning_rate": 0.0001,
      "loss": 0.5593,
      "step": 4490
    },
    {
      "epoch": 1.5609332609875204,
      "grad_norm": 0.08984375,
      "learning_rate": 0.0001,
      "loss": 0.5694,
      "step": 4495
    },
    {
      "epoch": 1.5626695604991863,
      "grad_norm": 0.09521484375,
      "learning_rate": 0.0001,
      "loss": 0.5424,
      "step": 4500
    },
    {
      "epoch": 1.5626695604991863,
      "step": 4500,
      "total_flos": 4.510419270260736e+18,
      "train_loss": 0.5749385200606452,
      "train_runtime": 199956.6124,
      "train_samples_per_second": 1.44,
      "train_steps_per_second": 0.023
    }
  ],
  "logging_steps": 5,
  "max_steps": 4500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 90,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.510419270260736e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}