{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 8969,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 6.601358890533447,
      "learning_rate": 0.001,
      "loss": 10.5249,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.328374981880188,
      "learning_rate": 0.001,
      "loss": 8.6782,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.0325572490692139,
      "learning_rate": 0.001,
      "loss": 7.3989,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.6852174997329712,
      "learning_rate": 0.001,
      "loss": 7.3653,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.656384289264679,
      "learning_rate": 0.001,
      "loss": 7.3625,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.7547395825386047,
      "learning_rate": 0.001,
      "loss": 7.3089,
      "step": 25
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.5359553694725037,
      "learning_rate": 0.001,
      "loss": 7.3326,
      "step": 30
    },
    {
      "epoch": 0.0,
      "grad_norm": 0.654456615447998,
      "learning_rate": 0.001,
      "loss": 7.279,
      "step": 35
    },
    {
      "epoch": 0.0,
      "grad_norm": 8.3489408493042,
      "learning_rate": 0.001,
      "loss": 7.1401,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7769318222999573,
      "learning_rate": 0.001,
      "loss": 7.0276,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.961713969707489,
      "learning_rate": 0.001,
      "loss": 6.8993,
      "step": 50
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5359512567520142,
      "learning_rate": 0.001,
      "loss": 6.7672,
      "step": 55
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7930881381034851,
      "learning_rate": 0.001,
      "loss": 6.6585,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6703987717628479,
      "learning_rate": 0.001,
      "loss": 6.6558,
      "step": 65
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6621621251106262,
      "learning_rate": 0.001,
      "loss": 6.3967,
      "step": 70
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8039225935935974,
      "learning_rate": 0.001,
      "loss": 6.3318,
      "step": 75
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5961666703224182,
      "learning_rate": 0.001,
      "loss": 6.3381,
      "step": 80
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5145270228385925,
      "learning_rate": 0.001,
      "loss": 6.1274,
      "step": 85
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7134149074554443,
      "learning_rate": 0.001,
      "loss": 6.1067,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5910749435424805,
      "learning_rate": 0.001,
      "loss": 6.0012,
      "step": 95
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.6277782917022705,
      "learning_rate": 0.001,
      "loss": 5.9658,
      "step": 100
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.44600120186805725,
      "learning_rate": 0.001,
      "loss": 5.9172,
      "step": 105
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.4796374440193176,
      "learning_rate": 0.001,
      "loss": 5.7681,
      "step": 110
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5527931451797485,
      "learning_rate": 0.001,
      "loss": 5.7746,
      "step": 115
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.7416689991950989,
      "learning_rate": 0.001,
      "loss": 5.6928,
      "step": 120
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5323288440704346,
      "learning_rate": 0.001,
      "loss": 5.6177,
      "step": 125
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.5754017233848572,
      "learning_rate": 0.001,
      "loss": 5.6344,
      "step": 130
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.690061092376709,
      "learning_rate": 0.001,
      "loss": 5.568,
      "step": 135
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6237902045249939,
      "learning_rate": 0.001,
      "loss": 5.4123,
      "step": 140
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.702131450176239,
      "learning_rate": 0.001,
      "loss": 5.4907,
      "step": 145
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6115543842315674,
      "learning_rate": 0.001,
      "loss": 5.4457,
      "step": 150
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5676571726799011,
      "learning_rate": 0.001,
      "loss": 5.3547,
      "step": 155
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6142025589942932,
      "learning_rate": 0.001,
      "loss": 5.3421,
      "step": 160
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5987421274185181,
      "learning_rate": 0.001,
      "loss": 5.3941,
      "step": 165
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6259231567382812,
      "learning_rate": 0.001,
      "loss": 5.231,
      "step": 170
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6100727319717407,
      "learning_rate": 0.001,
      "loss": 5.1879,
      "step": 175
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.064252495765686,
      "learning_rate": 0.001,
      "loss": 5.1959,
      "step": 180
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7423781156539917,
      "learning_rate": 0.001,
      "loss": 5.1457,
      "step": 185
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.788051426410675,
      "learning_rate": 0.001,
      "loss": 5.2344,
      "step": 190
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6291590332984924,
      "learning_rate": 0.001,
      "loss": 5.1488,
      "step": 195
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.6599926948547363,
      "learning_rate": 0.001,
      "loss": 5.0759,
      "step": 200
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5811977386474609,
      "learning_rate": 0.001,
      "loss": 5.0278,
      "step": 205
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.5062569379806519,
      "learning_rate": 0.001,
      "loss": 5.0043,
      "step": 210
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.47193270921707153,
      "learning_rate": 0.001,
      "loss": 4.8949,
      "step": 215
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.7771720290184021,
      "learning_rate": 0.001,
      "loss": 4.951,
      "step": 220
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5288494229316711,
      "learning_rate": 0.001,
      "loss": 4.9402,
      "step": 225
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5217804312705994,
      "learning_rate": 0.001,
      "loss": 4.8737,
      "step": 230
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.594836413860321,
      "learning_rate": 0.001,
      "loss": 4.8091,
      "step": 235
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6213844418525696,
      "learning_rate": 0.001,
      "loss": 4.83,
      "step": 240
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.595021665096283,
      "learning_rate": 0.001,
      "loss": 4.8749,
      "step": 245
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5776046514511108,
      "learning_rate": 0.001,
      "loss": 4.7623,
      "step": 250
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5899957418441772,
      "learning_rate": 0.001,
      "loss": 4.7488,
      "step": 255
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.598554253578186,
      "learning_rate": 0.001,
      "loss": 4.7859,
      "step": 260
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6169301271438599,
      "learning_rate": 0.001,
      "loss": 4.7067,
      "step": 265
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7676901817321777,
      "learning_rate": 0.001,
      "loss": 4.7561,
      "step": 270
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5868487358093262,
      "learning_rate": 0.001,
      "loss": 4.708,
      "step": 275
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.4976160526275635,
      "learning_rate": 0.001,
      "loss": 4.6998,
      "step": 280
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.5182490348815918,
      "learning_rate": 0.001,
      "loss": 4.5554,
      "step": 285
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6468132138252258,
      "learning_rate": 0.001,
      "loss": 4.6289,
      "step": 290
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7456973195075989,
      "learning_rate": 0.001,
      "loss": 4.5579,
      "step": 295
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6524637937545776,
      "learning_rate": 0.001,
      "loss": 4.5488,
      "step": 300
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.6689502596855164,
      "learning_rate": 0.001,
      "loss": 4.5551,
      "step": 305
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7009299397468567,
      "learning_rate": 0.001,
      "loss": 4.5232,
      "step": 310
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5543584227561951,
      "learning_rate": 0.001,
      "loss": 4.5442,
      "step": 315
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5917452573776245,
      "learning_rate": 0.001,
      "loss": 4.5228,
      "step": 320
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5315392017364502,
      "learning_rate": 0.001,
      "loss": 4.5275,
      "step": 325
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.603313684463501,
      "learning_rate": 0.001,
      "loss": 4.404,
      "step": 330
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.6234457492828369,
      "learning_rate": 0.001,
      "loss": 4.4923,
      "step": 335
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5459450483322144,
      "learning_rate": 0.001,
      "loss": 4.4246,
      "step": 340
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5857668519020081,
      "learning_rate": 0.001,
      "loss": 4.4323,
      "step": 345
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.723681628704071,
      "learning_rate": 0.001,
      "loss": 4.4539,
      "step": 350
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5971282124519348,
      "learning_rate": 0.001,
      "loss": 4.4693,
      "step": 355
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.6778796315193176,
      "learning_rate": 0.001,
      "loss": 4.3177,
      "step": 360
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.7114105820655823,
      "learning_rate": 0.001,
      "loss": 4.3722,
      "step": 365
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.6328292489051819,
      "learning_rate": 0.001,
      "loss": 4.4167,
      "step": 370
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5529919266700745,
      "learning_rate": 0.001,
      "loss": 4.2594,
      "step": 375
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5717706680297852,
      "learning_rate": 0.001,
      "loss": 4.2347,
      "step": 380
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.650095522403717,
      "learning_rate": 0.001,
      "loss": 4.401,
      "step": 385
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5702147483825684,
      "learning_rate": 0.001,
      "loss": 4.3214,
      "step": 390
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.5305518507957458,
      "learning_rate": 0.001,
      "loss": 4.3342,
      "step": 395
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.7027683258056641,
      "learning_rate": 0.001,
      "loss": 4.2095,
      "step": 400
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6951459050178528,
      "learning_rate": 0.001,
      "loss": 4.2882,
      "step": 405
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7413341403007507,
      "learning_rate": 0.001,
      "loss": 4.2696,
      "step": 410
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5163285136222839,
      "learning_rate": 0.001,
      "loss": 4.1847,
      "step": 415
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6258194446563721,
      "learning_rate": 0.001,
      "loss": 4.1892,
      "step": 420
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.50654536485672,
      "learning_rate": 0.001,
      "loss": 4.1689,
      "step": 425
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5324122905731201,
      "learning_rate": 0.001,
      "loss": 4.2116,
      "step": 430
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5191783905029297,
      "learning_rate": 0.001,
      "loss": 4.2668,
      "step": 435
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5304970145225525,
      "learning_rate": 0.001,
      "loss": 4.1726,
      "step": 440
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7103253602981567,
      "learning_rate": 0.001,
      "loss": 4.1596,
      "step": 445
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8040492534637451,
      "learning_rate": 0.001,
      "loss": 4.1499,
      "step": 450
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6443156599998474,
      "learning_rate": 0.001,
      "loss": 4.203,
      "step": 455
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5980875492095947,
      "learning_rate": 0.001,
      "loss": 4.0761,
      "step": 460
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5467867255210876,
      "learning_rate": 0.001,
      "loss": 4.0663,
      "step": 465
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5479826331138611,
      "learning_rate": 0.001,
      "loss": 4.0833,
      "step": 470
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5548881888389587,
      "learning_rate": 0.001,
      "loss": 4.0685,
      "step": 475
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.5504749417304993,
      "learning_rate": 0.001,
      "loss": 4.0026,
      "step": 480
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.586014449596405,
      "learning_rate": 0.001,
      "loss": 4.0989,
      "step": 485
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6106052398681641,
      "learning_rate": 0.001,
      "loss": 4.0259,
      "step": 490
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5603764653205872,
      "learning_rate": 0.001,
      "loss": 4.043,
      "step": 495
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5788707733154297,
      "learning_rate": 0.001,
      "loss": 3.9545,
      "step": 500
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6436803340911865,
      "learning_rate": 0.001,
      "loss": 3.9257,
      "step": 505
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5277396440505981,
      "learning_rate": 0.001,
      "loss": 4.0583,
      "step": 510
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6784444451332092,
      "learning_rate": 0.001,
      "loss": 4.0143,
      "step": 515
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.49947619438171387,
      "learning_rate": 0.001,
      "loss": 4.044,
      "step": 520
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6773266196250916,
      "learning_rate": 0.001,
      "loss": 3.9882,
      "step": 525
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.6483107209205627,
      "learning_rate": 0.001,
      "loss": 4.0046,
      "step": 530
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5643370747566223,
      "learning_rate": 0.001,
      "loss": 4.017,
      "step": 535
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5986424088478088,
      "learning_rate": 0.001,
      "loss": 4.0109,
      "step": 540
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.622429609298706,
      "learning_rate": 0.001,
      "loss": 3.9896,
      "step": 545
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5016334652900696,
      "learning_rate": 0.001,
      "loss": 3.9602,
      "step": 550
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5102951526641846,
      "learning_rate": 0.001,
      "loss": 4.0103,
      "step": 555
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.627109169960022,
      "learning_rate": 0.001,
      "loss": 3.9348,
      "step": 560
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4738805592060089,
      "learning_rate": 0.001,
      "loss": 4.0031,
      "step": 565
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.481922447681427,
      "learning_rate": 0.001,
      "loss": 3.7813,
      "step": 570
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5989673733711243,
      "learning_rate": 0.001,
      "loss": 3.8269,
      "step": 575
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.5524454116821289,
      "learning_rate": 0.001,
      "loss": 3.8124,
      "step": 580
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5364890098571777,
      "learning_rate": 0.001,
      "loss": 3.964,
      "step": 585
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5401674509048462,
      "learning_rate": 0.001,
      "loss": 3.857,
      "step": 590
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5142294764518738,
      "learning_rate": 0.001,
      "loss": 3.8363,
      "step": 595
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5484086275100708,
      "learning_rate": 0.001,
      "loss": 3.8658,
      "step": 600
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5811066627502441,
      "learning_rate": 0.001,
      "loss": 3.8457,
      "step": 605
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5071974396705627,
      "learning_rate": 0.001,
      "loss": 3.8835,
      "step": 610
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5145882964134216,
      "learning_rate": 0.001,
      "loss": 3.897,
      "step": 615
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6926450729370117,
      "learning_rate": 0.001,
      "loss": 3.8504,
      "step": 620
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5874271988868713,
      "learning_rate": 0.001,
      "loss": 3.8077,
      "step": 625
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5802385210990906,
      "learning_rate": 0.001,
      "loss": 3.7109,
      "step": 630
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.6140554547309875,
      "learning_rate": 0.001,
      "loss": 3.7658,
      "step": 635
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5724272727966309,
      "learning_rate": 0.001,
      "loss": 3.8777,
      "step": 640
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5731592178344727,
      "learning_rate": 0.001,
      "loss": 3.7584,
      "step": 645
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5535034537315369,
      "learning_rate": 0.001,
      "loss": 3.724,
      "step": 650
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5416511297225952,
      "learning_rate": 0.001,
      "loss": 3.77,
      "step": 655
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5728431344032288,
      "learning_rate": 0.001,
      "loss": 3.7719,
      "step": 660
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.5236804485321045,
      "learning_rate": 0.001,
      "loss": 3.6926,
      "step": 665
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.45392197370529175,
      "learning_rate": 0.001,
      "loss": 3.7736,
      "step": 670
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.7863075733184814,
      "learning_rate": 0.001,
      "loss": 3.7701,
      "step": 675
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5810198187828064,
      "learning_rate": 0.001,
      "loss": 3.8168,
      "step": 680
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5719917416572571,
      "learning_rate": 0.001,
      "loss": 3.752,
      "step": 685
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5634996294975281,
      "learning_rate": 0.001,
      "loss": 3.7123,
      "step": 690
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.49231231212615967,
      "learning_rate": 0.001,
      "loss": 3.8302,
      "step": 695
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.47931423783302307,
      "learning_rate": 0.001,
      "loss": 3.6192,
      "step": 700
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6042882204055786,
      "learning_rate": 0.001,
      "loss": 3.8259,
      "step": 705
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5771067142486572,
      "learning_rate": 0.001,
      "loss": 3.7338,
      "step": 710
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5514395236968994,
      "learning_rate": 0.001,
      "loss": 3.6775,
      "step": 715
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5697659850120544,
      "learning_rate": 0.001,
      "loss": 3.7105,
      "step": 720
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5804579257965088,
      "learning_rate": 0.001,
      "loss": 3.6896,
      "step": 725
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4920499622821808,
      "learning_rate": 0.001,
      "loss": 3.6511,
      "step": 730
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.523162841796875,
      "learning_rate": 0.001,
      "loss": 3.6019,
      "step": 735
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6031882166862488,
      "learning_rate": 0.001,
      "loss": 3.655,
      "step": 740
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6534960269927979,
      "learning_rate": 0.001,
      "loss": 3.7626,
      "step": 745
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.49020254611968994,
      "learning_rate": 0.001,
      "loss": 3.6494,
      "step": 750
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4870024621486664,
      "learning_rate": 0.001,
      "loss": 3.5926,
      "step": 755
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.48725512623786926,
      "learning_rate": 0.001,
      "loss": 3.714,
      "step": 760
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5335116982460022,
      "learning_rate": 0.001,
      "loss": 3.6698,
      "step": 765
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5708632469177246,
      "learning_rate": 0.001,
      "loss": 3.5812,
      "step": 770
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5631666779518127,
      "learning_rate": 0.001,
      "loss": 3.677,
      "step": 775
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5864231586456299,
      "learning_rate": 0.001,
      "loss": 3.5446,
      "step": 780
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5556133985519409,
      "learning_rate": 0.001,
      "loss": 3.529,
      "step": 785
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5345197319984436,
      "learning_rate": 0.001,
      "loss": 3.4959,
      "step": 790
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5649410486221313,
      "learning_rate": 0.001,
      "loss": 3.5892,
      "step": 795
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5137830376625061,
      "learning_rate": 0.001,
      "loss": 3.664,
      "step": 800
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.520827054977417,
      "learning_rate": 0.001,
      "loss": 3.6159,
      "step": 805
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5893942713737488,
      "learning_rate": 0.001,
      "loss": 3.5971,
      "step": 810
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.43705499172210693,
      "learning_rate": 0.001,
      "loss": 3.5704,
      "step": 815
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5601233243942261,
      "learning_rate": 0.001,
      "loss": 3.5379,
      "step": 820
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5474940538406372,
      "learning_rate": 0.001,
      "loss": 3.5245,
      "step": 825
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5808173418045044,
      "learning_rate": 0.001,
      "loss": 3.545,
      "step": 830
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5676250457763672,
      "learning_rate": 0.001,
      "loss": 3.5477,
      "step": 835
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.46730735898017883,
      "learning_rate": 0.001,
      "loss": 3.4416,
      "step": 840
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4779958128929138,
      "learning_rate": 0.001,
      "loss": 3.5729,
      "step": 845
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5286216735839844,
      "learning_rate": 0.001,
      "loss": 3.5727,
      "step": 850
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5003157258033752,
      "learning_rate": 0.001,
      "loss": 3.4412,
      "step": 855
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5391347408294678,
      "learning_rate": 0.001,
      "loss": 3.6227,
      "step": 860
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6406623721122742,
      "learning_rate": 0.001,
      "loss": 3.5755,
      "step": 865
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5844242572784424,
      "learning_rate": 0.001,
      "loss": 3.6629,
      "step": 870
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5803925395011902,
      "learning_rate": 0.001,
      "loss": 3.5022,
      "step": 875
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5397413372993469,
      "learning_rate": 0.001,
      "loss": 3.4458,
      "step": 880
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5229134559631348,
      "learning_rate": 0.001,
      "loss": 3.4075,
      "step": 885
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5586813688278198,
      "learning_rate": 0.001,
      "loss": 3.4069,
      "step": 890
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.457573801279068,
      "learning_rate": 0.001,
      "loss": 3.4071,
      "step": 895
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.511417031288147,
      "learning_rate": 0.001,
      "loss": 3.5552,
      "step": 900
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5040103197097778,
      "learning_rate": 0.001,
      "loss": 3.3518,
      "step": 905
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5554264187812805,
      "learning_rate": 0.001,
      "loss": 3.4197,
      "step": 910
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5565446019172668,
      "learning_rate": 0.001,
      "loss": 3.4614,
      "step": 915
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.518672525882721,
      "learning_rate": 0.001,
      "loss": 3.4518,
      "step": 920
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.5770706534385681,
      "learning_rate": 0.001,
      "loss": 3.3687,
      "step": 925
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.4561229646205902,
      "learning_rate": 0.001,
      "loss": 3.4112,
      "step": 930
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.49880415201187134,
      "learning_rate": 0.001,
      "loss": 3.4854,
      "step": 935
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.49336493015289307,
      "learning_rate": 0.001,
      "loss": 3.5201,
      "step": 940
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6361943483352661,
      "learning_rate": 0.001,
      "loss": 3.4105,
      "step": 945
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.49244019389152527,
      "learning_rate": 0.001,
      "loss": 3.4827,
      "step": 950
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5152875781059265,
      "learning_rate": 0.001,
      "loss": 3.4756,
      "step": 955
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5768316984176636,
      "learning_rate": 0.001,
      "loss": 3.3638,
      "step": 960
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5528039336204529,
      "learning_rate": 0.001,
      "loss": 3.3977,
      "step": 965
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5571544170379639,
      "learning_rate": 0.001,
      "loss": 3.4294,
      "step": 970
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.7128731608390808,
      "learning_rate": 0.001,
      "loss": 3.4117,
      "step": 975
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5800164341926575,
      "learning_rate": 0.001,
      "loss": 3.4575,
      "step": 980
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5743172764778137,
      "learning_rate": 0.001,
      "loss": 3.4217,
      "step": 985
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.526035726070404,
      "learning_rate": 0.001,
      "loss": 3.3472,
      "step": 990
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.47266894578933716,
      "learning_rate": 0.001,
      "loss": 3.3167,
      "step": 995
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5033674240112305,
      "learning_rate": 0.001,
      "loss": 3.2838,
      "step": 1000
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.46661603450775146,
      "learning_rate": 0.001,
      "loss": 3.2839,
      "step": 1005
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5624775886535645,
      "learning_rate": 0.001,
      "loss": 3.4624,
      "step": 1010
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5073076486587524,
      "learning_rate": 0.001,
      "loss": 3.3753,
      "step": 1015
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5086501240730286,
      "learning_rate": 0.001,
      "loss": 3.3817,
      "step": 1020
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.5310328602790833,
      "learning_rate": 0.001,
      "loss": 3.4449,
      "step": 1025
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.49915993213653564,
      "learning_rate": 0.001,
      "loss": 3.363,
      "step": 1030
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.628730058670044,
      "learning_rate": 0.001,
      "loss": 3.2733,
      "step": 1035
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5633178949356079,
      "learning_rate": 0.001,
      "loss": 3.3406,
      "step": 1040
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5721198320388794,
      "learning_rate": 0.001,
      "loss": 3.3102,
      "step": 1045
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.4597437381744385,
      "learning_rate": 0.001,
      "loss": 3.4369,
      "step": 1050
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5563680529594421,
      "learning_rate": 0.001,
      "loss": 3.3399,
      "step": 1055
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5712825059890747,
      "learning_rate": 0.001,
      "loss": 3.3234,
      "step": 1060
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5371583104133606,
      "learning_rate": 0.001,
      "loss": 3.3818,
      "step": 1065
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5913522243499756,
      "learning_rate": 0.001,
      "loss": 3.4158,
      "step": 1070
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.552200436592102,
      "learning_rate": 0.001,
      "loss": 3.3218,
      "step": 1075
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5645663142204285,
      "learning_rate": 0.001,
      "loss": 3.2884,
      "step": 1080
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.44569239020347595,
      "learning_rate": 0.001,
      "loss": 3.3263,
      "step": 1085
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8671941161155701,
      "learning_rate": 0.001,
      "loss": 3.3146,
      "step": 1090
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5645720362663269,
      "learning_rate": 0.001,
      "loss": 3.2641,
      "step": 1095
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.8040971755981445,
      "learning_rate": 0.001,
      "loss": 3.3728,
      "step": 1100
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5215751528739929,
      "learning_rate": 0.001,
      "loss": 3.2835,
      "step": 1105
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5733566284179688,
      "learning_rate": 0.001,
      "loss": 3.346,
      "step": 1110
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5822365880012512,
      "learning_rate": 0.001,
      "loss": 3.315,
      "step": 1115
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.5686581134796143,
      "learning_rate": 0.001,
      "loss": 3.3394,
      "step": 1120
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4820461571216583,
      "learning_rate": 0.001,
      "loss": 3.2288,
      "step": 1125
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5278226733207703,
      "learning_rate": 0.001,
      "loss": 3.2776,
      "step": 1130
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.45684191584587097,
      "learning_rate": 0.001,
      "loss": 3.2376,
      "step": 1135
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5594207644462585,
      "learning_rate": 0.001,
      "loss": 3.2892,
      "step": 1140
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5143517255783081,
      "learning_rate": 0.001,
      "loss": 3.2246,
      "step": 1145
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5187541842460632,
      "learning_rate": 0.001,
      "loss": 3.2442,
      "step": 1150
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5015732049942017,
      "learning_rate": 0.001,
      "loss": 3.1744,
      "step": 1155
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4809734523296356,
      "learning_rate": 0.001,
      "loss": 3.1277,
      "step": 1160
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.43246883153915405,
      "learning_rate": 0.001,
      "loss": 3.1667,
      "step": 1165
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.465455561876297,
      "learning_rate": 0.001,
      "loss": 3.1439,
      "step": 1170
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5171617865562439,
      "learning_rate": 0.001,
      "loss": 3.2541,
      "step": 1175
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.49306464195251465,
      "learning_rate": 0.001,
      "loss": 3.1276,
      "step": 1180
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.49295562505722046,
      "learning_rate": 0.001,
      "loss": 3.2369,
      "step": 1185
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5079097151756287,
      "learning_rate": 0.001,
      "loss": 3.2598,
      "step": 1190
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5303475260734558,
      "learning_rate": 0.001,
      "loss": 3.2031,
      "step": 1195
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5068909525871277,
      "learning_rate": 0.001,
      "loss": 3.2112,
      "step": 1200
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.509250819683075,
      "learning_rate": 0.001,
      "loss": 3.1927,
      "step": 1205
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.4846991002559662,
      "learning_rate": 0.001,
      "loss": 3.1834,
      "step": 1210
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5194084644317627,
      "learning_rate": 0.001,
      "loss": 3.2082,
      "step": 1215
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.45398786664009094,
      "learning_rate": 0.001,
      "loss": 3.1926,
      "step": 1220
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4347745478153229,
      "learning_rate": 0.001,
      "loss": 3.1371,
      "step": 1225
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5177764892578125,
      "learning_rate": 0.001,
      "loss": 3.1192,
      "step": 1230
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.41405558586120605,
      "learning_rate": 0.001,
      "loss": 3.1038,
      "step": 1235
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5231586694717407,
      "learning_rate": 0.001,
      "loss": 3.1101,
      "step": 1240
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.45989879965782166,
      "learning_rate": 0.001,
      "loss": 3.1313,
      "step": 1245
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4475898742675781,
      "learning_rate": 0.001,
      "loss": 3.164,
      "step": 1250
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4773527681827545,
      "learning_rate": 0.001,
      "loss": 3.0951,
      "step": 1255
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4193638265132904,
      "learning_rate": 0.001,
      "loss": 3.1184,
      "step": 1260
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5214313864707947,
      "learning_rate": 0.001,
      "loss": 3.0871,
      "step": 1265
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4901462197303772,
      "learning_rate": 0.001,
      "loss": 3.1454,
      "step": 1270
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.46214476227760315,
      "learning_rate": 0.001,
      "loss": 3.0632,
      "step": 1275
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4842498302459717,
      "learning_rate": 0.001,
      "loss": 3.0784,
      "step": 1280
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4946521520614624,
      "learning_rate": 0.001,
      "loss": 3.0605,
      "step": 1285
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.4697374999523163,
      "learning_rate": 0.001,
      "loss": 3.0256,
      "step": 1290
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6137799024581909,
      "learning_rate": 0.001,
      "loss": 3.1273,
      "step": 1295
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.45613402128219604,
      "learning_rate": 0.001,
      "loss": 3.0416,
      "step": 1300
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5034254789352417,
      "learning_rate": 0.001,
      "loss": 3.0875,
      "step": 1305
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5534226894378662,
      "learning_rate": 0.001,
      "loss": 3.1326,
      "step": 1310
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4972108006477356,
      "learning_rate": 0.001,
      "loss": 3.0417,
      "step": 1315
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.49605464935302734,
      "learning_rate": 0.001,
      "loss": 3.0207,
      "step": 1320
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4623696804046631,
      "learning_rate": 0.001,
      "loss": 3.0571,
      "step": 1325
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.5162575244903564,
      "learning_rate": 0.001,
      "loss": 3.0163,
      "step": 1330
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4853294789791107,
      "learning_rate": 0.001,
      "loss": 3.0887,
      "step": 1335
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.49405863881111145,
      "learning_rate": 0.001,
      "loss": 3.0364,
      "step": 1340
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.43189990520477295,
      "learning_rate": 0.001,
      "loss": 2.9356,
      "step": 1345
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4488712251186371,
      "learning_rate": 0.001,
      "loss": 3.0452,
      "step": 1350
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.44629403948783875,
      "learning_rate": 0.001,
      "loss": 2.9449,
      "step": 1355
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4826550781726837,
      "learning_rate": 0.001,
      "loss": 3.0303,
      "step": 1360
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.48896515369415283,
      "learning_rate": 0.001,
      "loss": 3.0951,
      "step": 1365
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4570922553539276,
      "learning_rate": 0.001,
      "loss": 3.0345,
      "step": 1370
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4907234013080597,
      "learning_rate": 0.001,
      "loss": 2.9406,
      "step": 1375
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.48720821738243103,
      "learning_rate": 0.001,
      "loss": 3.0605,
      "step": 1380
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.46225032210350037,
      "learning_rate": 0.001,
      "loss": 3.0262,
      "step": 1385
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.45703181624412537,
      "learning_rate": 0.001,
      "loss": 3.0684,
      "step": 1390
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.43968960642814636,
      "learning_rate": 0.001,
      "loss": 3.0707,
      "step": 1395
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5045078992843628,
      "learning_rate": 0.001,
      "loss": 3.0192,
      "step": 1400
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.49028080701828003,
      "learning_rate": 0.001,
      "loss": 2.9843,
      "step": 1405
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5171729326248169,
      "learning_rate": 0.001,
      "loss": 2.9823,
      "step": 1410
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4980088770389557,
      "learning_rate": 0.001,
      "loss": 3.0282,
      "step": 1415
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.47428199648857117,
      "learning_rate": 0.001,
      "loss": 3.0371,
      "step": 1420
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.49677643179893494,
      "learning_rate": 0.001,
      "loss": 2.8951,
      "step": 1425
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4643387198448181,
      "learning_rate": 0.001,
      "loss": 2.9398,
      "step": 1430
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4817788004875183,
      "learning_rate": 0.001,
      "loss": 3.0048,
      "step": 1435
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.46131885051727295,
      "learning_rate": 0.001,
      "loss": 3.0599,
      "step": 1440
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.48160094022750854,
      "learning_rate": 0.001,
      "loss": 2.9546,
      "step": 1445
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4246445894241333,
      "learning_rate": 0.001,
      "loss": 2.9025,
      "step": 1450
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4259074032306671,
      "learning_rate": 0.001,
      "loss": 2.9804,
      "step": 1455
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.4976622462272644,
      "learning_rate": 0.001,
      "loss": 2.9303,
      "step": 1460
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.40564754605293274,
      "learning_rate": 0.001,
      "loss": 2.9505,
      "step": 1465
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5583494305610657,
      "learning_rate": 0.001,
      "loss": 2.9898,
      "step": 1470
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.45296064019203186,
      "learning_rate": 0.001,
      "loss": 3.0225,
      "step": 1475
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.480425089597702,
      "learning_rate": 0.001,
      "loss": 2.9436,
      "step": 1480
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.47425466775894165,
      "learning_rate": 0.001,
      "loss": 3.0388,
      "step": 1485
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4673284888267517,
      "learning_rate": 0.001,
      "loss": 3.0704,
      "step": 1490
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.43329888582229614,
      "learning_rate": 0.001,
      "loss": 2.9181,
      "step": 1495
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.44485437870025635,
      "learning_rate": 0.001,
      "loss": 2.9171,
      "step": 1500
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.47537484765052795,
      "learning_rate": 0.001,
      "loss": 2.9375,
      "step": 1505
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.45429906249046326,
      "learning_rate": 0.001,
      "loss": 2.9171,
      "step": 1510
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4643488824367523,
      "learning_rate": 0.001,
      "loss": 2.9756,
      "step": 1515
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5488468408584595,
      "learning_rate": 0.001,
      "loss": 2.9335,
      "step": 1520
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5614462494850159,
      "learning_rate": 0.001,
      "loss": 2.9381,
      "step": 1525
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.5060487985610962,
      "learning_rate": 0.001,
      "loss": 2.9668,
      "step": 1530
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.46779105067253113,
      "learning_rate": 0.001,
      "loss": 2.9549,
      "step": 1535
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.43047505617141724,
      "learning_rate": 0.001,
      "loss": 2.9327,
      "step": 1540
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.42190369963645935,
      "learning_rate": 0.001,
      "loss": 2.9513,
      "step": 1545
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.44506198167800903,
      "learning_rate": 0.001,
      "loss": 2.896,
      "step": 1550
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4417656362056732,
      "learning_rate": 0.001,
      "loss": 2.8853,
      "step": 1555
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4769698977470398,
      "learning_rate": 0.001,
      "loss": 2.8021,
      "step": 1560
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.49687662720680237,
      "learning_rate": 0.001,
      "loss": 2.9001,
      "step": 1565
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4586490988731384,
      "learning_rate": 0.001,
      "loss": 2.9464,
      "step": 1570
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.45702454447746277,
      "learning_rate": 0.001,
      "loss": 2.9214,
      "step": 1575
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.44937050342559814,
      "learning_rate": 0.001,
      "loss": 2.96,
      "step": 1580
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4406932592391968,
      "learning_rate": 0.001,
      "loss": 2.8779,
      "step": 1585
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.41652610898017883,
      "learning_rate": 0.001,
      "loss": 2.9108,
      "step": 1590
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.47193869948387146,
      "learning_rate": 0.001,
      "loss": 2.9891,
      "step": 1595
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.44580379128456116,
      "learning_rate": 0.001,
      "loss": 2.8617,
      "step": 1600
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4470444619655609,
      "learning_rate": 0.001,
      "loss": 2.9058,
      "step": 1605
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.42569637298583984,
      "learning_rate": 0.001,
      "loss": 2.8462,
      "step": 1610
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.42253032326698303,
      "learning_rate": 0.001,
      "loss": 2.8579,
      "step": 1615
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4351730942726135,
      "learning_rate": 0.001,
      "loss": 2.9778,
      "step": 1620
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4730110168457031,
      "learning_rate": 0.001,
      "loss": 2.9028,
      "step": 1625
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.48326411843299866,
      "learning_rate": 0.001,
      "loss": 2.8306,
      "step": 1630
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4432814419269562,
      "learning_rate": 0.001,
      "loss": 2.8039,
      "step": 1635
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4231279194355011,
      "learning_rate": 0.001,
      "loss": 2.9011,
      "step": 1640
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4152611494064331,
      "learning_rate": 0.001,
      "loss": 2.9133,
      "step": 1645
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4555603861808777,
      "learning_rate": 0.001,
      "loss": 2.931,
      "step": 1650
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.4598769545555115,
      "learning_rate": 0.001,
      "loss": 2.8745,
      "step": 1655
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4184294044971466,
      "learning_rate": 0.001,
      "loss": 2.861,
      "step": 1660
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4301551282405853,
      "learning_rate": 0.001,
      "loss": 2.8547,
      "step": 1665
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4487614929676056,
      "learning_rate": 0.001,
      "loss": 2.9478,
      "step": 1670
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.47245121002197266,
      "learning_rate": 0.001,
      "loss": 2.8881,
      "step": 1675
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4665432572364807,
      "learning_rate": 0.001,
      "loss": 2.8708,
      "step": 1680
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4603164792060852,
      "learning_rate": 0.001,
      "loss": 2.7779,
      "step": 1685
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.39915090799331665,
      "learning_rate": 0.001,
      "loss": 2.8522,
      "step": 1690
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5415810346603394,
      "learning_rate": 0.001,
      "loss": 2.7831,
      "step": 1695
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.407660573720932,
      "learning_rate": 0.001,
      "loss": 2.7968,
      "step": 1700
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.41527602076530457,
      "learning_rate": 0.001,
      "loss": 2.9232,
      "step": 1705
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5355004072189331,
      "learning_rate": 0.001,
      "loss": 2.8865,
      "step": 1710
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.49530091881752014,
      "learning_rate": 0.001,
      "loss": 2.8188,
      "step": 1715
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.595953643321991,
      "learning_rate": 0.001,
      "loss": 2.8213,
      "step": 1720
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.44346776604652405,
      "learning_rate": 0.001,
      "loss": 2.9774,
      "step": 1725
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.4346086084842682,
      "learning_rate": 0.001,
      "loss": 2.7909,
      "step": 1730
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5057465434074402,
      "learning_rate": 0.001,
      "loss": 2.8719,
      "step": 1735
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.422921746969223,
      "learning_rate": 0.001,
      "loss": 2.8014,
      "step": 1740
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.41217803955078125,
      "learning_rate": 0.001,
      "loss": 2.8014,
      "step": 1745
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4260920584201813,
      "learning_rate": 0.001,
      "loss": 2.8018,
      "step": 1750
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3754720985889435,
      "learning_rate": 0.001,
      "loss": 2.7298,
      "step": 1755
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.45944565534591675,
      "learning_rate": 0.001,
      "loss": 2.825,
      "step": 1760
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.42768698930740356,
      "learning_rate": 0.001,
      "loss": 2.7903,
      "step": 1765
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4293604791164398,
      "learning_rate": 0.001,
      "loss": 2.7113,
      "step": 1770
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.43066537380218506,
      "learning_rate": 0.001,
      "loss": 2.8701,
      "step": 1775
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4278125762939453,
      "learning_rate": 0.001,
      "loss": 2.83,
      "step": 1780
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.430522084236145,
      "learning_rate": 0.001,
      "loss": 2.7792,
      "step": 1785
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.43876639008522034,
      "learning_rate": 0.001,
      "loss": 2.7494,
      "step": 1790
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.44354987144470215,
      "learning_rate": 0.001,
      "loss": 2.9732,
      "step": 1795
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.40760278701782227,
      "learning_rate": 0.001,
      "loss": 2.7543,
      "step": 1800
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4250537157058716,
      "learning_rate": 0.001,
      "loss": 2.8073,
      "step": 1805
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4468509256839752,
      "learning_rate": 0.001,
      "loss": 2.9124,
      "step": 1810
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4494674801826477,
      "learning_rate": 0.001,
      "loss": 2.7521,
      "step": 1815
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4005078077316284,
      "learning_rate": 0.001,
      "loss": 2.8788,
      "step": 1820
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4095415771007538,
      "learning_rate": 0.001,
      "loss": 2.8621,
      "step": 1825
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.5409905910491943,
      "learning_rate": 0.001,
      "loss": 2.8566,
      "step": 1830
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.49218934774398804,
      "learning_rate": 0.001,
      "loss": 2.8128,
      "step": 1835
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.45364123582839966,
      "learning_rate": 0.001,
      "loss": 2.8731,
      "step": 1840
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.43244045972824097,
      "learning_rate": 0.001,
      "loss": 2.7162,
      "step": 1845
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4195582866668701,
      "learning_rate": 0.001,
      "loss": 2.7579,
      "step": 1850
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4952572286128998,
      "learning_rate": 0.001,
      "loss": 2.7702,
      "step": 1855
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.44089606404304504,
      "learning_rate": 0.001,
      "loss": 2.7582,
      "step": 1860
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.39600706100463867,
      "learning_rate": 0.001,
      "loss": 2.7592,
      "step": 1865
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.43046873807907104,
      "learning_rate": 0.001,
      "loss": 2.8407,
      "step": 1870
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4137684404850006,
      "learning_rate": 0.001,
      "loss": 2.7494,
      "step": 1875
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.393900066614151,
      "learning_rate": 0.001,
      "loss": 2.7848,
      "step": 1880
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4853545129299164,
      "learning_rate": 0.001,
      "loss": 2.7849,
      "step": 1885
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4614884555339813,
      "learning_rate": 0.001,
      "loss": 2.8006,
      "step": 1890
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.41348981857299805,
      "learning_rate": 0.001,
      "loss": 2.7833,
      "step": 1895
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.46095848083496094,
      "learning_rate": 0.001,
      "loss": 2.7932,
      "step": 1900
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4020320773124695,
      "learning_rate": 0.001,
      "loss": 2.8505,
      "step": 1905
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.36754411458969116,
      "learning_rate": 0.001,
      "loss": 2.7562,
      "step": 1910
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4239796996116638,
      "learning_rate": 0.001,
      "loss": 2.8662,
      "step": 1915
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.40584829449653625,
      "learning_rate": 0.001,
      "loss": 2.8202,
      "step": 1920
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.39396342635154724,
      "learning_rate": 0.001,
      "loss": 2.8148,
      "step": 1925
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4060746729373932,
      "learning_rate": 0.001,
      "loss": 2.7949,
      "step": 1930
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.43469181656837463,
      "learning_rate": 0.001,
      "loss": 2.7653,
      "step": 1935
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.43287193775177,
      "learning_rate": 0.001,
      "loss": 2.8815,
      "step": 1940
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4579710066318512,
      "learning_rate": 0.001,
      "loss": 2.8147,
      "step": 1945
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4058091938495636,
      "learning_rate": 0.001,
      "loss": 2.8827,
      "step": 1950
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4505409300327301,
      "learning_rate": 0.001,
      "loss": 2.7635,
      "step": 1955
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.45017167925834656,
      "learning_rate": 0.001,
      "loss": 2.7809,
      "step": 1960
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.394645094871521,
      "learning_rate": 0.001,
      "loss": 2.6977,
      "step": 1965
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3600626587867737,
      "learning_rate": 0.001,
      "loss": 2.8135,
      "step": 1970
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4112387001514435,
      "learning_rate": 0.001,
      "loss": 2.7718,
      "step": 1975
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4643021523952484,
      "learning_rate": 0.001,
      "loss": 2.7279,
      "step": 1980
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4055080711841583,
      "learning_rate": 0.001,
      "loss": 2.8101,
      "step": 1985
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.4093475639820099,
      "learning_rate": 0.001,
      "loss": 2.7491,
      "step": 1990
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.43606457114219666,
      "learning_rate": 0.001,
      "loss": 2.7979,
      "step": 1995
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.38987255096435547,
      "learning_rate": 0.001,
      "loss": 2.7067,
      "step": 2000
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.44878268241882324,
      "learning_rate": 0.001,
      "loss": 2.7457,
      "step": 2005
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3909992277622223,
      "learning_rate": 0.001,
      "loss": 2.6962,
      "step": 2010
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3962748050689697,
      "learning_rate": 0.001,
      "loss": 2.7063,
      "step": 2015
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3859250843524933,
      "learning_rate": 0.001,
      "loss": 2.7814,
      "step": 2020
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4393429756164551,
      "learning_rate": 0.001,
      "loss": 2.7013,
      "step": 2025
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3963346481323242,
      "learning_rate": 0.001,
      "loss": 2.7637,
      "step": 2030
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.40981605648994446,
      "learning_rate": 0.001,
      "loss": 2.7556,
      "step": 2035
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.435086727142334,
      "learning_rate": 0.001,
      "loss": 2.7513,
      "step": 2040
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4393368661403656,
      "learning_rate": 0.001,
      "loss": 2.7835,
      "step": 2045
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3997877240180969,
      "learning_rate": 0.001,
      "loss": 2.6522,
      "step": 2050
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3757670521736145,
      "learning_rate": 0.001,
      "loss": 2.8765,
      "step": 2055
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3874036967754364,
      "learning_rate": 0.001,
      "loss": 2.7991,
      "step": 2060
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3644055128097534,
      "learning_rate": 0.001,
      "loss": 2.7529,
      "step": 2065
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3817943036556244,
      "learning_rate": 0.001,
      "loss": 2.703,
      "step": 2070
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3962540030479431,
      "learning_rate": 0.001,
      "loss": 2.672,
      "step": 2075
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.41838338971138,
      "learning_rate": 0.001,
      "loss": 2.7528,
      "step": 2080
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4394073784351349,
      "learning_rate": 0.001,
      "loss": 2.7653,
      "step": 2085
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4334862232208252,
      "learning_rate": 0.001,
      "loss": 2.7096,
      "step": 2090
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.41093143820762634,
      "learning_rate": 0.001,
      "loss": 2.7438,
      "step": 2095
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.35660019516944885,
      "learning_rate": 0.001,
      "loss": 2.8409,
      "step": 2100
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.373526394367218,
      "learning_rate": 0.001,
      "loss": 2.7351,
      "step": 2105
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4302496612071991,
      "learning_rate": 0.001,
      "loss": 2.8068,
      "step": 2110
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.38068363070487976,
      "learning_rate": 0.001,
      "loss": 2.6477,
      "step": 2115
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3733093738555908,
      "learning_rate": 0.001,
      "loss": 2.7554,
      "step": 2120
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4144055247306824,
      "learning_rate": 0.001,
      "loss": 2.6993,
      "step": 2125
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4264536499977112,
      "learning_rate": 0.001,
      "loss": 2.6572,
      "step": 2130
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4430059492588043,
      "learning_rate": 0.001,
      "loss": 2.6923,
      "step": 2135
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.42207950353622437,
      "learning_rate": 0.001,
      "loss": 2.7822,
      "step": 2140
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.35369545221328735,
      "learning_rate": 0.001,
      "loss": 2.7023,
      "step": 2145
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.47198721766471863,
      "learning_rate": 0.001,
      "loss": 2.6973,
      "step": 2150
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3969089388847351,
      "learning_rate": 0.001,
      "loss": 2.6911,
      "step": 2155
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.40460652112960815,
      "learning_rate": 0.001,
      "loss": 2.6883,
      "step": 2160
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.41705337166786194,
      "learning_rate": 0.001,
      "loss": 2.6843,
      "step": 2165
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.37905824184417725,
      "learning_rate": 0.001,
      "loss": 2.6169,
      "step": 2170
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3908678889274597,
      "learning_rate": 0.001,
      "loss": 2.6789,
      "step": 2175
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4016311466693878,
      "learning_rate": 0.001,
      "loss": 2.6935,
      "step": 2180
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.39033347368240356,
      "learning_rate": 0.001,
      "loss": 2.7653,
      "step": 2185
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4098024368286133,
      "learning_rate": 0.001,
      "loss": 2.6963,
      "step": 2190
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4102167785167694,
      "learning_rate": 0.001,
      "loss": 2.7452,
      "step": 2195
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3752821981906891,
      "learning_rate": 0.001,
      "loss": 2.6807,
      "step": 2200
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.412567675113678,
      "learning_rate": 0.001,
      "loss": 2.7763,
      "step": 2205
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4417850375175476,
      "learning_rate": 0.001,
      "loss": 2.6831,
      "step": 2210
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.41310781240463257,
      "learning_rate": 0.001,
      "loss": 2.6983,
      "step": 2215
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4148240089416504,
      "learning_rate": 0.001,
      "loss": 2.6653,
      "step": 2220
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.35075607895851135,
      "learning_rate": 0.001,
      "loss": 2.6965,
      "step": 2225
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3815755546092987,
      "learning_rate": 0.001,
      "loss": 2.6334,
      "step": 2230
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.35728955268859863,
      "learning_rate": 0.001,
      "loss": 2.7119,
      "step": 2235
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.4024193584918976,
      "learning_rate": 0.001,
      "loss": 2.7637,
      "step": 2240
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.38046032190322876,
      "learning_rate": 0.001,
      "loss": 2.7379,
      "step": 2245
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.39821740984916687,
      "learning_rate": 0.001,
      "loss": 2.6981,
      "step": 2250
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3438183665275574,
      "learning_rate": 0.001,
      "loss": 2.738,
      "step": 2255
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.39580243825912476,
      "learning_rate": 0.001,
      "loss": 2.662,
      "step": 2260
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3691055178642273,
      "learning_rate": 0.001,
      "loss": 2.7056,
      "step": 2265
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.36736342310905457,
      "learning_rate": 0.001,
      "loss": 2.6655,
      "step": 2270
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.36667105555534363,
      "learning_rate": 0.001,
      "loss": 2.7,
      "step": 2275
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.5083950757980347,
      "learning_rate": 0.001,
      "loss": 2.7651,
      "step": 2280
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.40544840693473816,
      "learning_rate": 0.001,
      "loss": 2.6053,
      "step": 2285
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3980036675930023,
      "learning_rate": 0.001,
      "loss": 2.64,
      "step": 2290
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.39068588614463806,
      "learning_rate": 0.001,
      "loss": 2.6498,
      "step": 2295
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.39119189977645874,
      "learning_rate": 0.001,
      "loss": 2.7336,
      "step": 2300
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3884316086769104,
      "learning_rate": 0.001,
      "loss": 2.6745,
      "step": 2305
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4157238006591797,
      "learning_rate": 0.001,
      "loss": 2.7262,
      "step": 2310
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.38538315892219543,
      "learning_rate": 0.001,
      "loss": 2.7375,
      "step": 2315
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.37677937746047974,
      "learning_rate": 0.001,
      "loss": 2.7534,
      "step": 2320
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.35606712102890015,
      "learning_rate": 0.001,
      "loss": 2.671,
      "step": 2325
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.37418973445892334,
      "learning_rate": 0.001,
      "loss": 2.7513,
      "step": 2330
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.36637210845947266,
      "learning_rate": 0.001,
      "loss": 2.6842,
      "step": 2335
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4047016501426697,
      "learning_rate": 0.001,
      "loss": 2.6538,
      "step": 2340
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.42498499155044556,
      "learning_rate": 0.001,
      "loss": 2.5852,
      "step": 2345
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4067090153694153,
      "learning_rate": 0.001,
      "loss": 2.6661,
      "step": 2350
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4145916998386383,
      "learning_rate": 0.001,
      "loss": 2.713,
      "step": 2355
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3945256173610687,
      "learning_rate": 0.001,
      "loss": 2.6483,
      "step": 2360
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4013041853904724,
      "learning_rate": 0.001,
      "loss": 2.6534,
      "step": 2365
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4283403754234314,
      "learning_rate": 0.001,
      "loss": 2.7215,
      "step": 2370
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.4242289066314697,
      "learning_rate": 0.001,
      "loss": 2.6233,
      "step": 2375
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3863242566585541,
      "learning_rate": 0.001,
      "loss": 2.674,
      "step": 2380
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.37469300627708435,
      "learning_rate": 0.001,
      "loss": 2.6534,
      "step": 2385
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.36515700817108154,
      "learning_rate": 0.001,
      "loss": 2.6909,
      "step": 2390
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3611971437931061,
      "learning_rate": 0.001,
      "loss": 2.6615,
      "step": 2395
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3939352035522461,
      "learning_rate": 0.001,
      "loss": 2.6824,
      "step": 2400
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3572635054588318,
      "learning_rate": 0.001,
      "loss": 2.6378,
      "step": 2405
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3500962257385254,
      "learning_rate": 0.001,
      "loss": 2.6074,
      "step": 2410
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3555634021759033,
      "learning_rate": 0.001,
      "loss": 2.588,
      "step": 2415
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3631855547428131,
      "learning_rate": 0.001,
      "loss": 2.7935,
      "step": 2420
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.35220634937286377,
      "learning_rate": 0.001,
      "loss": 2.7258,
      "step": 2425
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3524405360221863,
      "learning_rate": 0.001,
      "loss": 2.6489,
      "step": 2430
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.3426003158092499,
      "learning_rate": 0.001,
      "loss": 2.6484,
      "step": 2435
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.36606019735336304,
      "learning_rate": 0.001,
      "loss": 2.6621,
      "step": 2440
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.37562504410743713,
      "learning_rate": 0.001,
      "loss": 2.6919,
      "step": 2445
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.35107216238975525,
      "learning_rate": 0.001,
      "loss": 2.6227,
      "step": 2450
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.35221317410469055,
      "learning_rate": 0.001,
      "loss": 2.6016,
      "step": 2455
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.35811102390289307,
      "learning_rate": 0.001,
      "loss": 2.6422,
      "step": 2460
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.37652724981307983,
      "learning_rate": 0.001,
      "loss": 2.6598,
      "step": 2465
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3458632826805115,
      "learning_rate": 0.001,
      "loss": 2.7,
      "step": 2470
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.373162180185318,
      "learning_rate": 0.001,
      "loss": 2.6561,
      "step": 2475
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3542190492153168,
      "learning_rate": 0.001,
      "loss": 2.644,
      "step": 2480
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.34209930896759033,
      "learning_rate": 0.001,
      "loss": 2.6474,
      "step": 2485
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3443635106086731,
      "learning_rate": 0.001,
      "loss": 2.6696,
      "step": 2490
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.35547831654548645,
      "learning_rate": 0.001,
      "loss": 2.6165,
      "step": 2495
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4041297435760498,
      "learning_rate": 0.001,
      "loss": 2.729,
      "step": 2500
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.38527968525886536,
      "learning_rate": 0.001,
      "loss": 2.5177,
      "step": 2505
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3569463789463043,
      "learning_rate": 0.001,
      "loss": 2.5567,
      "step": 2510
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.35062989592552185,
      "learning_rate": 0.001,
      "loss": 2.6211,
      "step": 2515
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.375937819480896,
      "learning_rate": 0.001,
      "loss": 2.5725,
      "step": 2520
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.36163225769996643,
      "learning_rate": 0.001,
      "loss": 2.5525,
      "step": 2525
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3515682518482208,
      "learning_rate": 0.001,
      "loss": 2.5238,
      "step": 2530
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.36847323179244995,
      "learning_rate": 0.001,
      "loss": 2.6618,
      "step": 2535
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.39339104294776917,
      "learning_rate": 0.001,
      "loss": 2.5975,
      "step": 2540
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.37011218070983887,
      "learning_rate": 0.001,
      "loss": 2.6396,
      "step": 2545
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.31246432662010193,
      "learning_rate": 0.001,
      "loss": 2.5913,
      "step": 2550
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.4024028778076172,
      "learning_rate": 0.001,
      "loss": 2.6436,
      "step": 2555
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.41265320777893066,
      "learning_rate": 0.001,
      "loss": 2.6862,
      "step": 2560
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.35133570432662964,
      "learning_rate": 0.001,
      "loss": 2.6341,
      "step": 2565
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3641490340232849,
      "learning_rate": 0.001,
      "loss": 2.6138,
      "step": 2570
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.37135565280914307,
      "learning_rate": 0.001,
      "loss": 2.55,
      "step": 2575
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.8955615758895874,
      "learning_rate": 0.001,
      "loss": 2.5981,
      "step": 2580
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3701138198375702,
      "learning_rate": 0.001,
      "loss": 2.6368,
      "step": 2585
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.35533228516578674,
      "learning_rate": 0.001,
      "loss": 2.6313,
      "step": 2590
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3595050871372223,
      "learning_rate": 0.001,
      "loss": 2.5497,
      "step": 2595
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3303188383579254,
      "learning_rate": 0.001,
      "loss": 2.6971,
      "step": 2600
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3368977904319763,
      "learning_rate": 0.001,
      "loss": 2.6039,
      "step": 2605
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.35667169094085693,
      "learning_rate": 0.001,
      "loss": 2.6295,
      "step": 2610
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.343133807182312,
      "learning_rate": 0.001,
      "loss": 2.6248,
      "step": 2615
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.39267104864120483,
      "learning_rate": 0.001,
      "loss": 2.6834,
      "step": 2620
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.37460845708847046,
      "learning_rate": 0.001,
      "loss": 2.6252,
      "step": 2625
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.37292206287384033,
      "learning_rate": 0.001,
      "loss": 2.5632,
      "step": 2630
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3470982015132904,
      "learning_rate": 0.001,
      "loss": 2.5831,
      "step": 2635
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.33724093437194824,
      "learning_rate": 0.001,
      "loss": 2.6858,
      "step": 2640
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3472803831100464,
      "learning_rate": 0.001,
      "loss": 2.6371,
      "step": 2645
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5333802103996277,
      "learning_rate": 0.001,
      "loss": 2.6897,
      "step": 2650
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.325080007314682,
      "learning_rate": 0.001,
      "loss": 2.5439,
      "step": 2655
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3787167966365814,
      "learning_rate": 0.001,
      "loss": 2.6045,
      "step": 2660
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.36667078733444214,
      "learning_rate": 0.001,
      "loss": 2.6575,
      "step": 2665
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.36500290036201477,
      "learning_rate": 0.001,
      "loss": 2.5465,
      "step": 2670
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3717701733112335,
      "learning_rate": 0.001,
      "loss": 2.6718,
      "step": 2675
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.369501531124115,
      "learning_rate": 0.001,
      "loss": 2.6297,
      "step": 2680
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.33758407831192017,
      "learning_rate": 0.001,
      "loss": 2.6133,
      "step": 2685
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.383821964263916,
      "learning_rate": 0.001,
      "loss": 2.6689,
      "step": 2690
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3367357552051544,
      "learning_rate": 0.001,
      "loss": 2.636,
      "step": 2695
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3320029079914093,
      "learning_rate": 0.001,
      "loss": 2.6745,
      "step": 2700
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.33185863494873047,
      "learning_rate": 0.001,
      "loss": 2.6448,
      "step": 2705
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.35053524374961853,
      "learning_rate": 0.001,
      "loss": 2.7042,
      "step": 2710
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3864387571811676,
      "learning_rate": 0.001,
      "loss": 2.5968,
      "step": 2715
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.357790470123291,
      "learning_rate": 0.001,
      "loss": 2.6152,
      "step": 2720
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.37656790018081665,
      "learning_rate": 0.001,
      "loss": 2.608,
      "step": 2725
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3897472620010376,
      "learning_rate": 0.001,
      "loss": 2.6525,
      "step": 2730
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3524196147918701,
      "learning_rate": 0.001,
      "loss": 2.5498,
      "step": 2735
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3699324429035187,
      "learning_rate": 0.001,
      "loss": 2.5373,
      "step": 2740
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3401677906513214,
      "learning_rate": 0.001,
      "loss": 2.5847,
      "step": 2745
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3693203032016754,
      "learning_rate": 0.001,
      "loss": 2.5425,
      "step": 2750
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.36268529295921326,
      "learning_rate": 0.001,
      "loss": 2.5404,
      "step": 2755
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.365995854139328,
      "learning_rate": 0.001,
      "loss": 2.6491,
      "step": 2760
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3524530827999115,
      "learning_rate": 0.001,
      "loss": 2.5213,
      "step": 2765
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.35072949528694153,
      "learning_rate": 0.001,
      "loss": 2.6201,
      "step": 2770
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.34381788969039917,
      "learning_rate": 0.001,
      "loss": 2.6026,
      "step": 2775
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3447662591934204,
      "learning_rate": 0.001,
      "loss": 2.506,
      "step": 2780
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3632507622241974,
      "learning_rate": 0.001,
      "loss": 2.6266,
      "step": 2785
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3556525707244873,
      "learning_rate": 0.001,
      "loss": 2.5758,
      "step": 2790
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3616575300693512,
      "learning_rate": 0.001,
      "loss": 2.5722,
      "step": 2795
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.32523220777511597,
      "learning_rate": 0.001,
      "loss": 2.534,
      "step": 2800
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3450740873813629,
      "learning_rate": 0.001,
      "loss": 2.6071,
      "step": 2805
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3417651951313019,
      "learning_rate": 0.001,
      "loss": 2.5259,
      "step": 2810
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.33791735768318176,
      "learning_rate": 0.001,
      "loss": 2.541,
      "step": 2815
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.6335378885269165,
      "learning_rate": 0.001,
      "loss": 2.5412,
      "step": 2820
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.37333744764328003,
      "learning_rate": 0.001,
      "loss": 2.5993,
      "step": 2825
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.35744789242744446,
      "learning_rate": 0.001,
      "loss": 2.5405,
      "step": 2830
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3333902955055237,
      "learning_rate": 0.001,
      "loss": 2.6641,
      "step": 2835
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3603854775428772,
      "learning_rate": 0.001,
      "loss": 2.5494,
      "step": 2840
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.45514631271362305,
      "learning_rate": 0.001,
      "loss": 2.5488,
      "step": 2845
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3743136525154114,
      "learning_rate": 0.001,
      "loss": 2.6219,
      "step": 2850
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.34752535820007324,
      "learning_rate": 0.001,
      "loss": 2.6314,
      "step": 2855
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3930839002132416,
      "learning_rate": 0.001,
      "loss": 2.573,
      "step": 2860
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.36045753955841064,
      "learning_rate": 0.001,
      "loss": 2.5771,
      "step": 2865
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.32223668694496155,
      "learning_rate": 0.001,
      "loss": 2.6166,
      "step": 2870
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.33084774017333984,
      "learning_rate": 0.001,
      "loss": 2.6366,
      "step": 2875
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3649795651435852,
      "learning_rate": 0.001,
      "loss": 2.6151,
      "step": 2880
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3912632465362549,
      "learning_rate": 0.001,
      "loss": 2.6128,
      "step": 2885
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.40557408332824707,
      "learning_rate": 0.001,
      "loss": 2.5726,
      "step": 2890
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.36210837960243225,
      "learning_rate": 0.001,
      "loss": 2.603,
      "step": 2895
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3513619899749756,
      "learning_rate": 0.001,
      "loss": 2.5473,
      "step": 2900
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3209548592567444,
      "learning_rate": 0.001,
      "loss": 2.5749,
      "step": 2905
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.3491326570510864,
      "learning_rate": 0.001,
      "loss": 2.6812,
      "step": 2910
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.35135090351104736,
      "learning_rate": 0.001,
      "loss": 2.5705,
      "step": 2915
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3628234267234802,
      "learning_rate": 0.001,
      "loss": 2.5638,
      "step": 2920
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3620128333568573,
      "learning_rate": 0.001,
      "loss": 2.567,
      "step": 2925
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.35019025206565857,
      "learning_rate": 0.001,
      "loss": 2.5266,
      "step": 2930
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.36038318276405334,
      "learning_rate": 0.001,
      "loss": 2.5184,
      "step": 2935
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3591684401035309,
      "learning_rate": 0.001,
      "loss": 2.5427,
      "step": 2940
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3429654538631439,
      "learning_rate": 0.001,
      "loss": 2.5432,
      "step": 2945
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.35099634528160095,
      "learning_rate": 0.001,
      "loss": 2.5396,
      "step": 2950
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3457493484020233,
      "learning_rate": 0.001,
      "loss": 2.5672,
      "step": 2955
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.36206451058387756,
      "learning_rate": 0.001,
      "loss": 2.5766,
      "step": 2960
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3406093716621399,
      "learning_rate": 0.001,
      "loss": 2.6046,
      "step": 2965
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.35393887758255005,
      "learning_rate": 0.001,
      "loss": 2.5463,
      "step": 2970
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3434804081916809,
      "learning_rate": 0.001,
      "loss": 2.5328,
      "step": 2975
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.32725098729133606,
      "learning_rate": 0.001,
      "loss": 2.6078,
      "step": 2980
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3323499858379364,
      "learning_rate": 0.001,
      "loss": 2.5363,
      "step": 2985
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3031046390533447,
      "learning_rate": 0.001,
      "loss": 2.5489,
      "step": 2990
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3338392674922943,
      "learning_rate": 0.001,
      "loss": 2.5397,
      "step": 2995
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3185463547706604,
      "learning_rate": 0.001,
      "loss": 2.5963,
      "step": 3000
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.34104982018470764,
      "learning_rate": 0.001,
      "loss": 2.5777,
      "step": 3005
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.35556039214134216,
      "learning_rate": 0.001,
      "loss": 2.5151,
      "step": 3010
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3338676989078522,
      "learning_rate": 0.001,
      "loss": 2.501,
      "step": 3015
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.35118651390075684,
      "learning_rate": 0.001,
      "loss": 2.586,
      "step": 3020
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.32692933082580566,
      "learning_rate": 0.001,
      "loss": 2.6081,
      "step": 3025
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.34374868869781494,
      "learning_rate": 0.001,
      "loss": 2.6544,
      "step": 3030
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3365413546562195,
      "learning_rate": 0.001,
      "loss": 2.5178,
      "step": 3035
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.33963915705680847,
      "learning_rate": 0.001,
      "loss": 2.5041,
      "step": 3040
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3642427623271942,
      "learning_rate": 0.001,
      "loss": 2.5709,
      "step": 3045
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.33113953471183777,
      "learning_rate": 0.001,
      "loss": 2.62,
      "step": 3050
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.36132606863975525,
      "learning_rate": 0.001,
      "loss": 2.4981,
      "step": 3055
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.34095078706741333,
      "learning_rate": 0.001,
      "loss": 2.5128,
      "step": 3060
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.327793151140213,
      "learning_rate": 0.001,
      "loss": 2.4821,
      "step": 3065
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.33965983986854553,
      "learning_rate": 0.001,
      "loss": 2.6404,
      "step": 3070
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3375244438648224,
      "learning_rate": 0.001,
      "loss": 2.5655,
      "step": 3075
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.34356772899627686,
      "learning_rate": 0.001,
      "loss": 2.5837,
      "step": 3080
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.34157508611679077,
      "learning_rate": 0.001,
      "loss": 2.6517,
      "step": 3085
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3461490571498871,
      "learning_rate": 0.001,
      "loss": 2.598,
      "step": 3090
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3149019777774811,
      "learning_rate": 0.001,
      "loss": 2.5534,
      "step": 3095
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3143615126609802,
      "learning_rate": 0.001,
      "loss": 2.5102,
      "step": 3100
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3524455726146698,
      "learning_rate": 0.001,
      "loss": 2.5703,
      "step": 3105
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3223250210285187,
      "learning_rate": 0.001,
      "loss": 2.4897,
      "step": 3110
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3235696852207184,
      "learning_rate": 0.001,
      "loss": 2.5577,
      "step": 3115
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3154476284980774,
      "learning_rate": 0.001,
      "loss": 2.551,
      "step": 3120
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3305743634700775,
      "learning_rate": 0.001,
      "loss": 2.5497,
      "step": 3125
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3358781337738037,
      "learning_rate": 0.001,
      "loss": 2.4787,
      "step": 3130
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.33123084902763367,
      "learning_rate": 0.001,
      "loss": 2.5887,
      "step": 3135
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3147204518318176,
      "learning_rate": 0.001,
      "loss": 2.5253,
      "step": 3140
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.31788182258605957,
      "learning_rate": 0.001,
      "loss": 2.5734,
      "step": 3145
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3127879202365875,
      "learning_rate": 0.001,
      "loss": 2.4203,
      "step": 3150
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3299925625324249,
      "learning_rate": 0.001,
      "loss": 2.5275,
      "step": 3155
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3190239667892456,
      "learning_rate": 0.001,
      "loss": 2.4905,
      "step": 3160
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.34184321761131287,
      "learning_rate": 0.001,
      "loss": 2.4114,
      "step": 3165
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.31968846917152405,
      "learning_rate": 0.001,
      "loss": 2.5814,
      "step": 3170
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3431030213832855,
      "learning_rate": 0.001,
      "loss": 2.6748,
      "step": 3175
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3149561882019043,
      "learning_rate": 0.001,
      "loss": 2.5587,
      "step": 3180
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3136409521102905,
      "learning_rate": 0.001,
      "loss": 2.587,
      "step": 3185
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.31620776653289795,
      "learning_rate": 0.001,
      "loss": 2.5091,
      "step": 3190
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3681703805923462,
      "learning_rate": 0.001,
      "loss": 2.5609,
      "step": 3195
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.30699193477630615,
      "learning_rate": 0.001,
      "loss": 2.6231,
      "step": 3200
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.319044291973114,
      "learning_rate": 0.001,
      "loss": 2.652,
      "step": 3205
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3128690719604492,
      "learning_rate": 0.001,
      "loss": 2.4954,
      "step": 3210
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.328692764043808,
      "learning_rate": 0.001,
      "loss": 2.4826,
      "step": 3215
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3168734014034271,
      "learning_rate": 0.001,
      "loss": 2.4623,
      "step": 3220
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.31120923161506653,
      "learning_rate": 0.001,
      "loss": 2.5033,
      "step": 3225
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3216933608055115,
      "learning_rate": 0.001,
      "loss": 2.5691,
      "step": 3230
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.33055299520492554,
      "learning_rate": 0.001,
      "loss": 2.5753,
      "step": 3235
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.33167943358421326,
      "learning_rate": 0.001,
      "loss": 2.4651,
      "step": 3240
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3290786147117615,
      "learning_rate": 0.001,
      "loss": 2.557,
      "step": 3245
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3318544030189514,
      "learning_rate": 0.001,
      "loss": 2.5146,
      "step": 3250
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.31007036566734314,
      "learning_rate": 0.001,
      "loss": 2.463,
      "step": 3255
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.31148260831832886,
      "learning_rate": 0.001,
      "loss": 2.5997,
      "step": 3260
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3110112249851227,
      "learning_rate": 0.001,
      "loss": 2.5517,
      "step": 3265
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3323996663093567,
      "learning_rate": 0.001,
      "loss": 2.5036,
      "step": 3270
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3133048713207245,
      "learning_rate": 0.001,
      "loss": 2.4954,
      "step": 3275
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3396298289299011,
      "learning_rate": 0.001,
      "loss": 2.5113,
      "step": 3280
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3418448567390442,
      "learning_rate": 0.001,
      "loss": 2.5753,
      "step": 3285
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.34545567631721497,
      "learning_rate": 0.001,
      "loss": 2.5325,
      "step": 3290
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3261316418647766,
      "learning_rate": 0.001,
      "loss": 2.5489,
      "step": 3295
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3248863220214844,
      "learning_rate": 0.001,
      "loss": 2.4554,
      "step": 3300
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.30948251485824585,
      "learning_rate": 0.001,
      "loss": 2.497,
      "step": 3305
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.32646802067756653,
      "learning_rate": 0.001,
      "loss": 2.5322,
      "step": 3310
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3012072741985321,
      "learning_rate": 0.001,
      "loss": 2.4644,
      "step": 3315
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3290029764175415,
      "learning_rate": 0.001,
      "loss": 2.5015,
      "step": 3320
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3250294625759125,
      "learning_rate": 0.001,
      "loss": 2.4714,
      "step": 3325
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.2954888939857483,
      "learning_rate": 0.001,
      "loss": 2.5344,
      "step": 3330
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.31430017948150635,
      "learning_rate": 0.001,
      "loss": 2.518,
      "step": 3335
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.33378204703330994,
      "learning_rate": 0.001,
      "loss": 2.5019,
      "step": 3340
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.33612725138664246,
      "learning_rate": 0.001,
      "loss": 2.5037,
      "step": 3345
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.30687910318374634,
      "learning_rate": 0.001,
      "loss": 2.4912,
      "step": 3350
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.30845963954925537,
      "learning_rate": 0.001,
      "loss": 2.4837,
      "step": 3355
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.29248157143592834,
      "learning_rate": 0.001,
      "loss": 2.4905,
      "step": 3360
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2903001606464386,
      "learning_rate": 0.001,
      "loss": 2.4785,
      "step": 3365
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.31211957335472107,
      "learning_rate": 0.001,
      "loss": 2.484,
      "step": 3370
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3141609728336334,
      "learning_rate": 0.001,
      "loss": 2.6068,
      "step": 3375
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.30534839630126953,
      "learning_rate": 0.001,
      "loss": 2.4743,
      "step": 3380
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.31994548439979553,
      "learning_rate": 0.001,
      "loss": 2.521,
      "step": 3385
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.31002023816108704,
      "learning_rate": 0.001,
      "loss": 2.6565,
      "step": 3390
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3236083984375,
      "learning_rate": 0.001,
      "loss": 2.5602,
      "step": 3395
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3060869872570038,
      "learning_rate": 0.001,
      "loss": 2.4858,
      "step": 3400
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.28208038210868835,
      "learning_rate": 0.001,
      "loss": 2.4991,
      "step": 3405
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.33403393626213074,
      "learning_rate": 0.001,
      "loss": 2.4914,
      "step": 3410
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.382112592458725,
      "learning_rate": 0.001,
      "loss": 2.5804,
      "step": 3415
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2976379692554474,
      "learning_rate": 0.001,
      "loss": 2.4831,
      "step": 3420
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.29742851853370667,
      "learning_rate": 0.001,
      "loss": 2.5167,
      "step": 3425
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.27965688705444336,
      "learning_rate": 0.001,
      "loss": 2.4606,
      "step": 3430
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3177298307418823,
      "learning_rate": 0.001,
      "loss": 2.4297,
      "step": 3435
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3082432150840759,
      "learning_rate": 0.001,
      "loss": 2.4864,
      "step": 3440
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.30660197138786316,
      "learning_rate": 0.001,
      "loss": 2.4922,
      "step": 3445
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.3045155704021454,
      "learning_rate": 0.001,
      "loss": 2.5424,
      "step": 3450
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3095363974571228,
      "learning_rate": 0.001,
      "loss": 2.5205,
      "step": 3455
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.30330196022987366,
      "learning_rate": 0.001,
      "loss": 2.4601,
      "step": 3460
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.31325945258140564,
      "learning_rate": 0.001,
      "loss": 2.4919,
      "step": 3465
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2912018597126007,
      "learning_rate": 0.001,
      "loss": 2.4719,
      "step": 3470
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3126067519187927,
      "learning_rate": 0.001,
      "loss": 2.5373,
      "step": 3475
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.29371434450149536,
      "learning_rate": 0.001,
      "loss": 2.5213,
      "step": 3480
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.30860206484794617,
      "learning_rate": 0.001,
      "loss": 2.4439,
      "step": 3485
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.30670803785324097,
      "learning_rate": 0.001,
      "loss": 2.551,
      "step": 3490
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3091505765914917,
      "learning_rate": 0.001,
      "loss": 2.5826,
      "step": 3495
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3141428232192993,
      "learning_rate": 0.001,
      "loss": 2.4817,
      "step": 3500
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2985532283782959,
      "learning_rate": 0.001,
      "loss": 2.5519,
      "step": 3505
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3592018187046051,
      "learning_rate": 0.001,
      "loss": 2.4774,
      "step": 3510
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2912810444831848,
      "learning_rate": 0.001,
      "loss": 2.5316,
      "step": 3515
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3026273548603058,
      "learning_rate": 0.001,
      "loss": 2.451,
      "step": 3520
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.3669629693031311,
      "learning_rate": 0.001,
      "loss": 2.4407,
      "step": 3525
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.334534227848053,
      "learning_rate": 0.001,
      "loss": 2.5109,
      "step": 3530
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.33897146582603455,
      "learning_rate": 0.001,
      "loss": 2.4755,
      "step": 3535
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.30617454648017883,
      "learning_rate": 0.001,
      "loss": 2.5165,
      "step": 3540
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.30545178055763245,
      "learning_rate": 0.001,
      "loss": 2.5267,
      "step": 3545
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3039405345916748,
      "learning_rate": 0.001,
      "loss": 2.508,
      "step": 3550
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.31085509061813354,
      "learning_rate": 0.001,
      "loss": 2.5329,
      "step": 3555
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3017004132270813,
      "learning_rate": 0.001,
      "loss": 2.4686,
      "step": 3560
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2899893522262573,
      "learning_rate": 0.001,
      "loss": 2.5395,
      "step": 3565
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3174129128456116,
      "learning_rate": 0.001,
      "loss": 2.4909,
      "step": 3570
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.32396823167800903,
      "learning_rate": 0.001,
      "loss": 2.4498,
      "step": 3575
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.31135293841362,
      "learning_rate": 0.001,
      "loss": 2.4446,
      "step": 3580
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.32165661454200745,
      "learning_rate": 0.001,
      "loss": 2.4139,
      "step": 3585
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3169751465320587,
      "learning_rate": 0.001,
      "loss": 2.4799,
      "step": 3590
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.31132322549819946,
      "learning_rate": 0.001,
      "loss": 2.5442,
      "step": 3595
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3260294497013092,
      "learning_rate": 0.001,
      "loss": 2.4054,
      "step": 3600
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.29337745904922485,
      "learning_rate": 0.001,
      "loss": 2.4915,
      "step": 3605
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.31145504117012024,
      "learning_rate": 0.001,
      "loss": 2.4534,
      "step": 3610
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.299142450094223,
      "learning_rate": 0.001,
      "loss": 2.4935,
      "step": 3615
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.3256239593029022,
      "learning_rate": 0.001,
      "loss": 2.4649,
      "step": 3620
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.30987122654914856,
      "learning_rate": 0.001,
      "loss": 2.4172,
      "step": 3625
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.4387304484844208,
      "learning_rate": 0.001,
      "loss": 2.442,
      "step": 3630
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.28370875120162964,
      "learning_rate": 0.001,
      "loss": 2.4926,
      "step": 3635
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.27789434790611267,
      "learning_rate": 0.001,
      "loss": 2.4874,
      "step": 3640
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.298215389251709,
      "learning_rate": 0.001,
      "loss": 2.5084,
      "step": 3645
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.28470468521118164,
      "learning_rate": 0.001,
      "loss": 2.4685,
      "step": 3650
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.31583115458488464,
      "learning_rate": 0.001,
      "loss": 2.4241,
      "step": 3655
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3249650001525879,
      "learning_rate": 0.001,
      "loss": 2.415,
      "step": 3660
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.33161094784736633,
      "learning_rate": 0.001,
      "loss": 2.4805,
      "step": 3665
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.29928529262542725,
      "learning_rate": 0.001,
      "loss": 2.4966,
      "step": 3670
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.31636932492256165,
      "learning_rate": 0.001,
      "loss": 2.5237,
      "step": 3675
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3295460641384125,
      "learning_rate": 0.001,
      "loss": 2.4059,
      "step": 3680
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3235255777835846,
      "learning_rate": 0.001,
      "loss": 2.4589,
      "step": 3685
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.32697299122810364,
      "learning_rate": 0.001,
      "loss": 2.3683,
      "step": 3690
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3994431793689728,
      "learning_rate": 0.001,
      "loss": 2.4794,
      "step": 3695
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3653365671634674,
      "learning_rate": 0.001,
      "loss": 2.5027,
      "step": 3700
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3047979772090912,
      "learning_rate": 0.001,
      "loss": 2.4745,
      "step": 3705
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.4140577018260956,
      "learning_rate": 0.001,
      "loss": 2.5028,
      "step": 3710
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.3812958300113678,
      "learning_rate": 0.001,
      "loss": 2.4999,
      "step": 3715
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.32625457644462585,
      "learning_rate": 0.001,
      "loss": 2.4991,
      "step": 3720
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.30464789271354675,
      "learning_rate": 0.001,
      "loss": 2.4738,
      "step": 3725
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3046870827674866,
      "learning_rate": 0.001,
      "loss": 2.5732,
      "step": 3730
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.33171603083610535,
      "learning_rate": 0.001,
      "loss": 2.4796,
      "step": 3735
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2825039029121399,
      "learning_rate": 0.001,
      "loss": 2.4877,
      "step": 3740
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3741876184940338,
      "learning_rate": 0.001,
      "loss": 2.5031,
      "step": 3745
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.30283525586128235,
      "learning_rate": 0.001,
      "loss": 2.4659,
      "step": 3750
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.28069236874580383,
      "learning_rate": 0.001,
      "loss": 2.402,
      "step": 3755
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.273360937833786,
      "learning_rate": 0.001,
      "loss": 2.5441,
      "step": 3760
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3147111237049103,
      "learning_rate": 0.001,
      "loss": 2.5057,
      "step": 3765
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3091120719909668,
      "learning_rate": 0.001,
      "loss": 2.4444,
      "step": 3770
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3067731559276581,
      "learning_rate": 0.001,
      "loss": 2.5245,
      "step": 3775
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2935093939304352,
      "learning_rate": 0.001,
      "loss": 2.5006,
      "step": 3780
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3160226345062256,
      "learning_rate": 0.001,
      "loss": 2.4026,
      "step": 3785
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.306943416595459,
      "learning_rate": 0.001,
      "loss": 2.5193,
      "step": 3790
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.28538256883621216,
      "learning_rate": 0.001,
      "loss": 2.4627,
      "step": 3795
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2843259871006012,
      "learning_rate": 0.001,
      "loss": 2.3829,
      "step": 3800
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2874767482280731,
      "learning_rate": 0.001,
      "loss": 2.511,
      "step": 3805
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.26737627387046814,
      "learning_rate": 0.001,
      "loss": 2.439,
      "step": 3810
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.30951711535453796,
      "learning_rate": 0.001,
      "loss": 2.4584,
      "step": 3815
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2935316264629364,
      "learning_rate": 0.001,
      "loss": 2.4268,
      "step": 3820
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2819925546646118,
      "learning_rate": 0.001,
      "loss": 2.4609,
      "step": 3825
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.278676837682724,
      "learning_rate": 0.001,
      "loss": 2.3241,
      "step": 3830
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2986203134059906,
      "learning_rate": 0.001,
      "loss": 2.5406,
      "step": 3835
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2985837161540985,
      "learning_rate": 0.001,
      "loss": 2.441,
      "step": 3840
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2917584180831909,
      "learning_rate": 0.001,
      "loss": 2.4911,
      "step": 3845
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.29989174008369446,
      "learning_rate": 0.001,
      "loss": 2.5409,
      "step": 3850
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.30956900119781494,
      "learning_rate": 0.001,
      "loss": 2.55,
      "step": 3855
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3005373477935791,
      "learning_rate": 0.001,
      "loss": 2.5079,
      "step": 3860
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2777903079986572,
      "learning_rate": 0.001,
      "loss": 2.4218,
      "step": 3865
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3362637162208557,
      "learning_rate": 0.001,
      "loss": 2.5395,
      "step": 3870
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.299837201833725,
      "learning_rate": 0.001,
      "loss": 2.5473,
      "step": 3875
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.28119298815727234,
      "learning_rate": 0.001,
      "loss": 2.4785,
      "step": 3880
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.29247012734413147,
      "learning_rate": 0.001,
      "loss": 2.5776,
      "step": 3885
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.27871406078338623,
      "learning_rate": 0.001,
      "loss": 2.4137,
      "step": 3890
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.2752448320388794,
      "learning_rate": 0.001,
      "loss": 2.463,
      "step": 3895
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.291769802570343,
      "learning_rate": 0.001,
      "loss": 2.4766,
      "step": 3900
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.31150388717651367,
      "learning_rate": 0.001,
      "loss": 2.4992,
      "step": 3905
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2958334684371948,
      "learning_rate": 0.001,
      "loss": 2.4278,
      "step": 3910
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.30901768803596497,
      "learning_rate": 0.001,
      "loss": 2.4791,
      "step": 3915
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3076885938644409,
      "learning_rate": 0.001,
      "loss": 2.4737,
      "step": 3920
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.29541951417922974,
      "learning_rate": 0.001,
      "loss": 2.5097,
      "step": 3925
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.30541756749153137,
      "learning_rate": 0.001,
      "loss": 2.495,
      "step": 3930
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.29810360074043274,
      "learning_rate": 0.001,
      "loss": 2.3795,
      "step": 3935
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.30572912096977234,
      "learning_rate": 0.001,
      "loss": 2.4483,
      "step": 3940
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2885902523994446,
      "learning_rate": 0.001,
      "loss": 2.426,
      "step": 3945
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2982771396636963,
      "learning_rate": 0.001,
      "loss": 2.3061,
      "step": 3950
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3130563497543335,
      "learning_rate": 0.001,
      "loss": 2.4545,
      "step": 3955
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2932124137878418,
      "learning_rate": 0.001,
      "loss": 2.4065,
      "step": 3960
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.28204530477523804,
      "learning_rate": 0.001,
      "loss": 2.4234,
      "step": 3965
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.29289600253105164,
      "learning_rate": 0.001,
      "loss": 2.3772,
      "step": 3970
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2805161774158478,
      "learning_rate": 0.001,
      "loss": 2.5557,
      "step": 3975
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2907586097717285,
      "learning_rate": 0.001,
      "loss": 2.5438,
      "step": 3980
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2835649251937866,
      "learning_rate": 0.001,
      "loss": 2.4499,
      "step": 3985
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2827214300632477,
      "learning_rate": 0.001,
      "loss": 2.4512,
      "step": 3990
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.34279364347457886,
      "learning_rate": 0.001,
      "loss": 2.4925,
      "step": 3995
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2893143594264984,
      "learning_rate": 0.001,
      "loss": 2.4309,
      "step": 4000
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.27287694811820984,
      "learning_rate": 0.001,
      "loss": 2.417,
      "step": 4005
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.29163962602615356,
      "learning_rate": 0.001,
      "loss": 2.3912,
      "step": 4010
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3035675287246704,
      "learning_rate": 0.001,
      "loss": 2.4203,
      "step": 4015
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.28289881348609924,
      "learning_rate": 0.001,
      "loss": 2.3951,
      "step": 4020
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3087094724178314,
      "learning_rate": 0.001,
      "loss": 2.4091,
      "step": 4025
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.29908227920532227,
      "learning_rate": 0.001,
      "loss": 2.4428,
      "step": 4030
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.28541573882102966,
      "learning_rate": 0.001,
      "loss": 2.3673,
      "step": 4035
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.28469517827033997,
      "learning_rate": 0.001,
      "loss": 2.4308,
      "step": 4040
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3017624020576477,
      "learning_rate": 0.001,
      "loss": 2.417,
      "step": 4045
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3090055286884308,
      "learning_rate": 0.001,
      "loss": 2.3774,
      "step": 4050
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.30866459012031555,
      "learning_rate": 0.001,
      "loss": 2.4988,
      "step": 4055
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3137664794921875,
      "learning_rate": 0.001,
      "loss": 2.5087,
      "step": 4060
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2756161689758301,
      "learning_rate": 0.001,
      "loss": 2.4424,
      "step": 4065
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.28431424498558044,
      "learning_rate": 0.001,
      "loss": 2.4091,
      "step": 4070
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.2876141667366028,
      "learning_rate": 0.001,
      "loss": 2.4231,
      "step": 4075
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.27080801129341125,
      "learning_rate": 0.001,
      "loss": 2.4929,
      "step": 4080
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2989480197429657,
      "learning_rate": 0.001,
      "loss": 2.4463,
      "step": 4085
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.3767198622226715,
      "learning_rate": 0.001,
      "loss": 2.5178,
      "step": 4090
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.27639487385749817,
      "learning_rate": 0.001,
      "loss": 2.4472,
      "step": 4095
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.30378058552742004,
      "learning_rate": 0.001,
      "loss": 2.4917,
      "step": 4100
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2825583219528198,
      "learning_rate": 0.001,
      "loss": 2.4288,
      "step": 4105
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.26925694942474365,
      "learning_rate": 0.001,
      "loss": 2.4241,
      "step": 4110
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.287741482257843,
      "learning_rate": 0.001,
      "loss": 2.4622,
      "step": 4115
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.312418133020401,
      "learning_rate": 0.001,
      "loss": 2.4169,
      "step": 4120
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2728055417537689,
      "learning_rate": 0.001,
      "loss": 2.464,
      "step": 4125
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2967456877231598,
      "learning_rate": 0.001,
      "loss": 2.5317,
      "step": 4130
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2888936698436737,
      "learning_rate": 0.001,
      "loss": 2.4206,
      "step": 4135
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2940239906311035,
      "learning_rate": 0.001,
      "loss": 2.5697,
      "step": 4140
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.33775338530540466,
      "learning_rate": 0.001,
      "loss": 2.3773,
      "step": 4145
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2790275812149048,
      "learning_rate": 0.001,
      "loss": 2.3066,
      "step": 4150
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.26847320795059204,
      "learning_rate": 0.001,
      "loss": 2.4747,
      "step": 4155
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.3185865879058838,
      "learning_rate": 0.001,
      "loss": 2.4139,
      "step": 4160
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.278110533952713,
      "learning_rate": 0.001,
      "loss": 2.5095,
      "step": 4165
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.3306000828742981,
      "learning_rate": 0.001,
      "loss": 2.4582,
      "step": 4170
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2653733491897583,
      "learning_rate": 0.001,
      "loss": 2.4082,
      "step": 4175
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.27316001057624817,
      "learning_rate": 0.001,
      "loss": 2.4328,
      "step": 4180
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2686551809310913,
      "learning_rate": 0.001,
      "loss": 2.3893,
      "step": 4185
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2632417380809784,
      "learning_rate": 0.001,
      "loss": 2.4319,
      "step": 4190
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3223254084587097,
      "learning_rate": 0.001,
      "loss": 2.459,
      "step": 4195
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.31923434138298035,
      "learning_rate": 0.001,
      "loss": 2.4144,
      "step": 4200
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.28261134028434753,
      "learning_rate": 0.001,
      "loss": 2.4485,
      "step": 4205
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2726362943649292,
      "learning_rate": 0.001,
      "loss": 2.5626,
      "step": 4210
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2785876393318176,
      "learning_rate": 0.001,
      "loss": 2.4329,
      "step": 4215
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.28101497888565063,
      "learning_rate": 0.001,
      "loss": 2.4212,
      "step": 4220
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3232978284358978,
      "learning_rate": 0.001,
      "loss": 2.4316,
      "step": 4225
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.26223674416542053,
      "learning_rate": 0.001,
      "loss": 2.3672,
      "step": 4230
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.28561317920684814,
      "learning_rate": 0.001,
      "loss": 2.3806,
      "step": 4235
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2884681820869446,
      "learning_rate": 0.001,
      "loss": 2.4717,
      "step": 4240
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.27227768301963806,
      "learning_rate": 0.001,
      "loss": 2.4416,
      "step": 4245
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.272927850484848,
      "learning_rate": 0.001,
      "loss": 2.4401,
      "step": 4250
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.2963358461856842,
      "learning_rate": 0.001,
      "loss": 2.3382,
      "step": 4255
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.28761300444602966,
      "learning_rate": 0.001,
      "loss": 2.4503,
      "step": 4260
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.278731107711792,
      "learning_rate": 0.001,
      "loss": 2.4327,
      "step": 4265
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.25920960307121277,
      "learning_rate": 0.001,
      "loss": 2.3216,
      "step": 4270
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.31134849786758423,
      "learning_rate": 0.001,
      "loss": 2.3888,
      "step": 4275
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.30488523840904236,
      "learning_rate": 0.001,
      "loss": 2.4174,
      "step": 4280
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2729250192642212,
      "learning_rate": 0.001,
      "loss": 2.4526,
      "step": 4285
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.26827767491340637,
      "learning_rate": 0.001,
      "loss": 2.4263,
      "step": 4290
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.31564125418663025,
      "learning_rate": 0.001,
      "loss": 2.405,
      "step": 4295
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3122270405292511,
      "learning_rate": 0.001,
      "loss": 2.377,
      "step": 4300
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2852209508419037,
      "learning_rate": 0.001,
      "loss": 2.4195,
      "step": 4305
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2710750699043274,
      "learning_rate": 0.001,
      "loss": 2.4325,
      "step": 4310
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.26912280917167664,
      "learning_rate": 0.001,
      "loss": 2.3645,
      "step": 4315
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.32176169753074646,
      "learning_rate": 0.001,
      "loss": 2.4785,
      "step": 4320
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3187229633331299,
      "learning_rate": 0.001,
      "loss": 2.5005,
      "step": 4325
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3275264799594879,
      "learning_rate": 0.001,
      "loss": 2.4553,
      "step": 4330
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2660835087299347,
      "learning_rate": 0.001,
      "loss": 2.4242,
      "step": 4335
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.257010817527771,
      "learning_rate": 0.001,
      "loss": 2.5045,
      "step": 4340
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.274821013212204,
      "learning_rate": 0.001,
      "loss": 2.4028,
      "step": 4345
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2845490276813507,
      "learning_rate": 0.001,
      "loss": 2.4315,
      "step": 4350
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2754290997982025,
      "learning_rate": 0.001,
      "loss": 2.3878,
      "step": 4355
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2966209053993225,
      "learning_rate": 0.001,
      "loss": 2.4095,
      "step": 4360
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.29566988348960876,
      "learning_rate": 0.001,
      "loss": 2.4789,
      "step": 4365
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.27423152327537537,
      "learning_rate": 0.001,
      "loss": 2.3935,
      "step": 4370
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2520494759082794,
      "learning_rate": 0.001,
      "loss": 2.4523,
      "step": 4375
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2906912565231323,
      "learning_rate": 0.001,
      "loss": 2.4836,
      "step": 4380
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2802337408065796,
      "learning_rate": 0.001,
      "loss": 2.4883,
      "step": 4385
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2806122303009033,
      "learning_rate": 0.001,
      "loss": 2.4134,
      "step": 4390
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.27291759848594666,
      "learning_rate": 0.001,
      "loss": 2.4205,
      "step": 4395
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3132713735103607,
      "learning_rate": 0.001,
      "loss": 2.4312,
      "step": 4400
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3540079891681671,
      "learning_rate": 0.001,
      "loss": 2.42,
      "step": 4405
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.2768736183643341,
      "learning_rate": 0.001,
      "loss": 2.4446,
      "step": 4410
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.28580641746520996,
      "learning_rate": 0.001,
      "loss": 2.4914,
      "step": 4415
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.33273836970329285,
      "learning_rate": 0.001,
      "loss": 2.4596,
      "step": 4420
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.27512285113334656,
      "learning_rate": 0.001,
      "loss": 2.4612,
      "step": 4425
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.28600168228149414,
      "learning_rate": 0.001,
      "loss": 2.3727,
      "step": 4430
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.28686290979385376,
      "learning_rate": 0.001,
      "loss": 2.4356,
      "step": 4435
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30344080924987793,
      "learning_rate": 0.001,
      "loss": 2.5105,
      "step": 4440
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.34767141938209534,
      "learning_rate": 0.001,
      "loss": 2.4046,
      "step": 4445
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.28735655546188354,
      "learning_rate": 0.001,
      "loss": 2.3914,
      "step": 4450
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2845683693885803,
      "learning_rate": 0.001,
      "loss": 2.3604,
      "step": 4455
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3194609582424164,
      "learning_rate": 0.001,
      "loss": 2.4498,
      "step": 4460
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2727400064468384,
      "learning_rate": 0.001,
      "loss": 2.4558,
      "step": 4465
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.25680381059646606,
      "learning_rate": 0.001,
      "loss": 2.4769,
      "step": 4470
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.29016733169555664,
      "learning_rate": 0.001,
      "loss": 2.3797,
      "step": 4475
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.27482253313064575,
      "learning_rate": 0.001,
      "loss": 2.384,
      "step": 4480
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3203965723514557,
      "learning_rate": 0.001,
      "loss": 2.5246,
      "step": 4485
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.27123692631721497,
      "learning_rate": 0.001,
      "loss": 2.408,
      "step": 4490
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2830926477909088,
      "learning_rate": 0.001,
      "loss": 2.4685,
      "step": 4495
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2557945251464844,
      "learning_rate": 0.001,
      "loss": 2.4439,
      "step": 4500
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.30656105279922485,
      "learning_rate": 0.001,
      "loss": 2.3327,
      "step": 4505
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2598239481449127,
      "learning_rate": 0.001,
      "loss": 2.4493,
      "step": 4510
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.31256821751594543,
      "learning_rate": 0.001,
      "loss": 2.3842,
      "step": 4515
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3096126317977905,
      "learning_rate": 0.001,
      "loss": 2.4043,
      "step": 4520
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.29334813356399536,
      "learning_rate": 0.001,
      "loss": 2.4784,
      "step": 4525
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.270281046628952,
      "learning_rate": 0.001,
      "loss": 2.3645,
      "step": 4530
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2868260443210602,
      "learning_rate": 0.001,
      "loss": 2.4196,
      "step": 4535
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.26428335905075073,
      "learning_rate": 0.001,
      "loss": 2.3695,
      "step": 4540
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.28040027618408203,
      "learning_rate": 0.001,
      "loss": 2.4872,
      "step": 4545
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.26213884353637695,
      "learning_rate": 0.001,
      "loss": 2.3695,
      "step": 4550
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.27567023038864136,
      "learning_rate": 0.001,
      "loss": 2.3853,
      "step": 4555
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.29068320989608765,
      "learning_rate": 0.001,
      "loss": 2.3579,
      "step": 4560
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2767714262008667,
      "learning_rate": 0.001,
      "loss": 2.4266,
      "step": 4565
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.27447810769081116,
      "learning_rate": 0.001,
      "loss": 2.4005,
      "step": 4570
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.26019272208213806,
      "learning_rate": 0.001,
      "loss": 2.3552,
      "step": 4575
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.273270845413208,
      "learning_rate": 0.001,
      "loss": 2.3849,
      "step": 4580
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2733108699321747,
      "learning_rate": 0.001,
      "loss": 2.3321,
      "step": 4585
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.26178601384162903,
      "learning_rate": 0.001,
      "loss": 2.4505,
      "step": 4590
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.25937095284461975,
      "learning_rate": 0.001,
      "loss": 2.3453,
      "step": 4595
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.27965015172958374,
      "learning_rate": 0.001,
      "loss": 2.4615,
      "step": 4600
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2662375867366791,
      "learning_rate": 0.001,
      "loss": 2.3325,
      "step": 4605
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2801031470298767,
      "learning_rate": 0.001,
      "loss": 2.4223,
      "step": 4610
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.2957525849342346,
      "learning_rate": 0.001,
      "loss": 2.4036,
      "step": 4615
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.28883954882621765,
      "learning_rate": 0.001,
      "loss": 2.5068,
      "step": 4620
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2551986575126648,
      "learning_rate": 0.001,
      "loss": 2.405,
      "step": 4625
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2544376850128174,
      "learning_rate": 0.001,
      "loss": 2.4211,
      "step": 4630
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2596908211708069,
      "learning_rate": 0.001,
      "loss": 2.3166,
      "step": 4635
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.285861074924469,
      "learning_rate": 0.001,
      "loss": 2.3782,
      "step": 4640
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2784794270992279,
      "learning_rate": 0.001,
      "loss": 2.4322,
      "step": 4645
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2579992413520813,
      "learning_rate": 0.001,
      "loss": 2.4158,
      "step": 4650
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.27722522616386414,
      "learning_rate": 0.001,
      "loss": 2.4386,
      "step": 4655
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2655174732208252,
      "learning_rate": 0.001,
      "loss": 2.4496,
      "step": 4660
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.27302199602127075,
      "learning_rate": 0.001,
      "loss": 2.5056,
      "step": 4665
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.26454782485961914,
      "learning_rate": 0.001,
      "loss": 2.3612,
      "step": 4670
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.267187237739563,
      "learning_rate": 0.001,
      "loss": 2.4129,
      "step": 4675
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2863868772983551,
      "learning_rate": 0.001,
      "loss": 2.4469,
      "step": 4680
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.2754268944263458,
      "learning_rate": 0.001,
      "loss": 2.3578,
      "step": 4685
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.25754669308662415,
      "learning_rate": 0.001,
      "loss": 2.4504,
      "step": 4690
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.26824602484703064,
      "learning_rate": 0.001,
      "loss": 2.4302,
      "step": 4695
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.27403566241264343,
      "learning_rate": 0.001,
      "loss": 2.391,
      "step": 4700
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.248744934797287,
      "learning_rate": 0.001,
      "loss": 2.425,
      "step": 4705
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2514548599720001,
      "learning_rate": 0.001,
      "loss": 2.4186,
      "step": 4710
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2512078881263733,
      "learning_rate": 0.001,
      "loss": 2.3782,
      "step": 4715
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3069394528865814,
      "learning_rate": 0.001,
      "loss": 2.4268,
      "step": 4720
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3118564784526825,
      "learning_rate": 0.001,
      "loss": 2.4843,
      "step": 4725
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.33608558773994446,
      "learning_rate": 0.001,
      "loss": 2.3966,
      "step": 4730
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2619701027870178,
      "learning_rate": 0.001,
      "loss": 2.3888,
      "step": 4735
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.25927653908729553,
      "learning_rate": 0.001,
      "loss": 2.4065,
      "step": 4740
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.28429317474365234,
      "learning_rate": 0.001,
      "loss": 2.3549,
      "step": 4745
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.274682879447937,
      "learning_rate": 0.001,
      "loss": 2.2985,
      "step": 4750
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2774803638458252,
      "learning_rate": 0.001,
      "loss": 2.4104,
      "step": 4755
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.27073538303375244,
      "learning_rate": 0.001,
      "loss": 2.3434,
      "step": 4760
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.27918335795402527,
      "learning_rate": 0.001,
      "loss": 2.3986,
      "step": 4765
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2999061942100525,
      "learning_rate": 0.001,
      "loss": 2.3535,
      "step": 4770
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.26616016030311584,
      "learning_rate": 0.001,
      "loss": 2.3756,
      "step": 4775
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2599700093269348,
      "learning_rate": 0.001,
      "loss": 2.3859,
      "step": 4780
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2553490102291107,
      "learning_rate": 0.001,
      "loss": 2.3737,
      "step": 4785
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.329328328371048,
      "learning_rate": 0.001,
      "loss": 2.4573,
      "step": 4790
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.3233380913734436,
      "learning_rate": 0.001,
      "loss": 2.4289,
      "step": 4795
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3129531145095825,
      "learning_rate": 0.001,
      "loss": 2.4391,
      "step": 4800
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2748904228210449,
      "learning_rate": 0.001,
      "loss": 2.3729,
      "step": 4805
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2609279453754425,
      "learning_rate": 0.001,
      "loss": 2.461,
      "step": 4810
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.27427300810813904,
      "learning_rate": 0.001,
      "loss": 2.4313,
      "step": 4815
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2718832790851593,
      "learning_rate": 0.001,
      "loss": 2.3691,
      "step": 4820
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2804459035396576,
      "learning_rate": 0.001,
      "loss": 2.3728,
      "step": 4825
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.26752111315727234,
      "learning_rate": 0.001,
      "loss": 2.37,
      "step": 4830
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.27284350991249084,
      "learning_rate": 0.001,
      "loss": 2.4466,
      "step": 4835
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.24688100814819336,
      "learning_rate": 0.001,
      "loss": 2.4122,
      "step": 4840
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.27109193801879883,
      "learning_rate": 0.001,
      "loss": 2.3855,
      "step": 4845
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2605935335159302,
      "learning_rate": 0.001,
      "loss": 2.3628,
      "step": 4850
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2562985122203827,
      "learning_rate": 0.001,
      "loss": 2.4307,
      "step": 4855
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2749897837638855,
      "learning_rate": 0.001,
      "loss": 2.437,
      "step": 4860
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.23961135745048523,
      "learning_rate": 0.001,
      "loss": 2.2798,
      "step": 4865
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.25328007340431213,
      "learning_rate": 0.001,
      "loss": 2.3712,
      "step": 4870
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.23832069337368011,
      "learning_rate": 0.001,
      "loss": 2.3946,
      "step": 4875
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.2675718069076538,
      "learning_rate": 0.001,
      "loss": 2.304,
      "step": 4880
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.24774715304374695,
      "learning_rate": 0.001,
      "loss": 2.4343,
      "step": 4885
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.24921192228794098,
      "learning_rate": 0.001,
      "loss": 2.3984,
      "step": 4890
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3039065897464752,
      "learning_rate": 0.001,
      "loss": 2.3614,
      "step": 4895
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.27676519751548767,
      "learning_rate": 0.001,
      "loss": 2.3913,
      "step": 4900
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.26279640197753906,
      "learning_rate": 0.001,
      "loss": 2.3754,
      "step": 4905
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.27109283208847046,
      "learning_rate": 0.001,
      "loss": 2.3579,
      "step": 4910
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.3158806562423706,
      "learning_rate": 0.001,
      "loss": 2.3984,
      "step": 4915
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.26809123158454895,
      "learning_rate": 0.001,
      "loss": 2.3454,
      "step": 4920
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.24843958020210266,
      "learning_rate": 0.001,
      "loss": 2.3414,
      "step": 4925
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.23875649273395538,
      "learning_rate": 0.001,
      "loss": 2.2866,
      "step": 4930
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.25938940048217773,
      "learning_rate": 0.001,
      "loss": 2.2937,
      "step": 4935
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.27505698800086975,
      "learning_rate": 0.001,
      "loss": 2.3627,
      "step": 4940
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.29795336723327637,
      "learning_rate": 0.001,
      "loss": 2.3351,
      "step": 4945
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2589178681373596,
      "learning_rate": 0.001,
      "loss": 2.3451,
      "step": 4950
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2763518691062927,
      "learning_rate": 0.001,
      "loss": 2.5155,
      "step": 4955
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2927325367927551,
      "learning_rate": 0.001,
      "loss": 2.356,
      "step": 4960
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.26324424147605896,
      "learning_rate": 0.001,
      "loss": 2.3444,
      "step": 4965
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2642908990383148,
      "learning_rate": 0.001,
      "loss": 2.3295,
      "step": 4970
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.2945278286933899,
      "learning_rate": 0.001,
      "loss": 2.3377,
      "step": 4975
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2505294382572174,
      "learning_rate": 0.001,
      "loss": 2.3119,
      "step": 4980
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2456444352865219,
      "learning_rate": 0.001,
      "loss": 2.3599,
      "step": 4985
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.27345308661460876,
      "learning_rate": 0.001,
      "loss": 2.3649,
      "step": 4990
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.26777923107147217,
      "learning_rate": 0.001,
      "loss": 2.3936,
      "step": 4995
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2308703511953354,
      "learning_rate": 0.001,
      "loss": 2.3303,
      "step": 5000
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2660985589027405,
      "learning_rate": 0.001,
      "loss": 2.36,
      "step": 5005
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2525184750556946,
      "learning_rate": 0.001,
      "loss": 2.3628,
      "step": 5010
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2550134062767029,
      "learning_rate": 0.001,
      "loss": 2.4206,
      "step": 5015
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.24980038404464722,
      "learning_rate": 0.001,
      "loss": 2.3157,
      "step": 5020
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.27052590250968933,
      "learning_rate": 0.001,
      "loss": 2.3774,
      "step": 5025
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.26681166887283325,
      "learning_rate": 0.001,
      "loss": 2.4122,
      "step": 5030
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2528935670852661,
      "learning_rate": 0.001,
      "loss": 2.318,
      "step": 5035
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2519877552986145,
      "learning_rate": 0.001,
      "loss": 2.2664,
      "step": 5040
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2654362916946411,
      "learning_rate": 0.001,
      "loss": 2.2941,
      "step": 5045
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2554594576358795,
      "learning_rate": 0.001,
      "loss": 2.4286,
      "step": 5050
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.25834372639656067,
      "learning_rate": 0.001,
      "loss": 2.3451,
      "step": 5055
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.26128920912742615,
      "learning_rate": 0.001,
      "loss": 2.3545,
      "step": 5060
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2541617751121521,
      "learning_rate": 0.001,
      "loss": 2.3366,
      "step": 5065
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2685922384262085,
      "learning_rate": 0.001,
      "loss": 2.3235,
      "step": 5070
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2549839913845062,
      "learning_rate": 0.001,
      "loss": 2.4049,
      "step": 5075
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2501363158226013,
      "learning_rate": 0.001,
      "loss": 2.3329,
      "step": 5080
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2490965723991394,
      "learning_rate": 0.001,
      "loss": 2.4172,
      "step": 5085
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.23749250173568726,
      "learning_rate": 0.001,
      "loss": 2.2282,
      "step": 5090
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2824033498764038,
      "learning_rate": 0.001,
      "loss": 2.4628,
      "step": 5095
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2921418249607086,
      "learning_rate": 0.001,
      "loss": 2.341,
      "step": 5100
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3341614305973053,
      "learning_rate": 0.001,
      "loss": 2.3233,
      "step": 5105
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2599157392978668,
      "learning_rate": 0.001,
      "loss": 2.3244,
      "step": 5110
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2559368312358856,
      "learning_rate": 0.001,
      "loss": 2.4011,
      "step": 5115
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2784283757209778,
      "learning_rate": 0.001,
      "loss": 2.353,
      "step": 5120
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.25590980052948,
      "learning_rate": 0.001,
      "loss": 2.374,
      "step": 5125
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.29736170172691345,
      "learning_rate": 0.001,
      "loss": 2.3436,
      "step": 5130
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2628958225250244,
      "learning_rate": 0.001,
      "loss": 2.3212,
      "step": 5135
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.24243566393852234,
      "learning_rate": 0.001,
      "loss": 2.3772,
      "step": 5140
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.25928130745887756,
      "learning_rate": 0.001,
      "loss": 2.4252,
      "step": 5145
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.24278123676776886,
      "learning_rate": 0.001,
      "loss": 2.3692,
      "step": 5150
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.2492816299200058,
      "learning_rate": 0.001,
      "loss": 2.3512,
      "step": 5155
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.26422119140625,
      "learning_rate": 0.001,
      "loss": 2.3008,
      "step": 5160
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2409730702638626,
      "learning_rate": 0.001,
      "loss": 2.3404,
      "step": 5165
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.26327621936798096,
      "learning_rate": 0.001,
      "loss": 2.433,
      "step": 5170
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.3097192943096161,
      "learning_rate": 0.001,
      "loss": 2.3845,
      "step": 5175
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2360316962003708,
      "learning_rate": 0.001,
      "loss": 2.3833,
      "step": 5180
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.24150846898555756,
      "learning_rate": 0.001,
      "loss": 2.3128,
      "step": 5185
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.24739116430282593,
      "learning_rate": 0.001,
      "loss": 2.3132,
      "step": 5190
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2537684440612793,
      "learning_rate": 0.001,
      "loss": 2.4189,
      "step": 5195
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2589970529079437,
      "learning_rate": 0.001,
      "loss": 2.3695,
      "step": 5200
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.23109261691570282,
      "learning_rate": 0.001,
      "loss": 2.3342,
      "step": 5205
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.24093829095363617,
      "learning_rate": 0.001,
      "loss": 2.3929,
      "step": 5210
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.25576117634773254,
      "learning_rate": 0.001,
      "loss": 2.3288,
      "step": 5215
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.24346935749053955,
      "learning_rate": 0.001,
      "loss": 2.3516,
      "step": 5220
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.259819895029068,
      "learning_rate": 0.001,
      "loss": 2.3438,
      "step": 5225
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2514759302139282,
      "learning_rate": 0.001,
      "loss": 2.3943,
      "step": 5230
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.23569801449775696,
      "learning_rate": 0.001,
      "loss": 2.4217,
      "step": 5235
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2914365828037262,
      "learning_rate": 0.001,
      "loss": 2.4228,
      "step": 5240
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.2780640721321106,
      "learning_rate": 0.001,
      "loss": 2.4379,
      "step": 5245
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2788977026939392,
      "learning_rate": 0.001,
      "loss": 2.3052,
      "step": 5250
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.26812538504600525,
      "learning_rate": 0.001,
      "loss": 2.2674,
      "step": 5255
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.25950074195861816,
      "learning_rate": 0.001,
      "loss": 2.4089,
      "step": 5260
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.25828009843826294,
      "learning_rate": 0.001,
      "loss": 2.3674,
      "step": 5265
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.24739393591880798,
      "learning_rate": 0.001,
      "loss": 2.3111,
      "step": 5270
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2411363422870636,
      "learning_rate": 0.001,
      "loss": 2.4458,
      "step": 5275
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.23981954157352448,
      "learning_rate": 0.001,
      "loss": 2.3483,
      "step": 5280
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.24768058955669403,
      "learning_rate": 0.001,
      "loss": 2.3429,
      "step": 5285
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2422722429037094,
      "learning_rate": 0.001,
      "loss": 2.3193,
      "step": 5290
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.23856380581855774,
      "learning_rate": 0.001,
      "loss": 2.3493,
      "step": 5295
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2557425796985626,
      "learning_rate": 0.001,
      "loss": 2.2181,
      "step": 5300
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2540566623210907,
      "learning_rate": 0.001,
      "loss": 2.2677,
      "step": 5305
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.26815852522850037,
      "learning_rate": 0.001,
      "loss": 2.2979,
      "step": 5310
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.25368693470954895,
      "learning_rate": 0.001,
      "loss": 2.3508,
      "step": 5315
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2505514919757843,
      "learning_rate": 0.001,
      "loss": 2.3665,
      "step": 5320
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.27467647194862366,
      "learning_rate": 0.001,
      "loss": 2.2853,
      "step": 5325
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.25152960419654846,
      "learning_rate": 0.001,
      "loss": 2.3663,
      "step": 5330
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.24481120705604553,
      "learning_rate": 0.001,
      "loss": 2.4325,
      "step": 5335
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2353387176990509,
      "learning_rate": 0.001,
      "loss": 2.4131,
      "step": 5340
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.26953062415122986,
      "learning_rate": 0.001,
      "loss": 2.3549,
      "step": 5345
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.22940675914287567,
      "learning_rate": 0.001,
      "loss": 2.3911,
      "step": 5350
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.3119536340236664,
      "learning_rate": 0.001,
      "loss": 2.3921,
      "step": 5355
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.24042974412441254,
      "learning_rate": 0.001,
      "loss": 2.327,
      "step": 5360
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2505492866039276,
      "learning_rate": 0.001,
      "loss": 2.4174,
      "step": 5365
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2914923131465912,
      "learning_rate": 0.001,
      "loss": 2.2534,
      "step": 5370
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2598934769630432,
      "learning_rate": 0.001,
      "loss": 2.4125,
      "step": 5375
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.24274258315563202,
      "learning_rate": 0.001,
      "loss": 2.3404,
      "step": 5380
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.24509523808956146,
      "learning_rate": 0.001,
      "loss": 2.4063,
      "step": 5385
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.25125619769096375,
      "learning_rate": 0.001,
      "loss": 2.3473,
      "step": 5390
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2509467601776123,
      "learning_rate": 0.001,
      "loss": 2.3411,
      "step": 5395
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2541724145412445,
      "learning_rate": 0.001,
      "loss": 2.4116,
      "step": 5400
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.23472855985164642,
      "learning_rate": 0.001,
      "loss": 2.3242,
      "step": 5405
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.2244972288608551,
      "learning_rate": 0.001,
      "loss": 2.3185,
      "step": 5410
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.26157915592193604,
      "learning_rate": 0.001,
      "loss": 2.3682,
      "step": 5415
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.23372791707515717,
      "learning_rate": 0.001,
      "loss": 2.3788,
      "step": 5420
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.25492754578590393,
      "learning_rate": 0.001,
      "loss": 2.3125,
      "step": 5425
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2350878119468689,
      "learning_rate": 0.001,
      "loss": 2.3559,
      "step": 5430
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2786577343940735,
      "learning_rate": 0.001,
      "loss": 2.3551,
      "step": 5435
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.24445191025733948,
      "learning_rate": 0.001,
      "loss": 2.3195,
      "step": 5440
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2601463794708252,
      "learning_rate": 0.001,
      "loss": 2.3156,
      "step": 5445
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2302105575799942,
      "learning_rate": 0.001,
      "loss": 2.3936,
      "step": 5450
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.22858335077762604,
      "learning_rate": 0.001,
      "loss": 2.3565,
      "step": 5455
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.24476759135723114,
      "learning_rate": 0.001,
      "loss": 2.36,
      "step": 5460
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2723000943660736,
      "learning_rate": 0.001,
      "loss": 2.2823,
      "step": 5465
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.24212589859962463,
      "learning_rate": 0.001,
      "loss": 2.3353,
      "step": 5470
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.249808669090271,
      "learning_rate": 0.001,
      "loss": 2.361,
      "step": 5475
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.31722500920295715,
      "learning_rate": 0.001,
      "loss": 2.4057,
      "step": 5480
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2519189417362213,
      "learning_rate": 0.001,
      "loss": 2.3239,
      "step": 5485
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.24214482307434082,
      "learning_rate": 0.001,
      "loss": 2.3009,
      "step": 5490
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2553914785385132,
      "learning_rate": 0.001,
      "loss": 2.3704,
      "step": 5495
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2316461056470871,
      "learning_rate": 0.001,
      "loss": 2.3725,
      "step": 5500
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2363290637731552,
      "learning_rate": 0.001,
      "loss": 2.3258,
      "step": 5505
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2391221523284912,
      "learning_rate": 0.001,
      "loss": 2.3662,
      "step": 5510
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.25969377160072327,
      "learning_rate": 0.001,
      "loss": 2.344,
      "step": 5515
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.23705187439918518,
      "learning_rate": 0.001,
      "loss": 2.2742,
      "step": 5520
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.23285822570323944,
      "learning_rate": 0.001,
      "loss": 2.2341,
      "step": 5525
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.22560162842273712,
      "learning_rate": 0.001,
      "loss": 2.4168,
      "step": 5530
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.24206668138504028,
      "learning_rate": 0.001,
      "loss": 2.2389,
      "step": 5535
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.22780267894268036,
      "learning_rate": 0.001,
      "loss": 2.2827,
      "step": 5540
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.26254940032958984,
      "learning_rate": 0.001,
      "loss": 2.2634,
      "step": 5545
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2319927215576172,
      "learning_rate": 0.001,
      "loss": 2.3293,
      "step": 5550
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2692784070968628,
      "learning_rate": 0.001,
      "loss": 2.3751,
      "step": 5555
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.23394356667995453,
      "learning_rate": 0.001,
      "loss": 2.3735,
      "step": 5560
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.23533542454242706,
      "learning_rate": 0.001,
      "loss": 2.2469,
      "step": 5565
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.24933494627475739,
      "learning_rate": 0.001,
      "loss": 2.3927,
      "step": 5570
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.24868957698345184,
      "learning_rate": 0.001,
      "loss": 2.3305,
      "step": 5575
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.23358029127120972,
      "learning_rate": 0.001,
      "loss": 2.3149,
      "step": 5580
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2536250352859497,
      "learning_rate": 0.001,
      "loss": 2.26,
      "step": 5585
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2484007626771927,
      "learning_rate": 0.001,
      "loss": 2.4374,
      "step": 5590
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2672274708747864,
      "learning_rate": 0.001,
      "loss": 2.2977,
      "step": 5595
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.24046289920806885,
      "learning_rate": 0.001,
      "loss": 2.3444,
      "step": 5600
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.24538683891296387,
      "learning_rate": 0.001,
      "loss": 2.3603,
      "step": 5605
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.237552210688591,
      "learning_rate": 0.001,
      "loss": 2.3722,
      "step": 5610
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.22240518033504486,
      "learning_rate": 0.001,
      "loss": 2.2714,
      "step": 5615
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24909433722496033,
      "learning_rate": 0.001,
      "loss": 2.3317,
      "step": 5620
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.23186059296131134,
      "learning_rate": 0.001,
      "loss": 2.3821,
      "step": 5625
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.23126690089702606,
      "learning_rate": 0.001,
      "loss": 2.3512,
      "step": 5630
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2562063932418823,
      "learning_rate": 0.001,
      "loss": 2.3948,
      "step": 5635
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2485012412071228,
      "learning_rate": 0.001,
      "loss": 2.3371,
      "step": 5640
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24076102674007416,
      "learning_rate": 0.001,
      "loss": 2.395,
      "step": 5645
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2525179088115692,
      "learning_rate": 0.001,
      "loss": 2.4241,
      "step": 5650
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.23360827565193176,
      "learning_rate": 0.001,
      "loss": 2.3358,
      "step": 5655
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24136130511760712,
      "learning_rate": 0.001,
      "loss": 2.3938,
      "step": 5660
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2559976577758789,
      "learning_rate": 0.001,
      "loss": 2.3931,
      "step": 5665
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24539463222026825,
      "learning_rate": 0.001,
      "loss": 2.2946,
      "step": 5670
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24134691059589386,
      "learning_rate": 0.001,
      "loss": 2.3512,
      "step": 5675
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.22826063632965088,
      "learning_rate": 0.001,
      "loss": 2.3152,
      "step": 5680
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24264606833457947,
      "learning_rate": 0.001,
      "loss": 2.3635,
      "step": 5685
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.25732120871543884,
      "learning_rate": 0.001,
      "loss": 2.3409,
      "step": 5690
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.2741478681564331,
      "learning_rate": 0.001,
      "loss": 2.3171,
      "step": 5695
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.22876569628715515,
      "learning_rate": 0.001,
      "loss": 2.2749,
      "step": 5700
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2547457218170166,
      "learning_rate": 0.001,
      "loss": 2.3821,
      "step": 5705
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2326708734035492,
      "learning_rate": 0.001,
      "loss": 2.3686,
      "step": 5710
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2405540645122528,
      "learning_rate": 0.001,
      "loss": 2.3483,
      "step": 5715
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2408895492553711,
      "learning_rate": 0.001,
      "loss": 2.2978,
      "step": 5720
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.23036791384220123,
      "learning_rate": 0.001,
      "loss": 2.3084,
      "step": 5725
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.22999659180641174,
      "learning_rate": 0.001,
      "loss": 2.3285,
      "step": 5730
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.23441249132156372,
      "learning_rate": 0.001,
      "loss": 2.3568,
      "step": 5735
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2443058043718338,
      "learning_rate": 0.001,
      "loss": 2.3631,
      "step": 5740
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2505609393119812,
      "learning_rate": 0.001,
      "loss": 2.4046,
      "step": 5745
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.27520403265953064,
      "learning_rate": 0.001,
      "loss": 2.3677,
      "step": 5750
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.23651844263076782,
      "learning_rate": 0.001,
      "loss": 2.404,
      "step": 5755
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2234681099653244,
      "learning_rate": 0.001,
      "loss": 2.2937,
      "step": 5760
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2529265880584717,
      "learning_rate": 0.001,
      "loss": 2.2717,
      "step": 5765
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.22382284700870514,
      "learning_rate": 0.001,
      "loss": 2.3432,
      "step": 5770
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.23001866042613983,
      "learning_rate": 0.001,
      "loss": 2.3937,
      "step": 5775
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.22797128558158875,
      "learning_rate": 0.001,
      "loss": 2.3402,
      "step": 5780
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.23282378911972046,
      "learning_rate": 0.001,
      "loss": 2.3246,
      "step": 5785
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.233421191573143,
      "learning_rate": 0.001,
      "loss": 2.3537,
      "step": 5790
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.22940601408481598,
      "learning_rate": 0.001,
      "loss": 2.3019,
      "step": 5795
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2530461251735687,
      "learning_rate": 0.001,
      "loss": 2.3856,
      "step": 5800
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.22957460582256317,
      "learning_rate": 0.001,
      "loss": 2.3155,
      "step": 5805
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.23455286026000977,
      "learning_rate": 0.001,
      "loss": 2.3144,
      "step": 5810
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.23876363039016724,
      "learning_rate": 0.001,
      "loss": 2.324,
      "step": 5815
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.22128742933273315,
      "learning_rate": 0.001,
      "loss": 2.2817,
      "step": 5820
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.24940499663352966,
      "learning_rate": 0.001,
      "loss": 2.3068,
      "step": 5825
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2494128942489624,
      "learning_rate": 0.001,
      "loss": 2.2576,
      "step": 5830
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.22584503889083862,
      "learning_rate": 0.001,
      "loss": 2.3599,
      "step": 5835
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2631079852581024,
      "learning_rate": 0.001,
      "loss": 2.272,
      "step": 5840
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2861349284648895,
      "learning_rate": 0.001,
      "loss": 2.4105,
      "step": 5845
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.25299036502838135,
      "learning_rate": 0.001,
      "loss": 2.4458,
      "step": 5850
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2369646430015564,
      "learning_rate": 0.001,
      "loss": 2.3539,
      "step": 5855
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.2523862421512604,
      "learning_rate": 0.001,
      "loss": 2.3134,
      "step": 5860
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.21714548766613007,
      "learning_rate": 0.001,
      "loss": 2.3339,
      "step": 5865
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.26811763644218445,
      "learning_rate": 0.001,
      "loss": 2.3447,
      "step": 5870
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.23169931769371033,
      "learning_rate": 0.001,
      "loss": 2.3459,
      "step": 5875
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.25803160667419434,
      "learning_rate": 0.001,
      "loss": 2.2959,
      "step": 5880
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.23524072766304016,
      "learning_rate": 0.001,
      "loss": 2.3536,
      "step": 5885
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.23739273846149445,
      "learning_rate": 0.001,
      "loss": 2.2686,
      "step": 5890
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.22635112702846527,
      "learning_rate": 0.001,
      "loss": 2.3628,
      "step": 5895
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2273004949092865,
      "learning_rate": 0.001,
      "loss": 2.2628,
      "step": 5900
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2222699224948883,
      "learning_rate": 0.001,
      "loss": 2.3277,
      "step": 5905
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2190581113100052,
      "learning_rate": 0.001,
      "loss": 2.3848,
      "step": 5910
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2414412945508957,
      "learning_rate": 0.001,
      "loss": 2.3162,
      "step": 5915
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.23341503739356995,
      "learning_rate": 0.001,
      "loss": 2.4144,
      "step": 5920
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2254437506198883,
      "learning_rate": 0.001,
      "loss": 2.4501,
      "step": 5925
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.24044445157051086,
      "learning_rate": 0.001,
      "loss": 2.3355,
      "step": 5930
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.225622296333313,
      "learning_rate": 0.001,
      "loss": 2.4403,
      "step": 5935
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.22940073907375336,
      "learning_rate": 0.001,
      "loss": 2.3084,
      "step": 5940
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.23389093577861786,
      "learning_rate": 0.001,
      "loss": 2.3437,
      "step": 5945
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2822270393371582,
      "learning_rate": 0.001,
      "loss": 2.3956,
      "step": 5950
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.22801753878593445,
      "learning_rate": 0.001,
      "loss": 2.344,
      "step": 5955
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.2555668354034424,
      "learning_rate": 0.001,
      "loss": 2.2703,
      "step": 5960
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.24699978530406952,
      "learning_rate": 0.001,
      "loss": 2.3371,
      "step": 5965
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.23384138941764832,
      "learning_rate": 0.001,
      "loss": 2.3097,
      "step": 5970
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.255615234375,
      "learning_rate": 0.001,
      "loss": 2.3041,
      "step": 5975
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.22265081107616425,
      "learning_rate": 0.001,
      "loss": 2.3446,
      "step": 5980
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.238100066781044,
      "learning_rate": 0.001,
      "loss": 2.3431,
      "step": 5985
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.22238829731941223,
      "learning_rate": 0.001,
      "loss": 2.4082,
      "step": 5990
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.216730996966362,
      "learning_rate": 0.001,
      "loss": 2.3037,
      "step": 5995
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.22031597793102264,
      "learning_rate": 0.001,
      "loss": 2.3185,
      "step": 6000
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.21698999404907227,
      "learning_rate": 0.001,
      "loss": 2.3008,
      "step": 6005
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2661702632904053,
      "learning_rate": 0.001,
      "loss": 2.2595,
      "step": 6010
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2344740331172943,
      "learning_rate": 0.001,
      "loss": 2.3146,
      "step": 6015
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2401505559682846,
      "learning_rate": 0.001,
      "loss": 2.3003,
      "step": 6020
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2311105728149414,
      "learning_rate": 0.001,
      "loss": 2.3577,
      "step": 6025
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2426912486553192,
      "learning_rate": 0.001,
      "loss": 2.2576,
      "step": 6030
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.23958609998226166,
      "learning_rate": 0.001,
      "loss": 2.2899,
      "step": 6035
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.23275497555732727,
      "learning_rate": 0.001,
      "loss": 2.3389,
      "step": 6040
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2336193025112152,
      "learning_rate": 0.001,
      "loss": 2.2513,
      "step": 6045
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.23353838920593262,
      "learning_rate": 0.001,
      "loss": 2.3194,
      "step": 6050
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2304724007844925,
      "learning_rate": 0.001,
      "loss": 2.3508,
      "step": 6055
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2317655384540558,
      "learning_rate": 0.001,
      "loss": 2.2612,
      "step": 6060
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.25610777735710144,
      "learning_rate": 0.001,
      "loss": 2.4149,
      "step": 6065
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2321697473526001,
      "learning_rate": 0.001,
      "loss": 2.332,
      "step": 6070
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.23013491928577423,
      "learning_rate": 0.001,
      "loss": 2.2204,
      "step": 6075
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2328411042690277,
      "learning_rate": 0.001,
      "loss": 2.3794,
      "step": 6080
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2270471155643463,
      "learning_rate": 0.001,
      "loss": 2.2641,
      "step": 6085
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2320142686367035,
      "learning_rate": 0.001,
      "loss": 2.3091,
      "step": 6090
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.24872182309627533,
      "learning_rate": 0.001,
      "loss": 2.2783,
      "step": 6095
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.249810591340065,
      "learning_rate": 0.001,
      "loss": 2.3178,
      "step": 6100
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2650211453437805,
      "learning_rate": 0.001,
      "loss": 2.2551,
      "step": 6105
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.23337675631046295,
      "learning_rate": 0.001,
      "loss": 2.381,
      "step": 6110
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.23248009383678436,
      "learning_rate": 0.001,
      "loss": 2.3057,
      "step": 6115
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2519432604312897,
      "learning_rate": 0.001,
      "loss": 2.2995,
      "step": 6120
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.23362381756305695,
      "learning_rate": 0.001,
      "loss": 2.3268,
      "step": 6125
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.23035821318626404,
      "learning_rate": 0.001,
      "loss": 2.2877,
      "step": 6130
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2515948712825775,
      "learning_rate": 0.001,
      "loss": 2.2686,
      "step": 6135
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.22109408676624298,
      "learning_rate": 0.001,
      "loss": 2.3087,
      "step": 6140
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2815423905849457,
      "learning_rate": 0.001,
      "loss": 2.2718,
      "step": 6145
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.27658113837242126,
      "learning_rate": 0.001,
      "loss": 2.2606,
      "step": 6150
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2339669018983841,
      "learning_rate": 0.001,
      "loss": 2.2081,
      "step": 6155
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.22394266724586487,
      "learning_rate": 0.001,
      "loss": 2.2434,
      "step": 6160
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.274379700422287,
      "learning_rate": 0.001,
      "loss": 2.3479,
      "step": 6165
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.23967401683330536,
      "learning_rate": 0.001,
      "loss": 2.269,
      "step": 6170
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.22441859543323517,
      "learning_rate": 0.001,
      "loss": 2.4106,
      "step": 6175
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.22732356190681458,
      "learning_rate": 0.001,
      "loss": 2.2489,
      "step": 6180
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2360657900571823,
      "learning_rate": 0.001,
      "loss": 2.3203,
      "step": 6185
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.23805424571037292,
      "learning_rate": 0.001,
      "loss": 2.365,
      "step": 6190
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.23591791093349457,
      "learning_rate": 0.001,
      "loss": 2.3065,
      "step": 6195
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2685248553752899,
      "learning_rate": 0.001,
      "loss": 2.2762,
      "step": 6200
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.21762576699256897,
      "learning_rate": 0.001,
      "loss": 2.2622,
      "step": 6205
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.22365742921829224,
      "learning_rate": 0.001,
      "loss": 2.3262,
      "step": 6210
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2266640067100525,
      "learning_rate": 0.001,
      "loss": 2.1798,
      "step": 6215
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2408314198255539,
      "learning_rate": 0.001,
      "loss": 2.4467,
      "step": 6220
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2335503101348877,
      "learning_rate": 0.001,
      "loss": 2.2623,
      "step": 6225
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.23915047943592072,
      "learning_rate": 0.001,
      "loss": 2.3316,
      "step": 6230
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.23060446977615356,
      "learning_rate": 0.001,
      "loss": 2.2405,
      "step": 6235
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.23039406538009644,
      "learning_rate": 0.001,
      "loss": 2.3572,
      "step": 6240
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2337615042924881,
      "learning_rate": 0.001,
      "loss": 2.2782,
      "step": 6245
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.22629192471504211,
      "learning_rate": 0.001,
      "loss": 2.344,
      "step": 6250
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.24617376923561096,
      "learning_rate": 0.001,
      "loss": 2.3693,
      "step": 6255
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.22347915172576904,
      "learning_rate": 0.001,
      "loss": 2.2883,
      "step": 6260
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.22970449924468994,
      "learning_rate": 0.001,
      "loss": 2.3835,
      "step": 6265
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.23593968152999878,
      "learning_rate": 0.001,
      "loss": 2.3482,
      "step": 6270
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.22513240575790405,
      "learning_rate": 0.001,
      "loss": 2.2519,
      "step": 6275
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.21264764666557312,
      "learning_rate": 0.001,
      "loss": 2.3257,
      "step": 6280
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2323000133037567,
      "learning_rate": 0.001,
      "loss": 2.2972,
      "step": 6285
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.22790469229221344,
      "learning_rate": 0.001,
      "loss": 2.2954,
      "step": 6290
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.21939660608768463,
      "learning_rate": 0.001,
      "loss": 2.2985,
      "step": 6295
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2272881269454956,
      "learning_rate": 0.001,
      "loss": 2.2307,
      "step": 6300
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2288762778043747,
      "learning_rate": 0.001,
      "loss": 2.3405,
      "step": 6305
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.233880415558815,
      "learning_rate": 0.001,
      "loss": 2.2826,
      "step": 6310
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.22657474875450134,
      "learning_rate": 0.001,
      "loss": 2.3034,
      "step": 6315
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.2148173451423645,
      "learning_rate": 0.001,
      "loss": 2.307,
      "step": 6320
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2334657907485962,
      "learning_rate": 0.001,
      "loss": 2.3141,
      "step": 6325
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2156243771314621,
      "learning_rate": 0.001,
      "loss": 2.2665,
      "step": 6330
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2339843213558197,
      "learning_rate": 0.001,
      "loss": 2.2798,
      "step": 6335
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.22493353486061096,
      "learning_rate": 0.001,
      "loss": 2.372,
      "step": 6340
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.23672647774219513,
      "learning_rate": 0.001,
      "loss": 2.2306,
      "step": 6345
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.22123226523399353,
      "learning_rate": 0.001,
      "loss": 2.2672,
      "step": 6350
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.22304557263851166,
      "learning_rate": 0.001,
      "loss": 2.2178,
      "step": 6355
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2457527369260788,
      "learning_rate": 0.001,
      "loss": 2.2743,
      "step": 6360
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.21300479769706726,
      "learning_rate": 0.001,
      "loss": 2.2388,
      "step": 6365
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2059631049633026,
      "learning_rate": 0.001,
      "loss": 2.2377,
      "step": 6370
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.24631153047084808,
      "learning_rate": 0.001,
      "loss": 2.2768,
      "step": 6375
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.21497204899787903,
      "learning_rate": 0.001,
      "loss": 2.2844,
      "step": 6380
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.22930744290351868,
      "learning_rate": 0.001,
      "loss": 2.312,
      "step": 6385
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.23210054636001587,
      "learning_rate": 0.001,
      "loss": 2.2118,
      "step": 6390
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.21625350415706635,
      "learning_rate": 0.001,
      "loss": 2.3245,
      "step": 6395
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.23386377096176147,
      "learning_rate": 0.001,
      "loss": 2.3134,
      "step": 6400
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.21839016675949097,
      "learning_rate": 0.001,
      "loss": 2.3539,
      "step": 6405
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.2133498340845108,
      "learning_rate": 0.001,
      "loss": 2.2994,
      "step": 6410
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.21773594617843628,
      "learning_rate": 0.001,
      "loss": 2.3248,
      "step": 6415
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.23620709776878357,
      "learning_rate": 0.001,
      "loss": 2.2653,
      "step": 6420
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.22743947803974152,
      "learning_rate": 0.001,
      "loss": 2.2736,
      "step": 6425
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.2404935657978058,
      "learning_rate": 0.001,
      "loss": 2.3361,
      "step": 6430
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.24400784075260162,
      "learning_rate": 0.001,
      "loss": 2.3089,
      "step": 6435
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.23440074920654297,
      "learning_rate": 0.001,
      "loss": 2.2961,
      "step": 6440
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.22357019782066345,
      "learning_rate": 0.001,
      "loss": 2.3432,
      "step": 6445
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.2347543090581894,
      "learning_rate": 0.001,
      "loss": 2.304,
      "step": 6450
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.22234396636486053,
      "learning_rate": 0.001,
      "loss": 2.2524,
      "step": 6455
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.25612977147102356,
      "learning_rate": 0.001,
      "loss": 2.304,
      "step": 6460
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.23432089388370514,
      "learning_rate": 0.001,
      "loss": 2.3512,
      "step": 6465
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.23479154706001282,
      "learning_rate": 0.001,
      "loss": 2.243,
      "step": 6470
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.2278670221567154,
      "learning_rate": 0.001,
      "loss": 2.3402,
      "step": 6475
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.2140948623418808,
      "learning_rate": 0.001,
      "loss": 2.2892,
      "step": 6480
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.23153026401996613,
      "learning_rate": 0.001,
      "loss": 2.2849,
      "step": 6485
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.22485770285129547,
      "learning_rate": 0.001,
      "loss": 2.313,
      "step": 6490
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.21722319722175598,
      "learning_rate": 0.001,
      "loss": 2.3453,
      "step": 6495
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.21437281370162964,
      "learning_rate": 0.001,
      "loss": 2.2887,
      "step": 6500
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.214768648147583,
      "learning_rate": 0.001,
      "loss": 2.2759,
      "step": 6505
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.31841498613357544,
      "learning_rate": 0.001,
      "loss": 2.2481,
      "step": 6510
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.22883611917495728,
      "learning_rate": 0.001,
      "loss": 2.2809,
      "step": 6515
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.23380115628242493,
      "learning_rate": 0.001,
      "loss": 2.3128,
      "step": 6520
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.21790826320648193,
      "learning_rate": 0.001,
      "loss": 2.3074,
      "step": 6525
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.22439299523830414,
      "learning_rate": 0.001,
      "loss": 2.3287,
      "step": 6530
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.21973837912082672,
      "learning_rate": 0.001,
      "loss": 2.3249,
      "step": 6535
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.25569167733192444,
      "learning_rate": 0.001,
      "loss": 2.2878,
      "step": 6540
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.21167606115341187,
      "learning_rate": 0.001,
      "loss": 2.2877,
      "step": 6545
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2405991554260254,
      "learning_rate": 0.001,
      "loss": 2.3446,
      "step": 6550
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2218019813299179,
      "learning_rate": 0.001,
      "loss": 2.315,
      "step": 6555
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2425190806388855,
      "learning_rate": 0.001,
      "loss": 2.2657,
      "step": 6560
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.22708062827587128,
      "learning_rate": 0.001,
      "loss": 2.1857,
      "step": 6565
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2227758914232254,
      "learning_rate": 0.001,
      "loss": 2.2147,
      "step": 6570
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.22275353968143463,
      "learning_rate": 0.001,
      "loss": 2.2327,
      "step": 6575
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2267688810825348,
      "learning_rate": 0.001,
      "loss": 2.3549,
      "step": 6580
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2187042087316513,
      "learning_rate": 0.001,
      "loss": 2.302,
      "step": 6585
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.23258638381958008,
      "learning_rate": 0.001,
      "loss": 2.2425,
      "step": 6590
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.25112131237983704,
      "learning_rate": 0.001,
      "loss": 2.3606,
      "step": 6595
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2182377427816391,
      "learning_rate": 0.001,
      "loss": 2.2458,
      "step": 6600
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.24733804166316986,
      "learning_rate": 0.001,
      "loss": 2.3269,
      "step": 6605
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.20609302818775177,
      "learning_rate": 0.001,
      "loss": 2.2328,
      "step": 6610
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.20930394530296326,
      "learning_rate": 0.001,
      "loss": 2.3236,
      "step": 6615
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.22153881192207336,
      "learning_rate": 0.001,
      "loss": 2.2531,
      "step": 6620
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2152048796415329,
      "learning_rate": 0.001,
      "loss": 2.407,
      "step": 6625
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.255520224571228,
      "learning_rate": 0.001,
      "loss": 2.2682,
      "step": 6630
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.23848092555999756,
      "learning_rate": 0.001,
      "loss": 2.3644,
      "step": 6635
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.20903976261615753,
      "learning_rate": 0.001,
      "loss": 2.2142,
      "step": 6640
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.24808953702449799,
      "learning_rate": 0.001,
      "loss": 2.4207,
      "step": 6645
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2541098892688751,
      "learning_rate": 0.001,
      "loss": 2.3043,
      "step": 6650
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2177581936120987,
      "learning_rate": 0.001,
      "loss": 2.3392,
      "step": 6655
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.21430838108062744,
      "learning_rate": 0.001,
      "loss": 2.2786,
      "step": 6660
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.22286362946033478,
      "learning_rate": 0.001,
      "loss": 2.4054,
      "step": 6665
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.2035175859928131,
      "learning_rate": 0.001,
      "loss": 2.3528,
      "step": 6670
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.24235355854034424,
      "learning_rate": 0.001,
      "loss": 2.2776,
      "step": 6675
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.21710264682769775,
      "learning_rate": 0.001,
      "loss": 2.1995,
      "step": 6680
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.20540495216846466,
      "learning_rate": 0.001,
      "loss": 2.262,
      "step": 6685
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22396127879619598,
      "learning_rate": 0.001,
      "loss": 2.2918,
      "step": 6690
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2071615308523178,
      "learning_rate": 0.001,
      "loss": 2.2222,
      "step": 6695
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22722716629505157,
      "learning_rate": 0.001,
      "loss": 2.268,
      "step": 6700
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22743773460388184,
      "learning_rate": 0.001,
      "loss": 2.2728,
      "step": 6705
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.23747438192367554,
      "learning_rate": 0.001,
      "loss": 2.2671,
      "step": 6710
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.21889004111289978,
      "learning_rate": 0.001,
      "loss": 2.2856,
      "step": 6715
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.19960670173168182,
      "learning_rate": 0.001,
      "loss": 2.308,
      "step": 6720
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2208574414253235,
      "learning_rate": 0.001,
      "loss": 2.2424,
      "step": 6725
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2305884212255478,
      "learning_rate": 0.001,
      "loss": 2.2329,
      "step": 6730
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.2227226197719574,
      "learning_rate": 0.001,
      "loss": 2.3162,
      "step": 6735
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.21388469636440277,
      "learning_rate": 0.001,
      "loss": 2.2611,
      "step": 6740
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.21388593316078186,
      "learning_rate": 0.001,
      "loss": 2.2672,
      "step": 6745
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22708441317081451,
      "learning_rate": 0.001,
      "loss": 2.2176,
      "step": 6750
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.21498042345046997,
      "learning_rate": 0.001,
      "loss": 2.2735,
      "step": 6755
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.26390013098716736,
      "learning_rate": 0.001,
      "loss": 2.3951,
      "step": 6760
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22930343449115753,
      "learning_rate": 0.001,
      "loss": 2.2718,
      "step": 6765
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.22452136874198914,
      "learning_rate": 0.001,
      "loss": 2.2614,
      "step": 6770
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2295997589826584,
      "learning_rate": 0.001,
      "loss": 2.2575,
      "step": 6775
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2150869071483612,
      "learning_rate": 0.001,
      "loss": 2.2893,
      "step": 6780
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.22733640670776367,
      "learning_rate": 0.001,
      "loss": 2.2378,
      "step": 6785
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2284800410270691,
      "learning_rate": 0.001,
      "loss": 2.2126,
      "step": 6790
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.22430431842803955,
      "learning_rate": 0.001,
      "loss": 2.2428,
      "step": 6795
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2072402387857437,
      "learning_rate": 0.001,
      "loss": 2.278,
      "step": 6800
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.23210042715072632,
      "learning_rate": 0.001,
      "loss": 2.2302,
      "step": 6805
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.21655841171741486,
      "learning_rate": 0.001,
      "loss": 2.2909,
      "step": 6810
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.21356946229934692,
      "learning_rate": 0.001,
      "loss": 2.2277,
      "step": 6815
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.21072253584861755,
      "learning_rate": 0.001,
      "loss": 2.2721,
      "step": 6820
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.21372920274734497,
      "learning_rate": 0.001,
      "loss": 2.3177,
      "step": 6825
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.23465748131275177,
      "learning_rate": 0.001,
      "loss": 2.205,
      "step": 6830
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.20318353176116943,
      "learning_rate": 0.001,
      "loss": 2.2573,
      "step": 6835
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.22107534110546112,
      "learning_rate": 0.001,
      "loss": 2.2434,
      "step": 6840
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2000688910484314,
      "learning_rate": 0.001,
      "loss": 2.3421,
      "step": 6845
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.24129514396190643,
      "learning_rate": 0.001,
      "loss": 2.2897,
      "step": 6850
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.21887139976024628,
      "learning_rate": 0.001,
      "loss": 2.3154,
      "step": 6855
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.2198597937822342,
      "learning_rate": 0.001,
      "loss": 2.3218,
      "step": 6860
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.219230517745018,
      "learning_rate": 0.001,
      "loss": 2.3299,
      "step": 6865
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2002941071987152,
      "learning_rate": 0.001,
      "loss": 2.2028,
      "step": 6870
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.22916488349437714,
      "learning_rate": 0.001,
      "loss": 2.2602,
      "step": 6875
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.20966394245624542,
      "learning_rate": 0.001,
      "loss": 2.3592,
      "step": 6880
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.22105638682842255,
      "learning_rate": 0.001,
      "loss": 2.1895,
      "step": 6885
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.23469582200050354,
      "learning_rate": 0.001,
      "loss": 2.2608,
      "step": 6890
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.21237854659557343,
      "learning_rate": 0.001,
      "loss": 2.3013,
      "step": 6895
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.22927138209342957,
      "learning_rate": 0.001,
      "loss": 2.3076,
      "step": 6900
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.21441708505153656,
      "learning_rate": 0.001,
      "loss": 2.2768,
      "step": 6905
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.23027388751506805,
      "learning_rate": 0.001,
      "loss": 2.3414,
      "step": 6910
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.20745432376861572,
      "learning_rate": 0.001,
      "loss": 2.3512,
      "step": 6915
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.21277524530887604,
      "learning_rate": 0.001,
      "loss": 2.2667,
      "step": 6920
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.20173713564872742,
      "learning_rate": 0.001,
      "loss": 2.2563,
      "step": 6925
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.20181357860565186,
      "learning_rate": 0.001,
      "loss": 2.3219,
      "step": 6930
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.22798535227775574,
      "learning_rate": 0.001,
      "loss": 2.2603,
      "step": 6935
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2111693024635315,
      "learning_rate": 0.001,
      "loss": 2.2972,
      "step": 6940
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.225101500749588,
      "learning_rate": 0.001,
      "loss": 2.2476,
      "step": 6945
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.2258484959602356,
      "learning_rate": 0.001,
      "loss": 2.3323,
      "step": 6950
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2221709042787552,
      "learning_rate": 0.001,
      "loss": 2.3256,
      "step": 6955
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.21151743829250336,
      "learning_rate": 0.001,
      "loss": 2.2582,
      "step": 6960
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.22967129945755005,
      "learning_rate": 0.001,
      "loss": 2.2557,
      "step": 6965
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2297360748052597,
      "learning_rate": 0.001,
      "loss": 2.3392,
      "step": 6970
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.23820990324020386,
      "learning_rate": 0.001,
      "loss": 2.2955,
      "step": 6975
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2055855095386505,
      "learning_rate": 0.001,
      "loss": 2.2823,
      "step": 6980
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2106105536222458,
      "learning_rate": 0.001,
      "loss": 2.3535,
      "step": 6985
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2504548132419586,
      "learning_rate": 0.001,
      "loss": 2.3398,
      "step": 6990
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.20728346705436707,
      "learning_rate": 0.001,
      "loss": 2.2791,
      "step": 6995
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2041543871164322,
      "learning_rate": 0.001,
      "loss": 2.2221,
      "step": 7000
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.23212693631649017,
      "learning_rate": 0.001,
      "loss": 2.2897,
      "step": 7005
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.20440858602523804,
      "learning_rate": 0.001,
      "loss": 2.2188,
      "step": 7010
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.21270564198493958,
      "learning_rate": 0.001,
      "loss": 2.2856,
      "step": 7015
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.23255744576454163,
      "learning_rate": 0.001,
      "loss": 2.2796,
      "step": 7020
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.22522412240505219,
      "learning_rate": 0.001,
      "loss": 2.2653,
      "step": 7025
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.21376082301139832,
      "learning_rate": 0.001,
      "loss": 2.2828,
      "step": 7030
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2234581708908081,
      "learning_rate": 0.001,
      "loss": 2.2241,
      "step": 7035
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.22096699476242065,
      "learning_rate": 0.001,
      "loss": 2.2399,
      "step": 7040
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.23003298044204712,
      "learning_rate": 0.001,
      "loss": 2.3065,
      "step": 7045
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.22232045233249664,
      "learning_rate": 0.001,
      "loss": 2.1878,
      "step": 7050
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.21559427678585052,
      "learning_rate": 0.001,
      "loss": 2.1721,
      "step": 7055
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2325166016817093,
      "learning_rate": 0.001,
      "loss": 2.3431,
      "step": 7060
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2172939032316208,
      "learning_rate": 0.001,
      "loss": 2.2579,
      "step": 7065
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.21909840404987335,
      "learning_rate": 0.001,
      "loss": 2.2747,
      "step": 7070
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.21307264268398285,
      "learning_rate": 0.001,
      "loss": 2.2371,
      "step": 7075
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1988152116537094,
      "learning_rate": 0.001,
      "loss": 2.2615,
      "step": 7080
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.20012176036834717,
      "learning_rate": 0.001,
      "loss": 2.2191,
      "step": 7085
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.21884466707706451,
      "learning_rate": 0.001,
      "loss": 2.2541,
      "step": 7090
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2405450940132141,
      "learning_rate": 0.001,
      "loss": 2.1851,
      "step": 7095
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.20957379043102264,
      "learning_rate": 0.001,
      "loss": 2.29,
      "step": 7100
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2217535525560379,
      "learning_rate": 0.001,
      "loss": 2.1965,
      "step": 7105
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.19948221743106842,
      "learning_rate": 0.001,
      "loss": 2.1932,
      "step": 7110
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2090318202972412,
      "learning_rate": 0.001,
      "loss": 2.2536,
      "step": 7115
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.22281096875667572,
      "learning_rate": 0.001,
      "loss": 2.2951,
      "step": 7120
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.20826506614685059,
      "learning_rate": 0.001,
      "loss": 2.3136,
      "step": 7125
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2151436060667038,
      "learning_rate": 0.001,
      "loss": 2.4252,
      "step": 7130
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.213445782661438,
      "learning_rate": 0.001,
      "loss": 2.2703,
      "step": 7135
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.20548775792121887,
      "learning_rate": 0.001,
      "loss": 2.2282,
      "step": 7140
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.23199157416820526,
      "learning_rate": 0.001,
      "loss": 2.3681,
      "step": 7145
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.21607224643230438,
      "learning_rate": 0.001,
      "loss": 2.3197,
      "step": 7150
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.20880939066410065,
      "learning_rate": 0.001,
      "loss": 2.2826,
      "step": 7155
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.21946680545806885,
      "learning_rate": 0.001,
      "loss": 2.2451,
      "step": 7160
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.23108704388141632,
      "learning_rate": 0.001,
      "loss": 2.244,
      "step": 7165
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.19656045734882355,
      "learning_rate": 0.001,
      "loss": 2.2871,
      "step": 7170
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.22790350019931793,
      "learning_rate": 0.001,
      "loss": 2.3187,
      "step": 7175
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.23438240587711334,
      "learning_rate": 0.001,
      "loss": 2.2798,
      "step": 7180
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2033395767211914,
      "learning_rate": 0.001,
      "loss": 2.1845,
      "step": 7185
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.20368583500385284,
      "learning_rate": 0.001,
      "loss": 2.3395,
      "step": 7190
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.20549830794334412,
      "learning_rate": 0.001,
      "loss": 2.3311,
      "step": 7195
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.22930066287517548,
      "learning_rate": 0.001,
      "loss": 2.2961,
      "step": 7200
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.2134595811367035,
      "learning_rate": 0.001,
      "loss": 2.3287,
      "step": 7205
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.23215077817440033,
      "learning_rate": 0.001,
      "loss": 2.297,
      "step": 7210
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.19765566289424896,
      "learning_rate": 0.001,
      "loss": 2.2268,
      "step": 7215
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.24692144989967346,
      "learning_rate": 0.001,
      "loss": 2.3147,
      "step": 7220
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.30222055315971375,
      "learning_rate": 0.001,
      "loss": 2.2585,
      "step": 7225
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.19766630232334137,
      "learning_rate": 0.001,
      "loss": 2.327,
      "step": 7230
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.20692983269691467,
      "learning_rate": 0.001,
      "loss": 2.3028,
      "step": 7235
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.21192064881324768,
      "learning_rate": 0.001,
      "loss": 2.3742,
      "step": 7240
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.21428699791431427,
      "learning_rate": 0.001,
      "loss": 2.1737,
      "step": 7245
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.20385991036891937,
      "learning_rate": 0.001,
      "loss": 2.2721,
      "step": 7250
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.20821918547153473,
      "learning_rate": 0.001,
      "loss": 2.2602,
      "step": 7255
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.20553617179393768,
      "learning_rate": 0.001,
      "loss": 2.2719,
      "step": 7260
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.21833737194538116,
      "learning_rate": 0.001,
      "loss": 2.2495,
      "step": 7265
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.2193860560655594,
      "learning_rate": 0.001,
      "loss": 2.2582,
      "step": 7270
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.21469643712043762,
      "learning_rate": 0.001,
      "loss": 2.2987,
      "step": 7275
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.2157880663871765,
      "learning_rate": 0.001,
      "loss": 2.245,
      "step": 7280
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.20697729289531708,
      "learning_rate": 0.001,
      "loss": 2.3218,
      "step": 7285
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.23085835576057434,
      "learning_rate": 0.001,
      "loss": 2.2267,
      "step": 7290
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.21796423196792603,
      "learning_rate": 0.001,
      "loss": 2.2688,
      "step": 7295
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.24522145092487335,
      "learning_rate": 0.001,
      "loss": 2.3475,
      "step": 7300
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.203556627035141,
      "learning_rate": 0.001,
      "loss": 2.2861,
      "step": 7305
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2439495325088501,
      "learning_rate": 0.001,
      "loss": 2.2658,
      "step": 7310
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.22975510358810425,
      "learning_rate": 0.001,
      "loss": 2.1613,
      "step": 7315
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19879737496376038,
      "learning_rate": 0.001,
      "loss": 2.2234,
      "step": 7320
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2011747658252716,
      "learning_rate": 0.001,
      "loss": 2.2828,
      "step": 7325
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19531621038913727,
      "learning_rate": 0.001,
      "loss": 2.1786,
      "step": 7330
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2119741439819336,
      "learning_rate": 0.001,
      "loss": 2.2916,
      "step": 7335
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2018207460641861,
      "learning_rate": 0.001,
      "loss": 2.219,
      "step": 7340
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19718153774738312,
      "learning_rate": 0.001,
      "loss": 2.1588,
      "step": 7345
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19742237031459808,
      "learning_rate": 0.001,
      "loss": 2.2564,
      "step": 7350
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.20226089656352997,
      "learning_rate": 0.001,
      "loss": 2.289,
      "step": 7355
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19770477712154388,
      "learning_rate": 0.001,
      "loss": 2.2675,
      "step": 7360
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.20721545815467834,
      "learning_rate": 0.001,
      "loss": 2.1463,
      "step": 7365
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.18459871411323547,
      "learning_rate": 0.001,
      "loss": 2.1891,
      "step": 7370
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19960848987102509,
      "learning_rate": 0.001,
      "loss": 2.2649,
      "step": 7375
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2147081345319748,
      "learning_rate": 0.001,
      "loss": 2.2833,
      "step": 7380
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19740645587444305,
      "learning_rate": 0.001,
      "loss": 2.2549,
      "step": 7385
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19742794334888458,
      "learning_rate": 0.001,
      "loss": 2.2949,
      "step": 7390
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.19098269939422607,
      "learning_rate": 0.001,
      "loss": 2.2872,
      "step": 7395
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20648017525672913,
      "learning_rate": 0.001,
      "loss": 2.2777,
      "step": 7400
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.19683200120925903,
      "learning_rate": 0.001,
      "loss": 2.2212,
      "step": 7405
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.19511087238788605,
      "learning_rate": 0.001,
      "loss": 2.2753,
      "step": 7410
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20487697422504425,
      "learning_rate": 0.001,
      "loss": 2.2839,
      "step": 7415
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20138192176818848,
      "learning_rate": 0.001,
      "loss": 2.2486,
      "step": 7420
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20356020331382751,
      "learning_rate": 0.001,
      "loss": 2.1948,
      "step": 7425
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.21195770800113678,
      "learning_rate": 0.001,
      "loss": 2.2343,
      "step": 7430
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2272544503211975,
      "learning_rate": 0.001,
      "loss": 2.3143,
      "step": 7435
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20329448580741882,
      "learning_rate": 0.001,
      "loss": 2.2508,
      "step": 7440
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.19755227863788605,
      "learning_rate": 0.001,
      "loss": 2.3016,
      "step": 7445
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20580564439296722,
      "learning_rate": 0.001,
      "loss": 2.2072,
      "step": 7450
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.19712607562541962,
      "learning_rate": 0.001,
      "loss": 2.262,
      "step": 7455
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20621292293071747,
      "learning_rate": 0.001,
      "loss": 2.2982,
      "step": 7460
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2071629762649536,
      "learning_rate": 0.001,
      "loss": 2.2891,
      "step": 7465
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2150653451681137,
      "learning_rate": 0.001,
      "loss": 2.1865,
      "step": 7470
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.23920634388923645,
      "learning_rate": 0.001,
      "loss": 2.308,
      "step": 7475
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.21500001847743988,
      "learning_rate": 0.001,
      "loss": 2.2884,
      "step": 7480
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.21805007755756378,
      "learning_rate": 0.001,
      "loss": 2.1848,
      "step": 7485
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.1949496865272522,
      "learning_rate": 0.001,
      "loss": 2.1732,
      "step": 7490
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.21110284328460693,
      "learning_rate": 0.001,
      "loss": 2.2932,
      "step": 7495
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2048315703868866,
      "learning_rate": 0.001,
      "loss": 2.216,
      "step": 7500
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2010776698589325,
      "learning_rate": 0.001,
      "loss": 2.2891,
      "step": 7505
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.27239951491355896,
      "learning_rate": 0.001,
      "loss": 2.2777,
      "step": 7510
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2029271274805069,
      "learning_rate": 0.001,
      "loss": 2.2677,
      "step": 7515
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.21085187792778015,
      "learning_rate": 0.001,
      "loss": 2.2963,
      "step": 7520
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2122115045785904,
      "learning_rate": 0.001,
      "loss": 2.2493,
      "step": 7525
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.20089121162891388,
      "learning_rate": 0.001,
      "loss": 2.2474,
      "step": 7530
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.20639105141162872,
      "learning_rate": 0.001,
      "loss": 2.1943,
      "step": 7535
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.21445663273334503,
      "learning_rate": 0.001,
      "loss": 2.1288,
      "step": 7540
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.20437091588974,
      "learning_rate": 0.001,
      "loss": 2.1688,
      "step": 7545
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.23248904943466187,
      "learning_rate": 0.001,
      "loss": 2.2279,
      "step": 7550
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.34397122263908386,
      "learning_rate": 0.001,
      "loss": 2.326,
      "step": 7555
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2095007747411728,
      "learning_rate": 0.001,
      "loss": 2.2435,
      "step": 7560
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2087799459695816,
      "learning_rate": 0.001,
      "loss": 2.245,
      "step": 7565
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.18876142799854279,
      "learning_rate": 0.001,
      "loss": 2.1585,
      "step": 7570
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.19884787499904633,
      "learning_rate": 0.001,
      "loss": 2.2337,
      "step": 7575
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19892942905426025,
      "learning_rate": 0.001,
      "loss": 2.269,
      "step": 7580
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2039511501789093,
      "learning_rate": 0.001,
      "loss": 2.1756,
      "step": 7585
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2111709862947464,
      "learning_rate": 0.001,
      "loss": 2.2736,
      "step": 7590
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19605736434459686,
      "learning_rate": 0.001,
      "loss": 2.2364,
      "step": 7595
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.21033748984336853,
      "learning_rate": 0.001,
      "loss": 2.2014,
      "step": 7600
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2066594660282135,
      "learning_rate": 0.001,
      "loss": 2.1916,
      "step": 7605
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.21947403252124786,
      "learning_rate": 0.001,
      "loss": 2.2487,
      "step": 7610
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.20312318205833435,
      "learning_rate": 0.001,
      "loss": 2.2154,
      "step": 7615
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19939804077148438,
      "learning_rate": 0.001,
      "loss": 2.2305,
      "step": 7620
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19940166175365448,
      "learning_rate": 0.001,
      "loss": 2.2305,
      "step": 7625
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.21953539550304413,
      "learning_rate": 0.001,
      "loss": 2.1782,
      "step": 7630
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2001238316297531,
      "learning_rate": 0.001,
      "loss": 2.286,
      "step": 7635
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.20314432680606842,
      "learning_rate": 0.001,
      "loss": 2.2022,
      "step": 7640
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2054559886455536,
      "learning_rate": 0.001,
      "loss": 2.2359,
      "step": 7645
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.21550264954566956,
      "learning_rate": 0.001,
      "loss": 2.2974,
      "step": 7650
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.22751198709011078,
      "learning_rate": 0.001,
      "loss": 2.1827,
      "step": 7655
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.20444227755069733,
      "learning_rate": 0.001,
      "loss": 2.2039,
      "step": 7660
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.19839997589588165,
      "learning_rate": 0.001,
      "loss": 2.2031,
      "step": 7665
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2035912275314331,
      "learning_rate": 0.001,
      "loss": 2.2535,
      "step": 7670
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2072770595550537,
      "learning_rate": 0.001,
      "loss": 2.2198,
      "step": 7675
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1923142671585083,
      "learning_rate": 0.001,
      "loss": 2.247,
      "step": 7680
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2125130444765091,
      "learning_rate": 0.001,
      "loss": 2.3123,
      "step": 7685
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2017953097820282,
      "learning_rate": 0.001,
      "loss": 2.279,
      "step": 7690
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.21377520263195038,
      "learning_rate": 0.001,
      "loss": 2.2798,
      "step": 7695
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.23891973495483398,
      "learning_rate": 0.001,
      "loss": 2.2924,
      "step": 7700
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.22691906988620758,
      "learning_rate": 0.001,
      "loss": 2.2234,
      "step": 7705
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2150959074497223,
      "learning_rate": 0.001,
      "loss": 2.2182,
      "step": 7710
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.19418391585350037,
      "learning_rate": 0.001,
      "loss": 2.1508,
      "step": 7715
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.21434800326824188,
      "learning_rate": 0.001,
      "loss": 2.2382,
      "step": 7720
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.21070006489753723,
      "learning_rate": 0.001,
      "loss": 2.2943,
      "step": 7725
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.194297656416893,
      "learning_rate": 0.001,
      "loss": 2.2794,
      "step": 7730
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.220197856426239,
      "learning_rate": 0.001,
      "loss": 2.242,
      "step": 7735
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2022203654050827,
      "learning_rate": 0.001,
      "loss": 2.2955,
      "step": 7740
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2011207491159439,
      "learning_rate": 0.001,
      "loss": 2.1795,
      "step": 7745
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.18811127543449402,
      "learning_rate": 0.001,
      "loss": 2.255,
      "step": 7750
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.1934177577495575,
      "learning_rate": 0.001,
      "loss": 2.2592,
      "step": 7755
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2053891122341156,
      "learning_rate": 0.001,
      "loss": 2.1615,
      "step": 7760
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.19190414249897003,
      "learning_rate": 0.001,
      "loss": 2.2911,
      "step": 7765
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.19915464520454407,
      "learning_rate": 0.001,
      "loss": 2.2466,
      "step": 7770
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.20265983045101166,
      "learning_rate": 0.001,
      "loss": 2.2542,
      "step": 7775
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.184738427400589,
      "learning_rate": 0.001,
      "loss": 2.2625,
      "step": 7780
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2045259177684784,
      "learning_rate": 0.001,
      "loss": 2.2833,
      "step": 7785
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.21759961545467377,
      "learning_rate": 0.001,
      "loss": 2.2725,
      "step": 7790
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1975535899400711,
      "learning_rate": 0.001,
      "loss": 2.1523,
      "step": 7795
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.20412267744541168,
      "learning_rate": 0.001,
      "loss": 2.1857,
      "step": 7800
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.19706964492797852,
      "learning_rate": 0.001,
      "loss": 2.2819,
      "step": 7805
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1994798183441162,
      "learning_rate": 0.001,
      "loss": 2.1996,
      "step": 7810
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.19474805891513824,
      "learning_rate": 0.001,
      "loss": 2.2869,
      "step": 7815
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.20528604090213776,
      "learning_rate": 0.001,
      "loss": 2.2877,
      "step": 7820
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.23693878948688507,
      "learning_rate": 0.001,
      "loss": 2.3314,
      "step": 7825
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.2081892490386963,
      "learning_rate": 0.001,
      "loss": 2.2119,
      "step": 7830
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.19748859107494354,
      "learning_rate": 0.001,
      "loss": 2.2803,
      "step": 7835
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.22490812838077545,
      "learning_rate": 0.001,
      "loss": 2.2673,
      "step": 7840
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.1880314201116562,
      "learning_rate": 0.001,
      "loss": 2.2421,
      "step": 7845
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20123490691184998,
      "learning_rate": 0.001,
      "loss": 2.2754,
      "step": 7850
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20010483264923096,
      "learning_rate": 0.001,
      "loss": 2.2223,
      "step": 7855
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20877912640571594,
      "learning_rate": 0.001,
      "loss": 2.286,
      "step": 7860
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.1998739242553711,
      "learning_rate": 0.001,
      "loss": 2.2096,
      "step": 7865
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20182031393051147,
      "learning_rate": 0.001,
      "loss": 2.2369,
      "step": 7870
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.21211646497249603,
      "learning_rate": 0.001,
      "loss": 2.2279,
      "step": 7875
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.19902239739894867,
      "learning_rate": 0.001,
      "loss": 2.1724,
      "step": 7880
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2212173193693161,
      "learning_rate": 0.001,
      "loss": 2.1901,
      "step": 7885
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.18968746066093445,
      "learning_rate": 0.001,
      "loss": 2.2669,
      "step": 7890
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.19653159379959106,
      "learning_rate": 0.001,
      "loss": 2.157,
      "step": 7895
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.19996388256549835,
      "learning_rate": 0.001,
      "loss": 2.276,
      "step": 7900
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20356236398220062,
      "learning_rate": 0.001,
      "loss": 2.1891,
      "step": 7905
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.19595099985599518,
      "learning_rate": 0.001,
      "loss": 2.2522,
      "step": 7910
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.19440218806266785,
      "learning_rate": 0.001,
      "loss": 2.2183,
      "step": 7915
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.19151397049427032,
      "learning_rate": 0.001,
      "loss": 2.2508,
      "step": 7920
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.20077724754810333,
      "learning_rate": 0.001,
      "loss": 2.2832,
      "step": 7925
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.23355898261070251,
      "learning_rate": 0.001,
      "loss": 2.2868,
      "step": 7930
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.2240302562713623,
      "learning_rate": 0.001,
      "loss": 2.2996,
      "step": 7935
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.20705515146255493,
      "learning_rate": 0.001,
      "loss": 2.2632,
      "step": 7940
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.19102813303470612,
      "learning_rate": 0.001,
      "loss": 2.2122,
      "step": 7945
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2303714156150818,
      "learning_rate": 0.001,
      "loss": 2.2268,
      "step": 7950
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.19775277376174927,
      "learning_rate": 0.001,
      "loss": 2.2802,
      "step": 7955
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.21944767236709595,
      "learning_rate": 0.001,
      "loss": 2.2109,
      "step": 7960
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.20680417120456696,
      "learning_rate": 0.001,
      "loss": 2.2182,
      "step": 7965
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1878603845834732,
      "learning_rate": 0.001,
      "loss": 2.2759,
      "step": 7970
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.19500446319580078,
      "learning_rate": 0.001,
      "loss": 2.25,
      "step": 7975
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1924324333667755,
      "learning_rate": 0.001,
      "loss": 2.2426,
      "step": 7980
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.21792522072792053,
      "learning_rate": 0.001,
      "loss": 2.283,
      "step": 7985
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1874474287033081,
      "learning_rate": 0.001,
      "loss": 2.2719,
      "step": 7990
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.1898234784603119,
      "learning_rate": 0.001,
      "loss": 2.2499,
      "step": 7995
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.20445217192173004,
      "learning_rate": 0.001,
      "loss": 2.3275,
      "step": 8000
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.20067258179187775,
      "learning_rate": 0.001,
      "loss": 2.1847,
      "step": 8005
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.19402869045734406,
      "learning_rate": 0.001,
      "loss": 2.2531,
      "step": 8010
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.2005668431520462,
      "learning_rate": 0.001,
      "loss": 2.1642,
      "step": 8015
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.20115861296653748,
      "learning_rate": 0.001,
      "loss": 2.3334,
      "step": 8020
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.19401215016841888,
      "learning_rate": 0.001,
      "loss": 2.2377,
      "step": 8025
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19128204882144928,
      "learning_rate": 0.001,
      "loss": 2.319,
      "step": 8030
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.20478878915309906,
      "learning_rate": 0.001,
      "loss": 2.2586,
      "step": 8035
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.20956303179264069,
      "learning_rate": 0.001,
      "loss": 2.2173,
      "step": 8040
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19278179109096527,
      "learning_rate": 0.001,
      "loss": 2.2481,
      "step": 8045
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19514437019824982,
      "learning_rate": 0.001,
      "loss": 2.2832,
      "step": 8050
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2172226905822754,
      "learning_rate": 0.001,
      "loss": 2.2327,
      "step": 8055
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.18109092116355896,
      "learning_rate": 0.001,
      "loss": 2.1731,
      "step": 8060
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1890215128660202,
      "learning_rate": 0.001,
      "loss": 2.3154,
      "step": 8065
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19754280149936676,
      "learning_rate": 0.001,
      "loss": 2.2203,
      "step": 8070
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19780655205249786,
      "learning_rate": 0.001,
      "loss": 2.2826,
      "step": 8075
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.18817149102687836,
      "learning_rate": 0.001,
      "loss": 2.2465,
      "step": 8080
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.19357270002365112,
      "learning_rate": 0.001,
      "loss": 2.2105,
      "step": 8085
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1954236626625061,
      "learning_rate": 0.001,
      "loss": 2.2486,
      "step": 8090
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.18649014830589294,
      "learning_rate": 0.001,
      "loss": 2.2103,
      "step": 8095
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.20539119839668274,
      "learning_rate": 0.001,
      "loss": 2.1869,
      "step": 8100
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.18983592092990875,
      "learning_rate": 0.001,
      "loss": 2.1514,
      "step": 8105
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.1923411339521408,
      "learning_rate": 0.001,
      "loss": 2.2228,
      "step": 8110
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.18866506218910217,
      "learning_rate": 0.001,
      "loss": 2.1797,
      "step": 8115
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.19905278086662292,
      "learning_rate": 0.001,
      "loss": 2.2419,
      "step": 8120
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.19118526577949524,
      "learning_rate": 0.001,
      "loss": 2.2174,
      "step": 8125
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.20813317596912384,
      "learning_rate": 0.001,
      "loss": 2.2707,
      "step": 8130
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.20310407876968384,
      "learning_rate": 0.001,
      "loss": 2.2669,
      "step": 8135
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.20061352849006653,
      "learning_rate": 0.001,
      "loss": 2.2334,
      "step": 8140
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.19646792113780975,
      "learning_rate": 0.001,
      "loss": 2.2195,
      "step": 8145
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.21986974775791168,
      "learning_rate": 0.001,
      "loss": 2.2153,
      "step": 8150
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.17849063873291016,
      "learning_rate": 0.001,
      "loss": 2.2178,
      "step": 8155
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.19407208263874054,
      "learning_rate": 0.001,
      "loss": 2.1844,
      "step": 8160
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.1898379921913147,
      "learning_rate": 0.001,
      "loss": 2.2834,
      "step": 8165
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.20284302532672882,
      "learning_rate": 0.001,
      "loss": 2.304,
      "step": 8170
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.21038895845413208,
      "learning_rate": 0.001,
      "loss": 2.1693,
      "step": 8175
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2112080454826355,
      "learning_rate": 0.001,
      "loss": 2.1756,
      "step": 8180
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.21226562559604645,
      "learning_rate": 0.001,
      "loss": 2.2484,
      "step": 8185
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.18678170442581177,
      "learning_rate": 0.001,
      "loss": 2.2438,
      "step": 8190
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.22687122225761414,
      "learning_rate": 0.001,
      "loss": 2.2828,
      "step": 8195
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2054503709077835,
      "learning_rate": 0.001,
      "loss": 2.2264,
      "step": 8200
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.19257859885692596,
      "learning_rate": 0.001,
      "loss": 2.3202,
      "step": 8205
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.19209985435009003,
      "learning_rate": 0.001,
      "loss": 2.1799,
      "step": 8210
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2024695873260498,
      "learning_rate": 0.001,
      "loss": 2.1492,
      "step": 8215
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1906440705060959,
      "learning_rate": 0.001,
      "loss": 2.2483,
      "step": 8220
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1945749670267105,
      "learning_rate": 0.001,
      "loss": 2.2186,
      "step": 8225
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.21156622469425201,
      "learning_rate": 0.001,
      "loss": 2.2877,
      "step": 8230
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.18485133349895477,
      "learning_rate": 0.001,
      "loss": 2.3249,
      "step": 8235
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.18751317262649536,
      "learning_rate": 0.001,
      "loss": 2.1932,
      "step": 8240
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1993100941181183,
      "learning_rate": 0.001,
      "loss": 2.2184,
      "step": 8245
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2515455186367035,
      "learning_rate": 0.001,
      "loss": 2.2294,
      "step": 8250
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.19279171526432037,
      "learning_rate": 0.001,
      "loss": 2.2683,
      "step": 8255
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.27863675355911255,
      "learning_rate": 0.001,
      "loss": 2.2202,
      "step": 8260
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.20159432291984558,
      "learning_rate": 0.001,
      "loss": 2.2538,
      "step": 8265
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2060721218585968,
      "learning_rate": 0.001,
      "loss": 2.2585,
      "step": 8270
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.19700761139392853,
      "learning_rate": 0.001,
      "loss": 2.1878,
      "step": 8275
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.19809959828853607,
      "learning_rate": 0.001,
      "loss": 2.2367,
      "step": 8280
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.2082154005765915,
      "learning_rate": 0.001,
      "loss": 2.1445,
      "step": 8285
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.1983455866575241,
      "learning_rate": 0.001,
      "loss": 2.2885,
      "step": 8290
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.19476141035556793,
      "learning_rate": 0.001,
      "loss": 2.2266,
      "step": 8295
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.20148876309394836,
      "learning_rate": 0.001,
      "loss": 2.1522,
      "step": 8300
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2016109824180603,
      "learning_rate": 0.001,
      "loss": 2.2349,
      "step": 8305
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.23375721275806427,
      "learning_rate": 0.001,
      "loss": 2.1436,
      "step": 8310
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.19160257279872894,
      "learning_rate": 0.001,
      "loss": 2.1778,
      "step": 8315
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.18439500033855438,
      "learning_rate": 0.001,
      "loss": 2.1937,
      "step": 8320
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.19977669417858124,
      "learning_rate": 0.001,
      "loss": 2.2322,
      "step": 8325
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.20961971580982208,
      "learning_rate": 0.001,
      "loss": 2.2496,
      "step": 8330
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.18950814008712769,
      "learning_rate": 0.001,
      "loss": 2.3144,
      "step": 8335
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.18986515700817108,
      "learning_rate": 0.001,
      "loss": 2.2437,
      "step": 8340
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.19119036197662354,
      "learning_rate": 0.001,
      "loss": 2.2634,
      "step": 8345
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.18153807520866394,
      "learning_rate": 0.001,
      "loss": 2.1471,
      "step": 8350
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.18010474741458893,
      "learning_rate": 0.001,
      "loss": 2.1389,
      "step": 8355
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.19723130762577057,
      "learning_rate": 0.001,
      "loss": 2.2782,
      "step": 8360
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2162715643644333,
      "learning_rate": 0.001,
      "loss": 2.1351,
      "step": 8365
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.1888781636953354,
      "learning_rate": 0.001,
      "loss": 2.2107,
      "step": 8370
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2023809850215912,
      "learning_rate": 0.001,
      "loss": 2.1802,
      "step": 8375
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.18921934068202972,
      "learning_rate": 0.001,
      "loss": 2.1953,
      "step": 8380
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.19186371564865112,
      "learning_rate": 0.001,
      "loss": 2.2515,
      "step": 8385
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2036081999540329,
      "learning_rate": 0.001,
      "loss": 2.2583,
      "step": 8390
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.21693940460681915,
      "learning_rate": 0.001,
      "loss": 2.2567,
      "step": 8395
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.20340251922607422,
      "learning_rate": 0.001,
      "loss": 2.1928,
      "step": 8400
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1897629350423813,
      "learning_rate": 0.001,
      "loss": 2.1738,
      "step": 8405
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.19895780086517334,
      "learning_rate": 0.001,
      "loss": 2.1841,
      "step": 8410
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.19295699894428253,
      "learning_rate": 0.001,
      "loss": 2.194,
      "step": 8415
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.19201216101646423,
      "learning_rate": 0.001,
      "loss": 2.2543,
      "step": 8420
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.18177099525928497,
      "learning_rate": 0.001,
      "loss": 2.1576,
      "step": 8425
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2475714534521103,
      "learning_rate": 0.001,
      "loss": 2.1712,
      "step": 8430
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.18510274589061737,
      "learning_rate": 0.001,
      "loss": 2.2584,
      "step": 8435
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.19641967117786407,
      "learning_rate": 0.001,
      "loss": 2.2638,
      "step": 8440
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.18982478976249695,
      "learning_rate": 0.001,
      "loss": 2.1609,
      "step": 8445
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.18552134931087494,
      "learning_rate": 0.001,
      "loss": 2.1553,
      "step": 8450
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.198546901345253,
      "learning_rate": 0.001,
      "loss": 2.133,
      "step": 8455
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.19203849136829376,
      "learning_rate": 0.001,
      "loss": 2.2076,
      "step": 8460
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.22504118084907532,
      "learning_rate": 0.001,
      "loss": 2.2539,
      "step": 8465
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.18785686790943146,
      "learning_rate": 0.001,
      "loss": 2.2552,
      "step": 8470
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.1946154534816742,
      "learning_rate": 0.001,
      "loss": 2.1435,
      "step": 8475
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1913706511259079,
      "learning_rate": 0.001,
      "loss": 2.2436,
      "step": 8480
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.17650993168354034,
      "learning_rate": 0.001,
      "loss": 2.2237,
      "step": 8485
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.183892622590065,
      "learning_rate": 0.001,
      "loss": 2.2427,
      "step": 8490
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.2386084794998169,
      "learning_rate": 0.001,
      "loss": 2.2556,
      "step": 8495
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.21051189303398132,
      "learning_rate": 0.001,
      "loss": 2.1528,
      "step": 8500
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1852414757013321,
      "learning_rate": 0.001,
      "loss": 2.2319,
      "step": 8505
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.19965070486068726,
      "learning_rate": 0.001,
      "loss": 2.1629,
      "step": 8510
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.18793615698814392,
      "learning_rate": 0.001,
      "loss": 2.273,
      "step": 8515
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.194820836186409,
      "learning_rate": 0.001,
      "loss": 2.2136,
      "step": 8520
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.25766825675964355,
      "learning_rate": 0.001,
      "loss": 2.1721,
      "step": 8525
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.22273294627666473,
      "learning_rate": 0.001,
      "loss": 2.1595,
      "step": 8530
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.22989904880523682,
      "learning_rate": 0.001,
      "loss": 2.2719,
      "step": 8535
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.21977895498275757,
      "learning_rate": 0.001,
      "loss": 2.2358,
      "step": 8540
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.20909038186073303,
      "learning_rate": 0.001,
      "loss": 2.2797,
      "step": 8545
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.17358806729316711,
      "learning_rate": 0.001,
      "loss": 2.2295,
      "step": 8550
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.19235506653785706,
      "learning_rate": 0.001,
      "loss": 2.2181,
      "step": 8555
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.20009635388851166,
      "learning_rate": 0.001,
      "loss": 2.2965,
      "step": 8560
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1902618110179901,
      "learning_rate": 0.001,
      "loss": 2.2593,
      "step": 8565
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.20280583202838898,
      "learning_rate": 0.001,
      "loss": 2.1473,
      "step": 8570
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.19154559075832367,
      "learning_rate": 0.001,
      "loss": 2.2924,
      "step": 8575
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.20159044861793518,
      "learning_rate": 0.001,
      "loss": 2.214,
      "step": 8580
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.23447231948375702,
      "learning_rate": 0.001,
      "loss": 2.2696,
      "step": 8585
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18314625322818756,
      "learning_rate": 0.001,
      "loss": 2.1882,
      "step": 8590
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.1972244381904602,
      "learning_rate": 0.001,
      "loss": 2.2503,
      "step": 8595
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18517006933689117,
      "learning_rate": 0.001,
      "loss": 2.1281,
      "step": 8600
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18247254192829132,
      "learning_rate": 0.001,
      "loss": 2.2121,
      "step": 8605
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18291489779949188,
      "learning_rate": 0.001,
      "loss": 2.1932,
      "step": 8610
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.20838534832000732,
      "learning_rate": 0.001,
      "loss": 2.1754,
      "step": 8615
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18746349215507507,
      "learning_rate": 0.001,
      "loss": 2.1692,
      "step": 8620
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.20475895702838898,
      "learning_rate": 0.001,
      "loss": 2.158,
      "step": 8625
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.19268833100795746,
      "learning_rate": 0.001,
      "loss": 2.1658,
      "step": 8630
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18458105623722076,
      "learning_rate": 0.001,
      "loss": 2.1426,
      "step": 8635
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.18215860426425934,
      "learning_rate": 0.001,
      "loss": 2.2663,
      "step": 8640
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.20346863567829132,
      "learning_rate": 0.001,
      "loss": 2.1614,
      "step": 8645
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.20370927453041077,
      "learning_rate": 0.001,
      "loss": 2.1839,
      "step": 8650
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.19011425971984863,
      "learning_rate": 0.001,
      "loss": 2.1797,
      "step": 8655
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.19454734027385712,
      "learning_rate": 0.001,
      "loss": 2.2564,
      "step": 8660
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.18686948716640472,
      "learning_rate": 0.001,
      "loss": 2.1875,
      "step": 8665
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.19724272191524506,
      "learning_rate": 0.001,
      "loss": 2.2285,
      "step": 8670
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.18537306785583496,
      "learning_rate": 0.001,
      "loss": 2.1898,
      "step": 8675
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.18430756032466888,
      "learning_rate": 0.001,
      "loss": 2.1661,
      "step": 8680
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.18186238408088684,
      "learning_rate": 0.001,
      "loss": 2.1505,
      "step": 8685
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2050350457429886,
      "learning_rate": 0.001,
      "loss": 2.19,
      "step": 8690
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.22094227373600006,
      "learning_rate": 0.001,
      "loss": 2.231,
      "step": 8695
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.17757384479045868,
      "learning_rate": 0.001,
      "loss": 2.201,
      "step": 8700
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1837390959262848,
      "learning_rate": 0.001,
      "loss": 2.1807,
      "step": 8705
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.18438223004341125,
      "learning_rate": 0.001,
      "loss": 2.2083,
      "step": 8710
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.20233049988746643,
      "learning_rate": 0.001,
      "loss": 2.2958,
      "step": 8715
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.21688610315322876,
      "learning_rate": 0.001,
      "loss": 2.2159,
      "step": 8720
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.1822272092103958,
      "learning_rate": 0.001,
      "loss": 2.2242,
      "step": 8725
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.21040040254592896,
      "learning_rate": 0.001,
      "loss": 2.2087,
      "step": 8730
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.17634455859661102,
      "learning_rate": 0.001,
      "loss": 2.223,
      "step": 8735
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.17778807878494263,
      "learning_rate": 0.001,
      "loss": 2.317,
      "step": 8740
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.27437514066696167,
      "learning_rate": 0.001,
      "loss": 2.2154,
      "step": 8745
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.20957055687904358,
      "learning_rate": 0.001,
      "loss": 2.1963,
      "step": 8750
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.19060905277729034,
      "learning_rate": 0.001,
      "loss": 2.1631,
      "step": 8755
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18315349519252777,
      "learning_rate": 0.001,
      "loss": 2.1854,
      "step": 8760
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.20113958418369293,
      "learning_rate": 0.001,
      "loss": 2.1993,
      "step": 8765
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.19529324769973755,
      "learning_rate": 0.001,
      "loss": 2.1558,
      "step": 8770
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.19535817205905914,
      "learning_rate": 0.001,
      "loss": 2.2532,
      "step": 8775
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18067863583564758,
      "learning_rate": 0.001,
      "loss": 2.1366,
      "step": 8780
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18170098960399628,
      "learning_rate": 0.001,
      "loss": 2.2162,
      "step": 8785
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18915463984012604,
      "learning_rate": 0.001,
      "loss": 2.2737,
      "step": 8790
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.20589640736579895,
      "learning_rate": 0.001,
      "loss": 2.2656,
      "step": 8795
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.21199221909046173,
      "learning_rate": 0.001,
      "loss": 2.2281,
      "step": 8800
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.21070407330989838,
      "learning_rate": 0.001,
      "loss": 2.2422,
      "step": 8805
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1806258261203766,
      "learning_rate": 0.001,
      "loss": 2.1982,
      "step": 8810
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1874959021806717,
      "learning_rate": 0.001,
      "loss": 2.2128,
      "step": 8815
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2044542133808136,
      "learning_rate": 0.001,
      "loss": 2.1611,
      "step": 8820
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18704530596733093,
      "learning_rate": 0.001,
      "loss": 2.0984,
      "step": 8825
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.18768665194511414,
      "learning_rate": 0.001,
      "loss": 2.2514,
      "step": 8830
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.20076428353786469,
      "learning_rate": 0.001,
      "loss": 2.212,
      "step": 8835
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.19733686745166779,
      "learning_rate": 0.001,
      "loss": 2.21,
      "step": 8840
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.17910441756248474,
      "learning_rate": 0.001,
      "loss": 2.237,
      "step": 8845
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18687348067760468,
      "learning_rate": 0.001,
      "loss": 2.2015,
      "step": 8850
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18390965461730957,
      "learning_rate": 0.001,
      "loss": 2.1904,
      "step": 8855
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18445414304733276,
      "learning_rate": 0.001,
      "loss": 2.2473,
      "step": 8860
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.21044015884399414,
      "learning_rate": 0.001,
      "loss": 2.1581,
      "step": 8865
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.20410551130771637,
      "learning_rate": 0.001,
      "loss": 2.2001,
      "step": 8870
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.17719769477844238,
      "learning_rate": 0.001,
      "loss": 2.187,
      "step": 8875
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.17933371663093567,
      "learning_rate": 0.001,
      "loss": 2.1969,
      "step": 8880
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1819513589143753,
      "learning_rate": 0.001,
      "loss": 2.2288,
      "step": 8885
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1850760281085968,
      "learning_rate": 0.001,
      "loss": 2.1351,
      "step": 8890
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1834917962551117,
      "learning_rate": 0.001,
      "loss": 2.2392,
      "step": 8895
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.183976948261261,
      "learning_rate": 0.001,
      "loss": 2.2063,
      "step": 8900
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18732498586177826,
      "learning_rate": 0.001,
      "loss": 2.2379,
      "step": 8905
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18440058827400208,
      "learning_rate": 0.001,
      "loss": 2.1353,
      "step": 8910
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.18112874031066895,
      "learning_rate": 0.001,
      "loss": 2.219,
      "step": 8915
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.1814718097448349,
      "learning_rate": 0.001,
      "loss": 2.1945,
      "step": 8920
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.17877763509750366,
      "learning_rate": 0.001,
      "loss": 2.2328,
      "step": 8925
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.1817835569381714,
      "learning_rate": 0.001,
      "loss": 2.1293,
      "step": 8930
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.19791525602340698,
      "learning_rate": 0.001,
      "loss": 2.2204,
      "step": 8935
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.19058136641979218,
      "learning_rate": 0.001,
      "loss": 2.2824,
      "step": 8940
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.18067437410354614,
      "learning_rate": 0.001,
      "loss": 2.1005,
      "step": 8945
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.17680200934410095,
      "learning_rate": 0.001,
      "loss": 2.0916,
      "step": 8950
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.17984385788440704,
      "learning_rate": 0.001,
      "loss": 2.1737,
      "step": 8955
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.20398467779159546,
      "learning_rate": 0.001,
      "loss": 2.2369,
      "step": 8960
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.186319962143898,
      "learning_rate": 0.001,
      "loss": 2.249,
      "step": 8965
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.2163608074188232,
      "eval_runtime": 830.4777,
      "eval_samples_per_second": 19.124,
      "eval_steps_per_second": 2.391,
      "step": 8969
    },
    {
      "epoch": 1.0,
      "step": 8969,
      "total_flos": 1.7053077415447757e+17,
      "train_loss": 2.6744996346363004,
      "train_runtime": 27779.5691,
      "train_samples_per_second": 5.165,
      "train_steps_per_second": 0.323
    }
  ],
  "logging_steps": 5,
  "max_steps": 8969,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 1.7053077415447757e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}