{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9002671415850401,
  "eval_steps": 500,
  "global_step": 9099,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 3.023737668991089,
      "learning_rate": 4.998021173444148e-05,
      "loss": 10.2295,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.7276954650878906,
      "learning_rate": 4.9960423468882955e-05,
      "loss": 9.2429,
      "step": 8
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.202863931655884,
      "learning_rate": 4.9940635203324435e-05,
      "loss": 8.8335,
      "step": 12
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.1424026489257812,
      "learning_rate": 4.992084693776591e-05,
      "loss": 8.5825,
      "step": 16
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.1057238578796387,
      "learning_rate": 4.990105867220739e-05,
      "loss": 8.2023,
      "step": 20
    },
    {
      "epoch": 0.0,
      "grad_norm": 2.0647401809692383,
      "learning_rate": 4.988127040664886e-05,
      "loss": 7.7905,
      "step": 24
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.845037817955017,
      "learning_rate": 4.986148214109034e-05,
      "loss": 7.5639,
      "step": 28
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.766366958618164,
      "learning_rate": 4.984169387553181e-05,
      "loss": 7.3338,
      "step": 32
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.7267919778823853,
      "learning_rate": 4.982190560997329e-05,
      "loss": 7.3396,
      "step": 36
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.3584438562393188,
      "learning_rate": 4.9802117344414766e-05,
      "loss": 7.2269,
      "step": 40
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.5067508220672607,
      "learning_rate": 4.978232907885624e-05,
      "loss": 6.8696,
      "step": 44
    },
    {
      "epoch": 0.0,
      "grad_norm": 1.3194962739944458,
      "learning_rate": 4.976254081329771e-05,
      "loss": 6.7392,
      "step": 48
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.4466848373413086,
      "learning_rate": 4.974275254773919e-05,
      "loss": 6.5025,
      "step": 52
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.3488733768463135,
      "learning_rate": 4.9722964282180664e-05,
      "loss": 6.4044,
      "step": 56
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.301389217376709,
      "learning_rate": 4.9703176016622144e-05,
      "loss": 6.2985,
      "step": 60
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.354225754737854,
      "learning_rate": 4.9683387751063624e-05,
      "loss": 6.1545,
      "step": 64
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1548670530319214,
      "learning_rate": 4.9663599485505096e-05,
      "loss": 6.1522,
      "step": 68
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2819015979766846,
      "learning_rate": 4.9643811219946576e-05,
      "loss": 5.9876,
      "step": 72
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9923768639564514,
      "learning_rate": 4.962402295438805e-05,
      "loss": 5.8414,
      "step": 76
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.005074381828308,
      "learning_rate": 4.960423468882953e-05,
      "loss": 5.7825,
      "step": 80
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9730919599533081,
      "learning_rate": 4.9584446423271e-05,
      "loss": 5.675,
      "step": 84
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.032512903213501,
      "learning_rate": 4.956465815771248e-05,
      "loss": 5.644,
      "step": 88
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.2836546897888184,
      "learning_rate": 4.9544869892153954e-05,
      "loss": 5.7789,
      "step": 92
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9425686001777649,
      "learning_rate": 4.9525081626595434e-05,
      "loss": 5.429,
      "step": 96
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8604844212532043,
      "learning_rate": 4.950529336103691e-05,
      "loss": 5.4202,
      "step": 100
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8457801342010498,
      "learning_rate": 4.948550509547839e-05,
      "loss": 5.2959,
      "step": 104
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9246242046356201,
      "learning_rate": 4.946571682991986e-05,
      "loss": 5.4305,
      "step": 108
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9483153223991394,
      "learning_rate": 4.944592856436134e-05,
      "loss": 5.382,
      "step": 112
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8151757717132568,
      "learning_rate": 4.942614029880281e-05,
      "loss": 5.2487,
      "step": 116
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8092381954193115,
      "learning_rate": 4.940635203324429e-05,
      "loss": 5.4251,
      "step": 120
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.0206758975982666,
      "learning_rate": 4.9386563767685765e-05,
      "loss": 5.245,
      "step": 124
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.9263086318969727,
      "learning_rate": 4.936677550212724e-05,
      "loss": 5.3408,
      "step": 128
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8204173445701599,
      "learning_rate": 4.934698723656872e-05,
      "loss": 5.3163,
      "step": 132
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.817661702632904,
      "learning_rate": 4.932719897101019e-05,
      "loss": 5.2058,
      "step": 136
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8175525069236755,
      "learning_rate": 4.930741070545167e-05,
      "loss": 5.0683,
      "step": 140
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.8956682682037354,
      "learning_rate": 4.928762243989314e-05,
      "loss": 5.285,
      "step": 144
    },
    {
      "epoch": 0.01,
      "grad_norm": 1.1034317016601562,
      "learning_rate": 4.926783417433462e-05,
      "loss": 5.2186,
      "step": 148
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8372359275817871,
      "learning_rate": 4.9248045908776096e-05,
      "loss": 5.0079,
      "step": 152
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8574570417404175,
      "learning_rate": 4.9228257643217576e-05,
      "loss": 5.1058,
      "step": 156
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0270808935165405,
      "learning_rate": 4.920846937765905e-05,
      "loss": 5.3269,
      "step": 160
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8500723242759705,
      "learning_rate": 4.918868111210053e-05,
      "loss": 5.1533,
      "step": 164
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8077535033226013,
      "learning_rate": 4.9168892846542e-05,
      "loss": 4.9847,
      "step": 168
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9095591902732849,
      "learning_rate": 4.914910458098348e-05,
      "loss": 4.8715,
      "step": 172
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0022218227386475,
      "learning_rate": 4.9129316315424954e-05,
      "loss": 4.9394,
      "step": 176
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0106725692749023,
      "learning_rate": 4.9109528049866433e-05,
      "loss": 5.1306,
      "step": 180
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.980566143989563,
      "learning_rate": 4.9089739784307906e-05,
      "loss": 5.0595,
      "step": 184
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0247553586959839,
      "learning_rate": 4.9069951518749386e-05,
      "loss": 5.0103,
      "step": 188
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.925584077835083,
      "learning_rate": 4.905016325319086e-05,
      "loss": 4.7923,
      "step": 192
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9778655171394348,
      "learning_rate": 4.903037498763234e-05,
      "loss": 4.8465,
      "step": 196
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8761100172996521,
      "learning_rate": 4.901058672207381e-05,
      "loss": 4.856,
      "step": 200
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8272585272789001,
      "learning_rate": 4.899079845651529e-05,
      "loss": 4.8365,
      "step": 204
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8929518461227417,
      "learning_rate": 4.8971010190956764e-05,
      "loss": 4.7708,
      "step": 208
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0088162422180176,
      "learning_rate": 4.895122192539824e-05,
      "loss": 4.8667,
      "step": 212
    },
    {
      "epoch": 0.02,
      "grad_norm": 1.0307660102844238,
      "learning_rate": 4.893143365983972e-05,
      "loss": 4.9297,
      "step": 216
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8324564099311829,
      "learning_rate": 4.891164539428119e-05,
      "loss": 4.6895,
      "step": 220
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9149929285049438,
      "learning_rate": 4.889185712872267e-05,
      "loss": 5.0596,
      "step": 224
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9188018441200256,
      "learning_rate": 4.887206886316414e-05,
      "loss": 4.7217,
      "step": 228
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8771291971206665,
      "learning_rate": 4.885228059760562e-05,
      "loss": 4.5893,
      "step": 232
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9374723434448242,
      "learning_rate": 4.8832492332047095e-05,
      "loss": 4.7205,
      "step": 236
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9590020179748535,
      "learning_rate": 4.8812704066488575e-05,
      "loss": 4.6909,
      "step": 240
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.8914527893066406,
      "learning_rate": 4.879291580093005e-05,
      "loss": 4.7799,
      "step": 244
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.96151202917099,
      "learning_rate": 4.877312753537153e-05,
      "loss": 4.6818,
      "step": 248
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9983839988708496,
      "learning_rate": 4.8753339269813e-05,
      "loss": 4.9377,
      "step": 252
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3818860054016113,
      "learning_rate": 4.873355100425448e-05,
      "loss": 4.8015,
      "step": 256
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8486915826797485,
      "learning_rate": 4.871376273869595e-05,
      "loss": 4.6542,
      "step": 260
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9783093333244324,
      "learning_rate": 4.869397447313743e-05,
      "loss": 4.635,
      "step": 264
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9911454916000366,
      "learning_rate": 4.8674186207578906e-05,
      "loss": 4.7617,
      "step": 268
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8163164258003235,
      "learning_rate": 4.8654397942020385e-05,
      "loss": 4.6394,
      "step": 272
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9522330164909363,
      "learning_rate": 4.8634609676461865e-05,
      "loss": 4.661,
      "step": 276
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0596697330474854,
      "learning_rate": 4.861482141090334e-05,
      "loss": 4.5832,
      "step": 280
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9185882806777954,
      "learning_rate": 4.859503314534482e-05,
      "loss": 4.5482,
      "step": 284
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.980002760887146,
      "learning_rate": 4.857524487978629e-05,
      "loss": 4.6268,
      "step": 288
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.7909073829650879,
      "learning_rate": 4.855545661422777e-05,
      "loss": 4.735,
      "step": 292
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9752382636070251,
      "learning_rate": 4.853566834866924e-05,
      "loss": 4.6384,
      "step": 296
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.0073622465133667,
      "learning_rate": 4.8515880083110716e-05,
      "loss": 4.5566,
      "step": 300
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.1823135614395142,
      "learning_rate": 4.849609181755219e-05,
      "loss": 4.6323,
      "step": 304
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8075012564659119,
      "learning_rate": 4.847630355199367e-05,
      "loss": 4.4955,
      "step": 308
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9888399839401245,
      "learning_rate": 4.845651528643514e-05,
      "loss": 4.5405,
      "step": 312
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9254989624023438,
      "learning_rate": 4.843672702087662e-05,
      "loss": 4.4871,
      "step": 316
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.068520426750183,
      "learning_rate": 4.8416938755318095e-05,
      "loss": 4.7617,
      "step": 320
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8666436672210693,
      "learning_rate": 4.8397150489759574e-05,
      "loss": 4.6256,
      "step": 324
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9162269830703735,
      "learning_rate": 4.837736222420105e-05,
      "loss": 4.3914,
      "step": 328
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8688908219337463,
      "learning_rate": 4.835757395864253e-05,
      "loss": 4.469,
      "step": 332
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9131215214729309,
      "learning_rate": 4.8337785693084007e-05,
      "loss": 4.5437,
      "step": 336
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.847606360912323,
      "learning_rate": 4.831799742752548e-05,
      "loss": 4.6195,
      "step": 340
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9658231735229492,
      "learning_rate": 4.829820916196696e-05,
      "loss": 4.5982,
      "step": 344
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.8523685336112976,
      "learning_rate": 4.827842089640843e-05,
      "loss": 4.42,
      "step": 348
    },
    {
      "epoch": 0.03,
      "grad_norm": 0.9430307149887085,
      "learning_rate": 4.825863263084991e-05,
      "loss": 4.5254,
      "step": 352
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.079181432723999,
      "learning_rate": 4.8238844365291385e-05,
      "loss": 4.76,
      "step": 356
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9053409099578857,
      "learning_rate": 4.8219056099732865e-05,
      "loss": 4.3891,
      "step": 360
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0313445329666138,
      "learning_rate": 4.819926783417434e-05,
      "loss": 4.6794,
      "step": 364
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9754608273506165,
      "learning_rate": 4.817947956861582e-05,
      "loss": 4.3483,
      "step": 368
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8928918838500977,
      "learning_rate": 4.815969130305729e-05,
      "loss": 4.3506,
      "step": 372
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9732823967933655,
      "learning_rate": 4.813990303749877e-05,
      "loss": 4.4418,
      "step": 376
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9762834906578064,
      "learning_rate": 4.8120114771940236e-05,
      "loss": 4.4823,
      "step": 380
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0655723810195923,
      "learning_rate": 4.8100326506381716e-05,
      "loss": 4.3974,
      "step": 384
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9439470767974854,
      "learning_rate": 4.808053824082319e-05,
      "loss": 4.5157,
      "step": 388
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9962547421455383,
      "learning_rate": 4.806074997526467e-05,
      "loss": 4.3487,
      "step": 392
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0140503644943237,
      "learning_rate": 4.804096170970615e-05,
      "loss": 4.2984,
      "step": 396
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9673125147819519,
      "learning_rate": 4.802117344414762e-05,
      "loss": 4.361,
      "step": 400
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0120090246200562,
      "learning_rate": 4.80013851785891e-05,
      "loss": 4.4019,
      "step": 404
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.902720034122467,
      "learning_rate": 4.7981596913030574e-05,
      "loss": 4.392,
      "step": 408
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8929727077484131,
      "learning_rate": 4.796180864747205e-05,
      "loss": 4.7008,
      "step": 412
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.045725703239441,
      "learning_rate": 4.7942020381913526e-05,
      "loss": 4.4655,
      "step": 416
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9211822748184204,
      "learning_rate": 4.7922232116355006e-05,
      "loss": 4.464,
      "step": 420
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9875116348266602,
      "learning_rate": 4.790244385079648e-05,
      "loss": 4.3493,
      "step": 424
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9416815042495728,
      "learning_rate": 4.788265558523796e-05,
      "loss": 4.2362,
      "step": 428
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9395657777786255,
      "learning_rate": 4.786286731967943e-05,
      "loss": 4.3159,
      "step": 432
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8677350878715515,
      "learning_rate": 4.784307905412091e-05,
      "loss": 4.3343,
      "step": 436
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9983429908752441,
      "learning_rate": 4.7823290788562384e-05,
      "loss": 4.3666,
      "step": 440
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.076252818107605,
      "learning_rate": 4.7803502523003864e-05,
      "loss": 4.4674,
      "step": 444
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8906400203704834,
      "learning_rate": 4.778371425744534e-05,
      "loss": 4.3918,
      "step": 448
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9424994587898254,
      "learning_rate": 4.7763925991886817e-05,
      "loss": 4.3358,
      "step": 452
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9982520341873169,
      "learning_rate": 4.774413772632829e-05,
      "loss": 4.4196,
      "step": 456
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9584160447120667,
      "learning_rate": 4.772434946076977e-05,
      "loss": 4.2583,
      "step": 460
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9916421175003052,
      "learning_rate": 4.770456119521124e-05,
      "loss": 4.4004,
      "step": 464
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0897949934005737,
      "learning_rate": 4.7684772929652715e-05,
      "loss": 4.3993,
      "step": 468
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.3799035549163818,
      "learning_rate": 4.7664984664094195e-05,
      "loss": 4.3915,
      "step": 472
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0133105516433716,
      "learning_rate": 4.764519639853567e-05,
      "loss": 4.0608,
      "step": 476
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1998016834259033,
      "learning_rate": 4.762540813297715e-05,
      "loss": 4.4948,
      "step": 480
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1928074359893799,
      "learning_rate": 4.760561986741862e-05,
      "loss": 4.1837,
      "step": 484
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0764780044555664,
      "learning_rate": 4.75858316018601e-05,
      "loss": 4.3355,
      "step": 488
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0785058736801147,
      "learning_rate": 4.756604333630157e-05,
      "loss": 4.3751,
      "step": 492
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0299701690673828,
      "learning_rate": 4.754625507074305e-05,
      "loss": 4.2652,
      "step": 496
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9715741276741028,
      "learning_rate": 4.7526466805184526e-05,
      "loss": 4.2677,
      "step": 500
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1598002910614014,
      "learning_rate": 4.7506678539626005e-05,
      "loss": 4.2647,
      "step": 504
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9881719946861267,
      "learning_rate": 4.748689027406748e-05,
      "loss": 4.2163,
      "step": 508
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0998167991638184,
      "learning_rate": 4.746710200850896e-05,
      "loss": 4.1632,
      "step": 512
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.055464267730713,
      "learning_rate": 4.744731374295043e-05,
      "loss": 4.2287,
      "step": 516
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1115161180496216,
      "learning_rate": 4.742752547739191e-05,
      "loss": 4.3554,
      "step": 520
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.8956658840179443,
      "learning_rate": 4.7407737211833384e-05,
      "loss": 3.9656,
      "step": 524
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.2546039819717407,
      "learning_rate": 4.738794894627486e-05,
      "loss": 4.2604,
      "step": 528
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1724745035171509,
      "learning_rate": 4.7368160680716336e-05,
      "loss": 4.291,
      "step": 532
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.068641185760498,
      "learning_rate": 4.7348372415157816e-05,
      "loss": 4.0505,
      "step": 536
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0498847961425781,
      "learning_rate": 4.732858414959929e-05,
      "loss": 4.2438,
      "step": 540
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.0670949220657349,
      "learning_rate": 4.730879588404077e-05,
      "loss": 4.1264,
      "step": 544
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.076369285583496,
      "learning_rate": 4.728900761848225e-05,
      "loss": 4.2795,
      "step": 548
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.1954712867736816,
      "learning_rate": 4.7269219352923714e-05,
      "loss": 4.2029,
      "step": 552
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9923123717308044,
      "learning_rate": 4.7249431087365194e-05,
      "loss": 4.2668,
      "step": 556
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1335583925247192,
      "learning_rate": 4.722964282180667e-05,
      "loss": 3.9782,
      "step": 560
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0587942600250244,
      "learning_rate": 4.720985455624815e-05,
      "loss": 4.0898,
      "step": 564
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9978137016296387,
      "learning_rate": 4.719006629068962e-05,
      "loss": 3.9701,
      "step": 568
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0784775018692017,
      "learning_rate": 4.71702780251311e-05,
      "loss": 4.0768,
      "step": 572
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1316741704940796,
      "learning_rate": 4.715048975957257e-05,
      "loss": 4.1185,
      "step": 576
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1574583053588867,
      "learning_rate": 4.713070149401405e-05,
      "loss": 4.1627,
      "step": 580
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.4153751134872437,
      "learning_rate": 4.7110913228455525e-05,
      "loss": 4.4233,
      "step": 584
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0705913305282593,
      "learning_rate": 4.7091124962897005e-05,
      "loss": 4.1492,
      "step": 588
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.3845056295394897,
      "learning_rate": 4.707133669733848e-05,
      "loss": 4.1557,
      "step": 592
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.989825427532196,
      "learning_rate": 4.705154843177996e-05,
      "loss": 4.1221,
      "step": 596
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1429495811462402,
      "learning_rate": 4.703176016622143e-05,
      "loss": 4.1611,
      "step": 600
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1858593225479126,
      "learning_rate": 4.701197190066291e-05,
      "loss": 4.121,
      "step": 604
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9340394735336304,
      "learning_rate": 4.699218363510439e-05,
      "loss": 4.0173,
      "step": 608
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0611008405685425,
      "learning_rate": 4.697239536954586e-05,
      "loss": 4.0978,
      "step": 612
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.1144658327102661,
      "learning_rate": 4.695260710398734e-05,
      "loss": 3.9948,
      "step": 616
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0355262756347656,
      "learning_rate": 4.6932818838428815e-05,
      "loss": 3.9703,
      "step": 620
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0302252769470215,
      "learning_rate": 4.6913030572870295e-05,
      "loss": 4.055,
      "step": 624
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9846487641334534,
      "learning_rate": 4.689324230731177e-05,
      "loss": 4.0066,
      "step": 628
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.2877755165100098,
      "learning_rate": 4.687345404175325e-05,
      "loss": 4.1539,
      "step": 632
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.046048641204834,
      "learning_rate": 4.6853665776194714e-05,
      "loss": 3.8136,
      "step": 636
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0364129543304443,
      "learning_rate": 4.6833877510636193e-05,
      "loss": 4.1889,
      "step": 640
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.102586030960083,
      "learning_rate": 4.6814089245077666e-05,
      "loss": 3.9993,
      "step": 644
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0004945993423462,
      "learning_rate": 4.6794300979519146e-05,
      "loss": 3.9929,
      "step": 648
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0600558519363403,
      "learning_rate": 4.677451271396062e-05,
      "loss": 4.0179,
      "step": 652
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.0851588249206543,
      "learning_rate": 4.67547244484021e-05,
      "loss": 3.9701,
      "step": 656
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2407004833221436,
      "learning_rate": 4.673493618284357e-05,
      "loss": 4.1936,
      "step": 660
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.4276139736175537,
      "learning_rate": 4.671514791728505e-05,
      "loss": 4.0371,
      "step": 664
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1554222106933594,
      "learning_rate": 4.669535965172653e-05,
      "loss": 4.1037,
      "step": 668
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.062843918800354,
      "learning_rate": 4.6675571386168004e-05,
      "loss": 3.8831,
      "step": 672
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2357606887817383,
      "learning_rate": 4.6655783120609484e-05,
      "loss": 4.0885,
      "step": 676
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1083883047103882,
      "learning_rate": 4.663599485505096e-05,
      "loss": 3.8199,
      "step": 680
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0669097900390625,
      "learning_rate": 4.6616206589492436e-05,
      "loss": 4.1267,
      "step": 684
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0153981447219849,
      "learning_rate": 4.659641832393391e-05,
      "loss": 3.833,
      "step": 688
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2568185329437256,
      "learning_rate": 4.657663005837539e-05,
      "loss": 4.1119,
      "step": 692
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2034963369369507,
      "learning_rate": 4.655684179281686e-05,
      "loss": 3.7792,
      "step": 696
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.192213773727417,
      "learning_rate": 4.653705352725834e-05,
      "loss": 4.0965,
      "step": 700
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1414371728897095,
      "learning_rate": 4.6517265261699815e-05,
      "loss": 3.8453,
      "step": 704
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2463369369506836,
      "learning_rate": 4.6497476996141294e-05,
      "loss": 3.9088,
      "step": 708
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.062708854675293,
      "learning_rate": 4.647768873058277e-05,
      "loss": 3.9295,
      "step": 712
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.05110764503479,
      "learning_rate": 4.645790046502425e-05,
      "loss": 3.9937,
      "step": 716
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0769037008285522,
      "learning_rate": 4.643811219946571e-05,
      "loss": 3.9416,
      "step": 720
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1095948219299316,
      "learning_rate": 4.641832393390719e-05,
      "loss": 3.7524,
      "step": 724
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1786772012710571,
      "learning_rate": 4.639853566834867e-05,
      "loss": 3.7961,
      "step": 728
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1711845397949219,
      "learning_rate": 4.6378747402790145e-05,
      "loss": 3.9437,
      "step": 732
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.0217204093933105,
      "learning_rate": 4.6358959137231625e-05,
      "loss": 3.8099,
      "step": 736
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1621912717819214,
      "learning_rate": 4.63391708716731e-05,
      "loss": 3.9684,
      "step": 740
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2898588180541992,
      "learning_rate": 4.631938260611458e-05,
      "loss": 3.8558,
      "step": 744
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2988104820251465,
      "learning_rate": 4.629959434055605e-05,
      "loss": 3.8503,
      "step": 748
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.1195327043533325,
      "learning_rate": 4.627980607499753e-05,
      "loss": 3.8956,
      "step": 752
    },
    {
      "epoch": 0.07,
      "grad_norm": 1.2289355993270874,
      "learning_rate": 4.6260017809439003e-05,
      "loss": 3.6909,
      "step": 756
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.2389601469039917,
      "learning_rate": 4.624022954388048e-05,
      "loss": 3.9477,
      "step": 760
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.2713934183120728,
      "learning_rate": 4.6220441278321956e-05,
      "loss": 3.9466,
      "step": 764
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.19290292263031,
      "learning_rate": 4.6200653012763436e-05,
      "loss": 3.7048,
      "step": 768
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.107513666152954,
      "learning_rate": 4.618086474720491e-05,
      "loss": 3.8013,
      "step": 772
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9974806904792786,
      "learning_rate": 4.616107648164639e-05,
      "loss": 3.7578,
      "step": 776
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.180731177330017,
      "learning_rate": 4.614128821608786e-05,
      "loss": 3.9304,
      "step": 780
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.5762641429901123,
      "learning_rate": 4.612149995052934e-05,
      "loss": 3.9271,
      "step": 784
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1390140056610107,
      "learning_rate": 4.6101711684970814e-05,
      "loss": 4.0947,
      "step": 788
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.2429100275039673,
      "learning_rate": 4.6081923419412294e-05,
      "loss": 3.9939,
      "step": 792
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1867088079452515,
      "learning_rate": 4.606213515385377e-05,
      "loss": 3.8247,
      "step": 796
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0802291631698608,
      "learning_rate": 4.6042346888295246e-05,
      "loss": 3.6472,
      "step": 800
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3064371347427368,
      "learning_rate": 4.602255862273672e-05,
      "loss": 3.785,
      "step": 804
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3858779668807983,
      "learning_rate": 4.600277035717819e-05,
      "loss": 4.0065,
      "step": 808
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.09247887134552,
      "learning_rate": 4.598298209161967e-05,
      "loss": 3.8801,
      "step": 812
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1438707113265991,
      "learning_rate": 4.5963193826061145e-05,
      "loss": 3.7864,
      "step": 816
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.107773780822754,
      "learning_rate": 4.5943405560502625e-05,
      "loss": 3.6436,
      "step": 820
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0622732639312744,
      "learning_rate": 4.59236172949441e-05,
      "loss": 3.8597,
      "step": 824
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0812435150146484,
      "learning_rate": 4.590382902938558e-05,
      "loss": 3.8085,
      "step": 828
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1576310396194458,
      "learning_rate": 4.588404076382705e-05,
      "loss": 3.6459,
      "step": 832
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1088974475860596,
      "learning_rate": 4.586425249826853e-05,
      "loss": 3.8514,
      "step": 836
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.1565968990325928,
      "learning_rate": 4.584446423271e-05,
      "loss": 3.7429,
      "step": 840
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.226541519165039,
      "learning_rate": 4.582467596715148e-05,
      "loss": 3.9345,
      "step": 844
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.3901216983795166,
      "learning_rate": 4.5804887701592955e-05,
      "loss": 4.0038,
      "step": 848
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0757673978805542,
      "learning_rate": 4.5785099436034435e-05,
      "loss": 3.6689,
      "step": 852
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0806201696395874,
      "learning_rate": 4.576531117047591e-05,
      "loss": 3.8034,
      "step": 856
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0815987586975098,
      "learning_rate": 4.574552290491739e-05,
      "loss": 3.5688,
      "step": 860
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.2539079189300537,
      "learning_rate": 4.572573463935886e-05,
      "loss": 3.7314,
      "step": 864
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.2489246129989624,
      "learning_rate": 4.570594637380034e-05,
      "loss": 3.7358,
      "step": 868
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.188781976699829,
      "learning_rate": 4.568615810824181e-05,
      "loss": 3.8216,
      "step": 872
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1977195739746094,
      "learning_rate": 4.566636984268329e-05,
      "loss": 3.7418,
      "step": 876
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3144996166229248,
      "learning_rate": 4.564658157712477e-05,
      "loss": 3.8569,
      "step": 880
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.194718360900879,
      "learning_rate": 4.5626793311566246e-05,
      "loss": 3.9339,
      "step": 884
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1180779933929443,
      "learning_rate": 4.560700504600772e-05,
      "loss": 3.7593,
      "step": 888
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.2359132766723633,
      "learning_rate": 4.558721678044919e-05,
      "loss": 3.7554,
      "step": 892
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.456741213798523,
      "learning_rate": 4.556742851489067e-05,
      "loss": 3.8741,
      "step": 896
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.21878981590271,
      "learning_rate": 4.5547640249332144e-05,
      "loss": 3.7214,
      "step": 900
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.2724573612213135,
      "learning_rate": 4.5527851983773624e-05,
      "loss": 3.7083,
      "step": 904
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.3674492835998535,
      "learning_rate": 4.55080637182151e-05,
      "loss": 3.7782,
      "step": 908
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1962213516235352,
      "learning_rate": 4.5488275452656577e-05,
      "loss": 3.8053,
      "step": 912
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.204624056816101,
      "learning_rate": 4.546848718709805e-05,
      "loss": 3.8807,
      "step": 916
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.162916660308838,
      "learning_rate": 4.544869892153953e-05,
      "loss": 3.62,
      "step": 920
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1495506763458252,
      "learning_rate": 4.5428910655981e-05,
      "loss": 3.6001,
      "step": 924
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1593899726867676,
      "learning_rate": 4.540912239042248e-05,
      "loss": 4.0318,
      "step": 928
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1468162536621094,
      "learning_rate": 4.5389334124863955e-05,
      "loss": 3.6793,
      "step": 932
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.0815014839172363,
      "learning_rate": 4.5369545859305434e-05,
      "loss": 3.8516,
      "step": 936
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1018073558807373,
      "learning_rate": 4.5349757593746914e-05,
      "loss": 3.7732,
      "step": 940
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1128356456756592,
      "learning_rate": 4.532996932818839e-05,
      "loss": 3.5784,
      "step": 944
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.449112057685852,
      "learning_rate": 4.531018106262987e-05,
      "loss": 3.8318,
      "step": 948
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1925861835479736,
      "learning_rate": 4.529039279707134e-05,
      "loss": 3.7108,
      "step": 952
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.1462700366973877,
      "learning_rate": 4.527060453151282e-05,
      "loss": 3.6801,
      "step": 956
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.218001127243042,
      "learning_rate": 4.525081626595429e-05,
      "loss": 3.7581,
      "step": 960
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.09965980052948,
      "learning_rate": 4.523102800039577e-05,
      "loss": 3.661,
      "step": 964
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1705689430236816,
      "learning_rate": 4.5211239734837245e-05,
      "loss": 3.8608,
      "step": 968
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2540526390075684,
      "learning_rate": 4.519145146927872e-05,
      "loss": 3.5236,
      "step": 972
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1072841882705688,
      "learning_rate": 4.517166320372019e-05,
      "loss": 3.5754,
      "step": 976
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.0101158618927002,
      "learning_rate": 4.515187493816167e-05,
      "loss": 3.7373,
      "step": 980
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3311535120010376,
      "learning_rate": 4.5132086672603144e-05,
      "loss": 3.7958,
      "step": 984
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1845935583114624,
      "learning_rate": 4.511229840704462e-05,
      "loss": 3.8895,
      "step": 988
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1750415563583374,
      "learning_rate": 4.5092510141486096e-05,
      "loss": 3.8246,
      "step": 992
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.423757553100586,
      "learning_rate": 4.5072721875927576e-05,
      "loss": 3.5447,
      "step": 996
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1389232873916626,
      "learning_rate": 4.5052933610369056e-05,
      "loss": 3.5603,
      "step": 1000
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1800868511199951,
      "learning_rate": 4.503314534481053e-05,
      "loss": 3.9599,
      "step": 1004
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.163454294204712,
      "learning_rate": 4.501335707925201e-05,
      "loss": 3.7299,
      "step": 1008
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1138209104537964,
      "learning_rate": 4.499356881369348e-05,
      "loss": 3.5834,
      "step": 1012
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2494182586669922,
      "learning_rate": 4.497378054813496e-05,
      "loss": 3.7193,
      "step": 1016
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1176248788833618,
      "learning_rate": 4.4953992282576434e-05,
      "loss": 3.5979,
      "step": 1020
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1829919815063477,
      "learning_rate": 4.4934204017017914e-05,
      "loss": 3.6594,
      "step": 1024
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1001139879226685,
      "learning_rate": 4.4914415751459386e-05,
      "loss": 3.65,
      "step": 1028
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2622652053833008,
      "learning_rate": 4.4894627485900866e-05,
      "loss": 3.6189,
      "step": 1032
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.2330125570297241,
      "learning_rate": 4.487483922034234e-05,
      "loss": 3.7518,
      "step": 1036
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.5254411697387695,
      "learning_rate": 4.485505095478382e-05,
      "loss": 3.6814,
      "step": 1040
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.231075644493103,
      "learning_rate": 4.483526268922529e-05,
      "loss": 3.6836,
      "step": 1044
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.404321551322937,
      "learning_rate": 4.481547442366677e-05,
      "loss": 3.6597,
      "step": 1048
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.1372672319412231,
      "learning_rate": 4.4795686158108244e-05,
      "loss": 3.7171,
      "step": 1052
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.3405723571777344,
      "learning_rate": 4.477589789254972e-05,
      "loss": 3.7671,
      "step": 1056
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.246029019355774,
      "learning_rate": 4.47561096269912e-05,
      "loss": 4.1252,
      "step": 1060
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.395970344543457,
      "learning_rate": 4.473632136143267e-05,
      "loss": 3.554,
      "step": 1064
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.4145135879516602,
      "learning_rate": 4.471653309587415e-05,
      "loss": 3.7314,
      "step": 1068
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1709779500961304,
      "learning_rate": 4.469674483031562e-05,
      "loss": 3.4601,
      "step": 1072
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.4412345886230469,
      "learning_rate": 4.46769565647571e-05,
      "loss": 3.4616,
      "step": 1076
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3158358335494995,
      "learning_rate": 4.4657168299198575e-05,
      "loss": 3.5625,
      "step": 1080
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1684952974319458,
      "learning_rate": 4.4637380033640055e-05,
      "loss": 3.4344,
      "step": 1084
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3412691354751587,
      "learning_rate": 4.461759176808153e-05,
      "loss": 3.6387,
      "step": 1088
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2844157218933105,
      "learning_rate": 4.459780350252301e-05,
      "loss": 3.5851,
      "step": 1092
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.217627763748169,
      "learning_rate": 4.457801523696448e-05,
      "loss": 3.5215,
      "step": 1096
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3247570991516113,
      "learning_rate": 4.455822697140596e-05,
      "loss": 3.5599,
      "step": 1100
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.402744174003601,
      "learning_rate": 4.453843870584743e-05,
      "loss": 3.6444,
      "step": 1104
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2150934934616089,
      "learning_rate": 4.451865044028891e-05,
      "loss": 3.5712,
      "step": 1108
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1183940172195435,
      "learning_rate": 4.4498862174730386e-05,
      "loss": 3.6094,
      "step": 1112
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2478612661361694,
      "learning_rate": 4.4479073909171866e-05,
      "loss": 3.3976,
      "step": 1116
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2105813026428223,
      "learning_rate": 4.445928564361334e-05,
      "loss": 3.5313,
      "step": 1120
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1932976245880127,
      "learning_rate": 4.443949737805482e-05,
      "loss": 3.4661,
      "step": 1124
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3227620124816895,
      "learning_rate": 4.441970911249629e-05,
      "loss": 3.5443,
      "step": 1128
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1562778949737549,
      "learning_rate": 4.439992084693777e-05,
      "loss": 3.5925,
      "step": 1132
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3220064640045166,
      "learning_rate": 4.4380132581379244e-05,
      "loss": 3.5389,
      "step": 1136
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2712664604187012,
      "learning_rate": 4.436034431582072e-05,
      "loss": 3.6925,
      "step": 1140
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2455075979232788,
      "learning_rate": 4.4340556050262196e-05,
      "loss": 3.5731,
      "step": 1144
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.1540822982788086,
      "learning_rate": 4.432076778470367e-05,
      "loss": 3.4803,
      "step": 1148
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.2069053649902344,
      "learning_rate": 4.430097951914515e-05,
      "loss": 3.5102,
      "step": 1152
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.3267216682434082,
      "learning_rate": 4.428119125358662e-05,
      "loss": 3.7385,
      "step": 1156
    },
    {
      "epoch": 0.11,
      "grad_norm": 1.311906099319458,
      "learning_rate": 4.42614029880281e-05,
      "loss": 3.6113,
      "step": 1160
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1880320310592651,
      "learning_rate": 4.4241614722469575e-05,
      "loss": 3.4399,
      "step": 1164
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5865764617919922,
      "learning_rate": 4.4221826456911054e-05,
      "loss": 3.7722,
      "step": 1168
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.119573950767517,
      "learning_rate": 4.420203819135253e-05,
      "loss": 3.5,
      "step": 1172
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2991716861724854,
      "learning_rate": 4.418224992579401e-05,
      "loss": 3.5364,
      "step": 1176
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1628044843673706,
      "learning_rate": 4.416246166023548e-05,
      "loss": 3.3195,
      "step": 1180
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0280978679656982,
      "learning_rate": 4.414267339467696e-05,
      "loss": 3.4377,
      "step": 1184
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0991793870925903,
      "learning_rate": 4.412288512911843e-05,
      "loss": 3.5634,
      "step": 1188
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0973697900772095,
      "learning_rate": 4.410309686355991e-05,
      "loss": 3.5387,
      "step": 1192
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5861742496490479,
      "learning_rate": 4.4083308598001385e-05,
      "loss": 3.5671,
      "step": 1196
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.3537582159042358,
      "learning_rate": 4.4063520332442865e-05,
      "loss": 3.7076,
      "step": 1200
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1639502048492432,
      "learning_rate": 4.404373206688434e-05,
      "loss": 3.577,
      "step": 1204
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0819730758666992,
      "learning_rate": 4.402394380132582e-05,
      "loss": 3.5282,
      "step": 1208
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2191344499588013,
      "learning_rate": 4.40041555357673e-05,
      "loss": 3.447,
      "step": 1212
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1389458179473877,
      "learning_rate": 4.398436727020877e-05,
      "loss": 3.4366,
      "step": 1216
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2213441133499146,
      "learning_rate": 4.396457900465025e-05,
      "loss": 3.5216,
      "step": 1220
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.4312680959701538,
      "learning_rate": 4.3944790739091716e-05,
      "loss": 3.5432,
      "step": 1224
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.196575403213501,
      "learning_rate": 4.3925002473533196e-05,
      "loss": 3.3649,
      "step": 1228
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1721031665802002,
      "learning_rate": 4.390521420797467e-05,
      "loss": 3.3323,
      "step": 1232
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1894913911819458,
      "learning_rate": 4.388542594241615e-05,
      "loss": 3.431,
      "step": 1236
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.127156138420105,
      "learning_rate": 4.386563767685762e-05,
      "loss": 3.5986,
      "step": 1240
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.2058907747268677,
      "learning_rate": 4.38458494112991e-05,
      "loss": 3.4094,
      "step": 1244
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1665937900543213,
      "learning_rate": 4.3826061145740574e-05,
      "loss": 3.5399,
      "step": 1248
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.0757410526275635,
      "learning_rate": 4.3806272880182054e-05,
      "loss": 3.3555,
      "step": 1252
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1566282510757446,
      "learning_rate": 4.378648461462353e-05,
      "loss": 3.5372,
      "step": 1256
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.1599324941635132,
      "learning_rate": 4.3766696349065006e-05,
      "loss": 3.6885,
      "step": 1260
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2669498920440674,
      "learning_rate": 4.374690808350648e-05,
      "loss": 3.5088,
      "step": 1264
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2056463956832886,
      "learning_rate": 4.372711981794796e-05,
      "loss": 3.4207,
      "step": 1268
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.285954475402832,
      "learning_rate": 4.370733155238944e-05,
      "loss": 3.4905,
      "step": 1272
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.5418660640716553,
      "learning_rate": 4.368754328683091e-05,
      "loss": 3.46,
      "step": 1276
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3269129991531372,
      "learning_rate": 4.366775502127239e-05,
      "loss": 3.651,
      "step": 1280
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.206346869468689,
      "learning_rate": 4.3647966755713864e-05,
      "loss": 3.4792,
      "step": 1284
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.5066856145858765,
      "learning_rate": 4.3628178490155344e-05,
      "loss": 3.7163,
      "step": 1288
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.660408616065979,
      "learning_rate": 4.360839022459682e-05,
      "loss": 3.3533,
      "step": 1292
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3185946941375732,
      "learning_rate": 4.35886019590383e-05,
      "loss": 3.6158,
      "step": 1296
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2295507192611694,
      "learning_rate": 4.356881369347977e-05,
      "loss": 3.3683,
      "step": 1300
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2218221426010132,
      "learning_rate": 4.354902542792125e-05,
      "loss": 3.3305,
      "step": 1304
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4210175275802612,
      "learning_rate": 4.352923716236272e-05,
      "loss": 3.4825,
      "step": 1308
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4661409854888916,
      "learning_rate": 4.3509448896804195e-05,
      "loss": 3.4458,
      "step": 1312
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2703677415847778,
      "learning_rate": 4.348966063124567e-05,
      "loss": 3.2625,
      "step": 1316
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4040093421936035,
      "learning_rate": 4.346987236568715e-05,
      "loss": 3.5308,
      "step": 1320
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1892530918121338,
      "learning_rate": 4.345008410012862e-05,
      "loss": 3.4657,
      "step": 1324
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3836407661437988,
      "learning_rate": 4.34302958345701e-05,
      "loss": 3.4017,
      "step": 1328
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1580448150634766,
      "learning_rate": 4.341050756901158e-05,
      "loss": 3.367,
      "step": 1332
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1472089290618896,
      "learning_rate": 4.339071930345305e-05,
      "loss": 3.4479,
      "step": 1336
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2365572452545166,
      "learning_rate": 4.337093103789453e-05,
      "loss": 3.4469,
      "step": 1340
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2386647462844849,
      "learning_rate": 4.3351142772336006e-05,
      "loss": 3.6044,
      "step": 1344
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.3203320503234863,
      "learning_rate": 4.3331354506777485e-05,
      "loss": 3.5549,
      "step": 1348
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.4202224016189575,
      "learning_rate": 4.331156624121896e-05,
      "loss": 3.6182,
      "step": 1352
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1654404401779175,
      "learning_rate": 4.329177797566044e-05,
      "loss": 3.4088,
      "step": 1356
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.2231131792068481,
      "learning_rate": 4.327198971010191e-05,
      "loss": 3.4264,
      "step": 1360
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.1695729494094849,
      "learning_rate": 4.325220144454339e-05,
      "loss": 3.4689,
      "step": 1364
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.392952561378479,
      "learning_rate": 4.3232413178984864e-05,
      "loss": 3.4129,
      "step": 1368
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.9919403791427612,
      "learning_rate": 4.321262491342634e-05,
      "loss": 3.7645,
      "step": 1372
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3091975450515747,
      "learning_rate": 4.3192836647867816e-05,
      "loss": 3.4731,
      "step": 1376
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.440266489982605,
      "learning_rate": 4.3173048382309296e-05,
      "loss": 3.3394,
      "step": 1380
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3461400270462036,
      "learning_rate": 4.315326011675077e-05,
      "loss": 3.3563,
      "step": 1384
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4940065145492554,
      "learning_rate": 4.313347185119225e-05,
      "loss": 3.4845,
      "step": 1388
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.401813268661499,
      "learning_rate": 4.311368358563372e-05,
      "loss": 3.3725,
      "step": 1392
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.2095842361450195,
      "learning_rate": 4.3093895320075194e-05,
      "loss": 3.5393,
      "step": 1396
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.553962230682373,
      "learning_rate": 4.3074107054516674e-05,
      "loss": 3.6136,
      "step": 1400
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1763421297073364,
      "learning_rate": 4.305431878895815e-05,
      "loss": 3.3757,
      "step": 1404
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.5987308025360107,
      "learning_rate": 4.303453052339963e-05,
      "loss": 3.5515,
      "step": 1408
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3491801023483276,
      "learning_rate": 4.30147422578411e-05,
      "loss": 3.5995,
      "step": 1412
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.194741129875183,
      "learning_rate": 4.299495399228258e-05,
      "loss": 3.3052,
      "step": 1416
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3502442836761475,
      "learning_rate": 4.297516572672405e-05,
      "loss": 3.3727,
      "step": 1420
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.4272186756134033,
      "learning_rate": 4.295537746116553e-05,
      "loss": 3.3057,
      "step": 1424
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.1727451086044312,
      "learning_rate": 4.2935589195607005e-05,
      "loss": 3.4478,
      "step": 1428
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.367772102355957,
      "learning_rate": 4.2915800930048485e-05,
      "loss": 3.3518,
      "step": 1432
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3269891738891602,
      "learning_rate": 4.289601266448996e-05,
      "loss": 3.3426,
      "step": 1436
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.188839316368103,
      "learning_rate": 4.287622439893144e-05,
      "loss": 3.4735,
      "step": 1440
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3754055500030518,
      "learning_rate": 4.285643613337291e-05,
      "loss": 3.3271,
      "step": 1444
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.2622300386428833,
      "learning_rate": 4.283664786781439e-05,
      "loss": 3.7099,
      "step": 1448
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.2850700616836548,
      "learning_rate": 4.281685960225586e-05,
      "loss": 3.3448,
      "step": 1452
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.3063966035842896,
      "learning_rate": 4.279707133669734e-05,
      "loss": 3.4525,
      "step": 1456
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.2147005796432495,
      "learning_rate": 4.2777283071138816e-05,
      "loss": 3.3215,
      "step": 1460
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.347546100616455,
      "learning_rate": 4.2757494805580295e-05,
      "loss": 3.5355,
      "step": 1464
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.228791356086731,
      "learning_rate": 4.273770654002177e-05,
      "loss": 3.3136,
      "step": 1468
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.5091530084609985,
      "learning_rate": 4.271791827446325e-05,
      "loss": 3.321,
      "step": 1472
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1982457637786865,
      "learning_rate": 4.269813000890472e-05,
      "loss": 3.3217,
      "step": 1476
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2437925338745117,
      "learning_rate": 4.2678341743346194e-05,
      "loss": 3.666,
      "step": 1480
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.508147120475769,
      "learning_rate": 4.2658553477787674e-05,
      "loss": 3.1524,
      "step": 1484
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1853328943252563,
      "learning_rate": 4.2638765212229146e-05,
      "loss": 3.3933,
      "step": 1488
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.410609245300293,
      "learning_rate": 4.2618976946670626e-05,
      "loss": 3.6355,
      "step": 1492
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2669347524642944,
      "learning_rate": 4.25991886811121e-05,
      "loss": 3.3659,
      "step": 1496
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1890374422073364,
      "learning_rate": 4.257940041555358e-05,
      "loss": 3.4131,
      "step": 1500
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2087454795837402,
      "learning_rate": 4.255961214999505e-05,
      "loss": 3.437,
      "step": 1504
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2528936862945557,
      "learning_rate": 4.253982388443653e-05,
      "loss": 3.3099,
      "step": 1508
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.1920204162597656,
      "learning_rate": 4.2520035618878004e-05,
      "loss": 3.4506,
      "step": 1512
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.289352297782898,
      "learning_rate": 4.2500247353319484e-05,
      "loss": 3.5612,
      "step": 1516
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.3303660154342651,
      "learning_rate": 4.248045908776096e-05,
      "loss": 3.4003,
      "step": 1520
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2178258895874023,
      "learning_rate": 4.246067082220244e-05,
      "loss": 3.09,
      "step": 1524
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2495688199996948,
      "learning_rate": 4.244088255664391e-05,
      "loss": 3.362,
      "step": 1528
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2434120178222656,
      "learning_rate": 4.242109429108539e-05,
      "loss": 3.2701,
      "step": 1532
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4451349973678589,
      "learning_rate": 4.240130602552686e-05,
      "loss": 3.3683,
      "step": 1536
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4097487926483154,
      "learning_rate": 4.238151775996834e-05,
      "loss": 3.2426,
      "step": 1540
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2663370370864868,
      "learning_rate": 4.236172949440982e-05,
      "loss": 3.177,
      "step": 1544
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4020271301269531,
      "learning_rate": 4.2341941228851295e-05,
      "loss": 3.3555,
      "step": 1548
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.355491042137146,
      "learning_rate": 4.2322152963292774e-05,
      "loss": 3.2385,
      "step": 1552
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.521461844444275,
      "learning_rate": 4.230236469773425e-05,
      "loss": 3.153,
      "step": 1556
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.278916597366333,
      "learning_rate": 4.228257643217573e-05,
      "loss": 3.3441,
      "step": 1560
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.2973657846450806,
      "learning_rate": 4.226278816661719e-05,
      "loss": 3.3513,
      "step": 1564
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.254542350769043,
      "learning_rate": 4.224299990105867e-05,
      "loss": 3.3336,
      "step": 1568
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.475539207458496,
      "learning_rate": 4.2223211635500146e-05,
      "loss": 3.2832,
      "step": 1572
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3780232667922974,
      "learning_rate": 4.2203423369941626e-05,
      "loss": 3.4761,
      "step": 1576
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.28725266456604,
      "learning_rate": 4.21836351043831e-05,
      "loss": 3.3786,
      "step": 1580
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1149132251739502,
      "learning_rate": 4.216384683882458e-05,
      "loss": 3.4289,
      "step": 1584
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2939023971557617,
      "learning_rate": 4.214405857326605e-05,
      "loss": 3.3372,
      "step": 1588
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.5904688835144043,
      "learning_rate": 4.212427030770753e-05,
      "loss": 3.3784,
      "step": 1592
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3073712587356567,
      "learning_rate": 4.2104482042149004e-05,
      "loss": 3.4208,
      "step": 1596
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3639119863510132,
      "learning_rate": 4.2084693776590483e-05,
      "loss": 3.2751,
      "step": 1600
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1963696479797363,
      "learning_rate": 4.206490551103196e-05,
      "loss": 3.3383,
      "step": 1604
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.503912091255188,
      "learning_rate": 4.2045117245473436e-05,
      "loss": 3.3637,
      "step": 1608
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4482710361480713,
      "learning_rate": 4.2025328979914916e-05,
      "loss": 3.5253,
      "step": 1612
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.301410436630249,
      "learning_rate": 4.200554071435639e-05,
      "loss": 3.3642,
      "step": 1616
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2556508779525757,
      "learning_rate": 4.198575244879787e-05,
      "loss": 3.2129,
      "step": 1620
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2119685411453247,
      "learning_rate": 4.196596418323934e-05,
      "loss": 3.2957,
      "step": 1624
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.4289143085479736,
      "learning_rate": 4.194617591768082e-05,
      "loss": 3.5322,
      "step": 1628
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.314368486404419,
      "learning_rate": 4.1926387652122294e-05,
      "loss": 3.1358,
      "step": 1632
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2228057384490967,
      "learning_rate": 4.1906599386563774e-05,
      "loss": 3.4493,
      "step": 1636
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2432242631912231,
      "learning_rate": 4.188681112100525e-05,
      "loss": 3.2671,
      "step": 1640
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.34592866897583,
      "learning_rate": 4.1867022855446726e-05,
      "loss": 3.3529,
      "step": 1644
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.360672116279602,
      "learning_rate": 4.184723458988819e-05,
      "loss": 3.1993,
      "step": 1648
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.199696660041809,
      "learning_rate": 4.182744632432967e-05,
      "loss": 3.376,
      "step": 1652
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.3399593830108643,
      "learning_rate": 4.180765805877115e-05,
      "loss": 3.4037,
      "step": 1656
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.1756970882415771,
      "learning_rate": 4.1787869793212625e-05,
      "loss": 3.3409,
      "step": 1660
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.292723298072815,
      "learning_rate": 4.1768081527654105e-05,
      "loss": 3.3704,
      "step": 1664
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.324936032295227,
      "learning_rate": 4.174829326209558e-05,
      "loss": 3.3542,
      "step": 1668
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.5456078052520752,
      "learning_rate": 4.172850499653706e-05,
      "loss": 3.2134,
      "step": 1672
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.175718903541565,
      "learning_rate": 4.170871673097853e-05,
      "loss": 3.1983,
      "step": 1676
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1541227102279663,
      "learning_rate": 4.168892846542001e-05,
      "loss": 3.1252,
      "step": 1680
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4465668201446533,
      "learning_rate": 4.166914019986148e-05,
      "loss": 3.2788,
      "step": 1684
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.340494155883789,
      "learning_rate": 4.164935193430296e-05,
      "loss": 3.5389,
      "step": 1688
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2862470149993896,
      "learning_rate": 4.1629563668744435e-05,
      "loss": 3.3313,
      "step": 1692
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4783071279525757,
      "learning_rate": 4.1609775403185915e-05,
      "loss": 3.2759,
      "step": 1696
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.1228585243225098,
      "learning_rate": 4.158998713762739e-05,
      "loss": 3.3506,
      "step": 1700
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2861573696136475,
      "learning_rate": 4.157019887206887e-05,
      "loss": 3.3374,
      "step": 1704
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3766535520553589,
      "learning_rate": 4.155041060651034e-05,
      "loss": 3.264,
      "step": 1708
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4737287759780884,
      "learning_rate": 4.153062234095182e-05,
      "loss": 3.4986,
      "step": 1712
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3399109840393066,
      "learning_rate": 4.1510834075393293e-05,
      "loss": 3.5896,
      "step": 1716
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3930294513702393,
      "learning_rate": 4.149104580983477e-05,
      "loss": 3.2878,
      "step": 1720
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2427089214324951,
      "learning_rate": 4.1471257544276246e-05,
      "loss": 3.3721,
      "step": 1724
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3937251567840576,
      "learning_rate": 4.1451469278717726e-05,
      "loss": 3.3628,
      "step": 1728
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.4732547998428345,
      "learning_rate": 4.14316810131592e-05,
      "loss": 3.3437,
      "step": 1732
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2372941970825195,
      "learning_rate": 4.141189274760067e-05,
      "loss": 3.2885,
      "step": 1736
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3826080560684204,
      "learning_rate": 4.139210448204215e-05,
      "loss": 3.2253,
      "step": 1740
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3599334955215454,
      "learning_rate": 4.1372316216483624e-05,
      "loss": 3.3872,
      "step": 1744
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.286747694015503,
      "learning_rate": 4.1352527950925104e-05,
      "loss": 3.4709,
      "step": 1748
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3777498006820679,
      "learning_rate": 4.133273968536658e-05,
      "loss": 3.2032,
      "step": 1752
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.3295429944992065,
      "learning_rate": 4.131295141980806e-05,
      "loss": 3.4612,
      "step": 1756
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2335968017578125,
      "learning_rate": 4.129316315424953e-05,
      "loss": 3.2492,
      "step": 1760
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.282143235206604,
      "learning_rate": 4.127337488869101e-05,
      "loss": 3.3274,
      "step": 1764
    },
    {
      "epoch": 0.17,
      "grad_norm": 1.2488516569137573,
      "learning_rate": 4.125358662313248e-05,
      "loss": 3.3067,
      "step": 1768
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3893166780471802,
      "learning_rate": 4.123379835757396e-05,
      "loss": 3.1867,
      "step": 1772
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3564475774765015,
      "learning_rate": 4.1214010092015435e-05,
      "loss": 3.2346,
      "step": 1776
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2899243831634521,
      "learning_rate": 4.1194221826456915e-05,
      "loss": 3.3261,
      "step": 1780
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.25933837890625,
      "learning_rate": 4.117443356089839e-05,
      "loss": 3.3999,
      "step": 1784
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4917421340942383,
      "learning_rate": 4.115464529533987e-05,
      "loss": 3.5482,
      "step": 1788
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2294834852218628,
      "learning_rate": 4.113485702978134e-05,
      "loss": 3.1481,
      "step": 1792
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3130604028701782,
      "learning_rate": 4.111506876422282e-05,
      "loss": 3.5579,
      "step": 1796
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.234863519668579,
      "learning_rate": 4.109528049866429e-05,
      "loss": 3.2204,
      "step": 1800
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.606605052947998,
      "learning_rate": 4.107549223310577e-05,
      "loss": 3.3306,
      "step": 1804
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2031803131103516,
      "learning_rate": 4.1055703967547245e-05,
      "loss": 3.3349,
      "step": 1808
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.354557991027832,
      "learning_rate": 4.1035915701988725e-05,
      "loss": 3.1236,
      "step": 1812
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3591339588165283,
      "learning_rate": 4.10161274364302e-05,
      "loss": 3.1869,
      "step": 1816
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.277288556098938,
      "learning_rate": 4.099633917087167e-05,
      "loss": 3.0739,
      "step": 1820
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4283719062805176,
      "learning_rate": 4.097655090531315e-05,
      "loss": 3.3606,
      "step": 1824
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3387796878814697,
      "learning_rate": 4.0956762639754624e-05,
      "loss": 3.3761,
      "step": 1828
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3217228651046753,
      "learning_rate": 4.09369743741961e-05,
      "loss": 3.3802,
      "step": 1832
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2054520845413208,
      "learning_rate": 4.0917186108637576e-05,
      "loss": 3.1521,
      "step": 1836
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.6184253692626953,
      "learning_rate": 4.0897397843079056e-05,
      "loss": 3.3772,
      "step": 1840
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.4544942378997803,
      "learning_rate": 4.087760957752053e-05,
      "loss": 3.2255,
      "step": 1844
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.247625708580017,
      "learning_rate": 4.085782131196201e-05,
      "loss": 3.2605,
      "step": 1848
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2736012935638428,
      "learning_rate": 4.083803304640348e-05,
      "loss": 3.1793,
      "step": 1852
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2771648168563843,
      "learning_rate": 4.081824478084496e-05,
      "loss": 3.1654,
      "step": 1856
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.1745537519454956,
      "learning_rate": 4.0798456515286434e-05,
      "loss": 3.1628,
      "step": 1860
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.2880799770355225,
      "learning_rate": 4.0778668249727914e-05,
      "loss": 3.2518,
      "step": 1864
    },
    {
      "epoch": 0.18,
      "grad_norm": 1.3577632904052734,
      "learning_rate": 4.075887998416939e-05,
      "loss": 3.171,
      "step": 1868
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3156131505966187,
      "learning_rate": 4.0739091718610867e-05,
      "loss": 3.4448,
      "step": 1872
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2552664279937744,
      "learning_rate": 4.0719303453052346e-05,
      "loss": 3.3617,
      "step": 1876
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3123071193695068,
      "learning_rate": 4.069951518749382e-05,
      "loss": 3.271,
      "step": 1880
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3457539081573486,
      "learning_rate": 4.06797269219353e-05,
      "loss": 3.2374,
      "step": 1884
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3469061851501465,
      "learning_rate": 4.065993865637677e-05,
      "loss": 3.2502,
      "step": 1888
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2933361530303955,
      "learning_rate": 4.064015039081825e-05,
      "loss": 3.1439,
      "step": 1892
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2762932777404785,
      "learning_rate": 4.0620362125259724e-05,
      "loss": 3.1837,
      "step": 1896
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3552213907241821,
      "learning_rate": 4.06005738597012e-05,
      "loss": 3.1888,
      "step": 1900
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3573856353759766,
      "learning_rate": 4.058078559414267e-05,
      "loss": 3.1358,
      "step": 1904
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.308092474937439,
      "learning_rate": 4.056099732858415e-05,
      "loss": 3.3578,
      "step": 1908
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4167733192443848,
      "learning_rate": 4.054120906302562e-05,
      "loss": 3.4611,
      "step": 1912
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3537653684616089,
      "learning_rate": 4.05214207974671e-05,
      "loss": 3.224,
      "step": 1916
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4007222652435303,
      "learning_rate": 4.0501632531908576e-05,
      "loss": 3.2109,
      "step": 1920
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3303008079528809,
      "learning_rate": 4.0481844266350055e-05,
      "loss": 3.4183,
      "step": 1924
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.5393402576446533,
      "learning_rate": 4.0462056000791535e-05,
      "loss": 3.1778,
      "step": 1928
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2925307750701904,
      "learning_rate": 4.044226773523301e-05,
      "loss": 3.3246,
      "step": 1932
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.2742151021957397,
      "learning_rate": 4.042247946967449e-05,
      "loss": 3.2618,
      "step": 1936
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.341132402420044,
      "learning_rate": 4.040269120411596e-05,
      "loss": 3.2285,
      "step": 1940
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.279293179512024,
      "learning_rate": 4.038290293855744e-05,
      "loss": 3.2013,
      "step": 1944
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.304842233657837,
      "learning_rate": 4.036311467299891e-05,
      "loss": 3.3386,
      "step": 1948
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.4351922273635864,
      "learning_rate": 4.034332640744039e-05,
      "loss": 3.3236,
      "step": 1952
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3184465169906616,
      "learning_rate": 4.0323538141881866e-05,
      "loss": 3.3001,
      "step": 1956
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.358441948890686,
      "learning_rate": 4.0303749876323346e-05,
      "loss": 3.3153,
      "step": 1960
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.3413522243499756,
      "learning_rate": 4.028396161076482e-05,
      "loss": 3.2905,
      "step": 1964
    },
    {
      "epoch": 0.19,
      "grad_norm": 1.5526001453399658,
      "learning_rate": 4.02641733452063e-05,
      "loss": 3.2665,
      "step": 1968
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4052098989486694,
      "learning_rate": 4.024438507964777e-05,
      "loss": 3.0877,
      "step": 1972
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.362823724746704,
      "learning_rate": 4.022459681408925e-05,
      "loss": 3.0379,
      "step": 1976
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3832095861434937,
      "learning_rate": 4.0204808548530724e-05,
      "loss": 3.2831,
      "step": 1980
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3092079162597656,
      "learning_rate": 4.01850202829722e-05,
      "loss": 3.2422,
      "step": 1984
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3156044483184814,
      "learning_rate": 4.0165232017413676e-05,
      "loss": 3.2832,
      "step": 1988
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2616153955459595,
      "learning_rate": 4.014544375185515e-05,
      "loss": 3.2577,
      "step": 1992
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3820970058441162,
      "learning_rate": 4.012565548629663e-05,
      "loss": 3.1686,
      "step": 1996
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3397316932678223,
      "learning_rate": 4.01058672207381e-05,
      "loss": 3.1746,
      "step": 2000
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2949211597442627,
      "learning_rate": 4.008607895517958e-05,
      "loss": 3.1891,
      "step": 2004
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.457186222076416,
      "learning_rate": 4.0066290689621055e-05,
      "loss": 3.2126,
      "step": 2008
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3785138130187988,
      "learning_rate": 4.0046502424062534e-05,
      "loss": 3.3084,
      "step": 2012
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2410658597946167,
      "learning_rate": 4.002671415850401e-05,
      "loss": 3.2643,
      "step": 2016
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4620486497879028,
      "learning_rate": 4.000692589294549e-05,
      "loss": 3.1098,
      "step": 2020
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2891669273376465,
      "learning_rate": 3.998713762738696e-05,
      "loss": 3.1495,
      "step": 2024
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4822944402694702,
      "learning_rate": 3.996734936182844e-05,
      "loss": 3.2116,
      "step": 2028
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3625112771987915,
      "learning_rate": 3.994756109626991e-05,
      "loss": 3.2135,
      "step": 2032
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4246516227722168,
      "learning_rate": 3.992777283071139e-05,
      "loss": 3.0766,
      "step": 2036
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4355683326721191,
      "learning_rate": 3.9907984565152865e-05,
      "loss": 3.1343,
      "step": 2040
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3231664896011353,
      "learning_rate": 3.9888196299594345e-05,
      "loss": 2.9431,
      "step": 2044
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5020241737365723,
      "learning_rate": 3.986840803403582e-05,
      "loss": 3.3893,
      "step": 2048
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5987601280212402,
      "learning_rate": 3.98486197684773e-05,
      "loss": 3.3362,
      "step": 2052
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.5132125616073608,
      "learning_rate": 3.982883150291877e-05,
      "loss": 3.5167,
      "step": 2056
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.2821168899536133,
      "learning_rate": 3.980904323736025e-05,
      "loss": 3.2242,
      "step": 2060
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.3775780200958252,
      "learning_rate": 3.978925497180172e-05,
      "loss": 3.2939,
      "step": 2064
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.260902762413025,
      "learning_rate": 3.9769466706243196e-05,
      "loss": 3.0705,
      "step": 2068
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3593419790267944,
      "learning_rate": 3.9749678440684676e-05,
      "loss": 3.3769,
      "step": 2072
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4187499284744263,
      "learning_rate": 3.972989017512615e-05,
      "loss": 3.0845,
      "step": 2076
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3207932710647583,
      "learning_rate": 3.971010190956763e-05,
      "loss": 3.062,
      "step": 2080
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2566810846328735,
      "learning_rate": 3.96903136440091e-05,
      "loss": 3.2373,
      "step": 2084
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4184499979019165,
      "learning_rate": 3.967052537845058e-05,
      "loss": 3.243,
      "step": 2088
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.358438491821289,
      "learning_rate": 3.9650737112892054e-05,
      "loss": 3.1513,
      "step": 2092
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.1665936708450317,
      "learning_rate": 3.9630948847333534e-05,
      "loss": 3.1287,
      "step": 2096
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3452047109603882,
      "learning_rate": 3.961116058177501e-05,
      "loss": 3.2622,
      "step": 2100
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.434053659439087,
      "learning_rate": 3.9591372316216486e-05,
      "loss": 3.3259,
      "step": 2104
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3744926452636719,
      "learning_rate": 3.957158405065796e-05,
      "loss": 3.2474,
      "step": 2108
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.656575322151184,
      "learning_rate": 3.955179578509944e-05,
      "loss": 3.1016,
      "step": 2112
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3220798969268799,
      "learning_rate": 3.953200751954091e-05,
      "loss": 3.167,
      "step": 2116
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.250577449798584,
      "learning_rate": 3.951221925398239e-05,
      "loss": 3.0891,
      "step": 2120
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4257659912109375,
      "learning_rate": 3.9492430988423865e-05,
      "loss": 3.2264,
      "step": 2124
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.6712124347686768,
      "learning_rate": 3.9472642722865344e-05,
      "loss": 3.2472,
      "step": 2128
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.240073323249817,
      "learning_rate": 3.945285445730682e-05,
      "loss": 3.2035,
      "step": 2132
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3379197120666504,
      "learning_rate": 3.94330661917483e-05,
      "loss": 3.1321,
      "step": 2136
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.395456314086914,
      "learning_rate": 3.941327792618978e-05,
      "loss": 3.2416,
      "step": 2140
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4799938201904297,
      "learning_rate": 3.939348966063125e-05,
      "loss": 3.1631,
      "step": 2144
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4667563438415527,
      "learning_rate": 3.937370139507273e-05,
      "loss": 3.247,
      "step": 2148
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.5173306465148926,
      "learning_rate": 3.93539131295142e-05,
      "loss": 3.242,
      "step": 2152
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4049328565597534,
      "learning_rate": 3.9334124863955675e-05,
      "loss": 3.2837,
      "step": 2156
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.4764662981033325,
      "learning_rate": 3.931433659839715e-05,
      "loss": 3.201,
      "step": 2160
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3218275308609009,
      "learning_rate": 3.929454833283863e-05,
      "loss": 3.0854,
      "step": 2164
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.3813103437423706,
      "learning_rate": 3.92747600672801e-05,
      "loss": 3.1529,
      "step": 2168
    },
    {
      "epoch": 0.21,
      "grad_norm": 1.2131150960922241,
      "learning_rate": 3.925497180172158e-05,
      "loss": 3.0118,
      "step": 2172
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3104761838912964,
      "learning_rate": 3.9235183536163053e-05,
      "loss": 3.0601,
      "step": 2176
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6685932874679565,
      "learning_rate": 3.921539527060453e-05,
      "loss": 3.1585,
      "step": 2180
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3816349506378174,
      "learning_rate": 3.9195607005046006e-05,
      "loss": 3.2128,
      "step": 2184
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.293393611907959,
      "learning_rate": 3.9175818739487486e-05,
      "loss": 3.2281,
      "step": 2188
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4393213987350464,
      "learning_rate": 3.915603047392896e-05,
      "loss": 3.298,
      "step": 2192
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3526524305343628,
      "learning_rate": 3.913624220837044e-05,
      "loss": 3.0613,
      "step": 2196
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4308154582977295,
      "learning_rate": 3.911645394281192e-05,
      "loss": 3.2307,
      "step": 2200
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.688053011894226,
      "learning_rate": 3.909666567725339e-05,
      "loss": 3.1716,
      "step": 2204
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.5624258518218994,
      "learning_rate": 3.907687741169487e-05,
      "loss": 3.1115,
      "step": 2208
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3107057809829712,
      "learning_rate": 3.9057089146136344e-05,
      "loss": 3.3096,
      "step": 2212
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3054726123809814,
      "learning_rate": 3.9037300880577823e-05,
      "loss": 3.0534,
      "step": 2216
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2478541135787964,
      "learning_rate": 3.9017512615019296e-05,
      "loss": 3.2014,
      "step": 2220
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2822325229644775,
      "learning_rate": 3.8997724349460776e-05,
      "loss": 3.1986,
      "step": 2224
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.383755087852478,
      "learning_rate": 3.897793608390225e-05,
      "loss": 3.1122,
      "step": 2228
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3420864343643188,
      "learning_rate": 3.895814781834373e-05,
      "loss": 3.0604,
      "step": 2232
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3208297491073608,
      "learning_rate": 3.89383595527852e-05,
      "loss": 3.0086,
      "step": 2236
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3529295921325684,
      "learning_rate": 3.8918571287226675e-05,
      "loss": 3.146,
      "step": 2240
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4611573219299316,
      "learning_rate": 3.889878302166815e-05,
      "loss": 3.1959,
      "step": 2244
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.354971170425415,
      "learning_rate": 3.887899475610963e-05,
      "loss": 2.8893,
      "step": 2248
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.6441867351531982,
      "learning_rate": 3.88592064905511e-05,
      "loss": 3.2257,
      "step": 2252
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3111629486083984,
      "learning_rate": 3.883941822499258e-05,
      "loss": 3.0993,
      "step": 2256
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.3077386617660522,
      "learning_rate": 3.881962995943406e-05,
      "loss": 3.088,
      "step": 2260
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.331356167793274,
      "learning_rate": 3.879984169387553e-05,
      "loss": 3.31,
      "step": 2264
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.4217604398727417,
      "learning_rate": 3.878005342831701e-05,
      "loss": 3.1174,
      "step": 2268
    },
    {
      "epoch": 0.22,
      "grad_norm": 1.2443788051605225,
      "learning_rate": 3.8760265162758485e-05,
      "loss": 3.11,
      "step": 2272
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.288475751876831,
      "learning_rate": 3.8740476897199965e-05,
      "loss": 3.2074,
      "step": 2276
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.427825689315796,
      "learning_rate": 3.872068863164144e-05,
      "loss": 3.0412,
      "step": 2280
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4200148582458496,
      "learning_rate": 3.870090036608292e-05,
      "loss": 3.1752,
      "step": 2284
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.460418939590454,
      "learning_rate": 3.868111210052439e-05,
      "loss": 3.2623,
      "step": 2288
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.38273024559021,
      "learning_rate": 3.866132383496587e-05,
      "loss": 3.0194,
      "step": 2292
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5908851623535156,
      "learning_rate": 3.864153556940734e-05,
      "loss": 3.1621,
      "step": 2296
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4536399841308594,
      "learning_rate": 3.862174730384882e-05,
      "loss": 3.1168,
      "step": 2300
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3599523305892944,
      "learning_rate": 3.8601959038290296e-05,
      "loss": 3.2446,
      "step": 2304
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3411145210266113,
      "learning_rate": 3.8582170772731775e-05,
      "loss": 3.0645,
      "step": 2308
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4459635019302368,
      "learning_rate": 3.856238250717325e-05,
      "loss": 3.0084,
      "step": 2312
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.358909249305725,
      "learning_rate": 3.854259424161473e-05,
      "loss": 3.1009,
      "step": 2316
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5240005254745483,
      "learning_rate": 3.85228059760562e-05,
      "loss": 3.1275,
      "step": 2320
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4764708280563354,
      "learning_rate": 3.8503017710497674e-05,
      "loss": 3.0676,
      "step": 2324
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3841642141342163,
      "learning_rate": 3.8483229444939154e-05,
      "loss": 3.066,
      "step": 2328
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3241703510284424,
      "learning_rate": 3.8463441179380627e-05,
      "loss": 3.1452,
      "step": 2332
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4375591278076172,
      "learning_rate": 3.8443652913822106e-05,
      "loss": 2.9545,
      "step": 2336
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4808233976364136,
      "learning_rate": 3.842386464826358e-05,
      "loss": 3.2496,
      "step": 2340
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2886080741882324,
      "learning_rate": 3.840407638270506e-05,
      "loss": 3.1226,
      "step": 2344
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.4224919080734253,
      "learning_rate": 3.838428811714653e-05,
      "loss": 3.196,
      "step": 2348
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2736425399780273,
      "learning_rate": 3.836449985158801e-05,
      "loss": 3.1041,
      "step": 2352
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2044873237609863,
      "learning_rate": 3.8344711586029485e-05,
      "loss": 3.1979,
      "step": 2356
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.2397358417510986,
      "learning_rate": 3.8324923320470964e-05,
      "loss": 3.0942,
      "step": 2360
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.3050447702407837,
      "learning_rate": 3.830513505491244e-05,
      "loss": 3.1241,
      "step": 2364
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.239399790763855,
      "learning_rate": 3.828534678935392e-05,
      "loss": 3.0856,
      "step": 2368
    },
    {
      "epoch": 0.23,
      "grad_norm": 1.5021380186080933,
      "learning_rate": 3.826555852379539e-05,
      "loss": 3.0362,
      "step": 2372
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2373955249786377,
      "learning_rate": 3.824577025823687e-05,
      "loss": 3.0351,
      "step": 2376
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3865991830825806,
      "learning_rate": 3.822598199267834e-05,
      "loss": 3.1084,
      "step": 2380
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2980425357818604,
      "learning_rate": 3.820619372711982e-05,
      "loss": 3.1217,
      "step": 2384
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.5192391872406006,
      "learning_rate": 3.8186405461561295e-05,
      "loss": 3.1863,
      "step": 2388
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4014787673950195,
      "learning_rate": 3.8166617196002775e-05,
      "loss": 2.9426,
      "step": 2392
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4827275276184082,
      "learning_rate": 3.814682893044425e-05,
      "loss": 3.0311,
      "step": 2396
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.7020756006240845,
      "learning_rate": 3.812704066488573e-05,
      "loss": 3.229,
      "step": 2400
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3895395994186401,
      "learning_rate": 3.81072523993272e-05,
      "loss": 3.1177,
      "step": 2404
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.469364881515503,
      "learning_rate": 3.808746413376867e-05,
      "loss": 3.0913,
      "step": 2408
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2792705297470093,
      "learning_rate": 3.806767586821015e-05,
      "loss": 3.0318,
      "step": 2412
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4794644117355347,
      "learning_rate": 3.8047887602651626e-05,
      "loss": 3.1361,
      "step": 2416
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.6199162006378174,
      "learning_rate": 3.8028099337093106e-05,
      "loss": 3.0314,
      "step": 2420
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.303435206413269,
      "learning_rate": 3.800831107153458e-05,
      "loss": 3.0878,
      "step": 2424
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4103130102157593,
      "learning_rate": 3.798852280597606e-05,
      "loss": 3.1757,
      "step": 2428
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2540159225463867,
      "learning_rate": 3.796873454041753e-05,
      "loss": 3.2547,
      "step": 2432
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3388727903366089,
      "learning_rate": 3.794894627485901e-05,
      "loss": 3.2623,
      "step": 2436
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.2804844379425049,
      "learning_rate": 3.7929158009300484e-05,
      "loss": 3.0305,
      "step": 2440
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4325406551361084,
      "learning_rate": 3.7909369743741964e-05,
      "loss": 3.1235,
      "step": 2444
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4079090356826782,
      "learning_rate": 3.7889581478183437e-05,
      "loss": 3.0626,
      "step": 2448
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3477818965911865,
      "learning_rate": 3.7869793212624916e-05,
      "loss": 3.0653,
      "step": 2452
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.4144619703292847,
      "learning_rate": 3.785000494706639e-05,
      "loss": 3.0494,
      "step": 2456
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.279802680015564,
      "learning_rate": 3.783021668150787e-05,
      "loss": 3.0711,
      "step": 2460
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.32356858253479,
      "learning_rate": 3.781042841594934e-05,
      "loss": 3.0562,
      "step": 2464
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.3349965810775757,
      "learning_rate": 3.779064015039082e-05,
      "loss": 3.0004,
      "step": 2468
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.32090425491333,
      "learning_rate": 3.77708518848323e-05,
      "loss": 3.2252,
      "step": 2472
    },
    {
      "epoch": 0.24,
      "grad_norm": 1.335302472114563,
      "learning_rate": 3.7751063619273774e-05,
      "loss": 3.2227,
      "step": 2476
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4220561981201172,
      "learning_rate": 3.7731275353715254e-05,
      "loss": 3.0286,
      "step": 2480
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3459542989730835,
      "learning_rate": 3.771148708815673e-05,
      "loss": 3.0613,
      "step": 2484
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3894877433776855,
      "learning_rate": 3.7691698822598206e-05,
      "loss": 3.1524,
      "step": 2488
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3688431978225708,
      "learning_rate": 3.767191055703967e-05,
      "loss": 3.2377,
      "step": 2492
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2429877519607544,
      "learning_rate": 3.765212229148115e-05,
      "loss": 3.0412,
      "step": 2496
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6966149806976318,
      "learning_rate": 3.7632334025922625e-05,
      "loss": 2.9925,
      "step": 2500
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4553391933441162,
      "learning_rate": 3.7612545760364105e-05,
      "loss": 3.1891,
      "step": 2504
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3509441614151,
      "learning_rate": 3.759275749480558e-05,
      "loss": 3.0536,
      "step": 2508
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3694161176681519,
      "learning_rate": 3.757296922924706e-05,
      "loss": 3.2306,
      "step": 2512
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.6577290296554565,
      "learning_rate": 3.755318096368853e-05,
      "loss": 3.1636,
      "step": 2516
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3860623836517334,
      "learning_rate": 3.753339269813001e-05,
      "loss": 3.1207,
      "step": 2520
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5849056243896484,
      "learning_rate": 3.751360443257148e-05,
      "loss": 3.0021,
      "step": 2524
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4191337823867798,
      "learning_rate": 3.749381616701296e-05,
      "loss": 3.1331,
      "step": 2528
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3399465084075928,
      "learning_rate": 3.747402790145444e-05,
      "loss": 3.2503,
      "step": 2532
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.290664553642273,
      "learning_rate": 3.7454239635895916e-05,
      "loss": 3.114,
      "step": 2536
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2612792253494263,
      "learning_rate": 3.7434451370337395e-05,
      "loss": 3.1496,
      "step": 2540
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4211294651031494,
      "learning_rate": 3.741466310477887e-05,
      "loss": 3.0861,
      "step": 2544
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4412263631820679,
      "learning_rate": 3.739487483922035e-05,
      "loss": 3.1606,
      "step": 2548
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3415204286575317,
      "learning_rate": 3.737508657366182e-05,
      "loss": 3.0834,
      "step": 2552
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.5815222263336182,
      "learning_rate": 3.73552983081033e-05,
      "loss": 2.9571,
      "step": 2556
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.2796626091003418,
      "learning_rate": 3.7335510042544774e-05,
      "loss": 3.0611,
      "step": 2560
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.4062576293945312,
      "learning_rate": 3.731572177698625e-05,
      "loss": 3.1557,
      "step": 2564
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3327045440673828,
      "learning_rate": 3.7295933511427726e-05,
      "loss": 2.9271,
      "step": 2568
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.3985902070999146,
      "learning_rate": 3.7276145245869206e-05,
      "loss": 3.0211,
      "step": 2572
    },
    {
      "epoch": 0.25,
      "grad_norm": 1.494410514831543,
      "learning_rate": 3.725635698031067e-05,
      "loss": 2.9367,
      "step": 2576
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3023014068603516,
      "learning_rate": 3.723656871475215e-05,
      "loss": 3.0077,
      "step": 2580
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4200507402420044,
      "learning_rate": 3.7216780449193625e-05,
      "loss": 3.1887,
      "step": 2584
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4320076704025269,
      "learning_rate": 3.7196992183635104e-05,
      "loss": 2.9131,
      "step": 2588
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2773874998092651,
      "learning_rate": 3.7177203918076584e-05,
      "loss": 2.9756,
      "step": 2592
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3249197006225586,
      "learning_rate": 3.715741565251806e-05,
      "loss": 3.0856,
      "step": 2596
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.574328899383545,
      "learning_rate": 3.713762738695954e-05,
      "loss": 2.9265,
      "step": 2600
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2989048957824707,
      "learning_rate": 3.711783912140101e-05,
      "loss": 3.1064,
      "step": 2604
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2440621852874756,
      "learning_rate": 3.709805085584249e-05,
      "loss": 3.096,
      "step": 2608
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2897485494613647,
      "learning_rate": 3.707826259028396e-05,
      "loss": 2.9686,
      "step": 2612
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4826507568359375,
      "learning_rate": 3.705847432472544e-05,
      "loss": 3.0006,
      "step": 2616
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1983760595321655,
      "learning_rate": 3.7038686059166915e-05,
      "loss": 3.1168,
      "step": 2620
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.8693993091583252,
      "learning_rate": 3.7018897793608395e-05,
      "loss": 3.1812,
      "step": 2624
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.3697383403778076,
      "learning_rate": 3.699910952804987e-05,
      "loss": 3.0964,
      "step": 2628
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4124512672424316,
      "learning_rate": 3.697932126249135e-05,
      "loss": 3.0129,
      "step": 2632
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.1903616189956665,
      "learning_rate": 3.695953299693282e-05,
      "loss": 2.9507,
      "step": 2636
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.5959855318069458,
      "learning_rate": 3.69397447313743e-05,
      "loss": 3.0499,
      "step": 2640
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.506282091140747,
      "learning_rate": 3.691995646581577e-05,
      "loss": 3.0591,
      "step": 2644
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.6031521558761597,
      "learning_rate": 3.690016820025725e-05,
      "loss": 2.9865,
      "step": 2648
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.400120735168457,
      "learning_rate": 3.6880379934698726e-05,
      "loss": 3.0445,
      "step": 2652
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.300150752067566,
      "learning_rate": 3.6860591669140205e-05,
      "loss": 3.0002,
      "step": 2656
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2600607872009277,
      "learning_rate": 3.684080340358168e-05,
      "loss": 3.0244,
      "step": 2660
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.358891487121582,
      "learning_rate": 3.682101513802315e-05,
      "loss": 3.1586,
      "step": 2664
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.4360027313232422,
      "learning_rate": 3.680122687246463e-05,
      "loss": 3.2983,
      "step": 2668
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2610455751419067,
      "learning_rate": 3.6781438606906104e-05,
      "loss": 3.0377,
      "step": 2672
    },
    {
      "epoch": 0.26,
      "grad_norm": 1.2880274057388306,
      "learning_rate": 3.6761650341347583e-05,
      "loss": 3.0465,
      "step": 2676
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.603733777999878,
      "learning_rate": 3.6741862075789056e-05,
      "loss": 2.9944,
      "step": 2680
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4317169189453125,
      "learning_rate": 3.6722073810230536e-05,
      "loss": 3.1036,
      "step": 2684
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2459261417388916,
      "learning_rate": 3.670228554467201e-05,
      "loss": 2.8453,
      "step": 2688
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4440115690231323,
      "learning_rate": 3.668249727911349e-05,
      "loss": 3.0867,
      "step": 2692
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3969016075134277,
      "learning_rate": 3.666270901355496e-05,
      "loss": 3.0322,
      "step": 2696
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2696243524551392,
      "learning_rate": 3.664292074799644e-05,
      "loss": 3.1173,
      "step": 2700
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.5136852264404297,
      "learning_rate": 3.6623132482437914e-05,
      "loss": 3.021,
      "step": 2704
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4152220487594604,
      "learning_rate": 3.6603344216879394e-05,
      "loss": 2.925,
      "step": 2708
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.7811540365219116,
      "learning_rate": 3.658355595132087e-05,
      "loss": 3.0219,
      "step": 2712
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.384444236755371,
      "learning_rate": 3.656376768576235e-05,
      "loss": 3.036,
      "step": 2716
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4472016096115112,
      "learning_rate": 3.654397942020382e-05,
      "loss": 3.1301,
      "step": 2720
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2954609394073486,
      "learning_rate": 3.65241911546453e-05,
      "loss": 3.0274,
      "step": 2724
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3119304180145264,
      "learning_rate": 3.650440288908677e-05,
      "loss": 3.0739,
      "step": 2728
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.405961513519287,
      "learning_rate": 3.648461462352825e-05,
      "loss": 2.9857,
      "step": 2732
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.532257318496704,
      "learning_rate": 3.6464826357969725e-05,
      "loss": 3.0947,
      "step": 2736
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4909120798110962,
      "learning_rate": 3.6445038092411205e-05,
      "loss": 3.1899,
      "step": 2740
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2964402437210083,
      "learning_rate": 3.642524982685268e-05,
      "loss": 2.9136,
      "step": 2744
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3999255895614624,
      "learning_rate": 3.640546156129415e-05,
      "loss": 3.1076,
      "step": 2748
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.310899257659912,
      "learning_rate": 3.638567329573563e-05,
      "loss": 3.2241,
      "step": 2752
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.2926506996154785,
      "learning_rate": 3.63658850301771e-05,
      "loss": 2.9607,
      "step": 2756
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4596292972564697,
      "learning_rate": 3.634609676461858e-05,
      "loss": 2.9604,
      "step": 2760
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.4320791959762573,
      "learning_rate": 3.6326308499060056e-05,
      "loss": 3.1414,
      "step": 2764
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3382930755615234,
      "learning_rate": 3.6306520233501535e-05,
      "loss": 2.8968,
      "step": 2768
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.3211889266967773,
      "learning_rate": 3.628673196794301e-05,
      "loss": 2.8903,
      "step": 2772
    },
    {
      "epoch": 0.27,
      "grad_norm": 1.323534369468689,
      "learning_rate": 3.626694370238449e-05,
      "loss": 2.9924,
      "step": 2776
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.314077377319336,
      "learning_rate": 3.624715543682596e-05,
      "loss": 2.9786,
      "step": 2780
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.7178696393966675,
      "learning_rate": 3.622736717126744e-05,
      "loss": 3.1973,
      "step": 2784
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5973659753799438,
      "learning_rate": 3.6207578905708914e-05,
      "loss": 2.9483,
      "step": 2788
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.4203895330429077,
      "learning_rate": 3.618779064015039e-05,
      "loss": 2.9089,
      "step": 2792
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.486110806465149,
      "learning_rate": 3.6168002374591866e-05,
      "loss": 3.0919,
      "step": 2796
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.4070860147476196,
      "learning_rate": 3.6148214109033346e-05,
      "loss": 3.0393,
      "step": 2800
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.4570317268371582,
      "learning_rate": 3.6128425843474826e-05,
      "loss": 3.0708,
      "step": 2804
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.8382024765014648,
      "learning_rate": 3.61086375779163e-05,
      "loss": 3.1437,
      "step": 2808
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3467711210250854,
      "learning_rate": 3.608884931235778e-05,
      "loss": 3.047,
      "step": 2812
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.209226369857788,
      "learning_rate": 3.606906104679925e-05,
      "loss": 2.9569,
      "step": 2816
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.458509087562561,
      "learning_rate": 3.604927278124073e-05,
      "loss": 3.0049,
      "step": 2820
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.4587996006011963,
      "learning_rate": 3.6029484515682204e-05,
      "loss": 2.9481,
      "step": 2824
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2717212438583374,
      "learning_rate": 3.600969625012368e-05,
      "loss": 2.9896,
      "step": 2828
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3680723905563354,
      "learning_rate": 3.598990798456515e-05,
      "loss": 3.0087,
      "step": 2832
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3884750604629517,
      "learning_rate": 3.597011971900663e-05,
      "loss": 3.0768,
      "step": 2836
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5019512176513672,
      "learning_rate": 3.59503314534481e-05,
      "loss": 2.9614,
      "step": 2840
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.706018328666687,
      "learning_rate": 3.593054318788958e-05,
      "loss": 2.9818,
      "step": 2844
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3418842554092407,
      "learning_rate": 3.5910754922331055e-05,
      "loss": 2.9335,
      "step": 2848
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.6864101886749268,
      "learning_rate": 3.5890966656772535e-05,
      "loss": 3.1771,
      "step": 2852
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.488398551940918,
      "learning_rate": 3.587117839121401e-05,
      "loss": 2.9231,
      "step": 2856
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.43323814868927,
      "learning_rate": 3.585139012565549e-05,
      "loss": 2.9957,
      "step": 2860
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.4537419080734253,
      "learning_rate": 3.583160186009697e-05,
      "loss": 2.9353,
      "step": 2864
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.3649722337722778,
      "learning_rate": 3.581181359453844e-05,
      "loss": 2.9827,
      "step": 2868
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.420401692390442,
      "learning_rate": 3.579202532897992e-05,
      "loss": 3.0755,
      "step": 2872
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.5330480337142944,
      "learning_rate": 3.577223706342139e-05,
      "loss": 3.0098,
      "step": 2876
    },
    {
      "epoch": 0.28,
      "grad_norm": 1.2478176355361938,
      "learning_rate": 3.575244879786287e-05,
      "loss": 3.0078,
      "step": 2880
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4124771356582642,
      "learning_rate": 3.5732660532304345e-05,
      "loss": 3.0649,
      "step": 2884
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.5072100162506104,
      "learning_rate": 3.5712872266745825e-05,
      "loss": 3.1248,
      "step": 2888
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3434092998504639,
      "learning_rate": 3.56930840011873e-05,
      "loss": 3.1075,
      "step": 2892
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3640666007995605,
      "learning_rate": 3.567329573562878e-05,
      "loss": 2.9797,
      "step": 2896
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4746694564819336,
      "learning_rate": 3.565350747007025e-05,
      "loss": 2.9813,
      "step": 2900
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4786782264709473,
      "learning_rate": 3.563371920451173e-05,
      "loss": 3.111,
      "step": 2904
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.471091628074646,
      "learning_rate": 3.56139309389532e-05,
      "loss": 2.9263,
      "step": 2908
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.4228413105010986,
      "learning_rate": 3.5594142673394676e-05,
      "loss": 3.1006,
      "step": 2912
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.2829701900482178,
      "learning_rate": 3.557435440783615e-05,
      "loss": 2.9774,
      "step": 2916
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3806427717208862,
      "learning_rate": 3.555456614227763e-05,
      "loss": 3.0992,
      "step": 2920
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.5112413167953491,
      "learning_rate": 3.553477787671911e-05,
      "loss": 2.8947,
      "step": 2924
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3634873628616333,
      "learning_rate": 3.551498961116058e-05,
      "loss": 3.0911,
      "step": 2928
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.299893856048584,
      "learning_rate": 3.549520134560206e-05,
      "loss": 3.0254,
      "step": 2932
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.524451494216919,
      "learning_rate": 3.5475413080043534e-05,
      "loss": 2.9585,
      "step": 2936
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3160494565963745,
      "learning_rate": 3.5455624814485014e-05,
      "loss": 2.9677,
      "step": 2940
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.422784686088562,
      "learning_rate": 3.543583654892649e-05,
      "loss": 3.0225,
      "step": 2944
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3876713514328003,
      "learning_rate": 3.5416048283367967e-05,
      "loss": 2.9325,
      "step": 2948
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3324222564697266,
      "learning_rate": 3.539626001780944e-05,
      "loss": 3.0236,
      "step": 2952
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.5146019458770752,
      "learning_rate": 3.537647175225092e-05,
      "loss": 3.0919,
      "step": 2956
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.576621174812317,
      "learning_rate": 3.535668348669239e-05,
      "loss": 3.0733,
      "step": 2960
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.651235580444336,
      "learning_rate": 3.533689522113387e-05,
      "loss": 3.2117,
      "step": 2964
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.377480387687683,
      "learning_rate": 3.5317106955575345e-05,
      "loss": 2.8459,
      "step": 2968
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.462600588798523,
      "learning_rate": 3.5297318690016824e-05,
      "loss": 2.9828,
      "step": 2972
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.320587396621704,
      "learning_rate": 3.52775304244583e-05,
      "loss": 2.9569,
      "step": 2976
    },
    {
      "epoch": 0.29,
      "grad_norm": 1.3665657043457031,
      "learning_rate": 3.525774215889978e-05,
      "loss": 2.9964,
      "step": 2980
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4614739418029785,
      "learning_rate": 3.523795389334125e-05,
      "loss": 2.9624,
      "step": 2984
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.582934021949768,
      "learning_rate": 3.521816562778273e-05,
      "loss": 2.9726,
      "step": 2988
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4328818321228027,
      "learning_rate": 3.51983773622242e-05,
      "loss": 3.0368,
      "step": 2992
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3142259120941162,
      "learning_rate": 3.517858909666568e-05,
      "loss": 2.8137,
      "step": 2996
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4220584630966187,
      "learning_rate": 3.5158800831107155e-05,
      "loss": 3.0077,
      "step": 3000
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.5171457529067993,
      "learning_rate": 3.513901256554863e-05,
      "loss": 2.9489,
      "step": 3004
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.649688720703125,
      "learning_rate": 3.511922429999011e-05,
      "loss": 3.092,
      "step": 3008
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4517191648483276,
      "learning_rate": 3.509943603443158e-05,
      "loss": 2.8212,
      "step": 3012
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4763597249984741,
      "learning_rate": 3.507964776887306e-05,
      "loss": 3.145,
      "step": 3016
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3676544427871704,
      "learning_rate": 3.5059859503314534e-05,
      "loss": 3.0658,
      "step": 3020
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.242475152015686,
      "learning_rate": 3.504007123775601e-05,
      "loss": 2.9705,
      "step": 3024
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.509164810180664,
      "learning_rate": 3.5020282972197486e-05,
      "loss": 3.0112,
      "step": 3028
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3613548278808594,
      "learning_rate": 3.5000494706638966e-05,
      "loss": 2.9177,
      "step": 3032
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3977431058883667,
      "learning_rate": 3.498070644108044e-05,
      "loss": 3.0567,
      "step": 3036
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3139115571975708,
      "learning_rate": 3.496091817552192e-05,
      "loss": 2.9182,
      "step": 3040
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3715747594833374,
      "learning_rate": 3.494112990996339e-05,
      "loss": 3.0362,
      "step": 3044
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.2754712104797363,
      "learning_rate": 3.492134164440487e-05,
      "loss": 2.8626,
      "step": 3048
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.5584040880203247,
      "learning_rate": 3.4901553378846344e-05,
      "loss": 2.7835,
      "step": 3052
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.359312891960144,
      "learning_rate": 3.4881765113287824e-05,
      "loss": 2.9583,
      "step": 3056
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4567357301712036,
      "learning_rate": 3.48619768477293e-05,
      "loss": 2.8088,
      "step": 3060
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3756351470947266,
      "learning_rate": 3.4842188582170776e-05,
      "loss": 2.9143,
      "step": 3064
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4864003658294678,
      "learning_rate": 3.482240031661225e-05,
      "loss": 3.0454,
      "step": 3068
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.4282584190368652,
      "learning_rate": 3.480261205105373e-05,
      "loss": 2.863,
      "step": 3072
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3447134494781494,
      "learning_rate": 3.478282378549521e-05,
      "loss": 3.1251,
      "step": 3076
    },
    {
      "epoch": 0.3,
      "grad_norm": 1.3359801769256592,
      "learning_rate": 3.476303551993668e-05,
      "loss": 3.1131,
      "step": 3080
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.6050612926483154,
      "learning_rate": 3.4743247254378155e-05,
      "loss": 2.8146,
      "step": 3084
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4369301795959473,
      "learning_rate": 3.472345898881963e-05,
      "loss": 3.0191,
      "step": 3088
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4939532279968262,
      "learning_rate": 3.470367072326111e-05,
      "loss": 2.8275,
      "step": 3092
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4874589443206787,
      "learning_rate": 3.468388245770258e-05,
      "loss": 2.9788,
      "step": 3096
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4248600006103516,
      "learning_rate": 3.466409419214406e-05,
      "loss": 2.8865,
      "step": 3100
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3846864700317383,
      "learning_rate": 3.464430592658553e-05,
      "loss": 2.6846,
      "step": 3104
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.660626769065857,
      "learning_rate": 3.462451766102701e-05,
      "loss": 3.0301,
      "step": 3108
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3925195932388306,
      "learning_rate": 3.4604729395468486e-05,
      "loss": 2.893,
      "step": 3112
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4249762296676636,
      "learning_rate": 3.4584941129909965e-05,
      "loss": 2.9668,
      "step": 3116
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4822063446044922,
      "learning_rate": 3.456515286435144e-05,
      "loss": 2.7548,
      "step": 3120
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3979532718658447,
      "learning_rate": 3.454536459879292e-05,
      "loss": 2.9613,
      "step": 3124
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3357387781143188,
      "learning_rate": 3.452557633323439e-05,
      "loss": 3.0212,
      "step": 3128
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.5871549844741821,
      "learning_rate": 3.450578806767587e-05,
      "loss": 2.8224,
      "step": 3132
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.32512366771698,
      "learning_rate": 3.448599980211735e-05,
      "loss": 2.808,
      "step": 3136
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.383267879486084,
      "learning_rate": 3.446621153655882e-05,
      "loss": 2.877,
      "step": 3140
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4139374494552612,
      "learning_rate": 3.44464232710003e-05,
      "loss": 2.7878,
      "step": 3144
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4128443002700806,
      "learning_rate": 3.4426635005441776e-05,
      "loss": 2.7679,
      "step": 3148
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4073232412338257,
      "learning_rate": 3.4406846739883256e-05,
      "loss": 2.8843,
      "step": 3152
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4618744850158691,
      "learning_rate": 3.438705847432473e-05,
      "loss": 3.0241,
      "step": 3156
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.5047119855880737,
      "learning_rate": 3.436727020876621e-05,
      "loss": 3.0856,
      "step": 3160
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.426459789276123,
      "learning_rate": 3.434748194320768e-05,
      "loss": 3.0243,
      "step": 3164
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3475228548049927,
      "learning_rate": 3.4327693677649154e-05,
      "loss": 2.9228,
      "step": 3168
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.3559972047805786,
      "learning_rate": 3.430790541209063e-05,
      "loss": 2.9625,
      "step": 3172
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.4422370195388794,
      "learning_rate": 3.428811714653211e-05,
      "loss": 2.8789,
      "step": 3176
    },
    {
      "epoch": 0.31,
      "grad_norm": 1.404948115348816,
      "learning_rate": 3.426832888097358e-05,
      "loss": 3.017,
      "step": 3180
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3570481538772583,
      "learning_rate": 3.424854061541506e-05,
      "loss": 2.7785,
      "step": 3184
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.295659065246582,
      "learning_rate": 3.422875234985653e-05,
      "loss": 2.7838,
      "step": 3188
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3800030946731567,
      "learning_rate": 3.420896408429801e-05,
      "loss": 2.9911,
      "step": 3192
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3445378541946411,
      "learning_rate": 3.418917581873949e-05,
      "loss": 2.9278,
      "step": 3196
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3806275129318237,
      "learning_rate": 3.4169387553180965e-05,
      "loss": 2.9189,
      "step": 3200
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3769384622573853,
      "learning_rate": 3.4149599287622444e-05,
      "loss": 2.924,
      "step": 3204
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5868099927902222,
      "learning_rate": 3.412981102206392e-05,
      "loss": 2.9324,
      "step": 3208
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3903356790542603,
      "learning_rate": 3.41100227565054e-05,
      "loss": 2.7864,
      "step": 3212
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3730672597885132,
      "learning_rate": 3.409023449094687e-05,
      "loss": 2.934,
      "step": 3216
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3796464204788208,
      "learning_rate": 3.407044622538835e-05,
      "loss": 2.9562,
      "step": 3220
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.617128849029541,
      "learning_rate": 3.405065795982982e-05,
      "loss": 3.0319,
      "step": 3224
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5228337049484253,
      "learning_rate": 3.40308696942713e-05,
      "loss": 2.875,
      "step": 3228
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3389008045196533,
      "learning_rate": 3.4011081428712775e-05,
      "loss": 2.7897,
      "step": 3232
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3617353439331055,
      "learning_rate": 3.3991293163154255e-05,
      "loss": 2.9357,
      "step": 3236
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3685081005096436,
      "learning_rate": 3.397150489759573e-05,
      "loss": 2.9993,
      "step": 3240
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.4341316223144531,
      "learning_rate": 3.395171663203721e-05,
      "loss": 3.1416,
      "step": 3244
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3631387948989868,
      "learning_rate": 3.393192836647868e-05,
      "loss": 2.9022,
      "step": 3248
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3081270456314087,
      "learning_rate": 3.391214010092015e-05,
      "loss": 3.0646,
      "step": 3252
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3105672597885132,
      "learning_rate": 3.389235183536163e-05,
      "loss": 2.8765,
      "step": 3256
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.881330966949463,
      "learning_rate": 3.3872563569803106e-05,
      "loss": 2.818,
      "step": 3260
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.445777416229248,
      "learning_rate": 3.3852775304244586e-05,
      "loss": 2.786,
      "step": 3264
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.4523110389709473,
      "learning_rate": 3.383298703868606e-05,
      "loss": 3.077,
      "step": 3268
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.3693511486053467,
      "learning_rate": 3.381319877312754e-05,
      "loss": 2.9497,
      "step": 3272
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.4519122838974,
      "learning_rate": 3.379341050756901e-05,
      "loss": 2.8887,
      "step": 3276
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.2643674612045288,
      "learning_rate": 3.377362224201049e-05,
      "loss": 2.8879,
      "step": 3280
    },
    {
      "epoch": 0.32,
      "grad_norm": 1.5315420627593994,
      "learning_rate": 3.3753833976451964e-05,
      "loss": 2.8984,
      "step": 3284
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4942806959152222,
      "learning_rate": 3.3734045710893444e-05,
      "loss": 3.0528,
      "step": 3288
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.8208317756652832,
      "learning_rate": 3.3714257445334917e-05,
      "loss": 2.8531,
      "step": 3292
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.6653791666030884,
      "learning_rate": 3.3694469179776396e-05,
      "loss": 2.9974,
      "step": 3296
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4518672227859497,
      "learning_rate": 3.367468091421787e-05,
      "loss": 2.8841,
      "step": 3300
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.373217225074768,
      "learning_rate": 3.365489264865935e-05,
      "loss": 2.8967,
      "step": 3304
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.567406415939331,
      "learning_rate": 3.363510438310082e-05,
      "loss": 3.0588,
      "step": 3308
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.296132206916809,
      "learning_rate": 3.36153161175423e-05,
      "loss": 2.9882,
      "step": 3312
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3636564016342163,
      "learning_rate": 3.3595527851983775e-05,
      "loss": 2.8311,
      "step": 3316
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4256870746612549,
      "learning_rate": 3.3575739586425254e-05,
      "loss": 2.8547,
      "step": 3320
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5391736030578613,
      "learning_rate": 3.355595132086673e-05,
      "loss": 3.0205,
      "step": 3324
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5316370725631714,
      "learning_rate": 3.353616305530821e-05,
      "loss": 2.8755,
      "step": 3328
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5849004983901978,
      "learning_rate": 3.351637478974968e-05,
      "loss": 2.8707,
      "step": 3332
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5995546579360962,
      "learning_rate": 3.349658652419115e-05,
      "loss": 2.839,
      "step": 3336
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3852463960647583,
      "learning_rate": 3.347679825863263e-05,
      "loss": 2.8356,
      "step": 3340
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.2991340160369873,
      "learning_rate": 3.3457009993074105e-05,
      "loss": 2.8408,
      "step": 3344
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4714194536209106,
      "learning_rate": 3.3437221727515585e-05,
      "loss": 2.898,
      "step": 3348
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.466454267501831,
      "learning_rate": 3.341743346195706e-05,
      "loss": 3.1129,
      "step": 3352
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3871067762374878,
      "learning_rate": 3.339764519639854e-05,
      "loss": 2.8544,
      "step": 3356
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4386447668075562,
      "learning_rate": 3.337785693084001e-05,
      "loss": 2.8696,
      "step": 3360
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4348955154418945,
      "learning_rate": 3.335806866528149e-05,
      "loss": 3.0176,
      "step": 3364
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.3820732831954956,
      "learning_rate": 3.333828039972296e-05,
      "loss": 2.8816,
      "step": 3368
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4179422855377197,
      "learning_rate": 3.331849213416444e-05,
      "loss": 2.8338,
      "step": 3372
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.338954210281372,
      "learning_rate": 3.3298703868605916e-05,
      "loss": 2.8356,
      "step": 3376
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.5031296014785767,
      "learning_rate": 3.3278915603047396e-05,
      "loss": 3.0582,
      "step": 3380
    },
    {
      "epoch": 0.33,
      "grad_norm": 1.4210484027862549,
      "learning_rate": 3.325912733748887e-05,
      "loss": 2.7175,
      "step": 3384
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.301172137260437,
      "learning_rate": 3.323933907193035e-05,
      "loss": 2.8864,
      "step": 3388
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3612853288650513,
      "learning_rate": 3.321955080637182e-05,
      "loss": 2.9919,
      "step": 3392
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.8065041303634644,
      "learning_rate": 3.31997625408133e-05,
      "loss": 2.9198,
      "step": 3396
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.6538633108139038,
      "learning_rate": 3.3179974275254774e-05,
      "loss": 2.9522,
      "step": 3400
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.440852403640747,
      "learning_rate": 3.3160186009696254e-05,
      "loss": 3.0007,
      "step": 3404
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3337926864624023,
      "learning_rate": 3.314039774413773e-05,
      "loss": 3.0386,
      "step": 3408
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.6055779457092285,
      "learning_rate": 3.3120609478579206e-05,
      "loss": 2.7901,
      "step": 3412
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7633087635040283,
      "learning_rate": 3.3100821213020686e-05,
      "loss": 2.9472,
      "step": 3416
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3165653944015503,
      "learning_rate": 3.308103294746215e-05,
      "loss": 2.8783,
      "step": 3420
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4195142984390259,
      "learning_rate": 3.306124468190363e-05,
      "loss": 2.985,
      "step": 3424
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3331257104873657,
      "learning_rate": 3.3041456416345105e-05,
      "loss": 2.7907,
      "step": 3428
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7462310791015625,
      "learning_rate": 3.3021668150786584e-05,
      "loss": 2.9509,
      "step": 3432
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.522599220275879,
      "learning_rate": 3.300187988522806e-05,
      "loss": 2.8467,
      "step": 3436
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.349449872970581,
      "learning_rate": 3.298209161966954e-05,
      "loss": 3.0331,
      "step": 3440
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.2950477600097656,
      "learning_rate": 3.296230335411101e-05,
      "loss": 2.8946,
      "step": 3444
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.456489086151123,
      "learning_rate": 3.294251508855249e-05,
      "loss": 2.9786,
      "step": 3448
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4597196578979492,
      "learning_rate": 3.292272682299396e-05,
      "loss": 2.8709,
      "step": 3452
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.3608200550079346,
      "learning_rate": 3.290293855743544e-05,
      "loss": 2.8589,
      "step": 3456
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4015833139419556,
      "learning_rate": 3.2883150291876915e-05,
      "loss": 2.9041,
      "step": 3460
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.5029760599136353,
      "learning_rate": 3.2863362026318395e-05,
      "loss": 2.9721,
      "step": 3464
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.54849374294281,
      "learning_rate": 3.2843573760759875e-05,
      "loss": 2.9531,
      "step": 3468
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4009016752243042,
      "learning_rate": 3.282378549520135e-05,
      "loss": 2.8118,
      "step": 3472
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4506298303604126,
      "learning_rate": 3.280399722964283e-05,
      "loss": 2.8969,
      "step": 3476
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.4417248964309692,
      "learning_rate": 3.27842089640843e-05,
      "loss": 2.8477,
      "step": 3480
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.5801173448562622,
      "learning_rate": 3.276442069852578e-05,
      "loss": 3.0832,
      "step": 3484
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4388275146484375,
      "learning_rate": 3.274463243296725e-05,
      "loss": 2.938,
      "step": 3488
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5307226181030273,
      "learning_rate": 3.272484416740873e-05,
      "loss": 2.934,
      "step": 3492
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3734909296035767,
      "learning_rate": 3.2705055901850206e-05,
      "loss": 2.8592,
      "step": 3496
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4125416278839111,
      "learning_rate": 3.2685267636291685e-05,
      "loss": 2.8323,
      "step": 3500
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3895879983901978,
      "learning_rate": 3.266547937073315e-05,
      "loss": 3.025,
      "step": 3504
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3139837980270386,
      "learning_rate": 3.264569110517463e-05,
      "loss": 2.9754,
      "step": 3508
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4398940801620483,
      "learning_rate": 3.2625902839616104e-05,
      "loss": 2.7895,
      "step": 3512
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.474226951599121,
      "learning_rate": 3.2606114574057584e-05,
      "loss": 2.7609,
      "step": 3516
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.440250039100647,
      "learning_rate": 3.258632630849906e-05,
      "loss": 2.8932,
      "step": 3520
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4383162260055542,
      "learning_rate": 3.2566538042940536e-05,
      "loss": 2.8672,
      "step": 3524
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4906786680221558,
      "learning_rate": 3.2546749777382016e-05,
      "loss": 2.8692,
      "step": 3528
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.699826717376709,
      "learning_rate": 3.252696151182349e-05,
      "loss": 2.7653,
      "step": 3532
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4905091524124146,
      "learning_rate": 3.250717324626497e-05,
      "loss": 2.9857,
      "step": 3536
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.6657990217208862,
      "learning_rate": 3.248738498070644e-05,
      "loss": 3.0826,
      "step": 3540
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.464480996131897,
      "learning_rate": 3.246759671514792e-05,
      "loss": 2.833,
      "step": 3544
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5183041095733643,
      "learning_rate": 3.2447808449589394e-05,
      "loss": 2.8327,
      "step": 3548
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3862593173980713,
      "learning_rate": 3.2428020184030874e-05,
      "loss": 2.7403,
      "step": 3552
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.560221552848816,
      "learning_rate": 3.240823191847235e-05,
      "loss": 3.0822,
      "step": 3556
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.4953581094741821,
      "learning_rate": 3.238844365291383e-05,
      "loss": 2.873,
      "step": 3560
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3668919801712036,
      "learning_rate": 3.23686553873553e-05,
      "loss": 2.8884,
      "step": 3564
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5106139183044434,
      "learning_rate": 3.234886712179678e-05,
      "loss": 2.9894,
      "step": 3568
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5733063220977783,
      "learning_rate": 3.232907885623825e-05,
      "loss": 2.6832,
      "step": 3572
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.3589937686920166,
      "learning_rate": 3.230929059067973e-05,
      "loss": 2.8609,
      "step": 3576
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.5051617622375488,
      "learning_rate": 3.2289502325121205e-05,
      "loss": 3.0103,
      "step": 3580
    },
    {
      "epoch": 0.35,
      "grad_norm": 1.423579454421997,
      "learning_rate": 3.2269714059562685e-05,
      "loss": 2.8885,
      "step": 3584
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.4297691583633423,
      "learning_rate": 3.224992579400416e-05,
      "loss": 2.8898,
      "step": 3588
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5489140748977661,
      "learning_rate": 3.223013752844563e-05,
      "loss": 2.836,
      "step": 3592
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.4612491130828857,
      "learning_rate": 3.221034926288711e-05,
      "loss": 2.7572,
      "step": 3596
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.412541389465332,
      "learning_rate": 3.219056099732858e-05,
      "loss": 2.9328,
      "step": 3600
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5174839496612549,
      "learning_rate": 3.217077273177006e-05,
      "loss": 2.7265,
      "step": 3604
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.732851505279541,
      "learning_rate": 3.2150984466211536e-05,
      "loss": 2.8555,
      "step": 3608
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.3930217027664185,
      "learning_rate": 3.2131196200653016e-05,
      "loss": 2.8495,
      "step": 3612
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.9050196409225464,
      "learning_rate": 3.211140793509449e-05,
      "loss": 2.7914,
      "step": 3616
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.423581600189209,
      "learning_rate": 3.209161966953597e-05,
      "loss": 2.8277,
      "step": 3620
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.490317702293396,
      "learning_rate": 3.207183140397744e-05,
      "loss": 2.8735,
      "step": 3624
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.434913992881775,
      "learning_rate": 3.205204313841892e-05,
      "loss": 2.8089,
      "step": 3628
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.666888952255249,
      "learning_rate": 3.2032254872860394e-05,
      "loss": 2.791,
      "step": 3632
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.7636072635650635,
      "learning_rate": 3.2012466607301873e-05,
      "loss": 2.8418,
      "step": 3636
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.707155704498291,
      "learning_rate": 3.1992678341743346e-05,
      "loss": 2.7568,
      "step": 3640
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5901198387145996,
      "learning_rate": 3.1972890076184826e-05,
      "loss": 2.7845,
      "step": 3644
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.4522852897644043,
      "learning_rate": 3.19531018106263e-05,
      "loss": 2.8471,
      "step": 3648
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.348705768585205,
      "learning_rate": 3.193331354506778e-05,
      "loss": 3.13,
      "step": 3652
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.645375370979309,
      "learning_rate": 3.191352527950925e-05,
      "loss": 2.8477,
      "step": 3656
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5673881769180298,
      "learning_rate": 3.189373701395073e-05,
      "loss": 2.6724,
      "step": 3660
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.3590004444122314,
      "learning_rate": 3.1873948748392204e-05,
      "loss": 2.8683,
      "step": 3664
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.4108936786651611,
      "learning_rate": 3.1854160482833684e-05,
      "loss": 2.8131,
      "step": 3668
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.6254466772079468,
      "learning_rate": 3.183437221727516e-05,
      "loss": 2.9229,
      "step": 3672
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.460128903388977,
      "learning_rate": 3.181458395171663e-05,
      "loss": 2.7728,
      "step": 3676
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.473266363143921,
      "learning_rate": 3.179479568615811e-05,
      "loss": 2.8825,
      "step": 3680
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.422609567642212,
      "learning_rate": 3.177500742059958e-05,
      "loss": 2.8954,
      "step": 3684
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5324984788894653,
      "learning_rate": 3.175521915504106e-05,
      "loss": 2.8591,
      "step": 3688
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.472915530204773,
      "learning_rate": 3.1735430889482535e-05,
      "loss": 2.7172,
      "step": 3692
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3716154098510742,
      "learning_rate": 3.1715642623924015e-05,
      "loss": 2.7983,
      "step": 3696
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.2956911325454712,
      "learning_rate": 3.169585435836549e-05,
      "loss": 2.8473,
      "step": 3700
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3868067264556885,
      "learning_rate": 3.167606609280697e-05,
      "loss": 2.7424,
      "step": 3704
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.554824709892273,
      "learning_rate": 3.165627782724844e-05,
      "loss": 2.9807,
      "step": 3708
    },
    {
      "epoch": 0.37,
      "grad_norm": 2.6641464233398438,
      "learning_rate": 3.163648956168992e-05,
      "loss": 3.1424,
      "step": 3712
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.5256997346878052,
      "learning_rate": 3.161670129613139e-05,
      "loss": 2.7976,
      "step": 3716
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4320893287658691,
      "learning_rate": 3.159691303057287e-05,
      "loss": 2.7902,
      "step": 3720
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3281612396240234,
      "learning_rate": 3.1577124765014346e-05,
      "loss": 2.673,
      "step": 3724
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.5591663122177124,
      "learning_rate": 3.1557336499455825e-05,
      "loss": 2.6409,
      "step": 3728
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4519292116165161,
      "learning_rate": 3.15375482338973e-05,
      "loss": 2.9247,
      "step": 3732
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.34962797164917,
      "learning_rate": 3.151775996833878e-05,
      "loss": 2.8384,
      "step": 3736
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.7062203884124756,
      "learning_rate": 3.149797170278026e-05,
      "loss": 2.7923,
      "step": 3740
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.5387376546859741,
      "learning_rate": 3.147818343722173e-05,
      "loss": 3.0167,
      "step": 3744
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3127942085266113,
      "learning_rate": 3.145839517166321e-05,
      "loss": 2.9422,
      "step": 3748
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4698482751846313,
      "learning_rate": 3.143860690610468e-05,
      "loss": 2.9139,
      "step": 3752
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.472330927848816,
      "learning_rate": 3.1418818640546156e-05,
      "loss": 2.8468,
      "step": 3756
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.507949948310852,
      "learning_rate": 3.139903037498763e-05,
      "loss": 2.8117,
      "step": 3760
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.3722580671310425,
      "learning_rate": 3.137924210942911e-05,
      "loss": 3.0344,
      "step": 3764
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.6335688829421997,
      "learning_rate": 3.135945384387058e-05,
      "loss": 2.9591,
      "step": 3768
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.608832836151123,
      "learning_rate": 3.133966557831206e-05,
      "loss": 2.9696,
      "step": 3772
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4067091941833496,
      "learning_rate": 3.1319877312753535e-05,
      "loss": 2.797,
      "step": 3776
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.439083218574524,
      "learning_rate": 3.1300089047195014e-05,
      "loss": 2.9332,
      "step": 3780
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.4436142444610596,
      "learning_rate": 3.128030078163649e-05,
      "loss": 2.9546,
      "step": 3784
    },
    {
      "epoch": 0.37,
      "grad_norm": 1.5676676034927368,
      "learning_rate": 3.126051251607797e-05,
      "loss": 2.9537,
      "step": 3788
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.564913272857666,
      "learning_rate": 3.124072425051944e-05,
      "loss": 3.0004,
      "step": 3792
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5197542905807495,
      "learning_rate": 3.122093598496092e-05,
      "loss": 2.9121,
      "step": 3796
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.467848539352417,
      "learning_rate": 3.12011477194024e-05,
      "loss": 2.8398,
      "step": 3800
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.415881872177124,
      "learning_rate": 3.118135945384387e-05,
      "loss": 2.8903,
      "step": 3804
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3919771909713745,
      "learning_rate": 3.116157118828535e-05,
      "loss": 2.862,
      "step": 3808
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5213255882263184,
      "learning_rate": 3.1141782922726825e-05,
      "loss": 2.876,
      "step": 3812
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4630500078201294,
      "learning_rate": 3.1121994657168305e-05,
      "loss": 2.9567,
      "step": 3816
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4120430946350098,
      "learning_rate": 3.110220639160978e-05,
      "loss": 2.7293,
      "step": 3820
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.336676001548767,
      "learning_rate": 3.108241812605126e-05,
      "loss": 2.8274,
      "step": 3824
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5005687475204468,
      "learning_rate": 3.106262986049273e-05,
      "loss": 2.9078,
      "step": 3828
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5053759813308716,
      "learning_rate": 3.104284159493421e-05,
      "loss": 2.8964,
      "step": 3832
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.764050841331482,
      "learning_rate": 3.102305332937568e-05,
      "loss": 2.7096,
      "step": 3836
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.731428623199463,
      "learning_rate": 3.100326506381716e-05,
      "loss": 2.8654,
      "step": 3840
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5615731477737427,
      "learning_rate": 3.098347679825863e-05,
      "loss": 2.7207,
      "step": 3844
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3528928756713867,
      "learning_rate": 3.096368853270011e-05,
      "loss": 2.7666,
      "step": 3848
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4369572401046753,
      "learning_rate": 3.094390026714158e-05,
      "loss": 3.0385,
      "step": 3852
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.7137233018875122,
      "learning_rate": 3.092411200158306e-05,
      "loss": 2.7059,
      "step": 3856
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4573568105697632,
      "learning_rate": 3.090432373602454e-05,
      "loss": 2.6826,
      "step": 3860
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.6437382698059082,
      "learning_rate": 3.0884535470466014e-05,
      "loss": 2.8949,
      "step": 3864
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5055683851242065,
      "learning_rate": 3.086474720490749e-05,
      "loss": 2.7713,
      "step": 3868
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.3726445436477661,
      "learning_rate": 3.0844958939348966e-05,
      "loss": 2.7203,
      "step": 3872
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.4541021585464478,
      "learning_rate": 3.0825170673790446e-05,
      "loss": 2.6903,
      "step": 3876
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.6873700618743896,
      "learning_rate": 3.080538240823192e-05,
      "loss": 2.6891,
      "step": 3880
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5541645288467407,
      "learning_rate": 3.07855941426734e-05,
      "loss": 2.7473,
      "step": 3884
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.5389822721481323,
      "learning_rate": 3.076580587711487e-05,
      "loss": 2.7178,
      "step": 3888
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3678780794143677,
      "learning_rate": 3.074601761155635e-05,
      "loss": 2.914,
      "step": 3892
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.504895806312561,
      "learning_rate": 3.0726229345997824e-05,
      "loss": 2.635,
      "step": 3896
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5606142282485962,
      "learning_rate": 3.0706441080439304e-05,
      "loss": 2.8645,
      "step": 3900
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5420433282852173,
      "learning_rate": 3.068665281488078e-05,
      "loss": 3.0066,
      "step": 3904
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.494759202003479,
      "learning_rate": 3.0666864549322257e-05,
      "loss": 2.8071,
      "step": 3908
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.386305570602417,
      "learning_rate": 3.064707628376373e-05,
      "loss": 2.9353,
      "step": 3912
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5420827865600586,
      "learning_rate": 3.062728801820521e-05,
      "loss": 2.9036,
      "step": 3916
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4949458837509155,
      "learning_rate": 3.060749975264668e-05,
      "loss": 2.5632,
      "step": 3920
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.6547415256500244,
      "learning_rate": 3.058771148708816e-05,
      "loss": 2.8553,
      "step": 3924
    },
    {
      "epoch": 0.39,
      "grad_norm": 2.1204848289489746,
      "learning_rate": 3.0567923221529635e-05,
      "loss": 2.9203,
      "step": 3928
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4536625146865845,
      "learning_rate": 3.054813495597111e-05,
      "loss": 2.7916,
      "step": 3932
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4841111898422241,
      "learning_rate": 3.052834669041259e-05,
      "loss": 2.7978,
      "step": 3936
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5112613439559937,
      "learning_rate": 3.050855842485406e-05,
      "loss": 2.7701,
      "step": 3940
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4966477155685425,
      "learning_rate": 3.0488770159295537e-05,
      "loss": 2.6693,
      "step": 3944
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.3681539297103882,
      "learning_rate": 3.0468981893737013e-05,
      "loss": 2.7811,
      "step": 3948
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4147274494171143,
      "learning_rate": 3.044919362817849e-05,
      "loss": 2.7958,
      "step": 3952
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.451531171798706,
      "learning_rate": 3.0429405362619966e-05,
      "loss": 2.8591,
      "step": 3956
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4994832277297974,
      "learning_rate": 3.0409617097061442e-05,
      "loss": 2.7079,
      "step": 3960
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4195365905761719,
      "learning_rate": 3.038982883150292e-05,
      "loss": 2.8642,
      "step": 3964
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4782614707946777,
      "learning_rate": 3.0370040565944398e-05,
      "loss": 2.8319,
      "step": 3968
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.6576443910598755,
      "learning_rate": 3.0350252300385874e-05,
      "loss": 2.6855,
      "step": 3972
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.8916692733764648,
      "learning_rate": 3.033046403482735e-05,
      "loss": 2.8664,
      "step": 3976
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.4349229335784912,
      "learning_rate": 3.0310675769268827e-05,
      "loss": 2.7252,
      "step": 3980
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.5039000511169434,
      "learning_rate": 3.0290887503710303e-05,
      "loss": 2.8368,
      "step": 3984
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.552319884300232,
      "learning_rate": 3.027109923815178e-05,
      "loss": 2.7585,
      "step": 3988
    },
    {
      "epoch": 0.39,
      "grad_norm": 1.618980884552002,
      "learning_rate": 3.0251310972593256e-05,
      "loss": 2.8395,
      "step": 3992
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.370418667793274,
      "learning_rate": 3.0231522707034732e-05,
      "loss": 2.8074,
      "step": 3996
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.3681377172470093,
      "learning_rate": 3.021173444147621e-05,
      "loss": 2.8299,
      "step": 4000
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4201312065124512,
      "learning_rate": 3.0191946175917685e-05,
      "loss": 2.8259,
      "step": 4004
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.8192863464355469,
      "learning_rate": 3.017215791035916e-05,
      "loss": 2.9578,
      "step": 4008
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5189778804779053,
      "learning_rate": 3.015236964480063e-05,
      "loss": 2.7655,
      "step": 4012
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.401796579360962,
      "learning_rate": 3.0132581379242107e-05,
      "loss": 2.7549,
      "step": 4016
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5482715368270874,
      "learning_rate": 3.0112793113683583e-05,
      "loss": 2.5986,
      "step": 4020
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.446227788925171,
      "learning_rate": 3.0093004848125063e-05,
      "loss": 3.0037,
      "step": 4024
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5745714902877808,
      "learning_rate": 3.007321658256654e-05,
      "loss": 2.9434,
      "step": 4028
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4580100774765015,
      "learning_rate": 3.0053428317008016e-05,
      "loss": 2.7611,
      "step": 4032
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6320377588272095,
      "learning_rate": 3.0033640051449492e-05,
      "loss": 2.7802,
      "step": 4036
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5840559005737305,
      "learning_rate": 3.001385178589097e-05,
      "loss": 2.8459,
      "step": 4040
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4639407396316528,
      "learning_rate": 2.9994063520332445e-05,
      "loss": 2.8289,
      "step": 4044
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4280468225479126,
      "learning_rate": 2.997427525477392e-05,
      "loss": 2.6556,
      "step": 4048
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4254354238510132,
      "learning_rate": 2.9954486989215397e-05,
      "loss": 2.8135,
      "step": 4052
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5892139673233032,
      "learning_rate": 2.9934698723656874e-05,
      "loss": 2.8655,
      "step": 4056
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.422926425933838,
      "learning_rate": 2.991491045809835e-05,
      "loss": 2.9126,
      "step": 4060
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5437527894973755,
      "learning_rate": 2.9895122192539826e-05,
      "loss": 2.7983,
      "step": 4064
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6389095783233643,
      "learning_rate": 2.9875333926981303e-05,
      "loss": 2.8345,
      "step": 4068
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.9963791370391846,
      "learning_rate": 2.985554566142278e-05,
      "loss": 2.8074,
      "step": 4072
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4846253395080566,
      "learning_rate": 2.9835757395864255e-05,
      "loss": 2.8744,
      "step": 4076
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.5970510244369507,
      "learning_rate": 2.981596913030573e-05,
      "loss": 2.8114,
      "step": 4080
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.6389985084533691,
      "learning_rate": 2.9796180864747208e-05,
      "loss": 2.6592,
      "step": 4084
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4449435472488403,
      "learning_rate": 2.9776392599188684e-05,
      "loss": 2.7522,
      "step": 4088
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.3930374383926392,
      "learning_rate": 2.975660433363016e-05,
      "loss": 2.6228,
      "step": 4092
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4793918132781982,
      "learning_rate": 2.9736816068071633e-05,
      "loss": 2.7966,
      "step": 4096
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.457022786140442,
      "learning_rate": 2.971702780251311e-05,
      "loss": 2.8147,
      "step": 4100
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3806568384170532,
      "learning_rate": 2.9697239536954586e-05,
      "loss": 2.7444,
      "step": 4104
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.445950984954834,
      "learning_rate": 2.9677451271396062e-05,
      "loss": 2.6564,
      "step": 4108
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4374557733535767,
      "learning_rate": 2.965766300583754e-05,
      "loss": 2.8971,
      "step": 4112
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4371333122253418,
      "learning_rate": 2.9637874740279015e-05,
      "loss": 2.9071,
      "step": 4116
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5124506950378418,
      "learning_rate": 2.961808647472049e-05,
      "loss": 2.9047,
      "step": 4120
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5280250310897827,
      "learning_rate": 2.9598298209161968e-05,
      "loss": 2.8296,
      "step": 4124
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3567237854003906,
      "learning_rate": 2.9578509943603444e-05,
      "loss": 2.757,
      "step": 4128
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5457067489624023,
      "learning_rate": 2.955872167804492e-05,
      "loss": 2.8907,
      "step": 4132
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4458136558532715,
      "learning_rate": 2.9538933412486397e-05,
      "loss": 2.7835,
      "step": 4136
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4487221240997314,
      "learning_rate": 2.9519145146927873e-05,
      "loss": 2.7497,
      "step": 4140
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5274755954742432,
      "learning_rate": 2.949935688136935e-05,
      "loss": 2.7221,
      "step": 4144
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5352815389633179,
      "learning_rate": 2.9479568615810826e-05,
      "loss": 2.8334,
      "step": 4148
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5403484106063843,
      "learning_rate": 2.9459780350252302e-05,
      "loss": 2.7065,
      "step": 4152
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4784417152404785,
      "learning_rate": 2.9439992084693778e-05,
      "loss": 2.6853,
      "step": 4156
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4991648197174072,
      "learning_rate": 2.9420203819135255e-05,
      "loss": 2.6927,
      "step": 4160
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5489524602890015,
      "learning_rate": 2.940041555357673e-05,
      "loss": 2.8819,
      "step": 4164
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5099831819534302,
      "learning_rate": 2.9380627288018207e-05,
      "loss": 2.7281,
      "step": 4168
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.400437355041504,
      "learning_rate": 2.9360839022459684e-05,
      "loss": 2.706,
      "step": 4172
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4548041820526123,
      "learning_rate": 2.9341050756901163e-05,
      "loss": 2.7494,
      "step": 4176
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.3302807807922363,
      "learning_rate": 2.9321262491342633e-05,
      "loss": 2.9415,
      "step": 4180
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.5597517490386963,
      "learning_rate": 2.930147422578411e-05,
      "loss": 2.8837,
      "step": 4184
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.456819772720337,
      "learning_rate": 2.9281685960225585e-05,
      "loss": 2.7019,
      "step": 4188
    },
    {
      "epoch": 0.41,
      "grad_norm": 1.4813134670257568,
      "learning_rate": 2.9261897694667062e-05,
      "loss": 2.7553,
      "step": 4192
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4024277925491333,
      "learning_rate": 2.9242109429108538e-05,
      "loss": 2.7829,
      "step": 4196
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.5183507204055786,
      "learning_rate": 2.9222321163550014e-05,
      "loss": 2.6526,
      "step": 4200
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.73296320438385,
      "learning_rate": 2.920253289799149e-05,
      "loss": 2.8376,
      "step": 4204
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4100421667099,
      "learning_rate": 2.9182744632432967e-05,
      "loss": 2.8162,
      "step": 4208
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.3729844093322754,
      "learning_rate": 2.9162956366874443e-05,
      "loss": 2.6933,
      "step": 4212
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.381945252418518,
      "learning_rate": 2.914316810131592e-05,
      "loss": 2.6763,
      "step": 4216
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4343644380569458,
      "learning_rate": 2.9123379835757396e-05,
      "loss": 2.6836,
      "step": 4220
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4591138362884521,
      "learning_rate": 2.9103591570198872e-05,
      "loss": 2.6811,
      "step": 4224
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4086369276046753,
      "learning_rate": 2.908380330464035e-05,
      "loss": 2.7654,
      "step": 4228
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.3589953184127808,
      "learning_rate": 2.9064015039081825e-05,
      "loss": 2.9029,
      "step": 4232
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4481515884399414,
      "learning_rate": 2.9044226773523305e-05,
      "loss": 2.8348,
      "step": 4236
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4589226245880127,
      "learning_rate": 2.902443850796478e-05,
      "loss": 2.7427,
      "step": 4240
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.3930855989456177,
      "learning_rate": 2.9004650242406257e-05,
      "loss": 2.8394,
      "step": 4244
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.670841097831726,
      "learning_rate": 2.8984861976847734e-05,
      "loss": 2.8234,
      "step": 4248
    },
    {
      "epoch": 0.42,
      "grad_norm": 2.0551598072052,
      "learning_rate": 2.896507371128921e-05,
      "loss": 2.8201,
      "step": 4252
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6182607412338257,
      "learning_rate": 2.8945285445730686e-05,
      "loss": 2.7768,
      "step": 4256
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6682448387145996,
      "learning_rate": 2.8925497180172163e-05,
      "loss": 2.7996,
      "step": 4260
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.5060317516326904,
      "learning_rate": 2.8905708914613632e-05,
      "loss": 2.9885,
      "step": 4264
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.6912871599197388,
      "learning_rate": 2.888592064905511e-05,
      "loss": 2.7473,
      "step": 4268
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4391567707061768,
      "learning_rate": 2.8866132383496585e-05,
      "loss": 2.7497,
      "step": 4272
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4954835176467896,
      "learning_rate": 2.884634411793806e-05,
      "loss": 2.724,
      "step": 4276
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4786354303359985,
      "learning_rate": 2.8826555852379537e-05,
      "loss": 2.9594,
      "step": 4280
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.5478767156600952,
      "learning_rate": 2.8806767586821014e-05,
      "loss": 2.7347,
      "step": 4284
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.488110899925232,
      "learning_rate": 2.878697932126249e-05,
      "loss": 2.736,
      "step": 4288
    },
    {
      "epoch": 0.42,
      "grad_norm": 1.4168086051940918,
      "learning_rate": 2.8767191055703966e-05,
      "loss": 2.6479,
      "step": 4292
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.413324236869812,
      "learning_rate": 2.8747402790145446e-05,
      "loss": 2.7392,
      "step": 4296
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.40789794921875,
      "learning_rate": 2.8727614524586922e-05,
      "loss": 2.6474,
      "step": 4300
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.3412038087844849,
      "learning_rate": 2.87078262590284e-05,
      "loss": 2.6181,
      "step": 4304
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4228928089141846,
      "learning_rate": 2.8688037993469875e-05,
      "loss": 2.8285,
      "step": 4308
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.498350977897644,
      "learning_rate": 2.866824972791135e-05,
      "loss": 2.7477,
      "step": 4312
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5861128568649292,
      "learning_rate": 2.8648461462352828e-05,
      "loss": 2.7869,
      "step": 4316
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.544293761253357,
      "learning_rate": 2.8628673196794304e-05,
      "loss": 2.711,
      "step": 4320
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.576124906539917,
      "learning_rate": 2.860888493123578e-05,
      "loss": 2.8703,
      "step": 4324
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5502452850341797,
      "learning_rate": 2.8589096665677257e-05,
      "loss": 2.7711,
      "step": 4328
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.3446522951126099,
      "learning_rate": 2.8569308400118733e-05,
      "loss": 2.8535,
      "step": 4332
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4643163681030273,
      "learning_rate": 2.854952013456021e-05,
      "loss": 2.65,
      "step": 4336
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.6467000246047974,
      "learning_rate": 2.8529731869001686e-05,
      "loss": 2.8216,
      "step": 4340
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.387041687965393,
      "learning_rate": 2.8509943603443162e-05,
      "loss": 2.7273,
      "step": 4344
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4695614576339722,
      "learning_rate": 2.849015533788463e-05,
      "loss": 2.6167,
      "step": 4348
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.6175564527511597,
      "learning_rate": 2.847036707232611e-05,
      "loss": 2.761,
      "step": 4352
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5921447277069092,
      "learning_rate": 2.8450578806767588e-05,
      "loss": 2.7493,
      "step": 4356
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5560427904129028,
      "learning_rate": 2.8430790541209064e-05,
      "loss": 2.8581,
      "step": 4360
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5728269815444946,
      "learning_rate": 2.841100227565054e-05,
      "loss": 2.6552,
      "step": 4364
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.6176984310150146,
      "learning_rate": 2.8391214010092017e-05,
      "loss": 2.698,
      "step": 4368
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4440550804138184,
      "learning_rate": 2.8371425744533493e-05,
      "loss": 2.8222,
      "step": 4372
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.721948266029358,
      "learning_rate": 2.835163747897497e-05,
      "loss": 2.6922,
      "step": 4376
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.679369330406189,
      "learning_rate": 2.8331849213416445e-05,
      "loss": 2.8892,
      "step": 4380
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4985451698303223,
      "learning_rate": 2.8312060947857922e-05,
      "loss": 2.7287,
      "step": 4384
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4109586477279663,
      "learning_rate": 2.8292272682299398e-05,
      "loss": 2.7606,
      "step": 4388
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.5079187154769897,
      "learning_rate": 2.8272484416740874e-05,
      "loss": 2.6592,
      "step": 4392
    },
    {
      "epoch": 0.43,
      "grad_norm": 1.4183659553527832,
      "learning_rate": 2.825269615118235e-05,
      "loss": 2.7749,
      "step": 4396
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.3497616052627563,
      "learning_rate": 2.8232907885623827e-05,
      "loss": 2.8151,
      "step": 4400
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5650898218154907,
      "learning_rate": 2.8213119620065303e-05,
      "loss": 2.9046,
      "step": 4404
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5899450778961182,
      "learning_rate": 2.819333135450678e-05,
      "loss": 2.6898,
      "step": 4408
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4732282161712646,
      "learning_rate": 2.8173543088948256e-05,
      "loss": 2.7099,
      "step": 4412
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4923087358474731,
      "learning_rate": 2.8153754823389732e-05,
      "loss": 2.7293,
      "step": 4416
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.509701132774353,
      "learning_rate": 2.813396655783121e-05,
      "loss": 3.0249,
      "step": 4420
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.492567777633667,
      "learning_rate": 2.8114178292272685e-05,
      "loss": 2.779,
      "step": 4424
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.551950216293335,
      "learning_rate": 2.809439002671416e-05,
      "loss": 2.8592,
      "step": 4428
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5013272762298584,
      "learning_rate": 2.8074601761155634e-05,
      "loss": 2.6557,
      "step": 4432
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.757328987121582,
      "learning_rate": 2.805481349559711e-05,
      "loss": 2.917,
      "step": 4436
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4949613809585571,
      "learning_rate": 2.8035025230038587e-05,
      "loss": 2.8814,
      "step": 4440
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5500638484954834,
      "learning_rate": 2.8015236964480063e-05,
      "loss": 2.7875,
      "step": 4444
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.724369764328003,
      "learning_rate": 2.799544869892154e-05,
      "loss": 2.8852,
      "step": 4448
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6037100553512573,
      "learning_rate": 2.7975660433363016e-05,
      "loss": 2.8543,
      "step": 4452
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5066895484924316,
      "learning_rate": 2.7955872167804492e-05,
      "loss": 2.743,
      "step": 4456
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.412857174873352,
      "learning_rate": 2.793608390224597e-05,
      "loss": 2.6795,
      "step": 4460
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4990521669387817,
      "learning_rate": 2.7916295636687445e-05,
      "loss": 2.7017,
      "step": 4464
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7534558773040771,
      "learning_rate": 2.789650737112892e-05,
      "loss": 2.6926,
      "step": 4468
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6677272319793701,
      "learning_rate": 2.7876719105570398e-05,
      "loss": 2.7715,
      "step": 4472
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.4068310260772705,
      "learning_rate": 2.7856930840011874e-05,
      "loss": 2.7641,
      "step": 4476
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.6524847745895386,
      "learning_rate": 2.783714257445335e-05,
      "loss": 2.8915,
      "step": 4480
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.362470030784607,
      "learning_rate": 2.7817354308894826e-05,
      "loss": 2.7326,
      "step": 4484
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.7215932607650757,
      "learning_rate": 2.7797566043336303e-05,
      "loss": 2.9945,
      "step": 4488
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.5028218030929565,
      "learning_rate": 2.777777777777778e-05,
      "loss": 2.7282,
      "step": 4492
    },
    {
      "epoch": 0.44,
      "grad_norm": 1.628304123878479,
      "learning_rate": 2.7757989512219255e-05,
      "loss": 2.7148,
      "step": 4496
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5549771785736084,
      "learning_rate": 2.7738201246660732e-05,
      "loss": 2.7639,
      "step": 4500
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7809425592422485,
      "learning_rate": 2.7718412981102208e-05,
      "loss": 2.8016,
      "step": 4504
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5097914934158325,
      "learning_rate": 2.7698624715543688e-05,
      "loss": 2.7919,
      "step": 4508
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.6162583827972412,
      "learning_rate": 2.7678836449985164e-05,
      "loss": 2.7603,
      "step": 4512
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.3952536582946777,
      "learning_rate": 2.7659048184426634e-05,
      "loss": 2.8791,
      "step": 4516
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4246437549591064,
      "learning_rate": 2.763925991886811e-05,
      "loss": 2.7032,
      "step": 4520
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5354806184768677,
      "learning_rate": 2.7619471653309586e-05,
      "loss": 2.6509,
      "step": 4524
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.414692997932434,
      "learning_rate": 2.7599683387751063e-05,
      "loss": 2.8077,
      "step": 4528
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4562816619873047,
      "learning_rate": 2.757989512219254e-05,
      "loss": 2.6559,
      "step": 4532
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5134601593017578,
      "learning_rate": 2.7560106856634015e-05,
      "loss": 2.9881,
      "step": 4536
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4460077285766602,
      "learning_rate": 2.754031859107549e-05,
      "loss": 2.9132,
      "step": 4540
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.2956750392913818,
      "learning_rate": 2.7520530325516968e-05,
      "loss": 2.6041,
      "step": 4544
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.3744826316833496,
      "learning_rate": 2.7500742059958444e-05,
      "loss": 2.7083,
      "step": 4548
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.8652702569961548,
      "learning_rate": 2.748095379439992e-05,
      "loss": 2.7688,
      "step": 4552
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4958977699279785,
      "learning_rate": 2.7461165528841397e-05,
      "loss": 2.6397,
      "step": 4556
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.7164748907089233,
      "learning_rate": 2.7441377263282873e-05,
      "loss": 2.8242,
      "step": 4560
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.3987586498260498,
      "learning_rate": 2.742158899772435e-05,
      "loss": 2.7012,
      "step": 4564
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.44004487991333,
      "learning_rate": 2.740180073216583e-05,
      "loss": 2.6568,
      "step": 4568
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5092453956604004,
      "learning_rate": 2.7382012466607306e-05,
      "loss": 2.8711,
      "step": 4572
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4860295057296753,
      "learning_rate": 2.7362224201048782e-05,
      "loss": 2.6946,
      "step": 4576
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.4830384254455566,
      "learning_rate": 2.7342435935490258e-05,
      "loss": 2.6399,
      "step": 4580
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5015872716903687,
      "learning_rate": 2.7322647669931734e-05,
      "loss": 2.794,
      "step": 4584
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.514248013496399,
      "learning_rate": 2.730285940437321e-05,
      "loss": 2.8176,
      "step": 4588
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.651678204536438,
      "learning_rate": 2.7283071138814687e-05,
      "loss": 2.8214,
      "step": 4592
    },
    {
      "epoch": 0.45,
      "grad_norm": 1.5051383972167969,
      "learning_rate": 2.7263282873256163e-05,
      "loss": 2.7998,
      "step": 4596
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4921104907989502,
      "learning_rate": 2.7243494607697633e-05,
      "loss": 2.7712,
      "step": 4600
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4588526487350464,
      "learning_rate": 2.722370634213911e-05,
      "loss": 2.718,
      "step": 4604
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5112457275390625,
      "learning_rate": 2.7203918076580586e-05,
      "loss": 2.8277,
      "step": 4608
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5032070875167847,
      "learning_rate": 2.7184129811022062e-05,
      "loss": 2.6329,
      "step": 4612
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.3996548652648926,
      "learning_rate": 2.7164341545463538e-05,
      "loss": 2.5617,
      "step": 4616
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5661203861236572,
      "learning_rate": 2.7144553279905015e-05,
      "loss": 2.6544,
      "step": 4620
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4876741170883179,
      "learning_rate": 2.7124765014346494e-05,
      "loss": 2.7655,
      "step": 4624
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.550423264503479,
      "learning_rate": 2.710497674878797e-05,
      "loss": 2.7317,
      "step": 4628
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.477411150932312,
      "learning_rate": 2.7085188483229447e-05,
      "loss": 2.6941,
      "step": 4632
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4628989696502686,
      "learning_rate": 2.7065400217670923e-05,
      "loss": 2.791,
      "step": 4636
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4556288719177246,
      "learning_rate": 2.70456119521124e-05,
      "loss": 2.7437,
      "step": 4640
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.7020779848098755,
      "learning_rate": 2.7025823686553876e-05,
      "loss": 2.5774,
      "step": 4644
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.8932197093963623,
      "learning_rate": 2.7006035420995352e-05,
      "loss": 2.7657,
      "step": 4648
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.6236851215362549,
      "learning_rate": 2.698624715543683e-05,
      "loss": 2.7994,
      "step": 4652
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.6066538095474243,
      "learning_rate": 2.6966458889878305e-05,
      "loss": 2.8015,
      "step": 4656
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4361777305603027,
      "learning_rate": 2.694667062431978e-05,
      "loss": 2.7335,
      "step": 4660
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.6797633171081543,
      "learning_rate": 2.6926882358761258e-05,
      "loss": 2.7286,
      "step": 4664
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4555542469024658,
      "learning_rate": 2.6907094093202734e-05,
      "loss": 2.6814,
      "step": 4668
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4565298557281494,
      "learning_rate": 2.688730582764421e-05,
      "loss": 2.7205,
      "step": 4672
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4873201847076416,
      "learning_rate": 2.6867517562085687e-05,
      "loss": 2.866,
      "step": 4676
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.62325918674469,
      "learning_rate": 2.6847729296527163e-05,
      "loss": 2.7617,
      "step": 4680
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4098718166351318,
      "learning_rate": 2.682794103096864e-05,
      "loss": 2.6488,
      "step": 4684
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5865626335144043,
      "learning_rate": 2.6808152765410112e-05,
      "loss": 2.5687,
      "step": 4688
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.4347946643829346,
      "learning_rate": 2.678836449985159e-05,
      "loss": 2.7482,
      "step": 4692
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.5309094190597534,
      "learning_rate": 2.6768576234293065e-05,
      "loss": 2.8154,
      "step": 4696
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4870914220809937,
      "learning_rate": 2.674878796873454e-05,
      "loss": 2.7054,
      "step": 4700
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.362409234046936,
      "learning_rate": 2.6728999703176017e-05,
      "loss": 2.6008,
      "step": 4704
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.7659647464752197,
      "learning_rate": 2.6709211437617494e-05,
      "loss": 2.7721,
      "step": 4708
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5317940711975098,
      "learning_rate": 2.668942317205897e-05,
      "loss": 2.6605,
      "step": 4712
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5721125602722168,
      "learning_rate": 2.6669634906500446e-05,
      "loss": 2.9969,
      "step": 4716
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5925160646438599,
      "learning_rate": 2.6649846640941923e-05,
      "loss": 2.6141,
      "step": 4720
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5406255722045898,
      "learning_rate": 2.66300583753834e-05,
      "loss": 2.6511,
      "step": 4724
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5240296125411987,
      "learning_rate": 2.6610270109824875e-05,
      "loss": 2.6192,
      "step": 4728
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5159399509429932,
      "learning_rate": 2.659048184426635e-05,
      "loss": 2.6747,
      "step": 4732
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5311481952667236,
      "learning_rate": 2.6570693578707828e-05,
      "loss": 2.6445,
      "step": 4736
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.610335111618042,
      "learning_rate": 2.6550905313149304e-05,
      "loss": 2.7823,
      "step": 4740
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5212194919586182,
      "learning_rate": 2.653111704759078e-05,
      "loss": 2.7615,
      "step": 4744
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5369598865509033,
      "learning_rate": 2.6511328782032257e-05,
      "loss": 2.5804,
      "step": 4748
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4863191843032837,
      "learning_rate": 2.6491540516473733e-05,
      "loss": 2.8282,
      "step": 4752
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6071970462799072,
      "learning_rate": 2.647175225091521e-05,
      "loss": 2.7363,
      "step": 4756
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.543915867805481,
      "learning_rate": 2.6451963985356686e-05,
      "loss": 2.8169,
      "step": 4760
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.399947166442871,
      "learning_rate": 2.6432175719798162e-05,
      "loss": 2.7116,
      "step": 4764
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.3791353702545166,
      "learning_rate": 2.641238745423964e-05,
      "loss": 2.8288,
      "step": 4768
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5583422183990479,
      "learning_rate": 2.639259918868111e-05,
      "loss": 2.6392,
      "step": 4772
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4231597185134888,
      "learning_rate": 2.6372810923122588e-05,
      "loss": 2.6542,
      "step": 4776
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.5944623947143555,
      "learning_rate": 2.6353022657564064e-05,
      "loss": 2.7268,
      "step": 4780
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.586701512336731,
      "learning_rate": 2.633323439200554e-05,
      "loss": 2.8372,
      "step": 4784
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.9561891555786133,
      "learning_rate": 2.6313446126447017e-05,
      "loss": 2.6418,
      "step": 4788
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4587674140930176,
      "learning_rate": 2.6293657860888493e-05,
      "loss": 2.6083,
      "step": 4792
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.4791700839996338,
      "learning_rate": 2.627386959532997e-05,
      "loss": 2.7053,
      "step": 4796
    },
    {
      "epoch": 0.47,
      "grad_norm": 1.6022498607635498,
      "learning_rate": 2.6254081329771446e-05,
      "loss": 2.7117,
      "step": 4800
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.499300241470337,
      "learning_rate": 2.6234293064212922e-05,
      "loss": 2.879,
      "step": 4804
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4296376705169678,
      "learning_rate": 2.62145047986544e-05,
      "loss": 2.9341,
      "step": 4808
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4815930128097534,
      "learning_rate": 2.6194716533095875e-05,
      "loss": 2.505,
      "step": 4812
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.395080327987671,
      "learning_rate": 2.617492826753735e-05,
      "loss": 2.7681,
      "step": 4816
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5271371603012085,
      "learning_rate": 2.6155140001978827e-05,
      "loss": 2.7507,
      "step": 4820
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.522066593170166,
      "learning_rate": 2.6135351736420304e-05,
      "loss": 2.7203,
      "step": 4824
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4368566274642944,
      "learning_rate": 2.611556347086178e-05,
      "loss": 2.6228,
      "step": 4828
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.473230004310608,
      "learning_rate": 2.6095775205303256e-05,
      "loss": 2.726,
      "step": 4832
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4038417339324951,
      "learning_rate": 2.6075986939744736e-05,
      "loss": 2.6796,
      "step": 4836
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5209479331970215,
      "learning_rate": 2.6056198674186212e-05,
      "loss": 2.8567,
      "step": 4840
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.7008355855941772,
      "learning_rate": 2.603641040862769e-05,
      "loss": 2.8475,
      "step": 4844
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6162598133087158,
      "learning_rate": 2.6016622143069165e-05,
      "loss": 2.6449,
      "step": 4848
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4284156560897827,
      "learning_rate": 2.599683387751064e-05,
      "loss": 2.7213,
      "step": 4852
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3640947341918945,
      "learning_rate": 2.597704561195211e-05,
      "loss": 2.6178,
      "step": 4856
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6424022912979126,
      "learning_rate": 2.5957257346393587e-05,
      "loss": 2.6476,
      "step": 4860
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6182935237884521,
      "learning_rate": 2.5937469080835063e-05,
      "loss": 2.6933,
      "step": 4864
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3428994417190552,
      "learning_rate": 2.591768081527654e-05,
      "loss": 2.6471,
      "step": 4868
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.550610899925232,
      "learning_rate": 2.5897892549718016e-05,
      "loss": 2.651,
      "step": 4872
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.8785467147827148,
      "learning_rate": 2.5878104284159492e-05,
      "loss": 2.6447,
      "step": 4876
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5535615682601929,
      "learning_rate": 2.585831601860097e-05,
      "loss": 2.6525,
      "step": 4880
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5342074632644653,
      "learning_rate": 2.5838527753042445e-05,
      "loss": 2.7041,
      "step": 4884
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.6638293266296387,
      "learning_rate": 2.581873948748392e-05,
      "loss": 2.6759,
      "step": 4888
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.4886223077774048,
      "learning_rate": 2.5798951221925398e-05,
      "loss": 2.6616,
      "step": 4892
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.539483904838562,
      "learning_rate": 2.5779162956366877e-05,
      "loss": 2.6294,
      "step": 4896
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.415686845779419,
      "learning_rate": 2.5759374690808354e-05,
      "loss": 2.5885,
      "step": 4900
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6120994091033936,
      "learning_rate": 2.573958642524983e-05,
      "loss": 2.6169,
      "step": 4904
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.713667392730713,
      "learning_rate": 2.5719798159691306e-05,
      "loss": 2.7591,
      "step": 4908
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5422923564910889,
      "learning_rate": 2.5700009894132783e-05,
      "loss": 2.7633,
      "step": 4912
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5131072998046875,
      "learning_rate": 2.568022162857426e-05,
      "loss": 2.6749,
      "step": 4916
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5619624853134155,
      "learning_rate": 2.5660433363015735e-05,
      "loss": 2.5834,
      "step": 4920
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4721622467041016,
      "learning_rate": 2.564064509745721e-05,
      "loss": 2.6411,
      "step": 4924
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.424367070198059,
      "learning_rate": 2.5620856831898688e-05,
      "loss": 2.6848,
      "step": 4928
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.480597734451294,
      "learning_rate": 2.5601068566340164e-05,
      "loss": 2.5919,
      "step": 4932
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5287058353424072,
      "learning_rate": 2.558128030078164e-05,
      "loss": 2.7616,
      "step": 4936
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7165645360946655,
      "learning_rate": 2.556149203522311e-05,
      "loss": 2.7636,
      "step": 4940
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4173622131347656,
      "learning_rate": 2.5541703769664586e-05,
      "loss": 2.7301,
      "step": 4944
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4607168436050415,
      "learning_rate": 2.5521915504106063e-05,
      "loss": 2.609,
      "step": 4948
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4109200239181519,
      "learning_rate": 2.550212723854754e-05,
      "loss": 2.7442,
      "step": 4952
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5535697937011719,
      "learning_rate": 2.548233897298902e-05,
      "loss": 2.5371,
      "step": 4956
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.4528133869171143,
      "learning_rate": 2.5462550707430495e-05,
      "loss": 2.5971,
      "step": 4960
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7372537851333618,
      "learning_rate": 2.544276244187197e-05,
      "loss": 2.8321,
      "step": 4964
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6011797189712524,
      "learning_rate": 2.5422974176313448e-05,
      "loss": 2.7647,
      "step": 4968
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6484735012054443,
      "learning_rate": 2.5403185910754924e-05,
      "loss": 2.5399,
      "step": 4972
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5052474737167358,
      "learning_rate": 2.53833976451964e-05,
      "loss": 2.6719,
      "step": 4976
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5310105085372925,
      "learning_rate": 2.5363609379637877e-05,
      "loss": 2.6558,
      "step": 4980
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.469632863998413,
      "learning_rate": 2.5343821114079353e-05,
      "loss": 2.7404,
      "step": 4984
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.53377103805542,
      "learning_rate": 2.532403284852083e-05,
      "loss": 2.7085,
      "step": 4988
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.6303514242172241,
      "learning_rate": 2.5304244582962306e-05,
      "loss": 2.749,
      "step": 4992
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.5342525243759155,
      "learning_rate": 2.5284456317403782e-05,
      "loss": 2.8031,
      "step": 4996
    },
    {
      "epoch": 0.49,
      "grad_norm": 1.7527660131454468,
      "learning_rate": 2.526466805184526e-05,
      "loss": 2.6969,
      "step": 5000
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5336272716522217,
      "learning_rate": 2.5244879786286735e-05,
      "loss": 2.6488,
      "step": 5004
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6374998092651367,
      "learning_rate": 2.522509152072821e-05,
      "loss": 2.5628,
      "step": 5008
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4945316314697266,
      "learning_rate": 2.5205303255169687e-05,
      "loss": 2.8031,
      "step": 5012
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5056184530258179,
      "learning_rate": 2.5185514989611164e-05,
      "loss": 2.7111,
      "step": 5016
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.389101505279541,
      "learning_rate": 2.516572672405264e-05,
      "loss": 2.6636,
      "step": 5020
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.8190226554870605,
      "learning_rate": 2.5145938458494113e-05,
      "loss": 2.692,
      "step": 5024
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.553562879562378,
      "learning_rate": 2.512615019293559e-05,
      "loss": 2.6122,
      "step": 5028
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.710195779800415,
      "learning_rate": 2.5106361927377066e-05,
      "loss": 2.7634,
      "step": 5032
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.669340968132019,
      "learning_rate": 2.5086573661818542e-05,
      "loss": 2.6008,
      "step": 5036
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.7336004972457886,
      "learning_rate": 2.5066785396260018e-05,
      "loss": 2.6427,
      "step": 5040
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6529115438461304,
      "learning_rate": 2.5046997130701495e-05,
      "loss": 2.6774,
      "step": 5044
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.463104248046875,
      "learning_rate": 2.502720886514297e-05,
      "loss": 2.5906,
      "step": 5048
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.594033122062683,
      "learning_rate": 2.5007420599584447e-05,
      "loss": 2.5802,
      "step": 5052
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5680876970291138,
      "learning_rate": 2.4987632334025923e-05,
      "loss": 2.7835,
      "step": 5056
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.551176905632019,
      "learning_rate": 2.49678440684674e-05,
      "loss": 2.7352,
      "step": 5060
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6078176498413086,
      "learning_rate": 2.4948055802908876e-05,
      "loss": 2.6514,
      "step": 5064
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4909130334854126,
      "learning_rate": 2.4928267537350352e-05,
      "loss": 2.6863,
      "step": 5068
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.485734462738037,
      "learning_rate": 2.490847927179183e-05,
      "loss": 2.6079,
      "step": 5072
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.413549542427063,
      "learning_rate": 2.4888691006233305e-05,
      "loss": 2.7355,
      "step": 5076
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.5501868724822998,
      "learning_rate": 2.486890274067478e-05,
      "loss": 2.6522,
      "step": 5080
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4532791376113892,
      "learning_rate": 2.4849114475116258e-05,
      "loss": 2.7269,
      "step": 5084
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6605114936828613,
      "learning_rate": 2.482932620955773e-05,
      "loss": 2.6386,
      "step": 5088
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.6736440658569336,
      "learning_rate": 2.480953794399921e-05,
      "loss": 2.8607,
      "step": 5092
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.7026129961013794,
      "learning_rate": 2.4789749678440687e-05,
      "loss": 2.7838,
      "step": 5096
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.4763104915618896,
      "learning_rate": 2.4769961412882163e-05,
      "loss": 2.5136,
      "step": 5100
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.555331826210022,
      "learning_rate": 2.475017314732364e-05,
      "loss": 2.5966,
      "step": 5104
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.4764961004257202,
      "learning_rate": 2.4730384881765116e-05,
      "loss": 2.7076,
      "step": 5108
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5390756130218506,
      "learning_rate": 2.4710596616206592e-05,
      "loss": 2.7615,
      "step": 5112
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.503321647644043,
      "learning_rate": 2.4690808350648068e-05,
      "loss": 2.62,
      "step": 5116
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5027825832366943,
      "learning_rate": 2.4671020085089545e-05,
      "loss": 2.9407,
      "step": 5120
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.4294583797454834,
      "learning_rate": 2.465123181953102e-05,
      "loss": 2.8738,
      "step": 5124
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5709290504455566,
      "learning_rate": 2.4631443553972494e-05,
      "loss": 2.6954,
      "step": 5128
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6391737461090088,
      "learning_rate": 2.461165528841397e-05,
      "loss": 2.8727,
      "step": 5132
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5847351551055908,
      "learning_rate": 2.4591867022855447e-05,
      "loss": 2.5896,
      "step": 5136
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.4603075981140137,
      "learning_rate": 2.4572078757296923e-05,
      "loss": 2.7727,
      "step": 5140
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.8148667812347412,
      "learning_rate": 2.45522904917384e-05,
      "loss": 2.518,
      "step": 5144
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6938127279281616,
      "learning_rate": 2.4532502226179875e-05,
      "loss": 2.575,
      "step": 5148
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.495416522026062,
      "learning_rate": 2.4512713960621352e-05,
      "loss": 2.7263,
      "step": 5152
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5447607040405273,
      "learning_rate": 2.4492925695062828e-05,
      "loss": 2.6644,
      "step": 5156
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.594707727432251,
      "learning_rate": 2.4473137429504304e-05,
      "loss": 2.5444,
      "step": 5160
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.613542914390564,
      "learning_rate": 2.445334916394578e-05,
      "loss": 2.5996,
      "step": 5164
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6237636804580688,
      "learning_rate": 2.443356089838726e-05,
      "loss": 2.6802,
      "step": 5168
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5020370483398438,
      "learning_rate": 2.4413772632828733e-05,
      "loss": 2.6467,
      "step": 5172
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.441662311553955,
      "learning_rate": 2.439398436727021e-05,
      "loss": 2.6707,
      "step": 5176
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.626299500465393,
      "learning_rate": 2.4374196101711686e-05,
      "loss": 2.5029,
      "step": 5180
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.5790330171585083,
      "learning_rate": 2.4354407836153162e-05,
      "loss": 2.6478,
      "step": 5184
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.4696558713912964,
      "learning_rate": 2.433461957059464e-05,
      "loss": 2.6289,
      "step": 5188
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.3996411561965942,
      "learning_rate": 2.4314831305036115e-05,
      "loss": 2.5142,
      "step": 5192
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.630058765411377,
      "learning_rate": 2.429504303947759e-05,
      "loss": 2.6655,
      "step": 5196
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.648018717765808,
      "learning_rate": 2.4275254773919068e-05,
      "loss": 2.425,
      "step": 5200
    },
    {
      "epoch": 0.51,
      "grad_norm": 1.6513872146606445,
      "learning_rate": 2.4255466508360544e-05,
      "loss": 2.7594,
      "step": 5204
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.592578649520874,
      "learning_rate": 2.423567824280202e-05,
      "loss": 2.5232,
      "step": 5208
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.6589707136154175,
      "learning_rate": 2.4215889977243493e-05,
      "loss": 2.6596,
      "step": 5212
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5870932340621948,
      "learning_rate": 2.419610171168497e-05,
      "loss": 2.5496,
      "step": 5216
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.6115508079528809,
      "learning_rate": 2.4176313446126446e-05,
      "loss": 2.7024,
      "step": 5220
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5552654266357422,
      "learning_rate": 2.4156525180567922e-05,
      "loss": 2.6954,
      "step": 5224
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.530119776725769,
      "learning_rate": 2.4136736915009402e-05,
      "loss": 2.6178,
      "step": 5228
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5883251428604126,
      "learning_rate": 2.4116948649450878e-05,
      "loss": 2.7307,
      "step": 5232
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.6557033061981201,
      "learning_rate": 2.4097160383892355e-05,
      "loss": 2.7623,
      "step": 5236
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.6295868158340454,
      "learning_rate": 2.407737211833383e-05,
      "loss": 2.5603,
      "step": 5240
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4882088899612427,
      "learning_rate": 2.4057583852775307e-05,
      "loss": 2.7022,
      "step": 5244
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.559869647026062,
      "learning_rate": 2.4037795587216784e-05,
      "loss": 2.8356,
      "step": 5248
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5058995485305786,
      "learning_rate": 2.401800732165826e-05,
      "loss": 2.8771,
      "step": 5252
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5419751405715942,
      "learning_rate": 2.3998219056099733e-05,
      "loss": 2.6347,
      "step": 5256
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5456287860870361,
      "learning_rate": 2.397843079054121e-05,
      "loss": 2.7462,
      "step": 5260
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.6244553327560425,
      "learning_rate": 2.3958642524982685e-05,
      "loss": 2.7263,
      "step": 5264
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5929856300354004,
      "learning_rate": 2.3938854259424162e-05,
      "loss": 2.6368,
      "step": 5268
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.1067686080932617,
      "learning_rate": 2.3919065993865638e-05,
      "loss": 2.784,
      "step": 5272
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.4130641222000122,
      "learning_rate": 2.3899277728307114e-05,
      "loss": 2.6074,
      "step": 5276
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.516021490097046,
      "learning_rate": 2.387948946274859e-05,
      "loss": 2.7861,
      "step": 5280
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.6379324197769165,
      "learning_rate": 2.3859701197190067e-05,
      "loss": 2.6482,
      "step": 5284
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.7864313125610352,
      "learning_rate": 2.3839912931631543e-05,
      "loss": 2.6253,
      "step": 5288
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.6106088161468506,
      "learning_rate": 2.382012466607302e-05,
      "loss": 2.5256,
      "step": 5292
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5859754085540771,
      "learning_rate": 2.3800336400514496e-05,
      "loss": 2.6219,
      "step": 5296
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.5231300592422485,
      "learning_rate": 2.3780548134955972e-05,
      "loss": 2.7499,
      "step": 5300
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.472907543182373,
      "learning_rate": 2.376075986939745e-05,
      "loss": 2.5311,
      "step": 5304
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.8170127868652344,
      "learning_rate": 2.3740971603838925e-05,
      "loss": 2.5569,
      "step": 5308
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5469038486480713,
      "learning_rate": 2.37211833382804e-05,
      "loss": 2.6323,
      "step": 5312
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.594907283782959,
      "learning_rate": 2.3701395072721878e-05,
      "loss": 2.8168,
      "step": 5316
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5616902112960815,
      "learning_rate": 2.3681606807163354e-05,
      "loss": 2.7958,
      "step": 5320
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5545421838760376,
      "learning_rate": 2.366181854160483e-05,
      "loss": 2.6999,
      "step": 5324
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4910705089569092,
      "learning_rate": 2.3642030276046307e-05,
      "loss": 2.6652,
      "step": 5328
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.510505199432373,
      "learning_rate": 2.3622242010487783e-05,
      "loss": 2.6772,
      "step": 5332
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5067452192306519,
      "learning_rate": 2.360245374492926e-05,
      "loss": 2.765,
      "step": 5336
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6641408205032349,
      "learning_rate": 2.3582665479370732e-05,
      "loss": 2.5982,
      "step": 5340
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6969571113586426,
      "learning_rate": 2.356287721381221e-05,
      "loss": 2.8366,
      "step": 5344
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5563101768493652,
      "learning_rate": 2.3543088948253685e-05,
      "loss": 2.6941,
      "step": 5348
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5236217975616455,
      "learning_rate": 2.352330068269516e-05,
      "loss": 2.6801,
      "step": 5352
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6700359582901,
      "learning_rate": 2.3503512417136637e-05,
      "loss": 2.8281,
      "step": 5356
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5832792520523071,
      "learning_rate": 2.3483724151578114e-05,
      "loss": 2.6856,
      "step": 5360
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7960867881774902,
      "learning_rate": 2.3463935886019593e-05,
      "loss": 2.6418,
      "step": 5364
    },
    {
      "epoch": 0.53,
      "grad_norm": 2.743746042251587,
      "learning_rate": 2.344414762046107e-05,
      "loss": 2.7925,
      "step": 5368
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.7574864625930786,
      "learning_rate": 2.3424359354902546e-05,
      "loss": 2.8191,
      "step": 5372
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4211304187774658,
      "learning_rate": 2.3404571089344022e-05,
      "loss": 2.6055,
      "step": 5376
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6393153667449951,
      "learning_rate": 2.33847828237855e-05,
      "loss": 2.541,
      "step": 5380
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.6058558225631714,
      "learning_rate": 2.336499455822697e-05,
      "loss": 2.5992,
      "step": 5384
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.4362446069717407,
      "learning_rate": 2.3345206292668448e-05,
      "loss": 2.6894,
      "step": 5388
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.5442392826080322,
      "learning_rate": 2.3325418027109924e-05,
      "loss": 2.6369,
      "step": 5392
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.654779314994812,
      "learning_rate": 2.33056297615514e-05,
      "loss": 2.7932,
      "step": 5396
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.576923131942749,
      "learning_rate": 2.3285841495992877e-05,
      "loss": 2.6261,
      "step": 5400
    },
    {
      "epoch": 0.53,
      "grad_norm": 1.595237374305725,
      "learning_rate": 2.3266053230434353e-05,
      "loss": 2.5116,
      "step": 5404
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7787840366363525,
      "learning_rate": 2.324626496487583e-05,
      "loss": 2.7103,
      "step": 5408
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4982911348342896,
      "learning_rate": 2.3226476699317306e-05,
      "loss": 2.587,
      "step": 5412
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5009403228759766,
      "learning_rate": 2.3206688433758782e-05,
      "loss": 2.7021,
      "step": 5416
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5201959609985352,
      "learning_rate": 2.318690016820026e-05,
      "loss": 2.5477,
      "step": 5420
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.480794906616211,
      "learning_rate": 2.3167111902641735e-05,
      "loss": 2.725,
      "step": 5424
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5079982280731201,
      "learning_rate": 2.314732363708321e-05,
      "loss": 2.4785,
      "step": 5428
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.601902723312378,
      "learning_rate": 2.3127535371524688e-05,
      "loss": 2.5547,
      "step": 5432
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5328859090805054,
      "learning_rate": 2.3107747105966164e-05,
      "loss": 2.6307,
      "step": 5436
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5834133625030518,
      "learning_rate": 2.308795884040764e-05,
      "loss": 2.656,
      "step": 5440
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6440998315811157,
      "learning_rate": 2.3068170574849116e-05,
      "loss": 2.7507,
      "step": 5444
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4416550397872925,
      "learning_rate": 2.3048382309290593e-05,
      "loss": 2.8631,
      "step": 5448
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5241260528564453,
      "learning_rate": 2.302859404373207e-05,
      "loss": 2.815,
      "step": 5452
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.8453205823898315,
      "learning_rate": 2.3008805778173545e-05,
      "loss": 2.6714,
      "step": 5456
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5847132205963135,
      "learning_rate": 2.2989017512615022e-05,
      "loss": 2.6486,
      "step": 5460
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6181340217590332,
      "learning_rate": 2.2969229247056498e-05,
      "loss": 2.5461,
      "step": 5464
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4716991186141968,
      "learning_rate": 2.294944098149797e-05,
      "loss": 2.6739,
      "step": 5468
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4331846237182617,
      "learning_rate": 2.2929652715939447e-05,
      "loss": 2.5011,
      "step": 5472
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.7750493288040161,
      "learning_rate": 2.2909864450380924e-05,
      "loss": 2.683,
      "step": 5476
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.6344447135925293,
      "learning_rate": 2.28900761848224e-05,
      "loss": 2.8271,
      "step": 5480
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4624501466751099,
      "learning_rate": 2.2870287919263876e-05,
      "loss": 2.6828,
      "step": 5484
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.3957983255386353,
      "learning_rate": 2.2850499653705353e-05,
      "loss": 2.5376,
      "step": 5488
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4633480310440063,
      "learning_rate": 2.283071138814683e-05,
      "loss": 2.6159,
      "step": 5492
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.607138752937317,
      "learning_rate": 2.2810923122588305e-05,
      "loss": 2.5313,
      "step": 5496
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.5315511226654053,
      "learning_rate": 2.2791134857029785e-05,
      "loss": 2.6017,
      "step": 5500
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.3674393892288208,
      "learning_rate": 2.277134659147126e-05,
      "loss": 2.5482,
      "step": 5504
    },
    {
      "epoch": 0.54,
      "grad_norm": 1.4765186309814453,
      "learning_rate": 2.2751558325912734e-05,
      "loss": 2.5593,
      "step": 5508
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5527360439300537,
      "learning_rate": 2.273177006035421e-05,
      "loss": 2.7501,
      "step": 5512
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7403366565704346,
      "learning_rate": 2.2711981794795687e-05,
      "loss": 2.6449,
      "step": 5516
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.3762569427490234,
      "learning_rate": 2.2692193529237163e-05,
      "loss": 2.5219,
      "step": 5520
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5803818702697754,
      "learning_rate": 2.267240526367864e-05,
      "loss": 2.5245,
      "step": 5524
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.6432803869247437,
      "learning_rate": 2.2652616998120116e-05,
      "loss": 2.657,
      "step": 5528
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4227681159973145,
      "learning_rate": 2.2632828732561592e-05,
      "loss": 2.6232,
      "step": 5532
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.6246461868286133,
      "learning_rate": 2.261304046700307e-05,
      "loss": 2.7291,
      "step": 5536
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5243287086486816,
      "learning_rate": 2.2593252201444545e-05,
      "loss": 2.828,
      "step": 5540
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4307969808578491,
      "learning_rate": 2.257346393588602e-05,
      "loss": 2.6103,
      "step": 5544
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4919029474258423,
      "learning_rate": 2.2553675670327497e-05,
      "loss": 2.652,
      "step": 5548
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.633585810661316,
      "learning_rate": 2.253388740476897e-05,
      "loss": 2.5161,
      "step": 5552
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5435733795166016,
      "learning_rate": 2.2514099139210447e-05,
      "loss": 2.5211,
      "step": 5556
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.804867148399353,
      "learning_rate": 2.2494310873651926e-05,
      "loss": 2.6671,
      "step": 5560
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.7000728845596313,
      "learning_rate": 2.2474522608093403e-05,
      "loss": 2.5494,
      "step": 5564
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.597815752029419,
      "learning_rate": 2.245473434253488e-05,
      "loss": 2.7588,
      "step": 5568
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.824008584022522,
      "learning_rate": 2.2434946076976355e-05,
      "loss": 2.6684,
      "step": 5572
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4856268167495728,
      "learning_rate": 2.2415157811417832e-05,
      "loss": 2.5801,
      "step": 5576
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4334465265274048,
      "learning_rate": 2.2395369545859308e-05,
      "loss": 2.6735,
      "step": 5580
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.529638648033142,
      "learning_rate": 2.2375581280300784e-05,
      "loss": 2.4979,
      "step": 5584
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5678493976593018,
      "learning_rate": 2.235579301474226e-05,
      "loss": 2.458,
      "step": 5588
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4850939512252808,
      "learning_rate": 2.2336004749183734e-05,
      "loss": 2.5068,
      "step": 5592
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.8440673351287842,
      "learning_rate": 2.231621648362521e-05,
      "loss": 2.6083,
      "step": 5596
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4649991989135742,
      "learning_rate": 2.2296428218066686e-05,
      "loss": 2.7386,
      "step": 5600
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.5300774574279785,
      "learning_rate": 2.2276639952508163e-05,
      "loss": 2.6624,
      "step": 5604
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.4183005094528198,
      "learning_rate": 2.225685168694964e-05,
      "loss": 2.5477,
      "step": 5608
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5531784296035767,
      "learning_rate": 2.2237063421391115e-05,
      "loss": 2.5075,
      "step": 5612
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6235988140106201,
      "learning_rate": 2.221727515583259e-05,
      "loss": 2.5529,
      "step": 5616
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6714985370635986,
      "learning_rate": 2.2197486890274068e-05,
      "loss": 2.7613,
      "step": 5620
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5148706436157227,
      "learning_rate": 2.2177698624715544e-05,
      "loss": 2.8314,
      "step": 5624
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6229020357131958,
      "learning_rate": 2.215791035915702e-05,
      "loss": 2.5539,
      "step": 5628
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7232576608657837,
      "learning_rate": 2.2138122093598497e-05,
      "loss": 2.756,
      "step": 5632
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4733608961105347,
      "learning_rate": 2.2118333828039973e-05,
      "loss": 2.6255,
      "step": 5636
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.748321294784546,
      "learning_rate": 2.209854556248145e-05,
      "loss": 2.675,
      "step": 5640
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5811595916748047,
      "learning_rate": 2.2078757296922926e-05,
      "loss": 2.7092,
      "step": 5644
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.3841744661331177,
      "learning_rate": 2.2058969031364402e-05,
      "loss": 2.5024,
      "step": 5648
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.558066964149475,
      "learning_rate": 2.203918076580588e-05,
      "loss": 2.5734,
      "step": 5652
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4873226881027222,
      "learning_rate": 2.2019392500247355e-05,
      "loss": 2.7111,
      "step": 5656
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5889838933944702,
      "learning_rate": 2.199960423468883e-05,
      "loss": 2.6065,
      "step": 5660
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4627254009246826,
      "learning_rate": 2.1979815969130307e-05,
      "loss": 2.6233,
      "step": 5664
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6303706169128418,
      "learning_rate": 2.1960027703571784e-05,
      "loss": 2.7411,
      "step": 5668
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7008352279663086,
      "learning_rate": 2.194023943801326e-05,
      "loss": 2.5536,
      "step": 5672
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.538040041923523,
      "learning_rate": 2.1920451172454733e-05,
      "loss": 2.6556,
      "step": 5676
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5172909498214722,
      "learning_rate": 2.190066290689621e-05,
      "loss": 2.6931,
      "step": 5680
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.3563427925109863,
      "learning_rate": 2.1880874641337686e-05,
      "loss": 2.738,
      "step": 5684
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6162517070770264,
      "learning_rate": 2.1861086375779162e-05,
      "loss": 2.6649,
      "step": 5688
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.6602249145507812,
      "learning_rate": 2.1841298110220638e-05,
      "loss": 2.6552,
      "step": 5692
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.4889838695526123,
      "learning_rate": 2.1821509844662118e-05,
      "loss": 2.4506,
      "step": 5696
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5726191997528076,
      "learning_rate": 2.1801721579103594e-05,
      "loss": 2.696,
      "step": 5700
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.7843947410583496,
      "learning_rate": 2.178193331354507e-05,
      "loss": 2.6979,
      "step": 5704
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.5087847709655762,
      "learning_rate": 2.1762145047986547e-05,
      "loss": 2.5383,
      "step": 5708
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6036449670791626,
      "learning_rate": 2.1742356782428023e-05,
      "loss": 2.5809,
      "step": 5712
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.578401803970337,
      "learning_rate": 2.17225685168695e-05,
      "loss": 2.7029,
      "step": 5716
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.510796308517456,
      "learning_rate": 2.1702780251310972e-05,
      "loss": 2.4606,
      "step": 5720
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6872867345809937,
      "learning_rate": 2.168299198575245e-05,
      "loss": 2.6832,
      "step": 5724
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6195570230484009,
      "learning_rate": 2.1663203720193925e-05,
      "loss": 2.6499,
      "step": 5728
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4994075298309326,
      "learning_rate": 2.16434154546354e-05,
      "loss": 2.7302,
      "step": 5732
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5541733503341675,
      "learning_rate": 2.1623627189076878e-05,
      "loss": 2.6049,
      "step": 5736
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5584886074066162,
      "learning_rate": 2.1603838923518354e-05,
      "loss": 2.5976,
      "step": 5740
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5630571842193604,
      "learning_rate": 2.158405065795983e-05,
      "loss": 2.5042,
      "step": 5744
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4705463647842407,
      "learning_rate": 2.1564262392401307e-05,
      "loss": 2.6125,
      "step": 5748
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5950294733047485,
      "learning_rate": 2.1544474126842783e-05,
      "loss": 2.5934,
      "step": 5752
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4994343519210815,
      "learning_rate": 2.152468586128426e-05,
      "loss": 2.6526,
      "step": 5756
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4799262285232544,
      "learning_rate": 2.1504897595725736e-05,
      "loss": 2.6464,
      "step": 5760
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.4674098491668701,
      "learning_rate": 2.1485109330167212e-05,
      "loss": 2.6307,
      "step": 5764
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.519846796989441,
      "learning_rate": 2.146532106460869e-05,
      "loss": 2.6029,
      "step": 5768
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6543153524398804,
      "learning_rate": 2.1445532799050165e-05,
      "loss": 2.5781,
      "step": 5772
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.8287581205368042,
      "learning_rate": 2.142574453349164e-05,
      "loss": 2.6109,
      "step": 5776
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6444793939590454,
      "learning_rate": 2.1405956267933117e-05,
      "loss": 2.7402,
      "step": 5780
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7062760591506958,
      "learning_rate": 2.1386168002374594e-05,
      "loss": 2.8537,
      "step": 5784
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.794710636138916,
      "learning_rate": 2.136637973681607e-05,
      "loss": 2.8538,
      "step": 5788
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.6265027523040771,
      "learning_rate": 2.1346591471257546e-05,
      "loss": 2.7576,
      "step": 5792
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.637753963470459,
      "learning_rate": 2.1326803205699023e-05,
      "loss": 2.6573,
      "step": 5796
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5703392028808594,
      "learning_rate": 2.13070149401405e-05,
      "loss": 2.5352,
      "step": 5800
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.7320841550827026,
      "learning_rate": 2.1287226674581972e-05,
      "loss": 2.6481,
      "step": 5804
    },
    {
      "epoch": 0.57,
      "grad_norm": 1.5625312328338623,
      "learning_rate": 2.1267438409023448e-05,
      "loss": 2.5449,
      "step": 5808
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4894498586654663,
      "learning_rate": 2.1247650143464924e-05,
      "loss": 2.7572,
      "step": 5812
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.9753198623657227,
      "learning_rate": 2.12278618779064e-05,
      "loss": 2.6719,
      "step": 5816
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6790626049041748,
      "learning_rate": 2.1208073612347877e-05,
      "loss": 2.6557,
      "step": 5820
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5721454620361328,
      "learning_rate": 2.1188285346789353e-05,
      "loss": 2.5762,
      "step": 5824
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.654955506324768,
      "learning_rate": 2.116849708123083e-05,
      "loss": 2.6838,
      "step": 5828
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5922191143035889,
      "learning_rate": 2.114870881567231e-05,
      "loss": 2.8244,
      "step": 5832
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5361075401306152,
      "learning_rate": 2.1128920550113786e-05,
      "loss": 2.6246,
      "step": 5836
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4784355163574219,
      "learning_rate": 2.1109132284555262e-05,
      "loss": 2.5971,
      "step": 5840
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.9370943307876587,
      "learning_rate": 2.108934401899674e-05,
      "loss": 2.4188,
      "step": 5844
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6138865947723389,
      "learning_rate": 2.106955575343821e-05,
      "loss": 2.7327,
      "step": 5848
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6351326704025269,
      "learning_rate": 2.1049767487879688e-05,
      "loss": 2.6568,
      "step": 5852
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.4805229902267456,
      "learning_rate": 2.1029979222321164e-05,
      "loss": 2.7057,
      "step": 5856
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7400959730148315,
      "learning_rate": 2.101019095676264e-05,
      "loss": 2.7251,
      "step": 5860
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5572409629821777,
      "learning_rate": 2.0990402691204117e-05,
      "loss": 2.5459,
      "step": 5864
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6387975215911865,
      "learning_rate": 2.0970614425645593e-05,
      "loss": 2.7208,
      "step": 5868
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6002295017242432,
      "learning_rate": 2.095082616008707e-05,
      "loss": 2.5598,
      "step": 5872
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6020545959472656,
      "learning_rate": 2.0931037894528546e-05,
      "loss": 2.6175,
      "step": 5876
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5802550315856934,
      "learning_rate": 2.0911249628970022e-05,
      "loss": 2.7237,
      "step": 5880
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6119384765625,
      "learning_rate": 2.0891461363411498e-05,
      "loss": 2.7419,
      "step": 5884
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.748727798461914,
      "learning_rate": 2.087167309785297e-05,
      "loss": 2.5774,
      "step": 5888
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.707313060760498,
      "learning_rate": 2.085188483229445e-05,
      "loss": 2.764,
      "step": 5892
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5973708629608154,
      "learning_rate": 2.0832096566735927e-05,
      "loss": 2.5781,
      "step": 5896
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5735652446746826,
      "learning_rate": 2.0812308301177404e-05,
      "loss": 2.6071,
      "step": 5900
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.6533702611923218,
      "learning_rate": 2.079252003561888e-05,
      "loss": 2.7356,
      "step": 5904
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.7489056587219238,
      "learning_rate": 2.0772731770060356e-05,
      "loss": 2.6429,
      "step": 5908
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.793553352355957,
      "learning_rate": 2.0752943504501833e-05,
      "loss": 2.7273,
      "step": 5912
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7507030963897705,
      "learning_rate": 2.073315523894331e-05,
      "loss": 2.6684,
      "step": 5916
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6355698108673096,
      "learning_rate": 2.0713366973384785e-05,
      "loss": 2.5185,
      "step": 5920
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.636668086051941,
      "learning_rate": 2.069357870782626e-05,
      "loss": 2.6138,
      "step": 5924
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.161924123764038,
      "learning_rate": 2.0673790442267738e-05,
      "loss": 2.7274,
      "step": 5928
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6482921838760376,
      "learning_rate": 2.065400217670921e-05,
      "loss": 2.5224,
      "step": 5932
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7071282863616943,
      "learning_rate": 2.0634213911150687e-05,
      "loss": 2.514,
      "step": 5936
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6767092943191528,
      "learning_rate": 2.0614425645592163e-05,
      "loss": 2.5072,
      "step": 5940
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5524712800979614,
      "learning_rate": 2.059463738003364e-05,
      "loss": 2.402,
      "step": 5944
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.7386505603790283,
      "learning_rate": 2.0574849114475116e-05,
      "loss": 2.5022,
      "step": 5948
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5279655456542969,
      "learning_rate": 2.0555060848916592e-05,
      "loss": 2.5708,
      "step": 5952
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6289914846420288,
      "learning_rate": 2.053527258335807e-05,
      "loss": 2.6689,
      "step": 5956
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.536629557609558,
      "learning_rate": 2.0515484317799545e-05,
      "loss": 2.5802,
      "step": 5960
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6183369159698486,
      "learning_rate": 2.049569605224102e-05,
      "loss": 2.7748,
      "step": 5964
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5437310934066772,
      "learning_rate": 2.04759077866825e-05,
      "loss": 2.5857,
      "step": 5968
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.4670217037200928,
      "learning_rate": 2.0456119521123974e-05,
      "loss": 2.7475,
      "step": 5972
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.4012467861175537,
      "learning_rate": 2.043633125556545e-05,
      "loss": 2.4938,
      "step": 5976
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6246941089630127,
      "learning_rate": 2.0416542990006927e-05,
      "loss": 2.589,
      "step": 5980
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5825214385986328,
      "learning_rate": 2.0396754724448403e-05,
      "loss": 2.7124,
      "step": 5984
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.598771572113037,
      "learning_rate": 2.037696645888988e-05,
      "loss": 2.604,
      "step": 5988
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.6887823343276978,
      "learning_rate": 2.0357178193331356e-05,
      "loss": 2.463,
      "step": 5992
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5830531120300293,
      "learning_rate": 2.0337389927772832e-05,
      "loss": 2.5329,
      "step": 5996
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.5762251615524292,
      "learning_rate": 2.0317601662214308e-05,
      "loss": 2.5809,
      "step": 6000
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.8028018474578857,
      "learning_rate": 2.0297813396655785e-05,
      "loss": 2.5708,
      "step": 6004
    },
    {
      "epoch": 0.59,
      "grad_norm": 1.4725638628005981,
      "learning_rate": 2.027802513109726e-05,
      "loss": 2.7582,
      "step": 6008
    },
    {
      "epoch": 0.59,
      "grad_norm": 2.119457244873047,
      "learning_rate": 2.0258236865538737e-05,
      "loss": 2.7722,
      "step": 6012
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.713395357131958,
      "learning_rate": 2.023844859998021e-05,
      "loss": 2.5151,
      "step": 6016
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5877728462219238,
      "learning_rate": 2.0218660334421686e-05,
      "loss": 2.647,
      "step": 6020
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6302303075790405,
      "learning_rate": 2.0198872068863163e-05,
      "loss": 2.6777,
      "step": 6024
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.739067554473877,
      "learning_rate": 2.0179083803304642e-05,
      "loss": 2.736,
      "step": 6028
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.683003306388855,
      "learning_rate": 2.015929553774612e-05,
      "loss": 2.5068,
      "step": 6032
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5176719427108765,
      "learning_rate": 2.0139507272187595e-05,
      "loss": 2.4223,
      "step": 6036
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7050352096557617,
      "learning_rate": 2.011971900662907e-05,
      "loss": 2.7499,
      "step": 6040
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6220310926437378,
      "learning_rate": 2.0099930741070548e-05,
      "loss": 2.6881,
      "step": 6044
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.61378014087677,
      "learning_rate": 2.0080142475512024e-05,
      "loss": 2.7385,
      "step": 6048
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.709804892539978,
      "learning_rate": 2.00603542099535e-05,
      "loss": 2.4782,
      "step": 6052
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5907514095306396,
      "learning_rate": 2.0040565944394973e-05,
      "loss": 2.6041,
      "step": 6056
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.587908148765564,
      "learning_rate": 2.002077767883645e-05,
      "loss": 2.6641,
      "step": 6060
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4979619979858398,
      "learning_rate": 2.0000989413277926e-05,
      "loss": 2.6292,
      "step": 6064
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6397794485092163,
      "learning_rate": 1.9981201147719402e-05,
      "loss": 2.4687,
      "step": 6068
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5526165962219238,
      "learning_rate": 1.996141288216088e-05,
      "loss": 2.5717,
      "step": 6072
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6154141426086426,
      "learning_rate": 1.9941624616602355e-05,
      "loss": 2.4732,
      "step": 6076
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5267953872680664,
      "learning_rate": 1.992183635104383e-05,
      "loss": 2.5728,
      "step": 6080
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5741758346557617,
      "learning_rate": 1.9902048085485308e-05,
      "loss": 2.6762,
      "step": 6084
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4406962394714355,
      "learning_rate": 1.9882259819926784e-05,
      "loss": 2.4681,
      "step": 6088
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5490901470184326,
      "learning_rate": 1.986247155436826e-05,
      "loss": 2.7022,
      "step": 6092
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6361629962921143,
      "learning_rate": 1.9842683288809737e-05,
      "loss": 2.4699,
      "step": 6096
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.4828660488128662,
      "learning_rate": 1.9822895023251213e-05,
      "loss": 2.4622,
      "step": 6100
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.5388118028640747,
      "learning_rate": 1.980310675769269e-05,
      "loss": 2.5757,
      "step": 6104
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.7506879568099976,
      "learning_rate": 1.9783318492134165e-05,
      "loss": 2.5815,
      "step": 6108
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.6351765394210815,
      "learning_rate": 1.9763530226575642e-05,
      "loss": 2.7742,
      "step": 6112
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.708899736404419,
      "learning_rate": 1.9743741961017118e-05,
      "loss": 2.5452,
      "step": 6116
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6561249494552612,
      "learning_rate": 1.9723953695458594e-05,
      "loss": 2.5306,
      "step": 6120
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5460957288742065,
      "learning_rate": 1.970416542990007e-05,
      "loss": 2.529,
      "step": 6124
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5808017253875732,
      "learning_rate": 1.9684377164341547e-05,
      "loss": 2.587,
      "step": 6128
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5812370777130127,
      "learning_rate": 1.9664588898783023e-05,
      "loss": 2.6806,
      "step": 6132
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6108824014663696,
      "learning_rate": 1.96448006332245e-05,
      "loss": 2.7478,
      "step": 6136
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.501314401626587,
      "learning_rate": 1.9625012367665976e-05,
      "loss": 2.7401,
      "step": 6140
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5375041961669922,
      "learning_rate": 1.960522410210745e-05,
      "loss": 2.6419,
      "step": 6144
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5446549654006958,
      "learning_rate": 1.9585435836548925e-05,
      "loss": 2.4504,
      "step": 6148
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6299073696136475,
      "learning_rate": 1.95656475709904e-05,
      "loss": 2.79,
      "step": 6152
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5294485092163086,
      "learning_rate": 1.9545859305431878e-05,
      "loss": 2.5702,
      "step": 6156
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.7888283729553223,
      "learning_rate": 1.9526071039873354e-05,
      "loss": 2.7481,
      "step": 6160
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5239052772521973,
      "learning_rate": 1.9506282774314834e-05,
      "loss": 2.7299,
      "step": 6164
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5439248085021973,
      "learning_rate": 1.948649450875631e-05,
      "loss": 2.4492,
      "step": 6168
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5575426816940308,
      "learning_rate": 1.9466706243197787e-05,
      "loss": 2.5791,
      "step": 6172
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6688716411590576,
      "learning_rate": 1.9446917977639263e-05,
      "loss": 2.5319,
      "step": 6176
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6760215759277344,
      "learning_rate": 1.942712971208074e-05,
      "loss": 2.7329,
      "step": 6180
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5455204248428345,
      "learning_rate": 1.9407341446522212e-05,
      "loss": 2.6252,
      "step": 6184
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5902513265609741,
      "learning_rate": 1.938755318096369e-05,
      "loss": 2.5223,
      "step": 6188
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5582561492919922,
      "learning_rate": 1.9367764915405165e-05,
      "loss": 2.5884,
      "step": 6192
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6034671068191528,
      "learning_rate": 1.934797664984664e-05,
      "loss": 2.6888,
      "step": 6196
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6465849876403809,
      "learning_rate": 1.9328188384288117e-05,
      "loss": 2.4756,
      "step": 6200
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6276373863220215,
      "learning_rate": 1.9308400118729594e-05,
      "loss": 2.5836,
      "step": 6204
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.5838865041732788,
      "learning_rate": 1.928861185317107e-05,
      "loss": 2.6465,
      "step": 6208
    },
    {
      "epoch": 0.61,
      "grad_norm": 1.6422873735427856,
      "learning_rate": 1.9268823587612546e-05,
      "loss": 2.604,
      "step": 6212
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.591674566268921,
      "learning_rate": 1.9249035322054023e-05,
      "loss": 2.5624,
      "step": 6216
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.670748233795166,
      "learning_rate": 1.92292470564955e-05,
      "loss": 2.5349,
      "step": 6220
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5904476642608643,
      "learning_rate": 1.9209458790936975e-05,
      "loss": 2.4124,
      "step": 6224
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5372642278671265,
      "learning_rate": 1.9189670525378452e-05,
      "loss": 2.8056,
      "step": 6228
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6156257390975952,
      "learning_rate": 1.9169882259819928e-05,
      "loss": 2.3666,
      "step": 6232
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5358694791793823,
      "learning_rate": 1.9150093994261404e-05,
      "loss": 2.5517,
      "step": 6236
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6131166219711304,
      "learning_rate": 1.913030572870288e-05,
      "loss": 2.699,
      "step": 6240
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7722057104110718,
      "learning_rate": 1.9110517463144357e-05,
      "loss": 2.6431,
      "step": 6244
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7471140623092651,
      "learning_rate": 1.9090729197585833e-05,
      "loss": 2.7873,
      "step": 6248
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.698923945426941,
      "learning_rate": 1.907094093202731e-05,
      "loss": 2.6642,
      "step": 6252
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5123364925384521,
      "learning_rate": 1.9051152666468786e-05,
      "loss": 2.6528,
      "step": 6256
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5572673082351685,
      "learning_rate": 1.9031364400910262e-05,
      "loss": 2.4895,
      "step": 6260
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.8590190410614014,
      "learning_rate": 1.901157613535174e-05,
      "loss": 2.6208,
      "step": 6264
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.452551245689392,
      "learning_rate": 1.899178786979321e-05,
      "loss": 2.5474,
      "step": 6268
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5120790004730225,
      "learning_rate": 1.8971999604234688e-05,
      "loss": 2.4449,
      "step": 6272
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.7119253873825073,
      "learning_rate": 1.8952211338676164e-05,
      "loss": 2.5996,
      "step": 6276
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4548217058181763,
      "learning_rate": 1.893242307311764e-05,
      "loss": 2.6629,
      "step": 6280
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6181526184082031,
      "learning_rate": 1.8912634807559117e-05,
      "loss": 2.5607,
      "step": 6284
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5881283283233643,
      "learning_rate": 1.8892846542000593e-05,
      "loss": 2.7344,
      "step": 6288
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.5534508228302002,
      "learning_rate": 1.887305827644207e-05,
      "loss": 2.5413,
      "step": 6292
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4985178709030151,
      "learning_rate": 1.8853270010883546e-05,
      "loss": 2.6648,
      "step": 6296
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.698745846748352,
      "learning_rate": 1.8833481745325026e-05,
      "loss": 2.5784,
      "step": 6300
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.42083740234375,
      "learning_rate": 1.8813693479766502e-05,
      "loss": 2.7346,
      "step": 6304
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6639600992202759,
      "learning_rate": 1.8793905214207978e-05,
      "loss": 2.7116,
      "step": 6308
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6525336503982544,
      "learning_rate": 1.877411694864945e-05,
      "loss": 2.696,
      "step": 6312
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.6521825790405273,
      "learning_rate": 1.8754328683090927e-05,
      "loss": 2.4546,
      "step": 6316
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7254414558410645,
      "learning_rate": 1.8734540417532404e-05,
      "loss": 2.6165,
      "step": 6320
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.47000253200531,
      "learning_rate": 1.871475215197388e-05,
      "loss": 2.4961,
      "step": 6324
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6371259689331055,
      "learning_rate": 1.8694963886415356e-05,
      "loss": 2.4793,
      "step": 6328
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.928614854812622,
      "learning_rate": 1.8675175620856833e-05,
      "loss": 2.8014,
      "step": 6332
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5594316720962524,
      "learning_rate": 1.865538735529831e-05,
      "loss": 2.4129,
      "step": 6336
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5634496212005615,
      "learning_rate": 1.8635599089739785e-05,
      "loss": 2.5991,
      "step": 6340
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7114262580871582,
      "learning_rate": 1.861581082418126e-05,
      "loss": 2.6678,
      "step": 6344
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5006263256072998,
      "learning_rate": 1.8596022558622738e-05,
      "loss": 2.6402,
      "step": 6348
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6565409898757935,
      "learning_rate": 1.857623429306421e-05,
      "loss": 2.6237,
      "step": 6352
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.792551875114441,
      "learning_rate": 1.855644602750569e-05,
      "loss": 2.4968,
      "step": 6356
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.714000940322876,
      "learning_rate": 1.8536657761947167e-05,
      "loss": 2.5924,
      "step": 6360
    },
    {
      "epoch": 0.63,
      "grad_norm": 2.0721774101257324,
      "learning_rate": 1.8516869496388643e-05,
      "loss": 2.7513,
      "step": 6364
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5976616144180298,
      "learning_rate": 1.849708123083012e-05,
      "loss": 2.5484,
      "step": 6368
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5501139163970947,
      "learning_rate": 1.8477292965271596e-05,
      "loss": 2.6753,
      "step": 6372
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.590151309967041,
      "learning_rate": 1.8457504699713072e-05,
      "loss": 2.5246,
      "step": 6376
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6754822731018066,
      "learning_rate": 1.843771643415455e-05,
      "loss": 2.6071,
      "step": 6380
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.625549554824829,
      "learning_rate": 1.8417928168596025e-05,
      "loss": 2.7181,
      "step": 6384
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7473890781402588,
      "learning_rate": 1.83981399030375e-05,
      "loss": 2.5226,
      "step": 6388
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6598209142684937,
      "learning_rate": 1.8378351637478978e-05,
      "loss": 2.5456,
      "step": 6392
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5726118087768555,
      "learning_rate": 1.835856337192045e-05,
      "loss": 2.661,
      "step": 6396
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.6404038667678833,
      "learning_rate": 1.8338775106361927e-05,
      "loss": 2.451,
      "step": 6400
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.9309062957763672,
      "learning_rate": 1.8318986840803403e-05,
      "loss": 2.495,
      "step": 6404
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.5471394062042236,
      "learning_rate": 1.829919857524488e-05,
      "loss": 2.6116,
      "step": 6408
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.7516142129898071,
      "learning_rate": 1.8279410309686356e-05,
      "loss": 2.7225,
      "step": 6412
    },
    {
      "epoch": 0.63,
      "grad_norm": 1.534562110900879,
      "learning_rate": 1.8259622044127832e-05,
      "loss": 2.7206,
      "step": 6416
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.9905803203582764,
      "learning_rate": 1.823983377856931e-05,
      "loss": 2.6701,
      "step": 6420
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5553815364837646,
      "learning_rate": 1.8220045513010785e-05,
      "loss": 2.5828,
      "step": 6424
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5264707803726196,
      "learning_rate": 1.820025724745226e-05,
      "loss": 2.5069,
      "step": 6428
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.8345638513565063,
      "learning_rate": 1.8180468981893737e-05,
      "loss": 2.7473,
      "step": 6432
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6059725284576416,
      "learning_rate": 1.8160680716335214e-05,
      "loss": 2.6397,
      "step": 6436
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5249567031860352,
      "learning_rate": 1.814089245077669e-05,
      "loss": 2.6957,
      "step": 6440
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6217684745788574,
      "learning_rate": 1.8121104185218166e-05,
      "loss": 2.506,
      "step": 6444
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7154309749603271,
      "learning_rate": 1.8101315919659643e-05,
      "loss": 2.6848,
      "step": 6448
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.553000807762146,
      "learning_rate": 1.808152765410112e-05,
      "loss": 2.4827,
      "step": 6452
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7927863597869873,
      "learning_rate": 1.8061739388542595e-05,
      "loss": 2.5527,
      "step": 6456
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.4993387460708618,
      "learning_rate": 1.804195112298407e-05,
      "loss": 2.732,
      "step": 6460
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5990402698516846,
      "learning_rate": 1.8022162857425548e-05,
      "loss": 2.6775,
      "step": 6464
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5883631706237793,
      "learning_rate": 1.8002374591867024e-05,
      "loss": 2.5361,
      "step": 6468
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.4568392038345337,
      "learning_rate": 1.79825863263085e-05,
      "loss": 2.7561,
      "step": 6472
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5289112329483032,
      "learning_rate": 1.7962798060749977e-05,
      "loss": 2.4471,
      "step": 6476
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6319639682769775,
      "learning_rate": 1.794300979519145e-05,
      "loss": 2.6101,
      "step": 6480
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5441527366638184,
      "learning_rate": 1.7923221529632926e-05,
      "loss": 2.5532,
      "step": 6484
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.6211721897125244,
      "learning_rate": 1.7903433264074402e-05,
      "loss": 2.6647,
      "step": 6488
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.587514877319336,
      "learning_rate": 1.7883644998515882e-05,
      "loss": 2.6054,
      "step": 6492
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.764523983001709,
      "learning_rate": 1.786385673295736e-05,
      "loss": 2.5837,
      "step": 6496
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.555456519126892,
      "learning_rate": 1.7844068467398835e-05,
      "loss": 2.6482,
      "step": 6500
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5590410232543945,
      "learning_rate": 1.782428020184031e-05,
      "loss": 2.834,
      "step": 6504
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.5642460584640503,
      "learning_rate": 1.7804491936281787e-05,
      "loss": 2.6191,
      "step": 6508
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.7200212478637695,
      "learning_rate": 1.7784703670723264e-05,
      "loss": 2.6683,
      "step": 6512
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.682166576385498,
      "learning_rate": 1.776491540516474e-05,
      "loss": 2.5589,
      "step": 6516
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6353708505630493,
      "learning_rate": 1.7745127139606213e-05,
      "loss": 2.5668,
      "step": 6520
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5286115407943726,
      "learning_rate": 1.772533887404769e-05,
      "loss": 2.4588,
      "step": 6524
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6446589231491089,
      "learning_rate": 1.7705550608489166e-05,
      "loss": 2.6396,
      "step": 6528
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5995491743087769,
      "learning_rate": 1.7685762342930642e-05,
      "loss": 2.4982,
      "step": 6532
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5419460535049438,
      "learning_rate": 1.766597407737212e-05,
      "loss": 2.6596,
      "step": 6536
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5353944301605225,
      "learning_rate": 1.7646185811813595e-05,
      "loss": 2.5443,
      "step": 6540
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.683626651763916,
      "learning_rate": 1.762639754625507e-05,
      "loss": 2.6385,
      "step": 6544
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5726920366287231,
      "learning_rate": 1.7606609280696547e-05,
      "loss": 2.7459,
      "step": 6548
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7215642929077148,
      "learning_rate": 1.7586821015138024e-05,
      "loss": 2.4393,
      "step": 6552
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5964510440826416,
      "learning_rate": 1.75670327495795e-05,
      "loss": 2.7072,
      "step": 6556
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.8928728103637695,
      "learning_rate": 1.7547244484020976e-05,
      "loss": 2.664,
      "step": 6560
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6763553619384766,
      "learning_rate": 1.7527456218462453e-05,
      "loss": 2.4276,
      "step": 6564
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.570327639579773,
      "learning_rate": 1.750766795290393e-05,
      "loss": 2.621,
      "step": 6568
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5460453033447266,
      "learning_rate": 1.7487879687345405e-05,
      "loss": 2.5846,
      "step": 6572
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5720558166503906,
      "learning_rate": 1.746809142178688e-05,
      "loss": 2.6405,
      "step": 6576
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.6330138444900513,
      "learning_rate": 1.7448303156228358e-05,
      "loss": 2.6873,
      "step": 6580
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.7572230100631714,
      "learning_rate": 1.7428514890669834e-05,
      "loss": 2.6039,
      "step": 6584
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.5484024286270142,
      "learning_rate": 1.740872662511131e-05,
      "loss": 2.5413,
      "step": 6588
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.8520413637161255,
      "learning_rate": 1.7388938359552787e-05,
      "loss": 2.6781,
      "step": 6592
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.4541682004928589,
      "learning_rate": 1.7369150093994263e-05,
      "loss": 2.48,
      "step": 6596
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.503838062286377,
      "learning_rate": 1.734936182843574e-05,
      "loss": 2.5952,
      "step": 6600
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.73090660572052,
      "learning_rate": 1.7329573562877216e-05,
      "loss": 2.5526,
      "step": 6604
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.8160631656646729,
      "learning_rate": 1.730978529731869e-05,
      "loss": 2.8937,
      "step": 6608
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.646761178970337,
      "learning_rate": 1.7289997031760165e-05,
      "loss": 2.6505,
      "step": 6612
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.8188313245773315,
      "learning_rate": 1.727020876620164e-05,
      "loss": 2.5133,
      "step": 6616
    },
    {
      "epoch": 0.65,
      "grad_norm": 1.4916778802871704,
      "learning_rate": 1.7250420500643118e-05,
      "loss": 2.7022,
      "step": 6620
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6285133361816406,
      "learning_rate": 1.7230632235084594e-05,
      "loss": 2.8066,
      "step": 6624
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.8589283227920532,
      "learning_rate": 1.7210843969526074e-05,
      "loss": 2.6522,
      "step": 6628
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.4183986186981201,
      "learning_rate": 1.719105570396755e-05,
      "loss": 2.4366,
      "step": 6632
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6243681907653809,
      "learning_rate": 1.7171267438409026e-05,
      "loss": 2.4741,
      "step": 6636
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6193912029266357,
      "learning_rate": 1.7151479172850503e-05,
      "loss": 2.5141,
      "step": 6640
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7768150568008423,
      "learning_rate": 1.713169090729198e-05,
      "loss": 2.5269,
      "step": 6644
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7248855829238892,
      "learning_rate": 1.7111902641733452e-05,
      "loss": 2.6412,
      "step": 6648
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6122647523880005,
      "learning_rate": 1.7092114376174928e-05,
      "loss": 2.6304,
      "step": 6652
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7491798400878906,
      "learning_rate": 1.7072326110616405e-05,
      "loss": 2.7254,
      "step": 6656
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5350676774978638,
      "learning_rate": 1.705253784505788e-05,
      "loss": 2.6792,
      "step": 6660
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5873485803604126,
      "learning_rate": 1.7032749579499357e-05,
      "loss": 2.7011,
      "step": 6664
    },
    {
      "epoch": 0.66,
      "grad_norm": 2.228262186050415,
      "learning_rate": 1.7012961313940834e-05,
      "loss": 2.4938,
      "step": 6668
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5721497535705566,
      "learning_rate": 1.699317304838231e-05,
      "loss": 2.502,
      "step": 6672
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7099807262420654,
      "learning_rate": 1.6973384782823786e-05,
      "loss": 2.5514,
      "step": 6676
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7094590663909912,
      "learning_rate": 1.6953596517265263e-05,
      "loss": 2.5132,
      "step": 6680
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.536875605583191,
      "learning_rate": 1.693380825170674e-05,
      "loss": 2.5599,
      "step": 6684
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6018167734146118,
      "learning_rate": 1.6914019986148215e-05,
      "loss": 2.5827,
      "step": 6688
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.582042932510376,
      "learning_rate": 1.689423172058969e-05,
      "loss": 2.6136,
      "step": 6692
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5590091943740845,
      "learning_rate": 1.6874443455031168e-05,
      "loss": 2.7667,
      "step": 6696
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6942031383514404,
      "learning_rate": 1.6854655189472644e-05,
      "loss": 2.6389,
      "step": 6700
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.9794580936431885,
      "learning_rate": 1.683486692391412e-05,
      "loss": 2.4073,
      "step": 6704
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.7205188274383545,
      "learning_rate": 1.6815078658355597e-05,
      "loss": 2.6566,
      "step": 6708
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6077735424041748,
      "learning_rate": 1.6795290392797073e-05,
      "loss": 2.5308,
      "step": 6712
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.5309977531433105,
      "learning_rate": 1.677550212723855e-05,
      "loss": 2.6203,
      "step": 6716
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.6211974620819092,
      "learning_rate": 1.6755713861680026e-05,
      "loss": 2.6303,
      "step": 6720
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.4833593368530273,
      "learning_rate": 1.6735925596121502e-05,
      "loss": 2.4867,
      "step": 6724
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.793488621711731,
      "learning_rate": 1.671613733056298e-05,
      "loss": 2.5169,
      "step": 6728
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5642069578170776,
      "learning_rate": 1.669634906500445e-05,
      "loss": 2.5192,
      "step": 6732
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.84517240524292,
      "learning_rate": 1.6676560799445928e-05,
      "loss": 2.5577,
      "step": 6736
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5825902223587036,
      "learning_rate": 1.6656772533887404e-05,
      "loss": 2.5124,
      "step": 6740
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.642572045326233,
      "learning_rate": 1.663698426832888e-05,
      "loss": 2.5925,
      "step": 6744
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5614920854568481,
      "learning_rate": 1.6617196002770357e-05,
      "loss": 2.64,
      "step": 6748
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5319268703460693,
      "learning_rate": 1.6597407737211833e-05,
      "loss": 2.3939,
      "step": 6752
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.553107738494873,
      "learning_rate": 1.657761947165331e-05,
      "loss": 2.5153,
      "step": 6756
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5315675735473633,
      "learning_rate": 1.6557831206094786e-05,
      "loss": 2.4994,
      "step": 6760
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5716221332550049,
      "learning_rate": 1.6538042940536265e-05,
      "loss": 2.4707,
      "step": 6764
    },
    {
      "epoch": 0.67,
      "grad_norm": 2.1599626541137695,
      "learning_rate": 1.651825467497774e-05,
      "loss": 2.4953,
      "step": 6768
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.762488603591919,
      "learning_rate": 1.6498466409419218e-05,
      "loss": 2.5602,
      "step": 6772
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.584439754486084,
      "learning_rate": 1.647867814386069e-05,
      "loss": 2.5398,
      "step": 6776
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.592529535293579,
      "learning_rate": 1.6458889878302167e-05,
      "loss": 2.504,
      "step": 6780
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6364370584487915,
      "learning_rate": 1.6439101612743643e-05,
      "loss": 2.5175,
      "step": 6784
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.4882373809814453,
      "learning_rate": 1.641931334718512e-05,
      "loss": 2.506,
      "step": 6788
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6269882917404175,
      "learning_rate": 1.6399525081626596e-05,
      "loss": 2.5361,
      "step": 6792
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6206145286560059,
      "learning_rate": 1.6379736816068072e-05,
      "loss": 2.5575,
      "step": 6796
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5736054182052612,
      "learning_rate": 1.635994855050955e-05,
      "loss": 2.6325,
      "step": 6800
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.6561203002929688,
      "learning_rate": 1.6340160284951025e-05,
      "loss": 2.4781,
      "step": 6804
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.612779974937439,
      "learning_rate": 1.63203720193925e-05,
      "loss": 2.6126,
      "step": 6808
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7301238775253296,
      "learning_rate": 1.6300583753833978e-05,
      "loss": 2.6143,
      "step": 6812
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.7517168521881104,
      "learning_rate": 1.628079548827545e-05,
      "loss": 2.6524,
      "step": 6816
    },
    {
      "epoch": 0.67,
      "grad_norm": 1.5563359260559082,
      "learning_rate": 1.6261007222716927e-05,
      "loss": 2.5678,
      "step": 6820
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7279319763183594,
      "learning_rate": 1.6241218957158407e-05,
      "loss": 2.6764,
      "step": 6824
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6466785669326782,
      "learning_rate": 1.6221430691599883e-05,
      "loss": 2.4621,
      "step": 6828
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.578447937965393,
      "learning_rate": 1.620164242604136e-05,
      "loss": 2.6362,
      "step": 6832
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7554636001586914,
      "learning_rate": 1.6181854160482836e-05,
      "loss": 2.6348,
      "step": 6836
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8078508377075195,
      "learning_rate": 1.6162065894924312e-05,
      "loss": 2.4432,
      "step": 6840
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6505666971206665,
      "learning_rate": 1.6142277629365788e-05,
      "loss": 2.5918,
      "step": 6844
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6466803550720215,
      "learning_rate": 1.6122489363807265e-05,
      "loss": 2.7365,
      "step": 6848
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.682653546333313,
      "learning_rate": 1.610270109824874e-05,
      "loss": 2.6347,
      "step": 6852
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5635805130004883,
      "learning_rate": 1.6082912832690217e-05,
      "loss": 2.4988,
      "step": 6856
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6750497817993164,
      "learning_rate": 1.606312456713169e-05,
      "loss": 2.4018,
      "step": 6860
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5516955852508545,
      "learning_rate": 1.6043336301573167e-05,
      "loss": 2.5583,
      "step": 6864
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7607613801956177,
      "learning_rate": 1.6023548036014643e-05,
      "loss": 2.4629,
      "step": 6868
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6372771263122559,
      "learning_rate": 1.600375977045612e-05,
      "loss": 2.7317,
      "step": 6872
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5994541645050049,
      "learning_rate": 1.5983971504897595e-05,
      "loss": 2.4516,
      "step": 6876
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6625773906707764,
      "learning_rate": 1.5964183239339072e-05,
      "loss": 2.5038,
      "step": 6880
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.8105216026306152,
      "learning_rate": 1.5944394973780548e-05,
      "loss": 2.4708,
      "step": 6884
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7151480913162231,
      "learning_rate": 1.5924606708222024e-05,
      "loss": 2.4271,
      "step": 6888
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6709083318710327,
      "learning_rate": 1.59048184426635e-05,
      "loss": 2.7086,
      "step": 6892
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6832211017608643,
      "learning_rate": 1.5885030177104977e-05,
      "loss": 2.6396,
      "step": 6896
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6037044525146484,
      "learning_rate": 1.5865241911546453e-05,
      "loss": 2.3786,
      "step": 6900
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.6449296474456787,
      "learning_rate": 1.584545364598793e-05,
      "loss": 2.6229,
      "step": 6904
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7927825450897217,
      "learning_rate": 1.5825665380429406e-05,
      "loss": 2.5644,
      "step": 6908
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5636272430419922,
      "learning_rate": 1.5805877114870882e-05,
      "loss": 2.5509,
      "step": 6912
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.5654444694519043,
      "learning_rate": 1.578608884931236e-05,
      "loss": 2.4836,
      "step": 6916
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.7937673330307007,
      "learning_rate": 1.5766300583753835e-05,
      "loss": 2.669,
      "step": 6920
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6205195188522339,
      "learning_rate": 1.574651231819531e-05,
      "loss": 2.7591,
      "step": 6924
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5670419931411743,
      "learning_rate": 1.5726724052636788e-05,
      "loss": 2.7567,
      "step": 6928
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5903569459915161,
      "learning_rate": 1.5706935787078264e-05,
      "loss": 2.505,
      "step": 6932
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7012027502059937,
      "learning_rate": 1.568714752151974e-05,
      "loss": 2.4375,
      "step": 6936
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5950981378555298,
      "learning_rate": 1.5667359255961217e-05,
      "loss": 2.6312,
      "step": 6940
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.647019624710083,
      "learning_rate": 1.564757099040269e-05,
      "loss": 2.547,
      "step": 6944
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6310683488845825,
      "learning_rate": 1.5627782724844166e-05,
      "loss": 2.5012,
      "step": 6948
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5516809225082397,
      "learning_rate": 1.5607994459285642e-05,
      "loss": 2.6731,
      "step": 6952
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5987101793289185,
      "learning_rate": 1.558820619372712e-05,
      "loss": 2.4518,
      "step": 6956
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5934362411499023,
      "learning_rate": 1.5568417928168598e-05,
      "loss": 2.4448,
      "step": 6960
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6797455549240112,
      "learning_rate": 1.5548629662610075e-05,
      "loss": 2.6067,
      "step": 6964
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.9424878358840942,
      "learning_rate": 1.552884139705155e-05,
      "loss": 2.7723,
      "step": 6968
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.659338355064392,
      "learning_rate": 1.5509053131493027e-05,
      "loss": 2.5958,
      "step": 6972
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5043983459472656,
      "learning_rate": 1.5489264865934504e-05,
      "loss": 2.5618,
      "step": 6976
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7023625373840332,
      "learning_rate": 1.546947660037598e-05,
      "loss": 2.5923,
      "step": 6980
    },
    {
      "epoch": 0.69,
      "grad_norm": 2.293471097946167,
      "learning_rate": 1.5449688334817456e-05,
      "loss": 2.7371,
      "step": 6984
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6672747135162354,
      "learning_rate": 1.542990006925893e-05,
      "loss": 2.5335,
      "step": 6988
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.706365704536438,
      "learning_rate": 1.5410111803700405e-05,
      "loss": 2.4963,
      "step": 6992
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.7993139028549194,
      "learning_rate": 1.5390323538141882e-05,
      "loss": 2.5698,
      "step": 6996
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6543632745742798,
      "learning_rate": 1.5370535272583358e-05,
      "loss": 2.5999,
      "step": 7000
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.583425760269165,
      "learning_rate": 1.5350747007024834e-05,
      "loss": 2.585,
      "step": 7004
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6158406734466553,
      "learning_rate": 1.533095874146631e-05,
      "loss": 2.6547,
      "step": 7008
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.6902023553848267,
      "learning_rate": 1.5311170475907787e-05,
      "loss": 2.5624,
      "step": 7012
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.5671786069869995,
      "learning_rate": 1.5291382210349263e-05,
      "loss": 2.5003,
      "step": 7016
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.4708104133605957,
      "learning_rate": 1.527159394479074e-05,
      "loss": 2.4721,
      "step": 7020
    },
    {
      "epoch": 0.69,
      "grad_norm": 1.622897744178772,
      "learning_rate": 1.5251805679232218e-05,
      "loss": 2.6162,
      "step": 7024
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5571931600570679,
      "learning_rate": 1.523201741367369e-05,
      "loss": 2.5378,
      "step": 7028
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.4922298192977905,
      "learning_rate": 1.5212229148115167e-05,
      "loss": 2.5156,
      "step": 7032
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.0152604579925537,
      "learning_rate": 1.5192440882556643e-05,
      "loss": 2.4795,
      "step": 7036
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5429670810699463,
      "learning_rate": 1.517265261699812e-05,
      "loss": 2.5534,
      "step": 7040
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5789587497711182,
      "learning_rate": 1.5152864351439598e-05,
      "loss": 2.647,
      "step": 7044
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6377935409545898,
      "learning_rate": 1.5133076085881074e-05,
      "loss": 2.7136,
      "step": 7048
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6484628915786743,
      "learning_rate": 1.511328782032255e-05,
      "loss": 2.5635,
      "step": 7052
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7103774547576904,
      "learning_rate": 1.5093499554764027e-05,
      "loss": 2.767,
      "step": 7056
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6443290710449219,
      "learning_rate": 1.5073711289205503e-05,
      "loss": 2.5986,
      "step": 7060
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6503266096115112,
      "learning_rate": 1.505392302364698e-05,
      "loss": 2.4505,
      "step": 7064
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.656258225440979,
      "learning_rate": 1.5034134758088456e-05,
      "loss": 2.5093,
      "step": 7068
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5531201362609863,
      "learning_rate": 1.501434649252993e-05,
      "loss": 2.5512,
      "step": 7072
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5085419416427612,
      "learning_rate": 1.4994558226971406e-05,
      "loss": 2.4362,
      "step": 7076
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5943050384521484,
      "learning_rate": 1.4974769961412883e-05,
      "loss": 2.5309,
      "step": 7080
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7004754543304443,
      "learning_rate": 1.4954981695854359e-05,
      "loss": 2.573,
      "step": 7084
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6682156324386597,
      "learning_rate": 1.4935193430295835e-05,
      "loss": 2.5718,
      "step": 7088
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.489130973815918,
      "learning_rate": 1.4915405164737312e-05,
      "loss": 2.7728,
      "step": 7092
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.72616708278656,
      "learning_rate": 1.4895616899178788e-05,
      "loss": 2.6457,
      "step": 7096
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5968120098114014,
      "learning_rate": 1.4875828633620264e-05,
      "loss": 2.4886,
      "step": 7100
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.7065930366516113,
      "learning_rate": 1.485604036806174e-05,
      "loss": 2.6894,
      "step": 7104
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.8437420129776,
      "learning_rate": 1.4836252102503217e-05,
      "loss": 2.6103,
      "step": 7108
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5512233972549438,
      "learning_rate": 1.4816463836944692e-05,
      "loss": 2.4657,
      "step": 7112
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.6278142929077148,
      "learning_rate": 1.4796675571386168e-05,
      "loss": 2.7127,
      "step": 7116
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5600184202194214,
      "learning_rate": 1.4776887305827644e-05,
      "loss": 2.4707,
      "step": 7120
    },
    {
      "epoch": 0.7,
      "grad_norm": 1.5605080127716064,
      "learning_rate": 1.475709904026912e-05,
      "loss": 2.5009,
      "step": 7124
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6376603841781616,
      "learning_rate": 1.4737310774710597e-05,
      "loss": 2.4116,
      "step": 7128
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5430164337158203,
      "learning_rate": 1.4717522509152073e-05,
      "loss": 2.4315,
      "step": 7132
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.9249303340911865,
      "learning_rate": 1.469773424359355e-05,
      "loss": 2.5128,
      "step": 7136
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.914553165435791,
      "learning_rate": 1.4677945978035026e-05,
      "loss": 2.5726,
      "step": 7140
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6890403032302856,
      "learning_rate": 1.4658157712476502e-05,
      "loss": 2.5233,
      "step": 7144
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6215218305587769,
      "learning_rate": 1.463836944691798e-05,
      "loss": 2.4541,
      "step": 7148
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6023436784744263,
      "learning_rate": 1.4618581181359457e-05,
      "loss": 2.5825,
      "step": 7152
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6800665855407715,
      "learning_rate": 1.459879291580093e-05,
      "loss": 2.5275,
      "step": 7156
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5266000032424927,
      "learning_rate": 1.4579004650242406e-05,
      "loss": 2.6414,
      "step": 7160
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.660843014717102,
      "learning_rate": 1.4559216384683882e-05,
      "loss": 2.4934,
      "step": 7164
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5643970966339111,
      "learning_rate": 1.4539428119125358e-05,
      "loss": 2.4847,
      "step": 7168
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.4681652784347534,
      "learning_rate": 1.4519639853566835e-05,
      "loss": 2.423,
      "step": 7172
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5900237560272217,
      "learning_rate": 1.4499851588008311e-05,
      "loss": 2.5645,
      "step": 7176
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6541297435760498,
      "learning_rate": 1.4480063322449789e-05,
      "loss": 2.5434,
      "step": 7180
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5105384588241577,
      "learning_rate": 1.4460275056891265e-05,
      "loss": 2.6391,
      "step": 7184
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6248685121536255,
      "learning_rate": 1.4440486791332742e-05,
      "loss": 2.6772,
      "step": 7188
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.559591293334961,
      "learning_rate": 1.4420698525774218e-05,
      "loss": 2.5175,
      "step": 7192
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.73567533493042,
      "learning_rate": 1.4400910260215691e-05,
      "loss": 2.5799,
      "step": 7196
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.602149248123169,
      "learning_rate": 1.4381121994657167e-05,
      "loss": 2.5475,
      "step": 7200
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.6269313097000122,
      "learning_rate": 1.4361333729098644e-05,
      "loss": 2.5729,
      "step": 7204
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.5919511318206787,
      "learning_rate": 1.4341545463540122e-05,
      "loss": 2.4055,
      "step": 7208
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.9625754356384277,
      "learning_rate": 1.4321757197981598e-05,
      "loss": 2.6867,
      "step": 7212
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7777873277664185,
      "learning_rate": 1.4301968932423074e-05,
      "loss": 2.7518,
      "step": 7216
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.632881760597229,
      "learning_rate": 1.428218066686455e-05,
      "loss": 2.3954,
      "step": 7220
    },
    {
      "epoch": 0.71,
      "grad_norm": 1.7910826206207275,
      "learning_rate": 1.4262392401306027e-05,
      "loss": 2.5477,
      "step": 7224
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6172887086868286,
      "learning_rate": 1.4242604135747503e-05,
      "loss": 2.6665,
      "step": 7228
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7775293588638306,
      "learning_rate": 1.422281587018898e-05,
      "loss": 2.5237,
      "step": 7232
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6558274030685425,
      "learning_rate": 1.4203027604630456e-05,
      "loss": 2.5909,
      "step": 7236
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6271092891693115,
      "learning_rate": 1.418323933907193e-05,
      "loss": 2.5299,
      "step": 7240
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5768835544586182,
      "learning_rate": 1.4163451073513407e-05,
      "loss": 2.5376,
      "step": 7244
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.641741156578064,
      "learning_rate": 1.4143662807954883e-05,
      "loss": 2.5802,
      "step": 7248
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5866329669952393,
      "learning_rate": 1.412387454239636e-05,
      "loss": 2.4445,
      "step": 7252
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6858747005462646,
      "learning_rate": 1.4104086276837836e-05,
      "loss": 2.51,
      "step": 7256
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6350436210632324,
      "learning_rate": 1.4084298011279312e-05,
      "loss": 2.6297,
      "step": 7260
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7628724575042725,
      "learning_rate": 1.4064509745720788e-05,
      "loss": 2.6294,
      "step": 7264
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6731958389282227,
      "learning_rate": 1.4044721480162265e-05,
      "loss": 2.5577,
      "step": 7268
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7393099069595337,
      "learning_rate": 1.4024933214603741e-05,
      "loss": 2.6079,
      "step": 7272
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7639967203140259,
      "learning_rate": 1.4005144949045217e-05,
      "loss": 2.7389,
      "step": 7276
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5961958169937134,
      "learning_rate": 1.3985356683486692e-05,
      "loss": 2.5846,
      "step": 7280
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6526871919631958,
      "learning_rate": 1.3965568417928168e-05,
      "loss": 2.5965,
      "step": 7284
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.7771737575531006,
      "learning_rate": 1.3945780152369645e-05,
      "loss": 2.5905,
      "step": 7288
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.713395118713379,
      "learning_rate": 1.3925991886811121e-05,
      "loss": 2.5063,
      "step": 7292
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5294148921966553,
      "learning_rate": 1.3906203621252597e-05,
      "loss": 2.4081,
      "step": 7296
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6980494260787964,
      "learning_rate": 1.3886415355694074e-05,
      "loss": 2.4693,
      "step": 7300
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.703873634338379,
      "learning_rate": 1.386662709013555e-05,
      "loss": 2.4529,
      "step": 7304
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5093457698822021,
      "learning_rate": 1.3846838824577026e-05,
      "loss": 2.439,
      "step": 7308
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.6407102346420288,
      "learning_rate": 1.3827050559018503e-05,
      "loss": 2.4581,
      "step": 7312
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.485217809677124,
      "learning_rate": 1.380726229345998e-05,
      "loss": 2.5162,
      "step": 7316
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5687867403030396,
      "learning_rate": 1.3787474027901457e-05,
      "loss": 2.4868,
      "step": 7320
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5972037315368652,
      "learning_rate": 1.376768576234293e-05,
      "loss": 2.5713,
      "step": 7324
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6357609033584595,
      "learning_rate": 1.3747897496784406e-05,
      "loss": 2.5051,
      "step": 7328
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7274508476257324,
      "learning_rate": 1.3728109231225883e-05,
      "loss": 2.596,
      "step": 7332
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.9209715127944946,
      "learning_rate": 1.3708320965667359e-05,
      "loss": 2.7628,
      "step": 7336
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.5816470384597778,
      "learning_rate": 1.3688532700108835e-05,
      "loss": 2.4371,
      "step": 7340
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6190577745437622,
      "learning_rate": 1.3668744434550313e-05,
      "loss": 2.4256,
      "step": 7344
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.627508282661438,
      "learning_rate": 1.364895616899179e-05,
      "loss": 2.6653,
      "step": 7348
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6422009468078613,
      "learning_rate": 1.3629167903433266e-05,
      "loss": 2.6699,
      "step": 7352
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.546067476272583,
      "learning_rate": 1.3609379637874742e-05,
      "loss": 2.4263,
      "step": 7356
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.621297836303711,
      "learning_rate": 1.3589591372316218e-05,
      "loss": 2.4296,
      "step": 7360
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.643576741218567,
      "learning_rate": 1.3569803106757691e-05,
      "loss": 2.5806,
      "step": 7364
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6136977672576904,
      "learning_rate": 1.3550014841199168e-05,
      "loss": 2.4887,
      "step": 7368
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.704002857208252,
      "learning_rate": 1.3530226575640644e-05,
      "loss": 2.583,
      "step": 7372
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6694648265838623,
      "learning_rate": 1.3510438310082122e-05,
      "loss": 2.4127,
      "step": 7376
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6585307121276855,
      "learning_rate": 1.3490650044523598e-05,
      "loss": 2.4128,
      "step": 7380
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7365046739578247,
      "learning_rate": 1.3470861778965075e-05,
      "loss": 2.6269,
      "step": 7384
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6612383127212524,
      "learning_rate": 1.3451073513406551e-05,
      "loss": 2.5896,
      "step": 7388
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7151292562484741,
      "learning_rate": 1.3431285247848027e-05,
      "loss": 2.4705,
      "step": 7392
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.608148455619812,
      "learning_rate": 1.3411496982289504e-05,
      "loss": 2.4956,
      "step": 7396
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.9409472942352295,
      "learning_rate": 1.339170871673098e-05,
      "loss": 2.5163,
      "step": 7400
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6790823936462402,
      "learning_rate": 1.3371920451172456e-05,
      "loss": 2.5319,
      "step": 7404
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.833620548248291,
      "learning_rate": 1.3352132185613931e-05,
      "loss": 2.5467,
      "step": 7408
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7631961107254028,
      "learning_rate": 1.3332343920055407e-05,
      "loss": 2.5748,
      "step": 7412
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.5768530368804932,
      "learning_rate": 1.3312555654496884e-05,
      "loss": 2.4908,
      "step": 7416
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6658759117126465,
      "learning_rate": 1.329276738893836e-05,
      "loss": 2.5698,
      "step": 7420
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.7477085590362549,
      "learning_rate": 1.3272979123379836e-05,
      "loss": 2.3864,
      "step": 7424
    },
    {
      "epoch": 0.73,
      "grad_norm": 1.6787872314453125,
      "learning_rate": 1.3253190857821313e-05,
      "loss": 2.5065,
      "step": 7428
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6128181219100952,
      "learning_rate": 1.3233402592262789e-05,
      "loss": 2.6635,
      "step": 7432
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.4591010808944702,
      "learning_rate": 1.3213614326704265e-05,
      "loss": 2.4067,
      "step": 7436
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7754063606262207,
      "learning_rate": 1.3193826061145742e-05,
      "loss": 2.5308,
      "step": 7440
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.708154559135437,
      "learning_rate": 1.3174037795587218e-05,
      "loss": 2.4604,
      "step": 7444
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.539201021194458,
      "learning_rate": 1.3154249530028694e-05,
      "loss": 2.5372,
      "step": 7448
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8630115985870361,
      "learning_rate": 1.3134461264470169e-05,
      "loss": 2.5022,
      "step": 7452
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7771594524383545,
      "learning_rate": 1.3114672998911645e-05,
      "loss": 2.3826,
      "step": 7456
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8175123929977417,
      "learning_rate": 1.3094884733353121e-05,
      "loss": 2.5922,
      "step": 7460
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.4767266511917114,
      "learning_rate": 1.3075096467794598e-05,
      "loss": 2.5334,
      "step": 7464
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7710050344467163,
      "learning_rate": 1.3055308202236074e-05,
      "loss": 2.5901,
      "step": 7468
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.78278648853302,
      "learning_rate": 1.303551993667755e-05,
      "loss": 2.5621,
      "step": 7472
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.553341269493103,
      "learning_rate": 1.3015731671119027e-05,
      "loss": 2.5296,
      "step": 7476
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.4547873735427856,
      "learning_rate": 1.2995943405560505e-05,
      "loss": 2.5812,
      "step": 7480
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6034144163131714,
      "learning_rate": 1.2976155140001981e-05,
      "loss": 2.3507,
      "step": 7484
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6461652517318726,
      "learning_rate": 1.2956366874443457e-05,
      "loss": 2.6291,
      "step": 7488
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6804616451263428,
      "learning_rate": 1.293657860888493e-05,
      "loss": 2.4514,
      "step": 7492
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.5887171030044556,
      "learning_rate": 1.2916790343326407e-05,
      "loss": 2.386,
      "step": 7496
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.6213395595550537,
      "learning_rate": 1.2897002077767883e-05,
      "loss": 2.5569,
      "step": 7500
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7485177516937256,
      "learning_rate": 1.287721381220936e-05,
      "loss": 2.5454,
      "step": 7504
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.660869836807251,
      "learning_rate": 1.2857425546650836e-05,
      "loss": 2.5177,
      "step": 7508
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.674647331237793,
      "learning_rate": 1.2837637281092314e-05,
      "loss": 2.3855,
      "step": 7512
    },
    {
      "epoch": 0.74,
      "grad_norm": 2.053084373474121,
      "learning_rate": 1.281784901553379e-05,
      "loss": 2.7179,
      "step": 7516
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.5778189897537231,
      "learning_rate": 1.2798060749975266e-05,
      "loss": 2.5341,
      "step": 7520
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.670196294784546,
      "learning_rate": 1.2778272484416743e-05,
      "loss": 2.4186,
      "step": 7524
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.7191470861434937,
      "learning_rate": 1.2758484218858219e-05,
      "loss": 2.4801,
      "step": 7528
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6857608556747437,
      "learning_rate": 1.2738695953299695e-05,
      "loss": 2.4867,
      "step": 7532
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6951642036437988,
      "learning_rate": 1.2718907687741168e-05,
      "loss": 2.5551,
      "step": 7536
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.9836032390594482,
      "learning_rate": 1.2699119422182646e-05,
      "loss": 2.4202,
      "step": 7540
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6894361972808838,
      "learning_rate": 1.2679331156624122e-05,
      "loss": 2.5553,
      "step": 7544
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6373331546783447,
      "learning_rate": 1.2659542891065599e-05,
      "loss": 2.3549,
      "step": 7548
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5669854879379272,
      "learning_rate": 1.2639754625507075e-05,
      "loss": 2.4814,
      "step": 7552
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.697510838508606,
      "learning_rate": 1.2619966359948551e-05,
      "loss": 2.5292,
      "step": 7556
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6051623821258545,
      "learning_rate": 1.2600178094390028e-05,
      "loss": 2.4635,
      "step": 7560
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.550139307975769,
      "learning_rate": 1.2580389828831504e-05,
      "loss": 2.5052,
      "step": 7564
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8294285535812378,
      "learning_rate": 1.256060156327298e-05,
      "loss": 2.6251,
      "step": 7568
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.731762170791626,
      "learning_rate": 1.2540813297714457e-05,
      "loss": 2.6812,
      "step": 7572
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7276403903961182,
      "learning_rate": 1.2521025032155931e-05,
      "loss": 2.5587,
      "step": 7576
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8602039813995361,
      "learning_rate": 1.2501236766597408e-05,
      "loss": 2.5649,
      "step": 7580
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5579040050506592,
      "learning_rate": 1.2481448501038886e-05,
      "loss": 2.6513,
      "step": 7584
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.7131482362747192,
      "learning_rate": 1.246166023548036e-05,
      "loss": 2.4914,
      "step": 7588
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.68119215965271,
      "learning_rate": 1.2441871969921837e-05,
      "loss": 2.5319,
      "step": 7592
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.8896154165267944,
      "learning_rate": 1.2422083704363313e-05,
      "loss": 2.5683,
      "step": 7596
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5681391954421997,
      "learning_rate": 1.240229543880479e-05,
      "loss": 2.4047,
      "step": 7600
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.4728906154632568,
      "learning_rate": 1.2382507173246266e-05,
      "loss": 2.5071,
      "step": 7604
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.5875272750854492,
      "learning_rate": 1.2362718907687742e-05,
      "loss": 2.5415,
      "step": 7608
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6239516735076904,
      "learning_rate": 1.2342930642129218e-05,
      "loss": 2.4362,
      "step": 7612
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.812610149383545,
      "learning_rate": 1.2323142376570695e-05,
      "loss": 2.5034,
      "step": 7616
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6190868616104126,
      "learning_rate": 1.2303354111012171e-05,
      "loss": 2.4786,
      "step": 7620
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.6609690189361572,
      "learning_rate": 1.2283565845453647e-05,
      "loss": 2.372,
      "step": 7624
    },
    {
      "epoch": 0.75,
      "grad_norm": 1.700492024421692,
      "learning_rate": 1.2263777579895122e-05,
      "loss": 2.4527,
      "step": 7628
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8237019777297974,
      "learning_rate": 1.2243989314336598e-05,
      "loss": 2.548,
      "step": 7632
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.558048129081726,
      "learning_rate": 1.2224201048778074e-05,
      "loss": 2.5439,
      "step": 7636
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7001163959503174,
      "learning_rate": 1.220441278321955e-05,
      "loss": 2.5146,
      "step": 7640
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5829248428344727,
      "learning_rate": 1.2184624517661027e-05,
      "loss": 2.5473,
      "step": 7644
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5922554731369019,
      "learning_rate": 1.2164836252102505e-05,
      "loss": 2.5482,
      "step": 7648
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5552897453308105,
      "learning_rate": 1.214504798654398e-05,
      "loss": 2.7146,
      "step": 7652
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6641972064971924,
      "learning_rate": 1.2125259720985456e-05,
      "loss": 2.5317,
      "step": 7656
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.727538824081421,
      "learning_rate": 1.2105471455426932e-05,
      "loss": 2.511,
      "step": 7660
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7466334104537964,
      "learning_rate": 1.2085683189868409e-05,
      "loss": 2.6592,
      "step": 7664
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.578805923461914,
      "learning_rate": 1.2065894924309885e-05,
      "loss": 2.5405,
      "step": 7668
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7704498767852783,
      "learning_rate": 1.204610665875136e-05,
      "loss": 2.7104,
      "step": 7672
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.901253342628479,
      "learning_rate": 1.2026318393192838e-05,
      "loss": 2.68,
      "step": 7676
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6251075267791748,
      "learning_rate": 1.2006530127634314e-05,
      "loss": 2.5327,
      "step": 7680
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.9433711767196655,
      "learning_rate": 1.198674186207579e-05,
      "loss": 2.664,
      "step": 7684
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6232560873031616,
      "learning_rate": 1.1966953596517267e-05,
      "loss": 2.3881,
      "step": 7688
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.625562310218811,
      "learning_rate": 1.1947165330958741e-05,
      "loss": 2.3981,
      "step": 7692
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8203229904174805,
      "learning_rate": 1.1927377065400218e-05,
      "loss": 2.5587,
      "step": 7696
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.5125036239624023,
      "learning_rate": 1.1907588799841694e-05,
      "loss": 2.548,
      "step": 7700
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7927870750427246,
      "learning_rate": 1.188780053428317e-05,
      "loss": 2.5317,
      "step": 7704
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.853130578994751,
      "learning_rate": 1.1868012268724647e-05,
      "loss": 2.5032,
      "step": 7708
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.8347195386886597,
      "learning_rate": 1.1848224003166123e-05,
      "loss": 2.3406,
      "step": 7712
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6292403936386108,
      "learning_rate": 1.18284357376076e-05,
      "loss": 2.4874,
      "step": 7716
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.6553095579147339,
      "learning_rate": 1.1808647472049076e-05,
      "loss": 2.5065,
      "step": 7720
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.9831068515777588,
      "learning_rate": 1.1788859206490552e-05,
      "loss": 2.5079,
      "step": 7724
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7396299839019775,
      "learning_rate": 1.1769070940932028e-05,
      "loss": 2.5171,
      "step": 7728
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8842509984970093,
      "learning_rate": 1.1749282675373505e-05,
      "loss": 2.6561,
      "step": 7732
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5185445547103882,
      "learning_rate": 1.1729494409814979e-05,
      "loss": 2.6638,
      "step": 7736
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7892464399337769,
      "learning_rate": 1.1709706144256455e-05,
      "loss": 2.4509,
      "step": 7740
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6770362854003906,
      "learning_rate": 1.1689917878697933e-05,
      "loss": 2.536,
      "step": 7744
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6587915420532227,
      "learning_rate": 1.167012961313941e-05,
      "loss": 2.515,
      "step": 7748
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7976752519607544,
      "learning_rate": 1.1650341347580886e-05,
      "loss": 2.4524,
      "step": 7752
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7145442962646484,
      "learning_rate": 1.163055308202236e-05,
      "loss": 2.56,
      "step": 7756
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6737264394760132,
      "learning_rate": 1.1610764816463837e-05,
      "loss": 2.45,
      "step": 7760
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.683347225189209,
      "learning_rate": 1.1590976550905313e-05,
      "loss": 2.4508,
      "step": 7764
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6320801973342896,
      "learning_rate": 1.157118828534679e-05,
      "loss": 2.6438,
      "step": 7768
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8501825332641602,
      "learning_rate": 1.1551400019788266e-05,
      "loss": 2.4262,
      "step": 7772
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8582115173339844,
      "learning_rate": 1.1531611754229742e-05,
      "loss": 2.4314,
      "step": 7776
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5969737768173218,
      "learning_rate": 1.1511823488671219e-05,
      "loss": 2.4917,
      "step": 7780
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5447652339935303,
      "learning_rate": 1.1492035223112695e-05,
      "loss": 2.5134,
      "step": 7784
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7579408884048462,
      "learning_rate": 1.1472246957554171e-05,
      "loss": 2.4526,
      "step": 7788
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6958955526351929,
      "learning_rate": 1.1452458691995648e-05,
      "loss": 2.5719,
      "step": 7792
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6478573083877563,
      "learning_rate": 1.1432670426437124e-05,
      "loss": 2.5227,
      "step": 7796
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5438200235366821,
      "learning_rate": 1.1412882160878599e-05,
      "loss": 2.3835,
      "step": 7800
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.5718754529953003,
      "learning_rate": 1.1393093895320075e-05,
      "loss": 2.4568,
      "step": 7804
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.6531394720077515,
      "learning_rate": 1.1373305629761551e-05,
      "loss": 2.3707,
      "step": 7808
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8568472862243652,
      "learning_rate": 1.135351736420303e-05,
      "loss": 2.583,
      "step": 7812
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7812650203704834,
      "learning_rate": 1.1333729098644506e-05,
      "loss": 2.5865,
      "step": 7816
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.7314536571502686,
      "learning_rate": 1.131394083308598e-05,
      "loss": 2.5384,
      "step": 7820
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.8775840997695923,
      "learning_rate": 1.1294152567527457e-05,
      "loss": 2.6276,
      "step": 7824
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.3948659896850586,
      "learning_rate": 1.1274364301968933e-05,
      "loss": 2.5302,
      "step": 7828
    },
    {
      "epoch": 0.77,
      "grad_norm": 1.690691590309143,
      "learning_rate": 1.125457603641041e-05,
      "loss": 2.5569,
      "step": 7832
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.606754183769226,
      "learning_rate": 1.1234787770851885e-05,
      "loss": 2.3866,
      "step": 7836
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5848886966705322,
      "learning_rate": 1.1214999505293362e-05,
      "loss": 2.5407,
      "step": 7840
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5651729106903076,
      "learning_rate": 1.1195211239734838e-05,
      "loss": 2.2891,
      "step": 7844
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.704177975654602,
      "learning_rate": 1.1175422974176314e-05,
      "loss": 2.6963,
      "step": 7848
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.605334997177124,
      "learning_rate": 1.115563470861779e-05,
      "loss": 2.5576,
      "step": 7852
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.530823826789856,
      "learning_rate": 1.1135846443059267e-05,
      "loss": 2.4859,
      "step": 7856
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6835219860076904,
      "learning_rate": 1.1116058177500742e-05,
      "loss": 2.462,
      "step": 7860
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6420457363128662,
      "learning_rate": 1.1096269911942218e-05,
      "loss": 2.6175,
      "step": 7864
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6330996751785278,
      "learning_rate": 1.1076481646383694e-05,
      "loss": 2.6308,
      "step": 7868
    },
    {
      "epoch": 0.78,
      "grad_norm": 2.031244993209839,
      "learning_rate": 1.105669338082517e-05,
      "loss": 2.4599,
      "step": 7872
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5154794454574585,
      "learning_rate": 1.1036905115266647e-05,
      "loss": 2.4949,
      "step": 7876
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7397664785385132,
      "learning_rate": 1.1017116849708125e-05,
      "loss": 2.3962,
      "step": 7880
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.7765430212020874,
      "learning_rate": 1.09973285841496e-05,
      "loss": 2.4214,
      "step": 7884
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6177083253860474,
      "learning_rate": 1.0977540318591076e-05,
      "loss": 2.5221,
      "step": 7888
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6060839891433716,
      "learning_rate": 1.0957752053032552e-05,
      "loss": 2.6571,
      "step": 7892
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6707481145858765,
      "learning_rate": 1.0937963787474029e-05,
      "loss": 2.373,
      "step": 7896
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.8542253971099854,
      "learning_rate": 1.0918175521915505e-05,
      "loss": 2.4029,
      "step": 7900
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.6408201456069946,
      "learning_rate": 1.089838725635698e-05,
      "loss": 2.4065,
      "step": 7904
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5829615592956543,
      "learning_rate": 1.0878598990798458e-05,
      "loss": 2.6859,
      "step": 7908
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5252026319503784,
      "learning_rate": 1.0858810725239934e-05,
      "loss": 2.5326,
      "step": 7912
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5747402906417847,
      "learning_rate": 1.083902245968141e-05,
      "loss": 2.5311,
      "step": 7916
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.592838168144226,
      "learning_rate": 1.0819234194122887e-05,
      "loss": 2.3582,
      "step": 7920
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5652263164520264,
      "learning_rate": 1.0799445928564361e-05,
      "loss": 2.5613,
      "step": 7924
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.5245355367660522,
      "learning_rate": 1.0779657663005837e-05,
      "loss": 2.7002,
      "step": 7928
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.724204421043396,
      "learning_rate": 1.0759869397447314e-05,
      "loss": 2.9039,
      "step": 7932
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7106647491455078,
      "learning_rate": 1.074008113188879e-05,
      "loss": 2.5086,
      "step": 7936
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5160934925079346,
      "learning_rate": 1.0720292866330266e-05,
      "loss": 2.447,
      "step": 7940
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.833022952079773,
      "learning_rate": 1.0700504600771743e-05,
      "loss": 2.4753,
      "step": 7944
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7385578155517578,
      "learning_rate": 1.0680716335213219e-05,
      "loss": 2.6183,
      "step": 7948
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8667320013046265,
      "learning_rate": 1.0660928069654695e-05,
      "loss": 2.5605,
      "step": 7952
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5835785865783691,
      "learning_rate": 1.0641139804096172e-05,
      "loss": 2.4019,
      "step": 7956
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7108229398727417,
      "learning_rate": 1.0621351538537648e-05,
      "loss": 2.708,
      "step": 7960
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.674804925918579,
      "learning_rate": 1.0601563272979124e-05,
      "loss": 2.5102,
      "step": 7964
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.955705165863037,
      "learning_rate": 1.0581775007420599e-05,
      "loss": 2.5221,
      "step": 7968
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.740157961845398,
      "learning_rate": 1.0561986741862075e-05,
      "loss": 2.5232,
      "step": 7972
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6809014081954956,
      "learning_rate": 1.0542198476303553e-05,
      "loss": 2.3372,
      "step": 7976
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7483688592910767,
      "learning_rate": 1.052241021074503e-05,
      "loss": 2.4913,
      "step": 7980
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5279000997543335,
      "learning_rate": 1.0502621945186506e-05,
      "loss": 2.6386,
      "step": 7984
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7373197078704834,
      "learning_rate": 1.048283367962798e-05,
      "loss": 2.5823,
      "step": 7988
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6561895608901978,
      "learning_rate": 1.0463045414069457e-05,
      "loss": 2.4531,
      "step": 7992
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6324809789657593,
      "learning_rate": 1.0443257148510933e-05,
      "loss": 2.4943,
      "step": 7996
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.4932702779769897,
      "learning_rate": 1.042346888295241e-05,
      "loss": 2.5745,
      "step": 8000
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6975860595703125,
      "learning_rate": 1.0403680617393886e-05,
      "loss": 2.5195,
      "step": 8004
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.5818710327148438,
      "learning_rate": 1.0383892351835362e-05,
      "loss": 2.5146,
      "step": 8008
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6648439168930054,
      "learning_rate": 1.0364104086276839e-05,
      "loss": 2.5038,
      "step": 8012
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8664870262145996,
      "learning_rate": 1.0344315820718315e-05,
      "loss": 2.4395,
      "step": 8016
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6299059391021729,
      "learning_rate": 1.0324527555159791e-05,
      "loss": 2.632,
      "step": 8020
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.6017898321151733,
      "learning_rate": 1.0304739289601268e-05,
      "loss": 2.523,
      "step": 8024
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.8422410488128662,
      "learning_rate": 1.0284951024042744e-05,
      "loss": 2.5135,
      "step": 8028
    },
    {
      "epoch": 0.79,
      "grad_norm": 1.7916709184646606,
      "learning_rate": 1.0265162758484218e-05,
      "loss": 2.3953,
      "step": 8032
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6924244165420532,
      "learning_rate": 1.0245374492925695e-05,
      "loss": 2.5083,
      "step": 8036
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6407674551010132,
      "learning_rate": 1.0225586227367171e-05,
      "loss": 2.5489,
      "step": 8040
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.751863718032837,
      "learning_rate": 1.0205797961808649e-05,
      "loss": 2.5538,
      "step": 8044
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8018790483474731,
      "learning_rate": 1.0186009696250125e-05,
      "loss": 2.4931,
      "step": 8048
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6749464273452759,
      "learning_rate": 1.01662214306916e-05,
      "loss": 2.5116,
      "step": 8052
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8620166778564453,
      "learning_rate": 1.0146433165133076e-05,
      "loss": 2.4857,
      "step": 8056
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8505146503448486,
      "learning_rate": 1.0126644899574553e-05,
      "loss": 2.535,
      "step": 8060
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.4946340322494507,
      "learning_rate": 1.0106856634016029e-05,
      "loss": 2.3432,
      "step": 8064
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8005412817001343,
      "learning_rate": 1.0087068368457505e-05,
      "loss": 2.4651,
      "step": 8068
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.786773443222046,
      "learning_rate": 1.006728010289898e-05,
      "loss": 2.4242,
      "step": 8072
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7355870008468628,
      "learning_rate": 1.0047491837340458e-05,
      "loss": 2.4785,
      "step": 8076
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.677388310432434,
      "learning_rate": 1.0027703571781934e-05,
      "loss": 2.5572,
      "step": 8080
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7865822315216064,
      "learning_rate": 1.000791530622341e-05,
      "loss": 2.771,
      "step": 8084
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.8583182096481323,
      "learning_rate": 9.988127040664887e-06,
      "loss": 2.4894,
      "step": 8088
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6651523113250732,
      "learning_rate": 9.968338775106362e-06,
      "loss": 2.5608,
      "step": 8092
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5435396432876587,
      "learning_rate": 9.948550509547838e-06,
      "loss": 2.5702,
      "step": 8096
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6652929782867432,
      "learning_rate": 9.928762243989314e-06,
      "loss": 2.5155,
      "step": 8100
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.5997744798660278,
      "learning_rate": 9.90897397843079e-06,
      "loss": 2.412,
      "step": 8104
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.518398404121399,
      "learning_rate": 9.889185712872267e-06,
      "loss": 2.4749,
      "step": 8108
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6166468858718872,
      "learning_rate": 9.869397447313745e-06,
      "loss": 2.528,
      "step": 8112
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.7324260473251343,
      "learning_rate": 9.84960918175522e-06,
      "loss": 2.4854,
      "step": 8116
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6371716260910034,
      "learning_rate": 9.829820916196696e-06,
      "loss": 2.5037,
      "step": 8120
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.711093544960022,
      "learning_rate": 9.810032650638172e-06,
      "loss": 2.6155,
      "step": 8124
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6455191373825073,
      "learning_rate": 9.790244385079648e-06,
      "loss": 2.4078,
      "step": 8128
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.6366294622421265,
      "learning_rate": 9.770456119521125e-06,
      "loss": 2.4642,
      "step": 8132
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.668990969657898,
      "learning_rate": 9.7506678539626e-06,
      "loss": 2.5074,
      "step": 8136
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.5362697839736938,
      "learning_rate": 9.730879588404076e-06,
      "loss": 2.4896,
      "step": 8140
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8022223711013794,
      "learning_rate": 9.711091322845554e-06,
      "loss": 2.5353,
      "step": 8144
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6878408193588257,
      "learning_rate": 9.69130305728703e-06,
      "loss": 2.5586,
      "step": 8148
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.689186692237854,
      "learning_rate": 9.671514791728506e-06,
      "loss": 2.655,
      "step": 8152
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6996088027954102,
      "learning_rate": 9.651726526169981e-06,
      "loss": 2.7126,
      "step": 8156
    },
    {
      "epoch": 0.81,
      "grad_norm": 2.8080925941467285,
      "learning_rate": 9.631938260611457e-06,
      "loss": 2.7099,
      "step": 8160
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7663508653640747,
      "learning_rate": 9.612149995052934e-06,
      "loss": 2.4897,
      "step": 8164
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7868987321853638,
      "learning_rate": 9.59236172949441e-06,
      "loss": 2.4844,
      "step": 8168
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6591540575027466,
      "learning_rate": 9.572573463935886e-06,
      "loss": 2.4604,
      "step": 8172
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7677662372589111,
      "learning_rate": 9.552785198377363e-06,
      "loss": 2.5354,
      "step": 8176
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6505751609802246,
      "learning_rate": 9.532996932818839e-06,
      "loss": 2.5834,
      "step": 8180
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7923264503479004,
      "learning_rate": 9.513208667260315e-06,
      "loss": 2.5445,
      "step": 8184
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6527979373931885,
      "learning_rate": 9.493420401701792e-06,
      "loss": 2.3743,
      "step": 8188
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.490796446800232,
      "learning_rate": 9.473632136143268e-06,
      "loss": 2.4511,
      "step": 8192
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6742923259735107,
      "learning_rate": 9.453843870584744e-06,
      "loss": 2.5265,
      "step": 8196
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6631675958633423,
      "learning_rate": 9.434055605026219e-06,
      "loss": 2.3355,
      "step": 8200
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.9126391410827637,
      "learning_rate": 9.414267339467695e-06,
      "loss": 2.5204,
      "step": 8204
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6832706928253174,
      "learning_rate": 9.394479073909172e-06,
      "loss": 2.541,
      "step": 8208
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.8622807264328003,
      "learning_rate": 9.37469080835065e-06,
      "loss": 2.6101,
      "step": 8212
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.7272640466690063,
      "learning_rate": 9.354902542792126e-06,
      "loss": 2.7245,
      "step": 8216
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6699191331863403,
      "learning_rate": 9.3351142772336e-06,
      "loss": 2.4707,
      "step": 8220
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.599859595298767,
      "learning_rate": 9.315326011675077e-06,
      "loss": 2.5186,
      "step": 8224
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6403051614761353,
      "learning_rate": 9.295537746116553e-06,
      "loss": 2.6446,
      "step": 8228
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.6128116846084595,
      "learning_rate": 9.27574948055803e-06,
      "loss": 2.6126,
      "step": 8232
    },
    {
      "epoch": 0.81,
      "grad_norm": 1.588544487953186,
      "learning_rate": 9.255961214999506e-06,
      "loss": 2.6034,
      "step": 8236
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7381192445755005,
      "learning_rate": 9.236172949440982e-06,
      "loss": 2.6789,
      "step": 8240
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6580698490142822,
      "learning_rate": 9.216384683882458e-06,
      "loss": 2.6275,
      "step": 8244
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.779470443725586,
      "learning_rate": 9.196596418323935e-06,
      "loss": 2.6094,
      "step": 8248
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6793197393417358,
      "learning_rate": 9.176808152765411e-06,
      "loss": 2.3446,
      "step": 8252
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7751526832580566,
      "learning_rate": 9.157019887206887e-06,
      "loss": 2.5427,
      "step": 8256
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.687898874282837,
      "learning_rate": 9.137231621648364e-06,
      "loss": 2.5313,
      "step": 8260
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5011862516403198,
      "learning_rate": 9.117443356089838e-06,
      "loss": 2.4652,
      "step": 8264
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7039753198623657,
      "learning_rate": 9.097655090531315e-06,
      "loss": 2.5125,
      "step": 8268
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7907326221466064,
      "learning_rate": 9.077866824972791e-06,
      "loss": 2.4032,
      "step": 8272
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.689347267150879,
      "learning_rate": 9.058078559414267e-06,
      "loss": 2.4636,
      "step": 8276
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7220097780227661,
      "learning_rate": 9.038290293855745e-06,
      "loss": 2.4757,
      "step": 8280
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.8284951448440552,
      "learning_rate": 9.01850202829722e-06,
      "loss": 2.473,
      "step": 8284
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.7803459167480469,
      "learning_rate": 8.998713762738696e-06,
      "loss": 2.4781,
      "step": 8288
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6893064975738525,
      "learning_rate": 8.978925497180173e-06,
      "loss": 2.5115,
      "step": 8292
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.8316203355789185,
      "learning_rate": 8.959137231621649e-06,
      "loss": 2.5759,
      "step": 8296
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6714107990264893,
      "learning_rate": 8.939348966063125e-06,
      "loss": 2.5676,
      "step": 8300
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5850313901901245,
      "learning_rate": 8.9195607005046e-06,
      "loss": 2.3758,
      "step": 8304
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.595004677772522,
      "learning_rate": 8.899772434946078e-06,
      "loss": 2.4782,
      "step": 8308
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.756579041481018,
      "learning_rate": 8.879984169387554e-06,
      "loss": 2.4742,
      "step": 8312
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.601770043373108,
      "learning_rate": 8.86019590382903e-06,
      "loss": 2.3767,
      "step": 8316
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.8463127613067627,
      "learning_rate": 8.840407638270507e-06,
      "loss": 2.5466,
      "step": 8320
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.8509941101074219,
      "learning_rate": 8.820619372711983e-06,
      "loss": 2.606,
      "step": 8324
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.6294124126434326,
      "learning_rate": 8.800831107153458e-06,
      "loss": 2.4114,
      "step": 8328
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.5632402896881104,
      "learning_rate": 8.781042841594934e-06,
      "loss": 2.3389,
      "step": 8332
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.602977991104126,
      "learning_rate": 8.76125457603641e-06,
      "loss": 2.5443,
      "step": 8336
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.8994452953338623,
      "learning_rate": 8.741466310477887e-06,
      "loss": 2.4973,
      "step": 8340
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.703715205192566,
      "learning_rate": 8.721678044919363e-06,
      "loss": 2.3838,
      "step": 8344
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7744590044021606,
      "learning_rate": 8.70188977936084e-06,
      "loss": 2.2877,
      "step": 8348
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7415697574615479,
      "learning_rate": 8.682101513802316e-06,
      "loss": 2.5678,
      "step": 8352
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.6912627220153809,
      "learning_rate": 8.662313248243792e-06,
      "loss": 2.3746,
      "step": 8356
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.591155767440796,
      "learning_rate": 8.642524982685268e-06,
      "loss": 2.4637,
      "step": 8360
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.625097393989563,
      "learning_rate": 8.622736717126745e-06,
      "loss": 2.4071,
      "step": 8364
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7765227556228638,
      "learning_rate": 8.60294845156822e-06,
      "loss": 2.4339,
      "step": 8368
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7714086771011353,
      "learning_rate": 8.583160186009696e-06,
      "loss": 2.4657,
      "step": 8372
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5393376350402832,
      "learning_rate": 8.563371920451174e-06,
      "loss": 2.4295,
      "step": 8376
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.6363029479980469,
      "learning_rate": 8.54358365489265e-06,
      "loss": 2.3262,
      "step": 8380
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.6590903997421265,
      "learning_rate": 8.523795389334126e-06,
      "loss": 2.5488,
      "step": 8384
    },
    {
      "epoch": 0.83,
      "grad_norm": 2.057309150695801,
      "learning_rate": 8.504007123775601e-06,
      "loss": 2.4571,
      "step": 8388
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.643871784210205,
      "learning_rate": 8.484218858217077e-06,
      "loss": 2.5755,
      "step": 8392
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5784872770309448,
      "learning_rate": 8.464430592658554e-06,
      "loss": 2.3636,
      "step": 8396
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5051548480987549,
      "learning_rate": 8.44464232710003e-06,
      "loss": 2.3239,
      "step": 8400
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7668542861938477,
      "learning_rate": 8.424854061541506e-06,
      "loss": 2.4491,
      "step": 8404
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.7257317304611206,
      "learning_rate": 8.405065795982982e-06,
      "loss": 2.4014,
      "step": 8408
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.6491328477859497,
      "learning_rate": 8.385277530424459e-06,
      "loss": 2.5093,
      "step": 8412
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5918080806732178,
      "learning_rate": 8.365489264865935e-06,
      "loss": 2.7333,
      "step": 8416
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.709262490272522,
      "learning_rate": 8.345700999307411e-06,
      "loss": 2.3297,
      "step": 8420
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5516706705093384,
      "learning_rate": 8.325912733748888e-06,
      "loss": 2.3839,
      "step": 8424
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.711029291152954,
      "learning_rate": 8.306124468190364e-06,
      "loss": 2.4985,
      "step": 8428
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.5779407024383545,
      "learning_rate": 8.286336202631839e-06,
      "loss": 2.4944,
      "step": 8432
    },
    {
      "epoch": 0.83,
      "grad_norm": 1.8659144639968872,
      "learning_rate": 8.266547937073315e-06,
      "loss": 2.4244,
      "step": 8436
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.93182373046875,
      "learning_rate": 8.246759671514791e-06,
      "loss": 2.4782,
      "step": 8440
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.5297750234603882,
      "learning_rate": 8.22697140595627e-06,
      "loss": 2.5646,
      "step": 8444
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7578099966049194,
      "learning_rate": 8.207183140397746e-06,
      "loss": 2.5986,
      "step": 8448
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7178421020507812,
      "learning_rate": 8.18739487483922e-06,
      "loss": 2.4336,
      "step": 8452
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6756671667099,
      "learning_rate": 8.167606609280697e-06,
      "loss": 2.5326,
      "step": 8456
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.681978464126587,
      "learning_rate": 8.147818343722173e-06,
      "loss": 2.4505,
      "step": 8460
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.777003288269043,
      "learning_rate": 8.12803007816365e-06,
      "loss": 2.4595,
      "step": 8464
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7001906633377075,
      "learning_rate": 8.108241812605126e-06,
      "loss": 2.3165,
      "step": 8468
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6168255805969238,
      "learning_rate": 8.0884535470466e-06,
      "loss": 2.4513,
      "step": 8472
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7238038778305054,
      "learning_rate": 8.068665281488078e-06,
      "loss": 2.5386,
      "step": 8476
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7556039094924927,
      "learning_rate": 8.048877015929555e-06,
      "loss": 2.5683,
      "step": 8480
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.8705509901046753,
      "learning_rate": 8.029088750371031e-06,
      "loss": 2.411,
      "step": 8484
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.5241179466247559,
      "learning_rate": 8.009300484812507e-06,
      "loss": 2.424,
      "step": 8488
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6394929885864258,
      "learning_rate": 7.989512219253984e-06,
      "loss": 2.4855,
      "step": 8492
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7148469686508179,
      "learning_rate": 7.969723953695458e-06,
      "loss": 2.4967,
      "step": 8496
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7260617017745972,
      "learning_rate": 7.949935688136934e-06,
      "loss": 2.5231,
      "step": 8500
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.84506356716156,
      "learning_rate": 7.93014742257841e-06,
      "loss": 2.3639,
      "step": 8504
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.5255876779556274,
      "learning_rate": 7.910359157019887e-06,
      "loss": 2.4189,
      "step": 8508
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7980495691299438,
      "learning_rate": 7.890570891461365e-06,
      "loss": 2.4907,
      "step": 8512
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7311155796051025,
      "learning_rate": 7.87078262590284e-06,
      "loss": 2.5981,
      "step": 8516
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6992567777633667,
      "learning_rate": 7.850994360344316e-06,
      "loss": 2.5617,
      "step": 8520
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6677074432373047,
      "learning_rate": 7.831206094785792e-06,
      "loss": 2.5869,
      "step": 8524
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.717405080795288,
      "learning_rate": 7.811417829227269e-06,
      "loss": 2.534,
      "step": 8528
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7169833183288574,
      "learning_rate": 7.791629563668745e-06,
      "loss": 2.5644,
      "step": 8532
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.6315467357635498,
      "learning_rate": 7.77184129811022e-06,
      "loss": 2.5054,
      "step": 8536
    },
    {
      "epoch": 0.84,
      "grad_norm": 1.7004671096801758,
      "learning_rate": 7.752053032551696e-06,
      "loss": 2.3946,
      "step": 8540
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7185847759246826,
      "learning_rate": 7.732264766993174e-06,
      "loss": 2.6361,
      "step": 8544
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7826788425445557,
      "learning_rate": 7.71247650143465e-06,
      "loss": 2.4693,
      "step": 8548
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5494498014450073,
      "learning_rate": 7.692688235876127e-06,
      "loss": 2.475,
      "step": 8552
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6112264394760132,
      "learning_rate": 7.672899970317603e-06,
      "loss": 2.685,
      "step": 8556
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.583173394203186,
      "learning_rate": 7.653111704759078e-06,
      "loss": 2.4246,
      "step": 8560
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6651097536087036,
      "learning_rate": 7.633323439200554e-06,
      "loss": 2.4226,
      "step": 8564
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6389166116714478,
      "learning_rate": 7.613535173642031e-06,
      "loss": 2.4979,
      "step": 8568
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7426066398620605,
      "learning_rate": 7.5937469080835074e-06,
      "loss": 2.5513,
      "step": 8572
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6744520664215088,
      "learning_rate": 7.573958642524984e-06,
      "loss": 2.4503,
      "step": 8576
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7158594131469727,
      "learning_rate": 7.554170376966458e-06,
      "loss": 2.4517,
      "step": 8580
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6604433059692383,
      "learning_rate": 7.5343821114079355e-06,
      "loss": 2.4247,
      "step": 8584
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.636860728263855,
      "learning_rate": 7.514593845849412e-06,
      "loss": 2.2244,
      "step": 8588
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6336443424224854,
      "learning_rate": 7.494805580290888e-06,
      "loss": 2.5156,
      "step": 8592
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.8386069536209106,
      "learning_rate": 7.4750173147323645e-06,
      "loss": 2.4954,
      "step": 8596
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6216977834701538,
      "learning_rate": 7.45522904917384e-06,
      "loss": 2.4894,
      "step": 8600
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7700227499008179,
      "learning_rate": 7.435440783615316e-06,
      "loss": 2.6143,
      "step": 8604
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6691062450408936,
      "learning_rate": 7.415652518056793e-06,
      "loss": 2.4329,
      "step": 8608
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.517760992050171,
      "learning_rate": 7.395864252498269e-06,
      "loss": 2.402,
      "step": 8612
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5939360857009888,
      "learning_rate": 7.376075986939745e-06,
      "loss": 2.4902,
      "step": 8616
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.6323703527450562,
      "learning_rate": 7.356287721381221e-06,
      "loss": 2.6149,
      "step": 8620
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.5932726860046387,
      "learning_rate": 7.336499455822697e-06,
      "loss": 2.4341,
      "step": 8624
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7558265924453735,
      "learning_rate": 7.316711190264173e-06,
      "loss": 2.6402,
      "step": 8628
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.7289772033691406,
      "learning_rate": 7.29692292470565e-06,
      "loss": 2.3923,
      "step": 8632
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.800723910331726,
      "learning_rate": 7.277134659147127e-06,
      "loss": 2.4707,
      "step": 8636
    },
    {
      "epoch": 0.85,
      "grad_norm": 1.672587275505066,
      "learning_rate": 7.257346393588603e-06,
      "loss": 2.526,
      "step": 8640
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.686955451965332,
      "learning_rate": 7.237558128030078e-06,
      "loss": 2.4433,
      "step": 8644
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.5995945930480957,
      "learning_rate": 7.217769862471554e-06,
      "loss": 2.609,
      "step": 8648
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6582813262939453,
      "learning_rate": 7.197981596913031e-06,
      "loss": 2.3675,
      "step": 8652
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6642727851867676,
      "learning_rate": 7.178193331354508e-06,
      "loss": 2.4762,
      "step": 8656
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6755313873291016,
      "learning_rate": 7.158405065795984e-06,
      "loss": 2.4811,
      "step": 8660
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.787563681602478,
      "learning_rate": 7.138616800237459e-06,
      "loss": 2.4481,
      "step": 8664
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.5694202184677124,
      "learning_rate": 7.118828534678936e-06,
      "loss": 2.44,
      "step": 8668
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6753268241882324,
      "learning_rate": 7.099040269120412e-06,
      "loss": 2.664,
      "step": 8672
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8283534049987793,
      "learning_rate": 7.079252003561888e-06,
      "loss": 2.4761,
      "step": 8676
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.730433702468872,
      "learning_rate": 7.059463738003365e-06,
      "loss": 2.5623,
      "step": 8680
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7397688627243042,
      "learning_rate": 7.03967547244484e-06,
      "loss": 2.6005,
      "step": 8684
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7049192190170288,
      "learning_rate": 7.0198872068863165e-06,
      "loss": 2.4976,
      "step": 8688
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7921533584594727,
      "learning_rate": 7.000098941327793e-06,
      "loss": 2.5239,
      "step": 8692
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7059839963912964,
      "learning_rate": 6.980310675769269e-06,
      "loss": 2.403,
      "step": 8696
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.636533260345459,
      "learning_rate": 6.9605224102107455e-06,
      "loss": 2.4023,
      "step": 8700
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6458524465560913,
      "learning_rate": 6.940734144652221e-06,
      "loss": 2.6021,
      "step": 8704
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.8312097787857056,
      "learning_rate": 6.920945879093697e-06,
      "loss": 2.4073,
      "step": 8708
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6643646955490112,
      "learning_rate": 6.901157613535174e-06,
      "loss": 2.4315,
      "step": 8712
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7501451969146729,
      "learning_rate": 6.88136934797665e-06,
      "loss": 2.4771,
      "step": 8716
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.6172442436218262,
      "learning_rate": 6.861581082418127e-06,
      "loss": 2.4944,
      "step": 8720
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7708154916763306,
      "learning_rate": 6.841792816859603e-06,
      "loss": 2.5976,
      "step": 8724
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.640822410583496,
      "learning_rate": 6.822004551301078e-06,
      "loss": 2.4854,
      "step": 8728
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7448416948318481,
      "learning_rate": 6.802216285742554e-06,
      "loss": 2.5023,
      "step": 8732
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7678117752075195,
      "learning_rate": 6.7824280201840315e-06,
      "loss": 2.3739,
      "step": 8736
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.7742412090301514,
      "learning_rate": 6.762639754625508e-06,
      "loss": 2.5075,
      "step": 8740
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6153099536895752,
      "learning_rate": 6.742851489066984e-06,
      "loss": 2.3971,
      "step": 8744
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.1363136768341064,
      "learning_rate": 6.72306322350846e-06,
      "loss": 2.5768,
      "step": 8748
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.694061279296875,
      "learning_rate": 6.703274957949936e-06,
      "loss": 2.4628,
      "step": 8752
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5482895374298096,
      "learning_rate": 6.683486692391412e-06,
      "loss": 2.3695,
      "step": 8756
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6727862358093262,
      "learning_rate": 6.663698426832889e-06,
      "loss": 2.5452,
      "step": 8760
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.0361154079437256,
      "learning_rate": 6.643910161274365e-06,
      "loss": 2.7449,
      "step": 8764
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6247440576553345,
      "learning_rate": 6.62412189571584e-06,
      "loss": 2.5772,
      "step": 8768
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6015691757202148,
      "learning_rate": 6.604333630157317e-06,
      "loss": 2.5729,
      "step": 8772
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.7053395509719849,
      "learning_rate": 6.584545364598793e-06,
      "loss": 2.541,
      "step": 8776
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6078673601150513,
      "learning_rate": 6.564757099040269e-06,
      "loss": 2.5443,
      "step": 8780
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6850202083587646,
      "learning_rate": 6.544968833481746e-06,
      "loss": 2.6706,
      "step": 8784
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6219288110733032,
      "learning_rate": 6.525180567923223e-06,
      "loss": 2.5243,
      "step": 8788
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8155920505523682,
      "learning_rate": 6.5053923023646975e-06,
      "loss": 2.5544,
      "step": 8792
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.7686444520950317,
      "learning_rate": 6.485604036806174e-06,
      "loss": 2.3905,
      "step": 8796
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.9808334112167358,
      "learning_rate": 6.46581577124765e-06,
      "loss": 2.5068,
      "step": 8800
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.7255302667617798,
      "learning_rate": 6.446027505689127e-06,
      "loss": 2.4938,
      "step": 8804
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6764934062957764,
      "learning_rate": 6.426239240130604e-06,
      "loss": 2.3194,
      "step": 8808
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5869959592819214,
      "learning_rate": 6.406450974572078e-06,
      "loss": 2.4587,
      "step": 8812
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5677440166473389,
      "learning_rate": 6.386662709013555e-06,
      "loss": 2.4172,
      "step": 8816
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6823347806930542,
      "learning_rate": 6.366874443455032e-06,
      "loss": 2.4559,
      "step": 8820
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6375516653060913,
      "learning_rate": 6.347086177896508e-06,
      "loss": 2.5198,
      "step": 8824
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.629103422164917,
      "learning_rate": 6.327297912337984e-06,
      "loss": 2.5148,
      "step": 8828
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.5278247594833374,
      "learning_rate": 6.30750964677946e-06,
      "loss": 2.4411,
      "step": 8832
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.8897967338562012,
      "learning_rate": 6.287721381220936e-06,
      "loss": 2.4329,
      "step": 8836
    },
    {
      "epoch": 0.87,
      "grad_norm": 1.6970409154891968,
      "learning_rate": 6.2679331156624125e-06,
      "loss": 2.7446,
      "step": 8840
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.5253138542175293,
      "learning_rate": 6.248144850103889e-06,
      "loss": 2.4237,
      "step": 8844
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6197317838668823,
      "learning_rate": 6.228356584545364e-06,
      "loss": 2.5055,
      "step": 8848
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.5637788772583008,
      "learning_rate": 6.2085683189868415e-06,
      "loss": 2.3537,
      "step": 8852
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.662288784980774,
      "learning_rate": 6.188780053428318e-06,
      "loss": 2.3261,
      "step": 8856
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.642146348953247,
      "learning_rate": 6.168991787869793e-06,
      "loss": 2.4903,
      "step": 8860
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6560384035110474,
      "learning_rate": 6.1492035223112696e-06,
      "loss": 2.5563,
      "step": 8864
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7499784231185913,
      "learning_rate": 6.129415256752746e-06,
      "loss": 2.5078,
      "step": 8868
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.8147586584091187,
      "learning_rate": 6.109626991194222e-06,
      "loss": 2.3585,
      "step": 8872
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6541246175765991,
      "learning_rate": 6.0898387256356985e-06,
      "loss": 2.5436,
      "step": 8876
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.8309890031814575,
      "learning_rate": 6.070050460077174e-06,
      "loss": 2.6095,
      "step": 8880
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.725372552871704,
      "learning_rate": 6.050262194518651e-06,
      "loss": 2.5652,
      "step": 8884
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6648577451705933,
      "learning_rate": 6.0304739289601275e-06,
      "loss": 2.4624,
      "step": 8888
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6224662065505981,
      "learning_rate": 6.010685663401603e-06,
      "loss": 2.4616,
      "step": 8892
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.535581111907959,
      "learning_rate": 5.990897397843079e-06,
      "loss": 2.3335,
      "step": 8896
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.5054175853729248,
      "learning_rate": 5.971109132284556e-06,
      "loss": 2.3097,
      "step": 8900
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7295256853103638,
      "learning_rate": 5.951320866726032e-06,
      "loss": 2.538,
      "step": 8904
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6823656558990479,
      "learning_rate": 5.931532601167508e-06,
      "loss": 2.4051,
      "step": 8908
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.702989935874939,
      "learning_rate": 5.911744335608984e-06,
      "loss": 2.3926,
      "step": 8912
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6187750101089478,
      "learning_rate": 5.89195607005046e-06,
      "loss": 2.2627,
      "step": 8916
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6742513179779053,
      "learning_rate": 5.872167804491937e-06,
      "loss": 2.3604,
      "step": 8920
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7713173627853394,
      "learning_rate": 5.852379538933413e-06,
      "loss": 2.557,
      "step": 8924
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.624006748199463,
      "learning_rate": 5.832591273374889e-06,
      "loss": 2.4546,
      "step": 8928
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.8159652948379517,
      "learning_rate": 5.812803007816365e-06,
      "loss": 2.4153,
      "step": 8932
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6506266593933105,
      "learning_rate": 5.793014742257842e-06,
      "loss": 2.6401,
      "step": 8936
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.7533698081970215,
      "learning_rate": 5.773226476699318e-06,
      "loss": 2.5884,
      "step": 8940
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.6324323415756226,
      "learning_rate": 5.7534382111407935e-06,
      "loss": 2.2965,
      "step": 8944
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5913504362106323,
      "learning_rate": 5.73364994558227e-06,
      "loss": 2.4982,
      "step": 8948
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.9502431154251099,
      "learning_rate": 5.713861680023746e-06,
      "loss": 2.3815,
      "step": 8952
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5532846450805664,
      "learning_rate": 5.694073414465222e-06,
      "loss": 2.4715,
      "step": 8956
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6778051853179932,
      "learning_rate": 5.674285148906699e-06,
      "loss": 2.2548,
      "step": 8960
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5361812114715576,
      "learning_rate": 5.654496883348174e-06,
      "loss": 2.4623,
      "step": 8964
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6158546209335327,
      "learning_rate": 5.634708617789651e-06,
      "loss": 2.4732,
      "step": 8968
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.7343107461929321,
      "learning_rate": 5.614920352231128e-06,
      "loss": 2.4489,
      "step": 8972
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6918219327926636,
      "learning_rate": 5.595132086672603e-06,
      "loss": 2.4923,
      "step": 8976
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.725468397140503,
      "learning_rate": 5.5753438211140795e-06,
      "loss": 2.5126,
      "step": 8980
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6052392721176147,
      "learning_rate": 5.555555555555556e-06,
      "loss": 2.3489,
      "step": 8984
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6850762367248535,
      "learning_rate": 5.535767289997032e-06,
      "loss": 2.5302,
      "step": 8988
    },
    {
      "epoch": 0.89,
      "grad_norm": 2.009579658508301,
      "learning_rate": 5.5159790244385085e-06,
      "loss": 2.4038,
      "step": 8992
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.787927269935608,
      "learning_rate": 5.496190758879984e-06,
      "loss": 2.4092,
      "step": 8996
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.583129644393921,
      "learning_rate": 5.476402493321461e-06,
      "loss": 2.5511,
      "step": 9000
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5953203439712524,
      "learning_rate": 5.4566142277629374e-06,
      "loss": 2.3661,
      "step": 9004
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6569699048995972,
      "learning_rate": 5.436825962204413e-06,
      "loss": 2.3866,
      "step": 9008
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5289307832717896,
      "learning_rate": 5.417037696645889e-06,
      "loss": 2.4527,
      "step": 9012
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5741091966629028,
      "learning_rate": 5.3972494310873655e-06,
      "loss": 2.5182,
      "step": 9016
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.672529697418213,
      "learning_rate": 5.377461165528842e-06,
      "loss": 2.4246,
      "step": 9020
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.5960274934768677,
      "learning_rate": 5.357672899970318e-06,
      "loss": 2.3121,
      "step": 9024
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.9494187831878662,
      "learning_rate": 5.337884634411794e-06,
      "loss": 2.3272,
      "step": 9028
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.8283107280731201,
      "learning_rate": 5.31809636885327e-06,
      "loss": 2.4936,
      "step": 9032
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6064289808273315,
      "learning_rate": 5.298308103294747e-06,
      "loss": 2.5352,
      "step": 9036
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.6828787326812744,
      "learning_rate": 5.278519837736223e-06,
      "loss": 2.2623,
      "step": 9040
    },
    {
      "epoch": 0.89,
      "grad_norm": 1.93602454662323,
      "learning_rate": 5.258731572177699e-06,
      "loss": 2.4931,
      "step": 9044
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6569011211395264,
      "learning_rate": 5.238943306619174e-06,
      "loss": 2.4178,
      "step": 9048
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6328504085540771,
      "learning_rate": 5.219155041060652e-06,
      "loss": 2.6022,
      "step": 9052
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7538514137268066,
      "learning_rate": 5.199366775502128e-06,
      "loss": 2.3894,
      "step": 9056
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7447439432144165,
      "learning_rate": 5.179578509943603e-06,
      "loss": 2.3315,
      "step": 9060
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.793074607849121,
      "learning_rate": 5.15979024438508e-06,
      "loss": 2.4952,
      "step": 9064
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6562350988388062,
      "learning_rate": 5.140001978826556e-06,
      "loss": 2.3996,
      "step": 9068
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.7558027505874634,
      "learning_rate": 5.120213713268032e-06,
      "loss": 2.4778,
      "step": 9072
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6544588804244995,
      "learning_rate": 5.100425447709509e-06,
      "loss": 2.5229,
      "step": 9076
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.775094985961914,
      "learning_rate": 5.080637182150984e-06,
      "loss": 2.3536,
      "step": 9080
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.9515008926391602,
      "learning_rate": 5.060848916592461e-06,
      "loss": 2.4411,
      "step": 9084
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.8084925413131714,
      "learning_rate": 5.041060651033938e-06,
      "loss": 2.3756,
      "step": 9088
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6303074359893799,
      "learning_rate": 5.021272385475413e-06,
      "loss": 2.3407,
      "step": 9092
    },
    {
      "epoch": 0.9,
      "grad_norm": 1.6556874513626099,
      "learning_rate": 5.0014841199168894e-06,
      "loss": 2.4766,
      "step": 9096
    }
  ],
  "logging_steps": 4,
  "max_steps": 10107,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1011,
  "total_flos": 1.1128110498761933e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}