{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 3750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.004,
      "grad_norm": 2.46875,
      "learning_rate": 5e-06,
      "loss": 0.5096,
      "step": 5
    },
    {
      "epoch": 0.008,
      "grad_norm": 14.8125,
      "learning_rate": 1e-05,
      "loss": 0.5303,
      "step": 10
    },
    {
      "epoch": 0.012,
      "grad_norm": 14.5,
      "learning_rate": 1.5e-05,
      "loss": 0.8669,
      "step": 15
    },
    {
      "epoch": 0.016,
      "grad_norm": 2.765625,
      "learning_rate": 2e-05,
      "loss": 0.5856,
      "step": 20
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.75,
      "learning_rate": 2.5e-05,
      "loss": 0.466,
      "step": 25
    },
    {
      "epoch": 0.024,
      "grad_norm": 24.125,
      "learning_rate": 3e-05,
      "loss": 0.6848,
      "step": 30
    },
    {
      "epoch": 0.028,
      "grad_norm": 4.84375,
      "learning_rate": 3.5e-05,
      "loss": 0.5357,
      "step": 35
    },
    {
      "epoch": 0.032,
      "grad_norm": 6.15625,
      "learning_rate": 4e-05,
      "loss": 0.5755,
      "step": 40
    },
    {
      "epoch": 0.036,
      "grad_norm": 4.21875,
      "learning_rate": 4.5e-05,
      "loss": 0.4748,
      "step": 45
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.0,
      "learning_rate": 5e-05,
      "loss": 0.3974,
      "step": 50
    },
    {
      "epoch": 0.044,
      "grad_norm": 9.1875,
      "learning_rate": 5.500000000000001e-05,
      "loss": 0.5292,
      "step": 55
    },
    {
      "epoch": 0.048,
      "grad_norm": 5.59375,
      "learning_rate": 6e-05,
      "loss": 0.3294,
      "step": 60
    },
    {
      "epoch": 0.052,
      "grad_norm": 1.890625,
      "learning_rate": 6.500000000000001e-05,
      "loss": 0.4085,
      "step": 65
    },
    {
      "epoch": 0.056,
      "grad_norm": 16.5,
      "learning_rate": 7e-05,
      "loss": 0.4794,
      "step": 70
    },
    {
      "epoch": 0.06,
      "grad_norm": 1.2734375,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.265,
      "step": 75
    },
    {
      "epoch": 0.064,
      "grad_norm": 6.125,
      "learning_rate": 8e-05,
      "loss": 0.2528,
      "step": 80
    },
    {
      "epoch": 0.068,
      "grad_norm": 14.1875,
      "learning_rate": 8.5e-05,
      "loss": 0.4565,
      "step": 85
    },
    {
      "epoch": 0.072,
      "grad_norm": 10.5,
      "learning_rate": 9e-05,
      "loss": 0.1783,
      "step": 90
    },
    {
      "epoch": 0.076,
      "grad_norm": 0.67578125,
      "learning_rate": 9.5e-05,
      "loss": 0.1872,
      "step": 95
    },
    {
      "epoch": 0.08,
      "grad_norm": 8.8125,
      "learning_rate": 0.0001,
      "loss": 0.22,
      "step": 100
    },
    {
      "epoch": 0.084,
      "grad_norm": 8.0625,
      "learning_rate": 9.99995416169414e-05,
      "loss": 0.2013,
      "step": 105
    },
    {
      "epoch": 0.088,
      "grad_norm": 8.625,
      "learning_rate": 9.999816647625514e-05,
      "loss": 0.1776,
      "step": 110
    },
    {
      "epoch": 0.092,
      "grad_norm": 4.125,
      "learning_rate": 9.999587460340953e-05,
      "loss": 0.2741,
      "step": 115
    },
    {
      "epoch": 0.096,
      "grad_norm": 3.09375,
      "learning_rate": 9.999266604085126e-05,
      "loss": 0.2833,
      "step": 120
    },
    {
      "epoch": 0.1,
      "grad_norm": 7.15625,
      "learning_rate": 9.998854084800461e-05,
      "loss": 0.2947,
      "step": 125
    },
    {
      "epoch": 0.104,
      "grad_norm": 8.3125,
      "learning_rate": 9.99834991012703e-05,
      "loss": 0.369,
      "step": 130
    },
    {
      "epoch": 0.108,
      "grad_norm": 2.171875,
      "learning_rate": 9.997754089402419e-05,
      "loss": 0.2246,
      "step": 135
    },
    {
      "epoch": 0.112,
      "grad_norm": 2.421875,
      "learning_rate": 9.997066633661538e-05,
      "loss": 0.223,
      "step": 140
    },
    {
      "epoch": 0.116,
      "grad_norm": 4.0625,
      "learning_rate": 9.996287555636434e-05,
      "loss": 0.2326,
      "step": 145
    },
    {
      "epoch": 0.12,
      "grad_norm": 2.125,
      "learning_rate": 9.99541686975604e-05,
      "loss": 0.3262,
      "step": 150
    },
    {
      "epoch": 0.124,
      "grad_norm": 4.03125,
      "learning_rate": 9.994454592145917e-05,
      "loss": 0.3315,
      "step": 155
    },
    {
      "epoch": 0.128,
      "grad_norm": 3.171875,
      "learning_rate": 9.993400740627956e-05,
      "loss": 0.19,
      "step": 160
    },
    {
      "epoch": 0.132,
      "grad_norm": 2.953125,
      "learning_rate": 9.992255334720044e-05,
      "loss": 0.2665,
      "step": 165
    },
    {
      "epoch": 0.136,
      "grad_norm": 5.34375,
      "learning_rate": 9.991018395635701e-05,
      "loss": 0.2359,
      "step": 170
    },
    {
      "epoch": 0.14,
      "grad_norm": 14.8125,
      "learning_rate": 9.989689946283693e-05,
      "loss": 0.2083,
      "step": 175
    },
    {
      "epoch": 0.144,
      "grad_norm": 5.53125,
      "learning_rate": 9.9882700112676e-05,
      "loss": 0.2664,
      "step": 180
    },
    {
      "epoch": 0.148,
      "grad_norm": 10.1875,
      "learning_rate": 9.98675861688537e-05,
      "loss": 0.3044,
      "step": 185
    },
    {
      "epoch": 0.152,
      "grad_norm": 6.46875,
      "learning_rate": 9.985155791128826e-05,
      "loss": 0.2775,
      "step": 190
    },
    {
      "epoch": 0.156,
      "grad_norm": 5.21875,
      "learning_rate": 9.983461563683142e-05,
      "loss": 0.2734,
      "step": 195
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.1875,
      "learning_rate": 9.981675965926309e-05,
      "loss": 0.3062,
      "step": 200
    },
    {
      "epoch": 0.164,
      "grad_norm": 6.71875,
      "learning_rate": 9.979799030928537e-05,
      "loss": 0.238,
      "step": 205
    },
    {
      "epoch": 0.168,
      "grad_norm": 8.8125,
      "learning_rate": 9.977830793451653e-05,
      "loss": 0.3094,
      "step": 210
    },
    {
      "epoch": 0.172,
      "grad_norm": 12.125,
      "learning_rate": 9.975771289948454e-05,
      "loss": 0.2416,
      "step": 215
    },
    {
      "epoch": 0.176,
      "grad_norm": 3.515625,
      "learning_rate": 9.97362055856203e-05,
      "loss": 0.2148,
      "step": 220
    },
    {
      "epoch": 0.18,
      "grad_norm": 6.5,
      "learning_rate": 9.971378639125062e-05,
      "loss": 0.2539,
      "step": 225
    },
    {
      "epoch": 0.184,
      "grad_norm": 1.6171875,
      "learning_rate": 9.969045573159081e-05,
      "loss": 0.3035,
      "step": 230
    },
    {
      "epoch": 0.188,
      "grad_norm": 7.5625,
      "learning_rate": 9.966621403873702e-05,
      "loss": 0.1952,
      "step": 235
    },
    {
      "epoch": 0.192,
      "grad_norm": 2.234375,
      "learning_rate": 9.964106176165814e-05,
      "loss": 0.2455,
      "step": 240
    },
    {
      "epoch": 0.196,
      "grad_norm": 8.9375,
      "learning_rate": 9.961499936618766e-05,
      "loss": 0.3092,
      "step": 245
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.9453125,
      "learning_rate": 9.958802733501485e-05,
      "loss": 0.1766,
      "step": 250
    },
    {
      "epoch": 0.204,
      "grad_norm": 8.6875,
      "learning_rate": 9.9560146167676e-05,
      "loss": 0.1742,
      "step": 255
    },
    {
      "epoch": 0.208,
      "grad_norm": 4.5625,
      "learning_rate": 9.953135638054502e-05,
      "loss": 0.3154,
      "step": 260
    },
    {
      "epoch": 0.212,
      "grad_norm": 3.65625,
      "learning_rate": 9.950165850682397e-05,
      "loss": 0.2545,
      "step": 265
    },
    {
      "epoch": 0.216,
      "grad_norm": 9.6875,
      "learning_rate": 9.947105309653313e-05,
      "loss": 0.3673,
      "step": 270
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.9375,
      "learning_rate": 9.943954071650085e-05,
      "loss": 0.2252,
      "step": 275
    },
    {
      "epoch": 0.224,
      "grad_norm": 6.34375,
      "learning_rate": 9.940712195035303e-05,
      "loss": 0.3008,
      "step": 280
    },
    {
      "epoch": 0.228,
      "grad_norm": 2.734375,
      "learning_rate": 9.937379739850233e-05,
      "loss": 0.1844,
      "step": 285
    },
    {
      "epoch": 0.232,
      "grad_norm": 3.5625,
      "learning_rate": 9.933956767813702e-05,
      "loss": 0.1963,
      "step": 290
    },
    {
      "epoch": 0.236,
      "grad_norm": 2.3125,
      "learning_rate": 9.93044334232096e-05,
      "loss": 0.2386,
      "step": 295
    },
    {
      "epoch": 0.24,
      "grad_norm": 4.0625,
      "learning_rate": 9.926839528442501e-05,
      "loss": 0.18,
      "step": 300
    },
    {
      "epoch": 0.244,
      "grad_norm": 3.609375,
      "learning_rate": 9.923145392922857e-05,
      "loss": 0.3359,
      "step": 305
    },
    {
      "epoch": 0.248,
      "grad_norm": 22.125,
      "learning_rate": 9.919361004179369e-05,
      "loss": 0.2545,
      "step": 310
    },
    {
      "epoch": 0.252,
      "grad_norm": 8.0,
      "learning_rate": 9.915486432300913e-05,
      "loss": 0.2692,
      "step": 315
    },
    {
      "epoch": 0.256,
      "grad_norm": 3.453125,
      "learning_rate": 9.911521749046603e-05,
      "loss": 0.3142,
      "step": 320
    },
    {
      "epoch": 0.26,
      "grad_norm": 4.65625,
      "learning_rate": 9.907467027844465e-05,
      "loss": 0.2871,
      "step": 325
    },
    {
      "epoch": 0.264,
      "grad_norm": 22.125,
      "learning_rate": 9.90332234379008e-05,
      "loss": 0.4123,
      "step": 330
    },
    {
      "epoch": 0.268,
      "grad_norm": 2.734375,
      "learning_rate": 9.899087773645176e-05,
      "loss": 0.2588,
      "step": 335
    },
    {
      "epoch": 0.272,
      "grad_norm": 2.921875,
      "learning_rate": 9.89476339583623e-05,
      "loss": 0.1691,
      "step": 340
    },
    {
      "epoch": 0.276,
      "grad_norm": 4.1875,
      "learning_rate": 9.890349290453003e-05,
      "loss": 0.6462,
      "step": 345
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.3125,
      "learning_rate": 9.88584553924705e-05,
      "loss": 0.2532,
      "step": 350
    },
    {
      "epoch": 0.284,
      "grad_norm": 2.015625,
      "learning_rate": 9.881252225630223e-05,
      "loss": 0.1776,
      "step": 355
    },
    {
      "epoch": 0.288,
      "grad_norm": 2.546875,
      "learning_rate": 9.876569434673115e-05,
      "loss": 0.2845,
      "step": 360
    },
    {
      "epoch": 0.292,
      "grad_norm": 4.15625,
      "learning_rate": 9.871797253103483e-05,
      "loss": 0.2373,
      "step": 365
    },
    {
      "epoch": 0.296,
      "grad_norm": 5.46875,
      "learning_rate": 9.866935769304647e-05,
      "loss": 0.6995,
      "step": 370
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.453125,
      "learning_rate": 9.861985073313854e-05,
      "loss": 0.2509,
      "step": 375
    },
    {
      "epoch": 0.304,
      "grad_norm": 4.21875,
      "learning_rate": 9.856945256820605e-05,
      "loss": 0.2091,
      "step": 380
    },
    {
      "epoch": 0.308,
      "grad_norm": 9.6875,
      "learning_rate": 9.85181641316496e-05,
      "loss": 0.2073,
      "step": 385
    },
    {
      "epoch": 0.312,
      "grad_norm": 6.5,
      "learning_rate": 9.846598637335811e-05,
      "loss": 0.229,
      "step": 390
    },
    {
      "epoch": 0.316,
      "grad_norm": 16.0,
      "learning_rate": 9.841292025969118e-05,
      "loss": 0.1623,
      "step": 395
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.0,
      "learning_rate": 9.835896677346124e-05,
      "loss": 0.1968,
      "step": 400
    },
    {
      "epoch": 0.324,
      "grad_norm": 9.0,
      "learning_rate": 9.830412691391532e-05,
      "loss": 0.2493,
      "step": 405
    },
    {
      "epoch": 0.328,
      "grad_norm": 6.0625,
      "learning_rate": 9.824840169671655e-05,
      "loss": 0.2011,
      "step": 410
    },
    {
      "epoch": 0.332,
      "grad_norm": 5.46875,
      "learning_rate": 9.819179215392536e-05,
      "loss": 0.2407,
      "step": 415
    },
    {
      "epoch": 0.336,
      "grad_norm": 11.875,
      "learning_rate": 9.813429933398037e-05,
      "loss": 0.4208,
      "step": 420
    },
    {
      "epoch": 0.34,
      "grad_norm": 14.375,
      "learning_rate": 9.80759243016789e-05,
      "loss": 0.2376,
      "step": 425
    },
    {
      "epoch": 0.344,
      "grad_norm": 8.625,
      "learning_rate": 9.801666813815737e-05,
      "loss": 0.2432,
      "step": 430
    },
    {
      "epoch": 0.348,
      "grad_norm": 2.15625,
      "learning_rate": 9.795653194087119e-05,
      "loss": 0.192,
      "step": 435
    },
    {
      "epoch": 0.352,
      "grad_norm": 1.96875,
      "learning_rate": 9.789551682357445e-05,
      "loss": 0.2133,
      "step": 440
    },
    {
      "epoch": 0.356,
      "grad_norm": 8.8125,
      "learning_rate": 9.783362391629933e-05,
      "loss": 0.2349,
      "step": 445
    },
    {
      "epoch": 0.36,
      "grad_norm": 2.78125,
      "learning_rate": 9.777085436533515e-05,
      "loss": 0.2599,
      "step": 450
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.921875,
      "learning_rate": 9.770720933320706e-05,
      "loss": 0.2073,
      "step": 455
    },
    {
      "epoch": 0.368,
      "grad_norm": 3.53125,
      "learning_rate": 9.764268999865467e-05,
      "loss": 0.2317,
      "step": 460
    },
    {
      "epoch": 0.372,
      "grad_norm": 10.625,
      "learning_rate": 9.757729755661012e-05,
      "loss": 0.2824,
      "step": 465
    },
    {
      "epoch": 0.376,
      "grad_norm": 8.5625,
      "learning_rate": 9.751103321817589e-05,
      "loss": 0.1577,
      "step": 470
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.453125,
      "learning_rate": 9.744389821060249e-05,
      "loss": 0.2222,
      "step": 475
    },
    {
      "epoch": 0.384,
      "grad_norm": 9.375,
      "learning_rate": 9.737589377726573e-05,
      "loss": 0.2174,
      "step": 480
    },
    {
      "epoch": 0.388,
      "grad_norm": 2.828125,
      "learning_rate": 9.730702117764355e-05,
      "loss": 0.2527,
      "step": 485
    },
    {
      "epoch": 0.392,
      "grad_norm": 4.75,
      "learning_rate": 9.723728168729288e-05,
      "loss": 0.2732,
      "step": 490
    },
    {
      "epoch": 0.396,
      "grad_norm": 12.125,
      "learning_rate": 9.716667659782583e-05,
      "loss": 0.255,
      "step": 495
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.15625,
      "learning_rate": 9.709520721688593e-05,
      "loss": 0.1869,
      "step": 500
    },
    {
      "epoch": 0.404,
      "grad_norm": 1.4609375,
      "learning_rate": 9.702287486812384e-05,
      "loss": 0.2911,
      "step": 505
    },
    {
      "epoch": 0.408,
      "grad_norm": 3.78125,
      "learning_rate": 9.694968089117277e-05,
      "loss": 0.2023,
      "step": 510
    },
    {
      "epoch": 0.412,
      "grad_norm": 5.46875,
      "learning_rate": 9.687562664162387e-05,
      "loss": 0.1832,
      "step": 515
    },
    {
      "epoch": 0.416,
      "grad_norm": 6.75,
      "learning_rate": 9.680071349100085e-05,
      "loss": 0.2707,
      "step": 520
    },
    {
      "epoch": 0.42,
      "grad_norm": 6.0,
      "learning_rate": 9.672494282673479e-05,
      "loss": 0.4463,
      "step": 525
    },
    {
      "epoch": 0.424,
      "grad_norm": 2.15625,
      "learning_rate": 9.664831605213843e-05,
      "loss": 0.2574,
      "step": 530
    },
    {
      "epoch": 0.428,
      "grad_norm": 8.6875,
      "learning_rate": 9.657083458637998e-05,
      "loss": 0.3508,
      "step": 535
    },
    {
      "epoch": 0.432,
      "grad_norm": 4.625,
      "learning_rate": 9.649249986445713e-05,
      "loss": 0.2382,
      "step": 540
    },
    {
      "epoch": 0.436,
      "grad_norm": 2.1875,
      "learning_rate": 9.641331333717023e-05,
      "loss": 0.3205,
      "step": 545
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.25,
      "learning_rate": 9.633327647109558e-05,
      "loss": 0.266,
      "step": 550
    },
    {
      "epoch": 0.444,
      "grad_norm": 6.3125,
      "learning_rate": 9.625239074855813e-05,
      "loss": 0.283,
      "step": 555
    },
    {
      "epoch": 0.448,
      "grad_norm": 6.28125,
      "learning_rate": 9.617065766760414e-05,
      "loss": 0.2706,
      "step": 560
    },
    {
      "epoch": 0.452,
      "grad_norm": 6.375,
      "learning_rate": 9.608807874197342e-05,
      "loss": 0.2414,
      "step": 565
    },
    {
      "epoch": 0.456,
      "grad_norm": 5.71875,
      "learning_rate": 9.600465550107121e-05,
      "loss": 0.3608,
      "step": 570
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.125,
      "learning_rate": 9.592038948993999e-05,
      "loss": 0.1853,
      "step": 575
    },
    {
      "epoch": 0.464,
      "grad_norm": 3.3125,
      "learning_rate": 9.583528226923071e-05,
      "loss": 0.202,
      "step": 580
    },
    {
      "epoch": 0.468,
      "grad_norm": 3.234375,
      "learning_rate": 9.574933541517402e-05,
      "loss": 0.2194,
      "step": 585
    },
    {
      "epoch": 0.472,
      "grad_norm": 6.8125,
      "learning_rate": 9.566255051955101e-05,
      "loss": 0.2598,
      "step": 590
    },
    {
      "epoch": 0.476,
      "grad_norm": 6.5625,
      "learning_rate": 9.55749291896637e-05,
      "loss": 0.145,
      "step": 595
    },
    {
      "epoch": 0.48,
      "grad_norm": 4.53125,
      "learning_rate": 9.54864730483054e-05,
      "loss": 0.2518,
      "step": 600
    },
    {
      "epoch": 0.484,
      "grad_norm": 17.125,
      "learning_rate": 9.53971837337305e-05,
      "loss": 0.2188,
      "step": 605
    },
    {
      "epoch": 0.488,
      "grad_norm": 6.3125,
      "learning_rate": 9.530706289962418e-05,
      "loss": 0.2255,
      "step": 610
    },
    {
      "epoch": 0.492,
      "grad_norm": 3.9375,
      "learning_rate": 9.521611221507189e-05,
      "loss": 0.271,
      "step": 615
    },
    {
      "epoch": 0.496,
      "grad_norm": 5.5,
      "learning_rate": 9.512433336452828e-05,
      "loss": 0.2186,
      "step": 620
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.1875,
      "learning_rate": 9.503172804778604e-05,
      "loss": 0.2948,
      "step": 625
    },
    {
      "epoch": 0.504,
      "grad_norm": 5.90625,
      "learning_rate": 9.493829797994458e-05,
      "loss": 0.3803,
      "step": 630
    },
    {
      "epoch": 0.508,
      "grad_norm": 7.09375,
      "learning_rate": 9.4844044891378e-05,
      "loss": 0.2141,
      "step": 635
    },
    {
      "epoch": 0.512,
      "grad_norm": 6.5,
      "learning_rate": 9.474897052770322e-05,
      "loss": 0.2283,
      "step": 640
    },
    {
      "epoch": 0.516,
      "grad_norm": 2.09375,
      "learning_rate": 9.465307664974766e-05,
      "loss": 0.2791,
      "step": 645
    },
    {
      "epoch": 0.52,
      "grad_norm": 2.8125,
      "learning_rate": 9.45563650335165e-05,
      "loss": 0.2621,
      "step": 650
    },
    {
      "epoch": 0.524,
      "grad_norm": 6.125,
      "learning_rate": 9.445883747015992e-05,
      "loss": 0.3292,
      "step": 655
    },
    {
      "epoch": 0.528,
      "grad_norm": 5.0,
      "learning_rate": 9.436049576593984e-05,
      "loss": 0.3498,
      "step": 660
    },
    {
      "epoch": 0.532,
      "grad_norm": 1.359375,
      "learning_rate": 9.42613417421965e-05,
      "loss": 0.1882,
      "step": 665
    },
    {
      "epoch": 0.536,
      "grad_norm": 2.78125,
      "learning_rate": 9.416137723531476e-05,
      "loss": 0.3146,
      "step": 670
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.53125,
      "learning_rate": 9.406060409669002e-05,
      "loss": 0.1735,
      "step": 675
    },
    {
      "epoch": 0.544,
      "grad_norm": 9.3125,
      "learning_rate": 9.395902419269396e-05,
      "loss": 0.4134,
      "step": 680
    },
    {
      "epoch": 0.548,
      "grad_norm": 2.0,
      "learning_rate": 9.385663940464004e-05,
      "loss": 0.3328,
      "step": 685
    },
    {
      "epoch": 0.552,
      "grad_norm": 5.3125,
      "learning_rate": 9.37534516287485e-05,
      "loss": 0.2154,
      "step": 690
    },
    {
      "epoch": 0.556,
      "grad_norm": 5.0625,
      "learning_rate": 9.364946277611144e-05,
      "loss": 0.2152,
      "step": 695
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.625,
      "learning_rate": 9.354467477265725e-05,
      "loss": 0.1554,
      "step": 700
    },
    {
      "epoch": 0.564,
      "grad_norm": 3.3125,
      "learning_rate": 9.343908955911506e-05,
      "loss": 0.2418,
      "step": 705
    },
    {
      "epoch": 0.568,
      "grad_norm": 9.5,
      "learning_rate": 9.33327090909787e-05,
      "loss": 0.2844,
      "step": 710
    },
    {
      "epoch": 0.572,
      "grad_norm": 5.5,
      "learning_rate": 9.322553533847062e-05,
      "loss": 0.2158,
      "step": 715
    },
    {
      "epoch": 0.576,
      "grad_norm": 11.5,
      "learning_rate": 9.311757028650522e-05,
      "loss": 0.2493,
      "step": 720
    },
    {
      "epoch": 0.58,
      "grad_norm": 14.6875,
      "learning_rate": 9.300881593465226e-05,
      "loss": 0.2881,
      "step": 725
    },
    {
      "epoch": 0.584,
      "grad_norm": 2.953125,
      "learning_rate": 9.289927429709967e-05,
      "loss": 0.212,
      "step": 730
    },
    {
      "epoch": 0.588,
      "grad_norm": 2.40625,
      "learning_rate": 9.278894740261643e-05,
      "loss": 0.1999,
      "step": 735
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.640625,
      "learning_rate": 9.267783729451482e-05,
      "loss": 0.1648,
      "step": 740
    },
    {
      "epoch": 0.596,
      "grad_norm": 4.53125,
      "learning_rate": 9.256594603061263e-05,
      "loss": 0.2467,
      "step": 745
    },
    {
      "epoch": 0.6,
      "grad_norm": 6.6875,
      "learning_rate": 9.245327568319515e-05,
      "loss": 0.3942,
      "step": 750
    },
    {
      "epoch": 0.604,
      "grad_norm": 11.0625,
      "learning_rate": 9.233982833897663e-05,
      "loss": 0.2589,
      "step": 755
    },
    {
      "epoch": 0.608,
      "grad_norm": 4.71875,
      "learning_rate": 9.222560609906176e-05,
      "loss": 0.1867,
      "step": 760
    },
    {
      "epoch": 0.612,
      "grad_norm": 2.453125,
      "learning_rate": 9.211061107890667e-05,
      "loss": 0.1956,
      "step": 765
    },
    {
      "epoch": 0.616,
      "grad_norm": 5.5,
      "learning_rate": 9.199484540827984e-05,
      "loss": 0.1597,
      "step": 770
    },
    {
      "epoch": 0.62,
      "grad_norm": 4.03125,
      "learning_rate": 9.187831123122252e-05,
      "loss": 0.3174,
      "step": 775
    },
    {
      "epoch": 0.624,
      "grad_norm": 4.5625,
      "learning_rate": 9.176101070600922e-05,
      "loss": 0.223,
      "step": 780
    },
    {
      "epoch": 0.628,
      "grad_norm": 9.1875,
      "learning_rate": 9.164294600510749e-05,
      "loss": 0.3052,
      "step": 785
    },
    {
      "epoch": 0.632,
      "grad_norm": 2.53125,
      "learning_rate": 9.152411931513793e-05,
      "loss": 0.3569,
      "step": 790
    },
    {
      "epoch": 0.636,
      "grad_norm": 1.703125,
      "learning_rate": 9.140453283683353e-05,
      "loss": 0.2379,
      "step": 795
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.4375,
      "learning_rate": 9.128418878499893e-05,
      "loss": 0.2192,
      "step": 800
    },
    {
      "epoch": 0.644,
      "grad_norm": 4.59375,
      "learning_rate": 9.116308938846951e-05,
      "loss": 0.3337,
      "step": 805
    },
    {
      "epoch": 0.648,
      "grad_norm": 3.734375,
      "learning_rate": 9.104123689006996e-05,
      "loss": 0.1982,
      "step": 810
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.5234375,
      "learning_rate": 9.091863354657284e-05,
      "loss": 0.1729,
      "step": 815
    },
    {
      "epoch": 0.656,
      "grad_norm": 4.09375,
      "learning_rate": 9.07952816286568e-05,
      "loss": 0.255,
      "step": 820
    },
    {
      "epoch": 0.66,
      "grad_norm": 6.34375,
      "learning_rate": 9.067118342086437e-05,
      "loss": 0.2031,
      "step": 825
    },
    {
      "epoch": 0.664,
      "grad_norm": 3.71875,
      "learning_rate": 9.05463412215599e-05,
      "loss": 0.2625,
      "step": 830
    },
    {
      "epoch": 0.668,
      "grad_norm": 3.359375,
      "learning_rate": 9.042075734288676e-05,
      "loss": 0.2042,
      "step": 835
    },
    {
      "epoch": 0.672,
      "grad_norm": 3.03125,
      "learning_rate": 9.029443411072466e-05,
      "loss": 0.2612,
      "step": 840
    },
    {
      "epoch": 0.676,
      "grad_norm": 4.75,
      "learning_rate": 9.01673738646465e-05,
      "loss": 0.314,
      "step": 845
    },
    {
      "epoch": 0.68,
      "grad_norm": 2.40625,
      "learning_rate": 9.003957895787507e-05,
      "loss": 0.2208,
      "step": 850
    },
    {
      "epoch": 0.684,
      "grad_norm": 8.75,
      "learning_rate": 8.991105175723947e-05,
      "loss": 0.205,
      "step": 855
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.93359375,
      "learning_rate": 8.97817946431313e-05,
      "loss": 0.2096,
      "step": 860
    },
    {
      "epoch": 0.692,
      "grad_norm": 2.078125,
      "learning_rate": 8.965181000946048e-05,
      "loss": 0.269,
      "step": 865
    },
    {
      "epoch": 0.696,
      "grad_norm": 10.5625,
      "learning_rate": 8.952110026361101e-05,
      "loss": 0.2153,
      "step": 870
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.6875,
      "learning_rate": 8.938966782639636e-05,
      "loss": 0.1852,
      "step": 875
    },
    {
      "epoch": 0.704,
      "grad_norm": 3.984375,
      "learning_rate": 8.925751513201462e-05,
      "loss": 0.1756,
      "step": 880
    },
    {
      "epoch": 0.708,
      "grad_norm": 2.375,
      "learning_rate": 8.91246446280034e-05,
      "loss": 0.2608,
      "step": 885
    },
    {
      "epoch": 0.712,
      "grad_norm": 4.53125,
      "learning_rate": 8.899105877519454e-05,
      "loss": 0.2052,
      "step": 890
    },
    {
      "epoch": 0.716,
      "grad_norm": 3.109375,
      "learning_rate": 8.885676004766851e-05,
      "loss": 0.3208,
      "step": 895
    },
    {
      "epoch": 0.72,
      "grad_norm": 4.1875,
      "learning_rate": 8.872175093270864e-05,
      "loss": 0.1732,
      "step": 900
    },
    {
      "epoch": 0.724,
      "grad_norm": 6.71875,
      "learning_rate": 8.858603393075494e-05,
      "loss": 0.3626,
      "step": 905
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.66015625,
      "learning_rate": 8.844961155535789e-05,
      "loss": 0.1916,
      "step": 910
    },
    {
      "epoch": 0.732,
      "grad_norm": 2.546875,
      "learning_rate": 8.831248633313186e-05,
      "loss": 0.2105,
      "step": 915
    },
    {
      "epoch": 0.736,
      "grad_norm": 6.09375,
      "learning_rate": 8.817466080370834e-05,
      "loss": 0.2288,
      "step": 920
    },
    {
      "epoch": 0.74,
      "grad_norm": 12.9375,
      "learning_rate": 8.803613751968886e-05,
      "loss": 0.331,
      "step": 925
    },
    {
      "epoch": 0.744,
      "grad_norm": 7.25,
      "learning_rate": 8.789691904659768e-05,
      "loss": 0.3312,
      "step": 930
    },
    {
      "epoch": 0.748,
      "grad_norm": 8.1875,
      "learning_rate": 8.775700796283442e-05,
      "loss": 0.2068,
      "step": 935
    },
    {
      "epoch": 0.752,
      "grad_norm": 10.5,
      "learning_rate": 8.761640685962615e-05,
      "loss": 0.2967,
      "step": 940
    },
    {
      "epoch": 0.756,
      "grad_norm": 4.28125,
      "learning_rate": 8.747511834097948e-05,
      "loss": 0.3089,
      "step": 945
    },
    {
      "epoch": 0.76,
      "grad_norm": 4.53125,
      "learning_rate": 8.733314502363236e-05,
      "loss": 0.2597,
      "step": 950
    },
    {
      "epoch": 0.764,
      "grad_norm": 10.5625,
      "learning_rate": 8.719048953700551e-05,
      "loss": 0.3472,
      "step": 955
    },
    {
      "epoch": 0.768,
      "grad_norm": 8.8125,
      "learning_rate": 8.704715452315379e-05,
      "loss": 0.2246,
      "step": 960
    },
    {
      "epoch": 0.772,
      "grad_norm": 18.5,
      "learning_rate": 8.690314263671729e-05,
      "loss": 0.3343,
      "step": 965
    },
    {
      "epoch": 0.776,
      "grad_norm": 16.25,
      "learning_rate": 8.675845654487216e-05,
      "loss": 0.3376,
      "step": 970
    },
    {
      "epoch": 0.78,
      "grad_norm": 6.46875,
      "learning_rate": 8.661309892728114e-05,
      "loss": 0.1762,
      "step": 975
    },
    {
      "epoch": 0.784,
      "grad_norm": 5.84375,
      "learning_rate": 8.6467072476044e-05,
      "loss": 0.2071,
      "step": 980
    },
    {
      "epoch": 0.788,
      "grad_norm": 9.6875,
      "learning_rate": 8.632037989564767e-05,
      "loss": 0.1698,
      "step": 985
    },
    {
      "epoch": 0.792,
      "grad_norm": 9.4375,
      "learning_rate": 8.617302390291612e-05,
      "loss": 0.2083,
      "step": 990
    },
    {
      "epoch": 0.796,
      "grad_norm": 5.21875,
      "learning_rate": 8.602500722696012e-05,
      "loss": 0.1968,
      "step": 995
    },
    {
      "epoch": 0.8,
      "grad_norm": 8.375,
      "learning_rate": 8.587633260912658e-05,
      "loss": 0.3003,
      "step": 1000
    },
    {
      "epoch": 0.804,
      "grad_norm": 10.75,
      "learning_rate": 8.572700280294786e-05,
      "loss": 0.2759,
      "step": 1005
    },
    {
      "epoch": 0.808,
      "grad_norm": 10.4375,
      "learning_rate": 8.557702057409077e-05,
      "loss": 0.1903,
      "step": 1010
    },
    {
      "epoch": 0.812,
      "grad_norm": 3.078125,
      "learning_rate": 8.542638870030532e-05,
      "loss": 0.2064,
      "step": 1015
    },
    {
      "epoch": 0.816,
      "grad_norm": 8.875,
      "learning_rate": 8.527510997137327e-05,
      "loss": 0.2949,
      "step": 1020
    },
    {
      "epoch": 0.82,
      "grad_norm": 4.4375,
      "learning_rate": 8.512318718905652e-05,
      "loss": 0.2654,
      "step": 1025
    },
    {
      "epoch": 0.824,
      "grad_norm": 4.90625,
      "learning_rate": 8.497062316704518e-05,
      "loss": 0.3562,
      "step": 1030
    },
    {
      "epoch": 0.828,
      "grad_norm": 3.5625,
      "learning_rate": 8.481742073090538e-05,
      "loss": 0.1737,
      "step": 1035
    },
    {
      "epoch": 0.832,
      "grad_norm": 7.90625,
      "learning_rate": 8.466358271802711e-05,
      "loss": 0.2945,
      "step": 1040
    },
    {
      "epoch": 0.836,
      "grad_norm": 4.28125,
      "learning_rate": 8.450911197757152e-05,
      "loss": 0.3186,
      "step": 1045
    },
    {
      "epoch": 0.84,
      "grad_norm": 6.59375,
      "learning_rate": 8.435401137041824e-05,
      "loss": 0.1641,
      "step": 1050
    },
    {
      "epoch": 0.844,
      "grad_norm": 1.015625,
      "learning_rate": 8.41982837691123e-05,
      "loss": 0.2608,
      "step": 1055
    },
    {
      "epoch": 0.848,
      "grad_norm": 6.34375,
      "learning_rate": 8.404193205781112e-05,
      "loss": 0.2191,
      "step": 1060
    },
    {
      "epoch": 0.852,
      "grad_norm": 7.8125,
      "learning_rate": 8.38849591322308e-05,
      "loss": 0.2822,
      "step": 1065
    },
    {
      "epoch": 0.856,
      "grad_norm": 2.875,
      "learning_rate": 8.372736789959281e-05,
      "loss": 0.1995,
      "step": 1070
    },
    {
      "epoch": 0.86,
      "grad_norm": 7.84375,
      "learning_rate": 8.356916127856991e-05,
      "loss": 0.2957,
      "step": 1075
    },
    {
      "epoch": 0.864,
      "grad_norm": 4.3125,
      "learning_rate": 8.34103421992322e-05,
      "loss": 0.1796,
      "step": 1080
    },
    {
      "epoch": 0.868,
      "grad_norm": 6.78125,
      "learning_rate": 8.325091360299278e-05,
      "loss": 0.2447,
      "step": 1085
    },
    {
      "epoch": 0.872,
      "grad_norm": 11.75,
      "learning_rate": 8.309087844255342e-05,
      "loss": 0.2302,
      "step": 1090
    },
    {
      "epoch": 0.876,
      "grad_norm": 6.71875,
      "learning_rate": 8.293023968184972e-05,
      "loss": 0.2064,
      "step": 1095
    },
    {
      "epoch": 0.88,
      "grad_norm": 2.75,
      "learning_rate": 8.276900029599627e-05,
      "loss": 0.2153,
      "step": 1100
    },
    {
      "epoch": 0.884,
      "grad_norm": 14.1875,
      "learning_rate": 8.26071632712316e-05,
      "loss": 0.2526,
      "step": 1105
    },
    {
      "epoch": 0.888,
      "grad_norm": 4.09375,
      "learning_rate": 8.244473160486275e-05,
      "loss": 0.2503,
      "step": 1110
    },
    {
      "epoch": 0.892,
      "grad_norm": 3.140625,
      "learning_rate": 8.228170830520989e-05,
      "loss": 0.2594,
      "step": 1115
    },
    {
      "epoch": 0.896,
      "grad_norm": 5.3125,
      "learning_rate": 8.21180963915506e-05,
      "loss": 0.2233,
      "step": 1120
    },
    {
      "epoch": 0.9,
      "grad_norm": 5.6875,
      "learning_rate": 8.195389889406379e-05,
      "loss": 0.2687,
      "step": 1125
    },
    {
      "epoch": 0.904,
      "grad_norm": 1.9375,
      "learning_rate": 8.178911885377376e-05,
      "loss": 0.2107,
      "step": 1130
    },
    {
      "epoch": 0.908,
      "grad_norm": 3.328125,
      "learning_rate": 8.162375932249381e-05,
      "loss": 0.2649,
      "step": 1135
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.4921875,
      "learning_rate": 8.14578233627697e-05,
      "loss": 0.1771,
      "step": 1140
    },
    {
      "epoch": 0.916,
      "grad_norm": 3.671875,
      "learning_rate": 8.129131404782295e-05,
      "loss": 0.2173,
      "step": 1145
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.8125,
      "learning_rate": 8.112423446149393e-05,
      "loss": 0.2796,
      "step": 1150
    },
    {
      "epoch": 0.924,
      "grad_norm": 5.75,
      "learning_rate": 8.095658769818473e-05,
      "loss": 0.2349,
      "step": 1155
    },
    {
      "epoch": 0.928,
      "grad_norm": 6.90625,
      "learning_rate": 8.078837686280186e-05,
      "loss": 0.2175,
      "step": 1160
    },
    {
      "epoch": 0.932,
      "grad_norm": 5.84375,
      "learning_rate": 8.061960507069874e-05,
      "loss": 0.1354,
      "step": 1165
    },
    {
      "epoch": 0.936,
      "grad_norm": 2.609375,
      "learning_rate": 8.045027544761803e-05,
      "loss": 0.269,
      "step": 1170
    },
    {
      "epoch": 0.94,
      "grad_norm": 4.3125,
      "learning_rate": 8.028039112963366e-05,
      "loss": 0.2465,
      "step": 1175
    },
    {
      "epoch": 0.944,
      "grad_norm": 3.734375,
      "learning_rate": 8.010995526309287e-05,
      "loss": 0.2158,
      "step": 1180
    },
    {
      "epoch": 0.948,
      "grad_norm": 4.34375,
      "learning_rate": 7.99389710045578e-05,
      "loss": 0.2117,
      "step": 1185
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.60546875,
      "learning_rate": 7.976744152074712e-05,
      "loss": 0.2254,
      "step": 1190
    },
    {
      "epoch": 0.956,
      "grad_norm": 10.9375,
      "learning_rate": 7.959536998847742e-05,
      "loss": 0.2781,
      "step": 1195
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.1875,
      "learning_rate": 7.942275959460424e-05,
      "loss": 0.3292,
      "step": 1200
    },
    {
      "epoch": 0.964,
      "grad_norm": 1.4453125,
      "learning_rate": 7.924961353596315e-05,
      "loss": 0.1948,
      "step": 1205
    },
    {
      "epoch": 0.968,
      "grad_norm": 6.8125,
      "learning_rate": 7.90759350193105e-05,
      "loss": 0.2398,
      "step": 1210
    },
    {
      "epoch": 0.972,
      "grad_norm": 8.125,
      "learning_rate": 7.890172726126409e-05,
      "loss": 0.2422,
      "step": 1215
    },
    {
      "epoch": 0.976,
      "grad_norm": 2.265625,
      "learning_rate": 7.872699348824347e-05,
      "loss": 0.1944,
      "step": 1220
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.1875,
      "learning_rate": 7.855173693641032e-05,
      "loss": 0.2578,
      "step": 1225
    },
    {
      "epoch": 0.984,
      "grad_norm": 10.5,
      "learning_rate": 7.837596085160846e-05,
      "loss": 0.2683,
      "step": 1230
    },
    {
      "epoch": 0.988,
      "grad_norm": 1.4296875,
      "learning_rate": 7.819966848930375e-05,
      "loss": 0.3894,
      "step": 1235
    },
    {
      "epoch": 0.992,
      "grad_norm": 16.5,
      "learning_rate": 7.80228631145237e-05,
      "loss": 0.2195,
      "step": 1240
    },
    {
      "epoch": 0.996,
      "grad_norm": 2.84375,
      "learning_rate": 7.784554800179719e-05,
      "loss": 0.27,
      "step": 1245
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.546875,
      "learning_rate": 7.76677264350936e-05,
      "loss": 0.1809,
      "step": 1250
    },
    {
      "epoch": 1.004,
      "grad_norm": 16.0,
      "learning_rate": 7.748940170776223e-05,
      "loss": 0.274,
      "step": 1255
    },
    {
      "epoch": 1.008,
      "grad_norm": 1.8515625,
      "learning_rate": 7.731057712247109e-05,
      "loss": 0.2227,
      "step": 1260
    },
    {
      "epoch": 1.012,
      "grad_norm": 2.296875,
      "learning_rate": 7.713125599114583e-05,
      "loss": 0.2452,
      "step": 1265
    },
    {
      "epoch": 1.016,
      "grad_norm": 2.234375,
      "learning_rate": 7.695144163490847e-05,
      "loss": 0.228,
      "step": 1270
    },
    {
      "epoch": 1.02,
      "grad_norm": 7.25,
      "learning_rate": 7.67711373840157e-05,
      "loss": 0.2814,
      "step": 1275
    },
    {
      "epoch": 1.024,
      "grad_norm": 10.625,
      "learning_rate": 7.659034657779742e-05,
      "loss": 0.2426,
      "step": 1280
    },
    {
      "epoch": 1.028,
      "grad_norm": 6.53125,
      "learning_rate": 7.640907256459475e-05,
      "loss": 0.3028,
      "step": 1285
    },
    {
      "epoch": 1.032,
      "grad_norm": 9.9375,
      "learning_rate": 7.622731870169804e-05,
      "loss": 0.2303,
      "step": 1290
    },
    {
      "epoch": 1.036,
      "grad_norm": 1.3984375,
      "learning_rate": 7.604508835528473e-05,
      "loss": 0.2146,
      "step": 1295
    },
    {
      "epoch": 1.04,
      "grad_norm": 9.75,
      "learning_rate": 7.586238490035697e-05,
      "loss": 0.2326,
      "step": 1300
    },
    {
      "epoch": 1.044,
      "grad_norm": 9.0,
      "learning_rate": 7.567921172067917e-05,
      "loss": 0.2065,
      "step": 1305
    },
    {
      "epoch": 1.048,
      "grad_norm": 2.625,
      "learning_rate": 7.549557220871526e-05,
      "loss": 0.2621,
      "step": 1310
    },
    {
      "epoch": 1.052,
      "grad_norm": 8.375,
      "learning_rate": 7.53114697655659e-05,
      "loss": 0.2799,
      "step": 1315
    },
    {
      "epoch": 1.056,
      "grad_norm": 2.765625,
      "learning_rate": 7.512690780090545e-05,
      "loss": 0.28,
      "step": 1320
    },
    {
      "epoch": 1.06,
      "grad_norm": 5.4375,
      "learning_rate": 7.494188973291888e-05,
      "loss": 0.1886,
      "step": 1325
    },
    {
      "epoch": 1.064,
      "grad_norm": 1.671875,
      "learning_rate": 7.475641898823843e-05,
      "loss": 0.2508,
      "step": 1330
    },
    {
      "epoch": 1.068,
      "grad_norm": 16.625,
      "learning_rate": 7.457049900188017e-05,
      "loss": 0.2467,
      "step": 1335
    },
    {
      "epoch": 1.072,
      "grad_norm": 4.96875,
      "learning_rate": 7.438413321718031e-05,
      "loss": 0.2887,
      "step": 1340
    },
    {
      "epoch": 1.076,
      "grad_norm": 5.59375,
      "learning_rate": 7.419732508573154e-05,
      "loss": 0.1902,
      "step": 1345
    },
    {
      "epoch": 1.08,
      "grad_norm": 6.53125,
      "learning_rate": 7.4010078067319e-05,
      "loss": 0.4433,
      "step": 1350
    },
    {
      "epoch": 1.084,
      "grad_norm": 7.6875,
      "learning_rate": 7.382239562985627e-05,
      "loss": 0.3347,
      "step": 1355
    },
    {
      "epoch": 1.088,
      "grad_norm": 8.4375,
      "learning_rate": 7.363428124932112e-05,
      "loss": 0.231,
      "step": 1360
    },
    {
      "epoch": 1.092,
      "grad_norm": 3.875,
      "learning_rate": 7.344573840969113e-05,
      "loss": 0.2181,
      "step": 1365
    },
    {
      "epoch": 1.096,
      "grad_norm": 10.75,
      "learning_rate": 7.325677060287916e-05,
      "loss": 0.2058,
      "step": 1370
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.578125,
      "learning_rate": 7.306738132866871e-05,
      "loss": 0.2332,
      "step": 1375
    },
    {
      "epoch": 1.104,
      "grad_norm": 11.625,
      "learning_rate": 7.287757409464907e-05,
      "loss": 0.2881,
      "step": 1380
    },
    {
      "epoch": 1.108,
      "grad_norm": 4.90625,
      "learning_rate": 7.268735241615033e-05,
      "loss": 0.2106,
      "step": 1385
    },
    {
      "epoch": 1.112,
      "grad_norm": 2.375,
      "learning_rate": 7.249671981617837e-05,
      "loss": 0.1391,
      "step": 1390
    },
    {
      "epoch": 1.116,
      "grad_norm": 11.125,
      "learning_rate": 7.23056798253495e-05,
      "loss": 0.1426,
      "step": 1395
    },
    {
      "epoch": 1.12,
      "grad_norm": 4.6875,
      "learning_rate": 7.211423598182518e-05,
      "loss": 0.2658,
      "step": 1400
    },
    {
      "epoch": 1.124,
      "grad_norm": 17.0,
      "learning_rate": 7.192239183124634e-05,
      "loss": 0.2405,
      "step": 1405
    },
    {
      "epoch": 1.1280000000000001,
      "grad_norm": 2.390625,
      "learning_rate": 7.173015092666794e-05,
      "loss": 0.1683,
      "step": 1410
    },
    {
      "epoch": 1.1320000000000001,
      "grad_norm": 9.3125,
      "learning_rate": 7.153751682849293e-05,
      "loss": 0.217,
      "step": 1415
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 2.96875,
      "learning_rate": 7.134449310440643e-05,
      "loss": 0.189,
      "step": 1420
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 2.671875,
      "learning_rate": 7.115108332930969e-05,
      "loss": 0.3817,
      "step": 1425
    },
    {
      "epoch": 1.144,
      "grad_norm": 2.609375,
      "learning_rate": 7.095729108525377e-05,
      "loss": 0.2601,
      "step": 1430
    },
    {
      "epoch": 1.148,
      "grad_norm": 8.875,
      "learning_rate": 7.076311996137327e-05,
      "loss": 0.1535,
      "step": 1435
    },
    {
      "epoch": 1.152,
      "grad_norm": 1.5625,
      "learning_rate": 7.056857355381987e-05,
      "loss": 0.1995,
      "step": 1440
    },
    {
      "epoch": 1.156,
      "grad_norm": 2.90625,
      "learning_rate": 7.037365546569566e-05,
      "loss": 0.2107,
      "step": 1445
    },
    {
      "epoch": 1.16,
      "grad_norm": 4.625,
      "learning_rate": 7.017836930698654e-05,
      "loss": 0.3804,
      "step": 1450
    },
    {
      "epoch": 1.164,
      "grad_norm": 12.4375,
      "learning_rate": 6.99827186944951e-05,
      "loss": 0.2097,
      "step": 1455
    },
    {
      "epoch": 1.168,
      "grad_norm": 5.46875,
      "learning_rate": 6.9786707251774e-05,
      "loss": 0.2312,
      "step": 1460
    },
    {
      "epoch": 1.172,
      "grad_norm": 4.78125,
      "learning_rate": 6.959033860905856e-05,
      "loss": 0.2914,
      "step": 1465
    },
    {
      "epoch": 1.176,
      "grad_norm": 2.234375,
      "learning_rate": 6.939361640319962e-05,
      "loss": 0.2099,
      "step": 1470
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.4296875,
      "learning_rate": 6.919654427759627e-05,
      "loss": 0.2169,
      "step": 1475
    },
    {
      "epoch": 1.184,
      "grad_norm": 3.171875,
      "learning_rate": 6.899912588212822e-05,
      "loss": 0.2766,
      "step": 1480
    },
    {
      "epoch": 1.188,
      "grad_norm": 1.1875,
      "learning_rate": 6.880136487308832e-05,
      "loss": 0.1754,
      "step": 1485
    },
    {
      "epoch": 1.192,
      "grad_norm": 1.5078125,
      "learning_rate": 6.860326491311484e-05,
      "loss": 0.2075,
      "step": 1490
    },
    {
      "epoch": 1.196,
      "grad_norm": 3.015625,
      "learning_rate": 6.840482967112349e-05,
      "loss": 0.1331,
      "step": 1495
    },
    {
      "epoch": 1.2,
      "grad_norm": 11.1875,
      "learning_rate": 6.820606282223974e-05,
      "loss": 0.2355,
      "step": 1500
    },
    {
      "epoch": 1.204,
      "grad_norm": 6.59375,
      "learning_rate": 6.80069680477304e-05,
      "loss": 0.176,
      "step": 1505
    },
    {
      "epoch": 1.208,
      "grad_norm": 5.625,
      "learning_rate": 6.780754903493581e-05,
      "loss": 0.2066,
      "step": 1510
    },
    {
      "epoch": 1.212,
      "grad_norm": 2.265625,
      "learning_rate": 6.76078094772013e-05,
      "loss": 0.2216,
      "step": 1515
    },
    {
      "epoch": 1.216,
      "grad_norm": 8.1875,
      "learning_rate": 6.740775307380885e-05,
      "loss": 0.196,
      "step": 1520
    },
    {
      "epoch": 1.22,
      "grad_norm": 1.875,
      "learning_rate": 6.72073835299086e-05,
      "loss": 0.2984,
      "step": 1525
    },
    {
      "epoch": 1.224,
      "grad_norm": 10.6875,
      "learning_rate": 6.700670455645027e-05,
      "loss": 0.2543,
      "step": 1530
    },
    {
      "epoch": 1.228,
      "grad_norm": 3.046875,
      "learning_rate": 6.680571987011425e-05,
      "loss": 0.1873,
      "step": 1535
    },
    {
      "epoch": 1.232,
      "grad_norm": 10.8125,
      "learning_rate": 6.660443319324304e-05,
      "loss": 0.2808,
      "step": 1540
    },
    {
      "epoch": 1.236,
      "grad_norm": 5.59375,
      "learning_rate": 6.640284825377204e-05,
      "loss": 0.2627,
      "step": 1545
    },
    {
      "epoch": 1.24,
      "grad_norm": 6.34375,
      "learning_rate": 6.620096878516076e-05,
      "loss": 0.2201,
      "step": 1550
    },
    {
      "epoch": 1.244,
      "grad_norm": 3.28125,
      "learning_rate": 6.599879852632342e-05,
      "loss": 0.2456,
      "step": 1555
    },
    {
      "epoch": 1.248,
      "grad_norm": 3.625,
      "learning_rate": 6.57963412215599e-05,
      "loss": 0.2369,
      "step": 1560
    },
    {
      "epoch": 1.252,
      "grad_norm": 2.984375,
      "learning_rate": 6.559360062048631e-05,
      "loss": 0.2565,
      "step": 1565
    },
    {
      "epoch": 1.256,
      "grad_norm": 5.3125,
      "learning_rate": 6.539058047796555e-05,
      "loss": 0.2659,
      "step": 1570
    },
    {
      "epoch": 1.26,
      "grad_norm": 4.0,
      "learning_rate": 6.518728455403777e-05,
      "loss": 0.1814,
      "step": 1575
    },
    {
      "epoch": 1.264,
      "grad_norm": 8.625,
      "learning_rate": 6.498371661385076e-05,
      "loss": 0.2128,
      "step": 1580
    },
    {
      "epoch": 1.268,
      "grad_norm": 3.703125,
      "learning_rate": 6.477988042759015e-05,
      "loss": 0.2713,
      "step": 1585
    },
    {
      "epoch": 1.272,
      "grad_norm": 9.5,
      "learning_rate": 6.45757797704097e-05,
      "loss": 0.2641,
      "step": 1590
    },
    {
      "epoch": 1.276,
      "grad_norm": 8.8125,
      "learning_rate": 6.437141842236121e-05,
      "loss": 0.2591,
      "step": 1595
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.859375,
      "learning_rate": 6.416680016832473e-05,
      "loss": 0.2093,
      "step": 1600
    },
    {
      "epoch": 1.284,
      "grad_norm": 8.9375,
      "learning_rate": 6.396192879793822e-05,
      "loss": 0.1974,
      "step": 1605
    },
    {
      "epoch": 1.288,
      "grad_norm": 4.3125,
      "learning_rate": 6.375680810552762e-05,
      "loss": 0.2671,
      "step": 1610
    },
    {
      "epoch": 1.292,
      "grad_norm": 5.5,
      "learning_rate": 6.355144189003636e-05,
      "loss": 0.1532,
      "step": 1615
    },
    {
      "epoch": 1.296,
      "grad_norm": 10.25,
      "learning_rate": 6.334583395495508e-05,
      "loss": 0.3238,
      "step": 1620
    },
    {
      "epoch": 1.3,
      "grad_norm": 13.9375,
      "learning_rate": 6.313998810825126e-05,
      "loss": 0.2038,
      "step": 1625
    },
    {
      "epoch": 1.304,
      "grad_norm": 4.71875,
      "learning_rate": 6.293390816229857e-05,
      "loss": 0.2973,
      "step": 1630
    },
    {
      "epoch": 1.308,
      "grad_norm": 5.15625,
      "learning_rate": 6.272759793380632e-05,
      "loss": 0.383,
      "step": 1635
    },
    {
      "epoch": 1.312,
      "grad_norm": 1.921875,
      "learning_rate": 6.252106124374882e-05,
      "loss": 0.2139,
      "step": 1640
    },
    {
      "epoch": 1.316,
      "grad_norm": 3.734375,
      "learning_rate": 6.231430191729454e-05,
      "loss": 0.1553,
      "step": 1645
    },
    {
      "epoch": 1.32,
      "grad_norm": 2.1875,
      "learning_rate": 6.210732378373528e-05,
      "loss": 0.1988,
      "step": 1650
    },
    {
      "epoch": 1.324,
      "grad_norm": 7.59375,
      "learning_rate": 6.19001306764153e-05,
      "loss": 0.246,
      "step": 1655
    },
    {
      "epoch": 1.328,
      "grad_norm": 8.375,
      "learning_rate": 6.169272643266026e-05,
      "loss": 0.1977,
      "step": 1660
    },
    {
      "epoch": 1.332,
      "grad_norm": 10.5625,
      "learning_rate": 6.148511489370619e-05,
      "loss": 0.25,
      "step": 1665
    },
    {
      "epoch": 1.336,
      "grad_norm": 4.1875,
      "learning_rate": 6.127729990462834e-05,
      "loss": 0.2309,
      "step": 1670
    },
    {
      "epoch": 1.34,
      "grad_norm": 8.6875,
      "learning_rate": 6.106928531426991e-05,
      "loss": 0.1464,
      "step": 1675
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 7.34375,
      "learning_rate": 6.086107497517091e-05,
      "loss": 0.2663,
      "step": 1680
    },
    {
      "epoch": 1.3479999999999999,
      "grad_norm": 3.5,
      "learning_rate": 6.065267274349666e-05,
      "loss": 0.2519,
      "step": 1685
    },
    {
      "epoch": 1.3519999999999999,
      "grad_norm": 3.671875,
      "learning_rate": 6.044408247896646e-05,
      "loss": 0.3303,
      "step": 1690
    },
    {
      "epoch": 1.3559999999999999,
      "grad_norm": 12.6875,
      "learning_rate": 6.023530804478203e-05,
      "loss": 0.2372,
      "step": 1695
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 6.46875,
      "learning_rate": 6.002635330755608e-05,
      "loss": 0.2537,
      "step": 1700
    },
    {
      "epoch": 1.3639999999999999,
      "grad_norm": 4.90625,
      "learning_rate": 5.981722213724059e-05,
      "loss": 0.2679,
      "step": 1705
    },
    {
      "epoch": 1.3679999999999999,
      "grad_norm": 3.921875,
      "learning_rate": 5.960791840705517e-05,
      "loss": 0.2491,
      "step": 1710
    },
    {
      "epoch": 1.3719999999999999,
      "grad_norm": 0.9609375,
      "learning_rate": 5.939844599341533e-05,
      "loss": 0.2479,
      "step": 1715
    },
    {
      "epoch": 1.376,
      "grad_norm": 3.796875,
      "learning_rate": 5.9188808775860706e-05,
      "loss": 0.1989,
      "step": 1720
    },
    {
      "epoch": 1.38,
      "grad_norm": 6.375,
      "learning_rate": 5.897901063698315e-05,
      "loss": 0.1695,
      "step": 1725
    },
    {
      "epoch": 1.384,
      "grad_norm": 1.03125,
      "learning_rate": 5.87690554623549e-05,
      "loss": 0.3052,
      "step": 1730
    },
    {
      "epoch": 1.388,
      "grad_norm": 4.375,
      "learning_rate": 5.855894714045659e-05,
      "loss": 0.2052,
      "step": 1735
    },
    {
      "epoch": 1.392,
      "grad_norm": 5.96875,
      "learning_rate": 5.834868956260516e-05,
      "loss": 0.2204,
      "step": 1740
    },
    {
      "epoch": 1.396,
      "grad_norm": 6.34375,
      "learning_rate": 5.8138286622881934e-05,
      "loss": 0.3653,
      "step": 1745
    },
    {
      "epoch": 1.4,
      "grad_norm": 4.90625,
      "learning_rate": 5.792774221806031e-05,
      "loss": 0.2078,
      "step": 1750
    },
    {
      "epoch": 1.404,
      "grad_norm": 4.6875,
      "learning_rate": 5.771706024753379e-05,
      "loss": 0.3274,
      "step": 1755
    },
    {
      "epoch": 1.408,
      "grad_norm": 7.5,
      "learning_rate": 5.750624461324364e-05,
      "loss": 0.2377,
      "step": 1760
    },
    {
      "epoch": 1.412,
      "grad_norm": 6.78125,
      "learning_rate": 5.7295299219606604e-05,
      "loss": 0.2139,
      "step": 1765
    },
    {
      "epoch": 1.416,
      "grad_norm": 4.5625,
      "learning_rate": 5.7084227973442706e-05,
      "loss": 0.224,
      "step": 1770
    },
    {
      "epoch": 1.42,
      "grad_norm": 3.4375,
      "learning_rate": 5.687303478390276e-05,
      "loss": 0.2305,
      "step": 1775
    },
    {
      "epoch": 1.424,
      "grad_norm": 3.40625,
      "learning_rate": 5.666172356239607e-05,
      "loss": 0.3676,
      "step": 1780
    },
    {
      "epoch": 1.428,
      "grad_norm": 10.4375,
      "learning_rate": 5.645029822251796e-05,
      "loss": 0.2504,
      "step": 1785
    },
    {
      "epoch": 1.432,
      "grad_norm": 8.8125,
      "learning_rate": 5.6238762679977285e-05,
      "loss": 0.327,
      "step": 1790
    },
    {
      "epoch": 1.436,
      "grad_norm": 0.9375,
      "learning_rate": 5.602712085252388e-05,
      "loss": 0.2593,
      "step": 1795
    },
    {
      "epoch": 1.44,
      "grad_norm": 6.09375,
      "learning_rate": 5.581537665987607e-05,
      "loss": 0.2903,
      "step": 1800
    },
    {
      "epoch": 1.444,
      "grad_norm": 2.9375,
      "learning_rate": 5.560353402364803e-05,
      "loss": 0.218,
      "step": 1805
    },
    {
      "epoch": 1.448,
      "grad_norm": 9.0625,
      "learning_rate": 5.539159686727714e-05,
      "loss": 0.1839,
      "step": 1810
    },
    {
      "epoch": 1.452,
      "grad_norm": 7.84375,
      "learning_rate": 5.5179569115951354e-05,
      "loss": 0.2493,
      "step": 1815
    },
    {
      "epoch": 1.456,
      "grad_norm": 4.34375,
      "learning_rate": 5.496745469653654e-05,
      "loss": 0.2812,
      "step": 1820
    },
    {
      "epoch": 1.46,
      "grad_norm": 12.6875,
      "learning_rate": 5.475525753750361e-05,
      "loss": 0.1647,
      "step": 1825
    },
    {
      "epoch": 1.464,
      "grad_norm": 2.140625,
      "learning_rate": 5.454298156885592e-05,
      "loss": 0.1994,
      "step": 1830
    },
    {
      "epoch": 1.468,
      "grad_norm": 9.125,
      "learning_rate": 5.433063072205643e-05,
      "loss": 0.2761,
      "step": 1835
    },
    {
      "epoch": 1.472,
      "grad_norm": 6.375,
      "learning_rate": 5.411820892995484e-05,
      "loss": 0.4962,
      "step": 1840
    },
    {
      "epoch": 1.476,
      "grad_norm": 7.03125,
      "learning_rate": 5.3905720126714844e-05,
      "loss": 0.1545,
      "step": 1845
    },
    {
      "epoch": 1.48,
      "grad_norm": 15.0625,
      "learning_rate": 5.369316824774117e-05,
      "loss": 0.1802,
      "step": 1850
    },
    {
      "epoch": 1.484,
      "grad_norm": 7.25,
      "learning_rate": 5.3480557229606766e-05,
      "loss": 0.257,
      "step": 1855
    },
    {
      "epoch": 1.488,
      "grad_norm": 5.15625,
      "learning_rate": 5.3267891009979865e-05,
      "loss": 0.186,
      "step": 1860
    },
    {
      "epoch": 1.492,
      "grad_norm": 11.0,
      "learning_rate": 5.305517352755107e-05,
      "loss": 0.255,
      "step": 1865
    },
    {
      "epoch": 1.496,
      "grad_norm": 3.0,
      "learning_rate": 5.284240872196038e-05,
      "loss": 0.1839,
      "step": 1870
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.6015625,
      "learning_rate": 5.262960053372426e-05,
      "loss": 0.2419,
      "step": 1875
    },
    {
      "epoch": 1.504,
      "grad_norm": 5.125,
      "learning_rate": 5.241675290416265e-05,
      "loss": 0.1991,
      "step": 1880
    },
    {
      "epoch": 1.508,
      "grad_norm": 7.34375,
      "learning_rate": 5.220386977532592e-05,
      "loss": 0.2844,
      "step": 1885
    },
    {
      "epoch": 1.512,
      "grad_norm": 1.6328125,
      "learning_rate": 5.199095508992197e-05,
      "loss": 0.256,
      "step": 1890
    },
    {
      "epoch": 1.516,
      "grad_norm": 7.125,
      "learning_rate": 5.177801279124309e-05,
      "loss": 0.2767,
      "step": 1895
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.296875,
      "learning_rate": 5.156504682309301e-05,
      "loss": 0.2309,
      "step": 1900
    },
    {
      "epoch": 1.524,
      "grad_norm": 2.421875,
      "learning_rate": 5.1352061129713825e-05,
      "loss": 0.3085,
      "step": 1905
    },
    {
      "epoch": 1.528,
      "grad_norm": 6.15625,
      "learning_rate": 5.113905965571295e-05,
      "loss": 0.185,
      "step": 1910
    },
    {
      "epoch": 1.532,
      "grad_norm": 3.03125,
      "learning_rate": 5.0926046345990075e-05,
      "loss": 0.2313,
      "step": 1915
    },
    {
      "epoch": 1.536,
      "grad_norm": 7.90625,
      "learning_rate": 5.071302514566408e-05,
      "loss": 0.2945,
      "step": 1920
    },
    {
      "epoch": 1.54,
      "grad_norm": 8.375,
      "learning_rate": 5.05e-05,
      "loss": 0.2543,
      "step": 1925
    },
    {
      "epoch": 1.544,
      "grad_norm": 6.0,
      "learning_rate": 5.028697485433591e-05,
      "loss": 0.2708,
      "step": 1930
    },
    {
      "epoch": 1.548,
      "grad_norm": 3.359375,
      "learning_rate": 5.0073953654009934e-05,
      "loss": 0.1828,
      "step": 1935
    },
    {
      "epoch": 1.552,
      "grad_norm": 18.0,
      "learning_rate": 4.9860940344287057e-05,
      "loss": 0.2411,
      "step": 1940
    },
    {
      "epoch": 1.556,
      "grad_norm": 4.15625,
      "learning_rate": 4.964793887028619e-05,
      "loss": 0.2136,
      "step": 1945
    },
    {
      "epoch": 1.56,
      "grad_norm": 5.5625,
      "learning_rate": 4.943495317690701e-05,
      "loss": 0.2893,
      "step": 1950
    },
    {
      "epoch": 1.564,
      "grad_norm": 7.6875,
      "learning_rate": 4.922198720875694e-05,
      "loss": 0.1786,
      "step": 1955
    },
    {
      "epoch": 1.568,
      "grad_norm": 10.875,
      "learning_rate": 4.900904491007805e-05,
      "loss": 0.1553,
      "step": 1960
    },
    {
      "epoch": 1.572,
      "grad_norm": 9.4375,
      "learning_rate": 4.87961302246741e-05,
      "loss": 0.2255,
      "step": 1965
    },
    {
      "epoch": 1.576,
      "grad_norm": 1.6796875,
      "learning_rate": 4.8583247095837375e-05,
      "loss": 0.2767,
      "step": 1970
    },
    {
      "epoch": 1.58,
      "grad_norm": 2.4375,
      "learning_rate": 4.837039946627573e-05,
      "loss": 0.2157,
      "step": 1975
    },
    {
      "epoch": 1.584,
      "grad_norm": 1.9375,
      "learning_rate": 4.8157591278039625e-05,
      "loss": 0.2643,
      "step": 1980
    },
    {
      "epoch": 1.588,
      "grad_norm": 2.734375,
      "learning_rate": 4.794482647244894e-05,
      "loss": 0.1261,
      "step": 1985
    },
    {
      "epoch": 1.592,
      "grad_norm": 6.375,
      "learning_rate": 4.7732108990020144e-05,
      "loss": 0.1302,
      "step": 1990
    },
    {
      "epoch": 1.596,
      "grad_norm": 2.453125,
      "learning_rate": 4.751944277039325e-05,
      "loss": 0.1881,
      "step": 1995
    },
    {
      "epoch": 1.6,
      "grad_norm": 5.65625,
      "learning_rate": 4.730683175225885e-05,
      "loss": 0.2541,
      "step": 2000
    },
    {
      "epoch": 1.604,
      "grad_norm": 3.71875,
      "learning_rate": 4.709427987328517e-05,
      "loss": 0.214,
      "step": 2005
    },
    {
      "epoch": 1.608,
      "grad_norm": 3.28125,
      "learning_rate": 4.688179107004517e-05,
      "loss": 0.2405,
      "step": 2010
    },
    {
      "epoch": 1.612,
      "grad_norm": 5.84375,
      "learning_rate": 4.666936927794358e-05,
      "loss": 0.2371,
      "step": 2015
    },
    {
      "epoch": 1.616,
      "grad_norm": 9.5625,
      "learning_rate": 4.6457018431144095e-05,
      "loss": 0.25,
      "step": 2020
    },
    {
      "epoch": 1.62,
      "grad_norm": 3.34375,
      "learning_rate": 4.624474246249641e-05,
      "loss": 0.1999,
      "step": 2025
    },
    {
      "epoch": 1.624,
      "grad_norm": 12.25,
      "learning_rate": 4.603254530346347e-05,
      "loss": 0.1821,
      "step": 2030
    },
    {
      "epoch": 1.6280000000000001,
      "grad_norm": 2.796875,
      "learning_rate": 4.582043088404864e-05,
      "loss": 0.1656,
      "step": 2035
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 2.828125,
      "learning_rate": 4.560840313272286e-05,
      "loss": 0.1749,
      "step": 2040
    },
    {
      "epoch": 1.6360000000000001,
      "grad_norm": 0.66796875,
      "learning_rate": 4.539646597635199e-05,
      "loss": 0.1391,
      "step": 2045
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 3.828125,
      "learning_rate": 4.5184623340123936e-05,
      "loss": 0.2042,
      "step": 2050
    },
    {
      "epoch": 1.6440000000000001,
      "grad_norm": 7.0,
      "learning_rate": 4.4972879147476134e-05,
      "loss": 0.2436,
      "step": 2055
    },
    {
      "epoch": 1.6480000000000001,
      "grad_norm": 9.4375,
      "learning_rate": 4.476123732002273e-05,
      "loss": 0.7847,
      "step": 2060
    },
    {
      "epoch": 1.6520000000000001,
      "grad_norm": 8.9375,
      "learning_rate": 4.4549701777482056e-05,
      "loss": 0.1875,
      "step": 2065
    },
    {
      "epoch": 1.6560000000000001,
      "grad_norm": 10.3125,
      "learning_rate": 4.4338276437603936e-05,
      "loss": 0.231,
      "step": 2070
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 3.609375,
      "learning_rate": 4.4126965216097245e-05,
      "loss": 0.1835,
      "step": 2075
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 4.375,
      "learning_rate": 4.39157720265573e-05,
      "loss": 0.244,
      "step": 2080
    },
    {
      "epoch": 1.6680000000000001,
      "grad_norm": 5.28125,
      "learning_rate": 4.370470078039339e-05,
      "loss": 0.323,
      "step": 2085
    },
    {
      "epoch": 1.6720000000000002,
      "grad_norm": 1.8046875,
      "learning_rate": 4.3493755386756366e-05,
      "loss": 0.4006,
      "step": 2090
    },
    {
      "epoch": 1.6760000000000002,
      "grad_norm": 2.53125,
      "learning_rate": 4.32829397524662e-05,
      "loss": 0.2249,
      "step": 2095
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 4.4375,
      "learning_rate": 4.30722577819397e-05,
      "loss": 0.2631,
      "step": 2100
    },
    {
      "epoch": 1.6840000000000002,
      "grad_norm": 8.1875,
      "learning_rate": 4.286171337711809e-05,
      "loss": 0.2091,
      "step": 2105
    },
    {
      "epoch": 1.688,
      "grad_norm": 2.546875,
      "learning_rate": 4.265131043739485e-05,
      "loss": 0.2495,
      "step": 2110
    },
    {
      "epoch": 1.692,
      "grad_norm": 3.203125,
      "learning_rate": 4.2441052859543425e-05,
      "loss": 0.2168,
      "step": 2115
    },
    {
      "epoch": 1.696,
      "grad_norm": 9.875,
      "learning_rate": 4.223094453764511e-05,
      "loss": 0.2249,
      "step": 2120
    },
    {
      "epoch": 1.7,
      "grad_norm": 7.65625,
      "learning_rate": 4.202098936301686e-05,
      "loss": 0.1825,
      "step": 2125
    },
    {
      "epoch": 1.704,
      "grad_norm": 2.609375,
      "learning_rate": 4.181119122413931e-05,
      "loss": 0.2758,
      "step": 2130
    },
    {
      "epoch": 1.708,
      "grad_norm": 8.5,
      "learning_rate": 4.1601554006584684e-05,
      "loss": 0.2053,
      "step": 2135
    },
    {
      "epoch": 1.712,
      "grad_norm": 3.3125,
      "learning_rate": 4.1392081592944836e-05,
      "loss": 0.1751,
      "step": 2140
    },
    {
      "epoch": 1.716,
      "grad_norm": 6.5,
      "learning_rate": 4.118277786275942e-05,
      "loss": 0.1973,
      "step": 2145
    },
    {
      "epoch": 1.72,
      "grad_norm": 3.890625,
      "learning_rate": 4.097364669244392e-05,
      "loss": 0.1493,
      "step": 2150
    },
    {
      "epoch": 1.724,
      "grad_norm": 2.328125,
      "learning_rate": 4.076469195521799e-05,
      "loss": 0.1627,
      "step": 2155
    },
    {
      "epoch": 1.728,
      "grad_norm": 3.0,
      "learning_rate": 4.0555917521033556e-05,
      "loss": 0.2246,
      "step": 2160
    },
    {
      "epoch": 1.732,
      "grad_norm": 0.83203125,
      "learning_rate": 4.034732725650336e-05,
      "loss": 0.1962,
      "step": 2165
    },
    {
      "epoch": 1.736,
      "grad_norm": 2.15625,
      "learning_rate": 4.0138925024829095e-05,
      "loss": 0.2245,
      "step": 2170
    },
    {
      "epoch": 1.74,
      "grad_norm": 5.375,
      "learning_rate": 3.993071468573009e-05,
      "loss": 0.24,
      "step": 2175
    },
    {
      "epoch": 1.744,
      "grad_norm": 6.71875,
      "learning_rate": 3.972270009537168e-05,
      "loss": 0.2077,
      "step": 2180
    },
    {
      "epoch": 1.748,
      "grad_norm": 7.125,
      "learning_rate": 3.95148851062938e-05,
      "loss": 0.2049,
      "step": 2185
    },
    {
      "epoch": 1.752,
      "grad_norm": 5.0,
      "learning_rate": 3.9307273567339736e-05,
      "loss": 0.1957,
      "step": 2190
    },
    {
      "epoch": 1.756,
      "grad_norm": 3.890625,
      "learning_rate": 3.9099869323584695e-05,
      "loss": 0.1964,
      "step": 2195
    },
    {
      "epoch": 1.76,
      "grad_norm": 8.9375,
      "learning_rate": 3.8892676216264725e-05,
      "loss": 0.2341,
      "step": 2200
    },
    {
      "epoch": 1.764,
      "grad_norm": 1.40625,
      "learning_rate": 3.868569808270547e-05,
      "loss": 0.1902,
      "step": 2205
    },
    {
      "epoch": 1.768,
      "grad_norm": 31.0,
      "learning_rate": 3.84789387562512e-05,
      "loss": 0.3612,
      "step": 2210
    },
    {
      "epoch": 1.772,
      "grad_norm": 7.3125,
      "learning_rate": 3.827240206619369e-05,
      "loss": 0.2431,
      "step": 2215
    },
    {
      "epoch": 1.776,
      "grad_norm": 1.125,
      "learning_rate": 3.8066091837701456e-05,
      "loss": 0.2447,
      "step": 2220
    },
    {
      "epoch": 1.78,
      "grad_norm": 11.5625,
      "learning_rate": 3.786001189174875e-05,
      "loss": 0.3919,
      "step": 2225
    },
    {
      "epoch": 1.784,
      "grad_norm": 5.875,
      "learning_rate": 3.765416604504492e-05,
      "loss": 0.196,
      "step": 2230
    },
    {
      "epoch": 1.788,
      "grad_norm": 8.1875,
      "learning_rate": 3.744855810996365e-05,
      "loss": 0.2437,
      "step": 2235
    },
    {
      "epoch": 1.792,
      "grad_norm": 4.625,
      "learning_rate": 3.724319189447238e-05,
      "loss": 0.1862,
      "step": 2240
    },
    {
      "epoch": 1.796,
      "grad_norm": 1.5390625,
      "learning_rate": 3.7038071202061784e-05,
      "loss": 0.2062,
      "step": 2245
    },
    {
      "epoch": 1.8,
      "grad_norm": 5.3125,
      "learning_rate": 3.6833199831675294e-05,
      "loss": 0.2972,
      "step": 2250
    },
    {
      "epoch": 1.804,
      "grad_norm": 3.65625,
      "learning_rate": 3.662858157763881e-05,
      "loss": 0.2292,
      "step": 2255
    },
    {
      "epoch": 1.808,
      "grad_norm": 2.296875,
      "learning_rate": 3.6424220229590324e-05,
      "loss": 0.2717,
      "step": 2260
    },
    {
      "epoch": 1.812,
      "grad_norm": 5.28125,
      "learning_rate": 3.622011957240987e-05,
      "loss": 0.1763,
      "step": 2265
    },
    {
      "epoch": 1.8159999999999998,
      "grad_norm": 3.5,
      "learning_rate": 3.601628338614926e-05,
      "loss": 0.1896,
      "step": 2270
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 2.109375,
      "learning_rate": 3.581271544596223e-05,
      "loss": 0.1179,
      "step": 2275
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 5.1875,
      "learning_rate": 3.560941952203445e-05,
      "loss": 0.23,
      "step": 2280
    },
    {
      "epoch": 1.8279999999999998,
      "grad_norm": 5.375,
      "learning_rate": 3.5406399379513686e-05,
      "loss": 0.2536,
      "step": 2285
    },
    {
      "epoch": 1.8319999999999999,
      "grad_norm": 4.875,
      "learning_rate": 3.5203658778440106e-05,
      "loss": 0.1533,
      "step": 2290
    },
    {
      "epoch": 1.8359999999999999,
      "grad_norm": 4.3125,
      "learning_rate": 3.500120147367659e-05,
      "loss": 0.2593,
      "step": 2295
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 8.9375,
      "learning_rate": 3.4799031214839256e-05,
      "loss": 0.3177,
      "step": 2300
    },
    {
      "epoch": 1.8439999999999999,
      "grad_norm": 9.375,
      "learning_rate": 3.4597151746227963e-05,
      "loss": 0.1905,
      "step": 2305
    },
    {
      "epoch": 1.8479999999999999,
      "grad_norm": 9.0,
      "learning_rate": 3.439556680675699e-05,
      "loss": 0.2412,
      "step": 2310
    },
    {
      "epoch": 1.8519999999999999,
      "grad_norm": 6.5,
      "learning_rate": 3.419428012988577e-05,
      "loss": 0.301,
      "step": 2315
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 3.296875,
      "learning_rate": 3.3993295443549757e-05,
      "loss": 0.2301,
      "step": 2320
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 1.8203125,
      "learning_rate": 3.3792616470091396e-05,
      "loss": 0.2638,
      "step": 2325
    },
    {
      "epoch": 1.8639999999999999,
      "grad_norm": 1.2265625,
      "learning_rate": 3.359224692619115e-05,
      "loss": 0.2584,
      "step": 2330
    },
    {
      "epoch": 1.8679999999999999,
      "grad_norm": 7.3125,
      "learning_rate": 3.339219052279871e-05,
      "loss": 0.3116,
      "step": 2335
    },
    {
      "epoch": 1.8719999999999999,
      "grad_norm": 12.3125,
      "learning_rate": 3.31924509650642e-05,
      "loss": 0.291,
      "step": 2340
    },
    {
      "epoch": 1.876,
      "grad_norm": 4.875,
      "learning_rate": 3.299303195226962e-05,
      "loss": 0.2807,
      "step": 2345
    },
    {
      "epoch": 1.88,
      "grad_norm": 2.84375,
      "learning_rate": 3.2793937177760285e-05,
      "loss": 0.4022,
      "step": 2350
    },
    {
      "epoch": 1.884,
      "grad_norm": 16.25,
      "learning_rate": 3.259517032887652e-05,
      "loss": 0.253,
      "step": 2355
    },
    {
      "epoch": 1.888,
      "grad_norm": 5.46875,
      "learning_rate": 3.2396735086885174e-05,
      "loss": 0.2212,
      "step": 2360
    },
    {
      "epoch": 1.892,
      "grad_norm": 6.875,
      "learning_rate": 3.2198635126911695e-05,
      "loss": 0.1525,
      "step": 2365
    },
    {
      "epoch": 1.896,
      "grad_norm": 0.55078125,
      "learning_rate": 3.2000874117871786e-05,
      "loss": 0.2008,
      "step": 2370
    },
    {
      "epoch": 1.9,
      "grad_norm": 5.0,
      "learning_rate": 3.1803455722403736e-05,
      "loss": 0.2134,
      "step": 2375
    },
    {
      "epoch": 1.904,
      "grad_norm": 5.125,
      "learning_rate": 3.160638359680038e-05,
      "loss": 0.2858,
      "step": 2380
    },
    {
      "epoch": 1.908,
      "grad_norm": 5.03125,
      "learning_rate": 3.140966139094145e-05,
      "loss": 0.1827,
      "step": 2385
    },
    {
      "epoch": 1.912,
      "grad_norm": 5.3125,
      "learning_rate": 3.1213292748226005e-05,
      "loss": 0.2133,
      "step": 2390
    },
    {
      "epoch": 1.916,
      "grad_norm": 1.53125,
      "learning_rate": 3.1017281305504896e-05,
      "loss": 0.1609,
      "step": 2395
    },
    {
      "epoch": 1.92,
      "grad_norm": 3.0,
      "learning_rate": 3.082163069301349e-05,
      "loss": 0.2338,
      "step": 2400
    },
    {
      "epoch": 1.924,
      "grad_norm": 6.0,
      "learning_rate": 3.062634453430433e-05,
      "loss": 0.2946,
      "step": 2405
    },
    {
      "epoch": 1.928,
      "grad_norm": 1.9453125,
      "learning_rate": 3.043142644618015e-05,
      "loss": 0.2776,
      "step": 2410
    },
    {
      "epoch": 1.932,
      "grad_norm": 1.21875,
      "learning_rate": 3.0236880038626746e-05,
      "loss": 0.2709,
      "step": 2415
    },
    {
      "epoch": 1.936,
      "grad_norm": 7.28125,
      "learning_rate": 3.0042708914746243e-05,
      "loss": 0.3116,
      "step": 2420
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.5625,
      "learning_rate": 2.984891667069033e-05,
      "loss": 0.3037,
      "step": 2425
    },
    {
      "epoch": 1.944,
      "grad_norm": 7.625,
      "learning_rate": 2.965550689559356e-05,
      "loss": 0.2285,
      "step": 2430
    },
    {
      "epoch": 1.948,
      "grad_norm": 13.0,
      "learning_rate": 2.9462483171507077e-05,
      "loss": 0.3158,
      "step": 2435
    },
    {
      "epoch": 1.952,
      "grad_norm": 2.203125,
      "learning_rate": 2.9269849073332067e-05,
      "loss": 0.2553,
      "step": 2440
    },
    {
      "epoch": 1.956,
      "grad_norm": 9.625,
      "learning_rate": 2.9077608168753662e-05,
      "loss": 0.2483,
      "step": 2445
    },
    {
      "epoch": 1.96,
      "grad_norm": 6.46875,
      "learning_rate": 2.888576401817484e-05,
      "loss": 0.184,
      "step": 2450
    },
    {
      "epoch": 1.964,
      "grad_norm": 4.125,
      "learning_rate": 2.8694320174650508e-05,
      "loss": 0.212,
      "step": 2455
    },
    {
      "epoch": 1.968,
      "grad_norm": 9.0625,
      "learning_rate": 2.8503280183821644e-05,
      "loss": 0.2636,
      "step": 2460
    },
    {
      "epoch": 1.972,
      "grad_norm": 1.9140625,
      "learning_rate": 2.831264758384969e-05,
      "loss": 0.2095,
      "step": 2465
    },
    {
      "epoch": 1.976,
      "grad_norm": 1.2890625,
      "learning_rate": 2.8122425905350947e-05,
      "loss": 0.1762,
      "step": 2470
    },
    {
      "epoch": 1.98,
      "grad_norm": 4.15625,
      "learning_rate": 2.79326186713313e-05,
      "loss": 0.2045,
      "step": 2475
    },
    {
      "epoch": 1.984,
      "grad_norm": 1.796875,
      "learning_rate": 2.774322939712085e-05,
      "loss": 0.1516,
      "step": 2480
    },
    {
      "epoch": 1.988,
      "grad_norm": 2.84375,
      "learning_rate": 2.7554261590308877e-05,
      "loss": 0.2111,
      "step": 2485
    },
    {
      "epoch": 1.992,
      "grad_norm": 8.9375,
      "learning_rate": 2.7365718750678898e-05,
      "loss": 0.2379,
      "step": 2490
    },
    {
      "epoch": 1.996,
      "grad_norm": 1.296875,
      "learning_rate": 2.7177604370143738e-05,
      "loss": 0.3091,
      "step": 2495
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.4140625,
      "learning_rate": 2.6989921932681027e-05,
      "loss": 0.1719,
      "step": 2500
    },
    {
      "epoch": 2.004,
      "grad_norm": 4.09375,
      "learning_rate": 2.6802674914268477e-05,
      "loss": 0.206,
      "step": 2505
    },
    {
      "epoch": 2.008,
      "grad_norm": 12.5,
      "learning_rate": 2.6615866782819706e-05,
      "loss": 0.2479,
      "step": 2510
    },
    {
      "epoch": 2.012,
      "grad_norm": 5.03125,
      "learning_rate": 2.6429500998119832e-05,
      "loss": 0.233,
      "step": 2515
    },
    {
      "epoch": 2.016,
      "grad_norm": 13.1875,
      "learning_rate": 2.624358101176157e-05,
      "loss": 0.2007,
      "step": 2520
    },
    {
      "epoch": 2.02,
      "grad_norm": 1.4609375,
      "learning_rate": 2.605811026708113e-05,
      "loss": 0.2813,
      "step": 2525
    },
    {
      "epoch": 2.024,
      "grad_norm": 6.53125,
      "learning_rate": 2.5873092199094555e-05,
      "loss": 0.3496,
      "step": 2530
    },
    {
      "epoch": 2.028,
      "grad_norm": 1.7109375,
      "learning_rate": 2.568853023443411e-05,
      "loss": 0.2694,
      "step": 2535
    },
    {
      "epoch": 2.032,
      "grad_norm": 2.15625,
      "learning_rate": 2.5504427791284742e-05,
      "loss": 0.1851,
      "step": 2540
    },
    {
      "epoch": 2.036,
      "grad_norm": 0.890625,
      "learning_rate": 2.532078827932084e-05,
      "loss": 0.1534,
      "step": 2545
    },
    {
      "epoch": 2.04,
      "grad_norm": 6.28125,
      "learning_rate": 2.513761509964303e-05,
      "loss": 0.2738,
      "step": 2550
    },
    {
      "epoch": 2.044,
      "grad_norm": 1.9609375,
      "learning_rate": 2.4954911644715293e-05,
      "loss": 0.2147,
      "step": 2555
    },
    {
      "epoch": 2.048,
      "grad_norm": 7.53125,
      "learning_rate": 2.4772681298301987e-05,
      "loss": 0.2417,
      "step": 2560
    },
    {
      "epoch": 2.052,
      "grad_norm": 4.46875,
      "learning_rate": 2.4590927435405275e-05,
      "loss": 0.1669,
      "step": 2565
    },
    {
      "epoch": 2.056,
      "grad_norm": 10.5625,
      "learning_rate": 2.440965342220258e-05,
      "loss": 0.2052,
      "step": 2570
    },
    {
      "epoch": 2.06,
      "grad_norm": 4.28125,
      "learning_rate": 2.4228862615984288e-05,
      "loss": 0.284,
      "step": 2575
    },
    {
      "epoch": 2.064,
      "grad_norm": 2.59375,
      "learning_rate": 2.404855836509156e-05,
      "loss": 0.2411,
      "step": 2580
    },
    {
      "epoch": 2.068,
      "grad_norm": 0.71875,
      "learning_rate": 2.3868744008854172e-05,
      "loss": 0.1225,
      "step": 2585
    },
    {
      "epoch": 2.072,
      "grad_norm": 5.96875,
      "learning_rate": 2.368942287752893e-05,
      "loss": 0.2731,
      "step": 2590
    },
    {
      "epoch": 2.076,
      "grad_norm": 2.609375,
      "learning_rate": 2.3510598292237767e-05,
      "loss": 0.2515,
      "step": 2595
    },
    {
      "epoch": 2.08,
      "grad_norm": 5.09375,
      "learning_rate": 2.333227356490641e-05,
      "loss": 0.2434,
      "step": 2600
    },
    {
      "epoch": 2.084,
      "grad_norm": 1.8828125,
      "learning_rate": 2.3154451998202832e-05,
      "loss": 0.1641,
      "step": 2605
    },
    {
      "epoch": 2.088,
      "grad_norm": 12.3125,
      "learning_rate": 2.297713688547631e-05,
      "loss": 0.2349,
      "step": 2610
    },
    {
      "epoch": 2.092,
      "grad_norm": 4.5625,
      "learning_rate": 2.2800331510696253e-05,
      "loss": 0.1863,
      "step": 2615
    },
    {
      "epoch": 2.096,
      "grad_norm": 0.8671875,
      "learning_rate": 2.2624039148391528e-05,
      "loss": 0.2267,
      "step": 2620
    },
    {
      "epoch": 2.1,
      "grad_norm": 6.71875,
      "learning_rate": 2.2448263063589683e-05,
      "loss": 0.2269,
      "step": 2625
    },
    {
      "epoch": 2.104,
      "grad_norm": 4.90625,
      "learning_rate": 2.2273006511756536e-05,
      "loss": 0.292,
      "step": 2630
    },
    {
      "epoch": 2.108,
      "grad_norm": 4.75,
      "learning_rate": 2.2098272738735924e-05,
      "loss": 0.252,
      "step": 2635
    },
    {
      "epoch": 2.112,
      "grad_norm": 7.40625,
      "learning_rate": 2.1924064980689505e-05,
      "loss": 0.3024,
      "step": 2640
    },
    {
      "epoch": 2.116,
      "grad_norm": 1.5859375,
      "learning_rate": 2.175038646403687e-05,
      "loss": 0.2723,
      "step": 2645
    },
    {
      "epoch": 2.12,
      "grad_norm": 5.53125,
      "learning_rate": 2.1577240405395762e-05,
      "loss": 0.2735,
      "step": 2650
    },
    {
      "epoch": 2.124,
      "grad_norm": 11.375,
      "learning_rate": 2.1404630011522586e-05,
      "loss": 0.2396,
      "step": 2655
    },
    {
      "epoch": 2.128,
      "grad_norm": 1.6015625,
      "learning_rate": 2.1232558479252884e-05,
      "loss": 0.3201,
      "step": 2660
    },
    {
      "epoch": 2.132,
      "grad_norm": 2.640625,
      "learning_rate": 2.1061028995442214e-05,
      "loss": 0.2316,
      "step": 2665
    },
    {
      "epoch": 2.136,
      "grad_norm": 9.4375,
      "learning_rate": 2.089004473690714e-05,
      "loss": 0.2702,
      "step": 2670
    },
    {
      "epoch": 2.14,
      "grad_norm": 6.90625,
      "learning_rate": 2.0719608870366326e-05,
      "loss": 0.2258,
      "step": 2675
    },
    {
      "epoch": 2.144,
      "grad_norm": 1.140625,
      "learning_rate": 2.0549724552381986e-05,
      "loss": 0.181,
      "step": 2680
    },
    {
      "epoch": 2.148,
      "grad_norm": 3.53125,
      "learning_rate": 2.0380394929301255e-05,
      "loss": 0.2431,
      "step": 2685
    },
    {
      "epoch": 2.152,
      "grad_norm": 3.59375,
      "learning_rate": 2.021162313719815e-05,
      "loss": 0.2014,
      "step": 2690
    },
    {
      "epoch": 2.156,
      "grad_norm": 2.265625,
      "learning_rate": 2.0043412301815272e-05,
      "loss": 0.1898,
      "step": 2695
    },
    {
      "epoch": 2.16,
      "grad_norm": 10.4375,
      "learning_rate": 1.9875765538506097e-05,
      "loss": 0.241,
      "step": 2700
    },
    {
      "epoch": 2.164,
      "grad_norm": 2.484375,
      "learning_rate": 1.9708685952177067e-05,
      "loss": 0.1836,
      "step": 2705
    },
    {
      "epoch": 2.168,
      "grad_norm": 11.75,
      "learning_rate": 1.9542176637230318e-05,
      "loss": 0.2239,
      "step": 2710
    },
    {
      "epoch": 2.172,
      "grad_norm": 3.953125,
      "learning_rate": 1.937624067750619e-05,
      "loss": 0.2001,
      "step": 2715
    },
    {
      "epoch": 2.176,
      "grad_norm": 13.875,
      "learning_rate": 1.9210881146226243e-05,
      "loss": 0.2218,
      "step": 2720
    },
    {
      "epoch": 2.18,
      "grad_norm": 6.4375,
      "learning_rate": 1.9046101105936222e-05,
      "loss": 0.2144,
      "step": 2725
    },
    {
      "epoch": 2.184,
      "grad_norm": 0.76171875,
      "learning_rate": 1.8881903608449405e-05,
      "loss": 0.2392,
      "step": 2730
    },
    {
      "epoch": 2.188,
      "grad_norm": 3.140625,
      "learning_rate": 1.8718291694790104e-05,
      "loss": 0.2392,
      "step": 2735
    },
    {
      "epoch": 2.192,
      "grad_norm": 2.765625,
      "learning_rate": 1.855526839513727e-05,
      "loss": 0.2394,
      "step": 2740
    },
    {
      "epoch": 2.196,
      "grad_norm": 4.25,
      "learning_rate": 1.8392836728768428e-05,
      "loss": 0.1935,
      "step": 2745
    },
    {
      "epoch": 2.2,
      "grad_norm": 3.59375,
      "learning_rate": 1.823099970400373e-05,
      "loss": 0.1952,
      "step": 2750
    },
    {
      "epoch": 2.204,
      "grad_norm": 1.9765625,
      "learning_rate": 1.8069760318150286e-05,
      "loss": 0.3847,
      "step": 2755
    },
    {
      "epoch": 2.208,
      "grad_norm": 4.75,
      "learning_rate": 1.7909121557446587e-05,
      "loss": 0.2983,
      "step": 2760
    },
    {
      "epoch": 2.212,
      "grad_norm": 4.375,
      "learning_rate": 1.774908639700722e-05,
      "loss": 0.2924,
      "step": 2765
    },
    {
      "epoch": 2.216,
      "grad_norm": 5.59375,
      "learning_rate": 1.7589657800767826e-05,
      "loss": 0.2679,
      "step": 2770
    },
    {
      "epoch": 2.22,
      "grad_norm": 5.75,
      "learning_rate": 1.7430838721430084e-05,
      "loss": 0.2053,
      "step": 2775
    },
    {
      "epoch": 2.224,
      "grad_norm": 3.265625,
      "learning_rate": 1.7272632100407203e-05,
      "loss": 0.2297,
      "step": 2780
    },
    {
      "epoch": 2.228,
      "grad_norm": 2.953125,
      "learning_rate": 1.7115040867769202e-05,
      "loss": 0.2341,
      "step": 2785
    },
    {
      "epoch": 2.232,
      "grad_norm": 3.3125,
      "learning_rate": 1.6958067942188906e-05,
      "loss": 0.237,
      "step": 2790
    },
    {
      "epoch": 2.2359999999999998,
      "grad_norm": 5.40625,
      "learning_rate": 1.680171623088769e-05,
      "loss": 0.1687,
      "step": 2795
    },
    {
      "epoch": 2.24,
      "grad_norm": 18.625,
      "learning_rate": 1.6645988629581792e-05,
      "loss": 0.3214,
      "step": 2800
    },
    {
      "epoch": 2.2439999999999998,
      "grad_norm": 3.9375,
      "learning_rate": 1.649088802242849e-05,
      "loss": 0.104,
      "step": 2805
    },
    {
      "epoch": 2.248,
      "grad_norm": 2.5,
      "learning_rate": 1.6336417281972896e-05,
      "loss": 0.1931,
      "step": 2810
    },
    {
      "epoch": 2.252,
      "grad_norm": 2.0625,
      "learning_rate": 1.618257926909461e-05,
      "loss": 0.2171,
      "step": 2815
    },
    {
      "epoch": 2.2560000000000002,
      "grad_norm": 15.625,
      "learning_rate": 1.602937683295483e-05,
      "loss": 0.2504,
      "step": 2820
    },
    {
      "epoch": 2.26,
      "grad_norm": 5.1875,
      "learning_rate": 1.5876812810943482e-05,
      "loss": 0.2822,
      "step": 2825
    },
    {
      "epoch": 2.2640000000000002,
      "grad_norm": 3.0625,
      "learning_rate": 1.5724890028626727e-05,
      "loss": 0.2462,
      "step": 2830
    },
    {
      "epoch": 2.268,
      "grad_norm": 1.2265625,
      "learning_rate": 1.55736112996947e-05,
      "loss": 0.2166,
      "step": 2835
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 7.0625,
      "learning_rate": 1.5422979425909254e-05,
      "loss": 0.2026,
      "step": 2840
    },
    {
      "epoch": 2.276,
      "grad_norm": 1.1484375,
      "learning_rate": 1.5272997197052162e-05,
      "loss": 0.2885,
      "step": 2845
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 1.6875,
      "learning_rate": 1.5123667390873437e-05,
      "loss": 0.172,
      "step": 2850
    },
    {
      "epoch": 2.284,
      "grad_norm": 1.546875,
      "learning_rate": 1.4974992773039889e-05,
      "loss": 0.2386,
      "step": 2855
    },
    {
      "epoch": 2.288,
      "grad_norm": 4.59375,
      "learning_rate": 1.4826976097083886e-05,
      "loss": 0.3006,
      "step": 2860
    },
    {
      "epoch": 2.292,
      "grad_norm": 4.65625,
      "learning_rate": 1.4679620104352337e-05,
      "loss": 0.2628,
      "step": 2865
    },
    {
      "epoch": 2.296,
      "grad_norm": 14.625,
      "learning_rate": 1.453292752395601e-05,
      "loss": 0.3401,
      "step": 2870
    },
    {
      "epoch": 2.3,
      "grad_norm": 13.6875,
      "learning_rate": 1.4386901072718862e-05,
      "loss": 0.2556,
      "step": 2875
    },
    {
      "epoch": 2.304,
      "grad_norm": 4.5,
      "learning_rate": 1.4241543455127862e-05,
      "loss": 0.2243,
      "step": 2880
    },
    {
      "epoch": 2.308,
      "grad_norm": 13.0,
      "learning_rate": 1.4096857363282715e-05,
      "loss": 0.2837,
      "step": 2885
    },
    {
      "epoch": 2.312,
      "grad_norm": 6.625,
      "learning_rate": 1.3952845476846233e-05,
      "loss": 0.2373,
      "step": 2890
    },
    {
      "epoch": 2.316,
      "grad_norm": 9.25,
      "learning_rate": 1.3809510462994507e-05,
      "loss": 0.2668,
      "step": 2895
    },
    {
      "epoch": 2.32,
      "grad_norm": 5.28125,
      "learning_rate": 1.3666854976367664e-05,
      "loss": 0.2492,
      "step": 2900
    },
    {
      "epoch": 2.324,
      "grad_norm": 4.375,
      "learning_rate": 1.352488165902052e-05,
      "loss": 0.2805,
      "step": 2905
    },
    {
      "epoch": 2.328,
      "grad_norm": 4.28125,
      "learning_rate": 1.3383593140373869e-05,
      "loss": 0.2389,
      "step": 2910
    },
    {
      "epoch": 2.332,
      "grad_norm": 3.09375,
      "learning_rate": 1.3242992037165588e-05,
      "loss": 0.1966,
      "step": 2915
    },
    {
      "epoch": 2.336,
      "grad_norm": 5.25,
      "learning_rate": 1.3103080953402327e-05,
      "loss": 0.2806,
      "step": 2920
    },
    {
      "epoch": 2.34,
      "grad_norm": 5.71875,
      "learning_rate": 1.2963862480311165e-05,
      "loss": 0.2288,
      "step": 2925
    },
    {
      "epoch": 2.344,
      "grad_norm": 12.375,
      "learning_rate": 1.2825339196291653e-05,
      "loss": 0.2908,
      "step": 2930
    },
    {
      "epoch": 2.348,
      "grad_norm": 1.4375,
      "learning_rate": 1.2687513666868139e-05,
      "loss": 0.1973,
      "step": 2935
    },
    {
      "epoch": 2.352,
      "grad_norm": 3.9375,
      "learning_rate": 1.255038844464213e-05,
      "loss": 0.2176,
      "step": 2940
    },
    {
      "epoch": 2.356,
      "grad_norm": 9.375,
      "learning_rate": 1.2413966069245084e-05,
      "loss": 0.2179,
      "step": 2945
    },
    {
      "epoch": 2.36,
      "grad_norm": 8.125,
      "learning_rate": 1.227824906729137e-05,
      "loss": 0.2217,
      "step": 2950
    },
    {
      "epoch": 2.364,
      "grad_norm": 4.65625,
      "learning_rate": 1.2143239952331492e-05,
      "loss": 0.3206,
      "step": 2955
    },
    {
      "epoch": 2.368,
      "grad_norm": 0.93359375,
      "learning_rate": 1.2008941224805478e-05,
      "loss": 0.2215,
      "step": 2960
    },
    {
      "epoch": 2.372,
      "grad_norm": 5.0625,
      "learning_rate": 1.1875355371996608e-05,
      "loss": 0.2568,
      "step": 2965
    },
    {
      "epoch": 2.376,
      "grad_norm": 2.5,
      "learning_rate": 1.174248486798539e-05,
      "loss": 0.1783,
      "step": 2970
    },
    {
      "epoch": 2.38,
      "grad_norm": 3.90625,
      "learning_rate": 1.161033217360363e-05,
      "loss": 0.2066,
      "step": 2975
    },
    {
      "epoch": 2.384,
      "grad_norm": 4.53125,
      "learning_rate": 1.1478899736389001e-05,
      "loss": 0.2633,
      "step": 2980
    },
    {
      "epoch": 2.388,
      "grad_norm": 2.421875,
      "learning_rate": 1.1348189990539528e-05,
      "loss": 0.2126,
      "step": 2985
    },
    {
      "epoch": 2.392,
      "grad_norm": 22.0,
      "learning_rate": 1.1218205356868708e-05,
      "loss": 0.2356,
      "step": 2990
    },
    {
      "epoch": 2.396,
      "grad_norm": 0.7578125,
      "learning_rate": 1.1088948242760521e-05,
      "loss": 0.1807,
      "step": 2995
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.5078125,
      "learning_rate": 1.0960421042124948e-05,
      "loss": 0.2545,
      "step": 3000
    },
    {
      "epoch": 2.404,
      "grad_norm": 8.9375,
      "learning_rate": 1.083262613535351e-05,
      "loss": 0.4624,
      "step": 3005
    },
    {
      "epoch": 2.408,
      "grad_norm": 5.09375,
      "learning_rate": 1.0705565889275353e-05,
      "loss": 0.1927,
      "step": 3010
    },
    {
      "epoch": 2.412,
      "grad_norm": 7.15625,
      "learning_rate": 1.057924265711324e-05,
      "loss": 0.1696,
      "step": 3015
    },
    {
      "epoch": 2.416,
      "grad_norm": 4.6875,
      "learning_rate": 1.0453658778440107e-05,
      "loss": 0.2262,
      "step": 3020
    },
    {
      "epoch": 2.42,
      "grad_norm": 2.78125,
      "learning_rate": 1.0328816579135636e-05,
      "loss": 0.2476,
      "step": 3025
    },
    {
      "epoch": 2.424,
      "grad_norm": 1.6875,
      "learning_rate": 1.0204718371343215e-05,
      "loss": 0.2705,
      "step": 3030
    },
    {
      "epoch": 2.428,
      "grad_norm": 2.796875,
      "learning_rate": 1.008136645342716e-05,
      "loss": 0.2505,
      "step": 3035
    },
    {
      "epoch": 2.432,
      "grad_norm": 0.86328125,
      "learning_rate": 9.958763109930053e-06,
      "loss": 0.2551,
      "step": 3040
    },
    {
      "epoch": 2.436,
      "grad_norm": 1.15625,
      "learning_rate": 9.836910611530501e-06,
      "loss": 0.1786,
      "step": 3045
    },
    {
      "epoch": 2.44,
      "grad_norm": 4.0,
      "learning_rate": 9.715811215001068e-06,
      "loss": 0.2821,
      "step": 3050
    },
    {
      "epoch": 2.444,
      "grad_norm": 10.1875,
      "learning_rate": 9.595467163166482e-06,
      "loss": 0.2646,
      "step": 3055
    },
    {
      "epoch": 2.448,
      "grad_norm": 10.0,
      "learning_rate": 9.475880684862076e-06,
      "loss": 0.2521,
      "step": 3060
    },
    {
      "epoch": 2.452,
      "grad_norm": 1.4453125,
      "learning_rate": 9.357053994892506e-06,
      "loss": 0.186,
      "step": 3065
    },
    {
      "epoch": 2.456,
      "grad_norm": 4.75,
      "learning_rate": 9.238989293990796e-06,
      "loss": 0.289,
      "step": 3070
    },
    {
      "epoch": 2.46,
      "grad_norm": 11.9375,
      "learning_rate": 9.121688768777465e-06,
      "loss": 0.1989,
      "step": 3075
    },
    {
      "epoch": 2.464,
      "grad_norm": 5.5,
      "learning_rate": 9.005154591720184e-06,
      "loss": 0.294,
      "step": 3080
    },
    {
      "epoch": 2.468,
      "grad_norm": 1.2421875,
      "learning_rate": 8.889388921093334e-06,
      "loss": 0.2522,
      "step": 3085
    },
    {
      "epoch": 2.472,
      "grad_norm": 5.25,
      "learning_rate": 8.774393900938256e-06,
      "loss": 0.2692,
      "step": 3090
    },
    {
      "epoch": 2.476,
      "grad_norm": 2.8125,
      "learning_rate": 8.660171661023371e-06,
      "loss": 0.2109,
      "step": 3095
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.8046875,
      "learning_rate": 8.54672431680487e-06,
      "loss": 0.1933,
      "step": 3100
    },
    {
      "epoch": 2.484,
      "grad_norm": 8.25,
      "learning_rate": 8.434053969387374e-06,
      "loss": 0.1165,
      "step": 3105
    },
    {
      "epoch": 2.488,
      "grad_norm": 2.140625,
      "learning_rate": 8.32216270548519e-06,
      "loss": 0.2155,
      "step": 3110
    },
    {
      "epoch": 2.492,
      "grad_norm": 2.6875,
      "learning_rate": 8.211052597383563e-06,
      "loss": 0.2232,
      "step": 3115
    },
    {
      "epoch": 2.496,
      "grad_norm": 13.5625,
      "learning_rate": 8.100725702900322e-06,
      "loss": 0.1882,
      "step": 3120
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.75,
      "learning_rate": 7.991184065347759e-06,
      "loss": 0.1386,
      "step": 3125
    },
    {
      "epoch": 2.504,
      "grad_norm": 4.03125,
      "learning_rate": 7.882429713494781e-06,
      "loss": 0.1606,
      "step": 3130
    },
    {
      "epoch": 2.508,
      "grad_norm": 1.7890625,
      "learning_rate": 7.77446466152939e-06,
      "loss": 0.2935,
      "step": 3135
    },
    {
      "epoch": 2.512,
      "grad_norm": 3.6875,
      "learning_rate": 7.667290909021303e-06,
      "loss": 0.2613,
      "step": 3140
    },
    {
      "epoch": 2.516,
      "grad_norm": 2.21875,
      "learning_rate": 7.560910440884964e-06,
      "loss": 0.1835,
      "step": 3145
    },
    {
      "epoch": 2.52,
      "grad_norm": 6.34375,
      "learning_rate": 7.455325227342755e-06,
      "loss": 0.1829,
      "step": 3150
    },
    {
      "epoch": 2.524,
      "grad_norm": 2.6875,
      "learning_rate": 7.350537223888571e-06,
      "loss": 0.181,
      "step": 3155
    },
    {
      "epoch": 2.528,
      "grad_norm": 1.96875,
      "learning_rate": 7.2465483712515106e-06,
      "loss": 0.2518,
      "step": 3160
    },
    {
      "epoch": 2.532,
      "grad_norm": 2.203125,
      "learning_rate": 7.1433605953599725e-06,
      "loss": 0.1037,
      "step": 3165
    },
    {
      "epoch": 2.536,
      "grad_norm": 2.5,
      "learning_rate": 7.040975807306042e-06,
      "loss": 0.2237,
      "step": 3170
    },
    {
      "epoch": 2.54,
      "grad_norm": 29.625,
      "learning_rate": 6.939395903309986e-06,
      "loss": 0.3153,
      "step": 3175
    },
    {
      "epoch": 2.544,
      "grad_norm": 2.609375,
      "learning_rate": 6.838622764685256e-06,
      "loss": 0.2746,
      "step": 3180
    },
    {
      "epoch": 2.548,
      "grad_norm": 4.90625,
      "learning_rate": 6.738658257803502e-06,
      "loss": 0.2115,
      "step": 3185
    },
    {
      "epoch": 2.552,
      "grad_norm": 3.734375,
      "learning_rate": 6.639504234060176e-06,
      "loss": 0.1637,
      "step": 3190
    },
    {
      "epoch": 2.556,
      "grad_norm": 1.6953125,
      "learning_rate": 6.5411625298400865e-06,
      "loss": 0.2835,
      "step": 3195
    },
    {
      "epoch": 2.56,
      "grad_norm": 6.65625,
      "learning_rate": 6.4436349664835055e-06,
      "loss": 0.2081,
      "step": 3200
    },
    {
      "epoch": 2.564,
      "grad_norm": 9.5,
      "learning_rate": 6.346923350252354e-06,
      "loss": 0.161,
      "step": 3205
    },
    {
      "epoch": 2.568,
      "grad_norm": 15.9375,
      "learning_rate": 6.251029472296774e-06,
      "loss": 0.3144,
      "step": 3210
    },
    {
      "epoch": 2.572,
      "grad_norm": 3.21875,
      "learning_rate": 6.15595510862201e-06,
      "loss": 0.2441,
      "step": 3215
    },
    {
      "epoch": 2.576,
      "grad_norm": 9.625,
      "learning_rate": 6.061702020055431e-06,
      "loss": 0.2224,
      "step": 3220
    },
    {
      "epoch": 2.58,
      "grad_norm": 10.1875,
      "learning_rate": 5.96827195221396e-06,
      "loss": 0.2609,
      "step": 3225
    },
    {
      "epoch": 2.584,
      "grad_norm": 11.125,
      "learning_rate": 5.875666635471741e-06,
      "loss": 0.2013,
      "step": 3230
    },
    {
      "epoch": 2.588,
      "grad_norm": 3.09375,
      "learning_rate": 5.783887784928123e-06,
      "loss": 0.2377,
      "step": 3235
    },
    {
      "epoch": 2.592,
      "grad_norm": 1.2734375,
      "learning_rate": 5.692937100375828e-06,
      "loss": 0.1448,
      "step": 3240
    },
    {
      "epoch": 2.596,
      "grad_norm": 2.390625,
      "learning_rate": 5.602816266269527e-06,
      "loss": 0.2167,
      "step": 3245
    },
    {
      "epoch": 2.6,
      "grad_norm": 5.3125,
      "learning_rate": 5.513526951694604e-06,
      "loss": 0.2259,
      "step": 3250
    },
    {
      "epoch": 2.604,
      "grad_norm": 6.21875,
      "learning_rate": 5.425070810336297e-06,
      "loss": 0.4208,
      "step": 3255
    },
    {
      "epoch": 2.608,
      "grad_norm": 9.0,
      "learning_rate": 5.337449480449004e-06,
      "loss": 0.2434,
      "step": 3260
    },
    {
      "epoch": 2.612,
      "grad_norm": 10.4375,
      "learning_rate": 5.250664584825976e-06,
      "loss": 0.2231,
      "step": 3265
    },
    {
      "epoch": 2.616,
      "grad_norm": 5.5,
      "learning_rate": 5.16471773076929e-06,
      "loss": 0.2414,
      "step": 3270
    },
    {
      "epoch": 2.62,
      "grad_norm": 4.25,
      "learning_rate": 5.079610510060007e-06,
      "loss": 0.1683,
      "step": 3275
    },
    {
      "epoch": 2.624,
      "grad_norm": 3.890625,
      "learning_rate": 4.995344498928789e-06,
      "loss": 0.2159,
      "step": 3280
    },
    {
      "epoch": 2.628,
      "grad_norm": 1.0859375,
      "learning_rate": 4.911921258026592e-06,
      "loss": 0.311,
      "step": 3285
    },
    {
      "epoch": 2.632,
      "grad_norm": 5.9375,
      "learning_rate": 4.82934233239587e-06,
      "loss": 0.1862,
      "step": 3290
    },
    {
      "epoch": 2.636,
      "grad_norm": 10.8125,
      "learning_rate": 4.7476092514418826e-06,
      "loss": 0.3641,
      "step": 3295
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.8359375,
      "learning_rate": 4.666723528904435e-06,
      "loss": 0.2512,
      "step": 3300
    },
    {
      "epoch": 2.644,
      "grad_norm": 1.984375,
      "learning_rate": 4.586686662829771e-06,
      "loss": 0.2266,
      "step": 3305
    },
    {
      "epoch": 2.648,
      "grad_norm": 9.5625,
      "learning_rate": 4.507500135542872e-06,
      "loss": 0.3025,
      "step": 3310
    },
    {
      "epoch": 2.652,
      "grad_norm": 3.015625,
      "learning_rate": 4.429165413620022e-06,
      "loss": 0.18,
      "step": 3315
    },
    {
      "epoch": 2.656,
      "grad_norm": 8.75,
      "learning_rate": 4.351683947861591e-06,
      "loss": 0.2319,
      "step": 3320
    },
    {
      "epoch": 2.66,
      "grad_norm": 2.0,
      "learning_rate": 4.275057173265208e-06,
      "loss": 0.2251,
      "step": 3325
    },
    {
      "epoch": 2.664,
      "grad_norm": 1.2578125,
      "learning_rate": 4.199286508999162e-06,
      "loss": 0.2188,
      "step": 3330
    },
    {
      "epoch": 2.668,
      "grad_norm": 2.265625,
      "learning_rate": 4.124373358376145e-06,
      "loss": 0.2296,
      "step": 3335
    },
    {
      "epoch": 2.672,
      "grad_norm": 1.9921875,
      "learning_rate": 4.050319108827227e-06,
      "loss": 0.1519,
      "step": 3340
    },
    {
      "epoch": 2.676,
      "grad_norm": 4.0625,
      "learning_rate": 3.977125131876182e-06,
      "loss": 0.1885,
      "step": 3345
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.484375,
      "learning_rate": 3.90479278311407e-06,
      "loss": 0.2313,
      "step": 3350
    },
    {
      "epoch": 2.684,
      "grad_norm": 1.3125,
      "learning_rate": 3.83332340217418e-06,
      "loss": 0.1558,
      "step": 3355
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 5.9375,
      "learning_rate": 3.7627183127071335e-06,
      "loss": 0.2425,
      "step": 3360
    },
    {
      "epoch": 2.692,
      "grad_norm": 5.3125,
      "learning_rate": 3.6929788223564426e-06,
      "loss": 0.2278,
      "step": 3365
    },
    {
      "epoch": 2.6959999999999997,
      "grad_norm": 0.57421875,
      "learning_rate": 3.624106222734278e-06,
      "loss": 0.2796,
      "step": 3370
    },
    {
      "epoch": 2.7,
      "grad_norm": 2.703125,
      "learning_rate": 3.5561017893975043e-06,
      "loss": 0.2617,
      "step": 3375
    },
    {
      "epoch": 2.7039999999999997,
      "grad_norm": 6.75,
      "learning_rate": 3.4889667818241267e-06,
      "loss": 0.2075,
      "step": 3380
    },
    {
      "epoch": 2.708,
      "grad_norm": 0.80859375,
      "learning_rate": 3.4227024433899e-06,
      "loss": 0.2581,
      "step": 3385
    },
    {
      "epoch": 2.7119999999999997,
      "grad_norm": 2.15625,
      "learning_rate": 3.357310001345329e-06,
      "loss": 0.1813,
      "step": 3390
    },
    {
      "epoch": 2.716,
      "grad_norm": 3.640625,
      "learning_rate": 3.2927906667929434e-06,
      "loss": 0.3513,
      "step": 3395
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 9.5,
      "learning_rate": 3.2291456346648695e-06,
      "loss": 0.1969,
      "step": 3400
    },
    {
      "epoch": 2.724,
      "grad_norm": 5.40625,
      "learning_rate": 3.1663760837006716e-06,
      "loss": 0.2736,
      "step": 3405
    },
    {
      "epoch": 2.7279999999999998,
      "grad_norm": 0.859375,
      "learning_rate": 3.1044831764255566e-06,
      "loss": 0.1771,
      "step": 3410
    },
    {
      "epoch": 2.732,
      "grad_norm": 1.96875,
      "learning_rate": 3.0434680591288274e-06,
      "loss": 0.2071,
      "step": 3415
    },
    {
      "epoch": 2.7359999999999998,
      "grad_norm": 8.625,
      "learning_rate": 2.983331861842645e-06,
      "loss": 0.235,
      "step": 3420
    },
    {
      "epoch": 2.74,
      "grad_norm": 2.34375,
      "learning_rate": 2.9240756983211185e-06,
      "loss": 0.4819,
      "step": 3425
    },
    {
      "epoch": 2.7439999999999998,
      "grad_norm": 5.34375,
      "learning_rate": 2.865700666019643e-06,
      "loss": 0.2551,
      "step": 3430
    },
    {
      "epoch": 2.748,
      "grad_norm": 5.28125,
      "learning_rate": 2.8082078460746333e-06,
      "loss": 0.3602,
      "step": 3435
    },
    {
      "epoch": 2.752,
      "grad_norm": 1.3046875,
      "learning_rate": 2.7515983032834506e-06,
      "loss": 0.2711,
      "step": 3440
    },
    {
      "epoch": 2.7560000000000002,
      "grad_norm": 10.3125,
      "learning_rate": 2.6958730860846883e-06,
      "loss": 0.4155,
      "step": 3445
    },
    {
      "epoch": 2.76,
      "grad_norm": 4.71875,
      "learning_rate": 2.641033226538764e-06,
      "loss": 0.2659,
      "step": 3450
    },
    {
      "epoch": 2.7640000000000002,
      "grad_norm": 4.28125,
      "learning_rate": 2.5870797403088167e-06,
      "loss": 0.2592,
      "step": 3455
    },
    {
      "epoch": 2.768,
      "grad_norm": 0.80859375,
      "learning_rate": 2.5340136266418992e-06,
      "loss": 0.2555,
      "step": 3460
    },
    {
      "epoch": 2.7720000000000002,
      "grad_norm": 12.125,
      "learning_rate": 2.481835868350402e-06,
      "loss": 0.3904,
      "step": 3465
    },
    {
      "epoch": 2.776,
      "grad_norm": 4.9375,
      "learning_rate": 2.430547431793958e-06,
      "loss": 0.1825,
      "step": 3470
    },
    {
      "epoch": 2.7800000000000002,
      "grad_norm": 5.9375,
      "learning_rate": 2.38014926686146e-06,
      "loss": 0.1701,
      "step": 3475
    },
    {
      "epoch": 2.784,
      "grad_norm": 1.59375,
      "learning_rate": 2.3306423069535356e-06,
      "loss": 0.3218,
      "step": 3480
    },
    {
      "epoch": 2.7880000000000003,
      "grad_norm": 9.5625,
      "learning_rate": 2.282027468965185e-06,
      "loss": 0.2203,
      "step": 3485
    },
    {
      "epoch": 2.792,
      "grad_norm": 1.90625,
      "learning_rate": 2.2343056532688622e-06,
      "loss": 0.1943,
      "step": 3490
    },
    {
      "epoch": 2.7960000000000003,
      "grad_norm": 5.375,
      "learning_rate": 2.187477743697769e-06,
      "loss": 0.2201,
      "step": 3495
    },
    {
      "epoch": 2.8,
      "grad_norm": 7.4375,
      "learning_rate": 2.141544607529506e-06,
      "loss": 0.1929,
      "step": 3500
    },
    {
      "epoch": 2.8040000000000003,
      "grad_norm": 3.640625,
      "learning_rate": 2.0965070954699883e-06,
      "loss": 0.195,
      "step": 3505
    },
    {
      "epoch": 2.808,
      "grad_norm": 3.625,
      "learning_rate": 2.0523660416376947e-06,
      "loss": 0.2314,
      "step": 3510
    },
    {
      "epoch": 2.8120000000000003,
      "grad_norm": 1.2265625,
      "learning_rate": 2.009122263548242e-06,
      "loss": 0.1832,
      "step": 3515
    },
    {
      "epoch": 2.816,
      "grad_norm": 0.66015625,
      "learning_rate": 1.9667765620992154e-06,
      "loss": 0.1949,
      "step": 3520
    },
    {
      "epoch": 2.82,
      "grad_norm": 4.8125,
      "learning_rate": 1.925329721555338e-06,
      "loss": 0.149,
      "step": 3525
    },
    {
      "epoch": 2.824,
      "grad_norm": 3.625,
      "learning_rate": 1.8847825095339714e-06,
      "loss": 0.2328,
      "step": 3530
    },
    {
      "epoch": 2.828,
      "grad_norm": 2.9375,
      "learning_rate": 1.845135676990879e-06,
      "loss": 0.2937,
      "step": 3535
    },
    {
      "epoch": 2.832,
      "grad_norm": 1.6171875,
      "learning_rate": 1.8063899582063192e-06,
      "loss": 0.1943,
      "step": 3540
    },
    {
      "epoch": 2.836,
      "grad_norm": 4.8125,
      "learning_rate": 1.768546070771435e-06,
      "loss": 0.1837,
      "step": 3545
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.0703125,
      "learning_rate": 1.731604715574998e-06,
      "loss": 0.2379,
      "step": 3550
    },
    {
      "epoch": 2.844,
      "grad_norm": 3.5625,
      "learning_rate": 1.695566576790395e-06,
      "loss": 0.1816,
      "step": 3555
    },
    {
      "epoch": 2.848,
      "grad_norm": 1.0234375,
      "learning_rate": 1.6604323218629765e-06,
      "loss": 0.1572,
      "step": 3560
    },
    {
      "epoch": 2.852,
      "grad_norm": 2.71875,
      "learning_rate": 1.6262026014976748e-06,
      "loss": 0.2665,
      "step": 3565
    },
    {
      "epoch": 2.856,
      "grad_norm": 6.8125,
      "learning_rate": 1.5928780496469707e-06,
      "loss": 0.1376,
      "step": 3570
    },
    {
      "epoch": 2.86,
      "grad_norm": 5.40625,
      "learning_rate": 1.5604592834991465e-06,
      "loss": 0.1662,
      "step": 3575
    },
    {
      "epoch": 2.864,
      "grad_norm": 1.40625,
      "learning_rate": 1.5289469034668686e-06,
      "loss": 0.2482,
      "step": 3580
    },
    {
      "epoch": 2.868,
      "grad_norm": 3.640625,
      "learning_rate": 1.4983414931760246e-06,
      "loss": 0.1876,
      "step": 3585
    },
    {
      "epoch": 2.872,
      "grad_norm": 6.90625,
      "learning_rate": 1.4686436194549742e-06,
      "loss": 0.2189,
      "step": 3590
    },
    {
      "epoch": 2.876,
      "grad_norm": 4.0625,
      "learning_rate": 1.4398538323239924e-06,
      "loss": 0.1386,
      "step": 3595
    },
    {
      "epoch": 2.88,
      "grad_norm": 3.0625,
      "learning_rate": 1.411972664985146e-06,
      "loss": 0.2588,
      "step": 3600
    },
    {
      "epoch": 2.884,
      "grad_norm": 6.15625,
      "learning_rate": 1.3850006338123528e-06,
      "loss": 0.2615,
      "step": 3605
    },
    {
      "epoch": 2.888,
      "grad_norm": 0.7421875,
      "learning_rate": 1.3589382383418625e-06,
      "loss": 0.1873,
      "step": 3610
    },
    {
      "epoch": 2.892,
      "grad_norm": 3.84375,
      "learning_rate": 1.3337859612629915e-06,
      "loss": 0.2128,
      "step": 3615
    },
    {
      "epoch": 2.896,
      "grad_norm": 2.453125,
      "learning_rate": 1.3095442684091874e-06,
      "loss": 0.1934,
      "step": 3620
    },
    {
      "epoch": 2.9,
      "grad_norm": 3.921875,
      "learning_rate": 1.2862136087493837e-06,
      "loss": 0.2291,
      "step": 3625
    },
    {
      "epoch": 2.904,
      "grad_norm": 12.5625,
      "learning_rate": 1.2637944143797072e-06,
      "loss": 0.2081,
      "step": 3630
    },
    {
      "epoch": 2.908,
      "grad_norm": 4.46875,
      "learning_rate": 1.242287100515471e-06,
      "loss": 0.2323,
      "step": 3635
    },
    {
      "epoch": 2.912,
      "grad_norm": 10.8125,
      "learning_rate": 1.2216920654834752e-06,
      "loss": 0.1759,
      "step": 3640
    },
    {
      "epoch": 2.916,
      "grad_norm": 6.65625,
      "learning_rate": 1.202009690714637e-06,
      "loss": 0.2895,
      "step": 3645
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.3125,
      "learning_rate": 1.1832403407369131e-06,
      "loss": 0.1355,
      "step": 3650
    },
    {
      "epoch": 2.924,
      "grad_norm": 8.125,
      "learning_rate": 1.1653843631685772e-06,
      "loss": 0.2134,
      "step": 3655
    },
    {
      "epoch": 2.928,
      "grad_norm": 0.953125,
      "learning_rate": 1.1484420887117532e-06,
      "loss": 0.2601,
      "step": 3660
    },
    {
      "epoch": 2.932,
      "grad_norm": 7.46875,
      "learning_rate": 1.1324138311463032e-06,
      "loss": 0.3084,
      "step": 3665
    },
    {
      "epoch": 2.936,
      "grad_norm": 4.125,
      "learning_rate": 1.1172998873240079e-06,
      "loss": 0.1525,
      "step": 3670
    },
    {
      "epoch": 2.94,
      "grad_norm": 4.3125,
      "learning_rate": 1.1031005371630873e-06,
      "loss": 0.2672,
      "step": 3675
    },
    {
      "epoch": 2.944,
      "grad_norm": 10.6875,
      "learning_rate": 1.0898160436429967e-06,
      "loss": 0.3325,
      "step": 3680
    },
    {
      "epoch": 2.948,
      "grad_norm": 4.375,
      "learning_rate": 1.0774466527995685e-06,
      "loss": 0.2523,
      "step": 3685
    },
    {
      "epoch": 2.952,
      "grad_norm": 11.5,
      "learning_rate": 1.0659925937204445e-06,
      "loss": 0.2778,
      "step": 3690
    },
    {
      "epoch": 2.956,
      "grad_norm": 2.28125,
      "learning_rate": 1.055454078540835e-06,
      "loss": 0.446,
      "step": 3695
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.234375,
      "learning_rate": 1.0458313024396157e-06,
      "loss": 0.1925,
      "step": 3700
    },
    {
      "epoch": 2.964,
      "grad_norm": 3.796875,
      "learning_rate": 1.0371244436356671e-06,
      "loss": 0.2656,
      "step": 3705
    },
    {
      "epoch": 2.968,
      "grad_norm": 6.96875,
      "learning_rate": 1.0293336633846123e-06,
      "loss": 0.1901,
      "step": 3710
    },
    {
      "epoch": 2.972,
      "grad_norm": 11.5,
      "learning_rate": 1.022459105975814e-06,
      "loss": 0.1938,
      "step": 3715
    },
    {
      "epoch": 2.976,
      "grad_norm": 3.265625,
      "learning_rate": 1.0165008987296937e-06,
      "loss": 0.2682,
      "step": 3720
    },
    {
      "epoch": 2.98,
      "grad_norm": 4.21875,
      "learning_rate": 1.0114591519954021e-06,
      "loss": 0.4105,
      "step": 3725
    },
    {
      "epoch": 2.984,
      "grad_norm": 5.03125,
      "learning_rate": 1.0073339591487457e-06,
      "loss": 0.1863,
      "step": 3730
    },
    {
      "epoch": 2.988,
      "grad_norm": 6.03125,
      "learning_rate": 1.0041253965904744e-06,
      "loss": 0.1519,
      "step": 3735
    },
    {
      "epoch": 2.992,
      "grad_norm": 7.59375,
      "learning_rate": 1.0018335237448558e-06,
      "loss": 0.2199,
      "step": 3740
    },
    {
      "epoch": 2.996,
      "grad_norm": 8.3125,
      "learning_rate": 1.0004583830585885e-06,
      "loss": 0.3814,
      "step": 3745
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.625,
      "learning_rate": 1e-06,
      "loss": 0.2395,
      "step": 3750
    }
  ],
  "logging_steps": 5,
  "max_steps": 3750,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}