{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9996810207336523,
  "eval_steps": 500,
  "global_step": 1567,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006379585326953748,
      "grad_norm": 682.9805297851562,
      "learning_rate": 1.910828025477707e-06,
      "loss": 77.2852,
      "step": 1
    },
    {
      "epoch": 0.003189792663476874,
      "grad_norm": 520.5723266601562,
      "learning_rate": 9.554140127388534e-06,
      "loss": 77.5951,
      "step": 5
    },
    {
      "epoch": 0.006379585326953748,
      "grad_norm": 189.88075256347656,
      "learning_rate": 1.9108280254777068e-05,
      "loss": 64.9391,
      "step": 10
    },
    {
      "epoch": 0.009569377990430622,
      "grad_norm": 71.3049087524414,
      "learning_rate": 2.8662420382165606e-05,
      "loss": 55.1326,
      "step": 15
    },
    {
      "epoch": 0.012759170653907496,
      "grad_norm": 19.6275577545166,
      "learning_rate": 3.8216560509554137e-05,
      "loss": 52.7711,
      "step": 20
    },
    {
      "epoch": 0.01594896331738437,
      "grad_norm": 10.905985832214355,
      "learning_rate": 4.777070063694267e-05,
      "loss": 51.6146,
      "step": 25
    },
    {
      "epoch": 0.019138755980861243,
      "grad_norm": 13.994418144226074,
      "learning_rate": 5.732484076433121e-05,
      "loss": 50.0751,
      "step": 30
    },
    {
      "epoch": 0.022328548644338118,
      "grad_norm": 35.459774017333984,
      "learning_rate": 6.687898089171974e-05,
      "loss": 47.8509,
      "step": 35
    },
    {
      "epoch": 0.025518341307814992,
      "grad_norm": 112.83365631103516,
      "learning_rate": 7.643312101910827e-05,
      "loss": 43.2866,
      "step": 40
    },
    {
      "epoch": 0.028708133971291867,
      "grad_norm": 11.995097160339355,
      "learning_rate": 8.59872611464968e-05,
      "loss": 35.0235,
      "step": 45
    },
    {
      "epoch": 0.03189792663476874,
      "grad_norm": 5.324495792388916,
      "learning_rate": 9.554140127388533e-05,
      "loss": 34.1333,
      "step": 50
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 3.124239444732666,
      "learning_rate": 0.00010509554140127387,
      "loss": 33.767,
      "step": 55
    },
    {
      "epoch": 0.03827751196172249,
      "grad_norm": 2.8586628437042236,
      "learning_rate": 0.00011464968152866242,
      "loss": 33.6247,
      "step": 60
    },
    {
      "epoch": 0.04146730462519936,
      "grad_norm": 5.365510940551758,
      "learning_rate": 0.00012420382165605095,
      "loss": 33.4812,
      "step": 65
    },
    {
      "epoch": 0.044657097288676235,
      "grad_norm": 3.8166890144348145,
      "learning_rate": 0.00013375796178343948,
      "loss": 33.3342,
      "step": 70
    },
    {
      "epoch": 0.04784688995215311,
      "grad_norm": 4.826046466827393,
      "learning_rate": 0.00014331210191082802,
      "loss": 33.3135,
      "step": 75
    },
    {
      "epoch": 0.051036682615629984,
      "grad_norm": 2.350775957107544,
      "learning_rate": 0.00015286624203821655,
      "loss": 33.2625,
      "step": 80
    },
    {
      "epoch": 0.05422647527910686,
      "grad_norm": 4.356655120849609,
      "learning_rate": 0.00016242038216560508,
      "loss": 33.2039,
      "step": 85
    },
    {
      "epoch": 0.05741626794258373,
      "grad_norm": 5.798741817474365,
      "learning_rate": 0.0001719745222929936,
      "loss": 33.1815,
      "step": 90
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 7.0147705078125,
      "learning_rate": 0.00018152866242038214,
      "loss": 33.1506,
      "step": 95
    },
    {
      "epoch": 0.06379585326953748,
      "grad_norm": 5.96809720993042,
      "learning_rate": 0.00019108280254777067,
      "loss": 33.121,
      "step": 100
    },
    {
      "epoch": 0.06698564593301436,
      "grad_norm": 7.044692516326904,
      "learning_rate": 0.0002006369426751592,
      "loss": 33.0931,
      "step": 105
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 3.396991729736328,
      "learning_rate": 0.00021019108280254773,
      "loss": 33.0729,
      "step": 110
    },
    {
      "epoch": 0.0733652312599681,
      "grad_norm": 1.8412262201309204,
      "learning_rate": 0.00021974522292993626,
      "loss": 33.0843,
      "step": 115
    },
    {
      "epoch": 0.07655502392344497,
      "grad_norm": 2.1479506492614746,
      "learning_rate": 0.00022929936305732485,
      "loss": 33.069,
      "step": 120
    },
    {
      "epoch": 0.07974481658692185,
      "grad_norm": 4.789393901824951,
      "learning_rate": 0.00023885350318471338,
      "loss": 33.0806,
      "step": 125
    },
    {
      "epoch": 0.08293460925039872,
      "grad_norm": 11.621363639831543,
      "learning_rate": 0.0002484076433121019,
      "loss": 33.1734,
      "step": 130
    },
    {
      "epoch": 0.0861244019138756,
      "grad_norm": 73.14874267578125,
      "learning_rate": 0.00025796178343949044,
      "loss": 33.1719,
      "step": 135
    },
    {
      "epoch": 0.08931419457735247,
      "grad_norm": 1.356505274772644,
      "learning_rate": 0.00026751592356687897,
      "loss": 33.0876,
      "step": 140
    },
    {
      "epoch": 0.09250398724082935,
      "grad_norm": 1.3292990922927856,
      "learning_rate": 0.0002770700636942675,
      "loss": 33.0986,
      "step": 145
    },
    {
      "epoch": 0.09569377990430622,
      "grad_norm": 27.297182083129883,
      "learning_rate": 0.00028662420382165603,
      "loss": 33.0662,
      "step": 150
    },
    {
      "epoch": 0.09888357256778309,
      "grad_norm": 3.962141275405884,
      "learning_rate": 0.00029617834394904456,
      "loss": 33.103,
      "step": 155
    },
    {
      "epoch": 0.10207336523125997,
      "grad_norm": 2.6187081336975098,
      "learning_rate": 0.0002999966490829603,
      "loss": 33.0548,
      "step": 160
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 2.732147693634033,
      "learning_rate": 0.00029997617179878324,
      "loss": 33.0522,
      "step": 165
    },
    {
      "epoch": 0.10845295055821372,
      "grad_norm": 1.41147780418396,
      "learning_rate": 0.0002999370813893047,
      "loss": 33.0373,
      "step": 170
    },
    {
      "epoch": 0.11164274322169059,
      "grad_norm": 6.294727802276611,
      "learning_rate": 0.00029987938270592676,
      "loss": 33.0836,
      "step": 175
    },
    {
      "epoch": 0.11483253588516747,
      "grad_norm": 5.494381904602051,
      "learning_rate": 0.0002998030829094724,
      "loss": 33.0821,
      "step": 180
    },
    {
      "epoch": 0.11802232854864433,
      "grad_norm": 2.7319934368133545,
      "learning_rate": 0.00029970819146929694,
      "loss": 33.0523,
      "step": 185
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 1.471151351928711,
      "learning_rate": 0.0002995947201621131,
      "loss": 33.0452,
      "step": 190
    },
    {
      "epoch": 0.12440191387559808,
      "grad_norm": 2.9658069610595703,
      "learning_rate": 0.00029946268307052886,
      "loss": 33.0214,
      "step": 195
    },
    {
      "epoch": 0.12759170653907495,
      "grad_norm": 6.95133113861084,
      "learning_rate": 0.0002993120965813003,
      "loss": 33.011,
      "step": 200
    },
    {
      "epoch": 0.13078149920255183,
      "grad_norm": 2.7009997367858887,
      "learning_rate": 0.0002991429793832975,
      "loss": 33.0114,
      "step": 205
    },
    {
      "epoch": 0.1339712918660287,
      "grad_norm": 2.7211108207702637,
      "learning_rate": 0.00029895535246518524,
      "loss": 33.0345,
      "step": 210
    },
    {
      "epoch": 0.1371610845295056,
      "grad_norm": 1.866679072380066,
      "learning_rate": 0.0002987492391128182,
      "loss": 33.01,
      "step": 215
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 2.352771759033203,
      "learning_rate": 0.0002985246649063509,
      "loss": 32.9968,
      "step": 220
    },
    {
      "epoch": 0.14354066985645933,
      "grad_norm": 1.8509665727615356,
      "learning_rate": 0.0002982816577170631,
      "loss": 33.0003,
      "step": 225
    },
    {
      "epoch": 0.1467304625199362,
      "grad_norm": 2.1403417587280273,
      "learning_rate": 0.00029802024770390087,
      "loss": 33.0033,
      "step": 230
    },
    {
      "epoch": 0.14992025518341306,
      "grad_norm": 1.2801026105880737,
      "learning_rate": 0.00029774046730973334,
      "loss": 32.9999,
      "step": 235
    },
    {
      "epoch": 0.15311004784688995,
      "grad_norm": 1.6241658926010132,
      "learning_rate": 0.00029744235125732664,
      "loss": 32.9961,
      "step": 240
    },
    {
      "epoch": 0.15629984051036683,
      "grad_norm": 10.04849910736084,
      "learning_rate": 0.0002971259365450344,
      "loss": 33.0072,
      "step": 245
    },
    {
      "epoch": 0.1594896331738437,
      "grad_norm": 6.750959873199463,
      "learning_rate": 0.00029679126244220596,
      "loss": 32.986,
      "step": 250
    },
    {
      "epoch": 0.16267942583732056,
      "grad_norm": 1.4271435737609863,
      "learning_rate": 0.00029643837048431293,
      "loss": 33.0021,
      "step": 255
    },
    {
      "epoch": 0.16586921850079744,
      "grad_norm": 12.441051483154297,
      "learning_rate": 0.0002960673044677939,
      "loss": 33.0079,
      "step": 260
    },
    {
      "epoch": 0.16905901116427433,
      "grad_norm": 1.6060891151428223,
      "learning_rate": 0.00029567811044461977,
      "loss": 32.9779,
      "step": 265
    },
    {
      "epoch": 0.1722488038277512,
      "grad_norm": 2.5357677936553955,
      "learning_rate": 0.00029527083671657746,
      "loss": 33.0001,
      "step": 270
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 1.4731814861297607,
      "learning_rate": 0.00029484553382927594,
      "loss": 32.9766,
      "step": 275
    },
    {
      "epoch": 0.17862838915470494,
      "grad_norm": 1.6727865934371948,
      "learning_rate": 0.000294402254565873,
      "loss": 32.9865,
      "step": 280
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.122405529022217,
      "learning_rate": 0.00029394105394052434,
      "loss": 32.9938,
      "step": 285
    },
    {
      "epoch": 0.1850079744816587,
      "grad_norm": 1.9852848052978516,
      "learning_rate": 0.00029346198919155616,
      "loss": 33.0059,
      "step": 290
    },
    {
      "epoch": 0.18819776714513556,
      "grad_norm": 6.7702178955078125,
      "learning_rate": 0.00029296511977436107,
      "loss": 32.9958,
      "step": 295
    },
    {
      "epoch": 0.19138755980861244,
      "grad_norm": 2.152496576309204,
      "learning_rate": 0.0002924505073540198,
      "loss": 32.9785,
      "step": 300
    },
    {
      "epoch": 0.19457735247208932,
      "grad_norm": 1.5221149921417236,
      "learning_rate": 0.0002919182157976476,
      "loss": 32.9521,
      "step": 305
    },
    {
      "epoch": 0.19776714513556617,
      "grad_norm": 1.7718403339385986,
      "learning_rate": 0.00029136831116646815,
      "loss": 32.9534,
      "step": 310
    },
    {
      "epoch": 0.20095693779904306,
      "grad_norm": 1.301917552947998,
      "learning_rate": 0.000290800861707615,
      "loss": 32.9369,
      "step": 315
    },
    {
      "epoch": 0.20414673046251994,
      "grad_norm": 1.795551061630249,
      "learning_rate": 0.00029021593784566113,
      "loss": 32.9692,
      "step": 320
    },
    {
      "epoch": 0.20733652312599682,
      "grad_norm": 1.36078679561615,
      "learning_rate": 0.0002896136121738793,
      "loss": 32.9731,
      "step": 325
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 1.4519957304000854,
      "learning_rate": 0.0002889939594452323,
      "loss": 32.9647,
      "step": 330
    },
    {
      "epoch": 0.21371610845295055,
      "grad_norm": 1.334033489227295,
      "learning_rate": 0.00028835705656309583,
      "loss": 32.9777,
      "step": 335
    },
    {
      "epoch": 0.21690590111642744,
      "grad_norm": 1.4112995862960815,
      "learning_rate": 0.0002877029825717142,
      "loss": 32.9767,
      "step": 340
    },
    {
      "epoch": 0.22009569377990432,
      "grad_norm": 3.528616428375244,
      "learning_rate": 0.0002870318186463901,
      "loss": 32.9354,
      "step": 345
    },
    {
      "epoch": 0.22328548644338117,
      "grad_norm": 3.2746047973632812,
      "learning_rate": 0.0002863436480834105,
      "loss": 32.9897,
      "step": 350
    },
    {
      "epoch": 0.22647527910685805,
      "grad_norm": 1.641404151916504,
      "learning_rate": 0.00028563855628970886,
      "loss": 32.9667,
      "step": 355
    },
    {
      "epoch": 0.22966507177033493,
      "grad_norm": 1.5198549032211304,
      "learning_rate": 0.0002849166307722653,
      "loss": 32.9846,
      "step": 360
    },
    {
      "epoch": 0.23285486443381181,
      "grad_norm": 2.942479372024536,
      "learning_rate": 0.00028417796112724684,
      "loss": 32.9962,
      "step": 365
    },
    {
      "epoch": 0.23604465709728867,
      "grad_norm": 1.860962986946106,
      "learning_rate": 0.0002834226390288873,
      "loss": 32.9502,
      "step": 370
    },
    {
      "epoch": 0.23923444976076555,
      "grad_norm": 1.3557507991790771,
      "learning_rate": 0.0002826507582181103,
      "loss": 32.9439,
      "step": 375
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 2.8216800689697266,
      "learning_rate": 0.00028186241449089524,
      "loss": 32.9648,
      "step": 380
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 1.4041414260864258,
      "learning_rate": 0.000281057705686388,
      "loss": 32.9507,
      "step": 385
    },
    {
      "epoch": 0.24880382775119617,
      "grad_norm": 1.0593173503875732,
      "learning_rate": 0.0002802367316747589,
      "loss": 32.9521,
      "step": 390
    },
    {
      "epoch": 0.25199362041467305,
      "grad_norm": 1.5495184659957886,
      "learning_rate": 0.0002793995943448078,
      "loss": 32.9331,
      "step": 395
    },
    {
      "epoch": 0.2551834130781499,
      "grad_norm": 1.3299857378005981,
      "learning_rate": 0.00027854639759131893,
      "loss": 32.9373,
      "step": 400
    },
    {
      "epoch": 0.2583732057416268,
      "grad_norm": 4.341277122497559,
      "learning_rate": 0.00027767724730216696,
      "loss": 32.9751,
      "step": 405
    },
    {
      "epoch": 0.26156299840510366,
      "grad_norm": 2.890934944152832,
      "learning_rate": 0.0002767922513451754,
      "loss": 32.9733,
      "step": 410
    },
    {
      "epoch": 0.2647527910685805,
      "grad_norm": 1.1594318151474,
      "learning_rate": 0.00027589151955472965,
      "loss": 32.96,
      "step": 415
    },
    {
      "epoch": 0.2679425837320574,
      "grad_norm": 1.5049182176589966,
      "learning_rate": 0.00027497516371814543,
      "loss": 32.9408,
      "step": 420
    },
    {
      "epoch": 0.2711323763955343,
      "grad_norm": 1.2559763193130493,
      "learning_rate": 0.00027404329756179537,
      "loss": 32.949,
      "step": 425
    },
    {
      "epoch": 0.2743221690590112,
      "grad_norm": 3.811816453933716,
      "learning_rate": 0.0002730960367369949,
      "loss": 32.9946,
      "step": 430
    },
    {
      "epoch": 0.27751196172248804,
      "grad_norm": 1.7879546880722046,
      "learning_rate": 0.00027213349880564873,
      "loss": 32.9684,
      "step": 435
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 1.7998263835906982,
      "learning_rate": 0.0002711558032256607,
      "loss": 32.9474,
      "step": 440
    },
    {
      "epoch": 0.2838915470494418,
      "grad_norm": 1.705267071723938,
      "learning_rate": 0.0002701630713361085,
      "loss": 32.9497,
      "step": 445
    },
    {
      "epoch": 0.28708133971291866,
      "grad_norm": 1.5598078966140747,
      "learning_rate": 0.00026915542634218403,
      "loss": 32.9503,
      "step": 450
    },
    {
      "epoch": 0.2902711323763955,
      "grad_norm": 1.2488480806350708,
      "learning_rate": 0.00026813299329990335,
      "loss": 32.9554,
      "step": 455
    },
    {
      "epoch": 0.2934609250398724,
      "grad_norm": 1.4403730630874634,
      "learning_rate": 0.0002670958991005859,
      "loss": 32.9399,
      "step": 460
    },
    {
      "epoch": 0.2966507177033493,
      "grad_norm": 1.6609704494476318,
      "learning_rate": 0.0002660442724551065,
      "loss": 32.9369,
      "step": 465
    },
    {
      "epoch": 0.29984051036682613,
      "grad_norm": 1.792934536933899,
      "learning_rate": 0.00026497824387792146,
      "loss": 32.9469,
      "step": 470
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 1.1857388019561768,
      "learning_rate": 0.00026389794567087085,
      "loss": 32.947,
      "step": 475
    },
    {
      "epoch": 0.3062200956937799,
      "grad_norm": 3.0106189250946045,
      "learning_rate": 0.0002628035119067586,
      "loss": 32.9288,
      "step": 480
    },
    {
      "epoch": 0.3094098883572568,
      "grad_norm": 1.5340945720672607,
      "learning_rate": 0.0002616950784127135,
      "loss": 32.9379,
      "step": 485
    },
    {
      "epoch": 0.31259968102073366,
      "grad_norm": 14.717156410217285,
      "learning_rate": 0.00026057278275333165,
      "loss": 32.9656,
      "step": 490
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 1.2622380256652832,
      "learning_rate": 0.00025943676421360395,
      "loss": 32.9441,
      "step": 495
    },
    {
      "epoch": 0.3189792663476874,
      "grad_norm": 2.068324089050293,
      "learning_rate": 0.00025828716378163,
      "loss": 32.9572,
      "step": 500
    },
    {
      "epoch": 0.32216905901116427,
      "grad_norm": 1.5519779920578003,
      "learning_rate": 0.00025712412413112006,
      "loss": 32.9444,
      "step": 505
    },
    {
      "epoch": 0.3253588516746411,
      "grad_norm": 5.29410982131958,
      "learning_rate": 0.00025594778960368844,
      "loss": 32.9654,
      "step": 510
    },
    {
      "epoch": 0.32854864433811803,
      "grad_norm": 4.592552185058594,
      "learning_rate": 0.0002547583061909396,
      "loss": 32.9438,
      "step": 515
    },
    {
      "epoch": 0.3317384370015949,
      "grad_norm": 2.020326614379883,
      "learning_rate": 0.00025355582151634956,
      "loss": 32.9603,
      "step": 520
    },
    {
      "epoch": 0.3349282296650718,
      "grad_norm": 10.48231029510498,
      "learning_rate": 0.00025234048481694477,
      "loss": 32.9279,
      "step": 525
    },
    {
      "epoch": 0.33811802232854865,
      "grad_norm": 2.3022897243499756,
      "learning_rate": 0.0002511124469247809,
      "loss": 32.9371,
      "step": 530
    },
    {
      "epoch": 0.3413078149920255,
      "grad_norm": 1.5293172597885132,
      "learning_rate": 0.00024987186024822295,
      "loss": 32.917,
      "step": 535
    },
    {
      "epoch": 0.3444976076555024,
      "grad_norm": 1.6426869630813599,
      "learning_rate": 0.0002486188787530309,
      "loss": 32.9352,
      "step": 540
    },
    {
      "epoch": 0.34768740031897927,
      "grad_norm": 1.266628623008728,
      "learning_rate": 0.00024735365794325117,
      "loss": 32.9531,
      "step": 545
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 1.558414340019226,
      "learning_rate": 0.0002460763548419172,
      "loss": 32.9282,
      "step": 550
    },
    {
      "epoch": 0.35406698564593303,
      "grad_norm": 1.2409437894821167,
      "learning_rate": 0.0002447871279715624,
      "loss": 32.9508,
      "step": 555
    },
    {
      "epoch": 0.3572567783094099,
      "grad_norm": 1.6733072996139526,
      "learning_rate": 0.00024348613733454565,
      "loss": 32.9415,
      "step": 560
    },
    {
      "epoch": 0.36044657097288674,
      "grad_norm": 4.34638786315918,
      "learning_rate": 0.00024217354439319427,
      "loss": 32.9175,
      "step": 565
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 1.4273444414138794,
      "learning_rate": 0.00024084951204976528,
      "loss": 32.9106,
      "step": 570
    },
    {
      "epoch": 0.3668261562998405,
      "grad_norm": 2.605161190032959,
      "learning_rate": 0.0002395142046262281,
      "loss": 32.9428,
      "step": 575
    },
    {
      "epoch": 0.3700159489633174,
      "grad_norm": 4.460203647613525,
      "learning_rate": 0.00023816778784387094,
      "loss": 32.923,
      "step": 580
    },
    {
      "epoch": 0.37320574162679426,
      "grad_norm": 1.232483983039856,
      "learning_rate": 0.0002368104288027336,
      "loss": 32.9239,
      "step": 585
    },
    {
      "epoch": 0.3763955342902711,
      "grad_norm": 10.035587310791016,
      "learning_rate": 0.0002354422959608692,
      "loss": 32.9288,
      "step": 590
    },
    {
      "epoch": 0.379585326953748,
      "grad_norm": 1.621656894683838,
      "learning_rate": 0.00023406355911343717,
      "loss": 32.9499,
      "step": 595
    },
    {
      "epoch": 0.3827751196172249,
      "grad_norm": 1.5809745788574219,
      "learning_rate": 0.00023267438937163077,
      "loss": 32.9158,
      "step": 600
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 2.634787082672119,
      "learning_rate": 0.00023127495914144051,
      "loss": 32.936,
      "step": 605
    },
    {
      "epoch": 0.38915470494417864,
      "grad_norm": 1.3653333187103271,
      "learning_rate": 0.00022986544210225774,
      "loss": 32.9167,
      "step": 610
    },
    {
      "epoch": 0.3923444976076555,
      "grad_norm": 1.6077592372894287,
      "learning_rate": 0.00022844601318531955,
      "loss": 32.9071,
      "step": 615
    },
    {
      "epoch": 0.39553429027113235,
      "grad_norm": 2.3299481868743896,
      "learning_rate": 0.00022701684855199857,
      "loss": 32.9561,
      "step": 620
    },
    {
      "epoch": 0.39872408293460926,
      "grad_norm": 1.0616035461425781,
      "learning_rate": 0.00022557812557194,
      "loss": 32.904,
      "step": 625
    },
    {
      "epoch": 0.4019138755980861,
      "grad_norm": 1.510393500328064,
      "learning_rate": 0.00022413002280104915,
      "loss": 32.8996,
      "step": 630
    },
    {
      "epoch": 0.405103668261563,
      "grad_norm": 1.354466438293457,
      "learning_rate": 0.00022267271995933074,
      "loss": 32.9283,
      "step": 635
    },
    {
      "epoch": 0.4082934609250399,
      "grad_norm": 1.6015855073928833,
      "learning_rate": 0.00022120639790858482,
      "loss": 32.9179,
      "step": 640
    },
    {
      "epoch": 0.41148325358851673,
      "grad_norm": 1.3500322103500366,
      "learning_rate": 0.00021973123862996044,
      "loss": 32.9267,
      "step": 645
    },
    {
      "epoch": 0.41467304625199364,
      "grad_norm": 1.5598021745681763,
      "learning_rate": 0.00021824742520137026,
      "loss": 32.9303,
      "step": 650
    },
    {
      "epoch": 0.4178628389154705,
      "grad_norm": 1.3947721719741821,
      "learning_rate": 0.00021675514177476945,
      "loss": 32.9193,
      "step": 655
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 1.5799477100372314,
      "learning_rate": 0.0002152545735533012,
      "loss": 32.9358,
      "step": 660
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 2.971487045288086,
      "learning_rate": 0.00021374590676831136,
      "loss": 32.9176,
      "step": 665
    },
    {
      "epoch": 0.4274322169059011,
      "grad_norm": 1.6111208200454712,
      "learning_rate": 0.00021222932865623605,
      "loss": 32.9286,
      "step": 670
    },
    {
      "epoch": 0.430622009569378,
      "grad_norm": 1.521829605102539,
      "learning_rate": 0.00021070502743536414,
      "loss": 32.8895,
      "step": 675
    },
    {
      "epoch": 0.43381180223285487,
      "grad_norm": 2.311312675476074,
      "learning_rate": 0.00020917319228247805,
      "loss": 32.8955,
      "step": 680
    },
    {
      "epoch": 0.4370015948963317,
      "grad_norm": 1.3634883165359497,
      "learning_rate": 0.00020763401330937555,
      "loss": 32.9083,
      "step": 685
    },
    {
      "epoch": 0.44019138755980863,
      "grad_norm": 1.6512216329574585,
      "learning_rate": 0.00020608768153927546,
      "loss": 32.9185,
      "step": 690
    },
    {
      "epoch": 0.4433811802232855,
      "grad_norm": 1.24175226688385,
      "learning_rate": 0.00020453438888311042,
      "loss": 32.903,
      "step": 695
    },
    {
      "epoch": 0.44657097288676234,
      "grad_norm": 1.2090448141098022,
      "learning_rate": 0.00020297432811570916,
      "loss": 32.9181,
      "step": 700
    },
    {
      "epoch": 0.44976076555023925,
      "grad_norm": 1.4872608184814453,
      "learning_rate": 0.00020140769285187187,
      "loss": 32.8974,
      "step": 705
    },
    {
      "epoch": 0.4529505582137161,
      "grad_norm": 1.8509513139724731,
      "learning_rate": 0.00019983467752234132,
      "loss": 32.8918,
      "step": 710
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 1.2917256355285645,
      "learning_rate": 0.0001982554773496723,
      "loss": 32.9323,
      "step": 715
    },
    {
      "epoch": 0.45933014354066987,
      "grad_norm": 1.2343769073486328,
      "learning_rate": 0.00019667028832400345,
      "loss": 32.8814,
      "step": 720
    },
    {
      "epoch": 0.4625199362041467,
      "grad_norm": 1.5799022912979126,
      "learning_rate": 0.00019507930717873313,
      "loss": 32.9088,
      "step": 725
    },
    {
      "epoch": 0.46570972886762363,
      "grad_norm": 2.4687390327453613,
      "learning_rate": 0.00019348273136610364,
      "loss": 32.9116,
      "step": 730
    },
    {
      "epoch": 0.4688995215311005,
      "grad_norm": 3.06180477142334,
      "learning_rate": 0.00019188075903269587,
      "loss": 32.9153,
      "step": 735
    },
    {
      "epoch": 0.47208931419457734,
      "grad_norm": 1.394356608390808,
      "learning_rate": 0.00019027358899483776,
      "loss": 32.9041,
      "step": 740
    },
    {
      "epoch": 0.47527910685805425,
      "grad_norm": 2.034818410873413,
      "learning_rate": 0.00018866142071393013,
      "loss": 32.9037,
      "step": 745
    },
    {
      "epoch": 0.4784688995215311,
      "grad_norm": 1.3607292175292969,
      "learning_rate": 0.00018704445427169156,
      "loss": 32.9138,
      "step": 750
    },
    {
      "epoch": 0.48165869218500795,
      "grad_norm": 1.2347332239151,
      "learning_rate": 0.00018542289034532733,
      "loss": 32.9023,
      "step": 755
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.3349498510360718,
      "learning_rate": 0.00018379693018262349,
      "loss": 32.9105,
      "step": 760
    },
    {
      "epoch": 0.4880382775119617,
      "grad_norm": 1.068160891532898,
      "learning_rate": 0.00018216677557697083,
      "loss": 32.9054,
      "step": 765
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 1.386084794998169,
      "learning_rate": 0.00018053262884232078,
      "loss": 32.8915,
      "step": 770
    },
    {
      "epoch": 0.4944178628389155,
      "grad_norm": 1.655228614807129,
      "learning_rate": 0.0001788946927880768,
      "loss": 32.9004,
      "step": 775
    },
    {
      "epoch": 0.49760765550239233,
      "grad_norm": 1.4305061101913452,
      "learning_rate": 0.00017725317069392418,
      "loss": 32.8904,
      "step": 780
    },
    {
      "epoch": 0.5007974481658692,
      "grad_norm": 1.2984495162963867,
      "learning_rate": 0.00017560826628460182,
      "loss": 32.8991,
      "step": 785
    },
    {
      "epoch": 0.5039872408293461,
      "grad_norm": 1.272292971611023,
      "learning_rate": 0.00017396018370461808,
      "loss": 32.9138,
      "step": 790
    },
    {
      "epoch": 0.507177033492823,
      "grad_norm": 2.4082605838775635,
      "learning_rate": 0.00017230912749291547,
      "loss": 32.9214,
      "step": 795
    },
    {
      "epoch": 0.5103668261562998,
      "grad_norm": 1.3513133525848389,
      "learning_rate": 0.00017065530255748557,
      "loss": 32.9055,
      "step": 800
    },
    {
      "epoch": 0.5135566188197768,
      "grad_norm": 1.1499751806259155,
      "learning_rate": 0.00016899891414993854,
      "loss": 32.9026,
      "step": 805
    },
    {
      "epoch": 0.5167464114832536,
      "grad_norm": 1.3720273971557617,
      "learning_rate": 0.00016734016784002994,
      "loss": 32.9024,
      "step": 810
    },
    {
      "epoch": 0.5199362041467305,
      "grad_norm": 1.423717737197876,
      "learning_rate": 0.000165679269490148,
      "loss": 32.8942,
      "step": 815
    },
    {
      "epoch": 0.5231259968102073,
      "grad_norm": 1.2919470071792603,
      "learning_rate": 0.0001640164252297648,
      "loss": 32.9024,
      "step": 820
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 1.6845824718475342,
      "learning_rate": 0.00016235184142985367,
      "loss": 32.8939,
      "step": 825
    },
    {
      "epoch": 0.529505582137161,
      "grad_norm": 1.803841471672058,
      "learning_rate": 0.00016068572467727762,
      "loss": 32.9116,
      "step": 830
    },
    {
      "epoch": 0.532695374800638,
      "grad_norm": 1.202531099319458,
      "learning_rate": 0.00015901828174915005,
      "loss": 32.9034,
      "step": 835
    },
    {
      "epoch": 0.5358851674641149,
      "grad_norm": 1.2204720973968506,
      "learning_rate": 0.00015734971958717228,
      "loss": 32.9029,
      "step": 840
    },
    {
      "epoch": 0.5390749601275917,
      "grad_norm": 1.5371347665786743,
      "learning_rate": 0.00015568024527195067,
      "loss": 32.9148,
      "step": 845
    },
    {
      "epoch": 0.5422647527910686,
      "grad_norm": 1.621806263923645,
      "learning_rate": 0.0001540100659972963,
      "loss": 32.8704,
      "step": 850
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.5353915691375732,
      "learning_rate": 0.000152339389044511,
      "loss": 32.9093,
      "step": 855
    },
    {
      "epoch": 0.5486443381180224,
      "grad_norm": 1.5789800882339478,
      "learning_rate": 0.00015066842175666186,
      "loss": 32.8946,
      "step": 860
    },
    {
      "epoch": 0.5518341307814992,
      "grad_norm": 1.1287790536880493,
      "learning_rate": 0.0001489973715128487,
      "loss": 32.8768,
      "step": 865
    },
    {
      "epoch": 0.5550239234449761,
      "grad_norm": 2.7509007453918457,
      "learning_rate": 0.00014732644570246675,
      "loss": 32.8786,
      "step": 870
    },
    {
      "epoch": 0.5582137161084529,
      "grad_norm": 1.461751937866211,
      "learning_rate": 0.00014565585169946824,
      "loss": 32.8935,
      "step": 875
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 1.2226332426071167,
      "learning_rate": 0.0001439857968366256,
      "loss": 32.8947,
      "step": 880
    },
    {
      "epoch": 0.5645933014354066,
      "grad_norm": 1.1934312582015991,
      "learning_rate": 0.00014231648837980022,
      "loss": 32.8689,
      "step": 885
    },
    {
      "epoch": 0.5677830940988836,
      "grad_norm": 1.0734748840332031,
      "learning_rate": 0.00014064813350221894,
      "loss": 32.9223,
      "step": 890
    },
    {
      "epoch": 0.5709728867623605,
      "grad_norm": 1.6119298934936523,
      "learning_rate": 0.00013898093925876267,
      "loss": 32.8659,
      "step": 895
    },
    {
      "epoch": 0.5741626794258373,
      "grad_norm": 1.1905502080917358,
      "learning_rate": 0.00013731511256026913,
      "loss": 32.876,
      "step": 900
    },
    {
      "epoch": 0.5773524720893142,
      "grad_norm": 1.1675747632980347,
      "learning_rate": 0.00013565086014785406,
      "loss": 32.8991,
      "step": 905
    },
    {
      "epoch": 0.580542264752791,
      "grad_norm": 1.0815011262893677,
      "learning_rate": 0.00013398838856725257,
      "loss": 32.8836,
      "step": 910
    },
    {
      "epoch": 0.583732057416268,
      "grad_norm": 1.3470786809921265,
      "learning_rate": 0.00013232790414318608,
      "loss": 32.8723,
      "step": 915
    },
    {
      "epoch": 0.5869218500797448,
      "grad_norm": 1.1291898488998413,
      "learning_rate": 0.0001306696129537553,
      "loss": 32.8919,
      "step": 920
    },
    {
      "epoch": 0.5901116427432217,
      "grad_norm": 1.2471877336502075,
      "learning_rate": 0.00012901372080486472,
      "loss": 32.8716,
      "step": 925
    },
    {
      "epoch": 0.5933014354066986,
      "grad_norm": 1.2810490131378174,
      "learning_rate": 0.00012736043320468073,
      "loss": 32.8679,
      "step": 930
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 1.1581730842590332,
      "learning_rate": 0.0001257099553381262,
      "loss": 32.8792,
      "step": 935
    },
    {
      "epoch": 0.5996810207336523,
      "grad_norm": 1.0867726802825928,
      "learning_rate": 0.00012406249204141603,
      "loss": 32.8761,
      "step": 940
    },
    {
      "epoch": 0.6028708133971292,
      "grad_norm": 1.279363989830017,
      "learning_rate": 0.000122418247776635,
      "loss": 32.881,
      "step": 945
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.2689155340194702,
      "learning_rate": 0.00012077742660636299,
      "loss": 32.8814,
      "step": 950
    },
    {
      "epoch": 0.6092503987240829,
      "grad_norm": 0.9977864027023315,
      "learning_rate": 0.00011914023216834904,
      "loss": 32.8547,
      "step": 955
    },
    {
      "epoch": 0.6124401913875598,
      "grad_norm": 1.4236512184143066,
      "learning_rate": 0.0001175068676502386,
      "loss": 32.8553,
      "step": 960
    },
    {
      "epoch": 0.6156299840510366,
      "grad_norm": 1.0569161176681519,
      "learning_rate": 0.00011587753576435634,
      "loss": 32.8551,
      "step": 965
    },
    {
      "epoch": 0.6188197767145136,
      "grad_norm": 1.3193695545196533,
      "learning_rate": 0.00011425243872254835,
      "loss": 32.8631,
      "step": 970
    },
    {
      "epoch": 0.6220095693779905,
      "grad_norm": 1.0458968877792358,
      "learning_rate": 0.00011263177821108573,
      "loss": 32.8778,
      "step": 975
    },
    {
      "epoch": 0.6251993620414673,
      "grad_norm": 1.097130298614502,
      "learning_rate": 0.00011101575536563433,
      "loss": 32.8635,
      "step": 980
    },
    {
      "epoch": 0.6283891547049442,
      "grad_norm": 1.4956417083740234,
      "learning_rate": 0.00010940457074629217,
      "loss": 32.8633,
      "step": 985
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 1.9950311183929443,
      "learning_rate": 0.00010779842431269843,
      "loss": 32.8721,
      "step": 990
    },
    {
      "epoch": 0.6347687400318979,
      "grad_norm": 1.3927932977676392,
      "learning_rate": 0.0001061975153992172,
      "loss": 32.8595,
      "step": 995
    },
    {
      "epoch": 0.6379585326953748,
      "grad_norm": 1.089880347251892,
      "learning_rate": 0.00010460204269019829,
      "loss": 32.8519,
      "step": 1000
    },
    {
      "epoch": 0.6411483253588517,
      "grad_norm": 1.4656952619552612,
      "learning_rate": 0.0001030122041953196,
      "loss": 32.8695,
      "step": 1005
    },
    {
      "epoch": 0.6443381180223285,
      "grad_norm": 1.343774437904358,
      "learning_rate": 0.0001014281972250121,
      "loss": 32.8688,
      "step": 1010
    },
    {
      "epoch": 0.6475279106858054,
      "grad_norm": 1.0298471450805664,
      "learning_rate": 9.985021836597273e-05,
      "loss": 32.8466,
      "step": 1015
    },
    {
      "epoch": 0.6507177033492823,
      "grad_norm": 1.0400863885879517,
      "learning_rate": 9.827846345676614e-05,
      "loss": 32.8529,
      "step": 1020
    },
    {
      "epoch": 0.6539074960127592,
      "grad_norm": 1.187381625175476,
      "learning_rate": 9.671312756351998e-05,
      "loss": 32.8502,
      "step": 1025
    },
    {
      "epoch": 0.6570972886762361,
      "grad_norm": 1.0974117517471313,
      "learning_rate": 9.515440495571569e-05,
      "loss": 32.8592,
      "step": 1030
    },
    {
      "epoch": 0.6602870813397129,
      "grad_norm": 1.0766081809997559,
      "learning_rate": 9.360248908207813e-05,
      "loss": 32.8525,
      "step": 1035
    },
    {
      "epoch": 0.6634768740031898,
      "grad_norm": 1.3774298429489136,
      "learning_rate": 9.205757254656755e-05,
      "loss": 32.836,
      "step": 1040
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.1813623905181885,
      "learning_rate": 9.05198470844756e-05,
      "loss": 32.8414,
      "step": 1045
    },
    {
      "epoch": 0.6698564593301436,
      "grad_norm": 1.556418538093567,
      "learning_rate": 8.898950353862998e-05,
      "loss": 32.8629,
      "step": 1050
    },
    {
      "epoch": 0.6730462519936204,
      "grad_norm": 1.267234444618225,
      "learning_rate": 8.746673183570923e-05,
      "loss": 32.8663,
      "step": 1055
    },
    {
      "epoch": 0.6762360446570973,
      "grad_norm": 1.3755508661270142,
      "learning_rate": 8.595172096267157e-05,
      "loss": 32.8899,
      "step": 1060
    },
    {
      "epoch": 0.6794258373205742,
      "grad_norm": 1.3215252161026,
      "learning_rate": 8.444465894330024e-05,
      "loss": 32.838,
      "step": 1065
    },
    {
      "epoch": 0.682615629984051,
      "grad_norm": 1.0346448421478271,
      "learning_rate": 8.294573281486828e-05,
      "loss": 32.849,
      "step": 1070
    },
    {
      "epoch": 0.6858054226475279,
      "grad_norm": 1.00798761844635,
      "learning_rate": 8.145512860492596e-05,
      "loss": 32.8436,
      "step": 1075
    },
    {
      "epoch": 0.6889952153110048,
      "grad_norm": 1.5634852647781372,
      "learning_rate": 7.997303130821362e-05,
      "loss": 32.8626,
      "step": 1080
    },
    {
      "epoch": 0.6921850079744817,
      "grad_norm": 1.0482699871063232,
      "learning_rate": 7.849962486370206e-05,
      "loss": 32.8506,
      "step": 1085
    },
    {
      "epoch": 0.6953748006379585,
      "grad_norm": 1.077114224433899,
      "learning_rate": 7.703509213176451e-05,
      "loss": 32.8685,
      "step": 1090
    },
    {
      "epoch": 0.6985645933014354,
      "grad_norm": 1.0572760105133057,
      "learning_rate": 7.557961487148272e-05,
      "loss": 32.8712,
      "step": 1095
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.9180437922477722,
      "learning_rate": 7.413337371808884e-05,
      "loss": 32.8458,
      "step": 1100
    },
    {
      "epoch": 0.7049441786283892,
      "grad_norm": 1.1923459768295288,
      "learning_rate": 7.269654816054756e-05,
      "loss": 32.8429,
      "step": 1105
    },
    {
      "epoch": 0.7081339712918661,
      "grad_norm": 1.5775012969970703,
      "learning_rate": 7.126931651928012e-05,
      "loss": 32.8366,
      "step": 1110
    },
    {
      "epoch": 0.7113237639553429,
      "grad_norm": 1.0124591588974,
      "learning_rate": 6.985185592403367e-05,
      "loss": 32.8589,
      "step": 1115
    },
    {
      "epoch": 0.7145135566188198,
      "grad_norm": 1.6170281171798706,
      "learning_rate": 6.844434229189787e-05,
      "loss": 32.8525,
      "step": 1120
    },
    {
      "epoch": 0.7177033492822966,
      "grad_norm": 1.2746340036392212,
      "learning_rate": 6.704695030547252e-05,
      "loss": 32.8356,
      "step": 1125
    },
    {
      "epoch": 0.7208931419457735,
      "grad_norm": 1.6307950019836426,
      "learning_rate": 6.56598533911881e-05,
      "loss": 32.8478,
      "step": 1130
    },
    {
      "epoch": 0.7240829346092504,
      "grad_norm": 1.007103443145752,
      "learning_rate": 6.428322369778254e-05,
      "loss": 32.8573,
      "step": 1135
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 1.051518201828003,
      "learning_rate": 6.291723207493577e-05,
      "loss": 32.8216,
      "step": 1140
    },
    {
      "epoch": 0.7304625199362041,
      "grad_norm": 1.5083210468292236,
      "learning_rate": 6.15620480520666e-05,
      "loss": 32.8387,
      "step": 1145
    },
    {
      "epoch": 0.733652312599681,
      "grad_norm": 1.4043292999267578,
      "learning_rate": 6.0217839817292765e-05,
      "loss": 32.8262,
      "step": 1150
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 0.9952952861785889,
      "learning_rate": 5.888477419655733e-05,
      "loss": 32.854,
      "step": 1155
    },
    {
      "epoch": 0.7400318979266348,
      "grad_norm": 1.1356273889541626,
      "learning_rate": 5.7563016632924555e-05,
      "loss": 32.8296,
      "step": 1160
    },
    {
      "epoch": 0.7432216905901117,
      "grad_norm": 1.1537383794784546,
      "learning_rate": 5.625273116604728e-05,
      "loss": 32.8622,
      "step": 1165
    },
    {
      "epoch": 0.7464114832535885,
      "grad_norm": 1.2923429012298584,
      "learning_rate": 5.495408041180829e-05,
      "loss": 32.866,
      "step": 1170
    },
    {
      "epoch": 0.7496012759170654,
      "grad_norm": 1.1965166330337524,
      "learning_rate": 5.3667225542138507e-05,
      "loss": 32.8139,
      "step": 1175
    },
    {
      "epoch": 0.7527910685805422,
      "grad_norm": 0.9976691007614136,
      "learning_rate": 5.239232626501464e-05,
      "loss": 32.8625,
      "step": 1180
    },
    {
      "epoch": 0.7559808612440191,
      "grad_norm": 1.6022858619689941,
      "learning_rate": 5.1129540804637906e-05,
      "loss": 32.8465,
      "step": 1185
    },
    {
      "epoch": 0.759170653907496,
      "grad_norm": 1.2044793367385864,
      "learning_rate": 4.9879025881797466e-05,
      "loss": 32.8336,
      "step": 1190
    },
    {
      "epoch": 0.7623604465709729,
      "grad_norm": 1.1089948415756226,
      "learning_rate": 4.8640936694420245e-05,
      "loss": 32.8279,
      "step": 1195
    },
    {
      "epoch": 0.7655502392344498,
      "grad_norm": 1.0508229732513428,
      "learning_rate": 4.7415426898309704e-05,
      "loss": 32.8212,
      "step": 1200
    },
    {
      "epoch": 0.7687400318979266,
      "grad_norm": 1.102999210357666,
      "learning_rate": 4.62026485880761e-05,
      "loss": 32.8288,
      "step": 1205
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 1.1736469268798828,
      "learning_rate": 4.50027522782603e-05,
      "loss": 32.8339,
      "step": 1210
    },
    {
      "epoch": 0.7751196172248804,
      "grad_norm": 1.3593286275863647,
      "learning_rate": 4.3815886884654136e-05,
      "loss": 32.8248,
      "step": 1215
    },
    {
      "epoch": 0.7783094098883573,
      "grad_norm": 1.3843350410461426,
      "learning_rate": 4.264219970581854e-05,
      "loss": 32.8508,
      "step": 1220
    },
    {
      "epoch": 0.7814992025518341,
      "grad_norm": 1.1827675104141235,
      "learning_rate": 4.148183640480293e-05,
      "loss": 32.8585,
      "step": 1225
    },
    {
      "epoch": 0.784688995215311,
      "grad_norm": 1.0259432792663574,
      "learning_rate": 4.0334940991067276e-05,
      "loss": 32.846,
      "step": 1230
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9389122724533081,
      "learning_rate": 3.920165580260973e-05,
      "loss": 32.8456,
      "step": 1235
    },
    {
      "epoch": 0.7910685805422647,
      "grad_norm": 0.9574523568153381,
      "learning_rate": 3.808212148830095e-05,
      "loss": 32.8244,
      "step": 1240
    },
    {
      "epoch": 0.7942583732057417,
      "grad_norm": 1.0494085550308228,
      "learning_rate": 3.697647699042918e-05,
      "loss": 32.8273,
      "step": 1245
    },
    {
      "epoch": 0.7974481658692185,
      "grad_norm": 1.1580866575241089,
      "learning_rate": 3.5884859527455995e-05,
      "loss": 32.8352,
      "step": 1250
    },
    {
      "epoch": 0.8006379585326954,
      "grad_norm": 1.2154430150985718,
      "learning_rate": 3.48074045769868e-05,
      "loss": 32.8394,
      "step": 1255
    },
    {
      "epoch": 0.8038277511961722,
      "grad_norm": 1.0774935483932495,
      "learning_rate": 3.3744245858956967e-05,
      "loss": 32.8679,
      "step": 1260
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 1.0384732484817505,
      "learning_rate": 3.2695515319036186e-05,
      "loss": 32.8412,
      "step": 1265
    },
    {
      "epoch": 0.810207336523126,
      "grad_norm": 0.9459766149520874,
      "learning_rate": 3.1661343112253304e-05,
      "loss": 32.8298,
      "step": 1270
    },
    {
      "epoch": 0.8133971291866029,
      "grad_norm": 1.0627589225769043,
      "learning_rate": 3.064185758684265e-05,
      "loss": 32.813,
      "step": 1275
    },
    {
      "epoch": 0.8165869218500797,
      "grad_norm": 0.955592691898346,
      "learning_rate": 2.96371852683157e-05,
      "loss": 32.8367,
      "step": 1280
    },
    {
      "epoch": 0.8197767145135566,
      "grad_norm": 1.0569384098052979,
      "learning_rate": 2.8647450843757897e-05,
      "loss": 32.8481,
      "step": 1285
    },
    {
      "epoch": 0.8229665071770335,
      "grad_norm": 1.1720012426376343,
      "learning_rate": 2.7672777146354246e-05,
      "loss": 32.8415,
      "step": 1290
    },
    {
      "epoch": 0.8261562998405104,
      "grad_norm": 1.108202576637268,
      "learning_rate": 2.6713285140144802e-05,
      "loss": 32.8471,
      "step": 1295
    },
    {
      "epoch": 0.8293460925039873,
      "grad_norm": 1.3925212621688843,
      "learning_rate": 2.5769093905012333e-05,
      "loss": 32.8317,
      "step": 1300
    },
    {
      "epoch": 0.8325358851674641,
      "grad_norm": 1.017764687538147,
      "learning_rate": 2.4840320621903253e-05,
      "loss": 32.8244,
      "step": 1305
    },
    {
      "epoch": 0.835725677830941,
      "grad_norm": 1.1128332614898682,
      "learning_rate": 2.392708055828495e-05,
      "loss": 32.8317,
      "step": 1310
    },
    {
      "epoch": 0.8389154704944178,
      "grad_norm": 1.0427892208099365,
      "learning_rate": 2.3029487053840295e-05,
      "loss": 32.8307,
      "step": 1315
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.95958411693573,
      "learning_rate": 2.214765150640108e-05,
      "loss": 32.8408,
      "step": 1320
    },
    {
      "epoch": 0.8452950558213717,
      "grad_norm": 1.0608030557632446,
      "learning_rate": 2.1281683358122996e-05,
      "loss": 32.8259,
      "step": 1325
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.8998379707336426,
      "learning_rate": 2.043169008190289e-05,
      "loss": 32.8468,
      "step": 1330
    },
    {
      "epoch": 0.8516746411483254,
      "grad_norm": 0.9696708917617798,
      "learning_rate": 1.9597777168040872e-05,
      "loss": 32.8465,
      "step": 1335
    },
    {
      "epoch": 0.8548644338118022,
      "grad_norm": 0.9957761764526367,
      "learning_rate": 1.8780048111147776e-05,
      "loss": 32.8459,
      "step": 1340
    },
    {
      "epoch": 0.8580542264752791,
      "grad_norm": 1.0414236783981323,
      "learning_rate": 1.797860439730126e-05,
      "loss": 32.7988,
      "step": 1345
    },
    {
      "epoch": 0.861244019138756,
      "grad_norm": 1.0067895650863647,
      "learning_rate": 1.7193545491450183e-05,
      "loss": 32.8391,
      "step": 1350
    },
    {
      "epoch": 0.8644338118022329,
      "grad_norm": 0.9359643459320068,
      "learning_rate": 1.6424968825070567e-05,
      "loss": 32.8153,
      "step": 1355
    },
    {
      "epoch": 0.8676236044657097,
      "grad_norm": 1.0279393196105957,
      "learning_rate": 1.567296978407353e-05,
      "loss": 32.8272,
      "step": 1360
    },
    {
      "epoch": 0.8708133971291866,
      "grad_norm": 0.9850777387619019,
      "learning_rate": 1.4937641696967245e-05,
      "loss": 32.8449,
      "step": 1365
    },
    {
      "epoch": 0.8740031897926634,
      "grad_norm": 1.01199209690094,
      "learning_rate": 1.4219075823274251e-05,
      "loss": 32.8309,
      "step": 1370
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.9314611554145813,
      "learning_rate": 1.3517361342205295e-05,
      "loss": 32.8226,
      "step": 1375
    },
    {
      "epoch": 0.8803827751196173,
      "grad_norm": 0.9059179425239563,
      "learning_rate": 1.2832585341591844e-05,
      "loss": 32.8015,
      "step": 1380
    },
    {
      "epoch": 0.8835725677830941,
      "grad_norm": 1.2151134014129639,
      "learning_rate": 1.2164832807077585e-05,
      "loss": 32.833,
      "step": 1385
    },
    {
      "epoch": 0.886762360446571,
      "grad_norm": 1.0453729629516602,
      "learning_rate": 1.151418661157122e-05,
      "loss": 32.824,
      "step": 1390
    },
    {
      "epoch": 0.8899521531100478,
      "grad_norm": 0.9662382006645203,
      "learning_rate": 1.0880727504961339e-05,
      "loss": 32.8157,
      "step": 1395
    },
    {
      "epoch": 0.8931419457735247,
      "grad_norm": 1.0310821533203125,
      "learning_rate": 1.0264534104094812e-05,
      "loss": 32.836,
      "step": 1400
    },
    {
      "epoch": 0.8963317384370016,
      "grad_norm": 0.9117385745048523,
      "learning_rate": 9.665682883019732e-06,
      "loss": 32.8227,
      "step": 1405
    },
    {
      "epoch": 0.8995215311004785,
      "grad_norm": 0.8850731253623962,
      "learning_rate": 9.084248163494511e-06,
      "loss": 32.8276,
      "step": 1410
    },
    {
      "epoch": 0.9027113237639554,
      "grad_norm": 1.0270276069641113,
      "learning_rate": 8.520302105764148e-06,
      "loss": 32.8239,
      "step": 1415
    },
    {
      "epoch": 0.9059011164274322,
      "grad_norm": 0.9645489454269409,
      "learning_rate": 7.973914699604367e-06,
      "loss": 32.8235,
      "step": 1420
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 1.0180425643920898,
      "learning_rate": 7.445153755635569e-06,
      "loss": 32.834,
      "step": 1425
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.9252223372459412,
      "learning_rate": 6.934084896906983e-06,
      "loss": 32.8312,
      "step": 1430
    },
    {
      "epoch": 0.9154704944178629,
      "grad_norm": 1.1004729270935059,
      "learning_rate": 6.440771550752377e-06,
      "loss": 32.8397,
      "step": 1435
    },
    {
      "epoch": 0.9186602870813397,
      "grad_norm": 0.9907602071762085,
      "learning_rate": 5.965274940918274e-06,
      "loss": 32.8319,
      "step": 1440
    },
    {
      "epoch": 0.9218500797448166,
      "grad_norm": 0.9156322479248047,
      "learning_rate": 5.507654079965612e-06,
      "loss": 32.8098,
      "step": 1445
    },
    {
      "epoch": 0.9250398724082934,
      "grad_norm": 1.07799232006073,
      "learning_rate": 5.067965761945869e-06,
      "loss": 32.8379,
      "step": 1450
    },
    {
      "epoch": 0.9282296650717703,
      "grad_norm": 0.8984599113464355,
      "learning_rate": 4.646264555352586e-06,
      "loss": 32.8193,
      "step": 1455
    },
    {
      "epoch": 0.9314194577352473,
      "grad_norm": 0.9720312356948853,
      "learning_rate": 4.242602796348915e-06,
      "loss": 32.8124,
      "step": 1460
    },
    {
      "epoch": 0.9346092503987241,
      "grad_norm": 1.0742676258087158,
      "learning_rate": 3.857030582272369e-06,
      "loss": 32.8109,
      "step": 1465
    },
    {
      "epoch": 0.937799043062201,
      "grad_norm": 1.0529791116714478,
      "learning_rate": 3.489595765417441e-06,
      "loss": 32.7938,
      "step": 1470
    },
    {
      "epoch": 0.9409888357256778,
      "grad_norm": 0.9238316416740417,
      "learning_rate": 3.140343947096624e-06,
      "loss": 32.7966,
      "step": 1475
    },
    {
      "epoch": 0.9441786283891547,
      "grad_norm": 0.9737393260002136,
      "learning_rate": 2.80931847198117e-06,
      "loss": 32.8523,
      "step": 1480
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 1.0758775472640991,
      "learning_rate": 2.4965604227215774e-06,
      "loss": 32.8347,
      "step": 1485
    },
    {
      "epoch": 0.9505582137161085,
      "grad_norm": 0.958068311214447,
      "learning_rate": 2.202108614848885e-06,
      "loss": 32.8136,
      "step": 1490
    },
    {
      "epoch": 0.9537480063795853,
      "grad_norm": 0.9180477857589722,
      "learning_rate": 1.925999591957561e-06,
      "loss": 32.813,
      "step": 1495
    },
    {
      "epoch": 0.9569377990430622,
      "grad_norm": 0.9640750288963318,
      "learning_rate": 1.6682676211700107e-06,
      "loss": 32.8268,
      "step": 1500
    },
    {
      "epoch": 0.960127591706539,
      "grad_norm": 0.9315332174301147,
      "learning_rate": 1.4289446888838652e-06,
      "loss": 32.8336,
      "step": 1505
    },
    {
      "epoch": 0.9633173843700159,
      "grad_norm": 1.051619529724121,
      "learning_rate": 1.2080604968022378e-06,
      "loss": 32.8047,
      "step": 1510
    },
    {
      "epoch": 0.9665071770334929,
      "grad_norm": 0.9639779925346375,
      "learning_rate": 1.0056424582474575e-06,
      "loss": 32.8118,
      "step": 1515
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.9129908680915833,
      "learning_rate": 8.217156947590064e-07,
      "loss": 32.8089,
      "step": 1520
    },
    {
      "epoch": 0.9728867623604466,
      "grad_norm": 1.0344650745391846,
      "learning_rate": 6.563030329755969e-07,
      "loss": 32.8023,
      "step": 1525
    },
    {
      "epoch": 0.9760765550239234,
      "grad_norm": 0.9805324077606201,
      "learning_rate": 5.094250018023715e-07,
      "loss": 32.8376,
      "step": 1530
    },
    {
      "epoch": 0.9792663476874003,
      "grad_norm": 0.9154828190803528,
      "learning_rate": 3.8109982986300747e-07,
      "loss": 32.8375,
      "step": 1535
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.9236858487129211,
      "learning_rate": 2.7134344323747616e-07,
      "loss": 32.808,
      "step": 1540
    },
    {
      "epoch": 0.9856459330143541,
      "grad_norm": 0.9518983364105225,
      "learning_rate": 1.801694634854578e-07,
      "loss": 32.8399,
      "step": 1545
    },
    {
      "epoch": 0.988835725677831,
      "grad_norm": 0.9915338158607483,
      "learning_rate": 1.075892059558603e-07,
      "loss": 32.8379,
      "step": 1550
    },
    {
      "epoch": 0.9920255183413078,
      "grad_norm": 0.9686278700828552,
      "learning_rate": 5.3611678382442516e-08,
      "loss": 32.8208,
      "step": 1555
    },
    {
      "epoch": 0.9952153110047847,
      "grad_norm": 0.9356927871704102,
      "learning_rate": 1.824357976594193e-08,
      "loss": 32.7948,
      "step": 1560
    },
    {
      "epoch": 0.9984051036682615,
      "grad_norm": 0.9721894264221191,
      "learning_rate": 1.4892995426396548e-09,
      "loss": 32.8196,
      "step": 1565
    },
    {
      "epoch": 0.9996810207336523,
      "eval_loss": 34.31958770751953,
      "eval_runtime": 2.9694,
      "eval_samples_per_second": 3.368,
      "eval_steps_per_second": 0.674,
      "step": 1567
    },
    {
      "epoch": 0.9996810207336523,
      "step": 1567,
      "total_flos": 6.008655759456338e+17,
      "train_loss": 33.50232638594598,
      "train_runtime": 9663.2811,
      "train_samples_per_second": 1.297,
      "train_steps_per_second": 0.162
    }
  ],
  "logging_steps": 5,
  "max_steps": 1567,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 6.008655759456338e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}