{
  "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": 683.0049438476562,
      "learning_rate": 1.910828025477707e-06,
      "loss": 45.2944,
      "step": 1
    },
    {
      "epoch": 0.003189792663476874,
      "grad_norm": 519.6649780273438,
      "learning_rate": 9.554140127388534e-06,
      "loss": 45.6093,
      "step": 5
    },
    {
      "epoch": 0.006379585326953748,
      "grad_norm": 189.5433349609375,
      "learning_rate": 1.9108280254777068e-05,
      "loss": 32.9532,
      "step": 10
    },
    {
      "epoch": 0.009569377990430622,
      "grad_norm": 63.46455001831055,
      "learning_rate": 2.8662420382165606e-05,
      "loss": 23.1174,
      "step": 15
    },
    {
      "epoch": 0.012759170653907496,
      "grad_norm": 18.510099411010742,
      "learning_rate": 3.8216560509554137e-05,
      "loss": 20.6226,
      "step": 20
    },
    {
      "epoch": 0.01594896331738437,
      "grad_norm": 10.55661678314209,
      "learning_rate": 4.777070063694267e-05,
      "loss": 19.5559,
      "step": 25
    },
    {
      "epoch": 0.019138755980861243,
      "grad_norm": 14.438572883605957,
      "learning_rate": 5.732484076433121e-05,
      "loss": 18.0152,
      "step": 30
    },
    {
      "epoch": 0.022328548644338118,
      "grad_norm": 36.82223129272461,
      "learning_rate": 6.687898089171974e-05,
      "loss": 15.7287,
      "step": 35
    },
    {
      "epoch": 0.025518341307814992,
      "grad_norm": 75.36656951904297,
      "learning_rate": 7.643312101910827e-05,
      "loss": 10.9057,
      "step": 40
    },
    {
      "epoch": 0.028708133971291867,
      "grad_norm": 9.559320449829102,
      "learning_rate": 8.59872611464968e-05,
      "loss": 2.8213,
      "step": 45
    },
    {
      "epoch": 0.03189792663476874,
      "grad_norm": 4.798595428466797,
      "learning_rate": 9.554140127388533e-05,
      "loss": 2.1012,
      "step": 50
    },
    {
      "epoch": 0.03508771929824561,
      "grad_norm": 3.041489601135254,
      "learning_rate": 0.00010509554140127387,
      "loss": 1.751,
      "step": 55
    },
    {
      "epoch": 0.03827751196172249,
      "grad_norm": 10.074992179870605,
      "learning_rate": 0.00011464968152866242,
      "loss": 1.629,
      "step": 60
    },
    {
      "epoch": 0.04146730462519936,
      "grad_norm": 4.232330322265625,
      "learning_rate": 0.00012420382165605095,
      "loss": 1.5111,
      "step": 65
    },
    {
      "epoch": 0.044657097288676235,
      "grad_norm": 3.0990259647369385,
      "learning_rate": 0.00013375796178343948,
      "loss": 1.3541,
      "step": 70
    },
    {
      "epoch": 0.04784688995215311,
      "grad_norm": 3.949800968170166,
      "learning_rate": 0.00014331210191082802,
      "loss": 1.314,
      "step": 75
    },
    {
      "epoch": 0.051036682615629984,
      "grad_norm": 8.635440826416016,
      "learning_rate": 0.00015286624203821655,
      "loss": 1.28,
      "step": 80
    },
    {
      "epoch": 0.05422647527910686,
      "grad_norm": 2.358060359954834,
      "learning_rate": 0.00016242038216560508,
      "loss": 1.231,
      "step": 85
    },
    {
      "epoch": 0.05741626794258373,
      "grad_norm": 7.330495357513428,
      "learning_rate": 0.0001719745222929936,
      "loss": 1.1927,
      "step": 90
    },
    {
      "epoch": 0.06060606060606061,
      "grad_norm": 3.866813898086548,
      "learning_rate": 0.00018152866242038214,
      "loss": 1.1379,
      "step": 95
    },
    {
      "epoch": 0.06379585326953748,
      "grad_norm": 2.174020528793335,
      "learning_rate": 0.00019108280254777067,
      "loss": 1.1335,
      "step": 100
    },
    {
      "epoch": 0.06698564593301436,
      "grad_norm": 1.3060088157653809,
      "learning_rate": 0.0002006369426751592,
      "loss": 1.1068,
      "step": 105
    },
    {
      "epoch": 0.07017543859649122,
      "grad_norm": 4.537840843200684,
      "learning_rate": 0.00021019108280254773,
      "loss": 1.0773,
      "step": 110
    },
    {
      "epoch": 0.0733652312599681,
      "grad_norm": 51.75442886352539,
      "learning_rate": 0.00021974522292993626,
      "loss": 1.1051,
      "step": 115
    },
    {
      "epoch": 0.07655502392344497,
      "grad_norm": 1.6401653289794922,
      "learning_rate": 0.00022929936305732485,
      "loss": 1.0971,
      "step": 120
    },
    {
      "epoch": 0.07974481658692185,
      "grad_norm": 1.8242731094360352,
      "learning_rate": 0.00023885350318471338,
      "loss": 1.0969,
      "step": 125
    },
    {
      "epoch": 0.08293460925039872,
      "grad_norm": 3.504967212677002,
      "learning_rate": 0.0002484076433121019,
      "loss": 1.1196,
      "step": 130
    },
    {
      "epoch": 0.0861244019138756,
      "grad_norm": 11.742030143737793,
      "learning_rate": 0.00025796178343949044,
      "loss": 1.0801,
      "step": 135
    },
    {
      "epoch": 0.08931419457735247,
      "grad_norm": 8.536433219909668,
      "learning_rate": 0.00026751592356687897,
      "loss": 1.0728,
      "step": 140
    },
    {
      "epoch": 0.09250398724082935,
      "grad_norm": 2.490248441696167,
      "learning_rate": 0.0002770700636942675,
      "loss": 1.0875,
      "step": 145
    },
    {
      "epoch": 0.09569377990430622,
      "grad_norm": 2.524031639099121,
      "learning_rate": 0.00028662420382165603,
      "loss": 1.0371,
      "step": 150
    },
    {
      "epoch": 0.09888357256778309,
      "grad_norm": 3.3123767375946045,
      "learning_rate": 0.00029617834394904456,
      "loss": 1.049,
      "step": 155
    },
    {
      "epoch": 0.10207336523125997,
      "grad_norm": 1.9446417093276978,
      "learning_rate": 0.0002999966490829603,
      "loss": 1.0379,
      "step": 160
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 1.845177173614502,
      "learning_rate": 0.00029997617179878324,
      "loss": 1.0484,
      "step": 165
    },
    {
      "epoch": 0.10845295055821372,
      "grad_norm": 4.8916015625,
      "learning_rate": 0.0002999370813893047,
      "loss": 1.0585,
      "step": 170
    },
    {
      "epoch": 0.11164274322169059,
      "grad_norm": 23.282398223876953,
      "learning_rate": 0.00029987938270592676,
      "loss": 1.0488,
      "step": 175
    },
    {
      "epoch": 0.11483253588516747,
      "grad_norm": 4.743477821350098,
      "learning_rate": 0.0002998030829094724,
      "loss": 1.0479,
      "step": 180
    },
    {
      "epoch": 0.11802232854864433,
      "grad_norm": 2.9644861221313477,
      "learning_rate": 0.00029970819146929694,
      "loss": 1.0125,
      "step": 185
    },
    {
      "epoch": 0.12121212121212122,
      "grad_norm": 1.3036428689956665,
      "learning_rate": 0.0002995947201621131,
      "loss": 1.0429,
      "step": 190
    },
    {
      "epoch": 0.12440191387559808,
      "grad_norm": 12.80103588104248,
      "learning_rate": 0.00029946268307052886,
      "loss": 1.031,
      "step": 195
    },
    {
      "epoch": 0.12759170653907495,
      "grad_norm": 3.6281776428222656,
      "learning_rate": 0.0002993120965813003,
      "loss": 1.046,
      "step": 200
    },
    {
      "epoch": 0.13078149920255183,
      "grad_norm": 6.867593765258789,
      "learning_rate": 0.0002991429793832975,
      "loss": 1.0034,
      "step": 205
    },
    {
      "epoch": 0.1339712918660287,
      "grad_norm": 4.86781120300293,
      "learning_rate": 0.00029895535246518524,
      "loss": 1.0558,
      "step": 210
    },
    {
      "epoch": 0.1371610845295056,
      "grad_norm": 1.8913637399673462,
      "learning_rate": 0.0002987492391128182,
      "loss": 1.0404,
      "step": 215
    },
    {
      "epoch": 0.14035087719298245,
      "grad_norm": 1.4582362174987793,
      "learning_rate": 0.0002985246649063509,
      "loss": 1.0121,
      "step": 220
    },
    {
      "epoch": 0.14354066985645933,
      "grad_norm": 4.474733352661133,
      "learning_rate": 0.0002982816577170631,
      "loss": 0.9997,
      "step": 225
    },
    {
      "epoch": 0.1467304625199362,
      "grad_norm": 1.3253235816955566,
      "learning_rate": 0.00029802024770390087,
      "loss": 1.0034,
      "step": 230
    },
    {
      "epoch": 0.14992025518341306,
      "grad_norm": 1.452958106994629,
      "learning_rate": 0.00029774046730973334,
      "loss": 0.9983,
      "step": 235
    },
    {
      "epoch": 0.15311004784688995,
      "grad_norm": 2.8578474521636963,
      "learning_rate": 0.00029744235125732664,
      "loss": 1.0123,
      "step": 240
    },
    {
      "epoch": 0.15629984051036683,
      "grad_norm": 4.637836933135986,
      "learning_rate": 0.0002971259365450344,
      "loss": 1.0427,
      "step": 245
    },
    {
      "epoch": 0.1594896331738437,
      "grad_norm": 1.3336002826690674,
      "learning_rate": 0.00029679126244220596,
      "loss": 0.9828,
      "step": 250
    },
    {
      "epoch": 0.16267942583732056,
      "grad_norm": 2.1084749698638916,
      "learning_rate": 0.00029643837048431293,
      "loss": 0.9907,
      "step": 255
    },
    {
      "epoch": 0.16586921850079744,
      "grad_norm": 1.7289551496505737,
      "learning_rate": 0.0002960673044677939,
      "loss": 1.0017,
      "step": 260
    },
    {
      "epoch": 0.16905901116427433,
      "grad_norm": 1.0768038034439087,
      "learning_rate": 0.00029567811044461977,
      "loss": 0.9639,
      "step": 265
    },
    {
      "epoch": 0.1722488038277512,
      "grad_norm": 3.015089750289917,
      "learning_rate": 0.00029527083671657746,
      "loss": 0.9858,
      "step": 270
    },
    {
      "epoch": 0.17543859649122806,
      "grad_norm": 3.8819258213043213,
      "learning_rate": 0.00029484553382927594,
      "loss": 0.9907,
      "step": 275
    },
    {
      "epoch": 0.17862838915470494,
      "grad_norm": 1.3600151538848877,
      "learning_rate": 0.000294402254565873,
      "loss": 1.0016,
      "step": 280
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 2.409559488296509,
      "learning_rate": 0.00029394105394052434,
      "loss": 1.0042,
      "step": 285
    },
    {
      "epoch": 0.1850079744816587,
      "grad_norm": 2.906012773513794,
      "learning_rate": 0.00029346198919155616,
      "loss": 1.0192,
      "step": 290
    },
    {
      "epoch": 0.18819776714513556,
      "grad_norm": 3.697054386138916,
      "learning_rate": 0.00029296511977436107,
      "loss": 1.0157,
      "step": 295
    },
    {
      "epoch": 0.19138755980861244,
      "grad_norm": 2.336791753768921,
      "learning_rate": 0.0002924505073540198,
      "loss": 0.992,
      "step": 300
    },
    {
      "epoch": 0.19457735247208932,
      "grad_norm": 3.517256259918213,
      "learning_rate": 0.0002919182157976476,
      "loss": 0.9753,
      "step": 305
    },
    {
      "epoch": 0.19776714513556617,
      "grad_norm": 4.080871105194092,
      "learning_rate": 0.00029136831116646815,
      "loss": 0.9739,
      "step": 310
    },
    {
      "epoch": 0.20095693779904306,
      "grad_norm": 1.9315543174743652,
      "learning_rate": 0.000290800861707615,
      "loss": 0.959,
      "step": 315
    },
    {
      "epoch": 0.20414673046251994,
      "grad_norm": 4.471941947937012,
      "learning_rate": 0.00029021593784566113,
      "loss": 0.9718,
      "step": 320
    },
    {
      "epoch": 0.20733652312599682,
      "grad_norm": 4.151180744171143,
      "learning_rate": 0.0002896136121738793,
      "loss": 0.9705,
      "step": 325
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 2.165132999420166,
      "learning_rate": 0.0002889939594452323,
      "loss": 0.9575,
      "step": 330
    },
    {
      "epoch": 0.21371610845295055,
      "grad_norm": 1.3790881633758545,
      "learning_rate": 0.00028835705656309583,
      "loss": 0.9773,
      "step": 335
    },
    {
      "epoch": 0.21690590111642744,
      "grad_norm": 1.334600806236267,
      "learning_rate": 0.0002877029825717142,
      "loss": 0.9818,
      "step": 340
    },
    {
      "epoch": 0.22009569377990432,
      "grad_norm": 2.155672788619995,
      "learning_rate": 0.0002870318186463901,
      "loss": 0.9487,
      "step": 345
    },
    {
      "epoch": 0.22328548644338117,
      "grad_norm": 2.6333436965942383,
      "learning_rate": 0.0002863436480834105,
      "loss": 0.9999,
      "step": 350
    },
    {
      "epoch": 0.22647527910685805,
      "grad_norm": 1.2809535264968872,
      "learning_rate": 0.00028563855628970886,
      "loss": 0.967,
      "step": 355
    },
    {
      "epoch": 0.22966507177033493,
      "grad_norm": 2.015334367752075,
      "learning_rate": 0.0002849166307722653,
      "loss": 0.9965,
      "step": 360
    },
    {
      "epoch": 0.23285486443381181,
      "grad_norm": 1.6532715559005737,
      "learning_rate": 0.00028417796112724684,
      "loss": 0.9949,
      "step": 365
    },
    {
      "epoch": 0.23604465709728867,
      "grad_norm": 1.243087649345398,
      "learning_rate": 0.0002834226390288873,
      "loss": 0.9407,
      "step": 370
    },
    {
      "epoch": 0.23923444976076555,
      "grad_norm": 2.077500104904175,
      "learning_rate": 0.0002826507582181103,
      "loss": 0.9508,
      "step": 375
    },
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 1.656862497329712,
      "learning_rate": 0.00028186241449089524,
      "loss": 0.9701,
      "step": 380
    },
    {
      "epoch": 0.24561403508771928,
      "grad_norm": 1.4174304008483887,
      "learning_rate": 0.000281057705686388,
      "loss": 0.9649,
      "step": 385
    },
    {
      "epoch": 0.24880382775119617,
      "grad_norm": 2.1187663078308105,
      "learning_rate": 0.0002802367316747589,
      "loss": 0.9655,
      "step": 390
    },
    {
      "epoch": 0.25199362041467305,
      "grad_norm": 2.494663953781128,
      "learning_rate": 0.0002793995943448078,
      "loss": 0.9422,
      "step": 395
    },
    {
      "epoch": 0.2551834130781499,
      "grad_norm": 2.778376579284668,
      "learning_rate": 0.00027854639759131893,
      "loss": 0.9518,
      "step": 400
    },
    {
      "epoch": 0.2583732057416268,
      "grad_norm": 1.4692327976226807,
      "learning_rate": 0.00027767724730216696,
      "loss": 0.9773,
      "step": 405
    },
    {
      "epoch": 0.26156299840510366,
      "grad_norm": 2.749265670776367,
      "learning_rate": 0.0002767922513451754,
      "loss": 0.9687,
      "step": 410
    },
    {
      "epoch": 0.2647527910685805,
      "grad_norm": 1.0992218255996704,
      "learning_rate": 0.00027589151955472965,
      "loss": 0.9535,
      "step": 415
    },
    {
      "epoch": 0.2679425837320574,
      "grad_norm": 1.787949800491333,
      "learning_rate": 0.00027497516371814543,
      "loss": 0.9435,
      "step": 420
    },
    {
      "epoch": 0.2711323763955343,
      "grad_norm": 5.4542412757873535,
      "learning_rate": 0.00027404329756179537,
      "loss": 0.9498,
      "step": 425
    },
    {
      "epoch": 0.2743221690590112,
      "grad_norm": 2.4093711376190186,
      "learning_rate": 0.0002730960367369949,
      "loss": 1.0197,
      "step": 430
    },
    {
      "epoch": 0.27751196172248804,
      "grad_norm": 2.159757375717163,
      "learning_rate": 0.00027213349880564873,
      "loss": 0.9865,
      "step": 435
    },
    {
      "epoch": 0.2807017543859649,
      "grad_norm": 2.393873691558838,
      "learning_rate": 0.0002711558032256607,
      "loss": 0.9587,
      "step": 440
    },
    {
      "epoch": 0.2838915470494418,
      "grad_norm": 2.645077705383301,
      "learning_rate": 0.0002701630713361085,
      "loss": 0.9645,
      "step": 445
    },
    {
      "epoch": 0.28708133971291866,
      "grad_norm": 1.9091484546661377,
      "learning_rate": 0.00026915542634218403,
      "loss": 0.9629,
      "step": 450
    },
    {
      "epoch": 0.2902711323763955,
      "grad_norm": 1.352075457572937,
      "learning_rate": 0.00026813299329990335,
      "loss": 0.9761,
      "step": 455
    },
    {
      "epoch": 0.2934609250398724,
      "grad_norm": 2.6942131519317627,
      "learning_rate": 0.0002670958991005859,
      "loss": 0.9531,
      "step": 460
    },
    {
      "epoch": 0.2966507177033493,
      "grad_norm": 2.5837950706481934,
      "learning_rate": 0.0002660442724551065,
      "loss": 0.9489,
      "step": 465
    },
    {
      "epoch": 0.29984051036682613,
      "grad_norm": 2.5745198726654053,
      "learning_rate": 0.00026497824387792146,
      "loss": 0.9677,
      "step": 470
    },
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 1.8516491651535034,
      "learning_rate": 0.00026389794567087085,
      "loss": 0.9664,
      "step": 475
    },
    {
      "epoch": 0.3062200956937799,
      "grad_norm": 2.2351913452148438,
      "learning_rate": 0.0002628035119067586,
      "loss": 0.9425,
      "step": 480
    },
    {
      "epoch": 0.3094098883572568,
      "grad_norm": 1.51954185962677,
      "learning_rate": 0.0002616950784127135,
      "loss": 0.9261,
      "step": 485
    },
    {
      "epoch": 0.31259968102073366,
      "grad_norm": 1.722433090209961,
      "learning_rate": 0.00026057278275333165,
      "loss": 0.973,
      "step": 490
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 6.133448600769043,
      "learning_rate": 0.00025943676421360395,
      "loss": 0.9492,
      "step": 495
    },
    {
      "epoch": 0.3189792663476874,
      "grad_norm": 2.187849283218384,
      "learning_rate": 0.00025828716378163,
      "loss": 0.9612,
      "step": 500
    },
    {
      "epoch": 0.32216905901116427,
      "grad_norm": 2.13073468208313,
      "learning_rate": 0.00025712412413112006,
      "loss": 0.9485,
      "step": 505
    },
    {
      "epoch": 0.3253588516746411,
      "grad_norm": 13.925029754638672,
      "learning_rate": 0.00025594778960368844,
      "loss": 0.9697,
      "step": 510
    },
    {
      "epoch": 0.32854864433811803,
      "grad_norm": 1.2951607704162598,
      "learning_rate": 0.0002547583061909396,
      "loss": 0.9433,
      "step": 515
    },
    {
      "epoch": 0.3317384370015949,
      "grad_norm": 1.386167287826538,
      "learning_rate": 0.00025355582151634956,
      "loss": 0.9613,
      "step": 520
    },
    {
      "epoch": 0.3349282296650718,
      "grad_norm": 3.0757944583892822,
      "learning_rate": 0.00025234048481694477,
      "loss": 0.9238,
      "step": 525
    },
    {
      "epoch": 0.33811802232854865,
      "grad_norm": 2.6696460247039795,
      "learning_rate": 0.0002511124469247809,
      "loss": 0.956,
      "step": 530
    },
    {
      "epoch": 0.3413078149920255,
      "grad_norm": 1.5839197635650635,
      "learning_rate": 0.00024987186024822295,
      "loss": 0.9467,
      "step": 535
    },
    {
      "epoch": 0.3444976076555024,
      "grad_norm": 1.6664620637893677,
      "learning_rate": 0.0002486188787530309,
      "loss": 0.9578,
      "step": 540
    },
    {
      "epoch": 0.34768740031897927,
      "grad_norm": 1.2985708713531494,
      "learning_rate": 0.00024735365794325117,
      "loss": 0.9704,
      "step": 545
    },
    {
      "epoch": 0.3508771929824561,
      "grad_norm": 2.3431661128997803,
      "learning_rate": 0.0002460763548419172,
      "loss": 0.9408,
      "step": 550
    },
    {
      "epoch": 0.35406698564593303,
      "grad_norm": 1.8065603971481323,
      "learning_rate": 0.0002447871279715624,
      "loss": 0.9543,
      "step": 555
    },
    {
      "epoch": 0.3572567783094099,
      "grad_norm": 2.896728277206421,
      "learning_rate": 0.00024348613733454565,
      "loss": 0.948,
      "step": 560
    },
    {
      "epoch": 0.36044657097288674,
      "grad_norm": 2.205735206604004,
      "learning_rate": 0.00024217354439319427,
      "loss": 0.9263,
      "step": 565
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 2.1975135803222656,
      "learning_rate": 0.00024084951204976528,
      "loss": 0.9091,
      "step": 570
    },
    {
      "epoch": 0.3668261562998405,
      "grad_norm": 1.9090759754180908,
      "learning_rate": 0.0002395142046262281,
      "loss": 0.9426,
      "step": 575
    },
    {
      "epoch": 0.3700159489633174,
      "grad_norm": 2.041628360748291,
      "learning_rate": 0.00023816778784387094,
      "loss": 0.9296,
      "step": 580
    },
    {
      "epoch": 0.37320574162679426,
      "grad_norm": 1.0868958234786987,
      "learning_rate": 0.0002368104288027336,
      "loss": 0.9225,
      "step": 585
    },
    {
      "epoch": 0.3763955342902711,
      "grad_norm": 1.3771106004714966,
      "learning_rate": 0.0002354422959608692,
      "loss": 0.9267,
      "step": 590
    },
    {
      "epoch": 0.379585326953748,
      "grad_norm": 1.0883209705352783,
      "learning_rate": 0.00023406355911343717,
      "loss": 0.9482,
      "step": 595
    },
    {
      "epoch": 0.3827751196172249,
      "grad_norm": 4.788959980010986,
      "learning_rate": 0.00023267438937163077,
      "loss": 0.9281,
      "step": 600
    },
    {
      "epoch": 0.38596491228070173,
      "grad_norm": 1.3784050941467285,
      "learning_rate": 0.00023127495914144051,
      "loss": 0.9551,
      "step": 605
    },
    {
      "epoch": 0.38915470494417864,
      "grad_norm": 3.7434804439544678,
      "learning_rate": 0.00022986544210225774,
      "loss": 0.931,
      "step": 610
    },
    {
      "epoch": 0.3923444976076555,
      "grad_norm": 1.4201207160949707,
      "learning_rate": 0.00022844601318531955,
      "loss": 0.9154,
      "step": 615
    },
    {
      "epoch": 0.39553429027113235,
      "grad_norm": 2.0786445140838623,
      "learning_rate": 0.00022701684855199857,
      "loss": 0.9615,
      "step": 620
    },
    {
      "epoch": 0.39872408293460926,
      "grad_norm": 1.343225121498108,
      "learning_rate": 0.00022557812557194,
      "loss": 0.9164,
      "step": 625
    },
    {
      "epoch": 0.4019138755980861,
      "grad_norm": 1.6810039281845093,
      "learning_rate": 0.00022413002280104915,
      "loss": 0.9112,
      "step": 630
    },
    {
      "epoch": 0.405103668261563,
      "grad_norm": 1.2535432577133179,
      "learning_rate": 0.00022267271995933074,
      "loss": 0.9467,
      "step": 635
    },
    {
      "epoch": 0.4082934609250399,
      "grad_norm": 1.297658085823059,
      "learning_rate": 0.00022120639790858482,
      "loss": 0.927,
      "step": 640
    },
    {
      "epoch": 0.41148325358851673,
      "grad_norm": 1.5415024757385254,
      "learning_rate": 0.00021973123862996044,
      "loss": 0.9447,
      "step": 645
    },
    {
      "epoch": 0.41467304625199364,
      "grad_norm": 1.305788278579712,
      "learning_rate": 0.00021824742520137026,
      "loss": 0.9376,
      "step": 650
    },
    {
      "epoch": 0.4178628389154705,
      "grad_norm": 1.043959617614746,
      "learning_rate": 0.00021675514177476945,
      "loss": 0.9255,
      "step": 655
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 1.2613155841827393,
      "learning_rate": 0.0002152545735533012,
      "loss": 0.9493,
      "step": 660
    },
    {
      "epoch": 0.42424242424242425,
      "grad_norm": 2.3178012371063232,
      "learning_rate": 0.00021374590676831136,
      "loss": 0.9336,
      "step": 665
    },
    {
      "epoch": 0.4274322169059011,
      "grad_norm": 1.5422348976135254,
      "learning_rate": 0.00021222932865623605,
      "loss": 0.9411,
      "step": 670
    },
    {
      "epoch": 0.430622009569378,
      "grad_norm": 2.053426504135132,
      "learning_rate": 0.00021070502743536414,
      "loss": 0.9086,
      "step": 675
    },
    {
      "epoch": 0.43381180223285487,
      "grad_norm": 1.4567965269088745,
      "learning_rate": 0.00020917319228247805,
      "loss": 0.9135,
      "step": 680
    },
    {
      "epoch": 0.4370015948963317,
      "grad_norm": 1.7360142469406128,
      "learning_rate": 0.00020763401330937555,
      "loss": 0.9208,
      "step": 685
    },
    {
      "epoch": 0.44019138755980863,
      "grad_norm": 3.2345566749572754,
      "learning_rate": 0.00020608768153927546,
      "loss": 0.9281,
      "step": 690
    },
    {
      "epoch": 0.4433811802232855,
      "grad_norm": 2.1315088272094727,
      "learning_rate": 0.00020453438888311042,
      "loss": 0.9203,
      "step": 695
    },
    {
      "epoch": 0.44657097288676234,
      "grad_norm": 2.226001262664795,
      "learning_rate": 0.00020297432811570916,
      "loss": 0.9319,
      "step": 700
    },
    {
      "epoch": 0.44976076555023925,
      "grad_norm": 2.0687644481658936,
      "learning_rate": 0.00020140769285187187,
      "loss": 0.9079,
      "step": 705
    },
    {
      "epoch": 0.4529505582137161,
      "grad_norm": 1.8376455307006836,
      "learning_rate": 0.00019983467752234132,
      "loss": 0.898,
      "step": 710
    },
    {
      "epoch": 0.45614035087719296,
      "grad_norm": 1.3176559209823608,
      "learning_rate": 0.0001982554773496723,
      "loss": 0.9483,
      "step": 715
    },
    {
      "epoch": 0.45933014354066987,
      "grad_norm": 1.6489756107330322,
      "learning_rate": 0.00019667028832400345,
      "loss": 0.8925,
      "step": 720
    },
    {
      "epoch": 0.4625199362041467,
      "grad_norm": 1.6719319820404053,
      "learning_rate": 0.00019507930717873313,
      "loss": 0.9151,
      "step": 725
    },
    {
      "epoch": 0.46570972886762363,
      "grad_norm": 2.6439127922058105,
      "learning_rate": 0.00019348273136610364,
      "loss": 0.9222,
      "step": 730
    },
    {
      "epoch": 0.4688995215311005,
      "grad_norm": 2.219003677368164,
      "learning_rate": 0.00019188075903269587,
      "loss": 0.9322,
      "step": 735
    },
    {
      "epoch": 0.47208931419457734,
      "grad_norm": 3.740412712097168,
      "learning_rate": 0.00019027358899483776,
      "loss": 0.9102,
      "step": 740
    },
    {
      "epoch": 0.47527910685805425,
      "grad_norm": 1.2723886966705322,
      "learning_rate": 0.00018866142071393013,
      "loss": 0.9193,
      "step": 745
    },
    {
      "epoch": 0.4784688995215311,
      "grad_norm": 2.935451030731201,
      "learning_rate": 0.00018704445427169156,
      "loss": 0.9242,
      "step": 750
    },
    {
      "epoch": 0.48165869218500795,
      "grad_norm": 3.1072025299072266,
      "learning_rate": 0.00018542289034532733,
      "loss": 0.9203,
      "step": 755
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": 1.3583980798721313,
      "learning_rate": 0.00018379693018262349,
      "loss": 0.9202,
      "step": 760
    },
    {
      "epoch": 0.4880382775119617,
      "grad_norm": 1.6865596771240234,
      "learning_rate": 0.00018216677557697083,
      "loss": 0.9088,
      "step": 765
    },
    {
      "epoch": 0.49122807017543857,
      "grad_norm": 1.3977638483047485,
      "learning_rate": 0.00018053262884232078,
      "loss": 0.8964,
      "step": 770
    },
    {
      "epoch": 0.4944178628389155,
      "grad_norm": 1.32778000831604,
      "learning_rate": 0.0001788946927880768,
      "loss": 0.9019,
      "step": 775
    },
    {
      "epoch": 0.49760765550239233,
      "grad_norm": 1.1595113277435303,
      "learning_rate": 0.00017725317069392418,
      "loss": 0.8887,
      "step": 780
    },
    {
      "epoch": 0.5007974481658692,
      "grad_norm": 1.3176358938217163,
      "learning_rate": 0.00017560826628460182,
      "loss": 0.8992,
      "step": 785
    },
    {
      "epoch": 0.5039872408293461,
      "grad_norm": 2.6832964420318604,
      "learning_rate": 0.00017396018370461808,
      "loss": 0.922,
      "step": 790
    },
    {
      "epoch": 0.507177033492823,
      "grad_norm": 2.5020227432250977,
      "learning_rate": 0.00017230912749291547,
      "loss": 0.9321,
      "step": 795
    },
    {
      "epoch": 0.5103668261562998,
      "grad_norm": 2.304370403289795,
      "learning_rate": 0.00017065530255748557,
      "loss": 0.9174,
      "step": 800
    },
    {
      "epoch": 0.5135566188197768,
      "grad_norm": 1.984249472618103,
      "learning_rate": 0.00016899891414993854,
      "loss": 0.9184,
      "step": 805
    },
    {
      "epoch": 0.5167464114832536,
      "grad_norm": 1.0693037509918213,
      "learning_rate": 0.00016734016784002994,
      "loss": 0.9239,
      "step": 810
    },
    {
      "epoch": 0.5199362041467305,
      "grad_norm": 1.1777863502502441,
      "learning_rate": 0.000165679269490148,
      "loss": 0.9004,
      "step": 815
    },
    {
      "epoch": 0.5231259968102073,
      "grad_norm": 1.2900274991989136,
      "learning_rate": 0.0001640164252297648,
      "loss": 0.9034,
      "step": 820
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 1.350598692893982,
      "learning_rate": 0.00016235184142985367,
      "loss": 0.8938,
      "step": 825
    },
    {
      "epoch": 0.529505582137161,
      "grad_norm": 1.7906914949417114,
      "learning_rate": 0.00016068572467727762,
      "loss": 0.93,
      "step": 830
    },
    {
      "epoch": 0.532695374800638,
      "grad_norm": 1.3412362337112427,
      "learning_rate": 0.00015901828174915005,
      "loss": 0.9181,
      "step": 835
    },
    {
      "epoch": 0.5358851674641149,
      "grad_norm": 2.151066780090332,
      "learning_rate": 0.00015734971958717228,
      "loss": 0.9143,
      "step": 840
    },
    {
      "epoch": 0.5390749601275917,
      "grad_norm": 1.6764332056045532,
      "learning_rate": 0.00015568024527195067,
      "loss": 0.9275,
      "step": 845
    },
    {
      "epoch": 0.5422647527910686,
      "grad_norm": 2.056030750274658,
      "learning_rate": 0.0001540100659972963,
      "loss": 0.8855,
      "step": 850
    },
    {
      "epoch": 0.5454545454545454,
      "grad_norm": 1.733733892440796,
      "learning_rate": 0.000152339389044511,
      "loss": 0.9206,
      "step": 855
    },
    {
      "epoch": 0.5486443381180224,
      "grad_norm": 1.7512574195861816,
      "learning_rate": 0.00015066842175666186,
      "loss": 0.9079,
      "step": 860
    },
    {
      "epoch": 0.5518341307814992,
      "grad_norm": 1.1346197128295898,
      "learning_rate": 0.0001489973715128487,
      "loss": 0.8834,
      "step": 865
    },
    {
      "epoch": 0.5550239234449761,
      "grad_norm": 1.2773138284683228,
      "learning_rate": 0.00014732644570246675,
      "loss": 0.889,
      "step": 870
    },
    {
      "epoch": 0.5582137161084529,
      "grad_norm": 1.3865187168121338,
      "learning_rate": 0.00014565585169946824,
      "loss": 0.904,
      "step": 875
    },
    {
      "epoch": 0.5614035087719298,
      "grad_norm": 1.5330930948257446,
      "learning_rate": 0.0001439857968366256,
      "loss": 0.903,
      "step": 880
    },
    {
      "epoch": 0.5645933014354066,
      "grad_norm": 1.0300610065460205,
      "learning_rate": 0.00014231648837980022,
      "loss": 0.8837,
      "step": 885
    },
    {
      "epoch": 0.5677830940988836,
      "grad_norm": 1.1675297021865845,
      "learning_rate": 0.00014064813350221894,
      "loss": 0.9344,
      "step": 890
    },
    {
      "epoch": 0.5709728867623605,
      "grad_norm": 1.1978455781936646,
      "learning_rate": 0.00013898093925876267,
      "loss": 0.8725,
      "step": 895
    },
    {
      "epoch": 0.5741626794258373,
      "grad_norm": 1.0207343101501465,
      "learning_rate": 0.00013731511256026913,
      "loss": 0.8891,
      "step": 900
    },
    {
      "epoch": 0.5773524720893142,
      "grad_norm": 1.192966103553772,
      "learning_rate": 0.00013565086014785406,
      "loss": 0.9068,
      "step": 905
    },
    {
      "epoch": 0.580542264752791,
      "grad_norm": 1.3042093515396118,
      "learning_rate": 0.00013398838856725257,
      "loss": 0.9001,
      "step": 910
    },
    {
      "epoch": 0.583732057416268,
      "grad_norm": 1.0508038997650146,
      "learning_rate": 0.00013232790414318608,
      "loss": 0.8838,
      "step": 915
    },
    {
      "epoch": 0.5869218500797448,
      "grad_norm": 1.1323915719985962,
      "learning_rate": 0.0001306696129537553,
      "loss": 0.9089,
      "step": 920
    },
    {
      "epoch": 0.5901116427432217,
      "grad_norm": 1.5706528425216675,
      "learning_rate": 0.00012901372080486472,
      "loss": 0.8854,
      "step": 925
    },
    {
      "epoch": 0.5933014354066986,
      "grad_norm": 0.9833298325538635,
      "learning_rate": 0.00012736043320468073,
      "loss": 0.8725,
      "step": 930
    },
    {
      "epoch": 0.5964912280701754,
      "grad_norm": 1.0308219194412231,
      "learning_rate": 0.0001257099553381262,
      "loss": 0.8864,
      "step": 935
    },
    {
      "epoch": 0.5996810207336523,
      "grad_norm": 1.3172023296356201,
      "learning_rate": 0.00012406249204141603,
      "loss": 0.8876,
      "step": 940
    },
    {
      "epoch": 0.6028708133971292,
      "grad_norm": 1.1759730577468872,
      "learning_rate": 0.000122418247776635,
      "loss": 0.8923,
      "step": 945
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 1.251922369003296,
      "learning_rate": 0.00012077742660636299,
      "loss": 0.8914,
      "step": 950
    },
    {
      "epoch": 0.6092503987240829,
      "grad_norm": 1.0379743576049805,
      "learning_rate": 0.00011914023216834904,
      "loss": 0.8637,
      "step": 955
    },
    {
      "epoch": 0.6124401913875598,
      "grad_norm": 1.1909719705581665,
      "learning_rate": 0.0001175068676502386,
      "loss": 0.8646,
      "step": 960
    },
    {
      "epoch": 0.6156299840510366,
      "grad_norm": 1.0347654819488525,
      "learning_rate": 0.00011587753576435634,
      "loss": 0.8648,
      "step": 965
    },
    {
      "epoch": 0.6188197767145136,
      "grad_norm": 1.340886116027832,
      "learning_rate": 0.00011425243872254835,
      "loss": 0.8753,
      "step": 970
    },
    {
      "epoch": 0.6220095693779905,
      "grad_norm": 1.0791716575622559,
      "learning_rate": 0.00011263177821108573,
      "loss": 0.8905,
      "step": 975
    },
    {
      "epoch": 0.6251993620414673,
      "grad_norm": 1.0844085216522217,
      "learning_rate": 0.00011101575536563433,
      "loss": 0.8742,
      "step": 980
    },
    {
      "epoch": 0.6283891547049442,
      "grad_norm": 1.2681643962860107,
      "learning_rate": 0.00010940457074629217,
      "loss": 0.8707,
      "step": 985
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 1.3830182552337646,
      "learning_rate": 0.00010779842431269843,
      "loss": 0.8816,
      "step": 990
    },
    {
      "epoch": 0.6347687400318979,
      "grad_norm": 1.7231197357177734,
      "learning_rate": 0.0001061975153992172,
      "loss": 0.8691,
      "step": 995
    },
    {
      "epoch": 0.6379585326953748,
      "grad_norm": 1.4428848028182983,
      "learning_rate": 0.00010460204269019829,
      "loss": 0.8635,
      "step": 1000
    },
    {
      "epoch": 0.6411483253588517,
      "grad_norm": 1.5241450071334839,
      "learning_rate": 0.0001030122041953196,
      "loss": 0.8796,
      "step": 1005
    },
    {
      "epoch": 0.6443381180223285,
      "grad_norm": 1.2227340936660767,
      "learning_rate": 0.0001014281972250121,
      "loss": 0.8776,
      "step": 1010
    },
    {
      "epoch": 0.6475279106858054,
      "grad_norm": 1.110999584197998,
      "learning_rate": 9.985021836597273e-05,
      "loss": 0.8541,
      "step": 1015
    },
    {
      "epoch": 0.6507177033492823,
      "grad_norm": 1.0933470726013184,
      "learning_rate": 9.827846345676614e-05,
      "loss": 0.8668,
      "step": 1020
    },
    {
      "epoch": 0.6539074960127592,
      "grad_norm": 1.448918342590332,
      "learning_rate": 9.671312756351998e-05,
      "loss": 0.8672,
      "step": 1025
    },
    {
      "epoch": 0.6570972886762361,
      "grad_norm": 1.1408368349075317,
      "learning_rate": 9.515440495571569e-05,
      "loss": 0.8688,
      "step": 1030
    },
    {
      "epoch": 0.6602870813397129,
      "grad_norm": 0.9816156625747681,
      "learning_rate": 9.360248908207813e-05,
      "loss": 0.8623,
      "step": 1035
    },
    {
      "epoch": 0.6634768740031898,
      "grad_norm": 1.3135740756988525,
      "learning_rate": 9.205757254656755e-05,
      "loss": 0.8462,
      "step": 1040
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.1900348663330078,
      "learning_rate": 9.05198470844756e-05,
      "loss": 0.8489,
      "step": 1045
    },
    {
      "epoch": 0.6698564593301436,
      "grad_norm": 1.4275152683258057,
      "learning_rate": 8.898950353862998e-05,
      "loss": 0.8703,
      "step": 1050
    },
    {
      "epoch": 0.6730462519936204,
      "grad_norm": 1.202156901359558,
      "learning_rate": 8.746673183570923e-05,
      "loss": 0.8671,
      "step": 1055
    },
    {
      "epoch": 0.6762360446570973,
      "grad_norm": 1.4236383438110352,
      "learning_rate": 8.595172096267157e-05,
      "loss": 0.8948,
      "step": 1060
    },
    {
      "epoch": 0.6794258373205742,
      "grad_norm": 1.1701873540878296,
      "learning_rate": 8.444465894330024e-05,
      "loss": 0.8417,
      "step": 1065
    },
    {
      "epoch": 0.682615629984051,
      "grad_norm": 1.1106019020080566,
      "learning_rate": 8.294573281486828e-05,
      "loss": 0.8628,
      "step": 1070
    },
    {
      "epoch": 0.6858054226475279,
      "grad_norm": 1.0223338603973389,
      "learning_rate": 8.145512860492596e-05,
      "loss": 0.8571,
      "step": 1075
    },
    {
      "epoch": 0.6889952153110048,
      "grad_norm": 1.305443525314331,
      "learning_rate": 7.997303130821362e-05,
      "loss": 0.872,
      "step": 1080
    },
    {
      "epoch": 0.6921850079744817,
      "grad_norm": 0.9939827919006348,
      "learning_rate": 7.849962486370206e-05,
      "loss": 0.8614,
      "step": 1085
    },
    {
      "epoch": 0.6953748006379585,
      "grad_norm": 1.004202961921692,
      "learning_rate": 7.703509213176451e-05,
      "loss": 0.879,
      "step": 1090
    },
    {
      "epoch": 0.6985645933014354,
      "grad_norm": 0.9680610299110413,
      "learning_rate": 7.557961487148272e-05,
      "loss": 0.8788,
      "step": 1095
    },
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 0.912745475769043,
      "learning_rate": 7.413337371808884e-05,
      "loss": 0.8599,
      "step": 1100
    },
    {
      "epoch": 0.7049441786283892,
      "grad_norm": 1.167737603187561,
      "learning_rate": 7.269654816054756e-05,
      "loss": 0.8488,
      "step": 1105
    },
    {
      "epoch": 0.7081339712918661,
      "grad_norm": 1.242891788482666,
      "learning_rate": 7.126931651928012e-05,
      "loss": 0.8472,
      "step": 1110
    },
    {
      "epoch": 0.7113237639553429,
      "grad_norm": 1.010082721710205,
      "learning_rate": 6.985185592403367e-05,
      "loss": 0.8675,
      "step": 1115
    },
    {
      "epoch": 0.7145135566188198,
      "grad_norm": 1.4250268936157227,
      "learning_rate": 6.844434229189787e-05,
      "loss": 0.8626,
      "step": 1120
    },
    {
      "epoch": 0.7177033492822966,
      "grad_norm": 1.2300432920455933,
      "learning_rate": 6.704695030547252e-05,
      "loss": 0.8422,
      "step": 1125
    },
    {
      "epoch": 0.7208931419457735,
      "grad_norm": 1.642844319343567,
      "learning_rate": 6.56598533911881e-05,
      "loss": 0.8562,
      "step": 1130
    },
    {
      "epoch": 0.7240829346092504,
      "grad_norm": 1.089853048324585,
      "learning_rate": 6.428322369778254e-05,
      "loss": 0.8725,
      "step": 1135
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 0.9274362921714783,
      "learning_rate": 6.291723207493577e-05,
      "loss": 0.8328,
      "step": 1140
    },
    {
      "epoch": 0.7304625199362041,
      "grad_norm": 1.2909477949142456,
      "learning_rate": 6.15620480520666e-05,
      "loss": 0.8529,
      "step": 1145
    },
    {
      "epoch": 0.733652312599681,
      "grad_norm": 1.4193487167358398,
      "learning_rate": 6.0217839817292765e-05,
      "loss": 0.8341,
      "step": 1150
    },
    {
      "epoch": 0.7368421052631579,
      "grad_norm": 1.0142663717269897,
      "learning_rate": 5.888477419655733e-05,
      "loss": 0.8574,
      "step": 1155
    },
    {
      "epoch": 0.7400318979266348,
      "grad_norm": 1.0169768333435059,
      "learning_rate": 5.7563016632924555e-05,
      "loss": 0.8384,
      "step": 1160
    },
    {
      "epoch": 0.7432216905901117,
      "grad_norm": 1.1476997137069702,
      "learning_rate": 5.625273116604728e-05,
      "loss": 0.8734,
      "step": 1165
    },
    {
      "epoch": 0.7464114832535885,
      "grad_norm": 1.1595078706741333,
      "learning_rate": 5.495408041180829e-05,
      "loss": 0.8681,
      "step": 1170
    },
    {
      "epoch": 0.7496012759170654,
      "grad_norm": 1.3419498205184937,
      "learning_rate": 5.3667225542138507e-05,
      "loss": 0.8247,
      "step": 1175
    },
    {
      "epoch": 0.7527910685805422,
      "grad_norm": 1.0906846523284912,
      "learning_rate": 5.239232626501464e-05,
      "loss": 0.8724,
      "step": 1180
    },
    {
      "epoch": 0.7559808612440191,
      "grad_norm": 1.4378533363342285,
      "learning_rate": 5.1129540804637906e-05,
      "loss": 0.8566,
      "step": 1185
    },
    {
      "epoch": 0.759170653907496,
      "grad_norm": 1.0742905139923096,
      "learning_rate": 4.9879025881797466e-05,
      "loss": 0.8414,
      "step": 1190
    },
    {
      "epoch": 0.7623604465709729,
      "grad_norm": 1.0578699111938477,
      "learning_rate": 4.8640936694420245e-05,
      "loss": 0.839,
      "step": 1195
    },
    {
      "epoch": 0.7655502392344498,
      "grad_norm": 1.0157151222229004,
      "learning_rate": 4.7415426898309704e-05,
      "loss": 0.8279,
      "step": 1200
    },
    {
      "epoch": 0.7687400318979266,
      "grad_norm": 1.1732484102249146,
      "learning_rate": 4.62026485880761e-05,
      "loss": 0.8406,
      "step": 1205
    },
    {
      "epoch": 0.7719298245614035,
      "grad_norm": 1.163568139076233,
      "learning_rate": 4.50027522782603e-05,
      "loss": 0.8474,
      "step": 1210
    },
    {
      "epoch": 0.7751196172248804,
      "grad_norm": 1.4089716672897339,
      "learning_rate": 4.3815886884654136e-05,
      "loss": 0.8321,
      "step": 1215
    },
    {
      "epoch": 0.7783094098883573,
      "grad_norm": 1.2893257141113281,
      "learning_rate": 4.264219970581854e-05,
      "loss": 0.8578,
      "step": 1220
    },
    {
      "epoch": 0.7814992025518341,
      "grad_norm": 1.2172292470932007,
      "learning_rate": 4.148183640480293e-05,
      "loss": 0.8627,
      "step": 1225
    },
    {
      "epoch": 0.784688995215311,
      "grad_norm": 1.0065255165100098,
      "learning_rate": 4.0334940991067276e-05,
      "loss": 0.8574,
      "step": 1230
    },
    {
      "epoch": 0.7878787878787878,
      "grad_norm": 0.9151318669319153,
      "learning_rate": 3.920165580260973e-05,
      "loss": 0.8571,
      "step": 1235
    },
    {
      "epoch": 0.7910685805422647,
      "grad_norm": 1.065956473350525,
      "learning_rate": 3.808212148830095e-05,
      "loss": 0.8367,
      "step": 1240
    },
    {
      "epoch": 0.7942583732057417,
      "grad_norm": 1.0603456497192383,
      "learning_rate": 3.697647699042918e-05,
      "loss": 0.8336,
      "step": 1245
    },
    {
      "epoch": 0.7974481658692185,
      "grad_norm": 1.1699525117874146,
      "learning_rate": 3.5884859527455995e-05,
      "loss": 0.8453,
      "step": 1250
    },
    {
      "epoch": 0.8006379585326954,
      "grad_norm": 1.0553467273712158,
      "learning_rate": 3.48074045769868e-05,
      "loss": 0.8519,
      "step": 1255
    },
    {
      "epoch": 0.8038277511961722,
      "grad_norm": 1.0572487115859985,
      "learning_rate": 3.3744245858956967e-05,
      "loss": 0.8773,
      "step": 1260
    },
    {
      "epoch": 0.8070175438596491,
      "grad_norm": 1.0170570611953735,
      "learning_rate": 3.2695515319036186e-05,
      "loss": 0.856,
      "step": 1265
    },
    {
      "epoch": 0.810207336523126,
      "grad_norm": 0.894329309463501,
      "learning_rate": 3.1661343112253304e-05,
      "loss": 0.8405,
      "step": 1270
    },
    {
      "epoch": 0.8133971291866029,
      "grad_norm": 1.1681292057037354,
      "learning_rate": 3.064185758684265e-05,
      "loss": 0.8217,
      "step": 1275
    },
    {
      "epoch": 0.8165869218500797,
      "grad_norm": 1.047298550605774,
      "learning_rate": 2.96371852683157e-05,
      "loss": 0.847,
      "step": 1280
    },
    {
      "epoch": 0.8197767145135566,
      "grad_norm": 1.0787161588668823,
      "learning_rate": 2.8647450843757897e-05,
      "loss": 0.8541,
      "step": 1285
    },
    {
      "epoch": 0.8229665071770335,
      "grad_norm": 1.1034142971038818,
      "learning_rate": 2.7672777146354246e-05,
      "loss": 0.8522,
      "step": 1290
    },
    {
      "epoch": 0.8261562998405104,
      "grad_norm": 1.006184458732605,
      "learning_rate": 2.6713285140144802e-05,
      "loss": 0.8551,
      "step": 1295
    },
    {
      "epoch": 0.8293460925039873,
      "grad_norm": 1.2398364543914795,
      "learning_rate": 2.5769093905012333e-05,
      "loss": 0.8363,
      "step": 1300
    },
    {
      "epoch": 0.8325358851674641,
      "grad_norm": 1.0368655920028687,
      "learning_rate": 2.4840320621903253e-05,
      "loss": 0.837,
      "step": 1305
    },
    {
      "epoch": 0.835725677830941,
      "grad_norm": 1.1005960702896118,
      "learning_rate": 2.392708055828495e-05,
      "loss": 0.8372,
      "step": 1310
    },
    {
      "epoch": 0.8389154704944178,
      "grad_norm": 1.031446099281311,
      "learning_rate": 2.3029487053840295e-05,
      "loss": 0.838,
      "step": 1315
    },
    {
      "epoch": 0.8421052631578947,
      "grad_norm": 0.9887619614601135,
      "learning_rate": 2.214765150640108e-05,
      "loss": 0.8521,
      "step": 1320
    },
    {
      "epoch": 0.8452950558213717,
      "grad_norm": 1.0324478149414062,
      "learning_rate": 2.1281683358122996e-05,
      "loss": 0.8391,
      "step": 1325
    },
    {
      "epoch": 0.8484848484848485,
      "grad_norm": 0.9141530394554138,
      "learning_rate": 2.043169008190289e-05,
      "loss": 0.8526,
      "step": 1330
    },
    {
      "epoch": 0.8516746411483254,
      "grad_norm": 0.9525683522224426,
      "learning_rate": 1.9597777168040872e-05,
      "loss": 0.8589,
      "step": 1335
    },
    {
      "epoch": 0.8548644338118022,
      "grad_norm": 0.9357597231864929,
      "learning_rate": 1.8780048111147776e-05,
      "loss": 0.8565,
      "step": 1340
    },
    {
      "epoch": 0.8580542264752791,
      "grad_norm": 0.989803671836853,
      "learning_rate": 1.797860439730126e-05,
      "loss": 0.8097,
      "step": 1345
    },
    {
      "epoch": 0.861244019138756,
      "grad_norm": 0.9911389946937561,
      "learning_rate": 1.7193545491450183e-05,
      "loss": 0.85,
      "step": 1350
    },
    {
      "epoch": 0.8644338118022329,
      "grad_norm": 0.9085144996643066,
      "learning_rate": 1.6424968825070567e-05,
      "loss": 0.83,
      "step": 1355
    },
    {
      "epoch": 0.8676236044657097,
      "grad_norm": 0.9999231696128845,
      "learning_rate": 1.567296978407353e-05,
      "loss": 0.8378,
      "step": 1360
    },
    {
      "epoch": 0.8708133971291866,
      "grad_norm": 0.9475440979003906,
      "learning_rate": 1.4937641696967245e-05,
      "loss": 0.8531,
      "step": 1365
    },
    {
      "epoch": 0.8740031897926634,
      "grad_norm": 0.985124945640564,
      "learning_rate": 1.4219075823274251e-05,
      "loss": 0.8431,
      "step": 1370
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.9646289348602295,
      "learning_rate": 1.3517361342205295e-05,
      "loss": 0.8361,
      "step": 1375
    },
    {
      "epoch": 0.8803827751196173,
      "grad_norm": 0.9125611186027527,
      "learning_rate": 1.2832585341591844e-05,
      "loss": 0.8106,
      "step": 1380
    },
    {
      "epoch": 0.8835725677830941,
      "grad_norm": 1.214565634727478,
      "learning_rate": 1.2164832807077585e-05,
      "loss": 0.8441,
      "step": 1385
    },
    {
      "epoch": 0.886762360446571,
      "grad_norm": 1.085919976234436,
      "learning_rate": 1.151418661157122e-05,
      "loss": 0.8271,
      "step": 1390
    },
    {
      "epoch": 0.8899521531100478,
      "grad_norm": 0.9682278633117676,
      "learning_rate": 1.0880727504961339e-05,
      "loss": 0.8293,
      "step": 1395
    },
    {
      "epoch": 0.8931419457735247,
      "grad_norm": 1.015742540359497,
      "learning_rate": 1.0264534104094812e-05,
      "loss": 0.8452,
      "step": 1400
    },
    {
      "epoch": 0.8963317384370016,
      "grad_norm": 0.9043238162994385,
      "learning_rate": 9.665682883019732e-06,
      "loss": 0.8274,
      "step": 1405
    },
    {
      "epoch": 0.8995215311004785,
      "grad_norm": 0.9109312891960144,
      "learning_rate": 9.084248163494511e-06,
      "loss": 0.8332,
      "step": 1410
    },
    {
      "epoch": 0.9027113237639554,
      "grad_norm": 1.0157884359359741,
      "learning_rate": 8.520302105764148e-06,
      "loss": 0.8304,
      "step": 1415
    },
    {
      "epoch": 0.9059011164274322,
      "grad_norm": 0.9898545742034912,
      "learning_rate": 7.973914699604367e-06,
      "loss": 0.8346,
      "step": 1420
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 0.914945662021637,
      "learning_rate": 7.445153755635569e-06,
      "loss": 0.8443,
      "step": 1425
    },
    {
      "epoch": 0.9122807017543859,
      "grad_norm": 0.9307130575180054,
      "learning_rate": 6.934084896906983e-06,
      "loss": 0.8429,
      "step": 1430
    },
    {
      "epoch": 0.9154704944178629,
      "grad_norm": 1.0683540105819702,
      "learning_rate": 6.440771550752377e-06,
      "loss": 0.8504,
      "step": 1435
    },
    {
      "epoch": 0.9186602870813397,
      "grad_norm": 0.9779647588729858,
      "learning_rate": 5.965274940918274e-06,
      "loss": 0.8422,
      "step": 1440
    },
    {
      "epoch": 0.9218500797448166,
      "grad_norm": 0.9050576686859131,
      "learning_rate": 5.507654079965612e-06,
      "loss": 0.8186,
      "step": 1445
    },
    {
      "epoch": 0.9250398724082934,
      "grad_norm": 1.057937502861023,
      "learning_rate": 5.067965761945869e-06,
      "loss": 0.8447,
      "step": 1450
    },
    {
      "epoch": 0.9282296650717703,
      "grad_norm": 0.9198818206787109,
      "learning_rate": 4.646264555352586e-06,
      "loss": 0.8252,
      "step": 1455
    },
    {
      "epoch": 0.9314194577352473,
      "grad_norm": 1.075323462486267,
      "learning_rate": 4.242602796348915e-06,
      "loss": 0.8197,
      "step": 1460
    },
    {
      "epoch": 0.9346092503987241,
      "grad_norm": 1.1147295236587524,
      "learning_rate": 3.857030582272369e-06,
      "loss": 0.8224,
      "step": 1465
    },
    {
      "epoch": 0.937799043062201,
      "grad_norm": 1.0000038146972656,
      "learning_rate": 3.489595765417441e-06,
      "loss": 0.7997,
      "step": 1470
    },
    {
      "epoch": 0.9409888357256778,
      "grad_norm": 0.9286321401596069,
      "learning_rate": 3.140343947096624e-06,
      "loss": 0.8055,
      "step": 1475
    },
    {
      "epoch": 0.9441786283891547,
      "grad_norm": 0.9910067915916443,
      "learning_rate": 2.80931847198117e-06,
      "loss": 0.8619,
      "step": 1480
    },
    {
      "epoch": 0.9473684210526315,
      "grad_norm": 1.0662553310394287,
      "learning_rate": 2.4965604227215774e-06,
      "loss": 0.8424,
      "step": 1485
    },
    {
      "epoch": 0.9505582137161085,
      "grad_norm": 0.9524323344230652,
      "learning_rate": 2.202108614848885e-06,
      "loss": 0.8281,
      "step": 1490
    },
    {
      "epoch": 0.9537480063795853,
      "grad_norm": 0.910805344581604,
      "learning_rate": 1.925999591957561e-06,
      "loss": 0.8206,
      "step": 1495
    },
    {
      "epoch": 0.9569377990430622,
      "grad_norm": 0.9802441596984863,
      "learning_rate": 1.6682676211700107e-06,
      "loss": 0.8394,
      "step": 1500
    },
    {
      "epoch": 0.960127591706539,
      "grad_norm": 0.907018780708313,
      "learning_rate": 1.4289446888838652e-06,
      "loss": 0.8453,
      "step": 1505
    },
    {
      "epoch": 0.9633173843700159,
      "grad_norm": 1.0577489137649536,
      "learning_rate": 1.2080604968022378e-06,
      "loss": 0.8119,
      "step": 1510
    },
    {
      "epoch": 0.9665071770334929,
      "grad_norm": 0.9329078197479248,
      "learning_rate": 1.0056424582474575e-06,
      "loss": 0.8241,
      "step": 1515
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 0.9109911322593689,
      "learning_rate": 8.217156947590064e-07,
      "loss": 0.8181,
      "step": 1520
    },
    {
      "epoch": 0.9728867623604466,
      "grad_norm": 0.9339520931243896,
      "learning_rate": 6.563030329755969e-07,
      "loss": 0.805,
      "step": 1525
    },
    {
      "epoch": 0.9760765550239234,
      "grad_norm": 1.0125936269760132,
      "learning_rate": 5.094250018023715e-07,
      "loss": 0.85,
      "step": 1530
    },
    {
      "epoch": 0.9792663476874003,
      "grad_norm": 0.9169021844863892,
      "learning_rate": 3.8109982986300747e-07,
      "loss": 0.8446,
      "step": 1535
    },
    {
      "epoch": 0.9824561403508771,
      "grad_norm": 0.9390758872032166,
      "learning_rate": 2.7134344323747616e-07,
      "loss": 0.8234,
      "step": 1540
    },
    {
      "epoch": 0.9856459330143541,
      "grad_norm": 0.9467016458511353,
      "learning_rate": 1.801694634854578e-07,
      "loss": 0.8461,
      "step": 1545
    },
    {
      "epoch": 0.988835725677831,
      "grad_norm": 1.0382705926895142,
      "learning_rate": 1.075892059558603e-07,
      "loss": 0.8478,
      "step": 1550
    },
    {
      "epoch": 0.9920255183413078,
      "grad_norm": 0.9475221633911133,
      "learning_rate": 5.3611678382442516e-08,
      "loss": 0.8307,
      "step": 1555
    },
    {
      "epoch": 0.9952153110047847,
      "grad_norm": 0.9038047790527344,
      "learning_rate": 1.824357976594193e-08,
      "loss": 0.8018,
      "step": 1560
    },
    {
      "epoch": 0.9984051036682615,
      "grad_norm": 0.977640688419342,
      "learning_rate": 1.4892995426396548e-09,
      "loss": 0.8292,
      "step": 1565
    },
    {
      "epoch": 0.9996810207336523,
      "eval_loss": 2.336217164993286,
      "eval_runtime": 2.955,
      "eval_samples_per_second": 3.384,
      "eval_steps_per_second": 0.677,
      "step": 1567
    },
    {
      "epoch": 0.9996810207336523,
      "step": 1567,
      "total_flos": 6.008655759456338e+17,
      "train_loss": 1.5068606398405355,
      "train_runtime": 9641.8347,
      "train_samples_per_second": 1.3,
      "train_steps_per_second": 0.163
    }
  ],
  "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
}