{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.988452655889146,
  "eval_steps": 500,
  "global_step": 4325,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.023094688221709007,
      "grad_norm": 10.050519943237305,
      "learning_rate": 9.216589861751153e-06,
      "loss": 1.0509,
      "step": 10
    },
    {
      "epoch": 0.046189376443418015,
      "grad_norm": 2.694617986679077,
      "learning_rate": 1.8433179723502307e-05,
      "loss": 0.6955,
      "step": 20
    },
    {
      "epoch": 0.06928406466512702,
      "grad_norm": 2.627452850341797,
      "learning_rate": 2.764976958525346e-05,
      "loss": 0.3561,
      "step": 30
    },
    {
      "epoch": 0.09237875288683603,
      "grad_norm": 1.1433345079421997,
      "learning_rate": 3.6866359447004614e-05,
      "loss": 0.2924,
      "step": 40
    },
    {
      "epoch": 0.11547344110854503,
      "grad_norm": 1.5595552921295166,
      "learning_rate": 4.608294930875576e-05,
      "loss": 0.2146,
      "step": 50
    },
    {
      "epoch": 0.13856812933025403,
      "grad_norm": 2.5837883949279785,
      "learning_rate": 5.529953917050692e-05,
      "loss": 0.1891,
      "step": 60
    },
    {
      "epoch": 0.16166281755196305,
      "grad_norm": 0.9749034643173218,
      "learning_rate": 6.451612903225807e-05,
      "loss": 0.1584,
      "step": 70
    },
    {
      "epoch": 0.18475750577367206,
      "grad_norm": 0.9678087830543518,
      "learning_rate": 7.373271889400923e-05,
      "loss": 0.1337,
      "step": 80
    },
    {
      "epoch": 0.20785219399538107,
      "grad_norm": 1.4949584007263184,
      "learning_rate": 8.294930875576037e-05,
      "loss": 0.1235,
      "step": 90
    },
    {
      "epoch": 0.23094688221709006,
      "grad_norm": 0.9887685179710388,
      "learning_rate": 9.216589861751152e-05,
      "loss": 0.1061,
      "step": 100
    },
    {
      "epoch": 0.2540415704387991,
      "grad_norm": 0.6556084752082825,
      "learning_rate": 0.00010138248847926268,
      "loss": 0.1055,
      "step": 110
    },
    {
      "epoch": 0.27713625866050806,
      "grad_norm": 0.6040177345275879,
      "learning_rate": 0.00011059907834101384,
      "loss": 0.0893,
      "step": 120
    },
    {
      "epoch": 0.3002309468822171,
      "grad_norm": 0.5599756240844727,
      "learning_rate": 0.00011981566820276497,
      "loss": 0.0784,
      "step": 130
    },
    {
      "epoch": 0.3233256351039261,
      "grad_norm": 0.5788015723228455,
      "learning_rate": 0.00012903225806451613,
      "loss": 0.0857,
      "step": 140
    },
    {
      "epoch": 0.3464203233256351,
      "grad_norm": 0.49766597151756287,
      "learning_rate": 0.00013824884792626728,
      "loss": 0.0811,
      "step": 150
    },
    {
      "epoch": 0.3695150115473441,
      "grad_norm": 1.4221667051315308,
      "learning_rate": 0.00014746543778801845,
      "loss": 0.0786,
      "step": 160
    },
    {
      "epoch": 0.39260969976905313,
      "grad_norm": 0.906166136264801,
      "learning_rate": 0.0001566820276497696,
      "loss": 0.0734,
      "step": 170
    },
    {
      "epoch": 0.41570438799076215,
      "grad_norm": 0.7869032025337219,
      "learning_rate": 0.00016589861751152075,
      "loss": 0.0666,
      "step": 180
    },
    {
      "epoch": 0.4387990762124711,
      "grad_norm": 0.7141720652580261,
      "learning_rate": 0.0001751152073732719,
      "loss": 0.0667,
      "step": 190
    },
    {
      "epoch": 0.4618937644341801,
      "grad_norm": 0.538352370262146,
      "learning_rate": 0.00018433179723502304,
      "loss": 0.0659,
      "step": 200
    },
    {
      "epoch": 0.48498845265588914,
      "grad_norm": 0.5275980234146118,
      "learning_rate": 0.00019354838709677422,
      "loss": 0.0636,
      "step": 210
    },
    {
      "epoch": 0.5080831408775982,
      "grad_norm": 0.6203627586364746,
      "learning_rate": 0.00019999973682102037,
      "loss": 0.0619,
      "step": 220
    },
    {
      "epoch": 0.5311778290993071,
      "grad_norm": 0.48453882336616516,
      "learning_rate": 0.00019999505812214085,
      "loss": 0.062,
      "step": 230
    },
    {
      "epoch": 0.5542725173210161,
      "grad_norm": 0.46422210335731506,
      "learning_rate": 0.00019998453131645004,
      "loss": 0.0623,
      "step": 240
    },
    {
      "epoch": 0.5773672055427251,
      "grad_norm": 0.8341119289398193,
      "learning_rate": 0.00019996815701959716,
      "loss": 0.0545,
      "step": 250
    },
    {
      "epoch": 0.6004618937644342,
      "grad_norm": 0.6169431805610657,
      "learning_rate": 0.00019994593618921592,
      "loss": 0.0594,
      "step": 260
    },
    {
      "epoch": 0.6235565819861432,
      "grad_norm": 0.6459381580352783,
      "learning_rate": 0.00019991787012486843,
      "loss": 0.0562,
      "step": 270
    },
    {
      "epoch": 0.6466512702078522,
      "grad_norm": 0.5904170274734497,
      "learning_rate": 0.0001998839604679692,
      "loss": 0.0529,
      "step": 280
    },
    {
      "epoch": 0.6697459584295612,
      "grad_norm": 0.5025797486305237,
      "learning_rate": 0.00019984420920168928,
      "loss": 0.0501,
      "step": 290
    },
    {
      "epoch": 0.6928406466512702,
      "grad_norm": 0.6631583571434021,
      "learning_rate": 0.00019979861865083995,
      "loss": 0.0583,
      "step": 300
    },
    {
      "epoch": 0.7159353348729792,
      "grad_norm": 0.5108867883682251,
      "learning_rate": 0.00019974719148173716,
      "loss": 0.0551,
      "step": 310
    },
    {
      "epoch": 0.7390300230946882,
      "grad_norm": 0.30322328209877014,
      "learning_rate": 0.00019968993070204534,
      "loss": 0.0444,
      "step": 320
    },
    {
      "epoch": 0.7621247113163973,
      "grad_norm": 0.5695562958717346,
      "learning_rate": 0.0001996268396606015,
      "loss": 0.0503,
      "step": 330
    },
    {
      "epoch": 0.7852193995381063,
      "grad_norm": 0.5680237412452698,
      "learning_rate": 0.0001995579220472195,
      "loss": 0.0607,
      "step": 340
    },
    {
      "epoch": 0.8083140877598153,
      "grad_norm": 0.6361756920814514,
      "learning_rate": 0.00019948318189247415,
      "loss": 0.0534,
      "step": 350
    },
    {
      "epoch": 0.8314087759815243,
      "grad_norm": 0.30916863679885864,
      "learning_rate": 0.00019940262356746554,
      "loss": 0.0545,
      "step": 360
    },
    {
      "epoch": 0.8545034642032333,
      "grad_norm": 0.6371238827705383,
      "learning_rate": 0.00019931625178356343,
      "loss": 0.0536,
      "step": 370
    },
    {
      "epoch": 0.8775981524249422,
      "grad_norm": 0.42440804839134216,
      "learning_rate": 0.0001992240715921316,
      "loss": 0.0476,
      "step": 380
    },
    {
      "epoch": 0.9006928406466512,
      "grad_norm": 0.42693522572517395,
      "learning_rate": 0.0001991260883842325,
      "loss": 0.049,
      "step": 390
    },
    {
      "epoch": 0.9237875288683602,
      "grad_norm": 0.36199280619621277,
      "learning_rate": 0.00019902230789031207,
      "loss": 0.0479,
      "step": 400
    },
    {
      "epoch": 0.9468822170900693,
      "grad_norm": 0.3671620786190033,
      "learning_rate": 0.0001989127361798643,
      "loss": 0.0496,
      "step": 410
    },
    {
      "epoch": 0.9699769053117783,
      "grad_norm": 0.524524450302124,
      "learning_rate": 0.00019879737966107654,
      "loss": 0.0493,
      "step": 420
    },
    {
      "epoch": 0.9930715935334873,
      "grad_norm": 0.5563727021217346,
      "learning_rate": 0.0001986762450804547,
      "loss": 0.0501,
      "step": 430
    },
    {
      "epoch": 1.0161662817551964,
      "grad_norm": 0.4217710793018341,
      "learning_rate": 0.00019854933952242844,
      "loss": 0.0432,
      "step": 440
    },
    {
      "epoch": 1.0392609699769053,
      "grad_norm": 0.3949870765209198,
      "learning_rate": 0.00019841667040893722,
      "loss": 0.0467,
      "step": 450
    },
    {
      "epoch": 1.0623556581986142,
      "grad_norm": 0.4789281189441681,
      "learning_rate": 0.0001982782454989959,
      "loss": 0.0441,
      "step": 460
    },
    {
      "epoch": 1.0854503464203233,
      "grad_norm": 0.42184802889823914,
      "learning_rate": 0.00019813407288824113,
      "loss": 0.043,
      "step": 470
    },
    {
      "epoch": 1.1085450346420322,
      "grad_norm": 0.40358513593673706,
      "learning_rate": 0.0001979841610084579,
      "loss": 0.0449,
      "step": 480
    },
    {
      "epoch": 1.1316397228637414,
      "grad_norm": 0.353373259305954,
      "learning_rate": 0.00019782851862708634,
      "loss": 0.0454,
      "step": 490
    },
    {
      "epoch": 1.1547344110854503,
      "grad_norm": 0.35266897082328796,
      "learning_rate": 0.00019766715484670894,
      "loss": 0.0496,
      "step": 500
    },
    {
      "epoch": 1.1778290993071594,
      "grad_norm": 0.4049600660800934,
      "learning_rate": 0.00019750007910451838,
      "loss": 0.0422,
      "step": 510
    },
    {
      "epoch": 1.2009237875288683,
      "grad_norm": 0.383989155292511,
      "learning_rate": 0.00019732730117176533,
      "loss": 0.0459,
      "step": 520
    },
    {
      "epoch": 1.2240184757505774,
      "grad_norm": 0.28790971636772156,
      "learning_rate": 0.0001971488311531873,
      "loss": 0.0473,
      "step": 530
    },
    {
      "epoch": 1.2471131639722863,
      "grad_norm": 0.42718982696533203,
      "learning_rate": 0.00019696467948641732,
      "loss": 0.0427,
      "step": 540
    },
    {
      "epoch": 1.2702078521939955,
      "grad_norm": 0.4704771041870117,
      "learning_rate": 0.00019677485694137388,
      "loss": 0.0395,
      "step": 550
    },
    {
      "epoch": 1.2933025404157044,
      "grad_norm": 0.3732239007949829,
      "learning_rate": 0.00019657937461963073,
      "loss": 0.0437,
      "step": 560
    },
    {
      "epoch": 1.3163972286374133,
      "grad_norm": 0.41071411967277527,
      "learning_rate": 0.0001963782439537678,
      "loss": 0.0495,
      "step": 570
    },
    {
      "epoch": 1.3394919168591224,
      "grad_norm": 0.4559623897075653,
      "learning_rate": 0.00019617147670670268,
      "loss": 0.0432,
      "step": 580
    },
    {
      "epoch": 1.3625866050808315,
      "grad_norm": 0.48421376943588257,
      "learning_rate": 0.00019595908497100235,
      "loss": 0.0435,
      "step": 590
    },
    {
      "epoch": 1.3856812933025404,
      "grad_norm": 0.5113271474838257,
      "learning_rate": 0.00019574108116817622,
      "loss": 0.0412,
      "step": 600
    },
    {
      "epoch": 1.4087759815242493,
      "grad_norm": 0.38500213623046875,
      "learning_rate": 0.00019551747804794967,
      "loss": 0.0428,
      "step": 610
    },
    {
      "epoch": 1.4318706697459584,
      "grad_norm": 0.27223697304725647,
      "learning_rate": 0.00019528828868751818,
      "loss": 0.0434,
      "step": 620
    },
    {
      "epoch": 1.4549653579676676,
      "grad_norm": 0.3776547312736511,
      "learning_rate": 0.00019505352649078282,
      "loss": 0.043,
      "step": 630
    },
    {
      "epoch": 1.4780600461893765,
      "grad_norm": 0.359902024269104,
      "learning_rate": 0.00019481320518756608,
      "loss": 0.043,
      "step": 640
    },
    {
      "epoch": 1.5011547344110854,
      "grad_norm": 0.413602352142334,
      "learning_rate": 0.0001945673388328091,
      "loss": 0.046,
      "step": 650
    },
    {
      "epoch": 1.5242494226327945,
      "grad_norm": 0.3547694981098175,
      "learning_rate": 0.00019431594180574944,
      "loss": 0.0437,
      "step": 660
    },
    {
      "epoch": 1.5473441108545036,
      "grad_norm": 0.3197380602359772,
      "learning_rate": 0.0001940590288090804,
      "loss": 0.0434,
      "step": 670
    },
    {
      "epoch": 1.5704387990762125,
      "grad_norm": 0.296234130859375,
      "learning_rate": 0.00019379661486809094,
      "loss": 0.0383,
      "step": 680
    },
    {
      "epoch": 1.5935334872979214,
      "grad_norm": 0.38352492451667786,
      "learning_rate": 0.00019352871532978712,
      "loss": 0.037,
      "step": 690
    },
    {
      "epoch": 1.6166281755196303,
      "grad_norm": 0.3082719147205353,
      "learning_rate": 0.00019325534586199424,
      "loss": 0.0368,
      "step": 700
    },
    {
      "epoch": 1.6397228637413395,
      "grad_norm": 0.3575062155723572,
      "learning_rate": 0.000192976522452441,
      "loss": 0.0414,
      "step": 710
    },
    {
      "epoch": 1.6628175519630486,
      "grad_norm": 0.2746741771697998,
      "learning_rate": 0.00019269226140782402,
      "loss": 0.04,
      "step": 720
    },
    {
      "epoch": 1.6859122401847575,
      "grad_norm": 0.35034051537513733,
      "learning_rate": 0.00019240257935285434,
      "loss": 0.0411,
      "step": 730
    },
    {
      "epoch": 1.7090069284064664,
      "grad_norm": 0.36941060423851013,
      "learning_rate": 0.0001921074932292852,
      "loss": 0.0417,
      "step": 740
    },
    {
      "epoch": 1.7321016166281755,
      "grad_norm": 0.4880543351173401,
      "learning_rate": 0.00019180702029492118,
      "loss": 0.0381,
      "step": 750
    },
    {
      "epoch": 1.7551963048498846,
      "grad_norm": 0.5262073278427124,
      "learning_rate": 0.00019150117812260882,
      "loss": 0.045,
      "step": 760
    },
    {
      "epoch": 1.7782909930715936,
      "grad_norm": 0.3127468526363373,
      "learning_rate": 0.00019118998459920902,
      "loss": 0.0437,
      "step": 770
    },
    {
      "epoch": 1.8013856812933025,
      "grad_norm": 0.21983791887760162,
      "learning_rate": 0.0001908734579245508,
      "loss": 0.0403,
      "step": 780
    },
    {
      "epoch": 1.8244803695150116,
      "grad_norm": 0.3254013955593109,
      "learning_rate": 0.000190551616610367,
      "loss": 0.0351,
      "step": 790
    },
    {
      "epoch": 1.8475750577367207,
      "grad_norm": 0.4744585156440735,
      "learning_rate": 0.00019022447947921167,
      "loss": 0.04,
      "step": 800
    },
    {
      "epoch": 1.8706697459584296,
      "grad_norm": 0.42813506722450256,
      "learning_rate": 0.00018989206566335907,
      "loss": 0.0371,
      "step": 810
    },
    {
      "epoch": 1.8937644341801385,
      "grad_norm": 0.3789767622947693,
      "learning_rate": 0.00018955439460368502,
      "loss": 0.0383,
      "step": 820
    },
    {
      "epoch": 1.9168591224018474,
      "grad_norm": 0.35561323165893555,
      "learning_rate": 0.00018921148604852971,
      "loss": 0.0381,
      "step": 830
    },
    {
      "epoch": 1.9399538106235565,
      "grad_norm": 0.3015558123588562,
      "learning_rate": 0.00018886336005254274,
      "loss": 0.0393,
      "step": 840
    },
    {
      "epoch": 1.9630484988452657,
      "grad_norm": 0.27170488238334656,
      "learning_rate": 0.00018851003697551038,
      "loss": 0.0391,
      "step": 850
    },
    {
      "epoch": 1.9861431870669746,
      "grad_norm": 0.4331890940666199,
      "learning_rate": 0.0001881515374811647,
      "loss": 0.0401,
      "step": 860
    },
    {
      "epoch": 2.0092378752886835,
      "grad_norm": 0.5359442234039307,
      "learning_rate": 0.00018778788253597523,
      "loss": 0.039,
      "step": 870
    },
    {
      "epoch": 2.032332563510393,
      "grad_norm": 0.4329531192779541,
      "learning_rate": 0.00018741909340792262,
      "loss": 0.0394,
      "step": 880
    },
    {
      "epoch": 2.0554272517321017,
      "grad_norm": 0.29946601390838623,
      "learning_rate": 0.00018704519166525482,
      "loss": 0.0394,
      "step": 890
    },
    {
      "epoch": 2.0785219399538106,
      "grad_norm": 0.45591849088668823,
      "learning_rate": 0.00018666619917522584,
      "loss": 0.0424,
      "step": 900
    },
    {
      "epoch": 2.1016166281755195,
      "grad_norm": 0.27476781606674194,
      "learning_rate": 0.00018628213810281658,
      "loss": 0.0365,
      "step": 910
    },
    {
      "epoch": 2.1247113163972284,
      "grad_norm": 0.3706781566143036,
      "learning_rate": 0.00018589303090943887,
      "loss": 0.035,
      "step": 920
    },
    {
      "epoch": 2.147806004618938,
      "grad_norm": 0.38271817564964294,
      "learning_rate": 0.00018549890035162155,
      "loss": 0.0372,
      "step": 930
    },
    {
      "epoch": 2.1709006928406467,
      "grad_norm": 0.4124270975589752,
      "learning_rate": 0.00018509976947967975,
      "loss": 0.0406,
      "step": 940
    },
    {
      "epoch": 2.1939953810623556,
      "grad_norm": 0.3650427460670471,
      "learning_rate": 0.00018469566163636673,
      "loss": 0.038,
      "step": 950
    },
    {
      "epoch": 2.2170900692840645,
      "grad_norm": 0.2632494568824768,
      "learning_rate": 0.00018428660045550875,
      "loss": 0.0358,
      "step": 960
    },
    {
      "epoch": 2.240184757505774,
      "grad_norm": 0.2518555819988251,
      "learning_rate": 0.0001838726098606229,
      "loss": 0.0384,
      "step": 970
    },
    {
      "epoch": 2.2632794457274827,
      "grad_norm": 0.35301706194877625,
      "learning_rate": 0.00018345371406351782,
      "loss": 0.0368,
      "step": 980
    },
    {
      "epoch": 2.2863741339491916,
      "grad_norm": 0.2957130968570709,
      "learning_rate": 0.00018302993756287795,
      "loss": 0.0375,
      "step": 990
    },
    {
      "epoch": 2.3094688221709005,
      "grad_norm": 0.33476200699806213,
      "learning_rate": 0.00018260130514283036,
      "loss": 0.0377,
      "step": 1000
    },
    {
      "epoch": 2.3325635103926095,
      "grad_norm": 0.3259429931640625,
      "learning_rate": 0.0001821678418714957,
      "loss": 0.0333,
      "step": 1010
    },
    {
      "epoch": 2.355658198614319,
      "grad_norm": 0.4147096872329712,
      "learning_rate": 0.00018172957309952188,
      "loss": 0.0375,
      "step": 1020
    },
    {
      "epoch": 2.3787528868360277,
      "grad_norm": 0.29218655824661255,
      "learning_rate": 0.00018128652445860145,
      "loss": 0.0345,
      "step": 1030
    },
    {
      "epoch": 2.4018475750577366,
      "grad_norm": 0.32885679602622986,
      "learning_rate": 0.00018083872185997274,
      "loss": 0.0368,
      "step": 1040
    },
    {
      "epoch": 2.424942263279446,
      "grad_norm": 0.37419071793556213,
      "learning_rate": 0.00018038619149290415,
      "loss": 0.03,
      "step": 1050
    },
    {
      "epoch": 2.448036951501155,
      "grad_norm": 0.3435191810131073,
      "learning_rate": 0.0001799289598231629,
      "loss": 0.0343,
      "step": 1060
    },
    {
      "epoch": 2.4711316397228638,
      "grad_norm": 0.46311062574386597,
      "learning_rate": 0.00017946705359146684,
      "loss": 0.0334,
      "step": 1070
    },
    {
      "epoch": 2.4942263279445727,
      "grad_norm": 0.4619911015033722,
      "learning_rate": 0.00017900049981192081,
      "loss": 0.0382,
      "step": 1080
    },
    {
      "epoch": 2.5173210161662816,
      "grad_norm": 0.30530160665512085,
      "learning_rate": 0.00017852932577043658,
      "loss": 0.0352,
      "step": 1090
    },
    {
      "epoch": 2.540415704387991,
      "grad_norm": 0.23941746354103088,
      "learning_rate": 0.0001780535590231372,
      "loss": 0.0349,
      "step": 1100
    },
    {
      "epoch": 2.5635103926097,
      "grad_norm": 0.2063404619693756,
      "learning_rate": 0.00017757322739474525,
      "loss": 0.0304,
      "step": 1110
    },
    {
      "epoch": 2.5866050808314087,
      "grad_norm": 0.33000999689102173,
      "learning_rate": 0.0001770883589769557,
      "loss": 0.0329,
      "step": 1120
    },
    {
      "epoch": 2.6096997690531176,
      "grad_norm": 0.34663257002830505,
      "learning_rate": 0.0001765989821267928,
      "loss": 0.0336,
      "step": 1130
    },
    {
      "epoch": 2.6327944572748265,
      "grad_norm": 0.39627474546432495,
      "learning_rate": 0.00017610512546495195,
      "loss": 0.034,
      "step": 1140
    },
    {
      "epoch": 2.655889145496536,
      "grad_norm": 0.28975868225097656,
      "learning_rate": 0.00017560681787412552,
      "loss": 0.0331,
      "step": 1150
    },
    {
      "epoch": 2.678983833718245,
      "grad_norm": 0.3867139518260956,
      "learning_rate": 0.00017510408849731392,
      "loss": 0.0349,
      "step": 1160
    },
    {
      "epoch": 2.7020785219399537,
      "grad_norm": 0.23290959000587463,
      "learning_rate": 0.00017459696673612104,
      "loss": 0.0382,
      "step": 1170
    },
    {
      "epoch": 2.725173210161663,
      "grad_norm": 0.4149172902107239,
      "learning_rate": 0.00017408548224903488,
      "loss": 0.0375,
      "step": 1180
    },
    {
      "epoch": 2.748267898383372,
      "grad_norm": 0.23780949413776398,
      "learning_rate": 0.00017356966494969277,
      "loss": 0.0297,
      "step": 1190
    },
    {
      "epoch": 2.771362586605081,
      "grad_norm": 0.27576741576194763,
      "learning_rate": 0.00017304954500513224,
      "loss": 0.0365,
      "step": 1200
    },
    {
      "epoch": 2.7944572748267897,
      "grad_norm": 0.2264527529478073,
      "learning_rate": 0.00017252515283402636,
      "loss": 0.0307,
      "step": 1210
    },
    {
      "epoch": 2.8175519630484986,
      "grad_norm": 0.35455092787742615,
      "learning_rate": 0.0001719965191049051,
      "loss": 0.0315,
      "step": 1220
    },
    {
      "epoch": 2.840646651270208,
      "grad_norm": 0.2472825050354004,
      "learning_rate": 0.00017146367473436138,
      "loss": 0.0354,
      "step": 1230
    },
    {
      "epoch": 2.863741339491917,
      "grad_norm": 0.2657906115055084,
      "learning_rate": 0.00017092665088524315,
      "loss": 0.0361,
      "step": 1240
    },
    {
      "epoch": 2.886836027713626,
      "grad_norm": 0.36936143040657043,
      "learning_rate": 0.00017038547896483083,
      "loss": 0.0319,
      "step": 1250
    },
    {
      "epoch": 2.909930715935335,
      "grad_norm": 0.34996992349624634,
      "learning_rate": 0.0001698401906230005,
      "loss": 0.0349,
      "step": 1260
    },
    {
      "epoch": 2.9330254041570436,
      "grad_norm": 0.2003779411315918,
      "learning_rate": 0.00016929081775037276,
      "loss": 0.0301,
      "step": 1270
    },
    {
      "epoch": 2.956120092378753,
      "grad_norm": 0.2517768442630768,
      "learning_rate": 0.00016873739247644785,
      "loss": 0.0342,
      "step": 1280
    },
    {
      "epoch": 2.979214780600462,
      "grad_norm": 0.26178500056266785,
      "learning_rate": 0.00016817994716772635,
      "loss": 0.0339,
      "step": 1290
    },
    {
      "epoch": 3.0023094688221708,
      "grad_norm": 0.30486026406288147,
      "learning_rate": 0.0001676185144258165,
      "loss": 0.0352,
      "step": 1300
    },
    {
      "epoch": 3.02540415704388,
      "grad_norm": 0.26309579610824585,
      "learning_rate": 0.0001670531270855274,
      "loss": 0.0317,
      "step": 1310
    },
    {
      "epoch": 3.048498845265589,
      "grad_norm": 0.23204182088375092,
      "learning_rate": 0.00016648381821294863,
      "loss": 0.0298,
      "step": 1320
    },
    {
      "epoch": 3.071593533487298,
      "grad_norm": 0.23810264468193054,
      "learning_rate": 0.00016591062110351662,
      "loss": 0.0351,
      "step": 1330
    },
    {
      "epoch": 3.094688221709007,
      "grad_norm": 0.36545702815055847,
      "learning_rate": 0.00016533356928006722,
      "loss": 0.0347,
      "step": 1340
    },
    {
      "epoch": 3.1177829099307157,
      "grad_norm": 0.2644146978855133,
      "learning_rate": 0.0001647526964908752,
      "loss": 0.0319,
      "step": 1350
    },
    {
      "epoch": 3.140877598152425,
      "grad_norm": 0.29174649715423584,
      "learning_rate": 0.00016416803670768056,
      "loss": 0.0329,
      "step": 1360
    },
    {
      "epoch": 3.163972286374134,
      "grad_norm": 0.25455746054649353,
      "learning_rate": 0.0001635796241237017,
      "loss": 0.0327,
      "step": 1370
    },
    {
      "epoch": 3.187066974595843,
      "grad_norm": 0.36102402210235596,
      "learning_rate": 0.00016298749315163567,
      "loss": 0.0291,
      "step": 1380
    },
    {
      "epoch": 3.2101616628175518,
      "grad_norm": 0.4295611083507538,
      "learning_rate": 0.00016239167842164548,
      "loss": 0.0305,
      "step": 1390
    },
    {
      "epoch": 3.233256351039261,
      "grad_norm": 0.2627660632133484,
      "learning_rate": 0.0001617922147793351,
      "loss": 0.0294,
      "step": 1400
    },
    {
      "epoch": 3.25635103926097,
      "grad_norm": 0.2932153046131134,
      "learning_rate": 0.00016118913728371107,
      "loss": 0.0287,
      "step": 1410
    },
    {
      "epoch": 3.279445727482679,
      "grad_norm": 0.32482001185417175,
      "learning_rate": 0.0001605824812051326,
      "loss": 0.0298,
      "step": 1420
    },
    {
      "epoch": 3.302540415704388,
      "grad_norm": 0.3234105408191681,
      "learning_rate": 0.0001599722820232484,
      "loss": 0.0335,
      "step": 1430
    },
    {
      "epoch": 3.325635103926097,
      "grad_norm": 0.28580203652381897,
      "learning_rate": 0.0001593585754249221,
      "loss": 0.0301,
      "step": 1440
    },
    {
      "epoch": 3.348729792147806,
      "grad_norm": 0.44926854968070984,
      "learning_rate": 0.00015874139730214477,
      "loss": 0.0317,
      "step": 1450
    },
    {
      "epoch": 3.371824480369515,
      "grad_norm": 0.3214055895805359,
      "learning_rate": 0.00015812078374993603,
      "loss": 0.0338,
      "step": 1460
    },
    {
      "epoch": 3.394919168591224,
      "grad_norm": 0.2814524471759796,
      "learning_rate": 0.000157496771064233,
      "loss": 0.0314,
      "step": 1470
    },
    {
      "epoch": 3.418013856812933,
      "grad_norm": 0.22763226926326752,
      "learning_rate": 0.0001568693957397675,
      "loss": 0.035,
      "step": 1480
    },
    {
      "epoch": 3.441108545034642,
      "grad_norm": 0.46513620018959045,
      "learning_rate": 0.000156238694467932,
      "loss": 0.0354,
      "step": 1490
    },
    {
      "epoch": 3.464203233256351,
      "grad_norm": 0.3795129060745239,
      "learning_rate": 0.0001556047041346333,
      "loss": 0.0285,
      "step": 1500
    },
    {
      "epoch": 3.48729792147806,
      "grad_norm": 0.22874844074249268,
      "learning_rate": 0.00015496746181813565,
      "loss": 0.0329,
      "step": 1510
    },
    {
      "epoch": 3.5103926096997693,
      "grad_norm": 0.43440544605255127,
      "learning_rate": 0.0001543270047868921,
      "loss": 0.0292,
      "step": 1520
    },
    {
      "epoch": 3.533487297921478,
      "grad_norm": 0.328204482793808,
      "learning_rate": 0.00015368337049736502,
      "loss": 0.0298,
      "step": 1530
    },
    {
      "epoch": 3.556581986143187,
      "grad_norm": 0.2836969196796417,
      "learning_rate": 0.00015303659659183534,
      "loss": 0.0326,
      "step": 1540
    },
    {
      "epoch": 3.579676674364896,
      "grad_norm": 0.2993479073047638,
      "learning_rate": 0.0001523867208962012,
      "loss": 0.0314,
      "step": 1550
    },
    {
      "epoch": 3.602771362586605,
      "grad_norm": 0.24445673823356628,
      "learning_rate": 0.00015173378141776568,
      "loss": 0.034,
      "step": 1560
    },
    {
      "epoch": 3.6258660508083143,
      "grad_norm": 0.2466069757938385,
      "learning_rate": 0.00015107781634301408,
      "loss": 0.0332,
      "step": 1570
    },
    {
      "epoch": 3.648960739030023,
      "grad_norm": 0.23315438628196716,
      "learning_rate": 0.0001504188640353804,
      "loss": 0.0279,
      "step": 1580
    },
    {
      "epoch": 3.672055427251732,
      "grad_norm": 0.2599908709526062,
      "learning_rate": 0.00014975696303300398,
      "loss": 0.0292,
      "step": 1590
    },
    {
      "epoch": 3.695150115473441,
      "grad_norm": 0.25828924775123596,
      "learning_rate": 0.00014909215204647548,
      "loss": 0.0312,
      "step": 1600
    },
    {
      "epoch": 3.71824480369515,
      "grad_norm": 0.32230719923973083,
      "learning_rate": 0.0001484244699565729,
      "loss": 0.0297,
      "step": 1610
    },
    {
      "epoch": 3.741339491916859,
      "grad_norm": 0.29874908924102783,
      "learning_rate": 0.00014775395581198778,
      "loss": 0.0298,
      "step": 1620
    },
    {
      "epoch": 3.764434180138568,
      "grad_norm": 0.2458525151014328,
      "learning_rate": 0.00014708064882704134,
      "loss": 0.0283,
      "step": 1630
    },
    {
      "epoch": 3.787528868360277,
      "grad_norm": 0.23176351189613342,
      "learning_rate": 0.00014640458837939129,
      "loss": 0.0272,
      "step": 1640
    },
    {
      "epoch": 3.8106235565819864,
      "grad_norm": 0.3253209590911865,
      "learning_rate": 0.00014572581400772863,
      "loss": 0.0275,
      "step": 1650
    },
    {
      "epoch": 3.8337182448036953,
      "grad_norm": 0.3006962537765503,
      "learning_rate": 0.00014504436540946548,
      "loss": 0.0264,
      "step": 1660
    },
    {
      "epoch": 3.856812933025404,
      "grad_norm": 0.3920678496360779,
      "learning_rate": 0.00014436028243841316,
      "loss": 0.0293,
      "step": 1670
    },
    {
      "epoch": 3.879907621247113,
      "grad_norm": 0.3614073097705841,
      "learning_rate": 0.0001436736051024517,
      "loss": 0.029,
      "step": 1680
    },
    {
      "epoch": 3.903002309468822,
      "grad_norm": 0.36879438161849976,
      "learning_rate": 0.00014298437356118982,
      "loss": 0.0333,
      "step": 1690
    },
    {
      "epoch": 3.9260969976905313,
      "grad_norm": 0.2536484897136688,
      "learning_rate": 0.00014229262812361622,
      "loss": 0.0294,
      "step": 1700
    },
    {
      "epoch": 3.9491916859122402,
      "grad_norm": 0.28662723302841187,
      "learning_rate": 0.00014159840924574222,
      "loss": 0.0297,
      "step": 1710
    },
    {
      "epoch": 3.972286374133949,
      "grad_norm": 0.2127179205417633,
      "learning_rate": 0.00014090175752823572,
      "loss": 0.0288,
      "step": 1720
    },
    {
      "epoch": 3.995381062355658,
      "grad_norm": 0.25283893942832947,
      "learning_rate": 0.00014020271371404682,
      "loss": 0.035,
      "step": 1730
    },
    {
      "epoch": 4.018475750577367,
      "grad_norm": 0.18742461502552032,
      "learning_rate": 0.0001395013186860247,
      "loss": 0.0262,
      "step": 1740
    },
    {
      "epoch": 4.041570438799076,
      "grad_norm": 0.36856940388679504,
      "learning_rate": 0.00013879761346452703,
      "loss": 0.0326,
      "step": 1750
    },
    {
      "epoch": 4.064665127020786,
      "grad_norm": 0.281253457069397,
      "learning_rate": 0.0001380916392050206,
      "loss": 0.0285,
      "step": 1760
    },
    {
      "epoch": 4.087759815242494,
      "grad_norm": 0.24782918393611908,
      "learning_rate": 0.00013738343719567464,
      "loss": 0.0272,
      "step": 1770
    },
    {
      "epoch": 4.1108545034642034,
      "grad_norm": 0.25319260358810425,
      "learning_rate": 0.00013667304885494588,
      "loss": 0.0315,
      "step": 1780
    },
    {
      "epoch": 4.133949191685912,
      "grad_norm": 0.288912296295166,
      "learning_rate": 0.0001359605157291565,
      "loss": 0.0308,
      "step": 1790
    },
    {
      "epoch": 4.157043879907621,
      "grad_norm": 0.2902250587940216,
      "learning_rate": 0.000135245879490064,
      "loss": 0.0279,
      "step": 1800
    },
    {
      "epoch": 4.180138568129331,
      "grad_norm": 0.2891808748245239,
      "learning_rate": 0.00013452918193242457,
      "loss": 0.0263,
      "step": 1810
    },
    {
      "epoch": 4.203233256351039,
      "grad_norm": 0.3770907521247864,
      "learning_rate": 0.00013381046497154816,
      "loss": 0.0308,
      "step": 1820
    },
    {
      "epoch": 4.226327944572748,
      "grad_norm": 0.31093716621398926,
      "learning_rate": 0.00013308977064084763,
      "loss": 0.0298,
      "step": 1830
    },
    {
      "epoch": 4.249422632794457,
      "grad_norm": 0.31624069809913635,
      "learning_rate": 0.00013236714108938013,
      "loss": 0.0296,
      "step": 1840
    },
    {
      "epoch": 4.272517321016166,
      "grad_norm": 0.376964807510376,
      "learning_rate": 0.00013164261857938228,
      "loss": 0.0339,
      "step": 1850
    },
    {
      "epoch": 4.295612009237876,
      "grad_norm": 0.48355865478515625,
      "learning_rate": 0.0001309162454837983,
      "loss": 0.0328,
      "step": 1860
    },
    {
      "epoch": 4.318706697459584,
      "grad_norm": 0.2842826545238495,
      "learning_rate": 0.00013018806428380207,
      "loss": 0.0264,
      "step": 1870
    },
    {
      "epoch": 4.341801385681293,
      "grad_norm": 0.39021459221839905,
      "learning_rate": 0.00012945811756631255,
      "loss": 0.029,
      "step": 1880
    },
    {
      "epoch": 4.364896073903003,
      "grad_norm": 0.20523010194301605,
      "learning_rate": 0.0001287264480215031,
      "loss": 0.0312,
      "step": 1890
    },
    {
      "epoch": 4.387990762124711,
      "grad_norm": 0.2571987509727478,
      "learning_rate": 0.00012799309844030497,
      "loss": 0.0282,
      "step": 1900
    },
    {
      "epoch": 4.4110854503464205,
      "grad_norm": 0.18886256217956543,
      "learning_rate": 0.00012725811171190437,
      "loss": 0.0271,
      "step": 1910
    },
    {
      "epoch": 4.434180138568129,
      "grad_norm": 0.27965831756591797,
      "learning_rate": 0.00012652153082123456,
      "loss": 0.0284,
      "step": 1920
    },
    {
      "epoch": 4.457274826789838,
      "grad_norm": 0.26982808113098145,
      "learning_rate": 0.00012578339884646166,
      "loss": 0.0274,
      "step": 1930
    },
    {
      "epoch": 4.480369515011548,
      "grad_norm": 0.17887461185455322,
      "learning_rate": 0.00012504375895646533,
      "loss": 0.0318,
      "step": 1940
    },
    {
      "epoch": 4.503464203233256,
      "grad_norm": 0.24898460507392883,
      "learning_rate": 0.00012430265440831396,
      "loss": 0.0284,
      "step": 1950
    },
    {
      "epoch": 4.5265588914549655,
      "grad_norm": 0.20980703830718994,
      "learning_rate": 0.0001235601285447352,
      "loss": 0.0276,
      "step": 1960
    },
    {
      "epoch": 4.549653579676674,
      "grad_norm": 0.21943718194961548,
      "learning_rate": 0.00012281622479158062,
      "loss": 0.0257,
      "step": 1970
    },
    {
      "epoch": 4.572748267898383,
      "grad_norm": 0.19166800379753113,
      "learning_rate": 0.00012207098665528636,
      "loss": 0.0264,
      "step": 1980
    },
    {
      "epoch": 4.595842956120093,
      "grad_norm": 0.23728342354297638,
      "learning_rate": 0.00012132445772032843,
      "loss": 0.0273,
      "step": 1990
    },
    {
      "epoch": 4.618937644341801,
      "grad_norm": 0.20198193192481995,
      "learning_rate": 0.00012057668164667406,
      "loss": 0.0257,
      "step": 2000
    },
    {
      "epoch": 4.64203233256351,
      "grad_norm": 0.21189096570014954,
      "learning_rate": 0.00011982770216722789,
      "loss": 0.0256,
      "step": 2010
    },
    {
      "epoch": 4.665127020785219,
      "grad_norm": 0.25154975056648254,
      "learning_rate": 0.0001190775630852746,
      "loss": 0.0294,
      "step": 2020
    },
    {
      "epoch": 4.688221709006928,
      "grad_norm": 0.24083566665649414,
      "learning_rate": 0.00011832630827191705,
      "loss": 0.0276,
      "step": 2030
    },
    {
      "epoch": 4.711316397228638,
      "grad_norm": 0.18492062389850616,
      "learning_rate": 0.00011757398166351038,
      "loss": 0.0285,
      "step": 2040
    },
    {
      "epoch": 4.734411085450346,
      "grad_norm": 0.2945506274700165,
      "learning_rate": 0.00011682062725909258,
      "loss": 0.0268,
      "step": 2050
    },
    {
      "epoch": 4.757505773672055,
      "grad_norm": 0.3341239392757416,
      "learning_rate": 0.00011606628911781123,
      "loss": 0.029,
      "step": 2060
    },
    {
      "epoch": 4.780600461893765,
      "grad_norm": 0.20814575254917145,
      "learning_rate": 0.0001153110113563468,
      "loss": 0.027,
      "step": 2070
    },
    {
      "epoch": 4.803695150115473,
      "grad_norm": 0.26448166370391846,
      "learning_rate": 0.00011455483814633238,
      "loss": 0.0281,
      "step": 2080
    },
    {
      "epoch": 4.826789838337183,
      "grad_norm": 0.24523232877254486,
      "learning_rate": 0.0001137978137117705,
      "loss": 0.0299,
      "step": 2090
    },
    {
      "epoch": 4.849884526558892,
      "grad_norm": 0.2897242605686188,
      "learning_rate": 0.00011303998232644657,
      "loss": 0.0308,
      "step": 2100
    },
    {
      "epoch": 4.8729792147806,
      "grad_norm": 0.23827065527439117,
      "learning_rate": 0.00011228138831133978,
      "loss": 0.0238,
      "step": 2110
    },
    {
      "epoch": 4.89607390300231,
      "grad_norm": 0.2748923897743225,
      "learning_rate": 0.00011152207603203088,
      "loss": 0.0255,
      "step": 2120
    },
    {
      "epoch": 4.919168591224018,
      "grad_norm": 0.2012598067522049,
      "learning_rate": 0.00011076208989610761,
      "loss": 0.0297,
      "step": 2130
    },
    {
      "epoch": 4.9422632794457275,
      "grad_norm": 0.21421535313129425,
      "learning_rate": 0.00011000147435056742,
      "loss": 0.0231,
      "step": 2140
    },
    {
      "epoch": 4.965357967667437,
      "grad_norm": 0.22895927727222443,
      "learning_rate": 0.00010924027387921818,
      "loss": 0.0265,
      "step": 2150
    },
    {
      "epoch": 4.988452655889145,
      "grad_norm": 0.2699049413204193,
      "learning_rate": 0.00010847853300007653,
      "loss": 0.0251,
      "step": 2160
    },
    {
      "epoch": 5.011547344110855,
      "grad_norm": 0.3293992877006531,
      "learning_rate": 0.00010771629626276428,
      "loss": 0.0229,
      "step": 2170
    },
    {
      "epoch": 5.034642032332563,
      "grad_norm": 0.2511574327945709,
      "learning_rate": 0.00010695360824590303,
      "loss": 0.0234,
      "step": 2180
    },
    {
      "epoch": 5.0577367205542725,
      "grad_norm": 0.3633870780467987,
      "learning_rate": 0.00010619051355450696,
      "loss": 0.0237,
      "step": 2190
    },
    {
      "epoch": 5.080831408775982,
      "grad_norm": 0.27033141255378723,
      "learning_rate": 0.00010542705681737422,
      "loss": 0.0278,
      "step": 2200
    },
    {
      "epoch": 5.10392609699769,
      "grad_norm": 0.33227813243865967,
      "learning_rate": 0.00010466328268447674,
      "loss": 0.028,
      "step": 2210
    },
    {
      "epoch": 5.1270207852194,
      "grad_norm": 0.19936224818229675,
      "learning_rate": 0.00010389923582434913,
      "loss": 0.0266,
      "step": 2220
    },
    {
      "epoch": 5.150115473441108,
      "grad_norm": 0.22531768679618835,
      "learning_rate": 0.0001031349609214761,
      "loss": 0.0279,
      "step": 2230
    },
    {
      "epoch": 5.173210161662817,
      "grad_norm": 0.2059541642665863,
      "learning_rate": 0.00010237050267367921,
      "loss": 0.0243,
      "step": 2240
    },
    {
      "epoch": 5.196304849884527,
      "grad_norm": 0.3016744554042816,
      "learning_rate": 0.00010160590578950273,
      "loss": 0.0287,
      "step": 2250
    },
    {
      "epoch": 5.219399538106235,
      "grad_norm": 0.3229542672634125,
      "learning_rate": 0.00010084121498559902,
      "loss": 0.028,
      "step": 2260
    },
    {
      "epoch": 5.242494226327945,
      "grad_norm": 0.28354501724243164,
      "learning_rate": 0.00010007647498411313,
      "loss": 0.0239,
      "step": 2270
    },
    {
      "epoch": 5.265588914549654,
      "grad_norm": 0.24493178725242615,
      "learning_rate": 9.931173051006746e-05,
      "loss": 0.0255,
      "step": 2280
    },
    {
      "epoch": 5.288683602771362,
      "grad_norm": 0.21446578204631805,
      "learning_rate": 9.854702628874596e-05,
      "loss": 0.0227,
      "step": 2290
    },
    {
      "epoch": 5.311778290993072,
      "grad_norm": 0.21182934939861298,
      "learning_rate": 9.778240704307844e-05,
      "loss": 0.0261,
      "step": 2300
    },
    {
      "epoch": 5.33487297921478,
      "grad_norm": 0.19283224642276764,
      "learning_rate": 9.701791749102495e-05,
      "loss": 0.0257,
      "step": 2310
    },
    {
      "epoch": 5.35796766743649,
      "grad_norm": 0.31905102729797363,
      "learning_rate": 9.625360234296065e-05,
      "loss": 0.0254,
      "step": 2320
    },
    {
      "epoch": 5.381062355658199,
      "grad_norm": 0.17523027956485748,
      "learning_rate": 9.548950629906077e-05,
      "loss": 0.0284,
      "step": 2330
    },
    {
      "epoch": 5.404157043879907,
      "grad_norm": 0.3119193911552429,
      "learning_rate": 9.47256740466865e-05,
      "loss": 0.0274,
      "step": 2340
    },
    {
      "epoch": 5.427251732101617,
      "grad_norm": 0.18288664519786835,
      "learning_rate": 9.396215025777139e-05,
      "loss": 0.0243,
      "step": 2350
    },
    {
      "epoch": 5.450346420323326,
      "grad_norm": 0.3371618092060089,
      "learning_rate": 9.3198979586209e-05,
      "loss": 0.0263,
      "step": 2360
    },
    {
      "epoch": 5.4734411085450345,
      "grad_norm": 0.1650622934103012,
      "learning_rate": 9.243620666524099e-05,
      "loss": 0.0238,
      "step": 2370
    },
    {
      "epoch": 5.496535796766744,
      "grad_norm": 0.2566249370574951,
      "learning_rate": 9.167387610484712e-05,
      "loss": 0.0233,
      "step": 2380
    },
    {
      "epoch": 5.519630484988452,
      "grad_norm": 0.312350869178772,
      "learning_rate": 9.091203248913607e-05,
      "loss": 0.0265,
      "step": 2390
    },
    {
      "epoch": 5.542725173210162,
      "grad_norm": 0.14863193035125732,
      "learning_rate": 9.015072037373816e-05,
      "loss": 0.0235,
      "step": 2400
    },
    {
      "epoch": 5.565819861431871,
      "grad_norm": 0.2509799003601074,
      "learning_rate": 8.938998428319937e-05,
      "loss": 0.0278,
      "step": 2410
    },
    {
      "epoch": 5.5889145496535795,
      "grad_norm": 0.1843155473470688,
      "learning_rate": 8.862986870837753e-05,
      "loss": 0.0284,
      "step": 2420
    },
    {
      "epoch": 5.612009237875289,
      "grad_norm": 0.3865255117416382,
      "learning_rate": 8.787041810384019e-05,
      "loss": 0.0256,
      "step": 2430
    },
    {
      "epoch": 5.635103926096997,
      "grad_norm": 0.2363845258951187,
      "learning_rate": 8.711167688526493e-05,
      "loss": 0.0271,
      "step": 2440
    },
    {
      "epoch": 5.658198614318707,
      "grad_norm": 0.209372416138649,
      "learning_rate": 8.635368942684153e-05,
      "loss": 0.0248,
      "step": 2450
    },
    {
      "epoch": 5.681293302540416,
      "grad_norm": 0.22078511118888855,
      "learning_rate": 8.559650005867684e-05,
      "loss": 0.0252,
      "step": 2460
    },
    {
      "epoch": 5.704387990762124,
      "grad_norm": 0.20760229229927063,
      "learning_rate": 8.484015306420242e-05,
      "loss": 0.0235,
      "step": 2470
    },
    {
      "epoch": 5.727482678983834,
      "grad_norm": 0.18812189996242523,
      "learning_rate": 8.408469267758432e-05,
      "loss": 0.0238,
      "step": 2480
    },
    {
      "epoch": 5.750577367205542,
      "grad_norm": 0.1914282590150833,
      "learning_rate": 8.333016308113629e-05,
      "loss": 0.0255,
      "step": 2490
    },
    {
      "epoch": 5.773672055427252,
      "grad_norm": 0.288441926240921,
      "learning_rate": 8.257660840273579e-05,
      "loss": 0.0271,
      "step": 2500
    },
    {
      "epoch": 5.796766743648961,
      "grad_norm": 0.23392659425735474,
      "learning_rate": 8.182407271324326e-05,
      "loss": 0.0279,
      "step": 2510
    },
    {
      "epoch": 5.819861431870669,
      "grad_norm": 0.25330981612205505,
      "learning_rate": 8.107260002392457e-05,
      "loss": 0.0248,
      "step": 2520
    },
    {
      "epoch": 5.842956120092379,
      "grad_norm": 0.22893041372299194,
      "learning_rate": 8.032223428387719e-05,
      "loss": 0.0233,
      "step": 2530
    },
    {
      "epoch": 5.866050808314088,
      "grad_norm": 0.18465740978717804,
      "learning_rate": 7.95730193774597e-05,
      "loss": 0.0227,
      "step": 2540
    },
    {
      "epoch": 5.8891454965357966,
      "grad_norm": 0.17142651975154877,
      "learning_rate": 7.882499912172557e-05,
      "loss": 0.024,
      "step": 2550
    },
    {
      "epoch": 5.912240184757506,
      "grad_norm": 0.19113211333751678,
      "learning_rate": 7.807821726386022e-05,
      "loss": 0.0231,
      "step": 2560
    },
    {
      "epoch": 5.935334872979215,
      "grad_norm": 0.20570556819438934,
      "learning_rate": 7.733271747862265e-05,
      "loss": 0.0225,
      "step": 2570
    },
    {
      "epoch": 5.958429561200924,
      "grad_norm": 0.21998938918113708,
      "learning_rate": 7.65885433657913e-05,
      "loss": 0.0226,
      "step": 2580
    },
    {
      "epoch": 5.981524249422633,
      "grad_norm": 0.35705238580703735,
      "learning_rate": 7.584573844761393e-05,
      "loss": 0.0197,
      "step": 2590
    },
    {
      "epoch": 6.0046189376443415,
      "grad_norm": 0.27228960394859314,
      "learning_rate": 7.510434616626243e-05,
      "loss": 0.0252,
      "step": 2600
    },
    {
      "epoch": 6.027713625866051,
      "grad_norm": 0.19615407288074493,
      "learning_rate": 7.4364409881292e-05,
      "loss": 0.026,
      "step": 2610
    },
    {
      "epoch": 6.05080831408776,
      "grad_norm": 0.27591854333877563,
      "learning_rate": 7.362597286710562e-05,
      "loss": 0.0262,
      "step": 2620
    },
    {
      "epoch": 6.073903002309469,
      "grad_norm": 0.26121532917022705,
      "learning_rate": 7.288907831042279e-05,
      "loss": 0.0206,
      "step": 2630
    },
    {
      "epoch": 6.096997690531178,
      "grad_norm": 0.365448921918869,
      "learning_rate": 7.215376930775403e-05,
      "loss": 0.0264,
      "step": 2640
    },
    {
      "epoch": 6.1200923787528865,
      "grad_norm": 0.24886159598827362,
      "learning_rate": 7.14200888628804e-05,
      "loss": 0.0225,
      "step": 2650
    },
    {
      "epoch": 6.143187066974596,
      "grad_norm": 0.23590990900993347,
      "learning_rate": 7.06880798843385e-05,
      "loss": 0.0214,
      "step": 2660
    },
    {
      "epoch": 6.166281755196305,
      "grad_norm": 0.34595030546188354,
      "learning_rate": 6.995778518291089e-05,
      "loss": 0.024,
      "step": 2670
    },
    {
      "epoch": 6.189376443418014,
      "grad_norm": 0.2593909502029419,
      "learning_rate": 6.922924746912245e-05,
      "loss": 0.0227,
      "step": 2680
    },
    {
      "epoch": 6.212471131639723,
      "grad_norm": 0.30012357234954834,
      "learning_rate": 6.850250935074243e-05,
      "loss": 0.0237,
      "step": 2690
    },
    {
      "epoch": 6.235565819861431,
      "grad_norm": 0.219698965549469,
      "learning_rate": 6.777761333029275e-05,
      "loss": 0.0211,
      "step": 2700
    },
    {
      "epoch": 6.258660508083141,
      "grad_norm": 0.20809656381607056,
      "learning_rate": 6.705460180256199e-05,
      "loss": 0.0216,
      "step": 2710
    },
    {
      "epoch": 6.28175519630485,
      "grad_norm": 0.19182954728603363,
      "learning_rate": 6.633351705212617e-05,
      "loss": 0.0228,
      "step": 2720
    },
    {
      "epoch": 6.304849884526559,
      "grad_norm": 0.171848326921463,
      "learning_rate": 6.561440125087587e-05,
      "loss": 0.0232,
      "step": 2730
    },
    {
      "epoch": 6.327944572748268,
      "grad_norm": 0.2251206785440445,
      "learning_rate": 6.489729645554959e-05,
      "loss": 0.023,
      "step": 2740
    },
    {
      "epoch": 6.351039260969977,
      "grad_norm": 0.20676329731941223,
      "learning_rate": 6.418224460527428e-05,
      "loss": 0.0216,
      "step": 2750
    },
    {
      "epoch": 6.374133949191686,
      "grad_norm": 0.1491789072751999,
      "learning_rate": 6.346928751911255e-05,
      "loss": 0.0196,
      "step": 2760
    },
    {
      "epoch": 6.397228637413395,
      "grad_norm": 0.1880139261484146,
      "learning_rate": 6.275846689361693e-05,
      "loss": 0.0251,
      "step": 2770
    },
    {
      "epoch": 6.4203233256351036,
      "grad_norm": 0.24984711408615112,
      "learning_rate": 6.204982430039124e-05,
      "loss": 0.0218,
      "step": 2780
    },
    {
      "epoch": 6.443418013856813,
      "grad_norm": 0.1841384768486023,
      "learning_rate": 6.134340118365936e-05,
      "loss": 0.0224,
      "step": 2790
    },
    {
      "epoch": 6.466512702078522,
      "grad_norm": 0.2632406949996948,
      "learning_rate": 6.063923885784138e-05,
      "loss": 0.021,
      "step": 2800
    },
    {
      "epoch": 6.489607390300231,
      "grad_norm": 0.21054299175739288,
      "learning_rate": 5.9937378505137476e-05,
      "loss": 0.0232,
      "step": 2810
    },
    {
      "epoch": 6.51270207852194,
      "grad_norm": 0.18742221593856812,
      "learning_rate": 5.9237861173119234e-05,
      "loss": 0.0218,
      "step": 2820
    },
    {
      "epoch": 6.535796766743649,
      "grad_norm": 0.16454491019248962,
      "learning_rate": 5.854072777232914e-05,
      "loss": 0.02,
      "step": 2830
    },
    {
      "epoch": 6.558891454965358,
      "grad_norm": 0.12906990945339203,
      "learning_rate": 5.7846019073887904e-05,
      "loss": 0.0233,
      "step": 2840
    },
    {
      "epoch": 6.581986143187067,
      "grad_norm": 0.1799432933330536,
      "learning_rate": 5.715377570711019e-05,
      "loss": 0.0225,
      "step": 2850
    },
    {
      "epoch": 6.605080831408776,
      "grad_norm": 0.25904062390327454,
      "learning_rate": 5.6464038157128106e-05,
      "loss": 0.0241,
      "step": 2860
    },
    {
      "epoch": 6.628175519630485,
      "grad_norm": 0.14367002248764038,
      "learning_rate": 5.577684676252384e-05,
      "loss": 0.022,
      "step": 2870
    },
    {
      "epoch": 6.651270207852194,
      "grad_norm": 0.20157106220722198,
      "learning_rate": 5.509224171297027e-05,
      "loss": 0.0215,
      "step": 2880
    },
    {
      "epoch": 6.674364896073903,
      "grad_norm": 0.18459396064281464,
      "learning_rate": 5.4410263046880524e-05,
      "loss": 0.0252,
      "step": 2890
    },
    {
      "epoch": 6.697459584295612,
      "grad_norm": 0.17748424410820007,
      "learning_rate": 5.373095064906657e-05,
      "loss": 0.024,
      "step": 2900
    },
    {
      "epoch": 6.720554272517321,
      "grad_norm": 0.19444628059864044,
      "learning_rate": 5.305434424840623e-05,
      "loss": 0.0202,
      "step": 2910
    },
    {
      "epoch": 6.74364896073903,
      "grad_norm": 0.21914024651050568,
      "learning_rate": 5.2380483415520196e-05,
      "loss": 0.0229,
      "step": 2920
    },
    {
      "epoch": 6.766743648960739,
      "grad_norm": 0.22515784204006195,
      "learning_rate": 5.170940756045725e-05,
      "loss": 0.0241,
      "step": 2930
    },
    {
      "epoch": 6.789838337182448,
      "grad_norm": 0.2511735260486603,
      "learning_rate": 5.104115593038976e-05,
      "loss": 0.0202,
      "step": 2940
    },
    {
      "epoch": 6.812933025404157,
      "grad_norm": 0.22903645038604736,
      "learning_rate": 5.0375767607318106e-05,
      "loss": 0.0229,
      "step": 2950
    },
    {
      "epoch": 6.836027713625866,
      "grad_norm": 0.24204668402671814,
      "learning_rate": 4.97132815057854e-05,
      "loss": 0.0209,
      "step": 2960
    },
    {
      "epoch": 6.859122401847575,
      "grad_norm": 0.3427051901817322,
      "learning_rate": 4.905373637060108e-05,
      "loss": 0.0226,
      "step": 2970
    },
    {
      "epoch": 6.882217090069284,
      "grad_norm": 0.189266636967659,
      "learning_rate": 4.8397170774575394e-05,
      "loss": 0.0183,
      "step": 2980
    },
    {
      "epoch": 6.905311778290993,
      "grad_norm": 0.17002785205841064,
      "learning_rate": 4.7743623116263245e-05,
      "loss": 0.0187,
      "step": 2990
    },
    {
      "epoch": 6.928406466512702,
      "grad_norm": 0.21065160632133484,
      "learning_rate": 4.709313161771867e-05,
      "loss": 0.0237,
      "step": 3000
    },
    {
      "epoch": 6.951501154734411,
      "grad_norm": 0.14337095618247986,
      "learning_rate": 4.644573432225939e-05,
      "loss": 0.0205,
      "step": 3010
    },
    {
      "epoch": 6.97459584295612,
      "grad_norm": 0.14885887503623962,
      "learning_rate": 4.580146909224173e-05,
      "loss": 0.0204,
      "step": 3020
    },
    {
      "epoch": 6.997690531177829,
      "grad_norm": 0.1806829869747162,
      "learning_rate": 4.516037360684673e-05,
      "loss": 0.0186,
      "step": 3030
    },
    {
      "epoch": 7.020785219399538,
      "grad_norm": 0.1977592259645462,
      "learning_rate": 4.452248535987588e-05,
      "loss": 0.0203,
      "step": 3040
    },
    {
      "epoch": 7.043879907621247,
      "grad_norm": 0.17835497856140137,
      "learning_rate": 4.388784165755894e-05,
      "loss": 0.0208,
      "step": 3050
    },
    {
      "epoch": 7.066974595842956,
      "grad_norm": 0.29322361946105957,
      "learning_rate": 4.3256479616371636e-05,
      "loss": 0.0184,
      "step": 3060
    },
    {
      "epoch": 7.090069284064665,
      "grad_norm": 0.17030254006385803,
      "learning_rate": 4.262843616086534e-05,
      "loss": 0.02,
      "step": 3070
    },
    {
      "epoch": 7.113163972286374,
      "grad_norm": 0.18691743910312653,
      "learning_rate": 4.2003748021507336e-05,
      "loss": 0.019,
      "step": 3080
    },
    {
      "epoch": 7.1362586605080836,
      "grad_norm": 0.23188500106334686,
      "learning_rate": 4.1382451732532665e-05,
      "loss": 0.0189,
      "step": 3090
    },
    {
      "epoch": 7.159353348729792,
      "grad_norm": 0.20676599442958832,
      "learning_rate": 4.076458362980764e-05,
      "loss": 0.0209,
      "step": 3100
    },
    {
      "epoch": 7.182448036951501,
      "grad_norm": 0.22161860764026642,
      "learning_rate": 4.0150179848704614e-05,
      "loss": 0.0203,
      "step": 3110
    },
    {
      "epoch": 7.20554272517321,
      "grad_norm": 0.17719246447086334,
      "learning_rate": 3.9539276321988764e-05,
      "loss": 0.0201,
      "step": 3120
    },
    {
      "epoch": 7.228637413394919,
      "grad_norm": 0.2837867736816406,
      "learning_rate": 3.89319087777164e-05,
      "loss": 0.0226,
      "step": 3130
    },
    {
      "epoch": 7.2517321016166285,
      "grad_norm": 0.2676970362663269,
      "learning_rate": 3.832811273714569e-05,
      "loss": 0.0219,
      "step": 3140
    },
    {
      "epoch": 7.274826789838337,
      "grad_norm": 0.14641159772872925,
      "learning_rate": 3.772792351265917e-05,
      "loss": 0.0202,
      "step": 3150
    },
    {
      "epoch": 7.297921478060046,
      "grad_norm": 0.13401910662651062,
      "learning_rate": 3.713137620569833e-05,
      "loss": 0.0186,
      "step": 3160
    },
    {
      "epoch": 7.321016166281755,
      "grad_norm": 0.20314963161945343,
      "learning_rate": 3.653850570471106e-05,
      "loss": 0.0195,
      "step": 3170
    },
    {
      "epoch": 7.344110854503464,
      "grad_norm": 0.2034531831741333,
      "learning_rate": 3.594934668311101e-05,
      "loss": 0.0174,
      "step": 3180
    },
    {
      "epoch": 7.3672055427251735,
      "grad_norm": 0.22085122764110565,
      "learning_rate": 3.536393359724989e-05,
      "loss": 0.0202,
      "step": 3190
    },
    {
      "epoch": 7.390300230946882,
      "grad_norm": 0.20668219029903412,
      "learning_rate": 3.4782300684402134e-05,
      "loss": 0.0215,
      "step": 3200
    },
    {
      "epoch": 7.413394919168591,
      "grad_norm": 0.20111581683158875,
      "learning_rate": 3.420448196076282e-05,
      "loss": 0.0194,
      "step": 3210
    },
    {
      "epoch": 7.436489607390301,
      "grad_norm": 0.2311002016067505,
      "learning_rate": 3.363051121945809e-05,
      "loss": 0.0182,
      "step": 3220
    },
    {
      "epoch": 7.459584295612009,
      "grad_norm": 0.2030162215232849,
      "learning_rate": 3.30604220285689e-05,
      "loss": 0.0166,
      "step": 3230
    },
    {
      "epoch": 7.482678983833718,
      "grad_norm": 0.1821678727865219,
      "learning_rate": 3.249424772916765e-05,
      "loss": 0.02,
      "step": 3240
    },
    {
      "epoch": 7.505773672055427,
      "grad_norm": 0.38144606351852417,
      "learning_rate": 3.19320214333685e-05,
      "loss": 0.0194,
      "step": 3250
    },
    {
      "epoch": 7.528868360277136,
      "grad_norm": 0.26288631558418274,
      "learning_rate": 3.1373776022390735e-05,
      "loss": 0.0201,
      "step": 3260
    },
    {
      "epoch": 7.551963048498846,
      "grad_norm": 0.26908478140830994,
      "learning_rate": 3.081954414463564e-05,
      "loss": 0.0203,
      "step": 3270
    },
    {
      "epoch": 7.575057736720554,
      "grad_norm": 0.1838100701570511,
      "learning_rate": 3.0269358213777276e-05,
      "loss": 0.02,
      "step": 3280
    },
    {
      "epoch": 7.598152424942263,
      "grad_norm": 0.19634702801704407,
      "learning_rate": 2.97232504068667e-05,
      "loss": 0.024,
      "step": 3290
    },
    {
      "epoch": 7.621247113163973,
      "grad_norm": 0.17007845640182495,
      "learning_rate": 2.9181252662450153e-05,
      "loss": 0.0209,
      "step": 3300
    },
    {
      "epoch": 7.644341801385681,
      "grad_norm": 0.1375173032283783,
      "learning_rate": 2.8643396678701073e-05,
      "loss": 0.0195,
      "step": 3310
    },
    {
      "epoch": 7.6674364896073905,
      "grad_norm": 0.1866098791360855,
      "learning_rate": 2.8109713911566428e-05,
      "loss": 0.0185,
      "step": 3320
    },
    {
      "epoch": 7.690531177829099,
      "grad_norm": 0.18592612445354462,
      "learning_rate": 2.758023557292695e-05,
      "loss": 0.0178,
      "step": 3330
    },
    {
      "epoch": 7.713625866050808,
      "grad_norm": 0.15712429583072662,
      "learning_rate": 2.7054992628771668e-05,
      "loss": 0.0196,
      "step": 3340
    },
    {
      "epoch": 7.736720554272518,
      "grad_norm": 0.1279047578573227,
      "learning_rate": 2.65340157973871e-05,
      "loss": 0.0197,
      "step": 3350
    },
    {
      "epoch": 7.759815242494226,
      "grad_norm": 0.22441239655017853,
      "learning_rate": 2.6017335547560452e-05,
      "loss": 0.0191,
      "step": 3360
    },
    {
      "epoch": 7.7829099307159355,
      "grad_norm": 0.20676462352275848,
      "learning_rate": 2.5504982096798025e-05,
      "loss": 0.0199,
      "step": 3370
    },
    {
      "epoch": 7.806004618937644,
      "grad_norm": 0.2775067985057831,
      "learning_rate": 2.4996985409557595e-05,
      "loss": 0.0185,
      "step": 3380
    },
    {
      "epoch": 7.829099307159353,
      "grad_norm": 0.2614492177963257,
      "learning_rate": 2.4493375195496292e-05,
      "loss": 0.0169,
      "step": 3390
    },
    {
      "epoch": 7.852193995381063,
      "grad_norm": 0.2875981330871582,
      "learning_rate": 2.3994180907732856e-05,
      "loss": 0.018,
      "step": 3400
    },
    {
      "epoch": 7.875288683602771,
      "grad_norm": 0.18182726204395294,
      "learning_rate": 2.349943174112521e-05,
      "loss": 0.0194,
      "step": 3410
    },
    {
      "epoch": 7.8983833718244805,
      "grad_norm": 0.18898028135299683,
      "learning_rate": 2.300915663056302e-05,
      "loss": 0.0157,
      "step": 3420
    },
    {
      "epoch": 7.921478060046189,
      "grad_norm": 0.5363194346427917,
      "learning_rate": 2.2523384249275347e-05,
      "loss": 0.0178,
      "step": 3430
    },
    {
      "epoch": 7.944572748267898,
      "grad_norm": 0.1563793420791626,
      "learning_rate": 2.2042143007153994e-05,
      "loss": 0.0183,
      "step": 3440
    },
    {
      "epoch": 7.967667436489608,
      "grad_norm": 0.1951935738325119,
      "learning_rate": 2.156546104909163e-05,
      "loss": 0.0216,
      "step": 3450
    },
    {
      "epoch": 7.990762124711316,
      "grad_norm": 0.19452138245105743,
      "learning_rate": 2.1093366253336066e-05,
      "loss": 0.0191,
      "step": 3460
    },
    {
      "epoch": 8.013856812933025,
      "grad_norm": 0.20190148055553436,
      "learning_rate": 2.0625886229859582e-05,
      "loss": 0.0188,
      "step": 3470
    },
    {
      "epoch": 8.036951501154734,
      "grad_norm": 0.2394581139087677,
      "learning_rate": 2.0163048318744493e-05,
      "loss": 0.0181,
      "step": 3480
    },
    {
      "epoch": 8.060046189376443,
      "grad_norm": 0.17284128069877625,
      "learning_rate": 1.9704879588583836e-05,
      "loss": 0.0172,
      "step": 3490
    },
    {
      "epoch": 8.083140877598153,
      "grad_norm": 0.1908102035522461,
      "learning_rate": 1.9251406834898556e-05,
      "loss": 0.0175,
      "step": 3500
    },
    {
      "epoch": 8.106235565819862,
      "grad_norm": 0.25007396936416626,
      "learning_rate": 1.8802656578570244e-05,
      "loss": 0.016,
      "step": 3510
    },
    {
      "epoch": 8.129330254041571,
      "grad_norm": 0.36770346760749817,
      "learning_rate": 1.835865506429021e-05,
      "loss": 0.0175,
      "step": 3520
    },
    {
      "epoch": 8.152424942263279,
      "grad_norm": 0.20147453248500824,
      "learning_rate": 1.7919428259024518e-05,
      "loss": 0.0213,
      "step": 3530
    },
    {
      "epoch": 8.175519630484988,
      "grad_norm": 0.1591079831123352,
      "learning_rate": 1.748500185049532e-05,
      "loss": 0.0166,
      "step": 3540
    },
    {
      "epoch": 8.198614318706698,
      "grad_norm": 0.2231104075908661,
      "learning_rate": 1.705540124567858e-05,
      "loss": 0.0173,
      "step": 3550
    },
    {
      "epoch": 8.221709006928407,
      "grad_norm": 0.20360170304775238,
      "learning_rate": 1.663065156931819e-05,
      "loss": 0.0181,
      "step": 3560
    },
    {
      "epoch": 8.244803695150116,
      "grad_norm": 0.20085148513317108,
      "learning_rate": 1.621077766245652e-05,
      "loss": 0.0173,
      "step": 3570
    },
    {
      "epoch": 8.267898383371824,
      "grad_norm": 0.1734970211982727,
      "learning_rate": 1.57958040809816e-05,
      "loss": 0.0196,
      "step": 3580
    },
    {
      "epoch": 8.290993071593533,
      "grad_norm": 0.17935489118099213,
      "learning_rate": 1.5385755094191102e-05,
      "loss": 0.0177,
      "step": 3590
    },
    {
      "epoch": 8.314087759815243,
      "grad_norm": 0.19008755683898926,
      "learning_rate": 1.4980654683372886e-05,
      "loss": 0.0193,
      "step": 3600
    },
    {
      "epoch": 8.337182448036952,
      "grad_norm": 0.17935799062252045,
      "learning_rate": 1.4580526540402461e-05,
      "loss": 0.0165,
      "step": 3610
    },
    {
      "epoch": 8.360277136258661,
      "grad_norm": 0.2387223094701767,
      "learning_rate": 1.4185394066357483e-05,
      "loss": 0.018,
      "step": 3620
    },
    {
      "epoch": 8.383371824480369,
      "grad_norm": 0.1772770881652832,
      "learning_rate": 1.37952803701491e-05,
      "loss": 0.0173,
      "step": 3630
    },
    {
      "epoch": 8.406466512702078,
      "grad_norm": 0.30018919706344604,
      "learning_rate": 1.3410208267170466e-05,
      "loss": 0.0186,
      "step": 3640
    },
    {
      "epoch": 8.429561200923787,
      "grad_norm": 0.3067221939563751,
      "learning_rate": 1.3030200277962369e-05,
      "loss": 0.0175,
      "step": 3650
    },
    {
      "epoch": 8.452655889145497,
      "grad_norm": 0.3162385821342468,
      "learning_rate": 1.2655278626896227e-05,
      "loss": 0.0181,
      "step": 3660
    },
    {
      "epoch": 8.475750577367206,
      "grad_norm": 0.16854941844940186,
      "learning_rate": 1.2285465240874283e-05,
      "loss": 0.0183,
      "step": 3670
    },
    {
      "epoch": 8.498845265588914,
      "grad_norm": 0.1159670352935791,
      "learning_rate": 1.192078174804715e-05,
      "loss": 0.016,
      "step": 3680
    },
    {
      "epoch": 8.521939953810623,
      "grad_norm": 0.12184049934148788,
      "learning_rate": 1.1561249476549052e-05,
      "loss": 0.0157,
      "step": 3690
    },
    {
      "epoch": 8.545034642032332,
      "grad_norm": 0.15742027759552002,
      "learning_rate": 1.1206889453250313e-05,
      "loss": 0.0184,
      "step": 3700
    },
    {
      "epoch": 8.568129330254042,
      "grad_norm": 0.2727457582950592,
      "learning_rate": 1.0857722402527847e-05,
      "loss": 0.016,
      "step": 3710
    },
    {
      "epoch": 8.591224018475751,
      "grad_norm": 0.17201822996139526,
      "learning_rate": 1.0513768745052843e-05,
      "loss": 0.0174,
      "step": 3720
    },
    {
      "epoch": 8.61431870669746,
      "grad_norm": 0.17287689447402954,
      "learning_rate": 1.0175048596596682e-05,
      "loss": 0.0167,
      "step": 3730
    },
    {
      "epoch": 8.637413394919168,
      "grad_norm": 0.1511421501636505,
      "learning_rate": 9.841581766854401e-06,
      "loss": 0.0184,
      "step": 3740
    },
    {
      "epoch": 8.660508083140877,
      "grad_norm": 0.1608920693397522,
      "learning_rate": 9.513387758286196e-06,
      "loss": 0.0179,
      "step": 3750
    },
    {
      "epoch": 8.683602771362587,
      "grad_norm": 0.20927520096302032,
      "learning_rate": 9.190485764976753e-06,
      "loss": 0.0159,
      "step": 3760
    },
    {
      "epoch": 8.706697459584296,
      "grad_norm": 0.21599853038787842,
      "learning_rate": 8.872894671512799e-06,
      "loss": 0.0199,
      "step": 3770
    },
    {
      "epoch": 8.729792147806005,
      "grad_norm": 0.19188404083251953,
      "learning_rate": 8.560633051878631e-06,
      "loss": 0.0158,
      "step": 3780
    },
    {
      "epoch": 8.752886836027713,
      "grad_norm": 0.1599506288766861,
      "learning_rate": 8.253719168369768e-06,
      "loss": 0.0161,
      "step": 3790
    },
    {
      "epoch": 8.775981524249422,
      "grad_norm": 0.1853979527950287,
      "learning_rate": 7.952170970524985e-06,
      "loss": 0.0191,
      "step": 3800
    },
    {
      "epoch": 8.799076212471132,
      "grad_norm": 0.2847438156604767,
      "learning_rate": 7.656006094076506e-06,
      "loss": 0.0178,
      "step": 3810
    },
    {
      "epoch": 8.822170900692841,
      "grad_norm": 0.13585425913333893,
      "learning_rate": 7.365241859918659e-06,
      "loss": 0.0166,
      "step": 3820
    },
    {
      "epoch": 8.84526558891455,
      "grad_norm": 0.19831091165542603,
      "learning_rate": 7.07989527309475e-06,
      "loss": 0.019,
      "step": 3830
    },
    {
      "epoch": 8.868360277136258,
      "grad_norm": 0.3466614782810211,
      "learning_rate": 6.799983021802692e-06,
      "loss": 0.0164,
      "step": 3840
    },
    {
      "epoch": 8.891454965357967,
      "grad_norm": 0.344882607460022,
      "learning_rate": 6.525521476418939e-06,
      "loss": 0.0162,
      "step": 3850
    },
    {
      "epoch": 8.914549653579677,
      "grad_norm": 0.2650250792503357,
      "learning_rate": 6.256526688541043e-06,
      "loss": 0.0158,
      "step": 3860
    },
    {
      "epoch": 8.937644341801386,
      "grad_norm": 0.15144085884094238,
      "learning_rate": 5.99301439004899e-06,
      "loss": 0.0176,
      "step": 3870
    },
    {
      "epoch": 8.960739030023095,
      "grad_norm": 0.11926089227199554,
      "learning_rate": 5.734999992185008e-06,
      "loss": 0.0143,
      "step": 3880
    },
    {
      "epoch": 8.983833718244803,
      "grad_norm": 0.43381285667419434,
      "learning_rate": 5.482498584652462e-06,
      "loss": 0.0163,
      "step": 3890
    },
    {
      "epoch": 9.006928406466512,
      "grad_norm": 0.19410696625709534,
      "learning_rate": 5.235524934733094e-06,
      "loss": 0.0182,
      "step": 3900
    },
    {
      "epoch": 9.030023094688222,
      "grad_norm": 0.2376604527235031,
      "learning_rate": 4.99409348642359e-06,
      "loss": 0.0164,
      "step": 3910
    },
    {
      "epoch": 9.053117782909931,
      "grad_norm": 0.18821699917316437,
      "learning_rate": 4.7582183595906715e-06,
      "loss": 0.0214,
      "step": 3920
    },
    {
      "epoch": 9.07621247113164,
      "grad_norm": 0.25581300258636475,
      "learning_rate": 4.527913349145441e-06,
      "loss": 0.0166,
      "step": 3930
    },
    {
      "epoch": 9.099307159353348,
      "grad_norm": 0.24828465282917023,
      "learning_rate": 4.303191924236538e-06,
      "loss": 0.0169,
      "step": 3940
    },
    {
      "epoch": 9.122401847575057,
      "grad_norm": 0.19727648794651031,
      "learning_rate": 4.0840672274623955e-06,
      "loss": 0.0161,
      "step": 3950
    },
    {
      "epoch": 9.145496535796767,
      "grad_norm": 0.1969439536333084,
      "learning_rate": 3.870552074102662e-06,
      "loss": 0.0164,
      "step": 3960
    },
    {
      "epoch": 9.168591224018476,
      "grad_norm": 0.28042295575141907,
      "learning_rate": 3.6626589513686473e-06,
      "loss": 0.0169,
      "step": 3970
    },
    {
      "epoch": 9.191685912240185,
      "grad_norm": 0.2694367468357086,
      "learning_rate": 3.4604000176731044e-06,
      "loss": 0.0177,
      "step": 3980
    },
    {
      "epoch": 9.214780600461895,
      "grad_norm": 0.27931103110313416,
      "learning_rate": 3.263787101919036e-06,
      "loss": 0.0204,
      "step": 3990
    },
    {
      "epoch": 9.237875288683602,
      "grad_norm": 0.29098454117774963,
      "learning_rate": 3.0728317028080657e-06,
      "loss": 0.02,
      "step": 4000
    },
    {
      "epoch": 9.260969976905312,
      "grad_norm": 0.1651170551776886,
      "learning_rate": 2.8875449881677676e-06,
      "loss": 0.0155,
      "step": 4010
    },
    {
      "epoch": 9.28406466512702,
      "grad_norm": 0.213905468583107,
      "learning_rate": 2.7079377942986427e-06,
      "loss": 0.0179,
      "step": 4020
    },
    {
      "epoch": 9.30715935334873,
      "grad_norm": 0.1109078899025917,
      "learning_rate": 2.5340206253403276e-06,
      "loss": 0.0132,
      "step": 4030
    },
    {
      "epoch": 9.33025404157044,
      "grad_norm": 0.19865043461322784,
      "learning_rate": 2.3658036526572726e-06,
      "loss": 0.0175,
      "step": 4040
    },
    {
      "epoch": 9.353348729792147,
      "grad_norm": 0.15456566214561462,
      "learning_rate": 2.2032967142439297e-06,
      "loss": 0.0149,
      "step": 4050
    },
    {
      "epoch": 9.376443418013857,
      "grad_norm": 0.11047529429197311,
      "learning_rate": 2.0465093141492696e-06,
      "loss": 0.0152,
      "step": 4060
    },
    {
      "epoch": 9.399538106235566,
      "grad_norm": 0.13611409068107605,
      "learning_rate": 1.895450621921091e-06,
      "loss": 0.0145,
      "step": 4070
    },
    {
      "epoch": 9.422632794457275,
      "grad_norm": 0.2234010398387909,
      "learning_rate": 1.7501294720696504e-06,
      "loss": 0.0189,
      "step": 4080
    },
    {
      "epoch": 9.445727482678985,
      "grad_norm": 0.26444539427757263,
      "learning_rate": 1.6105543635510202e-06,
      "loss": 0.0157,
      "step": 4090
    },
    {
      "epoch": 9.468822170900692,
      "grad_norm": 0.1748080998659134,
      "learning_rate": 1.4767334592700188e-06,
      "loss": 0.0173,
      "step": 4100
    },
    {
      "epoch": 9.491916859122401,
      "grad_norm": 0.16052615642547607,
      "learning_rate": 1.3486745856028381e-06,
      "loss": 0.0161,
      "step": 4110
    },
    {
      "epoch": 9.51501154734411,
      "grad_norm": 0.1584330052137375,
      "learning_rate": 1.2263852319393309e-06,
      "loss": 0.0169,
      "step": 4120
    },
    {
      "epoch": 9.53810623556582,
      "grad_norm": 0.14805783331394196,
      "learning_rate": 1.109872550244917e-06,
      "loss": 0.0149,
      "step": 4130
    },
    {
      "epoch": 9.56120092378753,
      "grad_norm": 0.228169247508049,
      "learning_rate": 9.991433546424512e-07,
      "loss": 0.0203,
      "step": 4140
    },
    {
      "epoch": 9.584295612009237,
      "grad_norm": 0.19104620814323425,
      "learning_rate": 8.942041210135755e-07,
      "loss": 0.0164,
      "step": 4150
    },
    {
      "epoch": 9.607390300230946,
      "grad_norm": 0.163773313164711,
      "learning_rate": 7.950609866200665e-07,
      "loss": 0.0165,
      "step": 4160
    },
    {
      "epoch": 9.630484988452656,
      "grad_norm": 0.1500762552022934,
      "learning_rate": 7.017197497448557e-07,
      "loss": 0.0174,
      "step": 4170
    },
    {
      "epoch": 9.653579676674365,
      "grad_norm": 0.2414599508047104,
      "learning_rate": 6.141858693529457e-07,
      "loss": 0.0155,
      "step": 4180
    },
    {
      "epoch": 9.676674364896074,
      "grad_norm": 0.1968170553445816,
      "learning_rate": 5.324644647721755e-07,
      "loss": 0.013,
      "step": 4190
    },
    {
      "epoch": 9.699769053117784,
      "grad_norm": 0.18126633763313293,
      "learning_rate": 4.565603153937281e-07,
      "loss": 0.0167,
      "step": 4200
    },
    {
      "epoch": 9.722863741339491,
      "grad_norm": 0.18440669775009155,
      "learning_rate": 3.8647786039273103e-07,
      "loss": 0.0145,
      "step": 4210
    },
    {
      "epoch": 9.7459584295612,
      "grad_norm": 0.2203577756881714,
      "learning_rate": 3.222211984685641e-07,
      "loss": 0.0148,
      "step": 4220
    },
    {
      "epoch": 9.76905311778291,
      "grad_norm": 0.15016986429691315,
      "learning_rate": 2.6379408760516254e-07,
      "loss": 0.0148,
      "step": 4230
    },
    {
      "epoch": 9.79214780600462,
      "grad_norm": 0.23910097777843475,
      "learning_rate": 2.1119994485123695e-07,
      "loss": 0.0146,
      "step": 4240
    },
    {
      "epoch": 9.815242494226329,
      "grad_norm": 0.3367011249065399,
      "learning_rate": 1.6444184612044444e-07,
      "loss": 0.0207,
      "step": 4250
    },
    {
      "epoch": 9.838337182448036,
      "grad_norm": 0.20982211828231812,
      "learning_rate": 1.2352252601147697e-07,
      "loss": 0.0143,
      "step": 4260
    },
    {
      "epoch": 9.861431870669746,
      "grad_norm": 0.19607298076152802,
      "learning_rate": 8.844437764815583e-08,
      "loss": 0.0143,
      "step": 4270
    },
    {
      "epoch": 9.884526558891455,
      "grad_norm": 0.21889039874076843,
      "learning_rate": 5.9209452539432664e-08,
      "loss": 0.017,
      "step": 4280
    },
    {
      "epoch": 9.907621247113164,
      "grad_norm": 0.21648484468460083,
      "learning_rate": 3.581946045947415e-08,
      "loss": 0.0163,
      "step": 4290
    },
    {
      "epoch": 9.930715935334874,
      "grad_norm": 0.343557745218277,
      "learning_rate": 1.8275769347575467e-08,
      "loss": 0.0156,
      "step": 4300
    },
    {
      "epoch": 9.953810623556581,
      "grad_norm": 0.18986669182777405,
      "learning_rate": 6.579405228257507e-09,
      "loss": 0.0181,
      "step": 4310
    },
    {
      "epoch": 9.97690531177829,
      "grad_norm": 0.17259903252124786,
      "learning_rate": 7.310521511705304e-10,
      "loss": 0.0163,
      "step": 4320
    },
    {
      "epoch": 9.988452655889146,
      "step": 4325,
      "total_flos": 5.8174995935902e+17,
      "train_loss": 0.036179359453606466,
      "train_runtime": 4740.7562,
      "train_samples_per_second": 58.387,
      "train_steps_per_second": 0.912
    }
  ],
  "logging_steps": 10,
  "max_steps": 4325,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.8174995935902e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}