{
  "best_metric": 1.5991134643554688,
  "best_model_checkpoint": "tiny_bert_bc_km_5_v1/checkpoint-470000",
  "epoch": 25.0,
  "eval_steps": 10000,
  "global_step": 593075,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.021076592336551025,
      "grad_norm": 2.3582303524017334,
      "learning_rate": 5e-06,
      "loss": 10.4379,
      "step": 500
    },
    {
      "epoch": 0.04215318467310205,
      "grad_norm": 1.0591377019882202,
      "learning_rate": 1e-05,
      "loss": 7.9051,
      "step": 1000
    },
    {
      "epoch": 0.06322977700965308,
      "grad_norm": 0.9644157290458679,
      "learning_rate": 1.5e-05,
      "loss": 6.6902,
      "step": 1500
    },
    {
      "epoch": 0.0843063693462041,
      "grad_norm": 2.34735369682312,
      "learning_rate": 2e-05,
      "loss": 6.4553,
      "step": 2000
    },
    {
      "epoch": 0.10538296168275513,
      "grad_norm": 2.04646897315979,
      "learning_rate": 2.5e-05,
      "loss": 6.3169,
      "step": 2500
    },
    {
      "epoch": 0.12645955401930617,
      "grad_norm": 8.083847999572754,
      "learning_rate": 3e-05,
      "loss": 6.234,
      "step": 3000
    },
    {
      "epoch": 0.14753614635585718,
      "grad_norm": 2.9417366981506348,
      "learning_rate": 3.5e-05,
      "loss": 6.1691,
      "step": 3500
    },
    {
      "epoch": 0.1686127386924082,
      "grad_norm": 3.839484930038452,
      "learning_rate": 4e-05,
      "loss": 6.1198,
      "step": 4000
    },
    {
      "epoch": 0.18968933102895924,
      "grad_norm": 1.40390944480896,
      "learning_rate": 4.5e-05,
      "loss": 6.0747,
      "step": 4500
    },
    {
      "epoch": 0.21076592336551025,
      "grad_norm": 2.5994393825531006,
      "learning_rate": 5e-05,
      "loss": 6.0449,
      "step": 5000
    },
    {
      "epoch": 0.2318425157020613,
      "grad_norm": 4.612061977386475,
      "learning_rate": 5.500000000000001e-05,
      "loss": 6.0252,
      "step": 5500
    },
    {
      "epoch": 0.25291910803861234,
      "grad_norm": 1.329725980758667,
      "learning_rate": 6e-05,
      "loss": 5.9901,
      "step": 6000
    },
    {
      "epoch": 0.27399570037516335,
      "grad_norm": 1.8742631673812866,
      "learning_rate": 6.500000000000001e-05,
      "loss": 5.9729,
      "step": 6500
    },
    {
      "epoch": 0.29507229271171437,
      "grad_norm": 2.0280282497406006,
      "learning_rate": 7e-05,
      "loss": 5.9519,
      "step": 7000
    },
    {
      "epoch": 0.3161488850482654,
      "grad_norm": 3.695600748062134,
      "learning_rate": 7.500000000000001e-05,
      "loss": 5.9317,
      "step": 7500
    },
    {
      "epoch": 0.3372254773848164,
      "grad_norm": 1.662867784500122,
      "learning_rate": 8e-05,
      "loss": 5.9164,
      "step": 8000
    },
    {
      "epoch": 0.35830206972136747,
      "grad_norm": 1.9238191843032837,
      "learning_rate": 8.5e-05,
      "loss": 5.9004,
      "step": 8500
    },
    {
      "epoch": 0.3793786620579185,
      "grad_norm": 0.8860997557640076,
      "learning_rate": 9e-05,
      "loss": 5.8813,
      "step": 9000
    },
    {
      "epoch": 0.4004552543944695,
      "grad_norm": 2.660971164703369,
      "learning_rate": 9.5e-05,
      "loss": 5.8814,
      "step": 9500
    },
    {
      "epoch": 0.4215318467310205,
      "grad_norm": 2.1609928607940674,
      "learning_rate": 0.0001,
      "loss": 5.868,
      "step": 10000
    },
    {
      "epoch": 0.4215318467310205,
      "eval_accuracy": 0.15871994025282504,
      "eval_loss": 5.639025688171387,
      "eval_runtime": 169.4949,
      "eval_samples_per_second": 712.151,
      "eval_steps_per_second": 7.422,
      "step": 10000
    },
    {
      "epoch": 0.4426084390675716,
      "grad_norm": 2.7797799110412598,
      "learning_rate": 9.99142477382841e-05,
      "loss": 5.8534,
      "step": 10500
    },
    {
      "epoch": 0.4636850314041226,
      "grad_norm": 1.957123875617981,
      "learning_rate": 9.98284954765682e-05,
      "loss": 5.8455,
      "step": 11000
    },
    {
      "epoch": 0.4847616237406736,
      "grad_norm": 0.8883283734321594,
      "learning_rate": 9.97427432148523e-05,
      "loss": 5.8465,
      "step": 11500
    },
    {
      "epoch": 0.5058382160772247,
      "grad_norm": 1.1426115036010742,
      "learning_rate": 9.96569909531364e-05,
      "loss": 5.8336,
      "step": 12000
    },
    {
      "epoch": 0.5269148084137757,
      "grad_norm": 3.301488161087036,
      "learning_rate": 9.957123869142048e-05,
      "loss": 5.8245,
      "step": 12500
    },
    {
      "epoch": 0.5479914007503267,
      "grad_norm": 1.8311688899993896,
      "learning_rate": 9.94854864297046e-05,
      "loss": 5.8215,
      "step": 13000
    },
    {
      "epoch": 0.5690679930868777,
      "grad_norm": 2.6868529319763184,
      "learning_rate": 9.939973416798869e-05,
      "loss": 5.8169,
      "step": 13500
    },
    {
      "epoch": 0.5901445854234287,
      "grad_norm": 2.140174388885498,
      "learning_rate": 9.931398190627278e-05,
      "loss": 5.8109,
      "step": 14000
    },
    {
      "epoch": 0.6112211777599798,
      "grad_norm": 1.331060767173767,
      "learning_rate": 9.922822964455688e-05,
      "loss": 5.8028,
      "step": 14500
    },
    {
      "epoch": 0.6322977700965308,
      "grad_norm": 2.5419280529022217,
      "learning_rate": 9.914247738284099e-05,
      "loss": 5.8027,
      "step": 15000
    },
    {
      "epoch": 0.6533743624330818,
      "grad_norm": 1.681976079940796,
      "learning_rate": 9.905672512112507e-05,
      "loss": 5.7893,
      "step": 15500
    },
    {
      "epoch": 0.6744509547696328,
      "grad_norm": 0.9209150075912476,
      "learning_rate": 9.897097285940917e-05,
      "loss": 5.7874,
      "step": 16000
    },
    {
      "epoch": 0.6955275471061839,
      "grad_norm": 2.21614146232605,
      "learning_rate": 9.888522059769327e-05,
      "loss": 5.7822,
      "step": 16500
    },
    {
      "epoch": 0.7166041394427349,
      "grad_norm": 0.9894614815711975,
      "learning_rate": 9.879946833597737e-05,
      "loss": 5.7756,
      "step": 17000
    },
    {
      "epoch": 0.737680731779286,
      "grad_norm": 1.2416489124298096,
      "learning_rate": 9.871371607426147e-05,
      "loss": 5.7734,
      "step": 17500
    },
    {
      "epoch": 0.758757324115837,
      "grad_norm": 1.1923880577087402,
      "learning_rate": 9.862796381254556e-05,
      "loss": 5.7785,
      "step": 18000
    },
    {
      "epoch": 0.779833916452388,
      "grad_norm": 1.710436463356018,
      "learning_rate": 9.854221155082965e-05,
      "loss": 5.7681,
      "step": 18500
    },
    {
      "epoch": 0.800910508788939,
      "grad_norm": 1.2766753435134888,
      "learning_rate": 9.845645928911376e-05,
      "loss": 5.7673,
      "step": 19000
    },
    {
      "epoch": 0.82198710112549,
      "grad_norm": 2.2700886726379395,
      "learning_rate": 9.837070702739786e-05,
      "loss": 5.7657,
      "step": 19500
    },
    {
      "epoch": 0.843063693462041,
      "grad_norm": 1.2782379388809204,
      "learning_rate": 9.828495476568194e-05,
      "loss": 5.757,
      "step": 20000
    },
    {
      "epoch": 0.843063693462041,
      "eval_accuracy": 0.1631331267760251,
      "eval_loss": 5.5610857009887695,
      "eval_runtime": 169.7422,
      "eval_samples_per_second": 711.114,
      "eval_steps_per_second": 7.411,
      "step": 20000
    },
    {
      "epoch": 0.864140285798592,
      "grad_norm": 1.768823504447937,
      "learning_rate": 9.819920250396604e-05,
      "loss": 5.7581,
      "step": 20500
    },
    {
      "epoch": 0.8852168781351432,
      "grad_norm": 1.6019325256347656,
      "learning_rate": 9.811345024225015e-05,
      "loss": 5.7523,
      "step": 21000
    },
    {
      "epoch": 0.9062934704716942,
      "grad_norm": 0.9834634065628052,
      "learning_rate": 9.802769798053424e-05,
      "loss": 5.7513,
      "step": 21500
    },
    {
      "epoch": 0.9273700628082452,
      "grad_norm": 0.9205341935157776,
      "learning_rate": 9.794194571881834e-05,
      "loss": 5.7417,
      "step": 22000
    },
    {
      "epoch": 0.9484466551447962,
      "grad_norm": 1.746811032295227,
      "learning_rate": 9.785619345710244e-05,
      "loss": 5.7446,
      "step": 22500
    },
    {
      "epoch": 0.9695232474813472,
      "grad_norm": 1.1801968812942505,
      "learning_rate": 9.777044119538653e-05,
      "loss": 5.7398,
      "step": 23000
    },
    {
      "epoch": 0.9905998398178982,
      "grad_norm": 1.2588390111923218,
      "learning_rate": 9.768468893367063e-05,
      "loss": 5.7345,
      "step": 23500
    },
    {
      "epoch": 1.0116764321544494,
      "grad_norm": 1.574872612953186,
      "learning_rate": 9.759893667195473e-05,
      "loss": 5.7295,
      "step": 24000
    },
    {
      "epoch": 1.0327530244910004,
      "grad_norm": 1.6482529640197754,
      "learning_rate": 9.751318441023883e-05,
      "loss": 5.7228,
      "step": 24500
    },
    {
      "epoch": 1.0538296168275514,
      "grad_norm": 0.8604603409767151,
      "learning_rate": 9.742743214852293e-05,
      "loss": 5.7247,
      "step": 25000
    },
    {
      "epoch": 1.0749062091641024,
      "grad_norm": 0.7659223675727844,
      "learning_rate": 9.734167988680703e-05,
      "loss": 5.7147,
      "step": 25500
    },
    {
      "epoch": 1.0959828015006534,
      "grad_norm": 1.4631717205047607,
      "learning_rate": 9.725592762509111e-05,
      "loss": 5.7129,
      "step": 26000
    },
    {
      "epoch": 1.1170593938372044,
      "grad_norm": 0.8461015820503235,
      "learning_rate": 9.717017536337522e-05,
      "loss": 5.7087,
      "step": 26500
    },
    {
      "epoch": 1.1381359861737554,
      "grad_norm": 1.1580955982208252,
      "learning_rate": 9.708442310165932e-05,
      "loss": 5.7141,
      "step": 27000
    },
    {
      "epoch": 1.1592125785103065,
      "grad_norm": 0.7580386996269226,
      "learning_rate": 9.69986708399434e-05,
      "loss": 5.7061,
      "step": 27500
    },
    {
      "epoch": 1.1802891708468575,
      "grad_norm": 0.8994073271751404,
      "learning_rate": 9.69129185782275e-05,
      "loss": 5.7085,
      "step": 28000
    },
    {
      "epoch": 1.2013657631834085,
      "grad_norm": 2.149761915206909,
      "learning_rate": 9.68271663165116e-05,
      "loss": 5.7018,
      "step": 28500
    },
    {
      "epoch": 1.2224423555199595,
      "grad_norm": 0.7585108280181885,
      "learning_rate": 9.67414140547957e-05,
      "loss": 5.697,
      "step": 29000
    },
    {
      "epoch": 1.2435189478565105,
      "grad_norm": 2.1072428226470947,
      "learning_rate": 9.66556617930798e-05,
      "loss": 5.6696,
      "step": 29500
    },
    {
      "epoch": 1.2645955401930615,
      "grad_norm": 1.1866896152496338,
      "learning_rate": 9.656990953136388e-05,
      "loss": 5.6406,
      "step": 30000
    },
    {
      "epoch": 1.2645955401930615,
      "eval_accuracy": 0.17822696652705597,
      "eval_loss": 5.417064189910889,
      "eval_runtime": 169.0622,
      "eval_samples_per_second": 713.974,
      "eval_steps_per_second": 7.441,
      "step": 30000
    },
    {
      "epoch": 1.2856721325296125,
      "grad_norm": 1.5498770475387573,
      "learning_rate": 9.6484157269648e-05,
      "loss": 5.5947,
      "step": 30500
    },
    {
      "epoch": 1.3067487248661636,
      "grad_norm": 2.014603614807129,
      "learning_rate": 9.639840500793209e-05,
      "loss": 5.5589,
      "step": 31000
    },
    {
      "epoch": 1.3278253172027146,
      "grad_norm": 1.8741014003753662,
      "learning_rate": 9.631265274621618e-05,
      "loss": 5.5052,
      "step": 31500
    },
    {
      "epoch": 1.3489019095392658,
      "grad_norm": 2.2822089195251465,
      "learning_rate": 9.622690048450028e-05,
      "loss": 5.4306,
      "step": 32000
    },
    {
      "epoch": 1.3699785018758166,
      "grad_norm": 2.2035715579986572,
      "learning_rate": 9.614114822278439e-05,
      "loss": 5.117,
      "step": 32500
    },
    {
      "epoch": 1.3910550942123678,
      "grad_norm": 2.3333983421325684,
      "learning_rate": 9.605539596106847e-05,
      "loss": 4.7643,
      "step": 33000
    },
    {
      "epoch": 1.4121316865489189,
      "grad_norm": 1.9687154293060303,
      "learning_rate": 9.596964369935257e-05,
      "loss": 4.523,
      "step": 33500
    },
    {
      "epoch": 1.4332082788854699,
      "grad_norm": 2.046740770339966,
      "learning_rate": 9.588389143763667e-05,
      "loss": 4.3688,
      "step": 34000
    },
    {
      "epoch": 1.4542848712220209,
      "grad_norm": 3.8666162490844727,
      "learning_rate": 9.579813917592077e-05,
      "loss": 4.2356,
      "step": 34500
    },
    {
      "epoch": 1.475361463558572,
      "grad_norm": 1.9613947868347168,
      "learning_rate": 9.571238691420487e-05,
      "loss": 4.0601,
      "step": 35000
    },
    {
      "epoch": 1.496438055895123,
      "grad_norm": 2.2484829425811768,
      "learning_rate": 9.562663465248896e-05,
      "loss": 3.8998,
      "step": 35500
    },
    {
      "epoch": 1.517514648231674,
      "grad_norm": 2.5443923473358154,
      "learning_rate": 9.554088239077306e-05,
      "loss": 3.7879,
      "step": 36000
    },
    {
      "epoch": 1.538591240568225,
      "grad_norm": 1.9094462394714355,
      "learning_rate": 9.545513012905716e-05,
      "loss": 3.7018,
      "step": 36500
    },
    {
      "epoch": 1.559667832904776,
      "grad_norm": 3.810319423675537,
      "learning_rate": 9.536937786734126e-05,
      "loss": 3.6351,
      "step": 37000
    },
    {
      "epoch": 1.580744425241327,
      "grad_norm": 2.441361427307129,
      "learning_rate": 9.528362560562534e-05,
      "loss": 3.5593,
      "step": 37500
    },
    {
      "epoch": 1.601821017577878,
      "grad_norm": 2.561350107192993,
      "learning_rate": 9.519787334390946e-05,
      "loss": 3.5058,
      "step": 38000
    },
    {
      "epoch": 1.622897609914429,
      "grad_norm": 2.576829195022583,
      "learning_rate": 9.511212108219355e-05,
      "loss": 3.4526,
      "step": 38500
    },
    {
      "epoch": 1.64397420225098,
      "grad_norm": 1.6511363983154297,
      "learning_rate": 9.502636882047764e-05,
      "loss": 3.4064,
      "step": 39000
    },
    {
      "epoch": 1.6650507945875312,
      "grad_norm": 1.6814790964126587,
      "learning_rate": 9.494061655876174e-05,
      "loss": 3.3668,
      "step": 39500
    },
    {
      "epoch": 1.686127386924082,
      "grad_norm": 1.68977689743042,
      "learning_rate": 9.485486429704585e-05,
      "loss": 3.3457,
      "step": 40000
    },
    {
      "epoch": 1.686127386924082,
      "eval_accuracy": 0.4865069437056656,
      "eval_loss": 2.9073569774627686,
      "eval_runtime": 169.1211,
      "eval_samples_per_second": 713.725,
      "eval_steps_per_second": 7.438,
      "step": 40000
    },
    {
      "epoch": 1.7072039792606333,
      "grad_norm": 2.217505693435669,
      "learning_rate": 9.476911203532993e-05,
      "loss": 3.3017,
      "step": 40500
    },
    {
      "epoch": 1.728280571597184,
      "grad_norm": 2.724113702774048,
      "learning_rate": 9.468335977361403e-05,
      "loss": 3.2688,
      "step": 41000
    },
    {
      "epoch": 1.7493571639337353,
      "grad_norm": 2.0390663146972656,
      "learning_rate": 9.459760751189813e-05,
      "loss": 3.2469,
      "step": 41500
    },
    {
      "epoch": 1.770433756270286,
      "grad_norm": 1.5367827415466309,
      "learning_rate": 9.451185525018223e-05,
      "loss": 3.2149,
      "step": 42000
    },
    {
      "epoch": 1.7915103486068373,
      "grad_norm": 1.9959536790847778,
      "learning_rate": 9.442610298846633e-05,
      "loss": 3.1835,
      "step": 42500
    },
    {
      "epoch": 1.8125869409433881,
      "grad_norm": 1.3324756622314453,
      "learning_rate": 9.434035072675043e-05,
      "loss": 3.1604,
      "step": 43000
    },
    {
      "epoch": 1.8336635332799394,
      "grad_norm": 1.993525743484497,
      "learning_rate": 9.425459846503451e-05,
      "loss": 3.1297,
      "step": 43500
    },
    {
      "epoch": 1.8547401256164904,
      "grad_norm": 1.4063893556594849,
      "learning_rate": 9.416884620331862e-05,
      "loss": 3.1065,
      "step": 44000
    },
    {
      "epoch": 1.8758167179530414,
      "grad_norm": 2.0264713764190674,
      "learning_rate": 9.408309394160272e-05,
      "loss": 3.0875,
      "step": 44500
    },
    {
      "epoch": 1.8968933102895924,
      "grad_norm": 1.5733736753463745,
      "learning_rate": 9.39973416798868e-05,
      "loss": 3.0562,
      "step": 45000
    },
    {
      "epoch": 1.9179699026261434,
      "grad_norm": 2.629793643951416,
      "learning_rate": 9.39115894181709e-05,
      "loss": 3.0379,
      "step": 45500
    },
    {
      "epoch": 1.9390464949626944,
      "grad_norm": 2.5393755435943604,
      "learning_rate": 9.382583715645502e-05,
      "loss": 3.0085,
      "step": 46000
    },
    {
      "epoch": 1.9601230872992454,
      "grad_norm": 2.2285664081573486,
      "learning_rate": 9.37400848947391e-05,
      "loss": 2.9838,
      "step": 46500
    },
    {
      "epoch": 1.9811996796357965,
      "grad_norm": 2.466031074523926,
      "learning_rate": 9.36543326330232e-05,
      "loss": 2.9719,
      "step": 47000
    },
    {
      "epoch": 2.0022762719723475,
      "grad_norm": 1.1686755418777466,
      "learning_rate": 9.35685803713073e-05,
      "loss": 2.9573,
      "step": 47500
    },
    {
      "epoch": 2.0233528643088987,
      "grad_norm": 1.3075785636901855,
      "learning_rate": 9.34828281095914e-05,
      "loss": 2.921,
      "step": 48000
    },
    {
      "epoch": 2.0444294566454495,
      "grad_norm": 2.009345054626465,
      "learning_rate": 9.339707584787549e-05,
      "loss": 2.9056,
      "step": 48500
    },
    {
      "epoch": 2.0655060489820007,
      "grad_norm": 1.7309980392456055,
      "learning_rate": 9.331132358615959e-05,
      "loss": 2.8792,
      "step": 49000
    },
    {
      "epoch": 2.0865826413185515,
      "grad_norm": 1.2748831510543823,
      "learning_rate": 9.322557132444369e-05,
      "loss": 2.8688,
      "step": 49500
    },
    {
      "epoch": 2.1076592336551028,
      "grad_norm": 1.445093035697937,
      "learning_rate": 9.313981906272779e-05,
      "loss": 2.8544,
      "step": 50000
    },
    {
      "epoch": 2.1076592336551028,
      "eval_accuracy": 0.5422852943848815,
      "eval_loss": 2.4902899265289307,
      "eval_runtime": 168.925,
      "eval_samples_per_second": 714.554,
      "eval_steps_per_second": 7.447,
      "step": 50000
    },
    {
      "epoch": 2.1287358259916536,
      "grad_norm": 1.4034790992736816,
      "learning_rate": 9.305406680101189e-05,
      "loss": 2.84,
      "step": 50500
    },
    {
      "epoch": 2.149812418328205,
      "grad_norm": 1.3959338665008545,
      "learning_rate": 9.296831453929597e-05,
      "loss": 2.8234,
      "step": 51000
    },
    {
      "epoch": 2.1708890106647556,
      "grad_norm": 1.3900226354599,
      "learning_rate": 9.288256227758008e-05,
      "loss": 2.8036,
      "step": 51500
    },
    {
      "epoch": 2.191965603001307,
      "grad_norm": 1.6682796478271484,
      "learning_rate": 9.279681001586418e-05,
      "loss": 2.7876,
      "step": 52000
    },
    {
      "epoch": 2.2130421953378576,
      "grad_norm": 1.4198817014694214,
      "learning_rate": 9.271105775414827e-05,
      "loss": 2.7718,
      "step": 52500
    },
    {
      "epoch": 2.234118787674409,
      "grad_norm": 1.4925236701965332,
      "learning_rate": 9.262530549243236e-05,
      "loss": 2.7609,
      "step": 53000
    },
    {
      "epoch": 2.2551953800109596,
      "grad_norm": 1.4061663150787354,
      "learning_rate": 9.253955323071646e-05,
      "loss": 2.7468,
      "step": 53500
    },
    {
      "epoch": 2.276271972347511,
      "grad_norm": 1.5086860656738281,
      "learning_rate": 9.245380096900056e-05,
      "loss": 2.7358,
      "step": 54000
    },
    {
      "epoch": 2.2973485646840617,
      "grad_norm": 1.9930882453918457,
      "learning_rate": 9.236804870728466e-05,
      "loss": 2.7157,
      "step": 54500
    },
    {
      "epoch": 2.318425157020613,
      "grad_norm": 1.3227754831314087,
      "learning_rate": 9.228229644556876e-05,
      "loss": 2.7025,
      "step": 55000
    },
    {
      "epoch": 2.339501749357164,
      "grad_norm": 1.5141241550445557,
      "learning_rate": 9.219654418385286e-05,
      "loss": 2.6842,
      "step": 55500
    },
    {
      "epoch": 2.360578341693715,
      "grad_norm": 1.3763973712921143,
      "learning_rate": 9.211079192213695e-05,
      "loss": 2.6821,
      "step": 56000
    },
    {
      "epoch": 2.381654934030266,
      "grad_norm": 1.6842782497406006,
      "learning_rate": 9.202503966042105e-05,
      "loss": 2.6626,
      "step": 56500
    },
    {
      "epoch": 2.402731526366817,
      "grad_norm": 1.6818408966064453,
      "learning_rate": 9.193928739870514e-05,
      "loss": 2.6494,
      "step": 57000
    },
    {
      "epoch": 2.423808118703368,
      "grad_norm": 1.2991440296173096,
      "learning_rate": 9.185353513698925e-05,
      "loss": 2.6404,
      "step": 57500
    },
    {
      "epoch": 2.444884711039919,
      "grad_norm": 1.4616061449050903,
      "learning_rate": 9.176778287527335e-05,
      "loss": 2.6229,
      "step": 58000
    },
    {
      "epoch": 2.4659613033764702,
      "grad_norm": 1.6595426797866821,
      "learning_rate": 9.168203061355743e-05,
      "loss": 2.6148,
      "step": 58500
    },
    {
      "epoch": 2.487037895713021,
      "grad_norm": 1.2576252222061157,
      "learning_rate": 9.159627835184153e-05,
      "loss": 2.6054,
      "step": 59000
    },
    {
      "epoch": 2.5081144880495723,
      "grad_norm": 2.556204319000244,
      "learning_rate": 9.151052609012564e-05,
      "loss": 2.5985,
      "step": 59500
    },
    {
      "epoch": 2.529191080386123,
      "grad_norm": 1.1448605060577393,
      "learning_rate": 9.142477382840973e-05,
      "loss": 2.5821,
      "step": 60000
    },
    {
      "epoch": 2.529191080386123,
      "eval_accuracy": 0.5786292544191414,
      "eval_loss": 2.2187814712524414,
      "eval_runtime": 169.0954,
      "eval_samples_per_second": 713.834,
      "eval_steps_per_second": 7.44,
      "step": 60000
    },
    {
      "epoch": 2.5502676727226743,
      "grad_norm": 1.6365042924880981,
      "learning_rate": 9.133902156669383e-05,
      "loss": 2.5719,
      "step": 60500
    },
    {
      "epoch": 2.571344265059225,
      "grad_norm": 1.4883947372436523,
      "learning_rate": 9.125326930497792e-05,
      "loss": 2.5663,
      "step": 61000
    },
    {
      "epoch": 2.5924208573957763,
      "grad_norm": 2.672994613647461,
      "learning_rate": 9.116751704326202e-05,
      "loss": 2.5596,
      "step": 61500
    },
    {
      "epoch": 2.613497449732327,
      "grad_norm": 1.1591687202453613,
      "learning_rate": 9.108176478154612e-05,
      "loss": 2.5488,
      "step": 62000
    },
    {
      "epoch": 2.6345740420688784,
      "grad_norm": 1.4181878566741943,
      "learning_rate": 9.09960125198302e-05,
      "loss": 2.5466,
      "step": 62500
    },
    {
      "epoch": 2.655650634405429,
      "grad_norm": 1.2093358039855957,
      "learning_rate": 9.091026025811432e-05,
      "loss": 2.5297,
      "step": 63000
    },
    {
      "epoch": 2.6767272267419804,
      "grad_norm": 2.295060634613037,
      "learning_rate": 9.082450799639842e-05,
      "loss": 2.5256,
      "step": 63500
    },
    {
      "epoch": 2.6978038190785316,
      "grad_norm": 1.3373055458068848,
      "learning_rate": 9.07387557346825e-05,
      "loss": 2.5198,
      "step": 64000
    },
    {
      "epoch": 2.7188804114150824,
      "grad_norm": 1.3538049459457397,
      "learning_rate": 9.06530034729666e-05,
      "loss": 2.5155,
      "step": 64500
    },
    {
      "epoch": 2.739957003751633,
      "grad_norm": 1.2021138668060303,
      "learning_rate": 9.05672512112507e-05,
      "loss": 2.5047,
      "step": 65000
    },
    {
      "epoch": 2.7610335960881844,
      "grad_norm": 2.336235761642456,
      "learning_rate": 9.04814989495348e-05,
      "loss": 2.4947,
      "step": 65500
    },
    {
      "epoch": 2.7821101884247357,
      "grad_norm": 1.2530579566955566,
      "learning_rate": 9.03957466878189e-05,
      "loss": 2.4944,
      "step": 66000
    },
    {
      "epoch": 2.8031867807612865,
      "grad_norm": 1.556249976158142,
      "learning_rate": 9.030999442610299e-05,
      "loss": 2.4809,
      "step": 66500
    },
    {
      "epoch": 2.8242633730978377,
      "grad_norm": 2.2047195434570312,
      "learning_rate": 9.022424216438709e-05,
      "loss": 2.4716,
      "step": 67000
    },
    {
      "epoch": 2.8453399654343885,
      "grad_norm": 2.193751096725464,
      "learning_rate": 9.013848990267119e-05,
      "loss": 2.4627,
      "step": 67500
    },
    {
      "epoch": 2.8664165577709397,
      "grad_norm": 1.9192605018615723,
      "learning_rate": 9.005273764095529e-05,
      "loss": 2.4615,
      "step": 68000
    },
    {
      "epoch": 2.8874931501074905,
      "grad_norm": 1.36215341091156,
      "learning_rate": 8.996698537923937e-05,
      "loss": 2.4595,
      "step": 68500
    },
    {
      "epoch": 2.9085697424440418,
      "grad_norm": 1.5078896284103394,
      "learning_rate": 8.988123311752348e-05,
      "loss": 2.4471,
      "step": 69000
    },
    {
      "epoch": 2.9296463347805926,
      "grad_norm": 1.5379080772399902,
      "learning_rate": 8.979548085580758e-05,
      "loss": 2.4455,
      "step": 69500
    },
    {
      "epoch": 2.950722927117144,
      "grad_norm": 1.246437668800354,
      "learning_rate": 8.970972859409167e-05,
      "loss": 2.4361,
      "step": 70000
    },
    {
      "epoch": 2.950722927117144,
      "eval_accuracy": 0.598794356373302,
      "eval_loss": 2.0898706912994385,
      "eval_runtime": 168.962,
      "eval_samples_per_second": 714.397,
      "eval_steps_per_second": 7.445,
      "step": 70000
    },
    {
      "epoch": 2.9717995194536946,
      "grad_norm": 1.6131919622421265,
      "learning_rate": 8.962397633237576e-05,
      "loss": 2.4351,
      "step": 70500
    },
    {
      "epoch": 2.992876111790246,
      "grad_norm": 1.2681609392166138,
      "learning_rate": 8.953822407065988e-05,
      "loss": 2.4251,
      "step": 71000
    },
    {
      "epoch": 3.0139527041267966,
      "grad_norm": 1.4036773443222046,
      "learning_rate": 8.945247180894396e-05,
      "loss": 2.4127,
      "step": 71500
    },
    {
      "epoch": 3.035029296463348,
      "grad_norm": 1.4352750778198242,
      "learning_rate": 8.936671954722806e-05,
      "loss": 2.408,
      "step": 72000
    },
    {
      "epoch": 3.0561058887998986,
      "grad_norm": 1.6349232196807861,
      "learning_rate": 8.928096728551216e-05,
      "loss": 2.4079,
      "step": 72500
    },
    {
      "epoch": 3.07718248113645,
      "grad_norm": 1.1338483095169067,
      "learning_rate": 8.919521502379626e-05,
      "loss": 2.4018,
      "step": 73000
    },
    {
      "epoch": 3.0982590734730007,
      "grad_norm": 1.7271078824996948,
      "learning_rate": 8.910946276208035e-05,
      "loss": 2.398,
      "step": 73500
    },
    {
      "epoch": 3.119335665809552,
      "grad_norm": 2.9701766967773438,
      "learning_rate": 8.902371050036445e-05,
      "loss": 2.3868,
      "step": 74000
    },
    {
      "epoch": 3.140412258146103,
      "grad_norm": 4.18891716003418,
      "learning_rate": 8.893795823864855e-05,
      "loss": 2.3885,
      "step": 74500
    },
    {
      "epoch": 3.161488850482654,
      "grad_norm": 1.327147364616394,
      "learning_rate": 8.885220597693265e-05,
      "loss": 2.3857,
      "step": 75000
    },
    {
      "epoch": 3.182565442819205,
      "grad_norm": 1.657575249671936,
      "learning_rate": 8.876645371521675e-05,
      "loss": 2.3764,
      "step": 75500
    },
    {
      "epoch": 3.203642035155756,
      "grad_norm": 1.7480978965759277,
      "learning_rate": 8.868070145350083e-05,
      "loss": 2.3727,
      "step": 76000
    },
    {
      "epoch": 3.224718627492307,
      "grad_norm": 1.398889422416687,
      "learning_rate": 8.859494919178493e-05,
      "loss": 2.3717,
      "step": 76500
    },
    {
      "epoch": 3.245795219828858,
      "grad_norm": 1.2654310464859009,
      "learning_rate": 8.850919693006904e-05,
      "loss": 2.3673,
      "step": 77000
    },
    {
      "epoch": 3.2668718121654092,
      "grad_norm": 1.1807788610458374,
      "learning_rate": 8.842344466835313e-05,
      "loss": 2.3596,
      "step": 77500
    },
    {
      "epoch": 3.28794840450196,
      "grad_norm": 1.7121933698654175,
      "learning_rate": 8.833769240663723e-05,
      "loss": 2.352,
      "step": 78000
    },
    {
      "epoch": 3.3090249968385113,
      "grad_norm": 1.498852014541626,
      "learning_rate": 8.825194014492132e-05,
      "loss": 2.3501,
      "step": 78500
    },
    {
      "epoch": 3.330101589175062,
      "grad_norm": 1.3500930070877075,
      "learning_rate": 8.816618788320542e-05,
      "loss": 2.3457,
      "step": 79000
    },
    {
      "epoch": 3.3511781815116133,
      "grad_norm": 1.4794244766235352,
      "learning_rate": 8.808043562148952e-05,
      "loss": 2.3435,
      "step": 79500
    },
    {
      "epoch": 3.372254773848164,
      "grad_norm": 1.3416305780410767,
      "learning_rate": 8.799468335977362e-05,
      "loss": 2.3431,
      "step": 80000
    },
    {
      "epoch": 3.372254773848164,
      "eval_accuracy": 0.6117633743534212,
      "eval_loss": 2.0070223808288574,
      "eval_runtime": 169.0401,
      "eval_samples_per_second": 714.067,
      "eval_steps_per_second": 7.442,
      "step": 80000
    },
    {
      "epoch": 3.3933313661847153,
      "grad_norm": 2.674189567565918,
      "learning_rate": 8.790893109805772e-05,
      "loss": 2.3402,
      "step": 80500
    },
    {
      "epoch": 3.414407958521266,
      "grad_norm": 1.1912676095962524,
      "learning_rate": 8.782317883634182e-05,
      "loss": 2.331,
      "step": 81000
    },
    {
      "epoch": 3.4354845508578173,
      "grad_norm": 1.1836739778518677,
      "learning_rate": 8.773742657462591e-05,
      "loss": 2.3247,
      "step": 81500
    },
    {
      "epoch": 3.456561143194368,
      "grad_norm": 1.3293665647506714,
      "learning_rate": 8.765167431291e-05,
      "loss": 2.3257,
      "step": 82000
    },
    {
      "epoch": 3.4776377355309194,
      "grad_norm": 1.2547948360443115,
      "learning_rate": 8.756592205119411e-05,
      "loss": 2.3229,
      "step": 82500
    },
    {
      "epoch": 3.4987143278674706,
      "grad_norm": 1.3380351066589355,
      "learning_rate": 8.748016978947821e-05,
      "loss": 2.3189,
      "step": 83000
    },
    {
      "epoch": 3.5197909202040214,
      "grad_norm": 2.5370569229125977,
      "learning_rate": 8.73944175277623e-05,
      "loss": 2.3189,
      "step": 83500
    },
    {
      "epoch": 3.540867512540572,
      "grad_norm": 1.4062000513076782,
      "learning_rate": 8.730866526604639e-05,
      "loss": 2.3165,
      "step": 84000
    },
    {
      "epoch": 3.5619441048771234,
      "grad_norm": 1.9580022096633911,
      "learning_rate": 8.72229130043305e-05,
      "loss": 2.3086,
      "step": 84500
    },
    {
      "epoch": 3.5830206972136747,
      "grad_norm": 1.2954455614089966,
      "learning_rate": 8.713716074261459e-05,
      "loss": 2.306,
      "step": 85000
    },
    {
      "epoch": 3.6040972895502255,
      "grad_norm": 1.2455974817276,
      "learning_rate": 8.705140848089869e-05,
      "loss": 2.3054,
      "step": 85500
    },
    {
      "epoch": 3.6251738818867767,
      "grad_norm": 1.831058144569397,
      "learning_rate": 8.696565621918278e-05,
      "loss": 2.2948,
      "step": 86000
    },
    {
      "epoch": 3.6462504742233275,
      "grad_norm": 1.303527593612671,
      "learning_rate": 8.687990395746688e-05,
      "loss": 2.2979,
      "step": 86500
    },
    {
      "epoch": 3.6673270665598787,
      "grad_norm": 1.3147029876708984,
      "learning_rate": 8.679415169575098e-05,
      "loss": 2.295,
      "step": 87000
    },
    {
      "epoch": 3.6884036588964295,
      "grad_norm": 1.1574718952178955,
      "learning_rate": 8.670839943403508e-05,
      "loss": 2.2877,
      "step": 87500
    },
    {
      "epoch": 3.7094802512329808,
      "grad_norm": 1.3106043338775635,
      "learning_rate": 8.662264717231916e-05,
      "loss": 2.2901,
      "step": 88000
    },
    {
      "epoch": 3.7305568435695315,
      "grad_norm": 1.8105119466781616,
      "learning_rate": 8.653689491060328e-05,
      "loss": 2.2889,
      "step": 88500
    },
    {
      "epoch": 3.751633435906083,
      "grad_norm": 2.1772241592407227,
      "learning_rate": 8.645114264888737e-05,
      "loss": 2.2844,
      "step": 89000
    },
    {
      "epoch": 3.7727100282426336,
      "grad_norm": 1.1817958354949951,
      "learning_rate": 8.636539038717146e-05,
      "loss": 2.277,
      "step": 89500
    },
    {
      "epoch": 3.793786620579185,
      "grad_norm": 1.483210563659668,
      "learning_rate": 8.627963812545556e-05,
      "loss": 2.2752,
      "step": 90000
    },
    {
      "epoch": 3.793786620579185,
      "eval_accuracy": 0.6216132697069852,
      "eval_loss": 1.9472320079803467,
      "eval_runtime": 169.0581,
      "eval_samples_per_second": 713.991,
      "eval_steps_per_second": 7.441,
      "step": 90000
    },
    {
      "epoch": 3.8148632129157356,
      "grad_norm": 1.486264944076538,
      "learning_rate": 8.619388586373967e-05,
      "loss": 2.277,
      "step": 90500
    },
    {
      "epoch": 3.835939805252287,
      "grad_norm": 1.154267430305481,
      "learning_rate": 8.610813360202375e-05,
      "loss": 2.2696,
      "step": 91000
    },
    {
      "epoch": 3.857016397588838,
      "grad_norm": 1.2089483737945557,
      "learning_rate": 8.602238134030785e-05,
      "loss": 2.2692,
      "step": 91500
    },
    {
      "epoch": 3.878092989925389,
      "grad_norm": 2.1056156158447266,
      "learning_rate": 8.593662907859195e-05,
      "loss": 2.2618,
      "step": 92000
    },
    {
      "epoch": 3.8991695822619397,
      "grad_norm": 1.3647282123565674,
      "learning_rate": 8.585087681687605e-05,
      "loss": 2.2633,
      "step": 92500
    },
    {
      "epoch": 3.920246174598491,
      "grad_norm": 2.7009456157684326,
      "learning_rate": 8.576512455516015e-05,
      "loss": 2.2629,
      "step": 93000
    },
    {
      "epoch": 3.941322766935042,
      "grad_norm": 1.3317773342132568,
      "learning_rate": 8.567937229344425e-05,
      "loss": 2.2541,
      "step": 93500
    },
    {
      "epoch": 3.962399359271593,
      "grad_norm": 1.2492473125457764,
      "learning_rate": 8.559362003172834e-05,
      "loss": 2.2555,
      "step": 94000
    },
    {
      "epoch": 3.983475951608144,
      "grad_norm": 1.3696634769439697,
      "learning_rate": 8.550786777001244e-05,
      "loss": 2.2493,
      "step": 94500
    },
    {
      "epoch": 4.004552543944695,
      "grad_norm": 1.778423547744751,
      "learning_rate": 8.542211550829654e-05,
      "loss": 2.2457,
      "step": 95000
    },
    {
      "epoch": 4.025629136281246,
      "grad_norm": 1.7177969217300415,
      "learning_rate": 8.533636324658063e-05,
      "loss": 2.2291,
      "step": 95500
    },
    {
      "epoch": 4.046705728617797,
      "grad_norm": 2.0601108074188232,
      "learning_rate": 8.525061098486474e-05,
      "loss": 2.2344,
      "step": 96000
    },
    {
      "epoch": 4.067782320954348,
      "grad_norm": 1.3533416986465454,
      "learning_rate": 8.516485872314884e-05,
      "loss": 2.2239,
      "step": 96500
    },
    {
      "epoch": 4.088858913290899,
      "grad_norm": 1.148821473121643,
      "learning_rate": 8.507910646143292e-05,
      "loss": 2.2292,
      "step": 97000
    },
    {
      "epoch": 4.10993550562745,
      "grad_norm": 1.543866515159607,
      "learning_rate": 8.499335419971702e-05,
      "loss": 2.2247,
      "step": 97500
    },
    {
      "epoch": 4.1310120979640015,
      "grad_norm": 1.3425670862197876,
      "learning_rate": 8.490760193800112e-05,
      "loss": 2.2238,
      "step": 98000
    },
    {
      "epoch": 4.152088690300552,
      "grad_norm": 1.1997469663619995,
      "learning_rate": 8.482184967628522e-05,
      "loss": 2.2192,
      "step": 98500
    },
    {
      "epoch": 4.173165282637103,
      "grad_norm": 1.4936726093292236,
      "learning_rate": 8.473609741456931e-05,
      "loss": 2.2258,
      "step": 99000
    },
    {
      "epoch": 4.194241874973654,
      "grad_norm": 2.064547061920166,
      "learning_rate": 8.46503451528534e-05,
      "loss": 2.2237,
      "step": 99500
    },
    {
      "epoch": 4.2153184673102055,
      "grad_norm": 1.4626243114471436,
      "learning_rate": 8.456459289113751e-05,
      "loss": 2.2181,
      "step": 100000
    },
    {
      "epoch": 4.2153184673102055,
      "eval_accuracy": 0.6284274242345342,
      "eval_loss": 1.9011849164962769,
      "eval_runtime": 168.4235,
      "eval_samples_per_second": 716.681,
      "eval_steps_per_second": 7.469,
      "step": 100000
    },
    {
      "epoch": 4.236395059646756,
      "grad_norm": 1.18771231174469,
      "learning_rate": 8.447884062942161e-05,
      "loss": 2.2195,
      "step": 100500
    },
    {
      "epoch": 4.257471651983307,
      "grad_norm": 1.3363932371139526,
      "learning_rate": 8.43930883677057e-05,
      "loss": 2.2175,
      "step": 101000
    },
    {
      "epoch": 4.278548244319858,
      "grad_norm": 2.03973126411438,
      "learning_rate": 8.430733610598979e-05,
      "loss": 2.2099,
      "step": 101500
    },
    {
      "epoch": 4.29962483665641,
      "grad_norm": 1.3062665462493896,
      "learning_rate": 8.42215838442739e-05,
      "loss": 2.2055,
      "step": 102000
    },
    {
      "epoch": 4.32070142899296,
      "grad_norm": 1.1833539009094238,
      "learning_rate": 8.413583158255799e-05,
      "loss": 2.2103,
      "step": 102500
    },
    {
      "epoch": 4.341778021329511,
      "grad_norm": 1.0844303369522095,
      "learning_rate": 8.405007932084209e-05,
      "loss": 2.206,
      "step": 103000
    },
    {
      "epoch": 4.362854613666062,
      "grad_norm": 1.2328764200210571,
      "learning_rate": 8.396432705912619e-05,
      "loss": 2.2062,
      "step": 103500
    },
    {
      "epoch": 4.383931206002614,
      "grad_norm": 1.117450475692749,
      "learning_rate": 8.387857479741028e-05,
      "loss": 2.2068,
      "step": 104000
    },
    {
      "epoch": 4.405007798339165,
      "grad_norm": 2.347036838531494,
      "learning_rate": 8.379282253569438e-05,
      "loss": 2.2029,
      "step": 104500
    },
    {
      "epoch": 4.426084390675715,
      "grad_norm": 1.1019892692565918,
      "learning_rate": 8.370707027397848e-05,
      "loss": 2.203,
      "step": 105000
    },
    {
      "epoch": 4.4471609830122665,
      "grad_norm": 1.4269235134124756,
      "learning_rate": 8.362131801226258e-05,
      "loss": 2.1947,
      "step": 105500
    },
    {
      "epoch": 4.468237575348818,
      "grad_norm": 2.0702996253967285,
      "learning_rate": 8.353556575054668e-05,
      "loss": 2.1938,
      "step": 106000
    },
    {
      "epoch": 4.489314167685369,
      "grad_norm": 1.3875104188919067,
      "learning_rate": 8.344981348883077e-05,
      "loss": 2.1948,
      "step": 106500
    },
    {
      "epoch": 4.510390760021919,
      "grad_norm": 1.2226201295852661,
      "learning_rate": 8.336406122711486e-05,
      "loss": 2.1901,
      "step": 107000
    },
    {
      "epoch": 4.5314673523584705,
      "grad_norm": 1.2251983880996704,
      "learning_rate": 8.327830896539897e-05,
      "loss": 2.1919,
      "step": 107500
    },
    {
      "epoch": 4.552543944695022,
      "grad_norm": 1.2576063871383667,
      "learning_rate": 8.319255670368307e-05,
      "loss": 2.1898,
      "step": 108000
    },
    {
      "epoch": 4.573620537031573,
      "grad_norm": 2.3376500606536865,
      "learning_rate": 8.310680444196715e-05,
      "loss": 2.19,
      "step": 108500
    },
    {
      "epoch": 4.594697129368123,
      "grad_norm": 1.2776663303375244,
      "learning_rate": 8.302105218025125e-05,
      "loss": 2.1802,
      "step": 109000
    },
    {
      "epoch": 4.615773721704675,
      "grad_norm": 1.373831033706665,
      "learning_rate": 8.293529991853536e-05,
      "loss": 2.1849,
      "step": 109500
    },
    {
      "epoch": 4.636850314041226,
      "grad_norm": 1.317340612411499,
      "learning_rate": 8.284954765681945e-05,
      "loss": 2.1874,
      "step": 110000
    },
    {
      "epoch": 4.636850314041226,
      "eval_accuracy": 0.6334282353633865,
      "eval_loss": 1.8619428873062134,
      "eval_runtime": 169.325,
      "eval_samples_per_second": 712.866,
      "eval_steps_per_second": 7.429,
      "step": 110000
    },
    {
      "epoch": 4.657926906377777,
      "grad_norm": 2.7527825832366943,
      "learning_rate": 8.276379539510355e-05,
      "loss": 2.1807,
      "step": 110500
    },
    {
      "epoch": 4.679003498714328,
      "grad_norm": 1.2492730617523193,
      "learning_rate": 8.267804313338765e-05,
      "loss": 2.1816,
      "step": 111000
    },
    {
      "epoch": 4.700080091050879,
      "grad_norm": 2.0063440799713135,
      "learning_rate": 8.259229087167174e-05,
      "loss": 2.1767,
      "step": 111500
    },
    {
      "epoch": 4.72115668338743,
      "grad_norm": 1.6646426916122437,
      "learning_rate": 8.250653860995584e-05,
      "loss": 2.1737,
      "step": 112000
    },
    {
      "epoch": 4.742233275723981,
      "grad_norm": 1.195522665977478,
      "learning_rate": 8.242078634823994e-05,
      "loss": 2.1764,
      "step": 112500
    },
    {
      "epoch": 4.763309868060532,
      "grad_norm": 1.4563145637512207,
      "learning_rate": 8.233503408652403e-05,
      "loss": 2.1738,
      "step": 113000
    },
    {
      "epoch": 4.784386460397083,
      "grad_norm": 1.2892214059829712,
      "learning_rate": 8.224928182480814e-05,
      "loss": 2.1735,
      "step": 113500
    },
    {
      "epoch": 4.805463052733634,
      "grad_norm": 1.590841293334961,
      "learning_rate": 8.216352956309224e-05,
      "loss": 2.1639,
      "step": 114000
    },
    {
      "epoch": 4.826539645070185,
      "grad_norm": 1.2463469505310059,
      "learning_rate": 8.207777730137632e-05,
      "loss": 2.1683,
      "step": 114500
    },
    {
      "epoch": 4.847616237406736,
      "grad_norm": 1.4688940048217773,
      "learning_rate": 8.199202503966042e-05,
      "loss": 2.1662,
      "step": 115000
    },
    {
      "epoch": 4.868692829743287,
      "grad_norm": 1.5165785551071167,
      "learning_rate": 8.190627277794453e-05,
      "loss": 2.1662,
      "step": 115500
    },
    {
      "epoch": 4.889769422079838,
      "grad_norm": 1.1838643550872803,
      "learning_rate": 8.182052051622862e-05,
      "loss": 2.1617,
      "step": 116000
    },
    {
      "epoch": 4.910846014416389,
      "grad_norm": 2.648503303527832,
      "learning_rate": 8.173476825451271e-05,
      "loss": 2.1622,
      "step": 116500
    },
    {
      "epoch": 4.9319226067529405,
      "grad_norm": 2.5272367000579834,
      "learning_rate": 8.164901599279681e-05,
      "loss": 2.1601,
      "step": 117000
    },
    {
      "epoch": 4.952999199089491,
      "grad_norm": 1.8500511646270752,
      "learning_rate": 8.156326373108091e-05,
      "loss": 2.1592,
      "step": 117500
    },
    {
      "epoch": 4.974075791426042,
      "grad_norm": 1.3663266897201538,
      "learning_rate": 8.147751146936501e-05,
      "loss": 2.1572,
      "step": 118000
    },
    {
      "epoch": 4.995152383762593,
      "grad_norm": 1.4615684747695923,
      "learning_rate": 8.139175920764911e-05,
      "loss": 2.149,
      "step": 118500
    },
    {
      "epoch": 5.0162289760991445,
      "grad_norm": 1.4587327241897583,
      "learning_rate": 8.13060069459332e-05,
      "loss": 2.1483,
      "step": 119000
    },
    {
      "epoch": 5.037305568435695,
      "grad_norm": 1.3175787925720215,
      "learning_rate": 8.12202546842173e-05,
      "loss": 2.1405,
      "step": 119500
    },
    {
      "epoch": 5.058382160772246,
      "grad_norm": 1.4637924432754517,
      "learning_rate": 8.11345024225014e-05,
      "loss": 2.1438,
      "step": 120000
    },
    {
      "epoch": 5.058382160772246,
      "eval_accuracy": 0.6390799520761014,
      "eval_loss": 1.8210008144378662,
      "eval_runtime": 169.2085,
      "eval_samples_per_second": 713.357,
      "eval_steps_per_second": 7.435,
      "step": 120000
    },
    {
      "epoch": 5.079458753108797,
      "grad_norm": 2.7181005477905273,
      "learning_rate": 8.104875016078549e-05,
      "loss": 2.1356,
      "step": 120500
    },
    {
      "epoch": 5.100535345445349,
      "grad_norm": 1.1788932085037231,
      "learning_rate": 8.09629978990696e-05,
      "loss": 2.1361,
      "step": 121000
    },
    {
      "epoch": 5.1216119377819,
      "grad_norm": 2.8595821857452393,
      "learning_rate": 8.08772456373537e-05,
      "loss": 2.1381,
      "step": 121500
    },
    {
      "epoch": 5.14268853011845,
      "grad_norm": 2.147122621536255,
      "learning_rate": 8.079149337563778e-05,
      "loss": 2.1387,
      "step": 122000
    },
    {
      "epoch": 5.163765122455001,
      "grad_norm": 2.4605050086975098,
      "learning_rate": 8.070574111392188e-05,
      "loss": 2.1354,
      "step": 122500
    },
    {
      "epoch": 5.184841714791553,
      "grad_norm": 1.6390308141708374,
      "learning_rate": 8.061998885220599e-05,
      "loss": 2.131,
      "step": 123000
    },
    {
      "epoch": 5.205918307128104,
      "grad_norm": 1.443233609199524,
      "learning_rate": 8.053423659049008e-05,
      "loss": 2.1327,
      "step": 123500
    },
    {
      "epoch": 5.226994899464654,
      "grad_norm": 2.6145219802856445,
      "learning_rate": 8.044848432877418e-05,
      "loss": 2.1358,
      "step": 124000
    },
    {
      "epoch": 5.2480714918012055,
      "grad_norm": 1.1072486639022827,
      "learning_rate": 8.036273206705827e-05,
      "loss": 2.1377,
      "step": 124500
    },
    {
      "epoch": 5.269148084137757,
      "grad_norm": 1.130508303642273,
      "learning_rate": 8.027697980534237e-05,
      "loss": 2.1303,
      "step": 125000
    },
    {
      "epoch": 5.290224676474308,
      "grad_norm": 1.484932780265808,
      "learning_rate": 8.019122754362647e-05,
      "loss": 2.1299,
      "step": 125500
    },
    {
      "epoch": 5.311301268810858,
      "grad_norm": 1.2832825183868408,
      "learning_rate": 8.010547528191057e-05,
      "loss": 2.129,
      "step": 126000
    },
    {
      "epoch": 5.3323778611474095,
      "grad_norm": 1.7906153202056885,
      "learning_rate": 8.001972302019465e-05,
      "loss": 2.1249,
      "step": 126500
    },
    {
      "epoch": 5.353454453483961,
      "grad_norm": 2.8754312992095947,
      "learning_rate": 7.993397075847876e-05,
      "loss": 2.1233,
      "step": 127000
    },
    {
      "epoch": 5.374531045820512,
      "grad_norm": 1.5033295154571533,
      "learning_rate": 7.984821849676286e-05,
      "loss": 2.1249,
      "step": 127500
    },
    {
      "epoch": 5.395607638157063,
      "grad_norm": 1.4630507230758667,
      "learning_rate": 7.976246623504695e-05,
      "loss": 2.1254,
      "step": 128000
    },
    {
      "epoch": 5.416684230493614,
      "grad_norm": 1.5815629959106445,
      "learning_rate": 7.967671397333105e-05,
      "loss": 2.117,
      "step": 128500
    },
    {
      "epoch": 5.437760822830165,
      "grad_norm": 1.1827993392944336,
      "learning_rate": 7.959096171161516e-05,
      "loss": 2.1145,
      "step": 129000
    },
    {
      "epoch": 5.458837415166716,
      "grad_norm": 1.3161492347717285,
      "learning_rate": 7.950520944989924e-05,
      "loss": 2.1164,
      "step": 129500
    },
    {
      "epoch": 5.479914007503266,
      "grad_norm": 1.3320565223693848,
      "learning_rate": 7.941945718818334e-05,
      "loss": 2.117,
      "step": 130000
    },
    {
      "epoch": 5.479914007503266,
      "eval_accuracy": 0.642808509233094,
      "eval_loss": 1.7991453409194946,
      "eval_runtime": 168.3547,
      "eval_samples_per_second": 716.974,
      "eval_steps_per_second": 7.472,
      "step": 130000
    },
    {
      "epoch": 5.500990599839818,
      "grad_norm": 1.1708186864852905,
      "learning_rate": 7.933370492646744e-05,
      "loss": 2.1219,
      "step": 130500
    },
    {
      "epoch": 5.522067192176369,
      "grad_norm": 2.6440517902374268,
      "learning_rate": 7.924795266475154e-05,
      "loss": 2.1149,
      "step": 131000
    },
    {
      "epoch": 5.54314378451292,
      "grad_norm": 1.3327730894088745,
      "learning_rate": 7.916220040303564e-05,
      "loss": 2.1075,
      "step": 131500
    },
    {
      "epoch": 5.564220376849471,
      "grad_norm": 1.1900001764297485,
      "learning_rate": 7.907644814131972e-05,
      "loss": 2.1153,
      "step": 132000
    },
    {
      "epoch": 5.585296969186022,
      "grad_norm": 1.1786779165267944,
      "learning_rate": 7.899069587960383e-05,
      "loss": 2.1075,
      "step": 132500
    },
    {
      "epoch": 5.606373561522573,
      "grad_norm": 1.3604512214660645,
      "learning_rate": 7.890494361788793e-05,
      "loss": 2.1072,
      "step": 133000
    },
    {
      "epoch": 5.627450153859124,
      "grad_norm": 1.9362701177597046,
      "learning_rate": 7.881919135617202e-05,
      "loss": 2.1055,
      "step": 133500
    },
    {
      "epoch": 5.648526746195675,
      "grad_norm": 1.1473065614700317,
      "learning_rate": 7.873343909445611e-05,
      "loss": 2.1011,
      "step": 134000
    },
    {
      "epoch": 5.669603338532226,
      "grad_norm": 1.3713167905807495,
      "learning_rate": 7.864768683274023e-05,
      "loss": 2.1112,
      "step": 134500
    },
    {
      "epoch": 5.690679930868777,
      "grad_norm": 1.220912218093872,
      "learning_rate": 7.856193457102431e-05,
      "loss": 2.1103,
      "step": 135000
    },
    {
      "epoch": 5.711756523205328,
      "grad_norm": 1.3044885396957397,
      "learning_rate": 7.847618230930841e-05,
      "loss": 2.1067,
      "step": 135500
    },
    {
      "epoch": 5.7328331155418795,
      "grad_norm": 1.4253188371658325,
      "learning_rate": 7.839043004759251e-05,
      "loss": 2.1014,
      "step": 136000
    },
    {
      "epoch": 5.75390970787843,
      "grad_norm": 1.5199997425079346,
      "learning_rate": 7.83046777858766e-05,
      "loss": 2.1002,
      "step": 136500
    },
    {
      "epoch": 5.774986300214981,
      "grad_norm": 1.3262827396392822,
      "learning_rate": 7.82189255241607e-05,
      "loss": 2.102,
      "step": 137000
    },
    {
      "epoch": 5.796062892551532,
      "grad_norm": 1.5269746780395508,
      "learning_rate": 7.81331732624448e-05,
      "loss": 2.1025,
      "step": 137500
    },
    {
      "epoch": 5.8171394848880835,
      "grad_norm": 1.7843936681747437,
      "learning_rate": 7.804742100072889e-05,
      "loss": 2.0999,
      "step": 138000
    },
    {
      "epoch": 5.838216077224635,
      "grad_norm": 1.3698456287384033,
      "learning_rate": 7.7961668739013e-05,
      "loss": 2.1012,
      "step": 138500
    },
    {
      "epoch": 5.859292669561185,
      "grad_norm": 1.3630893230438232,
      "learning_rate": 7.78759164772971e-05,
      "loss": 2.0975,
      "step": 139000
    },
    {
      "epoch": 5.880369261897736,
      "grad_norm": 1.1533725261688232,
      "learning_rate": 7.779016421558118e-05,
      "loss": 2.0971,
      "step": 139500
    },
    {
      "epoch": 5.901445854234288,
      "grad_norm": 1.278883695602417,
      "learning_rate": 7.770441195386528e-05,
      "loss": 2.1013,
      "step": 140000
    },
    {
      "epoch": 5.901445854234288,
      "eval_accuracy": 0.6462912645777124,
      "eval_loss": 1.7722166776657104,
      "eval_runtime": 169.5356,
      "eval_samples_per_second": 711.98,
      "eval_steps_per_second": 7.42,
      "step": 140000
    },
    {
      "epoch": 5.922522446570838,
      "grad_norm": 1.522201418876648,
      "learning_rate": 7.761865969214939e-05,
      "loss": 2.0918,
      "step": 140500
    },
    {
      "epoch": 5.943599038907389,
      "grad_norm": 1.6708459854125977,
      "learning_rate": 7.753290743043348e-05,
      "loss": 2.0977,
      "step": 141000
    },
    {
      "epoch": 5.96467563124394,
      "grad_norm": 3.7320902347564697,
      "learning_rate": 7.744715516871758e-05,
      "loss": 2.0942,
      "step": 141500
    },
    {
      "epoch": 5.985752223580492,
      "grad_norm": 1.8618805408477783,
      "learning_rate": 7.736140290700167e-05,
      "loss": 2.0908,
      "step": 142000
    },
    {
      "epoch": 6.006828815917043,
      "grad_norm": 1.146378517150879,
      "learning_rate": 7.727565064528577e-05,
      "loss": 2.086,
      "step": 142500
    },
    {
      "epoch": 6.027905408253593,
      "grad_norm": 1.4122670888900757,
      "learning_rate": 7.718989838356987e-05,
      "loss": 2.0813,
      "step": 143000
    },
    {
      "epoch": 6.0489820005901445,
      "grad_norm": 1.691215991973877,
      "learning_rate": 7.710414612185397e-05,
      "loss": 2.08,
      "step": 143500
    },
    {
      "epoch": 6.070058592926696,
      "grad_norm": 1.5658148527145386,
      "learning_rate": 7.701839386013807e-05,
      "loss": 2.0778,
      "step": 144000
    },
    {
      "epoch": 6.091135185263247,
      "grad_norm": 1.596634030342102,
      "learning_rate": 7.693264159842217e-05,
      "loss": 2.0755,
      "step": 144500
    },
    {
      "epoch": 6.112211777599797,
      "grad_norm": 1.4554200172424316,
      "learning_rate": 7.684688933670626e-05,
      "loss": 2.0722,
      "step": 145000
    },
    {
      "epoch": 6.1332883699363485,
      "grad_norm": 1.6090881824493408,
      "learning_rate": 7.676113707499035e-05,
      "loss": 2.0726,
      "step": 145500
    },
    {
      "epoch": 6.1543649622729,
      "grad_norm": 1.2874369621276855,
      "learning_rate": 7.667538481327446e-05,
      "loss": 2.0717,
      "step": 146000
    },
    {
      "epoch": 6.175441554609451,
      "grad_norm": 1.55347740650177,
      "learning_rate": 7.658963255155856e-05,
      "loss": 2.0766,
      "step": 146500
    },
    {
      "epoch": 6.196518146946001,
      "grad_norm": 1.9679468870162964,
      "learning_rate": 7.650388028984264e-05,
      "loss": 2.0755,
      "step": 147000
    },
    {
      "epoch": 6.217594739282553,
      "grad_norm": 1.645655870437622,
      "learning_rate": 7.641812802812674e-05,
      "loss": 2.0702,
      "step": 147500
    },
    {
      "epoch": 6.238671331619104,
      "grad_norm": 1.470003604888916,
      "learning_rate": 7.633237576641084e-05,
      "loss": 2.0711,
      "step": 148000
    },
    {
      "epoch": 6.259747923955655,
      "grad_norm": 1.4030473232269287,
      "learning_rate": 7.624662350469494e-05,
      "loss": 2.0721,
      "step": 148500
    },
    {
      "epoch": 6.280824516292206,
      "grad_norm": 1.4137738943099976,
      "learning_rate": 7.616087124297904e-05,
      "loss": 2.0735,
      "step": 149000
    },
    {
      "epoch": 6.301901108628757,
      "grad_norm": 1.608684778213501,
      "learning_rate": 7.607511898126313e-05,
      "loss": 2.0708,
      "step": 149500
    },
    {
      "epoch": 6.322977700965308,
      "grad_norm": 2.005465269088745,
      "learning_rate": 7.598936671954723e-05,
      "loss": 2.0674,
      "step": 150000
    },
    {
      "epoch": 6.322977700965308,
      "eval_accuracy": 0.6491457309523007,
      "eval_loss": 1.7555503845214844,
      "eval_runtime": 168.5751,
      "eval_samples_per_second": 716.037,
      "eval_steps_per_second": 7.463,
      "step": 150000
    },
    {
      "epoch": 6.344054293301859,
      "grad_norm": 1.4014167785644531,
      "learning_rate": 7.590361445783133e-05,
      "loss": 2.0704,
      "step": 150500
    },
    {
      "epoch": 6.36513088563841,
      "grad_norm": 2.252483367919922,
      "learning_rate": 7.581786219611543e-05,
      "loss": 2.0707,
      "step": 151000
    },
    {
      "epoch": 6.386207477974961,
      "grad_norm": 1.6945236921310425,
      "learning_rate": 7.573210993439951e-05,
      "loss": 2.0701,
      "step": 151500
    },
    {
      "epoch": 6.407284070311512,
      "grad_norm": 1.3200337886810303,
      "learning_rate": 7.564635767268363e-05,
      "loss": 2.0697,
      "step": 152000
    },
    {
      "epoch": 6.428360662648063,
      "grad_norm": 1.8187270164489746,
      "learning_rate": 7.556060541096772e-05,
      "loss": 2.0642,
      "step": 152500
    },
    {
      "epoch": 6.449437254984614,
      "grad_norm": 1.771296739578247,
      "learning_rate": 7.547485314925181e-05,
      "loss": 2.0645,
      "step": 153000
    },
    {
      "epoch": 6.470513847321165,
      "grad_norm": 1.3185912370681763,
      "learning_rate": 7.538910088753591e-05,
      "loss": 2.0633,
      "step": 153500
    },
    {
      "epoch": 6.491590439657716,
      "grad_norm": 1.283721923828125,
      "learning_rate": 7.530334862582002e-05,
      "loss": 2.063,
      "step": 154000
    },
    {
      "epoch": 6.512667031994267,
      "grad_norm": 1.493251085281372,
      "learning_rate": 7.52175963641041e-05,
      "loss": 2.0664,
      "step": 154500
    },
    {
      "epoch": 6.5337436243308185,
      "grad_norm": 1.5226579904556274,
      "learning_rate": 7.51318441023882e-05,
      "loss": 2.0684,
      "step": 155000
    },
    {
      "epoch": 6.55482021666737,
      "grad_norm": 1.4216480255126953,
      "learning_rate": 7.50460918406723e-05,
      "loss": 2.0587,
      "step": 155500
    },
    {
      "epoch": 6.57589680900392,
      "grad_norm": 1.4682893753051758,
      "learning_rate": 7.49603395789564e-05,
      "loss": 2.0578,
      "step": 156000
    },
    {
      "epoch": 6.596973401340471,
      "grad_norm": 1.5490312576293945,
      "learning_rate": 7.48745873172405e-05,
      "loss": 2.0565,
      "step": 156500
    },
    {
      "epoch": 6.6180499936770225,
      "grad_norm": 1.2250019311904907,
      "learning_rate": 7.47888350555246e-05,
      "loss": 2.0534,
      "step": 157000
    },
    {
      "epoch": 6.639126586013573,
      "grad_norm": 1.4521032571792603,
      "learning_rate": 7.47030827938087e-05,
      "loss": 2.0585,
      "step": 157500
    },
    {
      "epoch": 6.660203178350124,
      "grad_norm": 1.6686809062957764,
      "learning_rate": 7.461733053209279e-05,
      "loss": 2.057,
      "step": 158000
    },
    {
      "epoch": 6.681279770686675,
      "grad_norm": 1.697413444519043,
      "learning_rate": 7.453157827037689e-05,
      "loss": 2.0607,
      "step": 158500
    },
    {
      "epoch": 6.702356363023227,
      "grad_norm": 1.6466525793075562,
      "learning_rate": 7.444582600866098e-05,
      "loss": 2.0522,
      "step": 159000
    },
    {
      "epoch": 6.723432955359778,
      "grad_norm": 1.6838945150375366,
      "learning_rate": 7.436007374694507e-05,
      "loss": 2.057,
      "step": 159500
    },
    {
      "epoch": 6.744509547696328,
      "grad_norm": 1.5388096570968628,
      "learning_rate": 7.427432148522919e-05,
      "loss": 2.0597,
      "step": 160000
    },
    {
      "epoch": 6.744509547696328,
      "eval_accuracy": 0.6515849059742659,
      "eval_loss": 1.741092324256897,
      "eval_runtime": 168.8232,
      "eval_samples_per_second": 714.985,
      "eval_steps_per_second": 7.452,
      "step": 160000
    },
    {
      "epoch": 6.765586140032879,
      "grad_norm": 1.4921259880065918,
      "learning_rate": 7.418856922351327e-05,
      "loss": 2.0559,
      "step": 160500
    },
    {
      "epoch": 6.786662732369431,
      "grad_norm": 1.3313426971435547,
      "learning_rate": 7.410281696179737e-05,
      "loss": 2.0513,
      "step": 161000
    },
    {
      "epoch": 6.807739324705982,
      "grad_norm": 1.4331045150756836,
      "learning_rate": 7.401706470008147e-05,
      "loss": 2.0479,
      "step": 161500
    },
    {
      "epoch": 6.828815917042532,
      "grad_norm": 1.5091363191604614,
      "learning_rate": 7.393131243836557e-05,
      "loss": 2.0492,
      "step": 162000
    },
    {
      "epoch": 6.8498925093790834,
      "grad_norm": 1.3586465120315552,
      "learning_rate": 7.384556017664966e-05,
      "loss": 2.0492,
      "step": 162500
    },
    {
      "epoch": 6.870969101715635,
      "grad_norm": 2.6906638145446777,
      "learning_rate": 7.375980791493376e-05,
      "loss": 2.046,
      "step": 163000
    },
    {
      "epoch": 6.892045694052186,
      "grad_norm": 1.37527334690094,
      "learning_rate": 7.367405565321786e-05,
      "loss": 2.0464,
      "step": 163500
    },
    {
      "epoch": 6.913122286388736,
      "grad_norm": 1.5771501064300537,
      "learning_rate": 7.358830339150196e-05,
      "loss": 2.0487,
      "step": 164000
    },
    {
      "epoch": 6.9341988787252875,
      "grad_norm": 1.4107612371444702,
      "learning_rate": 7.350255112978606e-05,
      "loss": 2.052,
      "step": 164500
    },
    {
      "epoch": 6.955275471061839,
      "grad_norm": 1.9493982791900635,
      "learning_rate": 7.341679886807014e-05,
      "loss": 2.0464,
      "step": 165000
    },
    {
      "epoch": 6.97635206339839,
      "grad_norm": 1.4127299785614014,
      "learning_rate": 7.333104660635425e-05,
      "loss": 2.0424,
      "step": 165500
    },
    {
      "epoch": 6.997428655734941,
      "grad_norm": 1.086712121963501,
      "learning_rate": 7.324529434463835e-05,
      "loss": 2.0425,
      "step": 166000
    },
    {
      "epoch": 7.018505248071492,
      "grad_norm": 1.436981439590454,
      "learning_rate": 7.315954208292244e-05,
      "loss": 2.0358,
      "step": 166500
    },
    {
      "epoch": 7.039581840408043,
      "grad_norm": 1.5045158863067627,
      "learning_rate": 7.307378982120653e-05,
      "loss": 2.029,
      "step": 167000
    },
    {
      "epoch": 7.060658432744594,
      "grad_norm": 1.7590110301971436,
      "learning_rate": 7.298803755949063e-05,
      "loss": 2.0259,
      "step": 167500
    },
    {
      "epoch": 7.081735025081145,
      "grad_norm": 1.339489459991455,
      "learning_rate": 7.290228529777473e-05,
      "loss": 2.029,
      "step": 168000
    },
    {
      "epoch": 7.102811617417696,
      "grad_norm": 1.3614423274993896,
      "learning_rate": 7.281653303605883e-05,
      "loss": 2.0342,
      "step": 168500
    },
    {
      "epoch": 7.123888209754247,
      "grad_norm": 1.4586058855056763,
      "learning_rate": 7.273078077434293e-05,
      "loss": 2.0362,
      "step": 169000
    },
    {
      "epoch": 7.144964802090798,
      "grad_norm": 1.29351007938385,
      "learning_rate": 7.264502851262703e-05,
      "loss": 2.0309,
      "step": 169500
    },
    {
      "epoch": 7.166041394427349,
      "grad_norm": 2.040558099746704,
      "learning_rate": 7.255927625091112e-05,
      "loss": 2.0254,
      "step": 170000
    },
    {
      "epoch": 7.166041394427349,
      "eval_accuracy": 0.6539346618386137,
      "eval_loss": 1.7298465967178345,
      "eval_runtime": 169.5891,
      "eval_samples_per_second": 711.756,
      "eval_steps_per_second": 7.418,
      "step": 170000
    },
    {
      "epoch": 7.1871179867639,
      "grad_norm": 1.46829092502594,
      "learning_rate": 7.247352398919521e-05,
      "loss": 2.032,
      "step": 170500
    },
    {
      "epoch": 7.208194579100451,
      "grad_norm": 1.797717571258545,
      "learning_rate": 7.238777172747931e-05,
      "loss": 2.0305,
      "step": 171000
    },
    {
      "epoch": 7.229271171437002,
      "grad_norm": 1.562495231628418,
      "learning_rate": 7.230201946576342e-05,
      "loss": 2.0308,
      "step": 171500
    },
    {
      "epoch": 7.250347763773553,
      "grad_norm": 1.9791704416275024,
      "learning_rate": 7.22162672040475e-05,
      "loss": 2.0313,
      "step": 172000
    },
    {
      "epoch": 7.271424356110104,
      "grad_norm": 1.4129750728607178,
      "learning_rate": 7.21305149423316e-05,
      "loss": 2.026,
      "step": 172500
    },
    {
      "epoch": 7.292500948446655,
      "grad_norm": 1.5719834566116333,
      "learning_rate": 7.20447626806157e-05,
      "loss": 2.0242,
      "step": 173000
    },
    {
      "epoch": 7.313577540783206,
      "grad_norm": 1.6413276195526123,
      "learning_rate": 7.19590104188998e-05,
      "loss": 2.0219,
      "step": 173500
    },
    {
      "epoch": 7.3346541331197574,
      "grad_norm": 2.2408313751220703,
      "learning_rate": 7.18732581571839e-05,
      "loss": 2.0281,
      "step": 174000
    },
    {
      "epoch": 7.355730725456308,
      "grad_norm": 1.6538187265396118,
      "learning_rate": 7.1787505895468e-05,
      "loss": 2.0265,
      "step": 174500
    },
    {
      "epoch": 7.376807317792859,
      "grad_norm": 1.55643892288208,
      "learning_rate": 7.17017536337521e-05,
      "loss": 2.0263,
      "step": 175000
    },
    {
      "epoch": 7.39788391012941,
      "grad_norm": 1.4459589719772339,
      "learning_rate": 7.161600137203619e-05,
      "loss": 2.0248,
      "step": 175500
    },
    {
      "epoch": 7.4189605024659615,
      "grad_norm": 1.9289194345474243,
      "learning_rate": 7.153024911032029e-05,
      "loss": 2.0195,
      "step": 176000
    },
    {
      "epoch": 7.440037094802513,
      "grad_norm": 2.1924948692321777,
      "learning_rate": 7.144449684860438e-05,
      "loss": 2.0232,
      "step": 176500
    },
    {
      "epoch": 7.461113687139063,
      "grad_norm": 2.435021162033081,
      "learning_rate": 7.135874458688849e-05,
      "loss": 2.0298,
      "step": 177000
    },
    {
      "epoch": 7.482190279475614,
      "grad_norm": 1.6398460865020752,
      "learning_rate": 7.127299232517259e-05,
      "loss": 2.024,
      "step": 177500
    },
    {
      "epoch": 7.503266871812166,
      "grad_norm": 1.5504934787750244,
      "learning_rate": 7.118724006345667e-05,
      "loss": 2.0241,
      "step": 178000
    },
    {
      "epoch": 7.524343464148717,
      "grad_norm": 1.4459161758422852,
      "learning_rate": 7.110148780174077e-05,
      "loss": 2.0197,
      "step": 178500
    },
    {
      "epoch": 7.545420056485267,
      "grad_norm": 1.365692138671875,
      "learning_rate": 7.101573554002488e-05,
      "loss": 2.0181,
      "step": 179000
    },
    {
      "epoch": 7.566496648821818,
      "grad_norm": 1.7721461057662964,
      "learning_rate": 7.092998327830897e-05,
      "loss": 2.0235,
      "step": 179500
    },
    {
      "epoch": 7.58757324115837,
      "grad_norm": 1.4443342685699463,
      "learning_rate": 7.084423101659306e-05,
      "loss": 2.0177,
      "step": 180000
    },
    {
      "epoch": 7.58757324115837,
      "eval_accuracy": 0.6553247301508573,
      "eval_loss": 1.718245506286621,
      "eval_runtime": 169.297,
      "eval_samples_per_second": 712.984,
      "eval_steps_per_second": 7.431,
      "step": 180000
    },
    {
      "epoch": 7.608649833494921,
      "grad_norm": 1.4858146905899048,
      "learning_rate": 7.075847875487716e-05,
      "loss": 2.0209,
      "step": 180500
    },
    {
      "epoch": 7.629726425831471,
      "grad_norm": 1.6458719968795776,
      "learning_rate": 7.067272649316126e-05,
      "loss": 2.0223,
      "step": 181000
    },
    {
      "epoch": 7.650803018168022,
      "grad_norm": 2.7739953994750977,
      "learning_rate": 7.058697423144536e-05,
      "loss": 2.0147,
      "step": 181500
    },
    {
      "epoch": 7.671879610504574,
      "grad_norm": 1.6043736934661865,
      "learning_rate": 7.050122196972946e-05,
      "loss": 2.0163,
      "step": 182000
    },
    {
      "epoch": 7.692956202841125,
      "grad_norm": 1.2854926586151123,
      "learning_rate": 7.041546970801354e-05,
      "loss": 2.0179,
      "step": 182500
    },
    {
      "epoch": 7.714032795177676,
      "grad_norm": 1.5477572679519653,
      "learning_rate": 7.032971744629765e-05,
      "loss": 2.0175,
      "step": 183000
    },
    {
      "epoch": 7.7351093875142265,
      "grad_norm": 1.2823032140731812,
      "learning_rate": 7.024396518458175e-05,
      "loss": 2.02,
      "step": 183500
    },
    {
      "epoch": 7.756185979850778,
      "grad_norm": 1.1444975137710571,
      "learning_rate": 7.015821292286584e-05,
      "loss": 2.0179,
      "step": 184000
    },
    {
      "epoch": 7.777262572187329,
      "grad_norm": 1.9443219900131226,
      "learning_rate": 7.007246066114993e-05,
      "loss": 2.0171,
      "step": 184500
    },
    {
      "epoch": 7.798339164523879,
      "grad_norm": 1.403972864151001,
      "learning_rate": 6.998670839943405e-05,
      "loss": 2.0141,
      "step": 185000
    },
    {
      "epoch": 7.8194157568604306,
      "grad_norm": 1.3283175230026245,
      "learning_rate": 6.990095613771813e-05,
      "loss": 2.0111,
      "step": 185500
    },
    {
      "epoch": 7.840492349196982,
      "grad_norm": 1.21708345413208,
      "learning_rate": 6.981520387600223e-05,
      "loss": 2.0138,
      "step": 186000
    },
    {
      "epoch": 7.861568941533533,
      "grad_norm": 1.5668725967407227,
      "learning_rate": 6.972945161428633e-05,
      "loss": 2.0146,
      "step": 186500
    },
    {
      "epoch": 7.882645533870084,
      "grad_norm": 2.1780078411102295,
      "learning_rate": 6.964369935257043e-05,
      "loss": 2.0141,
      "step": 187000
    },
    {
      "epoch": 7.903722126206635,
      "grad_norm": 1.3229855298995972,
      "learning_rate": 6.955794709085452e-05,
      "loss": 2.0155,
      "step": 187500
    },
    {
      "epoch": 7.924798718543186,
      "grad_norm": 2.8959453105926514,
      "learning_rate": 6.947219482913862e-05,
      "loss": 2.0021,
      "step": 188000
    },
    {
      "epoch": 7.945875310879737,
      "grad_norm": 1.5594075918197632,
      "learning_rate": 6.938644256742272e-05,
      "loss": 2.0139,
      "step": 188500
    },
    {
      "epoch": 7.966951903216288,
      "grad_norm": 1.6598451137542725,
      "learning_rate": 6.930069030570682e-05,
      "loss": 2.0096,
      "step": 189000
    },
    {
      "epoch": 7.988028495552839,
      "grad_norm": 2.01891827583313,
      "learning_rate": 6.921493804399092e-05,
      "loss": 2.008,
      "step": 189500
    },
    {
      "epoch": 8.00910508788939,
      "grad_norm": 1.7114694118499756,
      "learning_rate": 6.9129185782275e-05,
      "loss": 2.0039,
      "step": 190000
    },
    {
      "epoch": 8.00910508788939,
      "eval_accuracy": 0.6570838366705001,
      "eval_loss": 1.730424404144287,
      "eval_runtime": 170.4849,
      "eval_samples_per_second": 708.016,
      "eval_steps_per_second": 7.379,
      "step": 190000
    },
    {
      "epoch": 8.03018168022594,
      "grad_norm": 1.646298885345459,
      "learning_rate": 6.904343352055911e-05,
      "loss": 1.9945,
      "step": 190500
    },
    {
      "epoch": 8.051258272562492,
      "grad_norm": 1.4077107906341553,
      "learning_rate": 6.895768125884321e-05,
      "loss": 1.9979,
      "step": 191000
    },
    {
      "epoch": 8.072334864899043,
      "grad_norm": 1.0981172323226929,
      "learning_rate": 6.88719289971273e-05,
      "loss": 1.9966,
      "step": 191500
    },
    {
      "epoch": 8.093411457235595,
      "grad_norm": 1.2472593784332275,
      "learning_rate": 6.87861767354114e-05,
      "loss": 1.9896,
      "step": 192000
    },
    {
      "epoch": 8.114488049572145,
      "grad_norm": 2.859118938446045,
      "learning_rate": 6.870042447369551e-05,
      "loss": 1.9992,
      "step": 192500
    },
    {
      "epoch": 8.135564641908696,
      "grad_norm": 1.384700059890747,
      "learning_rate": 6.861467221197959e-05,
      "loss": 1.9958,
      "step": 193000
    },
    {
      "epoch": 8.156641234245248,
      "grad_norm": 1.4313791990280151,
      "learning_rate": 6.852891995026369e-05,
      "loss": 1.9977,
      "step": 193500
    },
    {
      "epoch": 8.177717826581798,
      "grad_norm": 1.6931345462799072,
      "learning_rate": 6.844316768854779e-05,
      "loss": 1.9989,
      "step": 194000
    },
    {
      "epoch": 8.19879441891835,
      "grad_norm": 1.599657416343689,
      "learning_rate": 6.835741542683189e-05,
      "loss": 1.9995,
      "step": 194500
    },
    {
      "epoch": 8.2198710112549,
      "grad_norm": 1.3218544721603394,
      "learning_rate": 6.827166316511599e-05,
      "loss": 1.9947,
      "step": 195000
    },
    {
      "epoch": 8.24094760359145,
      "grad_norm": 2.1590473651885986,
      "learning_rate": 6.818591090340008e-05,
      "loss": 1.9952,
      "step": 195500
    },
    {
      "epoch": 8.262024195928003,
      "grad_norm": 1.3900055885314941,
      "learning_rate": 6.810015864168417e-05,
      "loss": 1.9904,
      "step": 196000
    },
    {
      "epoch": 8.283100788264553,
      "grad_norm": 1.2849501371383667,
      "learning_rate": 6.801440637996828e-05,
      "loss": 1.9873,
      "step": 196500
    },
    {
      "epoch": 8.304177380601104,
      "grad_norm": 3.480060577392578,
      "learning_rate": 6.792865411825238e-05,
      "loss": 1.9967,
      "step": 197000
    },
    {
      "epoch": 8.325253972937656,
      "grad_norm": 2.1442606449127197,
      "learning_rate": 6.784290185653646e-05,
      "loss": 1.9923,
      "step": 197500
    },
    {
      "epoch": 8.346330565274206,
      "grad_norm": 1.4439301490783691,
      "learning_rate": 6.775714959482056e-05,
      "loss": 1.9933,
      "step": 198000
    },
    {
      "epoch": 8.367407157610758,
      "grad_norm": 1.5011173486709595,
      "learning_rate": 6.767139733310467e-05,
      "loss": 1.9937,
      "step": 198500
    },
    {
      "epoch": 8.388483749947309,
      "grad_norm": 1.6031461954116821,
      "learning_rate": 6.758564507138876e-05,
      "loss": 1.9967,
      "step": 199000
    },
    {
      "epoch": 8.409560342283859,
      "grad_norm": 1.9713538885116577,
      "learning_rate": 6.749989280967286e-05,
      "loss": 1.9948,
      "step": 199500
    },
    {
      "epoch": 8.430636934620411,
      "grad_norm": 1.4085208177566528,
      "learning_rate": 6.741414054795696e-05,
      "loss": 1.9926,
      "step": 200000
    },
    {
      "epoch": 8.430636934620411,
      "eval_accuracy": 0.6592788469597803,
      "eval_loss": 1.6948425769805908,
      "eval_runtime": 169.1762,
      "eval_samples_per_second": 713.493,
      "eval_steps_per_second": 7.436,
      "step": 200000
    },
    {
      "epoch": 8.451713526956961,
      "grad_norm": 1.624588966369629,
      "learning_rate": 6.732838828624105e-05,
      "loss": 1.99,
      "step": 200500
    },
    {
      "epoch": 8.472790119293512,
      "grad_norm": 1.7879573106765747,
      "learning_rate": 6.724263602452515e-05,
      "loss": 1.9949,
      "step": 201000
    },
    {
      "epoch": 8.493866711630064,
      "grad_norm": 1.4974569082260132,
      "learning_rate": 6.715688376280924e-05,
      "loss": 1.9883,
      "step": 201500
    },
    {
      "epoch": 8.514943303966614,
      "grad_norm": 1.3319435119628906,
      "learning_rate": 6.707113150109335e-05,
      "loss": 1.9881,
      "step": 202000
    },
    {
      "epoch": 8.536019896303166,
      "grad_norm": 1.5998083353042603,
      "learning_rate": 6.698537923937745e-05,
      "loss": 1.9957,
      "step": 202500
    },
    {
      "epoch": 8.557096488639717,
      "grad_norm": 1.4332315921783447,
      "learning_rate": 6.689962697766153e-05,
      "loss": 1.9891,
      "step": 203000
    },
    {
      "epoch": 8.578173080976267,
      "grad_norm": 1.632836937904358,
      "learning_rate": 6.681387471594563e-05,
      "loss": 1.9897,
      "step": 203500
    },
    {
      "epoch": 8.59924967331282,
      "grad_norm": 1.2590038776397705,
      "learning_rate": 6.672812245422974e-05,
      "loss": 1.9863,
      "step": 204000
    },
    {
      "epoch": 8.62032626564937,
      "grad_norm": 2.1204171180725098,
      "learning_rate": 6.664237019251383e-05,
      "loss": 1.9909,
      "step": 204500
    },
    {
      "epoch": 8.64140285798592,
      "grad_norm": 1.6717052459716797,
      "learning_rate": 6.655661793079792e-05,
      "loss": 1.99,
      "step": 205000
    },
    {
      "epoch": 8.662479450322472,
      "grad_norm": 1.4797532558441162,
      "learning_rate": 6.647086566908202e-05,
      "loss": 1.9829,
      "step": 205500
    },
    {
      "epoch": 8.683556042659022,
      "grad_norm": 1.3563510179519653,
      "learning_rate": 6.638511340736612e-05,
      "loss": 1.9863,
      "step": 206000
    },
    {
      "epoch": 8.704632634995574,
      "grad_norm": 1.5234061479568481,
      "learning_rate": 6.629936114565022e-05,
      "loss": 1.9801,
      "step": 206500
    },
    {
      "epoch": 8.725709227332125,
      "grad_norm": 1.2462966442108154,
      "learning_rate": 6.621360888393432e-05,
      "loss": 1.9841,
      "step": 207000
    },
    {
      "epoch": 8.746785819668675,
      "grad_norm": 1.5390892028808594,
      "learning_rate": 6.61278566222184e-05,
      "loss": 1.9789,
      "step": 207500
    },
    {
      "epoch": 8.767862412005227,
      "grad_norm": 1.1754952669143677,
      "learning_rate": 6.604210436050251e-05,
      "loss": 1.9866,
      "step": 208000
    },
    {
      "epoch": 8.788939004341778,
      "grad_norm": 1.1945759057998657,
      "learning_rate": 6.595635209878661e-05,
      "loss": 1.9851,
      "step": 208500
    },
    {
      "epoch": 8.81001559667833,
      "grad_norm": 1.5835222005844116,
      "learning_rate": 6.58705998370707e-05,
      "loss": 1.9864,
      "step": 209000
    },
    {
      "epoch": 8.83109218901488,
      "grad_norm": 1.5251295566558838,
      "learning_rate": 6.57848475753548e-05,
      "loss": 1.9858,
      "step": 209500
    },
    {
      "epoch": 8.85216878135143,
      "grad_norm": 2.2163751125335693,
      "learning_rate": 6.569909531363891e-05,
      "loss": 1.9789,
      "step": 210000
    },
    {
      "epoch": 8.85216878135143,
      "eval_accuracy": 0.6602911383503384,
      "eval_loss": 1.6922414302825928,
      "eval_runtime": 169.3555,
      "eval_samples_per_second": 712.738,
      "eval_steps_per_second": 7.428,
      "step": 210000
    },
    {
      "epoch": 8.873245373687983,
      "grad_norm": 1.4225358963012695,
      "learning_rate": 6.561334305192299e-05,
      "loss": 1.9824,
      "step": 210500
    },
    {
      "epoch": 8.894321966024533,
      "grad_norm": 1.6671773195266724,
      "learning_rate": 6.552759079020709e-05,
      "loss": 1.9877,
      "step": 211000
    },
    {
      "epoch": 8.915398558361083,
      "grad_norm": 2.346935987472534,
      "learning_rate": 6.544183852849119e-05,
      "loss": 1.98,
      "step": 211500
    },
    {
      "epoch": 8.936475150697635,
      "grad_norm": 1.5228843688964844,
      "learning_rate": 6.535608626677529e-05,
      "loss": 1.9886,
      "step": 212000
    },
    {
      "epoch": 8.957551743034186,
      "grad_norm": 1.397007942199707,
      "learning_rate": 6.527033400505939e-05,
      "loss": 1.9886,
      "step": 212500
    },
    {
      "epoch": 8.978628335370738,
      "grad_norm": 1.3140085935592651,
      "learning_rate": 6.518458174334348e-05,
      "loss": 1.9851,
      "step": 213000
    },
    {
      "epoch": 8.999704927707288,
      "grad_norm": 1.782447099685669,
      "learning_rate": 6.509882948162758e-05,
      "loss": 1.9835,
      "step": 213500
    },
    {
      "epoch": 9.020781520043839,
      "grad_norm": 1.1647506952285767,
      "learning_rate": 6.501307721991168e-05,
      "loss": 1.9675,
      "step": 214000
    },
    {
      "epoch": 9.04185811238039,
      "grad_norm": 1.7786974906921387,
      "learning_rate": 6.492732495819578e-05,
      "loss": 1.9667,
      "step": 214500
    },
    {
      "epoch": 9.062934704716941,
      "grad_norm": 1.5023516416549683,
      "learning_rate": 6.484157269647986e-05,
      "loss": 1.9698,
      "step": 215000
    },
    {
      "epoch": 9.084011297053493,
      "grad_norm": 1.4581481218338013,
      "learning_rate": 6.475582043476398e-05,
      "loss": 1.9688,
      "step": 215500
    },
    {
      "epoch": 9.105087889390044,
      "grad_norm": 1.502541422843933,
      "learning_rate": 6.467006817304807e-05,
      "loss": 1.9653,
      "step": 216000
    },
    {
      "epoch": 9.126164481726594,
      "grad_norm": 1.2884166240692139,
      "learning_rate": 6.458431591133216e-05,
      "loss": 1.9643,
      "step": 216500
    },
    {
      "epoch": 9.147241074063146,
      "grad_norm": 1.6011122465133667,
      "learning_rate": 6.449856364961626e-05,
      "loss": 1.9661,
      "step": 217000
    },
    {
      "epoch": 9.168317666399696,
      "grad_norm": 2.182887077331543,
      "learning_rate": 6.441281138790037e-05,
      "loss": 1.9654,
      "step": 217500
    },
    {
      "epoch": 9.189394258736247,
      "grad_norm": 1.2798610925674438,
      "learning_rate": 6.432705912618445e-05,
      "loss": 1.966,
      "step": 218000
    },
    {
      "epoch": 9.210470851072799,
      "grad_norm": 1.937672734260559,
      "learning_rate": 6.424130686446855e-05,
      "loss": 1.969,
      "step": 218500
    },
    {
      "epoch": 9.23154744340935,
      "grad_norm": 1.7348084449768066,
      "learning_rate": 6.415555460275265e-05,
      "loss": 1.9661,
      "step": 219000
    },
    {
      "epoch": 9.252624035745901,
      "grad_norm": 1.2335606813430786,
      "learning_rate": 6.406980234103675e-05,
      "loss": 1.9683,
      "step": 219500
    },
    {
      "epoch": 9.273700628082452,
      "grad_norm": 1.3807036876678467,
      "learning_rate": 6.398405007932085e-05,
      "loss": 1.9682,
      "step": 220000
    },
    {
      "epoch": 9.273700628082452,
      "eval_accuracy": 0.6616191565268488,
      "eval_loss": 1.6792075634002686,
      "eval_runtime": 169.2781,
      "eval_samples_per_second": 713.063,
      "eval_steps_per_second": 7.432,
      "step": 220000
    },
    {
      "epoch": 9.294777220419002,
      "grad_norm": 1.2907365560531616,
      "learning_rate": 6.389829781760495e-05,
      "loss": 1.9647,
      "step": 220500
    },
    {
      "epoch": 9.315853812755554,
      "grad_norm": 1.6088557243347168,
      "learning_rate": 6.381254555588903e-05,
      "loss": 1.9707,
      "step": 221000
    },
    {
      "epoch": 9.336930405092104,
      "grad_norm": 1.1321697235107422,
      "learning_rate": 6.372679329417314e-05,
      "loss": 1.9667,
      "step": 221500
    },
    {
      "epoch": 9.358006997428657,
      "grad_norm": 1.463560938835144,
      "learning_rate": 6.364104103245724e-05,
      "loss": 1.9637,
      "step": 222000
    },
    {
      "epoch": 9.379083589765207,
      "grad_norm": 1.4119582176208496,
      "learning_rate": 6.355528877074132e-05,
      "loss": 1.9673,
      "step": 222500
    },
    {
      "epoch": 9.400160182101757,
      "grad_norm": 1.5491584539413452,
      "learning_rate": 6.346953650902542e-05,
      "loss": 1.9649,
      "step": 223000
    },
    {
      "epoch": 9.42123677443831,
      "grad_norm": 2.4527153968811035,
      "learning_rate": 6.338378424730954e-05,
      "loss": 1.9645,
      "step": 223500
    },
    {
      "epoch": 9.44231336677486,
      "grad_norm": 1.5344223976135254,
      "learning_rate": 6.329803198559362e-05,
      "loss": 1.9629,
      "step": 224000
    },
    {
      "epoch": 9.46338995911141,
      "grad_norm": 2.139519453048706,
      "learning_rate": 6.321227972387772e-05,
      "loss": 1.9642,
      "step": 224500
    },
    {
      "epoch": 9.484466551447962,
      "grad_norm": 1.6216076612472534,
      "learning_rate": 6.312652746216182e-05,
      "loss": 1.9618,
      "step": 225000
    },
    {
      "epoch": 9.505543143784513,
      "grad_norm": 1.2440407276153564,
      "learning_rate": 6.304077520044591e-05,
      "loss": 1.9695,
      "step": 225500
    },
    {
      "epoch": 9.526619736121065,
      "grad_norm": 2.167052745819092,
      "learning_rate": 6.295502293873001e-05,
      "loss": 1.9607,
      "step": 226000
    },
    {
      "epoch": 9.547696328457615,
      "grad_norm": 1.456194519996643,
      "learning_rate": 6.286927067701411e-05,
      "loss": 1.9625,
      "step": 226500
    },
    {
      "epoch": 9.568772920794165,
      "grad_norm": 1.691516637802124,
      "learning_rate": 6.278351841529821e-05,
      "loss": 1.9624,
      "step": 227000
    },
    {
      "epoch": 9.589849513130718,
      "grad_norm": 1.498573899269104,
      "learning_rate": 6.269776615358231e-05,
      "loss": 1.9615,
      "step": 227500
    },
    {
      "epoch": 9.610926105467268,
      "grad_norm": 1.5033460855484009,
      "learning_rate": 6.26120138918664e-05,
      "loss": 1.958,
      "step": 228000
    },
    {
      "epoch": 9.63200269780382,
      "grad_norm": 1.2670778036117554,
      "learning_rate": 6.252626163015049e-05,
      "loss": 1.9598,
      "step": 228500
    },
    {
      "epoch": 9.65307929014037,
      "grad_norm": 1.184558391571045,
      "learning_rate": 6.24405093684346e-05,
      "loss": 1.9597,
      "step": 229000
    },
    {
      "epoch": 9.67415588247692,
      "grad_norm": 1.759701132774353,
      "learning_rate": 6.23547571067187e-05,
      "loss": 1.9624,
      "step": 229500
    },
    {
      "epoch": 9.695232474813473,
      "grad_norm": 1.7340415716171265,
      "learning_rate": 6.226900484500279e-05,
      "loss": 1.9548,
      "step": 230000
    },
    {
      "epoch": 9.695232474813473,
      "eval_accuracy": 0.6629032806449506,
      "eval_loss": 1.671502947807312,
      "eval_runtime": 169.0108,
      "eval_samples_per_second": 714.191,
      "eval_steps_per_second": 7.443,
      "step": 230000
    },
    {
      "epoch": 9.716309067150023,
      "grad_norm": 1.6396135091781616,
      "learning_rate": 6.218325258328688e-05,
      "loss": 1.9661,
      "step": 230500
    },
    {
      "epoch": 9.737385659486574,
      "grad_norm": 1.9293781518936157,
      "learning_rate": 6.209750032157098e-05,
      "loss": 1.964,
      "step": 231000
    },
    {
      "epoch": 9.758462251823126,
      "grad_norm": 1.8499342203140259,
      "learning_rate": 6.201174805985508e-05,
      "loss": 1.9612,
      "step": 231500
    },
    {
      "epoch": 9.779538844159676,
      "grad_norm": 2.275282859802246,
      "learning_rate": 6.192599579813918e-05,
      "loss": 1.957,
      "step": 232000
    },
    {
      "epoch": 9.800615436496226,
      "grad_norm": 1.2911840677261353,
      "learning_rate": 6.184024353642328e-05,
      "loss": 1.9572,
      "step": 232500
    },
    {
      "epoch": 9.821692028832778,
      "grad_norm": 1.2453526258468628,
      "learning_rate": 6.175449127470738e-05,
      "loss": 1.9564,
      "step": 233000
    },
    {
      "epoch": 9.842768621169329,
      "grad_norm": 1.502855896949768,
      "learning_rate": 6.166873901299147e-05,
      "loss": 1.9601,
      "step": 233500
    },
    {
      "epoch": 9.863845213505881,
      "grad_norm": 2.142183303833008,
      "learning_rate": 6.158298675127557e-05,
      "loss": 1.9542,
      "step": 234000
    },
    {
      "epoch": 9.884921805842431,
      "grad_norm": 1.496605634689331,
      "learning_rate": 6.149723448955966e-05,
      "loss": 1.9597,
      "step": 234500
    },
    {
      "epoch": 9.905998398178982,
      "grad_norm": 1.1376317739486694,
      "learning_rate": 6.141148222784377e-05,
      "loss": 1.9552,
      "step": 235000
    },
    {
      "epoch": 9.927074990515534,
      "grad_norm": 1.5101569890975952,
      "learning_rate": 6.132572996612787e-05,
      "loss": 1.9555,
      "step": 235500
    },
    {
      "epoch": 9.948151582852084,
      "grad_norm": 1.3703885078430176,
      "learning_rate": 6.123997770441195e-05,
      "loss": 1.9577,
      "step": 236000
    },
    {
      "epoch": 9.969228175188636,
      "grad_norm": 1.5332629680633545,
      "learning_rate": 6.115422544269605e-05,
      "loss": 1.961,
      "step": 236500
    },
    {
      "epoch": 9.990304767525187,
      "grad_norm": 1.4347738027572632,
      "learning_rate": 6.106847318098015e-05,
      "loss": 1.9593,
      "step": 237000
    },
    {
      "epoch": 10.011381359861737,
      "grad_norm": 1.7119171619415283,
      "learning_rate": 6.098272091926425e-05,
      "loss": 1.9493,
      "step": 237500
    },
    {
      "epoch": 10.032457952198289,
      "grad_norm": 1.5202869176864624,
      "learning_rate": 6.0896968657548345e-05,
      "loss": 1.9464,
      "step": 238000
    },
    {
      "epoch": 10.05353454453484,
      "grad_norm": 1.6234852075576782,
      "learning_rate": 6.081121639583244e-05,
      "loss": 1.9401,
      "step": 238500
    },
    {
      "epoch": 10.07461113687139,
      "grad_norm": 1.8282116651535034,
      "learning_rate": 6.072546413411654e-05,
      "loss": 1.9413,
      "step": 239000
    },
    {
      "epoch": 10.095687729207942,
      "grad_norm": 1.597546935081482,
      "learning_rate": 6.063971187240064e-05,
      "loss": 1.9436,
      "step": 239500
    },
    {
      "epoch": 10.116764321544492,
      "grad_norm": 1.9651734828948975,
      "learning_rate": 6.055395961068473e-05,
      "loss": 1.9364,
      "step": 240000
    },
    {
      "epoch": 10.116764321544492,
      "eval_accuracy": 0.6642122783009812,
      "eval_loss": 1.6793729066848755,
      "eval_runtime": 169.3504,
      "eval_samples_per_second": 712.759,
      "eval_steps_per_second": 7.428,
      "step": 240000
    },
    {
      "epoch": 10.137840913881044,
      "grad_norm": 1.4453253746032715,
      "learning_rate": 6.046820734896883e-05,
      "loss": 1.9416,
      "step": 240500
    },
    {
      "epoch": 10.158917506217595,
      "grad_norm": 1.6494587659835815,
      "learning_rate": 6.038245508725293e-05,
      "loss": 1.9434,
      "step": 241000
    },
    {
      "epoch": 10.179994098554145,
      "grad_norm": 1.5628561973571777,
      "learning_rate": 6.029670282553702e-05,
      "loss": 1.9425,
      "step": 241500
    },
    {
      "epoch": 10.201070690890697,
      "grad_norm": 1.2367149591445923,
      "learning_rate": 6.0210950563821125e-05,
      "loss": 1.9466,
      "step": 242000
    },
    {
      "epoch": 10.222147283227248,
      "grad_norm": 1.6526355743408203,
      "learning_rate": 6.0125198302105223e-05,
      "loss": 1.9415,
      "step": 242500
    },
    {
      "epoch": 10.2432238755638,
      "grad_norm": 2.3241097927093506,
      "learning_rate": 6.0039446040389315e-05,
      "loss": 1.9427,
      "step": 243000
    },
    {
      "epoch": 10.26430046790035,
      "grad_norm": 2.2323055267333984,
      "learning_rate": 5.995369377867341e-05,
      "loss": 1.9461,
      "step": 243500
    },
    {
      "epoch": 10.2853770602369,
      "grad_norm": 1.2214744091033936,
      "learning_rate": 5.986794151695752e-05,
      "loss": 1.9362,
      "step": 244000
    },
    {
      "epoch": 10.306453652573452,
      "grad_norm": 2.0853614807128906,
      "learning_rate": 5.97821892552416e-05,
      "loss": 1.9436,
      "step": 244500
    },
    {
      "epoch": 10.327530244910003,
      "grad_norm": 2.4493536949157715,
      "learning_rate": 5.969643699352571e-05,
      "loss": 1.9467,
      "step": 245000
    },
    {
      "epoch": 10.348606837246553,
      "grad_norm": 1.7084174156188965,
      "learning_rate": 5.9610684731809806e-05,
      "loss": 1.9437,
      "step": 245500
    },
    {
      "epoch": 10.369683429583105,
      "grad_norm": 2.1897871494293213,
      "learning_rate": 5.95249324700939e-05,
      "loss": 1.9387,
      "step": 246000
    },
    {
      "epoch": 10.390760021919656,
      "grad_norm": 2.57694149017334,
      "learning_rate": 5.9439180208377996e-05,
      "loss": 1.9417,
      "step": 246500
    },
    {
      "epoch": 10.411836614256208,
      "grad_norm": 1.7761352062225342,
      "learning_rate": 5.93534279466621e-05,
      "loss": 1.944,
      "step": 247000
    },
    {
      "epoch": 10.432913206592758,
      "grad_norm": 1.4742056131362915,
      "learning_rate": 5.9267675684946186e-05,
      "loss": 1.9451,
      "step": 247500
    },
    {
      "epoch": 10.453989798929308,
      "grad_norm": 1.3637433052062988,
      "learning_rate": 5.918192342323029e-05,
      "loss": 1.9366,
      "step": 248000
    },
    {
      "epoch": 10.47506639126586,
      "grad_norm": 1.5931974649429321,
      "learning_rate": 5.909617116151439e-05,
      "loss": 1.9393,
      "step": 248500
    },
    {
      "epoch": 10.496142983602411,
      "grad_norm": 1.2444241046905518,
      "learning_rate": 5.901041889979848e-05,
      "loss": 1.9393,
      "step": 249000
    },
    {
      "epoch": 10.517219575938963,
      "grad_norm": 1.328933835029602,
      "learning_rate": 5.892466663808258e-05,
      "loss": 1.9422,
      "step": 249500
    },
    {
      "epoch": 10.538296168275513,
      "grad_norm": 1.3554414510726929,
      "learning_rate": 5.8838914376366684e-05,
      "loss": 1.9411,
      "step": 250000
    },
    {
      "epoch": 10.538296168275513,
      "eval_accuracy": 0.665284821537416,
      "eval_loss": 1.6578097343444824,
      "eval_runtime": 169.1778,
      "eval_samples_per_second": 713.486,
      "eval_steps_per_second": 7.436,
      "step": 250000
    },
    {
      "epoch": 10.559372760612064,
      "grad_norm": 1.509717345237732,
      "learning_rate": 5.8753162114650776e-05,
      "loss": 1.948,
      "step": 250500
    },
    {
      "epoch": 10.580449352948616,
      "grad_norm": 1.9658507108688354,
      "learning_rate": 5.8667409852934874e-05,
      "loss": 1.9378,
      "step": 251000
    },
    {
      "epoch": 10.601525945285166,
      "grad_norm": 1.7557775974273682,
      "learning_rate": 5.858165759121897e-05,
      "loss": 1.9349,
      "step": 251500
    },
    {
      "epoch": 10.622602537621717,
      "grad_norm": 1.417623519897461,
      "learning_rate": 5.8495905329503064e-05,
      "loss": 1.94,
      "step": 252000
    },
    {
      "epoch": 10.643679129958269,
      "grad_norm": 1.5471104383468628,
      "learning_rate": 5.841015306778716e-05,
      "loss": 1.9401,
      "step": 252500
    },
    {
      "epoch": 10.664755722294819,
      "grad_norm": 1.5702890157699585,
      "learning_rate": 5.832440080607127e-05,
      "loss": 1.9381,
      "step": 253000
    },
    {
      "epoch": 10.685832314631371,
      "grad_norm": 1.580796480178833,
      "learning_rate": 5.823864854435536e-05,
      "loss": 1.9355,
      "step": 253500
    },
    {
      "epoch": 10.706908906967922,
      "grad_norm": 2.5696029663085938,
      "learning_rate": 5.815289628263946e-05,
      "loss": 1.9371,
      "step": 254000
    },
    {
      "epoch": 10.727985499304472,
      "grad_norm": 2.5817933082580566,
      "learning_rate": 5.8067144020923556e-05,
      "loss": 1.9398,
      "step": 254500
    },
    {
      "epoch": 10.749062091641024,
      "grad_norm": 1.4466750621795654,
      "learning_rate": 5.798139175920765e-05,
      "loss": 1.938,
      "step": 255000
    },
    {
      "epoch": 10.770138683977574,
      "grad_norm": 1.384232759475708,
      "learning_rate": 5.789563949749175e-05,
      "loss": 1.9405,
      "step": 255500
    },
    {
      "epoch": 10.791215276314126,
      "grad_norm": 2.1461946964263916,
      "learning_rate": 5.780988723577585e-05,
      "loss": 1.9383,
      "step": 256000
    },
    {
      "epoch": 10.812291868650677,
      "grad_norm": 2.096677303314209,
      "learning_rate": 5.772413497405994e-05,
      "loss": 1.9408,
      "step": 256500
    },
    {
      "epoch": 10.833368460987227,
      "grad_norm": 1.4637855291366577,
      "learning_rate": 5.763838271234404e-05,
      "loss": 1.9367,
      "step": 257000
    },
    {
      "epoch": 10.85444505332378,
      "grad_norm": 1.8804148435592651,
      "learning_rate": 5.7552630450628146e-05,
      "loss": 1.9371,
      "step": 257500
    },
    {
      "epoch": 10.87552164566033,
      "grad_norm": 2.038516044616699,
      "learning_rate": 5.746687818891223e-05,
      "loss": 1.9382,
      "step": 258000
    },
    {
      "epoch": 10.89659823799688,
      "grad_norm": 1.3879210948944092,
      "learning_rate": 5.7381125927196335e-05,
      "loss": 1.9358,
      "step": 258500
    },
    {
      "epoch": 10.917674830333432,
      "grad_norm": 1.3346680402755737,
      "learning_rate": 5.7295373665480434e-05,
      "loss": 1.9363,
      "step": 259000
    },
    {
      "epoch": 10.938751422669982,
      "grad_norm": 1.507238507270813,
      "learning_rate": 5.7209621403764525e-05,
      "loss": 1.935,
      "step": 259500
    },
    {
      "epoch": 10.959828015006533,
      "grad_norm": 2.036215305328369,
      "learning_rate": 5.7123869142048624e-05,
      "loss": 1.9335,
      "step": 260000
    },
    {
      "epoch": 10.959828015006533,
      "eval_accuracy": 0.6665323526699005,
      "eval_loss": 1.6482259035110474,
      "eval_runtime": 169.2534,
      "eval_samples_per_second": 713.167,
      "eval_steps_per_second": 7.433,
      "step": 260000
    },
    {
      "epoch": 10.980904607343085,
      "grad_norm": 1.334544062614441,
      "learning_rate": 5.703811688033273e-05,
      "loss": 1.9314,
      "step": 260500
    },
    {
      "epoch": 11.001981199679635,
      "grad_norm": 1.331507921218872,
      "learning_rate": 5.695236461861681e-05,
      "loss": 1.9312,
      "step": 261000
    },
    {
      "epoch": 11.023057792016187,
      "grad_norm": 1.4269322156906128,
      "learning_rate": 5.686661235690092e-05,
      "loss": 1.9185,
      "step": 261500
    },
    {
      "epoch": 11.044134384352738,
      "grad_norm": 1.2650457620620728,
      "learning_rate": 5.678086009518502e-05,
      "loss": 1.9263,
      "step": 262000
    },
    {
      "epoch": 11.065210976689288,
      "grad_norm": 1.9887479543685913,
      "learning_rate": 5.669510783346911e-05,
      "loss": 1.9237,
      "step": 262500
    },
    {
      "epoch": 11.08628756902584,
      "grad_norm": 1.2889591455459595,
      "learning_rate": 5.6609355571753207e-05,
      "loss": 1.921,
      "step": 263000
    },
    {
      "epoch": 11.10736416136239,
      "grad_norm": 2.539132595062256,
      "learning_rate": 5.652360331003731e-05,
      "loss": 1.9259,
      "step": 263500
    },
    {
      "epoch": 11.128440753698943,
      "grad_norm": 1.836949110031128,
      "learning_rate": 5.6437851048321396e-05,
      "loss": 1.921,
      "step": 264000
    },
    {
      "epoch": 11.149517346035493,
      "grad_norm": 1.9071457386016846,
      "learning_rate": 5.63520987866055e-05,
      "loss": 1.9235,
      "step": 264500
    },
    {
      "epoch": 11.170593938372043,
      "grad_norm": 2.346614360809326,
      "learning_rate": 5.62663465248896e-05,
      "loss": 1.9212,
      "step": 265000
    },
    {
      "epoch": 11.191670530708596,
      "grad_norm": 2.509633779525757,
      "learning_rate": 5.618059426317369e-05,
      "loss": 1.9208,
      "step": 265500
    },
    {
      "epoch": 11.212747123045146,
      "grad_norm": 1.6268631219863892,
      "learning_rate": 5.609484200145779e-05,
      "loss": 1.9233,
      "step": 266000
    },
    {
      "epoch": 11.233823715381696,
      "grad_norm": 1.5200402736663818,
      "learning_rate": 5.6009089739741895e-05,
      "loss": 1.9231,
      "step": 266500
    },
    {
      "epoch": 11.254900307718248,
      "grad_norm": 1.5498969554901123,
      "learning_rate": 5.5923337478025986e-05,
      "loss": 1.922,
      "step": 267000
    },
    {
      "epoch": 11.275976900054799,
      "grad_norm": 1.3444725275039673,
      "learning_rate": 5.5837585216310085e-05,
      "loss": 1.9261,
      "step": 267500
    },
    {
      "epoch": 11.29705349239135,
      "grad_norm": 1.5418438911437988,
      "learning_rate": 5.575183295459418e-05,
      "loss": 1.9191,
      "step": 268000
    },
    {
      "epoch": 11.318130084727901,
      "grad_norm": 2.069183111190796,
      "learning_rate": 5.5666080692878274e-05,
      "loss": 1.9219,
      "step": 268500
    },
    {
      "epoch": 11.339206677064452,
      "grad_norm": 1.2002756595611572,
      "learning_rate": 5.558032843116238e-05,
      "loss": 1.9223,
      "step": 269000
    },
    {
      "epoch": 11.360283269401004,
      "grad_norm": 2.047496795654297,
      "learning_rate": 5.549457616944648e-05,
      "loss": 1.9224,
      "step": 269500
    },
    {
      "epoch": 11.381359861737554,
      "grad_norm": 1.9832671880722046,
      "learning_rate": 5.540882390773057e-05,
      "loss": 1.9191,
      "step": 270000
    },
    {
      "epoch": 11.381359861737554,
      "eval_accuracy": 0.6670470449497112,
      "eval_loss": 1.6511168479919434,
      "eval_runtime": 169.2479,
      "eval_samples_per_second": 713.19,
      "eval_steps_per_second": 7.433,
      "step": 270000
    },
    {
      "epoch": 11.402436454074106,
      "grad_norm": 2.1018998622894287,
      "learning_rate": 5.532307164601467e-05,
      "loss": 1.9238,
      "step": 270500
    },
    {
      "epoch": 11.423513046410656,
      "grad_norm": 1.6960811614990234,
      "learning_rate": 5.5237319384298766e-05,
      "loss": 1.9193,
      "step": 271000
    },
    {
      "epoch": 11.444589638747207,
      "grad_norm": 1.4490071535110474,
      "learning_rate": 5.515156712258286e-05,
      "loss": 1.9188,
      "step": 271500
    },
    {
      "epoch": 11.465666231083759,
      "grad_norm": 1.5524721145629883,
      "learning_rate": 5.506581486086696e-05,
      "loss": 1.9204,
      "step": 272000
    },
    {
      "epoch": 11.48674282342031,
      "grad_norm": 1.5838063955307007,
      "learning_rate": 5.498006259915105e-05,
      "loss": 1.9184,
      "step": 272500
    },
    {
      "epoch": 11.50781941575686,
      "grad_norm": 1.3861750364303589,
      "learning_rate": 5.489431033743515e-05,
      "loss": 1.917,
      "step": 273000
    },
    {
      "epoch": 11.528896008093412,
      "grad_norm": 1.7460421323776245,
      "learning_rate": 5.480855807571925e-05,
      "loss": 1.9218,
      "step": 273500
    },
    {
      "epoch": 11.549972600429962,
      "grad_norm": 3.1658010482788086,
      "learning_rate": 5.472280581400334e-05,
      "loss": 1.921,
      "step": 274000
    },
    {
      "epoch": 11.571049192766514,
      "grad_norm": 1.8079376220703125,
      "learning_rate": 5.463705355228744e-05,
      "loss": 1.9164,
      "step": 274500
    },
    {
      "epoch": 11.592125785103065,
      "grad_norm": 3.3131113052368164,
      "learning_rate": 5.4551301290571546e-05,
      "loss": 1.9235,
      "step": 275000
    },
    {
      "epoch": 11.613202377439615,
      "grad_norm": 2.1756768226623535,
      "learning_rate": 5.446554902885563e-05,
      "loss": 1.9189,
      "step": 275500
    },
    {
      "epoch": 11.634278969776167,
      "grad_norm": 1.3388465642929077,
      "learning_rate": 5.4379796767139736e-05,
      "loss": 1.9158,
      "step": 276000
    },
    {
      "epoch": 11.655355562112717,
      "grad_norm": 1.4720878601074219,
      "learning_rate": 5.4294044505423834e-05,
      "loss": 1.9179,
      "step": 276500
    },
    {
      "epoch": 11.67643215444927,
      "grad_norm": 2.37357497215271,
      "learning_rate": 5.4208292243707925e-05,
      "loss": 1.919,
      "step": 277000
    },
    {
      "epoch": 11.69750874678582,
      "grad_norm": 1.411412000656128,
      "learning_rate": 5.4122539981992024e-05,
      "loss": 1.9205,
      "step": 277500
    },
    {
      "epoch": 11.71858533912237,
      "grad_norm": 1.7797832489013672,
      "learning_rate": 5.403678772027613e-05,
      "loss": 1.9147,
      "step": 278000
    },
    {
      "epoch": 11.739661931458922,
      "grad_norm": 1.4294768571853638,
      "learning_rate": 5.395103545856022e-05,
      "loss": 1.919,
      "step": 278500
    },
    {
      "epoch": 11.760738523795473,
      "grad_norm": 1.5599967241287231,
      "learning_rate": 5.386528319684432e-05,
      "loss": 1.917,
      "step": 279000
    },
    {
      "epoch": 11.781815116132023,
      "grad_norm": 2.1733498573303223,
      "learning_rate": 5.377953093512842e-05,
      "loss": 1.9148,
      "step": 279500
    },
    {
      "epoch": 11.802891708468575,
      "grad_norm": 1.4645706415176392,
      "learning_rate": 5.369377867341251e-05,
      "loss": 1.9194,
      "step": 280000
    },
    {
      "epoch": 11.802891708468575,
      "eval_accuracy": 0.6684710960756258,
      "eval_loss": 1.6416321992874146,
      "eval_runtime": 168.9745,
      "eval_samples_per_second": 714.344,
      "eval_steps_per_second": 7.445,
      "step": 280000
    },
    {
      "epoch": 11.823968300805126,
      "grad_norm": 1.4757664203643799,
      "learning_rate": 5.3608026411696614e-05,
      "loss": 1.9226,
      "step": 280500
    },
    {
      "epoch": 11.845044893141678,
      "grad_norm": 2.0871543884277344,
      "learning_rate": 5.352227414998071e-05,
      "loss": 1.9126,
      "step": 281000
    },
    {
      "epoch": 11.866121485478228,
      "grad_norm": 1.7234623432159424,
      "learning_rate": 5.34365218882648e-05,
      "loss": 1.9157,
      "step": 281500
    },
    {
      "epoch": 11.887198077814778,
      "grad_norm": 1.6794915199279785,
      "learning_rate": 5.33507696265489e-05,
      "loss": 1.9101,
      "step": 282000
    },
    {
      "epoch": 11.90827467015133,
      "grad_norm": 2.08085298538208,
      "learning_rate": 5.3265017364833e-05,
      "loss": 1.9208,
      "step": 282500
    },
    {
      "epoch": 11.92935126248788,
      "grad_norm": 1.239192008972168,
      "learning_rate": 5.317926510311709e-05,
      "loss": 1.9162,
      "step": 283000
    },
    {
      "epoch": 11.950427854824433,
      "grad_norm": 1.406118631362915,
      "learning_rate": 5.3093512841401197e-05,
      "loss": 1.9141,
      "step": 283500
    },
    {
      "epoch": 11.971504447160983,
      "grad_norm": 1.9544838666915894,
      "learning_rate": 5.3007760579685295e-05,
      "loss": 1.9201,
      "step": 284000
    },
    {
      "epoch": 11.992581039497534,
      "grad_norm": 1.8367359638214111,
      "learning_rate": 5.2922008317969386e-05,
      "loss": 1.9207,
      "step": 284500
    },
    {
      "epoch": 12.013657631834086,
      "grad_norm": 1.740451455116272,
      "learning_rate": 5.2836256056253485e-05,
      "loss": 1.9057,
      "step": 285000
    },
    {
      "epoch": 12.034734224170636,
      "grad_norm": 1.3087714910507202,
      "learning_rate": 5.275050379453759e-05,
      "loss": 1.9058,
      "step": 285500
    },
    {
      "epoch": 12.055810816507186,
      "grad_norm": 2.3969788551330566,
      "learning_rate": 5.2664751532821675e-05,
      "loss": 1.9022,
      "step": 286000
    },
    {
      "epoch": 12.076887408843739,
      "grad_norm": 1.760349988937378,
      "learning_rate": 5.257899927110578e-05,
      "loss": 1.9008,
      "step": 286500
    },
    {
      "epoch": 12.097964001180289,
      "grad_norm": 1.8947540521621704,
      "learning_rate": 5.249324700938988e-05,
      "loss": 1.8987,
      "step": 287000
    },
    {
      "epoch": 12.119040593516841,
      "grad_norm": 1.6933043003082275,
      "learning_rate": 5.240749474767397e-05,
      "loss": 1.9009,
      "step": 287500
    },
    {
      "epoch": 12.140117185853391,
      "grad_norm": 1.6944429874420166,
      "learning_rate": 5.232174248595807e-05,
      "loss": 1.9042,
      "step": 288000
    },
    {
      "epoch": 12.161193778189942,
      "grad_norm": 2.708965301513672,
      "learning_rate": 5.223599022424217e-05,
      "loss": 1.9045,
      "step": 288500
    },
    {
      "epoch": 12.182270370526494,
      "grad_norm": 1.8792153596878052,
      "learning_rate": 5.215023796252626e-05,
      "loss": 1.9073,
      "step": 289000
    },
    {
      "epoch": 12.203346962863044,
      "grad_norm": 1.8549127578735352,
      "learning_rate": 5.206448570081036e-05,
      "loss": 1.8995,
      "step": 289500
    },
    {
      "epoch": 12.224423555199595,
      "grad_norm": 2.6252317428588867,
      "learning_rate": 5.197873343909446e-05,
      "loss": 1.9004,
      "step": 290000
    },
    {
      "epoch": 12.224423555199595,
      "eval_accuracy": 0.6690299715472139,
      "eval_loss": 1.6557669639587402,
      "eval_runtime": 169.4484,
      "eval_samples_per_second": 712.347,
      "eval_steps_per_second": 7.424,
      "step": 290000
    },
    {
      "epoch": 12.245500147536147,
      "grad_norm": 1.9704797267913818,
      "learning_rate": 5.189298117737855e-05,
      "loss": 1.9002,
      "step": 290500
    },
    {
      "epoch": 12.266576739872697,
      "grad_norm": 1.6016666889190674,
      "learning_rate": 5.180722891566265e-05,
      "loss": 1.9002,
      "step": 291000
    },
    {
      "epoch": 12.28765333220925,
      "grad_norm": 2.216820001602173,
      "learning_rate": 5.1721476653946756e-05,
      "loss": 1.8979,
      "step": 291500
    },
    {
      "epoch": 12.3087299245458,
      "grad_norm": 2.360379219055176,
      "learning_rate": 5.163572439223085e-05,
      "loss": 1.9033,
      "step": 292000
    },
    {
      "epoch": 12.32980651688235,
      "grad_norm": 2.2686407566070557,
      "learning_rate": 5.1549972130514946e-05,
      "loss": 1.905,
      "step": 292500
    },
    {
      "epoch": 12.350883109218902,
      "grad_norm": 2.196948766708374,
      "learning_rate": 5.1464219868799044e-05,
      "loss": 1.9013,
      "step": 293000
    },
    {
      "epoch": 12.371959701555452,
      "grad_norm": 1.946319580078125,
      "learning_rate": 5.1378467607083136e-05,
      "loss": 1.8987,
      "step": 293500
    },
    {
      "epoch": 12.393036293892003,
      "grad_norm": 1.927105188369751,
      "learning_rate": 5.1292715345367234e-05,
      "loss": 1.9027,
      "step": 294000
    },
    {
      "epoch": 12.414112886228555,
      "grad_norm": 2.062389612197876,
      "learning_rate": 5.120696308365134e-05,
      "loss": 1.8976,
      "step": 294500
    },
    {
      "epoch": 12.435189478565105,
      "grad_norm": 2.0816409587860107,
      "learning_rate": 5.112121082193543e-05,
      "loss": 1.9051,
      "step": 295000
    },
    {
      "epoch": 12.456266070901657,
      "grad_norm": 1.4892915487289429,
      "learning_rate": 5.103545856021953e-05,
      "loss": 1.9034,
      "step": 295500
    },
    {
      "epoch": 12.477342663238208,
      "grad_norm": 1.8691354990005493,
      "learning_rate": 5.094970629850363e-05,
      "loss": 1.9014,
      "step": 296000
    },
    {
      "epoch": 12.498419255574758,
      "grad_norm": 1.5349853038787842,
      "learning_rate": 5.086395403678772e-05,
      "loss": 1.9016,
      "step": 296500
    },
    {
      "epoch": 12.51949584791131,
      "grad_norm": 1.5519059896469116,
      "learning_rate": 5.0778201775071824e-05,
      "loss": 1.9088,
      "step": 297000
    },
    {
      "epoch": 12.54057244024786,
      "grad_norm": 1.3968881368637085,
      "learning_rate": 5.069244951335592e-05,
      "loss": 1.9028,
      "step": 297500
    },
    {
      "epoch": 12.561649032584413,
      "grad_norm": 2.2018463611602783,
      "learning_rate": 5.0606697251640014e-05,
      "loss": 1.8987,
      "step": 298000
    },
    {
      "epoch": 12.582725624920963,
      "grad_norm": 2.5705435276031494,
      "learning_rate": 5.052094498992411e-05,
      "loss": 1.9056,
      "step": 298500
    },
    {
      "epoch": 12.603802217257513,
      "grad_norm": 1.6301090717315674,
      "learning_rate": 5.043519272820822e-05,
      "loss": 1.9068,
      "step": 299000
    },
    {
      "epoch": 12.624878809594065,
      "grad_norm": 1.7063623666763306,
      "learning_rate": 5.03494404664923e-05,
      "loss": 1.8979,
      "step": 299500
    },
    {
      "epoch": 12.645955401930616,
      "grad_norm": 1.8784030675888062,
      "learning_rate": 5.026368820477641e-05,
      "loss": 1.8987,
      "step": 300000
    },
    {
      "epoch": 12.645955401930616,
      "eval_accuracy": 0.6698791929444358,
      "eval_loss": 1.6351250410079956,
      "eval_runtime": 168.781,
      "eval_samples_per_second": 715.164,
      "eval_steps_per_second": 7.453,
      "step": 300000
    },
    {
      "epoch": 12.667031994267166,
      "grad_norm": 1.5910873413085938,
      "learning_rate": 5.0177935943060505e-05,
      "loss": 1.8947,
      "step": 300500
    },
    {
      "epoch": 12.688108586603718,
      "grad_norm": 1.5097805261611938,
      "learning_rate": 5.00921836813446e-05,
      "loss": 1.9044,
      "step": 301000
    },
    {
      "epoch": 12.709185178940269,
      "grad_norm": 1.6666632890701294,
      "learning_rate": 5.0006431419628695e-05,
      "loss": 1.901,
      "step": 301500
    },
    {
      "epoch": 12.73026177127682,
      "grad_norm": 1.591704249382019,
      "learning_rate": 4.992067915791279e-05,
      "loss": 1.9013,
      "step": 302000
    },
    {
      "epoch": 12.751338363613371,
      "grad_norm": 1.4160646200180054,
      "learning_rate": 4.9834926896196885e-05,
      "loss": 1.8968,
      "step": 302500
    },
    {
      "epoch": 12.772414955949921,
      "grad_norm": 1.357804536819458,
      "learning_rate": 4.974917463448099e-05,
      "loss": 1.9029,
      "step": 303000
    },
    {
      "epoch": 12.793491548286474,
      "grad_norm": 1.3279496431350708,
      "learning_rate": 4.966342237276508e-05,
      "loss": 1.9036,
      "step": 303500
    },
    {
      "epoch": 12.814568140623024,
      "grad_norm": 1.736592411994934,
      "learning_rate": 4.957767011104918e-05,
      "loss": 1.9006,
      "step": 304000
    },
    {
      "epoch": 12.835644732959576,
      "grad_norm": 1.715206265449524,
      "learning_rate": 4.949191784933328e-05,
      "loss": 1.9005,
      "step": 304500
    },
    {
      "epoch": 12.856721325296126,
      "grad_norm": 2.194636344909668,
      "learning_rate": 4.9406165587617376e-05,
      "loss": 1.897,
      "step": 305000
    },
    {
      "epoch": 12.877797917632677,
      "grad_norm": 1.7038402557373047,
      "learning_rate": 4.932041332590147e-05,
      "loss": 1.904,
      "step": 305500
    },
    {
      "epoch": 12.898874509969229,
      "grad_norm": 2.020873785018921,
      "learning_rate": 4.923466106418557e-05,
      "loss": 1.8958,
      "step": 306000
    },
    {
      "epoch": 12.91995110230578,
      "grad_norm": 1.5068098306655884,
      "learning_rate": 4.9148908802469665e-05,
      "loss": 1.8969,
      "step": 306500
    },
    {
      "epoch": 12.94102769464233,
      "grad_norm": 2.104900360107422,
      "learning_rate": 4.906315654075376e-05,
      "loss": 1.896,
      "step": 307000
    },
    {
      "epoch": 12.962104286978882,
      "grad_norm": 1.1670674085617065,
      "learning_rate": 4.897740427903786e-05,
      "loss": 1.9037,
      "step": 307500
    },
    {
      "epoch": 12.983180879315432,
      "grad_norm": 1.4832426309585571,
      "learning_rate": 4.889165201732196e-05,
      "loss": 1.9005,
      "step": 308000
    },
    {
      "epoch": 13.004257471651984,
      "grad_norm": 1.8848177194595337,
      "learning_rate": 4.880589975560606e-05,
      "loss": 1.8956,
      "step": 308500
    },
    {
      "epoch": 13.025334063988534,
      "grad_norm": 1.6777116060256958,
      "learning_rate": 4.8720147493890156e-05,
      "loss": 1.882,
      "step": 309000
    },
    {
      "epoch": 13.046410656325085,
      "grad_norm": 1.4964231252670288,
      "learning_rate": 4.8634395232174254e-05,
      "loss": 1.8886,
      "step": 309500
    },
    {
      "epoch": 13.067487248661637,
      "grad_norm": 3.3363170623779297,
      "learning_rate": 4.8548642970458346e-05,
      "loss": 1.8838,
      "step": 310000
    },
    {
      "epoch": 13.067487248661637,
      "eval_accuracy": 0.6707515217933873,
      "eval_loss": 1.63627290725708,
      "eval_runtime": 169.4465,
      "eval_samples_per_second": 712.355,
      "eval_steps_per_second": 7.424,
      "step": 310000
    },
    {
      "epoch": 13.088563840998187,
      "grad_norm": 1.2957303524017334,
      "learning_rate": 4.846289070874245e-05,
      "loss": 1.8848,
      "step": 310500
    },
    {
      "epoch": 13.109640433334738,
      "grad_norm": 1.8072482347488403,
      "learning_rate": 4.837713844702654e-05,
      "loss": 1.8854,
      "step": 311000
    },
    {
      "epoch": 13.13071702567129,
      "grad_norm": 2.4160025119781494,
      "learning_rate": 4.829138618531064e-05,
      "loss": 1.8825,
      "step": 311500
    },
    {
      "epoch": 13.15179361800784,
      "grad_norm": 1.816667079925537,
      "learning_rate": 4.820563392359474e-05,
      "loss": 1.8885,
      "step": 312000
    },
    {
      "epoch": 13.172870210344392,
      "grad_norm": 2.0105299949645996,
      "learning_rate": 4.811988166187884e-05,
      "loss": 1.8849,
      "step": 312500
    },
    {
      "epoch": 13.193946802680943,
      "grad_norm": 1.3057554960250854,
      "learning_rate": 4.803412940016293e-05,
      "loss": 1.8859,
      "step": 313000
    },
    {
      "epoch": 13.215023395017493,
      "grad_norm": 1.8101252317428589,
      "learning_rate": 4.7948377138447034e-05,
      "loss": 1.8879,
      "step": 313500
    },
    {
      "epoch": 13.236099987354045,
      "grad_norm": 2.265117883682251,
      "learning_rate": 4.7862624876731126e-05,
      "loss": 1.8828,
      "step": 314000
    },
    {
      "epoch": 13.257176579690595,
      "grad_norm": 1.527597427368164,
      "learning_rate": 4.7776872615015224e-05,
      "loss": 1.8849,
      "step": 314500
    },
    {
      "epoch": 13.278253172027147,
      "grad_norm": 2.7612435817718506,
      "learning_rate": 4.7691120353299315e-05,
      "loss": 1.8815,
      "step": 315000
    },
    {
      "epoch": 13.299329764363698,
      "grad_norm": 1.1332488059997559,
      "learning_rate": 4.760536809158342e-05,
      "loss": 1.8879,
      "step": 315500
    },
    {
      "epoch": 13.320406356700248,
      "grad_norm": 1.668847918510437,
      "learning_rate": 4.751961582986751e-05,
      "loss": 1.8903,
      "step": 316000
    },
    {
      "epoch": 13.3414829490368,
      "grad_norm": 2.150078773498535,
      "learning_rate": 4.743386356815161e-05,
      "loss": 1.8893,
      "step": 316500
    },
    {
      "epoch": 13.36255954137335,
      "grad_norm": 1.791129231452942,
      "learning_rate": 4.734811130643571e-05,
      "loss": 1.8847,
      "step": 317000
    },
    {
      "epoch": 13.383636133709901,
      "grad_norm": 1.5889842510223389,
      "learning_rate": 4.726235904471981e-05,
      "loss": 1.8838,
      "step": 317500
    },
    {
      "epoch": 13.404712726046453,
      "grad_norm": 2.180495262145996,
      "learning_rate": 4.71766067830039e-05,
      "loss": 1.8867,
      "step": 318000
    },
    {
      "epoch": 13.425789318383003,
      "grad_norm": 1.5035030841827393,
      "learning_rate": 4.7090854521288004e-05,
      "loss": 1.8887,
      "step": 318500
    },
    {
      "epoch": 13.446865910719556,
      "grad_norm": 1.3835151195526123,
      "learning_rate": 4.7005102259572095e-05,
      "loss": 1.8857,
      "step": 319000
    },
    {
      "epoch": 13.467942503056106,
      "grad_norm": 3.066915988922119,
      "learning_rate": 4.6919349997856193e-05,
      "loss": 1.8867,
      "step": 319500
    },
    {
      "epoch": 13.489019095392656,
      "grad_norm": 1.336668610572815,
      "learning_rate": 4.683359773614029e-05,
      "loss": 1.8851,
      "step": 320000
    },
    {
      "epoch": 13.489019095392656,
      "eval_accuracy": 0.6713279957108566,
      "eval_loss": 1.632880687713623,
      "eval_runtime": 169.6029,
      "eval_samples_per_second": 711.698,
      "eval_steps_per_second": 7.417,
      "step": 320000
    },
    {
      "epoch": 13.510095687729208,
      "grad_norm": 1.4284099340438843,
      "learning_rate": 4.674784547442439e-05,
      "loss": 1.8896,
      "step": 320500
    },
    {
      "epoch": 13.531172280065759,
      "grad_norm": 2.2746310234069824,
      "learning_rate": 4.666209321270849e-05,
      "loss": 1.8827,
      "step": 321000
    },
    {
      "epoch": 13.55224887240231,
      "grad_norm": 2.3596248626708984,
      "learning_rate": 4.657634095099259e-05,
      "loss": 1.8848,
      "step": 321500
    },
    {
      "epoch": 13.573325464738861,
      "grad_norm": 2.610645055770874,
      "learning_rate": 4.6490588689276685e-05,
      "loss": 1.8874,
      "step": 322000
    },
    {
      "epoch": 13.594402057075412,
      "grad_norm": 2.0139858722686768,
      "learning_rate": 4.6404836427560777e-05,
      "loss": 1.8867,
      "step": 322500
    },
    {
      "epoch": 13.615478649411964,
      "grad_norm": 2.2823054790496826,
      "learning_rate": 4.6319084165844875e-05,
      "loss": 1.8798,
      "step": 323000
    },
    {
      "epoch": 13.636555241748514,
      "grad_norm": 2.479999303817749,
      "learning_rate": 4.623333190412897e-05,
      "loss": 1.8873,
      "step": 323500
    },
    {
      "epoch": 13.657631834085064,
      "grad_norm": 1.4525541067123413,
      "learning_rate": 4.614757964241307e-05,
      "loss": 1.8801,
      "step": 324000
    },
    {
      "epoch": 13.678708426421617,
      "grad_norm": 1.5941494703292847,
      "learning_rate": 4.606182738069717e-05,
      "loss": 1.8845,
      "step": 324500
    },
    {
      "epoch": 13.699785018758167,
      "grad_norm": 1.5802496671676636,
      "learning_rate": 4.597607511898127e-05,
      "loss": 1.884,
      "step": 325000
    },
    {
      "epoch": 13.720861611094719,
      "grad_norm": 1.5109597444534302,
      "learning_rate": 4.589032285726536e-05,
      "loss": 1.8876,
      "step": 325500
    },
    {
      "epoch": 13.74193820343127,
      "grad_norm": 1.6784693002700806,
      "learning_rate": 4.5804570595549465e-05,
      "loss": 1.8841,
      "step": 326000
    },
    {
      "epoch": 13.76301479576782,
      "grad_norm": 2.3481030464172363,
      "learning_rate": 4.5718818333833556e-05,
      "loss": 1.8873,
      "step": 326500
    },
    {
      "epoch": 13.784091388104372,
      "grad_norm": 2.4491257667541504,
      "learning_rate": 4.5633066072117655e-05,
      "loss": 1.8875,
      "step": 327000
    },
    {
      "epoch": 13.805167980440922,
      "grad_norm": 1.8449732065200806,
      "learning_rate": 4.554731381040175e-05,
      "loss": 1.8787,
      "step": 327500
    },
    {
      "epoch": 13.826244572777473,
      "grad_norm": 2.731442928314209,
      "learning_rate": 4.546156154868585e-05,
      "loss": 1.8836,
      "step": 328000
    },
    {
      "epoch": 13.847321165114025,
      "grad_norm": 2.446242094039917,
      "learning_rate": 4.537580928696994e-05,
      "loss": 1.8854,
      "step": 328500
    },
    {
      "epoch": 13.868397757450575,
      "grad_norm": 1.314347505569458,
      "learning_rate": 4.529005702525405e-05,
      "loss": 1.882,
      "step": 329000
    },
    {
      "epoch": 13.889474349787127,
      "grad_norm": 2.160200357437134,
      "learning_rate": 4.520430476353814e-05,
      "loss": 1.8835,
      "step": 329500
    },
    {
      "epoch": 13.910550942123677,
      "grad_norm": 1.705217957496643,
      "learning_rate": 4.511855250182224e-05,
      "loss": 1.8829,
      "step": 330000
    },
    {
      "epoch": 13.910550942123677,
      "eval_accuracy": 0.6720260328136183,
      "eval_loss": 1.636759638786316,
      "eval_runtime": 169.0615,
      "eval_samples_per_second": 713.977,
      "eval_steps_per_second": 7.441,
      "step": 330000
    },
    {
      "epoch": 13.931627534460228,
      "grad_norm": 1.770910382270813,
      "learning_rate": 4.5032800240106336e-05,
      "loss": 1.8832,
      "step": 330500
    },
    {
      "epoch": 13.95270412679678,
      "grad_norm": 1.4513195753097534,
      "learning_rate": 4.4947047978390434e-05,
      "loss": 1.8898,
      "step": 331000
    },
    {
      "epoch": 13.97378071913333,
      "grad_norm": 2.5424728393554688,
      "learning_rate": 4.4861295716674526e-05,
      "loss": 1.8854,
      "step": 331500
    },
    {
      "epoch": 13.994857311469882,
      "grad_norm": 1.1714924573898315,
      "learning_rate": 4.4775543454958624e-05,
      "loss": 1.8846,
      "step": 332000
    },
    {
      "epoch": 14.015933903806433,
      "grad_norm": 1.7978134155273438,
      "learning_rate": 4.468979119324272e-05,
      "loss": 1.8732,
      "step": 332500
    },
    {
      "epoch": 14.037010496142983,
      "grad_norm": 1.5523267984390259,
      "learning_rate": 4.460403893152682e-05,
      "loss": 1.8724,
      "step": 333000
    },
    {
      "epoch": 14.058087088479535,
      "grad_norm": 3.443971633911133,
      "learning_rate": 4.451828666981092e-05,
      "loss": 1.8717,
      "step": 333500
    },
    {
      "epoch": 14.079163680816086,
      "grad_norm": 1.9500033855438232,
      "learning_rate": 4.443253440809502e-05,
      "loss": 1.8698,
      "step": 334000
    },
    {
      "epoch": 14.100240273152636,
      "grad_norm": 1.4595409631729126,
      "learning_rate": 4.4346782146379116e-05,
      "loss": 1.8724,
      "step": 334500
    },
    {
      "epoch": 14.121316865489188,
      "grad_norm": 1.6145015954971313,
      "learning_rate": 4.426102988466321e-05,
      "loss": 1.8677,
      "step": 335000
    },
    {
      "epoch": 14.142393457825738,
      "grad_norm": 2.1215124130249023,
      "learning_rate": 4.4175277622947305e-05,
      "loss": 1.8635,
      "step": 335500
    },
    {
      "epoch": 14.16347005016229,
      "grad_norm": 1.2612941265106201,
      "learning_rate": 4.4089525361231404e-05,
      "loss": 1.8714,
      "step": 336000
    },
    {
      "epoch": 14.184546642498841,
      "grad_norm": 1.718437910079956,
      "learning_rate": 4.40037730995155e-05,
      "loss": 1.8663,
      "step": 336500
    },
    {
      "epoch": 14.205623234835391,
      "grad_norm": 2.3869683742523193,
      "learning_rate": 4.39180208377996e-05,
      "loss": 1.8715,
      "step": 337000
    },
    {
      "epoch": 14.226699827171943,
      "grad_norm": 1.3405746221542358,
      "learning_rate": 4.38322685760837e-05,
      "loss": 1.8721,
      "step": 337500
    },
    {
      "epoch": 14.247776419508494,
      "grad_norm": 1.8297737836837769,
      "learning_rate": 4.374651631436779e-05,
      "loss": 1.8724,
      "step": 338000
    },
    {
      "epoch": 14.268853011845044,
      "grad_norm": 1.8604480028152466,
      "learning_rate": 4.3660764052651895e-05,
      "loss": 1.8733,
      "step": 338500
    },
    {
      "epoch": 14.289929604181596,
      "grad_norm": 1.6783016920089722,
      "learning_rate": 4.357501179093599e-05,
      "loss": 1.8686,
      "step": 339000
    },
    {
      "epoch": 14.311006196518147,
      "grad_norm": 1.8928593397140503,
      "learning_rate": 4.3489259529220085e-05,
      "loss": 1.87,
      "step": 339500
    },
    {
      "epoch": 14.332082788854699,
      "grad_norm": 1.4946208000183105,
      "learning_rate": 4.3403507267504183e-05,
      "loss": 1.8755,
      "step": 340000
    },
    {
      "epoch": 14.332082788854699,
      "eval_accuracy": 0.6730103035992883,
      "eval_loss": 1.6218472719192505,
      "eval_runtime": 169.579,
      "eval_samples_per_second": 711.798,
      "eval_steps_per_second": 7.418,
      "step": 340000
    },
    {
      "epoch": 14.353159381191249,
      "grad_norm": 2.053776741027832,
      "learning_rate": 4.331775500578828e-05,
      "loss": 1.8681,
      "step": 340500
    },
    {
      "epoch": 14.3742359735278,
      "grad_norm": 1.3982594013214111,
      "learning_rate": 4.323200274407237e-05,
      "loss": 1.8685,
      "step": 341000
    },
    {
      "epoch": 14.395312565864351,
      "grad_norm": 1.919197678565979,
      "learning_rate": 4.314625048235648e-05,
      "loss": 1.8683,
      "step": 341500
    },
    {
      "epoch": 14.416389158200902,
      "grad_norm": 2.465395927429199,
      "learning_rate": 4.306049822064057e-05,
      "loss": 1.8735,
      "step": 342000
    },
    {
      "epoch": 14.437465750537454,
      "grad_norm": 1.6815166473388672,
      "learning_rate": 4.297474595892467e-05,
      "loss": 1.8728,
      "step": 342500
    },
    {
      "epoch": 14.458542342874004,
      "grad_norm": 2.2970895767211914,
      "learning_rate": 4.2888993697208767e-05,
      "loss": 1.8663,
      "step": 343000
    },
    {
      "epoch": 14.479618935210555,
      "grad_norm": 2.4875316619873047,
      "learning_rate": 4.2803241435492865e-05,
      "loss": 1.8701,
      "step": 343500
    },
    {
      "epoch": 14.500695527547107,
      "grad_norm": 1.4955124855041504,
      "learning_rate": 4.2717489173776956e-05,
      "loss": 1.869,
      "step": 344000
    },
    {
      "epoch": 14.521772119883657,
      "grad_norm": 1.4000458717346191,
      "learning_rate": 4.263173691206106e-05,
      "loss": 1.8705,
      "step": 344500
    },
    {
      "epoch": 14.542848712220207,
      "grad_norm": 1.6072543859481812,
      "learning_rate": 4.254598465034515e-05,
      "loss": 1.8705,
      "step": 345000
    },
    {
      "epoch": 14.56392530455676,
      "grad_norm": 1.6247256994247437,
      "learning_rate": 4.246023238862925e-05,
      "loss": 1.8672,
      "step": 345500
    },
    {
      "epoch": 14.58500189689331,
      "grad_norm": 1.390052080154419,
      "learning_rate": 4.237448012691335e-05,
      "loss": 1.8712,
      "step": 346000
    },
    {
      "epoch": 14.606078489229862,
      "grad_norm": 1.1806817054748535,
      "learning_rate": 4.228872786519745e-05,
      "loss": 1.8702,
      "step": 346500
    },
    {
      "epoch": 14.627155081566412,
      "grad_norm": 2.0065128803253174,
      "learning_rate": 4.220297560348154e-05,
      "loss": 1.8732,
      "step": 347000
    },
    {
      "epoch": 14.648231673902963,
      "grad_norm": 2.2869880199432373,
      "learning_rate": 4.2117223341765645e-05,
      "loss": 1.8669,
      "step": 347500
    },
    {
      "epoch": 14.669308266239515,
      "grad_norm": 1.4001230001449585,
      "learning_rate": 4.2031471080049736e-05,
      "loss": 1.8703,
      "step": 348000
    },
    {
      "epoch": 14.690384858576065,
      "grad_norm": 1.6456128358840942,
      "learning_rate": 4.1945718818333834e-05,
      "loss": 1.8693,
      "step": 348500
    },
    {
      "epoch": 14.711461450912616,
      "grad_norm": 1.8096659183502197,
      "learning_rate": 4.185996655661793e-05,
      "loss": 1.8685,
      "step": 349000
    },
    {
      "epoch": 14.732538043249168,
      "grad_norm": 1.8272852897644043,
      "learning_rate": 4.177421429490203e-05,
      "loss": 1.8699,
      "step": 349500
    },
    {
      "epoch": 14.753614635585718,
      "grad_norm": 2.4391791820526123,
      "learning_rate": 4.168846203318613e-05,
      "loss": 1.8662,
      "step": 350000
    },
    {
      "epoch": 14.753614635585718,
      "eval_accuracy": 0.6732898154793913,
      "eval_loss": 1.6259301900863647,
      "eval_runtime": 169.4263,
      "eval_samples_per_second": 712.44,
      "eval_steps_per_second": 7.425,
      "step": 350000
    },
    {
      "epoch": 14.77469122792227,
      "grad_norm": 2.036801815032959,
      "learning_rate": 4.160270977147022e-05,
      "loss": 1.8679,
      "step": 350500
    },
    {
      "epoch": 14.79576782025882,
      "grad_norm": 2.732084274291992,
      "learning_rate": 4.1516957509754326e-05,
      "loss": 1.8733,
      "step": 351000
    },
    {
      "epoch": 14.81684441259537,
      "grad_norm": 1.9076359272003174,
      "learning_rate": 4.143120524803842e-05,
      "loss": 1.866,
      "step": 351500
    },
    {
      "epoch": 14.837921004931923,
      "grad_norm": 2.1984918117523193,
      "learning_rate": 4.1345452986322516e-05,
      "loss": 1.8691,
      "step": 352000
    },
    {
      "epoch": 14.858997597268473,
      "grad_norm": 1.1783524751663208,
      "learning_rate": 4.1259700724606614e-05,
      "loss": 1.867,
      "step": 352500
    },
    {
      "epoch": 14.880074189605025,
      "grad_norm": 1.6603800058364868,
      "learning_rate": 4.117394846289071e-05,
      "loss": 1.8679,
      "step": 353000
    },
    {
      "epoch": 14.901150781941576,
      "grad_norm": 2.517033576965332,
      "learning_rate": 4.1088196201174804e-05,
      "loss": 1.8692,
      "step": 353500
    },
    {
      "epoch": 14.922227374278126,
      "grad_norm": 2.3120429515838623,
      "learning_rate": 4.100244393945891e-05,
      "loss": 1.8687,
      "step": 354000
    },
    {
      "epoch": 14.943303966614678,
      "grad_norm": 1.7784377336502075,
      "learning_rate": 4.0916691677743e-05,
      "loss": 1.8752,
      "step": 354500
    },
    {
      "epoch": 14.964380558951229,
      "grad_norm": 2.3276593685150146,
      "learning_rate": 4.08309394160271e-05,
      "loss": 1.8659,
      "step": 355000
    },
    {
      "epoch": 14.985457151287779,
      "grad_norm": 1.8245887756347656,
      "learning_rate": 4.07451871543112e-05,
      "loss": 1.8693,
      "step": 355500
    },
    {
      "epoch": 15.006533743624331,
      "grad_norm": 2.0235772132873535,
      "learning_rate": 4.0659434892595295e-05,
      "loss": 1.8596,
      "step": 356000
    },
    {
      "epoch": 15.027610335960881,
      "grad_norm": 2.4570260047912598,
      "learning_rate": 4.057368263087939e-05,
      "loss": 1.8542,
      "step": 356500
    },
    {
      "epoch": 15.048686928297434,
      "grad_norm": 1.832669973373413,
      "learning_rate": 4.048793036916349e-05,
      "loss": 1.8557,
      "step": 357000
    },
    {
      "epoch": 15.069763520633984,
      "grad_norm": 1.1062568426132202,
      "learning_rate": 4.0402178107447584e-05,
      "loss": 1.8585,
      "step": 357500
    },
    {
      "epoch": 15.090840112970534,
      "grad_norm": 1.938820242881775,
      "learning_rate": 4.031642584573168e-05,
      "loss": 1.8554,
      "step": 358000
    },
    {
      "epoch": 15.111916705307086,
      "grad_norm": 2.091508150100708,
      "learning_rate": 4.023067358401578e-05,
      "loss": 1.8528,
      "step": 358500
    },
    {
      "epoch": 15.132993297643637,
      "grad_norm": 1.3219631910324097,
      "learning_rate": 4.014492132229988e-05,
      "loss": 1.8588,
      "step": 359000
    },
    {
      "epoch": 15.154069889980189,
      "grad_norm": 1.4405698776245117,
      "learning_rate": 4.005916906058397e-05,
      "loss": 1.8552,
      "step": 359500
    },
    {
      "epoch": 15.17514648231674,
      "grad_norm": 1.6316843032836914,
      "learning_rate": 3.9973416798868075e-05,
      "loss": 1.8555,
      "step": 360000
    },
    {
      "epoch": 15.17514648231674,
      "eval_accuracy": 0.673902387839351,
      "eval_loss": 1.6300581693649292,
      "eval_runtime": 169.1325,
      "eval_samples_per_second": 713.677,
      "eval_steps_per_second": 7.438,
      "step": 360000
    },
    {
      "epoch": 15.19622307465329,
      "grad_norm": 2.740753650665283,
      "learning_rate": 3.988766453715217e-05,
      "loss": 1.858,
      "step": 360500
    },
    {
      "epoch": 15.217299666989842,
      "grad_norm": 2.667917013168335,
      "learning_rate": 3.9801912275436265e-05,
      "loss": 1.8576,
      "step": 361000
    },
    {
      "epoch": 15.238376259326392,
      "grad_norm": 1.8298835754394531,
      "learning_rate": 3.971616001372036e-05,
      "loss": 1.8605,
      "step": 361500
    },
    {
      "epoch": 15.259452851662942,
      "grad_norm": 1.6575065851211548,
      "learning_rate": 3.963040775200446e-05,
      "loss": 1.8558,
      "step": 362000
    },
    {
      "epoch": 15.280529443999495,
      "grad_norm": 1.1588329076766968,
      "learning_rate": 3.954465549028856e-05,
      "loss": 1.8553,
      "step": 362500
    },
    {
      "epoch": 15.301606036336045,
      "grad_norm": 2.1532657146453857,
      "learning_rate": 3.945890322857266e-05,
      "loss": 1.8599,
      "step": 363000
    },
    {
      "epoch": 15.322682628672597,
      "grad_norm": 2.02732253074646,
      "learning_rate": 3.9373150966856756e-05,
      "loss": 1.8557,
      "step": 363500
    },
    {
      "epoch": 15.343759221009147,
      "grad_norm": 1.2668637037277222,
      "learning_rate": 3.928739870514085e-05,
      "loss": 1.8578,
      "step": 364000
    },
    {
      "epoch": 15.364835813345698,
      "grad_norm": 1.3955934047698975,
      "learning_rate": 3.920164644342495e-05,
      "loss": 1.8549,
      "step": 364500
    },
    {
      "epoch": 15.38591240568225,
      "grad_norm": 2.185978889465332,
      "learning_rate": 3.9115894181709045e-05,
      "loss": 1.8574,
      "step": 365000
    },
    {
      "epoch": 15.4069889980188,
      "grad_norm": 2.016780376434326,
      "learning_rate": 3.903014191999314e-05,
      "loss": 1.8516,
      "step": 365500
    },
    {
      "epoch": 15.42806559035535,
      "grad_norm": 1.8509719371795654,
      "learning_rate": 3.894438965827724e-05,
      "loss": 1.8579,
      "step": 366000
    },
    {
      "epoch": 15.449142182691903,
      "grad_norm": 2.1847355365753174,
      "learning_rate": 3.885863739656134e-05,
      "loss": 1.8573,
      "step": 366500
    },
    {
      "epoch": 15.470218775028453,
      "grad_norm": 1.503025770187378,
      "learning_rate": 3.877288513484543e-05,
      "loss": 1.8548,
      "step": 367000
    },
    {
      "epoch": 15.491295367365005,
      "grad_norm": 1.4371991157531738,
      "learning_rate": 3.868713287312953e-05,
      "loss": 1.8584,
      "step": 367500
    },
    {
      "epoch": 15.512371959701555,
      "grad_norm": 1.8988242149353027,
      "learning_rate": 3.860138061141363e-05,
      "loss": 1.8563,
      "step": 368000
    },
    {
      "epoch": 15.533448552038106,
      "grad_norm": 1.713152289390564,
      "learning_rate": 3.8515628349697726e-05,
      "loss": 1.8591,
      "step": 368500
    },
    {
      "epoch": 15.554525144374658,
      "grad_norm": 2.369403123855591,
      "learning_rate": 3.842987608798182e-05,
      "loss": 1.8574,
      "step": 369000
    },
    {
      "epoch": 15.575601736711208,
      "grad_norm": 2.9686832427978516,
      "learning_rate": 3.834412382626592e-05,
      "loss": 1.8528,
      "step": 369500
    },
    {
      "epoch": 15.596678329047759,
      "grad_norm": 1.2522916793823242,
      "learning_rate": 3.8258371564550014e-05,
      "loss": 1.8581,
      "step": 370000
    },
    {
      "epoch": 15.596678329047759,
      "eval_accuracy": 0.6747634164266969,
      "eval_loss": 1.6150254011154175,
      "eval_runtime": 169.2838,
      "eval_samples_per_second": 713.039,
      "eval_steps_per_second": 7.431,
      "step": 370000
    },
    {
      "epoch": 15.61775492138431,
      "grad_norm": 1.776956558227539,
      "learning_rate": 3.817261930283411e-05,
      "loss": 1.854,
      "step": 370500
    },
    {
      "epoch": 15.638831513720861,
      "grad_norm": 1.2868926525115967,
      "learning_rate": 3.808686704111821e-05,
      "loss": 1.8557,
      "step": 371000
    },
    {
      "epoch": 15.659908106057413,
      "grad_norm": 2.824329376220703,
      "learning_rate": 3.800111477940231e-05,
      "loss": 1.8556,
      "step": 371500
    },
    {
      "epoch": 15.680984698393964,
      "grad_norm": 2.0445375442504883,
      "learning_rate": 3.79153625176864e-05,
      "loss": 1.8585,
      "step": 372000
    },
    {
      "epoch": 15.702061290730514,
      "grad_norm": 1.4443556070327759,
      "learning_rate": 3.7829610255970506e-05,
      "loss": 1.8607,
      "step": 372500
    },
    {
      "epoch": 15.723137883067066,
      "grad_norm": 2.1355326175689697,
      "learning_rate": 3.77438579942546e-05,
      "loss": 1.8522,
      "step": 373000
    },
    {
      "epoch": 15.744214475403616,
      "grad_norm": 1.5137239694595337,
      "learning_rate": 3.7658105732538696e-05,
      "loss": 1.8513,
      "step": 373500
    },
    {
      "epoch": 15.765291067740169,
      "grad_norm": 1.4635331630706787,
      "learning_rate": 3.7572353470822794e-05,
      "loss": 1.8544,
      "step": 374000
    },
    {
      "epoch": 15.786367660076719,
      "grad_norm": 2.370457172393799,
      "learning_rate": 3.748660120910689e-05,
      "loss": 1.8534,
      "step": 374500
    },
    {
      "epoch": 15.80744425241327,
      "grad_norm": 2.064851760864258,
      "learning_rate": 3.740084894739099e-05,
      "loss": 1.8504,
      "step": 375000
    },
    {
      "epoch": 15.828520844749821,
      "grad_norm": 1.1207302808761597,
      "learning_rate": 3.731509668567509e-05,
      "loss": 1.8489,
      "step": 375500
    },
    {
      "epoch": 15.849597437086372,
      "grad_norm": 2.18172550201416,
      "learning_rate": 3.722934442395919e-05,
      "loss": 1.8564,
      "step": 376000
    },
    {
      "epoch": 15.870674029422922,
      "grad_norm": 1.5944300889968872,
      "learning_rate": 3.714359216224328e-05,
      "loss": 1.8562,
      "step": 376500
    },
    {
      "epoch": 15.891750621759474,
      "grad_norm": 2.6712453365325928,
      "learning_rate": 3.705783990052738e-05,
      "loss": 1.8585,
      "step": 377000
    },
    {
      "epoch": 15.912827214096025,
      "grad_norm": 2.0792593955993652,
      "learning_rate": 3.6972087638811475e-05,
      "loss": 1.8513,
      "step": 377500
    },
    {
      "epoch": 15.933903806432577,
      "grad_norm": 1.689400315284729,
      "learning_rate": 3.6886335377095574e-05,
      "loss": 1.8549,
      "step": 378000
    },
    {
      "epoch": 15.954980398769127,
      "grad_norm": 1.1678413152694702,
      "learning_rate": 3.680058311537967e-05,
      "loss": 1.8546,
      "step": 378500
    },
    {
      "epoch": 15.976056991105677,
      "grad_norm": 2.3108179569244385,
      "learning_rate": 3.671483085366377e-05,
      "loss": 1.8569,
      "step": 379000
    },
    {
      "epoch": 15.99713358344223,
      "grad_norm": 1.2608143091201782,
      "learning_rate": 3.662907859194786e-05,
      "loss": 1.8514,
      "step": 379500
    },
    {
      "epoch": 16.01821017577878,
      "grad_norm": 1.7369656562805176,
      "learning_rate": 3.654332633023197e-05,
      "loss": 1.8464,
      "step": 380000
    },
    {
      "epoch": 16.01821017577878,
      "eval_accuracy": 0.6752906411003048,
      "eval_loss": 1.622859001159668,
      "eval_runtime": 169.3881,
      "eval_samples_per_second": 712.6,
      "eval_steps_per_second": 7.427,
      "step": 380000
    },
    {
      "epoch": 16.039286768115332,
      "grad_norm": 1.6547441482543945,
      "learning_rate": 3.645757406851606e-05,
      "loss": 1.8439,
      "step": 380500
    },
    {
      "epoch": 16.06036336045188,
      "grad_norm": 2.3898234367370605,
      "learning_rate": 3.6371821806800157e-05,
      "loss": 1.8428,
      "step": 381000
    },
    {
      "epoch": 16.081439952788433,
      "grad_norm": 2.397005081176758,
      "learning_rate": 3.6286069545084255e-05,
      "loss": 1.8487,
      "step": 381500
    },
    {
      "epoch": 16.102516545124985,
      "grad_norm": 2.4906535148620605,
      "learning_rate": 3.620031728336835e-05,
      "loss": 1.8444,
      "step": 382000
    },
    {
      "epoch": 16.123593137461537,
      "grad_norm": 1.567560076713562,
      "learning_rate": 3.6114565021652445e-05,
      "loss": 1.8363,
      "step": 382500
    },
    {
      "epoch": 16.144669729798085,
      "grad_norm": 2.615719795227051,
      "learning_rate": 3.602881275993655e-05,
      "loss": 1.8401,
      "step": 383000
    },
    {
      "epoch": 16.165746322134638,
      "grad_norm": 1.9880684614181519,
      "learning_rate": 3.594306049822064e-05,
      "loss": 1.8453,
      "step": 383500
    },
    {
      "epoch": 16.18682291447119,
      "grad_norm": 2.511815309524536,
      "learning_rate": 3.585730823650474e-05,
      "loss": 1.8421,
      "step": 384000
    },
    {
      "epoch": 16.20789950680774,
      "grad_norm": 1.2609704732894897,
      "learning_rate": 3.577155597478883e-05,
      "loss": 1.8453,
      "step": 384500
    },
    {
      "epoch": 16.22897609914429,
      "grad_norm": 2.463542938232422,
      "learning_rate": 3.5685803713072936e-05,
      "loss": 1.8415,
      "step": 385000
    },
    {
      "epoch": 16.250052691480843,
      "grad_norm": 1.9402676820755005,
      "learning_rate": 3.560005145135703e-05,
      "loss": 1.8445,
      "step": 385500
    },
    {
      "epoch": 16.27112928381739,
      "grad_norm": 2.622596263885498,
      "learning_rate": 3.5514299189641126e-05,
      "loss": 1.8366,
      "step": 386000
    },
    {
      "epoch": 16.292205876153943,
      "grad_norm": 1.2724699974060059,
      "learning_rate": 3.5428546927925224e-05,
      "loss": 1.8455,
      "step": 386500
    },
    {
      "epoch": 16.313282468490495,
      "grad_norm": 2.1076819896698,
      "learning_rate": 3.534279466620932e-05,
      "loss": 1.8414,
      "step": 387000
    },
    {
      "epoch": 16.334359060827044,
      "grad_norm": 1.5682710409164429,
      "learning_rate": 3.525704240449342e-05,
      "loss": 1.8418,
      "step": 387500
    },
    {
      "epoch": 16.355435653163596,
      "grad_norm": 2.776425838470459,
      "learning_rate": 3.517129014277752e-05,
      "loss": 1.8406,
      "step": 388000
    },
    {
      "epoch": 16.376512245500148,
      "grad_norm": 2.2395098209381104,
      "learning_rate": 3.508553788106161e-05,
      "loss": 1.8415,
      "step": 388500
    },
    {
      "epoch": 16.3975888378367,
      "grad_norm": 1.2249066829681396,
      "learning_rate": 3.499978561934571e-05,
      "loss": 1.837,
      "step": 389000
    },
    {
      "epoch": 16.41866543017325,
      "grad_norm": 2.1931793689727783,
      "learning_rate": 3.491403335762981e-05,
      "loss": 1.8399,
      "step": 389500
    },
    {
      "epoch": 16.4397420225098,
      "grad_norm": 2.174604892730713,
      "learning_rate": 3.4828281095913906e-05,
      "loss": 1.8441,
      "step": 390000
    },
    {
      "epoch": 16.4397420225098,
      "eval_accuracy": 0.6757288964151689,
      "eval_loss": 1.6135960817337036,
      "eval_runtime": 169.0098,
      "eval_samples_per_second": 714.195,
      "eval_steps_per_second": 7.443,
      "step": 390000
    },
    {
      "epoch": 16.460818614846353,
      "grad_norm": 1.7218056917190552,
      "learning_rate": 3.4742528834198004e-05,
      "loss": 1.8415,
      "step": 390500
    },
    {
      "epoch": 16.4818952071829,
      "grad_norm": 1.8489373922348022,
      "learning_rate": 3.46567765724821e-05,
      "loss": 1.8439,
      "step": 391000
    },
    {
      "epoch": 16.502971799519454,
      "grad_norm": 2.2197482585906982,
      "learning_rate": 3.45710243107662e-05,
      "loss": 1.8433,
      "step": 391500
    },
    {
      "epoch": 16.524048391856006,
      "grad_norm": 1.6832969188690186,
      "learning_rate": 3.448527204905029e-05,
      "loss": 1.8452,
      "step": 392000
    },
    {
      "epoch": 16.545124984192555,
      "grad_norm": 1.4448823928833008,
      "learning_rate": 3.43995197873344e-05,
      "loss": 1.8395,
      "step": 392500
    },
    {
      "epoch": 16.566201576529107,
      "grad_norm": 2.4411988258361816,
      "learning_rate": 3.431376752561849e-05,
      "loss": 1.8453,
      "step": 393000
    },
    {
      "epoch": 16.58727816886566,
      "grad_norm": 1.9292789697647095,
      "learning_rate": 3.422801526390259e-05,
      "loss": 1.8479,
      "step": 393500
    },
    {
      "epoch": 16.608354761202207,
      "grad_norm": 2.1653401851654053,
      "learning_rate": 3.4142263002186686e-05,
      "loss": 1.8435,
      "step": 394000
    },
    {
      "epoch": 16.62943135353876,
      "grad_norm": 2.2345681190490723,
      "learning_rate": 3.4056510740470784e-05,
      "loss": 1.8489,
      "step": 394500
    },
    {
      "epoch": 16.65050794587531,
      "grad_norm": 2.4385154247283936,
      "learning_rate": 3.3970758478754875e-05,
      "loss": 1.8457,
      "step": 395000
    },
    {
      "epoch": 16.67158453821186,
      "grad_norm": 1.8488091230392456,
      "learning_rate": 3.388500621703898e-05,
      "loss": 1.8423,
      "step": 395500
    },
    {
      "epoch": 16.692661130548412,
      "grad_norm": 1.2518746852874756,
      "learning_rate": 3.379925395532307e-05,
      "loss": 1.8448,
      "step": 396000
    },
    {
      "epoch": 16.713737722884964,
      "grad_norm": 1.761244535446167,
      "learning_rate": 3.371350169360717e-05,
      "loss": 1.8421,
      "step": 396500
    },
    {
      "epoch": 16.734814315221517,
      "grad_norm": 1.7476589679718018,
      "learning_rate": 3.362774943189127e-05,
      "loss": 1.8459,
      "step": 397000
    },
    {
      "epoch": 16.755890907558065,
      "grad_norm": 2.0675172805786133,
      "learning_rate": 3.354199717017537e-05,
      "loss": 1.8424,
      "step": 397500
    },
    {
      "epoch": 16.776967499894617,
      "grad_norm": 1.1682469844818115,
      "learning_rate": 3.345624490845946e-05,
      "loss": 1.8433,
      "step": 398000
    },
    {
      "epoch": 16.79804409223117,
      "grad_norm": 2.4935712814331055,
      "learning_rate": 3.3370492646743564e-05,
      "loss": 1.8375,
      "step": 398500
    },
    {
      "epoch": 16.819120684567718,
      "grad_norm": 1.8945485353469849,
      "learning_rate": 3.3284740385027655e-05,
      "loss": 1.847,
      "step": 399000
    },
    {
      "epoch": 16.84019727690427,
      "grad_norm": 1.4658563137054443,
      "learning_rate": 3.319898812331175e-05,
      "loss": 1.8378,
      "step": 399500
    },
    {
      "epoch": 16.861273869240822,
      "grad_norm": 1.869767665863037,
      "learning_rate": 3.311323586159585e-05,
      "loss": 1.8378,
      "step": 400000
    },
    {
      "epoch": 16.861273869240822,
      "eval_accuracy": 0.6760976663357536,
      "eval_loss": 1.6065988540649414,
      "eval_runtime": 169.2191,
      "eval_samples_per_second": 713.312,
      "eval_steps_per_second": 7.434,
      "step": 400000
    },
    {
      "epoch": 16.88235046157737,
      "grad_norm": 1.8159115314483643,
      "learning_rate": 3.302748359987995e-05,
      "loss": 1.8384,
      "step": 400500
    },
    {
      "epoch": 16.903427053913923,
      "grad_norm": 2.3724896907806396,
      "learning_rate": 3.294173133816404e-05,
      "loss": 1.8422,
      "step": 401000
    },
    {
      "epoch": 16.924503646250475,
      "grad_norm": 1.5198428630828857,
      "learning_rate": 3.2855979076448147e-05,
      "loss": 1.8421,
      "step": 401500
    },
    {
      "epoch": 16.945580238587024,
      "grad_norm": 1.99824857711792,
      "learning_rate": 3.277022681473224e-05,
      "loss": 1.8406,
      "step": 402000
    },
    {
      "epoch": 16.966656830923576,
      "grad_norm": 1.9067180156707764,
      "learning_rate": 3.2684474553016336e-05,
      "loss": 1.8407,
      "step": 402500
    },
    {
      "epoch": 16.987733423260128,
      "grad_norm": 2.965806722640991,
      "learning_rate": 3.2598722291300435e-05,
      "loss": 1.8413,
      "step": 403000
    },
    {
      "epoch": 17.00881001559668,
      "grad_norm": 2.7578775882720947,
      "learning_rate": 3.251297002958453e-05,
      "loss": 1.8369,
      "step": 403500
    },
    {
      "epoch": 17.02988660793323,
      "grad_norm": 1.1781123876571655,
      "learning_rate": 3.242721776786863e-05,
      "loss": 1.8355,
      "step": 404000
    },
    {
      "epoch": 17.05096320026978,
      "grad_norm": 2.392967462539673,
      "learning_rate": 3.234146550615272e-05,
      "loss": 1.8325,
      "step": 404500
    },
    {
      "epoch": 17.072039792606333,
      "grad_norm": 2.1544382572174072,
      "learning_rate": 3.225571324443683e-05,
      "loss": 1.8313,
      "step": 405000
    },
    {
      "epoch": 17.09311638494288,
      "grad_norm": 2.4707798957824707,
      "learning_rate": 3.216996098272092e-05,
      "loss": 1.8298,
      "step": 405500
    },
    {
      "epoch": 17.114192977279433,
      "grad_norm": 1.6777886152267456,
      "learning_rate": 3.208420872100502e-05,
      "loss": 1.8327,
      "step": 406000
    },
    {
      "epoch": 17.135269569615986,
      "grad_norm": 1.5868972539901733,
      "learning_rate": 3.1998456459289116e-05,
      "loss": 1.8305,
      "step": 406500
    },
    {
      "epoch": 17.156346161952534,
      "grad_norm": 2.011024236679077,
      "learning_rate": 3.1912704197573214e-05,
      "loss": 1.8283,
      "step": 407000
    },
    {
      "epoch": 17.177422754289086,
      "grad_norm": 1.8350837230682373,
      "learning_rate": 3.1826951935857306e-05,
      "loss": 1.829,
      "step": 407500
    },
    {
      "epoch": 17.19849934662564,
      "grad_norm": 3.1748833656311035,
      "learning_rate": 3.174119967414141e-05,
      "loss": 1.8345,
      "step": 408000
    },
    {
      "epoch": 17.219575938962187,
      "grad_norm": 1.150686264038086,
      "learning_rate": 3.16554474124255e-05,
      "loss": 1.8302,
      "step": 408500
    },
    {
      "epoch": 17.24065253129874,
      "grad_norm": 1.8879340887069702,
      "learning_rate": 3.15696951507096e-05,
      "loss": 1.8331,
      "step": 409000
    },
    {
      "epoch": 17.26172912363529,
      "grad_norm": 1.507825493812561,
      "learning_rate": 3.14839428889937e-05,
      "loss": 1.8326,
      "step": 409500
    },
    {
      "epoch": 17.282805715971843,
      "grad_norm": 1.860090970993042,
      "learning_rate": 3.13981906272778e-05,
      "loss": 1.8364,
      "step": 410000
    },
    {
      "epoch": 17.282805715971843,
      "eval_accuracy": 0.6770460545040732,
      "eval_loss": 1.6129734516143799,
      "eval_runtime": 168.8029,
      "eval_samples_per_second": 715.071,
      "eval_steps_per_second": 7.452,
      "step": 410000
    },
    {
      "epoch": 17.303882308308392,
      "grad_norm": 1.627437949180603,
      "learning_rate": 3.131243836556189e-05,
      "loss": 1.8357,
      "step": 410500
    },
    {
      "epoch": 17.324958900644944,
      "grad_norm": 1.99100661277771,
      "learning_rate": 3.1226686103845994e-05,
      "loss": 1.8285,
      "step": 411000
    },
    {
      "epoch": 17.346035492981496,
      "grad_norm": 1.2576528787612915,
      "learning_rate": 3.1140933842130086e-05,
      "loss": 1.8316,
      "step": 411500
    },
    {
      "epoch": 17.367112085318045,
      "grad_norm": 1.25494384765625,
      "learning_rate": 3.1055181580414184e-05,
      "loss": 1.8344,
      "step": 412000
    },
    {
      "epoch": 17.388188677654597,
      "grad_norm": 1.621323585510254,
      "learning_rate": 3.096942931869828e-05,
      "loss": 1.8347,
      "step": 412500
    },
    {
      "epoch": 17.40926526999115,
      "grad_norm": 1.88888680934906,
      "learning_rate": 3.088367705698238e-05,
      "loss": 1.83,
      "step": 413000
    },
    {
      "epoch": 17.430341862327698,
      "grad_norm": 1.7120991945266724,
      "learning_rate": 3.079792479526647e-05,
      "loss": 1.8315,
      "step": 413500
    },
    {
      "epoch": 17.45141845466425,
      "grad_norm": 1.249309778213501,
      "learning_rate": 3.071217253355058e-05,
      "loss": 1.8306,
      "step": 414000
    },
    {
      "epoch": 17.472495047000802,
      "grad_norm": 1.5118231773376465,
      "learning_rate": 3.062642027183467e-05,
      "loss": 1.8302,
      "step": 414500
    },
    {
      "epoch": 17.49357163933735,
      "grad_norm": 1.246099591255188,
      "learning_rate": 3.054066801011877e-05,
      "loss": 1.8309,
      "step": 415000
    },
    {
      "epoch": 17.514648231673903,
      "grad_norm": 1.184031367301941,
      "learning_rate": 3.045491574840287e-05,
      "loss": 1.8326,
      "step": 415500
    },
    {
      "epoch": 17.535724824010455,
      "grad_norm": 2.407874345779419,
      "learning_rate": 3.0369163486686964e-05,
      "loss": 1.8291,
      "step": 416000
    },
    {
      "epoch": 17.556801416347007,
      "grad_norm": 2.1102893352508545,
      "learning_rate": 3.028341122497106e-05,
      "loss": 1.8312,
      "step": 416500
    },
    {
      "epoch": 17.577878008683555,
      "grad_norm": 1.7471858263015747,
      "learning_rate": 3.019765896325516e-05,
      "loss": 1.8342,
      "step": 417000
    },
    {
      "epoch": 17.598954601020107,
      "grad_norm": 2.2758498191833496,
      "learning_rate": 3.0111906701539255e-05,
      "loss": 1.8339,
      "step": 417500
    },
    {
      "epoch": 17.62003119335666,
      "grad_norm": 2.156168222427368,
      "learning_rate": 3.002615443982335e-05,
      "loss": 1.8275,
      "step": 418000
    },
    {
      "epoch": 17.641107785693208,
      "grad_norm": 1.4857163429260254,
      "learning_rate": 2.9940402178107452e-05,
      "loss": 1.8288,
      "step": 418500
    },
    {
      "epoch": 17.66218437802976,
      "grad_norm": 1.5087493658065796,
      "learning_rate": 2.9854649916391547e-05,
      "loss": 1.8298,
      "step": 419000
    },
    {
      "epoch": 17.683260970366312,
      "grad_norm": 1.503750205039978,
      "learning_rate": 2.976889765467564e-05,
      "loss": 1.8338,
      "step": 419500
    },
    {
      "epoch": 17.70433756270286,
      "grad_norm": 2.6227831840515137,
      "learning_rate": 2.9683145392959737e-05,
      "loss": 1.8323,
      "step": 420000
    },
    {
      "epoch": 17.70433756270286,
      "eval_accuracy": 0.6775551806929567,
      "eval_loss": 1.6022640466690063,
      "eval_runtime": 169.2599,
      "eval_samples_per_second": 713.14,
      "eval_steps_per_second": 7.432,
      "step": 420000
    },
    {
      "epoch": 17.725414155039413,
      "grad_norm": 1.1657729148864746,
      "learning_rate": 2.9597393131243838e-05,
      "loss": 1.8315,
      "step": 420500
    },
    {
      "epoch": 17.746490747375965,
      "grad_norm": 1.2400004863739014,
      "learning_rate": 2.9511640869527933e-05,
      "loss": 1.8263,
      "step": 421000
    },
    {
      "epoch": 17.767567339712514,
      "grad_norm": 2.4074459075927734,
      "learning_rate": 2.942588860781203e-05,
      "loss": 1.8289,
      "step": 421500
    },
    {
      "epoch": 17.788643932049066,
      "grad_norm": 2.747446060180664,
      "learning_rate": 2.934013634609613e-05,
      "loss": 1.8323,
      "step": 422000
    },
    {
      "epoch": 17.809720524385618,
      "grad_norm": 1.8731544017791748,
      "learning_rate": 2.9254384084380225e-05,
      "loss": 1.8336,
      "step": 422500
    },
    {
      "epoch": 17.830797116722167,
      "grad_norm": 2.475072145462036,
      "learning_rate": 2.9168631822664323e-05,
      "loss": 1.8306,
      "step": 423000
    },
    {
      "epoch": 17.85187370905872,
      "grad_norm": 1.3341856002807617,
      "learning_rate": 2.908287956094842e-05,
      "loss": 1.8339,
      "step": 423500
    },
    {
      "epoch": 17.87295030139527,
      "grad_norm": 2.1870877742767334,
      "learning_rate": 2.899712729923252e-05,
      "loss": 1.8317,
      "step": 424000
    },
    {
      "epoch": 17.894026893731823,
      "grad_norm": 1.134002685546875,
      "learning_rate": 2.8911375037516615e-05,
      "loss": 1.8295,
      "step": 424500
    },
    {
      "epoch": 17.91510348606837,
      "grad_norm": 1.40621018409729,
      "learning_rate": 2.8825622775800716e-05,
      "loss": 1.8313,
      "step": 425000
    },
    {
      "epoch": 17.936180078404924,
      "grad_norm": 2.2238516807556152,
      "learning_rate": 2.873987051408481e-05,
      "loss": 1.831,
      "step": 425500
    },
    {
      "epoch": 17.957256670741476,
      "grad_norm": 1.2630033493041992,
      "learning_rate": 2.8654118252368906e-05,
      "loss": 1.8285,
      "step": 426000
    },
    {
      "epoch": 17.978333263078024,
      "grad_norm": 1.7573308944702148,
      "learning_rate": 2.8568365990653008e-05,
      "loss": 1.8267,
      "step": 426500
    },
    {
      "epoch": 17.999409855414576,
      "grad_norm": 1.8199607133865356,
      "learning_rate": 2.8482613728937103e-05,
      "loss": 1.8344,
      "step": 427000
    },
    {
      "epoch": 18.02048644775113,
      "grad_norm": 2.149515390396118,
      "learning_rate": 2.8396861467221198e-05,
      "loss": 1.8217,
      "step": 427500
    },
    {
      "epoch": 18.041563040087677,
      "grad_norm": 2.074085235595703,
      "learning_rate": 2.83111092055053e-05,
      "loss": 1.8198,
      "step": 428000
    },
    {
      "epoch": 18.06263963242423,
      "grad_norm": 2.1428394317626953,
      "learning_rate": 2.8225356943789394e-05,
      "loss": 1.8222,
      "step": 428500
    },
    {
      "epoch": 18.08371622476078,
      "grad_norm": 2.040087938308716,
      "learning_rate": 2.813960468207349e-05,
      "loss": 1.8208,
      "step": 429000
    },
    {
      "epoch": 18.10479281709733,
      "grad_norm": 2.349421501159668,
      "learning_rate": 2.805385242035759e-05,
      "loss": 1.8199,
      "step": 429500
    },
    {
      "epoch": 18.125869409433882,
      "grad_norm": 1.9922237396240234,
      "learning_rate": 2.7968100158641686e-05,
      "loss": 1.8194,
      "step": 430000
    },
    {
      "epoch": 18.125869409433882,
      "eval_accuracy": 0.6777127347294049,
      "eval_loss": 1.6100091934204102,
      "eval_runtime": 169.2,
      "eval_samples_per_second": 713.392,
      "eval_steps_per_second": 7.435,
      "step": 430000
    },
    {
      "epoch": 18.146946001770434,
      "grad_norm": 1.4122517108917236,
      "learning_rate": 2.788234789692578e-05,
      "loss": 1.822,
      "step": 430500
    },
    {
      "epoch": 18.168022594106986,
      "grad_norm": 1.685200572013855,
      "learning_rate": 2.7796595635209882e-05,
      "loss": 1.8203,
      "step": 431000
    },
    {
      "epoch": 18.189099186443535,
      "grad_norm": 1.7541621923446655,
      "learning_rate": 2.7710843373493977e-05,
      "loss": 1.8241,
      "step": 431500
    },
    {
      "epoch": 18.210175778780087,
      "grad_norm": 3.204068899154663,
      "learning_rate": 2.7625091111778072e-05,
      "loss": 1.8229,
      "step": 432000
    },
    {
      "epoch": 18.23125237111664,
      "grad_norm": 1.1971794366836548,
      "learning_rate": 2.7539338850062174e-05,
      "loss": 1.8218,
      "step": 432500
    },
    {
      "epoch": 18.252328963453188,
      "grad_norm": 2.7981743812561035,
      "learning_rate": 2.745358658834627e-05,
      "loss": 1.823,
      "step": 433000
    },
    {
      "epoch": 18.27340555578974,
      "grad_norm": 1.5809448957443237,
      "learning_rate": 2.7367834326630364e-05,
      "loss": 1.8193,
      "step": 433500
    },
    {
      "epoch": 18.294482148126292,
      "grad_norm": 1.393281102180481,
      "learning_rate": 2.7282082064914465e-05,
      "loss": 1.8194,
      "step": 434000
    },
    {
      "epoch": 18.31555874046284,
      "grad_norm": 1.1647353172302246,
      "learning_rate": 2.719632980319856e-05,
      "loss": 1.8193,
      "step": 434500
    },
    {
      "epoch": 18.336635332799393,
      "grad_norm": 1.3115051984786987,
      "learning_rate": 2.7110577541482655e-05,
      "loss": 1.8196,
      "step": 435000
    },
    {
      "epoch": 18.357711925135945,
      "grad_norm": 3.4752583503723145,
      "learning_rate": 2.7024825279766757e-05,
      "loss": 1.8233,
      "step": 435500
    },
    {
      "epoch": 18.378788517472493,
      "grad_norm": 1.9354122877120972,
      "learning_rate": 2.6939073018050852e-05,
      "loss": 1.8246,
      "step": 436000
    },
    {
      "epoch": 18.399865109809046,
      "grad_norm": 1.5128587484359741,
      "learning_rate": 2.685332075633495e-05,
      "loss": 1.822,
      "step": 436500
    },
    {
      "epoch": 18.420941702145598,
      "grad_norm": 1.6362742185592651,
      "learning_rate": 2.6767568494619045e-05,
      "loss": 1.8214,
      "step": 437000
    },
    {
      "epoch": 18.44201829448215,
      "grad_norm": 2.7812399864196777,
      "learning_rate": 2.6681816232903143e-05,
      "loss": 1.819,
      "step": 437500
    },
    {
      "epoch": 18.4630948868187,
      "grad_norm": 1.2386285066604614,
      "learning_rate": 2.6596063971187242e-05,
      "loss": 1.817,
      "step": 438000
    },
    {
      "epoch": 18.48417147915525,
      "grad_norm": 2.225489854812622,
      "learning_rate": 2.6510311709471337e-05,
      "loss": 1.8198,
      "step": 438500
    },
    {
      "epoch": 18.505248071491803,
      "grad_norm": 1.6155035495758057,
      "learning_rate": 2.642455944775544e-05,
      "loss": 1.8207,
      "step": 439000
    },
    {
      "epoch": 18.52632466382835,
      "grad_norm": 3.0333433151245117,
      "learning_rate": 2.6338807186039533e-05,
      "loss": 1.8183,
      "step": 439500
    },
    {
      "epoch": 18.547401256164903,
      "grad_norm": 1.2076996564865112,
      "learning_rate": 2.6253054924323628e-05,
      "loss": 1.8169,
      "step": 440000
    },
    {
      "epoch": 18.547401256164903,
      "eval_accuracy": 0.6779192266463764,
      "eval_loss": 1.613216519355774,
      "eval_runtime": 169.0825,
      "eval_samples_per_second": 713.888,
      "eval_steps_per_second": 7.44,
      "step": 440000
    },
    {
      "epoch": 18.568477848501455,
      "grad_norm": 1.6616170406341553,
      "learning_rate": 2.616730266260773e-05,
      "loss": 1.8192,
      "step": 440500
    },
    {
      "epoch": 18.589554440838004,
      "grad_norm": 1.326160192489624,
      "learning_rate": 2.6081550400891825e-05,
      "loss": 1.8206,
      "step": 441000
    },
    {
      "epoch": 18.610631033174556,
      "grad_norm": 2.326592206954956,
      "learning_rate": 2.599579813917592e-05,
      "loss": 1.8242,
      "step": 441500
    },
    {
      "epoch": 18.63170762551111,
      "grad_norm": 1.3346434831619263,
      "learning_rate": 2.591004587746002e-05,
      "loss": 1.8211,
      "step": 442000
    },
    {
      "epoch": 18.652784217847657,
      "grad_norm": 2.4678409099578857,
      "learning_rate": 2.5824293615744116e-05,
      "loss": 1.8201,
      "step": 442500
    },
    {
      "epoch": 18.67386081018421,
      "grad_norm": 2.0234274864196777,
      "learning_rate": 2.573854135402821e-05,
      "loss": 1.8208,
      "step": 443000
    },
    {
      "epoch": 18.69493740252076,
      "grad_norm": 2.257481575012207,
      "learning_rate": 2.5652789092312313e-05,
      "loss": 1.8204,
      "step": 443500
    },
    {
      "epoch": 18.716013994857313,
      "grad_norm": 2.3827152252197266,
      "learning_rate": 2.5567036830596408e-05,
      "loss": 1.8204,
      "step": 444000
    },
    {
      "epoch": 18.737090587193862,
      "grad_norm": 1.1661750078201294,
      "learning_rate": 2.5481284568880503e-05,
      "loss": 1.8191,
      "step": 444500
    },
    {
      "epoch": 18.758167179530414,
      "grad_norm": 2.0620779991149902,
      "learning_rate": 2.5395532307164605e-05,
      "loss": 1.8167,
      "step": 445000
    },
    {
      "epoch": 18.779243771866966,
      "grad_norm": 2.0848565101623535,
      "learning_rate": 2.53097800454487e-05,
      "loss": 1.8246,
      "step": 445500
    },
    {
      "epoch": 18.800320364203515,
      "grad_norm": 1.6826413869857788,
      "learning_rate": 2.5224027783732794e-05,
      "loss": 1.8211,
      "step": 446000
    },
    {
      "epoch": 18.821396956540067,
      "grad_norm": 2.320129632949829,
      "learning_rate": 2.5138275522016896e-05,
      "loss": 1.8139,
      "step": 446500
    },
    {
      "epoch": 18.84247354887662,
      "grad_norm": 1.983734130859375,
      "learning_rate": 2.505252326030099e-05,
      "loss": 1.8164,
      "step": 447000
    },
    {
      "epoch": 18.863550141213167,
      "grad_norm": 1.8537651300430298,
      "learning_rate": 2.496677099858509e-05,
      "loss": 1.8163,
      "step": 447500
    },
    {
      "epoch": 18.88462673354972,
      "grad_norm": 2.1497092247009277,
      "learning_rate": 2.4881018736869184e-05,
      "loss": 1.8235,
      "step": 448000
    },
    {
      "epoch": 18.90570332588627,
      "grad_norm": 2.133894920349121,
      "learning_rate": 2.4795266475153283e-05,
      "loss": 1.8182,
      "step": 448500
    },
    {
      "epoch": 18.92677991822282,
      "grad_norm": 3.601257085800171,
      "learning_rate": 2.470951421343738e-05,
      "loss": 1.821,
      "step": 449000
    },
    {
      "epoch": 18.947856510559372,
      "grad_norm": 2.2144463062286377,
      "learning_rate": 2.4623761951721476e-05,
      "loss": 1.8151,
      "step": 449500
    },
    {
      "epoch": 18.968933102895924,
      "grad_norm": 2.1154873371124268,
      "learning_rate": 2.4538009690005574e-05,
      "loss": 1.8174,
      "step": 450000
    },
    {
      "epoch": 18.968933102895924,
      "eval_accuracy": 0.6788502383375498,
      "eval_loss": 1.6051075458526611,
      "eval_runtime": 170.4793,
      "eval_samples_per_second": 708.039,
      "eval_steps_per_second": 7.379,
      "step": 450000
    },
    {
      "epoch": 18.990009695232473,
      "grad_norm": 2.9789891242980957,
      "learning_rate": 2.4452257428289672e-05,
      "loss": 1.8167,
      "step": 450500
    },
    {
      "epoch": 19.011086287569025,
      "grad_norm": 3.796696662902832,
      "learning_rate": 2.436650516657377e-05,
      "loss": 1.814,
      "step": 451000
    },
    {
      "epoch": 19.032162879905577,
      "grad_norm": 1.3088061809539795,
      "learning_rate": 2.428075290485787e-05,
      "loss": 1.8101,
      "step": 451500
    },
    {
      "epoch": 19.05323947224213,
      "grad_norm": 1.3159023523330688,
      "learning_rate": 2.4195000643141964e-05,
      "loss": 1.8131,
      "step": 452000
    },
    {
      "epoch": 19.074316064578678,
      "grad_norm": 2.3627970218658447,
      "learning_rate": 2.4109248381426062e-05,
      "loss": 1.8151,
      "step": 452500
    },
    {
      "epoch": 19.09539265691523,
      "grad_norm": 2.7888543605804443,
      "learning_rate": 2.402349611971016e-05,
      "loss": 1.8113,
      "step": 453000
    },
    {
      "epoch": 19.116469249251782,
      "grad_norm": 1.2076860666275024,
      "learning_rate": 2.3937743857994255e-05,
      "loss": 1.8096,
      "step": 453500
    },
    {
      "epoch": 19.13754584158833,
      "grad_norm": 1.3177707195281982,
      "learning_rate": 2.3851991596278354e-05,
      "loss": 1.8125,
      "step": 454000
    },
    {
      "epoch": 19.158622433924883,
      "grad_norm": 1.9000552892684937,
      "learning_rate": 2.3766239334562452e-05,
      "loss": 1.8087,
      "step": 454500
    },
    {
      "epoch": 19.179699026261435,
      "grad_norm": 3.178394317626953,
      "learning_rate": 2.3680487072846547e-05,
      "loss": 1.8075,
      "step": 455000
    },
    {
      "epoch": 19.200775618597984,
      "grad_norm": 1.6848644018173218,
      "learning_rate": 2.3594734811130645e-05,
      "loss": 1.8079,
      "step": 455500
    },
    {
      "epoch": 19.221852210934536,
      "grad_norm": 1.9828680753707886,
      "learning_rate": 2.3508982549414744e-05,
      "loss": 1.8081,
      "step": 456000
    },
    {
      "epoch": 19.242928803271088,
      "grad_norm": 2.5300185680389404,
      "learning_rate": 2.342323028769884e-05,
      "loss": 1.8127,
      "step": 456500
    },
    {
      "epoch": 19.264005395607636,
      "grad_norm": 1.1643385887145996,
      "learning_rate": 2.3337478025982937e-05,
      "loss": 1.8091,
      "step": 457000
    },
    {
      "epoch": 19.28508198794419,
      "grad_norm": 2.0911662578582764,
      "learning_rate": 2.3251725764267035e-05,
      "loss": 1.8109,
      "step": 457500
    },
    {
      "epoch": 19.30615858028074,
      "grad_norm": 3.059058666229248,
      "learning_rate": 2.316597350255113e-05,
      "loss": 1.8146,
      "step": 458000
    },
    {
      "epoch": 19.327235172617293,
      "grad_norm": 2.0782341957092285,
      "learning_rate": 2.308022124083523e-05,
      "loss": 1.812,
      "step": 458500
    },
    {
      "epoch": 19.34831176495384,
      "grad_norm": 1.9521619081497192,
      "learning_rate": 2.2994468979119323e-05,
      "loss": 1.8148,
      "step": 459000
    },
    {
      "epoch": 19.369388357290394,
      "grad_norm": 2.1593618392944336,
      "learning_rate": 2.290871671740342e-05,
      "loss": 1.8094,
      "step": 459500
    },
    {
      "epoch": 19.390464949626946,
      "grad_norm": 2.5076372623443604,
      "learning_rate": 2.282296445568752e-05,
      "loss": 1.8105,
      "step": 460000
    },
    {
      "epoch": 19.390464949626946,
      "eval_accuracy": 0.6791289311722686,
      "eval_loss": 1.6219534873962402,
      "eval_runtime": 169.0895,
      "eval_samples_per_second": 713.859,
      "eval_steps_per_second": 7.44,
      "step": 460000
    },
    {
      "epoch": 19.411541541963494,
      "grad_norm": 2.0758657455444336,
      "learning_rate": 2.2737212193971615e-05,
      "loss": 1.8096,
      "step": 460500
    },
    {
      "epoch": 19.432618134300046,
      "grad_norm": 2.553884267807007,
      "learning_rate": 2.2651459932255713e-05,
      "loss": 1.8141,
      "step": 461000
    },
    {
      "epoch": 19.4536947266366,
      "grad_norm": 2.2717273235321045,
      "learning_rate": 2.256570767053981e-05,
      "loss": 1.8065,
      "step": 461500
    },
    {
      "epoch": 19.474771318973147,
      "grad_norm": 1.8689240217208862,
      "learning_rate": 2.2479955408823906e-05,
      "loss": 1.8075,
      "step": 462000
    },
    {
      "epoch": 19.4958479113097,
      "grad_norm": 1.5922154188156128,
      "learning_rate": 2.2394203147108005e-05,
      "loss": 1.8096,
      "step": 462500
    },
    {
      "epoch": 19.51692450364625,
      "grad_norm": 1.2837779521942139,
      "learning_rate": 2.2308450885392103e-05,
      "loss": 1.8078,
      "step": 463000
    },
    {
      "epoch": 19.5380010959828,
      "grad_norm": 2.4861838817596436,
      "learning_rate": 2.22226986236762e-05,
      "loss": 1.8066,
      "step": 463500
    },
    {
      "epoch": 19.559077688319352,
      "grad_norm": 1.4344888925552368,
      "learning_rate": 2.21369463619603e-05,
      "loss": 1.8099,
      "step": 464000
    },
    {
      "epoch": 19.580154280655904,
      "grad_norm": 2.50852632522583,
      "learning_rate": 2.2051194100244394e-05,
      "loss": 1.8093,
      "step": 464500
    },
    {
      "epoch": 19.601230872992456,
      "grad_norm": 1.3669157028198242,
      "learning_rate": 2.1965441838528493e-05,
      "loss": 1.8073,
      "step": 465000
    },
    {
      "epoch": 19.622307465329005,
      "grad_norm": 1.663144588470459,
      "learning_rate": 2.187968957681259e-05,
      "loss": 1.8086,
      "step": 465500
    },
    {
      "epoch": 19.643384057665557,
      "grad_norm": 1.93024742603302,
      "learning_rate": 2.179393731509669e-05,
      "loss": 1.812,
      "step": 466000
    },
    {
      "epoch": 19.66446065000211,
      "grad_norm": 3.2224364280700684,
      "learning_rate": 2.1708185053380784e-05,
      "loss": 1.8087,
      "step": 466500
    },
    {
      "epoch": 19.685537242338658,
      "grad_norm": 1.9516185522079468,
      "learning_rate": 2.1622432791664883e-05,
      "loss": 1.8095,
      "step": 467000
    },
    {
      "epoch": 19.70661383467521,
      "grad_norm": 2.701633930206299,
      "learning_rate": 2.1536680529948978e-05,
      "loss": 1.8106,
      "step": 467500
    },
    {
      "epoch": 19.727690427011762,
      "grad_norm": 2.376532793045044,
      "learning_rate": 2.1450928268233076e-05,
      "loss": 1.8078,
      "step": 468000
    },
    {
      "epoch": 19.74876701934831,
      "grad_norm": 1.4524612426757812,
      "learning_rate": 2.1365176006517174e-05,
      "loss": 1.8095,
      "step": 468500
    },
    {
      "epoch": 19.769843611684863,
      "grad_norm": 2.0169553756713867,
      "learning_rate": 2.127942374480127e-05,
      "loss": 1.8076,
      "step": 469000
    },
    {
      "epoch": 19.790920204021415,
      "grad_norm": 1.9969581365585327,
      "learning_rate": 2.1193671483085367e-05,
      "loss": 1.8138,
      "step": 469500
    },
    {
      "epoch": 19.811996796357963,
      "grad_norm": 2.916889190673828,
      "learning_rate": 2.1107919221369466e-05,
      "loss": 1.8115,
      "step": 470000
    },
    {
      "epoch": 19.811996796357963,
      "eval_accuracy": 0.6795063711887649,
      "eval_loss": 1.5991134643554688,
      "eval_runtime": 169.0604,
      "eval_samples_per_second": 713.982,
      "eval_steps_per_second": 7.441,
      "step": 470000
    },
    {
      "epoch": 19.833073388694515,
      "grad_norm": 1.1589467525482178,
      "learning_rate": 2.102216695965356e-05,
      "loss": 1.8107,
      "step": 470500
    },
    {
      "epoch": 19.854149981031068,
      "grad_norm": 1.2541656494140625,
      "learning_rate": 2.093641469793766e-05,
      "loss": 1.8069,
      "step": 471000
    },
    {
      "epoch": 19.87522657336762,
      "grad_norm": 1.1462225914001465,
      "learning_rate": 2.0850662436221757e-05,
      "loss": 1.8077,
      "step": 471500
    },
    {
      "epoch": 19.89630316570417,
      "grad_norm": 1.578399896621704,
      "learning_rate": 2.0764910174505852e-05,
      "loss": 1.806,
      "step": 472000
    },
    {
      "epoch": 19.91737975804072,
      "grad_norm": 1.9484221935272217,
      "learning_rate": 2.067915791278995e-05,
      "loss": 1.8097,
      "step": 472500
    },
    {
      "epoch": 19.938456350377272,
      "grad_norm": 5.217864990234375,
      "learning_rate": 2.059340565107405e-05,
      "loss": 1.8085,
      "step": 473000
    },
    {
      "epoch": 19.95953294271382,
      "grad_norm": 1.6523851156234741,
      "learning_rate": 2.0507653389358144e-05,
      "loss": 1.8073,
      "step": 473500
    },
    {
      "epoch": 19.980609535050373,
      "grad_norm": 1.9445120096206665,
      "learning_rate": 2.0421901127642242e-05,
      "loss": 1.8059,
      "step": 474000
    },
    {
      "epoch": 20.001686127386925,
      "grad_norm": 1.2479759454727173,
      "learning_rate": 2.033614886592634e-05,
      "loss": 1.814,
      "step": 474500
    },
    {
      "epoch": 20.022762719723474,
      "grad_norm": 1.269612193107605,
      "learning_rate": 2.0250396604210435e-05,
      "loss": 1.8035,
      "step": 475000
    },
    {
      "epoch": 20.043839312060026,
      "grad_norm": 1.865881323814392,
      "learning_rate": 2.0164644342494534e-05,
      "loss": 1.8045,
      "step": 475500
    },
    {
      "epoch": 20.064915904396578,
      "grad_norm": 1.1608577966690063,
      "learning_rate": 2.007889208077863e-05,
      "loss": 1.7979,
      "step": 476000
    },
    {
      "epoch": 20.085992496733127,
      "grad_norm": 1.7678890228271484,
      "learning_rate": 1.9993139819062727e-05,
      "loss": 1.8026,
      "step": 476500
    },
    {
      "epoch": 20.10706908906968,
      "grad_norm": 1.4839969873428345,
      "learning_rate": 1.9907387557346825e-05,
      "loss": 1.805,
      "step": 477000
    },
    {
      "epoch": 20.12814568140623,
      "grad_norm": 3.2681455612182617,
      "learning_rate": 1.9821635295630923e-05,
      "loss": 1.801,
      "step": 477500
    },
    {
      "epoch": 20.14922227374278,
      "grad_norm": 2.8914308547973633,
      "learning_rate": 1.9735883033915022e-05,
      "loss": 1.7973,
      "step": 478000
    },
    {
      "epoch": 20.17029886607933,
      "grad_norm": 1.8822407722473145,
      "learning_rate": 1.965013077219912e-05,
      "loss": 1.8038,
      "step": 478500
    },
    {
      "epoch": 20.191375458415884,
      "grad_norm": 3.769503116607666,
      "learning_rate": 1.9564378510483215e-05,
      "loss": 1.8021,
      "step": 479000
    },
    {
      "epoch": 20.212452050752436,
      "grad_norm": 1.6290580034255981,
      "learning_rate": 1.9478626248767313e-05,
      "loss": 1.8022,
      "step": 479500
    },
    {
      "epoch": 20.233528643088984,
      "grad_norm": 2.962827205657959,
      "learning_rate": 1.939287398705141e-05,
      "loss": 1.8015,
      "step": 480000
    },
    {
      "epoch": 20.233528643088984,
      "eval_accuracy": 0.6801228784019789,
      "eval_loss": 1.611912727355957,
      "eval_runtime": 169.2502,
      "eval_samples_per_second": 713.181,
      "eval_steps_per_second": 7.433,
      "step": 480000
    },
    {
      "epoch": 20.254605235425537,
      "grad_norm": 3.1885266304016113,
      "learning_rate": 1.9307121725335506e-05,
      "loss": 1.7981,
      "step": 480500
    },
    {
      "epoch": 20.27568182776209,
      "grad_norm": 1.4597065448760986,
      "learning_rate": 1.9221369463619605e-05,
      "loss": 1.7987,
      "step": 481000
    },
    {
      "epoch": 20.296758420098637,
      "grad_norm": 1.2021421194076538,
      "learning_rate": 1.9135617201903703e-05,
      "loss": 1.8009,
      "step": 481500
    },
    {
      "epoch": 20.31783501243519,
      "grad_norm": 2.738809108734131,
      "learning_rate": 1.9049864940187798e-05,
      "loss": 1.7986,
      "step": 482000
    },
    {
      "epoch": 20.33891160477174,
      "grad_norm": 2.4601268768310547,
      "learning_rate": 1.8964112678471896e-05,
      "loss": 1.8006,
      "step": 482500
    },
    {
      "epoch": 20.35998819710829,
      "grad_norm": 2.3304977416992188,
      "learning_rate": 1.8878360416755995e-05,
      "loss": 1.8033,
      "step": 483000
    },
    {
      "epoch": 20.381064789444842,
      "grad_norm": 1.3823086023330688,
      "learning_rate": 1.879260815504009e-05,
      "loss": 1.7994,
      "step": 483500
    },
    {
      "epoch": 20.402141381781394,
      "grad_norm": 2.5231456756591797,
      "learning_rate": 1.8706855893324188e-05,
      "loss": 1.7963,
      "step": 484000
    },
    {
      "epoch": 20.423217974117943,
      "grad_norm": 1.1569113731384277,
      "learning_rate": 1.8621103631608283e-05,
      "loss": 1.8016,
      "step": 484500
    },
    {
      "epoch": 20.444294566454495,
      "grad_norm": 1.8620951175689697,
      "learning_rate": 1.853535136989238e-05,
      "loss": 1.7984,
      "step": 485000
    },
    {
      "epoch": 20.465371158791047,
      "grad_norm": 2.247269868850708,
      "learning_rate": 1.844959910817648e-05,
      "loss": 1.7984,
      "step": 485500
    },
    {
      "epoch": 20.4864477511276,
      "grad_norm": 1.5692057609558105,
      "learning_rate": 1.8363846846460574e-05,
      "loss": 1.8027,
      "step": 486000
    },
    {
      "epoch": 20.507524343464148,
      "grad_norm": 2.07566499710083,
      "learning_rate": 1.8278094584744673e-05,
      "loss": 1.7999,
      "step": 486500
    },
    {
      "epoch": 20.5286009358007,
      "grad_norm": 3.668026924133301,
      "learning_rate": 1.819234232302877e-05,
      "loss": 1.8046,
      "step": 487000
    },
    {
      "epoch": 20.549677528137252,
      "grad_norm": 2.880370616912842,
      "learning_rate": 1.8106590061312866e-05,
      "loss": 1.8007,
      "step": 487500
    },
    {
      "epoch": 20.5707541204738,
      "grad_norm": 1.63459050655365,
      "learning_rate": 1.8020837799596964e-05,
      "loss": 1.8018,
      "step": 488000
    },
    {
      "epoch": 20.591830712810353,
      "grad_norm": 1.9449729919433594,
      "learning_rate": 1.7935085537881062e-05,
      "loss": 1.8026,
      "step": 488500
    },
    {
      "epoch": 20.612907305146905,
      "grad_norm": 1.455431342124939,
      "learning_rate": 1.7849333276165157e-05,
      "loss": 1.7975,
      "step": 489000
    },
    {
      "epoch": 20.633983897483454,
      "grad_norm": 1.456680417060852,
      "learning_rate": 1.7763581014449256e-05,
      "loss": 1.801,
      "step": 489500
    },
    {
      "epoch": 20.655060489820006,
      "grad_norm": 3.3171253204345703,
      "learning_rate": 1.7677828752733354e-05,
      "loss": 1.7975,
      "step": 490000
    },
    {
      "epoch": 20.655060489820006,
      "eval_accuracy": 0.6804385200200009,
      "eval_loss": 1.6027177572250366,
      "eval_runtime": 168.8762,
      "eval_samples_per_second": 714.76,
      "eval_steps_per_second": 7.449,
      "step": 490000
    },
    {
      "epoch": 20.676137082156558,
      "grad_norm": 1.4215635061264038,
      "learning_rate": 1.7592076491017452e-05,
      "loss": 1.7993,
      "step": 490500
    },
    {
      "epoch": 20.697213674493106,
      "grad_norm": 2.6899123191833496,
      "learning_rate": 1.7506324229301547e-05,
      "loss": 1.8033,
      "step": 491000
    },
    {
      "epoch": 20.71829026682966,
      "grad_norm": 1.2184942960739136,
      "learning_rate": 1.7420571967585646e-05,
      "loss": 1.8022,
      "step": 491500
    },
    {
      "epoch": 20.73936685916621,
      "grad_norm": 1.5315614938735962,
      "learning_rate": 1.7334819705869744e-05,
      "loss": 1.7976,
      "step": 492000
    },
    {
      "epoch": 20.760443451502763,
      "grad_norm": 2.1701629161834717,
      "learning_rate": 1.7249067444153842e-05,
      "loss": 1.8026,
      "step": 492500
    },
    {
      "epoch": 20.78152004383931,
      "grad_norm": 1.782426357269287,
      "learning_rate": 1.7163315182437937e-05,
      "loss": 1.7954,
      "step": 493000
    },
    {
      "epoch": 20.802596636175863,
      "grad_norm": 1.319637656211853,
      "learning_rate": 1.7077562920722035e-05,
      "loss": 1.7952,
      "step": 493500
    },
    {
      "epoch": 20.823673228512416,
      "grad_norm": 2.484625816345215,
      "learning_rate": 1.6991810659006134e-05,
      "loss": 1.7976,
      "step": 494000
    },
    {
      "epoch": 20.844749820848964,
      "grad_norm": 1.3538862466812134,
      "learning_rate": 1.690605839729023e-05,
      "loss": 1.7998,
      "step": 494500
    },
    {
      "epoch": 20.865826413185516,
      "grad_norm": 1.3715213537216187,
      "learning_rate": 1.6820306135574327e-05,
      "loss": 1.7994,
      "step": 495000
    },
    {
      "epoch": 20.88690300552207,
      "grad_norm": 2.301762819290161,
      "learning_rate": 1.6734553873858425e-05,
      "loss": 1.7995,
      "step": 495500
    },
    {
      "epoch": 20.907979597858617,
      "grad_norm": 2.703317642211914,
      "learning_rate": 1.664880161214252e-05,
      "loss": 1.7988,
      "step": 496000
    },
    {
      "epoch": 20.92905619019517,
      "grad_norm": 3.0928666591644287,
      "learning_rate": 1.656304935042662e-05,
      "loss": 1.8013,
      "step": 496500
    },
    {
      "epoch": 20.95013278253172,
      "grad_norm": 1.8786119222640991,
      "learning_rate": 1.6477297088710717e-05,
      "loss": 1.8,
      "step": 497000
    },
    {
      "epoch": 20.97120937486827,
      "grad_norm": 1.9498240947723389,
      "learning_rate": 1.639154482699481e-05,
      "loss": 1.7982,
      "step": 497500
    },
    {
      "epoch": 20.992285967204822,
      "grad_norm": 1.32588791847229,
      "learning_rate": 1.630579256527891e-05,
      "loss": 1.7987,
      "step": 498000
    },
    {
      "epoch": 21.013362559541374,
      "grad_norm": 1.2315385341644287,
      "learning_rate": 1.6220040303563008e-05,
      "loss": 1.798,
      "step": 498500
    },
    {
      "epoch": 21.034439151877926,
      "grad_norm": 1.7044378519058228,
      "learning_rate": 1.6134288041847103e-05,
      "loss": 1.7967,
      "step": 499000
    },
    {
      "epoch": 21.055515744214475,
      "grad_norm": 2.110351085662842,
      "learning_rate": 1.60485357801312e-05,
      "loss": 1.7914,
      "step": 499500
    },
    {
      "epoch": 21.076592336551027,
      "grad_norm": 1.9396986961364746,
      "learning_rate": 1.59627835184153e-05,
      "loss": 1.7918,
      "step": 500000
    },
    {
      "epoch": 21.076592336551027,
      "eval_accuracy": 0.6808371795763912,
      "eval_loss": 1.6093584299087524,
      "eval_runtime": 163.6158,
      "eval_samples_per_second": 737.74,
      "eval_steps_per_second": 7.689,
      "step": 500000
    },
    {
      "epoch": 21.09766892888758,
      "grad_norm": 3.097768783569336,
      "learning_rate": 1.5877031256699395e-05,
      "loss": 1.7917,
      "step": 500500
    },
    {
      "epoch": 21.118745521224128,
      "grad_norm": 2.0041847229003906,
      "learning_rate": 1.5791278994983493e-05,
      "loss": 1.795,
      "step": 501000
    },
    {
      "epoch": 21.13982211356068,
      "grad_norm": 2.7281696796417236,
      "learning_rate": 1.570552673326759e-05,
      "loss": 1.7906,
      "step": 501500
    },
    {
      "epoch": 21.16089870589723,
      "grad_norm": 2.8058207035064697,
      "learning_rate": 1.5619774471551686e-05,
      "loss": 1.7941,
      "step": 502000
    },
    {
      "epoch": 21.18197529823378,
      "grad_norm": 1.159959077835083,
      "learning_rate": 1.5534022209835785e-05,
      "loss": 1.7937,
      "step": 502500
    },
    {
      "epoch": 21.203051890570332,
      "grad_norm": 2.5489091873168945,
      "learning_rate": 1.544826994811988e-05,
      "loss": 1.7948,
      "step": 503000
    },
    {
      "epoch": 21.224128482906885,
      "grad_norm": 1.8744176626205444,
      "learning_rate": 1.5362517686403978e-05,
      "loss": 1.7916,
      "step": 503500
    },
    {
      "epoch": 21.245205075243433,
      "grad_norm": 2.3084676265716553,
      "learning_rate": 1.5276765424688076e-05,
      "loss": 1.7928,
      "step": 504000
    },
    {
      "epoch": 21.266281667579985,
      "grad_norm": 1.8601198196411133,
      "learning_rate": 1.5191013162972173e-05,
      "loss": 1.7915,
      "step": 504500
    },
    {
      "epoch": 21.287358259916537,
      "grad_norm": 1.5147699117660522,
      "learning_rate": 1.5105260901256271e-05,
      "loss": 1.793,
      "step": 505000
    },
    {
      "epoch": 21.308434852253086,
      "grad_norm": 1.4616293907165527,
      "learning_rate": 1.501950863954037e-05,
      "loss": 1.7942,
      "step": 505500
    },
    {
      "epoch": 21.329511444589638,
      "grad_norm": 3.0656189918518066,
      "learning_rate": 1.4933756377824464e-05,
      "loss": 1.7904,
      "step": 506000
    },
    {
      "epoch": 21.35058803692619,
      "grad_norm": 1.8124321699142456,
      "learning_rate": 1.4848004116108563e-05,
      "loss": 1.7876,
      "step": 506500
    },
    {
      "epoch": 21.371664629262742,
      "grad_norm": 1.3485218286514282,
      "learning_rate": 1.4762251854392661e-05,
      "loss": 1.7943,
      "step": 507000
    },
    {
      "epoch": 21.39274122159929,
      "grad_norm": 2.5013110637664795,
      "learning_rate": 1.4676499592676758e-05,
      "loss": 1.797,
      "step": 507500
    },
    {
      "epoch": 21.413817813935843,
      "grad_norm": 2.357736110687256,
      "learning_rate": 1.4590747330960856e-05,
      "loss": 1.791,
      "step": 508000
    },
    {
      "epoch": 21.434894406272395,
      "grad_norm": 1.3901863098144531,
      "learning_rate": 1.4504995069244952e-05,
      "loss": 1.7901,
      "step": 508500
    },
    {
      "epoch": 21.455970998608944,
      "grad_norm": 1.7388004064559937,
      "learning_rate": 1.4419242807529049e-05,
      "loss": 1.7912,
      "step": 509000
    },
    {
      "epoch": 21.477047590945496,
      "grad_norm": 2.541179895401001,
      "learning_rate": 1.4333490545813147e-05,
      "loss": 1.7928,
      "step": 509500
    },
    {
      "epoch": 21.498124183282048,
      "grad_norm": 1.2289456129074097,
      "learning_rate": 1.4247738284097246e-05,
      "loss": 1.7896,
      "step": 510000
    },
    {
      "epoch": 21.498124183282048,
      "eval_accuracy": 0.681210701677776,
      "eval_loss": 1.6054974794387817,
      "eval_runtime": 163.4307,
      "eval_samples_per_second": 738.576,
      "eval_steps_per_second": 7.697,
      "step": 510000
    },
    {
      "epoch": 21.519200775618597,
      "grad_norm": 2.8922178745269775,
      "learning_rate": 1.416198602238134e-05,
      "loss": 1.7948,
      "step": 510500
    },
    {
      "epoch": 21.54027736795515,
      "grad_norm": 1.5152218341827393,
      "learning_rate": 1.4076233760665439e-05,
      "loss": 1.7916,
      "step": 511000
    },
    {
      "epoch": 21.5613539602917,
      "grad_norm": 2.4753239154815674,
      "learning_rate": 1.3990481498949534e-05,
      "loss": 1.7909,
      "step": 511500
    },
    {
      "epoch": 21.58243055262825,
      "grad_norm": 2.015897274017334,
      "learning_rate": 1.3904729237233632e-05,
      "loss": 1.7916,
      "step": 512000
    },
    {
      "epoch": 21.6035071449648,
      "grad_norm": 2.0406131744384766,
      "learning_rate": 1.381897697551773e-05,
      "loss": 1.7907,
      "step": 512500
    },
    {
      "epoch": 21.624583737301354,
      "grad_norm": 1.3544546365737915,
      "learning_rate": 1.3733224713801825e-05,
      "loss": 1.794,
      "step": 513000
    },
    {
      "epoch": 21.645660329637906,
      "grad_norm": 1.5273191928863525,
      "learning_rate": 1.3647472452085924e-05,
      "loss": 1.7963,
      "step": 513500
    },
    {
      "epoch": 21.666736921974454,
      "grad_norm": 1.2591642141342163,
      "learning_rate": 1.3561720190370022e-05,
      "loss": 1.7929,
      "step": 514000
    },
    {
      "epoch": 21.687813514311006,
      "grad_norm": 2.816277503967285,
      "learning_rate": 1.3475967928654119e-05,
      "loss": 1.7905,
      "step": 514500
    },
    {
      "epoch": 21.70889010664756,
      "grad_norm": 2.991250991821289,
      "learning_rate": 1.3390215666938217e-05,
      "loss": 1.7939,
      "step": 515000
    },
    {
      "epoch": 21.729966698984107,
      "grad_norm": 2.66918683052063,
      "learning_rate": 1.3304463405222315e-05,
      "loss": 1.7921,
      "step": 515500
    },
    {
      "epoch": 21.75104329132066,
      "grad_norm": 1.236002802848816,
      "learning_rate": 1.321871114350641e-05,
      "loss": 1.7949,
      "step": 516000
    },
    {
      "epoch": 21.77211988365721,
      "grad_norm": 2.1234371662139893,
      "learning_rate": 1.3132958881790508e-05,
      "loss": 1.7906,
      "step": 516500
    },
    {
      "epoch": 21.79319647599376,
      "grad_norm": 1.916748046875,
      "learning_rate": 1.3047206620074607e-05,
      "loss": 1.79,
      "step": 517000
    },
    {
      "epoch": 21.814273068330312,
      "grad_norm": 2.3915822505950928,
      "learning_rate": 1.2961454358358702e-05,
      "loss": 1.7934,
      "step": 517500
    },
    {
      "epoch": 21.835349660666864,
      "grad_norm": 3.523082733154297,
      "learning_rate": 1.28757020966428e-05,
      "loss": 1.7884,
      "step": 518000
    },
    {
      "epoch": 21.856426253003413,
      "grad_norm": 2.0066192150115967,
      "learning_rate": 1.2789949834926898e-05,
      "loss": 1.7879,
      "step": 518500
    },
    {
      "epoch": 21.877502845339965,
      "grad_norm": 1.4498072862625122,
      "learning_rate": 1.2704197573210993e-05,
      "loss": 1.7968,
      "step": 519000
    },
    {
      "epoch": 21.898579437676517,
      "grad_norm": 1.8761427402496338,
      "learning_rate": 1.2618445311495092e-05,
      "loss": 1.7887,
      "step": 519500
    },
    {
      "epoch": 21.91965603001307,
      "grad_norm": 2.6337530612945557,
      "learning_rate": 1.2532693049779186e-05,
      "loss": 1.7885,
      "step": 520000
    },
    {
      "epoch": 21.91965603001307,
      "eval_accuracy": 0.6814033761039692,
      "eval_loss": 1.6037622690200806,
      "eval_runtime": 165.6529,
      "eval_samples_per_second": 728.668,
      "eval_steps_per_second": 7.594,
      "step": 520000
    },
    {
      "epoch": 21.940732622349618,
      "grad_norm": 2.6158289909362793,
      "learning_rate": 1.2446940788063285e-05,
      "loss": 1.7904,
      "step": 520500
    },
    {
      "epoch": 21.96180921468617,
      "grad_norm": 1.2750651836395264,
      "learning_rate": 1.2361188526347383e-05,
      "loss": 1.7918,
      "step": 521000
    },
    {
      "epoch": 21.982885807022722,
      "grad_norm": 1.223039984703064,
      "learning_rate": 1.227543626463148e-05,
      "loss": 1.7916,
      "step": 521500
    },
    {
      "epoch": 22.00396239935927,
      "grad_norm": 3.169057607650757,
      "learning_rate": 1.2189684002915578e-05,
      "loss": 1.7869,
      "step": 522000
    },
    {
      "epoch": 22.025038991695823,
      "grad_norm": 2.807615041732788,
      "learning_rate": 1.2103931741199675e-05,
      "loss": 1.7838,
      "step": 522500
    },
    {
      "epoch": 22.046115584032375,
      "grad_norm": 2.122692584991455,
      "learning_rate": 1.2018179479483773e-05,
      "loss": 1.7855,
      "step": 523000
    },
    {
      "epoch": 22.067192176368923,
      "grad_norm": 2.2609195709228516,
      "learning_rate": 1.193242721776787e-05,
      "loss": 1.7858,
      "step": 523500
    },
    {
      "epoch": 22.088268768705476,
      "grad_norm": 4.2342095375061035,
      "learning_rate": 1.1846674956051966e-05,
      "loss": 1.7883,
      "step": 524000
    },
    {
      "epoch": 22.109345361042028,
      "grad_norm": 3.2997336387634277,
      "learning_rate": 1.1760922694336064e-05,
      "loss": 1.7831,
      "step": 524500
    },
    {
      "epoch": 22.130421953378576,
      "grad_norm": 1.1845436096191406,
      "learning_rate": 1.1675170432620161e-05,
      "loss": 1.7882,
      "step": 525000
    },
    {
      "epoch": 22.15149854571513,
      "grad_norm": 3.630873203277588,
      "learning_rate": 1.1589418170904258e-05,
      "loss": 1.782,
      "step": 525500
    },
    {
      "epoch": 22.17257513805168,
      "grad_norm": 2.0185015201568604,
      "learning_rate": 1.1503665909188354e-05,
      "loss": 1.7835,
      "step": 526000
    },
    {
      "epoch": 22.193651730388233,
      "grad_norm": 2.547515869140625,
      "learning_rate": 1.1417913647472453e-05,
      "loss": 1.7852,
      "step": 526500
    },
    {
      "epoch": 22.21472832272478,
      "grad_norm": 2.034256935119629,
      "learning_rate": 1.133216138575655e-05,
      "loss": 1.7854,
      "step": 527000
    },
    {
      "epoch": 22.235804915061333,
      "grad_norm": 1.414648413658142,
      "learning_rate": 1.1246409124040646e-05,
      "loss": 1.7845,
      "step": 527500
    },
    {
      "epoch": 22.256881507397885,
      "grad_norm": 2.759115695953369,
      "learning_rate": 1.1160656862324744e-05,
      "loss": 1.7885,
      "step": 528000
    },
    {
      "epoch": 22.277958099734434,
      "grad_norm": 2.522829532623291,
      "learning_rate": 1.1074904600608842e-05,
      "loss": 1.7847,
      "step": 528500
    },
    {
      "epoch": 22.299034692070986,
      "grad_norm": 2.7815096378326416,
      "learning_rate": 1.0989152338892939e-05,
      "loss": 1.7837,
      "step": 529000
    },
    {
      "epoch": 22.320111284407538,
      "grad_norm": 1.1811890602111816,
      "learning_rate": 1.0903400077177037e-05,
      "loss": 1.7862,
      "step": 529500
    },
    {
      "epoch": 22.341187876744087,
      "grad_norm": 1.2759037017822266,
      "learning_rate": 1.0817647815461134e-05,
      "loss": 1.7882,
      "step": 530000
    },
    {
      "epoch": 22.341187876744087,
      "eval_accuracy": 0.6816172849305121,
      "eval_loss": 1.6067790985107422,
      "eval_runtime": 164.7784,
      "eval_samples_per_second": 732.535,
      "eval_steps_per_second": 7.634,
      "step": 530000
    },
    {
      "epoch": 22.36226446908064,
      "grad_norm": 2.3184103965759277,
      "learning_rate": 1.073189555374523e-05,
      "loss": 1.7877,
      "step": 530500
    },
    {
      "epoch": 22.38334106141719,
      "grad_norm": 2.0394628047943115,
      "learning_rate": 1.0646143292029327e-05,
      "loss": 1.7887,
      "step": 531000
    },
    {
      "epoch": 22.40441765375374,
      "grad_norm": 2.9373385906219482,
      "learning_rate": 1.0560391030313425e-05,
      "loss": 1.7814,
      "step": 531500
    },
    {
      "epoch": 22.42549424609029,
      "grad_norm": 2.4053289890289307,
      "learning_rate": 1.0474638768597522e-05,
      "loss": 1.7798,
      "step": 532000
    },
    {
      "epoch": 22.446570838426844,
      "grad_norm": 3.2647790908813477,
      "learning_rate": 1.0388886506881619e-05,
      "loss": 1.7841,
      "step": 532500
    },
    {
      "epoch": 22.467647430763392,
      "grad_norm": 1.577077865600586,
      "learning_rate": 1.0303134245165717e-05,
      "loss": 1.7863,
      "step": 533000
    },
    {
      "epoch": 22.488724023099945,
      "grad_norm": 1.844565987586975,
      "learning_rate": 1.0217381983449814e-05,
      "loss": 1.7832,
      "step": 533500
    },
    {
      "epoch": 22.509800615436497,
      "grad_norm": 1.7986823320388794,
      "learning_rate": 1.013162972173391e-05,
      "loss": 1.7858,
      "step": 534000
    },
    {
      "epoch": 22.53087720777305,
      "grad_norm": 2.3455095291137695,
      "learning_rate": 1.0045877460018009e-05,
      "loss": 1.7867,
      "step": 534500
    },
    {
      "epoch": 22.551953800109597,
      "grad_norm": 1.1820951700210571,
      "learning_rate": 9.960125198302105e-06,
      "loss": 1.7881,
      "step": 535000
    },
    {
      "epoch": 22.57303039244615,
      "grad_norm": 2.59078311920166,
      "learning_rate": 9.874372936586203e-06,
      "loss": 1.7838,
      "step": 535500
    },
    {
      "epoch": 22.5941069847827,
      "grad_norm": 2.28763747215271,
      "learning_rate": 9.7886206748703e-06,
      "loss": 1.7848,
      "step": 536000
    },
    {
      "epoch": 22.61518357711925,
      "grad_norm": 1.1617317199707031,
      "learning_rate": 9.702868413154398e-06,
      "loss": 1.7785,
      "step": 536500
    },
    {
      "epoch": 22.636260169455802,
      "grad_norm": 1.6714962720870972,
      "learning_rate": 9.617116151438495e-06,
      "loss": 1.7811,
      "step": 537000
    },
    {
      "epoch": 22.657336761792354,
      "grad_norm": 1.976488709449768,
      "learning_rate": 9.531363889722592e-06,
      "loss": 1.7842,
      "step": 537500
    },
    {
      "epoch": 22.678413354128903,
      "grad_norm": 2.936720132827759,
      "learning_rate": 9.44561162800669e-06,
      "loss": 1.7819,
      "step": 538000
    },
    {
      "epoch": 22.699489946465455,
      "grad_norm": 2.037707567214966,
      "learning_rate": 9.359859366290787e-06,
      "loss": 1.7858,
      "step": 538500
    },
    {
      "epoch": 22.720566538802007,
      "grad_norm": 2.8426458835601807,
      "learning_rate": 9.274107104574883e-06,
      "loss": 1.7862,
      "step": 539000
    },
    {
      "epoch": 22.741643131138556,
      "grad_norm": 2.507862091064453,
      "learning_rate": 9.18835484285898e-06,
      "loss": 1.7793,
      "step": 539500
    },
    {
      "epoch": 22.762719723475108,
      "grad_norm": 1.3730566501617432,
      "learning_rate": 9.102602581143078e-06,
      "loss": 1.7826,
      "step": 540000
    },
    {
      "epoch": 22.762719723475108,
      "eval_accuracy": 0.6819036482196468,
      "eval_loss": 1.6104704141616821,
      "eval_runtime": 170.0484,
      "eval_samples_per_second": 709.833,
      "eval_steps_per_second": 7.398,
      "step": 540000
    },
    {
      "epoch": 22.78379631581166,
      "grad_norm": 1.658623456954956,
      "learning_rate": 9.016850319427175e-06,
      "loss": 1.7869,
      "step": 540500
    },
    {
      "epoch": 22.804872908148212,
      "grad_norm": 2.5721659660339355,
      "learning_rate": 8.931098057711271e-06,
      "loss": 1.7832,
      "step": 541000
    },
    {
      "epoch": 22.82594950048476,
      "grad_norm": 1.4334765672683716,
      "learning_rate": 8.84534579599537e-06,
      "loss": 1.7818,
      "step": 541500
    },
    {
      "epoch": 22.847026092821313,
      "grad_norm": 1.5134004354476929,
      "learning_rate": 8.759593534279468e-06,
      "loss": 1.7862,
      "step": 542000
    },
    {
      "epoch": 22.868102685157865,
      "grad_norm": 2.4156453609466553,
      "learning_rate": 8.673841272563565e-06,
      "loss": 1.7819,
      "step": 542500
    },
    {
      "epoch": 22.889179277494414,
      "grad_norm": 1.1913425922393799,
      "learning_rate": 8.588089010847663e-06,
      "loss": 1.7852,
      "step": 543000
    },
    {
      "epoch": 22.910255869830966,
      "grad_norm": 3.83417010307312,
      "learning_rate": 8.50233674913176e-06,
      "loss": 1.7841,
      "step": 543500
    },
    {
      "epoch": 22.931332462167518,
      "grad_norm": 2.492175340652466,
      "learning_rate": 8.416584487415856e-06,
      "loss": 1.7806,
      "step": 544000
    },
    {
      "epoch": 22.952409054504066,
      "grad_norm": 2.8138508796691895,
      "learning_rate": 8.330832225699953e-06,
      "loss": 1.7782,
      "step": 544500
    },
    {
      "epoch": 22.97348564684062,
      "grad_norm": 1.9995532035827637,
      "learning_rate": 8.245079963984051e-06,
      "loss": 1.7786,
      "step": 545000
    },
    {
      "epoch": 22.99456223917717,
      "grad_norm": 1.2055670022964478,
      "learning_rate": 8.159327702268148e-06,
      "loss": 1.7849,
      "step": 545500
    },
    {
      "epoch": 23.01563883151372,
      "grad_norm": 7.2733659744262695,
      "learning_rate": 8.073575440552244e-06,
      "loss": 1.7803,
      "step": 546000
    },
    {
      "epoch": 23.03671542385027,
      "grad_norm": 2.3799593448638916,
      "learning_rate": 7.987823178836343e-06,
      "loss": 1.7798,
      "step": 546500
    },
    {
      "epoch": 23.057792016186824,
      "grad_norm": 1.2151885032653809,
      "learning_rate": 7.90207091712044e-06,
      "loss": 1.7769,
      "step": 547000
    },
    {
      "epoch": 23.078868608523376,
      "grad_norm": 1.172594666481018,
      "learning_rate": 7.816318655404536e-06,
      "loss": 1.7785,
      "step": 547500
    },
    {
      "epoch": 23.099945200859924,
      "grad_norm": 4.345755100250244,
      "learning_rate": 7.730566393688634e-06,
      "loss": 1.7831,
      "step": 548000
    },
    {
      "epoch": 23.121021793196476,
      "grad_norm": 2.249272108078003,
      "learning_rate": 7.64481413197273e-06,
      "loss": 1.7793,
      "step": 548500
    },
    {
      "epoch": 23.14209838553303,
      "grad_norm": 1.4738367795944214,
      "learning_rate": 7.559061870256828e-06,
      "loss": 1.7786,
      "step": 549000
    },
    {
      "epoch": 23.163174977869577,
      "grad_norm": 1.1760754585266113,
      "learning_rate": 7.473309608540925e-06,
      "loss": 1.7778,
      "step": 549500
    },
    {
      "epoch": 23.18425157020613,
      "grad_norm": 4.902884483337402,
      "learning_rate": 7.387557346825023e-06,
      "loss": 1.7816,
      "step": 550000
    },
    {
      "epoch": 23.18425157020613,
      "eval_accuracy": 0.6821593127784792,
      "eval_loss": 1.6100155115127563,
      "eval_runtime": 166.8874,
      "eval_samples_per_second": 723.278,
      "eval_steps_per_second": 7.538,
      "step": 550000
    },
    {
      "epoch": 23.20532816254268,
      "grad_norm": 3.02134108543396,
      "learning_rate": 7.3018050851091205e-06,
      "loss": 1.7762,
      "step": 550500
    },
    {
      "epoch": 23.22640475487923,
      "grad_norm": 3.4128761291503906,
      "learning_rate": 7.216052823393217e-06,
      "loss": 1.7802,
      "step": 551000
    },
    {
      "epoch": 23.247481347215782,
      "grad_norm": 3.9114174842834473,
      "learning_rate": 7.1303005616773155e-06,
      "loss": 1.7797,
      "step": 551500
    },
    {
      "epoch": 23.268557939552334,
      "grad_norm": 2.38464617729187,
      "learning_rate": 7.044548299961412e-06,
      "loss": 1.7846,
      "step": 552000
    },
    {
      "epoch": 23.289634531888883,
      "grad_norm": 1.2752312421798706,
      "learning_rate": 6.958796038245509e-06,
      "loss": 1.7712,
      "step": 552500
    },
    {
      "epoch": 23.310711124225435,
      "grad_norm": 2.189749002456665,
      "learning_rate": 6.873043776529606e-06,
      "loss": 1.7821,
      "step": 553000
    },
    {
      "epoch": 23.331787716561987,
      "grad_norm": 3.171915292739868,
      "learning_rate": 6.787291514813704e-06,
      "loss": 1.7776,
      "step": 553500
    },
    {
      "epoch": 23.35286430889854,
      "grad_norm": 1.2632099390029907,
      "learning_rate": 6.701539253097801e-06,
      "loss": 1.7771,
      "step": 554000
    },
    {
      "epoch": 23.373940901235088,
      "grad_norm": 2.149463415145874,
      "learning_rate": 6.615786991381898e-06,
      "loss": 1.7789,
      "step": 554500
    },
    {
      "epoch": 23.39501749357164,
      "grad_norm": 2.140155076980591,
      "learning_rate": 6.530034729665996e-06,
      "loss": 1.7796,
      "step": 555000
    },
    {
      "epoch": 23.416094085908192,
      "grad_norm": 1.2059754133224487,
      "learning_rate": 6.444282467950093e-06,
      "loss": 1.7761,
      "step": 555500
    },
    {
      "epoch": 23.43717067824474,
      "grad_norm": 1.9622489213943481,
      "learning_rate": 6.358530206234189e-06,
      "loss": 1.7764,
      "step": 556000
    },
    {
      "epoch": 23.458247270581293,
      "grad_norm": 1.1801263093948364,
      "learning_rate": 6.272777944518287e-06,
      "loss": 1.777,
      "step": 556500
    },
    {
      "epoch": 23.479323862917845,
      "grad_norm": 1.1938841342926025,
      "learning_rate": 6.187025682802384e-06,
      "loss": 1.7787,
      "step": 557000
    },
    {
      "epoch": 23.500400455254393,
      "grad_norm": 2.810182571411133,
      "learning_rate": 6.101273421086482e-06,
      "loss": 1.7788,
      "step": 557500
    },
    {
      "epoch": 23.521477047590945,
      "grad_norm": 1.3208647966384888,
      "learning_rate": 6.015521159370579e-06,
      "loss": 1.7766,
      "step": 558000
    },
    {
      "epoch": 23.542553639927497,
      "grad_norm": 2.456364154815674,
      "learning_rate": 5.929768897654676e-06,
      "loss": 1.7751,
      "step": 558500
    },
    {
      "epoch": 23.563630232264046,
      "grad_norm": 1.321045160293579,
      "learning_rate": 5.844016635938773e-06,
      "loss": 1.7767,
      "step": 559000
    },
    {
      "epoch": 23.584706824600598,
      "grad_norm": 2.2366201877593994,
      "learning_rate": 5.75826437422287e-06,
      "loss": 1.7771,
      "step": 559500
    },
    {
      "epoch": 23.60578341693715,
      "grad_norm": 3.362916946411133,
      "learning_rate": 5.672512112506967e-06,
      "loss": 1.7788,
      "step": 560000
    },
    {
      "epoch": 23.60578341693715,
      "eval_accuracy": 0.6828102130785378,
      "eval_loss": 1.6040875911712646,
      "eval_runtime": 165.2318,
      "eval_samples_per_second": 730.525,
      "eval_steps_per_second": 7.614,
      "step": 560000
    },
    {
      "epoch": 23.6268600092737,
      "grad_norm": 1.4292680025100708,
      "learning_rate": 5.5867598507910655e-06,
      "loss": 1.7779,
      "step": 560500
    },
    {
      "epoch": 23.64793660161025,
      "grad_norm": 2.126281499862671,
      "learning_rate": 5.501007589075162e-06,
      "loss": 1.7769,
      "step": 561000
    },
    {
      "epoch": 23.669013193946803,
      "grad_norm": 1.3507832288742065,
      "learning_rate": 5.41525532735926e-06,
      "loss": 1.7786,
      "step": 561500
    },
    {
      "epoch": 23.690089786283355,
      "grad_norm": 1.316178798675537,
      "learning_rate": 5.329503065643356e-06,
      "loss": 1.7727,
      "step": 562000
    },
    {
      "epoch": 23.711166378619904,
      "grad_norm": 1.48749577999115,
      "learning_rate": 5.243750803927454e-06,
      "loss": 1.775,
      "step": 562500
    },
    {
      "epoch": 23.732242970956456,
      "grad_norm": 5.032774925231934,
      "learning_rate": 5.157998542211551e-06,
      "loss": 1.7769,
      "step": 563000
    },
    {
      "epoch": 23.753319563293008,
      "grad_norm": 2.346156597137451,
      "learning_rate": 5.072246280495649e-06,
      "loss": 1.7737,
      "step": 563500
    },
    {
      "epoch": 23.774396155629557,
      "grad_norm": 1.3793861865997314,
      "learning_rate": 4.986494018779746e-06,
      "loss": 1.7772,
      "step": 564000
    },
    {
      "epoch": 23.79547274796611,
      "grad_norm": 2.513613700866699,
      "learning_rate": 4.900741757063843e-06,
      "loss": 1.7768,
      "step": 564500
    },
    {
      "epoch": 23.81654934030266,
      "grad_norm": 2.5091381072998047,
      "learning_rate": 4.81498949534794e-06,
      "loss": 1.7728,
      "step": 565000
    },
    {
      "epoch": 23.83762593263921,
      "grad_norm": 1.2946932315826416,
      "learning_rate": 4.729237233632038e-06,
      "loss": 1.7747,
      "step": 565500
    },
    {
      "epoch": 23.85870252497576,
      "grad_norm": 1.283813238143921,
      "learning_rate": 4.643484971916134e-06,
      "loss": 1.7775,
      "step": 566000
    },
    {
      "epoch": 23.879779117312314,
      "grad_norm": 4.075824737548828,
      "learning_rate": 4.557732710200232e-06,
      "loss": 1.7743,
      "step": 566500
    },
    {
      "epoch": 23.900855709648862,
      "grad_norm": 2.4480645656585693,
      "learning_rate": 4.471980448484329e-06,
      "loss": 1.7727,
      "step": 567000
    },
    {
      "epoch": 23.921932301985414,
      "grad_norm": 1.1916335821151733,
      "learning_rate": 4.386228186768427e-06,
      "loss": 1.7763,
      "step": 567500
    },
    {
      "epoch": 23.943008894321967,
      "grad_norm": 1.349599003791809,
      "learning_rate": 4.300475925052524e-06,
      "loss": 1.7748,
      "step": 568000
    },
    {
      "epoch": 23.96408548665852,
      "grad_norm": 2.2206592559814453,
      "learning_rate": 4.214723663336621e-06,
      "loss": 1.7799,
      "step": 568500
    },
    {
      "epoch": 23.985162078995067,
      "grad_norm": 2.4559051990509033,
      "learning_rate": 4.128971401620718e-06,
      "loss": 1.7743,
      "step": 569000
    },
    {
      "epoch": 24.00623867133162,
      "grad_norm": 1.3668975830078125,
      "learning_rate": 4.043219139904815e-06,
      "loss": 1.7757,
      "step": 569500
    },
    {
      "epoch": 24.02731526366817,
      "grad_norm": 3.144969940185547,
      "learning_rate": 3.957466878188912e-06,
      "loss": 1.772,
      "step": 570000
    },
    {
      "epoch": 24.02731526366817,
      "eval_accuracy": 0.6827260225116205,
      "eval_loss": 1.6117669343948364,
      "eval_runtime": 164.1689,
      "eval_samples_per_second": 735.255,
      "eval_steps_per_second": 7.663,
      "step": 570000
    },
    {
      "epoch": 24.04839185600472,
      "grad_norm": 2.380990982055664,
      "learning_rate": 3.87171461647301e-06,
      "loss": 1.776,
      "step": 570500
    },
    {
      "epoch": 24.069468448341272,
      "grad_norm": 1.421928882598877,
      "learning_rate": 3.7859623547571067e-06,
      "loss": 1.7702,
      "step": 571000
    },
    {
      "epoch": 24.090545040677824,
      "grad_norm": 2.5571649074554443,
      "learning_rate": 3.7002100930412046e-06,
      "loss": 1.7696,
      "step": 571500
    },
    {
      "epoch": 24.111621633014373,
      "grad_norm": 3.185093641281128,
      "learning_rate": 3.614457831325301e-06,
      "loss": 1.7783,
      "step": 572000
    },
    {
      "epoch": 24.132698225350925,
      "grad_norm": 1.5451873540878296,
      "learning_rate": 3.5287055696093987e-06,
      "loss": 1.7751,
      "step": 572500
    },
    {
      "epoch": 24.153774817687477,
      "grad_norm": 1.1850438117980957,
      "learning_rate": 3.4429533078934957e-06,
      "loss": 1.7748,
      "step": 573000
    },
    {
      "epoch": 24.174851410024026,
      "grad_norm": 1.6448487043380737,
      "learning_rate": 3.357201046177593e-06,
      "loss": 1.7771,
      "step": 573500
    },
    {
      "epoch": 24.195928002360578,
      "grad_norm": 4.053676128387451,
      "learning_rate": 3.2714487844616906e-06,
      "loss": 1.776,
      "step": 574000
    },
    {
      "epoch": 24.21700459469713,
      "grad_norm": 2.390244483947754,
      "learning_rate": 3.1856965227457872e-06,
      "loss": 1.7767,
      "step": 574500
    },
    {
      "epoch": 24.238081187033682,
      "grad_norm": 1.164123296737671,
      "learning_rate": 3.099944261029885e-06,
      "loss": 1.7775,
      "step": 575000
    },
    {
      "epoch": 24.25915777937023,
      "grad_norm": 2.965951919555664,
      "learning_rate": 3.014191999313982e-06,
      "loss": 1.7725,
      "step": 575500
    },
    {
      "epoch": 24.280234371706783,
      "grad_norm": 2.2808890342712402,
      "learning_rate": 2.928439737598079e-06,
      "loss": 1.7708,
      "step": 576000
    },
    {
      "epoch": 24.301310964043335,
      "grad_norm": 3.3673813343048096,
      "learning_rate": 2.8426874758821767e-06,
      "loss": 1.7752,
      "step": 576500
    },
    {
      "epoch": 24.322387556379883,
      "grad_norm": 1.2826939821243286,
      "learning_rate": 2.7569352141662737e-06,
      "loss": 1.7691,
      "step": 577000
    },
    {
      "epoch": 24.343464148716436,
      "grad_norm": 1.275975227355957,
      "learning_rate": 2.6711829524503707e-06,
      "loss": 1.7704,
      "step": 577500
    },
    {
      "epoch": 24.364540741052988,
      "grad_norm": 1.2287765741348267,
      "learning_rate": 2.585430690734468e-06,
      "loss": 1.7695,
      "step": 578000
    },
    {
      "epoch": 24.385617333389536,
      "grad_norm": 1.4020841121673584,
      "learning_rate": 2.4996784290185657e-06,
      "loss": 1.7734,
      "step": 578500
    },
    {
      "epoch": 24.40669392572609,
      "grad_norm": 3.6417651176452637,
      "learning_rate": 2.4139261673026627e-06,
      "loss": 1.7678,
      "step": 579000
    },
    {
      "epoch": 24.42777051806264,
      "grad_norm": 3.0534164905548096,
      "learning_rate": 2.3281739055867597e-06,
      "loss": 1.7704,
      "step": 579500
    },
    {
      "epoch": 24.44884711039919,
      "grad_norm": 3.4166922569274902,
      "learning_rate": 2.242421643870857e-06,
      "loss": 1.7736,
      "step": 580000
    },
    {
      "epoch": 24.44884711039919,
      "eval_accuracy": 0.6834791117630824,
      "eval_loss": 1.6020063161849976,
      "eval_runtime": 164.7626,
      "eval_samples_per_second": 732.606,
      "eval_steps_per_second": 7.635,
      "step": 580000
    },
    {
      "epoch": 24.46992370273574,
      "grad_norm": 3.35719895362854,
      "learning_rate": 2.1566693821549547e-06,
      "loss": 1.7691,
      "step": 580500
    },
    {
      "epoch": 24.491000295072293,
      "grad_norm": 3.078352928161621,
      "learning_rate": 2.0709171204390517e-06,
      "loss": 1.777,
      "step": 581000
    },
    {
      "epoch": 24.512076887408845,
      "grad_norm": 2.7058184146881104,
      "learning_rate": 1.985164858723149e-06,
      "loss": 1.7686,
      "step": 581500
    },
    {
      "epoch": 24.533153479745394,
      "grad_norm": 2.7066807746887207,
      "learning_rate": 1.8994125970072462e-06,
      "loss": 1.7742,
      "step": 582000
    },
    {
      "epoch": 24.554230072081946,
      "grad_norm": 1.1883063316345215,
      "learning_rate": 1.8136603352913432e-06,
      "loss": 1.7694,
      "step": 582500
    },
    {
      "epoch": 24.5753066644185,
      "grad_norm": 2.1207635402679443,
      "learning_rate": 1.7279080735754405e-06,
      "loss": 1.7748,
      "step": 583000
    },
    {
      "epoch": 24.596383256755047,
      "grad_norm": 2.5423316955566406,
      "learning_rate": 1.642155811859538e-06,
      "loss": 1.7759,
      "step": 583500
    },
    {
      "epoch": 24.6174598490916,
      "grad_norm": 1.3468568325042725,
      "learning_rate": 1.5564035501436352e-06,
      "loss": 1.7675,
      "step": 584000
    },
    {
      "epoch": 24.63853644142815,
      "grad_norm": 1.276308298110962,
      "learning_rate": 1.4706512884277324e-06,
      "loss": 1.7735,
      "step": 584500
    },
    {
      "epoch": 24.6596130337647,
      "grad_norm": 1.3233325481414795,
      "learning_rate": 1.3848990267118295e-06,
      "loss": 1.7701,
      "step": 585000
    },
    {
      "epoch": 24.680689626101252,
      "grad_norm": 1.4558098316192627,
      "learning_rate": 1.299146764995927e-06,
      "loss": 1.7707,
      "step": 585500
    },
    {
      "epoch": 24.701766218437804,
      "grad_norm": 4.167192459106445,
      "learning_rate": 1.213394503280024e-06,
      "loss": 1.7707,
      "step": 586000
    },
    {
      "epoch": 24.722842810774353,
      "grad_norm": 1.1916297674179077,
      "learning_rate": 1.1276422415641212e-06,
      "loss": 1.7707,
      "step": 586500
    },
    {
      "epoch": 24.743919403110905,
      "grad_norm": 2.7956340312957764,
      "learning_rate": 1.0418899798482187e-06,
      "loss": 1.7725,
      "step": 587000
    },
    {
      "epoch": 24.764995995447457,
      "grad_norm": 1.3004568815231323,
      "learning_rate": 9.561377181323157e-07,
      "loss": 1.7699,
      "step": 587500
    },
    {
      "epoch": 24.786072587784005,
      "grad_norm": 3.5255401134490967,
      "learning_rate": 8.70385456416413e-07,
      "loss": 1.7682,
      "step": 588000
    },
    {
      "epoch": 24.807149180120557,
      "grad_norm": 2.059431314468384,
      "learning_rate": 7.846331947005103e-07,
      "loss": 1.7738,
      "step": 588500
    },
    {
      "epoch": 24.82822577245711,
      "grad_norm": 1.1846002340316772,
      "learning_rate": 6.988809329846075e-07,
      "loss": 1.7666,
      "step": 589000
    },
    {
      "epoch": 24.84930236479366,
      "grad_norm": 1.4435313940048218,
      "learning_rate": 6.131286712687047e-07,
      "loss": 1.7704,
      "step": 589500
    },
    {
      "epoch": 24.87037895713021,
      "grad_norm": 3.019223213195801,
      "learning_rate": 5.27376409552802e-07,
      "loss": 1.7698,
      "step": 590000
    },
    {
      "epoch": 24.87037895713021,
      "eval_accuracy": 0.6829699607278547,
      "eval_loss": 1.6105071306228638,
      "eval_runtime": 165.1137,
      "eval_samples_per_second": 731.048,
      "eval_steps_per_second": 7.619,
      "step": 590000
    },
    {
      "epoch": 24.891455549466762,
      "grad_norm": 2.4003806114196777,
      "learning_rate": 4.4162414783689917e-07,
      "loss": 1.7712,
      "step": 590500
    },
    {
      "epoch": 24.912532141803315,
      "grad_norm": 1.176041841506958,
      "learning_rate": 3.558718861209964e-07,
      "loss": 1.7698,
      "step": 591000
    },
    {
      "epoch": 24.933608734139863,
      "grad_norm": 2.868479013442993,
      "learning_rate": 2.701196244050937e-07,
      "loss": 1.7708,
      "step": 591500
    },
    {
      "epoch": 24.954685326476415,
      "grad_norm": 3.838606834411621,
      "learning_rate": 1.8436736268919094e-07,
      "loss": 1.7716,
      "step": 592000
    },
    {
      "epoch": 24.975761918812967,
      "grad_norm": 2.015836238861084,
      "learning_rate": 9.861510097328816e-08,
      "loss": 1.7703,
      "step": 592500
    },
    {
      "epoch": 24.996838511149516,
      "grad_norm": 1.2224476337432861,
      "learning_rate": 1.2862839257385414e-08,
      "loss": 1.7712,
      "step": 593000
    },
    {
      "epoch": 25.0,
      "step": 593075,
      "total_flos": 3.037727852389325e+18,
      "train_loss": 2.208342132639112,
      "train_runtime": 127183.0636,
      "train_samples_per_second": 447.65,
      "train_steps_per_second": 4.663
    }
  ],
  "logging_steps": 500,
  "max_steps": 593075,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 25,
  "save_steps": 10000,
  "total_flos": 3.037727852389325e+18,
  "train_batch_size": 96,
  "trial_name": null,
  "trial_params": null
}