{
  "best_metric": 3.305659294128418,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_495/checkpoint-90000",
  "epoch": 10.0,
  "eval_steps": 1000,
  "global_step": 92910,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005381552039608223,
      "grad_norm": 1.9801093339920044,
      "learning_rate": 0.0003,
      "loss": 8.624,
      "step": 50
    },
    {
      "epoch": 0.010763104079216447,
      "grad_norm": 2.859027624130249,
      "learning_rate": 0.0006,
      "loss": 6.8551,
      "step": 100
    },
    {
      "epoch": 0.01614465611882467,
      "grad_norm": 2.4920058250427246,
      "learning_rate": 0.0005996767589699385,
      "loss": 6.4231,
      "step": 150
    },
    {
      "epoch": 0.021526208158432893,
      "grad_norm": 1.792466402053833,
      "learning_rate": 0.0005993535179398771,
      "loss": 6.1851,
      "step": 200
    },
    {
      "epoch": 0.026907760198041114,
      "grad_norm": 0.997410774230957,
      "learning_rate": 0.0005990302769098158,
      "loss": 6.0752,
      "step": 250
    },
    {
      "epoch": 0.03228931223764934,
      "grad_norm": 4.0290021896362305,
      "learning_rate": 0.0005987070358797543,
      "loss": 5.9368,
      "step": 300
    },
    {
      "epoch": 0.03767086427725756,
      "grad_norm": 1.1890099048614502,
      "learning_rate": 0.0005983837948496929,
      "loss": 5.881,
      "step": 350
    },
    {
      "epoch": 0.04305241631686579,
      "grad_norm": 1.0169402360916138,
      "learning_rate": 0.0005980605538196314,
      "loss": 5.7963,
      "step": 400
    },
    {
      "epoch": 0.048433968356474004,
      "grad_norm": 1.1916236877441406,
      "learning_rate": 0.0005977373127895701,
      "loss": 5.7266,
      "step": 450
    },
    {
      "epoch": 0.05381552039608223,
      "grad_norm": 1.1587761640548706,
      "learning_rate": 0.0005974140717595086,
      "loss": 5.6237,
      "step": 500
    },
    {
      "epoch": 0.05919707243569045,
      "grad_norm": 1.7533999681472778,
      "learning_rate": 0.0005970908307294472,
      "loss": 5.5532,
      "step": 550
    },
    {
      "epoch": 0.06457862447529868,
      "grad_norm": 1.3879553079605103,
      "learning_rate": 0.0005967675896993858,
      "loss": 5.506,
      "step": 600
    },
    {
      "epoch": 0.0699601765149069,
      "grad_norm": 1.314697504043579,
      "learning_rate": 0.0005964443486693243,
      "loss": 5.4294,
      "step": 650
    },
    {
      "epoch": 0.07534172855451512,
      "grad_norm": 1.3169344663619995,
      "learning_rate": 0.000596121107639263,
      "loss": 5.3783,
      "step": 700
    },
    {
      "epoch": 0.08072328059412334,
      "grad_norm": 1.2692291736602783,
      "learning_rate": 0.0005957978666092015,
      "loss": 5.297,
      "step": 750
    },
    {
      "epoch": 0.08610483263373157,
      "grad_norm": 0.7749007344245911,
      "learning_rate": 0.0005954746255791401,
      "loss": 5.2401,
      "step": 800
    },
    {
      "epoch": 0.09148638467333979,
      "grad_norm": 1.3382775783538818,
      "learning_rate": 0.0005951513845490787,
      "loss": 5.2177,
      "step": 850
    },
    {
      "epoch": 0.09686793671294801,
      "grad_norm": 1.1966224908828735,
      "learning_rate": 0.0005948281435190174,
      "loss": 5.1556,
      "step": 900
    },
    {
      "epoch": 0.10224948875255624,
      "grad_norm": 1.2725766897201538,
      "learning_rate": 0.0005945049024889559,
      "loss": 5.1106,
      "step": 950
    },
    {
      "epoch": 0.10763104079216446,
      "grad_norm": 1.2319713830947876,
      "learning_rate": 0.0005941816614588944,
      "loss": 5.057,
      "step": 1000
    },
    {
      "epoch": 0.10763104079216446,
      "eval_accuracy": 0.22853244391528074,
      "eval_loss": 5.010529041290283,
      "eval_runtime": 187.3434,
      "eval_samples_per_second": 96.139,
      "eval_steps_per_second": 6.01,
      "step": 1000
    },
    {
      "epoch": 0.11301259283177269,
      "grad_norm": 1.3787147998809814,
      "learning_rate": 0.000593858420428833,
      "loss": 5.0349,
      "step": 1050
    },
    {
      "epoch": 0.1183941448713809,
      "grad_norm": 1.5601210594177246,
      "learning_rate": 0.0005935351793987716,
      "loss": 4.9823,
      "step": 1100
    },
    {
      "epoch": 0.12377569691098914,
      "grad_norm": 1.1213598251342773,
      "learning_rate": 0.0005932119383687103,
      "loss": 4.9711,
      "step": 1150
    },
    {
      "epoch": 0.12915724895059735,
      "grad_norm": 1.1033381223678589,
      "learning_rate": 0.0005928886973386488,
      "loss": 4.916,
      "step": 1200
    },
    {
      "epoch": 0.13453880099020557,
      "grad_norm": 1.2041442394256592,
      "learning_rate": 0.0005925654563085874,
      "loss": 4.9033,
      "step": 1250
    },
    {
      "epoch": 0.1399203530298138,
      "grad_norm": 1.2366397380828857,
      "learning_rate": 0.000592242215278526,
      "loss": 4.9068,
      "step": 1300
    },
    {
      "epoch": 0.14530190506942203,
      "grad_norm": 1.0430039167404175,
      "learning_rate": 0.0005919189742484645,
      "loss": 4.8667,
      "step": 1350
    },
    {
      "epoch": 0.15068345710903025,
      "grad_norm": 1.1846684217453003,
      "learning_rate": 0.0005915957332184032,
      "loss": 4.8115,
      "step": 1400
    },
    {
      "epoch": 0.15606500914863847,
      "grad_norm": 1.0357134342193604,
      "learning_rate": 0.0005912724921883417,
      "loss": 4.7965,
      "step": 1450
    },
    {
      "epoch": 0.16144656118824668,
      "grad_norm": 1.0831513404846191,
      "learning_rate": 0.0005909492511582803,
      "loss": 4.7803,
      "step": 1500
    },
    {
      "epoch": 0.1668281132278549,
      "grad_norm": 0.9524378180503845,
      "learning_rate": 0.0005906260101282189,
      "loss": 4.7587,
      "step": 1550
    },
    {
      "epoch": 0.17220966526746315,
      "grad_norm": 1.0448411703109741,
      "learning_rate": 0.0005903027690981575,
      "loss": 4.7423,
      "step": 1600
    },
    {
      "epoch": 0.17759121730707136,
      "grad_norm": 1.3444677591323853,
      "learning_rate": 0.000589979528068096,
      "loss": 4.7112,
      "step": 1650
    },
    {
      "epoch": 0.18297276934667958,
      "grad_norm": 1.1023002862930298,
      "learning_rate": 0.0005896562870380347,
      "loss": 4.6949,
      "step": 1700
    },
    {
      "epoch": 0.1883543213862878,
      "grad_norm": 1.0242769718170166,
      "learning_rate": 0.0005893330460079732,
      "loss": 4.6725,
      "step": 1750
    },
    {
      "epoch": 0.19373587342589602,
      "grad_norm": 0.745267927646637,
      "learning_rate": 0.0005890098049779118,
      "loss": 4.6581,
      "step": 1800
    },
    {
      "epoch": 0.19911742546550426,
      "grad_norm": 1.1536999940872192,
      "learning_rate": 0.0005886865639478504,
      "loss": 4.6568,
      "step": 1850
    },
    {
      "epoch": 0.20449897750511248,
      "grad_norm": 1.0055886507034302,
      "learning_rate": 0.0005883633229177889,
      "loss": 4.6189,
      "step": 1900
    },
    {
      "epoch": 0.2098805295447207,
      "grad_norm": 1.636452317237854,
      "learning_rate": 0.0005880400818877276,
      "loss": 4.5954,
      "step": 1950
    },
    {
      "epoch": 0.2152620815843289,
      "grad_norm": 0.9592578411102295,
      "learning_rate": 0.0005877168408576662,
      "loss": 4.5755,
      "step": 2000
    },
    {
      "epoch": 0.2152620815843289,
      "eval_accuracy": 0.2712195558682827,
      "eval_loss": 4.50430154800415,
      "eval_runtime": 185.2595,
      "eval_samples_per_second": 97.22,
      "eval_steps_per_second": 6.078,
      "step": 2000
    },
    {
      "epoch": 0.22064363362393713,
      "grad_norm": 1.0056291818618774,
      "learning_rate": 0.0005873935998276048,
      "loss": 4.5635,
      "step": 2050
    },
    {
      "epoch": 0.22602518566354537,
      "grad_norm": 0.8652420043945312,
      "learning_rate": 0.0005870703587975433,
      "loss": 4.5397,
      "step": 2100
    },
    {
      "epoch": 0.2314067377031536,
      "grad_norm": 0.7207990288734436,
      "learning_rate": 0.0005867471177674818,
      "loss": 4.5214,
      "step": 2150
    },
    {
      "epoch": 0.2367882897427618,
      "grad_norm": 0.8635744452476501,
      "learning_rate": 0.0005864238767374205,
      "loss": 4.5113,
      "step": 2200
    },
    {
      "epoch": 0.24216984178237003,
      "grad_norm": 1.0317870378494263,
      "learning_rate": 0.0005861006357073591,
      "loss": 4.5058,
      "step": 2250
    },
    {
      "epoch": 0.24755139382197827,
      "grad_norm": 0.7938442826271057,
      "learning_rate": 0.0005857773946772977,
      "loss": 4.4739,
      "step": 2300
    },
    {
      "epoch": 0.2529329458615865,
      "grad_norm": 0.9807633757591248,
      "learning_rate": 0.0005854541536472362,
      "loss": 4.4825,
      "step": 2350
    },
    {
      "epoch": 0.2583144979011947,
      "grad_norm": 0.8235756158828735,
      "learning_rate": 0.0005851309126171749,
      "loss": 4.4583,
      "step": 2400
    },
    {
      "epoch": 0.2636960499408029,
      "grad_norm": 0.8362206816673279,
      "learning_rate": 0.0005848076715871134,
      "loss": 4.4043,
      "step": 2450
    },
    {
      "epoch": 0.26907760198041114,
      "grad_norm": 0.7393255233764648,
      "learning_rate": 0.000584484430557052,
      "loss": 4.4326,
      "step": 2500
    },
    {
      "epoch": 0.27445915402001936,
      "grad_norm": 0.7527334690093994,
      "learning_rate": 0.0005841611895269906,
      "loss": 4.4116,
      "step": 2550
    },
    {
      "epoch": 0.2798407060596276,
      "grad_norm": 0.8830164074897766,
      "learning_rate": 0.0005838379484969291,
      "loss": 4.3955,
      "step": 2600
    },
    {
      "epoch": 0.2852222580992358,
      "grad_norm": 0.7846371531486511,
      "learning_rate": 0.0005835147074668678,
      "loss": 4.3818,
      "step": 2650
    },
    {
      "epoch": 0.29060381013884407,
      "grad_norm": 0.7156797647476196,
      "learning_rate": 0.0005831914664368063,
      "loss": 4.3662,
      "step": 2700
    },
    {
      "epoch": 0.2959853621784523,
      "grad_norm": 0.9031903743743896,
      "learning_rate": 0.0005828682254067449,
      "loss": 4.3505,
      "step": 2750
    },
    {
      "epoch": 0.3013669142180605,
      "grad_norm": 0.7309237122535706,
      "learning_rate": 0.0005825449843766835,
      "loss": 4.3415,
      "step": 2800
    },
    {
      "epoch": 0.3067484662576687,
      "grad_norm": 0.8120318651199341,
      "learning_rate": 0.0005822217433466221,
      "loss": 4.3262,
      "step": 2850
    },
    {
      "epoch": 0.31213001829727693,
      "grad_norm": 0.7812032103538513,
      "learning_rate": 0.0005818985023165607,
      "loss": 4.3159,
      "step": 2900
    },
    {
      "epoch": 0.31751157033688515,
      "grad_norm": 0.8197738528251648,
      "learning_rate": 0.0005815752612864992,
      "loss": 4.3067,
      "step": 2950
    },
    {
      "epoch": 0.32289312237649337,
      "grad_norm": 0.7830599546432495,
      "learning_rate": 0.0005812520202564378,
      "loss": 4.2958,
      "step": 3000
    },
    {
      "epoch": 0.32289312237649337,
      "eval_accuracy": 0.2981761200292798,
      "eval_loss": 4.236570835113525,
      "eval_runtime": 185.3458,
      "eval_samples_per_second": 97.175,
      "eval_steps_per_second": 6.075,
      "step": 3000
    },
    {
      "epoch": 0.3282746744161016,
      "grad_norm": 0.7051645517349243,
      "learning_rate": 0.0005809287792263764,
      "loss": 4.3043,
      "step": 3050
    },
    {
      "epoch": 0.3336562264557098,
      "grad_norm": 0.7232017517089844,
      "learning_rate": 0.0005806055381963151,
      "loss": 4.2875,
      "step": 3100
    },
    {
      "epoch": 0.3390377784953181,
      "grad_norm": 0.7313389182090759,
      "learning_rate": 0.0005802822971662536,
      "loss": 4.2679,
      "step": 3150
    },
    {
      "epoch": 0.3444193305349263,
      "grad_norm": 0.5924382209777832,
      "learning_rate": 0.0005799590561361922,
      "loss": 4.2807,
      "step": 3200
    },
    {
      "epoch": 0.3498008825745345,
      "grad_norm": 0.7239092588424683,
      "learning_rate": 0.0005796358151061307,
      "loss": 4.2701,
      "step": 3250
    },
    {
      "epoch": 0.35518243461414273,
      "grad_norm": 0.6883352994918823,
      "learning_rate": 0.0005793125740760694,
      "loss": 4.2548,
      "step": 3300
    },
    {
      "epoch": 0.36056398665375095,
      "grad_norm": 0.8354396224021912,
      "learning_rate": 0.0005789893330460079,
      "loss": 4.2429,
      "step": 3350
    },
    {
      "epoch": 0.36594553869335916,
      "grad_norm": 0.7251695990562439,
      "learning_rate": 0.0005786660920159465,
      "loss": 4.244,
      "step": 3400
    },
    {
      "epoch": 0.3713270907329674,
      "grad_norm": 0.668481707572937,
      "learning_rate": 0.0005783428509858851,
      "loss": 4.2369,
      "step": 3450
    },
    {
      "epoch": 0.3767086427725756,
      "grad_norm": 0.6284920573234558,
      "learning_rate": 0.0005780196099558237,
      "loss": 4.2194,
      "step": 3500
    },
    {
      "epoch": 0.3820901948121838,
      "grad_norm": 0.6607083678245544,
      "learning_rate": 0.0005776963689257623,
      "loss": 4.2187,
      "step": 3550
    },
    {
      "epoch": 0.38747174685179203,
      "grad_norm": 0.7383949756622314,
      "learning_rate": 0.0005773731278957008,
      "loss": 4.2276,
      "step": 3600
    },
    {
      "epoch": 0.3928532988914003,
      "grad_norm": 0.644062340259552,
      "learning_rate": 0.0005770498868656394,
      "loss": 4.2046,
      "step": 3650
    },
    {
      "epoch": 0.3982348509310085,
      "grad_norm": 0.6715877056121826,
      "learning_rate": 0.000576726645835578,
      "loss": 4.1975,
      "step": 3700
    },
    {
      "epoch": 0.40361640297061674,
      "grad_norm": 0.8043977618217468,
      "learning_rate": 0.0005764034048055167,
      "loss": 4.2017,
      "step": 3750
    },
    {
      "epoch": 0.40899795501022496,
      "grad_norm": 0.7500972151756287,
      "learning_rate": 0.0005760801637754552,
      "loss": 4.1837,
      "step": 3800
    },
    {
      "epoch": 0.4143795070498332,
      "grad_norm": 0.6272939443588257,
      "learning_rate": 0.0005757569227453937,
      "loss": 4.1859,
      "step": 3850
    },
    {
      "epoch": 0.4197610590894414,
      "grad_norm": 0.6128783226013184,
      "learning_rate": 0.0005754336817153324,
      "loss": 4.1541,
      "step": 3900
    },
    {
      "epoch": 0.4251426111290496,
      "grad_norm": 0.8502875566482544,
      "learning_rate": 0.0005751104406852709,
      "loss": 4.1731,
      "step": 3950
    },
    {
      "epoch": 0.4305241631686578,
      "grad_norm": 1.4366008043289185,
      "learning_rate": 0.0005747871996552096,
      "loss": 4.1549,
      "step": 4000
    },
    {
      "epoch": 0.4305241631686578,
      "eval_accuracy": 0.31271050817933504,
      "eval_loss": 4.089645862579346,
      "eval_runtime": 185.3388,
      "eval_samples_per_second": 97.179,
      "eval_steps_per_second": 6.075,
      "step": 4000
    },
    {
      "epoch": 0.43590571520826604,
      "grad_norm": 0.6601532101631165,
      "learning_rate": 0.0005744639586251481,
      "loss": 4.1427,
      "step": 4050
    },
    {
      "epoch": 0.44128726724787426,
      "grad_norm": 0.7322743535041809,
      "learning_rate": 0.0005741407175950867,
      "loss": 4.1352,
      "step": 4100
    },
    {
      "epoch": 0.44666881928748253,
      "grad_norm": 0.8291831612586975,
      "learning_rate": 0.0005738174765650253,
      "loss": 4.1534,
      "step": 4150
    },
    {
      "epoch": 0.45205037132709075,
      "grad_norm": 0.6917663812637329,
      "learning_rate": 0.0005734942355349638,
      "loss": 4.1426,
      "step": 4200
    },
    {
      "epoch": 0.45743192336669897,
      "grad_norm": 0.7226840257644653,
      "learning_rate": 0.0005731709945049025,
      "loss": 4.1213,
      "step": 4250
    },
    {
      "epoch": 0.4628134754063072,
      "grad_norm": 0.7379767894744873,
      "learning_rate": 0.000572847753474841,
      "loss": 4.101,
      "step": 4300
    },
    {
      "epoch": 0.4681950274459154,
      "grad_norm": 0.6218189597129822,
      "learning_rate": 0.0005725245124447796,
      "loss": 4.129,
      "step": 4350
    },
    {
      "epoch": 0.4735765794855236,
      "grad_norm": 0.5983642935752869,
      "learning_rate": 0.0005722012714147182,
      "loss": 4.1061,
      "step": 4400
    },
    {
      "epoch": 0.47895813152513184,
      "grad_norm": 0.6328372955322266,
      "learning_rate": 0.0005718780303846568,
      "loss": 4.105,
      "step": 4450
    },
    {
      "epoch": 0.48433968356474005,
      "grad_norm": 0.671859860420227,
      "learning_rate": 0.0005715547893545953,
      "loss": 4.1134,
      "step": 4500
    },
    {
      "epoch": 0.48972123560434827,
      "grad_norm": 0.7485026717185974,
      "learning_rate": 0.000571231548324534,
      "loss": 4.1154,
      "step": 4550
    },
    {
      "epoch": 0.49510278764395654,
      "grad_norm": 0.6755270957946777,
      "learning_rate": 0.0005709083072944725,
      "loss": 4.0929,
      "step": 4600
    },
    {
      "epoch": 0.5004843396835648,
      "grad_norm": 0.6335451602935791,
      "learning_rate": 0.0005705850662644111,
      "loss": 4.0824,
      "step": 4650
    },
    {
      "epoch": 0.505865891723173,
      "grad_norm": 0.6942111849784851,
      "learning_rate": 0.0005702618252343497,
      "loss": 4.0921,
      "step": 4700
    },
    {
      "epoch": 0.5112474437627812,
      "grad_norm": 0.6951626539230347,
      "learning_rate": 0.0005699385842042882,
      "loss": 4.0955,
      "step": 4750
    },
    {
      "epoch": 0.5166289958023894,
      "grad_norm": 0.6223372220993042,
      "learning_rate": 0.0005696153431742269,
      "loss": 4.0921,
      "step": 4800
    },
    {
      "epoch": 0.5220105478419976,
      "grad_norm": 0.702920138835907,
      "learning_rate": 0.0005692921021441655,
      "loss": 4.0746,
      "step": 4850
    },
    {
      "epoch": 0.5273920998816058,
      "grad_norm": 0.7055066227912903,
      "learning_rate": 0.0005689688611141041,
      "loss": 4.0654,
      "step": 4900
    },
    {
      "epoch": 0.5327736519212141,
      "grad_norm": 0.6813945770263672,
      "learning_rate": 0.0005686456200840426,
      "loss": 4.0557,
      "step": 4950
    },
    {
      "epoch": 0.5381552039608223,
      "grad_norm": 0.603909969329834,
      "learning_rate": 0.0005683223790539811,
      "loss": 4.0527,
      "step": 5000
    },
    {
      "epoch": 0.5381552039608223,
      "eval_accuracy": 0.3219062366866258,
      "eval_loss": 3.9891772270202637,
      "eval_runtime": 185.3481,
      "eval_samples_per_second": 97.174,
      "eval_steps_per_second": 6.075,
      "step": 5000
    },
    {
      "epoch": 0.5435367560004305,
      "grad_norm": 0.6491262912750244,
      "learning_rate": 0.0005679991380239198,
      "loss": 4.0515,
      "step": 5050
    },
    {
      "epoch": 0.5489183080400387,
      "grad_norm": 0.6999713778495789,
      "learning_rate": 0.0005676758969938584,
      "loss": 4.0289,
      "step": 5100
    },
    {
      "epoch": 0.5542998600796469,
      "grad_norm": 0.6778654456138611,
      "learning_rate": 0.000567352655963797,
      "loss": 4.0356,
      "step": 5150
    },
    {
      "epoch": 0.5596814121192552,
      "grad_norm": 0.6406720280647278,
      "learning_rate": 0.0005670294149337355,
      "loss": 4.0621,
      "step": 5200
    },
    {
      "epoch": 0.5650629641588634,
      "grad_norm": 0.7699031829833984,
      "learning_rate": 0.0005667061739036742,
      "loss": 4.0314,
      "step": 5250
    },
    {
      "epoch": 0.5704445161984716,
      "grad_norm": 0.6253768801689148,
      "learning_rate": 0.0005663829328736127,
      "loss": 4.0586,
      "step": 5300
    },
    {
      "epoch": 0.5758260682380799,
      "grad_norm": 0.6249842047691345,
      "learning_rate": 0.0005660596918435512,
      "loss": 4.0421,
      "step": 5350
    },
    {
      "epoch": 0.5812076202776881,
      "grad_norm": 0.6070592999458313,
      "learning_rate": 0.0005657364508134899,
      "loss": 4.0113,
      "step": 5400
    },
    {
      "epoch": 0.5865891723172963,
      "grad_norm": 0.7027144432067871,
      "learning_rate": 0.0005654132097834284,
      "loss": 4.0403,
      "step": 5450
    },
    {
      "epoch": 0.5919707243569046,
      "grad_norm": 0.5996516346931458,
      "learning_rate": 0.0005650899687533671,
      "loss": 4.0223,
      "step": 5500
    },
    {
      "epoch": 0.5973522763965128,
      "grad_norm": 0.7442830801010132,
      "learning_rate": 0.0005647667277233056,
      "loss": 4.0055,
      "step": 5550
    },
    {
      "epoch": 0.602733828436121,
      "grad_norm": 0.7308864593505859,
      "learning_rate": 0.0005644434866932442,
      "loss": 4.0263,
      "step": 5600
    },
    {
      "epoch": 0.6081153804757292,
      "grad_norm": 0.6208139061927795,
      "learning_rate": 0.0005641202456631828,
      "loss": 4.0076,
      "step": 5650
    },
    {
      "epoch": 0.6134969325153374,
      "grad_norm": 0.6338828802108765,
      "learning_rate": 0.0005637970046331214,
      "loss": 3.9954,
      "step": 5700
    },
    {
      "epoch": 0.6188784845549457,
      "grad_norm": 0.7338832020759583,
      "learning_rate": 0.00056347376360306,
      "loss": 4.0017,
      "step": 5750
    },
    {
      "epoch": 0.6242600365945539,
      "grad_norm": 0.6001290678977966,
      "learning_rate": 0.0005631505225729985,
      "loss": 3.9903,
      "step": 5800
    },
    {
      "epoch": 0.6296415886341621,
      "grad_norm": 0.6113796234130859,
      "learning_rate": 0.0005628272815429371,
      "loss": 3.9839,
      "step": 5850
    },
    {
      "epoch": 0.6350231406737703,
      "grad_norm": 0.6154040098190308,
      "learning_rate": 0.0005625040405128757,
      "loss": 3.984,
      "step": 5900
    },
    {
      "epoch": 0.6404046927133785,
      "grad_norm": 0.6384268999099731,
      "learning_rate": 0.0005621807994828143,
      "loss": 3.9888,
      "step": 5950
    },
    {
      "epoch": 0.6457862447529867,
      "grad_norm": 0.5864706039428711,
      "learning_rate": 0.0005618575584527529,
      "loss": 3.9895,
      "step": 6000
    },
    {
      "epoch": 0.6457862447529867,
      "eval_accuracy": 0.32818648225519065,
      "eval_loss": 3.9195642471313477,
      "eval_runtime": 185.1346,
      "eval_samples_per_second": 97.286,
      "eval_steps_per_second": 6.082,
      "step": 6000
    },
    {
      "epoch": 0.651167796792595,
      "grad_norm": 0.6114383339881897,
      "learning_rate": 0.0005615343174226915,
      "loss": 3.9765,
      "step": 6050
    },
    {
      "epoch": 0.6565493488322032,
      "grad_norm": 0.5866206884384155,
      "learning_rate": 0.00056121107639263,
      "loss": 4.0004,
      "step": 6100
    },
    {
      "epoch": 0.6619309008718114,
      "grad_norm": 0.6376749873161316,
      "learning_rate": 0.0005608878353625687,
      "loss": 3.9844,
      "step": 6150
    },
    {
      "epoch": 0.6673124529114196,
      "grad_norm": 0.5870709419250488,
      "learning_rate": 0.0005605645943325072,
      "loss": 3.9584,
      "step": 6200
    },
    {
      "epoch": 0.6726940049510278,
      "grad_norm": 0.6266024708747864,
      "learning_rate": 0.0005602413533024458,
      "loss": 3.974,
      "step": 6250
    },
    {
      "epoch": 0.6780755569906362,
      "grad_norm": 0.6104730367660522,
      "learning_rate": 0.0005599181122723844,
      "loss": 3.9668,
      "step": 6300
    },
    {
      "epoch": 0.6834571090302444,
      "grad_norm": 0.6267789006233215,
      "learning_rate": 0.000559594871242323,
      "loss": 3.9744,
      "step": 6350
    },
    {
      "epoch": 0.6888386610698526,
      "grad_norm": 0.631796658039093,
      "learning_rate": 0.0005592780950328628,
      "loss": 3.965,
      "step": 6400
    },
    {
      "epoch": 0.6942202131094608,
      "grad_norm": 0.6150023937225342,
      "learning_rate": 0.0005589548540028014,
      "loss": 3.9666,
      "step": 6450
    },
    {
      "epoch": 0.699601765149069,
      "grad_norm": 0.6553137898445129,
      "learning_rate": 0.0005586316129727399,
      "loss": 3.9457,
      "step": 6500
    },
    {
      "epoch": 0.7049833171886772,
      "grad_norm": 0.6123223900794983,
      "learning_rate": 0.0005583083719426786,
      "loss": 3.9427,
      "step": 6550
    },
    {
      "epoch": 0.7103648692282855,
      "grad_norm": 0.6026431918144226,
      "learning_rate": 0.0005579851309126171,
      "loss": 3.9317,
      "step": 6600
    },
    {
      "epoch": 0.7157464212678937,
      "grad_norm": 0.5570296049118042,
      "learning_rate": 0.0005576618898825558,
      "loss": 3.9452,
      "step": 6650
    },
    {
      "epoch": 0.7211279733075019,
      "grad_norm": 0.6147912740707397,
      "learning_rate": 0.0005573386488524943,
      "loss": 3.9549,
      "step": 6700
    },
    {
      "epoch": 0.7265095253471101,
      "grad_norm": 0.8436711430549622,
      "learning_rate": 0.0005570154078224328,
      "loss": 3.9493,
      "step": 6750
    },
    {
      "epoch": 0.7318910773867183,
      "grad_norm": 0.615190863609314,
      "learning_rate": 0.0005566921667923715,
      "loss": 3.9392,
      "step": 6800
    },
    {
      "epoch": 0.7372726294263265,
      "grad_norm": 0.5037049651145935,
      "learning_rate": 0.00055636892576231,
      "loss": 3.9159,
      "step": 6850
    },
    {
      "epoch": 0.7426541814659348,
      "grad_norm": 0.5320571064949036,
      "learning_rate": 0.0005560456847322487,
      "loss": 3.9298,
      "step": 6900
    },
    {
      "epoch": 0.748035733505543,
      "grad_norm": 0.6132500767707825,
      "learning_rate": 0.0005557224437021872,
      "loss": 3.9308,
      "step": 6950
    },
    {
      "epoch": 0.7534172855451512,
      "grad_norm": 0.6210596561431885,
      "learning_rate": 0.0005553992026721258,
      "loss": 3.9326,
      "step": 7000
    },
    {
      "epoch": 0.7534172855451512,
      "eval_accuracy": 0.3332193926716452,
      "eval_loss": 3.8627383708953857,
      "eval_runtime": 185.2754,
      "eval_samples_per_second": 97.212,
      "eval_steps_per_second": 6.077,
      "step": 7000
    },
    {
      "epoch": 0.7587988375847594,
      "grad_norm": 0.5304657816886902,
      "learning_rate": 0.0005550759616420644,
      "loss": 3.9257,
      "step": 7050
    },
    {
      "epoch": 0.7641803896243676,
      "grad_norm": 0.7234488129615784,
      "learning_rate": 0.000554752720612003,
      "loss": 3.8996,
      "step": 7100
    },
    {
      "epoch": 0.7695619416639758,
      "grad_norm": 0.6656004190444946,
      "learning_rate": 0.0005544294795819415,
      "loss": 3.9271,
      "step": 7150
    },
    {
      "epoch": 0.7749434937035841,
      "grad_norm": 0.5789263844490051,
      "learning_rate": 0.0005541062385518801,
      "loss": 3.9016,
      "step": 7200
    },
    {
      "epoch": 0.7803250457431924,
      "grad_norm": 0.5520836710929871,
      "learning_rate": 0.0005537829975218188,
      "loss": 3.9054,
      "step": 7250
    },
    {
      "epoch": 0.7857065977828006,
      "grad_norm": 0.574553370475769,
      "learning_rate": 0.0005534597564917573,
      "loss": 3.8942,
      "step": 7300
    },
    {
      "epoch": 0.7910881498224088,
      "grad_norm": 0.572311520576477,
      "learning_rate": 0.0005531365154616959,
      "loss": 3.9114,
      "step": 7350
    },
    {
      "epoch": 0.796469701862017,
      "grad_norm": 0.6432904601097107,
      "learning_rate": 0.0005528132744316344,
      "loss": 3.9139,
      "step": 7400
    },
    {
      "epoch": 0.8018512539016253,
      "grad_norm": 0.5829313397407532,
      "learning_rate": 0.0005524900334015731,
      "loss": 3.9118,
      "step": 7450
    },
    {
      "epoch": 0.8072328059412335,
      "grad_norm": 0.5401226878166199,
      "learning_rate": 0.0005521667923715117,
      "loss": 3.8962,
      "step": 7500
    },
    {
      "epoch": 0.8126143579808417,
      "grad_norm": 0.6198664307594299,
      "learning_rate": 0.0005518435513414502,
      "loss": 3.9022,
      "step": 7550
    },
    {
      "epoch": 0.8179959100204499,
      "grad_norm": 0.6631735563278198,
      "learning_rate": 0.0005515203103113888,
      "loss": 3.8971,
      "step": 7600
    },
    {
      "epoch": 0.8233774620600581,
      "grad_norm": 0.6021295785903931,
      "learning_rate": 0.0005511970692813274,
      "loss": 3.912,
      "step": 7650
    },
    {
      "epoch": 0.8287590140996663,
      "grad_norm": 0.5508933067321777,
      "learning_rate": 0.000550873828251266,
      "loss": 3.9121,
      "step": 7700
    },
    {
      "epoch": 0.8341405661392746,
      "grad_norm": 0.5896097421646118,
      "learning_rate": 0.0005505505872212045,
      "loss": 3.8972,
      "step": 7750
    },
    {
      "epoch": 0.8395221181788828,
      "grad_norm": 0.7598628401756287,
      "learning_rate": 0.0005502273461911432,
      "loss": 3.9188,
      "step": 7800
    },
    {
      "epoch": 0.844903670218491,
      "grad_norm": 0.6770401000976562,
      "learning_rate": 0.0005499041051610817,
      "loss": 3.8951,
      "step": 7850
    },
    {
      "epoch": 0.8502852222580992,
      "grad_norm": 0.5735957622528076,
      "learning_rate": 0.0005495808641310204,
      "loss": 3.8904,
      "step": 7900
    },
    {
      "epoch": 0.8556667742977074,
      "grad_norm": 0.5606273412704468,
      "learning_rate": 0.0005492576231009589,
      "loss": 3.8575,
      "step": 7950
    },
    {
      "epoch": 0.8610483263373157,
      "grad_norm": 0.6246755123138428,
      "learning_rate": 0.0005489343820708974,
      "loss": 3.8965,
      "step": 8000
    },
    {
      "epoch": 0.8610483263373157,
      "eval_accuracy": 0.3376259191898493,
      "eval_loss": 3.816678524017334,
      "eval_runtime": 185.2734,
      "eval_samples_per_second": 97.213,
      "eval_steps_per_second": 6.078,
      "step": 8000
    },
    {
      "epoch": 0.8664298783769239,
      "grad_norm": 0.6273600459098816,
      "learning_rate": 0.0005486111410408361,
      "loss": 3.8801,
      "step": 8050
    },
    {
      "epoch": 0.8718114304165321,
      "grad_norm": 0.6346429586410522,
      "learning_rate": 0.0005482879000107746,
      "loss": 3.8892,
      "step": 8100
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.6020395159721375,
      "learning_rate": 0.0005479646589807133,
      "loss": 3.8638,
      "step": 8150
    },
    {
      "epoch": 0.8825745344957485,
      "grad_norm": 0.6787419319152832,
      "learning_rate": 0.0005476414179506518,
      "loss": 3.872,
      "step": 8200
    },
    {
      "epoch": 0.8879560865353568,
      "grad_norm": 0.5587417483329773,
      "learning_rate": 0.0005473181769205904,
      "loss": 3.8723,
      "step": 8250
    },
    {
      "epoch": 0.8933376385749651,
      "grad_norm": 0.5795436501502991,
      "learning_rate": 0.000546994935890529,
      "loss": 3.8719,
      "step": 8300
    },
    {
      "epoch": 0.8987191906145733,
      "grad_norm": 0.8941925764083862,
      "learning_rate": 0.0005466716948604677,
      "loss": 3.8702,
      "step": 8350
    },
    {
      "epoch": 0.9041007426541815,
      "grad_norm": 0.5755844712257385,
      "learning_rate": 0.0005463549186510073,
      "loss": 3.8721,
      "step": 8400
    },
    {
      "epoch": 0.9094822946937897,
      "grad_norm": 0.5602589845657349,
      "learning_rate": 0.000546031677620946,
      "loss": 3.8721,
      "step": 8450
    },
    {
      "epoch": 0.9148638467333979,
      "grad_norm": 0.5489923357963562,
      "learning_rate": 0.0005457084365908845,
      "loss": 3.8561,
      "step": 8500
    },
    {
      "epoch": 0.9202453987730062,
      "grad_norm": 0.5321398973464966,
      "learning_rate": 0.0005453851955608232,
      "loss": 3.8527,
      "step": 8550
    },
    {
      "epoch": 0.9256269508126144,
      "grad_norm": 0.5661652088165283,
      "learning_rate": 0.0005450619545307617,
      "loss": 3.8506,
      "step": 8600
    },
    {
      "epoch": 0.9310085028522226,
      "grad_norm": 0.5366695523262024,
      "learning_rate": 0.0005447387135007003,
      "loss": 3.8556,
      "step": 8650
    },
    {
      "epoch": 0.9363900548918308,
      "grad_norm": 0.5779429078102112,
      "learning_rate": 0.0005444154724706389,
      "loss": 3.853,
      "step": 8700
    },
    {
      "epoch": 0.941771606931439,
      "grad_norm": 0.8201158046722412,
      "learning_rate": 0.0005440922314405775,
      "loss": 3.8705,
      "step": 8750
    },
    {
      "epoch": 0.9471531589710472,
      "grad_norm": 0.5916135907173157,
      "learning_rate": 0.0005437689904105161,
      "loss": 3.8668,
      "step": 8800
    },
    {
      "epoch": 0.9525347110106555,
      "grad_norm": 0.5199770331382751,
      "learning_rate": 0.0005434457493804546,
      "loss": 3.8566,
      "step": 8850
    },
    {
      "epoch": 0.9579162630502637,
      "grad_norm": 0.628406286239624,
      "learning_rate": 0.0005431225083503932,
      "loss": 3.8451,
      "step": 8900
    },
    {
      "epoch": 0.9632978150898719,
      "grad_norm": 0.5287910103797913,
      "learning_rate": 0.0005427992673203318,
      "loss": 3.8502,
      "step": 8950
    },
    {
      "epoch": 0.9686793671294801,
      "grad_norm": 0.7501053214073181,
      "learning_rate": 0.0005424760262902704,
      "loss": 3.8452,
      "step": 9000
    },
    {
      "epoch": 0.9686793671294801,
      "eval_accuracy": 0.3411169364753286,
      "eval_loss": 3.7813615798950195,
      "eval_runtime": 184.8778,
      "eval_samples_per_second": 97.421,
      "eval_steps_per_second": 6.091,
      "step": 9000
    },
    {
      "epoch": 0.9740609191690883,
      "grad_norm": 0.5607249736785889,
      "learning_rate": 0.000542152785260209,
      "loss": 3.839,
      "step": 9050
    },
    {
      "epoch": 0.9794424712086965,
      "grad_norm": 0.6981595158576965,
      "learning_rate": 0.0005418295442301476,
      "loss": 3.8262,
      "step": 9100
    },
    {
      "epoch": 0.9848240232483048,
      "grad_norm": 0.5787471532821655,
      "learning_rate": 0.0005415063032000861,
      "loss": 3.8276,
      "step": 9150
    },
    {
      "epoch": 0.9902055752879131,
      "grad_norm": 0.5677970051765442,
      "learning_rate": 0.0005411830621700248,
      "loss": 3.8363,
      "step": 9200
    },
    {
      "epoch": 0.9955871273275213,
      "grad_norm": 0.5393511056900024,
      "learning_rate": 0.0005408598211399633,
      "loss": 3.8247,
      "step": 9250
    },
    {
      "epoch": 1.0009686793671295,
      "grad_norm": 0.5939765572547913,
      "learning_rate": 0.0005405365801099019,
      "loss": 3.8102,
      "step": 9300
    },
    {
      "epoch": 1.0063502314067376,
      "grad_norm": 0.5469721555709839,
      "learning_rate": 0.0005402133390798405,
      "loss": 3.7693,
      "step": 9350
    },
    {
      "epoch": 1.011731783446346,
      "grad_norm": 0.5030762553215027,
      "learning_rate": 0.000539890098049779,
      "loss": 3.7643,
      "step": 9400
    },
    {
      "epoch": 1.017113335485954,
      "grad_norm": 0.5492759346961975,
      "learning_rate": 0.0005395668570197177,
      "loss": 3.7637,
      "step": 9450
    },
    {
      "epoch": 1.0224948875255624,
      "grad_norm": 0.5253005623817444,
      "learning_rate": 0.0005392436159896562,
      "loss": 3.7705,
      "step": 9500
    },
    {
      "epoch": 1.0278764395651705,
      "grad_norm": 0.5683625936508179,
      "learning_rate": 0.0005389203749595948,
      "loss": 3.7519,
      "step": 9550
    },
    {
      "epoch": 1.0332579916047788,
      "grad_norm": 0.6160693764686584,
      "learning_rate": 0.0005385971339295334,
      "loss": 3.769,
      "step": 9600
    },
    {
      "epoch": 1.0386395436443872,
      "grad_norm": 0.57390296459198,
      "learning_rate": 0.000538273892899472,
      "loss": 3.7481,
      "step": 9650
    },
    {
      "epoch": 1.0440210956839953,
      "grad_norm": 0.5775496959686279,
      "learning_rate": 0.0005379506518694106,
      "loss": 3.7758,
      "step": 9700
    },
    {
      "epoch": 1.0494026477236036,
      "grad_norm": 0.5826638340950012,
      "learning_rate": 0.0005376274108393491,
      "loss": 3.7489,
      "step": 9750
    },
    {
      "epoch": 1.0547841997632117,
      "grad_norm": 0.5193576216697693,
      "learning_rate": 0.0005373041698092877,
      "loss": 3.7715,
      "step": 9800
    },
    {
      "epoch": 1.06016575180282,
      "grad_norm": 0.56683349609375,
      "learning_rate": 0.0005369809287792263,
      "loss": 3.7549,
      "step": 9850
    },
    {
      "epoch": 1.0655473038424281,
      "grad_norm": 0.6041780114173889,
      "learning_rate": 0.000536657687749165,
      "loss": 3.7564,
      "step": 9900
    },
    {
      "epoch": 1.0709288558820365,
      "grad_norm": 0.49289658665657043,
      "learning_rate": 0.0005363344467191035,
      "loss": 3.7664,
      "step": 9950
    },
    {
      "epoch": 1.0763104079216446,
      "grad_norm": 0.635551929473877,
      "learning_rate": 0.000536011205689042,
      "loss": 3.764,
      "step": 10000
    },
    {
      "epoch": 1.0763104079216446,
      "eval_accuracy": 0.34442661209104547,
      "eval_loss": 3.7476959228515625,
      "eval_runtime": 184.9845,
      "eval_samples_per_second": 97.365,
      "eval_steps_per_second": 6.087,
      "step": 10000
    },
    {
      "epoch": 1.081691959961253,
      "grad_norm": 0.5482255816459656,
      "learning_rate": 0.0005356879646589807,
      "loss": 3.7717,
      "step": 10050
    },
    {
      "epoch": 1.087073512000861,
      "grad_norm": 0.5536448955535889,
      "learning_rate": 0.0005353647236289192,
      "loss": 3.7666,
      "step": 10100
    },
    {
      "epoch": 1.0924550640404693,
      "grad_norm": 0.6858832240104675,
      "learning_rate": 0.0005350414825988579,
      "loss": 3.7624,
      "step": 10150
    },
    {
      "epoch": 1.0978366160800774,
      "grad_norm": 0.5250205993652344,
      "learning_rate": 0.0005347182415687964,
      "loss": 3.7447,
      "step": 10200
    },
    {
      "epoch": 1.1032181681196858,
      "grad_norm": 0.5747717022895813,
      "learning_rate": 0.000534395000538735,
      "loss": 3.7618,
      "step": 10250
    },
    {
      "epoch": 1.1085997201592939,
      "grad_norm": 0.5861135721206665,
      "learning_rate": 0.0005340717595086736,
      "loss": 3.7609,
      "step": 10300
    },
    {
      "epoch": 1.1139812721989022,
      "grad_norm": 0.5522782802581787,
      "learning_rate": 0.0005337485184786122,
      "loss": 3.7537,
      "step": 10350
    },
    {
      "epoch": 1.1193628242385103,
      "grad_norm": 0.5878082513809204,
      "learning_rate": 0.0005334252774485507,
      "loss": 3.7661,
      "step": 10400
    },
    {
      "epoch": 1.1247443762781186,
      "grad_norm": 0.5400363802909851,
      "learning_rate": 0.0005331020364184894,
      "loss": 3.7642,
      "step": 10450
    },
    {
      "epoch": 1.1301259283177267,
      "grad_norm": 0.5566787719726562,
      "learning_rate": 0.0005327787953884279,
      "loss": 3.7642,
      "step": 10500
    },
    {
      "epoch": 1.135507480357335,
      "grad_norm": 0.5809233784675598,
      "learning_rate": 0.0005324555543583665,
      "loss": 3.7652,
      "step": 10550
    },
    {
      "epoch": 1.1408890323969434,
      "grad_norm": 0.5614021420478821,
      "learning_rate": 0.0005321323133283051,
      "loss": 3.7699,
      "step": 10600
    },
    {
      "epoch": 1.1462705844365515,
      "grad_norm": 0.5702725052833557,
      "learning_rate": 0.0005318155371188449,
      "loss": 3.7726,
      "step": 10650
    },
    {
      "epoch": 1.1516521364761596,
      "grad_norm": 0.5950555205345154,
      "learning_rate": 0.0005314922960887834,
      "loss": 3.7595,
      "step": 10700
    },
    {
      "epoch": 1.157033688515768,
      "grad_norm": 0.518501341342926,
      "learning_rate": 0.0005311690550587221,
      "loss": 3.7587,
      "step": 10750
    },
    {
      "epoch": 1.1624152405553763,
      "grad_norm": 0.5681332945823669,
      "learning_rate": 0.0005308458140286607,
      "loss": 3.7554,
      "step": 10800
    },
    {
      "epoch": 1.1677967925949844,
      "grad_norm": 0.5314465165138245,
      "learning_rate": 0.0005305225729985993,
      "loss": 3.7504,
      "step": 10850
    },
    {
      "epoch": 1.1731783446345927,
      "grad_norm": 0.5260189175605774,
      "learning_rate": 0.0005301993319685378,
      "loss": 3.7521,
      "step": 10900
    },
    {
      "epoch": 1.1785598966742008,
      "grad_norm": 0.6054186224937439,
      "learning_rate": 0.0005298760909384765,
      "loss": 3.7538,
      "step": 10950
    },
    {
      "epoch": 1.1839414487138091,
      "grad_norm": 0.5630669593811035,
      "learning_rate": 0.000529552849908415,
      "loss": 3.7532,
      "step": 11000
    },
    {
      "epoch": 1.1839414487138091,
      "eval_accuracy": 0.34686119734830456,
      "eval_loss": 3.7236130237579346,
      "eval_runtime": 184.9646,
      "eval_samples_per_second": 97.375,
      "eval_steps_per_second": 6.088,
      "step": 11000
    },
    {
      "epoch": 1.1893230007534172,
      "grad_norm": 0.5823944807052612,
      "learning_rate": 0.0005292296088783535,
      "loss": 3.747,
      "step": 11050
    },
    {
      "epoch": 1.1947045527930256,
      "grad_norm": 0.5315054059028625,
      "learning_rate": 0.0005289063678482922,
      "loss": 3.7539,
      "step": 11100
    },
    {
      "epoch": 1.2000861048326337,
      "grad_norm": 0.5131509304046631,
      "learning_rate": 0.0005285831268182307,
      "loss": 3.7597,
      "step": 11150
    },
    {
      "epoch": 1.205467656872242,
      "grad_norm": 0.5881217122077942,
      "learning_rate": 0.0005282598857881694,
      "loss": 3.7435,
      "step": 11200
    },
    {
      "epoch": 1.21084920891185,
      "grad_norm": 0.5732146501541138,
      "learning_rate": 0.0005279366447581079,
      "loss": 3.743,
      "step": 11250
    },
    {
      "epoch": 1.2162307609514584,
      "grad_norm": 0.5278974771499634,
      "learning_rate": 0.0005276134037280465,
      "loss": 3.7225,
      "step": 11300
    },
    {
      "epoch": 1.2216123129910665,
      "grad_norm": 0.6023641228675842,
      "learning_rate": 0.0005272901626979851,
      "loss": 3.7489,
      "step": 11350
    },
    {
      "epoch": 1.2269938650306749,
      "grad_norm": 0.6074120998382568,
      "learning_rate": 0.0005269669216679236,
      "loss": 3.7493,
      "step": 11400
    },
    {
      "epoch": 1.232375417070283,
      "grad_norm": 0.6074919700622559,
      "learning_rate": 0.0005266436806378623,
      "loss": 3.7526,
      "step": 11450
    },
    {
      "epoch": 1.2377569691098913,
      "grad_norm": 0.6139492392539978,
      "learning_rate": 0.0005263204396078008,
      "loss": 3.7316,
      "step": 11500
    },
    {
      "epoch": 1.2431385211494996,
      "grad_norm": 0.5588337182998657,
      "learning_rate": 0.0005259971985777394,
      "loss": 3.7405,
      "step": 11550
    },
    {
      "epoch": 1.2485200731891077,
      "grad_norm": 0.6142476201057434,
      "learning_rate": 0.000525673957547678,
      "loss": 3.7467,
      "step": 11600
    },
    {
      "epoch": 1.2539016252287158,
      "grad_norm": 0.6412455439567566,
      "learning_rate": 0.0005253507165176167,
      "loss": 3.7455,
      "step": 11650
    },
    {
      "epoch": 1.2592831772683242,
      "grad_norm": 0.5440448522567749,
      "learning_rate": 0.0005250274754875552,
      "loss": 3.7402,
      "step": 11700
    },
    {
      "epoch": 1.2646647293079325,
      "grad_norm": 0.5540177226066589,
      "learning_rate": 0.0005247042344574938,
      "loss": 3.7444,
      "step": 11750
    },
    {
      "epoch": 1.2700462813475406,
      "grad_norm": 0.5741671919822693,
      "learning_rate": 0.0005243809934274323,
      "loss": 3.7444,
      "step": 11800
    },
    {
      "epoch": 1.275427833387149,
      "grad_norm": 0.601311445236206,
      "learning_rate": 0.0005240577523973709,
      "loss": 3.7318,
      "step": 11850
    },
    {
      "epoch": 1.280809385426757,
      "grad_norm": 0.5957517027854919,
      "learning_rate": 0.0005237345113673095,
      "loss": 3.7311,
      "step": 11900
    },
    {
      "epoch": 1.2861909374663654,
      "grad_norm": 0.6441899538040161,
      "learning_rate": 0.0005234112703372481,
      "loss": 3.7361,
      "step": 11950
    },
    {
      "epoch": 1.2915724895059735,
      "grad_norm": 0.6594961881637573,
      "learning_rate": 0.0005230880293071867,
      "loss": 3.7299,
      "step": 12000
    },
    {
      "epoch": 1.2915724895059735,
      "eval_accuracy": 0.3495079816954653,
      "eval_loss": 3.698274612426758,
      "eval_runtime": 184.7232,
      "eval_samples_per_second": 97.503,
      "eval_steps_per_second": 6.096,
      "step": 12000
    },
    {
      "epoch": 1.2969540415455818,
      "grad_norm": 0.5729615092277527,
      "learning_rate": 0.0005227647882771253,
      "loss": 3.7366,
      "step": 12050
    },
    {
      "epoch": 1.30233559358519,
      "grad_norm": 0.5904163718223572,
      "learning_rate": 0.0005224415472470639,
      "loss": 3.7393,
      "step": 12100
    },
    {
      "epoch": 1.3077171456247982,
      "grad_norm": 0.5671617388725281,
      "learning_rate": 0.0005221183062170024,
      "loss": 3.7603,
      "step": 12150
    },
    {
      "epoch": 1.3130986976644063,
      "grad_norm": 0.5398463010787964,
      "learning_rate": 0.0005217950651869409,
      "loss": 3.7206,
      "step": 12200
    },
    {
      "epoch": 1.3184802497040147,
      "grad_norm": 0.5277562737464905,
      "learning_rate": 0.0005214718241568796,
      "loss": 3.7159,
      "step": 12250
    },
    {
      "epoch": 1.3238618017436228,
      "grad_norm": 0.606819748878479,
      "learning_rate": 0.0005211485831268182,
      "loss": 3.7343,
      "step": 12300
    },
    {
      "epoch": 1.329243353783231,
      "grad_norm": 0.5795959234237671,
      "learning_rate": 0.0005208253420967568,
      "loss": 3.7255,
      "step": 12350
    },
    {
      "epoch": 1.3346249058228392,
      "grad_norm": 0.5494657754898071,
      "learning_rate": 0.0005205021010666953,
      "loss": 3.7103,
      "step": 12400
    },
    {
      "epoch": 1.3400064578624475,
      "grad_norm": 0.5491659641265869,
      "learning_rate": 0.0005201788600366339,
      "loss": 3.7417,
      "step": 12450
    },
    {
      "epoch": 1.3453880099020559,
      "grad_norm": 0.6019899249076843,
      "learning_rate": 0.0005198556190065725,
      "loss": 3.712,
      "step": 12500
    },
    {
      "epoch": 1.350769561941664,
      "grad_norm": 0.566986083984375,
      "learning_rate": 0.0005195323779765112,
      "loss": 3.7145,
      "step": 12550
    },
    {
      "epoch": 1.356151113981272,
      "grad_norm": 0.5378516316413879,
      "learning_rate": 0.0005192091369464497,
      "loss": 3.7042,
      "step": 12600
    },
    {
      "epoch": 1.3615326660208804,
      "grad_norm": 0.5663155317306519,
      "learning_rate": 0.0005188858959163882,
      "loss": 3.7142,
      "step": 12650
    },
    {
      "epoch": 1.3669142180604887,
      "grad_norm": 0.6072174906730652,
      "learning_rate": 0.000518569119706928,
      "loss": 3.72,
      "step": 12700
    },
    {
      "epoch": 1.3722957701000968,
      "grad_norm": 0.6151204109191895,
      "learning_rate": 0.0005182458786768667,
      "loss": 3.7181,
      "step": 12750
    },
    {
      "epoch": 1.3776773221397052,
      "grad_norm": 0.6041751503944397,
      "learning_rate": 0.0005179226376468052,
      "loss": 3.7157,
      "step": 12800
    },
    {
      "epoch": 1.3830588741793133,
      "grad_norm": 0.5433161854743958,
      "learning_rate": 0.0005175993966167438,
      "loss": 3.7149,
      "step": 12850
    },
    {
      "epoch": 1.3884404262189216,
      "grad_norm": 0.5453323125839233,
      "learning_rate": 0.0005172761555866824,
      "loss": 3.7329,
      "step": 12900
    },
    {
      "epoch": 1.3938219782585297,
      "grad_norm": 0.5270190238952637,
      "learning_rate": 0.0005169529145566211,
      "loss": 3.7299,
      "step": 12950
    },
    {
      "epoch": 1.399203530298138,
      "grad_norm": 0.6112902164459229,
      "learning_rate": 0.0005166296735265596,
      "loss": 3.7047,
      "step": 13000
    },
    {
      "epoch": 1.399203530298138,
      "eval_accuracy": 0.3511135454186999,
      "eval_loss": 3.6792566776275635,
      "eval_runtime": 184.3496,
      "eval_samples_per_second": 97.7,
      "eval_steps_per_second": 6.108,
      "step": 13000
    },
    {
      "epoch": 1.4045850823377461,
      "grad_norm": 0.6328718066215515,
      "learning_rate": 0.0005163064324964982,
      "loss": 3.7154,
      "step": 13050
    },
    {
      "epoch": 1.4099666343773545,
      "grad_norm": 0.5578170418739319,
      "learning_rate": 0.0005159831914664367,
      "loss": 3.7098,
      "step": 13100
    },
    {
      "epoch": 1.4153481864169626,
      "grad_norm": 0.5283219814300537,
      "learning_rate": 0.0005156599504363753,
      "loss": 3.7155,
      "step": 13150
    },
    {
      "epoch": 1.420729738456571,
      "grad_norm": 0.5568632483482361,
      "learning_rate": 0.000515336709406314,
      "loss": 3.7229,
      "step": 13200
    },
    {
      "epoch": 1.426111290496179,
      "grad_norm": 0.6089996695518494,
      "learning_rate": 0.0005150134683762525,
      "loss": 3.699,
      "step": 13250
    },
    {
      "epoch": 1.4314928425357873,
      "grad_norm": 0.5421528816223145,
      "learning_rate": 0.0005146902273461911,
      "loss": 3.7186,
      "step": 13300
    },
    {
      "epoch": 1.4368743945753955,
      "grad_norm": 0.5795373320579529,
      "learning_rate": 0.0005143669863161297,
      "loss": 3.6983,
      "step": 13350
    },
    {
      "epoch": 1.4422559466150038,
      "grad_norm": 0.5475091934204102,
      "learning_rate": 0.0005140437452860683,
      "loss": 3.6928,
      "step": 13400
    },
    {
      "epoch": 1.447637498654612,
      "grad_norm": 0.5778725743293762,
      "learning_rate": 0.0005137205042560069,
      "loss": 3.7071,
      "step": 13450
    },
    {
      "epoch": 1.4530190506942202,
      "grad_norm": 0.625730037689209,
      "learning_rate": 0.0005133972632259455,
      "loss": 3.704,
      "step": 13500
    },
    {
      "epoch": 1.4584006027338283,
      "grad_norm": 0.5876809358596802,
      "learning_rate": 0.000513074022195884,
      "loss": 3.6853,
      "step": 13550
    },
    {
      "epoch": 1.4637821547734367,
      "grad_norm": 0.6416075825691223,
      "learning_rate": 0.0005127507811658226,
      "loss": 3.7129,
      "step": 13600
    },
    {
      "epoch": 1.469163706813045,
      "grad_norm": 0.5492163896560669,
      "learning_rate": 0.0005124275401357612,
      "loss": 3.6995,
      "step": 13650
    },
    {
      "epoch": 1.474545258852653,
      "grad_norm": 0.5315779447555542,
      "learning_rate": 0.0005121042991056997,
      "loss": 3.7089,
      "step": 13700
    },
    {
      "epoch": 1.4799268108922612,
      "grad_norm": 0.5582946538925171,
      "learning_rate": 0.0005117810580756384,
      "loss": 3.7127,
      "step": 13750
    },
    {
      "epoch": 1.4853083629318695,
      "grad_norm": 0.5699086785316467,
      "learning_rate": 0.0005114578170455769,
      "loss": 3.6964,
      "step": 13800
    },
    {
      "epoch": 1.4906899149714778,
      "grad_norm": 0.594375729560852,
      "learning_rate": 0.0005111345760155156,
      "loss": 3.6946,
      "step": 13850
    },
    {
      "epoch": 1.496071467011086,
      "grad_norm": 0.8855546116828918,
      "learning_rate": 0.0005108113349854541,
      "loss": 3.6961,
      "step": 13900
    },
    {
      "epoch": 1.501453019050694,
      "grad_norm": 0.5943670868873596,
      "learning_rate": 0.0005104880939553926,
      "loss": 3.693,
      "step": 13950
    },
    {
      "epoch": 1.5068345710903024,
      "grad_norm": 0.5787729620933533,
      "learning_rate": 0.0005101648529253313,
      "loss": 3.7026,
      "step": 14000
    },
    {
      "epoch": 1.5068345710903024,
      "eval_accuracy": 0.35351607807405366,
      "eval_loss": 3.657789468765259,
      "eval_runtime": 184.1774,
      "eval_samples_per_second": 97.792,
      "eval_steps_per_second": 6.114,
      "step": 14000
    },
    {
      "epoch": 1.5122161231299107,
      "grad_norm": 0.5396068692207336,
      "learning_rate": 0.0005098416118952699,
      "loss": 3.6908,
      "step": 14050
    },
    {
      "epoch": 1.5175976751695188,
      "grad_norm": 0.6208797693252563,
      "learning_rate": 0.0005095183708652085,
      "loss": 3.7153,
      "step": 14100
    },
    {
      "epoch": 1.5229792272091272,
      "grad_norm": 0.5639035105705261,
      "learning_rate": 0.000509195129835147,
      "loss": 3.696,
      "step": 14150
    },
    {
      "epoch": 1.5283607792487355,
      "grad_norm": 0.5553680062294006,
      "learning_rate": 0.0005088718888050856,
      "loss": 3.6996,
      "step": 14200
    },
    {
      "epoch": 1.5337423312883436,
      "grad_norm": 0.5596705079078674,
      "learning_rate": 0.0005085486477750242,
      "loss": 3.6817,
      "step": 14250
    },
    {
      "epoch": 1.5391238833279517,
      "grad_norm": 0.5594382286071777,
      "learning_rate": 0.0005082254067449629,
      "loss": 3.6886,
      "step": 14300
    },
    {
      "epoch": 1.54450543536756,
      "grad_norm": 0.5481204986572266,
      "learning_rate": 0.0005079021657149014,
      "loss": 3.6832,
      "step": 14350
    },
    {
      "epoch": 1.5498869874071683,
      "grad_norm": 0.5734264254570007,
      "learning_rate": 0.0005075789246848399,
      "loss": 3.6921,
      "step": 14400
    },
    {
      "epoch": 1.5552685394467765,
      "grad_norm": 0.5279305577278137,
      "learning_rate": 0.0005072556836547785,
      "loss": 3.6929,
      "step": 14450
    },
    {
      "epoch": 1.5606500914863846,
      "grad_norm": 0.6012471914291382,
      "learning_rate": 0.0005069324426247171,
      "loss": 3.6822,
      "step": 14500
    },
    {
      "epoch": 1.566031643525993,
      "grad_norm": 0.588430643081665,
      "learning_rate": 0.0005066092015946557,
      "loss": 3.6901,
      "step": 14550
    },
    {
      "epoch": 1.5714131955656012,
      "grad_norm": 0.6223108172416687,
      "learning_rate": 0.0005062859605645943,
      "loss": 3.7083,
      "step": 14600
    },
    {
      "epoch": 1.5767947476052093,
      "grad_norm": 0.6168702244758606,
      "learning_rate": 0.0005059627195345329,
      "loss": 3.7105,
      "step": 14650
    },
    {
      "epoch": 1.5821762996448174,
      "grad_norm": 0.6164829730987549,
      "learning_rate": 0.0005056394785044715,
      "loss": 3.7006,
      "step": 14700
    },
    {
      "epoch": 1.5875578516844258,
      "grad_norm": 0.5340235829353333,
      "learning_rate": 0.00050531623747441,
      "loss": 3.6942,
      "step": 14750
    },
    {
      "epoch": 1.592939403724034,
      "grad_norm": 0.5426746606826782,
      "learning_rate": 0.0005049929964443486,
      "loss": 3.6889,
      "step": 14800
    },
    {
      "epoch": 1.5983209557636422,
      "grad_norm": 0.5669272541999817,
      "learning_rate": 0.0005046762202348884,
      "loss": 3.6671,
      "step": 14850
    },
    {
      "epoch": 1.6037025078032503,
      "grad_norm": 0.5241160988807678,
      "learning_rate": 0.000504352979204827,
      "loss": 3.6919,
      "step": 14900
    },
    {
      "epoch": 1.6090840598428586,
      "grad_norm": 0.5965842604637146,
      "learning_rate": 0.0005040297381747656,
      "loss": 3.697,
      "step": 14950
    },
    {
      "epoch": 1.614465611882467,
      "grad_norm": 0.5421640276908875,
      "learning_rate": 0.0005037064971447042,
      "loss": 3.6861,
      "step": 15000
    },
    {
      "epoch": 1.614465611882467,
      "eval_accuracy": 0.3554244573956272,
      "eval_loss": 3.6394846439361572,
      "eval_runtime": 184.2298,
      "eval_samples_per_second": 97.764,
      "eval_steps_per_second": 6.112,
      "step": 15000
    },
    {
      "epoch": 1.619847163922075,
      "grad_norm": 0.5675601363182068,
      "learning_rate": 0.0005033832561146428,
      "loss": 3.6957,
      "step": 15050
    },
    {
      "epoch": 1.6252287159616834,
      "grad_norm": 0.5630671381950378,
      "learning_rate": 0.0005030600150845813,
      "loss": 3.6725,
      "step": 15100
    },
    {
      "epoch": 1.6306102680012917,
      "grad_norm": 0.5368252396583557,
      "learning_rate": 0.00050273677405452,
      "loss": 3.6624,
      "step": 15150
    },
    {
      "epoch": 1.6359918200408998,
      "grad_norm": 0.557021975517273,
      "learning_rate": 0.0005024135330244585,
      "loss": 3.6653,
      "step": 15200
    },
    {
      "epoch": 1.641373372080508,
      "grad_norm": 0.5746883749961853,
      "learning_rate": 0.0005020902919943972,
      "loss": 3.6811,
      "step": 15250
    },
    {
      "epoch": 1.6467549241201163,
      "grad_norm": 0.539557695388794,
      "learning_rate": 0.0005017670509643357,
      "loss": 3.6759,
      "step": 15300
    },
    {
      "epoch": 1.6521364761597246,
      "grad_norm": 0.5449525117874146,
      "learning_rate": 0.0005014438099342743,
      "loss": 3.6887,
      "step": 15350
    },
    {
      "epoch": 1.6575180281993327,
      "grad_norm": 0.6430085897445679,
      "learning_rate": 0.0005011205689042129,
      "loss": 3.6815,
      "step": 15400
    },
    {
      "epoch": 1.6628995802389408,
      "grad_norm": 0.5697539448738098,
      "learning_rate": 0.0005007973278741514,
      "loss": 3.6711,
      "step": 15450
    },
    {
      "epoch": 1.6682811322785491,
      "grad_norm": 0.5965550541877747,
      "learning_rate": 0.00050047408684409,
      "loss": 3.6802,
      "step": 15500
    },
    {
      "epoch": 1.6736626843181575,
      "grad_norm": 0.5688132643699646,
      "learning_rate": 0.0005001508458140286,
      "loss": 3.6706,
      "step": 15550
    },
    {
      "epoch": 1.6790442363577656,
      "grad_norm": 0.565057098865509,
      "learning_rate": 0.0004998276047839673,
      "loss": 3.6932,
      "step": 15600
    },
    {
      "epoch": 1.6844257883973737,
      "grad_norm": 0.599248468875885,
      "learning_rate": 0.0004995043637539058,
      "loss": 3.6685,
      "step": 15650
    },
    {
      "epoch": 1.689807340436982,
      "grad_norm": 0.569948673248291,
      "learning_rate": 0.0004991811227238443,
      "loss": 3.6804,
      "step": 15700
    },
    {
      "epoch": 1.6951888924765903,
      "grad_norm": 0.6075010299682617,
      "learning_rate": 0.0004988643465143841,
      "loss": 3.6745,
      "step": 15750
    },
    {
      "epoch": 1.7005704445161984,
      "grad_norm": 0.5561265349388123,
      "learning_rate": 0.0004985411054843228,
      "loss": 3.6526,
      "step": 15800
    },
    {
      "epoch": 1.7059519965558065,
      "grad_norm": 0.5673480033874512,
      "learning_rate": 0.0004982178644542613,
      "loss": 3.6694,
      "step": 15850
    },
    {
      "epoch": 1.7113335485954149,
      "grad_norm": 0.6317378282546997,
      "learning_rate": 0.0004978946234242,
      "loss": 3.6777,
      "step": 15900
    },
    {
      "epoch": 1.7167151006350232,
      "grad_norm": 0.5389451384544373,
      "learning_rate": 0.0004975713823941385,
      "loss": 3.6692,
      "step": 15950
    },
    {
      "epoch": 1.7220966526746313,
      "grad_norm": 0.5190277695655823,
      "learning_rate": 0.0004972481413640771,
      "loss": 3.6623,
      "step": 16000
    },
    {
      "epoch": 1.7220966526746313,
      "eval_accuracy": 0.35677164455163896,
      "eval_loss": 3.621499538421631,
      "eval_runtime": 184.0771,
      "eval_samples_per_second": 97.845,
      "eval_steps_per_second": 6.117,
      "step": 16000
    },
    {
      "epoch": 1.7274782047142396,
      "grad_norm": 0.5854401588439941,
      "learning_rate": 0.0004969249003340157,
      "loss": 3.6558,
      "step": 16050
    },
    {
      "epoch": 1.732859756753848,
      "grad_norm": 0.5637298226356506,
      "learning_rate": 0.0004966016593039543,
      "loss": 3.6604,
      "step": 16100
    },
    {
      "epoch": 1.738241308793456,
      "grad_norm": 0.534237265586853,
      "learning_rate": 0.0004962784182738928,
      "loss": 3.6512,
      "step": 16150
    },
    {
      "epoch": 1.7436228608330642,
      "grad_norm": 0.5651060342788696,
      "learning_rate": 0.0004959551772438314,
      "loss": 3.671,
      "step": 16200
    },
    {
      "epoch": 1.7490044128726725,
      "grad_norm": 0.5263696312904358,
      "learning_rate": 0.0004956319362137701,
      "loss": 3.6597,
      "step": 16250
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 0.5691628456115723,
      "learning_rate": 0.0004953086951837086,
      "loss": 3.6613,
      "step": 16300
    },
    {
      "epoch": 1.759767516951889,
      "grad_norm": 0.5888131856918335,
      "learning_rate": 0.0004949854541536472,
      "loss": 3.6744,
      "step": 16350
    },
    {
      "epoch": 1.765149068991497,
      "grad_norm": 0.5408467650413513,
      "learning_rate": 0.0004946622131235857,
      "loss": 3.6785,
      "step": 16400
    },
    {
      "epoch": 1.7705306210311054,
      "grad_norm": 0.5659624934196472,
      "learning_rate": 0.0004943389720935244,
      "loss": 3.6521,
      "step": 16450
    },
    {
      "epoch": 1.7759121730707137,
      "grad_norm": 0.5278242826461792,
      "learning_rate": 0.000494015731063463,
      "loss": 3.6684,
      "step": 16500
    },
    {
      "epoch": 1.7812937251103218,
      "grad_norm": 0.5713918805122375,
      "learning_rate": 0.0004936924900334016,
      "loss": 3.6553,
      "step": 16550
    },
    {
      "epoch": 1.78667527714993,
      "grad_norm": 0.6127734184265137,
      "learning_rate": 0.0004933692490033401,
      "loss": 3.6751,
      "step": 16600
    },
    {
      "epoch": 1.7920568291895382,
      "grad_norm": 0.5428546071052551,
      "learning_rate": 0.0004930460079732786,
      "loss": 3.6366,
      "step": 16650
    },
    {
      "epoch": 1.7974383812291466,
      "grad_norm": 0.5874344110488892,
      "learning_rate": 0.0004927227669432173,
      "loss": 3.6531,
      "step": 16700
    },
    {
      "epoch": 1.8028199332687547,
      "grad_norm": 0.5529983639717102,
      "learning_rate": 0.0004923995259131558,
      "loss": 3.6795,
      "step": 16750
    },
    {
      "epoch": 1.8082014853083628,
      "grad_norm": 0.5577397346496582,
      "learning_rate": 0.0004920762848830945,
      "loss": 3.6571,
      "step": 16800
    },
    {
      "epoch": 1.813583037347971,
      "grad_norm": 0.6135964393615723,
      "learning_rate": 0.000491753043853033,
      "loss": 3.6622,
      "step": 16850
    },
    {
      "epoch": 1.8189645893875794,
      "grad_norm": 0.5189880728721619,
      "learning_rate": 0.0004914298028229717,
      "loss": 3.6603,
      "step": 16900
    },
    {
      "epoch": 1.8243461414271875,
      "grad_norm": 0.6014214158058167,
      "learning_rate": 0.0004911065617929102,
      "loss": 3.6549,
      "step": 16950
    },
    {
      "epoch": 1.8297276934667959,
      "grad_norm": 0.5619677901268005,
      "learning_rate": 0.0004907833207628487,
      "loss": 3.647,
      "step": 17000
    },
    {
      "epoch": 1.8297276934667959,
      "eval_accuracy": 0.35838405340680585,
      "eval_loss": 3.6053762435913086,
      "eval_runtime": 184.2908,
      "eval_samples_per_second": 97.731,
      "eval_steps_per_second": 6.11,
      "step": 17000
    },
    {
      "epoch": 1.8351092455064042,
      "grad_norm": 0.5847708582878113,
      "learning_rate": 0.0004904600797327874,
      "loss": 3.6395,
      "step": 17050
    },
    {
      "epoch": 1.8404907975460123,
      "grad_norm": 0.5379604697227478,
      "learning_rate": 0.0004901368387027259,
      "loss": 3.6662,
      "step": 17100
    },
    {
      "epoch": 1.8458723495856204,
      "grad_norm": 0.5651165843009949,
      "learning_rate": 0.0004898135976726646,
      "loss": 3.6599,
      "step": 17150
    },
    {
      "epoch": 1.8512539016252287,
      "grad_norm": 0.6006888151168823,
      "learning_rate": 0.0004894903566426031,
      "loss": 3.6618,
      "step": 17200
    },
    {
      "epoch": 1.856635453664837,
      "grad_norm": 0.5520943999290466,
      "learning_rate": 0.0004891671156125417,
      "loss": 3.6595,
      "step": 17250
    },
    {
      "epoch": 1.8620170057044452,
      "grad_norm": 0.541239857673645,
      "learning_rate": 0.0004888438745824803,
      "loss": 3.65,
      "step": 17300
    },
    {
      "epoch": 1.8673985577440533,
      "grad_norm": 0.5688744783401489,
      "learning_rate": 0.0004885206335524189,
      "loss": 3.6572,
      "step": 17350
    },
    {
      "epoch": 1.8727801097836616,
      "grad_norm": 0.5306487679481506,
      "learning_rate": 0.0004881973925223575,
      "loss": 3.6613,
      "step": 17400
    },
    {
      "epoch": 1.87816166182327,
      "grad_norm": 0.5473430752754211,
      "learning_rate": 0.00048787415149229604,
      "loss": 3.6403,
      "step": 17450
    },
    {
      "epoch": 1.883543213862878,
      "grad_norm": 0.5673529505729675,
      "learning_rate": 0.00048755091046223464,
      "loss": 3.6446,
      "step": 17500
    },
    {
      "epoch": 1.8889247659024861,
      "grad_norm": 0.5673733353614807,
      "learning_rate": 0.0004872276694321732,
      "loss": 3.6383,
      "step": 17550
    },
    {
      "epoch": 1.8943063179420945,
      "grad_norm": 0.5422235727310181,
      "learning_rate": 0.00048690442840211177,
      "loss": 3.6288,
      "step": 17600
    },
    {
      "epoch": 1.8996878699817028,
      "grad_norm": 0.6383090615272522,
      "learning_rate": 0.0004865811873720504,
      "loss": 3.6652,
      "step": 17650
    },
    {
      "epoch": 1.905069422021311,
      "grad_norm": 0.5502477288246155,
      "learning_rate": 0.00048625794634198896,
      "loss": 3.6513,
      "step": 17700
    },
    {
      "epoch": 1.910450974060919,
      "grad_norm": 0.5370838642120361,
      "learning_rate": 0.00048593470531192756,
      "loss": 3.6271,
      "step": 17750
    },
    {
      "epoch": 1.9158325261005273,
      "grad_norm": 0.5876355171203613,
      "learning_rate": 0.00048561146428186615,
      "loss": 3.6555,
      "step": 17800
    },
    {
      "epoch": 1.9212140781401357,
      "grad_norm": 0.5959500670433044,
      "learning_rate": 0.0004852882232518047,
      "loss": 3.6502,
      "step": 17850
    },
    {
      "epoch": 1.9265956301797438,
      "grad_norm": 0.5405507683753967,
      "learning_rate": 0.00048496498222174334,
      "loss": 3.6315,
      "step": 17900
    },
    {
      "epoch": 1.931977182219352,
      "grad_norm": 0.5238087177276611,
      "learning_rate": 0.00048464174119168193,
      "loss": 3.6462,
      "step": 17950
    },
    {
      "epoch": 1.9373587342589604,
      "grad_norm": 0.5805983543395996,
      "learning_rate": 0.0004843185001616205,
      "loss": 3.6213,
      "step": 18000
    },
    {
      "epoch": 1.9373587342589604,
      "eval_accuracy": 0.36002916678120495,
      "eval_loss": 3.593156099319458,
      "eval_runtime": 183.8245,
      "eval_samples_per_second": 97.979,
      "eval_steps_per_second": 6.125,
      "step": 18000
    },
    {
      "epoch": 1.9427402862985685,
      "grad_norm": 0.6219788193702698,
      "learning_rate": 0.00048399525913155907,
      "loss": 3.6193,
      "step": 18050
    },
    {
      "epoch": 1.9481218383381766,
      "grad_norm": 0.5761560797691345,
      "learning_rate": 0.0004836720181014976,
      "loss": 3.6411,
      "step": 18100
    },
    {
      "epoch": 1.953503390377785,
      "grad_norm": 0.5609335899353027,
      "learning_rate": 0.0004833487770714362,
      "loss": 3.6392,
      "step": 18150
    },
    {
      "epoch": 1.9588849424173933,
      "grad_norm": 0.5326209664344788,
      "learning_rate": 0.00048302553604137485,
      "loss": 3.6329,
      "step": 18200
    },
    {
      "epoch": 1.9642664944570014,
      "grad_norm": 0.580141007900238,
      "learning_rate": 0.0004827022950113134,
      "loss": 3.6419,
      "step": 18250
    },
    {
      "epoch": 1.9696480464966095,
      "grad_norm": 0.5702725052833557,
      "learning_rate": 0.000482379053981252,
      "loss": 3.6203,
      "step": 18300
    },
    {
      "epoch": 1.9750295985362178,
      "grad_norm": 0.5786581635475159,
      "learning_rate": 0.0004820558129511906,
      "loss": 3.6553,
      "step": 18350
    },
    {
      "epoch": 1.9804111505758262,
      "grad_norm": 0.5711109042167664,
      "learning_rate": 0.0004817325719211291,
      "loss": 3.6303,
      "step": 18400
    },
    {
      "epoch": 1.9857927026154343,
      "grad_norm": 0.5867487192153931,
      "learning_rate": 0.0004814093308910677,
      "loss": 3.6539,
      "step": 18450
    },
    {
      "epoch": 1.9911742546550424,
      "grad_norm": 0.5746527314186096,
      "learning_rate": 0.00048108608986100637,
      "loss": 3.617,
      "step": 18500
    },
    {
      "epoch": 1.9965558066946507,
      "grad_norm": 0.6316065192222595,
      "learning_rate": 0.0004807628488309449,
      "loss": 3.6253,
      "step": 18550
    },
    {
      "epoch": 2.001937358734259,
      "grad_norm": 0.6170855164527893,
      "learning_rate": 0.0004804396078008835,
      "loss": 3.5729,
      "step": 18600
    },
    {
      "epoch": 2.007318910773867,
      "grad_norm": 0.5935534834861755,
      "learning_rate": 0.00048011636677082204,
      "loss": 3.5357,
      "step": 18650
    },
    {
      "epoch": 2.0127004628134753,
      "grad_norm": 0.6016736626625061,
      "learning_rate": 0.00047979312574076064,
      "loss": 3.5417,
      "step": 18700
    },
    {
      "epoch": 2.018082014853084,
      "grad_norm": 0.5650578737258911,
      "learning_rate": 0.0004794698847106992,
      "loss": 3.5495,
      "step": 18750
    },
    {
      "epoch": 2.023463566892692,
      "grad_norm": 0.5854896903038025,
      "learning_rate": 0.0004791466436806378,
      "loss": 3.5542,
      "step": 18800
    },
    {
      "epoch": 2.0288451189323,
      "grad_norm": 0.5338824391365051,
      "learning_rate": 0.0004788234026505764,
      "loss": 3.5399,
      "step": 18850
    },
    {
      "epoch": 2.034226670971908,
      "grad_norm": 0.5451910495758057,
      "learning_rate": 0.00047850016162051496,
      "loss": 3.5429,
      "step": 18900
    },
    {
      "epoch": 2.0396082230115167,
      "grad_norm": 0.6403917074203491,
      "learning_rate": 0.00047817692059045356,
      "loss": 3.5485,
      "step": 18950
    },
    {
      "epoch": 2.044989775051125,
      "grad_norm": 0.6236400008201599,
      "learning_rate": 0.00047785367956039215,
      "loss": 3.5689,
      "step": 19000
    },
    {
      "epoch": 2.044989775051125,
      "eval_accuracy": 0.3610947256520015,
      "eval_loss": 3.584120273590088,
      "eval_runtime": 184.4198,
      "eval_samples_per_second": 97.663,
      "eval_steps_per_second": 6.106,
      "step": 19000
    },
    {
      "epoch": 2.050371327090733,
      "grad_norm": 0.6183650493621826,
      "learning_rate": 0.00047753043853033075,
      "loss": 3.5453,
      "step": 19050
    },
    {
      "epoch": 2.055752879130341,
      "grad_norm": 0.5917062163352966,
      "learning_rate": 0.00047720719750026934,
      "loss": 3.5494,
      "step": 19100
    },
    {
      "epoch": 2.0611344311699495,
      "grad_norm": 0.5498500466346741,
      "learning_rate": 0.00047688395647020793,
      "loss": 3.571,
      "step": 19150
    },
    {
      "epoch": 2.0665159832095576,
      "grad_norm": 0.531639039516449,
      "learning_rate": 0.0004765607154401465,
      "loss": 3.5618,
      "step": 19200
    },
    {
      "epoch": 2.0718975352491658,
      "grad_norm": 0.5593074560165405,
      "learning_rate": 0.00047623747441008507,
      "loss": 3.5534,
      "step": 19250
    },
    {
      "epoch": 2.0772790872887743,
      "grad_norm": 0.5986584424972534,
      "learning_rate": 0.0004759142333800236,
      "loss": 3.5434,
      "step": 19300
    },
    {
      "epoch": 2.0826606393283824,
      "grad_norm": 0.5871057510375977,
      "learning_rate": 0.00047559099234996226,
      "loss": 3.5727,
      "step": 19350
    },
    {
      "epoch": 2.0880421913679905,
      "grad_norm": 0.5127398371696472,
      "learning_rate": 0.00047526775131990085,
      "loss": 3.5729,
      "step": 19400
    },
    {
      "epoch": 2.0934237434075986,
      "grad_norm": 0.5388950705528259,
      "learning_rate": 0.0004749445102898394,
      "loss": 3.5437,
      "step": 19450
    },
    {
      "epoch": 2.098805295447207,
      "grad_norm": 0.5229042768478394,
      "learning_rate": 0.000474621269259778,
      "loss": 3.5625,
      "step": 19500
    },
    {
      "epoch": 2.1041868474868153,
      "grad_norm": 0.5500669479370117,
      "learning_rate": 0.0004742980282297166,
      "loss": 3.5555,
      "step": 19550
    },
    {
      "epoch": 2.1095683995264234,
      "grad_norm": 0.5531191229820251,
      "learning_rate": 0.0004739747871996551,
      "loss": 3.5753,
      "step": 19600
    },
    {
      "epoch": 2.1149499515660315,
      "grad_norm": 0.5585891008377075,
      "learning_rate": 0.00047365154616959377,
      "loss": 3.5857,
      "step": 19650
    },
    {
      "epoch": 2.12033150360564,
      "grad_norm": 0.5724694728851318,
      "learning_rate": 0.00047332830513953237,
      "loss": 3.5525,
      "step": 19700
    },
    {
      "epoch": 2.125713055645248,
      "grad_norm": 0.5514476299285889,
      "learning_rate": 0.0004730050641094709,
      "loss": 3.5673,
      "step": 19750
    },
    {
      "epoch": 2.1310946076848563,
      "grad_norm": 0.5218293070793152,
      "learning_rate": 0.00047268828790001077,
      "loss": 3.5643,
      "step": 19800
    },
    {
      "epoch": 2.1364761597244644,
      "grad_norm": 0.5710498690605164,
      "learning_rate": 0.0004723650468699493,
      "loss": 3.5711,
      "step": 19850
    },
    {
      "epoch": 2.141857711764073,
      "grad_norm": 0.5479194521903992,
      "learning_rate": 0.0004720418058398879,
      "loss": 3.5684,
      "step": 19900
    },
    {
      "epoch": 2.147239263803681,
      "grad_norm": 0.5603418350219727,
      "learning_rate": 0.00047171856480982644,
      "loss": 3.5563,
      "step": 19950
    },
    {
      "epoch": 2.152620815843289,
      "grad_norm": 0.6014620661735535,
      "learning_rate": 0.0004713953237797651,
      "loss": 3.5658,
      "step": 20000
    },
    {
      "epoch": 2.152620815843289,
      "eval_accuracy": 0.36233836660592605,
      "eval_loss": 3.5734927654266357,
      "eval_runtime": 183.8492,
      "eval_samples_per_second": 97.966,
      "eval_steps_per_second": 6.125,
      "step": 20000
    },
    {
      "epoch": 2.1580023678828972,
      "grad_norm": 0.5543132424354553,
      "learning_rate": 0.0004710720827497037,
      "loss": 3.5668,
      "step": 20050
    },
    {
      "epoch": 2.163383919922506,
      "grad_norm": 0.5795028209686279,
      "learning_rate": 0.00047074884171964223,
      "loss": 3.5501,
      "step": 20100
    },
    {
      "epoch": 2.168765471962114,
      "grad_norm": 0.649953305721283,
      "learning_rate": 0.0004704256006895808,
      "loss": 3.5761,
      "step": 20150
    },
    {
      "epoch": 2.174147024001722,
      "grad_norm": 0.569844126701355,
      "learning_rate": 0.00047010235965951936,
      "loss": 3.5548,
      "step": 20200
    },
    {
      "epoch": 2.1795285760413305,
      "grad_norm": 0.5314879417419434,
      "learning_rate": 0.00046977911862945796,
      "loss": 3.5583,
      "step": 20250
    },
    {
      "epoch": 2.1849101280809387,
      "grad_norm": 0.591710090637207,
      "learning_rate": 0.0004694558775993966,
      "loss": 3.5528,
      "step": 20300
    },
    {
      "epoch": 2.1902916801205468,
      "grad_norm": 0.5657521486282349,
      "learning_rate": 0.00046913263656933515,
      "loss": 3.5554,
      "step": 20350
    },
    {
      "epoch": 2.195673232160155,
      "grad_norm": 0.5465287566184998,
      "learning_rate": 0.00046880939553927374,
      "loss": 3.5574,
      "step": 20400
    },
    {
      "epoch": 2.2010547841997634,
      "grad_norm": 0.5874313712120056,
      "learning_rate": 0.00046848615450921234,
      "loss": 3.5713,
      "step": 20450
    },
    {
      "epoch": 2.2064363362393715,
      "grad_norm": 0.5441039800643921,
      "learning_rate": 0.0004681629134791509,
      "loss": 3.558,
      "step": 20500
    },
    {
      "epoch": 2.2118178882789796,
      "grad_norm": 0.5708215236663818,
      "learning_rate": 0.00046783967244908947,
      "loss": 3.5712,
      "step": 20550
    },
    {
      "epoch": 2.2171994403185877,
      "grad_norm": 0.6287713646888733,
      "learning_rate": 0.0004675164314190281,
      "loss": 3.5539,
      "step": 20600
    },
    {
      "epoch": 2.2225809923581963,
      "grad_norm": 0.5826866626739502,
      "learning_rate": 0.00046719319038896666,
      "loss": 3.5549,
      "step": 20650
    },
    {
      "epoch": 2.2279625443978044,
      "grad_norm": 0.6205884218215942,
      "learning_rate": 0.00046686994935890526,
      "loss": 3.5813,
      "step": 20700
    },
    {
      "epoch": 2.2333440964374125,
      "grad_norm": 0.6254109740257263,
      "learning_rate": 0.0004665467083288438,
      "loss": 3.5634,
      "step": 20750
    },
    {
      "epoch": 2.2387256484770206,
      "grad_norm": 0.5872119665145874,
      "learning_rate": 0.0004662234672987824,
      "loss": 3.5685,
      "step": 20800
    },
    {
      "epoch": 2.244107200516629,
      "grad_norm": 0.5845667123794556,
      "learning_rate": 0.00046590022626872104,
      "loss": 3.551,
      "step": 20850
    },
    {
      "epoch": 2.2494887525562373,
      "grad_norm": 0.5559894442558289,
      "learning_rate": 0.0004655769852386596,
      "loss": 3.5796,
      "step": 20900
    },
    {
      "epoch": 2.2548703045958454,
      "grad_norm": 0.5726771354675293,
      "learning_rate": 0.0004652537442085982,
      "loss": 3.5614,
      "step": 20950
    },
    {
      "epoch": 2.2602518566354535,
      "grad_norm": 0.5262014269828796,
      "learning_rate": 0.00046493050317853677,
      "loss": 3.566,
      "step": 21000
    },
    {
      "epoch": 2.2602518566354535,
      "eval_accuracy": 0.3635937420717346,
      "eval_loss": 3.563410520553589,
      "eval_runtime": 184.3029,
      "eval_samples_per_second": 97.725,
      "eval_steps_per_second": 6.11,
      "step": 21000
    },
    {
      "epoch": 2.265633408675062,
      "grad_norm": 0.6070656180381775,
      "learning_rate": 0.0004646072621484753,
      "loss": 3.5438,
      "step": 21050
    },
    {
      "epoch": 2.27101496071467,
      "grad_norm": 0.6696323156356812,
      "learning_rate": 0.0004642840211184139,
      "loss": 3.5596,
      "step": 21100
    },
    {
      "epoch": 2.2763965127542782,
      "grad_norm": 0.5441017150878906,
      "learning_rate": 0.00046396078008835255,
      "loss": 3.5596,
      "step": 21150
    },
    {
      "epoch": 2.281778064793887,
      "grad_norm": 0.6279102563858032,
      "learning_rate": 0.0004636375390582911,
      "loss": 3.5585,
      "step": 21200
    },
    {
      "epoch": 2.287159616833495,
      "grad_norm": 0.6477214097976685,
      "learning_rate": 0.0004633142980282297,
      "loss": 3.5443,
      "step": 21250
    },
    {
      "epoch": 2.292541168873103,
      "grad_norm": 0.5807666182518005,
      "learning_rate": 0.00046299105699816823,
      "loss": 3.5554,
      "step": 21300
    },
    {
      "epoch": 2.297922720912711,
      "grad_norm": 0.5846539735794067,
      "learning_rate": 0.0004626678159681068,
      "loss": 3.552,
      "step": 21350
    },
    {
      "epoch": 2.303304272952319,
      "grad_norm": 0.593036413192749,
      "learning_rate": 0.0004623445749380454,
      "loss": 3.5649,
      "step": 21400
    },
    {
      "epoch": 2.3086858249919278,
      "grad_norm": 0.5329724550247192,
      "learning_rate": 0.000462021333907984,
      "loss": 3.5584,
      "step": 21450
    },
    {
      "epoch": 2.314067377031536,
      "grad_norm": 0.5547925233840942,
      "learning_rate": 0.0004616980928779226,
      "loss": 3.5653,
      "step": 21500
    },
    {
      "epoch": 2.319448929071144,
      "grad_norm": 0.631915271282196,
      "learning_rate": 0.0004613748518478612,
      "loss": 3.5641,
      "step": 21550
    },
    {
      "epoch": 2.3248304811107525,
      "grad_norm": 0.5473701357841492,
      "learning_rate": 0.00046105161081779974,
      "loss": 3.5647,
      "step": 21600
    },
    {
      "epoch": 2.3302120331503606,
      "grad_norm": 0.6023962497711182,
      "learning_rate": 0.00046072836978773834,
      "loss": 3.5581,
      "step": 21650
    },
    {
      "epoch": 2.3355935851899687,
      "grad_norm": 0.5688156485557556,
      "learning_rate": 0.000460405128757677,
      "loss": 3.567,
      "step": 21700
    },
    {
      "epoch": 2.340975137229577,
      "grad_norm": 0.6036368012428284,
      "learning_rate": 0.0004600818877276155,
      "loss": 3.579,
      "step": 21750
    },
    {
      "epoch": 2.3463566892691854,
      "grad_norm": 0.5894560813903809,
      "learning_rate": 0.0004597651115181554,
      "loss": 3.5676,
      "step": 21800
    },
    {
      "epoch": 2.3517382413087935,
      "grad_norm": 0.595342755317688,
      "learning_rate": 0.00045944187048809393,
      "loss": 3.5457,
      "step": 21850
    },
    {
      "epoch": 2.3571197933484016,
      "grad_norm": 0.5520660877227783,
      "learning_rate": 0.0004591186294580325,
      "loss": 3.5564,
      "step": 21900
    },
    {
      "epoch": 2.3625013453880097,
      "grad_norm": 0.5292737483978271,
      "learning_rate": 0.00045879538842797106,
      "loss": 3.5484,
      "step": 21950
    },
    {
      "epoch": 2.3678828974276183,
      "grad_norm": 0.5825929045677185,
      "learning_rate": 0.00045847214739790966,
      "loss": 3.5658,
      "step": 22000
    },
    {
      "epoch": 2.3678828974276183,
      "eval_accuracy": 0.3643646343107783,
      "eval_loss": 3.553661584854126,
      "eval_runtime": 184.0601,
      "eval_samples_per_second": 97.854,
      "eval_steps_per_second": 6.118,
      "step": 22000
    },
    {
      "epoch": 2.3732644494672264,
      "grad_norm": 0.5640090703964233,
      "learning_rate": 0.0004581489063678482,
      "loss": 3.5509,
      "step": 22050
    },
    {
      "epoch": 2.3786460015068345,
      "grad_norm": 0.6931118369102478,
      "learning_rate": 0.00045782566533778685,
      "loss": 3.5325,
      "step": 22100
    },
    {
      "epoch": 2.384027553546443,
      "grad_norm": 0.6165780425071716,
      "learning_rate": 0.00045750242430772544,
      "loss": 3.5778,
      "step": 22150
    },
    {
      "epoch": 2.389409105586051,
      "grad_norm": 0.6399766206741333,
      "learning_rate": 0.000457179183277664,
      "loss": 3.5703,
      "step": 22200
    },
    {
      "epoch": 2.3947906576256592,
      "grad_norm": 0.6212499141693115,
      "learning_rate": 0.0004568559422476026,
      "loss": 3.553,
      "step": 22250
    },
    {
      "epoch": 2.4001722096652673,
      "grad_norm": 0.5530919432640076,
      "learning_rate": 0.00045653270121754117,
      "loss": 3.5612,
      "step": 22300
    },
    {
      "epoch": 2.4055537617048754,
      "grad_norm": 0.5530394315719604,
      "learning_rate": 0.0004562094601874797,
      "loss": 3.578,
      "step": 22350
    },
    {
      "epoch": 2.410935313744484,
      "grad_norm": 0.5988221168518066,
      "learning_rate": 0.00045588621915741836,
      "loss": 3.5377,
      "step": 22400
    },
    {
      "epoch": 2.416316865784092,
      "grad_norm": 0.5622323155403137,
      "learning_rate": 0.00045556297812735696,
      "loss": 3.5745,
      "step": 22450
    },
    {
      "epoch": 2.4216984178237,
      "grad_norm": 0.5235289335250854,
      "learning_rate": 0.0004552397370972955,
      "loss": 3.5349,
      "step": 22500
    },
    {
      "epoch": 2.4270799698633088,
      "grad_norm": 0.5971170663833618,
      "learning_rate": 0.0004549164960672341,
      "loss": 3.5585,
      "step": 22550
    },
    {
      "epoch": 2.432461521902917,
      "grad_norm": 0.6937503814697266,
      "learning_rate": 0.00045459325503717263,
      "loss": 3.5479,
      "step": 22600
    },
    {
      "epoch": 2.437843073942525,
      "grad_norm": 0.5887149572372437,
      "learning_rate": 0.0004542700140071113,
      "loss": 3.5507,
      "step": 22650
    },
    {
      "epoch": 2.443224625982133,
      "grad_norm": 0.5551357269287109,
      "learning_rate": 0.0004539467729770499,
      "loss": 3.5413,
      "step": 22700
    },
    {
      "epoch": 2.4486061780217416,
      "grad_norm": 0.6114067435264587,
      "learning_rate": 0.0004536235319469884,
      "loss": 3.5426,
      "step": 22750
    },
    {
      "epoch": 2.4539877300613497,
      "grad_norm": 0.5874900221824646,
      "learning_rate": 0.000453300290916927,
      "loss": 3.5662,
      "step": 22800
    },
    {
      "epoch": 2.459369282100958,
      "grad_norm": 0.557999849319458,
      "learning_rate": 0.0004529770498868656,
      "loss": 3.5557,
      "step": 22850
    },
    {
      "epoch": 2.464750834140566,
      "grad_norm": 0.5926379561424255,
      "learning_rate": 0.00045265380885680414,
      "loss": 3.5634,
      "step": 22900
    },
    {
      "epoch": 2.4701323861801745,
      "grad_norm": 0.6192485094070435,
      "learning_rate": 0.0004523305678267428,
      "loss": 3.5682,
      "step": 22950
    },
    {
      "epoch": 2.4755139382197826,
      "grad_norm": 0.6196098327636719,
      "learning_rate": 0.0004520073267966814,
      "loss": 3.563,
      "step": 23000
    },
    {
      "epoch": 2.4755139382197826,
      "eval_accuracy": 0.3652500466935785,
      "eval_loss": 3.5426125526428223,
      "eval_runtime": 184.3906,
      "eval_samples_per_second": 97.679,
      "eval_steps_per_second": 6.107,
      "step": 23000
    },
    {
      "epoch": 2.4808954902593907,
      "grad_norm": 0.5979634523391724,
      "learning_rate": 0.00045168408576661993,
      "loss": 3.5572,
      "step": 23050
    },
    {
      "epoch": 2.4862770422989993,
      "grad_norm": 0.563565731048584,
      "learning_rate": 0.0004513608447365585,
      "loss": 3.5484,
      "step": 23100
    },
    {
      "epoch": 2.4916585943386074,
      "grad_norm": 0.6143245697021484,
      "learning_rate": 0.00045103760370649706,
      "loss": 3.5519,
      "step": 23150
    },
    {
      "epoch": 2.4970401463782155,
      "grad_norm": 0.6232026219367981,
      "learning_rate": 0.00045071436267643566,
      "loss": 3.5435,
      "step": 23200
    },
    {
      "epoch": 2.5024216984178236,
      "grad_norm": 0.55870121717453,
      "learning_rate": 0.0004503911216463743,
      "loss": 3.5512,
      "step": 23250
    },
    {
      "epoch": 2.5078032504574317,
      "grad_norm": 0.5873537659645081,
      "learning_rate": 0.00045006788061631285,
      "loss": 3.5495,
      "step": 23300
    },
    {
      "epoch": 2.5131848024970402,
      "grad_norm": 0.5867093205451965,
      "learning_rate": 0.00044974463958625144,
      "loss": 3.5709,
      "step": 23350
    },
    {
      "epoch": 2.5185663545366483,
      "grad_norm": 0.5661380887031555,
      "learning_rate": 0.00044942139855619004,
      "loss": 3.5423,
      "step": 23400
    },
    {
      "epoch": 2.5239479065762565,
      "grad_norm": 0.5543169379234314,
      "learning_rate": 0.0004490981575261286,
      "loss": 3.5509,
      "step": 23450
    },
    {
      "epoch": 2.529329458615865,
      "grad_norm": 0.5908783078193665,
      "learning_rate": 0.0004487749164960672,
      "loss": 3.5607,
      "step": 23500
    },
    {
      "epoch": 2.534711010655473,
      "grad_norm": 0.5555139183998108,
      "learning_rate": 0.0004484516754660058,
      "loss": 3.5498,
      "step": 23550
    },
    {
      "epoch": 2.540092562695081,
      "grad_norm": 0.6041598916053772,
      "learning_rate": 0.00044812843443594436,
      "loss": 3.5396,
      "step": 23600
    },
    {
      "epoch": 2.5454741147346893,
      "grad_norm": 0.5496644377708435,
      "learning_rate": 0.00044780519340588296,
      "loss": 3.5558,
      "step": 23650
    },
    {
      "epoch": 2.550855666774298,
      "grad_norm": 0.5645830035209656,
      "learning_rate": 0.0004474819523758215,
      "loss": 3.5478,
      "step": 23700
    },
    {
      "epoch": 2.556237218813906,
      "grad_norm": 0.5833712816238403,
      "learning_rate": 0.0004471587113457601,
      "loss": 3.5437,
      "step": 23750
    },
    {
      "epoch": 2.561618770853514,
      "grad_norm": 0.6317982077598572,
      "learning_rate": 0.0004468419351362999,
      "loss": 3.541,
      "step": 23800
    },
    {
      "epoch": 2.567000322893122,
      "grad_norm": 0.5809699296951294,
      "learning_rate": 0.0004465186941062385,
      "loss": 3.5511,
      "step": 23850
    },
    {
      "epoch": 2.5723818749327307,
      "grad_norm": 0.5693453550338745,
      "learning_rate": 0.00044619545307617714,
      "loss": 3.5458,
      "step": 23900
    },
    {
      "epoch": 2.577763426972339,
      "grad_norm": 0.6099852323532104,
      "learning_rate": 0.0004458722120461157,
      "loss": 3.5349,
      "step": 23950
    },
    {
      "epoch": 2.583144979011947,
      "grad_norm": 0.5787376165390015,
      "learning_rate": 0.0004455489710160543,
      "loss": 3.5665,
      "step": 24000
    },
    {
      "epoch": 2.583144979011947,
      "eval_accuracy": 0.36613469850616404,
      "eval_loss": 3.5340969562530518,
      "eval_runtime": 183.9225,
      "eval_samples_per_second": 97.927,
      "eval_steps_per_second": 6.122,
      "step": 24000
    },
    {
      "epoch": 2.5885265310515555,
      "grad_norm": 0.5951037406921387,
      "learning_rate": 0.0004452257299859928,
      "loss": 3.5477,
      "step": 24050
    },
    {
      "epoch": 2.5939080830911636,
      "grad_norm": 0.633108913898468,
      "learning_rate": 0.0004449024889559314,
      "loss": 3.5457,
      "step": 24100
    },
    {
      "epoch": 2.5992896351307717,
      "grad_norm": 0.6345134973526001,
      "learning_rate": 0.00044457924792587,
      "loss": 3.5497,
      "step": 24150
    },
    {
      "epoch": 2.60467118717038,
      "grad_norm": 0.6203924417495728,
      "learning_rate": 0.0004442560068958086,
      "loss": 3.5236,
      "step": 24200
    },
    {
      "epoch": 2.610052739209988,
      "grad_norm": 0.5503219962120056,
      "learning_rate": 0.0004439327658657472,
      "loss": 3.5468,
      "step": 24250
    },
    {
      "epoch": 2.6154342912495965,
      "grad_norm": 0.6177816987037659,
      "learning_rate": 0.0004436095248356858,
      "loss": 3.5436,
      "step": 24300
    },
    {
      "epoch": 2.6208158432892046,
      "grad_norm": 0.5722578763961792,
      "learning_rate": 0.00044328628380562433,
      "loss": 3.5258,
      "step": 24350
    },
    {
      "epoch": 2.6261973953288127,
      "grad_norm": 0.611875057220459,
      "learning_rate": 0.0004429630427755629,
      "loss": 3.5442,
      "step": 24400
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.6074414849281311,
      "learning_rate": 0.0004426398017455016,
      "loss": 3.5477,
      "step": 24450
    },
    {
      "epoch": 2.6369604994080293,
      "grad_norm": 0.571189284324646,
      "learning_rate": 0.0004423165607154401,
      "loss": 3.5361,
      "step": 24500
    },
    {
      "epoch": 2.6423420514476375,
      "grad_norm": 0.6095614433288574,
      "learning_rate": 0.0004419933196853787,
      "loss": 3.5631,
      "step": 24550
    },
    {
      "epoch": 2.6477236034872456,
      "grad_norm": 0.5902724862098694,
      "learning_rate": 0.00044167007865531725,
      "loss": 3.5453,
      "step": 24600
    },
    {
      "epoch": 2.653105155526854,
      "grad_norm": 0.6402266025543213,
      "learning_rate": 0.00044134683762525584,
      "loss": 3.5387,
      "step": 24650
    },
    {
      "epoch": 2.658486707566462,
      "grad_norm": 0.6058180332183838,
      "learning_rate": 0.00044102359659519444,
      "loss": 3.5434,
      "step": 24700
    },
    {
      "epoch": 2.6638682596060703,
      "grad_norm": 0.5468237996101379,
      "learning_rate": 0.00044070035556513303,
      "loss": 3.5322,
      "step": 24750
    },
    {
      "epoch": 2.6692498116456784,
      "grad_norm": 0.5596649646759033,
      "learning_rate": 0.00044037711453507163,
      "loss": 3.547,
      "step": 24800
    },
    {
      "epoch": 2.674631363685287,
      "grad_norm": 0.5948994755744934,
      "learning_rate": 0.0004400538735050102,
      "loss": 3.5358,
      "step": 24850
    },
    {
      "epoch": 2.680012915724895,
      "grad_norm": 0.5811374187469482,
      "learning_rate": 0.00043973063247494876,
      "loss": 3.553,
      "step": 24900
    },
    {
      "epoch": 2.685394467764503,
      "grad_norm": 0.6123229265213013,
      "learning_rate": 0.00043940739144488736,
      "loss": 3.5437,
      "step": 24950
    },
    {
      "epoch": 2.6907760198041117,
      "grad_norm": 0.6782708168029785,
      "learning_rate": 0.0004390841504148259,
      "loss": 3.537,
      "step": 25000
    },
    {
      "epoch": 2.6907760198041117,
      "eval_accuracy": 0.36760922684669434,
      "eval_loss": 3.523207426071167,
      "eval_runtime": 184.4452,
      "eval_samples_per_second": 97.65,
      "eval_steps_per_second": 6.105,
      "step": 25000
    },
    {
      "epoch": 2.69615757184372,
      "grad_norm": 0.5596674680709839,
      "learning_rate": 0.00043876090938476455,
      "loss": 3.5569,
      "step": 25050
    },
    {
      "epoch": 2.701539123883328,
      "grad_norm": 0.5176482200622559,
      "learning_rate": 0.00043843766835470314,
      "loss": 3.5328,
      "step": 25100
    },
    {
      "epoch": 2.706920675922936,
      "grad_norm": 0.5768962502479553,
      "learning_rate": 0.0004381144273246417,
      "loss": 3.5506,
      "step": 25150
    },
    {
      "epoch": 2.712302227962544,
      "grad_norm": 0.6353698968887329,
      "learning_rate": 0.0004377911862945803,
      "loss": 3.551,
      "step": 25200
    },
    {
      "epoch": 2.7176837800021527,
      "grad_norm": 0.5625211000442505,
      "learning_rate": 0.00043746794526451887,
      "loss": 3.5369,
      "step": 25250
    },
    {
      "epoch": 2.723065332041761,
      "grad_norm": 0.5924413800239563,
      "learning_rate": 0.00043714470423445747,
      "loss": 3.5389,
      "step": 25300
    },
    {
      "epoch": 2.728446884081369,
      "grad_norm": 0.5794433951377869,
      "learning_rate": 0.00043682146320439606,
      "loss": 3.5541,
      "step": 25350
    },
    {
      "epoch": 2.7338284361209775,
      "grad_norm": 0.5895462036132812,
      "learning_rate": 0.00043649822217433466,
      "loss": 3.5359,
      "step": 25400
    },
    {
      "epoch": 2.7392099881605856,
      "grad_norm": 0.5819709300994873,
      "learning_rate": 0.0004361749811442732,
      "loss": 3.5474,
      "step": 25450
    },
    {
      "epoch": 2.7445915402001937,
      "grad_norm": 0.6035992503166199,
      "learning_rate": 0.0004358517401142118,
      "loss": 3.5515,
      "step": 25500
    },
    {
      "epoch": 2.749973092239802,
      "grad_norm": 0.5621617436408997,
      "learning_rate": 0.00043552849908415033,
      "loss": 3.5543,
      "step": 25550
    },
    {
      "epoch": 2.7553546442794103,
      "grad_norm": 0.5853516459465027,
      "learning_rate": 0.000435205258054089,
      "loss": 3.5174,
      "step": 25600
    },
    {
      "epoch": 2.7607361963190185,
      "grad_norm": 0.5553240180015564,
      "learning_rate": 0.0004348820170240276,
      "loss": 3.5281,
      "step": 25650
    },
    {
      "epoch": 2.7661177483586266,
      "grad_norm": 0.6032129526138306,
      "learning_rate": 0.0004345587759939661,
      "loss": 3.5529,
      "step": 25700
    },
    {
      "epoch": 2.7714993003982347,
      "grad_norm": 0.5731261968612671,
      "learning_rate": 0.0004342355349639047,
      "loss": 3.532,
      "step": 25750
    },
    {
      "epoch": 2.776880852437843,
      "grad_norm": 0.5767373442649841,
      "learning_rate": 0.0004339122939338433,
      "loss": 3.5536,
      "step": 25800
    },
    {
      "epoch": 2.7822624044774513,
      "grad_norm": 0.5856601595878601,
      "learning_rate": 0.0004335955177243831,
      "loss": 3.5332,
      "step": 25850
    },
    {
      "epoch": 2.7876439565170594,
      "grad_norm": 0.5949828624725342,
      "learning_rate": 0.00043327227669432165,
      "loss": 3.525,
      "step": 25900
    },
    {
      "epoch": 2.793025508556668,
      "grad_norm": 0.6371858716011047,
      "learning_rate": 0.00043294903566426025,
      "loss": 3.5449,
      "step": 25950
    },
    {
      "epoch": 2.798407060596276,
      "grad_norm": 0.5875661969184875,
      "learning_rate": 0.0004326257946341989,
      "loss": 3.5296,
      "step": 26000
    },
    {
      "epoch": 2.798407060596276,
      "eval_accuracy": 0.3682672287352989,
      "eval_loss": 3.5160295963287354,
      "eval_runtime": 183.9997,
      "eval_samples_per_second": 97.886,
      "eval_steps_per_second": 6.12,
      "step": 26000
    },
    {
      "epoch": 2.803788612635884,
      "grad_norm": 0.5926333665847778,
      "learning_rate": 0.00043230255360413744,
      "loss": 3.5295,
      "step": 26050
    },
    {
      "epoch": 2.8091701646754923,
      "grad_norm": 0.5747031569480896,
      "learning_rate": 0.00043197931257407603,
      "loss": 3.518,
      "step": 26100
    },
    {
      "epoch": 2.8145517167151004,
      "grad_norm": 0.561006486415863,
      "learning_rate": 0.0004316560715440146,
      "loss": 3.5353,
      "step": 26150
    },
    {
      "epoch": 2.819933268754709,
      "grad_norm": 0.5835769176483154,
      "learning_rate": 0.00043133283051395317,
      "loss": 3.526,
      "step": 26200
    },
    {
      "epoch": 2.825314820794317,
      "grad_norm": 0.5821505784988403,
      "learning_rate": 0.0004310095894838918,
      "loss": 3.5406,
      "step": 26250
    },
    {
      "epoch": 2.830696372833925,
      "grad_norm": 0.6165505647659302,
      "learning_rate": 0.0004306863484538304,
      "loss": 3.5351,
      "step": 26300
    },
    {
      "epoch": 2.8360779248735337,
      "grad_norm": 0.6276370882987976,
      "learning_rate": 0.00043036310742376895,
      "loss": 3.5383,
      "step": 26350
    },
    {
      "epoch": 2.841459476913142,
      "grad_norm": 0.6106695532798767,
      "learning_rate": 0.00043004633121430876,
      "loss": 3.5334,
      "step": 26400
    },
    {
      "epoch": 2.84684102895275,
      "grad_norm": 0.5662029981613159,
      "learning_rate": 0.00042972309018424735,
      "loss": 3.5446,
      "step": 26450
    },
    {
      "epoch": 2.852222580992358,
      "grad_norm": 0.5863426923751831,
      "learning_rate": 0.00042939984915418594,
      "loss": 3.5405,
      "step": 26500
    },
    {
      "epoch": 2.857604133031966,
      "grad_norm": 0.6597570776939392,
      "learning_rate": 0.0004290766081241245,
      "loss": 3.5122,
      "step": 26550
    },
    {
      "epoch": 2.8629856850715747,
      "grad_norm": 0.6128398776054382,
      "learning_rate": 0.0004287533670940631,
      "loss": 3.5593,
      "step": 26600
    },
    {
      "epoch": 2.868367237111183,
      "grad_norm": 0.6450544595718384,
      "learning_rate": 0.00042843012606400173,
      "loss": 3.526,
      "step": 26650
    },
    {
      "epoch": 2.873748789150791,
      "grad_norm": 0.5586770176887512,
      "learning_rate": 0.00042810688503394027,
      "loss": 3.5359,
      "step": 26700
    },
    {
      "epoch": 2.8791303411903995,
      "grad_norm": 0.576113224029541,
      "learning_rate": 0.00042778364400387886,
      "loss": 3.5216,
      "step": 26750
    },
    {
      "epoch": 2.8845118932300076,
      "grad_norm": 0.5750468969345093,
      "learning_rate": 0.0004274604029738174,
      "loss": 3.5345,
      "step": 26800
    },
    {
      "epoch": 2.8898934452696157,
      "grad_norm": 0.5640283823013306,
      "learning_rate": 0.000427137161943756,
      "loss": 3.5384,
      "step": 26850
    },
    {
      "epoch": 2.895274997309224,
      "grad_norm": 0.5613973140716553,
      "learning_rate": 0.00042681392091369465,
      "loss": 3.5314,
      "step": 26900
    },
    {
      "epoch": 2.9006565493488323,
      "grad_norm": 0.5923638343811035,
      "learning_rate": 0.0004264906798836332,
      "loss": 3.5356,
      "step": 26950
    },
    {
      "epoch": 2.9060381013884404,
      "grad_norm": 0.5590032339096069,
      "learning_rate": 0.0004261674388535718,
      "loss": 3.5247,
      "step": 27000
    },
    {
      "epoch": 2.9060381013884404,
      "eval_accuracy": 0.3691369951022538,
      "eval_loss": 3.5068440437316895,
      "eval_runtime": 184.2053,
      "eval_samples_per_second": 97.777,
      "eval_steps_per_second": 6.113,
      "step": 27000
    },
    {
      "epoch": 2.9114196534280485,
      "grad_norm": 0.573150098323822,
      "learning_rate": 0.0004258441978235104,
      "loss": 3.5257,
      "step": 27050
    },
    {
      "epoch": 2.9168012054676566,
      "grad_norm": 0.6146303415298462,
      "learning_rate": 0.0004255209567934489,
      "loss": 3.516,
      "step": 27100
    },
    {
      "epoch": 2.922182757507265,
      "grad_norm": 0.5765056014060974,
      "learning_rate": 0.0004251977157633875,
      "loss": 3.5299,
      "step": 27150
    },
    {
      "epoch": 2.9275643095468733,
      "grad_norm": 0.5774337649345398,
      "learning_rate": 0.00042487447473332616,
      "loss": 3.5398,
      "step": 27200
    },
    {
      "epoch": 2.9329458615864814,
      "grad_norm": 0.5678874254226685,
      "learning_rate": 0.0004245512337032647,
      "loss": 3.5336,
      "step": 27250
    },
    {
      "epoch": 2.93832741362609,
      "grad_norm": 0.5384457111358643,
      "learning_rate": 0.0004242279926732033,
      "loss": 3.541,
      "step": 27300
    },
    {
      "epoch": 2.943708965665698,
      "grad_norm": 0.5983695983886719,
      "learning_rate": 0.00042390475164314184,
      "loss": 3.5286,
      "step": 27350
    },
    {
      "epoch": 2.949090517705306,
      "grad_norm": 0.581414520740509,
      "learning_rate": 0.00042358151061308043,
      "loss": 3.5292,
      "step": 27400
    },
    {
      "epoch": 2.9544720697449143,
      "grad_norm": 0.6593036651611328,
      "learning_rate": 0.000423258269583019,
      "loss": 3.5279,
      "step": 27450
    },
    {
      "epoch": 2.9598536217845224,
      "grad_norm": 0.5813656449317932,
      "learning_rate": 0.0004229350285529576,
      "loss": 3.5494,
      "step": 27500
    },
    {
      "epoch": 2.965235173824131,
      "grad_norm": 0.5790393948554993,
      "learning_rate": 0.0004226117875228962,
      "loss": 3.539,
      "step": 27550
    },
    {
      "epoch": 2.970616725863739,
      "grad_norm": 0.6063200831413269,
      "learning_rate": 0.0004222885464928348,
      "loss": 3.5361,
      "step": 27600
    },
    {
      "epoch": 2.975998277903347,
      "grad_norm": 0.6125507354736328,
      "learning_rate": 0.00042196530546277335,
      "loss": 3.5305,
      "step": 27650
    },
    {
      "epoch": 2.9813798299429557,
      "grad_norm": 0.5802561044692993,
      "learning_rate": 0.00042164206443271195,
      "loss": 3.5324,
      "step": 27700
    },
    {
      "epoch": 2.986761381982564,
      "grad_norm": 0.5806757211685181,
      "learning_rate": 0.0004213188234026505,
      "loss": 3.5344,
      "step": 27750
    },
    {
      "epoch": 2.992142934022172,
      "grad_norm": 0.6241071224212646,
      "learning_rate": 0.00042099558237258914,
      "loss": 3.511,
      "step": 27800
    },
    {
      "epoch": 2.9975244860617805,
      "grad_norm": 0.5668007731437683,
      "learning_rate": 0.00042067234134252773,
      "loss": 3.5157,
      "step": 27850
    },
    {
      "epoch": 3.0029060381013886,
      "grad_norm": 0.6086427569389343,
      "learning_rate": 0.00042034910031246627,
      "loss": 3.4832,
      "step": 27900
    },
    {
      "epoch": 3.0082875901409967,
      "grad_norm": 0.5952228307723999,
      "learning_rate": 0.00042002585928240486,
      "loss": 3.432,
      "step": 27950
    },
    {
      "epoch": 3.0136691421806048,
      "grad_norm": 0.6184588670730591,
      "learning_rate": 0.00041970261825234346,
      "loss": 3.4305,
      "step": 28000
    },
    {
      "epoch": 3.0136691421806048,
      "eval_accuracy": 0.3701027019691489,
      "eval_loss": 3.5004260540008545,
      "eval_runtime": 183.7533,
      "eval_samples_per_second": 98.017,
      "eval_steps_per_second": 6.128,
      "step": 28000
    },
    {
      "epoch": 3.0190506942202133,
      "grad_norm": 0.5777584314346313,
      "learning_rate": 0.00041937937722228205,
      "loss": 3.4477,
      "step": 28050
    },
    {
      "epoch": 3.0244322462598214,
      "grad_norm": 0.5836468935012817,
      "learning_rate": 0.00041905613619222065,
      "loss": 3.4316,
      "step": 28100
    },
    {
      "epoch": 3.0298137982994295,
      "grad_norm": 0.6238242983818054,
      "learning_rate": 0.00041873289516215924,
      "loss": 3.4291,
      "step": 28150
    },
    {
      "epoch": 3.0351953503390376,
      "grad_norm": 0.5937563180923462,
      "learning_rate": 0.0004184096541320978,
      "loss": 3.4618,
      "step": 28200
    },
    {
      "epoch": 3.040576902378646,
      "grad_norm": 0.604397177696228,
      "learning_rate": 0.0004180864131020364,
      "loss": 3.4199,
      "step": 28250
    },
    {
      "epoch": 3.0459584544182543,
      "grad_norm": 0.6072754263877869,
      "learning_rate": 0.0004177631720719749,
      "loss": 3.464,
      "step": 28300
    },
    {
      "epoch": 3.0513400064578624,
      "grad_norm": 0.5920140743255615,
      "learning_rate": 0.00041743993104191357,
      "loss": 3.4444,
      "step": 28350
    },
    {
      "epoch": 3.0567215584974705,
      "grad_norm": 0.5765038132667542,
      "learning_rate": 0.00041711669001185216,
      "loss": 3.4338,
      "step": 28400
    },
    {
      "epoch": 3.062103110537079,
      "grad_norm": 0.607071578502655,
      "learning_rate": 0.0004167934489817907,
      "loss": 3.4439,
      "step": 28450
    },
    {
      "epoch": 3.067484662576687,
      "grad_norm": 0.5591879487037659,
      "learning_rate": 0.0004164702079517293,
      "loss": 3.4612,
      "step": 28500
    },
    {
      "epoch": 3.0728662146162953,
      "grad_norm": 0.6579169631004333,
      "learning_rate": 0.0004161469669216679,
      "loss": 3.4458,
      "step": 28550
    },
    {
      "epoch": 3.0782477666559034,
      "grad_norm": 0.6010002493858337,
      "learning_rate": 0.00041582372589160643,
      "loss": 3.4467,
      "step": 28600
    },
    {
      "epoch": 3.083629318695512,
      "grad_norm": 0.6072065234184265,
      "learning_rate": 0.0004155004848615451,
      "loss": 3.4572,
      "step": 28650
    },
    {
      "epoch": 3.08901087073512,
      "grad_norm": 0.6025614142417908,
      "learning_rate": 0.0004151772438314837,
      "loss": 3.4301,
      "step": 28700
    },
    {
      "epoch": 3.094392422774728,
      "grad_norm": 0.6360799074172974,
      "learning_rate": 0.0004148540028014222,
      "loss": 3.4507,
      "step": 28750
    },
    {
      "epoch": 3.0997739748143363,
      "grad_norm": 0.5660099983215332,
      "learning_rate": 0.0004145307617713608,
      "loss": 3.4559,
      "step": 28800
    },
    {
      "epoch": 3.105155526853945,
      "grad_norm": 0.5821103453636169,
      "learning_rate": 0.00041420752074129935,
      "loss": 3.4548,
      "step": 28850
    },
    {
      "epoch": 3.110537078893553,
      "grad_norm": 0.6310404539108276,
      "learning_rate": 0.000413884279711238,
      "loss": 3.4429,
      "step": 28900
    },
    {
      "epoch": 3.115918630933161,
      "grad_norm": 0.6091339588165283,
      "learning_rate": 0.0004135610386811766,
      "loss": 3.4502,
      "step": 28950
    },
    {
      "epoch": 3.121300182972769,
      "grad_norm": 0.6030012369155884,
      "learning_rate": 0.00041323779765111514,
      "loss": 3.4617,
      "step": 29000
    },
    {
      "epoch": 3.121300182972769,
      "eval_accuracy": 0.3702236326332864,
      "eval_loss": 3.4993674755096436,
      "eval_runtime": 184.1447,
      "eval_samples_per_second": 97.809,
      "eval_steps_per_second": 6.115,
      "step": 29000
    },
    {
      "epoch": 3.1266817350123777,
      "grad_norm": 0.6194199323654175,
      "learning_rate": 0.00041291455662105373,
      "loss": 3.4436,
      "step": 29050
    },
    {
      "epoch": 3.132063287051986,
      "grad_norm": 0.5936355590820312,
      "learning_rate": 0.0004125913155909923,
      "loss": 3.4531,
      "step": 29100
    },
    {
      "epoch": 3.137444839091594,
      "grad_norm": 0.56592857837677,
      "learning_rate": 0.00041226807456093087,
      "loss": 3.4583,
      "step": 29150
    },
    {
      "epoch": 3.1428263911312024,
      "grad_norm": 0.6247716546058655,
      "learning_rate": 0.0004119448335308695,
      "loss": 3.4475,
      "step": 29200
    },
    {
      "epoch": 3.1482079431708105,
      "grad_norm": 0.6644309163093567,
      "learning_rate": 0.0004116215925008081,
      "loss": 3.4745,
      "step": 29250
    },
    {
      "epoch": 3.1535894952104186,
      "grad_norm": 0.5529394149780273,
      "learning_rate": 0.00041129835147074665,
      "loss": 3.4486,
      "step": 29300
    },
    {
      "epoch": 3.1589710472500268,
      "grad_norm": 0.5880443453788757,
      "learning_rate": 0.00041097511044068524,
      "loss": 3.4525,
      "step": 29350
    },
    {
      "epoch": 3.1643525992896353,
      "grad_norm": 0.6674182415008545,
      "learning_rate": 0.0004106518694106238,
      "loss": 3.4421,
      "step": 29400
    },
    {
      "epoch": 3.1697341513292434,
      "grad_norm": 0.5733335018157959,
      "learning_rate": 0.0004103286283805624,
      "loss": 3.4708,
      "step": 29450
    },
    {
      "epoch": 3.1751157033688515,
      "grad_norm": 0.5916309952735901,
      "learning_rate": 0.00041000538735050103,
      "loss": 3.4535,
      "step": 29500
    },
    {
      "epoch": 3.1804972554084596,
      "grad_norm": 0.6200011372566223,
      "learning_rate": 0.00040968214632043957,
      "loss": 3.4564,
      "step": 29550
    },
    {
      "epoch": 3.185878807448068,
      "grad_norm": 0.6224344372749329,
      "learning_rate": 0.00040935890529037816,
      "loss": 3.4616,
      "step": 29600
    },
    {
      "epoch": 3.1912603594876763,
      "grad_norm": 0.5997889041900635,
      "learning_rate": 0.00040903566426031676,
      "loss": 3.4729,
      "step": 29650
    },
    {
      "epoch": 3.1966419115272844,
      "grad_norm": 0.610285222530365,
      "learning_rate": 0.0004087124232302553,
      "loss": 3.4552,
      "step": 29700
    },
    {
      "epoch": 3.2020234635668925,
      "grad_norm": 0.6233900785446167,
      "learning_rate": 0.0004083891822001939,
      "loss": 3.4356,
      "step": 29750
    },
    {
      "epoch": 3.207405015606501,
      "grad_norm": 0.5861396193504333,
      "learning_rate": 0.00040806594117013254,
      "loss": 3.4666,
      "step": 29800
    },
    {
      "epoch": 3.212786567646109,
      "grad_norm": 0.5982944965362549,
      "learning_rate": 0.0004077427001400711,
      "loss": 3.454,
      "step": 29850
    },
    {
      "epoch": 3.2181681196857173,
      "grad_norm": 0.6225701570510864,
      "learning_rate": 0.0004074194591100097,
      "loss": 3.4744,
      "step": 29900
    },
    {
      "epoch": 3.2235496717253254,
      "grad_norm": 0.6209411025047302,
      "learning_rate": 0.0004070962180799482,
      "loss": 3.4622,
      "step": 29950
    },
    {
      "epoch": 3.228931223764934,
      "grad_norm": 0.6087484955787659,
      "learning_rate": 0.0004067729770498868,
      "loss": 3.4523,
      "step": 30000
    },
    {
      "epoch": 3.228931223764934,
      "eval_accuracy": 0.3710652579022974,
      "eval_loss": 3.491379499435425,
      "eval_runtime": 183.8412,
      "eval_samples_per_second": 97.97,
      "eval_steps_per_second": 6.125,
      "step": 30000
    },
    {
      "epoch": 3.234312775804542,
      "grad_norm": 0.563022255897522,
      "learning_rate": 0.00040644973601982546,
      "loss": 3.4529,
      "step": 30050
    },
    {
      "epoch": 3.23969432784415,
      "grad_norm": 0.636675238609314,
      "learning_rate": 0.000406126494989764,
      "loss": 3.447,
      "step": 30100
    },
    {
      "epoch": 3.2450758798837587,
      "grad_norm": 0.6359305381774902,
      "learning_rate": 0.0004058032539597026,
      "loss": 3.4779,
      "step": 30150
    },
    {
      "epoch": 3.250457431923367,
      "grad_norm": 0.6125782132148743,
      "learning_rate": 0.00040548001292964114,
      "loss": 3.4612,
      "step": 30200
    },
    {
      "epoch": 3.255838983962975,
      "grad_norm": 0.5876561999320984,
      "learning_rate": 0.00040515677189957973,
      "loss": 3.4661,
      "step": 30250
    },
    {
      "epoch": 3.261220536002583,
      "grad_norm": 0.5789377689361572,
      "learning_rate": 0.0004048335308695183,
      "loss": 3.4939,
      "step": 30300
    },
    {
      "epoch": 3.2666020880421915,
      "grad_norm": 0.5683958530426025,
      "learning_rate": 0.0004045102898394569,
      "loss": 3.4499,
      "step": 30350
    },
    {
      "epoch": 3.2719836400817996,
      "grad_norm": 0.6178617477416992,
      "learning_rate": 0.0004041870488093955,
      "loss": 3.4683,
      "step": 30400
    },
    {
      "epoch": 3.2773651921214078,
      "grad_norm": 0.5633194446563721,
      "learning_rate": 0.0004038638077793341,
      "loss": 3.4634,
      "step": 30450
    },
    {
      "epoch": 3.282746744161016,
      "grad_norm": Infinity,
      "learning_rate": 0.0004035470315698739,
      "loss": 3.4446,
      "step": 30500
    },
    {
      "epoch": 3.2881282962006244,
      "grad_norm": 0.5629279613494873,
      "learning_rate": 0.0004032237905398125,
      "loss": 3.4439,
      "step": 30550
    },
    {
      "epoch": 3.2935098482402325,
      "grad_norm": 0.621898889541626,
      "learning_rate": 0.00040290054950975105,
      "loss": 3.4689,
      "step": 30600
    },
    {
      "epoch": 3.2988914002798406,
      "grad_norm": 0.6087920665740967,
      "learning_rate": 0.00040257730847968965,
      "loss": 3.4557,
      "step": 30650
    },
    {
      "epoch": 3.304272952319449,
      "grad_norm": 0.5668652057647705,
      "learning_rate": 0.0004022540674496283,
      "loss": 3.4534,
      "step": 30700
    },
    {
      "epoch": 3.3096545043590573,
      "grad_norm": 0.6095494031906128,
      "learning_rate": 0.00040193082641956684,
      "loss": 3.4721,
      "step": 30750
    },
    {
      "epoch": 3.3150360563986654,
      "grad_norm": 0.621385395526886,
      "learning_rate": 0.00040160758538950543,
      "loss": 3.4684,
      "step": 30800
    },
    {
      "epoch": 3.3204176084382735,
      "grad_norm": 0.5755494236946106,
      "learning_rate": 0.00040128434435944397,
      "loss": 3.4652,
      "step": 30850
    },
    {
      "epoch": 3.3257991604778816,
      "grad_norm": 0.6060776710510254,
      "learning_rate": 0.00040096110332938257,
      "loss": 3.4776,
      "step": 30900
    },
    {
      "epoch": 3.33118071251749,
      "grad_norm": 0.5666420459747314,
      "learning_rate": 0.00040063786229932116,
      "loss": 3.4667,
      "step": 30950
    },
    {
      "epoch": 3.3365622645570983,
      "grad_norm": 0.7270104289054871,
      "learning_rate": 0.00040031462126925975,
      "loss": 3.463,
      "step": 31000
    },
    {
      "epoch": 3.3365622645570983,
      "eval_accuracy": 0.3720799672215968,
      "eval_loss": 3.486278533935547,
      "eval_runtime": 184.2811,
      "eval_samples_per_second": 97.737,
      "eval_steps_per_second": 6.11,
      "step": 31000
    },
    {
      "epoch": 3.3419438165967064,
      "grad_norm": 0.69792240858078,
      "learning_rate": 0.00039999138023919835,
      "loss": 3.4517,
      "step": 31050
    },
    {
      "epoch": 3.347325368636315,
      "grad_norm": 0.6014659404754639,
      "learning_rate": 0.00039966813920913694,
      "loss": 3.4484,
      "step": 31100
    },
    {
      "epoch": 3.352706920675923,
      "grad_norm": 0.6934936046600342,
      "learning_rate": 0.0003993448981790755,
      "loss": 3.4876,
      "step": 31150
    },
    {
      "epoch": 3.358088472715531,
      "grad_norm": 0.5674849152565002,
      "learning_rate": 0.0003990216571490141,
      "loss": 3.4499,
      "step": 31200
    },
    {
      "epoch": 3.3634700247551392,
      "grad_norm": 0.5648531913757324,
      "learning_rate": 0.0003986984161189526,
      "loss": 3.4542,
      "step": 31250
    },
    {
      "epoch": 3.368851576794748,
      "grad_norm": 0.5894842743873596,
      "learning_rate": 0.00039837517508889127,
      "loss": 3.4587,
      "step": 31300
    },
    {
      "epoch": 3.374233128834356,
      "grad_norm": 0.6235491633415222,
      "learning_rate": 0.00039805193405882986,
      "loss": 3.4592,
      "step": 31350
    },
    {
      "epoch": 3.379614680873964,
      "grad_norm": 0.5868720412254333,
      "learning_rate": 0.0003977286930287684,
      "loss": 3.4583,
      "step": 31400
    },
    {
      "epoch": 3.384996232913572,
      "grad_norm": 0.6123641133308411,
      "learning_rate": 0.000397405451998707,
      "loss": 3.4609,
      "step": 31450
    },
    {
      "epoch": 3.3903777849531807,
      "grad_norm": 0.6174355149269104,
      "learning_rate": 0.00039708221096864554,
      "loss": 3.4727,
      "step": 31500
    },
    {
      "epoch": 3.3957593369927888,
      "grad_norm": 0.6319126486778259,
      "learning_rate": 0.00039675896993858413,
      "loss": 3.4701,
      "step": 31550
    },
    {
      "epoch": 3.401140889032397,
      "grad_norm": 0.5984129309654236,
      "learning_rate": 0.0003964357289085228,
      "loss": 3.4613,
      "step": 31600
    },
    {
      "epoch": 3.4065224410720054,
      "grad_norm": 0.6103127598762512,
      "learning_rate": 0.0003961124878784613,
      "loss": 3.4529,
      "step": 31650
    },
    {
      "epoch": 3.4119039931116135,
      "grad_norm": 0.6017516851425171,
      "learning_rate": 0.0003957892468483999,
      "loss": 3.4687,
      "step": 31700
    },
    {
      "epoch": 3.4172855451512216,
      "grad_norm": 0.5514386296272278,
      "learning_rate": 0.0003954660058183385,
      "loss": 3.4386,
      "step": 31750
    },
    {
      "epoch": 3.4226670971908297,
      "grad_norm": 0.5874953866004944,
      "learning_rate": 0.00039514276478827705,
      "loss": 3.4526,
      "step": 31800
    },
    {
      "epoch": 3.428048649230438,
      "grad_norm": 0.5859862565994263,
      "learning_rate": 0.0003948195237582157,
      "loss": 3.4586,
      "step": 31850
    },
    {
      "epoch": 3.4334302012700464,
      "grad_norm": 0.6138542294502258,
      "learning_rate": 0.0003944962827281543,
      "loss": 3.4695,
      "step": 31900
    },
    {
      "epoch": 3.4388117533096545,
      "grad_norm": 0.600860595703125,
      "learning_rate": 0.00039417304169809284,
      "loss": 3.4582,
      "step": 31950
    },
    {
      "epoch": 3.4441933053492626,
      "grad_norm": 0.5975131988525391,
      "learning_rate": 0.00039384980066803143,
      "loss": 3.4803,
      "step": 32000
    },
    {
      "epoch": 3.4441933053492626,
      "eval_accuracy": 0.3724557975605471,
      "eval_loss": 3.480112314224243,
      "eval_runtime": 183.9468,
      "eval_samples_per_second": 97.914,
      "eval_steps_per_second": 6.121,
      "step": 32000
    },
    {
      "epoch": 3.449574857388871,
      "grad_norm": 0.5837554335594177,
      "learning_rate": 0.00039352655963796997,
      "loss": 3.4701,
      "step": 32050
    },
    {
      "epoch": 3.4549564094284793,
      "grad_norm": 0.5798618793487549,
      "learning_rate": 0.00039320331860790857,
      "loss": 3.4666,
      "step": 32100
    },
    {
      "epoch": 3.4603379614680874,
      "grad_norm": 0.5829716920852661,
      "learning_rate": 0.0003928800775778472,
      "loss": 3.4716,
      "step": 32150
    },
    {
      "epoch": 3.4657195135076955,
      "grad_norm": 0.6208363771438599,
      "learning_rate": 0.00039255683654778576,
      "loss": 3.4553,
      "step": 32200
    },
    {
      "epoch": 3.471101065547304,
      "grad_norm": 0.5943921804428101,
      "learning_rate": 0.00039223359551772435,
      "loss": 3.4635,
      "step": 32250
    },
    {
      "epoch": 3.476482617586912,
      "grad_norm": 0.6104678511619568,
      "learning_rate": 0.00039191035448766294,
      "loss": 3.476,
      "step": 32300
    },
    {
      "epoch": 3.4818641696265202,
      "grad_norm": 0.5945563912391663,
      "learning_rate": 0.0003915871134576015,
      "loss": 3.4507,
      "step": 32350
    },
    {
      "epoch": 3.4872457216661283,
      "grad_norm": 0.5837498903274536,
      "learning_rate": 0.0003912638724275401,
      "loss": 3.4554,
      "step": 32400
    },
    {
      "epoch": 3.492627273705737,
      "grad_norm": 0.5934209227561951,
      "learning_rate": 0.00039094063139747873,
      "loss": 3.4461,
      "step": 32450
    },
    {
      "epoch": 3.498008825745345,
      "grad_norm": 0.5918319225311279,
      "learning_rate": 0.00039061739036741727,
      "loss": 3.4565,
      "step": 32500
    },
    {
      "epoch": 3.503390377784953,
      "grad_norm": 0.5761519074440002,
      "learning_rate": 0.00039029414933735586,
      "loss": 3.4546,
      "step": 32550
    },
    {
      "epoch": 3.5087719298245617,
      "grad_norm": 0.6044222712516785,
      "learning_rate": 0.0003899709083072944,
      "loss": 3.4494,
      "step": 32600
    },
    {
      "epoch": 3.5141534818641698,
      "grad_norm": 0.5812404751777649,
      "learning_rate": 0.00038965413209783426,
      "loss": 3.4521,
      "step": 32650
    },
    {
      "epoch": 3.519535033903778,
      "grad_norm": 0.6203511953353882,
      "learning_rate": 0.0003893308910677728,
      "loss": 3.4574,
      "step": 32700
    },
    {
      "epoch": 3.524916585943386,
      "grad_norm": 0.6171815395355225,
      "learning_rate": 0.0003890076500377114,
      "loss": 3.4533,
      "step": 32750
    },
    {
      "epoch": 3.530298137982994,
      "grad_norm": 0.5955988168716431,
      "learning_rate": 0.00038868440900765005,
      "loss": 3.4657,
      "step": 32800
    },
    {
      "epoch": 3.5356796900226026,
      "grad_norm": 0.6277480125427246,
      "learning_rate": 0.0003883611679775886,
      "loss": 3.4643,
      "step": 32850
    },
    {
      "epoch": 3.5410612420622107,
      "grad_norm": 0.5849340558052063,
      "learning_rate": 0.0003880379269475272,
      "loss": 3.45,
      "step": 32900
    },
    {
      "epoch": 3.546442794101819,
      "grad_norm": 0.6264828443527222,
      "learning_rate": 0.0003877146859174657,
      "loss": 3.4678,
      "step": 32950
    },
    {
      "epoch": 3.5518243461414274,
      "grad_norm": 0.611932635307312,
      "learning_rate": 0.0003873914448874043,
      "loss": 3.4668,
      "step": 33000
    },
    {
      "epoch": 3.5518243461414274,
      "eval_accuracy": 0.37325819913705705,
      "eval_loss": 3.4734108448028564,
      "eval_runtime": 184.1437,
      "eval_samples_per_second": 97.809,
      "eval_steps_per_second": 6.115,
      "step": 33000
    },
    {
      "epoch": 3.5572058981810355,
      "grad_norm": 0.6945005059242249,
      "learning_rate": 0.0003870682038573429,
      "loss": 3.4592,
      "step": 33050
    },
    {
      "epoch": 3.5625874502206436,
      "grad_norm": 0.5690363645553589,
      "learning_rate": 0.0003867449628272815,
      "loss": 3.4557,
      "step": 33100
    },
    {
      "epoch": 3.5679690022602517,
      "grad_norm": 0.581488847732544,
      "learning_rate": 0.0003864217217972201,
      "loss": 3.4596,
      "step": 33150
    },
    {
      "epoch": 3.57335055429986,
      "grad_norm": 0.6117942929267883,
      "learning_rate": 0.0003860984807671587,
      "loss": 3.4644,
      "step": 33200
    },
    {
      "epoch": 3.5787321063394684,
      "grad_norm": 0.6318809390068054,
      "learning_rate": 0.00038577523973709724,
      "loss": 3.4514,
      "step": 33250
    },
    {
      "epoch": 3.5841136583790765,
      "grad_norm": 0.6746701002120972,
      "learning_rate": 0.00038545199870703583,
      "loss": 3.4632,
      "step": 33300
    },
    {
      "epoch": 3.5894952104186846,
      "grad_norm": 0.6160062551498413,
      "learning_rate": 0.0003851287576769744,
      "loss": 3.4524,
      "step": 33350
    },
    {
      "epoch": 3.594876762458293,
      "grad_norm": 0.6030224561691284,
      "learning_rate": 0.000384805516646913,
      "loss": 3.4901,
      "step": 33400
    },
    {
      "epoch": 3.6002583144979012,
      "grad_norm": 0.601292610168457,
      "learning_rate": 0.0003844822756168516,
      "loss": 3.4773,
      "step": 33450
    },
    {
      "epoch": 3.6056398665375093,
      "grad_norm": 0.6335893869400024,
      "learning_rate": 0.00038415903458679016,
      "loss": 3.4543,
      "step": 33500
    },
    {
      "epoch": 3.611021418577118,
      "grad_norm": 0.578372061252594,
      "learning_rate": 0.00038383579355672875,
      "loss": 3.4612,
      "step": 33550
    },
    {
      "epoch": 3.616402970616726,
      "grad_norm": 0.5897672176361084,
      "learning_rate": 0.00038351255252666735,
      "loss": 3.463,
      "step": 33600
    },
    {
      "epoch": 3.621784522656334,
      "grad_norm": 0.6561678647994995,
      "learning_rate": 0.00038318931149660594,
      "loss": 3.4569,
      "step": 33650
    },
    {
      "epoch": 3.627166074695942,
      "grad_norm": 0.5887959003448486,
      "learning_rate": 0.00038286607046654454,
      "loss": 3.462,
      "step": 33700
    },
    {
      "epoch": 3.6325476267355503,
      "grad_norm": 0.5961508750915527,
      "learning_rate": 0.00038254282943648313,
      "loss": 3.4697,
      "step": 33750
    },
    {
      "epoch": 3.637929178775159,
      "grad_norm": 0.6062142252922058,
      "learning_rate": 0.00038221958840642167,
      "loss": 3.4746,
      "step": 33800
    },
    {
      "epoch": 3.643310730814767,
      "grad_norm": 0.6095653176307678,
      "learning_rate": 0.00038189634737636027,
      "loss": 3.4457,
      "step": 33850
    },
    {
      "epoch": 3.648692282854375,
      "grad_norm": 0.5974946022033691,
      "learning_rate": 0.0003815731063462988,
      "loss": 3.4774,
      "step": 33900
    },
    {
      "epoch": 3.6540738348939836,
      "grad_norm": 0.5606889128684998,
      "learning_rate": 0.00038124986531623745,
      "loss": 3.4711,
      "step": 33950
    },
    {
      "epoch": 3.6594553869335917,
      "grad_norm": 0.570770800113678,
      "learning_rate": 0.00038092662428617605,
      "loss": 3.4535,
      "step": 34000
    },
    {
      "epoch": 3.6594553869335917,
      "eval_accuracy": 0.3737298613230597,
      "eval_loss": 3.4663772583007812,
      "eval_runtime": 184.3793,
      "eval_samples_per_second": 97.684,
      "eval_steps_per_second": 6.107,
      "step": 34000
    },
    {
      "epoch": 3.6648369389732,
      "grad_norm": 0.6170541048049927,
      "learning_rate": 0.0003806033832561146,
      "loss": 3.4722,
      "step": 34050
    },
    {
      "epoch": 3.670218491012808,
      "grad_norm": 0.5918223857879639,
      "learning_rate": 0.0003802801422260532,
      "loss": 3.4546,
      "step": 34100
    },
    {
      "epoch": 3.675600043052416,
      "grad_norm": 0.6568109393119812,
      "learning_rate": 0.0003799569011959918,
      "loss": 3.4606,
      "step": 34150
    },
    {
      "epoch": 3.6809815950920246,
      "grad_norm": 0.5959489941596985,
      "learning_rate": 0.0003796336601659303,
      "loss": 3.4598,
      "step": 34200
    },
    {
      "epoch": 3.6863631471316327,
      "grad_norm": 0.5992705225944519,
      "learning_rate": 0.00037931041913586897,
      "loss": 3.4666,
      "step": 34250
    },
    {
      "epoch": 3.691744699171241,
      "grad_norm": 0.5846099257469177,
      "learning_rate": 0.00037898717810580756,
      "loss": 3.4678,
      "step": 34300
    },
    {
      "epoch": 3.6971262512108494,
      "grad_norm": 0.5915592312812805,
      "learning_rate": 0.0003786639370757461,
      "loss": 3.462,
      "step": 34350
    },
    {
      "epoch": 3.7025078032504575,
      "grad_norm": 0.5719226598739624,
      "learning_rate": 0.0003783406960456847,
      "loss": 3.457,
      "step": 34400
    },
    {
      "epoch": 3.7078893552900656,
      "grad_norm": 0.614751398563385,
      "learning_rate": 0.00037801745501562324,
      "loss": 3.4643,
      "step": 34450
    },
    {
      "epoch": 3.713270907329674,
      "grad_norm": 0.5760737061500549,
      "learning_rate": 0.0003776942139855619,
      "loss": 3.454,
      "step": 34500
    },
    {
      "epoch": 3.7186524593692822,
      "grad_norm": 0.6911596059799194,
      "learning_rate": 0.0003773709729555005,
      "loss": 3.4618,
      "step": 34550
    },
    {
      "epoch": 3.7240340114088903,
      "grad_norm": 0.5906473994255066,
      "learning_rate": 0.000377047731925439,
      "loss": 3.4483,
      "step": 34600
    },
    {
      "epoch": 3.7294155634484984,
      "grad_norm": 0.6351513862609863,
      "learning_rate": 0.0003767244908953776,
      "loss": 3.4556,
      "step": 34650
    },
    {
      "epoch": 3.7347971154881066,
      "grad_norm": 0.5966063737869263,
      "learning_rate": 0.0003764077146859174,
      "loss": 3.4498,
      "step": 34700
    },
    {
      "epoch": 3.740178667527715,
      "grad_norm": 0.6996062994003296,
      "learning_rate": 0.000376084473655856,
      "loss": 3.4631,
      "step": 34750
    },
    {
      "epoch": 3.745560219567323,
      "grad_norm": 0.6367126107215881,
      "learning_rate": 0.00037576123262579456,
      "loss": 3.4698,
      "step": 34800
    },
    {
      "epoch": 3.7509417716069313,
      "grad_norm": 0.611136794090271,
      "learning_rate": 0.00037543799159573315,
      "loss": 3.469,
      "step": 34850
    },
    {
      "epoch": 3.75632332364654,
      "grad_norm": 0.6161131858825684,
      "learning_rate": 0.0003751147505656718,
      "loss": 3.4493,
      "step": 34900
    },
    {
      "epoch": 3.761704875686148,
      "grad_norm": 0.6409106254577637,
      "learning_rate": 0.00037479150953561034,
      "loss": 3.4366,
      "step": 34950
    },
    {
      "epoch": 3.767086427725756,
      "grad_norm": 0.6586318016052246,
      "learning_rate": 0.00037446826850554894,
      "loss": 3.4434,
      "step": 35000
    },
    {
      "epoch": 3.767086427725756,
      "eval_accuracy": 0.37432256282608767,
      "eval_loss": 3.4600696563720703,
      "eval_runtime": 184.4474,
      "eval_samples_per_second": 97.648,
      "eval_steps_per_second": 6.105,
      "step": 35000
    },
    {
      "epoch": 3.772467979765364,
      "grad_norm": 0.640187680721283,
      "learning_rate": 0.00037414502747548753,
      "loss": 3.456,
      "step": 35050
    },
    {
      "epoch": 3.7778495318049723,
      "grad_norm": 0.6228316426277161,
      "learning_rate": 0.00037382178644542607,
      "loss": 3.4643,
      "step": 35100
    },
    {
      "epoch": 3.783231083844581,
      "grad_norm": 0.6194892525672913,
      "learning_rate": 0.00037349854541536467,
      "loss": 3.4568,
      "step": 35150
    },
    {
      "epoch": 3.788612635884189,
      "grad_norm": 0.5966285467147827,
      "learning_rate": 0.0003731753043853033,
      "loss": 3.4626,
      "step": 35200
    },
    {
      "epoch": 3.793994187923797,
      "grad_norm": 0.5912230610847473,
      "learning_rate": 0.00037285206335524186,
      "loss": 3.4755,
      "step": 35250
    },
    {
      "epoch": 3.7993757399634056,
      "grad_norm": 0.5575622320175171,
      "learning_rate": 0.00037252882232518045,
      "loss": 3.4487,
      "step": 35300
    },
    {
      "epoch": 3.8047572920030137,
      "grad_norm": 0.625277042388916,
      "learning_rate": 0.000372205581295119,
      "loss": 3.4581,
      "step": 35350
    },
    {
      "epoch": 3.810138844042622,
      "grad_norm": 0.6495728492736816,
      "learning_rate": 0.0003718823402650576,
      "loss": 3.461,
      "step": 35400
    },
    {
      "epoch": 3.8155203960822304,
      "grad_norm": 0.5841511487960815,
      "learning_rate": 0.00037155909923499624,
      "loss": 3.4412,
      "step": 35450
    },
    {
      "epoch": 3.8209019481218385,
      "grad_norm": 0.6009453535079956,
      "learning_rate": 0.0003712358582049348,
      "loss": 3.4672,
      "step": 35500
    },
    {
      "epoch": 3.8262835001614466,
      "grad_norm": 0.5865740776062012,
      "learning_rate": 0.00037091261717487337,
      "loss": 3.4327,
      "step": 35550
    },
    {
      "epoch": 3.8316650522010547,
      "grad_norm": 0.6665526628494263,
      "learning_rate": 0.00037058937614481197,
      "loss": 3.4607,
      "step": 35600
    },
    {
      "epoch": 3.837046604240663,
      "grad_norm": 0.6233394145965576,
      "learning_rate": 0.0003702661351147505,
      "loss": 3.4627,
      "step": 35650
    },
    {
      "epoch": 3.8424281562802713,
      "grad_norm": 0.5696569085121155,
      "learning_rate": 0.0003699428940846891,
      "loss": 3.4368,
      "step": 35700
    },
    {
      "epoch": 3.8478097083198795,
      "grad_norm": 0.5669475197792053,
      "learning_rate": 0.00036961965305462775,
      "loss": 3.4578,
      "step": 35750
    },
    {
      "epoch": 3.8531912603594876,
      "grad_norm": 0.6642895936965942,
      "learning_rate": 0.0003692964120245663,
      "loss": 3.4614,
      "step": 35800
    },
    {
      "epoch": 3.858572812399096,
      "grad_norm": 0.6008265018463135,
      "learning_rate": 0.0003689796358151061,
      "loss": 3.4643,
      "step": 35850
    },
    {
      "epoch": 3.863954364438704,
      "grad_norm": 0.6085975170135498,
      "learning_rate": 0.0003686563947850447,
      "loss": 3.4369,
      "step": 35900
    },
    {
      "epoch": 3.8693359164783123,
      "grad_norm": 0.6562402248382568,
      "learning_rate": 0.0003683331537549833,
      "loss": 3.4736,
      "step": 35950
    },
    {
      "epoch": 3.8747174685179204,
      "grad_norm": 0.6528108716011047,
      "learning_rate": 0.0003680099127249218,
      "loss": 3.4518,
      "step": 36000
    },
    {
      "epoch": 3.8747174685179204,
      "eval_accuracy": 0.375415828183277,
      "eval_loss": 3.453058958053589,
      "eval_runtime": 184.0017,
      "eval_samples_per_second": 97.885,
      "eval_steps_per_second": 6.12,
      "step": 36000
    },
    {
      "epoch": 3.8800990205575285,
      "grad_norm": 0.658398449420929,
      "learning_rate": 0.0003676866716948604,
      "loss": 3.4609,
      "step": 36050
    },
    {
      "epoch": 3.885480572597137,
      "grad_norm": 0.6022287011146545,
      "learning_rate": 0.00036736343066479907,
      "loss": 3.4347,
      "step": 36100
    },
    {
      "epoch": 3.890862124636745,
      "grad_norm": 0.6196445822715759,
      "learning_rate": 0.0003670401896347376,
      "loss": 3.4794,
      "step": 36150
    },
    {
      "epoch": 3.8962436766763533,
      "grad_norm": 0.6056769490242004,
      "learning_rate": 0.0003667169486046762,
      "loss": 3.4622,
      "step": 36200
    },
    {
      "epoch": 3.901625228715962,
      "grad_norm": 0.6219421029090881,
      "learning_rate": 0.00036639370757461475,
      "loss": 3.4373,
      "step": 36250
    },
    {
      "epoch": 3.90700678075557,
      "grad_norm": 0.5989059209823608,
      "learning_rate": 0.00036607046654455334,
      "loss": 3.4589,
      "step": 36300
    },
    {
      "epoch": 3.912388332795178,
      "grad_norm": 0.5721907615661621,
      "learning_rate": 0.00036574722551449193,
      "loss": 3.4476,
      "step": 36350
    },
    {
      "epoch": 3.9177698848347866,
      "grad_norm": 0.6650936007499695,
      "learning_rate": 0.00036542398448443053,
      "loss": 3.4837,
      "step": 36400
    },
    {
      "epoch": 3.9231514368743947,
      "grad_norm": 0.6627401113510132,
      "learning_rate": 0.0003651007434543691,
      "loss": 3.4507,
      "step": 36450
    },
    {
      "epoch": 3.928532988914003,
      "grad_norm": 0.5737228989601135,
      "learning_rate": 0.0003647775024243077,
      "loss": 3.4505,
      "step": 36500
    },
    {
      "epoch": 3.933914540953611,
      "grad_norm": 0.6006911993026733,
      "learning_rate": 0.00036445426139424626,
      "loss": 3.4485,
      "step": 36550
    },
    {
      "epoch": 3.939296092993219,
      "grad_norm": 0.6139994263648987,
      "learning_rate": 0.00036413102036418485,
      "loss": 3.4422,
      "step": 36600
    },
    {
      "epoch": 3.9446776450328276,
      "grad_norm": 0.6549245119094849,
      "learning_rate": 0.0003638077793341234,
      "loss": 3.4326,
      "step": 36650
    },
    {
      "epoch": 3.9500591970724357,
      "grad_norm": 0.6006163954734802,
      "learning_rate": 0.00036348453830406204,
      "loss": 3.4325,
      "step": 36700
    },
    {
      "epoch": 3.955440749112044,
      "grad_norm": 0.608538031578064,
      "learning_rate": 0.00036316129727400064,
      "loss": 3.4531,
      "step": 36750
    },
    {
      "epoch": 3.9608223011516523,
      "grad_norm": 0.6129742860794067,
      "learning_rate": 0.0003628380562439392,
      "loss": 3.4542,
      "step": 36800
    },
    {
      "epoch": 3.9662038531912605,
      "grad_norm": 0.6515746116638184,
      "learning_rate": 0.00036251481521387777,
      "loss": 3.4488,
      "step": 36850
    },
    {
      "epoch": 3.9715854052308686,
      "grad_norm": 0.6117689609527588,
      "learning_rate": 0.00036219157418381637,
      "loss": 3.4564,
      "step": 36900
    },
    {
      "epoch": 3.9769669572704767,
      "grad_norm": 0.5867621302604675,
      "learning_rate": 0.0003618683331537549,
      "loss": 3.4538,
      "step": 36950
    },
    {
      "epoch": 3.9823485093100848,
      "grad_norm": 0.6559283137321472,
      "learning_rate": 0.00036154509212369356,
      "loss": 3.4354,
      "step": 37000
    },
    {
      "epoch": 3.9823485093100848,
      "eval_accuracy": 0.37597951936525853,
      "eval_loss": 3.448546886444092,
      "eval_runtime": 184.3936,
      "eval_samples_per_second": 97.677,
      "eval_steps_per_second": 6.107,
      "step": 37000
    },
    {
      "epoch": 3.9877300613496933,
      "grad_norm": 0.6801021099090576,
      "learning_rate": 0.00036122185109363215,
      "loss": 3.4591,
      "step": 37050
    },
    {
      "epoch": 3.9931116133893014,
      "grad_norm": 0.5816856026649475,
      "learning_rate": 0.0003608986100635707,
      "loss": 3.4553,
      "step": 37100
    },
    {
      "epoch": 3.9984931654289095,
      "grad_norm": 0.642889678478241,
      "learning_rate": 0.0003605753690335093,
      "loss": 3.4379,
      "step": 37150
    },
    {
      "epoch": 4.003874717468518,
      "grad_norm": 0.6150274872779846,
      "learning_rate": 0.0003602521280034478,
      "loss": 3.3775,
      "step": 37200
    },
    {
      "epoch": 4.009256269508126,
      "grad_norm": 0.6354314088821411,
      "learning_rate": 0.0003599288869733865,
      "loss": 3.3552,
      "step": 37250
    },
    {
      "epoch": 4.014637821547734,
      "grad_norm": 0.6061065793037415,
      "learning_rate": 0.00035960564594332507,
      "loss": 3.3468,
      "step": 37300
    },
    {
      "epoch": 4.020019373587343,
      "grad_norm": 0.6188650131225586,
      "learning_rate": 0.0003592824049132636,
      "loss": 3.3678,
      "step": 37350
    },
    {
      "epoch": 4.0254009256269505,
      "grad_norm": 0.6503527164459229,
      "learning_rate": 0.0003589591638832022,
      "loss": 3.3557,
      "step": 37400
    },
    {
      "epoch": 4.030782477666559,
      "grad_norm": 0.6220972537994385,
      "learning_rate": 0.0003586359228531408,
      "loss": 3.3714,
      "step": 37450
    },
    {
      "epoch": 4.036164029706168,
      "grad_norm": 0.5954411625862122,
      "learning_rate": 0.00035831268182307934,
      "loss": 3.3669,
      "step": 37500
    },
    {
      "epoch": 4.041545581745775,
      "grad_norm": 0.6053151488304138,
      "learning_rate": 0.000357989440793018,
      "loss": 3.3759,
      "step": 37550
    },
    {
      "epoch": 4.046927133785384,
      "grad_norm": 0.6040852665901184,
      "learning_rate": 0.0003576661997629566,
      "loss": 3.3815,
      "step": 37600
    },
    {
      "epoch": 4.0523086858249915,
      "grad_norm": 0.5903968214988708,
      "learning_rate": 0.0003573429587328951,
      "loss": 3.3498,
      "step": 37650
    },
    {
      "epoch": 4.0576902378646,
      "grad_norm": 0.6199638843536377,
      "learning_rate": 0.0003570197177028337,
      "loss": 3.3596,
      "step": 37700
    },
    {
      "epoch": 4.063071789904209,
      "grad_norm": 0.6323930025100708,
      "learning_rate": 0.00035669647667277226,
      "loss": 3.3769,
      "step": 37750
    },
    {
      "epoch": 4.068453341943816,
      "grad_norm": 0.6106070280075073,
      "learning_rate": 0.00035637323564271085,
      "loss": 3.366,
      "step": 37800
    },
    {
      "epoch": 4.073834893983425,
      "grad_norm": 0.6530179381370544,
      "learning_rate": 0.0003560499946126495,
      "loss": 3.3535,
      "step": 37850
    },
    {
      "epoch": 4.079216446023033,
      "grad_norm": 0.6518010497093201,
      "learning_rate": 0.00035572675358258804,
      "loss": 3.3781,
      "step": 37900
    },
    {
      "epoch": 4.084597998062641,
      "grad_norm": 0.6036293506622314,
      "learning_rate": 0.00035540351255252664,
      "loss": 3.3654,
      "step": 37950
    },
    {
      "epoch": 4.08997955010225,
      "grad_norm": 0.6176841259002686,
      "learning_rate": 0.00035508027152246523,
      "loss": 3.3653,
      "step": 38000
    },
    {
      "epoch": 4.08997955010225,
      "eval_accuracy": 0.3758944441540998,
      "eval_loss": 3.4506988525390625,
      "eval_runtime": 183.9799,
      "eval_samples_per_second": 97.897,
      "eval_steps_per_second": 6.12,
      "step": 38000
    },
    {
      "epoch": 4.095361102141858,
      "grad_norm": 0.6339088678359985,
      "learning_rate": 0.0003547570304924038,
      "loss": 3.3755,
      "step": 38050
    },
    {
      "epoch": 4.100742654181466,
      "grad_norm": 0.5995780229568481,
      "learning_rate": 0.0003544337894623424,
      "loss": 3.3907,
      "step": 38100
    },
    {
      "epoch": 4.106124206221074,
      "grad_norm": 0.5926193594932556,
      "learning_rate": 0.000354110548432281,
      "loss": 3.3701,
      "step": 38150
    },
    {
      "epoch": 4.111505758260682,
      "grad_norm": 0.6507819890975952,
      "learning_rate": 0.00035378730740221956,
      "loss": 3.3711,
      "step": 38200
    },
    {
      "epoch": 4.1168873103002905,
      "grad_norm": 0.642949104309082,
      "learning_rate": 0.00035346406637215815,
      "loss": 3.3694,
      "step": 38250
    },
    {
      "epoch": 4.122268862339899,
      "grad_norm": 0.6337674260139465,
      "learning_rate": 0.0003531408253420967,
      "loss": 3.3579,
      "step": 38300
    },
    {
      "epoch": 4.127650414379507,
      "grad_norm": 0.6153765916824341,
      "learning_rate": 0.0003528175843120353,
      "loss": 3.3886,
      "step": 38350
    },
    {
      "epoch": 4.133031966419115,
      "grad_norm": 0.6312838196754456,
      "learning_rate": 0.00035249434328197394,
      "loss": 3.3786,
      "step": 38400
    },
    {
      "epoch": 4.138413518458724,
      "grad_norm": 0.6568372845649719,
      "learning_rate": 0.0003521711022519125,
      "loss": 3.3806,
      "step": 38450
    },
    {
      "epoch": 4.1437950704983315,
      "grad_norm": 0.6215652227401733,
      "learning_rate": 0.00035184786122185107,
      "loss": 3.3844,
      "step": 38500
    },
    {
      "epoch": 4.14917662253794,
      "grad_norm": 0.6149177551269531,
      "learning_rate": 0.00035152462019178967,
      "loss": 3.3939,
      "step": 38550
    },
    {
      "epoch": 4.154558174577549,
      "grad_norm": 0.6630335450172424,
      "learning_rate": 0.0003512013791617282,
      "loss": 3.4038,
      "step": 38600
    },
    {
      "epoch": 4.159939726617156,
      "grad_norm": 0.7132558822631836,
      "learning_rate": 0.0003508781381316668,
      "loss": 3.3818,
      "step": 38650
    },
    {
      "epoch": 4.165321278656765,
      "grad_norm": 0.6255067586898804,
      "learning_rate": 0.00035055489710160545,
      "loss": 3.3675,
      "step": 38700
    },
    {
      "epoch": 4.1707028306963725,
      "grad_norm": 0.6831744313240051,
      "learning_rate": 0.000350231656071544,
      "loss": 3.3883,
      "step": 38750
    },
    {
      "epoch": 4.176084382735981,
      "grad_norm": 0.6473819017410278,
      "learning_rate": 0.0003499084150414826,
      "loss": 3.3884,
      "step": 38800
    },
    {
      "epoch": 4.18146593477559,
      "grad_norm": 0.6523324251174927,
      "learning_rate": 0.0003495851740114211,
      "loss": 3.3931,
      "step": 38850
    },
    {
      "epoch": 4.186847486815197,
      "grad_norm": 0.6294863820075989,
      "learning_rate": 0.0003492619329813597,
      "loss": 3.4002,
      "step": 38900
    },
    {
      "epoch": 4.192229038854806,
      "grad_norm": 0.6406927704811096,
      "learning_rate": 0.0003489386919512983,
      "loss": 3.3733,
      "step": 38950
    },
    {
      "epoch": 4.197610590894414,
      "grad_norm": 0.6410805583000183,
      "learning_rate": 0.0003486154509212369,
      "loss": 3.3676,
      "step": 39000
    },
    {
      "epoch": 4.197610590894414,
      "eval_accuracy": 0.37683896370787107,
      "eval_loss": 3.4443728923797607,
      "eval_runtime": 184.3071,
      "eval_samples_per_second": 97.723,
      "eval_steps_per_second": 6.109,
      "step": 39000
    },
    {
      "epoch": 4.202992142934022,
      "grad_norm": 0.6174502372741699,
      "learning_rate": 0.0003482922098911755,
      "loss": 3.3747,
      "step": 39050
    },
    {
      "epoch": 4.208373694973631,
      "grad_norm": 0.6655563116073608,
      "learning_rate": 0.0003479689688611141,
      "loss": 3.3754,
      "step": 39100
    },
    {
      "epoch": 4.213755247013238,
      "grad_norm": 0.6777321100234985,
      "learning_rate": 0.00034764572783105264,
      "loss": 3.4173,
      "step": 39150
    },
    {
      "epoch": 4.219136799052847,
      "grad_norm": 0.6062920689582825,
      "learning_rate": 0.00034732248680099123,
      "loss": 3.4046,
      "step": 39200
    },
    {
      "epoch": 4.224518351092455,
      "grad_norm": 0.6097728610038757,
      "learning_rate": 0.0003469992457709299,
      "loss": 3.3879,
      "step": 39250
    },
    {
      "epoch": 4.229899903132063,
      "grad_norm": 0.6410273313522339,
      "learning_rate": 0.0003466760047408684,
      "loss": 3.3904,
      "step": 39300
    },
    {
      "epoch": 4.2352814551716715,
      "grad_norm": 0.6393591165542603,
      "learning_rate": 0.000346352763710807,
      "loss": 3.3986,
      "step": 39350
    },
    {
      "epoch": 4.24066300721128,
      "grad_norm": 0.6021938323974609,
      "learning_rate": 0.00034602952268074556,
      "loss": 3.3914,
      "step": 39400
    },
    {
      "epoch": 4.246044559250888,
      "grad_norm": 0.5947967767715454,
      "learning_rate": 0.00034570628165068415,
      "loss": 3.3783,
      "step": 39450
    },
    {
      "epoch": 4.251426111290496,
      "grad_norm": 0.5978347063064575,
      "learning_rate": 0.0003453830406206227,
      "loss": 3.3923,
      "step": 39500
    },
    {
      "epoch": 4.256807663330104,
      "grad_norm": 0.6636889576911926,
      "learning_rate": 0.00034505979959056134,
      "loss": 3.3822,
      "step": 39550
    },
    {
      "epoch": 4.2621892153697125,
      "grad_norm": 0.6549127697944641,
      "learning_rate": 0.00034473655856049994,
      "loss": 3.3942,
      "step": 39600
    },
    {
      "epoch": 4.267570767409321,
      "grad_norm": 0.6724443435668945,
      "learning_rate": 0.0003444133175304385,
      "loss": 3.3909,
      "step": 39650
    },
    {
      "epoch": 4.272952319448929,
      "grad_norm": 0.6483871340751648,
      "learning_rate": 0.00034409007650037707,
      "loss": 3.4082,
      "step": 39700
    },
    {
      "epoch": 4.278333871488537,
      "grad_norm": 0.5984669327735901,
      "learning_rate": 0.00034376683547031567,
      "loss": 3.3686,
      "step": 39750
    },
    {
      "epoch": 4.283715423528146,
      "grad_norm": 0.6218523383140564,
      "learning_rate": 0.0003434435944402542,
      "loss": 3.3774,
      "step": 39800
    },
    {
      "epoch": 4.2890969755677535,
      "grad_norm": 0.6709349751472473,
      "learning_rate": 0.00034312035341019286,
      "loss": 3.3813,
      "step": 39850
    },
    {
      "epoch": 4.294478527607362,
      "grad_norm": 0.6159582138061523,
      "learning_rate": 0.00034280357720073266,
      "loss": 3.4031,
      "step": 39900
    },
    {
      "epoch": 4.299860079646971,
      "grad_norm": 0.6412374377250671,
      "learning_rate": 0.00034248033617067126,
      "loss": 3.3988,
      "step": 39950
    },
    {
      "epoch": 4.305241631686578,
      "grad_norm": 0.6323827505111694,
      "learning_rate": 0.00034215709514060985,
      "loss": 3.3825,
      "step": 40000
    },
    {
      "epoch": 4.305241631686578,
      "eval_accuracy": 0.3769284936874302,
      "eval_loss": 3.4414515495300293,
      "eval_runtime": 184.025,
      "eval_samples_per_second": 97.873,
      "eval_steps_per_second": 6.119,
      "step": 40000
    },
    {
      "epoch": 4.310623183726187,
      "grad_norm": 0.6323182582855225,
      "learning_rate": 0.0003418338541105484,
      "loss": 3.394,
      "step": 40050
    },
    {
      "epoch": 4.3160047357657945,
      "grad_norm": 0.6397620439529419,
      "learning_rate": 0.000341510613080487,
      "loss": 3.4039,
      "step": 40100
    },
    {
      "epoch": 4.321386287805403,
      "grad_norm": 0.6485004425048828,
      "learning_rate": 0.0003411873720504255,
      "loss": 3.3782,
      "step": 40150
    },
    {
      "epoch": 4.326767839845012,
      "grad_norm": 0.6374657154083252,
      "learning_rate": 0.0003408641310203642,
      "loss": 3.3768,
      "step": 40200
    },
    {
      "epoch": 4.332149391884619,
      "grad_norm": 0.6392961740493774,
      "learning_rate": 0.00034054088999030277,
      "loss": 3.4142,
      "step": 40250
    },
    {
      "epoch": 4.337530943924228,
      "grad_norm": 0.6229289770126343,
      "learning_rate": 0.0003402176489602413,
      "loss": 3.4056,
      "step": 40300
    },
    {
      "epoch": 4.342912495963836,
      "grad_norm": 0.6302274465560913,
      "learning_rate": 0.0003398944079301799,
      "loss": 3.3819,
      "step": 40350
    },
    {
      "epoch": 4.348294048003444,
      "grad_norm": 0.6370872855186462,
      "learning_rate": 0.0003395711669001185,
      "loss": 3.3847,
      "step": 40400
    },
    {
      "epoch": 4.3536756000430525,
      "grad_norm": 0.6489060521125793,
      "learning_rate": 0.00033924792587005704,
      "loss": 3.3862,
      "step": 40450
    },
    {
      "epoch": 4.359057152082661,
      "grad_norm": 0.5757327675819397,
      "learning_rate": 0.0003389246848399957,
      "loss": 3.3863,
      "step": 40500
    },
    {
      "epoch": 4.364438704122269,
      "grad_norm": 0.6207980513572693,
      "learning_rate": 0.0003386014438099343,
      "loss": 3.3982,
      "step": 40550
    },
    {
      "epoch": 4.369820256161877,
      "grad_norm": 0.6486942768096924,
      "learning_rate": 0.0003382782027798728,
      "loss": 3.3753,
      "step": 40600
    },
    {
      "epoch": 4.375201808201485,
      "grad_norm": 0.6791574358940125,
      "learning_rate": 0.0003379549617498114,
      "loss": 3.3852,
      "step": 40650
    },
    {
      "epoch": 4.3805833602410935,
      "grad_norm": 0.6521627306938171,
      "learning_rate": 0.00033763172071974996,
      "loss": 3.4018,
      "step": 40700
    },
    {
      "epoch": 4.385964912280702,
      "grad_norm": 0.6518412828445435,
      "learning_rate": 0.00033730847968968855,
      "loss": 3.3954,
      "step": 40750
    },
    {
      "epoch": 4.39134646432031,
      "grad_norm": 0.6499926447868347,
      "learning_rate": 0.0003369852386596272,
      "loss": 3.4053,
      "step": 40800
    },
    {
      "epoch": 4.396728016359918,
      "grad_norm": 0.580716609954834,
      "learning_rate": 0.00033666199762956574,
      "loss": 3.3932,
      "step": 40850
    },
    {
      "epoch": 4.402109568399527,
      "grad_norm": 0.6994695663452148,
      "learning_rate": 0.00033633875659950434,
      "loss": 3.4005,
      "step": 40900
    },
    {
      "epoch": 4.4074911204391345,
      "grad_norm": 0.6295654773712158,
      "learning_rate": 0.0003360155155694429,
      "loss": 3.378,
      "step": 40950
    },
    {
      "epoch": 4.412872672478743,
      "grad_norm": 0.6220209002494812,
      "learning_rate": 0.0003356922745393815,
      "loss": 3.4014,
      "step": 41000
    },
    {
      "epoch": 4.412872672478743,
      "eval_accuracy": 0.3773276843972606,
      "eval_loss": 3.4385666847229004,
      "eval_runtime": 184.2414,
      "eval_samples_per_second": 97.758,
      "eval_steps_per_second": 6.112,
      "step": 41000
    },
    {
      "epoch": 4.418254224518351,
      "grad_norm": 0.624264121055603,
      "learning_rate": 0.0003353690335093201,
      "loss": 3.3994,
      "step": 41050
    },
    {
      "epoch": 4.423635776557959,
      "grad_norm": 0.626992404460907,
      "learning_rate": 0.0003350457924792587,
      "loss": 3.408,
      "step": 41100
    },
    {
      "epoch": 4.429017328597568,
      "grad_norm": 0.6352009177207947,
      "learning_rate": 0.00033472255144919726,
      "loss": 3.3896,
      "step": 41150
    },
    {
      "epoch": 4.4343988806371755,
      "grad_norm": 0.6544060707092285,
      "learning_rate": 0.00033439931041913585,
      "loss": 3.3747,
      "step": 41200
    },
    {
      "epoch": 4.439780432676784,
      "grad_norm": 0.6308299899101257,
      "learning_rate": 0.0003340760693890744,
      "loss": 3.3783,
      "step": 41250
    },
    {
      "epoch": 4.445161984716393,
      "grad_norm": 0.6337025165557861,
      "learning_rate": 0.000333752828359013,
      "loss": 3.3802,
      "step": 41300
    },
    {
      "epoch": 4.450543536756,
      "grad_norm": 0.6184533834457397,
      "learning_rate": 0.00033342958732895164,
      "loss": 3.3984,
      "step": 41350
    },
    {
      "epoch": 4.455925088795609,
      "grad_norm": 0.6686426401138306,
      "learning_rate": 0.0003331063462988902,
      "loss": 3.388,
      "step": 41400
    },
    {
      "epoch": 4.461306640835216,
      "grad_norm": 0.6282316446304321,
      "learning_rate": 0.00033278310526882877,
      "loss": 3.3885,
      "step": 41450
    },
    {
      "epoch": 4.466688192874825,
      "grad_norm": 0.6537407636642456,
      "learning_rate": 0.0003324598642387673,
      "loss": 3.3984,
      "step": 41500
    },
    {
      "epoch": 4.4720697449144335,
      "grad_norm": 0.6691215634346008,
      "learning_rate": 0.0003321366232087059,
      "loss": 3.4119,
      "step": 41550
    },
    {
      "epoch": 4.477451296954041,
      "grad_norm": 0.661702036857605,
      "learning_rate": 0.0003318133821786445,
      "loss": 3.389,
      "step": 41600
    },
    {
      "epoch": 4.48283284899365,
      "grad_norm": 0.6723365187644958,
      "learning_rate": 0.0003314901411485831,
      "loss": 3.3925,
      "step": 41650
    },
    {
      "epoch": 4.488214401033258,
      "grad_norm": 0.6523467302322388,
      "learning_rate": 0.0003311669001185217,
      "loss": 3.3907,
      "step": 41700
    },
    {
      "epoch": 4.493595953072866,
      "grad_norm": 0.6824305653572083,
      "learning_rate": 0.0003308436590884603,
      "loss": 3.3988,
      "step": 41750
    },
    {
      "epoch": 4.4989775051124745,
      "grad_norm": 0.6121111512184143,
      "learning_rate": 0.0003305204180583988,
      "loss": 3.3822,
      "step": 41800
    },
    {
      "epoch": 4.504359057152083,
      "grad_norm": 0.6316137909889221,
      "learning_rate": 0.0003301971770283374,
      "loss": 3.3759,
      "step": 41850
    },
    {
      "epoch": 4.509740609191691,
      "grad_norm": 0.672080934047699,
      "learning_rate": 0.0003298804008188772,
      "loss": 3.3984,
      "step": 41900
    },
    {
      "epoch": 4.515122161231299,
      "grad_norm": 0.6413889527320862,
      "learning_rate": 0.0003295571597888158,
      "loss": 3.3855,
      "step": 41950
    },
    {
      "epoch": 4.520503713270907,
      "grad_norm": 0.6120907068252563,
      "learning_rate": 0.00032923391875875447,
      "loss": 3.4053,
      "step": 42000
    },
    {
      "epoch": 4.520503713270907,
      "eval_accuracy": 0.37800698225187673,
      "eval_loss": 3.4314796924591064,
      "eval_runtime": 184.2191,
      "eval_samples_per_second": 97.769,
      "eval_steps_per_second": 6.112,
      "step": 42000
    },
    {
      "epoch": 4.5258852653105155,
      "grad_norm": 0.72449791431427,
      "learning_rate": 0.000328910677728693,
      "loss": 3.4036,
      "step": 42050
    },
    {
      "epoch": 4.531266817350124,
      "grad_norm": 0.6400125622749329,
      "learning_rate": 0.0003285874366986316,
      "loss": 3.3921,
      "step": 42100
    },
    {
      "epoch": 4.536648369389732,
      "grad_norm": 0.653170108795166,
      "learning_rate": 0.00032826419566857015,
      "loss": 3.3952,
      "step": 42150
    },
    {
      "epoch": 4.54202992142934,
      "grad_norm": 1.8366317749023438,
      "learning_rate": 0.00032794095463850874,
      "loss": 3.4014,
      "step": 42200
    },
    {
      "epoch": 4.547411473468949,
      "grad_norm": 0.6847442388534546,
      "learning_rate": 0.0003276177136084473,
      "loss": 3.3887,
      "step": 42250
    },
    {
      "epoch": 4.5527930255085565,
      "grad_norm": 0.677156925201416,
      "learning_rate": 0.00032729447257838593,
      "loss": 3.3922,
      "step": 42300
    },
    {
      "epoch": 4.558174577548165,
      "grad_norm": 0.6153601408004761,
      "learning_rate": 0.0003269776963689257,
      "loss": 3.4018,
      "step": 42350
    },
    {
      "epoch": 4.563556129587774,
      "grad_norm": 0.7138859033584595,
      "learning_rate": 0.00032665445533886433,
      "loss": 3.3989,
      "step": 42400
    },
    {
      "epoch": 4.568937681627381,
      "grad_norm": 0.650017261505127,
      "learning_rate": 0.0003263312143088029,
      "loss": 3.4075,
      "step": 42450
    },
    {
      "epoch": 4.57431923366699,
      "grad_norm": 0.6374242305755615,
      "learning_rate": 0.00032600797327874147,
      "loss": 3.4105,
      "step": 42500
    },
    {
      "epoch": 4.579700785706597,
      "grad_norm": 0.6373197436332703,
      "learning_rate": 0.00032568473224868006,
      "loss": 3.4061,
      "step": 42550
    },
    {
      "epoch": 4.585082337746206,
      "grad_norm": 0.6459031701087952,
      "learning_rate": 0.00032536149121861866,
      "loss": 3.3887,
      "step": 42600
    },
    {
      "epoch": 4.5904638897858145,
      "grad_norm": 0.689174234867096,
      "learning_rate": 0.00032503825018855725,
      "loss": 3.3896,
      "step": 42650
    },
    {
      "epoch": 4.595845441825422,
      "grad_norm": 0.6378280520439148,
      "learning_rate": 0.00032471500915849584,
      "loss": 3.3969,
      "step": 42700
    },
    {
      "epoch": 4.601226993865031,
      "grad_norm": 0.6298919916152954,
      "learning_rate": 0.00032439176812843444,
      "loss": 3.4001,
      "step": 42750
    },
    {
      "epoch": 4.606608545904638,
      "grad_norm": 0.720784604549408,
      "learning_rate": 0.000324068527098373,
      "loss": 3.3823,
      "step": 42800
    },
    {
      "epoch": 4.611990097944247,
      "grad_norm": 0.6367201805114746,
      "learning_rate": 0.0003237452860683116,
      "loss": 3.3951,
      "step": 42850
    },
    {
      "epoch": 4.6173716499838555,
      "grad_norm": 0.6843454837799072,
      "learning_rate": 0.0003234220450382501,
      "loss": 3.3998,
      "step": 42900
    },
    {
      "epoch": 4.622753202023463,
      "grad_norm": 0.6279202699661255,
      "learning_rate": 0.00032309880400818876,
      "loss": 3.3924,
      "step": 42950
    },
    {
      "epoch": 4.628134754063072,
      "grad_norm": 0.6313354969024658,
      "learning_rate": 0.00032277556297812736,
      "loss": 3.3822,
      "step": 43000
    },
    {
      "epoch": 4.628134754063072,
      "eval_accuracy": 0.3785127614446531,
      "eval_loss": 3.4267942905426025,
      "eval_runtime": 184.394,
      "eval_samples_per_second": 97.677,
      "eval_steps_per_second": 6.106,
      "step": 43000
    },
    {
      "epoch": 4.63351630610268,
      "grad_norm": 0.5984571576118469,
      "learning_rate": 0.0003224523219480659,
      "loss": 3.3979,
      "step": 43050
    },
    {
      "epoch": 4.638897858142288,
      "grad_norm": 0.611714243888855,
      "learning_rate": 0.0003221290809180045,
      "loss": 3.409,
      "step": 43100
    },
    {
      "epoch": 4.6442794101818965,
      "grad_norm": 0.6246342062950134,
      "learning_rate": 0.0003218058398879431,
      "loss": 3.4127,
      "step": 43150
    },
    {
      "epoch": 4.649660962221505,
      "grad_norm": 0.6131724715232849,
      "learning_rate": 0.00032148259885788163,
      "loss": 3.3924,
      "step": 43200
    },
    {
      "epoch": 4.655042514261113,
      "grad_norm": 0.6691821813583374,
      "learning_rate": 0.0003211593578278203,
      "loss": 3.394,
      "step": 43250
    },
    {
      "epoch": 4.660424066300721,
      "grad_norm": 0.7128880620002747,
      "learning_rate": 0.00032083611679775887,
      "loss": 3.4036,
      "step": 43300
    },
    {
      "epoch": 4.665805618340329,
      "grad_norm": 0.6468649506568909,
      "learning_rate": 0.0003205128757676974,
      "loss": 3.3947,
      "step": 43350
    },
    {
      "epoch": 4.6711871703799375,
      "grad_norm": 0.6241949796676636,
      "learning_rate": 0.000320189634737636,
      "loss": 3.3972,
      "step": 43400
    },
    {
      "epoch": 4.676568722419546,
      "grad_norm": 0.6339065432548523,
      "learning_rate": 0.00031986639370757455,
      "loss": 3.39,
      "step": 43450
    },
    {
      "epoch": 4.681950274459154,
      "grad_norm": 0.626457154750824,
      "learning_rate": 0.0003195431526775132,
      "loss": 3.3969,
      "step": 43500
    },
    {
      "epoch": 4.687331826498762,
      "grad_norm": 0.6233946681022644,
      "learning_rate": 0.0003192199116474518,
      "loss": 3.3966,
      "step": 43550
    },
    {
      "epoch": 4.692713378538371,
      "grad_norm": 0.5824977159500122,
      "learning_rate": 0.00031889667061739033,
      "loss": 3.4006,
      "step": 43600
    },
    {
      "epoch": 4.6980949305779784,
      "grad_norm": 0.6422542929649353,
      "learning_rate": 0.0003185734295873289,
      "loss": 3.4118,
      "step": 43650
    },
    {
      "epoch": 4.703476482617587,
      "grad_norm": 0.6130107641220093,
      "learning_rate": 0.00031825018855726747,
      "loss": 3.3822,
      "step": 43700
    },
    {
      "epoch": 4.7088580346571955,
      "grad_norm": 0.6715350151062012,
      "learning_rate": 0.00031792694752720606,
      "loss": 3.3828,
      "step": 43750
    },
    {
      "epoch": 4.714239586696803,
      "grad_norm": 0.6314501762390137,
      "learning_rate": 0.0003176037064971447,
      "loss": 3.3802,
      "step": 43800
    },
    {
      "epoch": 4.719621138736412,
      "grad_norm": 0.6555110812187195,
      "learning_rate": 0.0003172804654670833,
      "loss": 3.3969,
      "step": 43850
    },
    {
      "epoch": 4.725002690776019,
      "grad_norm": 0.6788358688354492,
      "learning_rate": 0.00031695722443702185,
      "loss": 3.3978,
      "step": 43900
    },
    {
      "epoch": 4.730384242815628,
      "grad_norm": 0.5993989706039429,
      "learning_rate": 0.00031663398340696044,
      "loss": 3.3878,
      "step": 43950
    },
    {
      "epoch": 4.7357657948552365,
      "grad_norm": 0.7137261629104614,
      "learning_rate": 0.000316310742376899,
      "loss": 3.392,
      "step": 44000
    },
    {
      "epoch": 4.7357657948552365,
      "eval_accuracy": 0.3790802554777082,
      "eval_loss": 3.4194159507751465,
      "eval_runtime": 184.4475,
      "eval_samples_per_second": 97.648,
      "eval_steps_per_second": 6.105,
      "step": 44000
    },
    {
      "epoch": 4.741147346894844,
      "grad_norm": 0.6539814472198486,
      "learning_rate": 0.0003159875013468376,
      "loss": 3.3988,
      "step": 44050
    },
    {
      "epoch": 4.746528898934453,
      "grad_norm": 0.6142458915710449,
      "learning_rate": 0.0003156642603167762,
      "loss": 3.3952,
      "step": 44100
    },
    {
      "epoch": 4.751910450974061,
      "grad_norm": 0.6481937170028687,
      "learning_rate": 0.00031534101928671476,
      "loss": 3.397,
      "step": 44150
    },
    {
      "epoch": 4.757292003013669,
      "grad_norm": 0.6322684288024902,
      "learning_rate": 0.00031501777825665336,
      "loss": 3.3964,
      "step": 44200
    },
    {
      "epoch": 4.7626735550532775,
      "grad_norm": 0.6843492388725281,
      "learning_rate": 0.0003146945372265919,
      "loss": 3.3975,
      "step": 44250
    },
    {
      "epoch": 4.768055107092886,
      "grad_norm": 0.6624308228492737,
      "learning_rate": 0.0003143712961965305,
      "loss": 3.3844,
      "step": 44300
    },
    {
      "epoch": 4.773436659132494,
      "grad_norm": 0.6903315186500549,
      "learning_rate": 0.0003140480551664691,
      "loss": 3.3856,
      "step": 44350
    },
    {
      "epoch": 4.778818211172102,
      "grad_norm": 0.6254662871360779,
      "learning_rate": 0.0003137248141364077,
      "loss": 3.3876,
      "step": 44400
    },
    {
      "epoch": 4.78419976321171,
      "grad_norm": 0.632722795009613,
      "learning_rate": 0.0003134015731063463,
      "loss": 3.3982,
      "step": 44450
    },
    {
      "epoch": 4.7895813152513185,
      "grad_norm": 0.6366475224494934,
      "learning_rate": 0.0003130783320762849,
      "loss": 3.3935,
      "step": 44500
    },
    {
      "epoch": 4.794962867290927,
      "grad_norm": 0.5925983786582947,
      "learning_rate": 0.0003127550910462234,
      "loss": 3.3836,
      "step": 44550
    },
    {
      "epoch": 4.800344419330535,
      "grad_norm": 0.6208972334861755,
      "learning_rate": 0.000312431850016162,
      "loss": 3.3946,
      "step": 44600
    },
    {
      "epoch": 4.805725971370143,
      "grad_norm": 0.653488039970398,
      "learning_rate": 0.00031210860898610066,
      "loss": 3.4078,
      "step": 44650
    },
    {
      "epoch": 4.811107523409751,
      "grad_norm": 0.6298935413360596,
      "learning_rate": 0.0003117853679560392,
      "loss": 3.4058,
      "step": 44700
    },
    {
      "epoch": 4.8164890754493594,
      "grad_norm": 0.6644455790519714,
      "learning_rate": 0.0003114621269259778,
      "loss": 3.3979,
      "step": 44750
    },
    {
      "epoch": 4.821870627488968,
      "grad_norm": 0.6411975622177124,
      "learning_rate": 0.00031113888589591633,
      "loss": 3.3932,
      "step": 44800
    },
    {
      "epoch": 4.827252179528576,
      "grad_norm": 0.5985714793205261,
      "learning_rate": 0.00031081564486585493,
      "loss": 3.3851,
      "step": 44850
    },
    {
      "epoch": 4.832633731568184,
      "grad_norm": 0.6659331321716309,
      "learning_rate": 0.0003104924038357935,
      "loss": 3.3995,
      "step": 44900
    },
    {
      "epoch": 4.838015283607793,
      "grad_norm": 0.630987823009491,
      "learning_rate": 0.0003101691628057321,
      "loss": 3.3903,
      "step": 44950
    },
    {
      "epoch": 4.8433968356474,
      "grad_norm": 0.6505104303359985,
      "learning_rate": 0.0003098459217756707,
      "loss": 3.3906,
      "step": 45000
    },
    {
      "epoch": 4.8433968356474,
      "eval_accuracy": 0.379598095141032,
      "eval_loss": 3.4169139862060547,
      "eval_runtime": 184.0453,
      "eval_samples_per_second": 97.862,
      "eval_steps_per_second": 6.118,
      "step": 45000
    },
    {
      "epoch": 4.848778387687009,
      "grad_norm": 0.754757285118103,
      "learning_rate": 0.0003095226807456093,
      "loss": 3.3863,
      "step": 45050
    },
    {
      "epoch": 4.8541599397266175,
      "grad_norm": 0.60341477394104,
      "learning_rate": 0.00030919943971554785,
      "loss": 3.4033,
      "step": 45100
    },
    {
      "epoch": 4.859541491766225,
      "grad_norm": 0.6391721367835999,
      "learning_rate": 0.00030887619868548644,
      "loss": 3.4038,
      "step": 45150
    },
    {
      "epoch": 4.864923043805834,
      "grad_norm": 0.6096944808959961,
      "learning_rate": 0.000308552957655425,
      "loss": 3.3717,
      "step": 45200
    },
    {
      "epoch": 4.870304595845441,
      "grad_norm": 0.6128942370414734,
      "learning_rate": 0.00030822971662536363,
      "loss": 3.3914,
      "step": 45250
    },
    {
      "epoch": 4.87568614788505,
      "grad_norm": 0.6303298473358154,
      "learning_rate": 0.0003079064755953022,
      "loss": 3.3827,
      "step": 45300
    },
    {
      "epoch": 4.8810676999246585,
      "grad_norm": 0.6172555088996887,
      "learning_rate": 0.00030758323456524077,
      "loss": 3.3792,
      "step": 45350
    },
    {
      "epoch": 4.886449251964266,
      "grad_norm": 0.6756383776664734,
      "learning_rate": 0.00030725999353517936,
      "loss": 3.3868,
      "step": 45400
    },
    {
      "epoch": 4.891830804003875,
      "grad_norm": 0.5975400805473328,
      "learning_rate": 0.00030693675250511795,
      "loss": 3.3847,
      "step": 45450
    },
    {
      "epoch": 4.897212356043483,
      "grad_norm": 0.6471746563911438,
      "learning_rate": 0.00030661351147505655,
      "loss": 3.3794,
      "step": 45500
    },
    {
      "epoch": 4.902593908083091,
      "grad_norm": 0.6886560320854187,
      "learning_rate": 0.00030629027044499514,
      "loss": 3.402,
      "step": 45550
    },
    {
      "epoch": 4.9079754601226995,
      "grad_norm": 0.6070012450218201,
      "learning_rate": 0.00030596702941493374,
      "loss": 3.3874,
      "step": 45600
    },
    {
      "epoch": 4.913357012162308,
      "grad_norm": 0.7431271076202393,
      "learning_rate": 0.0003056437883848723,
      "loss": 3.3905,
      "step": 45650
    },
    {
      "epoch": 4.918738564201916,
      "grad_norm": 0.698964536190033,
      "learning_rate": 0.0003053270121754121,
      "loss": 3.3852,
      "step": 45700
    },
    {
      "epoch": 4.924120116241524,
      "grad_norm": 0.6414524912834167,
      "learning_rate": 0.0003050037711453507,
      "loss": 3.4026,
      "step": 45750
    },
    {
      "epoch": 4.929501668281132,
      "grad_norm": 0.6788366436958313,
      "learning_rate": 0.0003046805301152893,
      "loss": 3.3862,
      "step": 45800
    },
    {
      "epoch": 4.9348832203207404,
      "grad_norm": 0.655552327632904,
      "learning_rate": 0.0003043572890852278,
      "loss": 3.3768,
      "step": 45850
    },
    {
      "epoch": 4.940264772360349,
      "grad_norm": 0.6916018724441528,
      "learning_rate": 0.00030403404805516646,
      "loss": 3.394,
      "step": 45900
    },
    {
      "epoch": 4.945646324399957,
      "grad_norm": 0.6448120474815369,
      "learning_rate": 0.00030371080702510506,
      "loss": 3.3786,
      "step": 45950
    },
    {
      "epoch": 4.951027876439565,
      "grad_norm": 0.6547817587852478,
      "learning_rate": 0.0003033875659950436,
      "loss": 3.3949,
      "step": 46000
    },
    {
      "epoch": 4.951027876439565,
      "eval_accuracy": 0.380340737629244,
      "eval_loss": 3.410447120666504,
      "eval_runtime": 184.293,
      "eval_samples_per_second": 97.73,
      "eval_steps_per_second": 6.11,
      "step": 46000
    },
    {
      "epoch": 4.956409428479174,
      "grad_norm": 0.6945672631263733,
      "learning_rate": 0.0003030643249649822,
      "loss": 3.4012,
      "step": 46050
    },
    {
      "epoch": 4.961790980518781,
      "grad_norm": 0.6346149444580078,
      "learning_rate": 0.00030274108393492073,
      "loss": 3.4051,
      "step": 46100
    },
    {
      "epoch": 4.96717253255839,
      "grad_norm": 0.6204469203948975,
      "learning_rate": 0.00030241784290485933,
      "loss": 3.3927,
      "step": 46150
    },
    {
      "epoch": 4.9725540845979985,
      "grad_norm": 0.7108182907104492,
      "learning_rate": 0.000302094601874798,
      "loss": 3.3966,
      "step": 46200
    },
    {
      "epoch": 4.977935636637606,
      "grad_norm": 0.6828261613845825,
      "learning_rate": 0.0003017713608447365,
      "loss": 3.3846,
      "step": 46250
    },
    {
      "epoch": 4.983317188677215,
      "grad_norm": 0.6654807925224304,
      "learning_rate": 0.0003014481198146751,
      "loss": 3.405,
      "step": 46300
    },
    {
      "epoch": 4.988698740716822,
      "grad_norm": 0.6794192790985107,
      "learning_rate": 0.0003011248787846137,
      "loss": 3.3889,
      "step": 46350
    },
    {
      "epoch": 4.994080292756431,
      "grad_norm": 0.7037332057952881,
      "learning_rate": 0.00030080163775455225,
      "loss": 3.4116,
      "step": 46400
    },
    {
      "epoch": 4.9994618447960395,
      "grad_norm": 0.6396273970603943,
      "learning_rate": 0.0003004783967244909,
      "loss": 3.3804,
      "step": 46450
    },
    {
      "epoch": 5.004843396835647,
      "grad_norm": 0.6637768745422363,
      "learning_rate": 0.0003001551556944295,
      "loss": 3.2996,
      "step": 46500
    },
    {
      "epoch": 5.010224948875256,
      "grad_norm": 0.6840579509735107,
      "learning_rate": 0.00029983191466436803,
      "loss": 3.3067,
      "step": 46550
    },
    {
      "epoch": 5.015606500914864,
      "grad_norm": 0.6343075633049011,
      "learning_rate": 0.0002995086736343066,
      "loss": 3.2992,
      "step": 46600
    },
    {
      "epoch": 5.020988052954472,
      "grad_norm": 0.6339579224586487,
      "learning_rate": 0.00029918543260424517,
      "loss": 3.3019,
      "step": 46650
    },
    {
      "epoch": 5.0263696049940805,
      "grad_norm": 0.6617447137832642,
      "learning_rate": 0.0002988621915741838,
      "loss": 3.2996,
      "step": 46700
    },
    {
      "epoch": 5.031751157033688,
      "grad_norm": 0.6230089068412781,
      "learning_rate": 0.00029853895054412236,
      "loss": 3.2893,
      "step": 46750
    },
    {
      "epoch": 5.037132709073297,
      "grad_norm": 0.6579585075378418,
      "learning_rate": 0.00029821570951406095,
      "loss": 3.3042,
      "step": 46800
    },
    {
      "epoch": 5.042514261112905,
      "grad_norm": 0.6669675707817078,
      "learning_rate": 0.00029789246848399955,
      "loss": 3.3109,
      "step": 46850
    },
    {
      "epoch": 5.047895813152513,
      "grad_norm": 0.6590904593467712,
      "learning_rate": 0.00029756922745393814,
      "loss": 3.2972,
      "step": 46900
    },
    {
      "epoch": 5.0532773651921215,
      "grad_norm": 0.6489461660385132,
      "learning_rate": 0.00029724598642387674,
      "loss": 3.3187,
      "step": 46950
    },
    {
      "epoch": 5.05865891723173,
      "grad_norm": 0.6401969194412231,
      "learning_rate": 0.0002969227453938153,
      "loss": 3.3016,
      "step": 47000
    },
    {
      "epoch": 5.05865891723173,
      "eval_accuracy": 0.38062388705488853,
      "eval_loss": 3.4139175415039062,
      "eval_runtime": 184.748,
      "eval_samples_per_second": 97.49,
      "eval_steps_per_second": 6.095,
      "step": 47000
    },
    {
      "epoch": 5.064040469271338,
      "grad_norm": 0.7038257718086243,
      "learning_rate": 0.00029659950436375387,
      "loss": 3.3118,
      "step": 47050
    },
    {
      "epoch": 5.069422021310946,
      "grad_norm": 0.6294475793838501,
      "learning_rate": 0.00029627626333369246,
      "loss": 3.302,
      "step": 47100
    },
    {
      "epoch": 5.074803573350554,
      "grad_norm": 0.6852615475654602,
      "learning_rate": 0.00029595302230363106,
      "loss": 3.3119,
      "step": 47150
    },
    {
      "epoch": 5.080185125390162,
      "grad_norm": 0.6621981263160706,
      "learning_rate": 0.0002956297812735696,
      "loss": 3.3019,
      "step": 47200
    },
    {
      "epoch": 5.085566677429771,
      "grad_norm": 0.6848009824752808,
      "learning_rate": 0.00029530654024350825,
      "loss": 3.3175,
      "step": 47250
    },
    {
      "epoch": 5.090948229469379,
      "grad_norm": 0.6857882142066956,
      "learning_rate": 0.0002949832992134468,
      "loss": 3.2997,
      "step": 47300
    },
    {
      "epoch": 5.096329781508987,
      "grad_norm": 0.6457169651985168,
      "learning_rate": 0.0002946600581833854,
      "loss": 3.2997,
      "step": 47350
    },
    {
      "epoch": 5.101711333548596,
      "grad_norm": 0.6649097800254822,
      "learning_rate": 0.000294336817153324,
      "loss": 3.3322,
      "step": 47400
    },
    {
      "epoch": 5.107092885588203,
      "grad_norm": 0.7111925482749939,
      "learning_rate": 0.0002940135761232626,
      "loss": 3.3109,
      "step": 47450
    },
    {
      "epoch": 5.112474437627812,
      "grad_norm": 0.6585880517959595,
      "learning_rate": 0.0002936903350932011,
      "loss": 3.3306,
      "step": 47500
    },
    {
      "epoch": 5.1178559896674205,
      "grad_norm": 0.6791628003120422,
      "learning_rate": 0.0002933670940631397,
      "loss": 3.3159,
      "step": 47550
    },
    {
      "epoch": 5.123237541707028,
      "grad_norm": 0.656025230884552,
      "learning_rate": 0.0002930438530330783,
      "loss": 3.302,
      "step": 47600
    },
    {
      "epoch": 5.128619093746637,
      "grad_norm": 0.6528550982475281,
      "learning_rate": 0.0002927206120030169,
      "loss": 3.3108,
      "step": 47650
    },
    {
      "epoch": 5.134000645786244,
      "grad_norm": 0.6754864454269409,
      "learning_rate": 0.0002923973709729555,
      "loss": 3.3212,
      "step": 47700
    },
    {
      "epoch": 5.139382197825853,
      "grad_norm": 0.6551601886749268,
      "learning_rate": 0.00029207412994289403,
      "loss": 3.3272,
      "step": 47750
    },
    {
      "epoch": 5.1447637498654615,
      "grad_norm": 0.6972602605819702,
      "learning_rate": 0.0002917508889128327,
      "loss": 3.3239,
      "step": 47800
    },
    {
      "epoch": 5.150145301905069,
      "grad_norm": 0.6529858112335205,
      "learning_rate": 0.0002914276478827712,
      "loss": 3.3298,
      "step": 47850
    },
    {
      "epoch": 5.155526853944678,
      "grad_norm": 0.6726899743080139,
      "learning_rate": 0.0002911044068527098,
      "loss": 3.3339,
      "step": 47900
    },
    {
      "epoch": 5.160908405984286,
      "grad_norm": 0.6551409959793091,
      "learning_rate": 0.0002907811658226484,
      "loss": 3.3214,
      "step": 47950
    },
    {
      "epoch": 5.166289958023894,
      "grad_norm": 0.6539642810821533,
      "learning_rate": 0.000290457924792587,
      "loss": 3.3206,
      "step": 48000
    },
    {
      "epoch": 5.166289958023894,
      "eval_accuracy": 0.38074025429773783,
      "eval_loss": 3.411975860595703,
      "eval_runtime": 184.2469,
      "eval_samples_per_second": 97.755,
      "eval_steps_per_second": 6.111,
      "step": 48000
    },
    {
      "epoch": 5.1716715100635025,
      "grad_norm": 0.6480507850646973,
      "learning_rate": 0.00029013468376252555,
      "loss": 3.3357,
      "step": 48050
    },
    {
      "epoch": 5.17705306210311,
      "grad_norm": 0.6803812980651855,
      "learning_rate": 0.00028981144273246414,
      "loss": 3.328,
      "step": 48100
    },
    {
      "epoch": 5.182434614142719,
      "grad_norm": 0.6463451385498047,
      "learning_rate": 0.00028948820170240274,
      "loss": 3.3254,
      "step": 48150
    },
    {
      "epoch": 5.187816166182327,
      "grad_norm": 0.6713252067565918,
      "learning_rate": 0.00028916496067234133,
      "loss": 3.3277,
      "step": 48200
    },
    {
      "epoch": 5.193197718221935,
      "grad_norm": 0.6666660308837891,
      "learning_rate": 0.0002888417196422799,
      "loss": 3.3248,
      "step": 48250
    },
    {
      "epoch": 5.198579270261543,
      "grad_norm": 0.647766649723053,
      "learning_rate": 0.00028851847861221847,
      "loss": 3.3298,
      "step": 48300
    },
    {
      "epoch": 5.203960822301152,
      "grad_norm": 0.6583616137504578,
      "learning_rate": 0.00028819523758215706,
      "loss": 3.3164,
      "step": 48350
    },
    {
      "epoch": 5.20934237434076,
      "grad_norm": 0.6577174067497253,
      "learning_rate": 0.00028787199655209566,
      "loss": 3.3342,
      "step": 48400
    },
    {
      "epoch": 5.214723926380368,
      "grad_norm": 0.6606809496879578,
      "learning_rate": 0.00028754875552203425,
      "loss": 3.3371,
      "step": 48450
    },
    {
      "epoch": 5.220105478419977,
      "grad_norm": 0.6682845950126648,
      "learning_rate": 0.0002872255144919728,
      "loss": 3.324,
      "step": 48500
    },
    {
      "epoch": 5.225487030459584,
      "grad_norm": 0.6789083480834961,
      "learning_rate": 0.00028690227346191144,
      "loss": 3.3225,
      "step": 48550
    },
    {
      "epoch": 5.230868582499193,
      "grad_norm": 0.6483572125434875,
      "learning_rate": 0.00028657903243185,
      "loss": 3.3331,
      "step": 48600
    },
    {
      "epoch": 5.236250134538801,
      "grad_norm": 0.6880369782447815,
      "learning_rate": 0.0002862557914017886,
      "loss": 3.3202,
      "step": 48650
    },
    {
      "epoch": 5.241631686578409,
      "grad_norm": 0.6623727679252625,
      "learning_rate": 0.00028593255037172717,
      "loss": 3.3419,
      "step": 48700
    },
    {
      "epoch": 5.247013238618018,
      "grad_norm": 0.699100911617279,
      "learning_rate": 0.00028560930934166576,
      "loss": 3.3384,
      "step": 48750
    },
    {
      "epoch": 5.252394790657625,
      "grad_norm": 0.6639148592948914,
      "learning_rate": 0.00028528606831160436,
      "loss": 3.3473,
      "step": 48800
    },
    {
      "epoch": 5.257776342697234,
      "grad_norm": 0.7329900860786438,
      "learning_rate": 0.0002849628272815429,
      "loss": 3.3251,
      "step": 48850
    },
    {
      "epoch": 5.2631578947368425,
      "grad_norm": 0.6515591144561768,
      "learning_rate": 0.0002846395862514815,
      "loss": 3.3292,
      "step": 48900
    },
    {
      "epoch": 5.26853944677645,
      "grad_norm": 0.6936444044113159,
      "learning_rate": 0.0002843163452214201,
      "loss": 3.3397,
      "step": 48950
    },
    {
      "epoch": 5.273920998816059,
      "grad_norm": 0.7181999683380127,
      "learning_rate": 0.0002839931041913587,
      "loss": 3.3362,
      "step": 49000
    },
    {
      "epoch": 5.273920998816059,
      "eval_accuracy": 0.38123223457376176,
      "eval_loss": 3.4094080924987793,
      "eval_runtime": 184.3555,
      "eval_samples_per_second": 97.697,
      "eval_steps_per_second": 6.108,
      "step": 49000
    },
    {
      "epoch": 5.279302550855666,
      "grad_norm": 0.6783355474472046,
      "learning_rate": 0.0002836698631612972,
      "loss": 3.3526,
      "step": 49050
    },
    {
      "epoch": 5.284684102895275,
      "grad_norm": 0.694335401058197,
      "learning_rate": 0.00028334662213123587,
      "loss": 3.3189,
      "step": 49100
    },
    {
      "epoch": 5.2900656549348835,
      "grad_norm": 0.6517884731292725,
      "learning_rate": 0.0002830233811011744,
      "loss": 3.3359,
      "step": 49150
    },
    {
      "epoch": 5.295447206974491,
      "grad_norm": 0.6741759777069092,
      "learning_rate": 0.000282700140071113,
      "loss": 3.3493,
      "step": 49200
    },
    {
      "epoch": 5.3008287590141,
      "grad_norm": 0.6531067490577698,
      "learning_rate": 0.0002823768990410516,
      "loss": 3.3366,
      "step": 49250
    },
    {
      "epoch": 5.306210311053708,
      "grad_norm": 0.6721645593643188,
      "learning_rate": 0.00028205365801099014,
      "loss": 3.3501,
      "step": 49300
    },
    {
      "epoch": 5.311591863093316,
      "grad_norm": 0.6623327136039734,
      "learning_rate": 0.00028173041698092874,
      "loss": 3.3131,
      "step": 49350
    },
    {
      "epoch": 5.316973415132924,
      "grad_norm": 0.773777425289154,
      "learning_rate": 0.00028140717595086733,
      "loss": 3.3408,
      "step": 49400
    },
    {
      "epoch": 5.322354967172533,
      "grad_norm": 0.6638360619544983,
      "learning_rate": 0.0002810839349208059,
      "loss": 3.3342,
      "step": 49450
    },
    {
      "epoch": 5.327736519212141,
      "grad_norm": 0.6455284953117371,
      "learning_rate": 0.00028076069389074447,
      "loss": 3.3309,
      "step": 49500
    },
    {
      "epoch": 5.333118071251749,
      "grad_norm": 0.6641452312469482,
      "learning_rate": 0.0002804374528606831,
      "loss": 3.3207,
      "step": 49550
    },
    {
      "epoch": 5.338499623291357,
      "grad_norm": 0.6540976166725159,
      "learning_rate": 0.00028011421183062166,
      "loss": 3.3308,
      "step": 49600
    },
    {
      "epoch": 5.343881175330965,
      "grad_norm": 0.6941928863525391,
      "learning_rate": 0.00027979097080056025,
      "loss": 3.3265,
      "step": 49650
    },
    {
      "epoch": 5.349262727370574,
      "grad_norm": 0.6998748779296875,
      "learning_rate": 0.00027946772977049885,
      "loss": 3.3356,
      "step": 49700
    },
    {
      "epoch": 5.354644279410182,
      "grad_norm": 0.671293318271637,
      "learning_rate": 0.00027915095356103865,
      "loss": 3.3312,
      "step": 49750
    },
    {
      "epoch": 5.36002583144979,
      "grad_norm": 0.6354805827140808,
      "learning_rate": 0.00027882771253097725,
      "loss": 3.3336,
      "step": 49800
    },
    {
      "epoch": 5.365407383489399,
      "grad_norm": 0.6770129203796387,
      "learning_rate": 0.00027850447150091584,
      "loss": 3.3313,
      "step": 49850
    },
    {
      "epoch": 5.370788935529006,
      "grad_norm": 0.6457779407501221,
      "learning_rate": 0.00027818769529145565,
      "loss": 3.3468,
      "step": 49900
    },
    {
      "epoch": 5.376170487568615,
      "grad_norm": 0.6751704216003418,
      "learning_rate": 0.0002778644542613942,
      "loss": 3.3508,
      "step": 49950
    },
    {
      "epoch": 5.3815520396082235,
      "grad_norm": 0.6823959350585938,
      "learning_rate": 0.00027754121323133284,
      "loss": 3.3247,
      "step": 50000
    },
    {
      "epoch": 5.3815520396082235,
      "eval_accuracy": 0.3813353461642977,
      "eval_loss": 3.40474796295166,
      "eval_runtime": 184.7208,
      "eval_samples_per_second": 97.504,
      "eval_steps_per_second": 6.096,
      "step": 50000
    },
    {
      "epoch": 5.386933591647831,
      "grad_norm": 0.7053428292274475,
      "learning_rate": 0.0002772179722012714,
      "loss": 3.3318,
      "step": 50050
    },
    {
      "epoch": 5.39231514368744,
      "grad_norm": 0.6931941509246826,
      "learning_rate": 0.00027689473117120997,
      "loss": 3.3369,
      "step": 50100
    },
    {
      "epoch": 5.397696695727047,
      "grad_norm": 0.6617245078086853,
      "learning_rate": 0.00027657149014114857,
      "loss": 3.3364,
      "step": 50150
    },
    {
      "epoch": 5.403078247766656,
      "grad_norm": 0.6629908084869385,
      "learning_rate": 0.00027624824911108716,
      "loss": 3.3274,
      "step": 50200
    },
    {
      "epoch": 5.4084597998062645,
      "grad_norm": 0.6928619742393494,
      "learning_rate": 0.0002759250080810257,
      "loss": 3.3372,
      "step": 50250
    },
    {
      "epoch": 5.413841351845872,
      "grad_norm": 0.726039469242096,
      "learning_rate": 0.0002756017670509643,
      "loss": 3.3426,
      "step": 50300
    },
    {
      "epoch": 5.419222903885481,
      "grad_norm": 0.7072526216506958,
      "learning_rate": 0.0002752785260209029,
      "loss": 3.339,
      "step": 50350
    },
    {
      "epoch": 5.424604455925088,
      "grad_norm": 0.669489860534668,
      "learning_rate": 0.0002749552849908415,
      "loss": 3.3274,
      "step": 50400
    },
    {
      "epoch": 5.429986007964697,
      "grad_norm": 0.696245014667511,
      "learning_rate": 0.0002746320439607801,
      "loss": 3.3314,
      "step": 50450
    },
    {
      "epoch": 5.435367560004305,
      "grad_norm": 0.666018545627594,
      "learning_rate": 0.0002743088029307186,
      "loss": 3.3554,
      "step": 50500
    },
    {
      "epoch": 5.440749112043913,
      "grad_norm": 0.6564254760742188,
      "learning_rate": 0.00027398556190065727,
      "loss": 3.3346,
      "step": 50550
    },
    {
      "epoch": 5.446130664083522,
      "grad_norm": 0.6895231604576111,
      "learning_rate": 0.0002736623208705958,
      "loss": 3.3526,
      "step": 50600
    },
    {
      "epoch": 5.45151221612313,
      "grad_norm": 0.6966924071311951,
      "learning_rate": 0.0002733390798405344,
      "loss": 3.3312,
      "step": 50650
    },
    {
      "epoch": 5.456893768162738,
      "grad_norm": 0.6883962154388428,
      "learning_rate": 0.000273015838810473,
      "loss": 3.3321,
      "step": 50700
    },
    {
      "epoch": 5.462275320202346,
      "grad_norm": 0.6729317903518677,
      "learning_rate": 0.0002726925977804116,
      "loss": 3.3222,
      "step": 50750
    },
    {
      "epoch": 5.467656872241955,
      "grad_norm": 0.6859122514724731,
      "learning_rate": 0.00027236935675035013,
      "loss": 3.3516,
      "step": 50800
    },
    {
      "epoch": 5.473038424281563,
      "grad_norm": 0.6682018637657166,
      "learning_rate": 0.00027204611572028873,
      "loss": 3.323,
      "step": 50850
    },
    {
      "epoch": 5.478419976321171,
      "grad_norm": 0.6889828443527222,
      "learning_rate": 0.0002717228746902273,
      "loss": 3.3458,
      "step": 50900
    },
    {
      "epoch": 5.483801528360779,
      "grad_norm": 0.6683465838432312,
      "learning_rate": 0.0002713996336601659,
      "loss": 3.3255,
      "step": 50950
    },
    {
      "epoch": 5.489183080400387,
      "grad_norm": 0.6655813455581665,
      "learning_rate": 0.0002710763926301045,
      "loss": 3.3329,
      "step": 51000
    },
    {
      "epoch": 5.489183080400387,
      "eval_accuracy": 0.38184916567077243,
      "eval_loss": 3.4012391567230225,
      "eval_runtime": 184.3226,
      "eval_samples_per_second": 97.715,
      "eval_steps_per_second": 6.109,
      "step": 51000
    },
    {
      "epoch": 5.494564632439996,
      "grad_norm": 0.6571836471557617,
      "learning_rate": 0.00027075315160004305,
      "loss": 3.3396,
      "step": 51050
    },
    {
      "epoch": 5.499946184479604,
      "grad_norm": 0.6146655678749084,
      "learning_rate": 0.00027042991056998165,
      "loss": 3.3316,
      "step": 51100
    },
    {
      "epoch": 5.505327736519212,
      "grad_norm": 0.6827743649482727,
      "learning_rate": 0.00027010666953992024,
      "loss": 3.3393,
      "step": 51150
    },
    {
      "epoch": 5.510709288558821,
      "grad_norm": 0.6813520789146423,
      "learning_rate": 0.00026978342850985884,
      "loss": 3.343,
      "step": 51200
    },
    {
      "epoch": 5.516090840598428,
      "grad_norm": 0.6529282331466675,
      "learning_rate": 0.0002694601874797974,
      "loss": 3.337,
      "step": 51250
    },
    {
      "epoch": 5.521472392638037,
      "grad_norm": 0.733112633228302,
      "learning_rate": 0.000269136946449736,
      "loss": 3.3409,
      "step": 51300
    },
    {
      "epoch": 5.5268539446776455,
      "grad_norm": 0.6543546915054321,
      "learning_rate": 0.00026881370541967457,
      "loss": 3.3307,
      "step": 51350
    },
    {
      "epoch": 5.532235496717253,
      "grad_norm": 0.6887438893318176,
      "learning_rate": 0.00026849046438961316,
      "loss": 3.3391,
      "step": 51400
    },
    {
      "epoch": 5.537617048756862,
      "grad_norm": 0.6722603440284729,
      "learning_rate": 0.00026816722335955176,
      "loss": 3.3356,
      "step": 51450
    },
    {
      "epoch": 5.542998600796469,
      "grad_norm": 0.6176708340644836,
      "learning_rate": 0.00026784398232949035,
      "loss": 3.3292,
      "step": 51500
    },
    {
      "epoch": 5.548380152836078,
      "grad_norm": 0.6900912523269653,
      "learning_rate": 0.00026752074129942895,
      "loss": 3.333,
      "step": 51550
    },
    {
      "epoch": 5.553761704875686,
      "grad_norm": 0.6852001547813416,
      "learning_rate": 0.0002671975002693675,
      "loss": 3.3251,
      "step": 51600
    },
    {
      "epoch": 5.559143256915294,
      "grad_norm": 0.6884951591491699,
      "learning_rate": 0.0002668742592393061,
      "loss": 3.351,
      "step": 51650
    },
    {
      "epoch": 5.564524808954903,
      "grad_norm": 0.6931823492050171,
      "learning_rate": 0.0002665510182092447,
      "loss": 3.3422,
      "step": 51700
    },
    {
      "epoch": 5.569906360994511,
      "grad_norm": 0.6956307888031006,
      "learning_rate": 0.00026622777717918327,
      "loss": 3.3231,
      "step": 51750
    },
    {
      "epoch": 5.575287913034119,
      "grad_norm": 0.6652266383171082,
      "learning_rate": 0.0002659045361491218,
      "loss": 3.3278,
      "step": 51800
    },
    {
      "epoch": 5.580669465073727,
      "grad_norm": 0.6874408721923828,
      "learning_rate": 0.00026558129511906046,
      "loss": 3.343,
      "step": 51850
    },
    {
      "epoch": 5.586051017113336,
      "grad_norm": 0.657923698425293,
      "learning_rate": 0.000265258054088999,
      "loss": 3.3353,
      "step": 51900
    },
    {
      "epoch": 5.591432569152944,
      "grad_norm": 0.6838491559028625,
      "learning_rate": 0.0002649348130589376,
      "loss": 3.3348,
      "step": 51950
    },
    {
      "epoch": 5.596814121192552,
      "grad_norm": 0.6994488835334778,
      "learning_rate": 0.0002646115720288762,
      "loss": 3.3394,
      "step": 52000
    },
    {
      "epoch": 5.596814121192552,
      "eval_accuracy": 0.3824370864467366,
      "eval_loss": 3.3962817192077637,
      "eval_runtime": 184.5442,
      "eval_samples_per_second": 97.597,
      "eval_steps_per_second": 6.102,
      "step": 52000
    },
    {
      "epoch": 5.60219567323216,
      "grad_norm": 0.6435365676879883,
      "learning_rate": 0.00026428833099881473,
      "loss": 3.3386,
      "step": 52050
    },
    {
      "epoch": 5.607577225271768,
      "grad_norm": 0.6442993879318237,
      "learning_rate": 0.0002639650899687533,
      "loss": 3.3266,
      "step": 52100
    },
    {
      "epoch": 5.612958777311377,
      "grad_norm": 0.6546722054481506,
      "learning_rate": 0.00026364831375929313,
      "loss": 3.3448,
      "step": 52150
    },
    {
      "epoch": 5.618340329350985,
      "grad_norm": 0.6626961827278137,
      "learning_rate": 0.0002633250727292318,
      "loss": 3.3235,
      "step": 52200
    },
    {
      "epoch": 5.623721881390593,
      "grad_norm": 0.7044293880462646,
      "learning_rate": 0.0002630018316991703,
      "loss": 3.3505,
      "step": 52250
    },
    {
      "epoch": 5.629103433430201,
      "grad_norm": 0.7244980931282043,
      "learning_rate": 0.0002626785906691089,
      "loss": 3.3387,
      "step": 52300
    },
    {
      "epoch": 5.634484985469809,
      "grad_norm": 0.7230950593948364,
      "learning_rate": 0.0002623553496390475,
      "loss": 3.3349,
      "step": 52350
    },
    {
      "epoch": 5.639866537509418,
      "grad_norm": 0.6979284286499023,
      "learning_rate": 0.0002620321086089861,
      "loss": 3.3402,
      "step": 52400
    },
    {
      "epoch": 5.645248089549026,
      "grad_norm": 0.6802158355712891,
      "learning_rate": 0.00026170886757892464,
      "loss": 3.3368,
      "step": 52450
    },
    {
      "epoch": 5.650629641588634,
      "grad_norm": 0.7164120674133301,
      "learning_rate": 0.00026138562654886324,
      "loss": 3.34,
      "step": 52500
    },
    {
      "epoch": 5.656011193628243,
      "grad_norm": 0.6808815002441406,
      "learning_rate": 0.00026106238551880183,
      "loss": 3.3578,
      "step": 52550
    },
    {
      "epoch": 5.66139274566785,
      "grad_norm": 0.7645325064659119,
      "learning_rate": 0.00026073914448874043,
      "loss": 3.3508,
      "step": 52600
    },
    {
      "epoch": 5.666774297707459,
      "grad_norm": 0.6486349701881409,
      "learning_rate": 0.000260415903458679,
      "loss": 3.3448,
      "step": 52650
    },
    {
      "epoch": 5.672155849747067,
      "grad_norm": 0.6709673404693604,
      "learning_rate": 0.00026009266242861756,
      "loss": 3.3407,
      "step": 52700
    },
    {
      "epoch": 5.677537401786675,
      "grad_norm": 0.6882966160774231,
      "learning_rate": 0.00025976942139855616,
      "loss": 3.3255,
      "step": 52750
    },
    {
      "epoch": 5.682918953826284,
      "grad_norm": 0.7009903788566589,
      "learning_rate": 0.00025944618036849475,
      "loss": 3.3511,
      "step": 52800
    },
    {
      "epoch": 5.688300505865891,
      "grad_norm": 0.6677101254463196,
      "learning_rate": 0.00025912293933843335,
      "loss": 3.3314,
      "step": 52850
    },
    {
      "epoch": 5.6936820579055,
      "grad_norm": 0.6308375597000122,
      "learning_rate": 0.0002587996983083719,
      "loss": 3.3381,
      "step": 52900
    },
    {
      "epoch": 5.699063609945108,
      "grad_norm": 0.6462803483009338,
      "learning_rate": 0.00025847645727831054,
      "loss": 3.3441,
      "step": 52950
    },
    {
      "epoch": 5.704445161984716,
      "grad_norm": 0.6592354774475098,
      "learning_rate": 0.0002581532162482491,
      "loss": 3.3535,
      "step": 53000
    },
    {
      "epoch": 5.704445161984716,
      "eval_accuracy": 0.382956881862041,
      "eval_loss": 3.391707181930542,
      "eval_runtime": 184.7719,
      "eval_samples_per_second": 97.477,
      "eval_steps_per_second": 6.094,
      "step": 53000
    },
    {
      "epoch": 5.709826714024325,
      "grad_norm": 0.6917129158973694,
      "learning_rate": 0.00025782997521818767,
      "loss": 3.3264,
      "step": 53050
    },
    {
      "epoch": 5.715208266063933,
      "grad_norm": 0.7247872948646545,
      "learning_rate": 0.00025750673418812627,
      "loss": 3.3521,
      "step": 53100
    },
    {
      "epoch": 5.720589818103541,
      "grad_norm": 0.6706468462944031,
      "learning_rate": 0.00025718349315806486,
      "loss": 3.3507,
      "step": 53150
    },
    {
      "epoch": 5.725971370143149,
      "grad_norm": 0.6482888460159302,
      "learning_rate": 0.00025686025212800346,
      "loss": 3.3457,
      "step": 53200
    },
    {
      "epoch": 5.731352922182758,
      "grad_norm": 0.6776821613311768,
      "learning_rate": 0.000256537011097942,
      "loss": 3.3573,
      "step": 53250
    },
    {
      "epoch": 5.736734474222366,
      "grad_norm": 0.6718127131462097,
      "learning_rate": 0.0002562137700678806,
      "loss": 3.3315,
      "step": 53300
    },
    {
      "epoch": 5.742116026261974,
      "grad_norm": 0.6800867915153503,
      "learning_rate": 0.0002558905290378192,
      "loss": 3.3415,
      "step": 53350
    },
    {
      "epoch": 5.747497578301582,
      "grad_norm": 0.7636213898658752,
      "learning_rate": 0.0002555672880077578,
      "loss": 3.342,
      "step": 53400
    },
    {
      "epoch": 5.75287913034119,
      "grad_norm": 0.659597635269165,
      "learning_rate": 0.0002552440469776963,
      "loss": 3.3375,
      "step": 53450
    },
    {
      "epoch": 5.758260682380799,
      "grad_norm": 0.6465020179748535,
      "learning_rate": 0.00025492080594763497,
      "loss": 3.3356,
      "step": 53500
    },
    {
      "epoch": 5.763642234420407,
      "grad_norm": 0.6734376549720764,
      "learning_rate": 0.0002545975649175735,
      "loss": 3.3197,
      "step": 53550
    },
    {
      "epoch": 5.769023786460015,
      "grad_norm": 0.6750368475914001,
      "learning_rate": 0.0002542743238875121,
      "loss": 3.3442,
      "step": 53600
    },
    {
      "epoch": 5.774405338499624,
      "grad_norm": 0.7284238338470459,
      "learning_rate": 0.0002539510828574507,
      "loss": 3.3184,
      "step": 53650
    },
    {
      "epoch": 5.779786890539231,
      "grad_norm": 0.6970507502555847,
      "learning_rate": 0.00025362784182738924,
      "loss": 3.3364,
      "step": 53700
    },
    {
      "epoch": 5.78516844257884,
      "grad_norm": 0.6999320983886719,
      "learning_rate": 0.00025330460079732783,
      "loss": 3.3638,
      "step": 53750
    },
    {
      "epoch": 5.790549994618448,
      "grad_norm": 0.6974489688873291,
      "learning_rate": 0.00025298135976726643,
      "loss": 3.3404,
      "step": 53800
    },
    {
      "epoch": 5.795931546658056,
      "grad_norm": 0.6947113275527954,
      "learning_rate": 0.000252658118737205,
      "loss": 3.3508,
      "step": 53850
    },
    {
      "epoch": 5.801313098697665,
      "grad_norm": 0.7016981840133667,
      "learning_rate": 0.00025233487770714356,
      "loss": 3.3414,
      "step": 53900
    },
    {
      "epoch": 5.806694650737272,
      "grad_norm": 0.7312738299369812,
      "learning_rate": 0.0002520116366770822,
      "loss": 3.3386,
      "step": 53950
    },
    {
      "epoch": 5.812076202776881,
      "grad_norm": 0.7036862373352051,
      "learning_rate": 0.00025168839564702075,
      "loss": 3.3201,
      "step": 54000
    },
    {
      "epoch": 5.812076202776881,
      "eval_accuracy": 0.38318222795136825,
      "eval_loss": 3.3879685401916504,
      "eval_runtime": 184.3491,
      "eval_samples_per_second": 97.7,
      "eval_steps_per_second": 6.108,
      "step": 54000
    },
    {
      "epoch": 5.817457754816489,
      "grad_norm": 0.67717045545578,
      "learning_rate": 0.00025136515461695935,
      "loss": 3.334,
      "step": 54050
    },
    {
      "epoch": 5.822839306856097,
      "grad_norm": 0.6660524010658264,
      "learning_rate": 0.00025104191358689794,
      "loss": 3.3369,
      "step": 54100
    },
    {
      "epoch": 5.828220858895706,
      "grad_norm": 0.735390305519104,
      "learning_rate": 0.00025071867255683654,
      "loss": 3.3326,
      "step": 54150
    },
    {
      "epoch": 5.833602410935313,
      "grad_norm": 0.7380104064941406,
      "learning_rate": 0.00025039543152677513,
      "loss": 3.3429,
      "step": 54200
    },
    {
      "epoch": 5.838983962974922,
      "grad_norm": 0.7150477170944214,
      "learning_rate": 0.0002500721904967137,
      "loss": 3.3174,
      "step": 54250
    },
    {
      "epoch": 5.84436551501453,
      "grad_norm": 0.6599828600883484,
      "learning_rate": 0.00024974894946665227,
      "loss": 3.3273,
      "step": 54300
    },
    {
      "epoch": 5.849747067054138,
      "grad_norm": 0.7460449934005737,
      "learning_rate": 0.00024942570843659086,
      "loss": 3.3255,
      "step": 54350
    },
    {
      "epoch": 5.855128619093747,
      "grad_norm": 0.6856649518013,
      "learning_rate": 0.00024910246740652946,
      "loss": 3.3505,
      "step": 54400
    },
    {
      "epoch": 5.860510171133355,
      "grad_norm": 0.7612301111221313,
      "learning_rate": 0.000248779226376468,
      "loss": 3.3358,
      "step": 54450
    },
    {
      "epoch": 5.865891723172963,
      "grad_norm": 0.7577794790267944,
      "learning_rate": 0.00024845598534640665,
      "loss": 3.3429,
      "step": 54500
    },
    {
      "epoch": 5.871273275212571,
      "grad_norm": 0.7047007083892822,
      "learning_rate": 0.0002481327443163452,
      "loss": 3.3431,
      "step": 54550
    },
    {
      "epoch": 5.87665482725218,
      "grad_norm": 0.7000530362129211,
      "learning_rate": 0.0002478095032862838,
      "loss": 3.3237,
      "step": 54600
    },
    {
      "epoch": 5.882036379291788,
      "grad_norm": 0.6977667808532715,
      "learning_rate": 0.0002474862622562224,
      "loss": 3.3564,
      "step": 54650
    },
    {
      "epoch": 5.887417931331396,
      "grad_norm": 0.7015401721000671,
      "learning_rate": 0.00024716302122616097,
      "loss": 3.3457,
      "step": 54700
    },
    {
      "epoch": 5.892799483371004,
      "grad_norm": 0.7044519782066345,
      "learning_rate": 0.0002468397801960995,
      "loss": 3.3315,
      "step": 54750
    },
    {
      "epoch": 5.898181035410612,
      "grad_norm": 0.6926525235176086,
      "learning_rate": 0.0002465165391660381,
      "loss": 3.3418,
      "step": 54800
    },
    {
      "epoch": 5.903562587450221,
      "grad_norm": 0.7459431290626526,
      "learning_rate": 0.0002461997629565779,
      "loss": 3.3326,
      "step": 54850
    },
    {
      "epoch": 5.9089441394898286,
      "grad_norm": 0.6971675157546997,
      "learning_rate": 0.0002458765219265165,
      "loss": 3.3521,
      "step": 54900
    },
    {
      "epoch": 5.914325691529437,
      "grad_norm": 0.6713802814483643,
      "learning_rate": 0.0002455532808964551,
      "loss": 3.3575,
      "step": 54950
    },
    {
      "epoch": 5.919707243569046,
      "grad_norm": 0.6946560144424438,
      "learning_rate": 0.0002452300398663937,
      "loss": 3.3292,
      "step": 55000
    },
    {
      "epoch": 5.919707243569046,
      "eval_accuracy": 0.38378536013162645,
      "eval_loss": 3.3823401927948,
      "eval_runtime": 184.6206,
      "eval_samples_per_second": 97.557,
      "eval_steps_per_second": 6.099,
      "step": 55000
    },
    {
      "epoch": 5.925088795608653,
      "grad_norm": 0.7011900544166565,
      "learning_rate": 0.0002449067988363323,
      "loss": 3.3305,
      "step": 55050
    },
    {
      "epoch": 5.930470347648262,
      "grad_norm": 0.6637669205665588,
      "learning_rate": 0.00024458355780627083,
      "loss": 3.3323,
      "step": 55100
    },
    {
      "epoch": 5.93585189968787,
      "grad_norm": 0.6542259454727173,
      "learning_rate": 0.0002442603167762094,
      "loss": 3.3464,
      "step": 55150
    },
    {
      "epoch": 5.941233451727478,
      "grad_norm": 0.6658049821853638,
      "learning_rate": 0.00024393707574614802,
      "loss": 3.3283,
      "step": 55200
    },
    {
      "epoch": 5.946615003767087,
      "grad_norm": 0.6897269487380981,
      "learning_rate": 0.0002436138347160866,
      "loss": 3.3443,
      "step": 55250
    },
    {
      "epoch": 5.951996555806694,
      "grad_norm": 0.6982568502426147,
      "learning_rate": 0.0002432905936860252,
      "loss": 3.3235,
      "step": 55300
    },
    {
      "epoch": 5.957378107846303,
      "grad_norm": 0.6752307415008545,
      "learning_rate": 0.00024296735265596378,
      "loss": 3.3331,
      "step": 55350
    },
    {
      "epoch": 5.962759659885911,
      "grad_norm": 0.6866647005081177,
      "learning_rate": 0.00024264411162590235,
      "loss": 3.3436,
      "step": 55400
    },
    {
      "epoch": 5.968141211925519,
      "grad_norm": 0.7104104161262512,
      "learning_rate": 0.00024232087059584097,
      "loss": 3.3421,
      "step": 55450
    },
    {
      "epoch": 5.973522763965128,
      "grad_norm": 0.6807641983032227,
      "learning_rate": 0.00024199762956577953,
      "loss": 3.3311,
      "step": 55500
    },
    {
      "epoch": 5.978904316004736,
      "grad_norm": 0.6734718084335327,
      "learning_rate": 0.0002416743885357181,
      "loss": 3.3311,
      "step": 55550
    },
    {
      "epoch": 5.984285868044344,
      "grad_norm": 0.703883945941925,
      "learning_rate": 0.0002413511475056567,
      "loss": 3.3366,
      "step": 55600
    },
    {
      "epoch": 5.989667420083952,
      "grad_norm": 0.675209641456604,
      "learning_rate": 0.0002410279064755953,
      "loss": 3.3378,
      "step": 55650
    },
    {
      "epoch": 5.995048972123561,
      "grad_norm": 0.7055371403694153,
      "learning_rate": 0.00024070466544553386,
      "loss": 3.3572,
      "step": 55700
    },
    {
      "epoch": 6.000430524163169,
      "grad_norm": 0.6902370452880859,
      "learning_rate": 0.00024038142441547245,
      "loss": 3.346,
      "step": 55750
    },
    {
      "epoch": 6.005812076202777,
      "grad_norm": 0.6972889304161072,
      "learning_rate": 0.00024005818338541102,
      "loss": 3.2616,
      "step": 55800
    },
    {
      "epoch": 6.011193628242385,
      "grad_norm": 0.7072308659553528,
      "learning_rate": 0.0002397349423553496,
      "loss": 3.2354,
      "step": 55850
    },
    {
      "epoch": 6.016575180281993,
      "grad_norm": 0.6985675692558289,
      "learning_rate": 0.0002394117013252882,
      "loss": 3.2471,
      "step": 55900
    },
    {
      "epoch": 6.021956732321602,
      "grad_norm": 0.6993648409843445,
      "learning_rate": 0.00023908846029522678,
      "loss": 3.2386,
      "step": 55950
    },
    {
      "epoch": 6.0273382843612096,
      "grad_norm": 0.707125723361969,
      "learning_rate": 0.00023876521926516537,
      "loss": 3.2594,
      "step": 56000
    },
    {
      "epoch": 6.0273382843612096,
      "eval_accuracy": 0.38364030852639414,
      "eval_loss": 3.3859715461730957,
      "eval_runtime": 184.6538,
      "eval_samples_per_second": 97.539,
      "eval_steps_per_second": 6.098,
      "step": 56000
    },
    {
      "epoch": 6.032719836400818,
      "grad_norm": 0.6912574172019958,
      "learning_rate": 0.00023844197823510397,
      "loss": 3.2554,
      "step": 56050
    },
    {
      "epoch": 6.038101388440427,
      "grad_norm": 0.6978053450584412,
      "learning_rate": 0.00023811873720504253,
      "loss": 3.2538,
      "step": 56100
    },
    {
      "epoch": 6.043482940480034,
      "grad_norm": 0.6996335387229919,
      "learning_rate": 0.00023779549617498113,
      "loss": 3.2647,
      "step": 56150
    },
    {
      "epoch": 6.048864492519643,
      "grad_norm": 0.719261646270752,
      "learning_rate": 0.0002374722551449197,
      "loss": 3.2531,
      "step": 56200
    },
    {
      "epoch": 6.0542460445592505,
      "grad_norm": 0.7113375663757324,
      "learning_rate": 0.0002371490141148583,
      "loss": 3.2639,
      "step": 56250
    },
    {
      "epoch": 6.059627596598859,
      "grad_norm": 0.7506604194641113,
      "learning_rate": 0.00023682577308479689,
      "loss": 3.2661,
      "step": 56300
    },
    {
      "epoch": 6.065009148638468,
      "grad_norm": 0.7061278820037842,
      "learning_rate": 0.00023650253205473545,
      "loss": 3.2758,
      "step": 56350
    },
    {
      "epoch": 6.070390700678075,
      "grad_norm": 0.7056647539138794,
      "learning_rate": 0.00023617929102467402,
      "loss": 3.2522,
      "step": 56400
    },
    {
      "epoch": 6.075772252717684,
      "grad_norm": 0.7045934796333313,
      "learning_rate": 0.00023585604999461264,
      "loss": 3.2555,
      "step": 56450
    },
    {
      "epoch": 6.081153804757292,
      "grad_norm": 0.7240189909934998,
      "learning_rate": 0.0002355328089645512,
      "loss": 3.2528,
      "step": 56500
    },
    {
      "epoch": 6.0865353567969,
      "grad_norm": 0.6941931247711182,
      "learning_rate": 0.00023520956793448978,
      "loss": 3.2755,
      "step": 56550
    },
    {
      "epoch": 6.091916908836509,
      "grad_norm": 0.6853485107421875,
      "learning_rate": 0.0002348863269044284,
      "loss": 3.2661,
      "step": 56600
    },
    {
      "epoch": 6.097298460876116,
      "grad_norm": 0.7292964458465576,
      "learning_rate": 0.00023456308587436697,
      "loss": 3.2766,
      "step": 56650
    },
    {
      "epoch": 6.102680012915725,
      "grad_norm": 0.7185317277908325,
      "learning_rate": 0.00023423984484430554,
      "loss": 3.2588,
      "step": 56700
    },
    {
      "epoch": 6.108061564955333,
      "grad_norm": 0.7048366665840149,
      "learning_rate": 0.00023391660381424413,
      "loss": 3.2755,
      "step": 56750
    },
    {
      "epoch": 6.113443116994941,
      "grad_norm": 0.7272534966468811,
      "learning_rate": 0.00023359336278418272,
      "loss": 3.2628,
      "step": 56800
    },
    {
      "epoch": 6.11882466903455,
      "grad_norm": 0.6993345618247986,
      "learning_rate": 0.0002332701217541213,
      "loss": 3.2733,
      "step": 56850
    },
    {
      "epoch": 6.124206221074158,
      "grad_norm": 0.7473379373550415,
      "learning_rate": 0.0002329468807240599,
      "loss": 3.2695,
      "step": 56900
    },
    {
      "epoch": 6.129587773113766,
      "grad_norm": 0.71280837059021,
      "learning_rate": 0.00023262363969399845,
      "loss": 3.2827,
      "step": 56950
    },
    {
      "epoch": 6.134969325153374,
      "grad_norm": 0.7106375694274902,
      "learning_rate": 0.00023230039866393708,
      "loss": 3.2761,
      "step": 57000
    },
    {
      "epoch": 6.134969325153374,
      "eval_accuracy": 0.38394910003356286,
      "eval_loss": 3.3853514194488525,
      "eval_runtime": 184.2581,
      "eval_samples_per_second": 97.749,
      "eval_steps_per_second": 6.111,
      "step": 57000
    },
    {
      "epoch": 6.140350877192983,
      "grad_norm": 0.6565351486206055,
      "learning_rate": 0.00023197715763387564,
      "loss": 3.2631,
      "step": 57050
    },
    {
      "epoch": 6.1457324292325906,
      "grad_norm": 0.738466203212738,
      "learning_rate": 0.0002316539166038142,
      "loss": 3.2477,
      "step": 57100
    },
    {
      "epoch": 6.151113981272199,
      "grad_norm": 0.7343447804450989,
      "learning_rate": 0.00023133067557375283,
      "loss": 3.2572,
      "step": 57150
    },
    {
      "epoch": 6.156495533311807,
      "grad_norm": 0.7094284892082214,
      "learning_rate": 0.0002310074345436914,
      "loss": 3.2594,
      "step": 57200
    },
    {
      "epoch": 6.161877085351415,
      "grad_norm": 0.7118187546730042,
      "learning_rate": 0.00023068419351362997,
      "loss": 3.2571,
      "step": 57250
    },
    {
      "epoch": 6.167258637391024,
      "grad_norm": 0.7504222393035889,
      "learning_rate": 0.00023036095248356856,
      "loss": 3.2626,
      "step": 57300
    },
    {
      "epoch": 6.1726401894306315,
      "grad_norm": 0.7145105004310608,
      "learning_rate": 0.00023003771145350716,
      "loss": 3.2675,
      "step": 57350
    },
    {
      "epoch": 6.17802174147024,
      "grad_norm": 0.7162860631942749,
      "learning_rate": 0.00022971447042344572,
      "loss": 3.2938,
      "step": 57400
    },
    {
      "epoch": 6.183403293509849,
      "grad_norm": 0.7450271248817444,
      "learning_rate": 0.00022939122939338432,
      "loss": 3.2844,
      "step": 57450
    },
    {
      "epoch": 6.188784845549456,
      "grad_norm": 0.7309553623199463,
      "learning_rate": 0.0002290679883633229,
      "loss": 3.2528,
      "step": 57500
    },
    {
      "epoch": 6.194166397589065,
      "grad_norm": 0.739514946937561,
      "learning_rate": 0.00022874474733326148,
      "loss": 3.2826,
      "step": 57550
    },
    {
      "epoch": 6.1995479496286725,
      "grad_norm": 0.73067706823349,
      "learning_rate": 0.00022842150630320008,
      "loss": 3.2763,
      "step": 57600
    },
    {
      "epoch": 6.204929501668281,
      "grad_norm": 0.6793087124824524,
      "learning_rate": 0.00022809826527313864,
      "loss": 3.2755,
      "step": 57650
    },
    {
      "epoch": 6.21031105370789,
      "grad_norm": 0.7127063870429993,
      "learning_rate": 0.0002277750242430772,
      "loss": 3.2669,
      "step": 57700
    },
    {
      "epoch": 6.215692605747497,
      "grad_norm": 0.7443158030509949,
      "learning_rate": 0.00022745178321301583,
      "loss": 3.2807,
      "step": 57750
    },
    {
      "epoch": 6.221074157787106,
      "grad_norm": 0.7019312381744385,
      "learning_rate": 0.0002271285421829544,
      "loss": 3.2779,
      "step": 57800
    },
    {
      "epoch": 6.226455709826714,
      "grad_norm": 0.7332218289375305,
      "learning_rate": 0.00022680530115289297,
      "loss": 3.2759,
      "step": 57850
    },
    {
      "epoch": 6.231837261866322,
      "grad_norm": 0.7171279788017273,
      "learning_rate": 0.0002264820601228316,
      "loss": 3.2717,
      "step": 57900
    },
    {
      "epoch": 6.237218813905931,
      "grad_norm": 0.7057476043701172,
      "learning_rate": 0.00022615881909277016,
      "loss": 3.2711,
      "step": 57950
    },
    {
      "epoch": 6.242600365945538,
      "grad_norm": 0.7647303342819214,
      "learning_rate": 0.00022583557806270875,
      "loss": 3.294,
      "step": 58000
    },
    {
      "epoch": 6.242600365945538,
      "eval_accuracy": 0.38429027010130035,
      "eval_loss": 3.3797404766082764,
      "eval_runtime": 184.4757,
      "eval_samples_per_second": 97.633,
      "eval_steps_per_second": 6.104,
      "step": 58000
    },
    {
      "epoch": 6.247981917985147,
      "grad_norm": 0.7837227582931519,
      "learning_rate": 0.00022551233703264732,
      "loss": 3.2747,
      "step": 58050
    },
    {
      "epoch": 6.253363470024755,
      "grad_norm": 0.695070743560791,
      "learning_rate": 0.0002251890960025859,
      "loss": 3.275,
      "step": 58100
    },
    {
      "epoch": 6.258745022064363,
      "grad_norm": 0.7820463180541992,
      "learning_rate": 0.0002248658549725245,
      "loss": 3.2708,
      "step": 58150
    },
    {
      "epoch": 6.264126574103972,
      "grad_norm": 0.7407938838005066,
      "learning_rate": 0.00022454261394246308,
      "loss": 3.2694,
      "step": 58200
    },
    {
      "epoch": 6.26950812614358,
      "grad_norm": 0.7725470662117004,
      "learning_rate": 0.00022421937291240164,
      "loss": 3.2987,
      "step": 58250
    },
    {
      "epoch": 6.274889678183188,
      "grad_norm": 0.6986104846000671,
      "learning_rate": 0.00022389613188234027,
      "loss": 3.2811,
      "step": 58300
    },
    {
      "epoch": 6.280271230222796,
      "grad_norm": 0.7266172170639038,
      "learning_rate": 0.00022357289085227883,
      "loss": 3.2621,
      "step": 58350
    },
    {
      "epoch": 6.285652782262405,
      "grad_norm": 0.7130420207977295,
      "learning_rate": 0.0002232496498222174,
      "loss": 3.2854,
      "step": 58400
    },
    {
      "epoch": 6.2910343343020125,
      "grad_norm": 0.7785718441009521,
      "learning_rate": 0.000222926408792156,
      "loss": 3.2968,
      "step": 58450
    },
    {
      "epoch": 6.296415886341621,
      "grad_norm": 0.7451063990592957,
      "learning_rate": 0.0002226031677620946,
      "loss": 3.2744,
      "step": 58500
    },
    {
      "epoch": 6.301797438381229,
      "grad_norm": 0.7651119232177734,
      "learning_rate": 0.00022227992673203316,
      "loss": 3.2693,
      "step": 58550
    },
    {
      "epoch": 6.307178990420837,
      "grad_norm": 0.6871006488800049,
      "learning_rate": 0.00022195668570197175,
      "loss": 3.2854,
      "step": 58600
    },
    {
      "epoch": 6.312560542460446,
      "grad_norm": 0.7136965394020081,
      "learning_rate": 0.00022163344467191032,
      "loss": 3.2816,
      "step": 58650
    },
    {
      "epoch": 6.3179420945000535,
      "grad_norm": 0.6924245953559875,
      "learning_rate": 0.00022131020364184891,
      "loss": 3.2792,
      "step": 58700
    },
    {
      "epoch": 6.323323646539662,
      "grad_norm": 0.7000690698623657,
      "learning_rate": 0.0002209869626117875,
      "loss": 3.27,
      "step": 58750
    },
    {
      "epoch": 6.328705198579271,
      "grad_norm": 0.7083388566970825,
      "learning_rate": 0.00022066372158172608,
      "loss": 3.2717,
      "step": 58800
    },
    {
      "epoch": 6.334086750618878,
      "grad_norm": 0.7490871548652649,
      "learning_rate": 0.00022034048055166464,
      "loss": 3.2644,
      "step": 58850
    },
    {
      "epoch": 6.339468302658487,
      "grad_norm": 0.8122921586036682,
      "learning_rate": 0.00022001723952160327,
      "loss": 3.2735,
      "step": 58900
    },
    {
      "epoch": 6.344849854698095,
      "grad_norm": 0.7660691738128662,
      "learning_rate": 0.00021970046331214307,
      "loss": 3.289,
      "step": 58950
    },
    {
      "epoch": 6.350231406737703,
      "grad_norm": 0.7373554706573486,
      "learning_rate": 0.00021937722228208167,
      "loss": 3.2875,
      "step": 59000
    },
    {
      "epoch": 6.350231406737703,
      "eval_accuracy": 0.3845901520716683,
      "eval_loss": 3.377363920211792,
      "eval_runtime": 184.6602,
      "eval_samples_per_second": 97.536,
      "eval_steps_per_second": 6.098,
      "step": 59000
    },
    {
      "epoch": 6.355612958777312,
      "grad_norm": 0.6990751028060913,
      "learning_rate": 0.00021905398125202024,
      "loss": 3.2777,
      "step": 59050
    },
    {
      "epoch": 6.360994510816919,
      "grad_norm": 0.6831377744674683,
      "learning_rate": 0.00021873074022195883,
      "loss": 3.2672,
      "step": 59100
    },
    {
      "epoch": 6.366376062856528,
      "grad_norm": 0.7702904343605042,
      "learning_rate": 0.0002184074991918974,
      "loss": 3.2795,
      "step": 59150
    },
    {
      "epoch": 6.371757614896136,
      "grad_norm": 0.7417288422584534,
      "learning_rate": 0.00021808425816183597,
      "loss": 3.2912,
      "step": 59200
    },
    {
      "epoch": 6.377139166935744,
      "grad_norm": 0.7117642760276794,
      "learning_rate": 0.0002177610171317746,
      "loss": 3.2837,
      "step": 59250
    },
    {
      "epoch": 6.382520718975353,
      "grad_norm": 0.708830714225769,
      "learning_rate": 0.00021743777610171315,
      "loss": 3.2765,
      "step": 59300
    },
    {
      "epoch": 6.387902271014961,
      "grad_norm": 0.7194830179214478,
      "learning_rate": 0.00021711453507165172,
      "loss": 3.2744,
      "step": 59350
    },
    {
      "epoch": 6.393283823054569,
      "grad_norm": 0.7376460433006287,
      "learning_rate": 0.00021679129404159034,
      "loss": 3.2752,
      "step": 59400
    },
    {
      "epoch": 6.398665375094177,
      "grad_norm": 0.7774999141693115,
      "learning_rate": 0.0002164680530115289,
      "loss": 3.2929,
      "step": 59450
    },
    {
      "epoch": 6.404046927133785,
      "grad_norm": 0.7298051118850708,
      "learning_rate": 0.00021614481198146748,
      "loss": 3.2803,
      "step": 59500
    },
    {
      "epoch": 6.4094284791733935,
      "grad_norm": 0.7311624884605408,
      "learning_rate": 0.00021582157095140607,
      "loss": 3.2915,
      "step": 59550
    },
    {
      "epoch": 6.414810031213002,
      "grad_norm": 0.7226830124855042,
      "learning_rate": 0.0002155047947419459,
      "loss": 3.2698,
      "step": 59600
    },
    {
      "epoch": 6.42019158325261,
      "grad_norm": 0.7621427774429321,
      "learning_rate": 0.00021518155371188447,
      "loss": 3.3053,
      "step": 59650
    },
    {
      "epoch": 6.425573135292218,
      "grad_norm": 0.7501605153083801,
      "learning_rate": 0.00021485831268182304,
      "loss": 3.2884,
      "step": 59700
    },
    {
      "epoch": 6.430954687331827,
      "grad_norm": 0.7090311050415039,
      "learning_rate": 0.00021453507165176166,
      "loss": 3.291,
      "step": 59750
    },
    {
      "epoch": 6.4363362393714345,
      "grad_norm": 0.7262114882469177,
      "learning_rate": 0.00021421183062170023,
      "loss": 3.2812,
      "step": 59800
    },
    {
      "epoch": 6.441717791411043,
      "grad_norm": 0.7257036566734314,
      "learning_rate": 0.0002138885895916388,
      "loss": 3.2759,
      "step": 59850
    },
    {
      "epoch": 6.447099343450651,
      "grad_norm": 0.7639397978782654,
      "learning_rate": 0.00021356534856157742,
      "loss": 3.2704,
      "step": 59900
    },
    {
      "epoch": 6.452480895490259,
      "grad_norm": 0.7122944593429565,
      "learning_rate": 0.000213242107531516,
      "loss": 3.2779,
      "step": 59950
    },
    {
      "epoch": 6.457862447529868,
      "grad_norm": 0.752813458442688,
      "learning_rate": 0.00021291886650145456,
      "loss": 3.2888,
      "step": 60000
    },
    {
      "epoch": 6.457862447529868,
      "eval_accuracy": 0.38499097257481596,
      "eval_loss": 3.374267339706421,
      "eval_runtime": 184.357,
      "eval_samples_per_second": 97.696,
      "eval_steps_per_second": 6.108,
      "step": 60000
    },
    {
      "epoch": 6.4632439995694755,
      "grad_norm": 0.7329573035240173,
      "learning_rate": 0.00021259562547139315,
      "loss": 3.2826,
      "step": 60050
    },
    {
      "epoch": 6.468625551609084,
      "grad_norm": 0.8175730109214783,
      "learning_rate": 0.00021227238444133175,
      "loss": 3.2989,
      "step": 60100
    },
    {
      "epoch": 6.474007103648693,
      "grad_norm": 0.7768701314926147,
      "learning_rate": 0.0002119491434112703,
      "loss": 3.3011,
      "step": 60150
    },
    {
      "epoch": 6.4793886556883,
      "grad_norm": 0.7943546175956726,
      "learning_rate": 0.0002116259023812089,
      "loss": 3.2931,
      "step": 60200
    },
    {
      "epoch": 6.484770207727909,
      "grad_norm": 0.7377960681915283,
      "learning_rate": 0.00021130266135114747,
      "loss": 3.2956,
      "step": 60250
    },
    {
      "epoch": 6.490151759767517,
      "grad_norm": 0.7520676851272583,
      "learning_rate": 0.00021097942032108607,
      "loss": 3.3043,
      "step": 60300
    },
    {
      "epoch": 6.495533311807125,
      "grad_norm": 0.7445512413978577,
      "learning_rate": 0.00021065617929102466,
      "loss": 3.2857,
      "step": 60350
    },
    {
      "epoch": 6.500914863846734,
      "grad_norm": 0.7369388937950134,
      "learning_rate": 0.00021033293826096323,
      "loss": 3.2871,
      "step": 60400
    },
    {
      "epoch": 6.506296415886341,
      "grad_norm": 0.7274868488311768,
      "learning_rate": 0.0002100096972309018,
      "loss": 3.2942,
      "step": 60450
    },
    {
      "epoch": 6.51167796792595,
      "grad_norm": 0.6820805668830872,
      "learning_rate": 0.00020968645620084042,
      "loss": 3.2786,
      "step": 60500
    },
    {
      "epoch": 6.517059519965558,
      "grad_norm": 0.7346232533454895,
      "learning_rate": 0.000209363215170779,
      "loss": 3.3005,
      "step": 60550
    },
    {
      "epoch": 6.522441072005166,
      "grad_norm": 0.7301453351974487,
      "learning_rate": 0.00020903997414071758,
      "loss": 3.2615,
      "step": 60600
    },
    {
      "epoch": 6.5278226240447745,
      "grad_norm": 0.731817901134491,
      "learning_rate": 0.00020871673311065618,
      "loss": 3.2904,
      "step": 60650
    },
    {
      "epoch": 6.533204176084383,
      "grad_norm": 0.7374353408813477,
      "learning_rate": 0.00020839349208059475,
      "loss": 3.2955,
      "step": 60700
    },
    {
      "epoch": 6.538585728123991,
      "grad_norm": 0.7328256368637085,
      "learning_rate": 0.00020807025105053334,
      "loss": 3.2936,
      "step": 60750
    },
    {
      "epoch": 6.543967280163599,
      "grad_norm": 0.7349701523780823,
      "learning_rate": 0.0002077470100204719,
      "loss": 3.2798,
      "step": 60800
    },
    {
      "epoch": 6.549348832203208,
      "grad_norm": 0.7573752403259277,
      "learning_rate": 0.00020742376899041048,
      "loss": 3.2768,
      "step": 60850
    },
    {
      "epoch": 6.5547303842428155,
      "grad_norm": 0.7667546272277832,
      "learning_rate": 0.0002071005279603491,
      "loss": 3.2734,
      "step": 60900
    },
    {
      "epoch": 6.560111936282424,
      "grad_norm": 0.7604581117630005,
      "learning_rate": 0.00020677728693028766,
      "loss": 3.3027,
      "step": 60950
    },
    {
      "epoch": 6.565493488322032,
      "grad_norm": 0.7706729173660278,
      "learning_rate": 0.00020645404590022623,
      "loss": 3.2793,
      "step": 61000
    },
    {
      "epoch": 6.565493488322032,
      "eval_accuracy": 0.38562528813387686,
      "eval_loss": 3.3712966442108154,
      "eval_runtime": 184.3467,
      "eval_samples_per_second": 97.702,
      "eval_steps_per_second": 6.108,
      "step": 61000
    },
    {
      "epoch": 6.57087504036164,
      "grad_norm": 0.7069858908653259,
      "learning_rate": 0.00020613080487016485,
      "loss": 3.3016,
      "step": 61050
    },
    {
      "epoch": 6.576256592401249,
      "grad_norm": 0.7463257312774658,
      "learning_rate": 0.00020580756384010342,
      "loss": 3.2881,
      "step": 61100
    },
    {
      "epoch": 6.5816381444408565,
      "grad_norm": 0.7532726526260376,
      "learning_rate": 0.000205484322810042,
      "loss": 3.3061,
      "step": 61150
    },
    {
      "epoch": 6.587019696480465,
      "grad_norm": 0.7295279502868652,
      "learning_rate": 0.00020516108177998058,
      "loss": 3.2828,
      "step": 61200
    },
    {
      "epoch": 6.592401248520073,
      "grad_norm": 0.7714658975601196,
      "learning_rate": 0.00020483784074991918,
      "loss": 3.2938,
      "step": 61250
    },
    {
      "epoch": 6.597782800559681,
      "grad_norm": 0.7401666641235352,
      "learning_rate": 0.00020451459971985775,
      "loss": 3.285,
      "step": 61300
    },
    {
      "epoch": 6.60316435259929,
      "grad_norm": 0.7363163232803345,
      "learning_rate": 0.00020419135868979634,
      "loss": 3.2999,
      "step": 61350
    },
    {
      "epoch": 6.608545904638898,
      "grad_norm": 0.7745422720909119,
      "learning_rate": 0.0002038681176597349,
      "loss": 3.2896,
      "step": 61400
    },
    {
      "epoch": 6.613927456678506,
      "grad_norm": 0.7607348561286926,
      "learning_rate": 0.0002035448766296735,
      "loss": 3.271,
      "step": 61450
    },
    {
      "epoch": 6.619309008718115,
      "grad_norm": 0.7640889883041382,
      "learning_rate": 0.0002032216355996121,
      "loss": 3.2854,
      "step": 61500
    },
    {
      "epoch": 6.624690560757722,
      "grad_norm": 0.7307209372520447,
      "learning_rate": 0.00020289839456955067,
      "loss": 3.2924,
      "step": 61550
    },
    {
      "epoch": 6.630072112797331,
      "grad_norm": 0.7490965127944946,
      "learning_rate": 0.0002025751535394893,
      "loss": 3.2879,
      "step": 61600
    },
    {
      "epoch": 6.635453664836939,
      "grad_norm": 0.7853986024856567,
      "learning_rate": 0.00020225191250942785,
      "loss": 3.303,
      "step": 61650
    },
    {
      "epoch": 6.640835216876547,
      "grad_norm": 0.7805899977684021,
      "learning_rate": 0.00020192867147936642,
      "loss": 3.3031,
      "step": 61700
    },
    {
      "epoch": 6.6462167689161555,
      "grad_norm": 0.7700685858726501,
      "learning_rate": 0.00020160543044930502,
      "loss": 3.2844,
      "step": 61750
    },
    {
      "epoch": 6.651598320955763,
      "grad_norm": 0.7245808243751526,
      "learning_rate": 0.0002012821894192436,
      "loss": 3.2975,
      "step": 61800
    },
    {
      "epoch": 6.656979872995372,
      "grad_norm": 0.7608835697174072,
      "learning_rate": 0.00020095894838918218,
      "loss": 3.2969,
      "step": 61850
    },
    {
      "epoch": 6.66236142503498,
      "grad_norm": 0.7222352623939514,
      "learning_rate": 0.00020063570735912077,
      "loss": 3.2865,
      "step": 61900
    },
    {
      "epoch": 6.667742977074588,
      "grad_norm": 0.7594550251960754,
      "learning_rate": 0.00020031246632905934,
      "loss": 3.28,
      "step": 61950
    },
    {
      "epoch": 6.6731245291141965,
      "grad_norm": 0.7762590646743774,
      "learning_rate": 0.00019998922529899794,
      "loss": 3.3076,
      "step": 62000
    },
    {
      "epoch": 6.6731245291141965,
      "eval_accuracy": 0.38575176009529294,
      "eval_loss": 3.3661062717437744,
      "eval_runtime": 184.5115,
      "eval_samples_per_second": 97.615,
      "eval_steps_per_second": 6.103,
      "step": 62000
    },
    {
      "epoch": 6.678506081153805,
      "grad_norm": 0.7811170220375061,
      "learning_rate": 0.00019966598426893653,
      "loss": 3.2865,
      "step": 62050
    },
    {
      "epoch": 6.683887633193413,
      "grad_norm": 0.7994932532310486,
      "learning_rate": 0.0001993427432388751,
      "loss": 3.2957,
      "step": 62100
    },
    {
      "epoch": 6.689269185233021,
      "grad_norm": 0.7882549166679382,
      "learning_rate": 0.00019901950220881367,
      "loss": 3.2879,
      "step": 62150
    },
    {
      "epoch": 6.69465073727263,
      "grad_norm": 0.754192054271698,
      "learning_rate": 0.0001986962611787523,
      "loss": 3.2916,
      "step": 62200
    },
    {
      "epoch": 6.7000322893122375,
      "grad_norm": 0.7306296229362488,
      "learning_rate": 0.00019837302014869085,
      "loss": 3.2847,
      "step": 62250
    },
    {
      "epoch": 6.705413841351846,
      "grad_norm": 0.7327289581298828,
      "learning_rate": 0.00019804977911862942,
      "loss": 3.2982,
      "step": 62300
    },
    {
      "epoch": 6.710795393391454,
      "grad_norm": 0.7537993788719177,
      "learning_rate": 0.00019772653808856804,
      "loss": 3.2966,
      "step": 62350
    },
    {
      "epoch": 6.716176945431062,
      "grad_norm": 0.7344412803649902,
      "learning_rate": 0.0001974032970585066,
      "loss": 3.2869,
      "step": 62400
    },
    {
      "epoch": 6.721558497470671,
      "grad_norm": 0.7439603209495544,
      "learning_rate": 0.00019708005602844518,
      "loss": 3.283,
      "step": 62450
    },
    {
      "epoch": 6.7269400495102785,
      "grad_norm": 0.7974243760108948,
      "learning_rate": 0.00019675681499838377,
      "loss": 3.2961,
      "step": 62500
    },
    {
      "epoch": 6.732321601549887,
      "grad_norm": 0.7178831100463867,
      "learning_rate": 0.00019643357396832237,
      "loss": 3.2921,
      "step": 62550
    },
    {
      "epoch": 6.737703153589496,
      "grad_norm": 0.766057014465332,
      "learning_rate": 0.00019611033293826096,
      "loss": 3.2831,
      "step": 62600
    },
    {
      "epoch": 6.743084705629103,
      "grad_norm": 0.707496702671051,
      "learning_rate": 0.00019578709190819953,
      "loss": 3.3025,
      "step": 62650
    },
    {
      "epoch": 6.748466257668712,
      "grad_norm": 0.7768601775169373,
      "learning_rate": 0.0001954638508781381,
      "loss": 3.2813,
      "step": 62700
    },
    {
      "epoch": 6.75384780970832,
      "grad_norm": 0.7563713192939758,
      "learning_rate": 0.00019514060984807672,
      "loss": 3.2937,
      "step": 62750
    },
    {
      "epoch": 6.759229361747928,
      "grad_norm": 0.7284691333770752,
      "learning_rate": 0.0001948173688180153,
      "loss": 3.2902,
      "step": 62800
    },
    {
      "epoch": 6.7646109137875365,
      "grad_norm": 0.7694335579872131,
      "learning_rate": 0.00019449412778795386,
      "loss": 3.2945,
      "step": 62850
    },
    {
      "epoch": 6.769992465827144,
      "grad_norm": 0.7515472769737244,
      "learning_rate": 0.00019417088675789245,
      "loss": 3.2992,
      "step": 62900
    },
    {
      "epoch": 6.775374017866753,
      "grad_norm": 0.7684220671653748,
      "learning_rate": 0.00019384764572783104,
      "loss": 3.2811,
      "step": 62950
    },
    {
      "epoch": 6.780755569906361,
      "grad_norm": 0.7146512866020203,
      "learning_rate": 0.0001935244046977696,
      "loss": 3.2747,
      "step": 63000
    },
    {
      "epoch": 6.780755569906361,
      "eval_accuracy": 0.3862579738996206,
      "eval_loss": 3.3628346920013428,
      "eval_runtime": 184.8889,
      "eval_samples_per_second": 97.415,
      "eval_steps_per_second": 6.09,
      "step": 63000
    },
    {
      "epoch": 6.786137121945969,
      "grad_norm": 0.7978770136833191,
      "learning_rate": 0.0001932011636677082,
      "loss": 3.2844,
      "step": 63050
    },
    {
      "epoch": 6.7915186739855775,
      "grad_norm": 0.736603319644928,
      "learning_rate": 0.00019287792263764677,
      "loss": 3.2882,
      "step": 63100
    },
    {
      "epoch": 6.796900226025185,
      "grad_norm": 0.7129955291748047,
      "learning_rate": 0.00019255468160758537,
      "loss": 3.3103,
      "step": 63150
    },
    {
      "epoch": 6.802281778064794,
      "grad_norm": 0.746450662612915,
      "learning_rate": 0.00019223144057752396,
      "loss": 3.2909,
      "step": 63200
    },
    {
      "epoch": 6.807663330104402,
      "grad_norm": 0.7554918527603149,
      "learning_rate": 0.00019190819954746253,
      "loss": 3.2843,
      "step": 63250
    },
    {
      "epoch": 6.813044882144011,
      "grad_norm": 0.7812954783439636,
      "learning_rate": 0.0001915849585174011,
      "loss": 3.2837,
      "step": 63300
    },
    {
      "epoch": 6.8184264341836185,
      "grad_norm": 0.6976150870323181,
      "learning_rate": 0.00019126171748733972,
      "loss": 3.2814,
      "step": 63350
    },
    {
      "epoch": 6.823807986223227,
      "grad_norm": 0.7288589477539062,
      "learning_rate": 0.0001909384764572783,
      "loss": 3.2857,
      "step": 63400
    },
    {
      "epoch": 6.829189538262835,
      "grad_norm": 0.7863096594810486,
      "learning_rate": 0.00019061523542721688,
      "loss": 3.2833,
      "step": 63450
    },
    {
      "epoch": 6.834571090302443,
      "grad_norm": 0.7787613272666931,
      "learning_rate": 0.00019029199439715548,
      "loss": 3.2926,
      "step": 63500
    },
    {
      "epoch": 6.839952642342052,
      "grad_norm": 0.7694011330604553,
      "learning_rate": 0.00018996875336709404,
      "loss": 3.2688,
      "step": 63550
    },
    {
      "epoch": 6.8453341943816595,
      "grad_norm": 0.725056529045105,
      "learning_rate": 0.00018965197715763385,
      "loss": 3.3016,
      "step": 63600
    },
    {
      "epoch": 6.850715746421268,
      "grad_norm": 0.7571734189987183,
      "learning_rate": 0.00018932873612757245,
      "loss": 3.2893,
      "step": 63650
    },
    {
      "epoch": 6.856097298460876,
      "grad_norm": 0.7747765183448792,
      "learning_rate": 0.00018900549509751104,
      "loss": 3.2775,
      "step": 63700
    },
    {
      "epoch": 6.861478850500484,
      "grad_norm": 0.727161169052124,
      "learning_rate": 0.0001886822540674496,
      "loss": 3.2792,
      "step": 63750
    },
    {
      "epoch": 6.866860402540093,
      "grad_norm": 0.7230259776115417,
      "learning_rate": 0.00018835901303738818,
      "loss": 3.283,
      "step": 63800
    },
    {
      "epoch": 6.8722419545797,
      "grad_norm": 0.7609695792198181,
      "learning_rate": 0.0001880357720073268,
      "loss": 3.2759,
      "step": 63850
    },
    {
      "epoch": 6.877623506619309,
      "grad_norm": 0.7278729677200317,
      "learning_rate": 0.00018771253097726537,
      "loss": 3.2971,
      "step": 63900
    },
    {
      "epoch": 6.8830050586589175,
      "grad_norm": 0.7795426249504089,
      "learning_rate": 0.00018738928994720393,
      "loss": 3.2935,
      "step": 63950
    },
    {
      "epoch": 6.888386610698525,
      "grad_norm": 0.7273575067520142,
      "learning_rate": 0.00018706604891714255,
      "loss": 3.2721,
      "step": 64000
    },
    {
      "epoch": 6.888386610698525,
      "eval_accuracy": 0.3869662820752832,
      "eval_loss": 3.35867977142334,
      "eval_runtime": 184.7144,
      "eval_samples_per_second": 97.507,
      "eval_steps_per_second": 6.096,
      "step": 64000
    },
    {
      "epoch": 6.893768162738134,
      "grad_norm": 0.7398075461387634,
      "learning_rate": 0.00018674280788708112,
      "loss": 3.3042,
      "step": 64050
    },
    {
      "epoch": 6.899149714777742,
      "grad_norm": 0.7296501398086548,
      "learning_rate": 0.0001864195668570197,
      "loss": 3.2869,
      "step": 64100
    },
    {
      "epoch": 6.90453126681735,
      "grad_norm": 0.7152851223945618,
      "learning_rate": 0.00018609632582695828,
      "loss": 3.2904,
      "step": 64150
    },
    {
      "epoch": 6.9099128188569585,
      "grad_norm": 0.7873846292495728,
      "learning_rate": 0.00018577308479689685,
      "loss": 3.2882,
      "step": 64200
    },
    {
      "epoch": 6.915294370896566,
      "grad_norm": 0.7342795729637146,
      "learning_rate": 0.00018544984376683545,
      "loss": 3.2858,
      "step": 64250
    },
    {
      "epoch": 6.920675922936175,
      "grad_norm": 0.7274847626686096,
      "learning_rate": 0.00018512660273677404,
      "loss": 3.2816,
      "step": 64300
    },
    {
      "epoch": 6.926057474975783,
      "grad_norm": 0.7213486433029175,
      "learning_rate": 0.0001848033617067126,
      "loss": 3.2917,
      "step": 64350
    },
    {
      "epoch": 6.931439027015391,
      "grad_norm": 0.7593924403190613,
      "learning_rate": 0.00018448012067665123,
      "loss": 3.2918,
      "step": 64400
    },
    {
      "epoch": 6.9368205790549995,
      "grad_norm": 0.7654011249542236,
      "learning_rate": 0.000184163344467191,
      "loss": 3.2951,
      "step": 64450
    },
    {
      "epoch": 6.942202131094608,
      "grad_norm": 0.7375518083572388,
      "learning_rate": 0.0001838401034371296,
      "loss": 3.2838,
      "step": 64500
    },
    {
      "epoch": 6.947583683134216,
      "grad_norm": 0.7767999172210693,
      "learning_rate": 0.0001835168624070682,
      "loss": 3.2713,
      "step": 64550
    },
    {
      "epoch": 6.952965235173824,
      "grad_norm": 0.7748505473136902,
      "learning_rate": 0.00018319362137700677,
      "loss": 3.2991,
      "step": 64600
    },
    {
      "epoch": 6.958346787213433,
      "grad_norm": 0.7966855764389038,
      "learning_rate": 0.00018287038034694536,
      "loss": 3.2939,
      "step": 64650
    },
    {
      "epoch": 6.9637283392530405,
      "grad_norm": 0.7504708170890808,
      "learning_rate": 0.00018254713931688393,
      "loss": 3.2935,
      "step": 64700
    },
    {
      "epoch": 6.969109891292649,
      "grad_norm": 0.7350914478302002,
      "learning_rate": 0.00018222389828682252,
      "loss": 3.2941,
      "step": 64750
    },
    {
      "epoch": 6.974491443332257,
      "grad_norm": 0.7610340118408203,
      "learning_rate": 0.00018190065725676112,
      "loss": 3.2865,
      "step": 64800
    },
    {
      "epoch": 6.979872995371865,
      "grad_norm": 0.7873901724815369,
      "learning_rate": 0.00018157741622669969,
      "loss": 3.2812,
      "step": 64850
    },
    {
      "epoch": 6.985254547411474,
      "grad_norm": 0.7452903389930725,
      "learning_rate": 0.00018125417519663825,
      "loss": 3.274,
      "step": 64900
    },
    {
      "epoch": 6.990636099451081,
      "grad_norm": 0.7723655104637146,
      "learning_rate": 0.00018093093416657687,
      "loss": 3.2738,
      "step": 64950
    },
    {
      "epoch": 6.99601765149069,
      "grad_norm": 0.7380433082580566,
      "learning_rate": 0.00018060769313651544,
      "loss": 3.3015,
      "step": 65000
    },
    {
      "epoch": 6.99601765149069,
      "eval_accuracy": 0.38695530813361395,
      "eval_loss": 3.3558216094970703,
      "eval_runtime": 184.6319,
      "eval_samples_per_second": 97.551,
      "eval_steps_per_second": 6.099,
      "step": 65000
    },
    {
      "epoch": 7.0013992035302985,
      "grad_norm": 0.7733202576637268,
      "learning_rate": 0.000180284452106454,
      "loss": 3.2541,
      "step": 65050
    },
    {
      "epoch": 7.006780755569906,
      "grad_norm": 0.7763499021530151,
      "learning_rate": 0.00017996121107639263,
      "loss": 3.2048,
      "step": 65100
    },
    {
      "epoch": 7.012162307609515,
      "grad_norm": 0.7320621013641357,
      "learning_rate": 0.0001796379700463312,
      "loss": 3.2076,
      "step": 65150
    },
    {
      "epoch": 7.017543859649122,
      "grad_norm": 0.8151242136955261,
      "learning_rate": 0.0001793147290162698,
      "loss": 3.215,
      "step": 65200
    },
    {
      "epoch": 7.022925411688731,
      "grad_norm": 0.7708078026771545,
      "learning_rate": 0.00017899148798620836,
      "loss": 3.2105,
      "step": 65250
    },
    {
      "epoch": 7.0283069637283395,
      "grad_norm": 0.8013386726379395,
      "learning_rate": 0.00017866824695614696,
      "loss": 3.2011,
      "step": 65300
    },
    {
      "epoch": 7.033688515767947,
      "grad_norm": 0.7393853664398193,
      "learning_rate": 0.00017834500592608555,
      "loss": 3.2074,
      "step": 65350
    },
    {
      "epoch": 7.039070067807556,
      "grad_norm": 0.7421411871910095,
      "learning_rate": 0.00017802176489602412,
      "loss": 3.2042,
      "step": 65400
    },
    {
      "epoch": 7.044451619847164,
      "grad_norm": 0.7495855689048767,
      "learning_rate": 0.00017769852386596269,
      "loss": 3.2167,
      "step": 65450
    },
    {
      "epoch": 7.049833171886772,
      "grad_norm": 0.7470657229423523,
      "learning_rate": 0.0001773752828359013,
      "loss": 3.1952,
      "step": 65500
    },
    {
      "epoch": 7.0552147239263805,
      "grad_norm": 0.8882248401641846,
      "learning_rate": 0.00017705204180583988,
      "loss": 3.2211,
      "step": 65550
    },
    {
      "epoch": 7.060596275965988,
      "grad_norm": 0.768017053604126,
      "learning_rate": 0.00017672880077577844,
      "loss": 3.2248,
      "step": 65600
    },
    {
      "epoch": 7.065977828005597,
      "grad_norm": 0.7792190313339233,
      "learning_rate": 0.00017640555974571704,
      "loss": 3.2227,
      "step": 65650
    },
    {
      "epoch": 7.071359380045205,
      "grad_norm": 0.8252404928207397,
      "learning_rate": 0.00017608231871565563,
      "loss": 3.2228,
      "step": 65700
    },
    {
      "epoch": 7.076740932084813,
      "grad_norm": 0.7655684351921082,
      "learning_rate": 0.0001757590776855942,
      "loss": 3.2197,
      "step": 65750
    },
    {
      "epoch": 7.0821224841244215,
      "grad_norm": 0.7804136276245117,
      "learning_rate": 0.0001754358366555328,
      "loss": 3.2213,
      "step": 65800
    },
    {
      "epoch": 7.08750403616403,
      "grad_norm": 0.7610635161399841,
      "learning_rate": 0.00017511259562547136,
      "loss": 3.2253,
      "step": 65850
    },
    {
      "epoch": 7.092885588203638,
      "grad_norm": 0.7551177740097046,
      "learning_rate": 0.00017478935459540996,
      "loss": 3.2124,
      "step": 65900
    },
    {
      "epoch": 7.098267140243246,
      "grad_norm": 0.7511612772941589,
      "learning_rate": 0.00017446611356534855,
      "loss": 3.1975,
      "step": 65950
    },
    {
      "epoch": 7.103648692282855,
      "grad_norm": 0.7450958490371704,
      "learning_rate": 0.00017414287253528712,
      "loss": 3.212,
      "step": 66000
    },
    {
      "epoch": 7.103648692282855,
      "eval_accuracy": 0.3868903337067009,
      "eval_loss": 3.361914873123169,
      "eval_runtime": 184.4591,
      "eval_samples_per_second": 97.642,
      "eval_steps_per_second": 6.104,
      "step": 66000
    },
    {
      "epoch": 7.109030244322462,
      "grad_norm": 0.7630361914634705,
      "learning_rate": 0.00017381963150522569,
      "loss": 3.215,
      "step": 66050
    },
    {
      "epoch": 7.114411796362071,
      "grad_norm": 0.7323626279830933,
      "learning_rate": 0.0001734963904751643,
      "loss": 3.2146,
      "step": 66100
    },
    {
      "epoch": 7.119793348401679,
      "grad_norm": 0.8025902509689331,
      "learning_rate": 0.00017317314944510288,
      "loss": 3.2304,
      "step": 66150
    },
    {
      "epoch": 7.125174900441287,
      "grad_norm": 0.738914966583252,
      "learning_rate": 0.00017284990841504147,
      "loss": 3.1982,
      "step": 66200
    },
    {
      "epoch": 7.130556452480896,
      "grad_norm": 0.7768139839172363,
      "learning_rate": 0.00017252666738498007,
      "loss": 3.2225,
      "step": 66250
    },
    {
      "epoch": 7.135938004520503,
      "grad_norm": 0.7497488856315613,
      "learning_rate": 0.00017220342635491863,
      "loss": 3.2186,
      "step": 66300
    },
    {
      "epoch": 7.141319556560112,
      "grad_norm": 0.8401795029640198,
      "learning_rate": 0.00017188018532485723,
      "loss": 3.2201,
      "step": 66350
    },
    {
      "epoch": 7.1467011085997205,
      "grad_norm": 0.7919259667396545,
      "learning_rate": 0.0001715569442947958,
      "loss": 3.2207,
      "step": 66400
    },
    {
      "epoch": 7.152082660639328,
      "grad_norm": 0.8087385892868042,
      "learning_rate": 0.0001712337032647344,
      "loss": 3.219,
      "step": 66450
    },
    {
      "epoch": 7.157464212678937,
      "grad_norm": 0.7992889881134033,
      "learning_rate": 0.00017091046223467298,
      "loss": 3.2403,
      "step": 66500
    },
    {
      "epoch": 7.162845764718545,
      "grad_norm": 0.7694268226623535,
      "learning_rate": 0.00017058722120461155,
      "loss": 3.2292,
      "step": 66550
    },
    {
      "epoch": 7.168227316758153,
      "grad_norm": 0.7614873051643372,
      "learning_rate": 0.00017026398017455012,
      "loss": 3.2384,
      "step": 66600
    },
    {
      "epoch": 7.1736088687977615,
      "grad_norm": 0.817666232585907,
      "learning_rate": 0.00016994073914448874,
      "loss": 3.2018,
      "step": 66650
    },
    {
      "epoch": 7.178990420837369,
      "grad_norm": 0.7415510416030884,
      "learning_rate": 0.0001696174981144273,
      "loss": 3.2306,
      "step": 66700
    },
    {
      "epoch": 7.184371972876978,
      "grad_norm": 0.7840098738670349,
      "learning_rate": 0.00016929425708436588,
      "loss": 3.2175,
      "step": 66750
    },
    {
      "epoch": 7.189753524916586,
      "grad_norm": 0.7869345545768738,
      "learning_rate": 0.0001689710160543045,
      "loss": 3.2276,
      "step": 66800
    },
    {
      "epoch": 7.195135076956194,
      "grad_norm": 0.8216828107833862,
      "learning_rate": 0.00016864777502424307,
      "loss": 3.2068,
      "step": 66850
    },
    {
      "epoch": 7.2005166289958025,
      "grad_norm": 0.7508730888366699,
      "learning_rate": 0.00016832453399418163,
      "loss": 3.2105,
      "step": 66900
    },
    {
      "epoch": 7.205898181035411,
      "grad_norm": 0.7955493927001953,
      "learning_rate": 0.00016800129296412023,
      "loss": 3.2426,
      "step": 66950
    },
    {
      "epoch": 7.211279733075019,
      "grad_norm": 0.8153558373451233,
      "learning_rate": 0.00016767805193405882,
      "loss": 3.2296,
      "step": 67000
    },
    {
      "epoch": 7.211279733075019,
      "eval_accuracy": 0.38737492558635345,
      "eval_loss": 3.3583383560180664,
      "eval_runtime": 184.4993,
      "eval_samples_per_second": 97.621,
      "eval_steps_per_second": 6.103,
      "step": 67000
    },
    {
      "epoch": 7.216661285114627,
      "grad_norm": 0.7690896391868591,
      "learning_rate": 0.0001673548109039974,
      "loss": 3.2248,
      "step": 67050
    },
    {
      "epoch": 7.222042837154235,
      "grad_norm": 0.8362789750099182,
      "learning_rate": 0.00016703156987393598,
      "loss": 3.2299,
      "step": 67100
    },
    {
      "epoch": 7.2274243891938434,
      "grad_norm": 0.7391707897186279,
      "learning_rate": 0.00016670832884387455,
      "loss": 3.2048,
      "step": 67150
    },
    {
      "epoch": 7.232805941233452,
      "grad_norm": 0.8125260472297668,
      "learning_rate": 0.00016638508781381317,
      "loss": 3.2238,
      "step": 67200
    },
    {
      "epoch": 7.23818749327306,
      "grad_norm": 0.8026116490364075,
      "learning_rate": 0.00016606184678375174,
      "loss": 3.2537,
      "step": 67250
    },
    {
      "epoch": 7.243569045312668,
      "grad_norm": 0.7808555960655212,
      "learning_rate": 0.0001657386057536903,
      "loss": 3.2316,
      "step": 67300
    },
    {
      "epoch": 7.248950597352277,
      "grad_norm": 0.7729163765907288,
      "learning_rate": 0.00016541536472362893,
      "loss": 3.203,
      "step": 67350
    },
    {
      "epoch": 7.254332149391884,
      "grad_norm": 0.7771885395050049,
      "learning_rate": 0.0001650921236935675,
      "loss": 3.213,
      "step": 67400
    },
    {
      "epoch": 7.259713701431493,
      "grad_norm": 0.7727643251419067,
      "learning_rate": 0.00016476888266350607,
      "loss": 3.2356,
      "step": 67450
    },
    {
      "epoch": 7.265095253471101,
      "grad_norm": 0.8023343682289124,
      "learning_rate": 0.00016444564163344466,
      "loss": 3.2202,
      "step": 67500
    },
    {
      "epoch": 7.270476805510709,
      "grad_norm": 0.8250343203544617,
      "learning_rate": 0.00016412240060338326,
      "loss": 3.2159,
      "step": 67550
    },
    {
      "epoch": 7.275858357550318,
      "grad_norm": 0.7585141062736511,
      "learning_rate": 0.00016379915957332182,
      "loss": 3.2372,
      "step": 67600
    },
    {
      "epoch": 7.281239909589925,
      "grad_norm": 0.7813882827758789,
      "learning_rate": 0.00016347591854326042,
      "loss": 3.2399,
      "step": 67650
    },
    {
      "epoch": 7.286621461629534,
      "grad_norm": 0.8016800284385681,
      "learning_rate": 0.00016315267751319898,
      "loss": 3.2188,
      "step": 67700
    },
    {
      "epoch": 7.2920030136691425,
      "grad_norm": 0.7547227144241333,
      "learning_rate": 0.00016282943648313755,
      "loss": 3.2164,
      "step": 67750
    },
    {
      "epoch": 7.29738456570875,
      "grad_norm": 0.8415718078613281,
      "learning_rate": 0.00016250619545307617,
      "loss": 3.2366,
      "step": 67800
    },
    {
      "epoch": 7.302766117748359,
      "grad_norm": 0.8185309767723083,
      "learning_rate": 0.00016218295442301474,
      "loss": 3.2294,
      "step": 67850
    },
    {
      "epoch": 7.308147669787967,
      "grad_norm": 0.7623180150985718,
      "learning_rate": 0.0001618597133929533,
      "loss": 3.2326,
      "step": 67900
    },
    {
      "epoch": 7.313529221827575,
      "grad_norm": 0.7978857755661011,
      "learning_rate": 0.00016153647236289193,
      "loss": 3.2278,
      "step": 67950
    },
    {
      "epoch": 7.3189107738671835,
      "grad_norm": 0.8266440629959106,
      "learning_rate": 0.0001612132313328305,
      "loss": 3.2317,
      "step": 68000
    },
    {
      "epoch": 7.3189107738671835,
      "eval_accuracy": 0.3874469624509745,
      "eval_loss": 3.356354236602783,
      "eval_runtime": 184.0898,
      "eval_samples_per_second": 97.838,
      "eval_steps_per_second": 6.117,
      "step": 68000
    },
    {
      "epoch": 7.324292325906791,
      "grad_norm": 0.8046820759773254,
      "learning_rate": 0.0001608899903027691,
      "loss": 3.2336,
      "step": 68050
    },
    {
      "epoch": 7.3296738779464,
      "grad_norm": 0.7737776637077332,
      "learning_rate": 0.0001605732140933089,
      "loss": 3.2317,
      "step": 68100
    },
    {
      "epoch": 7.335055429986008,
      "grad_norm": 0.7804083824157715,
      "learning_rate": 0.0001602499730632475,
      "loss": 3.242,
      "step": 68150
    },
    {
      "epoch": 7.340436982025616,
      "grad_norm": 0.8369970917701721,
      "learning_rate": 0.00015992673203318606,
      "loss": 3.2488,
      "step": 68200
    },
    {
      "epoch": 7.3458185340652244,
      "grad_norm": 0.7761555910110474,
      "learning_rate": 0.00015960349100312463,
      "loss": 3.2404,
      "step": 68250
    },
    {
      "epoch": 7.351200086104833,
      "grad_norm": 0.7949065566062927,
      "learning_rate": 0.00015928024997306325,
      "loss": 3.2401,
      "step": 68300
    },
    {
      "epoch": 7.356581638144441,
      "grad_norm": 0.786142110824585,
      "learning_rate": 0.00015895700894300182,
      "loss": 3.2246,
      "step": 68350
    },
    {
      "epoch": 7.361963190184049,
      "grad_norm": 0.8034423589706421,
      "learning_rate": 0.00015863376791294039,
      "loss": 3.2479,
      "step": 68400
    },
    {
      "epoch": 7.367344742223658,
      "grad_norm": 0.7865124344825745,
      "learning_rate": 0.000158310526882879,
      "loss": 3.2533,
      "step": 68450
    },
    {
      "epoch": 7.372726294263265,
      "grad_norm": 0.8077775835990906,
      "learning_rate": 0.00015798728585281758,
      "loss": 3.2212,
      "step": 68500
    },
    {
      "epoch": 7.378107846302874,
      "grad_norm": 0.7793838381767273,
      "learning_rate": 0.00015766404482275614,
      "loss": 3.2376,
      "step": 68550
    },
    {
      "epoch": 7.383489398342482,
      "grad_norm": 0.7663792967796326,
      "learning_rate": 0.00015734080379269474,
      "loss": 3.233,
      "step": 68600
    },
    {
      "epoch": 7.38887095038209,
      "grad_norm": 0.8004462122917175,
      "learning_rate": 0.00015701756276263333,
      "loss": 3.2285,
      "step": 68650
    },
    {
      "epoch": 7.394252502421699,
      "grad_norm": 0.8044269680976868,
      "learning_rate": 0.0001566943217325719,
      "loss": 3.2373,
      "step": 68700
    },
    {
      "epoch": 7.399634054461306,
      "grad_norm": 0.7555727362632751,
      "learning_rate": 0.0001563710807025105,
      "loss": 3.2488,
      "step": 68750
    },
    {
      "epoch": 7.405015606500915,
      "grad_norm": 0.7628423571586609,
      "learning_rate": 0.00015604783967244906,
      "loss": 3.2561,
      "step": 68800
    },
    {
      "epoch": 7.4103971585405235,
      "grad_norm": 0.8033735156059265,
      "learning_rate": 0.00015572459864238763,
      "loss": 3.2305,
      "step": 68850
    },
    {
      "epoch": 7.415778710580131,
      "grad_norm": 0.7518730163574219,
      "learning_rate": 0.00015540135761232625,
      "loss": 3.233,
      "step": 68900
    },
    {
      "epoch": 7.42116026261974,
      "grad_norm": 0.7811611294746399,
      "learning_rate": 0.00015507811658226482,
      "loss": 3.2354,
      "step": 68950
    },
    {
      "epoch": 7.426541814659347,
      "grad_norm": 0.7855293154716492,
      "learning_rate": 0.00015475487555220344,
      "loss": 3.2286,
      "step": 69000
    },
    {
      "epoch": 7.426541814659347,
      "eval_accuracy": 0.387994247046896,
      "eval_loss": 3.35193133354187,
      "eval_runtime": 184.3176,
      "eval_samples_per_second": 97.717,
      "eval_steps_per_second": 6.109,
      "step": 69000
    },
    {
      "epoch": 7.431923366698956,
      "grad_norm": 0.8192943334579468,
      "learning_rate": 0.000154431634522142,
      "loss": 3.2247,
      "step": 69050
    },
    {
      "epoch": 7.4373049187385645,
      "grad_norm": 0.8080233931541443,
      "learning_rate": 0.00015410839349208058,
      "loss": 3.2429,
      "step": 69100
    },
    {
      "epoch": 7.442686470778172,
      "grad_norm": 0.7587073445320129,
      "learning_rate": 0.00015378515246201917,
      "loss": 3.2347,
      "step": 69150
    },
    {
      "epoch": 7.448068022817781,
      "grad_norm": 0.7683467864990234,
      "learning_rate": 0.00015346191143195774,
      "loss": 3.2382,
      "step": 69200
    },
    {
      "epoch": 7.453449574857389,
      "grad_norm": 0.7774455547332764,
      "learning_rate": 0.00015313867040189633,
      "loss": 3.2234,
      "step": 69250
    },
    {
      "epoch": 7.458831126896997,
      "grad_norm": 0.7898199558258057,
      "learning_rate": 0.00015281542937183493,
      "loss": 3.2451,
      "step": 69300
    },
    {
      "epoch": 7.4642126789366054,
      "grad_norm": 0.7566768527030945,
      "learning_rate": 0.0001524921883417735,
      "loss": 3.2346,
      "step": 69350
    },
    {
      "epoch": 7.469594230976213,
      "grad_norm": 0.7716087698936462,
      "learning_rate": 0.00015216894731171206,
      "loss": 3.2341,
      "step": 69400
    },
    {
      "epoch": 7.474975783015822,
      "grad_norm": 0.7814199924468994,
      "learning_rate": 0.00015184570628165068,
      "loss": 3.2306,
      "step": 69450
    },
    {
      "epoch": 7.48035733505543,
      "grad_norm": 0.8122509121894836,
      "learning_rate": 0.00015152246525158925,
      "loss": 3.2184,
      "step": 69500
    },
    {
      "epoch": 7.485738887095038,
      "grad_norm": 0.7457565069198608,
      "learning_rate": 0.00015119922422152782,
      "loss": 3.2325,
      "step": 69550
    },
    {
      "epoch": 7.491120439134646,
      "grad_norm": 0.7829018235206604,
      "learning_rate": 0.00015087598319146644,
      "loss": 3.2243,
      "step": 69600
    },
    {
      "epoch": 7.496501991174255,
      "grad_norm": 0.8459808230400085,
      "learning_rate": 0.000150552742161405,
      "loss": 3.2302,
      "step": 69650
    },
    {
      "epoch": 7.501883543213863,
      "grad_norm": 0.8299519419670105,
      "learning_rate": 0.00015022950113134358,
      "loss": 3.2257,
      "step": 69700
    },
    {
      "epoch": 7.507265095253471,
      "grad_norm": 0.8326858282089233,
      "learning_rate": 0.00014990626010128217,
      "loss": 3.2396,
      "step": 69750
    },
    {
      "epoch": 7.51264664729308,
      "grad_norm": 0.7516535520553589,
      "learning_rate": 0.00014958301907122077,
      "loss": 3.2274,
      "step": 69800
    },
    {
      "epoch": 7.518028199332687,
      "grad_norm": 0.7791556119918823,
      "learning_rate": 0.00014925977804115933,
      "loss": 3.2422,
      "step": 69850
    },
    {
      "epoch": 7.523409751372296,
      "grad_norm": 0.8247345685958862,
      "learning_rate": 0.00014893653701109793,
      "loss": 3.2314,
      "step": 69900
    },
    {
      "epoch": 7.528791303411904,
      "grad_norm": 0.7909741997718811,
      "learning_rate": 0.0001486132959810365,
      "loss": 3.2438,
      "step": 69950
    },
    {
      "epoch": 7.534172855451512,
      "grad_norm": 0.8029667139053345,
      "learning_rate": 0.0001482900549509751,
      "loss": 3.233,
      "step": 70000
    },
    {
      "epoch": 7.534172855451512,
      "eval_accuracy": 0.3884731889763822,
      "eval_loss": 3.3487160205841064,
      "eval_runtime": 183.8927,
      "eval_samples_per_second": 97.943,
      "eval_steps_per_second": 6.123,
      "step": 70000
    },
    {
      "epoch": 7.539554407491121,
      "grad_norm": 0.7621917128562927,
      "learning_rate": 0.00014796681392091368,
      "loss": 3.2358,
      "step": 70050
    },
    {
      "epoch": 7.544935959530728,
      "grad_norm": 0.7857876420021057,
      "learning_rate": 0.00014764357289085228,
      "loss": 3.2597,
      "step": 70100
    },
    {
      "epoch": 7.550317511570337,
      "grad_norm": 0.7604870200157166,
      "learning_rate": 0.00014732033186079085,
      "loss": 3.2395,
      "step": 70150
    },
    {
      "epoch": 7.5556990636099455,
      "grad_norm": 0.7385656833648682,
      "learning_rate": 0.00014699709083072944,
      "loss": 3.23,
      "step": 70200
    },
    {
      "epoch": 7.561080615649553,
      "grad_norm": 0.8508681058883667,
      "learning_rate": 0.00014668031462126925,
      "loss": 3.2464,
      "step": 70250
    },
    {
      "epoch": 7.566462167689162,
      "grad_norm": 0.796556830406189,
      "learning_rate": 0.00014635707359120784,
      "loss": 3.2477,
      "step": 70300
    },
    {
      "epoch": 7.57184371972877,
      "grad_norm": 0.825899064540863,
      "learning_rate": 0.0001460338325611464,
      "loss": 3.2513,
      "step": 70350
    },
    {
      "epoch": 7.577225271768378,
      "grad_norm": 0.7872753143310547,
      "learning_rate": 0.00014571059153108498,
      "loss": 3.2477,
      "step": 70400
    },
    {
      "epoch": 7.5826068238079865,
      "grad_norm": 0.8007068634033203,
      "learning_rate": 0.00014538735050102357,
      "loss": 3.2435,
      "step": 70450
    },
    {
      "epoch": 7.587988375847594,
      "grad_norm": 0.7999556064605713,
      "learning_rate": 0.00014506410947096217,
      "loss": 3.2424,
      "step": 70500
    },
    {
      "epoch": 7.593369927887203,
      "grad_norm": 0.8032796382904053,
      "learning_rate": 0.00014474086844090076,
      "loss": 3.2531,
      "step": 70550
    },
    {
      "epoch": 7.598751479926811,
      "grad_norm": 0.8466640710830688,
      "learning_rate": 0.00014441762741083933,
      "loss": 3.2365,
      "step": 70600
    },
    {
      "epoch": 7.604133031966419,
      "grad_norm": 0.7856156229972839,
      "learning_rate": 0.00014409438638077792,
      "loss": 3.2433,
      "step": 70650
    },
    {
      "epoch": 7.609514584006027,
      "grad_norm": 0.9202490448951721,
      "learning_rate": 0.00014377114535071652,
      "loss": 3.2455,
      "step": 70700
    },
    {
      "epoch": 7.614896136045635,
      "grad_norm": 0.782346248626709,
      "learning_rate": 0.0001434479043206551,
      "loss": 3.2319,
      "step": 70750
    },
    {
      "epoch": 7.620277688085244,
      "grad_norm": 0.8108184933662415,
      "learning_rate": 0.00014312466329059368,
      "loss": 3.235,
      "step": 70800
    },
    {
      "epoch": 7.625659240124852,
      "grad_norm": 0.7564395666122437,
      "learning_rate": 0.00014280142226053225,
      "loss": 3.2335,
      "step": 70850
    },
    {
      "epoch": 7.63104079216446,
      "grad_norm": 0.7824851274490356,
      "learning_rate": 0.00014247818123047084,
      "loss": 3.2388,
      "step": 70900
    },
    {
      "epoch": 7.636422344204068,
      "grad_norm": 0.7763656377792358,
      "learning_rate": 0.0001421549402004094,
      "loss": 3.2411,
      "step": 70950
    },
    {
      "epoch": 7.641803896243677,
      "grad_norm": 0.7939996719360352,
      "learning_rate": 0.000141831699170348,
      "loss": 3.2331,
      "step": 71000
    },
    {
      "epoch": 7.641803896243677,
      "eval_accuracy": 0.3886686555215605,
      "eval_loss": 3.3441824913024902,
      "eval_runtime": 183.9195,
      "eval_samples_per_second": 97.929,
      "eval_steps_per_second": 6.122,
      "step": 71000
    },
    {
      "epoch": 7.647185448283285,
      "grad_norm": 0.8930923342704773,
      "learning_rate": 0.0001415084581402866,
      "loss": 3.243,
      "step": 71050
    },
    {
      "epoch": 7.652567000322893,
      "grad_norm": 0.7729410529136658,
      "learning_rate": 0.00014118521711022517,
      "loss": 3.2257,
      "step": 71100
    },
    {
      "epoch": 7.657948552362502,
      "grad_norm": 0.7841166257858276,
      "learning_rate": 0.00014086197608016376,
      "loss": 3.2449,
      "step": 71150
    },
    {
      "epoch": 7.663330104402109,
      "grad_norm": 0.8066850304603577,
      "learning_rate": 0.00014053873505010236,
      "loss": 3.2398,
      "step": 71200
    },
    {
      "epoch": 7.668711656441718,
      "grad_norm": 0.7879757881164551,
      "learning_rate": 0.00014021549402004092,
      "loss": 3.2541,
      "step": 71250
    },
    {
      "epoch": 7.674093208481326,
      "grad_norm": 0.8244560956954956,
      "learning_rate": 0.00013989225298997952,
      "loss": 3.2352,
      "step": 71300
    },
    {
      "epoch": 7.679474760520934,
      "grad_norm": 0.8272544145584106,
      "learning_rate": 0.00013956901195991811,
      "loss": 3.2471,
      "step": 71350
    },
    {
      "epoch": 7.684856312560543,
      "grad_norm": 0.7601016163825989,
      "learning_rate": 0.00013924577092985668,
      "loss": 3.2286,
      "step": 71400
    },
    {
      "epoch": 7.69023786460015,
      "grad_norm": 0.8125275373458862,
      "learning_rate": 0.00013892252989979528,
      "loss": 3.2627,
      "step": 71450
    },
    {
      "epoch": 7.695619416639759,
      "grad_norm": 0.7913317084312439,
      "learning_rate": 0.00013859928886973384,
      "loss": 3.2377,
      "step": 71500
    },
    {
      "epoch": 7.7010009686793675,
      "grad_norm": 0.8260928988456726,
      "learning_rate": 0.00013827604783967244,
      "loss": 3.2376,
      "step": 71550
    },
    {
      "epoch": 7.706382520718975,
      "grad_norm": 0.8788133263587952,
      "learning_rate": 0.000137952806809611,
      "loss": 3.2478,
      "step": 71600
    },
    {
      "epoch": 7.711764072758584,
      "grad_norm": 0.8264977931976318,
      "learning_rate": 0.0001376295657795496,
      "loss": 3.2384,
      "step": 71650
    },
    {
      "epoch": 7.717145624798192,
      "grad_norm": 0.7762619256973267,
      "learning_rate": 0.0001373063247494882,
      "loss": 3.2401,
      "step": 71700
    },
    {
      "epoch": 7.7225271768378,
      "grad_norm": 0.8169916272163391,
      "learning_rate": 0.00013698308371942676,
      "loss": 3.2624,
      "step": 71750
    },
    {
      "epoch": 7.727908728877408,
      "grad_norm": 0.8310092687606812,
      "learning_rate": 0.00013665984268936536,
      "loss": 3.2632,
      "step": 71800
    },
    {
      "epoch": 7.733290280917016,
      "grad_norm": 0.8596165776252747,
      "learning_rate": 0.00013633660165930395,
      "loss": 3.2462,
      "step": 71850
    },
    {
      "epoch": 7.738671832956625,
      "grad_norm": 0.8650221824645996,
      "learning_rate": 0.00013601336062924255,
      "loss": 3.2373,
      "step": 71900
    },
    {
      "epoch": 7.744053384996233,
      "grad_norm": 0.7710651755332947,
      "learning_rate": 0.00013569011959918111,
      "loss": 3.2236,
      "step": 71950
    },
    {
      "epoch": 7.749434937035841,
      "grad_norm": 0.8062445521354675,
      "learning_rate": 0.0001353668785691197,
      "loss": 3.2351,
      "step": 72000
    },
    {
      "epoch": 7.749434937035841,
      "eval_accuracy": 0.3893080777663487,
      "eval_loss": 3.340257406234741,
      "eval_runtime": 183.9994,
      "eval_samples_per_second": 97.886,
      "eval_steps_per_second": 6.12,
      "step": 72000
    },
    {
      "epoch": 7.754816489075449,
      "grad_norm": 0.8193359375,
      "learning_rate": 0.00013504363753905828,
      "loss": 3.2326,
      "step": 72050
    },
    {
      "epoch": 7.760198041115058,
      "grad_norm": 0.8223119974136353,
      "learning_rate": 0.00013472039650899687,
      "loss": 3.2475,
      "step": 72100
    },
    {
      "epoch": 7.765579593154666,
      "grad_norm": 0.8278833627700806,
      "learning_rate": 0.00013439715547893544,
      "loss": 3.2453,
      "step": 72150
    },
    {
      "epoch": 7.770961145194274,
      "grad_norm": 0.81866854429245,
      "learning_rate": 0.00013407391444887403,
      "loss": 3.2393,
      "step": 72200
    },
    {
      "epoch": 7.776342697233883,
      "grad_norm": 0.8339356184005737,
      "learning_rate": 0.0001337506734188126,
      "loss": 3.2256,
      "step": 72250
    },
    {
      "epoch": 7.78172424927349,
      "grad_norm": 0.8300743103027344,
      "learning_rate": 0.0001334274323887512,
      "loss": 3.2373,
      "step": 72300
    },
    {
      "epoch": 7.787105801313099,
      "grad_norm": 0.7977941036224365,
      "learning_rate": 0.0001331041913586898,
      "loss": 3.2321,
      "step": 72350
    },
    {
      "epoch": 7.792487353352707,
      "grad_norm": 0.7674457430839539,
      "learning_rate": 0.00013278095032862838,
      "loss": 3.2398,
      "step": 72400
    },
    {
      "epoch": 7.797868905392315,
      "grad_norm": 0.8473909497261047,
      "learning_rate": 0.00013245770929856695,
      "loss": 3.2443,
      "step": 72450
    },
    {
      "epoch": 7.803250457431924,
      "grad_norm": 0.7988379597663879,
      "learning_rate": 0.00013213446826850555,
      "loss": 3.2534,
      "step": 72500
    },
    {
      "epoch": 7.808632009471531,
      "grad_norm": 0.8200710415840149,
      "learning_rate": 0.00013181122723844411,
      "loss": 3.2418,
      "step": 72550
    },
    {
      "epoch": 7.81401356151114,
      "grad_norm": 0.7961328625679016,
      "learning_rate": 0.0001314879862083827,
      "loss": 3.2305,
      "step": 72600
    },
    {
      "epoch": 7.819395113550748,
      "grad_norm": 0.7949033379554749,
      "learning_rate": 0.00013116474517832128,
      "loss": 3.2171,
      "step": 72650
    },
    {
      "epoch": 7.824776665590356,
      "grad_norm": 0.8292585015296936,
      "learning_rate": 0.00013084150414825987,
      "loss": 3.2526,
      "step": 72700
    },
    {
      "epoch": 7.830158217629965,
      "grad_norm": 0.8255747556686401,
      "learning_rate": 0.00013051826311819844,
      "loss": 3.2476,
      "step": 72750
    },
    {
      "epoch": 7.835539769669572,
      "grad_norm": 0.7956199645996094,
      "learning_rate": 0.00013019502208813703,
      "loss": 3.2272,
      "step": 72800
    },
    {
      "epoch": 7.840921321709181,
      "grad_norm": 0.8041326999664307,
      "learning_rate": 0.00012987178105807563,
      "loss": 3.239,
      "step": 72850
    },
    {
      "epoch": 7.846302873748789,
      "grad_norm": 0.7850301265716553,
      "learning_rate": 0.00012954854002801422,
      "loss": 3.2418,
      "step": 72900
    },
    {
      "epoch": 7.851684425788397,
      "grad_norm": 0.8481513857841492,
      "learning_rate": 0.0001292252989979528,
      "loss": 3.2386,
      "step": 72950
    },
    {
      "epoch": 7.857065977828006,
      "grad_norm": 0.7845147252082825,
      "learning_rate": 0.00012890205796789139,
      "loss": 3.2302,
      "step": 73000
    },
    {
      "epoch": 7.857065977828006,
      "eval_accuracy": 0.3894613869910549,
      "eval_loss": 3.3379290103912354,
      "eval_runtime": 184.4624,
      "eval_samples_per_second": 97.64,
      "eval_steps_per_second": 6.104,
      "step": 73000
    },
    {
      "epoch": 7.862447529867614,
      "grad_norm": 0.856820285320282,
      "learning_rate": 0.00012857881693782998,
      "loss": 3.2399,
      "step": 73050
    },
    {
      "epoch": 7.867829081907222,
      "grad_norm": 0.8157970309257507,
      "learning_rate": 0.00012825557590776855,
      "loss": 3.2647,
      "step": 73100
    },
    {
      "epoch": 7.87321063394683,
      "grad_norm": 0.7562020421028137,
      "learning_rate": 0.00012793233487770714,
      "loss": 3.2523,
      "step": 73150
    },
    {
      "epoch": 7.878592185986438,
      "grad_norm": 0.787735641002655,
      "learning_rate": 0.0001276090938476457,
      "loss": 3.2244,
      "step": 73200
    },
    {
      "epoch": 7.883973738026047,
      "grad_norm": 0.8418269157409668,
      "learning_rate": 0.0001272858528175843,
      "loss": 3.2317,
      "step": 73250
    },
    {
      "epoch": 7.889355290065655,
      "grad_norm": 0.8080412149429321,
      "learning_rate": 0.00012696261178752287,
      "loss": 3.2469,
      "step": 73300
    },
    {
      "epoch": 7.894736842105263,
      "grad_norm": 0.8602666854858398,
      "learning_rate": 0.00012663937075746147,
      "loss": 3.2513,
      "step": 73350
    },
    {
      "epoch": 7.900118394144871,
      "grad_norm": 0.8057377934455872,
      "learning_rate": 0.00012631612972740006,
      "loss": 3.243,
      "step": 73400
    },
    {
      "epoch": 7.90549994618448,
      "grad_norm": 0.7962327599525452,
      "learning_rate": 0.00012599288869733863,
      "loss": 3.2494,
      "step": 73450
    },
    {
      "epoch": 7.910881498224088,
      "grad_norm": 0.8277156352996826,
      "learning_rate": 0.00012566964766727722,
      "loss": 3.2401,
      "step": 73500
    },
    {
      "epoch": 7.916263050263696,
      "grad_norm": 0.8099175095558167,
      "learning_rate": 0.00012534640663721582,
      "loss": 3.2389,
      "step": 73550
    },
    {
      "epoch": 7.921644602303305,
      "grad_norm": 0.8122092485427856,
      "learning_rate": 0.00012502316560715439,
      "loss": 3.2332,
      "step": 73600
    },
    {
      "epoch": 7.927026154342912,
      "grad_norm": 0.7989636659622192,
      "learning_rate": 0.00012469992457709298,
      "loss": 3.2237,
      "step": 73650
    },
    {
      "epoch": 7.932407706382521,
      "grad_norm": 0.8023879528045654,
      "learning_rate": 0.00012437668354703158,
      "loss": 3.2415,
      "step": 73700
    },
    {
      "epoch": 7.937789258422129,
      "grad_norm": 0.8016743063926697,
      "learning_rate": 0.00012405344251697014,
      "loss": 3.2513,
      "step": 73750
    },
    {
      "epoch": 7.943170810461737,
      "grad_norm": 0.7854738235473633,
      "learning_rate": 0.00012373020148690874,
      "loss": 3.2359,
      "step": 73800
    },
    {
      "epoch": 7.948552362501346,
      "grad_norm": 0.8327288627624512,
      "learning_rate": 0.0001234069604568473,
      "loss": 3.2498,
      "step": 73850
    },
    {
      "epoch": 7.953933914540953,
      "grad_norm": 0.8185678720474243,
      "learning_rate": 0.0001230837194267859,
      "loss": 3.2475,
      "step": 73900
    },
    {
      "epoch": 7.959315466580562,
      "grad_norm": 0.8182744979858398,
      "learning_rate": 0.00012276047839672447,
      "loss": 3.2302,
      "step": 73950
    },
    {
      "epoch": 7.96469701862017,
      "grad_norm": 0.8040328621864319,
      "learning_rate": 0.00012243723736666306,
      "loss": 3.2432,
      "step": 74000
    },
    {
      "epoch": 7.96469701862017,
      "eval_accuracy": 0.38971454821966267,
      "eval_loss": 3.3338217735290527,
      "eval_runtime": 184.4099,
      "eval_samples_per_second": 97.668,
      "eval_steps_per_second": 6.106,
      "step": 74000
    },
    {
      "epoch": 7.970078570659778,
      "grad_norm": 0.8335957527160645,
      "learning_rate": 0.00012211399633660166,
      "loss": 3.2301,
      "step": 74050
    },
    {
      "epoch": 7.975460122699387,
      "grad_norm": 0.8088896870613098,
      "learning_rate": 0.00012179075530654022,
      "loss": 3.2286,
      "step": 74100
    },
    {
      "epoch": 7.980841674738995,
      "grad_norm": 0.8292359709739685,
      "learning_rate": 0.00012146751427647882,
      "loss": 3.2493,
      "step": 74150
    },
    {
      "epoch": 7.986223226778603,
      "grad_norm": 0.7748470902442932,
      "learning_rate": 0.0001211442732464174,
      "loss": 3.22,
      "step": 74200
    },
    {
      "epoch": 7.991604778818211,
      "grad_norm": Infinity,
      "learning_rate": 0.0001208274970369572,
      "loss": 3.2319,
      "step": 74250
    },
    {
      "epoch": 7.996986330857819,
      "grad_norm": 0.8105632066726685,
      "learning_rate": 0.0001205042560068958,
      "loss": 3.2342,
      "step": 74300
    },
    {
      "epoch": 8.002367882897428,
      "grad_norm": 0.7788518071174622,
      "learning_rate": 0.0001201810149768344,
      "loss": 3.2038,
      "step": 74350
    },
    {
      "epoch": 8.007749434937036,
      "grad_norm": 0.9167351722717285,
      "learning_rate": 0.00011985777394677296,
      "loss": 3.1749,
      "step": 74400
    },
    {
      "epoch": 8.013130986976645,
      "grad_norm": 0.7819600701332092,
      "learning_rate": 0.00011953453291671154,
      "loss": 3.1353,
      "step": 74450
    },
    {
      "epoch": 8.018512539016251,
      "grad_norm": 0.8220762014389038,
      "learning_rate": 0.00011921129188665014,
      "loss": 3.155,
      "step": 74500
    },
    {
      "epoch": 8.02389409105586,
      "grad_norm": 0.8363442420959473,
      "learning_rate": 0.0001188880508565887,
      "loss": 3.1558,
      "step": 74550
    },
    {
      "epoch": 8.029275643095469,
      "grad_norm": 0.7991000413894653,
      "learning_rate": 0.0001185648098265273,
      "loss": 3.1509,
      "step": 74600
    },
    {
      "epoch": 8.034657195135077,
      "grad_norm": 0.8093554973602295,
      "learning_rate": 0.0001182415687964659,
      "loss": 3.1659,
      "step": 74650
    },
    {
      "epoch": 8.040038747174686,
      "grad_norm": 0.7979394197463989,
      "learning_rate": 0.00011791832776640448,
      "loss": 3.1792,
      "step": 74700
    },
    {
      "epoch": 8.045420299214294,
      "grad_norm": 0.8007020950317383,
      "learning_rate": 0.00011759508673634306,
      "loss": 3.1709,
      "step": 74750
    },
    {
      "epoch": 8.050801851253901,
      "grad_norm": 0.7973904013633728,
      "learning_rate": 0.00011727184570628164,
      "loss": 3.1665,
      "step": 74800
    },
    {
      "epoch": 8.05618340329351,
      "grad_norm": 0.7981194257736206,
      "learning_rate": 0.00011694860467622023,
      "loss": 3.1629,
      "step": 74850
    },
    {
      "epoch": 8.061564955333118,
      "grad_norm": 0.8133523464202881,
      "learning_rate": 0.0001166253636461588,
      "loss": 3.1575,
      "step": 74900
    },
    {
      "epoch": 8.066946507372727,
      "grad_norm": 0.7740296721458435,
      "learning_rate": 0.0001163021226160974,
      "loss": 3.1605,
      "step": 74950
    },
    {
      "epoch": 8.072328059412335,
      "grad_norm": 0.8136142492294312,
      "learning_rate": 0.00011597888158603598,
      "loss": 3.1648,
      "step": 75000
    },
    {
      "epoch": 8.072328059412335,
      "eval_accuracy": 0.38995803934125495,
      "eval_loss": 3.3387575149536133,
      "eval_runtime": 184.3105,
      "eval_samples_per_second": 97.721,
      "eval_steps_per_second": 6.109,
      "step": 75000
    },
    {
      "epoch": 8.077709611451942,
      "grad_norm": 0.8422960042953491,
      "learning_rate": 0.00011566210537657578,
      "loss": 3.1912,
      "step": 75050
    },
    {
      "epoch": 8.08309116349155,
      "grad_norm": 0.880723237991333,
      "learning_rate": 0.00011533886434651438,
      "loss": 3.1485,
      "step": 75100
    },
    {
      "epoch": 8.088472715531159,
      "grad_norm": 0.7921397089958191,
      "learning_rate": 0.00011501562331645296,
      "loss": 3.16,
      "step": 75150
    },
    {
      "epoch": 8.093854267570768,
      "grad_norm": 0.8067227602005005,
      "learning_rate": 0.00011469238228639154,
      "loss": 3.1756,
      "step": 75200
    },
    {
      "epoch": 8.099235819610376,
      "grad_norm": 0.8231829404830933,
      "learning_rate": 0.00011436914125633012,
      "loss": 3.1826,
      "step": 75250
    },
    {
      "epoch": 8.104617371649983,
      "grad_norm": 0.8055379986763,
      "learning_rate": 0.00011404590022626872,
      "loss": 3.1891,
      "step": 75300
    },
    {
      "epoch": 8.109998923689592,
      "grad_norm": 0.8581538200378418,
      "learning_rate": 0.00011372265919620728,
      "loss": 3.1865,
      "step": 75350
    },
    {
      "epoch": 8.1153804757292,
      "grad_norm": 0.8242756724357605,
      "learning_rate": 0.00011339941816614588,
      "loss": 3.1839,
      "step": 75400
    },
    {
      "epoch": 8.120762027768809,
      "grad_norm": 0.8146909475326538,
      "learning_rate": 0.00011307617713608447,
      "loss": 3.1571,
      "step": 75450
    },
    {
      "epoch": 8.126143579808417,
      "grad_norm": 0.8469723463058472,
      "learning_rate": 0.00011275293610602305,
      "loss": 3.1717,
      "step": 75500
    },
    {
      "epoch": 8.131525131848026,
      "grad_norm": 0.8159957528114319,
      "learning_rate": 0.00011242969507596164,
      "loss": 3.1675,
      "step": 75550
    },
    {
      "epoch": 8.136906683887632,
      "grad_norm": 0.7882417440414429,
      "learning_rate": 0.00011210645404590022,
      "loss": 3.1808,
      "step": 75600
    },
    {
      "epoch": 8.142288235927241,
      "grad_norm": 0.8272544741630554,
      "learning_rate": 0.00011178321301583881,
      "loss": 3.176,
      "step": 75650
    },
    {
      "epoch": 8.14766978796685,
      "grad_norm": 0.8206055164337158,
      "learning_rate": 0.00011145997198577738,
      "loss": 3.1458,
      "step": 75700
    },
    {
      "epoch": 8.153051340006458,
      "grad_norm": 0.7993244528770447,
      "learning_rate": 0.00011113673095571597,
      "loss": 3.1791,
      "step": 75750
    },
    {
      "epoch": 8.158432892046067,
      "grad_norm": 0.8186585307121277,
      "learning_rate": 0.00011081348992565455,
      "loss": 3.1814,
      "step": 75800
    },
    {
      "epoch": 8.163814444085673,
      "grad_norm": 0.8069587349891663,
      "learning_rate": 0.00011049024889559314,
      "loss": 3.1815,
      "step": 75850
    },
    {
      "epoch": 8.169195996125282,
      "grad_norm": 0.7979862689971924,
      "learning_rate": 0.00011016700786553172,
      "loss": 3.1671,
      "step": 75900
    },
    {
      "epoch": 8.17457754816489,
      "grad_norm": 0.8265254497528076,
      "learning_rate": 0.00010984376683547031,
      "loss": 3.1755,
      "step": 75950
    },
    {
      "epoch": 8.1799591002045,
      "grad_norm": 0.8612515926361084,
      "learning_rate": 0.00010952052580540889,
      "loss": 3.1602,
      "step": 76000
    },
    {
      "epoch": 8.1799591002045,
      "eval_accuracy": 0.39020946212365765,
      "eval_loss": 3.3369975090026855,
      "eval_runtime": 184.6466,
      "eval_samples_per_second": 97.543,
      "eval_steps_per_second": 6.098,
      "step": 76000
    },
    {
      "epoch": 8.185340652244108,
      "grad_norm": 0.8273695111274719,
      "learning_rate": 0.00010919728477534747,
      "loss": 3.1793,
      "step": 76050
    },
    {
      "epoch": 8.190722204283716,
      "grad_norm": 0.8961669206619263,
      "learning_rate": 0.00010887404374528605,
      "loss": 3.1681,
      "step": 76100
    },
    {
      "epoch": 8.196103756323323,
      "grad_norm": 0.8245730400085449,
      "learning_rate": 0.00010855080271522465,
      "loss": 3.1835,
      "step": 76150
    },
    {
      "epoch": 8.201485308362932,
      "grad_norm": 0.8162047863006592,
      "learning_rate": 0.00010822756168516322,
      "loss": 3.174,
      "step": 76200
    },
    {
      "epoch": 8.20686686040254,
      "grad_norm": 0.849535346031189,
      "learning_rate": 0.00010790432065510181,
      "loss": 3.2086,
      "step": 76250
    },
    {
      "epoch": 8.212248412442149,
      "grad_norm": 0.8092169761657715,
      "learning_rate": 0.0001075810796250404,
      "loss": 3.1832,
      "step": 76300
    },
    {
      "epoch": 8.217629964481757,
      "grad_norm": 0.8230430483818054,
      "learning_rate": 0.00010725783859497897,
      "loss": 3.1853,
      "step": 76350
    },
    {
      "epoch": 8.223011516521364,
      "grad_norm": 0.7821382284164429,
      "learning_rate": 0.00010693459756491757,
      "loss": 3.187,
      "step": 76400
    },
    {
      "epoch": 8.228393068560973,
      "grad_norm": 0.8055399060249329,
      "learning_rate": 0.00010661135653485615,
      "loss": 3.1708,
      "step": 76450
    },
    {
      "epoch": 8.233774620600581,
      "grad_norm": 0.806912899017334,
      "learning_rate": 0.00010628811550479474,
      "loss": 3.1802,
      "step": 76500
    },
    {
      "epoch": 8.23915617264019,
      "grad_norm": 0.8519048690795898,
      "learning_rate": 0.00010596487447473331,
      "loss": 3.1795,
      "step": 76550
    },
    {
      "epoch": 8.244537724679798,
      "grad_norm": 0.8232311010360718,
      "learning_rate": 0.0001056416334446719,
      "loss": 3.2047,
      "step": 76600
    },
    {
      "epoch": 8.249919276719407,
      "grad_norm": 0.8433911204338074,
      "learning_rate": 0.00010531839241461049,
      "loss": 3.1923,
      "step": 76650
    },
    {
      "epoch": 8.255300828759013,
      "grad_norm": 0.7914126515388489,
      "learning_rate": 0.00010499515138454907,
      "loss": 3.1656,
      "step": 76700
    },
    {
      "epoch": 8.260682380798622,
      "grad_norm": 0.8264209032058716,
      "learning_rate": 0.00010467191035448765,
      "loss": 3.1945,
      "step": 76750
    },
    {
      "epoch": 8.26606393283823,
      "grad_norm": 0.8277523517608643,
      "learning_rate": 0.00010434866932442624,
      "loss": 3.1846,
      "step": 76800
    },
    {
      "epoch": 8.27144548487784,
      "grad_norm": 0.7977397441864014,
      "learning_rate": 0.00010402542829436481,
      "loss": 3.1725,
      "step": 76850
    },
    {
      "epoch": 8.276827036917448,
      "grad_norm": 0.8298705220222473,
      "learning_rate": 0.0001037021872643034,
      "loss": 3.2021,
      "step": 76900
    },
    {
      "epoch": 8.282208588957054,
      "grad_norm": 0.8329617381095886,
      "learning_rate": 0.00010337894623424199,
      "loss": 3.1974,
      "step": 76950
    },
    {
      "epoch": 8.287590140996663,
      "grad_norm": 0.8346620798110962,
      "learning_rate": 0.00010305570520418058,
      "loss": 3.1843,
      "step": 77000
    },
    {
      "epoch": 8.287590140996663,
      "eval_accuracy": 0.3903277851184876,
      "eval_loss": 3.334428071975708,
      "eval_runtime": 184.4901,
      "eval_samples_per_second": 97.626,
      "eval_steps_per_second": 6.103,
      "step": 77000
    },
    {
      "epoch": 8.292971693036272,
      "grad_norm": 0.8602821230888367,
      "learning_rate": 0.00010273246417411915,
      "loss": 3.1815,
      "step": 77050
    },
    {
      "epoch": 8.29835324507588,
      "grad_norm": 0.7964959740638733,
      "learning_rate": 0.00010240922314405774,
      "loss": 3.1878,
      "step": 77100
    },
    {
      "epoch": 8.303734797115489,
      "grad_norm": 0.8557408452033997,
      "learning_rate": 0.00010208598211399634,
      "loss": 3.1858,
      "step": 77150
    },
    {
      "epoch": 8.309116349155097,
      "grad_norm": 0.7792113423347473,
      "learning_rate": 0.0001017627410839349,
      "loss": 3.1596,
      "step": 77200
    },
    {
      "epoch": 8.314497901194704,
      "grad_norm": 0.84828782081604,
      "learning_rate": 0.0001014395000538735,
      "loss": 3.1778,
      "step": 77250
    },
    {
      "epoch": 8.319879453234313,
      "grad_norm": 0.8216564059257507,
      "learning_rate": 0.00010111625902381208,
      "loss": 3.1955,
      "step": 77300
    },
    {
      "epoch": 8.325261005273921,
      "grad_norm": 0.8319216966629028,
      "learning_rate": 0.00010079301799375066,
      "loss": 3.1809,
      "step": 77350
    },
    {
      "epoch": 8.33064255731353,
      "grad_norm": 0.8257953524589539,
      "learning_rate": 0.00010046977696368924,
      "loss": 3.2053,
      "step": 77400
    },
    {
      "epoch": 8.336024109353138,
      "grad_norm": 0.8302624225616455,
      "learning_rate": 0.00010014653593362784,
      "loss": 3.1878,
      "step": 77450
    },
    {
      "epoch": 8.341405661392745,
      "grad_norm": 0.8428215384483337,
      "learning_rate": 9.982329490356642e-05,
      "loss": 3.1801,
      "step": 77500
    },
    {
      "epoch": 8.346787213432354,
      "grad_norm": 0.862400472164154,
      "learning_rate": 9.9500053873505e-05,
      "loss": 3.1861,
      "step": 77550
    },
    {
      "epoch": 8.352168765471962,
      "grad_norm": 0.8840287327766418,
      "learning_rate": 9.917681284344358e-05,
      "loss": 3.1761,
      "step": 77600
    },
    {
      "epoch": 8.35755031751157,
      "grad_norm": 0.9052258729934692,
      "learning_rate": 9.885357181338218e-05,
      "loss": 3.1939,
      "step": 77650
    },
    {
      "epoch": 8.36293186955118,
      "grad_norm": 0.8464065790176392,
      "learning_rate": 9.853033078332074e-05,
      "loss": 3.1734,
      "step": 77700
    },
    {
      "epoch": 8.368313421590786,
      "grad_norm": 0.8124513626098633,
      "learning_rate": 9.820708975325934e-05,
      "loss": 3.2015,
      "step": 77750
    },
    {
      "epoch": 8.373694973630395,
      "grad_norm": 0.8710871934890747,
      "learning_rate": 9.788384872319793e-05,
      "loss": 3.1968,
      "step": 77800
    },
    {
      "epoch": 8.379076525670003,
      "grad_norm": 0.8791707754135132,
      "learning_rate": 9.75606076931365e-05,
      "loss": 3.1893,
      "step": 77850
    },
    {
      "epoch": 8.384458077709612,
      "grad_norm": 0.8334492444992065,
      "learning_rate": 9.723736666307508e-05,
      "loss": 3.1827,
      "step": 77900
    },
    {
      "epoch": 8.38983962974922,
      "grad_norm": 0.8116469383239746,
      "learning_rate": 9.691412563301368e-05,
      "loss": 3.1898,
      "step": 77950
    },
    {
      "epoch": 8.395221181788829,
      "grad_norm": 0.9038192629814148,
      "learning_rate": 9.659088460295227e-05,
      "loss": 3.1962,
      "step": 78000
    },
    {
      "epoch": 8.395221181788829,
      "eval_accuracy": 0.39070709234984796,
      "eval_loss": 3.3316900730133057,
      "eval_runtime": 184.4709,
      "eval_samples_per_second": 97.636,
      "eval_steps_per_second": 6.104,
      "step": 78000
    },
    {
      "epoch": 8.400602733828435,
      "grad_norm": 0.807905912399292,
      "learning_rate": 9.626764357289084e-05,
      "loss": 3.18,
      "step": 78050
    },
    {
      "epoch": 8.405984285868044,
      "grad_norm": 0.8067154884338379,
      "learning_rate": 9.594440254282943e-05,
      "loss": 3.1812,
      "step": 78100
    },
    {
      "epoch": 8.411365837907653,
      "grad_norm": 0.8242512345314026,
      "learning_rate": 9.562116151276802e-05,
      "loss": 3.2013,
      "step": 78150
    },
    {
      "epoch": 8.416747389947261,
      "grad_norm": 0.8568668365478516,
      "learning_rate": 9.52979204827066e-05,
      "loss": 3.1984,
      "step": 78200
    },
    {
      "epoch": 8.42212894198687,
      "grad_norm": 0.791793704032898,
      "learning_rate": 9.497467945264518e-05,
      "loss": 3.193,
      "step": 78250
    },
    {
      "epoch": 8.427510494026476,
      "grad_norm": 0.840458333492279,
      "learning_rate": 9.465143842258377e-05,
      "loss": 3.1926,
      "step": 78300
    },
    {
      "epoch": 8.432892046066085,
      "grad_norm": 0.8061281442642212,
      "learning_rate": 9.432819739252234e-05,
      "loss": 3.1853,
      "step": 78350
    },
    {
      "epoch": 8.438273598105694,
      "grad_norm": 0.8072904944419861,
      "learning_rate": 9.400495636246093e-05,
      "loss": 3.1874,
      "step": 78400
    },
    {
      "epoch": 8.443655150145302,
      "grad_norm": 0.8812602162361145,
      "learning_rate": 9.368818015300075e-05,
      "loss": 3.195,
      "step": 78450
    },
    {
      "epoch": 8.44903670218491,
      "grad_norm": 0.8170022964477539,
      "learning_rate": 9.336493912293932e-05,
      "loss": 3.1897,
      "step": 78500
    },
    {
      "epoch": 8.45441825422452,
      "grad_norm": 0.7490344047546387,
      "learning_rate": 9.304169809287792e-05,
      "loss": 3.1699,
      "step": 78550
    },
    {
      "epoch": 8.459799806264126,
      "grad_norm": 0.8053438663482666,
      "learning_rate": 9.27184570628165e-05,
      "loss": 3.194,
      "step": 78600
    },
    {
      "epoch": 8.465181358303735,
      "grad_norm": 0.8496149778366089,
      "learning_rate": 9.239521603275508e-05,
      "loss": 3.1806,
      "step": 78650
    },
    {
      "epoch": 8.470562910343343,
      "grad_norm": 0.8424170613288879,
      "learning_rate": 9.207197500269366e-05,
      "loss": 3.1816,
      "step": 78700
    },
    {
      "epoch": 8.475944462382952,
      "grad_norm": 0.8264489769935608,
      "learning_rate": 9.174873397263225e-05,
      "loss": 3.1812,
      "step": 78750
    },
    {
      "epoch": 8.48132601442256,
      "grad_norm": 0.888958215713501,
      "learning_rate": 9.142549294257085e-05,
      "loss": 3.2029,
      "step": 78800
    },
    {
      "epoch": 8.486707566462167,
      "grad_norm": 0.826907753944397,
      "learning_rate": 9.110225191250942e-05,
      "loss": 3.1994,
      "step": 78850
    },
    {
      "epoch": 8.492089118501776,
      "grad_norm": 0.8486902713775635,
      "learning_rate": 9.077901088244801e-05,
      "loss": 3.1758,
      "step": 78900
    },
    {
      "epoch": 8.497470670541384,
      "grad_norm": 0.8913923501968384,
      "learning_rate": 9.045576985238659e-05,
      "loss": 3.1778,
      "step": 78950
    },
    {
      "epoch": 8.502852222580993,
      "grad_norm": 0.81163090467453,
      "learning_rate": 9.013252882232517e-05,
      "loss": 3.189,
      "step": 79000
    },
    {
      "epoch": 8.502852222580993,
      "eval_accuracy": 0.39104163459142877,
      "eval_loss": 3.32745623588562,
      "eval_runtime": 184.4942,
      "eval_samples_per_second": 97.624,
      "eval_steps_per_second": 6.103,
      "step": 79000
    },
    {
      "epoch": 8.508233774620601,
      "grad_norm": 0.8120415806770325,
      "learning_rate": 8.980928779226375e-05,
      "loss": 3.1895,
      "step": 79050
    },
    {
      "epoch": 8.513615326660208,
      "grad_norm": 0.8586983680725098,
      "learning_rate": 8.948604676220235e-05,
      "loss": 3.1932,
      "step": 79100
    },
    {
      "epoch": 8.518996878699816,
      "grad_norm": 0.8254155516624451,
      "learning_rate": 8.916280573214092e-05,
      "loss": 3.1717,
      "step": 79150
    },
    {
      "epoch": 8.524378430739425,
      "grad_norm": 0.8049229979515076,
      "learning_rate": 8.883956470207951e-05,
      "loss": 3.1787,
      "step": 79200
    },
    {
      "epoch": 8.529759982779034,
      "grad_norm": 0.8635473251342773,
      "learning_rate": 8.851632367201809e-05,
      "loss": 3.1851,
      "step": 79250
    },
    {
      "epoch": 8.535141534818642,
      "grad_norm": 0.8458271026611328,
      "learning_rate": 8.819308264195669e-05,
      "loss": 3.1931,
      "step": 79300
    },
    {
      "epoch": 8.54052308685825,
      "grad_norm": 0.8510284423828125,
      "learning_rate": 8.786984161189526e-05,
      "loss": 3.1948,
      "step": 79350
    },
    {
      "epoch": 8.545904638897857,
      "grad_norm": 0.8462408185005188,
      "learning_rate": 8.754660058183385e-05,
      "loss": 3.1951,
      "step": 79400
    },
    {
      "epoch": 8.551286190937466,
      "grad_norm": 0.8107200264930725,
      "learning_rate": 8.722335955177243e-05,
      "loss": 3.1983,
      "step": 79450
    },
    {
      "epoch": 8.556667742977075,
      "grad_norm": 0.8210574984550476,
      "learning_rate": 8.690011852171101e-05,
      "loss": 3.2164,
      "step": 79500
    },
    {
      "epoch": 8.562049295016683,
      "grad_norm": 0.8413606882095337,
      "learning_rate": 8.657687749164959e-05,
      "loss": 3.2008,
      "step": 79550
    },
    {
      "epoch": 8.567430847056292,
      "grad_norm": 0.8475111126899719,
      "learning_rate": 8.625363646158819e-05,
      "loss": 3.1821,
      "step": 79600
    },
    {
      "epoch": 8.572812399095898,
      "grad_norm": 1.098301887512207,
      "learning_rate": 8.593039543152676e-05,
      "loss": 3.1869,
      "step": 79650
    },
    {
      "epoch": 8.578193951135507,
      "grad_norm": 0.8468955755233765,
      "learning_rate": 8.560715440146535e-05,
      "loss": 3.193,
      "step": 79700
    },
    {
      "epoch": 8.583575503175116,
      "grad_norm": 0.8470736742019653,
      "learning_rate": 8.528391337140394e-05,
      "loss": 3.192,
      "step": 79750
    },
    {
      "epoch": 8.588957055214724,
      "grad_norm": 0.9154149889945984,
      "learning_rate": 8.496067234134253e-05,
      "loss": 3.2032,
      "step": 79800
    },
    {
      "epoch": 8.594338607254333,
      "grad_norm": 0.8389812707901001,
      "learning_rate": 8.46374313112811e-05,
      "loss": 3.1704,
      "step": 79850
    },
    {
      "epoch": 8.599720159293941,
      "grad_norm": 0.8794986605644226,
      "learning_rate": 8.431419028121969e-05,
      "loss": 3.1839,
      "step": 79900
    },
    {
      "epoch": 8.605101711333548,
      "grad_norm": 0.8579456806182861,
      "learning_rate": 8.399094925115828e-05,
      "loss": 3.1932,
      "step": 79950
    },
    {
      "epoch": 8.610483263373157,
      "grad_norm": 0.8823782205581665,
      "learning_rate": 8.366770822109685e-05,
      "loss": 3.1951,
      "step": 80000
    },
    {
      "epoch": 8.610483263373157,
      "eval_accuracy": 0.3912045052702626,
      "eval_loss": 3.3249828815460205,
      "eval_runtime": 186.6595,
      "eval_samples_per_second": 96.491,
      "eval_steps_per_second": 6.032,
      "step": 80000
    },
    {
      "epoch": 8.615864815412765,
      "grad_norm": 0.8854741454124451,
      "learning_rate": 8.334446719103544e-05,
      "loss": 3.1919,
      "step": 80050
    },
    {
      "epoch": 8.621246367452374,
      "grad_norm": 0.8204372525215149,
      "learning_rate": 8.302122616097403e-05,
      "loss": 3.2015,
      "step": 80100
    },
    {
      "epoch": 8.626627919491982,
      "grad_norm": 0.8447315096855164,
      "learning_rate": 8.269798513091261e-05,
      "loss": 3.1919,
      "step": 80150
    },
    {
      "epoch": 8.632009471531589,
      "grad_norm": 0.8891607522964478,
      "learning_rate": 8.237474410085119e-05,
      "loss": 3.1831,
      "step": 80200
    },
    {
      "epoch": 8.637391023571197,
      "grad_norm": 0.8633220791816711,
      "learning_rate": 8.205150307078978e-05,
      "loss": 3.208,
      "step": 80250
    },
    {
      "epoch": 8.642772575610806,
      "grad_norm": 0.8665201663970947,
      "learning_rate": 8.172826204072838e-05,
      "loss": 3.1912,
      "step": 80300
    },
    {
      "epoch": 8.648154127650415,
      "grad_norm": 0.8600960373878479,
      "learning_rate": 8.140502101066694e-05,
      "loss": 3.2051,
      "step": 80350
    },
    {
      "epoch": 8.653535679690023,
      "grad_norm": 0.8409681916236877,
      "learning_rate": 8.108177998060553e-05,
      "loss": 3.1702,
      "step": 80400
    },
    {
      "epoch": 8.658917231729632,
      "grad_norm": 0.8480629920959473,
      "learning_rate": 8.075853895054412e-05,
      "loss": 3.1897,
      "step": 80450
    },
    {
      "epoch": 8.664298783769238,
      "grad_norm": 0.8516739010810852,
      "learning_rate": 8.043529792048269e-05,
      "loss": 3.1834,
      "step": 80500
    },
    {
      "epoch": 8.669680335808847,
      "grad_norm": 0.8579591512680054,
      "learning_rate": 8.011205689042128e-05,
      "loss": 3.1832,
      "step": 80550
    },
    {
      "epoch": 8.675061887848456,
      "grad_norm": 0.881572961807251,
      "learning_rate": 7.978881586035988e-05,
      "loss": 3.1876,
      "step": 80600
    },
    {
      "epoch": 8.680443439888064,
      "grad_norm": 0.8350430727005005,
      "learning_rate": 7.946557483029845e-05,
      "loss": 3.2115,
      "step": 80650
    },
    {
      "epoch": 8.685824991927673,
      "grad_norm": 0.8533426523208618,
      "learning_rate": 7.914233380023704e-05,
      "loss": 3.1914,
      "step": 80700
    },
    {
      "epoch": 8.69120654396728,
      "grad_norm": 0.7919235229492188,
      "learning_rate": 7.881909277017562e-05,
      "loss": 3.1916,
      "step": 80750
    },
    {
      "epoch": 8.696588096006888,
      "grad_norm": 0.8782989978790283,
      "learning_rate": 7.849585174011422e-05,
      "loss": 3.1851,
      "step": 80800
    },
    {
      "epoch": 8.701969648046497,
      "grad_norm": 0.8633365035057068,
      "learning_rate": 7.817261071005278e-05,
      "loss": 3.1989,
      "step": 80850
    },
    {
      "epoch": 8.707351200086105,
      "grad_norm": 0.8320886492729187,
      "learning_rate": 7.784936967999138e-05,
      "loss": 3.1888,
      "step": 80900
    },
    {
      "epoch": 8.712732752125714,
      "grad_norm": 0.8777769804000854,
      "learning_rate": 7.752612864992996e-05,
      "loss": 3.1838,
      "step": 80950
    },
    {
      "epoch": 8.718114304165322,
      "grad_norm": 0.8315258622169495,
      "learning_rate": 7.720288761986854e-05,
      "loss": 3.1758,
      "step": 81000
    },
    {
      "epoch": 8.718114304165322,
      "eval_accuracy": 0.39172343146246463,
      "eval_loss": 3.322150707244873,
      "eval_runtime": 184.6599,
      "eval_samples_per_second": 97.536,
      "eval_steps_per_second": 6.098,
      "step": 81000
    },
    {
      "epoch": 8.723495856204929,
      "grad_norm": 0.8168962597846985,
      "learning_rate": 7.687964658980712e-05,
      "loss": 3.1993,
      "step": 81050
    },
    {
      "epoch": 8.728877408244538,
      "grad_norm": 0.8527893424034119,
      "learning_rate": 7.655640555974572e-05,
      "loss": 3.1974,
      "step": 81100
    },
    {
      "epoch": 8.734258960284146,
      "grad_norm": 0.8347617387771606,
      "learning_rate": 7.623316452968428e-05,
      "loss": 3.1846,
      "step": 81150
    },
    {
      "epoch": 8.739640512323755,
      "grad_norm": 0.8442969918251038,
      "learning_rate": 7.590992349962288e-05,
      "loss": 3.1884,
      "step": 81200
    },
    {
      "epoch": 8.745022064363363,
      "grad_norm": 0.8483829498291016,
      "learning_rate": 7.558668246956147e-05,
      "loss": 3.192,
      "step": 81250
    },
    {
      "epoch": 8.75040361640297,
      "grad_norm": 0.8468664288520813,
      "learning_rate": 7.526344143950005e-05,
      "loss": 3.1873,
      "step": 81300
    },
    {
      "epoch": 8.755785168442578,
      "grad_norm": 0.8552200198173523,
      "learning_rate": 7.494020040943862e-05,
      "loss": 3.188,
      "step": 81350
    },
    {
      "epoch": 8.761166720482187,
      "grad_norm": 0.8391306400299072,
      "learning_rate": 7.461695937937722e-05,
      "loss": 3.1817,
      "step": 81400
    },
    {
      "epoch": 8.766548272521796,
      "grad_norm": 0.8693702816963196,
      "learning_rate": 7.42937183493158e-05,
      "loss": 3.1716,
      "step": 81450
    },
    {
      "epoch": 8.771929824561404,
      "grad_norm": 0.8215579390525818,
      "learning_rate": 7.397047731925439e-05,
      "loss": 3.176,
      "step": 81500
    },
    {
      "epoch": 8.777311376601011,
      "grad_norm": 0.8595774173736572,
      "learning_rate": 7.364723628919297e-05,
      "loss": 3.1801,
      "step": 81550
    },
    {
      "epoch": 8.78269292864062,
      "grad_norm": 0.8401481509208679,
      "learning_rate": 7.332399525913155e-05,
      "loss": 3.1852,
      "step": 81600
    },
    {
      "epoch": 8.788074480680228,
      "grad_norm": 0.8775984644889832,
      "learning_rate": 7.300075422907013e-05,
      "loss": 3.188,
      "step": 81650
    },
    {
      "epoch": 8.793456032719837,
      "grad_norm": 0.8345686793327332,
      "learning_rate": 7.267751319900872e-05,
      "loss": 3.2003,
      "step": 81700
    },
    {
      "epoch": 8.798837584759445,
      "grad_norm": 0.8534639477729797,
      "learning_rate": 7.235427216894731e-05,
      "loss": 3.1929,
      "step": 81750
    },
    {
      "epoch": 8.804219136799054,
      "grad_norm": 0.8451040387153625,
      "learning_rate": 7.203103113888589e-05,
      "loss": 3.1792,
      "step": 81800
    },
    {
      "epoch": 8.80960068883866,
      "grad_norm": 0.8620363473892212,
      "learning_rate": 7.170779010882447e-05,
      "loss": 3.1908,
      "step": 81850
    },
    {
      "epoch": 8.814982240878269,
      "grad_norm": 0.8879191875457764,
      "learning_rate": 7.138454907876305e-05,
      "loss": 3.18,
      "step": 81900
    },
    {
      "epoch": 8.820363792917878,
      "grad_norm": 0.8199324011802673,
      "learning_rate": 7.106130804870164e-05,
      "loss": 3.1916,
      "step": 81950
    },
    {
      "epoch": 8.825745344957486,
      "grad_norm": 0.8415508270263672,
      "learning_rate": 7.073806701864023e-05,
      "loss": 3.1849,
      "step": 82000
    },
    {
      "epoch": 8.825745344957486,
      "eval_accuracy": 0.39201994519331035,
      "eval_loss": 3.3185806274414062,
      "eval_runtime": 184.9373,
      "eval_samples_per_second": 97.39,
      "eval_steps_per_second": 6.089,
      "step": 82000
    },
    {
      "epoch": 8.831126896997095,
      "grad_norm": 0.8347445726394653,
      "learning_rate": 7.041482598857881e-05,
      "loss": 3.1864,
      "step": 82050
    },
    {
      "epoch": 8.836508449036701,
      "grad_norm": 0.8316588997840881,
      "learning_rate": 7.009158495851739e-05,
      "loss": 3.2075,
      "step": 82100
    },
    {
      "epoch": 8.84189000107631,
      "grad_norm": 0.8623098731040955,
      "learning_rate": 6.976834392845599e-05,
      "loss": 3.189,
      "step": 82150
    },
    {
      "epoch": 8.847271553115919,
      "grad_norm": 0.8621864318847656,
      "learning_rate": 6.944510289839457e-05,
      "loss": 3.1859,
      "step": 82200
    },
    {
      "epoch": 8.852653105155527,
      "grad_norm": 0.8130858540534973,
      "learning_rate": 6.912186186833315e-05,
      "loss": 3.1817,
      "step": 82250
    },
    {
      "epoch": 8.858034657195136,
      "grad_norm": 0.9033706188201904,
      "learning_rate": 6.879862083827173e-05,
      "loss": 3.1834,
      "step": 82300
    },
    {
      "epoch": 8.863416209234742,
      "grad_norm": 0.8645665645599365,
      "learning_rate": 6.847537980821031e-05,
      "loss": 3.1939,
      "step": 82350
    },
    {
      "epoch": 8.868797761274351,
      "grad_norm": 0.8805700540542603,
      "learning_rate": 6.81521387781489e-05,
      "loss": 3.1944,
      "step": 82400
    },
    {
      "epoch": 8.87417931331396,
      "grad_norm": 0.8313035368919373,
      "learning_rate": 6.783536256868871e-05,
      "loss": 3.1976,
      "step": 82450
    },
    {
      "epoch": 8.879560865353568,
      "grad_norm": 0.854358971118927,
      "learning_rate": 6.75121215386273e-05,
      "loss": 3.1921,
      "step": 82500
    },
    {
      "epoch": 8.884942417393177,
      "grad_norm": 0.8628531098365784,
      "learning_rate": 6.718888050856589e-05,
      "loss": 3.1886,
      "step": 82550
    },
    {
      "epoch": 8.890323969432785,
      "grad_norm": 0.8644590973854065,
      "learning_rate": 6.686563947850447e-05,
      "loss": 3.194,
      "step": 82600
    },
    {
      "epoch": 8.895705521472392,
      "grad_norm": 0.8778955340385437,
      "learning_rate": 6.654239844844305e-05,
      "loss": 3.1841,
      "step": 82650
    },
    {
      "epoch": 8.901087073512,
      "grad_norm": 0.8454921245574951,
      "learning_rate": 6.621915741838163e-05,
      "loss": 3.1821,
      "step": 82700
    },
    {
      "epoch": 8.906468625551609,
      "grad_norm": 0.8296894431114197,
      "learning_rate": 6.589591638832021e-05,
      "loss": 3.1843,
      "step": 82750
    },
    {
      "epoch": 8.911850177591218,
      "grad_norm": 0.8557996153831482,
      "learning_rate": 6.557267535825881e-05,
      "loss": 3.2012,
      "step": 82800
    },
    {
      "epoch": 8.917231729630826,
      "grad_norm": 0.8527143001556396,
      "learning_rate": 6.524943432819739e-05,
      "loss": 3.1831,
      "step": 82850
    },
    {
      "epoch": 8.922613281670433,
      "grad_norm": 0.8428352475166321,
      "learning_rate": 6.492619329813597e-05,
      "loss": 3.1896,
      "step": 82900
    },
    {
      "epoch": 8.927994833710041,
      "grad_norm": 0.8830580711364746,
      "learning_rate": 6.460295226807455e-05,
      "loss": 3.1677,
      "step": 82950
    },
    {
      "epoch": 8.93337638574965,
      "grad_norm": 0.8274183869361877,
      "learning_rate": 6.427971123801313e-05,
      "loss": 3.1848,
      "step": 83000
    },
    {
      "epoch": 8.93337638574965,
      "eval_accuracy": 0.3923523143771348,
      "eval_loss": 3.3143954277038574,
      "eval_runtime": 184.8625,
      "eval_samples_per_second": 97.429,
      "eval_steps_per_second": 6.091,
      "step": 83000
    },
    {
      "epoch": 8.938757937789259,
      "grad_norm": 0.8386024832725525,
      "learning_rate": 6.395647020795173e-05,
      "loss": 3.1778,
      "step": 83050
    },
    {
      "epoch": 8.944139489828867,
      "grad_norm": 0.887488067150116,
      "learning_rate": 6.363322917789031e-05,
      "loss": 3.1879,
      "step": 83100
    },
    {
      "epoch": 8.949521041868476,
      "grad_norm": 0.8449468612670898,
      "learning_rate": 6.330998814782889e-05,
      "loss": 3.1816,
      "step": 83150
    },
    {
      "epoch": 8.954902593908082,
      "grad_norm": 0.8413586616516113,
      "learning_rate": 6.298674711776748e-05,
      "loss": 3.1822,
      "step": 83200
    },
    {
      "epoch": 8.960284145947691,
      "grad_norm": 0.8581786155700684,
      "learning_rate": 6.266350608770606e-05,
      "loss": 3.1807,
      "step": 83250
    },
    {
      "epoch": 8.9656656979873,
      "grad_norm": 0.8413141965866089,
      "learning_rate": 6.234026505764465e-05,
      "loss": 3.1761,
      "step": 83300
    },
    {
      "epoch": 8.971047250026908,
      "grad_norm": 0.860555112361908,
      "learning_rate": 6.201702402758323e-05,
      "loss": 3.1895,
      "step": 83350
    },
    {
      "epoch": 8.976428802066517,
      "grad_norm": 0.8005810379981995,
      "learning_rate": 6.169378299752181e-05,
      "loss": 3.1865,
      "step": 83400
    },
    {
      "epoch": 8.981810354106123,
      "grad_norm": 0.8426989316940308,
      "learning_rate": 6.13705419674604e-05,
      "loss": 3.1966,
      "step": 83450
    },
    {
      "epoch": 8.987191906145732,
      "grad_norm": 0.8380297422409058,
      "learning_rate": 6.104730093739898e-05,
      "loss": 3.1913,
      "step": 83500
    },
    {
      "epoch": 8.99257345818534,
      "grad_norm": 0.8661143779754639,
      "learning_rate": 6.072405990733757e-05,
      "loss": 3.1772,
      "step": 83550
    },
    {
      "epoch": 8.997955010224949,
      "grad_norm": 0.8457013368606567,
      "learning_rate": 6.040081887727615e-05,
      "loss": 3.1762,
      "step": 83600
    },
    {
      "epoch": 9.003336562264558,
      "grad_norm": 0.8705750107765198,
      "learning_rate": 6.007757784721473e-05,
      "loss": 3.1515,
      "step": 83650
    },
    {
      "epoch": 9.008718114304166,
      "grad_norm": 0.7868455648422241,
      "learning_rate": 5.975433681715332e-05,
      "loss": 3.1213,
      "step": 83700
    },
    {
      "epoch": 9.014099666343773,
      "grad_norm": 0.8370267152786255,
      "learning_rate": 5.94310957870919e-05,
      "loss": 3.1228,
      "step": 83750
    },
    {
      "epoch": 9.019481218383381,
      "grad_norm": 0.8520972728729248,
      "learning_rate": 5.910785475703049e-05,
      "loss": 3.1279,
      "step": 83800
    },
    {
      "epoch": 9.02486277042299,
      "grad_norm": 0.8587152361869812,
      "learning_rate": 5.878461372696907e-05,
      "loss": 3.1198,
      "step": 83850
    },
    {
      "epoch": 9.030244322462599,
      "grad_norm": 0.9143362045288086,
      "learning_rate": 5.846137269690765e-05,
      "loss": 3.1163,
      "step": 83900
    },
    {
      "epoch": 9.035625874502207,
      "grad_norm": 0.8209463953971863,
      "learning_rate": 5.813813166684624e-05,
      "loss": 3.1171,
      "step": 83950
    },
    {
      "epoch": 9.041007426541814,
      "grad_norm": 0.8472164869308472,
      "learning_rate": 5.781489063678482e-05,
      "loss": 3.1365,
      "step": 84000
    },
    {
      "epoch": 9.041007426541814,
      "eval_accuracy": 0.3923821939212838,
      "eval_loss": 3.3175294399261475,
      "eval_runtime": 184.9515,
      "eval_samples_per_second": 97.382,
      "eval_steps_per_second": 6.088,
      "step": 84000
    },
    {
      "epoch": 9.046388978581422,
      "grad_norm": 0.8457962274551392,
      "learning_rate": 5.749164960672341e-05,
      "loss": 3.1368,
      "step": 84050
    },
    {
      "epoch": 9.051770530621031,
      "grad_norm": 0.8350469470024109,
      "learning_rate": 5.716840857666199e-05,
      "loss": 3.1047,
      "step": 84100
    },
    {
      "epoch": 9.05715208266064,
      "grad_norm": 0.8469880819320679,
      "learning_rate": 5.684516754660057e-05,
      "loss": 3.1273,
      "step": 84150
    },
    {
      "epoch": 9.062533634700248,
      "grad_norm": 0.8550379276275635,
      "learning_rate": 5.652192651653916e-05,
      "loss": 3.1376,
      "step": 84200
    },
    {
      "epoch": 9.067915186739857,
      "grad_norm": 0.8022764921188354,
      "learning_rate": 5.619868548647774e-05,
      "loss": 3.1234,
      "step": 84250
    },
    {
      "epoch": 9.073296738779463,
      "grad_norm": 0.8484096527099609,
      "learning_rate": 5.5875444456416335e-05,
      "loss": 3.1086,
      "step": 84300
    },
    {
      "epoch": 9.078678290819072,
      "grad_norm": 0.891518771648407,
      "learning_rate": 5.5552203426354916e-05,
      "loss": 3.1393,
      "step": 84350
    },
    {
      "epoch": 9.08405984285868,
      "grad_norm": 0.8611201047897339,
      "learning_rate": 5.52289623962935e-05,
      "loss": 3.1161,
      "step": 84400
    },
    {
      "epoch": 9.089441394898289,
      "grad_norm": 0.8376824855804443,
      "learning_rate": 5.4905721366232085e-05,
      "loss": 3.1187,
      "step": 84450
    },
    {
      "epoch": 9.094822946937898,
      "grad_norm": 0.88045734167099,
      "learning_rate": 5.4582480336170666e-05,
      "loss": 3.1401,
      "step": 84500
    },
    {
      "epoch": 9.100204498977504,
      "grad_norm": 0.8772515058517456,
      "learning_rate": 5.426570412671048e-05,
      "loss": 3.129,
      "step": 84550
    },
    {
      "epoch": 9.105586051017113,
      "grad_norm": 0.8910012245178223,
      "learning_rate": 5.394246309664907e-05,
      "loss": 3.1328,
      "step": 84600
    },
    {
      "epoch": 9.110967603056721,
      "grad_norm": 0.8341832756996155,
      "learning_rate": 5.361922206658765e-05,
      "loss": 3.1367,
      "step": 84650
    },
    {
      "epoch": 9.11634915509633,
      "grad_norm": 0.8946830630302429,
      "learning_rate": 5.329598103652623e-05,
      "loss": 3.1334,
      "step": 84700
    },
    {
      "epoch": 9.121730707135939,
      "grad_norm": 0.8773099780082703,
      "learning_rate": 5.297274000646482e-05,
      "loss": 3.1358,
      "step": 84750
    },
    {
      "epoch": 9.127112259175545,
      "grad_norm": 0.8716827630996704,
      "learning_rate": 5.26494989764034e-05,
      "loss": 3.1283,
      "step": 84800
    },
    {
      "epoch": 9.132493811215154,
      "grad_norm": 0.8345746397972107,
      "learning_rate": 5.232625794634199e-05,
      "loss": 3.1305,
      "step": 84850
    },
    {
      "epoch": 9.137875363254762,
      "grad_norm": 0.8926873803138733,
      "learning_rate": 5.200301691628057e-05,
      "loss": 3.1288,
      "step": 84900
    },
    {
      "epoch": 9.143256915294371,
      "grad_norm": 0.8594902753829956,
      "learning_rate": 5.167977588621915e-05,
      "loss": 3.1269,
      "step": 84950
    },
    {
      "epoch": 9.14863846733398,
      "grad_norm": 0.8748249411582947,
      "learning_rate": 5.135653485615774e-05,
      "loss": 3.1278,
      "step": 85000
    },
    {
      "epoch": 9.14863846733398,
      "eval_accuracy": 0.39252116096479855,
      "eval_loss": 3.3163113594055176,
      "eval_runtime": 184.5419,
      "eval_samples_per_second": 97.598,
      "eval_steps_per_second": 6.102,
      "step": 85000
    },
    {
      "epoch": 9.154020019373588,
      "grad_norm": 0.8709080815315247,
      "learning_rate": 5.103329382609632e-05,
      "loss": 3.1363,
      "step": 85050
    },
    {
      "epoch": 9.159401571413195,
      "grad_norm": 0.8843692541122437,
      "learning_rate": 5.0710052796034906e-05,
      "loss": 3.1386,
      "step": 85100
    },
    {
      "epoch": 9.164783123452803,
      "grad_norm": 0.9347819089889526,
      "learning_rate": 5.038681176597349e-05,
      "loss": 3.1503,
      "step": 85150
    },
    {
      "epoch": 9.170164675492412,
      "grad_norm": 0.8654942512512207,
      "learning_rate": 5.006357073591207e-05,
      "loss": 3.1261,
      "step": 85200
    },
    {
      "epoch": 9.17554622753202,
      "grad_norm": 0.906558096408844,
      "learning_rate": 4.974032970585066e-05,
      "loss": 3.1424,
      "step": 85250
    },
    {
      "epoch": 9.180927779571629,
      "grad_norm": 0.8706866502761841,
      "learning_rate": 4.9417088675789244e-05,
      "loss": 3.1276,
      "step": 85300
    },
    {
      "epoch": 9.186309331611236,
      "grad_norm": 0.8573899269104004,
      "learning_rate": 4.909384764572783e-05,
      "loss": 3.1404,
      "step": 85350
    },
    {
      "epoch": 9.191690883650844,
      "grad_norm": 0.8437527418136597,
      "learning_rate": 4.877060661566641e-05,
      "loss": 3.1462,
      "step": 85400
    },
    {
      "epoch": 9.197072435690453,
      "grad_norm": 0.8533856868743896,
      "learning_rate": 4.8447365585604994e-05,
      "loss": 3.1523,
      "step": 85450
    },
    {
      "epoch": 9.202453987730062,
      "grad_norm": 0.8516963124275208,
      "learning_rate": 4.812412455554358e-05,
      "loss": 3.1247,
      "step": 85500
    },
    {
      "epoch": 9.20783553976967,
      "grad_norm": 0.8821370005607605,
      "learning_rate": 4.780088352548216e-05,
      "loss": 3.123,
      "step": 85550
    },
    {
      "epoch": 9.213217091809279,
      "grad_norm": 0.9131876230239868,
      "learning_rate": 4.747764249542075e-05,
      "loss": 3.137,
      "step": 85600
    },
    {
      "epoch": 9.218598643848885,
      "grad_norm": 0.8451762199401855,
      "learning_rate": 4.715440146535933e-05,
      "loss": 3.1243,
      "step": 85650
    },
    {
      "epoch": 9.223980195888494,
      "grad_norm": 0.8500200510025024,
      "learning_rate": 4.683116043529791e-05,
      "loss": 3.1301,
      "step": 85700
    },
    {
      "epoch": 9.229361747928102,
      "grad_norm": 0.8670781850814819,
      "learning_rate": 4.65079194052365e-05,
      "loss": 3.1391,
      "step": 85750
    },
    {
      "epoch": 9.234743299967711,
      "grad_norm": 0.8745872974395752,
      "learning_rate": 4.618467837517508e-05,
      "loss": 3.1359,
      "step": 85800
    },
    {
      "epoch": 9.24012485200732,
      "grad_norm": 0.8826162815093994,
      "learning_rate": 4.586143734511367e-05,
      "loss": 3.1217,
      "step": 85850
    },
    {
      "epoch": 9.245506404046926,
      "grad_norm": 0.8553646802902222,
      "learning_rate": 4.553819631505225e-05,
      "loss": 3.1452,
      "step": 85900
    },
    {
      "epoch": 9.250887956086535,
      "grad_norm": 0.8691889047622681,
      "learning_rate": 4.521495528499083e-05,
      "loss": 3.1509,
      "step": 85950
    },
    {
      "epoch": 9.256269508126143,
      "grad_norm": 0.8630021810531616,
      "learning_rate": 4.489171425492942e-05,
      "loss": 3.1425,
      "step": 86000
    },
    {
      "epoch": 9.256269508126143,
      "eval_accuracy": 0.39285896279301374,
      "eval_loss": 3.3147192001342773,
      "eval_runtime": 184.6967,
      "eval_samples_per_second": 97.517,
      "eval_steps_per_second": 6.096,
      "step": 86000
    },
    {
      "epoch": 9.261651060165752,
      "grad_norm": 0.890877366065979,
      "learning_rate": 4.4568473224868e-05,
      "loss": 3.1497,
      "step": 86050
    },
    {
      "epoch": 9.26703261220536,
      "grad_norm": 1.0040676593780518,
      "learning_rate": 4.4245232194806596e-05,
      "loss": 3.1154,
      "step": 86100
    },
    {
      "epoch": 9.272414164244967,
      "grad_norm": 0.8417730927467346,
      "learning_rate": 4.392199116474518e-05,
      "loss": 3.1267,
      "step": 86150
    },
    {
      "epoch": 9.277795716284576,
      "grad_norm": 0.8883495330810547,
      "learning_rate": 4.359875013468376e-05,
      "loss": 3.1384,
      "step": 86200
    },
    {
      "epoch": 9.283177268324184,
      "grad_norm": 0.8868676424026489,
      "learning_rate": 4.3275509104622346e-05,
      "loss": 3.1315,
      "step": 86250
    },
    {
      "epoch": 9.288558820363793,
      "grad_norm": 0.8812539577484131,
      "learning_rate": 4.295226807456093e-05,
      "loss": 3.1479,
      "step": 86300
    },
    {
      "epoch": 9.293940372403402,
      "grad_norm": 0.846923291683197,
      "learning_rate": 4.2629027044499515e-05,
      "loss": 3.1389,
      "step": 86350
    },
    {
      "epoch": 9.29932192444301,
      "grad_norm": 0.8891094923019409,
      "learning_rate": 4.2305786014438096e-05,
      "loss": 3.1262,
      "step": 86400
    },
    {
      "epoch": 9.304703476482617,
      "grad_norm": 0.8506091833114624,
      "learning_rate": 4.198254498437668e-05,
      "loss": 3.1405,
      "step": 86450
    },
    {
      "epoch": 9.310085028522225,
      "grad_norm": 0.884657621383667,
      "learning_rate": 4.1659303954315265e-05,
      "loss": 3.1399,
      "step": 86500
    },
    {
      "epoch": 9.315466580561834,
      "grad_norm": 0.8441677093505859,
      "learning_rate": 4.134252774485508e-05,
      "loss": 3.1429,
      "step": 86550
    },
    {
      "epoch": 9.320848132601443,
      "grad_norm": 0.8587312698364258,
      "learning_rate": 4.101928671479366e-05,
      "loss": 3.1336,
      "step": 86600
    },
    {
      "epoch": 9.326229684641051,
      "grad_norm": 0.8646252751350403,
      "learning_rate": 4.069604568473225e-05,
      "loss": 3.1336,
      "step": 86650
    },
    {
      "epoch": 9.331611236680658,
      "grad_norm": 0.8256805539131165,
      "learning_rate": 4.037280465467083e-05,
      "loss": 3.1314,
      "step": 86700
    },
    {
      "epoch": 9.336992788720266,
      "grad_norm": 0.8126273155212402,
      "learning_rate": 4.004956362460941e-05,
      "loss": 3.1356,
      "step": 86750
    },
    {
      "epoch": 9.342374340759875,
      "grad_norm": 0.8811523914337158,
      "learning_rate": 3.9726322594548e-05,
      "loss": 3.1268,
      "step": 86800
    },
    {
      "epoch": 9.347755892799483,
      "grad_norm": 0.9037714004516602,
      "learning_rate": 3.940308156448658e-05,
      "loss": 3.1267,
      "step": 86850
    },
    {
      "epoch": 9.353137444839092,
      "grad_norm": 0.8573242425918579,
      "learning_rate": 3.9079840534425166e-05,
      "loss": 3.1561,
      "step": 86900
    },
    {
      "epoch": 9.3585189968787,
      "grad_norm": 0.8658702373504639,
      "learning_rate": 3.875659950436375e-05,
      "loss": 3.1303,
      "step": 86950
    },
    {
      "epoch": 9.363900548918307,
      "grad_norm": 0.8456016182899475,
      "learning_rate": 3.843335847430233e-05,
      "loss": 3.1378,
      "step": 87000
    },
    {
      "epoch": 9.363900548918307,
      "eval_accuracy": 0.3931422208715461,
      "eval_loss": 3.3129429817199707,
      "eval_runtime": 184.8161,
      "eval_samples_per_second": 97.454,
      "eval_steps_per_second": 6.093,
      "step": 87000
    },
    {
      "epoch": 9.369282100957916,
      "grad_norm": 0.850412905216217,
      "learning_rate": 3.811011744424092e-05,
      "loss": 3.1402,
      "step": 87050
    },
    {
      "epoch": 9.374663652997524,
      "grad_norm": 0.8742541074752808,
      "learning_rate": 3.7786876414179504e-05,
      "loss": 3.1296,
      "step": 87100
    },
    {
      "epoch": 9.380045205037133,
      "grad_norm": 0.8422418236732483,
      "learning_rate": 3.7463635384118085e-05,
      "loss": 3.1395,
      "step": 87150
    },
    {
      "epoch": 9.385426757076742,
      "grad_norm": 0.8755067586898804,
      "learning_rate": 3.714039435405667e-05,
      "loss": 3.1312,
      "step": 87200
    },
    {
      "epoch": 9.390808309116348,
      "grad_norm": 0.8353830575942993,
      "learning_rate": 3.6817153323995254e-05,
      "loss": 3.1413,
      "step": 87250
    },
    {
      "epoch": 9.396189861155957,
      "grad_norm": 0.8410850167274475,
      "learning_rate": 3.649391229393384e-05,
      "loss": 3.1286,
      "step": 87300
    },
    {
      "epoch": 9.401571413195565,
      "grad_norm": 0.8951404690742493,
      "learning_rate": 3.617067126387242e-05,
      "loss": 3.1358,
      "step": 87350
    },
    {
      "epoch": 9.406952965235174,
      "grad_norm": 0.8594951629638672,
      "learning_rate": 3.5847430233811004e-05,
      "loss": 3.1352,
      "step": 87400
    },
    {
      "epoch": 9.412334517274783,
      "grad_norm": 0.8391014337539673,
      "learning_rate": 3.552418920374959e-05,
      "loss": 3.1471,
      "step": 87450
    },
    {
      "epoch": 9.417716069314391,
      "grad_norm": 0.8913040161132812,
      "learning_rate": 3.520094817368818e-05,
      "loss": 3.1331,
      "step": 87500
    },
    {
      "epoch": 9.423097621353998,
      "grad_norm": 0.8500414490699768,
      "learning_rate": 3.487770714362676e-05,
      "loss": 3.1476,
      "step": 87550
    },
    {
      "epoch": 9.428479173393606,
      "grad_norm": 0.8843672871589661,
      "learning_rate": 3.455446611356535e-05,
      "loss": 3.1498,
      "step": 87600
    },
    {
      "epoch": 9.433860725433215,
      "grad_norm": 0.8870358467102051,
      "learning_rate": 3.423122508350393e-05,
      "loss": 3.1357,
      "step": 87650
    },
    {
      "epoch": 9.439242277472824,
      "grad_norm": 0.8403342962265015,
      "learning_rate": 3.390798405344251e-05,
      "loss": 3.1628,
      "step": 87700
    },
    {
      "epoch": 9.444623829512432,
      "grad_norm": 0.8929362893104553,
      "learning_rate": 3.35847430233811e-05,
      "loss": 3.1288,
      "step": 87750
    },
    {
      "epoch": 9.450005381552039,
      "grad_norm": 0.8921561241149902,
      "learning_rate": 3.326150199331968e-05,
      "loss": 3.1232,
      "step": 87800
    },
    {
      "epoch": 9.455386933591647,
      "grad_norm": 0.8558990359306335,
      "learning_rate": 3.293826096325827e-05,
      "loss": 3.141,
      "step": 87850
    },
    {
      "epoch": 9.460768485631256,
      "grad_norm": 0.8367136716842651,
      "learning_rate": 3.261501993319685e-05,
      "loss": 3.1338,
      "step": 87900
    },
    {
      "epoch": 9.466150037670864,
      "grad_norm": 0.8506068587303162,
      "learning_rate": 3.229177890313544e-05,
      "loss": 3.1362,
      "step": 87950
    },
    {
      "epoch": 9.471531589710473,
      "grad_norm": 0.8316249847412109,
      "learning_rate": 3.196853787307402e-05,
      "loss": 3.1518,
      "step": 88000
    },
    {
      "epoch": 9.471531589710473,
      "eval_accuracy": 0.3935000148311192,
      "eval_loss": 3.3096585273742676,
      "eval_runtime": 184.9229,
      "eval_samples_per_second": 97.397,
      "eval_steps_per_second": 6.089,
      "step": 88000
    },
    {
      "epoch": 9.476913141750082,
      "grad_norm": 0.8826675415039062,
      "learning_rate": 3.1645296843012606e-05,
      "loss": 3.1324,
      "step": 88050
    },
    {
      "epoch": 9.482294693789688,
      "grad_norm": 0.8648707270622253,
      "learning_rate": 3.132205581295119e-05,
      "loss": 3.1291,
      "step": 88100
    },
    {
      "epoch": 9.487676245829297,
      "grad_norm": 0.8441886305809021,
      "learning_rate": 3.099881478288977e-05,
      "loss": 3.1342,
      "step": 88150
    },
    {
      "epoch": 9.493057797868905,
      "grad_norm": 0.8682032227516174,
      "learning_rate": 3.0675573752828356e-05,
      "loss": 3.1322,
      "step": 88200
    },
    {
      "epoch": 9.498439349908514,
      "grad_norm": 0.8759601712226868,
      "learning_rate": 3.035233272276694e-05,
      "loss": 3.147,
      "step": 88250
    },
    {
      "epoch": 9.503820901948123,
      "grad_norm": 0.8368282318115234,
      "learning_rate": 3.0029091692705525e-05,
      "loss": 3.1305,
      "step": 88300
    },
    {
      "epoch": 9.50920245398773,
      "grad_norm": 0.8676648736000061,
      "learning_rate": 2.970585066264411e-05,
      "loss": 3.1422,
      "step": 88350
    },
    {
      "epoch": 9.514584006027338,
      "grad_norm": 0.8752458095550537,
      "learning_rate": 2.938260963258269e-05,
      "loss": 3.1359,
      "step": 88400
    },
    {
      "epoch": 9.519965558066946,
      "grad_norm": 0.8628907799720764,
      "learning_rate": 2.9059368602521275e-05,
      "loss": 3.1378,
      "step": 88450
    },
    {
      "epoch": 9.525347110106555,
      "grad_norm": 0.9217661619186401,
      "learning_rate": 2.8736127572459863e-05,
      "loss": 3.1173,
      "step": 88500
    },
    {
      "epoch": 9.530728662146164,
      "grad_norm": 0.8665773868560791,
      "learning_rate": 2.8419351362999673e-05,
      "loss": 3.1332,
      "step": 88550
    },
    {
      "epoch": 9.536110214185772,
      "grad_norm": 0.8887430429458618,
      "learning_rate": 2.809611033293826e-05,
      "loss": 3.1445,
      "step": 88600
    },
    {
      "epoch": 9.541491766225379,
      "grad_norm": 0.8543220162391663,
      "learning_rate": 2.7772869302876845e-05,
      "loss": 3.1466,
      "step": 88650
    },
    {
      "epoch": 9.546873318264987,
      "grad_norm": 0.8621816039085388,
      "learning_rate": 2.7449628272815427e-05,
      "loss": 3.1472,
      "step": 88700
    },
    {
      "epoch": 9.552254870304596,
      "grad_norm": 0.8645244836807251,
      "learning_rate": 2.712638724275401e-05,
      "loss": 3.1459,
      "step": 88750
    },
    {
      "epoch": 9.557636422344205,
      "grad_norm": 0.8483282923698425,
      "learning_rate": 2.6803146212692596e-05,
      "loss": 3.1479,
      "step": 88800
    },
    {
      "epoch": 9.563017974383813,
      "grad_norm": 0.8576905727386475,
      "learning_rate": 2.647990518263118e-05,
      "loss": 3.1474,
      "step": 88850
    },
    {
      "epoch": 9.56839952642342,
      "grad_norm": 0.8047739863395691,
      "learning_rate": 2.6156664152569765e-05,
      "loss": 3.1341,
      "step": 88900
    },
    {
      "epoch": 9.573781078463028,
      "grad_norm": 0.895772397518158,
      "learning_rate": 2.5833423122508346e-05,
      "loss": 3.1356,
      "step": 88950
    },
    {
      "epoch": 9.579162630502637,
      "grad_norm": 0.8613327741622925,
      "learning_rate": 2.5510182092446934e-05,
      "loss": 3.1506,
      "step": 89000
    },
    {
      "epoch": 9.579162630502637,
      "eval_accuracy": 0.3937628461667424,
      "eval_loss": 3.3074944019317627,
      "eval_runtime": 184.351,
      "eval_samples_per_second": 97.7,
      "eval_steps_per_second": 6.108,
      "step": 89000
    },
    {
      "epoch": 9.584544182542245,
      "grad_norm": 0.9347299337387085,
      "learning_rate": 2.5186941062385518e-05,
      "loss": 3.1305,
      "step": 89050
    },
    {
      "epoch": 9.589925734581854,
      "grad_norm": 0.8868260383605957,
      "learning_rate": 2.4863700032324103e-05,
      "loss": 3.1347,
      "step": 89100
    },
    {
      "epoch": 9.59530728662146,
      "grad_norm": 0.8855026364326477,
      "learning_rate": 2.4540459002262687e-05,
      "loss": 3.1357,
      "step": 89150
    },
    {
      "epoch": 9.60068883866107,
      "grad_norm": 0.8583725690841675,
      "learning_rate": 2.4217217972201268e-05,
      "loss": 3.1468,
      "step": 89200
    },
    {
      "epoch": 9.606070390700678,
      "grad_norm": 0.9050620198249817,
      "learning_rate": 2.3893976942139853e-05,
      "loss": 3.1399,
      "step": 89250
    },
    {
      "epoch": 9.611451942740286,
      "grad_norm": 0.8982083201408386,
      "learning_rate": 2.3570735912078437e-05,
      "loss": 3.1416,
      "step": 89300
    },
    {
      "epoch": 9.616833494779895,
      "grad_norm": 0.8842360377311707,
      "learning_rate": 2.324749488201702e-05,
      "loss": 3.144,
      "step": 89350
    },
    {
      "epoch": 9.622215046819504,
      "grad_norm": 0.868691086769104,
      "learning_rate": 2.292425385195561e-05,
      "loss": 3.1363,
      "step": 89400
    },
    {
      "epoch": 9.62759659885911,
      "grad_norm": 0.8905691504478455,
      "learning_rate": 2.260101282189419e-05,
      "loss": 3.126,
      "step": 89450
    },
    {
      "epoch": 9.632978150898719,
      "grad_norm": 0.8737261295318604,
      "learning_rate": 2.2277771791832775e-05,
      "loss": 3.1435,
      "step": 89500
    },
    {
      "epoch": 9.638359702938327,
      "grad_norm": 0.8221793174743652,
      "learning_rate": 2.195453076177136e-05,
      "loss": 3.1392,
      "step": 89550
    },
    {
      "epoch": 9.643741254977936,
      "grad_norm": 0.9344587922096252,
      "learning_rate": 2.1631289731709944e-05,
      "loss": 3.1441,
      "step": 89600
    },
    {
      "epoch": 9.649122807017545,
      "grad_norm": 0.8708004355430603,
      "learning_rate": 2.130804870164853e-05,
      "loss": 3.1346,
      "step": 89650
    },
    {
      "epoch": 9.654504359057151,
      "grad_norm": 0.8501155376434326,
      "learning_rate": 2.098480767158711e-05,
      "loss": 3.1295,
      "step": 89700
    },
    {
      "epoch": 9.65988591109676,
      "grad_norm": 0.850181519985199,
      "learning_rate": 2.0661566641525694e-05,
      "loss": 3.1435,
      "step": 89750
    },
    {
      "epoch": 9.665267463136368,
      "grad_norm": 0.8723414540290833,
      "learning_rate": 2.0338325611464282e-05,
      "loss": 3.1391,
      "step": 89800
    },
    {
      "epoch": 9.670649015175977,
      "grad_norm": 0.8618516325950623,
      "learning_rate": 2.0015084581402867e-05,
      "loss": 3.127,
      "step": 89850
    },
    {
      "epoch": 9.676030567215586,
      "grad_norm": 0.8360557556152344,
      "learning_rate": 1.969184355134145e-05,
      "loss": 3.1378,
      "step": 89900
    },
    {
      "epoch": 9.681412119255192,
      "grad_norm": 0.8770431876182556,
      "learning_rate": 1.9368602521280032e-05,
      "loss": 3.1286,
      "step": 89950
    },
    {
      "epoch": 9.6867936712948,
      "grad_norm": 0.8881701827049255,
      "learning_rate": 1.9045361491218617e-05,
      "loss": 3.1299,
      "step": 90000
    },
    {
      "epoch": 9.6867936712948,
      "eval_accuracy": 0.3939788481077176,
      "eval_loss": 3.305659294128418,
      "eval_runtime": 184.7566,
      "eval_samples_per_second": 97.485,
      "eval_steps_per_second": 6.095,
      "step": 90000
    },
    {
      "epoch": 9.69217522333441,
      "grad_norm": 0.8480678796768188,
      "learning_rate": 1.87221204611572e-05,
      "loss": 3.1307,
      "step": 90050
    },
    {
      "epoch": 9.697556775374018,
      "grad_norm": 0.8607448935508728,
      "learning_rate": 1.8398879431095786e-05,
      "loss": 3.1351,
      "step": 90100
    },
    {
      "epoch": 9.702938327413626,
      "grad_norm": 0.8670057058334351,
      "learning_rate": 1.807563840103437e-05,
      "loss": 3.1457,
      "step": 90150
    },
    {
      "epoch": 9.708319879453235,
      "grad_norm": 0.8635824918746948,
      "learning_rate": 1.7752397370972955e-05,
      "loss": 3.1417,
      "step": 90200
    },
    {
      "epoch": 9.713701431492842,
      "grad_norm": 0.878579318523407,
      "learning_rate": 1.742915634091154e-05,
      "loss": 3.1265,
      "step": 90250
    },
    {
      "epoch": 9.71908298353245,
      "grad_norm": 0.8604105710983276,
      "learning_rate": 1.7105915310850124e-05,
      "loss": 3.1386,
      "step": 90300
    },
    {
      "epoch": 9.724464535572059,
      "grad_norm": 0.8821996450424194,
      "learning_rate": 1.6782674280788708e-05,
      "loss": 3.1368,
      "step": 90350
    },
    {
      "epoch": 9.729846087611667,
      "grad_norm": 0.8872800469398499,
      "learning_rate": 1.645943325072729e-05,
      "loss": 3.1367,
      "step": 90400
    },
    {
      "epoch": 9.735227639651276,
      "grad_norm": 0.879416286945343,
      "learning_rate": 1.6136192220665877e-05,
      "loss": 3.1379,
      "step": 90450
    },
    {
      "epoch": 9.740609191690883,
      "grad_norm": 0.862957775592804,
      "learning_rate": 1.5812951190604458e-05,
      "loss": 3.1519,
      "step": 90500
    },
    {
      "epoch": 9.745990743730491,
      "grad_norm": 0.8256456851959229,
      "learning_rate": 1.5489710160543043e-05,
      "loss": 3.1452,
      "step": 90550
    },
    {
      "epoch": 9.7513722957701,
      "grad_norm": 0.8845705389976501,
      "learning_rate": 1.5172933951082856e-05,
      "loss": 3.1252,
      "step": 90600
    },
    {
      "epoch": 9.756753847809708,
      "grad_norm": 0.8848972916603088,
      "learning_rate": 1.4849692921021439e-05,
      "loss": 3.1397,
      "step": 90650
    },
    {
      "epoch": 9.762135399849317,
      "grad_norm": 0.8965819478034973,
      "learning_rate": 1.4526451890960025e-05,
      "loss": 3.144,
      "step": 90700
    },
    {
      "epoch": 9.767516951888926,
      "grad_norm": 0.8574418425559998,
      "learning_rate": 1.420321086089861e-05,
      "loss": 3.1394,
      "step": 90750
    },
    {
      "epoch": 9.772898503928532,
      "grad_norm": 0.8599215149879456,
      "learning_rate": 1.3879969830837192e-05,
      "loss": 3.1226,
      "step": 90800
    },
    {
      "epoch": 9.77828005596814,
      "grad_norm": 1.027356743812561,
      "learning_rate": 1.3556728800775778e-05,
      "loss": 3.1412,
      "step": 90850
    },
    {
      "epoch": 9.78366160800775,
      "grad_norm": 0.8927180767059326,
      "learning_rate": 1.3233487770714361e-05,
      "loss": 3.1559,
      "step": 90900
    },
    {
      "epoch": 9.789043160047358,
      "grad_norm": 0.8471749424934387,
      "learning_rate": 1.2910246740652946e-05,
      "loss": 3.1254,
      "step": 90950
    },
    {
      "epoch": 9.794424712086967,
      "grad_norm": 0.9021408557891846,
      "learning_rate": 1.258700571059153e-05,
      "loss": 3.1302,
      "step": 91000
    },
    {
      "epoch": 9.794424712086967,
      "eval_accuracy": 0.3942114739405284,
      "eval_loss": 3.3036468029022217,
      "eval_runtime": 184.6554,
      "eval_samples_per_second": 97.538,
      "eval_steps_per_second": 6.098,
      "step": 91000
    },
    {
      "epoch": 9.799806264126573,
      "grad_norm": 0.8827714920043945,
      "learning_rate": 1.2263764680530113e-05,
      "loss": 3.1613,
      "step": 91050
    },
    {
      "epoch": 9.805187816166182,
      "grad_norm": 0.8419482707977295,
      "learning_rate": 1.19405236504687e-05,
      "loss": 3.1467,
      "step": 91100
    },
    {
      "epoch": 9.81056936820579,
      "grad_norm": 0.8355914950370789,
      "learning_rate": 1.1617282620407282e-05,
      "loss": 3.1314,
      "step": 91150
    },
    {
      "epoch": 9.815950920245399,
      "grad_norm": 0.8849267959594727,
      "learning_rate": 1.1294041590345867e-05,
      "loss": 3.1402,
      "step": 91200
    },
    {
      "epoch": 9.821332472285007,
      "grad_norm": 0.9032121896743774,
      "learning_rate": 1.0970800560284453e-05,
      "loss": 3.1429,
      "step": 91250
    },
    {
      "epoch": 9.826714024324616,
      "grad_norm": 0.8774483799934387,
      "learning_rate": 1.0647559530223035e-05,
      "loss": 3.1647,
      "step": 91300
    },
    {
      "epoch": 9.832095576364223,
      "grad_norm": 0.8687463998794556,
      "learning_rate": 1.032431850016162e-05,
      "loss": 3.1282,
      "step": 91350
    },
    {
      "epoch": 9.837477128403831,
      "grad_norm": 0.8409572839736938,
      "learning_rate": 1.0001077470100203e-05,
      "loss": 3.1399,
      "step": 91400
    },
    {
      "epoch": 9.84285868044344,
      "grad_norm": 0.8912094831466675,
      "learning_rate": 9.677836440038787e-06,
      "loss": 3.1383,
      "step": 91450
    },
    {
      "epoch": 9.848240232483048,
      "grad_norm": 0.8655765056610107,
      "learning_rate": 9.354595409977372e-06,
      "loss": 3.1459,
      "step": 91500
    },
    {
      "epoch": 9.853621784522657,
      "grad_norm": 0.8873117566108704,
      "learning_rate": 9.031354379915956e-06,
      "loss": 3.1471,
      "step": 91550
    },
    {
      "epoch": 9.859003336562264,
      "grad_norm": 0.8902177214622498,
      "learning_rate": 8.70811334985454e-06,
      "loss": 3.1333,
      "step": 91600
    },
    {
      "epoch": 9.864384888601872,
      "grad_norm": 0.8534007668495178,
      "learning_rate": 8.384872319793125e-06,
      "loss": 3.1252,
      "step": 91650
    },
    {
      "epoch": 9.869766440641481,
      "grad_norm": 0.8633430004119873,
      "learning_rate": 8.06163128973171e-06,
      "loss": 3.1333,
      "step": 91700
    },
    {
      "epoch": 9.87514799268109,
      "grad_norm": 0.8976386189460754,
      "learning_rate": 7.738390259670293e-06,
      "loss": 3.1314,
      "step": 91750
    },
    {
      "epoch": 9.880529544720698,
      "grad_norm": 0.9152008295059204,
      "learning_rate": 7.415149229608878e-06,
      "loss": 3.1542,
      "step": 91800
    },
    {
      "epoch": 9.885911096760307,
      "grad_norm": 0.8877971768379211,
      "learning_rate": 7.091908199547462e-06,
      "loss": 3.131,
      "step": 91850
    },
    {
      "epoch": 9.891292648799913,
      "grad_norm": 0.8462663292884827,
      "learning_rate": 6.768667169486046e-06,
      "loss": 3.1421,
      "step": 91900
    },
    {
      "epoch": 9.896674200839522,
      "grad_norm": 0.8960040211677551,
      "learning_rate": 6.4454261394246305e-06,
      "loss": 3.1398,
      "step": 91950
    },
    {
      "epoch": 9.90205575287913,
      "grad_norm": 0.8556128740310669,
      "learning_rate": 6.122185109363214e-06,
      "loss": 3.131,
      "step": 92000
    },
    {
      "epoch": 9.90205575287913,
      "eval_accuracy": 0.3943444650752133,
      "eval_loss": 3.30191969871521,
      "eval_runtime": 184.7232,
      "eval_samples_per_second": 97.503,
      "eval_steps_per_second": 6.096,
      "step": 92000
    },
    {
      "epoch": 9.907437304918739,
      "grad_norm": 0.8960968255996704,
      "learning_rate": 5.7989440793017995e-06,
      "loss": 3.1272,
      "step": 92050
    },
    {
      "epoch": 9.912818856958348,
      "grad_norm": 0.9231632947921753,
      "learning_rate": 5.475703049240383e-06,
      "loss": 3.1329,
      "step": 92100
    },
    {
      "epoch": 9.918200408997954,
      "grad_norm": 0.8854824900627136,
      "learning_rate": 5.152462019178968e-06,
      "loss": 3.1333,
      "step": 92150
    },
    {
      "epoch": 9.923581961037563,
      "grad_norm": 0.8688467741012573,
      "learning_rate": 4.829220989117551e-06,
      "loss": 3.1238,
      "step": 92200
    },
    {
      "epoch": 9.928963513077171,
      "grad_norm": 0.8522169589996338,
      "learning_rate": 4.505979959056136e-06,
      "loss": 3.1175,
      "step": 92250
    },
    {
      "epoch": 9.93434506511678,
      "grad_norm": 0.8435911536216736,
      "learning_rate": 4.18273892899472e-06,
      "loss": 3.151,
      "step": 92300
    },
    {
      "epoch": 9.939726617156388,
      "grad_norm": 0.8723344206809998,
      "learning_rate": 3.859497898933305e-06,
      "loss": 3.1187,
      "step": 92350
    },
    {
      "epoch": 9.945108169195997,
      "grad_norm": 0.8489755392074585,
      "learning_rate": 3.5362568688718884e-06,
      "loss": 3.1349,
      "step": 92400
    },
    {
      "epoch": 9.950489721235604,
      "grad_norm": 0.8661078810691833,
      "learning_rate": 3.2130158388104725e-06,
      "loss": 3.152,
      "step": 92450
    },
    {
      "epoch": 9.955871273275212,
      "grad_norm": 0.907784104347229,
      "learning_rate": 2.889774808749057e-06,
      "loss": 3.1343,
      "step": 92500
    },
    {
      "epoch": 9.961252825314821,
      "grad_norm": 1.4848933219909668,
      "learning_rate": 2.566533778687641e-06,
      "loss": 3.1212,
      "step": 92550
    },
    {
      "epoch": 9.96663437735443,
      "grad_norm": 0.883884608745575,
      "learning_rate": 2.249757569227454e-06,
      "loss": 3.1044,
      "step": 92600
    },
    {
      "epoch": 9.972015929394038,
      "grad_norm": 0.8627755641937256,
      "learning_rate": 1.926516539166038e-06,
      "loss": 3.1414,
      "step": 92650
    },
    {
      "epoch": 9.977397481433645,
      "grad_norm": 0.8736936450004578,
      "learning_rate": 1.603275509104622e-06,
      "loss": 3.1349,
      "step": 92700
    },
    {
      "epoch": 9.982779033473253,
      "grad_norm": 0.8741436004638672,
      "learning_rate": 1.2800344790432064e-06,
      "loss": 3.1541,
      "step": 92750
    },
    {
      "epoch": 9.988160585512862,
      "grad_norm": 0.8391033411026001,
      "learning_rate": 9.567934489817906e-07,
      "loss": 3.1273,
      "step": 92800
    },
    {
      "epoch": 9.99354213755247,
      "grad_norm": 0.9001469016075134,
      "learning_rate": 6.335524189203748e-07,
      "loss": 3.1254,
      "step": 92850
    },
    {
      "epoch": 9.998923689592079,
      "grad_norm": 0.8495814800262451,
      "learning_rate": 3.103113888589591e-07,
      "loss": 3.13,
      "step": 92900
    },
    {
      "epoch": 10.0,
      "step": 92910,
      "total_flos": 7.7681075945472e+17,
      "train_loss": 3.458772770117256,
      "train_runtime": 79806.4071,
      "train_samples_per_second": 37.252,
      "train_steps_per_second": 1.164
    }
  ],
  "logging_steps": 50,
  "max_steps": 92910,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.7681075945472e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}