{
  "best_metric": 3.301800489425659,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_6910/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": 12.853434562683105,
      "learning_rate": 0.0003,
      "loss": 8.5553,
      "step": 50
    },
    {
      "epoch": 0.010763104079216447,
      "grad_norm": 2.2888307571411133,
      "learning_rate": 0.0006,
      "loss": 6.9394,
      "step": 100
    },
    {
      "epoch": 0.01614465611882467,
      "grad_norm": 1.0520085096359253,
      "learning_rate": 0.0005996767589699385,
      "loss": 6.4987,
      "step": 150
    },
    {
      "epoch": 0.021526208158432893,
      "grad_norm": 2.049221992492676,
      "learning_rate": 0.0005993535179398771,
      "loss": 6.2279,
      "step": 200
    },
    {
      "epoch": 0.026907760198041114,
      "grad_norm": 1.6346766948699951,
      "learning_rate": 0.0005990302769098158,
      "loss": 6.0734,
      "step": 250
    },
    {
      "epoch": 0.03228931223764934,
      "grad_norm": 1.4647449254989624,
      "learning_rate": 0.0005987070358797543,
      "loss": 5.9608,
      "step": 300
    },
    {
      "epoch": 0.03767086427725756,
      "grad_norm": 1.140522837638855,
      "learning_rate": 0.0005983837948496929,
      "loss": 5.8542,
      "step": 350
    },
    {
      "epoch": 0.04305241631686579,
      "grad_norm": 1.6901541948318481,
      "learning_rate": 0.0005980605538196314,
      "loss": 5.7803,
      "step": 400
    },
    {
      "epoch": 0.048433968356474004,
      "grad_norm": 1.6334667205810547,
      "learning_rate": 0.0005977373127895701,
      "loss": 5.7209,
      "step": 450
    },
    {
      "epoch": 0.05381552039608223,
      "grad_norm": 1.9607793092727661,
      "learning_rate": 0.0005974140717595086,
      "loss": 5.639,
      "step": 500
    },
    {
      "epoch": 0.05919707243569045,
      "grad_norm": 1.286758542060852,
      "learning_rate": 0.0005970908307294472,
      "loss": 5.5708,
      "step": 550
    },
    {
      "epoch": 0.06457862447529868,
      "grad_norm": 1.0357468128204346,
      "learning_rate": 0.0005967675896993858,
      "loss": 5.5179,
      "step": 600
    },
    {
      "epoch": 0.0699601765149069,
      "grad_norm": 1.1820484399795532,
      "learning_rate": 0.0005964443486693243,
      "loss": 5.4594,
      "step": 650
    },
    {
      "epoch": 0.07534172855451512,
      "grad_norm": 0.9575048089027405,
      "learning_rate": 0.000596121107639263,
      "loss": 5.3791,
      "step": 700
    },
    {
      "epoch": 0.08072328059412334,
      "grad_norm": 1.2174663543701172,
      "learning_rate": 0.0005957978666092015,
      "loss": 5.3318,
      "step": 750
    },
    {
      "epoch": 0.08610483263373157,
      "grad_norm": 1.6399223804473877,
      "learning_rate": 0.0005954746255791401,
      "loss": 5.2423,
      "step": 800
    },
    {
      "epoch": 0.09148638467333979,
      "grad_norm": 0.8586328029632568,
      "learning_rate": 0.0005951513845490787,
      "loss": 5.1898,
      "step": 850
    },
    {
      "epoch": 0.09686793671294801,
      "grad_norm": 1.4032552242279053,
      "learning_rate": 0.0005948281435190174,
      "loss": 5.1885,
      "step": 900
    },
    {
      "epoch": 0.10224948875255624,
      "grad_norm": 1.364333152770996,
      "learning_rate": 0.0005945049024889559,
      "loss": 5.1196,
      "step": 950
    },
    {
      "epoch": 0.10763104079216446,
      "grad_norm": 1.1644333600997925,
      "learning_rate": 0.0005941816614588944,
      "loss": 5.1198,
      "step": 1000
    },
    {
      "epoch": 0.10763104079216446,
      "eval_accuracy": 0.22768223506813243,
      "eval_loss": 5.016661643981934,
      "eval_runtime": 191.8167,
      "eval_samples_per_second": 93.897,
      "eval_steps_per_second": 5.87,
      "step": 1000
    },
    {
      "epoch": 0.11301259283177269,
      "grad_norm": 1.3850427865982056,
      "learning_rate": 0.000593858420428833,
      "loss": 5.0633,
      "step": 1050
    },
    {
      "epoch": 0.1183941448713809,
      "grad_norm": 1.2087067365646362,
      "learning_rate": 0.0005935351793987716,
      "loss": 5.0493,
      "step": 1100
    },
    {
      "epoch": 0.12377569691098914,
      "grad_norm": 0.9972895979881287,
      "learning_rate": 0.0005932119383687103,
      "loss": 4.9925,
      "step": 1150
    },
    {
      "epoch": 0.12915724895059735,
      "grad_norm": 1.0618246793746948,
      "learning_rate": 0.0005928886973386488,
      "loss": 4.9241,
      "step": 1200
    },
    {
      "epoch": 0.13453880099020557,
      "grad_norm": 1.147125005722046,
      "learning_rate": 0.0005925654563085874,
      "loss": 4.9353,
      "step": 1250
    },
    {
      "epoch": 0.1399203530298138,
      "grad_norm": 0.9205951690673828,
      "learning_rate": 0.000592242215278526,
      "loss": 4.904,
      "step": 1300
    },
    {
      "epoch": 0.14530190506942203,
      "grad_norm": 1.1346954107284546,
      "learning_rate": 0.0005919189742484645,
      "loss": 4.9043,
      "step": 1350
    },
    {
      "epoch": 0.15068345710903025,
      "grad_norm": 0.9538147449493408,
      "learning_rate": 0.0005915957332184032,
      "loss": 4.8268,
      "step": 1400
    },
    {
      "epoch": 0.15606500914863847,
      "grad_norm": 0.993122398853302,
      "learning_rate": 0.0005912724921883417,
      "loss": 4.8241,
      "step": 1450
    },
    {
      "epoch": 0.16144656118824668,
      "grad_norm": 0.9443798065185547,
      "learning_rate": 0.0005909492511582803,
      "loss": 4.7935,
      "step": 1500
    },
    {
      "epoch": 0.1668281132278549,
      "grad_norm": 1.03653883934021,
      "learning_rate": 0.0005906260101282189,
      "loss": 4.7565,
      "step": 1550
    },
    {
      "epoch": 0.17220966526746315,
      "grad_norm": 1.085654854774475,
      "learning_rate": 0.0005903027690981575,
      "loss": 4.7377,
      "step": 1600
    },
    {
      "epoch": 0.17759121730707136,
      "grad_norm": 1.0298115015029907,
      "learning_rate": 0.000589979528068096,
      "loss": 4.7123,
      "step": 1650
    },
    {
      "epoch": 0.18297276934667958,
      "grad_norm": 0.7080234885215759,
      "learning_rate": 0.0005896562870380347,
      "loss": 4.7112,
      "step": 1700
    },
    {
      "epoch": 0.1883543213862878,
      "grad_norm": 0.9891972541809082,
      "learning_rate": 0.0005893330460079732,
      "loss": 4.6859,
      "step": 1750
    },
    {
      "epoch": 0.19373587342589602,
      "grad_norm": 0.9329166412353516,
      "learning_rate": 0.0005890098049779118,
      "loss": 4.6557,
      "step": 1800
    },
    {
      "epoch": 0.19911742546550426,
      "grad_norm": 0.9141618013381958,
      "learning_rate": 0.0005886865639478504,
      "loss": 4.6404,
      "step": 1850
    },
    {
      "epoch": 0.20449897750511248,
      "grad_norm": 0.9240756630897522,
      "learning_rate": 0.0005883633229177889,
      "loss": 4.6361,
      "step": 1900
    },
    {
      "epoch": 0.2098805295447207,
      "grad_norm": 1.1672885417938232,
      "learning_rate": 0.0005880400818877276,
      "loss": 4.6135,
      "step": 1950
    },
    {
      "epoch": 0.2152620815843289,
      "grad_norm": 1.0662496089935303,
      "learning_rate": 0.0005877168408576662,
      "loss": 4.5885,
      "step": 2000
    },
    {
      "epoch": 0.2152620815843289,
      "eval_accuracy": 0.2698795398028667,
      "eval_loss": 4.5171732902526855,
      "eval_runtime": 190.0727,
      "eval_samples_per_second": 94.758,
      "eval_steps_per_second": 5.924,
      "step": 2000
    },
    {
      "epoch": 0.22064363362393713,
      "grad_norm": 0.9489683508872986,
      "learning_rate": 0.0005873935998276048,
      "loss": 4.5521,
      "step": 2050
    },
    {
      "epoch": 0.22602518566354537,
      "grad_norm": 1.1695523262023926,
      "learning_rate": 0.0005870703587975433,
      "loss": 4.5315,
      "step": 2100
    },
    {
      "epoch": 0.2314067377031536,
      "grad_norm": 0.7433720827102661,
      "learning_rate": 0.0005867471177674818,
      "loss": 4.5271,
      "step": 2150
    },
    {
      "epoch": 0.2367882897427618,
      "grad_norm": 0.8087969422340393,
      "learning_rate": 0.0005864238767374205,
      "loss": 4.5019,
      "step": 2200
    },
    {
      "epoch": 0.24216984178237003,
      "grad_norm": 0.992636501789093,
      "learning_rate": 0.0005861006357073591,
      "loss": 4.493,
      "step": 2250
    },
    {
      "epoch": 0.24755139382197827,
      "grad_norm": 0.6901751160621643,
      "learning_rate": 0.0005857773946772977,
      "loss": 4.464,
      "step": 2300
    },
    {
      "epoch": 0.2529329458615865,
      "grad_norm": 1.005516529083252,
      "learning_rate": 0.0005854541536472362,
      "loss": 4.4568,
      "step": 2350
    },
    {
      "epoch": 0.2583144979011947,
      "grad_norm": 1.19230055809021,
      "learning_rate": 0.0005851309126171749,
      "loss": 4.4311,
      "step": 2400
    },
    {
      "epoch": 0.2636960499408029,
      "grad_norm": 0.8644243478775024,
      "learning_rate": 0.0005848076715871134,
      "loss": 4.4349,
      "step": 2450
    },
    {
      "epoch": 0.26907760198041114,
      "grad_norm": 0.8199976086616516,
      "learning_rate": 0.000584484430557052,
      "loss": 4.4301,
      "step": 2500
    },
    {
      "epoch": 0.27445915402001936,
      "grad_norm": 0.9369893670082092,
      "learning_rate": 0.0005841611895269906,
      "loss": 4.4209,
      "step": 2550
    },
    {
      "epoch": 0.2798407060596276,
      "grad_norm": 0.6631494164466858,
      "learning_rate": 0.0005838379484969291,
      "loss": 4.374,
      "step": 2600
    },
    {
      "epoch": 0.2852222580992358,
      "grad_norm": 0.8206772208213806,
      "learning_rate": 0.0005835147074668678,
      "loss": 4.378,
      "step": 2650
    },
    {
      "epoch": 0.29060381013884407,
      "grad_norm": 1.0306700468063354,
      "learning_rate": 0.0005831914664368063,
      "loss": 4.3651,
      "step": 2700
    },
    {
      "epoch": 0.2959853621784523,
      "grad_norm": 0.8099345564842224,
      "learning_rate": 0.0005828682254067449,
      "loss": 4.3605,
      "step": 2750
    },
    {
      "epoch": 0.3013669142180605,
      "grad_norm": 0.9737576842308044,
      "learning_rate": 0.0005825449843766835,
      "loss": 4.342,
      "step": 2800
    },
    {
      "epoch": 0.3067484662576687,
      "grad_norm": 1.0105592012405396,
      "learning_rate": 0.0005822217433466221,
      "loss": 4.3429,
      "step": 2850
    },
    {
      "epoch": 0.31213001829727693,
      "grad_norm": 0.8239940404891968,
      "learning_rate": 0.0005818985023165607,
      "loss": 4.3248,
      "step": 2900
    },
    {
      "epoch": 0.31751157033688515,
      "grad_norm": 0.8788948059082031,
      "learning_rate": 0.0005815752612864992,
      "loss": 4.3174,
      "step": 2950
    },
    {
      "epoch": 0.32289312237649337,
      "grad_norm": 0.7173974514007568,
      "learning_rate": 0.0005812520202564378,
      "loss": 4.3239,
      "step": 3000
    },
    {
      "epoch": 0.32289312237649337,
      "eval_accuracy": 0.29787058810874545,
      "eval_loss": 4.2347259521484375,
      "eval_runtime": 190.037,
      "eval_samples_per_second": 94.776,
      "eval_steps_per_second": 5.925,
      "step": 3000
    },
    {
      "epoch": 0.3282746744161016,
      "grad_norm": 1.0855518579483032,
      "learning_rate": 0.0005809287792263764,
      "loss": 4.2995,
      "step": 3050
    },
    {
      "epoch": 0.3336562264557098,
      "grad_norm": 0.7936787009239197,
      "learning_rate": 0.0005806055381963151,
      "loss": 4.2907,
      "step": 3100
    },
    {
      "epoch": 0.3390377784953181,
      "grad_norm": 0.701678991317749,
      "learning_rate": 0.0005802822971662536,
      "loss": 4.2695,
      "step": 3150
    },
    {
      "epoch": 0.3444193305349263,
      "grad_norm": 0.7440540194511414,
      "learning_rate": 0.0005799590561361922,
      "loss": 4.2827,
      "step": 3200
    },
    {
      "epoch": 0.3498008825745345,
      "grad_norm": 0.7642802000045776,
      "learning_rate": 0.0005796358151061307,
      "loss": 4.2395,
      "step": 3250
    },
    {
      "epoch": 0.35518243461414273,
      "grad_norm": 0.7320182919502258,
      "learning_rate": 0.0005793125740760694,
      "loss": 4.2654,
      "step": 3300
    },
    {
      "epoch": 0.36056398665375095,
      "grad_norm": 0.8758799433708191,
      "learning_rate": 0.0005789893330460079,
      "loss": 4.2597,
      "step": 3350
    },
    {
      "epoch": 0.36594553869335916,
      "grad_norm": 0.7132930159568787,
      "learning_rate": 0.0005786660920159465,
      "loss": 4.2342,
      "step": 3400
    },
    {
      "epoch": 0.3713270907329674,
      "grad_norm": 0.7813701629638672,
      "learning_rate": 0.0005783428509858851,
      "loss": 4.2476,
      "step": 3450
    },
    {
      "epoch": 0.3767086427725756,
      "grad_norm": 0.7079502940177917,
      "learning_rate": 0.0005780196099558237,
      "loss": 4.2239,
      "step": 3500
    },
    {
      "epoch": 0.3820901948121838,
      "grad_norm": 0.6457203030586243,
      "learning_rate": 0.0005776963689257623,
      "loss": 4.2229,
      "step": 3550
    },
    {
      "epoch": 0.38747174685179203,
      "grad_norm": 0.6975264549255371,
      "learning_rate": 0.0005773731278957008,
      "loss": 4.2215,
      "step": 3600
    },
    {
      "epoch": 0.3928532988914003,
      "grad_norm": 0.7320036292076111,
      "learning_rate": 0.0005770498868656394,
      "loss": 4.2053,
      "step": 3650
    },
    {
      "epoch": 0.3982348509310085,
      "grad_norm": 0.763953447341919,
      "learning_rate": 0.000576726645835578,
      "loss": 4.2095,
      "step": 3700
    },
    {
      "epoch": 0.40361640297061674,
      "grad_norm": 0.6851209402084351,
      "learning_rate": 0.0005764034048055167,
      "loss": 4.1695,
      "step": 3750
    },
    {
      "epoch": 0.40899795501022496,
      "grad_norm": 0.746885359287262,
      "learning_rate": 0.0005760801637754552,
      "loss": 4.2002,
      "step": 3800
    },
    {
      "epoch": 0.4143795070498332,
      "grad_norm": 0.7517973780632019,
      "learning_rate": 0.0005757569227453937,
      "loss": 4.208,
      "step": 3850
    },
    {
      "epoch": 0.4197610590894414,
      "grad_norm": 0.7405137419700623,
      "learning_rate": 0.0005754336817153324,
      "loss": 4.1665,
      "step": 3900
    },
    {
      "epoch": 0.4251426111290496,
      "grad_norm": 0.6628493666648865,
      "learning_rate": 0.0005751104406852709,
      "loss": 4.1592,
      "step": 3950
    },
    {
      "epoch": 0.4305241631686578,
      "grad_norm": 0.660922110080719,
      "learning_rate": 0.0005747871996552096,
      "loss": 4.1572,
      "step": 4000
    },
    {
      "epoch": 0.4305241631686578,
      "eval_accuracy": 0.31207032536433216,
      "eval_loss": 4.093014717102051,
      "eval_runtime": 190.3211,
      "eval_samples_per_second": 94.635,
      "eval_steps_per_second": 5.916,
      "step": 4000
    },
    {
      "epoch": 0.43590571520826604,
      "grad_norm": 0.6598483324050903,
      "learning_rate": 0.0005744639586251481,
      "loss": 4.1677,
      "step": 4050
    },
    {
      "epoch": 0.44128726724787426,
      "grad_norm": 0.7623891234397888,
      "learning_rate": 0.0005741407175950867,
      "loss": 4.1385,
      "step": 4100
    },
    {
      "epoch": 0.44666881928748253,
      "grad_norm": 0.6655964255332947,
      "learning_rate": 0.0005738174765650253,
      "loss": 4.1479,
      "step": 4150
    },
    {
      "epoch": 0.45205037132709075,
      "grad_norm": 0.6780499219894409,
      "learning_rate": 0.0005734942355349638,
      "loss": 4.1354,
      "step": 4200
    },
    {
      "epoch": 0.45743192336669897,
      "grad_norm": 0.695399284362793,
      "learning_rate": 0.0005731709945049025,
      "loss": 4.1459,
      "step": 4250
    },
    {
      "epoch": 0.4628134754063072,
      "grad_norm": 0.6070560812950134,
      "learning_rate": 0.000572847753474841,
      "loss": 4.1249,
      "step": 4300
    },
    {
      "epoch": 0.4681950274459154,
      "grad_norm": 0.7866783142089844,
      "learning_rate": 0.0005725245124447796,
      "loss": 4.1306,
      "step": 4350
    },
    {
      "epoch": 0.4735765794855236,
      "grad_norm": 0.6466276049613953,
      "learning_rate": 0.0005722012714147182,
      "loss": 4.13,
      "step": 4400
    },
    {
      "epoch": 0.47895813152513184,
      "grad_norm": 0.6519131064414978,
      "learning_rate": 0.0005718780303846568,
      "loss": 4.1253,
      "step": 4450
    },
    {
      "epoch": 0.48433968356474005,
      "grad_norm": 0.701987624168396,
      "learning_rate": 0.0005715547893545953,
      "loss": 4.1068,
      "step": 4500
    },
    {
      "epoch": 0.48972123560434827,
      "grad_norm": 0.725153386592865,
      "learning_rate": 0.000571231548324534,
      "loss": 4.102,
      "step": 4550
    },
    {
      "epoch": 0.49510278764395654,
      "grad_norm": 0.7474183440208435,
      "learning_rate": 0.0005709083072944725,
      "loss": 4.113,
      "step": 4600
    },
    {
      "epoch": 0.5004843396835648,
      "grad_norm": 0.6553453803062439,
      "learning_rate": 0.0005705850662644111,
      "loss": 4.0973,
      "step": 4650
    },
    {
      "epoch": 0.505865891723173,
      "grad_norm": 0.5860002636909485,
      "learning_rate": 0.0005702618252343497,
      "loss": 4.0819,
      "step": 4700
    },
    {
      "epoch": 0.5112474437627812,
      "grad_norm": 0.7331970930099487,
      "learning_rate": 0.0005699385842042882,
      "loss": 4.0847,
      "step": 4750
    },
    {
      "epoch": 0.5166289958023894,
      "grad_norm": 0.7375046014785767,
      "learning_rate": 0.0005696153431742269,
      "loss": 4.0899,
      "step": 4800
    },
    {
      "epoch": 0.5220105478419976,
      "grad_norm": 0.6551966071128845,
      "learning_rate": 0.0005692921021441655,
      "loss": 4.0733,
      "step": 4850
    },
    {
      "epoch": 0.5273920998816058,
      "grad_norm": 0.5899481177330017,
      "learning_rate": 0.0005689688611141041,
      "loss": 4.0628,
      "step": 4900
    },
    {
      "epoch": 0.5327736519212141,
      "grad_norm": 0.6506676077842712,
      "learning_rate": 0.0005686456200840426,
      "loss": 4.0713,
      "step": 4950
    },
    {
      "epoch": 0.5381552039608223,
      "grad_norm": 0.6364733576774597,
      "learning_rate": 0.0005683223790539811,
      "loss": 4.0457,
      "step": 5000
    },
    {
      "epoch": 0.5381552039608223,
      "eval_accuracy": 0.3211192638201856,
      "eval_loss": 3.9926953315734863,
      "eval_runtime": 190.1606,
      "eval_samples_per_second": 94.715,
      "eval_steps_per_second": 5.921,
      "step": 5000
    },
    {
      "epoch": 0.5435367560004305,
      "grad_norm": 0.6197501420974731,
      "learning_rate": 0.0005679991380239198,
      "loss": 4.0678,
      "step": 5050
    },
    {
      "epoch": 0.5489183080400387,
      "grad_norm": 0.6434921026229858,
      "learning_rate": 0.0005676758969938584,
      "loss": 4.062,
      "step": 5100
    },
    {
      "epoch": 0.5542998600796469,
      "grad_norm": 0.6450833678245544,
      "learning_rate": 0.000567352655963797,
      "loss": 4.0389,
      "step": 5150
    },
    {
      "epoch": 0.5596814121192552,
      "grad_norm": 0.5980785489082336,
      "learning_rate": 0.0005670294149337355,
      "loss": 4.0606,
      "step": 5200
    },
    {
      "epoch": 0.5650629641588634,
      "grad_norm": 0.7756244540214539,
      "learning_rate": 0.0005667061739036742,
      "loss": 4.0456,
      "step": 5250
    },
    {
      "epoch": 0.5704445161984716,
      "grad_norm": 0.620424747467041,
      "learning_rate": 0.0005663829328736127,
      "loss": 4.0386,
      "step": 5300
    },
    {
      "epoch": 0.5758260682380799,
      "grad_norm": 0.5773258805274963,
      "learning_rate": 0.0005660596918435512,
      "loss": 4.0324,
      "step": 5350
    },
    {
      "epoch": 0.5812076202776881,
      "grad_norm": 0.5738138556480408,
      "learning_rate": 0.0005657364508134899,
      "loss": 4.0211,
      "step": 5400
    },
    {
      "epoch": 0.5865891723172963,
      "grad_norm": 0.6829665899276733,
      "learning_rate": 0.0005654132097834284,
      "loss": 4.0491,
      "step": 5450
    },
    {
      "epoch": 0.5919707243569046,
      "grad_norm": 0.6426494121551514,
      "learning_rate": 0.0005650899687533671,
      "loss": 4.0318,
      "step": 5500
    },
    {
      "epoch": 0.5973522763965128,
      "grad_norm": 0.7136837840080261,
      "learning_rate": 0.0005647667277233056,
      "loss": 4.0282,
      "step": 5550
    },
    {
      "epoch": 0.602733828436121,
      "grad_norm": 0.5938235521316528,
      "learning_rate": 0.0005644434866932442,
      "loss": 3.9908,
      "step": 5600
    },
    {
      "epoch": 0.6081153804757292,
      "grad_norm": 0.6892990469932556,
      "learning_rate": 0.0005641202456631828,
      "loss": 4.0199,
      "step": 5650
    },
    {
      "epoch": 0.6134969325153374,
      "grad_norm": 0.634809672832489,
      "learning_rate": 0.0005637970046331214,
      "loss": 4.0156,
      "step": 5700
    },
    {
      "epoch": 0.6188784845549457,
      "grad_norm": 0.5722842812538147,
      "learning_rate": 0.00056347376360306,
      "loss": 3.9966,
      "step": 5750
    },
    {
      "epoch": 0.6242600365945539,
      "grad_norm": 0.5929718017578125,
      "learning_rate": 0.0005631505225729985,
      "loss": 3.9986,
      "step": 5800
    },
    {
      "epoch": 0.6296415886341621,
      "grad_norm": 0.6082379221916199,
      "learning_rate": 0.0005628272815429371,
      "loss": 3.9875,
      "step": 5850
    },
    {
      "epoch": 0.6350231406737703,
      "grad_norm": 0.6514686346054077,
      "learning_rate": 0.0005625040405128757,
      "loss": 3.9591,
      "step": 5900
    },
    {
      "epoch": 0.6404046927133785,
      "grad_norm": 0.5314525961875916,
      "learning_rate": 0.0005621807994828143,
      "loss": 3.9801,
      "step": 5950
    },
    {
      "epoch": 0.6457862447529867,
      "grad_norm": 0.4907394349575043,
      "learning_rate": 0.0005618575584527529,
      "loss": 3.9818,
      "step": 6000
    },
    {
      "epoch": 0.6457862447529867,
      "eval_accuracy": 0.32771308162298296,
      "eval_loss": 3.917905569076538,
      "eval_runtime": 189.6405,
      "eval_samples_per_second": 94.974,
      "eval_steps_per_second": 5.938,
      "step": 6000
    },
    {
      "epoch": 0.651167796792595,
      "grad_norm": 0.6379232406616211,
      "learning_rate": 0.0005615343174226915,
      "loss": 3.9764,
      "step": 6050
    },
    {
      "epoch": 0.6565493488322032,
      "grad_norm": 0.7434052228927612,
      "learning_rate": 0.00056121107639263,
      "loss": 3.9955,
      "step": 6100
    },
    {
      "epoch": 0.6619309008718114,
      "grad_norm": 0.5711482167243958,
      "learning_rate": 0.0005608878353625687,
      "loss": 3.9783,
      "step": 6150
    },
    {
      "epoch": 0.6673124529114196,
      "grad_norm": 0.6705279350280762,
      "learning_rate": 0.0005605645943325072,
      "loss": 3.9633,
      "step": 6200
    },
    {
      "epoch": 0.6726940049510278,
      "grad_norm": 0.5938876867294312,
      "learning_rate": 0.0005602413533024458,
      "loss": 3.961,
      "step": 6250
    },
    {
      "epoch": 0.6780755569906362,
      "grad_norm": 0.6743350625038147,
      "learning_rate": 0.0005599245770929855,
      "loss": 3.968,
      "step": 6300
    },
    {
      "epoch": 0.6834571090302444,
      "grad_norm": 0.6835203170776367,
      "learning_rate": 0.0005596013360629242,
      "loss": 3.9816,
      "step": 6350
    },
    {
      "epoch": 0.6888386610698526,
      "grad_norm": 0.6302648782730103,
      "learning_rate": 0.0005592780950328628,
      "loss": 3.9548,
      "step": 6400
    },
    {
      "epoch": 0.6942202131094608,
      "grad_norm": 0.5848674774169922,
      "learning_rate": 0.0005589548540028014,
      "loss": 3.9687,
      "step": 6450
    },
    {
      "epoch": 0.699601765149069,
      "grad_norm": 0.6802118420600891,
      "learning_rate": 0.0005586316129727399,
      "loss": 3.9581,
      "step": 6500
    },
    {
      "epoch": 0.7049833171886772,
      "grad_norm": 0.6318409442901611,
      "learning_rate": 0.0005583083719426786,
      "loss": 3.9456,
      "step": 6550
    },
    {
      "epoch": 0.7103648692282855,
      "grad_norm": 0.6047348976135254,
      "learning_rate": 0.0005579851309126171,
      "loss": 3.9536,
      "step": 6600
    },
    {
      "epoch": 0.7157464212678937,
      "grad_norm": 0.5617726445198059,
      "learning_rate": 0.0005576618898825558,
      "loss": 3.9542,
      "step": 6650
    },
    {
      "epoch": 0.7211279733075019,
      "grad_norm": 0.5408573746681213,
      "learning_rate": 0.0005573386488524943,
      "loss": 3.9567,
      "step": 6700
    },
    {
      "epoch": 0.7265095253471101,
      "grad_norm": 0.585254967212677,
      "learning_rate": 0.0005570154078224328,
      "loss": 3.9359,
      "step": 6750
    },
    {
      "epoch": 0.7318910773867183,
      "grad_norm": 0.5611248016357422,
      "learning_rate": 0.0005566921667923715,
      "loss": 3.9518,
      "step": 6800
    },
    {
      "epoch": 0.7372726294263265,
      "grad_norm": 0.5755305290222168,
      "learning_rate": 0.00055636892576231,
      "loss": 3.927,
      "step": 6850
    },
    {
      "epoch": 0.7426541814659348,
      "grad_norm": 0.5712345838546753,
      "learning_rate": 0.0005560456847322487,
      "loss": 3.9773,
      "step": 6900
    },
    {
      "epoch": 0.748035733505543,
      "grad_norm": 0.661975085735321,
      "learning_rate": 0.0005557224437021872,
      "loss": 3.9534,
      "step": 6950
    },
    {
      "epoch": 0.7534172855451512,
      "grad_norm": 0.673778235912323,
      "learning_rate": 0.0005553992026721258,
      "loss": 3.9485,
      "step": 7000
    },
    {
      "epoch": 0.7534172855451512,
      "eval_accuracy": 0.33348591820545415,
      "eval_loss": 3.8628029823303223,
      "eval_runtime": 190.5947,
      "eval_samples_per_second": 94.499,
      "eval_steps_per_second": 5.908,
      "step": 7000
    },
    {
      "epoch": 0.7587988375847594,
      "grad_norm": 0.5945296287536621,
      "learning_rate": 0.0005550759616420644,
      "loss": 3.929,
      "step": 7050
    },
    {
      "epoch": 0.7641803896243676,
      "grad_norm": 0.6192255616188049,
      "learning_rate": 0.000554752720612003,
      "loss": 3.9085,
      "step": 7100
    },
    {
      "epoch": 0.7695619416639758,
      "grad_norm": 0.6141802668571472,
      "learning_rate": 0.0005544294795819415,
      "loss": 3.9312,
      "step": 7150
    },
    {
      "epoch": 0.7749434937035841,
      "grad_norm": 0.6082919239997864,
      "learning_rate": 0.0005541062385518801,
      "loss": 3.8999,
      "step": 7200
    },
    {
      "epoch": 0.7803250457431924,
      "grad_norm": 0.5389159321784973,
      "learning_rate": 0.0005537829975218188,
      "loss": 3.9072,
      "step": 7250
    },
    {
      "epoch": 0.7857065977828006,
      "grad_norm": 0.5647724270820618,
      "learning_rate": 0.0005534597564917573,
      "loss": 3.9081,
      "step": 7300
    },
    {
      "epoch": 0.7910881498224088,
      "grad_norm": 0.6201077699661255,
      "learning_rate": 0.0005531365154616959,
      "loss": 3.9216,
      "step": 7350
    },
    {
      "epoch": 0.796469701862017,
      "grad_norm": 0.578396201133728,
      "learning_rate": 0.0005528132744316344,
      "loss": 3.9052,
      "step": 7400
    },
    {
      "epoch": 0.8018512539016253,
      "grad_norm": 0.5452779531478882,
      "learning_rate": 0.0005524900334015731,
      "loss": 3.9132,
      "step": 7450
    },
    {
      "epoch": 0.8072328059412335,
      "grad_norm": 0.5811088681221008,
      "learning_rate": 0.0005521667923715117,
      "loss": 3.886,
      "step": 7500
    },
    {
      "epoch": 0.8126143579808417,
      "grad_norm": 0.5738024115562439,
      "learning_rate": 0.0005518435513414502,
      "loss": 3.9123,
      "step": 7550
    },
    {
      "epoch": 0.8179959100204499,
      "grad_norm": 0.5562692284584045,
      "learning_rate": 0.0005515203103113888,
      "loss": 3.8996,
      "step": 7600
    },
    {
      "epoch": 0.8233774620600581,
      "grad_norm": 0.6202870011329651,
      "learning_rate": 0.0005511970692813274,
      "loss": 3.9049,
      "step": 7650
    },
    {
      "epoch": 0.8287590140996663,
      "grad_norm": 0.5561438202857971,
      "learning_rate": 0.000550873828251266,
      "loss": 3.9111,
      "step": 7700
    },
    {
      "epoch": 0.8341405661392746,
      "grad_norm": 0.599392294883728,
      "learning_rate": 0.0005505505872212045,
      "loss": 3.8809,
      "step": 7750
    },
    {
      "epoch": 0.8395221181788828,
      "grad_norm": 0.6705716252326965,
      "learning_rate": 0.0005502273461911432,
      "loss": 3.8804,
      "step": 7800
    },
    {
      "epoch": 0.844903670218491,
      "grad_norm": 0.6356175541877747,
      "learning_rate": 0.0005499041051610817,
      "loss": 3.8808,
      "step": 7850
    },
    {
      "epoch": 0.8502852222580992,
      "grad_norm": 0.5892457962036133,
      "learning_rate": 0.0005495808641310204,
      "loss": 3.8944,
      "step": 7900
    },
    {
      "epoch": 0.8556667742977074,
      "grad_norm": 0.6224340200424194,
      "learning_rate": 0.0005492576231009589,
      "loss": 3.8857,
      "step": 7950
    },
    {
      "epoch": 0.8610483263373157,
      "grad_norm": 0.6061555743217468,
      "learning_rate": 0.0005489343820708974,
      "loss": 3.9,
      "step": 8000
    },
    {
      "epoch": 0.8610483263373157,
      "eval_accuracy": 0.33765275645313947,
      "eval_loss": 3.815392255783081,
      "eval_runtime": 189.9949,
      "eval_samples_per_second": 94.797,
      "eval_steps_per_second": 5.926,
      "step": 8000
    },
    {
      "epoch": 0.8664298783769239,
      "grad_norm": 1.3503775596618652,
      "learning_rate": 0.0005486111410408361,
      "loss": 3.8918,
      "step": 8050
    },
    {
      "epoch": 0.8718114304165321,
      "grad_norm": 0.6002238392829895,
      "learning_rate": 0.0005482879000107746,
      "loss": 3.8702,
      "step": 8100
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.5893197655677795,
      "learning_rate": 0.0005479646589807133,
      "loss": 3.8738,
      "step": 8150
    },
    {
      "epoch": 0.8825745344957485,
      "grad_norm": 0.5278034210205078,
      "learning_rate": 0.0005476414179506518,
      "loss": 3.8751,
      "step": 8200
    },
    {
      "epoch": 0.8879560865353568,
      "grad_norm": 0.5492599010467529,
      "learning_rate": 0.0005473181769205904,
      "loss": 3.8864,
      "step": 8250
    },
    {
      "epoch": 0.8933376385749651,
      "grad_norm": 0.5954228639602661,
      "learning_rate": 0.000546994935890529,
      "loss": 3.8709,
      "step": 8300
    },
    {
      "epoch": 0.8987191906145733,
      "grad_norm": 0.531387448310852,
      "learning_rate": 0.0005466716948604677,
      "loss": 3.8718,
      "step": 8350
    },
    {
      "epoch": 0.9041007426541815,
      "grad_norm": 0.5963478684425354,
      "learning_rate": 0.0005463484538304062,
      "loss": 3.8667,
      "step": 8400
    },
    {
      "epoch": 0.9094822946937897,
      "grad_norm": 0.6018481254577637,
      "learning_rate": 0.0005460252128003447,
      "loss": 3.8588,
      "step": 8450
    },
    {
      "epoch": 0.9148638467333979,
      "grad_norm": 0.5060446262359619,
      "learning_rate": 0.0005457084365908845,
      "loss": 3.8727,
      "step": 8500
    },
    {
      "epoch": 0.9202453987730062,
      "grad_norm": 0.6362840533256531,
      "learning_rate": 0.0005453851955608232,
      "loss": 3.8484,
      "step": 8550
    },
    {
      "epoch": 0.9256269508126144,
      "grad_norm": 0.5442132353782654,
      "learning_rate": 0.0005450619545307617,
      "loss": 3.8627,
      "step": 8600
    },
    {
      "epoch": 0.9310085028522226,
      "grad_norm": 0.582240641117096,
      "learning_rate": 0.0005447387135007003,
      "loss": 3.847,
      "step": 8650
    },
    {
      "epoch": 0.9363900548918308,
      "grad_norm": 0.5463622212409973,
      "learning_rate": 0.0005444154724706389,
      "loss": 3.8341,
      "step": 8700
    },
    {
      "epoch": 0.941771606931439,
      "grad_norm": 0.5677204728126526,
      "learning_rate": 0.0005440922314405775,
      "loss": 3.8555,
      "step": 8750
    },
    {
      "epoch": 0.9471531589710472,
      "grad_norm": 0.5602720975875854,
      "learning_rate": 0.0005437689904105161,
      "loss": 3.8412,
      "step": 8800
    },
    {
      "epoch": 0.9525347110106555,
      "grad_norm": 0.6204741597175598,
      "learning_rate": 0.0005434457493804546,
      "loss": 3.8451,
      "step": 8850
    },
    {
      "epoch": 0.9579162630502637,
      "grad_norm": 0.541634738445282,
      "learning_rate": 0.0005431225083503932,
      "loss": 3.839,
      "step": 8900
    },
    {
      "epoch": 0.9632978150898719,
      "grad_norm": 0.5156535506248474,
      "learning_rate": 0.0005427992673203318,
      "loss": 3.8544,
      "step": 8950
    },
    {
      "epoch": 0.9686793671294801,
      "grad_norm": 0.5302301049232483,
      "learning_rate": 0.0005424760262902704,
      "loss": 3.8485,
      "step": 9000
    },
    {
      "epoch": 0.9686793671294801,
      "eval_accuracy": 0.3409820982415508,
      "eval_loss": 3.7786221504211426,
      "eval_runtime": 190.014,
      "eval_samples_per_second": 94.788,
      "eval_steps_per_second": 5.926,
      "step": 9000
    },
    {
      "epoch": 0.9740609191690883,
      "grad_norm": 0.5033555030822754,
      "learning_rate": 0.000542152785260209,
      "loss": 3.8472,
      "step": 9050
    },
    {
      "epoch": 0.9794424712086965,
      "grad_norm": 0.5836737751960754,
      "learning_rate": 0.0005418295442301476,
      "loss": 3.8217,
      "step": 9100
    },
    {
      "epoch": 0.9848240232483048,
      "grad_norm": 0.5942977070808411,
      "learning_rate": 0.0005415063032000861,
      "loss": 3.8188,
      "step": 9150
    },
    {
      "epoch": 0.9902055752879131,
      "grad_norm": 0.5461533069610596,
      "learning_rate": 0.0005411830621700248,
      "loss": 3.8337,
      "step": 9200
    },
    {
      "epoch": 0.9955871273275213,
      "grad_norm": 0.520717978477478,
      "learning_rate": 0.0005408598211399633,
      "loss": 3.8433,
      "step": 9250
    },
    {
      "epoch": 1.0009686793671295,
      "grad_norm": 0.636262059211731,
      "learning_rate": 0.0005405365801099019,
      "loss": 3.8245,
      "step": 9300
    },
    {
      "epoch": 1.0063502314067376,
      "grad_norm": 0.588956356048584,
      "learning_rate": 0.0005402133390798405,
      "loss": 3.7877,
      "step": 9350
    },
    {
      "epoch": 1.011731783446346,
      "grad_norm": 0.5915181636810303,
      "learning_rate": 0.000539890098049779,
      "loss": 3.768,
      "step": 9400
    },
    {
      "epoch": 1.017113335485954,
      "grad_norm": 0.5807539820671082,
      "learning_rate": 0.0005395668570197177,
      "loss": 3.77,
      "step": 9450
    },
    {
      "epoch": 1.0224948875255624,
      "grad_norm": 0.5824580788612366,
      "learning_rate": 0.0005392436159896562,
      "loss": 3.7704,
      "step": 9500
    },
    {
      "epoch": 1.0278764395651705,
      "grad_norm": 0.5871043801307678,
      "learning_rate": 0.0005389203749595948,
      "loss": 3.7728,
      "step": 9550
    },
    {
      "epoch": 1.0332579916047788,
      "grad_norm": 0.5721834301948547,
      "learning_rate": 0.0005385971339295334,
      "loss": 3.7646,
      "step": 9600
    },
    {
      "epoch": 1.0386395436443872,
      "grad_norm": 0.6355053782463074,
      "learning_rate": 0.000538273892899472,
      "loss": 3.7769,
      "step": 9650
    },
    {
      "epoch": 1.0440210956839953,
      "grad_norm": 0.6225118637084961,
      "learning_rate": 0.0005379506518694106,
      "loss": 3.761,
      "step": 9700
    },
    {
      "epoch": 1.0494026477236036,
      "grad_norm": 0.523883044719696,
      "learning_rate": 0.0005376274108393491,
      "loss": 3.7765,
      "step": 9750
    },
    {
      "epoch": 1.0547841997632117,
      "grad_norm": 0.5992940664291382,
      "learning_rate": 0.0005373106346298889,
      "loss": 3.7831,
      "step": 9800
    },
    {
      "epoch": 1.06016575180282,
      "grad_norm": 0.5649681091308594,
      "learning_rate": 0.0005369873935998276,
      "loss": 3.7596,
      "step": 9850
    },
    {
      "epoch": 1.0655473038424281,
      "grad_norm": 0.5120299458503723,
      "learning_rate": 0.0005366641525697661,
      "loss": 3.7549,
      "step": 9900
    },
    {
      "epoch": 1.0709288558820365,
      "grad_norm": 0.5953982472419739,
      "learning_rate": 0.0005363409115397048,
      "loss": 3.7681,
      "step": 9950
    },
    {
      "epoch": 1.0763104079216446,
      "grad_norm": 0.6468278169631958,
      "learning_rate": 0.0005360176705096433,
      "loss": 3.7602,
      "step": 10000
    },
    {
      "epoch": 1.0763104079216446,
      "eval_accuracy": 0.34376828424377753,
      "eval_loss": 3.7513997554779053,
      "eval_runtime": 190.2644,
      "eval_samples_per_second": 94.663,
      "eval_steps_per_second": 5.918,
      "step": 10000
    },
    {
      "epoch": 1.081691959961253,
      "grad_norm": 0.5653687715530396,
      "learning_rate": 0.0005356944294795819,
      "loss": 3.7664,
      "step": 10050
    },
    {
      "epoch": 1.087073512000861,
      "grad_norm": 0.5622637271881104,
      "learning_rate": 0.0005353711884495205,
      "loss": 3.7878,
      "step": 10100
    },
    {
      "epoch": 1.0924550640404693,
      "grad_norm": 0.6015283465385437,
      "learning_rate": 0.000535047947419459,
      "loss": 3.7737,
      "step": 10150
    },
    {
      "epoch": 1.0978366160800774,
      "grad_norm": 0.587753415107727,
      "learning_rate": 0.0005347247063893976,
      "loss": 3.7622,
      "step": 10200
    },
    {
      "epoch": 1.1032181681196858,
      "grad_norm": 0.6016691327095032,
      "learning_rate": 0.0005344014653593362,
      "loss": 3.7509,
      "step": 10250
    },
    {
      "epoch": 1.1085997201592939,
      "grad_norm": 0.585686981678009,
      "learning_rate": 0.0005340782243292748,
      "loss": 3.7491,
      "step": 10300
    },
    {
      "epoch": 1.1139812721989022,
      "grad_norm": 0.5483114719390869,
      "learning_rate": 0.0005337549832992134,
      "loss": 3.7545,
      "step": 10350
    },
    {
      "epoch": 1.1193628242385103,
      "grad_norm": 0.5673395991325378,
      "learning_rate": 0.000533431742269152,
      "loss": 3.77,
      "step": 10400
    },
    {
      "epoch": 1.1247443762781186,
      "grad_norm": 0.5739583969116211,
      "learning_rate": 0.0005331085012390905,
      "loss": 3.7689,
      "step": 10450
    },
    {
      "epoch": 1.1301259283177267,
      "grad_norm": 0.5223838090896606,
      "learning_rate": 0.0005327852602090292,
      "loss": 3.7639,
      "step": 10500
    },
    {
      "epoch": 1.135507480357335,
      "grad_norm": 0.5173977613449097,
      "learning_rate": 0.0005324620191789678,
      "loss": 3.7542,
      "step": 10550
    },
    {
      "epoch": 1.1408890323969434,
      "grad_norm": 0.5478479266166687,
      "learning_rate": 0.0005321387781489063,
      "loss": 3.7543,
      "step": 10600
    },
    {
      "epoch": 1.1462705844365515,
      "grad_norm": 0.5701403617858887,
      "learning_rate": 0.0005318155371188449,
      "loss": 3.7493,
      "step": 10650
    },
    {
      "epoch": 1.1516521364761596,
      "grad_norm": 0.5782787799835205,
      "learning_rate": 0.0005314922960887834,
      "loss": 3.7354,
      "step": 10700
    },
    {
      "epoch": 1.157033688515768,
      "grad_norm": 0.6041041016578674,
      "learning_rate": 0.0005311690550587221,
      "loss": 3.7273,
      "step": 10750
    },
    {
      "epoch": 1.1624152405553763,
      "grad_norm": 0.5966827273368835,
      "learning_rate": 0.0005308458140286607,
      "loss": 3.7588,
      "step": 10800
    },
    {
      "epoch": 1.1677967925949844,
      "grad_norm": 0.5381931662559509,
      "learning_rate": 0.0005305225729985993,
      "loss": 3.7631,
      "step": 10850
    },
    {
      "epoch": 1.1731783446345927,
      "grad_norm": 0.5918460488319397,
      "learning_rate": 0.0005301993319685378,
      "loss": 3.7495,
      "step": 10900
    },
    {
      "epoch": 1.1785598966742008,
      "grad_norm": 0.5490368604660034,
      "learning_rate": 0.0005298760909384765,
      "loss": 3.7387,
      "step": 10950
    },
    {
      "epoch": 1.1839414487138091,
      "grad_norm": 0.5570510625839233,
      "learning_rate": 0.000529552849908415,
      "loss": 3.7453,
      "step": 11000
    },
    {
      "epoch": 1.1839414487138091,
      "eval_accuracy": 0.34701005180461036,
      "eval_loss": 3.722733736038208,
      "eval_runtime": 189.9554,
      "eval_samples_per_second": 94.817,
      "eval_steps_per_second": 5.928,
      "step": 11000
    },
    {
      "epoch": 1.1893230007534172,
      "grad_norm": 0.5757861733436584,
      "learning_rate": 0.0005292296088783535,
      "loss": 3.7504,
      "step": 11050
    },
    {
      "epoch": 1.1947045527930256,
      "grad_norm": 0.5488125681877136,
      "learning_rate": 0.0005289063678482922,
      "loss": 3.7602,
      "step": 11100
    },
    {
      "epoch": 1.2000861048326337,
      "grad_norm": 0.5695418119430542,
      "learning_rate": 0.0005285831268182307,
      "loss": 3.7483,
      "step": 11150
    },
    {
      "epoch": 1.205467656872242,
      "grad_norm": 0.5144563913345337,
      "learning_rate": 0.0005282598857881694,
      "loss": 3.7588,
      "step": 11200
    },
    {
      "epoch": 1.21084920891185,
      "grad_norm": 0.5603275895118713,
      "learning_rate": 0.0005279366447581079,
      "loss": 3.7496,
      "step": 11250
    },
    {
      "epoch": 1.2162307609514584,
      "grad_norm": 0.5516204833984375,
      "learning_rate": 0.0005276134037280465,
      "loss": 3.7366,
      "step": 11300
    },
    {
      "epoch": 1.2216123129910665,
      "grad_norm": 0.5955469012260437,
      "learning_rate": 0.0005272901626979851,
      "loss": 3.7335,
      "step": 11350
    },
    {
      "epoch": 1.2269938650306749,
      "grad_norm": 0.5783482789993286,
      "learning_rate": 0.0005269669216679236,
      "loss": 3.7412,
      "step": 11400
    },
    {
      "epoch": 1.232375417070283,
      "grad_norm": 0.5967660546302795,
      "learning_rate": 0.0005266436806378623,
      "loss": 3.7337,
      "step": 11450
    },
    {
      "epoch": 1.2377569691098913,
      "grad_norm": 0.565900444984436,
      "learning_rate": 0.0005263204396078008,
      "loss": 3.7415,
      "step": 11500
    },
    {
      "epoch": 1.2431385211494996,
      "grad_norm": 0.5914108753204346,
      "learning_rate": 0.0005259971985777394,
      "loss": 3.7388,
      "step": 11550
    },
    {
      "epoch": 1.2485200731891077,
      "grad_norm": 0.5387791395187378,
      "learning_rate": 0.000525673957547678,
      "loss": 3.7342,
      "step": 11600
    },
    {
      "epoch": 1.2539016252287158,
      "grad_norm": 0.6433733701705933,
      "learning_rate": 0.0005253507165176167,
      "loss": 3.7379,
      "step": 11650
    },
    {
      "epoch": 1.2592831772683242,
      "grad_norm": 0.5274289846420288,
      "learning_rate": 0.0005250274754875552,
      "loss": 3.7251,
      "step": 11700
    },
    {
      "epoch": 1.2646647293079325,
      "grad_norm": 0.5553549528121948,
      "learning_rate": 0.0005247042344574938,
      "loss": 3.7497,
      "step": 11750
    },
    {
      "epoch": 1.2700462813475406,
      "grad_norm": 0.5412461757659912,
      "learning_rate": 0.0005243809934274323,
      "loss": 3.7307,
      "step": 11800
    },
    {
      "epoch": 1.275427833387149,
      "grad_norm": 0.525709331035614,
      "learning_rate": 0.0005240577523973709,
      "loss": 3.7278,
      "step": 11850
    },
    {
      "epoch": 1.280809385426757,
      "grad_norm": 0.5819870829582214,
      "learning_rate": 0.0005237345113673095,
      "loss": 3.7434,
      "step": 11900
    },
    {
      "epoch": 1.2861909374663654,
      "grad_norm": 0.5735135078430176,
      "learning_rate": 0.0005234112703372481,
      "loss": 3.7017,
      "step": 11950
    },
    {
      "epoch": 1.2915724895059735,
      "grad_norm": 0.5071877241134644,
      "learning_rate": 0.0005230880293071867,
      "loss": 3.735,
      "step": 12000
    },
    {
      "epoch": 1.2915724895059735,
      "eval_accuracy": 0.34927274819334697,
      "eval_loss": 3.698476791381836,
      "eval_runtime": 190.0816,
      "eval_samples_per_second": 94.754,
      "eval_steps_per_second": 5.924,
      "step": 12000
    },
    {
      "epoch": 1.2969540415455818,
      "grad_norm": 0.6387230157852173,
      "learning_rate": 0.0005227647882771253,
      "loss": 3.7406,
      "step": 12050
    },
    {
      "epoch": 1.30233559358519,
      "grad_norm": 0.5852469801902771,
      "learning_rate": 0.0005224415472470639,
      "loss": 3.7085,
      "step": 12100
    },
    {
      "epoch": 1.3077171456247982,
      "grad_norm": 0.5338030457496643,
      "learning_rate": 0.0005221183062170024,
      "loss": 3.7475,
      "step": 12150
    },
    {
      "epoch": 1.3130986976644063,
      "grad_norm": 0.6056957244873047,
      "learning_rate": 0.0005217950651869409,
      "loss": 3.7335,
      "step": 12200
    },
    {
      "epoch": 1.3184802497040147,
      "grad_norm": 0.5705059170722961,
      "learning_rate": 0.0005214718241568796,
      "loss": 3.7206,
      "step": 12250
    },
    {
      "epoch": 1.3238618017436228,
      "grad_norm": 0.7600420713424683,
      "learning_rate": 0.0005211485831268182,
      "loss": 3.7084,
      "step": 12300
    },
    {
      "epoch": 1.329243353783231,
      "grad_norm": 0.5453450083732605,
      "learning_rate": 0.0005208253420967568,
      "loss": 3.7121,
      "step": 12350
    },
    {
      "epoch": 1.3346249058228392,
      "grad_norm": 0.6082261204719543,
      "learning_rate": 0.0005205021010666953,
      "loss": 3.737,
      "step": 12400
    },
    {
      "epoch": 1.3400064578624475,
      "grad_norm": 0.567171573638916,
      "learning_rate": 0.0005201788600366339,
      "loss": 3.725,
      "step": 12450
    },
    {
      "epoch": 1.3453880099020559,
      "grad_norm": 0.4867621064186096,
      "learning_rate": 0.0005198556190065725,
      "loss": 3.7155,
      "step": 12500
    },
    {
      "epoch": 1.350769561941664,
      "grad_norm": 0.5987406373023987,
      "learning_rate": 0.0005195323779765112,
      "loss": 3.7151,
      "step": 12550
    },
    {
      "epoch": 1.356151113981272,
      "grad_norm": 0.5686656832695007,
      "learning_rate": 0.0005192091369464497,
      "loss": 3.7116,
      "step": 12600
    },
    {
      "epoch": 1.3615326660208804,
      "grad_norm": 0.5483375191688538,
      "learning_rate": 0.0005188858959163882,
      "loss": 3.7158,
      "step": 12650
    },
    {
      "epoch": 1.3669142180604887,
      "grad_norm": 0.5874713063240051,
      "learning_rate": 0.0005185626548863269,
      "loss": 3.7336,
      "step": 12700
    },
    {
      "epoch": 1.3722957701000968,
      "grad_norm": 0.6102232336997986,
      "learning_rate": 0.0005182394138562654,
      "loss": 3.7276,
      "step": 12750
    },
    {
      "epoch": 1.3776773221397052,
      "grad_norm": 0.5877566337585449,
      "learning_rate": 0.0005179161728262041,
      "loss": 3.7092,
      "step": 12800
    },
    {
      "epoch": 1.3830588741793133,
      "grad_norm": 0.6303877830505371,
      "learning_rate": 0.0005175929317961426,
      "loss": 3.7222,
      "step": 12850
    },
    {
      "epoch": 1.3884404262189216,
      "grad_norm": 0.5772109627723694,
      "learning_rate": 0.0005172696907660812,
      "loss": 3.708,
      "step": 12900
    },
    {
      "epoch": 1.3938219782585297,
      "grad_norm": 0.5728392601013184,
      "learning_rate": 0.0005169464497360198,
      "loss": 3.7129,
      "step": 12950
    },
    {
      "epoch": 1.399203530298138,
      "grad_norm": 0.586063027381897,
      "learning_rate": 0.0005166232087059583,
      "loss": 3.6957,
      "step": 13000
    },
    {
      "epoch": 1.399203530298138,
      "eval_accuracy": 0.351647031097869,
      "eval_loss": 3.674795150756836,
      "eval_runtime": 189.8549,
      "eval_samples_per_second": 94.867,
      "eval_steps_per_second": 5.931,
      "step": 13000
    },
    {
      "epoch": 1.4045850823377461,
      "grad_norm": 0.5242763757705688,
      "learning_rate": 0.0005162999676758969,
      "loss": 3.7197,
      "step": 13050
    },
    {
      "epoch": 1.4099666343773545,
      "grad_norm": 0.5695936679840088,
      "learning_rate": 0.0005159767266458355,
      "loss": 3.715,
      "step": 13100
    },
    {
      "epoch": 1.4153481864169626,
      "grad_norm": 0.5991773009300232,
      "learning_rate": 0.0005156534856157741,
      "loss": 3.7094,
      "step": 13150
    },
    {
      "epoch": 1.420729738456571,
      "grad_norm": 0.569190502166748,
      "learning_rate": 0.0005153302445857127,
      "loss": 3.7116,
      "step": 13200
    },
    {
      "epoch": 1.426111290496179,
      "grad_norm": 0.5540894865989685,
      "learning_rate": 0.0005150070035556513,
      "loss": 3.7228,
      "step": 13250
    },
    {
      "epoch": 1.4314928425357873,
      "grad_norm": 0.5972295999526978,
      "learning_rate": 0.0005146837625255898,
      "loss": 3.6857,
      "step": 13300
    },
    {
      "epoch": 1.4368743945753955,
      "grad_norm": 0.5575430989265442,
      "learning_rate": 0.0005143605214955285,
      "loss": 3.6994,
      "step": 13350
    },
    {
      "epoch": 1.4422559466150038,
      "grad_norm": 0.6386284232139587,
      "learning_rate": 0.0005140372804654671,
      "loss": 3.7066,
      "step": 13400
    },
    {
      "epoch": 1.447637498654612,
      "grad_norm": 0.5550312399864197,
      "learning_rate": 0.0005137140394354056,
      "loss": 3.6983,
      "step": 13450
    },
    {
      "epoch": 1.4530190506942202,
      "grad_norm": 0.6196916699409485,
      "learning_rate": 0.0005133907984053442,
      "loss": 3.6945,
      "step": 13500
    },
    {
      "epoch": 1.4584006027338283,
      "grad_norm": 0.5429996252059937,
      "learning_rate": 0.0005130675573752827,
      "loss": 3.7142,
      "step": 13550
    },
    {
      "epoch": 1.4637821547734367,
      "grad_norm": 0.5590792894363403,
      "learning_rate": 0.0005127443163452214,
      "loss": 3.6994,
      "step": 13600
    },
    {
      "epoch": 1.469163706813045,
      "grad_norm": 0.6075905561447144,
      "learning_rate": 0.00051242107531516,
      "loss": 3.702,
      "step": 13650
    },
    {
      "epoch": 1.474545258852653,
      "grad_norm": 0.5841047167778015,
      "learning_rate": 0.0005120978342850986,
      "loss": 3.7131,
      "step": 13700
    },
    {
      "epoch": 1.4799268108922612,
      "grad_norm": 0.661999523639679,
      "learning_rate": 0.0005117745932550371,
      "loss": 3.7007,
      "step": 13750
    },
    {
      "epoch": 1.4853083629318695,
      "grad_norm": 0.5407006740570068,
      "learning_rate": 0.0005114513522249758,
      "loss": 3.7063,
      "step": 13800
    },
    {
      "epoch": 1.4906899149714778,
      "grad_norm": 0.5758219361305237,
      "learning_rate": 0.0005111345760155156,
      "loss": 3.69,
      "step": 13850
    },
    {
      "epoch": 1.496071467011086,
      "grad_norm": 0.6450278759002686,
      "learning_rate": 0.0005108113349854541,
      "loss": 3.6903,
      "step": 13900
    },
    {
      "epoch": 1.501453019050694,
      "grad_norm": 0.5611175298690796,
      "learning_rate": 0.0005104880939553926,
      "loss": 3.6972,
      "step": 13950
    },
    {
      "epoch": 1.5068345710903024,
      "grad_norm": 0.5707948207855225,
      "learning_rate": 0.0005101648529253313,
      "loss": 3.7009,
      "step": 14000
    },
    {
      "epoch": 1.5068345710903024,
      "eval_accuracy": 0.3540759664049617,
      "eval_loss": 3.655988931655884,
      "eval_runtime": 190.1764,
      "eval_samples_per_second": 94.707,
      "eval_steps_per_second": 5.921,
      "step": 14000
    },
    {
      "epoch": 1.5122161231299107,
      "grad_norm": 0.5171223878860474,
      "learning_rate": 0.0005098416118952699,
      "loss": 3.7128,
      "step": 14050
    },
    {
      "epoch": 1.5175976751695188,
      "grad_norm": 0.5456389784812927,
      "learning_rate": 0.0005095183708652085,
      "loss": 3.7077,
      "step": 14100
    },
    {
      "epoch": 1.5229792272091272,
      "grad_norm": 0.594289243221283,
      "learning_rate": 0.000509195129835147,
      "loss": 3.6759,
      "step": 14150
    },
    {
      "epoch": 1.5283607792487355,
      "grad_norm": 0.5579487085342407,
      "learning_rate": 0.0005088718888050856,
      "loss": 3.6748,
      "step": 14200
    },
    {
      "epoch": 1.5337423312883436,
      "grad_norm": 0.5491787791252136,
      "learning_rate": 0.0005085486477750242,
      "loss": 3.6736,
      "step": 14250
    },
    {
      "epoch": 1.5391238833279517,
      "grad_norm": 0.5308236479759216,
      "learning_rate": 0.0005082254067449629,
      "loss": 3.7048,
      "step": 14300
    },
    {
      "epoch": 1.54450543536756,
      "grad_norm": 0.5362514853477478,
      "learning_rate": 0.0005079021657149014,
      "loss": 3.6836,
      "step": 14350
    },
    {
      "epoch": 1.5498869874071683,
      "grad_norm": 0.5341024398803711,
      "learning_rate": 0.0005075789246848399,
      "loss": 3.6718,
      "step": 14400
    },
    {
      "epoch": 1.5552685394467765,
      "grad_norm": 0.5298555493354797,
      "learning_rate": 0.0005072556836547785,
      "loss": 3.6958,
      "step": 14450
    },
    {
      "epoch": 1.5606500914863846,
      "grad_norm": 0.6174782514572144,
      "learning_rate": 0.0005069324426247171,
      "loss": 3.6862,
      "step": 14500
    },
    {
      "epoch": 1.566031643525993,
      "grad_norm": 0.5435296893119812,
      "learning_rate": 0.0005066092015946557,
      "loss": 3.6909,
      "step": 14550
    },
    {
      "epoch": 1.5714131955656012,
      "grad_norm": 0.5510305166244507,
      "learning_rate": 0.0005062859605645943,
      "loss": 3.6793,
      "step": 14600
    },
    {
      "epoch": 1.5767947476052093,
      "grad_norm": 0.5439825057983398,
      "learning_rate": 0.0005059627195345329,
      "loss": 3.67,
      "step": 14650
    },
    {
      "epoch": 1.5821762996448174,
      "grad_norm": 0.5861722826957703,
      "learning_rate": 0.0005056394785044715,
      "loss": 3.6833,
      "step": 14700
    },
    {
      "epoch": 1.5875578516844258,
      "grad_norm": 0.5586617588996887,
      "learning_rate": 0.00050531623747441,
      "loss": 3.6821,
      "step": 14750
    },
    {
      "epoch": 1.592939403724034,
      "grad_norm": 0.5112905502319336,
      "learning_rate": 0.0005049929964443486,
      "loss": 3.6978,
      "step": 14800
    },
    {
      "epoch": 1.5983209557636422,
      "grad_norm": 0.5185847282409668,
      "learning_rate": 0.0005046697554142871,
      "loss": 3.6978,
      "step": 14850
    },
    {
      "epoch": 1.6037025078032503,
      "grad_norm": 0.5247318744659424,
      "learning_rate": 0.0005043465143842258,
      "loss": 3.693,
      "step": 14900
    },
    {
      "epoch": 1.6090840598428586,
      "grad_norm": 0.48752453923225403,
      "learning_rate": 0.0005040232733541644,
      "loss": 3.689,
      "step": 14950
    },
    {
      "epoch": 1.614465611882467,
      "grad_norm": 0.5501242280006409,
      "learning_rate": 0.000503700032324103,
      "loss": 3.6716,
      "step": 15000
    },
    {
      "epoch": 1.614465611882467,
      "eval_accuracy": 0.3549987553811701,
      "eval_loss": 3.6379735469818115,
      "eval_runtime": 189.9927,
      "eval_samples_per_second": 94.798,
      "eval_steps_per_second": 5.927,
      "step": 15000
    },
    {
      "epoch": 1.619847163922075,
      "grad_norm": 0.5478789210319519,
      "learning_rate": 0.0005033767912940415,
      "loss": 3.6686,
      "step": 15050
    },
    {
      "epoch": 1.6252287159616834,
      "grad_norm": 0.543006181716919,
      "learning_rate": 0.0005030535502639802,
      "loss": 3.671,
      "step": 15100
    },
    {
      "epoch": 1.6306102680012917,
      "grad_norm": 0.5525874495506287,
      "learning_rate": 0.0005027303092339187,
      "loss": 3.6546,
      "step": 15150
    },
    {
      "epoch": 1.6359918200408998,
      "grad_norm": 0.5512278079986572,
      "learning_rate": 0.0005024070682038573,
      "loss": 3.6633,
      "step": 15200
    },
    {
      "epoch": 1.641373372080508,
      "grad_norm": 0.5621350407600403,
      "learning_rate": 0.0005020838271737959,
      "loss": 3.6938,
      "step": 15250
    },
    {
      "epoch": 1.6467549241201163,
      "grad_norm": 0.5889851450920105,
      "learning_rate": 0.0005017605861437344,
      "loss": 3.6734,
      "step": 15300
    },
    {
      "epoch": 1.6521364761597246,
      "grad_norm": 0.5883168578147888,
      "learning_rate": 0.0005014373451136731,
      "loss": 3.6728,
      "step": 15350
    },
    {
      "epoch": 1.6575180281993327,
      "grad_norm": 0.5759154558181763,
      "learning_rate": 0.0005011141040836116,
      "loss": 3.6677,
      "step": 15400
    },
    {
      "epoch": 1.6628995802389408,
      "grad_norm": 0.6026946902275085,
      "learning_rate": 0.0005007908630535503,
      "loss": 3.6692,
      "step": 15450
    },
    {
      "epoch": 1.6682811322785491,
      "grad_norm": 0.51389080286026,
      "learning_rate": 0.0005004676220234888,
      "loss": 3.6809,
      "step": 15500
    },
    {
      "epoch": 1.6736626843181575,
      "grad_norm": 0.5595884323120117,
      "learning_rate": 0.0005001443809934273,
      "loss": 3.6563,
      "step": 15550
    },
    {
      "epoch": 1.6790442363577656,
      "grad_norm": 0.5687103271484375,
      "learning_rate": 0.000499821139963366,
      "loss": 3.6582,
      "step": 15600
    },
    {
      "epoch": 1.6844257883973737,
      "grad_norm": 0.5701649785041809,
      "learning_rate": 0.0004994978989333045,
      "loss": 3.6647,
      "step": 15650
    },
    {
      "epoch": 1.689807340436982,
      "grad_norm": 0.5454140901565552,
      "learning_rate": 0.0004991746579032431,
      "loss": 3.6767,
      "step": 15700
    },
    {
      "epoch": 1.6951888924765903,
      "grad_norm": 0.5374668836593628,
      "learning_rate": 0.0004988514168731817,
      "loss": 3.6617,
      "step": 15750
    },
    {
      "epoch": 1.7005704445161984,
      "grad_norm": 0.561478316783905,
      "learning_rate": 0.0004985281758431204,
      "loss": 3.6751,
      "step": 15800
    },
    {
      "epoch": 1.7059519965558065,
      "grad_norm": 0.5445524454116821,
      "learning_rate": 0.0004982049348130589,
      "loss": 3.6799,
      "step": 15850
    },
    {
      "epoch": 1.7113335485954149,
      "grad_norm": 0.5765650868415833,
      "learning_rate": 0.0004978881586035987,
      "loss": 3.6691,
      "step": 15900
    },
    {
      "epoch": 1.7167151006350232,
      "grad_norm": 0.5499103665351868,
      "learning_rate": 0.0004975649175735373,
      "loss": 3.6859,
      "step": 15950
    },
    {
      "epoch": 1.7220966526746313,
      "grad_norm": 0.6657256484031677,
      "learning_rate": 0.0004972416765434759,
      "loss": 3.6712,
      "step": 16000
    },
    {
      "epoch": 1.7220966526746313,
      "eval_accuracy": 0.35714801815502834,
      "eval_loss": 3.6184983253479004,
      "eval_runtime": 189.8405,
      "eval_samples_per_second": 94.874,
      "eval_steps_per_second": 5.931,
      "step": 16000
    },
    {
      "epoch": 1.7274782047142396,
      "grad_norm": 0.5714435577392578,
      "learning_rate": 0.0004969184355134145,
      "loss": 3.658,
      "step": 16050
    },
    {
      "epoch": 1.732859756753848,
      "grad_norm": 0.5882607698440552,
      "learning_rate": 0.0004965951944833531,
      "loss": 3.6608,
      "step": 16100
    },
    {
      "epoch": 1.738241308793456,
      "grad_norm": 0.6153818368911743,
      "learning_rate": 0.0004962719534532916,
      "loss": 3.6552,
      "step": 16150
    },
    {
      "epoch": 1.7436228608330642,
      "grad_norm": 0.5522136092185974,
      "learning_rate": 0.0004959487124232302,
      "loss": 3.6536,
      "step": 16200
    },
    {
      "epoch": 1.7490044128726725,
      "grad_norm": 0.5768370628356934,
      "learning_rate": 0.0004956254713931688,
      "loss": 3.6641,
      "step": 16250
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 0.5250673890113831,
      "learning_rate": 0.0004953022303631074,
      "loss": 3.6526,
      "step": 16300
    },
    {
      "epoch": 1.759767516951889,
      "grad_norm": 0.5400654673576355,
      "learning_rate": 0.0004949789893330459,
      "loss": 3.6458,
      "step": 16350
    },
    {
      "epoch": 1.765149068991497,
      "grad_norm": 0.6002535223960876,
      "learning_rate": 0.0004946557483029846,
      "loss": 3.6566,
      "step": 16400
    },
    {
      "epoch": 1.7705306210311054,
      "grad_norm": 0.5564870834350586,
      "learning_rate": 0.0004943325072729231,
      "loss": 3.6464,
      "step": 16450
    },
    {
      "epoch": 1.7759121730707137,
      "grad_norm": 0.5639302730560303,
      "learning_rate": 0.0004940092662428617,
      "loss": 3.6785,
      "step": 16500
    },
    {
      "epoch": 1.7812937251103218,
      "grad_norm": 0.535802960395813,
      "learning_rate": 0.0004936860252128003,
      "loss": 3.65,
      "step": 16550
    },
    {
      "epoch": 1.78667527714993,
      "grad_norm": 0.5739794373512268,
      "learning_rate": 0.0004933627841827388,
      "loss": 3.6538,
      "step": 16600
    },
    {
      "epoch": 1.7920568291895382,
      "grad_norm": 0.5378957986831665,
      "learning_rate": 0.0004930395431526775,
      "loss": 3.654,
      "step": 16650
    },
    {
      "epoch": 1.7974383812291466,
      "grad_norm": 0.5327773690223694,
      "learning_rate": 0.0004927163021226161,
      "loss": 3.6532,
      "step": 16700
    },
    {
      "epoch": 1.8028199332687547,
      "grad_norm": 0.5824779868125916,
      "learning_rate": 0.0004923930610925547,
      "loss": 3.658,
      "step": 16750
    },
    {
      "epoch": 1.8082014853083628,
      "grad_norm": 0.5551158785820007,
      "learning_rate": 0.0004920698200624932,
      "loss": 3.6335,
      "step": 16800
    },
    {
      "epoch": 1.813583037347971,
      "grad_norm": 0.5858696699142456,
      "learning_rate": 0.0004917465790324317,
      "loss": 3.66,
      "step": 16850
    },
    {
      "epoch": 1.8189645893875794,
      "grad_norm": 0.5639346837997437,
      "learning_rate": 0.0004914233380023704,
      "loss": 3.6588,
      "step": 16900
    },
    {
      "epoch": 1.8243461414271875,
      "grad_norm": 0.5276885032653809,
      "learning_rate": 0.0004911000969723089,
      "loss": 3.6257,
      "step": 16950
    },
    {
      "epoch": 1.8297276934667959,
      "grad_norm": 0.558270275592804,
      "learning_rate": 0.0004907768559422476,
      "loss": 3.6428,
      "step": 17000
    },
    {
      "epoch": 1.8297276934667959,
      "eval_accuracy": 0.3587097947644737,
      "eval_loss": 3.603337049484253,
      "eval_runtime": 190.2401,
      "eval_samples_per_second": 94.675,
      "eval_steps_per_second": 5.919,
      "step": 17000
    },
    {
      "epoch": 1.8351092455064042,
      "grad_norm": 0.596117377281189,
      "learning_rate": 0.0004904536149121861,
      "loss": 3.6476,
      "step": 17050
    },
    {
      "epoch": 1.8404907975460123,
      "grad_norm": 0.5245980620384216,
      "learning_rate": 0.0004901303738821248,
      "loss": 3.628,
      "step": 17100
    },
    {
      "epoch": 1.8458723495856204,
      "grad_norm": 0.5581666231155396,
      "learning_rate": 0.0004898071328520633,
      "loss": 3.6554,
      "step": 17150
    },
    {
      "epoch": 1.8512539016252287,
      "grad_norm": 0.547304093837738,
      "learning_rate": 0.0004894838918220019,
      "loss": 3.6456,
      "step": 17200
    },
    {
      "epoch": 1.856635453664837,
      "grad_norm": 0.5650014281272888,
      "learning_rate": 0.0004891606507919405,
      "loss": 3.6434,
      "step": 17250
    },
    {
      "epoch": 1.8620170057044452,
      "grad_norm": 0.5709023475646973,
      "learning_rate": 0.000488837409761879,
      "loss": 3.6561,
      "step": 17300
    },
    {
      "epoch": 1.8673985577440533,
      "grad_norm": 0.5545434951782227,
      "learning_rate": 0.0004885141687318177,
      "loss": 3.6456,
      "step": 17350
    },
    {
      "epoch": 1.8727801097836616,
      "grad_norm": 0.5061001777648926,
      "learning_rate": 0.00048819092770175623,
      "loss": 3.6526,
      "step": 17400
    },
    {
      "epoch": 1.87816166182327,
      "grad_norm": 0.5813119411468506,
      "learning_rate": 0.0004878676866716948,
      "loss": 3.6265,
      "step": 17450
    },
    {
      "epoch": 1.883543213862878,
      "grad_norm": 0.6016594171524048,
      "learning_rate": 0.00048754444564163337,
      "loss": 3.6488,
      "step": 17500
    },
    {
      "epoch": 1.8889247659024861,
      "grad_norm": 0.5019450783729553,
      "learning_rate": 0.000487221204611572,
      "loss": 3.6287,
      "step": 17550
    },
    {
      "epoch": 1.8943063179420945,
      "grad_norm": 0.6088424324989319,
      "learning_rate": 0.00048689796358151056,
      "loss": 3.6322,
      "step": 17600
    },
    {
      "epoch": 1.8996878699817028,
      "grad_norm": 0.5747146010398865,
      "learning_rate": 0.00048657472255144915,
      "loss": 3.6463,
      "step": 17650
    },
    {
      "epoch": 1.905069422021311,
      "grad_norm": 0.5056893229484558,
      "learning_rate": 0.00048625148152138775,
      "loss": 3.6397,
      "step": 17700
    },
    {
      "epoch": 1.910450974060919,
      "grad_norm": 0.5647739171981812,
      "learning_rate": 0.0004859282404913263,
      "loss": 3.6556,
      "step": 17750
    },
    {
      "epoch": 1.9158325261005273,
      "grad_norm": 0.5460042953491211,
      "learning_rate": 0.0004856049994612649,
      "loss": 3.6388,
      "step": 17800
    },
    {
      "epoch": 1.9212140781401357,
      "grad_norm": 0.5859935283660889,
      "learning_rate": 0.00048528175843120353,
      "loss": 3.6283,
      "step": 17850
    },
    {
      "epoch": 1.9265956301797438,
      "grad_norm": 0.5831868052482605,
      "learning_rate": 0.00048496498222174334,
      "loss": 3.6554,
      "step": 17900
    },
    {
      "epoch": 1.931977182219352,
      "grad_norm": 0.5555105805397034,
      "learning_rate": 0.00048464174119168193,
      "loss": 3.64,
      "step": 17950
    },
    {
      "epoch": 1.9373587342589604,
      "grad_norm": 0.5119417905807495,
      "learning_rate": 0.0004843185001616205,
      "loss": 3.6334,
      "step": 18000
    },
    {
      "epoch": 1.9373587342589604,
      "eval_accuracy": 0.3602852616377836,
      "eval_loss": 3.588430881500244,
      "eval_runtime": 189.8367,
      "eval_samples_per_second": 94.876,
      "eval_steps_per_second": 5.931,
      "step": 18000
    },
    {
      "epoch": 1.9427402862985685,
      "grad_norm": 0.5583716630935669,
      "learning_rate": 0.00048399525913155907,
      "loss": 3.6355,
      "step": 18050
    },
    {
      "epoch": 1.9481218383381766,
      "grad_norm": 0.5842928290367126,
      "learning_rate": 0.0004836720181014976,
      "loss": 3.6222,
      "step": 18100
    },
    {
      "epoch": 1.953503390377785,
      "grad_norm": 0.5783905982971191,
      "learning_rate": 0.0004833487770714362,
      "loss": 3.6408,
      "step": 18150
    },
    {
      "epoch": 1.9588849424173933,
      "grad_norm": 0.5660985708236694,
      "learning_rate": 0.00048302553604137485,
      "loss": 3.6451,
      "step": 18200
    },
    {
      "epoch": 1.9642664944570014,
      "grad_norm": 0.5207471251487732,
      "learning_rate": 0.0004827022950113134,
      "loss": 3.6263,
      "step": 18250
    },
    {
      "epoch": 1.9696480464966095,
      "grad_norm": 0.5245326161384583,
      "learning_rate": 0.000482379053981252,
      "loss": 3.615,
      "step": 18300
    },
    {
      "epoch": 1.9750295985362178,
      "grad_norm": 0.5456293225288391,
      "learning_rate": 0.0004820558129511906,
      "loss": 3.6388,
      "step": 18350
    },
    {
      "epoch": 1.9804111505758262,
      "grad_norm": 0.5086528658866882,
      "learning_rate": 0.0004817325719211291,
      "loss": 3.6411,
      "step": 18400
    },
    {
      "epoch": 1.9857927026154343,
      "grad_norm": 0.5529159903526306,
      "learning_rate": 0.0004814093308910677,
      "loss": 3.6433,
      "step": 18450
    },
    {
      "epoch": 1.9911742546550424,
      "grad_norm": 0.5180243849754333,
      "learning_rate": 0.00048108608986100637,
      "loss": 3.6266,
      "step": 18500
    },
    {
      "epoch": 1.9965558066946507,
      "grad_norm": 0.5631664991378784,
      "learning_rate": 0.0004807628488309449,
      "loss": 3.6199,
      "step": 18550
    },
    {
      "epoch": 2.001937358734259,
      "grad_norm": 0.5551926493644714,
      "learning_rate": 0.0004804396078008835,
      "loss": 3.5799,
      "step": 18600
    },
    {
      "epoch": 2.007318910773867,
      "grad_norm": 0.5640063881874084,
      "learning_rate": 0.00048011636677082204,
      "loss": 3.5381,
      "step": 18650
    },
    {
      "epoch": 2.0127004628134753,
      "grad_norm": 0.5894219875335693,
      "learning_rate": 0.00047979312574076064,
      "loss": 3.5414,
      "step": 18700
    },
    {
      "epoch": 2.018082014853084,
      "grad_norm": 0.6182511448860168,
      "learning_rate": 0.0004794698847106992,
      "loss": 3.5331,
      "step": 18750
    },
    {
      "epoch": 2.023463566892692,
      "grad_norm": 0.5752744674682617,
      "learning_rate": 0.0004791466436806378,
      "loss": 3.5405,
      "step": 18800
    },
    {
      "epoch": 2.0288451189323,
      "grad_norm": 0.5756574273109436,
      "learning_rate": 0.0004788234026505764,
      "loss": 3.546,
      "step": 18850
    },
    {
      "epoch": 2.034226670971908,
      "grad_norm": 0.5627039670944214,
      "learning_rate": 0.00047850016162051496,
      "loss": 3.56,
      "step": 18900
    },
    {
      "epoch": 2.0396082230115167,
      "grad_norm": 0.548718273639679,
      "learning_rate": 0.00047817692059045356,
      "loss": 3.5278,
      "step": 18950
    },
    {
      "epoch": 2.044989775051125,
      "grad_norm": 0.6026332974433899,
      "learning_rate": 0.00047785367956039215,
      "loss": 3.5421,
      "step": 19000
    },
    {
      "epoch": 2.044989775051125,
      "eval_accuracy": 0.3617015520304454,
      "eval_loss": 3.578596830368042,
      "eval_runtime": 189.8598,
      "eval_samples_per_second": 94.865,
      "eval_steps_per_second": 5.931,
      "step": 19000
    },
    {
      "epoch": 2.050371327090733,
      "grad_norm": 0.5351270437240601,
      "learning_rate": 0.00047753043853033075,
      "loss": 3.5495,
      "step": 19050
    },
    {
      "epoch": 2.055752879130341,
      "grad_norm": 0.6178998351097107,
      "learning_rate": 0.00047720719750026934,
      "loss": 3.5441,
      "step": 19100
    },
    {
      "epoch": 2.0611344311699495,
      "grad_norm": 0.6207836270332336,
      "learning_rate": 0.00047688395647020793,
      "loss": 3.5471,
      "step": 19150
    },
    {
      "epoch": 2.0665159832095576,
      "grad_norm": 0.5897159576416016,
      "learning_rate": 0.0004765607154401465,
      "loss": 3.5546,
      "step": 19200
    },
    {
      "epoch": 2.0718975352491658,
      "grad_norm": 0.5344468355178833,
      "learning_rate": 0.00047623747441008507,
      "loss": 3.5524,
      "step": 19250
    },
    {
      "epoch": 2.0772790872887743,
      "grad_norm": 0.5814924836158752,
      "learning_rate": 0.0004759142333800236,
      "loss": 3.5618,
      "step": 19300
    },
    {
      "epoch": 2.0826606393283824,
      "grad_norm": 0.5519033074378967,
      "learning_rate": 0.00047559099234996226,
      "loss": 3.5498,
      "step": 19350
    },
    {
      "epoch": 2.0880421913679905,
      "grad_norm": 0.5164206027984619,
      "learning_rate": 0.00047526775131990085,
      "loss": 3.5524,
      "step": 19400
    },
    {
      "epoch": 2.0934237434075986,
      "grad_norm": 0.550093948841095,
      "learning_rate": 0.0004749445102898394,
      "loss": 3.5449,
      "step": 19450
    },
    {
      "epoch": 2.098805295447207,
      "grad_norm": 0.6058476567268372,
      "learning_rate": 0.000474621269259778,
      "loss": 3.5416,
      "step": 19500
    },
    {
      "epoch": 2.1041868474868153,
      "grad_norm": 0.5596462488174438,
      "learning_rate": 0.0004742980282297166,
      "loss": 3.5547,
      "step": 19550
    },
    {
      "epoch": 2.1095683995264234,
      "grad_norm": 0.5401041507720947,
      "learning_rate": 0.0004739747871996551,
      "loss": 3.5674,
      "step": 19600
    },
    {
      "epoch": 2.1149499515660315,
      "grad_norm": 0.5709035396575928,
      "learning_rate": 0.00047365154616959377,
      "loss": 3.5573,
      "step": 19650
    },
    {
      "epoch": 2.12033150360564,
      "grad_norm": 0.5438994765281677,
      "learning_rate": 0.00047332830513953237,
      "loss": 3.5488,
      "step": 19700
    },
    {
      "epoch": 2.125713055645248,
      "grad_norm": 0.589968740940094,
      "learning_rate": 0.0004730050641094709,
      "loss": 3.5513,
      "step": 19750
    },
    {
      "epoch": 2.1310946076848563,
      "grad_norm": 0.5390543341636658,
      "learning_rate": 0.0004726818230794095,
      "loss": 3.5648,
      "step": 19800
    },
    {
      "epoch": 2.1364761597244644,
      "grad_norm": 0.5575344562530518,
      "learning_rate": 0.00047235858204934804,
      "loss": 3.5597,
      "step": 19850
    },
    {
      "epoch": 2.141857711764073,
      "grad_norm": 0.5531691312789917,
      "learning_rate": 0.0004720353410192867,
      "loss": 3.5554,
      "step": 19900
    },
    {
      "epoch": 2.147239263803681,
      "grad_norm": 0.5608752965927124,
      "learning_rate": 0.00047171856480982644,
      "loss": 3.5409,
      "step": 19950
    },
    {
      "epoch": 2.152620815843289,
      "grad_norm": 0.5748964548110962,
      "learning_rate": 0.0004713953237797651,
      "loss": 3.5783,
      "step": 20000
    },
    {
      "epoch": 2.152620815843289,
      "eval_accuracy": 0.3626011979415493,
      "eval_loss": 3.5692453384399414,
      "eval_runtime": 190.2183,
      "eval_samples_per_second": 94.686,
      "eval_steps_per_second": 5.92,
      "step": 20000
    },
    {
      "epoch": 2.1580023678828972,
      "grad_norm": 0.5758733153343201,
      "learning_rate": 0.0004710720827497037,
      "loss": 3.568,
      "step": 20050
    },
    {
      "epoch": 2.163383919922506,
      "grad_norm": 0.5911745429039001,
      "learning_rate": 0.00047074884171964223,
      "loss": 3.5588,
      "step": 20100
    },
    {
      "epoch": 2.168765471962114,
      "grad_norm": 0.615014374256134,
      "learning_rate": 0.0004704256006895808,
      "loss": 3.5652,
      "step": 20150
    },
    {
      "epoch": 2.174147024001722,
      "grad_norm": 0.5349318385124207,
      "learning_rate": 0.00047010235965951936,
      "loss": 3.5454,
      "step": 20200
    },
    {
      "epoch": 2.1795285760413305,
      "grad_norm": 0.582699716091156,
      "learning_rate": 0.00046977911862945796,
      "loss": 3.5537,
      "step": 20250
    },
    {
      "epoch": 2.1849101280809387,
      "grad_norm": 0.5917876958847046,
      "learning_rate": 0.0004694558775993966,
      "loss": 3.5707,
      "step": 20300
    },
    {
      "epoch": 2.1902916801205468,
      "grad_norm": 0.5654390454292297,
      "learning_rate": 0.00046913263656933515,
      "loss": 3.5555,
      "step": 20350
    },
    {
      "epoch": 2.195673232160155,
      "grad_norm": 0.6025635004043579,
      "learning_rate": 0.00046880939553927374,
      "loss": 3.5669,
      "step": 20400
    },
    {
      "epoch": 2.2010547841997634,
      "grad_norm": 0.5900663137435913,
      "learning_rate": 0.00046848615450921234,
      "loss": 3.5409,
      "step": 20450
    },
    {
      "epoch": 2.2064363362393715,
      "grad_norm": 0.577644944190979,
      "learning_rate": 0.0004681629134791509,
      "loss": 3.5865,
      "step": 20500
    },
    {
      "epoch": 2.2118178882789796,
      "grad_norm": 0.5554080605506897,
      "learning_rate": 0.00046783967244908947,
      "loss": 3.5456,
      "step": 20550
    },
    {
      "epoch": 2.2171994403185877,
      "grad_norm": 0.6127049922943115,
      "learning_rate": 0.0004675164314190281,
      "loss": 3.5463,
      "step": 20600
    },
    {
      "epoch": 2.2225809923581963,
      "grad_norm": 0.6620156168937683,
      "learning_rate": 0.00046719319038896666,
      "loss": 3.5595,
      "step": 20650
    },
    {
      "epoch": 2.2279625443978044,
      "grad_norm": 0.5703527331352234,
      "learning_rate": 0.00046686994935890526,
      "loss": 3.5583,
      "step": 20700
    },
    {
      "epoch": 2.2333440964374125,
      "grad_norm": 0.6113893389701843,
      "learning_rate": 0.0004665467083288438,
      "loss": 3.5553,
      "step": 20750
    },
    {
      "epoch": 2.2387256484770206,
      "grad_norm": 0.5821236371994019,
      "learning_rate": 0.0004662234672987824,
      "loss": 3.5519,
      "step": 20800
    },
    {
      "epoch": 2.244107200516629,
      "grad_norm": 0.6120278239250183,
      "learning_rate": 0.00046590022626872104,
      "loss": 3.5583,
      "step": 20850
    },
    {
      "epoch": 2.2494887525562373,
      "grad_norm": 0.6116923093795776,
      "learning_rate": 0.0004655769852386596,
      "loss": 3.5639,
      "step": 20900
    },
    {
      "epoch": 2.2548703045958454,
      "grad_norm": 0.5908465385437012,
      "learning_rate": 0.0004652537442085982,
      "loss": 3.5449,
      "step": 20950
    },
    {
      "epoch": 2.2602518566354535,
      "grad_norm": 0.5203503370285034,
      "learning_rate": 0.00046493050317853677,
      "loss": 3.5742,
      "step": 21000
    },
    {
      "epoch": 2.2602518566354535,
      "eval_accuracy": 0.3637896432284641,
      "eval_loss": 3.557666540145874,
      "eval_runtime": 189.505,
      "eval_samples_per_second": 95.042,
      "eval_steps_per_second": 5.942,
      "step": 21000
    },
    {
      "epoch": 2.265633408675062,
      "grad_norm": 0.5375860333442688,
      "learning_rate": 0.0004646072621484753,
      "loss": 3.5484,
      "step": 21050
    },
    {
      "epoch": 2.27101496071467,
      "grad_norm": 0.5789048671722412,
      "learning_rate": 0.0004642840211184139,
      "loss": 3.5621,
      "step": 21100
    },
    {
      "epoch": 2.2763965127542782,
      "grad_norm": 0.6134610176086426,
      "learning_rate": 0.00046396078008835255,
      "loss": 3.5604,
      "step": 21150
    },
    {
      "epoch": 2.281778064793887,
      "grad_norm": 0.5613147616386414,
      "learning_rate": 0.0004636375390582911,
      "loss": 3.5573,
      "step": 21200
    },
    {
      "epoch": 2.287159616833495,
      "grad_norm": 0.5626756548881531,
      "learning_rate": 0.0004633142980282297,
      "loss": 3.5476,
      "step": 21250
    },
    {
      "epoch": 2.292541168873103,
      "grad_norm": 0.5917938351631165,
      "learning_rate": 0.00046299105699816823,
      "loss": 3.5564,
      "step": 21300
    },
    {
      "epoch": 2.297922720912711,
      "grad_norm": 0.6403683423995972,
      "learning_rate": 0.0004626678159681068,
      "loss": 3.5681,
      "step": 21350
    },
    {
      "epoch": 2.303304272952319,
      "grad_norm": 0.5359059572219849,
      "learning_rate": 0.0004623445749380454,
      "loss": 3.5374,
      "step": 21400
    },
    {
      "epoch": 2.3086858249919278,
      "grad_norm": 0.6532794833183289,
      "learning_rate": 0.000462021333907984,
      "loss": 3.5407,
      "step": 21450
    },
    {
      "epoch": 2.314067377031536,
      "grad_norm": 0.5737901329994202,
      "learning_rate": 0.0004616980928779226,
      "loss": 3.5469,
      "step": 21500
    },
    {
      "epoch": 2.319448929071144,
      "grad_norm": 0.6427896022796631,
      "learning_rate": 0.0004613748518478612,
      "loss": 3.5502,
      "step": 21550
    },
    {
      "epoch": 2.3248304811107525,
      "grad_norm": 0.539666473865509,
      "learning_rate": 0.00046105161081779974,
      "loss": 3.5552,
      "step": 21600
    },
    {
      "epoch": 2.3302120331503606,
      "grad_norm": 0.5903276205062866,
      "learning_rate": 0.00046072836978773834,
      "loss": 3.5687,
      "step": 21650
    },
    {
      "epoch": 2.3355935851899687,
      "grad_norm": 0.5565731525421143,
      "learning_rate": 0.000460405128757677,
      "loss": 3.5291,
      "step": 21700
    },
    {
      "epoch": 2.340975137229577,
      "grad_norm": 0.5656738877296448,
      "learning_rate": 0.0004600818877276155,
      "loss": 3.5506,
      "step": 21750
    },
    {
      "epoch": 2.3463566892691854,
      "grad_norm": 0.5375106334686279,
      "learning_rate": 0.0004597586466975541,
      "loss": 3.569,
      "step": 21800
    },
    {
      "epoch": 2.3517382413087935,
      "grad_norm": 0.585565447807312,
      "learning_rate": 0.00045943540566749266,
      "loss": 3.5602,
      "step": 21850
    },
    {
      "epoch": 2.3571197933484016,
      "grad_norm": 0.6016998291015625,
      "learning_rate": 0.00045911216463743126,
      "loss": 3.5668,
      "step": 21900
    },
    {
      "epoch": 2.3625013453880097,
      "grad_norm": 0.5457884073257446,
      "learning_rate": 0.00045878892360736985,
      "loss": 3.5591,
      "step": 21950
    },
    {
      "epoch": 2.3678828974276183,
      "grad_norm": 0.5570237040519714,
      "learning_rate": 0.00045847214739790966,
      "loss": 3.5623,
      "step": 22000
    },
    {
      "epoch": 2.3678828974276183,
      "eval_accuracy": 0.36518561553110457,
      "eval_loss": 3.54646372795105,
      "eval_runtime": 190.4088,
      "eval_samples_per_second": 94.591,
      "eval_steps_per_second": 5.914,
      "step": 22000
    },
    {
      "epoch": 2.3732644494672264,
      "grad_norm": 0.6019724607467651,
      "learning_rate": 0.0004581489063678482,
      "loss": 3.5546,
      "step": 22050
    },
    {
      "epoch": 2.3786460015068345,
      "grad_norm": 0.5821262001991272,
      "learning_rate": 0.00045782566533778685,
      "loss": 3.5505,
      "step": 22100
    },
    {
      "epoch": 2.384027553546443,
      "grad_norm": 0.5628877878189087,
      "learning_rate": 0.00045750242430772544,
      "loss": 3.5551,
      "step": 22150
    },
    {
      "epoch": 2.389409105586051,
      "grad_norm": 0.5676998496055603,
      "learning_rate": 0.000457179183277664,
      "loss": 3.5563,
      "step": 22200
    },
    {
      "epoch": 2.3947906576256592,
      "grad_norm": 0.6015381813049316,
      "learning_rate": 0.0004568559422476026,
      "loss": 3.5611,
      "step": 22250
    },
    {
      "epoch": 2.4001722096652673,
      "grad_norm": 0.5897294878959656,
      "learning_rate": 0.00045653270121754117,
      "loss": 3.5532,
      "step": 22300
    },
    {
      "epoch": 2.4055537617048754,
      "grad_norm": 0.5502423644065857,
      "learning_rate": 0.0004562094601874797,
      "loss": 3.5334,
      "step": 22350
    },
    {
      "epoch": 2.410935313744484,
      "grad_norm": 0.6234706044197083,
      "learning_rate": 0.00045588621915741836,
      "loss": 3.5598,
      "step": 22400
    },
    {
      "epoch": 2.416316865784092,
      "grad_norm": 0.5574871897697449,
      "learning_rate": 0.00045556297812735696,
      "loss": 3.5558,
      "step": 22450
    },
    {
      "epoch": 2.4216984178237,
      "grad_norm": 0.5500792264938354,
      "learning_rate": 0.0004552397370972955,
      "loss": 3.555,
      "step": 22500
    },
    {
      "epoch": 2.4270799698633088,
      "grad_norm": 0.5803767442703247,
      "learning_rate": 0.0004549164960672341,
      "loss": 3.5593,
      "step": 22550
    },
    {
      "epoch": 2.432461521902917,
      "grad_norm": 0.6330329179763794,
      "learning_rate": 0.00045459325503717263,
      "loss": 3.547,
      "step": 22600
    },
    {
      "epoch": 2.437843073942525,
      "grad_norm": 0.5570193529129028,
      "learning_rate": 0.0004542700140071113,
      "loss": 3.5432,
      "step": 22650
    },
    {
      "epoch": 2.443224625982133,
      "grad_norm": 0.5483201742172241,
      "learning_rate": 0.0004539467729770499,
      "loss": 3.5498,
      "step": 22700
    },
    {
      "epoch": 2.4486061780217416,
      "grad_norm": 0.6233546137809753,
      "learning_rate": 0.0004536235319469884,
      "loss": 3.5491,
      "step": 22750
    },
    {
      "epoch": 2.4539877300613497,
      "grad_norm": 0.5224318504333496,
      "learning_rate": 0.000453300290916927,
      "loss": 3.5488,
      "step": 22800
    },
    {
      "epoch": 2.459369282100958,
      "grad_norm": 0.5755411386489868,
      "learning_rate": 0.0004529770498868656,
      "loss": 3.5665,
      "step": 22850
    },
    {
      "epoch": 2.464750834140566,
      "grad_norm": 0.6066120862960815,
      "learning_rate": 0.00045265380885680414,
      "loss": 3.5557,
      "step": 22900
    },
    {
      "epoch": 2.4701323861801745,
      "grad_norm": 0.5412178039550781,
      "learning_rate": 0.0004523305678267428,
      "loss": 3.5353,
      "step": 22950
    },
    {
      "epoch": 2.4755139382197826,
      "grad_norm": 0.526543915271759,
      "learning_rate": 0.0004520073267966814,
      "loss": 3.5432,
      "step": 23000
    },
    {
      "epoch": 2.4755139382197826,
      "eval_accuracy": 0.36600040353682534,
      "eval_loss": 3.536306858062744,
      "eval_runtime": 189.7162,
      "eval_samples_per_second": 94.937,
      "eval_steps_per_second": 5.935,
      "step": 23000
    },
    {
      "epoch": 2.4808954902593907,
      "grad_norm": 0.5737837553024292,
      "learning_rate": 0.00045168408576661993,
      "loss": 3.5504,
      "step": 23050
    },
    {
      "epoch": 2.4862770422989993,
      "grad_norm": 0.6918070316314697,
      "learning_rate": 0.0004513608447365585,
      "loss": 3.5575,
      "step": 23100
    },
    {
      "epoch": 2.4916585943386074,
      "grad_norm": 0.6289458870887756,
      "learning_rate": 0.00045103760370649706,
      "loss": 3.5706,
      "step": 23150
    },
    {
      "epoch": 2.4970401463782155,
      "grad_norm": 0.605403482913971,
      "learning_rate": 0.00045071436267643566,
      "loss": 3.5448,
      "step": 23200
    },
    {
      "epoch": 2.5024216984178236,
      "grad_norm": 0.6292030811309814,
      "learning_rate": 0.0004503911216463743,
      "loss": 3.5505,
      "step": 23250
    },
    {
      "epoch": 2.5078032504574317,
      "grad_norm": 0.6145636439323425,
      "learning_rate": 0.00045006788061631285,
      "loss": 3.5531,
      "step": 23300
    },
    {
      "epoch": 2.5131848024970402,
      "grad_norm": 0.5649071335792542,
      "learning_rate": 0.00044974463958625144,
      "loss": 3.5292,
      "step": 23350
    },
    {
      "epoch": 2.5185663545366483,
      "grad_norm": 0.5961219668388367,
      "learning_rate": 0.00044942139855619004,
      "loss": 3.554,
      "step": 23400
    },
    {
      "epoch": 2.5239479065762565,
      "grad_norm": 0.5621718764305115,
      "learning_rate": 0.0004490981575261286,
      "loss": 3.5375,
      "step": 23450
    },
    {
      "epoch": 2.529329458615865,
      "grad_norm": 0.5819845199584961,
      "learning_rate": 0.0004487749164960672,
      "loss": 3.5435,
      "step": 23500
    },
    {
      "epoch": 2.534711010655473,
      "grad_norm": 0.5559252500534058,
      "learning_rate": 0.0004484516754660058,
      "loss": 3.5393,
      "step": 23550
    },
    {
      "epoch": 2.540092562695081,
      "grad_norm": 0.6207060813903809,
      "learning_rate": 0.00044812843443594436,
      "loss": 3.5437,
      "step": 23600
    },
    {
      "epoch": 2.5454741147346893,
      "grad_norm": 0.5800049901008606,
      "learning_rate": 0.00044780519340588296,
      "loss": 3.5499,
      "step": 23650
    },
    {
      "epoch": 2.550855666774298,
      "grad_norm": 0.6205190420150757,
      "learning_rate": 0.0004474819523758215,
      "loss": 3.5248,
      "step": 23700
    },
    {
      "epoch": 2.556237218813906,
      "grad_norm": 0.5872529149055481,
      "learning_rate": 0.0004471587113457601,
      "loss": 3.5573,
      "step": 23750
    },
    {
      "epoch": 2.561618770853514,
      "grad_norm": 0.5815856456756592,
      "learning_rate": 0.00044683547031569874,
      "loss": 3.5473,
      "step": 23800
    },
    {
      "epoch": 2.567000322893122,
      "grad_norm": 0.5739285349845886,
      "learning_rate": 0.0004465122292856373,
      "loss": 3.5485,
      "step": 23850
    },
    {
      "epoch": 2.5723818749327307,
      "grad_norm": 0.6001350283622742,
      "learning_rate": 0.0004461889882555759,
      "loss": 3.5385,
      "step": 23900
    },
    {
      "epoch": 2.577763426972339,
      "grad_norm": 0.5841218829154968,
      "learning_rate": 0.00044586574722551447,
      "loss": 3.5312,
      "step": 23950
    },
    {
      "epoch": 2.583144979011947,
      "grad_norm": 0.5895936489105225,
      "learning_rate": 0.0004455489710160543,
      "loss": 3.535,
      "step": 24000
    },
    {
      "epoch": 2.583144979011947,
      "eval_accuracy": 0.36708486801010165,
      "eval_loss": 3.5288374423980713,
      "eval_runtime": 190.536,
      "eval_samples_per_second": 94.528,
      "eval_steps_per_second": 5.91,
      "step": 24000
    },
    {
      "epoch": 2.5885265310515555,
      "grad_norm": 0.5898272395133972,
      "learning_rate": 0.0004452257299859928,
      "loss": 3.5419,
      "step": 24050
    },
    {
      "epoch": 2.5939080830911636,
      "grad_norm": 0.6210259199142456,
      "learning_rate": 0.0004449024889559314,
      "loss": 3.5412,
      "step": 24100
    },
    {
      "epoch": 2.5992896351307717,
      "grad_norm": 0.5567688345909119,
      "learning_rate": 0.00044457924792587,
      "loss": 3.524,
      "step": 24150
    },
    {
      "epoch": 2.60467118717038,
      "grad_norm": 0.6107264161109924,
      "learning_rate": 0.0004442560068958086,
      "loss": 3.5461,
      "step": 24200
    },
    {
      "epoch": 2.610052739209988,
      "grad_norm": 0.5399898886680603,
      "learning_rate": 0.0004439327658657472,
      "loss": 3.53,
      "step": 24250
    },
    {
      "epoch": 2.6154342912495965,
      "grad_norm": 0.6111960411071777,
      "learning_rate": 0.0004436095248356858,
      "loss": 3.5517,
      "step": 24300
    },
    {
      "epoch": 2.6208158432892046,
      "grad_norm": 0.5308720469474792,
      "learning_rate": 0.00044328628380562433,
      "loss": 3.5477,
      "step": 24350
    },
    {
      "epoch": 2.6261973953288127,
      "grad_norm": 0.5515609979629517,
      "learning_rate": 0.0004429630427755629,
      "loss": 3.5439,
      "step": 24400
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.5771924257278442,
      "learning_rate": 0.0004426398017455016,
      "loss": 3.5249,
      "step": 24450
    },
    {
      "epoch": 2.6369604994080293,
      "grad_norm": 0.5828399658203125,
      "learning_rate": 0.0004423165607154401,
      "loss": 3.5582,
      "step": 24500
    },
    {
      "epoch": 2.6423420514476375,
      "grad_norm": 0.6540676355361938,
      "learning_rate": 0.0004419933196853787,
      "loss": 3.5508,
      "step": 24550
    },
    {
      "epoch": 2.6477236034872456,
      "grad_norm": 0.5388296842575073,
      "learning_rate": 0.00044167007865531725,
      "loss": 3.5068,
      "step": 24600
    },
    {
      "epoch": 2.653105155526854,
      "grad_norm": 0.5928964018821716,
      "learning_rate": 0.00044134683762525584,
      "loss": 3.5529,
      "step": 24650
    },
    {
      "epoch": 2.658486707566462,
      "grad_norm": 0.5698816776275635,
      "learning_rate": 0.00044102359659519444,
      "loss": 3.546,
      "step": 24700
    },
    {
      "epoch": 2.6638682596060703,
      "grad_norm": 0.5650924444198608,
      "learning_rate": 0.00044070035556513303,
      "loss": 3.5354,
      "step": 24750
    },
    {
      "epoch": 2.6692498116456784,
      "grad_norm": 0.5819073915481567,
      "learning_rate": 0.00044037711453507163,
      "loss": 3.5468,
      "step": 24800
    },
    {
      "epoch": 2.674631363685287,
      "grad_norm": 0.5399896502494812,
      "learning_rate": 0.0004400538735050102,
      "loss": 3.5441,
      "step": 24850
    },
    {
      "epoch": 2.680012915724895,
      "grad_norm": 0.5661169290542603,
      "learning_rate": 0.00043973063247494876,
      "loss": 3.5284,
      "step": 24900
    },
    {
      "epoch": 2.685394467764503,
      "grad_norm": 0.5845262408256531,
      "learning_rate": 0.00043940739144488736,
      "loss": 3.5371,
      "step": 24950
    },
    {
      "epoch": 2.6907760198041117,
      "grad_norm": 0.6130923628807068,
      "learning_rate": 0.0004390841504148259,
      "loss": 3.5387,
      "step": 25000
    },
    {
      "epoch": 2.6907760198041117,
      "eval_accuracy": 0.36775992840209304,
      "eval_loss": 3.519505023956299,
      "eval_runtime": 190.2059,
      "eval_samples_per_second": 94.692,
      "eval_steps_per_second": 5.92,
      "step": 25000
    },
    {
      "epoch": 2.69615757184372,
      "grad_norm": 0.5774505734443665,
      "learning_rate": 0.00043876090938476455,
      "loss": 3.5409,
      "step": 25050
    },
    {
      "epoch": 2.701539123883328,
      "grad_norm": 0.5982598662376404,
      "learning_rate": 0.00043843766835470314,
      "loss": 3.55,
      "step": 25100
    },
    {
      "epoch": 2.706920675922936,
      "grad_norm": 0.5944047570228577,
      "learning_rate": 0.0004381144273246417,
      "loss": 3.549,
      "step": 25150
    },
    {
      "epoch": 2.712302227962544,
      "grad_norm": 0.5446933507919312,
      "learning_rate": 0.0004377911862945803,
      "loss": 3.5413,
      "step": 25200
    },
    {
      "epoch": 2.7176837800021527,
      "grad_norm": 0.6071139574050903,
      "learning_rate": 0.00043746794526451887,
      "loss": 3.534,
      "step": 25250
    },
    {
      "epoch": 2.723065332041761,
      "grad_norm": 0.6073064208030701,
      "learning_rate": 0.00043714470423445747,
      "loss": 3.5375,
      "step": 25300
    },
    {
      "epoch": 2.728446884081369,
      "grad_norm": 0.576163113117218,
      "learning_rate": 0.00043682146320439606,
      "loss": 3.5237,
      "step": 25350
    },
    {
      "epoch": 2.7338284361209775,
      "grad_norm": 0.5555893182754517,
      "learning_rate": 0.00043649822217433466,
      "loss": 3.5348,
      "step": 25400
    },
    {
      "epoch": 2.7392099881605856,
      "grad_norm": 0.6135348081588745,
      "learning_rate": 0.0004361749811442732,
      "loss": 3.5322,
      "step": 25450
    },
    {
      "epoch": 2.7445915402001937,
      "grad_norm": 0.570088267326355,
      "learning_rate": 0.0004358517401142118,
      "loss": 3.5202,
      "step": 25500
    },
    {
      "epoch": 2.749973092239802,
      "grad_norm": 0.598089337348938,
      "learning_rate": 0.00043552849908415033,
      "loss": 3.5317,
      "step": 25550
    },
    {
      "epoch": 2.7553546442794103,
      "grad_norm": 0.6342823505401611,
      "learning_rate": 0.000435205258054089,
      "loss": 3.5293,
      "step": 25600
    },
    {
      "epoch": 2.7607361963190185,
      "grad_norm": 0.599950909614563,
      "learning_rate": 0.0004348820170240276,
      "loss": 3.5387,
      "step": 25650
    },
    {
      "epoch": 2.7661177483586266,
      "grad_norm": 0.5143266916275024,
      "learning_rate": 0.0004345587759939661,
      "loss": 3.5348,
      "step": 25700
    },
    {
      "epoch": 2.7714993003982347,
      "grad_norm": 0.5734071731567383,
      "learning_rate": 0.0004342355349639047,
      "loss": 3.526,
      "step": 25750
    },
    {
      "epoch": 2.776880852437843,
      "grad_norm": 0.6031785607337952,
      "learning_rate": 0.0004339122939338433,
      "loss": 3.5182,
      "step": 25800
    },
    {
      "epoch": 2.7822624044774513,
      "grad_norm": 0.5556403994560242,
      "learning_rate": 0.00043358905290378184,
      "loss": 3.5134,
      "step": 25850
    },
    {
      "epoch": 2.7876439565170594,
      "grad_norm": 0.531978189945221,
      "learning_rate": 0.0004332658118737205,
      "loss": 3.53,
      "step": 25900
    },
    {
      "epoch": 2.793025508556668,
      "grad_norm": 0.6397167444229126,
      "learning_rate": 0.0004329425708436591,
      "loss": 3.5385,
      "step": 25950
    },
    {
      "epoch": 2.798407060596276,
      "grad_norm": 0.5569565892219543,
      "learning_rate": 0.00043261932981359763,
      "loss": 3.5256,
      "step": 26000
    },
    {
      "epoch": 2.798407060596276,
      "eval_accuracy": 0.36892544792967896,
      "eval_loss": 3.508462905883789,
      "eval_runtime": 190.1152,
      "eval_samples_per_second": 94.737,
      "eval_steps_per_second": 5.923,
      "step": 26000
    },
    {
      "epoch": 2.803788612635884,
      "grad_norm": 0.5564899444580078,
      "learning_rate": 0.00043230255360413744,
      "loss": 3.5426,
      "step": 26050
    },
    {
      "epoch": 2.8091701646754923,
      "grad_norm": 0.5744142532348633,
      "learning_rate": 0.00043197931257407603,
      "loss": 3.5441,
      "step": 26100
    },
    {
      "epoch": 2.8145517167151004,
      "grad_norm": 0.5524824261665344,
      "learning_rate": 0.0004316560715440146,
      "loss": 3.527,
      "step": 26150
    },
    {
      "epoch": 2.819933268754709,
      "grad_norm": 0.5708995461463928,
      "learning_rate": 0.00043133283051395317,
      "loss": 3.5306,
      "step": 26200
    },
    {
      "epoch": 2.825314820794317,
      "grad_norm": 0.6121014952659607,
      "learning_rate": 0.0004310095894838918,
      "loss": 3.5425,
      "step": 26250
    },
    {
      "epoch": 2.830696372833925,
      "grad_norm": 0.6067882180213928,
      "learning_rate": 0.0004306863484538304,
      "loss": 3.5187,
      "step": 26300
    },
    {
      "epoch": 2.8360779248735337,
      "grad_norm": 0.6431559324264526,
      "learning_rate": 0.00043036310742376895,
      "loss": 3.5378,
      "step": 26350
    },
    {
      "epoch": 2.841459476913142,
      "grad_norm": 0.5877844095230103,
      "learning_rate": 0.00043003986639370754,
      "loss": 3.5157,
      "step": 26400
    },
    {
      "epoch": 2.84684102895275,
      "grad_norm": 0.5687451958656311,
      "learning_rate": 0.0004297166253636461,
      "loss": 3.5265,
      "step": 26450
    },
    {
      "epoch": 2.852222580992358,
      "grad_norm": 0.5869263410568237,
      "learning_rate": 0.0004293933843335847,
      "loss": 3.5083,
      "step": 26500
    },
    {
      "epoch": 2.857604133031966,
      "grad_norm": 0.6137527823448181,
      "learning_rate": 0.00042907014330352333,
      "loss": 3.5378,
      "step": 26550
    },
    {
      "epoch": 2.8629856850715747,
      "grad_norm": 0.5713273286819458,
      "learning_rate": 0.00042874690227346187,
      "loss": 3.529,
      "step": 26600
    },
    {
      "epoch": 2.868367237111183,
      "grad_norm": 0.5419256687164307,
      "learning_rate": 0.00042842366124340046,
      "loss": 3.5192,
      "step": 26650
    },
    {
      "epoch": 2.873748789150791,
      "grad_norm": 0.5879168510437012,
      "learning_rate": 0.00042810042021333906,
      "loss": 3.5021,
      "step": 26700
    },
    {
      "epoch": 2.8791303411903995,
      "grad_norm": 0.5535828471183777,
      "learning_rate": 0.0004277771791832776,
      "loss": 3.5147,
      "step": 26750
    },
    {
      "epoch": 2.8845118932300076,
      "grad_norm": 0.5824949741363525,
      "learning_rate": 0.0004274539381532162,
      "loss": 3.5029,
      "step": 26800
    },
    {
      "epoch": 2.8898934452696157,
      "grad_norm": 0.5599977970123291,
      "learning_rate": 0.00042713069712315484,
      "loss": 3.5362,
      "step": 26850
    },
    {
      "epoch": 2.895274997309224,
      "grad_norm": 0.5262795686721802,
      "learning_rate": 0.0004268074560930934,
      "loss": 3.5372,
      "step": 26900
    },
    {
      "epoch": 2.9006565493488323,
      "grad_norm": 0.5597524046897888,
      "learning_rate": 0.000426484215063032,
      "loss": 3.5223,
      "step": 26950
    },
    {
      "epoch": 2.9060381013884404,
      "grad_norm": 0.582640528678894,
      "learning_rate": 0.0004261609740329705,
      "loss": 3.5445,
      "step": 27000
    },
    {
      "epoch": 2.9060381013884404,
      "eval_accuracy": 0.36968449700395095,
      "eval_loss": 3.5020923614501953,
      "eval_runtime": 190.2601,
      "eval_samples_per_second": 94.665,
      "eval_steps_per_second": 5.918,
      "step": 27000
    },
    {
      "epoch": 2.9114196534280485,
      "grad_norm": 0.5409820675849915,
      "learning_rate": 0.0004258377330029091,
      "loss": 3.5292,
      "step": 27050
    },
    {
      "epoch": 2.9168012054676566,
      "grad_norm": 0.5684029459953308,
      "learning_rate": 0.00042551449197284776,
      "loss": 3.5164,
      "step": 27100
    },
    {
      "epoch": 2.922182757507265,
      "grad_norm": 0.565455973148346,
      "learning_rate": 0.0004251912509427863,
      "loss": 3.5219,
      "step": 27150
    },
    {
      "epoch": 2.9275643095468733,
      "grad_norm": 0.5930619239807129,
      "learning_rate": 0.0004248680099127249,
      "loss": 3.5126,
      "step": 27200
    },
    {
      "epoch": 2.9329458615864814,
      "grad_norm": 0.6132225394248962,
      "learning_rate": 0.0004245447688826635,
      "loss": 3.5148,
      "step": 27250
    },
    {
      "epoch": 2.93832741362609,
      "grad_norm": 0.5916863083839417,
      "learning_rate": 0.00042422152785260203,
      "loss": 3.507,
      "step": 27300
    },
    {
      "epoch": 2.943708965665698,
      "grad_norm": 0.6536251306533813,
      "learning_rate": 0.0004238982868225406,
      "loss": 3.5338,
      "step": 27350
    },
    {
      "epoch": 2.949090517705306,
      "grad_norm": 0.5629107356071472,
      "learning_rate": 0.0004235750457924793,
      "loss": 3.5218,
      "step": 27400
    },
    {
      "epoch": 2.9544720697449143,
      "grad_norm": 0.590984046459198,
      "learning_rate": 0.0004232518047624178,
      "loss": 3.5115,
      "step": 27450
    },
    {
      "epoch": 2.9598536217845224,
      "grad_norm": 0.5421058535575867,
      "learning_rate": 0.0004229285637323564,
      "loss": 3.5289,
      "step": 27500
    },
    {
      "epoch": 2.965235173824131,
      "grad_norm": 0.6006096601486206,
      "learning_rate": 0.00042260532270229495,
      "loss": 3.5213,
      "step": 27550
    },
    {
      "epoch": 2.970616725863739,
      "grad_norm": 0.62128084897995,
      "learning_rate": 0.00042228208167223354,
      "loss": 3.5035,
      "step": 27600
    },
    {
      "epoch": 2.975998277903347,
      "grad_norm": 0.6029218435287476,
      "learning_rate": 0.00042195884064217214,
      "loss": 3.5262,
      "step": 27650
    },
    {
      "epoch": 2.9813798299429557,
      "grad_norm": 0.5777669548988342,
      "learning_rate": 0.00042163559961211073,
      "loss": 3.5318,
      "step": 27700
    },
    {
      "epoch": 2.986761381982564,
      "grad_norm": 0.5785093903541565,
      "learning_rate": 0.00042131235858204933,
      "loss": 3.5129,
      "step": 27750
    },
    {
      "epoch": 2.992142934022172,
      "grad_norm": 0.58424973487854,
      "learning_rate": 0.0004209891175519879,
      "loss": 3.5163,
      "step": 27800
    },
    {
      "epoch": 2.9975244860617805,
      "grad_norm": 0.5958218574523926,
      "learning_rate": 0.00042066587652192646,
      "loss": 3.5132,
      "step": 27850
    },
    {
      "epoch": 3.0029060381013886,
      "grad_norm": 0.598947286605835,
      "learning_rate": 0.00042034263549186506,
      "loss": 3.4755,
      "step": 27900
    },
    {
      "epoch": 3.0082875901409967,
      "grad_norm": 0.6067318916320801,
      "learning_rate": 0.0004200193944618036,
      "loss": 3.434,
      "step": 27950
    },
    {
      "epoch": 3.0136691421806048,
      "grad_norm": 0.6365818977355957,
      "learning_rate": 0.00041969615343174225,
      "loss": 3.4107,
      "step": 28000
    },
    {
      "epoch": 3.0136691421806048,
      "eval_accuracy": 0.37059598607982663,
      "eval_loss": 3.4984636306762695,
      "eval_runtime": 190.2647,
      "eval_samples_per_second": 94.663,
      "eval_steps_per_second": 5.918,
      "step": 28000
    },
    {
      "epoch": 3.0190506942202133,
      "grad_norm": 0.6193200945854187,
      "learning_rate": 0.00041937937722228205,
      "loss": 3.4429,
      "step": 28050
    },
    {
      "epoch": 3.0244322462598214,
      "grad_norm": 0.5575525164604187,
      "learning_rate": 0.00041905613619222065,
      "loss": 3.4201,
      "step": 28100
    },
    {
      "epoch": 3.0298137982994295,
      "grad_norm": 0.6336436867713928,
      "learning_rate": 0.00041873289516215924,
      "loss": 3.4075,
      "step": 28150
    },
    {
      "epoch": 3.0351953503390376,
      "grad_norm": 0.6951696276664734,
      "learning_rate": 0.0004184096541320978,
      "loss": 3.4136,
      "step": 28200
    },
    {
      "epoch": 3.040576902378646,
      "grad_norm": 0.5632196068763733,
      "learning_rate": 0.0004180864131020364,
      "loss": 3.4288,
      "step": 28250
    },
    {
      "epoch": 3.0459584544182543,
      "grad_norm": 0.60652756690979,
      "learning_rate": 0.0004177631720719749,
      "loss": 3.4253,
      "step": 28300
    },
    {
      "epoch": 3.0513400064578624,
      "grad_norm": 0.6440793871879578,
      "learning_rate": 0.00041743993104191357,
      "loss": 3.4486,
      "step": 28350
    },
    {
      "epoch": 3.0567215584974705,
      "grad_norm": 0.6193422079086304,
      "learning_rate": 0.00041711669001185216,
      "loss": 3.4353,
      "step": 28400
    },
    {
      "epoch": 3.062103110537079,
      "grad_norm": 0.5869994759559631,
      "learning_rate": 0.0004167934489817907,
      "loss": 3.4371,
      "step": 28450
    },
    {
      "epoch": 3.067484662576687,
      "grad_norm": 0.5577710270881653,
      "learning_rate": 0.0004164702079517293,
      "loss": 3.4412,
      "step": 28500
    },
    {
      "epoch": 3.0728662146162953,
      "grad_norm": 0.56831294298172,
      "learning_rate": 0.0004161469669216679,
      "loss": 3.448,
      "step": 28550
    },
    {
      "epoch": 3.0782477666559034,
      "grad_norm": 0.573702335357666,
      "learning_rate": 0.00041582372589160643,
      "loss": 3.4273,
      "step": 28600
    },
    {
      "epoch": 3.083629318695512,
      "grad_norm": 0.6112154722213745,
      "learning_rate": 0.0004155004848615451,
      "loss": 3.446,
      "step": 28650
    },
    {
      "epoch": 3.08901087073512,
      "grad_norm": 0.5853226780891418,
      "learning_rate": 0.0004151772438314837,
      "loss": 3.4437,
      "step": 28700
    },
    {
      "epoch": 3.094392422774728,
      "grad_norm": 0.577938437461853,
      "learning_rate": 0.0004148540028014222,
      "loss": 3.4488,
      "step": 28750
    },
    {
      "epoch": 3.0997739748143363,
      "grad_norm": 0.5792301297187805,
      "learning_rate": 0.0004145307617713608,
      "loss": 3.4295,
      "step": 28800
    },
    {
      "epoch": 3.105155526853945,
      "grad_norm": 0.6762536764144897,
      "learning_rate": 0.00041420752074129935,
      "loss": 3.4583,
      "step": 28850
    },
    {
      "epoch": 3.110537078893553,
      "grad_norm": 0.6302620768547058,
      "learning_rate": 0.000413884279711238,
      "loss": 3.4365,
      "step": 28900
    },
    {
      "epoch": 3.115918630933161,
      "grad_norm": 0.617859423160553,
      "learning_rate": 0.0004135610386811766,
      "loss": 3.4521,
      "step": 28950
    },
    {
      "epoch": 3.121300182972769,
      "grad_norm": 0.6463020443916321,
      "learning_rate": 0.00041323779765111514,
      "loss": 3.4583,
      "step": 29000
    },
    {
      "epoch": 3.121300182972769,
      "eval_accuracy": 0.3707458184121228,
      "eval_loss": 3.494701385498047,
      "eval_runtime": 190.0105,
      "eval_samples_per_second": 94.789,
      "eval_steps_per_second": 5.926,
      "step": 29000
    },
    {
      "epoch": 3.1266817350123777,
      "grad_norm": 0.6337493062019348,
      "learning_rate": 0.00041291455662105373,
      "loss": 3.449,
      "step": 29050
    },
    {
      "epoch": 3.132063287051986,
      "grad_norm": 0.611283540725708,
      "learning_rate": 0.0004125913155909923,
      "loss": 3.4658,
      "step": 29100
    },
    {
      "epoch": 3.137444839091594,
      "grad_norm": 0.620567798614502,
      "learning_rate": 0.00041226807456093087,
      "loss": 3.4357,
      "step": 29150
    },
    {
      "epoch": 3.1428263911312024,
      "grad_norm": 0.5975887179374695,
      "learning_rate": 0.0004119448335308695,
      "loss": 3.4526,
      "step": 29200
    },
    {
      "epoch": 3.1482079431708105,
      "grad_norm": 0.6219651699066162,
      "learning_rate": 0.0004116215925008081,
      "loss": 3.4545,
      "step": 29250
    },
    {
      "epoch": 3.1535894952104186,
      "grad_norm": 0.6166074872016907,
      "learning_rate": 0.00041129835147074665,
      "loss": 3.4703,
      "step": 29300
    },
    {
      "epoch": 3.1589710472500268,
      "grad_norm": 0.625161349773407,
      "learning_rate": 0.00041097511044068524,
      "loss": 3.4349,
      "step": 29350
    },
    {
      "epoch": 3.1643525992896353,
      "grad_norm": 0.6400313973426819,
      "learning_rate": 0.0004106518694106238,
      "loss": 3.451,
      "step": 29400
    },
    {
      "epoch": 3.1697341513292434,
      "grad_norm": 0.6128105521202087,
      "learning_rate": 0.0004103286283805624,
      "loss": 3.4552,
      "step": 29450
    },
    {
      "epoch": 3.1751157033688515,
      "grad_norm": 0.5921278595924377,
      "learning_rate": 0.00041000538735050103,
      "loss": 3.4474,
      "step": 29500
    },
    {
      "epoch": 3.1804972554084596,
      "grad_norm": 0.6078826189041138,
      "learning_rate": 0.00040968214632043957,
      "loss": 3.4499,
      "step": 29550
    },
    {
      "epoch": 3.185878807448068,
      "grad_norm": 0.5951222777366638,
      "learning_rate": 0.00040935890529037816,
      "loss": 3.4509,
      "step": 29600
    },
    {
      "epoch": 3.1912603594876763,
      "grad_norm": 0.5870286822319031,
      "learning_rate": 0.00040903566426031676,
      "loss": 3.453,
      "step": 29650
    },
    {
      "epoch": 3.1966419115272844,
      "grad_norm": 0.6397120356559753,
      "learning_rate": 0.0004087124232302553,
      "loss": 3.4569,
      "step": 29700
    },
    {
      "epoch": 3.2020234635668925,
      "grad_norm": 0.6286545991897583,
      "learning_rate": 0.0004083891822001939,
      "loss": 3.4714,
      "step": 29750
    },
    {
      "epoch": 3.207405015606501,
      "grad_norm": 0.5900658965110779,
      "learning_rate": 0.00040806594117013254,
      "loss": 3.4222,
      "step": 29800
    },
    {
      "epoch": 3.212786567646109,
      "grad_norm": 0.5930511951446533,
      "learning_rate": 0.0004077427001400711,
      "loss": 3.4558,
      "step": 29850
    },
    {
      "epoch": 3.2181681196857173,
      "grad_norm": 0.6059979796409607,
      "learning_rate": 0.0004074194591100097,
      "loss": 3.4193,
      "step": 29900
    },
    {
      "epoch": 3.2235496717253254,
      "grad_norm": 0.5822314023971558,
      "learning_rate": 0.0004070962180799482,
      "loss": 3.4665,
      "step": 29950
    },
    {
      "epoch": 3.228931223764934,
      "grad_norm": 0.6738995909690857,
      "learning_rate": 0.0004067729770498868,
      "loss": 3.454,
      "step": 30000
    },
    {
      "epoch": 3.228931223764934,
      "eval_accuracy": 0.37218981529117723,
      "eval_loss": 3.4861505031585693,
      "eval_runtime": 190.2872,
      "eval_samples_per_second": 94.652,
      "eval_steps_per_second": 5.917,
      "step": 30000
    },
    {
      "epoch": 3.234312775804542,
      "grad_norm": 0.6074284911155701,
      "learning_rate": 0.00040644973601982546,
      "loss": 3.4501,
      "step": 30050
    },
    {
      "epoch": 3.23969432784415,
      "grad_norm": 0.6389995813369751,
      "learning_rate": 0.0004061329598103652,
      "loss": 3.4374,
      "step": 30100
    },
    {
      "epoch": 3.2450758798837587,
      "grad_norm": 0.6276536583900452,
      "learning_rate": 0.00040580971878030386,
      "loss": 3.4236,
      "step": 30150
    },
    {
      "epoch": 3.250457431923367,
      "grad_norm": 0.6304341554641724,
      "learning_rate": 0.0004054864777502424,
      "loss": 3.4539,
      "step": 30200
    },
    {
      "epoch": 3.255838983962975,
      "grad_norm": 0.6120592355728149,
      "learning_rate": 0.000405163236720181,
      "loss": 3.471,
      "step": 30250
    },
    {
      "epoch": 3.261220536002583,
      "grad_norm": 0.5779737830162048,
      "learning_rate": 0.00040483999569011954,
      "loss": 3.4437,
      "step": 30300
    },
    {
      "epoch": 3.2666020880421915,
      "grad_norm": 0.6014507412910461,
      "learning_rate": 0.00040451675466005813,
      "loss": 3.4328,
      "step": 30350
    },
    {
      "epoch": 3.2719836400817996,
      "grad_norm": 0.5755355954170227,
      "learning_rate": 0.0004041935136299967,
      "loss": 3.438,
      "step": 30400
    },
    {
      "epoch": 3.2773651921214078,
      "grad_norm": 0.6800598502159119,
      "learning_rate": 0.0004038702725999353,
      "loss": 3.4285,
      "step": 30450
    },
    {
      "epoch": 3.282746744161016,
      "grad_norm": 0.5759975910186768,
      "learning_rate": 0.0004035470315698739,
      "loss": 3.4489,
      "step": 30500
    },
    {
      "epoch": 3.2881282962006244,
      "grad_norm": 0.6012275815010071,
      "learning_rate": 0.0004032237905398125,
      "loss": 3.4434,
      "step": 30550
    },
    {
      "epoch": 3.2935098482402325,
      "grad_norm": 0.5994857549667358,
      "learning_rate": 0.00040290054950975105,
      "loss": 3.4626,
      "step": 30600
    },
    {
      "epoch": 3.2988914002798406,
      "grad_norm": 0.5616105198860168,
      "learning_rate": 0.00040257730847968965,
      "loss": 3.4537,
      "step": 30650
    },
    {
      "epoch": 3.304272952319449,
      "grad_norm": 0.5422913432121277,
      "learning_rate": 0.0004022540674496283,
      "loss": 3.4573,
      "step": 30700
    },
    {
      "epoch": 3.3096545043590573,
      "grad_norm": 0.5708365440368652,
      "learning_rate": 0.00040193082641956684,
      "loss": 3.4404,
      "step": 30750
    },
    {
      "epoch": 3.3150360563986654,
      "grad_norm": 0.6706158518791199,
      "learning_rate": 0.00040160758538950543,
      "loss": 3.4579,
      "step": 30800
    },
    {
      "epoch": 3.3204176084382735,
      "grad_norm": 0.6198281049728394,
      "learning_rate": 0.00040128434435944397,
      "loss": 3.452,
      "step": 30850
    },
    {
      "epoch": 3.3257991604778816,
      "grad_norm": 0.5976844429969788,
      "learning_rate": 0.00040096110332938257,
      "loss": 3.4637,
      "step": 30900
    },
    {
      "epoch": 3.33118071251749,
      "grad_norm": 0.5780091881752014,
      "learning_rate": 0.00040063786229932116,
      "loss": 3.4263,
      "step": 30950
    },
    {
      "epoch": 3.3365622645570983,
      "grad_norm": 0.615962028503418,
      "learning_rate": 0.00040031462126925975,
      "loss": 3.4484,
      "step": 31000
    },
    {
      "epoch": 3.3365622645570983,
      "eval_accuracy": 0.37262116725580074,
      "eval_loss": 3.480499267578125,
      "eval_runtime": 189.9073,
      "eval_samples_per_second": 94.841,
      "eval_steps_per_second": 5.929,
      "step": 31000
    },
    {
      "epoch": 3.3419438165967064,
      "grad_norm": 0.5808433890342712,
      "learning_rate": 0.0003999978450597995,
      "loss": 3.4683,
      "step": 31050
    },
    {
      "epoch": 3.347325368636315,
      "grad_norm": 0.5867395401000977,
      "learning_rate": 0.00039967460402973816,
      "loss": 3.4526,
      "step": 31100
    },
    {
      "epoch": 3.352706920675923,
      "grad_norm": 0.5752953290939331,
      "learning_rate": 0.00039935136299967675,
      "loss": 3.4381,
      "step": 31150
    },
    {
      "epoch": 3.358088472715531,
      "grad_norm": 0.5619564652442932,
      "learning_rate": 0.0003990281219696153,
      "loss": 3.4508,
      "step": 31200
    },
    {
      "epoch": 3.3634700247551392,
      "grad_norm": 0.6469112038612366,
      "learning_rate": 0.0003987048809395539,
      "loss": 3.4497,
      "step": 31250
    },
    {
      "epoch": 3.368851576794748,
      "grad_norm": 0.5907962322235107,
      "learning_rate": 0.0003983816399094925,
      "loss": 3.4729,
      "step": 31300
    },
    {
      "epoch": 3.374233128834356,
      "grad_norm": 0.6403781175613403,
      "learning_rate": 0.000398058398879431,
      "loss": 3.4609,
      "step": 31350
    },
    {
      "epoch": 3.379614680873964,
      "grad_norm": 0.5692846179008484,
      "learning_rate": 0.00039773515784936967,
      "loss": 3.4642,
      "step": 31400
    },
    {
      "epoch": 3.384996232913572,
      "grad_norm": 0.5766311883926392,
      "learning_rate": 0.00039741191681930826,
      "loss": 3.4616,
      "step": 31450
    },
    {
      "epoch": 3.3903777849531807,
      "grad_norm": 0.6147029399871826,
      "learning_rate": 0.0003970886757892468,
      "loss": 3.4665,
      "step": 31500
    },
    {
      "epoch": 3.3957593369927888,
      "grad_norm": 0.6090274453163147,
      "learning_rate": 0.0003967654347591854,
      "loss": 3.462,
      "step": 31550
    },
    {
      "epoch": 3.401140889032397,
      "grad_norm": 0.6536292433738708,
      "learning_rate": 0.00039644219372912394,
      "loss": 3.4601,
      "step": 31600
    },
    {
      "epoch": 3.4065224410720054,
      "grad_norm": 0.5580132603645325,
      "learning_rate": 0.0003961189526990626,
      "loss": 3.4423,
      "step": 31650
    },
    {
      "epoch": 3.4119039931116135,
      "grad_norm": 0.5564467906951904,
      "learning_rate": 0.0003957957116690012,
      "loss": 3.4451,
      "step": 31700
    },
    {
      "epoch": 3.4172855451512216,
      "grad_norm": 0.5772013664245605,
      "learning_rate": 0.0003954724706389397,
      "loss": 3.4723,
      "step": 31750
    },
    {
      "epoch": 3.4226670971908297,
      "grad_norm": 0.5875008702278137,
      "learning_rate": 0.0003951492296088783,
      "loss": 3.4596,
      "step": 31800
    },
    {
      "epoch": 3.428048649230438,
      "grad_norm": 0.597587525844574,
      "learning_rate": 0.0003948259885788169,
      "loss": 3.4747,
      "step": 31850
    },
    {
      "epoch": 3.4334302012700464,
      "grad_norm": 0.6114709973335266,
      "learning_rate": 0.00039450274754875545,
      "loss": 3.4682,
      "step": 31900
    },
    {
      "epoch": 3.4388117533096545,
      "grad_norm": 0.5742499828338623,
      "learning_rate": 0.0003941795065186941,
      "loss": 3.4624,
      "step": 31950
    },
    {
      "epoch": 3.4441933053492626,
      "grad_norm": 0.5680760741233826,
      "learning_rate": 0.0003938562654886327,
      "loss": 3.4554,
      "step": 32000
    },
    {
      "epoch": 3.4441933053492626,
      "eval_accuracy": 0.373478438540657,
      "eval_loss": 3.4736874103546143,
      "eval_runtime": 190.1682,
      "eval_samples_per_second": 94.711,
      "eval_steps_per_second": 5.921,
      "step": 32000
    },
    {
      "epoch": 3.449574857388871,
      "grad_norm": 0.5946290493011475,
      "learning_rate": 0.00039353302445857124,
      "loss": 3.4468,
      "step": 32050
    },
    {
      "epoch": 3.4549564094284793,
      "grad_norm": 0.6221204996109009,
      "learning_rate": 0.00039320978342850983,
      "loss": 3.4514,
      "step": 32100
    },
    {
      "epoch": 3.4603379614680874,
      "grad_norm": 0.6781211495399475,
      "learning_rate": 0.00039288654239844837,
      "loss": 3.4443,
      "step": 32150
    },
    {
      "epoch": 3.4657195135076955,
      "grad_norm": 0.61839759349823,
      "learning_rate": 0.00039256330136838697,
      "loss": 3.4651,
      "step": 32200
    },
    {
      "epoch": 3.471101065547304,
      "grad_norm": 0.5934427380561829,
      "learning_rate": 0.0003922400603383256,
      "loss": 3.46,
      "step": 32250
    },
    {
      "epoch": 3.476482617586912,
      "grad_norm": 0.6592109799385071,
      "learning_rate": 0.00039191681930826416,
      "loss": 3.4629,
      "step": 32300
    },
    {
      "epoch": 3.4818641696265202,
      "grad_norm": 0.5891439318656921,
      "learning_rate": 0.00039159357827820275,
      "loss": 3.4637,
      "step": 32350
    },
    {
      "epoch": 3.4872457216661283,
      "grad_norm": 0.5820631980895996,
      "learning_rate": 0.00039127033724814135,
      "loss": 3.4514,
      "step": 32400
    },
    {
      "epoch": 3.492627273705737,
      "grad_norm": 0.6294196248054504,
      "learning_rate": 0.0003909470962180799,
      "loss": 3.4595,
      "step": 32450
    },
    {
      "epoch": 3.498008825745345,
      "grad_norm": 0.7152283191680908,
      "learning_rate": 0.0003906303200086197,
      "loss": 3.4668,
      "step": 32500
    },
    {
      "epoch": 3.503390377784953,
      "grad_norm": 0.5932292938232422,
      "learning_rate": 0.0003903070789785583,
      "loss": 3.4707,
      "step": 32550
    },
    {
      "epoch": 3.5087719298245617,
      "grad_norm": 0.6057958006858826,
      "learning_rate": 0.00038998383794849694,
      "loss": 3.4596,
      "step": 32600
    },
    {
      "epoch": 3.5141534818641698,
      "grad_norm": 0.6847020387649536,
      "learning_rate": 0.0003896605969184355,
      "loss": 3.4423,
      "step": 32650
    },
    {
      "epoch": 3.519535033903778,
      "grad_norm": 0.6164110898971558,
      "learning_rate": 0.00038933735588837407,
      "loss": 3.4495,
      "step": 32700
    },
    {
      "epoch": 3.524916585943386,
      "grad_norm": 0.6239364743232727,
      "learning_rate": 0.00038901411485831267,
      "loss": 3.4558,
      "step": 32750
    },
    {
      "epoch": 3.530298137982994,
      "grad_norm": 0.6085065603256226,
      "learning_rate": 0.0003886908738282512,
      "loss": 3.454,
      "step": 32800
    },
    {
      "epoch": 3.5356796900226026,
      "grad_norm": 0.6039212942123413,
      "learning_rate": 0.0003883676327981898,
      "loss": 3.4413,
      "step": 32850
    },
    {
      "epoch": 3.5410612420622107,
      "grad_norm": 0.6689512133598328,
      "learning_rate": 0.00038804439176812845,
      "loss": 3.4563,
      "step": 32900
    },
    {
      "epoch": 3.546442794101819,
      "grad_norm": 0.5615278482437134,
      "learning_rate": 0.000387721150738067,
      "loss": 3.4423,
      "step": 32950
    },
    {
      "epoch": 3.5518243461414274,
      "grad_norm": 0.6573343873023987,
      "learning_rate": 0.0003873979097080056,
      "loss": 3.454,
      "step": 33000
    },
    {
      "epoch": 3.5518243461414274,
      "eval_accuracy": 0.3741382875283543,
      "eval_loss": 3.4689416885375977,
      "eval_runtime": 190.3109,
      "eval_samples_per_second": 94.64,
      "eval_steps_per_second": 5.917,
      "step": 33000
    },
    {
      "epoch": 3.5572058981810355,
      "grad_norm": 0.6435312628746033,
      "learning_rate": 0.0003870746686779441,
      "loss": 3.4429,
      "step": 33050
    },
    {
      "epoch": 3.5625874502206436,
      "grad_norm": 0.6067267060279846,
      "learning_rate": 0.0003867514276478827,
      "loss": 3.4578,
      "step": 33100
    },
    {
      "epoch": 3.5679690022602517,
      "grad_norm": 0.6325441002845764,
      "learning_rate": 0.0003864281866178213,
      "loss": 3.4638,
      "step": 33150
    },
    {
      "epoch": 3.57335055429986,
      "grad_norm": 0.6111650466918945,
      "learning_rate": 0.0003861049455877599,
      "loss": 3.4758,
      "step": 33200
    },
    {
      "epoch": 3.5787321063394684,
      "grad_norm": 0.6026036739349365,
      "learning_rate": 0.0003857817045576985,
      "loss": 3.4493,
      "step": 33250
    },
    {
      "epoch": 3.5841136583790765,
      "grad_norm": 0.5603199601173401,
      "learning_rate": 0.0003854584635276371,
      "loss": 3.4328,
      "step": 33300
    },
    {
      "epoch": 3.5894952104186846,
      "grad_norm": 0.6166011691093445,
      "learning_rate": 0.00038513522249757564,
      "loss": 3.4578,
      "step": 33350
    },
    {
      "epoch": 3.594876762458293,
      "grad_norm": 0.6033289432525635,
      "learning_rate": 0.00038481198146751423,
      "loss": 3.4468,
      "step": 33400
    },
    {
      "epoch": 3.6002583144979012,
      "grad_norm": 0.627385139465332,
      "learning_rate": 0.0003844887404374529,
      "loss": 3.4607,
      "step": 33450
    },
    {
      "epoch": 3.6056398665375093,
      "grad_norm": 0.5884710550308228,
      "learning_rate": 0.00038417196422799264,
      "loss": 3.4524,
      "step": 33500
    },
    {
      "epoch": 3.611021418577118,
      "grad_norm": 0.6180612444877625,
      "learning_rate": 0.00038384872319793123,
      "loss": 3.4307,
      "step": 33550
    },
    {
      "epoch": 3.616402970616726,
      "grad_norm": 0.595292866230011,
      "learning_rate": 0.0003835254821678698,
      "loss": 3.458,
      "step": 33600
    },
    {
      "epoch": 3.621784522656334,
      "grad_norm": 0.6020985245704651,
      "learning_rate": 0.0003832022411378084,
      "loss": 3.4353,
      "step": 33650
    },
    {
      "epoch": 3.627166074695942,
      "grad_norm": 0.6147534847259521,
      "learning_rate": 0.00038287900010774696,
      "loss": 3.4458,
      "step": 33700
    },
    {
      "epoch": 3.6325476267355503,
      "grad_norm": 0.6300920844078064,
      "learning_rate": 0.00038255575907768555,
      "loss": 3.4676,
      "step": 33750
    },
    {
      "epoch": 3.637929178775159,
      "grad_norm": 0.5686904191970825,
      "learning_rate": 0.0003822325180476241,
      "loss": 3.4637,
      "step": 33800
    },
    {
      "epoch": 3.643310730814767,
      "grad_norm": 0.6037928462028503,
      "learning_rate": 0.00038190927701756274,
      "loss": 3.4534,
      "step": 33850
    },
    {
      "epoch": 3.648692282854375,
      "grad_norm": 0.5712344646453857,
      "learning_rate": 0.00038158603598750134,
      "loss": 3.4352,
      "step": 33900
    },
    {
      "epoch": 3.6540738348939836,
      "grad_norm": 0.6415552496910095,
      "learning_rate": 0.0003812627949574399,
      "loss": 3.4374,
      "step": 33950
    },
    {
      "epoch": 3.6594553869335917,
      "grad_norm": 0.6251377463340759,
      "learning_rate": 0.0003809395539273785,
      "loss": 3.456,
      "step": 34000
    },
    {
      "epoch": 3.6594553869335917,
      "eval_accuracy": 0.3745376955439604,
      "eval_loss": 3.4615447521209717,
      "eval_runtime": 190.4008,
      "eval_samples_per_second": 94.595,
      "eval_steps_per_second": 5.914,
      "step": 34000
    },
    {
      "epoch": 3.6648369389732,
      "grad_norm": 0.5713412761688232,
      "learning_rate": 0.00038061631289731707,
      "loss": 3.435,
      "step": 34050
    },
    {
      "epoch": 3.670218491012808,
      "grad_norm": 0.6113209128379822,
      "learning_rate": 0.00038029307186725566,
      "loss": 3.4515,
      "step": 34100
    },
    {
      "epoch": 3.675600043052416,
      "grad_norm": 0.6462851762771606,
      "learning_rate": 0.00037996983083719426,
      "loss": 3.4469,
      "step": 34150
    },
    {
      "epoch": 3.6809815950920246,
      "grad_norm": 0.5907116532325745,
      "learning_rate": 0.00037964658980713285,
      "loss": 3.4537,
      "step": 34200
    },
    {
      "epoch": 3.6863631471316327,
      "grad_norm": 0.6197351813316345,
      "learning_rate": 0.0003793233487770714,
      "loss": 3.4473,
      "step": 34250
    },
    {
      "epoch": 3.691744699171241,
      "grad_norm": 0.6316325068473816,
      "learning_rate": 0.00037900010774701,
      "loss": 3.4383,
      "step": 34300
    },
    {
      "epoch": 3.6971262512108494,
      "grad_norm": 0.6373358368873596,
      "learning_rate": 0.00037867686671694853,
      "loss": 3.457,
      "step": 34350
    },
    {
      "epoch": 3.7025078032504575,
      "grad_norm": 0.6204352974891663,
      "learning_rate": 0.0003783536256868872,
      "loss": 3.4432,
      "step": 34400
    },
    {
      "epoch": 3.7078893552900656,
      "grad_norm": 0.6196008920669556,
      "learning_rate": 0.00037803038465682577,
      "loss": 3.4646,
      "step": 34450
    },
    {
      "epoch": 3.713270907329674,
      "grad_norm": 0.6414692997932434,
      "learning_rate": 0.0003777071436267643,
      "loss": 3.4624,
      "step": 34500
    },
    {
      "epoch": 3.7186524593692822,
      "grad_norm": 0.6107802987098694,
      "learning_rate": 0.0003773839025967029,
      "loss": 3.4737,
      "step": 34550
    },
    {
      "epoch": 3.7240340114088903,
      "grad_norm": 0.58467036485672,
      "learning_rate": 0.0003770606615666415,
      "loss": 3.4526,
      "step": 34600
    },
    {
      "epoch": 3.7294155634484984,
      "grad_norm": 0.5918004512786865,
      "learning_rate": 0.00037673742053658004,
      "loss": 3.4523,
      "step": 34650
    },
    {
      "epoch": 3.7347971154881066,
      "grad_norm": 0.6468043923377991,
      "learning_rate": 0.0003764141795065187,
      "loss": 3.451,
      "step": 34700
    },
    {
      "epoch": 3.740178667527715,
      "grad_norm": 0.6421909332275391,
      "learning_rate": 0.0003760909384764573,
      "loss": 3.4576,
      "step": 34750
    },
    {
      "epoch": 3.745560219567323,
      "grad_norm": 0.5914976596832275,
      "learning_rate": 0.0003757676974463958,
      "loss": 3.4498,
      "step": 34800
    },
    {
      "epoch": 3.7509417716069313,
      "grad_norm": 0.5865017175674438,
      "learning_rate": 0.0003754444564163344,
      "loss": 3.4522,
      "step": 34850
    },
    {
      "epoch": 3.75632332364654,
      "grad_norm": 0.6309718489646912,
      "learning_rate": 0.00037512121538627296,
      "loss": 3.442,
      "step": 34900
    },
    {
      "epoch": 3.761704875686148,
      "grad_norm": 0.6371271014213562,
      "learning_rate": 0.00037479797435621155,
      "loss": 3.4382,
      "step": 34950
    },
    {
      "epoch": 3.767086427725756,
      "grad_norm": 0.6336926817893982,
      "learning_rate": 0.0003744747333261502,
      "loss": 3.4535,
      "step": 35000
    },
    {
      "epoch": 3.767086427725756,
      "eval_accuracy": 0.3756370454628673,
      "eval_loss": 3.45515513420105,
      "eval_runtime": 190.2068,
      "eval_samples_per_second": 94.692,
      "eval_steps_per_second": 5.92,
      "step": 35000
    },
    {
      "epoch": 3.772467979765364,
      "grad_norm": 0.6118438243865967,
      "learning_rate": 0.00037415149229608874,
      "loss": 3.4541,
      "step": 35050
    },
    {
      "epoch": 3.7778495318049723,
      "grad_norm": 0.6314883828163147,
      "learning_rate": 0.00037382825126602734,
      "loss": 3.4616,
      "step": 35100
    },
    {
      "epoch": 3.783231083844581,
      "grad_norm": 0.6189128756523132,
      "learning_rate": 0.00037350501023596593,
      "loss": 3.4662,
      "step": 35150
    },
    {
      "epoch": 3.788612635884189,
      "grad_norm": 0.5882723331451416,
      "learning_rate": 0.0003731817692059045,
      "loss": 3.4543,
      "step": 35200
    },
    {
      "epoch": 3.793994187923797,
      "grad_norm": 0.5908012986183167,
      "learning_rate": 0.0003728585281758431,
      "loss": 3.438,
      "step": 35250
    },
    {
      "epoch": 3.7993757399634056,
      "grad_norm": 0.6089473962783813,
      "learning_rate": 0.0003725352871457817,
      "loss": 3.437,
      "step": 35300
    },
    {
      "epoch": 3.8047572920030137,
      "grad_norm": 0.592900812625885,
      "learning_rate": 0.00037221204611572026,
      "loss": 3.4673,
      "step": 35350
    },
    {
      "epoch": 3.810138844042622,
      "grad_norm": 0.6114164590835571,
      "learning_rate": 0.00037188880508565885,
      "loss": 3.4505,
      "step": 35400
    },
    {
      "epoch": 3.8155203960822304,
      "grad_norm": 0.5772916078567505,
      "learning_rate": 0.0003715655640555974,
      "loss": 3.4779,
      "step": 35450
    },
    {
      "epoch": 3.8209019481218385,
      "grad_norm": 0.6408140659332275,
      "learning_rate": 0.000371242323025536,
      "loss": 3.4665,
      "step": 35500
    },
    {
      "epoch": 3.8262835001614466,
      "grad_norm": 0.6018230319023132,
      "learning_rate": 0.00037091908199547464,
      "loss": 3.4505,
      "step": 35550
    },
    {
      "epoch": 3.8316650522010547,
      "grad_norm": 0.6414809226989746,
      "learning_rate": 0.0003705958409654132,
      "loss": 3.4543,
      "step": 35600
    },
    {
      "epoch": 3.837046604240663,
      "grad_norm": 0.5931991338729858,
      "learning_rate": 0.00037027259993535177,
      "loss": 3.4557,
      "step": 35650
    },
    {
      "epoch": 3.8424281562802713,
      "grad_norm": 0.7092098593711853,
      "learning_rate": 0.0003699493589052903,
      "loss": 3.4472,
      "step": 35700
    },
    {
      "epoch": 3.8478097083198795,
      "grad_norm": 0.6029160618782043,
      "learning_rate": 0.0003696261178752289,
      "loss": 3.443,
      "step": 35750
    },
    {
      "epoch": 3.8531912603594876,
      "grad_norm": 0.6160264611244202,
      "learning_rate": 0.0003693028768451675,
      "loss": 3.4358,
      "step": 35800
    },
    {
      "epoch": 3.858572812399096,
      "grad_norm": 0.6086969375610352,
      "learning_rate": 0.0003689796358151061,
      "loss": 3.4517,
      "step": 35850
    },
    {
      "epoch": 3.863954364438704,
      "grad_norm": 0.5965676307678223,
      "learning_rate": 0.0003686563947850447,
      "loss": 3.4429,
      "step": 35900
    },
    {
      "epoch": 3.8693359164783123,
      "grad_norm": 0.5858458876609802,
      "learning_rate": 0.0003683331537549833,
      "loss": 3.4408,
      "step": 35950
    },
    {
      "epoch": 3.8747174685179204,
      "grad_norm": 0.6486075520515442,
      "learning_rate": 0.0003680099127249218,
      "loss": 3.433,
      "step": 36000
    },
    {
      "epoch": 3.8747174685179204,
      "eval_accuracy": 0.37605351198186016,
      "eval_loss": 3.4477474689483643,
      "eval_runtime": 189.9414,
      "eval_samples_per_second": 94.824,
      "eval_steps_per_second": 5.928,
      "step": 36000
    },
    {
      "epoch": 3.8800990205575285,
      "grad_norm": 0.6607269048690796,
      "learning_rate": 0.0003676866716948604,
      "loss": 3.4432,
      "step": 36050
    },
    {
      "epoch": 3.885480572597137,
      "grad_norm": 0.6403939127922058,
      "learning_rate": 0.00036736343066479907,
      "loss": 3.4527,
      "step": 36100
    },
    {
      "epoch": 3.890862124636745,
      "grad_norm": 0.615782618522644,
      "learning_rate": 0.0003670401896347376,
      "loss": 3.4667,
      "step": 36150
    },
    {
      "epoch": 3.8962436766763533,
      "grad_norm": 0.6064379811286926,
      "learning_rate": 0.0003667169486046762,
      "loss": 3.4332,
      "step": 36200
    },
    {
      "epoch": 3.901625228715962,
      "grad_norm": 0.6022096276283264,
      "learning_rate": 0.00036639370757461475,
      "loss": 3.4334,
      "step": 36250
    },
    {
      "epoch": 3.90700678075557,
      "grad_norm": 0.5906829237937927,
      "learning_rate": 0.00036607046654455334,
      "loss": 3.4363,
      "step": 36300
    },
    {
      "epoch": 3.912388332795178,
      "grad_norm": 0.6686057448387146,
      "learning_rate": 0.00036574722551449193,
      "loss": 3.4406,
      "step": 36350
    },
    {
      "epoch": 3.9177698848347866,
      "grad_norm": 0.629077672958374,
      "learning_rate": 0.00036542398448443053,
      "loss": 3.4385,
      "step": 36400
    },
    {
      "epoch": 3.9231514368743947,
      "grad_norm": 0.6432328820228577,
      "learning_rate": 0.0003651007434543691,
      "loss": 3.4439,
      "step": 36450
    },
    {
      "epoch": 3.928532988914003,
      "grad_norm": 0.6564190983772278,
      "learning_rate": 0.0003647775024243077,
      "loss": 3.461,
      "step": 36500
    },
    {
      "epoch": 3.933914540953611,
      "grad_norm": 0.6439425945281982,
      "learning_rate": 0.00036445426139424626,
      "loss": 3.4437,
      "step": 36550
    },
    {
      "epoch": 3.939296092993219,
      "grad_norm": 0.6073256731033325,
      "learning_rate": 0.00036413102036418485,
      "loss": 3.4506,
      "step": 36600
    },
    {
      "epoch": 3.9446776450328276,
      "grad_norm": 0.607316255569458,
      "learning_rate": 0.0003638077793341234,
      "loss": 3.4411,
      "step": 36650
    },
    {
      "epoch": 3.9500591970724357,
      "grad_norm": 0.6186468601226807,
      "learning_rate": 0.00036348453830406204,
      "loss": 3.4472,
      "step": 36700
    },
    {
      "epoch": 3.955440749112044,
      "grad_norm": 0.6006415486335754,
      "learning_rate": 0.00036316129727400064,
      "loss": 3.4528,
      "step": 36750
    },
    {
      "epoch": 3.9608223011516523,
      "grad_norm": 0.6272019147872925,
      "learning_rate": 0.0003628380562439392,
      "loss": 3.4366,
      "step": 36800
    },
    {
      "epoch": 3.9662038531912605,
      "grad_norm": 0.6135907173156738,
      "learning_rate": 0.00036251481521387777,
      "loss": 3.4435,
      "step": 36850
    },
    {
      "epoch": 3.9715854052308686,
      "grad_norm": 0.6131258010864258,
      "learning_rate": 0.00036219157418381637,
      "loss": 3.4515,
      "step": 36900
    },
    {
      "epoch": 3.9769669572704767,
      "grad_norm": 0.6010150909423828,
      "learning_rate": 0.0003618683331537549,
      "loss": 3.4516,
      "step": 36950
    },
    {
      "epoch": 3.9823485093100848,
      "grad_norm": 0.6117335557937622,
      "learning_rate": 0.00036154509212369356,
      "loss": 3.4352,
      "step": 37000
    },
    {
      "epoch": 3.9823485093100848,
      "eval_accuracy": 0.37687851335903555,
      "eval_loss": 3.4433538913726807,
      "eval_runtime": 190.3456,
      "eval_samples_per_second": 94.623,
      "eval_steps_per_second": 5.916,
      "step": 37000
    },
    {
      "epoch": 3.9877300613496933,
      "grad_norm": 0.6108750104904175,
      "learning_rate": 0.00036122185109363215,
      "loss": 3.4321,
      "step": 37050
    },
    {
      "epoch": 3.9931116133893014,
      "grad_norm": 0.6648726463317871,
      "learning_rate": 0.0003608986100635707,
      "loss": 3.4438,
      "step": 37100
    },
    {
      "epoch": 3.9984931654289095,
      "grad_norm": 0.6244668960571289,
      "learning_rate": 0.0003605753690335093,
      "loss": 3.4437,
      "step": 37150
    },
    {
      "epoch": 4.003874717468518,
      "grad_norm": 0.5718086957931519,
      "learning_rate": 0.0003602521280034478,
      "loss": 3.402,
      "step": 37200
    },
    {
      "epoch": 4.009256269508126,
      "grad_norm": 0.5737576484680176,
      "learning_rate": 0.0003599288869733865,
      "loss": 3.3428,
      "step": 37250
    },
    {
      "epoch": 4.014637821547734,
      "grad_norm": 0.588140606880188,
      "learning_rate": 0.00035960564594332507,
      "loss": 3.3544,
      "step": 37300
    },
    {
      "epoch": 4.020019373587343,
      "grad_norm": 0.6539402604103088,
      "learning_rate": 0.0003592824049132636,
      "loss": 3.3579,
      "step": 37350
    },
    {
      "epoch": 4.0254009256269505,
      "grad_norm": 0.5864378213882446,
      "learning_rate": 0.0003589591638832022,
      "loss": 3.3601,
      "step": 37400
    },
    {
      "epoch": 4.030782477666559,
      "grad_norm": 0.6314789652824402,
      "learning_rate": 0.0003586359228531408,
      "loss": 3.345,
      "step": 37450
    },
    {
      "epoch": 4.036164029706168,
      "grad_norm": 0.6276288032531738,
      "learning_rate": 0.00035831268182307934,
      "loss": 3.3525,
      "step": 37500
    },
    {
      "epoch": 4.041545581745775,
      "grad_norm": 0.5828260779380798,
      "learning_rate": 0.000357989440793018,
      "loss": 3.3732,
      "step": 37550
    },
    {
      "epoch": 4.046927133785384,
      "grad_norm": 0.6124811768531799,
      "learning_rate": 0.0003576661997629566,
      "loss": 3.3577,
      "step": 37600
    },
    {
      "epoch": 4.0523086858249915,
      "grad_norm": 0.6164279580116272,
      "learning_rate": 0.0003573429587328951,
      "loss": 3.3643,
      "step": 37650
    },
    {
      "epoch": 4.0576902378646,
      "grad_norm": 0.5956741571426392,
      "learning_rate": 0.0003570197177028337,
      "loss": 3.3762,
      "step": 37700
    },
    {
      "epoch": 4.063071789904209,
      "grad_norm": 0.6794794797897339,
      "learning_rate": 0.00035669647667277226,
      "loss": 3.3627,
      "step": 37750
    },
    {
      "epoch": 4.068453341943816,
      "grad_norm": 0.63214510679245,
      "learning_rate": 0.00035637323564271085,
      "loss": 3.3497,
      "step": 37800
    },
    {
      "epoch": 4.073834893983425,
      "grad_norm": 0.6622040867805481,
      "learning_rate": 0.0003560499946126495,
      "loss": 3.3709,
      "step": 37850
    },
    {
      "epoch": 4.079216446023033,
      "grad_norm": 0.5973483324050903,
      "learning_rate": 0.00035572675358258804,
      "loss": 3.3733,
      "step": 37900
    },
    {
      "epoch": 4.084597998062641,
      "grad_norm": 0.6202905774116516,
      "learning_rate": 0.00035540351255252664,
      "loss": 3.3411,
      "step": 37950
    },
    {
      "epoch": 4.08997955010225,
      "grad_norm": 0.6901636123657227,
      "learning_rate": 0.00035508027152246523,
      "loss": 3.3738,
      "step": 38000
    },
    {
      "epoch": 4.08997955010225,
      "eval_accuracy": 0.3766697911615439,
      "eval_loss": 3.444305658340454,
      "eval_runtime": 190.0812,
      "eval_samples_per_second": 94.754,
      "eval_steps_per_second": 5.924,
      "step": 38000
    },
    {
      "epoch": 4.095361102141858,
      "grad_norm": 0.6387552618980408,
      "learning_rate": 0.0003547570304924038,
      "loss": 3.353,
      "step": 38050
    },
    {
      "epoch": 4.100742654181466,
      "grad_norm": 0.677064061164856,
      "learning_rate": 0.0003544337894623424,
      "loss": 3.3739,
      "step": 38100
    },
    {
      "epoch": 4.106124206221074,
      "grad_norm": 0.619288980960846,
      "learning_rate": 0.000354110548432281,
      "loss": 3.3705,
      "step": 38150
    },
    {
      "epoch": 4.111505758260682,
      "grad_norm": 0.6704347133636475,
      "learning_rate": 0.00035378730740221956,
      "loss": 3.3621,
      "step": 38200
    },
    {
      "epoch": 4.1168873103002905,
      "grad_norm": 0.619070827960968,
      "learning_rate": 0.00035346406637215815,
      "loss": 3.3689,
      "step": 38250
    },
    {
      "epoch": 4.122268862339899,
      "grad_norm": 0.6347267627716064,
      "learning_rate": 0.0003531408253420967,
      "loss": 3.3791,
      "step": 38300
    },
    {
      "epoch": 4.127650414379507,
      "grad_norm": 0.6314436197280884,
      "learning_rate": 0.0003528175843120353,
      "loss": 3.3598,
      "step": 38350
    },
    {
      "epoch": 4.133031966419115,
      "grad_norm": 0.6617787480354309,
      "learning_rate": 0.00035249434328197394,
      "loss": 3.3765,
      "step": 38400
    },
    {
      "epoch": 4.138413518458724,
      "grad_norm": 0.6698671579360962,
      "learning_rate": 0.0003521711022519125,
      "loss": 3.3698,
      "step": 38450
    },
    {
      "epoch": 4.1437950704983315,
      "grad_norm": 0.6467815041542053,
      "learning_rate": 0.00035184786122185107,
      "loss": 3.3583,
      "step": 38500
    },
    {
      "epoch": 4.14917662253794,
      "grad_norm": 0.6683939695358276,
      "learning_rate": 0.00035152462019178967,
      "loss": 3.3734,
      "step": 38550
    },
    {
      "epoch": 4.154558174577549,
      "grad_norm": 0.6323290467262268,
      "learning_rate": 0.0003512013791617282,
      "loss": 3.3762,
      "step": 38600
    },
    {
      "epoch": 4.159939726617156,
      "grad_norm": 0.6349487900733948,
      "learning_rate": 0.0003508781381316668,
      "loss": 3.3819,
      "step": 38650
    },
    {
      "epoch": 4.165321278656765,
      "grad_norm": 0.6143118143081665,
      "learning_rate": 0.00035055489710160545,
      "loss": 3.3841,
      "step": 38700
    },
    {
      "epoch": 4.1707028306963725,
      "grad_norm": 0.6510623693466187,
      "learning_rate": 0.000350231656071544,
      "loss": 3.3649,
      "step": 38750
    },
    {
      "epoch": 4.176084382735981,
      "grad_norm": 0.627197265625,
      "learning_rate": 0.0003499084150414826,
      "loss": 3.3837,
      "step": 38800
    },
    {
      "epoch": 4.18146593477559,
      "grad_norm": 0.6500346660614014,
      "learning_rate": 0.0003495851740114211,
      "loss": 3.3611,
      "step": 38850
    },
    {
      "epoch": 4.186847486815197,
      "grad_norm": 0.6525781750679016,
      "learning_rate": 0.0003492619329813597,
      "loss": 3.3897,
      "step": 38900
    },
    {
      "epoch": 4.192229038854806,
      "grad_norm": 0.6362114548683167,
      "learning_rate": 0.0003489386919512983,
      "loss": 3.3904,
      "step": 38950
    },
    {
      "epoch": 4.197610590894414,
      "grad_norm": 0.6572438478469849,
      "learning_rate": 0.0003486154509212369,
      "loss": 3.3747,
      "step": 39000
    },
    {
      "epoch": 4.197610590894414,
      "eval_accuracy": 0.3771017950434943,
      "eval_loss": 3.4433939456939697,
      "eval_runtime": 190.2352,
      "eval_samples_per_second": 94.678,
      "eval_steps_per_second": 5.919,
      "step": 39000
    },
    {
      "epoch": 4.202992142934022,
      "grad_norm": 0.5828280448913574,
      "learning_rate": 0.0003482922098911755,
      "loss": 3.3864,
      "step": 39050
    },
    {
      "epoch": 4.208373694973631,
      "grad_norm": 0.6487306356430054,
      "learning_rate": 0.0003479689688611141,
      "loss": 3.4071,
      "step": 39100
    },
    {
      "epoch": 4.213755247013238,
      "grad_norm": 0.5984469652175903,
      "learning_rate": 0.00034764572783105264,
      "loss": 3.3732,
      "step": 39150
    },
    {
      "epoch": 4.219136799052847,
      "grad_norm": 0.648029625415802,
      "learning_rate": 0.00034732248680099123,
      "loss": 3.3789,
      "step": 39200
    },
    {
      "epoch": 4.224518351092455,
      "grad_norm": 0.6460610032081604,
      "learning_rate": 0.0003469992457709299,
      "loss": 3.3773,
      "step": 39250
    },
    {
      "epoch": 4.229899903132063,
      "grad_norm": 0.6132860779762268,
      "learning_rate": 0.0003466760047408684,
      "loss": 3.3572,
      "step": 39300
    },
    {
      "epoch": 4.2352814551716715,
      "grad_norm": 0.6209291815757751,
      "learning_rate": 0.000346352763710807,
      "loss": 3.3821,
      "step": 39350
    },
    {
      "epoch": 4.24066300721128,
      "grad_norm": 0.6748947501182556,
      "learning_rate": 0.00034602952268074556,
      "loss": 3.3818,
      "step": 39400
    },
    {
      "epoch": 4.246044559250888,
      "grad_norm": 0.6416504979133606,
      "learning_rate": 0.00034570628165068415,
      "loss": 3.3861,
      "step": 39450
    },
    {
      "epoch": 4.251426111290496,
      "grad_norm": 0.5997656583786011,
      "learning_rate": 0.0003453830406206227,
      "loss": 3.4062,
      "step": 39500
    },
    {
      "epoch": 4.256807663330104,
      "grad_norm": 0.651505172252655,
      "learning_rate": 0.00034505979959056134,
      "loss": 3.3819,
      "step": 39550
    },
    {
      "epoch": 4.2621892153697125,
      "grad_norm": 0.6245538592338562,
      "learning_rate": 0.00034473655856049994,
      "loss": 3.3728,
      "step": 39600
    },
    {
      "epoch": 4.267570767409321,
      "grad_norm": 0.6150814890861511,
      "learning_rate": 0.0003444133175304385,
      "loss": 3.3773,
      "step": 39650
    },
    {
      "epoch": 4.272952319448929,
      "grad_norm": 0.6090626120567322,
      "learning_rate": 0.00034409007650037707,
      "loss": 3.3848,
      "step": 39700
    },
    {
      "epoch": 4.278333871488537,
      "grad_norm": 0.6361591219902039,
      "learning_rate": 0.00034376683547031567,
      "loss": 3.3754,
      "step": 39750
    },
    {
      "epoch": 4.283715423528146,
      "grad_norm": 0.6771186590194702,
      "learning_rate": 0.0003434435944402542,
      "loss": 3.3524,
      "step": 39800
    },
    {
      "epoch": 4.2890969755677535,
      "grad_norm": 0.5855597257614136,
      "learning_rate": 0.00034312035341019286,
      "loss": 3.3914,
      "step": 39850
    },
    {
      "epoch": 4.294478527607362,
      "grad_norm": 0.6105331778526306,
      "learning_rate": 0.00034279711238013145,
      "loss": 3.4036,
      "step": 39900
    },
    {
      "epoch": 4.299860079646971,
      "grad_norm": 0.6054653525352478,
      "learning_rate": 0.00034247387135007,
      "loss": 3.378,
      "step": 39950
    },
    {
      "epoch": 4.305241631686578,
      "grad_norm": 0.6510891914367676,
      "learning_rate": 0.0003421506303200086,
      "loss": 3.38,
      "step": 40000
    },
    {
      "epoch": 4.305241631686578,
      "eval_accuracy": 0.3775370585120791,
      "eval_loss": 3.4385550022125244,
      "eval_runtime": 190.1728,
      "eval_samples_per_second": 94.709,
      "eval_steps_per_second": 5.921,
      "step": 40000
    },
    {
      "epoch": 4.310623183726187,
      "grad_norm": 0.6370204091072083,
      "learning_rate": 0.0003418273892899471,
      "loss": 3.3987,
      "step": 40050
    },
    {
      "epoch": 4.3160047357657945,
      "grad_norm": 0.5936571955680847,
      "learning_rate": 0.0003415041482598858,
      "loss": 3.3743,
      "step": 40100
    },
    {
      "epoch": 4.321386287805403,
      "grad_norm": 0.6611200571060181,
      "learning_rate": 0.00034118090722982437,
      "loss": 3.3794,
      "step": 40150
    },
    {
      "epoch": 4.326767839845012,
      "grad_norm": 0.6471608281135559,
      "learning_rate": 0.0003408576661997629,
      "loss": 3.3834,
      "step": 40200
    },
    {
      "epoch": 4.332149391884619,
      "grad_norm": 0.6454547047615051,
      "learning_rate": 0.0003405344251697015,
      "loss": 3.371,
      "step": 40250
    },
    {
      "epoch": 4.337530943924228,
      "grad_norm": 0.6713175773620605,
      "learning_rate": 0.0003402111841396401,
      "loss": 3.3847,
      "step": 40300
    },
    {
      "epoch": 4.342912495963836,
      "grad_norm": 0.6623990535736084,
      "learning_rate": 0.00033988794310957864,
      "loss": 3.3824,
      "step": 40350
    },
    {
      "epoch": 4.348294048003444,
      "grad_norm": 0.5979756116867065,
      "learning_rate": 0.0003395647020795173,
      "loss": 3.3961,
      "step": 40400
    },
    {
      "epoch": 4.3536756000430525,
      "grad_norm": 0.6444520354270935,
      "learning_rate": 0.0003392414610494559,
      "loss": 3.3971,
      "step": 40450
    },
    {
      "epoch": 4.359057152082661,
      "grad_norm": 0.6258366703987122,
      "learning_rate": 0.0003389182200193944,
      "loss": 3.3898,
      "step": 40500
    },
    {
      "epoch": 4.364438704122269,
      "grad_norm": 0.5918956398963928,
      "learning_rate": 0.000338594978989333,
      "loss": 3.3932,
      "step": 40550
    },
    {
      "epoch": 4.369820256161877,
      "grad_norm": 0.6441937685012817,
      "learning_rate": 0.00033827173795927156,
      "loss": 3.392,
      "step": 40600
    },
    {
      "epoch": 4.375201808201485,
      "grad_norm": 0.6300854086875916,
      "learning_rate": 0.00033794849692921015,
      "loss": 3.368,
      "step": 40650
    },
    {
      "epoch": 4.3805833602410935,
      "grad_norm": 0.6491185426712036,
      "learning_rate": 0.0003376252558991488,
      "loss": 3.3907,
      "step": 40700
    },
    {
      "epoch": 4.385964912280702,
      "grad_norm": 0.6179837584495544,
      "learning_rate": 0.00033730201486908734,
      "loss": 3.3602,
      "step": 40750
    },
    {
      "epoch": 4.39134646432031,
      "grad_norm": 0.6025059223175049,
      "learning_rate": 0.00033697877383902594,
      "loss": 3.3806,
      "step": 40800
    },
    {
      "epoch": 4.396728016359918,
      "grad_norm": 0.6713990569114685,
      "learning_rate": 0.00033665553280896453,
      "loss": 3.3648,
      "step": 40850
    },
    {
      "epoch": 4.402109568399527,
      "grad_norm": 0.6606966257095337,
      "learning_rate": 0.00033633875659950434,
      "loss": 3.3747,
      "step": 40900
    },
    {
      "epoch": 4.4074911204391345,
      "grad_norm": 0.6316465735435486,
      "learning_rate": 0.0003360155155694429,
      "loss": 3.3875,
      "step": 40950
    },
    {
      "epoch": 4.412872672478743,
      "grad_norm": 0.6216477751731873,
      "learning_rate": 0.0003356922745393815,
      "loss": 3.3825,
      "step": 41000
    },
    {
      "epoch": 4.412872672478743,
      "eval_accuracy": 0.3780088293509696,
      "eval_loss": 3.4331576824188232,
      "eval_runtime": 190.2549,
      "eval_samples_per_second": 94.668,
      "eval_steps_per_second": 5.918,
      "step": 41000
    },
    {
      "epoch": 4.418254224518351,
      "grad_norm": 0.6269261240959167,
      "learning_rate": 0.0003353690335093201,
      "loss": 3.387,
      "step": 41050
    },
    {
      "epoch": 4.423635776557959,
      "grad_norm": 0.644935131072998,
      "learning_rate": 0.0003350457924792587,
      "loss": 3.3786,
      "step": 41100
    },
    {
      "epoch": 4.429017328597568,
      "grad_norm": 0.671531081199646,
      "learning_rate": 0.00033472255144919726,
      "loss": 3.3846,
      "step": 41150
    },
    {
      "epoch": 4.4343988806371755,
      "grad_norm": 0.6102561354637146,
      "learning_rate": 0.00033439931041913585,
      "loss": 3.3862,
      "step": 41200
    },
    {
      "epoch": 4.439780432676784,
      "grad_norm": 0.6670401692390442,
      "learning_rate": 0.0003340760693890744,
      "loss": 3.3651,
      "step": 41250
    },
    {
      "epoch": 4.445161984716393,
      "grad_norm": 0.612514078617096,
      "learning_rate": 0.000333752828359013,
      "loss": 3.3763,
      "step": 41300
    },
    {
      "epoch": 4.450543536756,
      "grad_norm": 0.6115955114364624,
      "learning_rate": 0.00033342958732895164,
      "loss": 3.388,
      "step": 41350
    },
    {
      "epoch": 4.455925088795609,
      "grad_norm": 0.5957492589950562,
      "learning_rate": 0.0003331063462988902,
      "loss": 3.384,
      "step": 41400
    },
    {
      "epoch": 4.461306640835216,
      "grad_norm": 0.637829065322876,
      "learning_rate": 0.00033278310526882877,
      "loss": 3.3899,
      "step": 41450
    },
    {
      "epoch": 4.466688192874825,
      "grad_norm": 0.6799536347389221,
      "learning_rate": 0.0003324598642387673,
      "loss": 3.3982,
      "step": 41500
    },
    {
      "epoch": 4.4720697449144335,
      "grad_norm": 0.6371310949325562,
      "learning_rate": 0.0003321366232087059,
      "loss": 3.3957,
      "step": 41550
    },
    {
      "epoch": 4.477451296954041,
      "grad_norm": 0.6020154356956482,
      "learning_rate": 0.0003318133821786445,
      "loss": 3.3865,
      "step": 41600
    },
    {
      "epoch": 4.48283284899365,
      "grad_norm": 0.654555082321167,
      "learning_rate": 0.0003314901411485831,
      "loss": 3.3744,
      "step": 41650
    },
    {
      "epoch": 4.488214401033258,
      "grad_norm": 0.6542131900787354,
      "learning_rate": 0.0003311669001185217,
      "loss": 3.3707,
      "step": 41700
    },
    {
      "epoch": 4.493595953072866,
      "grad_norm": 0.5788448452949524,
      "learning_rate": 0.0003308436590884603,
      "loss": 3.3921,
      "step": 41750
    },
    {
      "epoch": 4.4989775051124745,
      "grad_norm": 0.6686775088310242,
      "learning_rate": 0.0003305204180583988,
      "loss": 3.3789,
      "step": 41800
    },
    {
      "epoch": 4.504359057152083,
      "grad_norm": 0.589725911617279,
      "learning_rate": 0.0003301971770283374,
      "loss": 3.3823,
      "step": 41850
    },
    {
      "epoch": 4.509740609191691,
      "grad_norm": 0.61907958984375,
      "learning_rate": 0.00032987393599827607,
      "loss": 3.3854,
      "step": 41900
    },
    {
      "epoch": 4.515122161231299,
      "grad_norm": 0.5937468409538269,
      "learning_rate": 0.0003295506949682146,
      "loss": 3.3946,
      "step": 41950
    },
    {
      "epoch": 4.520503713270907,
      "grad_norm": 0.6533059477806091,
      "learning_rate": 0.0003292274539381532,
      "loss": 3.3831,
      "step": 42000
    },
    {
      "epoch": 4.520503713270907,
      "eval_accuracy": 0.37835434553421965,
      "eval_loss": 3.427962064743042,
      "eval_runtime": 190.0044,
      "eval_samples_per_second": 94.793,
      "eval_steps_per_second": 5.926,
      "step": 42000
    },
    {
      "epoch": 4.5258852653105155,
      "grad_norm": 0.6122633218765259,
      "learning_rate": 0.00032890421290809174,
      "loss": 3.3928,
      "step": 42050
    },
    {
      "epoch": 4.531266817350124,
      "grad_norm": 0.6673691272735596,
      "learning_rate": 0.00032858097187803034,
      "loss": 3.3626,
      "step": 42100
    },
    {
      "epoch": 4.536648369389732,
      "grad_norm": 0.6171354651451111,
      "learning_rate": 0.00032825773084796893,
      "loss": 3.3802,
      "step": 42150
    },
    {
      "epoch": 4.54202992142934,
      "grad_norm": 0.5974791049957275,
      "learning_rate": 0.00032793448981790753,
      "loss": 3.3925,
      "step": 42200
    },
    {
      "epoch": 4.547411473468949,
      "grad_norm": 0.672253429889679,
      "learning_rate": 0.0003276112487878461,
      "loss": 3.3897,
      "step": 42250
    },
    {
      "epoch": 4.5527930255085565,
      "grad_norm": 0.6303079724311829,
      "learning_rate": 0.0003272880077577847,
      "loss": 3.3951,
      "step": 42300
    },
    {
      "epoch": 4.558174577548165,
      "grad_norm": 0.6007497310638428,
      "learning_rate": 0.00032696476672772326,
      "loss": 3.3951,
      "step": 42350
    },
    {
      "epoch": 4.563556129587774,
      "grad_norm": 0.6923205852508545,
      "learning_rate": 0.00032664152569766185,
      "loss": 3.3881,
      "step": 42400
    },
    {
      "epoch": 4.568937681627381,
      "grad_norm": 0.60187828540802,
      "learning_rate": 0.0003263182846676004,
      "loss": 3.4044,
      "step": 42450
    },
    {
      "epoch": 4.57431923366699,
      "grad_norm": 0.598962664604187,
      "learning_rate": 0.00032599504363753904,
      "loss": 3.3836,
      "step": 42500
    },
    {
      "epoch": 4.579700785706597,
      "grad_norm": 0.6940217614173889,
      "learning_rate": 0.00032567180260747764,
      "loss": 3.3951,
      "step": 42550
    },
    {
      "epoch": 4.585082337746206,
      "grad_norm": 0.6212169528007507,
      "learning_rate": 0.0003253485615774162,
      "loss": 3.4021,
      "step": 42600
    },
    {
      "epoch": 4.5904638897858145,
      "grad_norm": 0.6305793523788452,
      "learning_rate": 0.00032502532054735477,
      "loss": 3.3785,
      "step": 42650
    },
    {
      "epoch": 4.595845441825422,
      "grad_norm": 0.6519752740859985,
      "learning_rate": 0.00032470207951729337,
      "loss": 3.3783,
      "step": 42700
    },
    {
      "epoch": 4.601226993865031,
      "grad_norm": 0.6358832716941833,
      "learning_rate": 0.0003243788384872319,
      "loss": 3.3628,
      "step": 42750
    },
    {
      "epoch": 4.606608545904638,
      "grad_norm": 0.6865209341049194,
      "learning_rate": 0.00032405559745717056,
      "loss": 3.3822,
      "step": 42800
    },
    {
      "epoch": 4.611990097944247,
      "grad_norm": 0.6884016394615173,
      "learning_rate": 0.00032373235642710915,
      "loss": 3.3845,
      "step": 42850
    },
    {
      "epoch": 4.6173716499838555,
      "grad_norm": 0.6297987699508667,
      "learning_rate": 0.0003234091153970477,
      "loss": 3.4073,
      "step": 42900
    },
    {
      "epoch": 4.622753202023463,
      "grad_norm": 0.5868752598762512,
      "learning_rate": 0.0003230858743669863,
      "loss": 3.3696,
      "step": 42950
    },
    {
      "epoch": 4.628134754063072,
      "grad_norm": 0.637920081615448,
      "learning_rate": 0.0003227626333369248,
      "loss": 3.3838,
      "step": 43000
    },
    {
      "epoch": 4.628134754063072,
      "eval_accuracy": 0.3791500106316851,
      "eval_loss": 3.4227232933044434,
      "eval_runtime": 190.4399,
      "eval_samples_per_second": 94.576,
      "eval_steps_per_second": 5.913,
      "step": 43000
    },
    {
      "epoch": 4.63351630610268,
      "grad_norm": 0.6682226657867432,
      "learning_rate": 0.0003224393923068635,
      "loss": 3.3917,
      "step": 43050
    },
    {
      "epoch": 4.638897858142288,
      "grad_norm": 0.6613405346870422,
      "learning_rate": 0.00032211615127680207,
      "loss": 3.3964,
      "step": 43100
    },
    {
      "epoch": 4.6442794101818965,
      "grad_norm": 0.6426084041595459,
      "learning_rate": 0.0003217929102467406,
      "loss": 3.3817,
      "step": 43150
    },
    {
      "epoch": 4.649660962221505,
      "grad_norm": 0.6437234282493591,
      "learning_rate": 0.0003214696692166792,
      "loss": 3.3799,
      "step": 43200
    },
    {
      "epoch": 4.655042514261113,
      "grad_norm": 0.6341087818145752,
      "learning_rate": 0.0003211464281866178,
      "loss": 3.3747,
      "step": 43250
    },
    {
      "epoch": 4.660424066300721,
      "grad_norm": 0.6276434659957886,
      "learning_rate": 0.00032082318715655634,
      "loss": 3.3734,
      "step": 43300
    },
    {
      "epoch": 4.665805618340329,
      "grad_norm": 0.6735036969184875,
      "learning_rate": 0.000320499946126495,
      "loss": 3.4059,
      "step": 43350
    },
    {
      "epoch": 4.6711871703799375,
      "grad_norm": 0.6285054683685303,
      "learning_rate": 0.0003201767050964336,
      "loss": 3.3802,
      "step": 43400
    },
    {
      "epoch": 4.676568722419546,
      "grad_norm": 0.6021018624305725,
      "learning_rate": 0.0003198534640663721,
      "loss": 3.3892,
      "step": 43450
    },
    {
      "epoch": 4.681950274459154,
      "grad_norm": 0.7303967475891113,
      "learning_rate": 0.0003195302230363107,
      "loss": 3.3891,
      "step": 43500
    },
    {
      "epoch": 4.687331826498762,
      "grad_norm": 0.6451719403266907,
      "learning_rate": 0.00031920698200624926,
      "loss": 3.3844,
      "step": 43550
    },
    {
      "epoch": 4.692713378538371,
      "grad_norm": 0.6166881322860718,
      "learning_rate": 0.00031888374097618785,
      "loss": 3.3991,
      "step": 43600
    },
    {
      "epoch": 4.6980949305779784,
      "grad_norm": 0.641990602016449,
      "learning_rate": 0.0003185604999461265,
      "loss": 3.3997,
      "step": 43650
    },
    {
      "epoch": 4.703476482617587,
      "grad_norm": 0.5929169058799744,
      "learning_rate": 0.00031823725891606504,
      "loss": 3.3805,
      "step": 43700
    },
    {
      "epoch": 4.7088580346571955,
      "grad_norm": 0.6536329984664917,
      "learning_rate": 0.00031791401788600364,
      "loss": 3.3962,
      "step": 43750
    },
    {
      "epoch": 4.714239586696803,
      "grad_norm": 0.6582703590393066,
      "learning_rate": 0.00031759077685594223,
      "loss": 3.3837,
      "step": 43800
    },
    {
      "epoch": 4.719621138736412,
      "grad_norm": 0.6824732422828674,
      "learning_rate": 0.00031726753582588077,
      "loss": 3.3869,
      "step": 43850
    },
    {
      "epoch": 4.725002690776019,
      "grad_norm": 0.602605402469635,
      "learning_rate": 0.0003169442947958194,
      "loss": 3.3761,
      "step": 43900
    },
    {
      "epoch": 4.730384242815628,
      "grad_norm": 0.6119815111160278,
      "learning_rate": 0.000316621053765758,
      "loss": 3.3943,
      "step": 43950
    },
    {
      "epoch": 4.7357657948552365,
      "grad_norm": 0.6312377452850342,
      "learning_rate": 0.00031629781273569656,
      "loss": 3.4042,
      "step": 44000
    },
    {
      "epoch": 4.7357657948552365,
      "eval_accuracy": 0.37972467575533586,
      "eval_loss": 3.416722297668457,
      "eval_runtime": 190.0781,
      "eval_samples_per_second": 94.756,
      "eval_steps_per_second": 5.924,
      "step": 44000
    },
    {
      "epoch": 4.741147346894844,
      "grad_norm": 0.6086893081665039,
      "learning_rate": 0.00031597457170563515,
      "loss": 3.3884,
      "step": 44050
    },
    {
      "epoch": 4.746528898934453,
      "grad_norm": 0.6373066306114197,
      "learning_rate": 0.0003156513306755737,
      "loss": 3.3933,
      "step": 44100
    },
    {
      "epoch": 4.751910450974061,
      "grad_norm": 0.6317101120948792,
      "learning_rate": 0.0003153280896455123,
      "loss": 3.3983,
      "step": 44150
    },
    {
      "epoch": 4.757292003013669,
      "grad_norm": 0.6273621320724487,
      "learning_rate": 0.00031500484861545094,
      "loss": 3.3824,
      "step": 44200
    },
    {
      "epoch": 4.7626735550532775,
      "grad_norm": 0.6613987684249878,
      "learning_rate": 0.0003146816075853895,
      "loss": 3.3801,
      "step": 44250
    },
    {
      "epoch": 4.768055107092886,
      "grad_norm": 0.6187384128570557,
      "learning_rate": 0.00031435836655532807,
      "loss": 3.3887,
      "step": 44300
    },
    {
      "epoch": 4.773436659132494,
      "grad_norm": 0.6020026803016663,
      "learning_rate": 0.00031403512552526667,
      "loss": 3.3791,
      "step": 44350
    },
    {
      "epoch": 4.778818211172102,
      "grad_norm": 0.6246537566184998,
      "learning_rate": 0.0003137118844952052,
      "loss": 3.3947,
      "step": 44400
    },
    {
      "epoch": 4.78419976321171,
      "grad_norm": 0.6720128655433655,
      "learning_rate": 0.0003133886434651438,
      "loss": 3.381,
      "step": 44450
    },
    {
      "epoch": 4.7895813152513185,
      "grad_norm": 0.6206055879592896,
      "learning_rate": 0.00031306540243508245,
      "loss": 3.3935,
      "step": 44500
    },
    {
      "epoch": 4.794962867290927,
      "grad_norm": 0.6099329590797424,
      "learning_rate": 0.000312742161405021,
      "loss": 3.3881,
      "step": 44550
    },
    {
      "epoch": 4.800344419330535,
      "grad_norm": 0.5944895148277283,
      "learning_rate": 0.0003124189203749596,
      "loss": 3.3854,
      "step": 44600
    },
    {
      "epoch": 4.805725971370143,
      "grad_norm": 0.6910694241523743,
      "learning_rate": 0.0003120956793448981,
      "loss": 3.3742,
      "step": 44650
    },
    {
      "epoch": 4.811107523409751,
      "grad_norm": 0.6518204808235168,
      "learning_rate": 0.0003117724383148367,
      "loss": 3.3738,
      "step": 44700
    },
    {
      "epoch": 4.8164890754493594,
      "grad_norm": 0.6314294338226318,
      "learning_rate": 0.00031144919728477526,
      "loss": 3.381,
      "step": 44750
    },
    {
      "epoch": 4.821870627488968,
      "grad_norm": 0.6671834588050842,
      "learning_rate": 0.0003111259562547139,
      "loss": 3.3889,
      "step": 44800
    },
    {
      "epoch": 4.827252179528576,
      "grad_norm": 0.610420823097229,
      "learning_rate": 0.0003108027152246525,
      "loss": 3.3668,
      "step": 44850
    },
    {
      "epoch": 4.832633731568184,
      "grad_norm": 0.6934123635292053,
      "learning_rate": 0.0003104859390151923,
      "loss": 3.3965,
      "step": 44900
    },
    {
      "epoch": 4.838015283607793,
      "grad_norm": 0.6090071201324463,
      "learning_rate": 0.0003101626979851309,
      "loss": 3.3755,
      "step": 44950
    },
    {
      "epoch": 4.8433968356474,
      "grad_norm": 0.6591959595680237,
      "learning_rate": 0.00030983945695506945,
      "loss": 3.4036,
      "step": 45000
    },
    {
      "epoch": 4.8433968356474,
      "eval_accuracy": 0.3799556717948295,
      "eval_loss": 3.4122121334075928,
      "eval_runtime": 190.2985,
      "eval_samples_per_second": 94.646,
      "eval_steps_per_second": 5.917,
      "step": 45000
    },
    {
      "epoch": 4.848778387687009,
      "grad_norm": 0.6309224367141724,
      "learning_rate": 0.00030951621592500804,
      "loss": 3.393,
      "step": 45050
    },
    {
      "epoch": 4.8541599397266175,
      "grad_norm": 0.6031204462051392,
      "learning_rate": 0.00030919297489494663,
      "loss": 3.3962,
      "step": 45100
    },
    {
      "epoch": 4.859541491766225,
      "grad_norm": 0.6486292481422424,
      "learning_rate": 0.00030886973386488523,
      "loss": 3.3778,
      "step": 45150
    },
    {
      "epoch": 4.864923043805834,
      "grad_norm": 0.6576681733131409,
      "learning_rate": 0.0003085464928348238,
      "loss": 3.3739,
      "step": 45200
    },
    {
      "epoch": 4.870304595845441,
      "grad_norm": 0.6652286648750305,
      "learning_rate": 0.0003082232518047624,
      "loss": 3.3891,
      "step": 45250
    },
    {
      "epoch": 4.87568614788505,
      "grad_norm": 0.6884920597076416,
      "learning_rate": 0.00030790001077470096,
      "loss": 3.3853,
      "step": 45300
    },
    {
      "epoch": 4.8810676999246585,
      "grad_norm": 0.6760656237602234,
      "learning_rate": 0.00030757676974463955,
      "loss": 3.383,
      "step": 45350
    },
    {
      "epoch": 4.886449251964266,
      "grad_norm": 0.6775695085525513,
      "learning_rate": 0.0003072535287145781,
      "loss": 3.3977,
      "step": 45400
    },
    {
      "epoch": 4.891830804003875,
      "grad_norm": 0.6460280418395996,
      "learning_rate": 0.00030693028768451674,
      "loss": 3.3793,
      "step": 45450
    },
    {
      "epoch": 4.897212356043483,
      "grad_norm": 0.6144499182701111,
      "learning_rate": 0.00030660704665445534,
      "loss": 3.3798,
      "step": 45500
    },
    {
      "epoch": 4.902593908083091,
      "grad_norm": 0.6140328645706177,
      "learning_rate": 0.0003062838056243939,
      "loss": 3.3743,
      "step": 45550
    },
    {
      "epoch": 4.9079754601226995,
      "grad_norm": 0.6704367995262146,
      "learning_rate": 0.00030596056459433247,
      "loss": 3.3799,
      "step": 45600
    },
    {
      "epoch": 4.913357012162308,
      "grad_norm": 0.6094426512718201,
      "learning_rate": 0.00030563732356427107,
      "loss": 3.3813,
      "step": 45650
    },
    {
      "epoch": 4.918738564201916,
      "grad_norm": 0.6442837119102478,
      "learning_rate": 0.00030531408253420966,
      "loss": 3.3844,
      "step": 45700
    },
    {
      "epoch": 4.924120116241524,
      "grad_norm": 0.6164265275001526,
      "learning_rate": 0.00030499084150414826,
      "loss": 3.3752,
      "step": 45750
    },
    {
      "epoch": 4.929501668281132,
      "grad_norm": 0.5994932055473328,
      "learning_rate": 0.00030466760047408685,
      "loss": 3.3809,
      "step": 45800
    },
    {
      "epoch": 4.9348832203207404,
      "grad_norm": 0.6426939368247986,
      "learning_rate": 0.0003043443594440254,
      "loss": 3.3974,
      "step": 45850
    },
    {
      "epoch": 4.940264772360349,
      "grad_norm": 0.6147627234458923,
      "learning_rate": 0.000304021118413964,
      "loss": 3.3937,
      "step": 45900
    },
    {
      "epoch": 4.945646324399957,
      "grad_norm": 0.6684779524803162,
      "learning_rate": 0.0003036978773839025,
      "loss": 3.3732,
      "step": 45950
    },
    {
      "epoch": 4.951027876439565,
      "grad_norm": 0.6723201870918274,
      "learning_rate": 0.0003033746363538412,
      "loss": 3.3922,
      "step": 46000
    },
    {
      "epoch": 4.951027876439565,
      "eval_accuracy": 0.380636490789875,
      "eval_loss": 3.4082179069519043,
      "eval_runtime": 190.2161,
      "eval_samples_per_second": 94.687,
      "eval_steps_per_second": 5.92,
      "step": 46000
    },
    {
      "epoch": 4.956409428479174,
      "grad_norm": 0.6478368639945984,
      "learning_rate": 0.00030305139532377977,
      "loss": 3.3799,
      "step": 46050
    },
    {
      "epoch": 4.961790980518781,
      "grad_norm": 0.6748535633087158,
      "learning_rate": 0.0003027281542937183,
      "loss": 3.3928,
      "step": 46100
    },
    {
      "epoch": 4.96717253255839,
      "grad_norm": 0.6150487661361694,
      "learning_rate": 0.0003024049132636569,
      "loss": 3.3907,
      "step": 46150
    },
    {
      "epoch": 4.9725540845979985,
      "grad_norm": 0.6397275328636169,
      "learning_rate": 0.0003020816722335955,
      "loss": 3.4075,
      "step": 46200
    },
    {
      "epoch": 4.977935636637606,
      "grad_norm": 0.6354976296424866,
      "learning_rate": 0.00030175843120353404,
      "loss": 3.3782,
      "step": 46250
    },
    {
      "epoch": 4.983317188677215,
      "grad_norm": 0.6589702367782593,
      "learning_rate": 0.0003014351901734727,
      "loss": 3.383,
      "step": 46300
    },
    {
      "epoch": 4.988698740716822,
      "grad_norm": 0.5894850492477417,
      "learning_rate": 0.0003011119491434113,
      "loss": 3.381,
      "step": 46350
    },
    {
      "epoch": 4.994080292756431,
      "grad_norm": 0.6286249160766602,
      "learning_rate": 0.0003007887081133498,
      "loss": 3.3782,
      "step": 46400
    },
    {
      "epoch": 4.9994618447960395,
      "grad_norm": 0.6172303557395935,
      "learning_rate": 0.0003004654670832884,
      "loss": 3.3761,
      "step": 46450
    },
    {
      "epoch": 5.004843396835647,
      "grad_norm": 0.6467801332473755,
      "learning_rate": 0.00030014222605322696,
      "loss": 3.2933,
      "step": 46500
    },
    {
      "epoch": 5.010224948875256,
      "grad_norm": 0.6523343324661255,
      "learning_rate": 0.00029981898502316555,
      "loss": 3.2978,
      "step": 46550
    },
    {
      "epoch": 5.015606500914864,
      "grad_norm": 0.6472421288490295,
      "learning_rate": 0.00029949574399310415,
      "loss": 3.2995,
      "step": 46600
    },
    {
      "epoch": 5.020988052954472,
      "grad_norm": 0.692363440990448,
      "learning_rate": 0.00029917250296304274,
      "loss": 3.2779,
      "step": 46650
    },
    {
      "epoch": 5.0263696049940805,
      "grad_norm": 0.66607666015625,
      "learning_rate": 0.00029884926193298134,
      "loss": 3.2956,
      "step": 46700
    },
    {
      "epoch": 5.031751157033688,
      "grad_norm": 0.6656844019889832,
      "learning_rate": 0.0002985260209029199,
      "loss": 3.2984,
      "step": 46750
    },
    {
      "epoch": 5.037132709073297,
      "grad_norm": 0.6590693593025208,
      "learning_rate": 0.00029820277987285853,
      "loss": 3.2883,
      "step": 46800
    },
    {
      "epoch": 5.042514261112905,
      "grad_norm": 0.7028095722198486,
      "learning_rate": 0.00029787953884279707,
      "loss": 3.3185,
      "step": 46850
    },
    {
      "epoch": 5.047895813152513,
      "grad_norm": 0.6914402842521667,
      "learning_rate": 0.00029756276263333693,
      "loss": 3.3079,
      "step": 46900
    },
    {
      "epoch": 5.0532773651921215,
      "grad_norm": 0.6603579521179199,
      "learning_rate": 0.00029723952160327547,
      "loss": 3.3142,
      "step": 46950
    },
    {
      "epoch": 5.05865891723173,
      "grad_norm": 0.6558881402015686,
      "learning_rate": 0.00029691628057321406,
      "loss": 3.2894,
      "step": 47000
    },
    {
      "epoch": 5.05865891723173,
      "eval_accuracy": 0.3807744799573994,
      "eval_loss": 3.4128940105438232,
      "eval_runtime": 189.9502,
      "eval_samples_per_second": 94.82,
      "eval_steps_per_second": 5.928,
      "step": 47000
    },
    {
      "epoch": 5.064040469271338,
      "grad_norm": 0.6843393445014954,
      "learning_rate": 0.00029659303954315266,
      "loss": 3.2938,
      "step": 47050
    },
    {
      "epoch": 5.069422021310946,
      "grad_norm": 0.6472140550613403,
      "learning_rate": 0.00029626979851309125,
      "loss": 3.3054,
      "step": 47100
    },
    {
      "epoch": 5.074803573350554,
      "grad_norm": 0.627531886100769,
      "learning_rate": 0.00029594655748302985,
      "loss": 3.3064,
      "step": 47150
    },
    {
      "epoch": 5.080185125390162,
      "grad_norm": 0.6687477827072144,
      "learning_rate": 0.0002956233164529684,
      "loss": 3.285,
      "step": 47200
    },
    {
      "epoch": 5.085566677429771,
      "grad_norm": 0.6495850682258606,
      "learning_rate": 0.000295300075422907,
      "loss": 3.3041,
      "step": 47250
    },
    {
      "epoch": 5.090948229469379,
      "grad_norm": 0.637628436088562,
      "learning_rate": 0.0002949768343928456,
      "loss": 3.3134,
      "step": 47300
    },
    {
      "epoch": 5.096329781508987,
      "grad_norm": 0.6417757868766785,
      "learning_rate": 0.00029465359336278417,
      "loss": 3.3,
      "step": 47350
    },
    {
      "epoch": 5.101711333548596,
      "grad_norm": 0.6612465381622314,
      "learning_rate": 0.0002943303523327227,
      "loss": 3.2839,
      "step": 47400
    },
    {
      "epoch": 5.107092885588203,
      "grad_norm": 0.680810809135437,
      "learning_rate": 0.00029400711130266136,
      "loss": 3.3289,
      "step": 47450
    },
    {
      "epoch": 5.112474437627812,
      "grad_norm": 0.7323889136314392,
      "learning_rate": 0.0002936838702725999,
      "loss": 3.3161,
      "step": 47500
    },
    {
      "epoch": 5.1178559896674205,
      "grad_norm": 0.6742674112319946,
      "learning_rate": 0.0002933606292425385,
      "loss": 3.3134,
      "step": 47550
    },
    {
      "epoch": 5.123237541707028,
      "grad_norm": 0.6374590992927551,
      "learning_rate": 0.0002930373882124771,
      "loss": 3.3231,
      "step": 47600
    },
    {
      "epoch": 5.128619093746637,
      "grad_norm": 0.6212331056594849,
      "learning_rate": 0.0002927141471824157,
      "loss": 3.2949,
      "step": 47650
    },
    {
      "epoch": 5.134000645786244,
      "grad_norm": 0.6317815780639648,
      "learning_rate": 0.0002923909061523542,
      "loss": 3.3224,
      "step": 47700
    },
    {
      "epoch": 5.139382197825853,
      "grad_norm": 0.6304091215133667,
      "learning_rate": 0.0002920676651222928,
      "loss": 3.2972,
      "step": 47750
    },
    {
      "epoch": 5.1447637498654615,
      "grad_norm": 0.6608728170394897,
      "learning_rate": 0.0002917444240922314,
      "loss": 3.313,
      "step": 47800
    },
    {
      "epoch": 5.150145301905069,
      "grad_norm": 0.6596003174781799,
      "learning_rate": 0.00029142118306216996,
      "loss": 3.3159,
      "step": 47850
    },
    {
      "epoch": 5.155526853944678,
      "grad_norm": 0.6998940110206604,
      "learning_rate": 0.0002910979420321086,
      "loss": 3.3174,
      "step": 47900
    },
    {
      "epoch": 5.160908405984286,
      "grad_norm": 0.6569470167160034,
      "learning_rate": 0.00029077470100204715,
      "loss": 3.3091,
      "step": 47950
    },
    {
      "epoch": 5.166289958023894,
      "grad_norm": 0.6322058439254761,
      "learning_rate": 0.00029045145997198574,
      "loss": 3.3025,
      "step": 48000
    },
    {
      "epoch": 5.166289958023894,
      "eval_accuracy": 0.38140401478939645,
      "eval_loss": 3.408338785171509,
      "eval_runtime": 190.2616,
      "eval_samples_per_second": 94.664,
      "eval_steps_per_second": 5.918,
      "step": 48000
    },
    {
      "epoch": 5.1716715100635025,
      "grad_norm": 0.6784811615943909,
      "learning_rate": 0.00029012821894192433,
      "loss": 3.3128,
      "step": 48050
    },
    {
      "epoch": 5.17705306210311,
      "grad_norm": 0.7051177620887756,
      "learning_rate": 0.00028980497791186293,
      "loss": 3.3115,
      "step": 48100
    },
    {
      "epoch": 5.182434614142719,
      "grad_norm": 0.6922482252120972,
      "learning_rate": 0.0002894817368818015,
      "loss": 3.3106,
      "step": 48150
    },
    {
      "epoch": 5.187816166182327,
      "grad_norm": 0.6171578764915466,
      "learning_rate": 0.00028915849585174006,
      "loss": 3.3047,
      "step": 48200
    },
    {
      "epoch": 5.193197718221935,
      "grad_norm": 0.6317104697227478,
      "learning_rate": 0.00028883525482167866,
      "loss": 3.3077,
      "step": 48250
    },
    {
      "epoch": 5.198579270261543,
      "grad_norm": 0.7686342000961304,
      "learning_rate": 0.00028851201379161725,
      "loss": 3.334,
      "step": 48300
    },
    {
      "epoch": 5.203960822301152,
      "grad_norm": 0.6540053486824036,
      "learning_rate": 0.00028818877276155585,
      "loss": 3.2984,
      "step": 48350
    },
    {
      "epoch": 5.20934237434076,
      "grad_norm": 0.6591601371765137,
      "learning_rate": 0.0002878655317314944,
      "loss": 3.3177,
      "step": 48400
    },
    {
      "epoch": 5.214723926380368,
      "grad_norm": 0.6442411541938782,
      "learning_rate": 0.00028754229070143304,
      "loss": 3.3122,
      "step": 48450
    },
    {
      "epoch": 5.220105478419977,
      "grad_norm": 0.6508519649505615,
      "learning_rate": 0.0002872190496713716,
      "loss": 3.3051,
      "step": 48500
    },
    {
      "epoch": 5.225487030459584,
      "grad_norm": 0.6853311657905579,
      "learning_rate": 0.0002868958086413102,
      "loss": 3.3272,
      "step": 48550
    },
    {
      "epoch": 5.230868582499193,
      "grad_norm": 0.674636721611023,
      "learning_rate": 0.00028657256761124877,
      "loss": 3.3136,
      "step": 48600
    },
    {
      "epoch": 5.236250134538801,
      "grad_norm": 0.7199856638908386,
      "learning_rate": 0.00028624932658118736,
      "loss": 3.3287,
      "step": 48650
    },
    {
      "epoch": 5.241631686578409,
      "grad_norm": 0.6740505695343018,
      "learning_rate": 0.0002859260855511259,
      "loss": 3.3269,
      "step": 48700
    },
    {
      "epoch": 5.247013238618018,
      "grad_norm": 0.6485350728034973,
      "learning_rate": 0.0002856028445210645,
      "loss": 3.3185,
      "step": 48750
    },
    {
      "epoch": 5.252394790657625,
      "grad_norm": 0.6070066094398499,
      "learning_rate": 0.0002852796034910031,
      "loss": 3.3233,
      "step": 48800
    },
    {
      "epoch": 5.257776342697234,
      "grad_norm": 0.6313185095787048,
      "learning_rate": 0.0002849563624609417,
      "loss": 3.3378,
      "step": 48850
    },
    {
      "epoch": 5.2631578947368425,
      "grad_norm": 0.6935280561447144,
      "learning_rate": 0.0002846331214308803,
      "loss": 3.3371,
      "step": 48900
    },
    {
      "epoch": 5.26853944677645,
      "grad_norm": 0.6744174361228943,
      "learning_rate": 0.0002843163452214201,
      "loss": 3.3083,
      "step": 48950
    },
    {
      "epoch": 5.273920998816059,
      "grad_norm": 0.6619080305099487,
      "learning_rate": 0.0002839931041913587,
      "loss": 3.3297,
      "step": 49000
    },
    {
      "epoch": 5.273920998816059,
      "eval_accuracy": 0.38157568635214334,
      "eval_loss": 3.406853675842285,
      "eval_runtime": 189.811,
      "eval_samples_per_second": 94.889,
      "eval_steps_per_second": 5.932,
      "step": 49000
    },
    {
      "epoch": 5.279302550855666,
      "grad_norm": 0.6502954363822937,
      "learning_rate": 0.0002836698631612972,
      "loss": 3.308,
      "step": 49050
    },
    {
      "epoch": 5.284684102895275,
      "grad_norm": 0.6972982287406921,
      "learning_rate": 0.00028334662213123587,
      "loss": 3.3174,
      "step": 49100
    },
    {
      "epoch": 5.2900656549348835,
      "grad_norm": 0.6818156242370605,
      "learning_rate": 0.0002830233811011744,
      "loss": 3.3101,
      "step": 49150
    },
    {
      "epoch": 5.295447206974491,
      "grad_norm": 0.688595175743103,
      "learning_rate": 0.000282700140071113,
      "loss": 3.3305,
      "step": 49200
    },
    {
      "epoch": 5.3008287590141,
      "grad_norm": 0.6568655967712402,
      "learning_rate": 0.0002823768990410516,
      "loss": 3.3256,
      "step": 49250
    },
    {
      "epoch": 5.306210311053708,
      "grad_norm": 0.6636142730712891,
      "learning_rate": 0.00028205365801099014,
      "loss": 3.3177,
      "step": 49300
    },
    {
      "epoch": 5.311591863093316,
      "grad_norm": 0.7073917984962463,
      "learning_rate": 0.00028173041698092874,
      "loss": 3.3228,
      "step": 49350
    },
    {
      "epoch": 5.316973415132924,
      "grad_norm": 0.6609116196632385,
      "learning_rate": 0.00028140717595086733,
      "loss": 3.3097,
      "step": 49400
    },
    {
      "epoch": 5.322354967172533,
      "grad_norm": 0.9245494604110718,
      "learning_rate": 0.0002810839349208059,
      "loss": 3.3218,
      "step": 49450
    },
    {
      "epoch": 5.327736519212141,
      "grad_norm": 0.6439619660377502,
      "learning_rate": 0.00028076069389074447,
      "loss": 3.3371,
      "step": 49500
    },
    {
      "epoch": 5.333118071251749,
      "grad_norm": 0.7020370364189148,
      "learning_rate": 0.0002804374528606831,
      "loss": 3.3362,
      "step": 49550
    },
    {
      "epoch": 5.338499623291357,
      "grad_norm": 0.7692583203315735,
      "learning_rate": 0.00028011421183062166,
      "loss": 3.3295,
      "step": 49600
    },
    {
      "epoch": 5.343881175330965,
      "grad_norm": 0.6994715332984924,
      "learning_rate": 0.00027979097080056025,
      "loss": 3.3278,
      "step": 49650
    },
    {
      "epoch": 5.349262727370574,
      "grad_norm": 0.6521602272987366,
      "learning_rate": 0.00027946772977049885,
      "loss": 3.3467,
      "step": 49700
    },
    {
      "epoch": 5.354644279410182,
      "grad_norm": 0.635476291179657,
      "learning_rate": 0.00027914448874043744,
      "loss": 3.3291,
      "step": 49750
    },
    {
      "epoch": 5.36002583144979,
      "grad_norm": 0.6542438268661499,
      "learning_rate": 0.00027882124771037603,
      "loss": 3.3436,
      "step": 49800
    },
    {
      "epoch": 5.365407383489399,
      "grad_norm": 0.6599149703979492,
      "learning_rate": 0.0002784980066803146,
      "loss": 3.3114,
      "step": 49850
    },
    {
      "epoch": 5.370788935529006,
      "grad_norm": 0.6250675916671753,
      "learning_rate": 0.00027817476565025317,
      "loss": 3.3453,
      "step": 49900
    },
    {
      "epoch": 5.376170487568615,
      "grad_norm": 0.6826515793800354,
      "learning_rate": 0.00027785152462019176,
      "loss": 3.3307,
      "step": 49950
    },
    {
      "epoch": 5.3815520396082235,
      "grad_norm": 0.6683448553085327,
      "learning_rate": 0.00027752828359013036,
      "loss": 3.3347,
      "step": 50000
    },
    {
      "epoch": 5.3815520396082235,
      "eval_accuracy": 0.3819247880806913,
      "eval_loss": 3.4006173610687256,
      "eval_runtime": 190.2807,
      "eval_samples_per_second": 94.655,
      "eval_steps_per_second": 5.918,
      "step": 50000
    },
    {
      "epoch": 5.386933591647831,
      "grad_norm": 0.6442648768424988,
      "learning_rate": 0.0002772050425600689,
      "loss": 3.3459,
      "step": 50050
    },
    {
      "epoch": 5.39231514368744,
      "grad_norm": 0.6820710897445679,
      "learning_rate": 0.00027688180153000755,
      "loss": 3.329,
      "step": 50100
    },
    {
      "epoch": 5.397696695727047,
      "grad_norm": 0.6910305619239807,
      "learning_rate": 0.0002765585604999461,
      "loss": 3.3357,
      "step": 50150
    },
    {
      "epoch": 5.403078247766656,
      "grad_norm": 0.6926243305206299,
      "learning_rate": 0.0002762353194698847,
      "loss": 3.3114,
      "step": 50200
    },
    {
      "epoch": 5.4084597998062645,
      "grad_norm": 0.7306303977966309,
      "learning_rate": 0.0002759120784398233,
      "loss": 3.3243,
      "step": 50250
    },
    {
      "epoch": 5.413841351845872,
      "grad_norm": 0.695617139339447,
      "learning_rate": 0.00027558883740976187,
      "loss": 3.3225,
      "step": 50300
    },
    {
      "epoch": 5.419222903885481,
      "grad_norm": 0.6790807843208313,
      "learning_rate": 0.0002752655963797004,
      "loss": 3.3303,
      "step": 50350
    },
    {
      "epoch": 5.424604455925088,
      "grad_norm": 0.6384757161140442,
      "learning_rate": 0.000274942355349639,
      "loss": 3.3333,
      "step": 50400
    },
    {
      "epoch": 5.429986007964697,
      "grad_norm": 0.685840904712677,
      "learning_rate": 0.0002746191143195776,
      "loss": 3.3157,
      "step": 50450
    },
    {
      "epoch": 5.435367560004305,
      "grad_norm": 0.6745836138725281,
      "learning_rate": 0.0002742958732895162,
      "loss": 3.3404,
      "step": 50500
    },
    {
      "epoch": 5.440749112043913,
      "grad_norm": 0.6714034080505371,
      "learning_rate": 0.0002739726322594548,
      "loss": 3.3319,
      "step": 50550
    },
    {
      "epoch": 5.446130664083522,
      "grad_norm": 0.7188451290130615,
      "learning_rate": 0.00027364939122939333,
      "loss": 3.3436,
      "step": 50600
    },
    {
      "epoch": 5.45151221612313,
      "grad_norm": 0.6573353409767151,
      "learning_rate": 0.0002733261501993319,
      "loss": 3.3336,
      "step": 50650
    },
    {
      "epoch": 5.456893768162738,
      "grad_norm": 0.6462401747703552,
      "learning_rate": 0.0002730029091692705,
      "loss": 3.3437,
      "step": 50700
    },
    {
      "epoch": 5.462275320202346,
      "grad_norm": 0.6640335917472839,
      "learning_rate": 0.0002726796681392091,
      "loss": 3.3312,
      "step": 50750
    },
    {
      "epoch": 5.467656872241955,
      "grad_norm": 0.656714916229248,
      "learning_rate": 0.0002723564271091477,
      "loss": 3.3215,
      "step": 50800
    },
    {
      "epoch": 5.473038424281563,
      "grad_norm": 0.6289355754852295,
      "learning_rate": 0.0002720331860790863,
      "loss": 3.3266,
      "step": 50850
    },
    {
      "epoch": 5.478419976321171,
      "grad_norm": 0.661991536617279,
      "learning_rate": 0.00027170994504902485,
      "loss": 3.3208,
      "step": 50900
    },
    {
      "epoch": 5.483801528360779,
      "grad_norm": 0.6606703400611877,
      "learning_rate": 0.00027138670401896344,
      "loss": 3.3356,
      "step": 50950
    },
    {
      "epoch": 5.489183080400387,
      "grad_norm": 0.7058117985725403,
      "learning_rate": 0.00027106992780950325,
      "loss": 3.3325,
      "step": 51000
    },
    {
      "epoch": 5.489183080400387,
      "eval_accuracy": 0.3821605648472487,
      "eval_loss": 3.3974854946136475,
      "eval_runtime": 190.1283,
      "eval_samples_per_second": 94.731,
      "eval_steps_per_second": 5.922,
      "step": 51000
    },
    {
      "epoch": 5.494564632439996,
      "grad_norm": 0.7179932594299316,
      "learning_rate": 0.00027074668677944184,
      "loss": 3.3297,
      "step": 51050
    },
    {
      "epoch": 5.499946184479604,
      "grad_norm": 0.6589070558547974,
      "learning_rate": 0.00027042344574938044,
      "loss": 3.3301,
      "step": 51100
    },
    {
      "epoch": 5.505327736519212,
      "grad_norm": 0.663331925868988,
      "learning_rate": 0.000270100204719319,
      "loss": 3.3403,
      "step": 51150
    },
    {
      "epoch": 5.510709288558821,
      "grad_norm": 0.7381957769393921,
      "learning_rate": 0.0002697769636892576,
      "loss": 3.332,
      "step": 51200
    },
    {
      "epoch": 5.516090840598428,
      "grad_norm": 0.6564040184020996,
      "learning_rate": 0.00026945372265919617,
      "loss": 3.3178,
      "step": 51250
    },
    {
      "epoch": 5.521472392638037,
      "grad_norm": 0.6270856857299805,
      "learning_rate": 0.00026913048162913476,
      "loss": 3.3374,
      "step": 51300
    },
    {
      "epoch": 5.5268539446776455,
      "grad_norm": 0.6517801284790039,
      "learning_rate": 0.00026880724059907336,
      "loss": 3.3378,
      "step": 51350
    },
    {
      "epoch": 5.532235496717253,
      "grad_norm": 0.7353670597076416,
      "learning_rate": 0.00026848399956901195,
      "loss": 3.3211,
      "step": 51400
    },
    {
      "epoch": 5.537617048756862,
      "grad_norm": 0.6573992371559143,
      "learning_rate": 0.0002681607585389505,
      "loss": 3.3185,
      "step": 51450
    },
    {
      "epoch": 5.542998600796469,
      "grad_norm": 0.6705713868141174,
      "learning_rate": 0.0002678375175088891,
      "loss": 3.3226,
      "step": 51500
    },
    {
      "epoch": 5.548380152836078,
      "grad_norm": 0.7160366773605347,
      "learning_rate": 0.0002675142764788277,
      "loss": 3.3221,
      "step": 51550
    },
    {
      "epoch": 5.553761704875686,
      "grad_norm": 0.6563833951950073,
      "learning_rate": 0.0002671910354487663,
      "loss": 3.3367,
      "step": 51600
    },
    {
      "epoch": 5.559143256915294,
      "grad_norm": 0.6591808795928955,
      "learning_rate": 0.00026686779441870487,
      "loss": 3.3232,
      "step": 51650
    },
    {
      "epoch": 5.564524808954903,
      "grad_norm": 0.7058665156364441,
      "learning_rate": 0.0002665445533886434,
      "loss": 3.3415,
      "step": 51700
    },
    {
      "epoch": 5.569906360994511,
      "grad_norm": 0.6810771822929382,
      "learning_rate": 0.00026622131235858206,
      "loss": 3.3312,
      "step": 51750
    },
    {
      "epoch": 5.575287913034119,
      "grad_norm": 0.6352242231369019,
      "learning_rate": 0.0002658980713285206,
      "loss": 3.3427,
      "step": 51800
    },
    {
      "epoch": 5.580669465073727,
      "grad_norm": 0.727452278137207,
      "learning_rate": 0.0002655748302984592,
      "loss": 3.346,
      "step": 51850
    },
    {
      "epoch": 5.586051017113336,
      "grad_norm": 0.738322913646698,
      "learning_rate": 0.0002652515892683978,
      "loss": 3.3333,
      "step": 51900
    },
    {
      "epoch": 5.591432569152944,
      "grad_norm": 0.7183504700660706,
      "learning_rate": 0.0002649283482383364,
      "loss": 3.3285,
      "step": 51950
    },
    {
      "epoch": 5.596814121192552,
      "grad_norm": 0.6700738072395325,
      "learning_rate": 0.0002646051072082749,
      "loss": 3.3352,
      "step": 52000
    },
    {
      "epoch": 5.596814121192552,
      "eval_accuracy": 0.38317081939814773,
      "eval_loss": 3.3900537490844727,
      "eval_runtime": 190.2312,
      "eval_samples_per_second": 94.68,
      "eval_steps_per_second": 5.919,
      "step": 52000
    },
    {
      "epoch": 5.60219567323216,
      "grad_norm": 0.6733425259590149,
      "learning_rate": 0.0002642818661782135,
      "loss": 3.3392,
      "step": 52050
    },
    {
      "epoch": 5.607577225271768,
      "grad_norm": 0.6791594624519348,
      "learning_rate": 0.0002639586251481521,
      "loss": 3.3304,
      "step": 52100
    },
    {
      "epoch": 5.612958777311377,
      "grad_norm": 0.7923411130905151,
      "learning_rate": 0.0002636353841180907,
      "loss": 3.3283,
      "step": 52150
    },
    {
      "epoch": 5.618340329350985,
      "grad_norm": 0.6906782984733582,
      "learning_rate": 0.0002633121430880293,
      "loss": 3.3379,
      "step": 52200
    },
    {
      "epoch": 5.623721881390593,
      "grad_norm": 0.691555380821228,
      "learning_rate": 0.00026298890205796784,
      "loss": 3.3359,
      "step": 52250
    },
    {
      "epoch": 5.629103433430201,
      "grad_norm": 0.6687930226325989,
      "learning_rate": 0.00026266566102790644,
      "loss": 3.3255,
      "step": 52300
    },
    {
      "epoch": 5.634484985469809,
      "grad_norm": 0.6763534545898438,
      "learning_rate": 0.00026234241999784503,
      "loss": 3.3341,
      "step": 52350
    },
    {
      "epoch": 5.639866537509418,
      "grad_norm": 0.670367419719696,
      "learning_rate": 0.0002620191789677836,
      "loss": 3.3132,
      "step": 52400
    },
    {
      "epoch": 5.645248089549026,
      "grad_norm": 0.6834412217140198,
      "learning_rate": 0.00026169593793772217,
      "loss": 3.3286,
      "step": 52450
    },
    {
      "epoch": 5.650629641588634,
      "grad_norm": 0.6871232390403748,
      "learning_rate": 0.0002613726969076608,
      "loss": 3.3358,
      "step": 52500
    },
    {
      "epoch": 5.656011193628243,
      "grad_norm": 0.670904278755188,
      "learning_rate": 0.00026104945587759936,
      "loss": 3.345,
      "step": 52550
    },
    {
      "epoch": 5.66139274566785,
      "grad_norm": 0.6787131428718567,
      "learning_rate": 0.00026072621484753795,
      "loss": 3.3151,
      "step": 52600
    },
    {
      "epoch": 5.666774297707459,
      "grad_norm": 0.6872842311859131,
      "learning_rate": 0.00026040297381747655,
      "loss": 3.3393,
      "step": 52650
    },
    {
      "epoch": 5.672155849747067,
      "grad_norm": 0.6617214679718018,
      "learning_rate": 0.00026007973278741514,
      "loss": 3.3504,
      "step": 52700
    },
    {
      "epoch": 5.677537401786675,
      "grad_norm": 0.6898906230926514,
      "learning_rate": 0.00025975649175735373,
      "loss": 3.326,
      "step": 52750
    },
    {
      "epoch": 5.682918953826284,
      "grad_norm": 0.6670631170272827,
      "learning_rate": 0.0002594332507272923,
      "loss": 3.3107,
      "step": 52800
    },
    {
      "epoch": 5.688300505865891,
      "grad_norm": 0.6664484739303589,
      "learning_rate": 0.00025911000969723087,
      "loss": 3.3376,
      "step": 52850
    },
    {
      "epoch": 5.6936820579055,
      "grad_norm": 0.6903056502342224,
      "learning_rate": 0.00025878676866716946,
      "loss": 3.3449,
      "step": 52900
    },
    {
      "epoch": 5.699063609945108,
      "grad_norm": 0.6648628115653992,
      "learning_rate": 0.00025846352763710806,
      "loss": 3.3206,
      "step": 52950
    },
    {
      "epoch": 5.704445161984716,
      "grad_norm": 0.67125403881073,
      "learning_rate": 0.00025814675142764787,
      "loss": 3.3379,
      "step": 53000
    },
    {
      "epoch": 5.704445161984716,
      "eval_accuracy": 0.3832094998262097,
      "eval_loss": 3.389312982559204,
      "eval_runtime": 190.0622,
      "eval_samples_per_second": 94.764,
      "eval_steps_per_second": 5.924,
      "step": 53000
    },
    {
      "epoch": 5.709826714024325,
      "grad_norm": 0.6492651104927063,
      "learning_rate": 0.00025782351039758646,
      "loss": 3.3206,
      "step": 53050
    },
    {
      "epoch": 5.715208266063933,
      "grad_norm": 0.6747674345970154,
      "learning_rate": 0.000257500269367525,
      "loss": 3.3411,
      "step": 53100
    },
    {
      "epoch": 5.720589818103541,
      "grad_norm": 0.6961668133735657,
      "learning_rate": 0.0002571770283374636,
      "loss": 3.3278,
      "step": 53150
    },
    {
      "epoch": 5.725971370143149,
      "grad_norm": 0.6968915462493896,
      "learning_rate": 0.0002568537873074022,
      "loss": 3.3289,
      "step": 53200
    },
    {
      "epoch": 5.731352922182758,
      "grad_norm": 0.6421557664871216,
      "learning_rate": 0.0002565305462773408,
      "loss": 3.3507,
      "step": 53250
    },
    {
      "epoch": 5.736734474222366,
      "grad_norm": 0.6829367876052856,
      "learning_rate": 0.0002562073052472794,
      "loss": 3.3368,
      "step": 53300
    },
    {
      "epoch": 5.742116026261974,
      "grad_norm": 0.7034095525741577,
      "learning_rate": 0.0002558840642172179,
      "loss": 3.3272,
      "step": 53350
    },
    {
      "epoch": 5.747497578301582,
      "grad_norm": 0.6726918816566467,
      "learning_rate": 0.00025556082318715657,
      "loss": 3.3207,
      "step": 53400
    },
    {
      "epoch": 5.75287913034119,
      "grad_norm": 0.7066564559936523,
      "learning_rate": 0.0002552375821570951,
      "loss": 3.3198,
      "step": 53450
    },
    {
      "epoch": 5.758260682380799,
      "grad_norm": 0.7037680745124817,
      "learning_rate": 0.0002549143411270337,
      "loss": 3.3275,
      "step": 53500
    },
    {
      "epoch": 5.763642234420407,
      "grad_norm": 0.6634194850921631,
      "learning_rate": 0.0002545911000969723,
      "loss": 3.3369,
      "step": 53550
    },
    {
      "epoch": 5.769023786460015,
      "grad_norm": 0.6896076798439026,
      "learning_rate": 0.0002542678590669109,
      "loss": 3.3367,
      "step": 53600
    },
    {
      "epoch": 5.774405338499624,
      "grad_norm": 0.703079342842102,
      "learning_rate": 0.00025394461803684943,
      "loss": 3.3158,
      "step": 53650
    },
    {
      "epoch": 5.779786890539231,
      "grad_norm": 0.6666619181632996,
      "learning_rate": 0.00025362137700678803,
      "loss": 3.3219,
      "step": 53700
    },
    {
      "epoch": 5.78516844257884,
      "grad_norm": 0.7185107469558716,
      "learning_rate": 0.0002532981359767266,
      "loss": 3.3287,
      "step": 53750
    },
    {
      "epoch": 5.790549994618448,
      "grad_norm": 0.6933943033218384,
      "learning_rate": 0.0002529748949466652,
      "loss": 3.3482,
      "step": 53800
    },
    {
      "epoch": 5.795931546658056,
      "grad_norm": 0.6702208518981934,
      "learning_rate": 0.0002526516539166038,
      "loss": 3.3286,
      "step": 53850
    },
    {
      "epoch": 5.801313098697665,
      "grad_norm": 0.6634646058082581,
      "learning_rate": 0.00025232841288654235,
      "loss": 3.3196,
      "step": 53900
    },
    {
      "epoch": 5.806694650737272,
      "grad_norm": 0.6809012293815613,
      "learning_rate": 0.00025200517185648095,
      "loss": 3.3258,
      "step": 53950
    },
    {
      "epoch": 5.812076202776881,
      "grad_norm": 0.7039394974708557,
      "learning_rate": 0.00025168193082641954,
      "loss": 3.3314,
      "step": 54000
    },
    {
      "epoch": 5.812076202776881,
      "eval_accuracy": 0.3837249491260016,
      "eval_loss": 3.3817226886749268,
      "eval_runtime": 189.7408,
      "eval_samples_per_second": 94.924,
      "eval_steps_per_second": 5.934,
      "step": 54000
    },
    {
      "epoch": 5.817457754816489,
      "grad_norm": 0.6475569605827332,
      "learning_rate": 0.00025135868979635814,
      "loss": 3.3249,
      "step": 54050
    },
    {
      "epoch": 5.822839306856097,
      "grad_norm": 0.6935245394706726,
      "learning_rate": 0.0002510354487662967,
      "loss": 3.3218,
      "step": 54100
    },
    {
      "epoch": 5.828220858895706,
      "grad_norm": 0.7183129787445068,
      "learning_rate": 0.0002507122077362353,
      "loss": 3.3463,
      "step": 54150
    },
    {
      "epoch": 5.833602410935313,
      "grad_norm": 0.7094807028770447,
      "learning_rate": 0.00025038896670617387,
      "loss": 3.3243,
      "step": 54200
    },
    {
      "epoch": 5.838983962974922,
      "grad_norm": 0.7039912939071655,
      "learning_rate": 0.00025006572567611246,
      "loss": 3.3373,
      "step": 54250
    },
    {
      "epoch": 5.84436551501453,
      "grad_norm": 0.6808781623840332,
      "learning_rate": 0.00024974248464605106,
      "loss": 3.3043,
      "step": 54300
    },
    {
      "epoch": 5.849747067054138,
      "grad_norm": 0.6898743510246277,
      "learning_rate": 0.00024941924361598965,
      "loss": 3.3292,
      "step": 54350
    },
    {
      "epoch": 5.855128619093747,
      "grad_norm": 0.70346599817276,
      "learning_rate": 0.00024909600258592825,
      "loss": 3.3399,
      "step": 54400
    },
    {
      "epoch": 5.860510171133355,
      "grad_norm": 0.656381368637085,
      "learning_rate": 0.0002487727615558668,
      "loss": 3.324,
      "step": 54450
    },
    {
      "epoch": 5.865891723172963,
      "grad_norm": 0.687276303768158,
      "learning_rate": 0.0002484495205258054,
      "loss": 3.3338,
      "step": 54500
    },
    {
      "epoch": 5.871273275212571,
      "grad_norm": 0.683623194694519,
      "learning_rate": 0.000248126279495744,
      "loss": 3.3229,
      "step": 54550
    },
    {
      "epoch": 5.87665482725218,
      "grad_norm": 0.7214177846908569,
      "learning_rate": 0.00024780303846568257,
      "loss": 3.3395,
      "step": 54600
    },
    {
      "epoch": 5.882036379291788,
      "grad_norm": 0.6841638088226318,
      "learning_rate": 0.0002474797974356211,
      "loss": 3.3189,
      "step": 54650
    },
    {
      "epoch": 5.887417931331396,
      "grad_norm": 0.7175382375717163,
      "learning_rate": 0.00024715655640555976,
      "loss": 3.3381,
      "step": 54700
    },
    {
      "epoch": 5.892799483371004,
      "grad_norm": 0.7035430073738098,
      "learning_rate": 0.0002468333153754983,
      "loss": 3.3301,
      "step": 54750
    },
    {
      "epoch": 5.898181035410612,
      "grad_norm": 0.6584702730178833,
      "learning_rate": 0.0002465100743454369,
      "loss": 3.3266,
      "step": 54800
    },
    {
      "epoch": 5.903562587450221,
      "grad_norm": 0.706007719039917,
      "learning_rate": 0.0002461868333153755,
      "loss": 3.3307,
      "step": 54850
    },
    {
      "epoch": 5.9089441394898286,
      "grad_norm": 0.6674470901489258,
      "learning_rate": 0.0002458635922853141,
      "loss": 3.3279,
      "step": 54900
    },
    {
      "epoch": 5.914325691529437,
      "grad_norm": 0.6810092329978943,
      "learning_rate": 0.0002455403512552526,
      "loss": 3.3187,
      "step": 54950
    },
    {
      "epoch": 5.919707243569046,
      "grad_norm": 0.658345639705658,
      "learning_rate": 0.0002452171102251912,
      "loss": 3.331,
      "step": 55000
    },
    {
      "epoch": 5.919707243569046,
      "eval_accuracy": 0.38436697904009737,
      "eval_loss": 3.378103017807007,
      "eval_runtime": 189.945,
      "eval_samples_per_second": 94.822,
      "eval_steps_per_second": 5.928,
      "step": 55000
    },
    {
      "epoch": 5.925088795608653,
      "grad_norm": 0.6954408884048462,
      "learning_rate": 0.000244900334015731,
      "loss": 3.3296,
      "step": 55050
    },
    {
      "epoch": 5.930470347648262,
      "grad_norm": 0.7177773118019104,
      "learning_rate": 0.0002445770929856696,
      "loss": 3.3259,
      "step": 55100
    },
    {
      "epoch": 5.93585189968787,
      "grad_norm": 0.6974418759346008,
      "learning_rate": 0.0002442538519556082,
      "loss": 3.326,
      "step": 55150
    },
    {
      "epoch": 5.941233451727478,
      "grad_norm": 0.690752387046814,
      "learning_rate": 0.0002439306109255468,
      "loss": 3.3342,
      "step": 55200
    },
    {
      "epoch": 5.946615003767087,
      "grad_norm": 0.7155065536499023,
      "learning_rate": 0.00024360736989548538,
      "loss": 3.3203,
      "step": 55250
    },
    {
      "epoch": 5.951996555806694,
      "grad_norm": 0.7307837605476379,
      "learning_rate": 0.00024328412886542394,
      "loss": 3.3337,
      "step": 55300
    },
    {
      "epoch": 5.957378107846303,
      "grad_norm": 0.6956316232681274,
      "learning_rate": 0.00024296088783536257,
      "loss": 3.3386,
      "step": 55350
    },
    {
      "epoch": 5.962759659885911,
      "grad_norm": 0.7020690441131592,
      "learning_rate": 0.00024263764680530113,
      "loss": 3.315,
      "step": 55400
    },
    {
      "epoch": 5.968141211925519,
      "grad_norm": 0.7208767533302307,
      "learning_rate": 0.0002423144057752397,
      "loss": 3.3312,
      "step": 55450
    },
    {
      "epoch": 5.973522763965128,
      "grad_norm": 0.6836248636245728,
      "learning_rate": 0.00024199116474517832,
      "loss": 3.3435,
      "step": 55500
    },
    {
      "epoch": 5.978904316004736,
      "grad_norm": 0.7073252201080322,
      "learning_rate": 0.0002416679237151169,
      "loss": 3.3207,
      "step": 55550
    },
    {
      "epoch": 5.984285868044344,
      "grad_norm": 0.7180842757225037,
      "learning_rate": 0.00024134468268505546,
      "loss": 3.3244,
      "step": 55600
    },
    {
      "epoch": 5.989667420083952,
      "grad_norm": 0.7365034818649292,
      "learning_rate": 0.00024102144165499405,
      "loss": 3.3348,
      "step": 55650
    },
    {
      "epoch": 5.995048972123561,
      "grad_norm": 0.7047129273414612,
      "learning_rate": 0.00024069820062493265,
      "loss": 3.3524,
      "step": 55700
    },
    {
      "epoch": 6.000430524163169,
      "grad_norm": 0.6616429090499878,
      "learning_rate": 0.00024037495959487121,
      "loss": 3.3277,
      "step": 55750
    },
    {
      "epoch": 6.005812076202777,
      "grad_norm": 0.7128415703773499,
      "learning_rate": 0.0002400517185648098,
      "loss": 3.227,
      "step": 55800
    },
    {
      "epoch": 6.011193628242385,
      "grad_norm": 0.6549385190010071,
      "learning_rate": 0.00023972847753474838,
      "loss": 3.2359,
      "step": 55850
    },
    {
      "epoch": 6.016575180281993,
      "grad_norm": 0.7191146612167358,
      "learning_rate": 0.00023940523650468697,
      "loss": 3.2544,
      "step": 55900
    },
    {
      "epoch": 6.021956732321602,
      "grad_norm": 0.6856921911239624,
      "learning_rate": 0.00023908199547462557,
      "loss": 3.2425,
      "step": 55950
    },
    {
      "epoch": 6.0273382843612096,
      "grad_norm": 0.6903359293937683,
      "learning_rate": 0.00023875875444456413,
      "loss": 3.2461,
      "step": 56000
    },
    {
      "epoch": 6.0273382843612096,
      "eval_accuracy": 0.38410501692757665,
      "eval_loss": 3.3839967250823975,
      "eval_runtime": 189.7407,
      "eval_samples_per_second": 94.924,
      "eval_steps_per_second": 5.934,
      "step": 56000
    },
    {
      "epoch": 6.032719836400818,
      "grad_norm": 0.6743659973144531,
      "learning_rate": 0.0002384355134145027,
      "loss": 3.2299,
      "step": 56050
    },
    {
      "epoch": 6.038101388440427,
      "grad_norm": 0.6880258321762085,
      "learning_rate": 0.00023811227238444132,
      "loss": 3.2516,
      "step": 56100
    },
    {
      "epoch": 6.043482940480034,
      "grad_norm": 0.7418484091758728,
      "learning_rate": 0.0002377890313543799,
      "loss": 3.2601,
      "step": 56150
    },
    {
      "epoch": 6.048864492519643,
      "grad_norm": 0.6839198470115662,
      "learning_rate": 0.00023746579032431849,
      "loss": 3.2285,
      "step": 56200
    },
    {
      "epoch": 6.0542460445592505,
      "grad_norm": 0.6740571856498718,
      "learning_rate": 0.00023714254929425708,
      "loss": 3.2518,
      "step": 56250
    },
    {
      "epoch": 6.059627596598859,
      "grad_norm": 0.6967892050743103,
      "learning_rate": 0.00023681930826419565,
      "loss": 3.261,
      "step": 56300
    },
    {
      "epoch": 6.065009148638468,
      "grad_norm": 0.7453837394714355,
      "learning_rate": 0.00023649606723413424,
      "loss": 3.2363,
      "step": 56350
    },
    {
      "epoch": 6.070390700678075,
      "grad_norm": 0.6671978831291199,
      "learning_rate": 0.0002361728262040728,
      "loss": 3.249,
      "step": 56400
    },
    {
      "epoch": 6.075772252717684,
      "grad_norm": 0.6880906224250793,
      "learning_rate": 0.00023584958517401138,
      "loss": 3.2615,
      "step": 56450
    },
    {
      "epoch": 6.081153804757292,
      "grad_norm": 0.7420952916145325,
      "learning_rate": 0.00023552634414395,
      "loss": 3.2402,
      "step": 56500
    },
    {
      "epoch": 6.0865353567969,
      "grad_norm": 0.7020220160484314,
      "learning_rate": 0.00023520310311388857,
      "loss": 3.259,
      "step": 56550
    },
    {
      "epoch": 6.091916908836509,
      "grad_norm": 0.6925346851348877,
      "learning_rate": 0.00023487986208382713,
      "loss": 3.2814,
      "step": 56600
    },
    {
      "epoch": 6.097298460876116,
      "grad_norm": 0.6871832013130188,
      "learning_rate": 0.00023455662105376576,
      "loss": 3.2501,
      "step": 56650
    },
    {
      "epoch": 6.102680012915725,
      "grad_norm": 0.7279773354530334,
      "learning_rate": 0.00023423338002370432,
      "loss": 3.2528,
      "step": 56700
    },
    {
      "epoch": 6.108061564955333,
      "grad_norm": 0.6858897805213928,
      "learning_rate": 0.0002339101389936429,
      "loss": 3.2177,
      "step": 56750
    },
    {
      "epoch": 6.113443116994941,
      "grad_norm": 0.698556125164032,
      "learning_rate": 0.00023358689796358149,
      "loss": 3.2313,
      "step": 56800
    },
    {
      "epoch": 6.11882466903455,
      "grad_norm": 0.7234476804733276,
      "learning_rate": 0.00023326365693352008,
      "loss": 3.2706,
      "step": 56850
    },
    {
      "epoch": 6.124206221074158,
      "grad_norm": 0.681518018245697,
      "learning_rate": 0.00023294041590345865,
      "loss": 3.2725,
      "step": 56900
    },
    {
      "epoch": 6.129587773113766,
      "grad_norm": 0.7498565912246704,
      "learning_rate": 0.00023261717487339724,
      "loss": 3.2482,
      "step": 56950
    },
    {
      "epoch": 6.134969325153374,
      "grad_norm": 0.7866875529289246,
      "learning_rate": 0.0002322939338433358,
      "loss": 3.2544,
      "step": 57000
    },
    {
      "epoch": 6.134969325153374,
      "eval_accuracy": 0.3843896874936506,
      "eval_loss": 3.3818647861480713,
      "eval_runtime": 190.1579,
      "eval_samples_per_second": 94.716,
      "eval_steps_per_second": 5.921,
      "step": 57000
    },
    {
      "epoch": 6.140350877192983,
      "grad_norm": 0.6911661028862,
      "learning_rate": 0.0002319706928132744,
      "loss": 3.2537,
      "step": 57050
    },
    {
      "epoch": 6.1457324292325906,
      "grad_norm": 0.6852210164070129,
      "learning_rate": 0.0002316539166038142,
      "loss": 3.2504,
      "step": 57100
    },
    {
      "epoch": 6.151113981272199,
      "grad_norm": 0.6963878870010376,
      "learning_rate": 0.00023133067557375283,
      "loss": 3.2559,
      "step": 57150
    },
    {
      "epoch": 6.156495533311807,
      "grad_norm": 0.7116459608078003,
      "learning_rate": 0.0002310074345436914,
      "loss": 3.2567,
      "step": 57200
    },
    {
      "epoch": 6.161877085351415,
      "grad_norm": 0.7186686992645264,
      "learning_rate": 0.00023068419351362997,
      "loss": 3.2549,
      "step": 57250
    },
    {
      "epoch": 6.167258637391024,
      "grad_norm": 0.7681869268417358,
      "learning_rate": 0.00023036095248356856,
      "loss": 3.2553,
      "step": 57300
    },
    {
      "epoch": 6.1726401894306315,
      "grad_norm": 0.6896945238113403,
      "learning_rate": 0.00023003771145350716,
      "loss": 3.247,
      "step": 57350
    },
    {
      "epoch": 6.17802174147024,
      "grad_norm": 0.7051165699958801,
      "learning_rate": 0.00022971447042344572,
      "loss": 3.2718,
      "step": 57400
    },
    {
      "epoch": 6.183403293509849,
      "grad_norm": 0.7039802670478821,
      "learning_rate": 0.00022939122939338432,
      "loss": 3.2557,
      "step": 57450
    },
    {
      "epoch": 6.188784845549456,
      "grad_norm": 0.7361300587654114,
      "learning_rate": 0.0002290679883633229,
      "loss": 3.2514,
      "step": 57500
    },
    {
      "epoch": 6.194166397589065,
      "grad_norm": 0.7037554383277893,
      "learning_rate": 0.00022874474733326148,
      "loss": 3.2732,
      "step": 57550
    },
    {
      "epoch": 6.1995479496286725,
      "grad_norm": 0.7038314342498779,
      "learning_rate": 0.00022842150630320008,
      "loss": 3.2462,
      "step": 57600
    },
    {
      "epoch": 6.204929501668281,
      "grad_norm": 0.7034255862236023,
      "learning_rate": 0.00022809826527313864,
      "loss": 3.2901,
      "step": 57650
    },
    {
      "epoch": 6.21031105370789,
      "grad_norm": 0.6886646151542664,
      "learning_rate": 0.0002277750242430772,
      "loss": 3.2819,
      "step": 57700
    },
    {
      "epoch": 6.215692605747497,
      "grad_norm": 0.7283786535263062,
      "learning_rate": 0.00022745178321301583,
      "loss": 3.257,
      "step": 57750
    },
    {
      "epoch": 6.221074157787106,
      "grad_norm": 0.7100892066955566,
      "learning_rate": 0.0002271285421829544,
      "loss": 3.2512,
      "step": 57800
    },
    {
      "epoch": 6.226455709826714,
      "grad_norm": 0.727598249912262,
      "learning_rate": 0.00022680530115289297,
      "loss": 3.2715,
      "step": 57850
    },
    {
      "epoch": 6.231837261866322,
      "grad_norm": 0.6963880658149719,
      "learning_rate": 0.0002264820601228316,
      "loss": 3.2534,
      "step": 57900
    },
    {
      "epoch": 6.237218813905931,
      "grad_norm": 0.721747636795044,
      "learning_rate": 0.00022615881909277016,
      "loss": 3.2725,
      "step": 57950
    },
    {
      "epoch": 6.242600365945538,
      "grad_norm": 0.6808701157569885,
      "learning_rate": 0.00022583557806270875,
      "loss": 3.2673,
      "step": 58000
    },
    {
      "epoch": 6.242600365945538,
      "eval_accuracy": 0.3846008000546741,
      "eval_loss": 3.3768603801727295,
      "eval_runtime": 190.4815,
      "eval_samples_per_second": 94.555,
      "eval_steps_per_second": 5.911,
      "step": 58000
    },
    {
      "epoch": 6.247981917985147,
      "grad_norm": 1.011569857597351,
      "learning_rate": 0.00022551233703264732,
      "loss": 3.2694,
      "step": 58050
    },
    {
      "epoch": 6.253363470024755,
      "grad_norm": 0.7165616750717163,
      "learning_rate": 0.0002251890960025859,
      "loss": 3.2799,
      "step": 58100
    },
    {
      "epoch": 6.258745022064363,
      "grad_norm": 0.7355701327323914,
      "learning_rate": 0.0002248658549725245,
      "loss": 3.2697,
      "step": 58150
    },
    {
      "epoch": 6.264126574103972,
      "grad_norm": 0.6905155777931213,
      "learning_rate": 0.00022454261394246308,
      "loss": 3.2721,
      "step": 58200
    },
    {
      "epoch": 6.26950812614358,
      "grad_norm": 0.7252981662750244,
      "learning_rate": 0.00022421937291240164,
      "loss": 3.2813,
      "step": 58250
    },
    {
      "epoch": 6.274889678183188,
      "grad_norm": 0.7116219997406006,
      "learning_rate": 0.00022389613188234027,
      "loss": 3.2782,
      "step": 58300
    },
    {
      "epoch": 6.280271230222796,
      "grad_norm": 0.7018437385559082,
      "learning_rate": 0.00022357289085227883,
      "loss": 3.2626,
      "step": 58350
    },
    {
      "epoch": 6.285652782262405,
      "grad_norm": 0.7009330987930298,
      "learning_rate": 0.0002232496498222174,
      "loss": 3.2683,
      "step": 58400
    },
    {
      "epoch": 6.2910343343020125,
      "grad_norm": 0.7227058410644531,
      "learning_rate": 0.000222926408792156,
      "loss": 3.2753,
      "step": 58450
    },
    {
      "epoch": 6.296415886341621,
      "grad_norm": 0.7396367192268372,
      "learning_rate": 0.0002226031677620946,
      "loss": 3.2705,
      "step": 58500
    },
    {
      "epoch": 6.301797438381229,
      "grad_norm": 0.7184813618659973,
      "learning_rate": 0.00022227992673203316,
      "loss": 3.2724,
      "step": 58550
    },
    {
      "epoch": 6.307178990420837,
      "grad_norm": 0.7396824955940247,
      "learning_rate": 0.00022195668570197175,
      "loss": 3.2649,
      "step": 58600
    },
    {
      "epoch": 6.312560542460446,
      "grad_norm": 0.7457385659217834,
      "learning_rate": 0.00022163344467191032,
      "loss": 3.2821,
      "step": 58650
    },
    {
      "epoch": 6.3179420945000535,
      "grad_norm": 0.7198763489723206,
      "learning_rate": 0.00022131020364184891,
      "loss": 3.2687,
      "step": 58700
    },
    {
      "epoch": 6.323323646539662,
      "grad_norm": 0.7418332695960999,
      "learning_rate": 0.0002209869626117875,
      "loss": 3.2777,
      "step": 58750
    },
    {
      "epoch": 6.328705198579271,
      "grad_norm": 0.7484312057495117,
      "learning_rate": 0.00022066372158172608,
      "loss": 3.2756,
      "step": 58800
    },
    {
      "epoch": 6.334086750618878,
      "grad_norm": 0.7683062553405762,
      "learning_rate": 0.00022034048055166464,
      "loss": 3.2654,
      "step": 58850
    },
    {
      "epoch": 6.339468302658487,
      "grad_norm": 0.6564449071884155,
      "learning_rate": 0.00022001723952160327,
      "loss": 3.2431,
      "step": 58900
    },
    {
      "epoch": 6.344849854698095,
      "grad_norm": 0.742984414100647,
      "learning_rate": 0.00021969399849154183,
      "loss": 3.2791,
      "step": 58950
    },
    {
      "epoch": 6.350231406737703,
      "grad_norm": 0.7099165916442871,
      "learning_rate": 0.00021937075746148043,
      "loss": 3.2566,
      "step": 59000
    },
    {
      "epoch": 6.350231406737703,
      "eval_accuracy": 0.3853096514947758,
      "eval_loss": 3.3746886253356934,
      "eval_runtime": 189.6105,
      "eval_samples_per_second": 94.989,
      "eval_steps_per_second": 5.938,
      "step": 59000
    },
    {
      "epoch": 6.355612958777312,
      "grad_norm": 0.7248750925064087,
      "learning_rate": 0.00021904751643141902,
      "loss": 3.2569,
      "step": 59050
    },
    {
      "epoch": 6.360994510816919,
      "grad_norm": 0.7620924711227417,
      "learning_rate": 0.00021873074022195883,
      "loss": 3.2713,
      "step": 59100
    },
    {
      "epoch": 6.366376062856528,
      "grad_norm": 0.6913202404975891,
      "learning_rate": 0.0002184074991918974,
      "loss": 3.2727,
      "step": 59150
    },
    {
      "epoch": 6.371757614896136,
      "grad_norm": 0.7224660515785217,
      "learning_rate": 0.00021808425816183597,
      "loss": 3.2592,
      "step": 59200
    },
    {
      "epoch": 6.377139166935744,
      "grad_norm": 0.7816100716590881,
      "learning_rate": 0.0002177610171317746,
      "loss": 3.2917,
      "step": 59250
    },
    {
      "epoch": 6.382520718975353,
      "grad_norm": 0.7203670144081116,
      "learning_rate": 0.00021743777610171315,
      "loss": 3.2677,
      "step": 59300
    },
    {
      "epoch": 6.387902271014961,
      "grad_norm": 0.7703629732131958,
      "learning_rate": 0.00021711453507165172,
      "loss": 3.2676,
      "step": 59350
    },
    {
      "epoch": 6.393283823054569,
      "grad_norm": 0.7150661945343018,
      "learning_rate": 0.00021679129404159034,
      "loss": 3.283,
      "step": 59400
    },
    {
      "epoch": 6.398665375094177,
      "grad_norm": 0.6808066964149475,
      "learning_rate": 0.0002164680530115289,
      "loss": 3.2853,
      "step": 59450
    },
    {
      "epoch": 6.404046927133785,
      "grad_norm": 0.7157021760940552,
      "learning_rate": 0.00021614481198146748,
      "loss": 3.2858,
      "step": 59500
    },
    {
      "epoch": 6.4094284791733935,
      "grad_norm": 0.7176766395568848,
      "learning_rate": 0.00021582157095140607,
      "loss": 3.2623,
      "step": 59550
    },
    {
      "epoch": 6.414810031213002,
      "grad_norm": 0.7077521681785583,
      "learning_rate": 0.00021549832992134467,
      "loss": 3.2907,
      "step": 59600
    },
    {
      "epoch": 6.42019158325261,
      "grad_norm": 0.7326254844665527,
      "learning_rate": 0.00021517508889128324,
      "loss": 3.283,
      "step": 59650
    },
    {
      "epoch": 6.425573135292218,
      "grad_norm": 0.7177848219871521,
      "learning_rate": 0.00021485184786122183,
      "loss": 3.2828,
      "step": 59700
    },
    {
      "epoch": 6.430954687331827,
      "grad_norm": 0.7121162414550781,
      "learning_rate": 0.0002145286068311604,
      "loss": 3.2885,
      "step": 59750
    },
    {
      "epoch": 6.4363362393714345,
      "grad_norm": 0.7246776819229126,
      "learning_rate": 0.00021420536580109902,
      "loss": 3.2783,
      "step": 59800
    },
    {
      "epoch": 6.441717791411043,
      "grad_norm": 0.7581564784049988,
      "learning_rate": 0.0002138821247710376,
      "loss": 3.2796,
      "step": 59850
    },
    {
      "epoch": 6.447099343450651,
      "grad_norm": 0.7635112404823303,
      "learning_rate": 0.00021355888374097615,
      "loss": 3.2582,
      "step": 59900
    },
    {
      "epoch": 6.452480895490259,
      "grad_norm": 0.7147925496101379,
      "learning_rate": 0.00021323564271091478,
      "loss": 3.2755,
      "step": 59950
    },
    {
      "epoch": 6.457862447529868,
      "grad_norm": 0.7165830731391907,
      "learning_rate": 0.00021291240168085334,
      "loss": 3.2666,
      "step": 60000
    },
    {
      "epoch": 6.457862447529868,
      "eval_accuracy": 0.3857059085766352,
      "eval_loss": 3.3704819679260254,
      "eval_runtime": 190.1181,
      "eval_samples_per_second": 94.736,
      "eval_steps_per_second": 5.923,
      "step": 60000
    },
    {
      "epoch": 6.4632439995694755,
      "grad_norm": 0.7389584183692932,
      "learning_rate": 0.0002125891606507919,
      "loss": 3.277,
      "step": 60050
    },
    {
      "epoch": 6.468625551609084,
      "grad_norm": 0.7635201811790466,
      "learning_rate": 0.0002122659196207305,
      "loss": 3.2901,
      "step": 60100
    },
    {
      "epoch": 6.474007103648693,
      "grad_norm": 0.6967383027076721,
      "learning_rate": 0.0002119426785906691,
      "loss": 3.2836,
      "step": 60150
    },
    {
      "epoch": 6.4793886556883,
      "grad_norm": 0.7632560133934021,
      "learning_rate": 0.00021161943756060767,
      "loss": 3.2966,
      "step": 60200
    },
    {
      "epoch": 6.484770207727909,
      "grad_norm": 0.7013333439826965,
      "learning_rate": 0.00021129619653054626,
      "loss": 3.2908,
      "step": 60250
    },
    {
      "epoch": 6.490151759767517,
      "grad_norm": 0.708081066608429,
      "learning_rate": 0.00021097942032108607,
      "loss": 3.2837,
      "step": 60300
    },
    {
      "epoch": 6.495533311807125,
      "grad_norm": 0.7899771928787231,
      "learning_rate": 0.00021065617929102466,
      "loss": 3.285,
      "step": 60350
    },
    {
      "epoch": 6.500914863846734,
      "grad_norm": 0.7200154662132263,
      "learning_rate": 0.00021033293826096323,
      "loss": 3.2748,
      "step": 60400
    },
    {
      "epoch": 6.506296415886341,
      "grad_norm": 0.7418370246887207,
      "learning_rate": 0.0002100096972309018,
      "loss": 3.2702,
      "step": 60450
    },
    {
      "epoch": 6.51167796792595,
      "grad_norm": 0.7911790609359741,
      "learning_rate": 0.00020968645620084042,
      "loss": 3.2903,
      "step": 60500
    },
    {
      "epoch": 6.517059519965558,
      "grad_norm": 0.6987434029579163,
      "learning_rate": 0.000209363215170779,
      "loss": 3.2778,
      "step": 60550
    },
    {
      "epoch": 6.522441072005166,
      "grad_norm": 0.7445406317710876,
      "learning_rate": 0.00020903997414071758,
      "loss": 3.279,
      "step": 60600
    },
    {
      "epoch": 6.5278226240447745,
      "grad_norm": 0.7416291236877441,
      "learning_rate": 0.00020871673311065618,
      "loss": 3.2686,
      "step": 60650
    },
    {
      "epoch": 6.533204176084383,
      "grad_norm": 0.8053084015846252,
      "learning_rate": 0.00020839349208059475,
      "loss": 3.2793,
      "step": 60700
    },
    {
      "epoch": 6.538585728123991,
      "grad_norm": 0.7757678627967834,
      "learning_rate": 0.00020807025105053334,
      "loss": 3.2816,
      "step": 60750
    },
    {
      "epoch": 6.543967280163599,
      "grad_norm": 0.7882310152053833,
      "learning_rate": 0.0002077470100204719,
      "loss": 3.2937,
      "step": 60800
    },
    {
      "epoch": 6.549348832203208,
      "grad_norm": 0.7543111443519592,
      "learning_rate": 0.00020742376899041048,
      "loss": 3.2731,
      "step": 60850
    },
    {
      "epoch": 6.5547303842428155,
      "grad_norm": 0.7289414405822754,
      "learning_rate": 0.0002071005279603491,
      "loss": 3.2782,
      "step": 60900
    },
    {
      "epoch": 6.560111936282424,
      "grad_norm": 0.7349234223365784,
      "learning_rate": 0.00020677728693028766,
      "loss": 3.2807,
      "step": 60950
    },
    {
      "epoch": 6.565493488322032,
      "grad_norm": 0.7486505508422852,
      "learning_rate": 0.00020645404590022623,
      "loss": 3.2843,
      "step": 61000
    },
    {
      "epoch": 6.565493488322032,
      "eval_accuracy": 0.38620364745571334,
      "eval_loss": 3.366868734359741,
      "eval_runtime": 190.182,
      "eval_samples_per_second": 94.704,
      "eval_steps_per_second": 5.921,
      "step": 61000
    },
    {
      "epoch": 6.57087504036164,
      "grad_norm": 0.7355003952980042,
      "learning_rate": 0.00020613080487016485,
      "loss": 3.279,
      "step": 61050
    },
    {
      "epoch": 6.576256592401249,
      "grad_norm": 0.7191846966743469,
      "learning_rate": 0.00020580756384010342,
      "loss": 3.2847,
      "step": 61100
    },
    {
      "epoch": 6.5816381444408565,
      "grad_norm": 0.8277631998062134,
      "learning_rate": 0.000205484322810042,
      "loss": 3.2725,
      "step": 61150
    },
    {
      "epoch": 6.587019696480465,
      "grad_norm": 0.7376859188079834,
      "learning_rate": 0.00020516108177998058,
      "loss": 3.2544,
      "step": 61200
    },
    {
      "epoch": 6.592401248520073,
      "grad_norm": 0.7258176207542419,
      "learning_rate": 0.00020483784074991918,
      "loss": 3.2824,
      "step": 61250
    },
    {
      "epoch": 6.597782800559681,
      "grad_norm": 0.7444310188293457,
      "learning_rate": 0.00020451459971985775,
      "loss": 3.2932,
      "step": 61300
    },
    {
      "epoch": 6.60316435259929,
      "grad_norm": 0.7077869176864624,
      "learning_rate": 0.00020419135868979634,
      "loss": 3.2867,
      "step": 61350
    },
    {
      "epoch": 6.608545904638898,
      "grad_norm": 0.7489200830459595,
      "learning_rate": 0.0002038681176597349,
      "loss": 3.2814,
      "step": 61400
    },
    {
      "epoch": 6.613927456678506,
      "grad_norm": 0.7412776350975037,
      "learning_rate": 0.0002035448766296735,
      "loss": 3.283,
      "step": 61450
    },
    {
      "epoch": 6.619309008718115,
      "grad_norm": 0.7478265762329102,
      "learning_rate": 0.0002032216355996121,
      "loss": 3.2791,
      "step": 61500
    },
    {
      "epoch": 6.624690560757722,
      "grad_norm": 0.7470537424087524,
      "learning_rate": 0.00020289839456955067,
      "loss": 3.2644,
      "step": 61550
    },
    {
      "epoch": 6.630072112797331,
      "grad_norm": 0.7429487705230713,
      "learning_rate": 0.0002025751535394893,
      "loss": 3.2831,
      "step": 61600
    },
    {
      "epoch": 6.635453664836939,
      "grad_norm": 0.723760724067688,
      "learning_rate": 0.00020225191250942785,
      "loss": 3.2816,
      "step": 61650
    },
    {
      "epoch": 6.640835216876547,
      "grad_norm": 0.7412201762199402,
      "learning_rate": 0.00020192867147936642,
      "loss": 3.277,
      "step": 61700
    },
    {
      "epoch": 6.6462167689161555,
      "grad_norm": 0.7405781745910645,
      "learning_rate": 0.00020160543044930502,
      "loss": 3.2834,
      "step": 61750
    },
    {
      "epoch": 6.651598320955763,
      "grad_norm": 0.7630212306976318,
      "learning_rate": 0.0002012821894192436,
      "loss": 3.279,
      "step": 61800
    },
    {
      "epoch": 6.656979872995372,
      "grad_norm": 0.7422797083854675,
      "learning_rate": 0.00020095894838918218,
      "loss": 3.2746,
      "step": 61850
    },
    {
      "epoch": 6.66236142503498,
      "grad_norm": 0.7069374918937683,
      "learning_rate": 0.00020063570735912077,
      "loss": 3.2714,
      "step": 61900
    },
    {
      "epoch": 6.667742977074588,
      "grad_norm": 0.7747578620910645,
      "learning_rate": 0.00020031246632905934,
      "loss": 3.2721,
      "step": 61950
    },
    {
      "epoch": 6.6731245291141965,
      "grad_norm": 0.7529718279838562,
      "learning_rate": 0.00019998922529899794,
      "loss": 3.2763,
      "step": 62000
    },
    {
      "epoch": 6.6731245291141965,
      "eval_accuracy": 0.3865844758275031,
      "eval_loss": 3.363325834274292,
      "eval_runtime": 190.0168,
      "eval_samples_per_second": 94.786,
      "eval_steps_per_second": 5.926,
      "step": 62000
    },
    {
      "epoch": 6.678506081153805,
      "grad_norm": 0.7516931891441345,
      "learning_rate": 0.00019966598426893653,
      "loss": 3.2742,
      "step": 62050
    },
    {
      "epoch": 6.683887633193413,
      "grad_norm": 0.7729787230491638,
      "learning_rate": 0.0001993427432388751,
      "loss": 3.2672,
      "step": 62100
    },
    {
      "epoch": 6.689269185233021,
      "grad_norm": 0.7330310344696045,
      "learning_rate": 0.00019901950220881367,
      "loss": 3.2669,
      "step": 62150
    },
    {
      "epoch": 6.69465073727263,
      "grad_norm": 0.7198368906974792,
      "learning_rate": 0.0001986962611787523,
      "loss": 3.2907,
      "step": 62200
    },
    {
      "epoch": 6.7000322893122375,
      "grad_norm": 0.7286278605461121,
      "learning_rate": 0.00019837302014869085,
      "loss": 3.2855,
      "step": 62250
    },
    {
      "epoch": 6.705413841351846,
      "grad_norm": 0.7030113935470581,
      "learning_rate": 0.00019804977911862942,
      "loss": 3.2621,
      "step": 62300
    },
    {
      "epoch": 6.710795393391454,
      "grad_norm": 0.7173521518707275,
      "learning_rate": 0.00019772653808856804,
      "loss": 3.2769,
      "step": 62350
    },
    {
      "epoch": 6.716176945431062,
      "grad_norm": 0.8167335391044617,
      "learning_rate": 0.0001974032970585066,
      "loss": 3.284,
      "step": 62400
    },
    {
      "epoch": 6.721558497470671,
      "grad_norm": 0.7530385255813599,
      "learning_rate": 0.00019708005602844518,
      "loss": 3.2977,
      "step": 62450
    },
    {
      "epoch": 6.7269400495102785,
      "grad_norm": 0.8321440815925598,
      "learning_rate": 0.00019675681499838377,
      "loss": 3.2964,
      "step": 62500
    },
    {
      "epoch": 6.732321601549887,
      "grad_norm": 0.7328689694404602,
      "learning_rate": 0.00019643357396832237,
      "loss": 3.298,
      "step": 62550
    },
    {
      "epoch": 6.737703153589496,
      "grad_norm": 0.7410628795623779,
      "learning_rate": 0.00019611033293826096,
      "loss": 3.2931,
      "step": 62600
    },
    {
      "epoch": 6.743084705629103,
      "grad_norm": 0.7810050249099731,
      "learning_rate": 0.00019578709190819953,
      "loss": 3.2952,
      "step": 62650
    },
    {
      "epoch": 6.748466257668712,
      "grad_norm": 0.7138219475746155,
      "learning_rate": 0.0001954638508781381,
      "loss": 3.2866,
      "step": 62700
    },
    {
      "epoch": 6.75384780970832,
      "grad_norm": 0.7371669411659241,
      "learning_rate": 0.00019514060984807672,
      "loss": 3.282,
      "step": 62750
    },
    {
      "epoch": 6.759229361747928,
      "grad_norm": 0.7349672913551331,
      "learning_rate": 0.0001948173688180153,
      "loss": 3.289,
      "step": 62800
    },
    {
      "epoch": 6.7646109137875365,
      "grad_norm": 0.7345631718635559,
      "learning_rate": 0.00019449412778795386,
      "loss": 3.2832,
      "step": 62850
    },
    {
      "epoch": 6.769992465827144,
      "grad_norm": 0.793651819229126,
      "learning_rate": 0.00019417088675789245,
      "loss": 3.2866,
      "step": 62900
    },
    {
      "epoch": 6.775374017866753,
      "grad_norm": 0.7369472980499268,
      "learning_rate": 0.00019384764572783104,
      "loss": 3.2853,
      "step": 62950
    },
    {
      "epoch": 6.780755569906361,
      "grad_norm": 0.7100161910057068,
      "learning_rate": 0.0001935244046977696,
      "loss": 3.2883,
      "step": 63000
    },
    {
      "epoch": 6.780755569906361,
      "eval_accuracy": 0.38676875112523645,
      "eval_loss": 3.3582944869995117,
      "eval_runtime": 190.297,
      "eval_samples_per_second": 94.647,
      "eval_steps_per_second": 5.917,
      "step": 63000
    },
    {
      "epoch": 6.786137121945969,
      "grad_norm": 0.894089937210083,
      "learning_rate": 0.0001932011636677082,
      "loss": 3.2861,
      "step": 63050
    },
    {
      "epoch": 6.7915186739855775,
      "grad_norm": 0.7339184284210205,
      "learning_rate": 0.00019287792263764677,
      "loss": 3.2867,
      "step": 63100
    },
    {
      "epoch": 6.796900226025185,
      "grad_norm": 0.7779406905174255,
      "learning_rate": 0.00019255468160758537,
      "loss": 3.2731,
      "step": 63150
    },
    {
      "epoch": 6.802281778064794,
      "grad_norm": 0.7760416269302368,
      "learning_rate": 0.00019223144057752396,
      "loss": 3.2787,
      "step": 63200
    },
    {
      "epoch": 6.807663330104402,
      "grad_norm": 0.795799732208252,
      "learning_rate": 0.00019191466436806374,
      "loss": 3.2885,
      "step": 63250
    },
    {
      "epoch": 6.813044882144011,
      "grad_norm": 0.7552291750907898,
      "learning_rate": 0.00019159142333800236,
      "loss": 3.28,
      "step": 63300
    },
    {
      "epoch": 6.8184264341836185,
      "grad_norm": 0.7520554065704346,
      "learning_rate": 0.00019126818230794093,
      "loss": 3.2941,
      "step": 63350
    },
    {
      "epoch": 6.823807986223227,
      "grad_norm": 0.7361971139907837,
      "learning_rate": 0.00019094494127787953,
      "loss": 3.2775,
      "step": 63400
    },
    {
      "epoch": 6.829189538262835,
      "grad_norm": 0.6947896480560303,
      "learning_rate": 0.00019062170024781812,
      "loss": 3.2895,
      "step": 63450
    },
    {
      "epoch": 6.834571090302443,
      "grad_norm": 0.7636711001396179,
      "learning_rate": 0.0001902984592177567,
      "loss": 3.2916,
      "step": 63500
    },
    {
      "epoch": 6.839952642342052,
      "grad_norm": 0.7767258286476135,
      "learning_rate": 0.00018997521818769528,
      "loss": 3.2726,
      "step": 63550
    },
    {
      "epoch": 6.8453341943816595,
      "grad_norm": 0.7709169387817383,
      "learning_rate": 0.00018965197715763385,
      "loss": 3.293,
      "step": 63600
    },
    {
      "epoch": 6.850715746421268,
      "grad_norm": 0.7238304018974304,
      "learning_rate": 0.00018932873612757245,
      "loss": 3.2842,
      "step": 63650
    },
    {
      "epoch": 6.856097298460876,
      "grad_norm": 0.7447382211685181,
      "learning_rate": 0.00018900549509751104,
      "loss": 3.2932,
      "step": 63700
    },
    {
      "epoch": 6.861478850500484,
      "grad_norm": 0.7233273386955261,
      "learning_rate": 0.0001886822540674496,
      "loss": 3.2665,
      "step": 63750
    },
    {
      "epoch": 6.866860402540093,
      "grad_norm": 0.7737557291984558,
      "learning_rate": 0.00018835901303738818,
      "loss": 3.266,
      "step": 63800
    },
    {
      "epoch": 6.8722419545797,
      "grad_norm": 0.7991005182266235,
      "learning_rate": 0.0001880357720073268,
      "loss": 3.2729,
      "step": 63850
    },
    {
      "epoch": 6.877623506619309,
      "grad_norm": 0.7907280921936035,
      "learning_rate": 0.00018771253097726537,
      "loss": 3.2944,
      "step": 63900
    },
    {
      "epoch": 6.8830050586589175,
      "grad_norm": 0.7917770147323608,
      "learning_rate": 0.00018738928994720393,
      "loss": 3.2746,
      "step": 63950
    },
    {
      "epoch": 6.888386610698525,
      "grad_norm": 0.7574546337127686,
      "learning_rate": 0.00018706604891714255,
      "loss": 3.2877,
      "step": 64000
    },
    {
      "epoch": 6.888386610698525,
      "eval_accuracy": 0.38737101408239216,
      "eval_loss": 3.3532512187957764,
      "eval_runtime": 190.3175,
      "eval_samples_per_second": 94.637,
      "eval_steps_per_second": 5.916,
      "step": 64000
    },
    {
      "epoch": 6.893768162738134,
      "grad_norm": 0.7376413345336914,
      "learning_rate": 0.00018674280788708112,
      "loss": 3.2775,
      "step": 64050
    },
    {
      "epoch": 6.899149714777742,
      "grad_norm": 0.7031199336051941,
      "learning_rate": 0.0001864195668570197,
      "loss": 3.2613,
      "step": 64100
    },
    {
      "epoch": 6.90453126681735,
      "grad_norm": 0.7579556107521057,
      "learning_rate": 0.00018609632582695828,
      "loss": 3.294,
      "step": 64150
    },
    {
      "epoch": 6.9099128188569585,
      "grad_norm": 0.7226597666740417,
      "learning_rate": 0.00018577308479689685,
      "loss": 3.2921,
      "step": 64200
    },
    {
      "epoch": 6.915294370896566,
      "grad_norm": 0.7821003198623657,
      "learning_rate": 0.00018544984376683545,
      "loss": 3.2702,
      "step": 64250
    },
    {
      "epoch": 6.920675922936175,
      "grad_norm": 0.818265438079834,
      "learning_rate": 0.00018512660273677404,
      "loss": 3.2816,
      "step": 64300
    },
    {
      "epoch": 6.926057474975783,
      "grad_norm": 0.8839375972747803,
      "learning_rate": 0.0001848033617067126,
      "loss": 3.2821,
      "step": 64350
    },
    {
      "epoch": 6.931439027015391,
      "grad_norm": 0.7486461400985718,
      "learning_rate": 0.00018448012067665123,
      "loss": 3.2779,
      "step": 64400
    },
    {
      "epoch": 6.9368205790549995,
      "grad_norm": 0.7852596044540405,
      "learning_rate": 0.0001841568796465898,
      "loss": 3.2657,
      "step": 64450
    },
    {
      "epoch": 6.942202131094608,
      "grad_norm": 0.7576976418495178,
      "learning_rate": 0.00018383363861652837,
      "loss": 3.2812,
      "step": 64500
    },
    {
      "epoch": 6.947583683134216,
      "grad_norm": 0.7118873596191406,
      "learning_rate": 0.00018351039758646696,
      "loss": 3.2889,
      "step": 64550
    },
    {
      "epoch": 6.952965235173824,
      "grad_norm": 0.7383100390434265,
      "learning_rate": 0.00018318715655640555,
      "loss": 3.2833,
      "step": 64600
    },
    {
      "epoch": 6.958346787213433,
      "grad_norm": 0.7142897844314575,
      "learning_rate": 0.00018286391552634412,
      "loss": 3.2963,
      "step": 64650
    },
    {
      "epoch": 6.9637283392530405,
      "grad_norm": 0.7522610425949097,
      "learning_rate": 0.00018254067449628272,
      "loss": 3.2769,
      "step": 64700
    },
    {
      "epoch": 6.969109891292649,
      "grad_norm": 0.728192925453186,
      "learning_rate": 0.00018221743346622128,
      "loss": 3.2866,
      "step": 64750
    },
    {
      "epoch": 6.974491443332257,
      "grad_norm": 0.7720183730125427,
      "learning_rate": 0.00018189419243615988,
      "loss": 3.2778,
      "step": 64800
    },
    {
      "epoch": 6.979872995371865,
      "grad_norm": 0.7467803359031677,
      "learning_rate": 0.00018157095140609847,
      "loss": 3.2798,
      "step": 64850
    },
    {
      "epoch": 6.985254547411474,
      "grad_norm": 0.8028253316879272,
      "learning_rate": 0.00018124771037603704,
      "loss": 3.2707,
      "step": 64900
    },
    {
      "epoch": 6.990636099451081,
      "grad_norm": 0.7127172350883484,
      "learning_rate": 0.0001809244693459756,
      "loss": 3.2682,
      "step": 64950
    },
    {
      "epoch": 6.99601765149069,
      "grad_norm": 0.7452797293663025,
      "learning_rate": 0.00018060122831591423,
      "loss": 3.2976,
      "step": 65000
    },
    {
      "epoch": 6.99601765149069,
      "eval_accuracy": 0.3878374609297797,
      "eval_loss": 3.3518593311309814,
      "eval_runtime": 190.2774,
      "eval_samples_per_second": 94.657,
      "eval_steps_per_second": 5.918,
      "step": 65000
    },
    {
      "epoch": 7.0013992035302985,
      "grad_norm": 0.7658704519271851,
      "learning_rate": 0.0001802779872858528,
      "loss": 3.2585,
      "step": 65050
    },
    {
      "epoch": 7.006780755569906,
      "grad_norm": 0.7505213022232056,
      "learning_rate": 0.00017995474625579137,
      "loss": 3.202,
      "step": 65100
    },
    {
      "epoch": 7.012162307609515,
      "grad_norm": 0.767410397529602,
      "learning_rate": 0.00017963150522573,
      "loss": 3.1883,
      "step": 65150
    },
    {
      "epoch": 7.017543859649122,
      "grad_norm": 0.7873502373695374,
      "learning_rate": 0.00017930826419566856,
      "loss": 3.201,
      "step": 65200
    },
    {
      "epoch": 7.022925411688731,
      "grad_norm": 0.7236568927764893,
      "learning_rate": 0.00017898502316560712,
      "loss": 3.178,
      "step": 65250
    },
    {
      "epoch": 7.0283069637283395,
      "grad_norm": 0.7697278261184692,
      "learning_rate": 0.00017866178213554572,
      "loss": 3.2122,
      "step": 65300
    },
    {
      "epoch": 7.033688515767947,
      "grad_norm": 0.7726287841796875,
      "learning_rate": 0.0001783385411054843,
      "loss": 3.1944,
      "step": 65350
    },
    {
      "epoch": 7.039070067807556,
      "grad_norm": 0.7471922636032104,
      "learning_rate": 0.0001780153000754229,
      "loss": 3.2096,
      "step": 65400
    },
    {
      "epoch": 7.044451619847164,
      "grad_norm": 0.779881477355957,
      "learning_rate": 0.00017769205904536147,
      "loss": 3.1936,
      "step": 65450
    },
    {
      "epoch": 7.049833171886772,
      "grad_norm": 0.718411386013031,
      "learning_rate": 0.00017736881801530004,
      "loss": 3.1914,
      "step": 65500
    },
    {
      "epoch": 7.0552147239263805,
      "grad_norm": 0.7645953297615051,
      "learning_rate": 0.00017704557698523866,
      "loss": 3.2038,
      "step": 65550
    },
    {
      "epoch": 7.060596275965988,
      "grad_norm": 0.7848701477050781,
      "learning_rate": 0.00017672233595517723,
      "loss": 3.2111,
      "step": 65600
    },
    {
      "epoch": 7.065977828005597,
      "grad_norm": 0.7617135047912598,
      "learning_rate": 0.0001763990949251158,
      "loss": 3.2076,
      "step": 65650
    },
    {
      "epoch": 7.071359380045205,
      "grad_norm": 0.7438549995422363,
      "learning_rate": 0.00017607585389505442,
      "loss": 3.2109,
      "step": 65700
    },
    {
      "epoch": 7.076740932084813,
      "grad_norm": 0.7936565279960632,
      "learning_rate": 0.000175752612864993,
      "loss": 3.2091,
      "step": 65750
    },
    {
      "epoch": 7.0821224841244215,
      "grad_norm": 0.7821770906448364,
      "learning_rate": 0.00017542937183493156,
      "loss": 3.2017,
      "step": 65800
    },
    {
      "epoch": 7.08750403616403,
      "grad_norm": 0.7998859286308289,
      "learning_rate": 0.00017510613080487015,
      "loss": 3.2041,
      "step": 65850
    },
    {
      "epoch": 7.092885588203638,
      "grad_norm": 0.7892220616340637,
      "learning_rate": 0.00017478288977480874,
      "loss": 3.2177,
      "step": 65900
    },
    {
      "epoch": 7.098267140243246,
      "grad_norm": 0.7639216184616089,
      "learning_rate": 0.0001744596487447473,
      "loss": 3.2003,
      "step": 65950
    },
    {
      "epoch": 7.103648692282855,
      "grad_norm": 0.7874082326889038,
      "learning_rate": 0.0001741364077146859,
      "loss": 3.2249,
      "step": 66000
    },
    {
      "epoch": 7.103648692282855,
      "eval_accuracy": 0.38738763797422776,
      "eval_loss": 3.3583173751831055,
      "eval_runtime": 189.979,
      "eval_samples_per_second": 94.805,
      "eval_steps_per_second": 5.927,
      "step": 66000
    },
    {
      "epoch": 7.109030244322462,
      "grad_norm": 0.7661296725273132,
      "learning_rate": 0.00017381316668462447,
      "loss": 3.1967,
      "step": 66050
    },
    {
      "epoch": 7.114411796362071,
      "grad_norm": 0.7492254972457886,
      "learning_rate": 0.00017348992565456304,
      "loss": 3.2228,
      "step": 66100
    },
    {
      "epoch": 7.119793348401679,
      "grad_norm": 0.8151422142982483,
      "learning_rate": 0.00017316668462450166,
      "loss": 3.2147,
      "step": 66150
    },
    {
      "epoch": 7.125174900441287,
      "grad_norm": 0.7686956524848938,
      "learning_rate": 0.00017284344359444023,
      "loss": 3.2081,
      "step": 66200
    },
    {
      "epoch": 7.130556452480896,
      "grad_norm": 0.7949997186660767,
      "learning_rate": 0.00017252020256437885,
      "loss": 3.2164,
      "step": 66250
    },
    {
      "epoch": 7.135938004520503,
      "grad_norm": 0.855109453201294,
      "learning_rate": 0.00017219696153431742,
      "loss": 3.2115,
      "step": 66300
    },
    {
      "epoch": 7.141319556560112,
      "grad_norm": 0.7911727428436279,
      "learning_rate": 0.00017188018532485723,
      "loss": 3.2048,
      "step": 66350
    },
    {
      "epoch": 7.1467011085997205,
      "grad_norm": 0.7496804594993591,
      "learning_rate": 0.0001715569442947958,
      "loss": 3.2223,
      "step": 66400
    },
    {
      "epoch": 7.152082660639328,
      "grad_norm": 0.7830147743225098,
      "learning_rate": 0.0001712337032647344,
      "loss": 3.2149,
      "step": 66450
    },
    {
      "epoch": 7.157464212678937,
      "grad_norm": 0.7633971571922302,
      "learning_rate": 0.00017091046223467298,
      "loss": 3.2136,
      "step": 66500
    },
    {
      "epoch": 7.162845764718545,
      "grad_norm": 0.7498416900634766,
      "learning_rate": 0.00017058722120461155,
      "loss": 3.1984,
      "step": 66550
    },
    {
      "epoch": 7.168227316758153,
      "grad_norm": 0.7846064567565918,
      "learning_rate": 0.00017026398017455012,
      "loss": 3.2094,
      "step": 66600
    },
    {
      "epoch": 7.1736088687977615,
      "grad_norm": 0.7548909187316895,
      "learning_rate": 0.00016994720396508995,
      "loss": 3.2099,
      "step": 66650
    },
    {
      "epoch": 7.178990420837369,
      "grad_norm": 0.7729141712188721,
      "learning_rate": 0.00016962396293502852,
      "loss": 3.2053,
      "step": 66700
    },
    {
      "epoch": 7.184371972876978,
      "grad_norm": 0.816646933555603,
      "learning_rate": 0.00016930072190496714,
      "loss": 3.2249,
      "step": 66750
    },
    {
      "epoch": 7.189753524916586,
      "grad_norm": 0.750622034072876,
      "learning_rate": 0.0001689774808749057,
      "loss": 3.2165,
      "step": 66800
    },
    {
      "epoch": 7.195135076956194,
      "grad_norm": 0.7979385256767273,
      "learning_rate": 0.00016865423984484428,
      "loss": 3.2147,
      "step": 66850
    },
    {
      "epoch": 7.2005166289958025,
      "grad_norm": 0.7793417572975159,
      "learning_rate": 0.00016833099881478287,
      "loss": 3.2045,
      "step": 66900
    },
    {
      "epoch": 7.205898181035411,
      "grad_norm": 0.7790278196334839,
      "learning_rate": 0.00016800775778472144,
      "loss": 3.2069,
      "step": 66950
    },
    {
      "epoch": 7.211279733075019,
      "grad_norm": 0.7771934270858765,
      "learning_rate": 0.00016768451675466006,
      "loss": 3.2144,
      "step": 67000
    },
    {
      "epoch": 7.211279733075019,
      "eval_accuracy": 0.38788211726667143,
      "eval_loss": 3.3560354709625244,
      "eval_runtime": 190.3693,
      "eval_samples_per_second": 94.611,
      "eval_steps_per_second": 5.915,
      "step": 67000
    },
    {
      "epoch": 7.216661285114627,
      "grad_norm": 0.7584231495857239,
      "learning_rate": 0.00016736127572459863,
      "loss": 3.2255,
      "step": 67050
    },
    {
      "epoch": 7.222042837154235,
      "grad_norm": 0.8255765438079834,
      "learning_rate": 0.0001670380346945372,
      "loss": 3.2248,
      "step": 67100
    },
    {
      "epoch": 7.2274243891938434,
      "grad_norm": 0.8321775794029236,
      "learning_rate": 0.00016671479366447582,
      "loss": 3.2234,
      "step": 67150
    },
    {
      "epoch": 7.232805941233452,
      "grad_norm": 0.7867320775985718,
      "learning_rate": 0.00016639155263441439,
      "loss": 3.2265,
      "step": 67200
    },
    {
      "epoch": 7.23818749327306,
      "grad_norm": 0.7619370222091675,
      "learning_rate": 0.00016606831160435295,
      "loss": 3.2119,
      "step": 67250
    },
    {
      "epoch": 7.243569045312668,
      "grad_norm": 0.7555158734321594,
      "learning_rate": 0.00016574507057429155,
      "loss": 3.2035,
      "step": 67300
    },
    {
      "epoch": 7.248950597352277,
      "grad_norm": 0.7570140361785889,
      "learning_rate": 0.00016542182954423014,
      "loss": 3.2434,
      "step": 67350
    },
    {
      "epoch": 7.254332149391884,
      "grad_norm": 0.7532446384429932,
      "learning_rate": 0.0001650985885141687,
      "loss": 3.2135,
      "step": 67400
    },
    {
      "epoch": 7.259713701431493,
      "grad_norm": 0.8236835598945618,
      "learning_rate": 0.0001647753474841073,
      "loss": 3.2317,
      "step": 67450
    },
    {
      "epoch": 7.265095253471101,
      "grad_norm": 0.7405281662940979,
      "learning_rate": 0.00016445210645404587,
      "loss": 3.2159,
      "step": 67500
    },
    {
      "epoch": 7.270476805510709,
      "grad_norm": 0.7878313660621643,
      "learning_rate": 0.00016412886542398447,
      "loss": 3.2368,
      "step": 67550
    },
    {
      "epoch": 7.275858357550318,
      "grad_norm": 0.8011958599090576,
      "learning_rate": 0.00016380562439392306,
      "loss": 3.2305,
      "step": 67600
    },
    {
      "epoch": 7.281239909589925,
      "grad_norm": 0.8197931051254272,
      "learning_rate": 0.00016348238336386163,
      "loss": 3.2112,
      "step": 67650
    },
    {
      "epoch": 7.286621461629534,
      "grad_norm": 0.7783384323120117,
      "learning_rate": 0.0001631591423338002,
      "loss": 3.2309,
      "step": 67700
    },
    {
      "epoch": 7.2920030136691425,
      "grad_norm": 0.8042258024215698,
      "learning_rate": 0.00016283590130373882,
      "loss": 3.207,
      "step": 67750
    },
    {
      "epoch": 7.29738456570875,
      "grad_norm": 0.7969570755958557,
      "learning_rate": 0.00016251266027367739,
      "loss": 3.212,
      "step": 67800
    },
    {
      "epoch": 7.302766117748359,
      "grad_norm": 0.7729682326316833,
      "learning_rate": 0.00016218941924361595,
      "loss": 3.2285,
      "step": 67850
    },
    {
      "epoch": 7.308147669787967,
      "grad_norm": 0.8051829934120178,
      "learning_rate": 0.00016186617821355458,
      "loss": 3.2173,
      "step": 67900
    },
    {
      "epoch": 7.313529221827575,
      "grad_norm": 0.7974266409873962,
      "learning_rate": 0.00016154293718349314,
      "loss": 3.2278,
      "step": 67950
    },
    {
      "epoch": 7.3189107738671835,
      "grad_norm": 0.849501371383667,
      "learning_rate": 0.00016121969615343174,
      "loss": 3.2323,
      "step": 68000
    },
    {
      "epoch": 7.3189107738671835,
      "eval_accuracy": 0.38829901839721565,
      "eval_loss": 3.351057291030884,
      "eval_runtime": 190.0561,
      "eval_samples_per_second": 94.767,
      "eval_steps_per_second": 5.925,
      "step": 68000
    },
    {
      "epoch": 7.324292325906791,
      "grad_norm": 0.7351307272911072,
      "learning_rate": 0.0001608964551233703,
      "loss": 3.215,
      "step": 68050
    },
    {
      "epoch": 7.3296738779464,
      "grad_norm": 0.7568185925483704,
      "learning_rate": 0.0001605732140933089,
      "loss": 3.2267,
      "step": 68100
    },
    {
      "epoch": 7.335055429986008,
      "grad_norm": 0.7892095446586609,
      "learning_rate": 0.0001602499730632475,
      "loss": 3.2166,
      "step": 68150
    },
    {
      "epoch": 7.340436982025616,
      "grad_norm": 0.7952831387519836,
      "learning_rate": 0.00015992673203318606,
      "loss": 3.2311,
      "step": 68200
    },
    {
      "epoch": 7.3458185340652244,
      "grad_norm": 0.8136652708053589,
      "learning_rate": 0.00015960349100312463,
      "loss": 3.2352,
      "step": 68250
    },
    {
      "epoch": 7.351200086104833,
      "grad_norm": 0.7723620533943176,
      "learning_rate": 0.00015928024997306325,
      "loss": 3.2063,
      "step": 68300
    },
    {
      "epoch": 7.356581638144441,
      "grad_norm": 0.7793118953704834,
      "learning_rate": 0.00015895700894300182,
      "loss": 3.2254,
      "step": 68350
    },
    {
      "epoch": 7.361963190184049,
      "grad_norm": 0.7805073857307434,
      "learning_rate": 0.00015863376791294039,
      "loss": 3.205,
      "step": 68400
    },
    {
      "epoch": 7.367344742223658,
      "grad_norm": 0.773893415927887,
      "learning_rate": 0.000158310526882879,
      "loss": 3.2119,
      "step": 68450
    },
    {
      "epoch": 7.372726294263265,
      "grad_norm": 0.7920604348182678,
      "learning_rate": 0.00015798728585281758,
      "loss": 3.2137,
      "step": 68500
    },
    {
      "epoch": 7.378107846302874,
      "grad_norm": 0.7308557629585266,
      "learning_rate": 0.00015766404482275614,
      "loss": 3.2369,
      "step": 68550
    },
    {
      "epoch": 7.383489398342482,
      "grad_norm": 0.7705138921737671,
      "learning_rate": 0.00015734080379269474,
      "loss": 3.2333,
      "step": 68600
    },
    {
      "epoch": 7.38887095038209,
      "grad_norm": 0.855993390083313,
      "learning_rate": 0.00015701756276263333,
      "loss": 3.2318,
      "step": 68650
    },
    {
      "epoch": 7.394252502421699,
      "grad_norm": 0.7925063371658325,
      "learning_rate": 0.0001566943217325719,
      "loss": 3.234,
      "step": 68700
    },
    {
      "epoch": 7.399634054461306,
      "grad_norm": 0.7881805300712585,
      "learning_rate": 0.0001563710807025105,
      "loss": 3.2348,
      "step": 68750
    },
    {
      "epoch": 7.405015606500915,
      "grad_norm": 0.7927467226982117,
      "learning_rate": 0.00015604783967244906,
      "loss": 3.2191,
      "step": 68800
    },
    {
      "epoch": 7.4103971585405235,
      "grad_norm": 0.8324493169784546,
      "learning_rate": 0.00015572459864238763,
      "loss": 3.2317,
      "step": 68850
    },
    {
      "epoch": 7.415778710580131,
      "grad_norm": 0.8232147693634033,
      "learning_rate": 0.00015540135761232625,
      "loss": 3.2399,
      "step": 68900
    },
    {
      "epoch": 7.42116026261974,
      "grad_norm": 0.7550985217094421,
      "learning_rate": 0.00015507811658226482,
      "loss": 3.2289,
      "step": 68950
    },
    {
      "epoch": 7.426541814659347,
      "grad_norm": 0.803789496421814,
      "learning_rate": 0.00015475487555220344,
      "loss": 3.2143,
      "step": 69000
    },
    {
      "epoch": 7.426541814659347,
      "eval_accuracy": 0.388261315845144,
      "eval_loss": 3.351091146469116,
      "eval_runtime": 189.6922,
      "eval_samples_per_second": 94.949,
      "eval_steps_per_second": 5.936,
      "step": 69000
    },
    {
      "epoch": 7.431923366698956,
      "grad_norm": 0.7600038647651672,
      "learning_rate": 0.000154431634522142,
      "loss": 3.2342,
      "step": 69050
    },
    {
      "epoch": 7.4373049187385645,
      "grad_norm": 0.7752826809883118,
      "learning_rate": 0.00015410839349208058,
      "loss": 3.2347,
      "step": 69100
    },
    {
      "epoch": 7.442686470778172,
      "grad_norm": 0.7836809754371643,
      "learning_rate": 0.00015378515246201917,
      "loss": 3.204,
      "step": 69150
    },
    {
      "epoch": 7.448068022817781,
      "grad_norm": 0.8244750499725342,
      "learning_rate": 0.00015346191143195774,
      "loss": 3.2207,
      "step": 69200
    },
    {
      "epoch": 7.453449574857389,
      "grad_norm": 0.7681989669799805,
      "learning_rate": 0.00015313867040189633,
      "loss": 3.2302,
      "step": 69250
    },
    {
      "epoch": 7.458831126896997,
      "grad_norm": 0.7804749011993408,
      "learning_rate": 0.00015281542937183493,
      "loss": 3.2284,
      "step": 69300
    },
    {
      "epoch": 7.4642126789366054,
      "grad_norm": 0.8117129802703857,
      "learning_rate": 0.0001524921883417735,
      "loss": 3.2366,
      "step": 69350
    },
    {
      "epoch": 7.469594230976213,
      "grad_norm": 0.8194923996925354,
      "learning_rate": 0.00015216894731171206,
      "loss": 3.2415,
      "step": 69400
    },
    {
      "epoch": 7.474975783015822,
      "grad_norm": 0.8146264553070068,
      "learning_rate": 0.00015184570628165068,
      "loss": 3.2211,
      "step": 69450
    },
    {
      "epoch": 7.48035733505543,
      "grad_norm": 0.8037434816360474,
      "learning_rate": 0.00015152246525158925,
      "loss": 3.2275,
      "step": 69500
    },
    {
      "epoch": 7.485738887095038,
      "grad_norm": 0.7875609397888184,
      "learning_rate": 0.00015119922422152782,
      "loss": 3.2227,
      "step": 69550
    },
    {
      "epoch": 7.491120439134646,
      "grad_norm": 0.7770443558692932,
      "learning_rate": 0.00015087598319146644,
      "loss": 3.2253,
      "step": 69600
    },
    {
      "epoch": 7.496501991174255,
      "grad_norm": 0.7791767716407776,
      "learning_rate": 0.000150552742161405,
      "loss": 3.2341,
      "step": 69650
    },
    {
      "epoch": 7.501883543213863,
      "grad_norm": 0.809085488319397,
      "learning_rate": 0.00015022950113134358,
      "loss": 3.2128,
      "step": 69700
    },
    {
      "epoch": 7.507265095253471,
      "grad_norm": 0.781852126121521,
      "learning_rate": 0.00014990626010128217,
      "loss": 3.2239,
      "step": 69750
    },
    {
      "epoch": 7.51264664729308,
      "grad_norm": 0.8119115829467773,
      "learning_rate": 0.00014958301907122077,
      "loss": 3.2481,
      "step": 69800
    },
    {
      "epoch": 7.518028199332687,
      "grad_norm": 0.8209654092788696,
      "learning_rate": 0.00014925977804115933,
      "loss": 3.2396,
      "step": 69850
    },
    {
      "epoch": 7.523409751372296,
      "grad_norm": 0.8188632726669312,
      "learning_rate": 0.00014893653701109793,
      "loss": 3.2466,
      "step": 69900
    },
    {
      "epoch": 7.528791303411904,
      "grad_norm": 0.8054358959197998,
      "learning_rate": 0.0001486132959810365,
      "loss": 3.2162,
      "step": 69950
    },
    {
      "epoch": 7.534172855451512,
      "grad_norm": 0.7858330607414246,
      "learning_rate": 0.0001482900549509751,
      "loss": 3.245,
      "step": 70000
    },
    {
      "epoch": 7.534172855451512,
      "eval_accuracy": 0.38878013338445816,
      "eval_loss": 3.345130443572998,
      "eval_runtime": 190.2002,
      "eval_samples_per_second": 94.695,
      "eval_steps_per_second": 5.92,
      "step": 70000
    },
    {
      "epoch": 7.539554407491121,
      "grad_norm": 0.7718562483787537,
      "learning_rate": 0.00014796681392091368,
      "loss": 3.2183,
      "step": 70050
    },
    {
      "epoch": 7.544935959530728,
      "grad_norm": 0.8102341294288635,
      "learning_rate": 0.00014764357289085228,
      "loss": 3.252,
      "step": 70100
    },
    {
      "epoch": 7.550317511570337,
      "grad_norm": 0.7838711738586426,
      "learning_rate": 0.00014732033186079085,
      "loss": 3.2132,
      "step": 70150
    },
    {
      "epoch": 7.5556990636099455,
      "grad_norm": 0.830418050289154,
      "learning_rate": 0.00014699709083072944,
      "loss": 3.2271,
      "step": 70200
    },
    {
      "epoch": 7.561080615649553,
      "grad_norm": 0.7321741580963135,
      "learning_rate": 0.00014667384980066804,
      "loss": 3.2167,
      "step": 70250
    },
    {
      "epoch": 7.566462167689162,
      "grad_norm": 0.7650930881500244,
      "learning_rate": 0.0001463506087706066,
      "loss": 3.221,
      "step": 70300
    },
    {
      "epoch": 7.57184371972877,
      "grad_norm": 0.7693846225738525,
      "learning_rate": 0.0001460273677405452,
      "loss": 3.2232,
      "step": 70350
    },
    {
      "epoch": 7.577225271768378,
      "grad_norm": 0.768191933631897,
      "learning_rate": 0.00014570412671048377,
      "loss": 3.2379,
      "step": 70400
    },
    {
      "epoch": 7.5826068238079865,
      "grad_norm": 0.7954201102256775,
      "learning_rate": 0.00014538088568042236,
      "loss": 3.2352,
      "step": 70450
    },
    {
      "epoch": 7.587988375847594,
      "grad_norm": 0.7926831841468811,
      "learning_rate": 0.00014505764465036093,
      "loss": 3.225,
      "step": 70500
    },
    {
      "epoch": 7.593369927887203,
      "grad_norm": 0.7868989706039429,
      "learning_rate": 0.00014473440362029952,
      "loss": 3.2308,
      "step": 70550
    },
    {
      "epoch": 7.598751479926811,
      "grad_norm": 0.745983898639679,
      "learning_rate": 0.00014441116259023812,
      "loss": 3.2348,
      "step": 70600
    },
    {
      "epoch": 7.604133031966419,
      "grad_norm": 0.7672894597053528,
      "learning_rate": 0.00014408792156017669,
      "loss": 3.2342,
      "step": 70650
    },
    {
      "epoch": 7.609514584006027,
      "grad_norm": 0.752560555934906,
      "learning_rate": 0.00014376468053011528,
      "loss": 3.2087,
      "step": 70700
    },
    {
      "epoch": 7.614896136045635,
      "grad_norm": 0.798228919506073,
      "learning_rate": 0.00014344143950005387,
      "loss": 3.2363,
      "step": 70750
    },
    {
      "epoch": 7.620277688085244,
      "grad_norm": 0.8537287712097168,
      "learning_rate": 0.00014311819846999244,
      "loss": 3.2448,
      "step": 70800
    },
    {
      "epoch": 7.625659240124852,
      "grad_norm": 0.7980437874794006,
      "learning_rate": 0.00014279495743993104,
      "loss": 3.2296,
      "step": 70850
    },
    {
      "epoch": 7.63104079216446,
      "grad_norm": 0.7725957036018372,
      "learning_rate": 0.00014247171640986963,
      "loss": 3.2179,
      "step": 70900
    },
    {
      "epoch": 7.636422344204068,
      "grad_norm": 0.8028611540794373,
      "learning_rate": 0.0001421484753798082,
      "loss": 3.2151,
      "step": 70950
    },
    {
      "epoch": 7.641803896243677,
      "grad_norm": 0.8120908141136169,
      "learning_rate": 0.0001418252343497468,
      "loss": 3.2242,
      "step": 71000
    },
    {
      "epoch": 7.641803896243677,
      "eval_accuracy": 0.3895405949462717,
      "eval_loss": 3.3396248817443848,
      "eval_runtime": 190.2457,
      "eval_samples_per_second": 94.672,
      "eval_steps_per_second": 5.919,
      "step": 71000
    },
    {
      "epoch": 7.647185448283285,
      "grad_norm": 0.7809242606163025,
      "learning_rate": 0.00014150199331968536,
      "loss": 3.2267,
      "step": 71050
    },
    {
      "epoch": 7.652567000322893,
      "grad_norm": 0.8149883151054382,
      "learning_rate": 0.00014117875228962396,
      "loss": 3.2345,
      "step": 71100
    },
    {
      "epoch": 7.657948552362502,
      "grad_norm": 0.8204036951065063,
      "learning_rate": 0.00014085551125956252,
      "loss": 3.2166,
      "step": 71150
    },
    {
      "epoch": 7.663330104402109,
      "grad_norm": 0.8267731070518494,
      "learning_rate": 0.00014053227022950112,
      "loss": 3.2088,
      "step": 71200
    },
    {
      "epoch": 7.668711656441718,
      "grad_norm": 0.7837862968444824,
      "learning_rate": 0.0001402090291994397,
      "loss": 3.2314,
      "step": 71250
    },
    {
      "epoch": 7.674093208481326,
      "grad_norm": 0.7852527499198914,
      "learning_rate": 0.00013988578816937828,
      "loss": 3.2282,
      "step": 71300
    },
    {
      "epoch": 7.679474760520934,
      "grad_norm": 0.7429063320159912,
      "learning_rate": 0.00013956254713931687,
      "loss": 3.2376,
      "step": 71350
    },
    {
      "epoch": 7.684856312560543,
      "grad_norm": 0.7983697056770325,
      "learning_rate": 0.00013923930610925547,
      "loss": 3.2291,
      "step": 71400
    },
    {
      "epoch": 7.69023786460015,
      "grad_norm": 0.7264466285705566,
      "learning_rate": 0.00013891606507919404,
      "loss": 3.2356,
      "step": 71450
    },
    {
      "epoch": 7.695619416639759,
      "grad_norm": 0.7640838027000427,
      "learning_rate": 0.00013859282404913263,
      "loss": 3.2375,
      "step": 71500
    },
    {
      "epoch": 7.7010009686793675,
      "grad_norm": 0.8236213326454163,
      "learning_rate": 0.0001382695830190712,
      "loss": 3.2185,
      "step": 71550
    },
    {
      "epoch": 7.706382520718975,
      "grad_norm": 0.7898839116096497,
      "learning_rate": 0.0001379463419890098,
      "loss": 3.2156,
      "step": 71600
    },
    {
      "epoch": 7.711764072758584,
      "grad_norm": 0.766157329082489,
      "learning_rate": 0.00013762310095894836,
      "loss": 3.2376,
      "step": 71650
    },
    {
      "epoch": 7.717145624798192,
      "grad_norm": 0.7972264885902405,
      "learning_rate": 0.00013729985992888696,
      "loss": 3.2108,
      "step": 71700
    },
    {
      "epoch": 7.7225271768378,
      "grad_norm": 0.7563880085945129,
      "learning_rate": 0.00013697661889882555,
      "loss": 3.2374,
      "step": 71750
    },
    {
      "epoch": 7.727908728877408,
      "grad_norm": 0.8285365104675293,
      "learning_rate": 0.00013665337786876412,
      "loss": 3.231,
      "step": 71800
    },
    {
      "epoch": 7.733290280917016,
      "grad_norm": 0.7620691061019897,
      "learning_rate": 0.0001363301368387027,
      "loss": 3.236,
      "step": 71850
    },
    {
      "epoch": 7.738671832956625,
      "grad_norm": 0.796030580997467,
      "learning_rate": 0.0001360068958086413,
      "loss": 3.2342,
      "step": 71900
    },
    {
      "epoch": 7.744053384996233,
      "grad_norm": 0.802382230758667,
      "learning_rate": 0.00013568365477857988,
      "loss": 3.24,
      "step": 71950
    },
    {
      "epoch": 7.749434937035841,
      "grad_norm": 1.017646312713623,
      "learning_rate": 0.00013536041374851847,
      "loss": 3.232,
      "step": 72000
    },
    {
      "epoch": 7.749434937035841,
      "eval_accuracy": 0.38963262394225057,
      "eval_loss": 3.337721824645996,
      "eval_runtime": 189.8132,
      "eval_samples_per_second": 94.888,
      "eval_steps_per_second": 5.932,
      "step": 72000
    },
    {
      "epoch": 7.754816489075449,
      "grad_norm": 0.8234233260154724,
      "learning_rate": 0.00013503717271845706,
      "loss": 3.2352,
      "step": 72050
    },
    {
      "epoch": 7.760198041115058,
      "grad_norm": 0.7723364233970642,
      "learning_rate": 0.00013471393168839563,
      "loss": 3.2364,
      "step": 72100
    },
    {
      "epoch": 7.765579593154666,
      "grad_norm": 0.8264212012290955,
      "learning_rate": 0.00013439069065833423,
      "loss": 3.2308,
      "step": 72150
    },
    {
      "epoch": 7.770961145194274,
      "grad_norm": 0.7821183800697327,
      "learning_rate": 0.0001340674496282728,
      "loss": 3.225,
      "step": 72200
    },
    {
      "epoch": 7.776342697233883,
      "grad_norm": 0.8186830282211304,
      "learning_rate": 0.0001337442085982114,
      "loss": 3.2213,
      "step": 72250
    },
    {
      "epoch": 7.78172424927349,
      "grad_norm": 0.7722200155258179,
      "learning_rate": 0.00013342096756814996,
      "loss": 3.2322,
      "step": 72300
    },
    {
      "epoch": 7.787105801313099,
      "grad_norm": 0.8005505204200745,
      "learning_rate": 0.00013309772653808855,
      "loss": 3.2472,
      "step": 72350
    },
    {
      "epoch": 7.792487353352707,
      "grad_norm": 0.7402283549308777,
      "learning_rate": 0.00013277448550802715,
      "loss": 3.2305,
      "step": 72400
    },
    {
      "epoch": 7.797868905392315,
      "grad_norm": 0.759713351726532,
      "learning_rate": 0.0001324512444779657,
      "loss": 3.2507,
      "step": 72450
    },
    {
      "epoch": 7.803250457431924,
      "grad_norm": 0.7819993495941162,
      "learning_rate": 0.0001321280034479043,
      "loss": 3.2174,
      "step": 72500
    },
    {
      "epoch": 7.808632009471531,
      "grad_norm": 0.7440714836120605,
      "learning_rate": 0.0001318047624178429,
      "loss": 3.2099,
      "step": 72550
    },
    {
      "epoch": 7.81401356151114,
      "grad_norm": 0.7539764642715454,
      "learning_rate": 0.0001314815213877815,
      "loss": 3.237,
      "step": 72600
    },
    {
      "epoch": 7.819395113550748,
      "grad_norm": 0.8063895106315613,
      "learning_rate": 0.00013116474517832128,
      "loss": 3.2311,
      "step": 72650
    },
    {
      "epoch": 7.824776665590356,
      "grad_norm": 0.7834722995758057,
      "learning_rate": 0.00013084150414825987,
      "loss": 3.2351,
      "step": 72700
    },
    {
      "epoch": 7.830158217629965,
      "grad_norm": 0.8279936909675598,
      "learning_rate": 0.00013051826311819844,
      "loss": 3.2409,
      "step": 72750
    },
    {
      "epoch": 7.835539769669572,
      "grad_norm": 0.8224948048591614,
      "learning_rate": 0.00013019502208813703,
      "loss": 3.237,
      "step": 72800
    },
    {
      "epoch": 7.840921321709181,
      "grad_norm": 0.8062725067138672,
      "learning_rate": 0.00012987178105807563,
      "loss": 3.2411,
      "step": 72850
    },
    {
      "epoch": 7.846302873748789,
      "grad_norm": 0.802028477191925,
      "learning_rate": 0.00012954854002801422,
      "loss": 3.2308,
      "step": 72900
    },
    {
      "epoch": 7.851684425788397,
      "grad_norm": 0.843268096446991,
      "learning_rate": 0.0001292252989979528,
      "loss": 3.2455,
      "step": 72950
    },
    {
      "epoch": 7.857065977828006,
      "grad_norm": 0.7682187557220459,
      "learning_rate": 0.00012890205796789139,
      "loss": 3.2415,
      "step": 73000
    },
    {
      "epoch": 7.857065977828006,
      "eval_accuracy": 0.38995901721724524,
      "eval_loss": 3.3341541290283203,
      "eval_runtime": 190.2153,
      "eval_samples_per_second": 94.687,
      "eval_steps_per_second": 5.92,
      "step": 73000
    },
    {
      "epoch": 7.862447529867614,
      "grad_norm": 0.7943944931030273,
      "learning_rate": 0.00012857881693782998,
      "loss": 3.2359,
      "step": 73050
    },
    {
      "epoch": 7.867829081907222,
      "grad_norm": 0.8085488677024841,
      "learning_rate": 0.00012825557590776855,
      "loss": 3.2149,
      "step": 73100
    },
    {
      "epoch": 7.87321063394683,
      "grad_norm": 0.7818835377693176,
      "learning_rate": 0.00012793233487770714,
      "loss": 3.226,
      "step": 73150
    },
    {
      "epoch": 7.878592185986438,
      "grad_norm": 0.8501896858215332,
      "learning_rate": 0.0001276090938476457,
      "loss": 3.2309,
      "step": 73200
    },
    {
      "epoch": 7.883973738026047,
      "grad_norm": 0.7847088575363159,
      "learning_rate": 0.0001272858528175843,
      "loss": 3.2269,
      "step": 73250
    },
    {
      "epoch": 7.889355290065655,
      "grad_norm": 0.77759850025177,
      "learning_rate": 0.00012696261178752287,
      "loss": 3.2162,
      "step": 73300
    },
    {
      "epoch": 7.894736842105263,
      "grad_norm": 0.8463942408561707,
      "learning_rate": 0.00012663937075746147,
      "loss": 3.2261,
      "step": 73350
    },
    {
      "epoch": 7.900118394144871,
      "grad_norm": 0.8091511130332947,
      "learning_rate": 0.00012631612972740006,
      "loss": 3.2348,
      "step": 73400
    },
    {
      "epoch": 7.90549994618448,
      "grad_norm": 0.8184037804603577,
      "learning_rate": 0.00012599288869733863,
      "loss": 3.2207,
      "step": 73450
    },
    {
      "epoch": 7.910881498224088,
      "grad_norm": 0.8060580492019653,
      "learning_rate": 0.00012566964766727722,
      "loss": 3.2354,
      "step": 73500
    },
    {
      "epoch": 7.916263050263696,
      "grad_norm": 0.8405670523643494,
      "learning_rate": 0.00012534640663721582,
      "loss": 3.2404,
      "step": 73550
    },
    {
      "epoch": 7.921644602303305,
      "grad_norm": 0.793616533279419,
      "learning_rate": 0.00012502316560715439,
      "loss": 3.234,
      "step": 73600
    },
    {
      "epoch": 7.927026154342912,
      "grad_norm": 0.8910527229309082,
      "learning_rate": 0.00012469992457709298,
      "loss": 3.2345,
      "step": 73650
    },
    {
      "epoch": 7.932407706382521,
      "grad_norm": 0.7778403759002686,
      "learning_rate": 0.00012437668354703158,
      "loss": 3.2206,
      "step": 73700
    },
    {
      "epoch": 7.937789258422129,
      "grad_norm": 0.7960861921310425,
      "learning_rate": 0.00012405344251697014,
      "loss": 3.2347,
      "step": 73750
    },
    {
      "epoch": 7.943170810461737,
      "grad_norm": 0.8153463006019592,
      "learning_rate": 0.00012373020148690874,
      "loss": 3.2305,
      "step": 73800
    },
    {
      "epoch": 7.948552362501346,
      "grad_norm": 0.8068676590919495,
      "learning_rate": 0.00012341342527744854,
      "loss": 3.2158,
      "step": 73850
    },
    {
      "epoch": 7.953933914540953,
      "grad_norm": 0.812384307384491,
      "learning_rate": 0.0001230901842473871,
      "loss": 3.2476,
      "step": 73900
    },
    {
      "epoch": 7.959315466580562,
      "grad_norm": 0.7880306243896484,
      "learning_rate": 0.0001227669432173257,
      "loss": 3.249,
      "step": 73950
    },
    {
      "epoch": 7.96469701862017,
      "grad_norm": 0.8056456446647644,
      "learning_rate": 0.0001224437021872643,
      "loss": 3.2203,
      "step": 74000
    },
    {
      "epoch": 7.96469701862017,
      "eval_accuracy": 0.39043937163427306,
      "eval_loss": 3.3286335468292236,
      "eval_runtime": 189.9685,
      "eval_samples_per_second": 94.81,
      "eval_steps_per_second": 5.927,
      "step": 74000
    },
    {
      "epoch": 7.970078570659778,
      "grad_norm": 0.8301999568939209,
      "learning_rate": 0.00012212046115720287,
      "loss": 3.2124,
      "step": 74050
    },
    {
      "epoch": 7.975460122699387,
      "grad_norm": 0.8286321759223938,
      "learning_rate": 0.00012179722012714146,
      "loss": 3.2296,
      "step": 74100
    },
    {
      "epoch": 7.980841674738995,
      "grad_norm": 0.8427879214286804,
      "learning_rate": 0.00012147397909708004,
      "loss": 3.2238,
      "step": 74150
    },
    {
      "epoch": 7.986223226778603,
      "grad_norm": 0.7714815139770508,
      "learning_rate": 0.00012115073806701864,
      "loss": 3.2206,
      "step": 74200
    },
    {
      "epoch": 7.991604778818211,
      "grad_norm": 0.7950589060783386,
      "learning_rate": 0.0001208274970369572,
      "loss": 3.249,
      "step": 74250
    },
    {
      "epoch": 7.996986330857819,
      "grad_norm": 0.8017162084579468,
      "learning_rate": 0.0001205042560068958,
      "loss": 3.2427,
      "step": 74300
    },
    {
      "epoch": 8.002367882897428,
      "grad_norm": 0.780544638633728,
      "learning_rate": 0.0001201810149768344,
      "loss": 3.1826,
      "step": 74350
    },
    {
      "epoch": 8.007749434937036,
      "grad_norm": 0.7938651442527771,
      "learning_rate": 0.00011985777394677296,
      "loss": 3.1669,
      "step": 74400
    },
    {
      "epoch": 8.013130986976645,
      "grad_norm": 0.7893415093421936,
      "learning_rate": 0.00011953453291671154,
      "loss": 3.135,
      "step": 74450
    },
    {
      "epoch": 8.018512539016251,
      "grad_norm": 0.7795463800430298,
      "learning_rate": 0.00011921129188665014,
      "loss": 3.1442,
      "step": 74500
    },
    {
      "epoch": 8.02389409105586,
      "grad_norm": 0.7784271836280823,
      "learning_rate": 0.0001188880508565887,
      "loss": 3.1444,
      "step": 74550
    },
    {
      "epoch": 8.029275643095469,
      "grad_norm": 0.8579685091972351,
      "learning_rate": 0.0001185648098265273,
      "loss": 3.1508,
      "step": 74600
    },
    {
      "epoch": 8.034657195135077,
      "grad_norm": 0.8437514305114746,
      "learning_rate": 0.0001182415687964659,
      "loss": 3.1525,
      "step": 74650
    },
    {
      "epoch": 8.040038747174686,
      "grad_norm": 0.8004893660545349,
      "learning_rate": 0.00011791832776640448,
      "loss": 3.1644,
      "step": 74700
    },
    {
      "epoch": 8.045420299214294,
      "grad_norm": 0.7970762848854065,
      "learning_rate": 0.00011759508673634306,
      "loss": 3.1229,
      "step": 74750
    },
    {
      "epoch": 8.050801851253901,
      "grad_norm": 0.8060651421546936,
      "learning_rate": 0.00011727184570628164,
      "loss": 3.1711,
      "step": 74800
    },
    {
      "epoch": 8.05618340329351,
      "grad_norm": 0.7956426739692688,
      "learning_rate": 0.00011694860467622023,
      "loss": 3.1677,
      "step": 74850
    },
    {
      "epoch": 8.061564955333118,
      "grad_norm": 0.8106923699378967,
      "learning_rate": 0.0001166253636461588,
      "loss": 3.1552,
      "step": 74900
    },
    {
      "epoch": 8.066946507372727,
      "grad_norm": 0.8077366948127747,
      "learning_rate": 0.0001163021226160974,
      "loss": 3.1599,
      "step": 74950
    },
    {
      "epoch": 8.072328059412335,
      "grad_norm": 0.8373199701309204,
      "learning_rate": 0.00011597888158603598,
      "loss": 3.156,
      "step": 75000
    },
    {
      "epoch": 8.072328059412335,
      "eval_accuracy": 0.39020294295038876,
      "eval_loss": 3.33540678024292,
      "eval_runtime": 190.3165,
      "eval_samples_per_second": 94.637,
      "eval_steps_per_second": 5.916,
      "step": 75000
    },
    {
      "epoch": 8.077709611451942,
      "grad_norm": 0.8667137622833252,
      "learning_rate": 0.00011565564055597456,
      "loss": 3.1778,
      "step": 75050
    },
    {
      "epoch": 8.08309116349155,
      "grad_norm": 0.8201097249984741,
      "learning_rate": 0.00011533239952591314,
      "loss": 3.1664,
      "step": 75100
    },
    {
      "epoch": 8.088472715531159,
      "grad_norm": 0.8039436340332031,
      "learning_rate": 0.00011500915849585173,
      "loss": 3.1773,
      "step": 75150
    },
    {
      "epoch": 8.093854267570768,
      "grad_norm": 0.8173180818557739,
      "learning_rate": 0.00011468591746579033,
      "loss": 3.1523,
      "step": 75200
    },
    {
      "epoch": 8.099235819610376,
      "grad_norm": 0.8135039806365967,
      "learning_rate": 0.0001143626764357289,
      "loss": 3.1612,
      "step": 75250
    },
    {
      "epoch": 8.104617371649983,
      "grad_norm": 0.8241603970527649,
      "learning_rate": 0.00011403943540566749,
      "loss": 3.1841,
      "step": 75300
    },
    {
      "epoch": 8.109998923689592,
      "grad_norm": 0.8451812267303467,
      "learning_rate": 0.00011371619437560607,
      "loss": 3.158,
      "step": 75350
    },
    {
      "epoch": 8.1153804757292,
      "grad_norm": 0.8551275134086609,
      "learning_rate": 0.00011339295334554464,
      "loss": 3.1643,
      "step": 75400
    },
    {
      "epoch": 8.120762027768809,
      "grad_norm": 0.8072112202644348,
      "learning_rate": 0.00011306971231548323,
      "loss": 3.1647,
      "step": 75450
    },
    {
      "epoch": 8.126143579808417,
      "grad_norm": 0.7860527038574219,
      "learning_rate": 0.00011274647128542183,
      "loss": 3.1475,
      "step": 75500
    },
    {
      "epoch": 8.131525131848026,
      "grad_norm": 0.7687428593635559,
      "learning_rate": 0.0001124232302553604,
      "loss": 3.1774,
      "step": 75550
    },
    {
      "epoch": 8.136906683887632,
      "grad_norm": 0.8291406035423279,
      "learning_rate": 0.00011209998922529899,
      "loss": 3.1702,
      "step": 75600
    },
    {
      "epoch": 8.142288235927241,
      "grad_norm": 0.8310118913650513,
      "learning_rate": 0.00011177674819523757,
      "loss": 3.1612,
      "step": 75650
    },
    {
      "epoch": 8.14766978796685,
      "grad_norm": 0.8539298176765442,
      "learning_rate": 0.00011145350716517617,
      "loss": 3.1857,
      "step": 75700
    },
    {
      "epoch": 8.153051340006458,
      "grad_norm": 0.8130014538764954,
      "learning_rate": 0.00011113026613511473,
      "loss": 3.1589,
      "step": 75750
    },
    {
      "epoch": 8.158432892046067,
      "grad_norm": 0.8164442777633667,
      "learning_rate": 0.00011080702510505333,
      "loss": 3.1733,
      "step": 75800
    },
    {
      "epoch": 8.163814444085673,
      "grad_norm": 0.816186785697937,
      "learning_rate": 0.00011048378407499191,
      "loss": 3.1717,
      "step": 75850
    },
    {
      "epoch": 8.169195996125282,
      "grad_norm": 0.8147589564323425,
      "learning_rate": 0.00011016054304493049,
      "loss": 3.1696,
      "step": 75900
    },
    {
      "epoch": 8.17457754816489,
      "grad_norm": 0.8228213787078857,
      "learning_rate": 0.00010983730201486907,
      "loss": 3.1668,
      "step": 75950
    },
    {
      "epoch": 8.1799591002045,
      "grad_norm": 0.8431198596954346,
      "learning_rate": 0.00010951406098480767,
      "loss": 3.1603,
      "step": 76000
    },
    {
      "epoch": 8.1799591002045,
      "eval_accuracy": 0.39050543259006426,
      "eval_loss": 3.333915948867798,
      "eval_runtime": 190.0421,
      "eval_samples_per_second": 94.774,
      "eval_steps_per_second": 5.925,
      "step": 76000
    },
    {
      "epoch": 8.185340652244108,
      "grad_norm": 0.7798217535018921,
      "learning_rate": 0.00010919081995474623,
      "loss": 3.1584,
      "step": 76050
    },
    {
      "epoch": 8.190722204283716,
      "grad_norm": 0.830636203289032,
      "learning_rate": 0.00010886757892468483,
      "loss": 3.1734,
      "step": 76100
    },
    {
      "epoch": 8.196103756323323,
      "grad_norm": 0.8242005109786987,
      "learning_rate": 0.00010854433789462342,
      "loss": 3.1535,
      "step": 76150
    },
    {
      "epoch": 8.201485308362932,
      "grad_norm": 0.8283331394195557,
      "learning_rate": 0.000108221096864562,
      "loss": 3.1735,
      "step": 76200
    },
    {
      "epoch": 8.20686686040254,
      "grad_norm": 0.8457024693489075,
      "learning_rate": 0.00010789785583450059,
      "loss": 3.1722,
      "step": 76250
    },
    {
      "epoch": 8.212248412442149,
      "grad_norm": 0.8082210421562195,
      "learning_rate": 0.00010757461480443917,
      "loss": 3.172,
      "step": 76300
    },
    {
      "epoch": 8.217629964481757,
      "grad_norm": 0.8381291627883911,
      "learning_rate": 0.00010725137377437776,
      "loss": 3.1685,
      "step": 76350
    },
    {
      "epoch": 8.223011516521364,
      "grad_norm": 0.7815647125244141,
      "learning_rate": 0.00010692813274431633,
      "loss": 3.1683,
      "step": 76400
    },
    {
      "epoch": 8.228393068560973,
      "grad_norm": 0.8144623637199402,
      "learning_rate": 0.00010660489171425492,
      "loss": 3.1738,
      "step": 76450
    },
    {
      "epoch": 8.233774620600581,
      "grad_norm": 0.8301607966423035,
      "learning_rate": 0.0001062816506841935,
      "loss": 3.164,
      "step": 76500
    },
    {
      "epoch": 8.23915617264019,
      "grad_norm": 0.8335390686988831,
      "learning_rate": 0.00010595840965413209,
      "loss": 3.1694,
      "step": 76550
    },
    {
      "epoch": 8.244537724679798,
      "grad_norm": 0.8333656787872314,
      "learning_rate": 0.00010563516862407067,
      "loss": 3.1748,
      "step": 76600
    },
    {
      "epoch": 8.249919276719407,
      "grad_norm": 0.8568822741508484,
      "learning_rate": 0.00010531192759400926,
      "loss": 3.1889,
      "step": 76650
    },
    {
      "epoch": 8.255300828759013,
      "grad_norm": 0.82938152551651,
      "learning_rate": 0.00010498868656394784,
      "loss": 3.1831,
      "step": 76700
    },
    {
      "epoch": 8.260682380798622,
      "grad_norm": 0.813696026802063,
      "learning_rate": 0.00010466544553388642,
      "loss": 3.1615,
      "step": 76750
    },
    {
      "epoch": 8.26606393283823,
      "grad_norm": 0.8118577003479004,
      "learning_rate": 0.000104342204503825,
      "loss": 3.1658,
      "step": 76800
    },
    {
      "epoch": 8.27144548487784,
      "grad_norm": 0.7700328230857849,
      "learning_rate": 0.0001040189634737636,
      "loss": 3.1727,
      "step": 76850
    },
    {
      "epoch": 8.276827036917448,
      "grad_norm": 0.8473666906356812,
      "learning_rate": 0.00010369572244370217,
      "loss": 3.1597,
      "step": 76900
    },
    {
      "epoch": 8.282208588957054,
      "grad_norm": 0.8092920184135437,
      "learning_rate": 0.00010337248141364076,
      "loss": 3.1721,
      "step": 76950
    },
    {
      "epoch": 8.287590140996663,
      "grad_norm": 0.8102917075157166,
      "learning_rate": 0.00010304924038357936,
      "loss": 3.1911,
      "step": 77000
    },
    {
      "epoch": 8.287590140996663,
      "eval_accuracy": 0.3908290008899758,
      "eval_loss": 3.3315770626068115,
      "eval_runtime": 189.7719,
      "eval_samples_per_second": 94.909,
      "eval_steps_per_second": 5.933,
      "step": 77000
    },
    {
      "epoch": 8.292971693036272,
      "grad_norm": 0.8550979495048523,
      "learning_rate": 0.00010272599935351792,
      "loss": 3.1637,
      "step": 77050
    },
    {
      "epoch": 8.29835324507588,
      "grad_norm": 0.7964987754821777,
      "learning_rate": 0.00010240275832345652,
      "loss": 3.1637,
      "step": 77100
    },
    {
      "epoch": 8.303734797115489,
      "grad_norm": 0.8223913908004761,
      "learning_rate": 0.0001020795172933951,
      "loss": 3.1811,
      "step": 77150
    },
    {
      "epoch": 8.309116349155097,
      "grad_norm": 0.812432587146759,
      "learning_rate": 0.0001017562762633337,
      "loss": 3.1674,
      "step": 77200
    },
    {
      "epoch": 8.314497901194704,
      "grad_norm": 0.8130179643630981,
      "learning_rate": 0.00010143303523327226,
      "loss": 3.1804,
      "step": 77250
    },
    {
      "epoch": 8.319879453234313,
      "grad_norm": 0.8016122579574585,
      "learning_rate": 0.00010110979420321086,
      "loss": 3.1808,
      "step": 77300
    },
    {
      "epoch": 8.325261005273921,
      "grad_norm": 0.833431601524353,
      "learning_rate": 0.00010078655317314944,
      "loss": 3.1812,
      "step": 77350
    },
    {
      "epoch": 8.33064255731353,
      "grad_norm": 0.838892936706543,
      "learning_rate": 0.00010046331214308802,
      "loss": 3.1925,
      "step": 77400
    },
    {
      "epoch": 8.336024109353138,
      "grad_norm": 0.7732236981391907,
      "learning_rate": 0.0001001400711130266,
      "loss": 3.1758,
      "step": 77450
    },
    {
      "epoch": 8.341405661392745,
      "grad_norm": 0.8904444575309753,
      "learning_rate": 9.98168300829652e-05,
      "loss": 3.1759,
      "step": 77500
    },
    {
      "epoch": 8.346787213432354,
      "grad_norm": 0.8524021506309509,
      "learning_rate": 9.949358905290376e-05,
      "loss": 3.1514,
      "step": 77550
    },
    {
      "epoch": 8.352168765471962,
      "grad_norm": 0.8150763511657715,
      "learning_rate": 9.917034802284236e-05,
      "loss": 3.1758,
      "step": 77600
    },
    {
      "epoch": 8.35755031751157,
      "grad_norm": 0.8031736612319946,
      "learning_rate": 9.884710699278094e-05,
      "loss": 3.1516,
      "step": 77650
    },
    {
      "epoch": 8.36293186955118,
      "grad_norm": 0.8593947887420654,
      "learning_rate": 9.852386596271953e-05,
      "loss": 3.1794,
      "step": 77700
    },
    {
      "epoch": 8.368313421590786,
      "grad_norm": 0.8277099132537842,
      "learning_rate": 9.82006249326581e-05,
      "loss": 3.1719,
      "step": 77750
    },
    {
      "epoch": 8.373694973630395,
      "grad_norm": 0.8137666583061218,
      "learning_rate": 9.788384872319793e-05,
      "loss": 3.1698,
      "step": 77800
    },
    {
      "epoch": 8.379076525670003,
      "grad_norm": 0.8309085965156555,
      "learning_rate": 9.75606076931365e-05,
      "loss": 3.1919,
      "step": 77850
    },
    {
      "epoch": 8.384458077709612,
      "grad_norm": 0.8904669880867004,
      "learning_rate": 9.723736666307508e-05,
      "loss": 3.1698,
      "step": 77900
    },
    {
      "epoch": 8.38983962974922,
      "grad_norm": 0.8591392040252686,
      "learning_rate": 9.691412563301368e-05,
      "loss": 3.1868,
      "step": 77950
    },
    {
      "epoch": 8.395221181788829,
      "grad_norm": 0.8529878854751587,
      "learning_rate": 9.659088460295227e-05,
      "loss": 3.1844,
      "step": 78000
    },
    {
      "epoch": 8.395221181788829,
      "eval_accuracy": 0.391199615890311,
      "eval_loss": 3.326892614364624,
      "eval_runtime": 190.2408,
      "eval_samples_per_second": 94.675,
      "eval_steps_per_second": 5.919,
      "step": 78000
    },
    {
      "epoch": 8.400602733828435,
      "grad_norm": 0.830120861530304,
      "learning_rate": 9.626764357289084e-05,
      "loss": 3.1732,
      "step": 78050
    },
    {
      "epoch": 8.405984285868044,
      "grad_norm": 0.8445336222648621,
      "learning_rate": 9.594440254282943e-05,
      "loss": 3.1828,
      "step": 78100
    },
    {
      "epoch": 8.411365837907653,
      "grad_norm": 0.8456713557243347,
      "learning_rate": 9.562116151276802e-05,
      "loss": 3.1827,
      "step": 78150
    },
    {
      "epoch": 8.416747389947261,
      "grad_norm": 0.8553755283355713,
      "learning_rate": 9.52979204827066e-05,
      "loss": 3.1817,
      "step": 78200
    },
    {
      "epoch": 8.42212894198687,
      "grad_norm": 0.8498519659042358,
      "learning_rate": 9.497467945264518e-05,
      "loss": 3.1866,
      "step": 78250
    },
    {
      "epoch": 8.427510494026476,
      "grad_norm": 0.800201416015625,
      "learning_rate": 9.465143842258377e-05,
      "loss": 3.1888,
      "step": 78300
    },
    {
      "epoch": 8.432892046066085,
      "grad_norm": 0.805806040763855,
      "learning_rate": 9.432819739252234e-05,
      "loss": 3.1892,
      "step": 78350
    },
    {
      "epoch": 8.438273598105694,
      "grad_norm": 0.8127837181091309,
      "learning_rate": 9.400495636246093e-05,
      "loss": 3.1929,
      "step": 78400
    },
    {
      "epoch": 8.443655150145302,
      "grad_norm": 0.8124808073043823,
      "learning_rate": 9.368171533239952e-05,
      "loss": 3.1893,
      "step": 78450
    },
    {
      "epoch": 8.44903670218491,
      "grad_norm": 0.8213273882865906,
      "learning_rate": 9.335847430233811e-05,
      "loss": 3.1824,
      "step": 78500
    },
    {
      "epoch": 8.45441825422452,
      "grad_norm": 0.8079547882080078,
      "learning_rate": 9.303523327227668e-05,
      "loss": 3.1789,
      "step": 78550
    },
    {
      "epoch": 8.459799806264126,
      "grad_norm": 0.8386607766151428,
      "learning_rate": 9.271199224221527e-05,
      "loss": 3.1948,
      "step": 78600
    },
    {
      "epoch": 8.465181358303735,
      "grad_norm": 0.8313483595848083,
      "learning_rate": 9.238875121215387e-05,
      "loss": 3.1881,
      "step": 78650
    },
    {
      "epoch": 8.470562910343343,
      "grad_norm": 0.8849664926528931,
      "learning_rate": 9.206551018209243e-05,
      "loss": 3.1565,
      "step": 78700
    },
    {
      "epoch": 8.475944462382952,
      "grad_norm": 0.848349928855896,
      "learning_rate": 9.174226915203103e-05,
      "loss": 3.1625,
      "step": 78750
    },
    {
      "epoch": 8.48132601442256,
      "grad_norm": 0.8258644938468933,
      "learning_rate": 9.141902812196961e-05,
      "loss": 3.199,
      "step": 78800
    },
    {
      "epoch": 8.486707566462167,
      "grad_norm": 0.7915391325950623,
      "learning_rate": 9.109578709190818e-05,
      "loss": 3.1731,
      "step": 78850
    },
    {
      "epoch": 8.492089118501776,
      "grad_norm": 0.8821637034416199,
      "learning_rate": 9.077254606184677e-05,
      "loss": 3.1698,
      "step": 78900
    },
    {
      "epoch": 8.497470670541384,
      "grad_norm": 0.8676961660385132,
      "learning_rate": 9.044930503178537e-05,
      "loss": 3.1903,
      "step": 78950
    },
    {
      "epoch": 8.502852222580993,
      "grad_norm": 0.795990526676178,
      "learning_rate": 9.012606400172395e-05,
      "loss": 3.1873,
      "step": 79000
    },
    {
      "epoch": 8.502852222580993,
      "eval_accuracy": 0.3916445494659113,
      "eval_loss": 3.3241238594055176,
      "eval_runtime": 189.6529,
      "eval_samples_per_second": 94.968,
      "eval_steps_per_second": 5.937,
      "step": 79000
    },
    {
      "epoch": 8.508233774620601,
      "grad_norm": 0.8447521924972534,
      "learning_rate": 8.980282297166253e-05,
      "loss": 3.1787,
      "step": 79050
    },
    {
      "epoch": 8.513615326660208,
      "grad_norm": 0.8267421126365662,
      "learning_rate": 8.947958194160111e-05,
      "loss": 3.1733,
      "step": 79100
    },
    {
      "epoch": 8.518996878699816,
      "grad_norm": 0.849238932132721,
      "learning_rate": 8.91563409115397e-05,
      "loss": 3.1767,
      "step": 79150
    },
    {
      "epoch": 8.524378430739425,
      "grad_norm": 0.8542905449867249,
      "learning_rate": 8.883309988147827e-05,
      "loss": 3.1816,
      "step": 79200
    },
    {
      "epoch": 8.529759982779034,
      "grad_norm": 0.8247508406639099,
      "learning_rate": 8.850985885141687e-05,
      "loss": 3.1622,
      "step": 79250
    },
    {
      "epoch": 8.535141534818642,
      "grad_norm": 0.7993038892745972,
      "learning_rate": 8.818661782135545e-05,
      "loss": 3.165,
      "step": 79300
    },
    {
      "epoch": 8.54052308685825,
      "grad_norm": 0.8211236000061035,
      "learning_rate": 8.786337679129403e-05,
      "loss": 3.171,
      "step": 79350
    },
    {
      "epoch": 8.545904638897857,
      "grad_norm": 0.8362860083580017,
      "learning_rate": 8.754013576123261e-05,
      "loss": 3.1685,
      "step": 79400
    },
    {
      "epoch": 8.551286190937466,
      "grad_norm": 0.7837921977043152,
      "learning_rate": 8.72168947311712e-05,
      "loss": 3.1916,
      "step": 79450
    },
    {
      "epoch": 8.556667742977075,
      "grad_norm": 0.8205451965332031,
      "learning_rate": 8.68936537011098e-05,
      "loss": 3.1756,
      "step": 79500
    },
    {
      "epoch": 8.562049295016683,
      "grad_norm": 0.8241888284683228,
      "learning_rate": 8.657041267104837e-05,
      "loss": 3.1733,
      "step": 79550
    },
    {
      "epoch": 8.567430847056292,
      "grad_norm": 0.8107258677482605,
      "learning_rate": 8.624717164098696e-05,
      "loss": 3.1702,
      "step": 79600
    },
    {
      "epoch": 8.572812399095898,
      "grad_norm": 0.8234117031097412,
      "learning_rate": 8.592393061092554e-05,
      "loss": 3.178,
      "step": 79650
    },
    {
      "epoch": 8.578193951135507,
      "grad_norm": 0.8076245784759521,
      "learning_rate": 8.560068958086412e-05,
      "loss": 3.17,
      "step": 79700
    },
    {
      "epoch": 8.583575503175116,
      "grad_norm": 0.8706180453300476,
      "learning_rate": 8.52774485508027e-05,
      "loss": 3.1744,
      "step": 79750
    },
    {
      "epoch": 8.588957055214724,
      "grad_norm": 0.8408890962600708,
      "learning_rate": 8.49542075207413e-05,
      "loss": 3.1731,
      "step": 79800
    },
    {
      "epoch": 8.594338607254333,
      "grad_norm": 0.8679055571556091,
      "learning_rate": 8.463096649067987e-05,
      "loss": 3.186,
      "step": 79850
    },
    {
      "epoch": 8.599720159293941,
      "grad_norm": 0.8492652177810669,
      "learning_rate": 8.430772546061846e-05,
      "loss": 3.178,
      "step": 79900
    },
    {
      "epoch": 8.605101711333548,
      "grad_norm": 0.8724219799041748,
      "learning_rate": 8.398448443055704e-05,
      "loss": 3.1754,
      "step": 79950
    },
    {
      "epoch": 8.610483263373157,
      "grad_norm": 0.8504689931869507,
      "learning_rate": 8.366124340049564e-05,
      "loss": 3.1793,
      "step": 80000
    },
    {
      "epoch": 8.610483263373157,
      "eval_accuracy": 0.39190140489270475,
      "eval_loss": 3.3210363388061523,
      "eval_runtime": 190.1892,
      "eval_samples_per_second": 94.7,
      "eval_steps_per_second": 5.92,
      "step": 80000
    },
    {
      "epoch": 8.615864815412765,
      "grad_norm": 0.8523551225662231,
      "learning_rate": 8.33380023704342e-05,
      "loss": 3.1905,
      "step": 80050
    },
    {
      "epoch": 8.621246367452374,
      "grad_norm": 0.791414737701416,
      "learning_rate": 8.30147613403728e-05,
      "loss": 3.1662,
      "step": 80100
    },
    {
      "epoch": 8.626627919491982,
      "grad_norm": 0.8727241158485413,
      "learning_rate": 8.269152031031138e-05,
      "loss": 3.1811,
      "step": 80150
    },
    {
      "epoch": 8.632009471531589,
      "grad_norm": 0.9096086621284485,
      "learning_rate": 8.236827928024996e-05,
      "loss": 3.1804,
      "step": 80200
    },
    {
      "epoch": 8.637391023571197,
      "grad_norm": 0.8429610729217529,
      "learning_rate": 8.204503825018854e-05,
      "loss": 3.1711,
      "step": 80250
    },
    {
      "epoch": 8.642772575610806,
      "grad_norm": 0.7959967851638794,
      "learning_rate": 8.172179722012714e-05,
      "loss": 3.1894,
      "step": 80300
    },
    {
      "epoch": 8.648154127650415,
      "grad_norm": 0.8352919220924377,
      "learning_rate": 8.13985561900657e-05,
      "loss": 3.1665,
      "step": 80350
    },
    {
      "epoch": 8.653535679690023,
      "grad_norm": 0.8290886878967285,
      "learning_rate": 8.10753151600043e-05,
      "loss": 3.1865,
      "step": 80400
    },
    {
      "epoch": 8.658917231729632,
      "grad_norm": 0.8746451139450073,
      "learning_rate": 8.07520741299429e-05,
      "loss": 3.1717,
      "step": 80450
    },
    {
      "epoch": 8.664298783769238,
      "grad_norm": 0.8428299427032471,
      "learning_rate": 8.042883309988148e-05,
      "loss": 3.1847,
      "step": 80500
    },
    {
      "epoch": 8.669680335808847,
      "grad_norm": 0.8098769783973694,
      "learning_rate": 8.010559206982006e-05,
      "loss": 3.1725,
      "step": 80550
    },
    {
      "epoch": 8.675061887848456,
      "grad_norm": 0.8425123691558838,
      "learning_rate": 7.978235103975864e-05,
      "loss": 3.1805,
      "step": 80600
    },
    {
      "epoch": 8.680443439888064,
      "grad_norm": 0.8586872816085815,
      "learning_rate": 7.945911000969723e-05,
      "loss": 3.1738,
      "step": 80650
    },
    {
      "epoch": 8.685824991927673,
      "grad_norm": 0.8304820656776428,
      "learning_rate": 7.91358689796358e-05,
      "loss": 3.1983,
      "step": 80700
    },
    {
      "epoch": 8.69120654396728,
      "grad_norm": 0.8227179050445557,
      "learning_rate": 7.88126279495744e-05,
      "loss": 3.1884,
      "step": 80750
    },
    {
      "epoch": 8.696588096006888,
      "grad_norm": 0.8312771320343018,
      "learning_rate": 7.848938691951298e-05,
      "loss": 3.1918,
      "step": 80800
    },
    {
      "epoch": 8.701969648046497,
      "grad_norm": 0.8414456844329834,
      "learning_rate": 7.816614588945156e-05,
      "loss": 3.1785,
      "step": 80850
    },
    {
      "epoch": 8.707351200086105,
      "grad_norm": 0.8723168969154358,
      "learning_rate": 7.784290485939014e-05,
      "loss": 3.1784,
      "step": 80900
    },
    {
      "epoch": 8.712732752125714,
      "grad_norm": 0.8794534206390381,
      "learning_rate": 7.751966382932873e-05,
      "loss": 3.1905,
      "step": 80950
    },
    {
      "epoch": 8.718114304165322,
      "grad_norm": 0.8242953419685364,
      "learning_rate": 7.719642279926731e-05,
      "loss": 3.1808,
      "step": 81000
    },
    {
      "epoch": 8.718114304165322,
      "eval_accuracy": 0.39228864378487555,
      "eval_loss": 3.3183116912841797,
      "eval_runtime": 190.3059,
      "eval_samples_per_second": 94.642,
      "eval_steps_per_second": 5.917,
      "step": 81000
    },
    {
      "epoch": 8.723495856204929,
      "grad_norm": 0.8565047979354858,
      "learning_rate": 7.68731817692059e-05,
      "loss": 3.1873,
      "step": 81050
    },
    {
      "epoch": 8.728877408244538,
      "grad_norm": 0.8532828092575073,
      "learning_rate": 7.654994073914448e-05,
      "loss": 3.156,
      "step": 81100
    },
    {
      "epoch": 8.734258960284146,
      "grad_norm": 0.8730028867721558,
      "learning_rate": 7.622669970908307e-05,
      "loss": 3.1856,
      "step": 81150
    },
    {
      "epoch": 8.739640512323755,
      "grad_norm": 0.8408501148223877,
      "learning_rate": 7.590345867902164e-05,
      "loss": 3.1878,
      "step": 81200
    },
    {
      "epoch": 8.745022064363363,
      "grad_norm": 0.8017849922180176,
      "learning_rate": 7.558021764896023e-05,
      "loss": 3.1806,
      "step": 81250
    },
    {
      "epoch": 8.75040361640297,
      "grad_norm": 0.8024411201477051,
      "learning_rate": 7.525697661889883e-05,
      "loss": 3.1858,
      "step": 81300
    },
    {
      "epoch": 8.755785168442578,
      "grad_norm": 0.8082074522972107,
      "learning_rate": 7.493373558883741e-05,
      "loss": 3.1692,
      "step": 81350
    },
    {
      "epoch": 8.761166720482187,
      "grad_norm": 0.8132904171943665,
      "learning_rate": 7.461049455877599e-05,
      "loss": 3.1776,
      "step": 81400
    },
    {
      "epoch": 8.766548272521796,
      "grad_norm": 0.804818332195282,
      "learning_rate": 7.428725352871457e-05,
      "loss": 3.1755,
      "step": 81450
    },
    {
      "epoch": 8.771929824561404,
      "grad_norm": 0.8590632081031799,
      "learning_rate": 7.396401249865315e-05,
      "loss": 3.1636,
      "step": 81500
    },
    {
      "epoch": 8.777311376601011,
      "grad_norm": 0.9342987537384033,
      "learning_rate": 7.364077146859173e-05,
      "loss": 3.1618,
      "step": 81550
    },
    {
      "epoch": 8.78269292864062,
      "grad_norm": 0.8472897410392761,
      "learning_rate": 7.331753043853033e-05,
      "loss": 3.1742,
      "step": 81600
    },
    {
      "epoch": 8.788074480680228,
      "grad_norm": 0.8631134629249573,
      "learning_rate": 7.299428940846891e-05,
      "loss": 3.1907,
      "step": 81650
    },
    {
      "epoch": 8.793456032719837,
      "grad_norm": 0.8849440813064575,
      "learning_rate": 7.267104837840749e-05,
      "loss": 3.1814,
      "step": 81700
    },
    {
      "epoch": 8.798837584759445,
      "grad_norm": 0.799471914768219,
      "learning_rate": 7.234780734834607e-05,
      "loss": 3.1809,
      "step": 81750
    },
    {
      "epoch": 8.804219136799054,
      "grad_norm": 0.819151759147644,
      "learning_rate": 7.202456631828465e-05,
      "loss": 3.1945,
      "step": 81800
    },
    {
      "epoch": 8.80960068883866,
      "grad_norm": 0.8566768765449524,
      "learning_rate": 7.170779010882447e-05,
      "loss": 3.1669,
      "step": 81850
    },
    {
      "epoch": 8.814982240878269,
      "grad_norm": 0.8203052878379822,
      "learning_rate": 7.138454907876305e-05,
      "loss": 3.1767,
      "step": 81900
    },
    {
      "epoch": 8.820363792917878,
      "grad_norm": 0.8461906313896179,
      "learning_rate": 7.106130804870164e-05,
      "loss": 3.1907,
      "step": 81950
    },
    {
      "epoch": 8.825745344957486,
      "grad_norm": 0.8182411789894104,
      "learning_rate": 7.073806701864023e-05,
      "loss": 3.184,
      "step": 82000
    },
    {
      "epoch": 8.825745344957486,
      "eval_accuracy": 0.39260243332488376,
      "eval_loss": 3.3135645389556885,
      "eval_runtime": 190.05,
      "eval_samples_per_second": 94.77,
      "eval_steps_per_second": 5.925,
      "step": 82000
    },
    {
      "epoch": 8.831126896997095,
      "grad_norm": 0.9185845851898193,
      "learning_rate": 7.041482598857881e-05,
      "loss": 3.1655,
      "step": 82050
    },
    {
      "epoch": 8.836508449036701,
      "grad_norm": 0.8070856928825378,
      "learning_rate": 7.009804977911862e-05,
      "loss": 3.1813,
      "step": 82100
    },
    {
      "epoch": 8.84189000107631,
      "grad_norm": 0.8231473565101624,
      "learning_rate": 6.97748087490572e-05,
      "loss": 3.1735,
      "step": 82150
    },
    {
      "epoch": 8.847271553115919,
      "grad_norm": 0.8550606966018677,
      "learning_rate": 6.945156771899579e-05,
      "loss": 3.177,
      "step": 82200
    },
    {
      "epoch": 8.852653105155527,
      "grad_norm": 0.8077003359794617,
      "learning_rate": 6.912832668893437e-05,
      "loss": 3.1841,
      "step": 82250
    },
    {
      "epoch": 8.858034657195136,
      "grad_norm": 0.8284718990325928,
      "learning_rate": 6.880508565887297e-05,
      "loss": 3.1753,
      "step": 82300
    },
    {
      "epoch": 8.863416209234742,
      "grad_norm": 0.849946141242981,
      "learning_rate": 6.848184462881155e-05,
      "loss": 3.1666,
      "step": 82350
    },
    {
      "epoch": 8.868797761274351,
      "grad_norm": 0.8250811696052551,
      "learning_rate": 6.815860359875013e-05,
      "loss": 3.1868,
      "step": 82400
    },
    {
      "epoch": 8.87417931331396,
      "grad_norm": 0.8054350018501282,
      "learning_rate": 6.783536256868871e-05,
      "loss": 3.1743,
      "step": 82450
    },
    {
      "epoch": 8.879560865353568,
      "grad_norm": 0.9101194739341736,
      "learning_rate": 6.75121215386273e-05,
      "loss": 3.1733,
      "step": 82500
    },
    {
      "epoch": 8.884942417393177,
      "grad_norm": 0.850450873374939,
      "learning_rate": 6.718888050856589e-05,
      "loss": 3.183,
      "step": 82550
    },
    {
      "epoch": 8.890323969432785,
      "grad_norm": 0.8298824429512024,
      "learning_rate": 6.686563947850447e-05,
      "loss": 3.1947,
      "step": 82600
    },
    {
      "epoch": 8.895705521472392,
      "grad_norm": 0.8521466255187988,
      "learning_rate": 6.654239844844305e-05,
      "loss": 3.1704,
      "step": 82650
    },
    {
      "epoch": 8.901087073512,
      "grad_norm": 0.9018093347549438,
      "learning_rate": 6.621915741838163e-05,
      "loss": 3.1852,
      "step": 82700
    },
    {
      "epoch": 8.906468625551609,
      "grad_norm": 0.8763890266418457,
      "learning_rate": 6.589591638832021e-05,
      "loss": 3.1739,
      "step": 82750
    },
    {
      "epoch": 8.911850177591218,
      "grad_norm": 0.7935939431190491,
      "learning_rate": 6.557267535825881e-05,
      "loss": 3.1752,
      "step": 82800
    },
    {
      "epoch": 8.917231729630826,
      "grad_norm": 0.7923632264137268,
      "learning_rate": 6.524943432819739e-05,
      "loss": 3.183,
      "step": 82850
    },
    {
      "epoch": 8.922613281670433,
      "grad_norm": 0.8742818832397461,
      "learning_rate": 6.492619329813597e-05,
      "loss": 3.1782,
      "step": 82900
    },
    {
      "epoch": 8.927994833710041,
      "grad_norm": 0.7885937094688416,
      "learning_rate": 6.460295226807455e-05,
      "loss": 3.1513,
      "step": 82950
    },
    {
      "epoch": 8.93337638574965,
      "grad_norm": 0.8109541535377502,
      "learning_rate": 6.427971123801313e-05,
      "loss": 3.1707,
      "step": 83000
    },
    {
      "epoch": 8.93337638574965,
      "eval_accuracy": 0.39286874155291707,
      "eval_loss": 3.3116352558135986,
      "eval_runtime": 190.2742,
      "eval_samples_per_second": 94.658,
      "eval_steps_per_second": 5.918,
      "step": 83000
    },
    {
      "epoch": 8.938757937789259,
      "grad_norm": 0.8382235765457153,
      "learning_rate": 6.395647020795173e-05,
      "loss": 3.1807,
      "step": 83050
    },
    {
      "epoch": 8.944139489828867,
      "grad_norm": 0.8553214073181152,
      "learning_rate": 6.363322917789031e-05,
      "loss": 3.1595,
      "step": 83100
    },
    {
      "epoch": 8.949521041868476,
      "grad_norm": 0.8265877962112427,
      "learning_rate": 6.330998814782889e-05,
      "loss": 3.1902,
      "step": 83150
    },
    {
      "epoch": 8.954902593908082,
      "grad_norm": 0.8195887207984924,
      "learning_rate": 6.298674711776748e-05,
      "loss": 3.1713,
      "step": 83200
    },
    {
      "epoch": 8.960284145947691,
      "grad_norm": 0.8596498370170593,
      "learning_rate": 6.266350608770606e-05,
      "loss": 3.1549,
      "step": 83250
    },
    {
      "epoch": 8.9656656979873,
      "grad_norm": 0.8457815647125244,
      "learning_rate": 6.234026505764465e-05,
      "loss": 3.1965,
      "step": 83300
    },
    {
      "epoch": 8.971047250026908,
      "grad_norm": 0.8061408400535583,
      "learning_rate": 6.201702402758323e-05,
      "loss": 3.1929,
      "step": 83350
    },
    {
      "epoch": 8.976428802066517,
      "grad_norm": 0.8716901540756226,
      "learning_rate": 6.169378299752181e-05,
      "loss": 3.168,
      "step": 83400
    },
    {
      "epoch": 8.981810354106123,
      "grad_norm": 0.8381494283676147,
      "learning_rate": 6.13705419674604e-05,
      "loss": 3.1884,
      "step": 83450
    },
    {
      "epoch": 8.987191906145732,
      "grad_norm": 0.8308653831481934,
      "learning_rate": 6.104730093739898e-05,
      "loss": 3.1778,
      "step": 83500
    },
    {
      "epoch": 8.99257345818534,
      "grad_norm": 0.8332216739654541,
      "learning_rate": 6.072405990733757e-05,
      "loss": 3.1869,
      "step": 83550
    },
    {
      "epoch": 8.997955010224949,
      "grad_norm": 0.8378949761390686,
      "learning_rate": 6.040081887727615e-05,
      "loss": 3.1762,
      "step": 83600
    },
    {
      "epoch": 9.003336562264558,
      "grad_norm": 0.7988559007644653,
      "learning_rate": 6.007757784721473e-05,
      "loss": 3.1398,
      "step": 83650
    },
    {
      "epoch": 9.008718114304166,
      "grad_norm": 0.8361743688583374,
      "learning_rate": 5.975433681715332e-05,
      "loss": 3.1088,
      "step": 83700
    },
    {
      "epoch": 9.014099666343773,
      "grad_norm": 0.8134690523147583,
      "learning_rate": 5.94310957870919e-05,
      "loss": 3.1164,
      "step": 83750
    },
    {
      "epoch": 9.019481218383381,
      "grad_norm": 0.8642722964286804,
      "learning_rate": 5.910785475703049e-05,
      "loss": 3.1213,
      "step": 83800
    },
    {
      "epoch": 9.02486277042299,
      "grad_norm": 0.8392188549041748,
      "learning_rate": 5.878461372696907e-05,
      "loss": 3.1174,
      "step": 83850
    },
    {
      "epoch": 9.030244322462599,
      "grad_norm": 0.8445099592208862,
      "learning_rate": 5.846137269690765e-05,
      "loss": 3.1109,
      "step": 83900
    },
    {
      "epoch": 9.035625874502207,
      "grad_norm": 0.8090319633483887,
      "learning_rate": 5.813813166684624e-05,
      "loss": 3.127,
      "step": 83950
    },
    {
      "epoch": 9.041007426541814,
      "grad_norm": 0.8555005192756653,
      "learning_rate": 5.781489063678482e-05,
      "loss": 3.1079,
      "step": 84000
    },
    {
      "epoch": 9.041007426541814,
      "eval_accuracy": 0.3928621137267604,
      "eval_loss": 3.314976453781128,
      "eval_runtime": 189.6659,
      "eval_samples_per_second": 94.962,
      "eval_steps_per_second": 5.937,
      "step": 84000
    },
    {
      "epoch": 9.046388978581422,
      "grad_norm": 0.8304621577262878,
      "learning_rate": 5.749164960672341e-05,
      "loss": 3.1113,
      "step": 84050
    },
    {
      "epoch": 9.051770530621031,
      "grad_norm": 0.8350916504859924,
      "learning_rate": 5.716840857666199e-05,
      "loss": 3.1085,
      "step": 84100
    },
    {
      "epoch": 9.05715208266064,
      "grad_norm": 0.8908973932266235,
      "learning_rate": 5.684516754660057e-05,
      "loss": 3.1033,
      "step": 84150
    },
    {
      "epoch": 9.062533634700248,
      "grad_norm": 0.8434808254241943,
      "learning_rate": 5.652192651653916e-05,
      "loss": 3.1137,
      "step": 84200
    },
    {
      "epoch": 9.067915186739857,
      "grad_norm": 0.8735474944114685,
      "learning_rate": 5.619868548647774e-05,
      "loss": 3.1264,
      "step": 84250
    },
    {
      "epoch": 9.073296738779463,
      "grad_norm": 0.8355602622032166,
      "learning_rate": 5.5875444456416335e-05,
      "loss": 3.1179,
      "step": 84300
    },
    {
      "epoch": 9.078678290819072,
      "grad_norm": 0.8927472829818726,
      "learning_rate": 5.5552203426354916e-05,
      "loss": 3.1317,
      "step": 84350
    },
    {
      "epoch": 9.08405984285868,
      "grad_norm": 0.8360982537269592,
      "learning_rate": 5.52289623962935e-05,
      "loss": 3.1285,
      "step": 84400
    },
    {
      "epoch": 9.089441394898289,
      "grad_norm": 0.83144211769104,
      "learning_rate": 5.4905721366232085e-05,
      "loss": 3.1324,
      "step": 84450
    },
    {
      "epoch": 9.094822946937898,
      "grad_norm": 0.8676847219467163,
      "learning_rate": 5.4582480336170666e-05,
      "loss": 3.1091,
      "step": 84500
    },
    {
      "epoch": 9.100204498977504,
      "grad_norm": 0.8253108263015747,
      "learning_rate": 5.4259239306109254e-05,
      "loss": 3.1109,
      "step": 84550
    },
    {
      "epoch": 9.105586051017113,
      "grad_norm": 0.8409557938575745,
      "learning_rate": 5.3935998276047835e-05,
      "loss": 3.113,
      "step": 84600
    },
    {
      "epoch": 9.110967603056721,
      "grad_norm": 0.8219881057739258,
      "learning_rate": 5.3612757245986416e-05,
      "loss": 3.1058,
      "step": 84650
    },
    {
      "epoch": 9.11634915509633,
      "grad_norm": 0.8523204922676086,
      "learning_rate": 5.3289516215925004e-05,
      "loss": 3.1252,
      "step": 84700
    },
    {
      "epoch": 9.121730707135939,
      "grad_norm": 0.7867476344108582,
      "learning_rate": 5.2966275185863585e-05,
      "loss": 3.1111,
      "step": 84750
    },
    {
      "epoch": 9.127112259175545,
      "grad_norm": 0.836459755897522,
      "learning_rate": 5.264303415580217e-05,
      "loss": 3.1178,
      "step": 84800
    },
    {
      "epoch": 9.132493811215154,
      "grad_norm": 0.8583942651748657,
      "learning_rate": 5.2319793125740754e-05,
      "loss": 3.1498,
      "step": 84850
    },
    {
      "epoch": 9.137875363254762,
      "grad_norm": 0.9006932973861694,
      "learning_rate": 5.1996552095679336e-05,
      "loss": 3.1049,
      "step": 84900
    },
    {
      "epoch": 9.143256915294371,
      "grad_norm": 0.8754859566688538,
      "learning_rate": 5.1673311065617923e-05,
      "loss": 3.1452,
      "step": 84950
    },
    {
      "epoch": 9.14863846733398,
      "grad_norm": 0.8559167385101318,
      "learning_rate": 5.1350070035556505e-05,
      "loss": 3.1234,
      "step": 85000
    },
    {
      "epoch": 9.14863846733398,
      "eval_accuracy": 0.39311168941007024,
      "eval_loss": 3.3131887912750244,
      "eval_runtime": 190.0658,
      "eval_samples_per_second": 94.762,
      "eval_steps_per_second": 5.924,
      "step": 85000
    },
    {
      "epoch": 9.154020019373588,
      "grad_norm": 0.8461971282958984,
      "learning_rate": 5.102682900549509e-05,
      "loss": 3.1064,
      "step": 85050
    },
    {
      "epoch": 9.159401571413195,
      "grad_norm": 0.8278712630271912,
      "learning_rate": 5.0703587975433674e-05,
      "loss": 3.1183,
      "step": 85100
    },
    {
      "epoch": 9.164783123452803,
      "grad_norm": 0.8924991488456726,
      "learning_rate": 5.0380346945372255e-05,
      "loss": 3.1227,
      "step": 85150
    },
    {
      "epoch": 9.170164675492412,
      "grad_norm": 0.8371621966362,
      "learning_rate": 5.005710591531085e-05,
      "loss": 3.1216,
      "step": 85200
    },
    {
      "epoch": 9.17554622753202,
      "grad_norm": 0.8644618391990662,
      "learning_rate": 4.973386488524943e-05,
      "loss": 3.13,
      "step": 85250
    },
    {
      "epoch": 9.180927779571629,
      "grad_norm": 0.8401750326156616,
      "learning_rate": 4.941062385518802e-05,
      "loss": 3.1058,
      "step": 85300
    },
    {
      "epoch": 9.186309331611236,
      "grad_norm": 0.8433226943016052,
      "learning_rate": 4.90873828251266e-05,
      "loss": 3.1391,
      "step": 85350
    },
    {
      "epoch": 9.191690883650844,
      "grad_norm": 0.8032107353210449,
      "learning_rate": 4.876414179506518e-05,
      "loss": 3.1312,
      "step": 85400
    },
    {
      "epoch": 9.197072435690453,
      "grad_norm": 0.8685449361801147,
      "learning_rate": 4.844090076500377e-05,
      "loss": 3.1244,
      "step": 85450
    },
    {
      "epoch": 9.202453987730062,
      "grad_norm": 0.8396304845809937,
      "learning_rate": 4.811765973494235e-05,
      "loss": 3.1094,
      "step": 85500
    },
    {
      "epoch": 9.20783553976967,
      "grad_norm": 0.8684375286102295,
      "learning_rate": 4.779441870488094e-05,
      "loss": 3.1341,
      "step": 85550
    },
    {
      "epoch": 9.213217091809279,
      "grad_norm": 0.8813304305076599,
      "learning_rate": 4.747117767481952e-05,
      "loss": 3.1213,
      "step": 85600
    },
    {
      "epoch": 9.218598643848885,
      "grad_norm": 0.8566760420799255,
      "learning_rate": 4.71479366447581e-05,
      "loss": 3.1163,
      "step": 85650
    },
    {
      "epoch": 9.223980195888494,
      "grad_norm": 0.8688989877700806,
      "learning_rate": 4.682469561469669e-05,
      "loss": 3.1331,
      "step": 85700
    },
    {
      "epoch": 9.229361747928102,
      "grad_norm": 0.7989726662635803,
      "learning_rate": 4.650145458463527e-05,
      "loss": 3.1396,
      "step": 85750
    },
    {
      "epoch": 9.234743299967711,
      "grad_norm": 0.8513889908790588,
      "learning_rate": 4.6178213554573856e-05,
      "loss": 3.1168,
      "step": 85800
    },
    {
      "epoch": 9.24012485200732,
      "grad_norm": 0.8649311661720276,
      "learning_rate": 4.585497252451244e-05,
      "loss": 3.1319,
      "step": 85850
    },
    {
      "epoch": 9.245506404046926,
      "grad_norm": 0.8756608366966248,
      "learning_rate": 4.553173149445102e-05,
      "loss": 3.1269,
      "step": 85900
    },
    {
      "epoch": 9.250887956086535,
      "grad_norm": 0.8677488565444946,
      "learning_rate": 4.520849046438961e-05,
      "loss": 3.1357,
      "step": 85950
    },
    {
      "epoch": 9.256269508126143,
      "grad_norm": 0.8723577857017517,
      "learning_rate": 4.488524943432819e-05,
      "loss": 3.1071,
      "step": 86000
    },
    {
      "epoch": 9.256269508126143,
      "eval_accuracy": 0.3933045482859409,
      "eval_loss": 3.3104920387268066,
      "eval_runtime": 190.0748,
      "eval_samples_per_second": 94.757,
      "eval_steps_per_second": 5.924,
      "step": 86000
    },
    {
      "epoch": 9.261651060165752,
      "grad_norm": 0.864126443862915,
      "learning_rate": 4.456200840426678e-05,
      "loss": 3.128,
      "step": 86050
    },
    {
      "epoch": 9.26703261220536,
      "grad_norm": 0.8417099118232727,
      "learning_rate": 4.4238767374205363e-05,
      "loss": 3.1496,
      "step": 86100
    },
    {
      "epoch": 9.272414164244967,
      "grad_norm": 0.8650234937667847,
      "learning_rate": 4.3915526344143945e-05,
      "loss": 3.1263,
      "step": 86150
    },
    {
      "epoch": 9.277795716284576,
      "grad_norm": 0.8400073051452637,
      "learning_rate": 4.359228531408253e-05,
      "loss": 3.1133,
      "step": 86200
    },
    {
      "epoch": 9.283177268324184,
      "grad_norm": 0.8600847721099854,
      "learning_rate": 4.3269044284021114e-05,
      "loss": 3.1291,
      "step": 86250
    },
    {
      "epoch": 9.288558820363793,
      "grad_norm": 0.8189371824264526,
      "learning_rate": 4.295226807456093e-05,
      "loss": 3.138,
      "step": 86300
    },
    {
      "epoch": 9.293940372403402,
      "grad_norm": 0.8316525816917419,
      "learning_rate": 4.2629027044499515e-05,
      "loss": 3.133,
      "step": 86350
    },
    {
      "epoch": 9.29932192444301,
      "grad_norm": 0.848581850528717,
      "learning_rate": 4.2305786014438096e-05,
      "loss": 3.1306,
      "step": 86400
    },
    {
      "epoch": 9.304703476482617,
      "grad_norm": 0.8404590487480164,
      "learning_rate": 4.198254498437668e-05,
      "loss": 3.1026,
      "step": 86450
    },
    {
      "epoch": 9.310085028522225,
      "grad_norm": 0.8248619437217712,
      "learning_rate": 4.1659303954315265e-05,
      "loss": 3.1139,
      "step": 86500
    },
    {
      "epoch": 9.315466580561834,
      "grad_norm": 0.8430253863334656,
      "learning_rate": 4.1336062924253846e-05,
      "loss": 3.1354,
      "step": 86550
    },
    {
      "epoch": 9.320848132601443,
      "grad_norm": 0.8569157123565674,
      "learning_rate": 4.1012821894192434e-05,
      "loss": 3.1214,
      "step": 86600
    },
    {
      "epoch": 9.326229684641051,
      "grad_norm": 0.847260057926178,
      "learning_rate": 4.0689580864131015e-05,
      "loss": 3.1245,
      "step": 86650
    },
    {
      "epoch": 9.331611236680658,
      "grad_norm": 0.7993084192276001,
      "learning_rate": 4.0366339834069596e-05,
      "loss": 3.1293,
      "step": 86700
    },
    {
      "epoch": 9.336992788720266,
      "grad_norm": 0.8383640646934509,
      "learning_rate": 4.0043098804008184e-05,
      "loss": 3.1383,
      "step": 86750
    },
    {
      "epoch": 9.342374340759875,
      "grad_norm": 0.8617927432060242,
      "learning_rate": 3.9719857773946765e-05,
      "loss": 3.1332,
      "step": 86800
    },
    {
      "epoch": 9.347755892799483,
      "grad_norm": 0.875795841217041,
      "learning_rate": 3.939661674388536e-05,
      "loss": 3.1298,
      "step": 86850
    },
    {
      "epoch": 9.353137444839092,
      "grad_norm": 0.8759364485740662,
      "learning_rate": 3.9079840534425166e-05,
      "loss": 3.1301,
      "step": 86900
    },
    {
      "epoch": 9.3585189968787,
      "grad_norm": 0.9027230739593506,
      "learning_rate": 3.875659950436375e-05,
      "loss": 3.1224,
      "step": 86950
    },
    {
      "epoch": 9.363900548918307,
      "grad_norm": 0.8533556461334229,
      "learning_rate": 3.843335847430233e-05,
      "loss": 3.1274,
      "step": 87000
    },
    {
      "epoch": 9.363900548918307,
      "eval_accuracy": 0.39366712297257783,
      "eval_loss": 3.309715986251831,
      "eval_runtime": 189.9351,
      "eval_samples_per_second": 94.827,
      "eval_steps_per_second": 5.928,
      "step": 87000
    },
    {
      "epoch": 9.369282100957916,
      "grad_norm": 0.8597930073738098,
      "learning_rate": 3.811011744424092e-05,
      "loss": 3.1182,
      "step": 87050
    },
    {
      "epoch": 9.374663652997524,
      "grad_norm": 0.824130654335022,
      "learning_rate": 3.7786876414179504e-05,
      "loss": 3.1152,
      "step": 87100
    },
    {
      "epoch": 9.380045205037133,
      "grad_norm": 0.9420901536941528,
      "learning_rate": 3.7463635384118085e-05,
      "loss": 3.1219,
      "step": 87150
    },
    {
      "epoch": 9.385426757076742,
      "grad_norm": 0.867396891117096,
      "learning_rate": 3.714039435405667e-05,
      "loss": 3.1304,
      "step": 87200
    },
    {
      "epoch": 9.390808309116348,
      "grad_norm": 0.8354865908622742,
      "learning_rate": 3.6817153323995254e-05,
      "loss": 3.1241,
      "step": 87250
    },
    {
      "epoch": 9.396189861155957,
      "grad_norm": 0.8731216788291931,
      "learning_rate": 3.649391229393384e-05,
      "loss": 3.132,
      "step": 87300
    },
    {
      "epoch": 9.401571413195565,
      "grad_norm": 0.836256206035614,
      "learning_rate": 3.617067126387242e-05,
      "loss": 3.1377,
      "step": 87350
    },
    {
      "epoch": 9.406952965235174,
      "grad_norm": 0.8706014752388,
      "learning_rate": 3.5847430233811004e-05,
      "loss": 3.1254,
      "step": 87400
    },
    {
      "epoch": 9.412334517274783,
      "grad_norm": 0.8802419900894165,
      "learning_rate": 3.552418920374959e-05,
      "loss": 3.123,
      "step": 87450
    },
    {
      "epoch": 9.417716069314391,
      "grad_norm": 0.8520860075950623,
      "learning_rate": 3.520094817368818e-05,
      "loss": 3.1359,
      "step": 87500
    },
    {
      "epoch": 9.423097621353998,
      "grad_norm": 0.8559780120849609,
      "learning_rate": 3.487770714362676e-05,
      "loss": 3.1278,
      "step": 87550
    },
    {
      "epoch": 9.428479173393606,
      "grad_norm": 0.8601959943771362,
      "learning_rate": 3.455446611356535e-05,
      "loss": 3.1389,
      "step": 87600
    },
    {
      "epoch": 9.433860725433215,
      "grad_norm": 0.8391619920730591,
      "learning_rate": 3.423122508350393e-05,
      "loss": 3.1332,
      "step": 87650
    },
    {
      "epoch": 9.439242277472824,
      "grad_norm": 0.836699366569519,
      "learning_rate": 3.390798405344251e-05,
      "loss": 3.1241,
      "step": 87700
    },
    {
      "epoch": 9.444623829512432,
      "grad_norm": 0.8568204641342163,
      "learning_rate": 3.35847430233811e-05,
      "loss": 3.1295,
      "step": 87750
    },
    {
      "epoch": 9.450005381552039,
      "grad_norm": 0.852116584777832,
      "learning_rate": 3.326150199331968e-05,
      "loss": 3.1281,
      "step": 87800
    },
    {
      "epoch": 9.455386933591647,
      "grad_norm": 0.8235575556755066,
      "learning_rate": 3.293826096325827e-05,
      "loss": 3.1329,
      "step": 87850
    },
    {
      "epoch": 9.460768485631256,
      "grad_norm": 0.8330269455909729,
      "learning_rate": 3.261501993319685e-05,
      "loss": 3.1445,
      "step": 87900
    },
    {
      "epoch": 9.466150037670864,
      "grad_norm": 0.8570313453674316,
      "learning_rate": 3.229177890313544e-05,
      "loss": 3.1174,
      "step": 87950
    },
    {
      "epoch": 9.471531589710473,
      "grad_norm": 0.9033732414245605,
      "learning_rate": 3.196853787307402e-05,
      "loss": 3.1093,
      "step": 88000
    },
    {
      "epoch": 9.471531589710473,
      "eval_accuracy": 0.39385759148491667,
      "eval_loss": 3.306760311126709,
      "eval_runtime": 190.0676,
      "eval_samples_per_second": 94.761,
      "eval_steps_per_second": 5.924,
      "step": 88000
    },
    {
      "epoch": 9.476913141750082,
      "grad_norm": 0.8972622752189636,
      "learning_rate": 3.1645296843012606e-05,
      "loss": 3.132,
      "step": 88050
    },
    {
      "epoch": 9.482294693789688,
      "grad_norm": 0.8442611694335938,
      "learning_rate": 3.132205581295119e-05,
      "loss": 3.1208,
      "step": 88100
    },
    {
      "epoch": 9.487676245829297,
      "grad_norm": 0.8502057790756226,
      "learning_rate": 3.099881478288977e-05,
      "loss": 3.1412,
      "step": 88150
    },
    {
      "epoch": 9.493057797868905,
      "grad_norm": 0.8361905217170715,
      "learning_rate": 3.0675573752828356e-05,
      "loss": 3.1306,
      "step": 88200
    },
    {
      "epoch": 9.498439349908514,
      "grad_norm": 0.8201267719268799,
      "learning_rate": 3.035233272276694e-05,
      "loss": 3.1164,
      "step": 88250
    },
    {
      "epoch": 9.503820901948123,
      "grad_norm": 0.8912597298622131,
      "learning_rate": 3.0029091692705525e-05,
      "loss": 3.1229,
      "step": 88300
    },
    {
      "epoch": 9.50920245398773,
      "grad_norm": 0.860805869102478,
      "learning_rate": 2.970585066264411e-05,
      "loss": 3.1157,
      "step": 88350
    },
    {
      "epoch": 9.514584006027338,
      "grad_norm": 0.8319489359855652,
      "learning_rate": 2.938260963258269e-05,
      "loss": 3.1396,
      "step": 88400
    },
    {
      "epoch": 9.519965558066946,
      "grad_norm": 0.8614979982376099,
      "learning_rate": 2.9059368602521275e-05,
      "loss": 3.1263,
      "step": 88450
    },
    {
      "epoch": 9.525347110106555,
      "grad_norm": 0.8343462944030762,
      "learning_rate": 2.8736127572459863e-05,
      "loss": 3.1215,
      "step": 88500
    },
    {
      "epoch": 9.530728662146164,
      "grad_norm": 0.8397179245948792,
      "learning_rate": 2.8412886542398448e-05,
      "loss": 3.1153,
      "step": 88550
    },
    {
      "epoch": 9.536110214185772,
      "grad_norm": 0.8414610624313354,
      "learning_rate": 2.8089645512337032e-05,
      "loss": 3.1327,
      "step": 88600
    },
    {
      "epoch": 9.541491766225379,
      "grad_norm": 0.8669750094413757,
      "learning_rate": 2.7766404482275613e-05,
      "loss": 3.1362,
      "step": 88650
    },
    {
      "epoch": 9.546873318264987,
      "grad_norm": 0.8965323567390442,
      "learning_rate": 2.7443163452214198e-05,
      "loss": 3.1408,
      "step": 88700
    },
    {
      "epoch": 9.552254870304596,
      "grad_norm": 0.8437132239341736,
      "learning_rate": 2.7119922422152782e-05,
      "loss": 3.1243,
      "step": 88750
    },
    {
      "epoch": 9.557636422344205,
      "grad_norm": 0.8450977206230164,
      "learning_rate": 2.6796681392091367e-05,
      "loss": 3.1336,
      "step": 88800
    },
    {
      "epoch": 9.563017974383813,
      "grad_norm": 0.8513481616973877,
      "learning_rate": 2.6473440362029955e-05,
      "loss": 3.1359,
      "step": 88850
    },
    {
      "epoch": 9.56839952642342,
      "grad_norm": 0.9124820232391357,
      "learning_rate": 2.6150199331968536e-05,
      "loss": 3.1171,
      "step": 88900
    },
    {
      "epoch": 9.573781078463028,
      "grad_norm": 0.8457428216934204,
      "learning_rate": 2.582695830190712e-05,
      "loss": 3.1219,
      "step": 88950
    },
    {
      "epoch": 9.579162630502637,
      "grad_norm": 0.8853180408477783,
      "learning_rate": 2.5503717271845705e-05,
      "loss": 3.113,
      "step": 89000
    },
    {
      "epoch": 9.579162630502637,
      "eval_accuracy": 0.394095758615006,
      "eval_loss": 3.3056252002716064,
      "eval_runtime": 189.9647,
      "eval_samples_per_second": 94.812,
      "eval_steps_per_second": 5.927,
      "step": 89000
    },
    {
      "epoch": 9.584544182542245,
      "grad_norm": 0.8555046916007996,
      "learning_rate": 2.5186941062385518e-05,
      "loss": 3.133,
      "step": 89050
    },
    {
      "epoch": 9.589925734581854,
      "grad_norm": 0.8598777055740356,
      "learning_rate": 2.4863700032324103e-05,
      "loss": 3.1373,
      "step": 89100
    },
    {
      "epoch": 9.59530728662146,
      "grad_norm": 0.8013734221458435,
      "learning_rate": 2.4540459002262687e-05,
      "loss": 3.1255,
      "step": 89150
    },
    {
      "epoch": 9.60068883866107,
      "grad_norm": 0.8821795582771301,
      "learning_rate": 2.4217217972201268e-05,
      "loss": 3.1234,
      "step": 89200
    },
    {
      "epoch": 9.606070390700678,
      "grad_norm": 0.8431457877159119,
      "learning_rate": 2.3893976942139853e-05,
      "loss": 3.1264,
      "step": 89250
    },
    {
      "epoch": 9.611451942740286,
      "grad_norm": 0.8205217123031616,
      "learning_rate": 2.3570735912078437e-05,
      "loss": 3.1299,
      "step": 89300
    },
    {
      "epoch": 9.616833494779895,
      "grad_norm": 0.8434730172157288,
      "learning_rate": 2.324749488201702e-05,
      "loss": 3.1281,
      "step": 89350
    },
    {
      "epoch": 9.622215046819504,
      "grad_norm": 0.9389562010765076,
      "learning_rate": 2.292425385195561e-05,
      "loss": 3.1272,
      "step": 89400
    },
    {
      "epoch": 9.62759659885911,
      "grad_norm": 0.8556050062179565,
      "learning_rate": 2.260101282189419e-05,
      "loss": 3.1288,
      "step": 89450
    },
    {
      "epoch": 9.632978150898719,
      "grad_norm": 0.8911100625991821,
      "learning_rate": 2.2277771791832775e-05,
      "loss": 3.1402,
      "step": 89500
    },
    {
      "epoch": 9.638359702938327,
      "grad_norm": 0.8732587695121765,
      "learning_rate": 2.195453076177136e-05,
      "loss": 3.1307,
      "step": 89550
    },
    {
      "epoch": 9.643741254977936,
      "grad_norm": 0.8422968983650208,
      "learning_rate": 2.1631289731709944e-05,
      "loss": 3.1357,
      "step": 89600
    },
    {
      "epoch": 9.649122807017545,
      "grad_norm": 0.8346823453903198,
      "learning_rate": 2.130804870164853e-05,
      "loss": 3.1263,
      "step": 89650
    },
    {
      "epoch": 9.654504359057151,
      "grad_norm": 0.9292513132095337,
      "learning_rate": 2.098480767158711e-05,
      "loss": 3.1138,
      "step": 89700
    },
    {
      "epoch": 9.65988591109676,
      "grad_norm": 0.8220443725585938,
      "learning_rate": 2.0661566641525694e-05,
      "loss": 3.1267,
      "step": 89750
    },
    {
      "epoch": 9.665267463136368,
      "grad_norm": 0.8924703598022461,
      "learning_rate": 2.0338325611464282e-05,
      "loss": 3.118,
      "step": 89800
    },
    {
      "epoch": 9.670649015175977,
      "grad_norm": 0.8519506454467773,
      "learning_rate": 2.0015084581402867e-05,
      "loss": 3.1242,
      "step": 89850
    },
    {
      "epoch": 9.676030567215586,
      "grad_norm": 0.9123126864433289,
      "learning_rate": 1.969184355134145e-05,
      "loss": 3.1287,
      "step": 89900
    },
    {
      "epoch": 9.681412119255192,
      "grad_norm": 0.8319849967956543,
      "learning_rate": 1.9368602521280032e-05,
      "loss": 3.1121,
      "step": 89950
    },
    {
      "epoch": 9.6867936712948,
      "grad_norm": 0.8460592031478882,
      "learning_rate": 1.9045361491218617e-05,
      "loss": 3.1358,
      "step": 90000
    },
    {
      "epoch": 9.6867936712948,
      "eval_accuracy": 0.3944938627959582,
      "eval_loss": 3.301800489425659,
      "eval_runtime": 190.3536,
      "eval_samples_per_second": 94.619,
      "eval_steps_per_second": 5.915,
      "step": 90000
    },
    {
      "epoch": 9.69217522333441,
      "grad_norm": 0.8394421935081482,
      "learning_rate": 1.87221204611572e-05,
      "loss": 3.1336,
      "step": 90050
    },
    {
      "epoch": 9.697556775374018,
      "grad_norm": 0.8398434519767761,
      "learning_rate": 1.8398879431095786e-05,
      "loss": 3.1372,
      "step": 90100
    },
    {
      "epoch": 9.702938327413626,
      "grad_norm": 0.8492758870124817,
      "learning_rate": 1.807563840103437e-05,
      "loss": 3.1154,
      "step": 90150
    },
    {
      "epoch": 9.708319879453235,
      "grad_norm": 0.8635022640228271,
      "learning_rate": 1.7752397370972955e-05,
      "loss": 3.123,
      "step": 90200
    },
    {
      "epoch": 9.713701431492842,
      "grad_norm": 0.8306649923324585,
      "learning_rate": 1.742915634091154e-05,
      "loss": 3.132,
      "step": 90250
    },
    {
      "epoch": 9.71908298353245,
      "grad_norm": 0.8687105178833008,
      "learning_rate": 1.7105915310850124e-05,
      "loss": 3.1437,
      "step": 90300
    },
    {
      "epoch": 9.724464535572059,
      "grad_norm": 0.866300106048584,
      "learning_rate": 1.6782674280788708e-05,
      "loss": 3.1372,
      "step": 90350
    },
    {
      "epoch": 9.729846087611667,
      "grad_norm": 0.8367854952812195,
      "learning_rate": 1.645943325072729e-05,
      "loss": 3.1361,
      "step": 90400
    },
    {
      "epoch": 9.735227639651276,
      "grad_norm": 0.8828139305114746,
      "learning_rate": 1.6136192220665877e-05,
      "loss": 3.1254,
      "step": 90450
    },
    {
      "epoch": 9.740609191690883,
      "grad_norm": 0.8941412568092346,
      "learning_rate": 1.5812951190604458e-05,
      "loss": 3.1302,
      "step": 90500
    },
    {
      "epoch": 9.745990743730491,
      "grad_norm": 0.8547508716583252,
      "learning_rate": 1.5489710160543043e-05,
      "loss": 3.1411,
      "step": 90550
    },
    {
      "epoch": 9.7513722957701,
      "grad_norm": 0.934457540512085,
      "learning_rate": 1.5166469130481629e-05,
      "loss": 3.1158,
      "step": 90600
    },
    {
      "epoch": 9.756753847809708,
      "grad_norm": 0.8271318078041077,
      "learning_rate": 1.4843228100420212e-05,
      "loss": 3.1193,
      "step": 90650
    },
    {
      "epoch": 9.762135399849317,
      "grad_norm": 0.852136492729187,
      "learning_rate": 1.4519987070358796e-05,
      "loss": 3.1212,
      "step": 90700
    },
    {
      "epoch": 9.767516951888926,
      "grad_norm": 0.7947787642478943,
      "learning_rate": 1.419674604029738e-05,
      "loss": 3.1235,
      "step": 90750
    },
    {
      "epoch": 9.772898503928532,
      "grad_norm": 0.8722898364067078,
      "learning_rate": 1.3873505010235965e-05,
      "loss": 3.1233,
      "step": 90800
    },
    {
      "epoch": 9.77828005596814,
      "grad_norm": 0.8331910371780396,
      "learning_rate": 1.355026398017455e-05,
      "loss": 3.1266,
      "step": 90850
    },
    {
      "epoch": 9.78366160800775,
      "grad_norm": 0.8797382116317749,
      "learning_rate": 1.3227022950113132e-05,
      "loss": 3.1331,
      "step": 90900
    },
    {
      "epoch": 9.789043160047358,
      "grad_norm": 0.8739795684814453,
      "learning_rate": 1.2903781920051719e-05,
      "loss": 3.1392,
      "step": 90950
    },
    {
      "epoch": 9.794424712086967,
      "grad_norm": 0.875079870223999,
      "learning_rate": 1.2580540889990301e-05,
      "loss": 3.1167,
      "step": 91000
    },
    {
      "epoch": 9.794424712086967,
      "eval_accuracy": 0.39460055993179205,
      "eval_loss": 3.300231695175171,
      "eval_runtime": 190.4226,
      "eval_samples_per_second": 94.584,
      "eval_steps_per_second": 5.913,
      "step": 91000
    },
    {
      "epoch": 9.799806264126573,
      "grad_norm": 0.8346747159957886,
      "learning_rate": 1.2257299859928886e-05,
      "loss": 3.124,
      "step": 91050
    },
    {
      "epoch": 9.805187816166182,
      "grad_norm": 0.8464721441268921,
      "learning_rate": 1.193405882986747e-05,
      "loss": 3.1131,
      "step": 91100
    },
    {
      "epoch": 9.81056936820579,
      "grad_norm": 0.8748368620872498,
      "learning_rate": 1.1610817799806053e-05,
      "loss": 3.1294,
      "step": 91150
    },
    {
      "epoch": 9.815950920245399,
      "grad_norm": 0.9805678129196167,
      "learning_rate": 1.128757676974464e-05,
      "loss": 3.1281,
      "step": 91200
    },
    {
      "epoch": 9.821332472285007,
      "grad_norm": 0.8663314580917358,
      "learning_rate": 1.0964335739683222e-05,
      "loss": 3.1222,
      "step": 91250
    },
    {
      "epoch": 9.826714024324616,
      "grad_norm": 0.8092943429946899,
      "learning_rate": 1.0641094709621807e-05,
      "loss": 3.1147,
      "step": 91300
    },
    {
      "epoch": 9.832095576364223,
      "grad_norm": 0.8579533696174622,
      "learning_rate": 1.0317853679560393e-05,
      "loss": 3.1318,
      "step": 91350
    },
    {
      "epoch": 9.837477128403831,
      "grad_norm": 0.8417325019836426,
      "learning_rate": 9.994612649498976e-06,
      "loss": 3.13,
      "step": 91400
    },
    {
      "epoch": 9.84285868044344,
      "grad_norm": 0.8230124711990356,
      "learning_rate": 9.67137161943756e-06,
      "loss": 3.1251,
      "step": 91450
    },
    {
      "epoch": 9.848240232483048,
      "grad_norm": 0.8273383378982544,
      "learning_rate": 9.348130589376145e-06,
      "loss": 3.1219,
      "step": 91500
    },
    {
      "epoch": 9.853621784522657,
      "grad_norm": 0.8730839490890503,
      "learning_rate": 9.024889559314727e-06,
      "loss": 3.1383,
      "step": 91550
    },
    {
      "epoch": 9.859003336562264,
      "grad_norm": 0.8343610167503357,
      "learning_rate": 8.701648529253312e-06,
      "loss": 3.1193,
      "step": 91600
    },
    {
      "epoch": 9.864384888601872,
      "grad_norm": 0.841031551361084,
      "learning_rate": 8.378407499191896e-06,
      "loss": 3.1326,
      "step": 91650
    },
    {
      "epoch": 9.869766440641481,
      "grad_norm": 0.8363505601882935,
      "learning_rate": 8.055166469130481e-06,
      "loss": 3.1063,
      "step": 91700
    },
    {
      "epoch": 9.87514799268109,
      "grad_norm": 0.831885814666748,
      "learning_rate": 7.731925439069065e-06,
      "loss": 3.1336,
      "step": 91750
    },
    {
      "epoch": 9.880529544720698,
      "grad_norm": 0.8267219066619873,
      "learning_rate": 7.408684409007649e-06,
      "loss": 3.1453,
      "step": 91800
    },
    {
      "epoch": 9.885911096760307,
      "grad_norm": 0.8478328585624695,
      "learning_rate": 7.085443378946234e-06,
      "loss": 3.1343,
      "step": 91850
    },
    {
      "epoch": 9.891292648799913,
      "grad_norm": 0.8537425994873047,
      "learning_rate": 6.762202348884817e-06,
      "loss": 3.1174,
      "step": 91900
    },
    {
      "epoch": 9.896674200839522,
      "grad_norm": 0.8562872409820557,
      "learning_rate": 6.438961318823403e-06,
      "loss": 3.1165,
      "step": 91950
    },
    {
      "epoch": 9.90205575287913,
      "grad_norm": 0.82090824842453,
      "learning_rate": 6.115720288761986e-06,
      "loss": 3.111,
      "step": 92000
    },
    {
      "epoch": 9.90205575287913,
      "eval_accuracy": 0.3948433991360574,
      "eval_loss": 3.2990994453430176,
      "eval_runtime": 189.7158,
      "eval_samples_per_second": 94.937,
      "eval_steps_per_second": 5.935,
      "step": 92000
    },
    {
      "epoch": 9.907437304918739,
      "grad_norm": 0.8207184672355652,
      "learning_rate": 5.792479258700571e-06,
      "loss": 3.1236,
      "step": 92050
    },
    {
      "epoch": 9.912818856958348,
      "grad_norm": 0.8829610347747803,
      "learning_rate": 5.469238228639154e-06,
      "loss": 3.1241,
      "step": 92100
    },
    {
      "epoch": 9.918200408997954,
      "grad_norm": 0.841016948223114,
      "learning_rate": 5.145997198577739e-06,
      "loss": 3.1228,
      "step": 92150
    },
    {
      "epoch": 9.923581961037563,
      "grad_norm": 0.8278433680534363,
      "learning_rate": 4.822756168516323e-06,
      "loss": 3.1172,
      "step": 92200
    },
    {
      "epoch": 9.928963513077171,
      "grad_norm": 0.891380786895752,
      "learning_rate": 4.499515138454908e-06,
      "loss": 3.123,
      "step": 92250
    },
    {
      "epoch": 9.93434506511678,
      "grad_norm": 0.8488532304763794,
      "learning_rate": 4.1762741083934915e-06,
      "loss": 3.1094,
      "step": 92300
    },
    {
      "epoch": 9.939726617156388,
      "grad_norm": 0.8263406157493591,
      "learning_rate": 3.853033078332076e-06,
      "loss": 3.1469,
      "step": 92350
    },
    {
      "epoch": 9.945108169195997,
      "grad_norm": 0.8684765696525574,
      "learning_rate": 3.52979204827066e-06,
      "loss": 3.1383,
      "step": 92400
    },
    {
      "epoch": 9.950489721235604,
      "grad_norm": 0.8202494382858276,
      "learning_rate": 3.206551018209244e-06,
      "loss": 3.1068,
      "step": 92450
    },
    {
      "epoch": 9.955871273275212,
      "grad_norm": 0.894985020160675,
      "learning_rate": 2.889774808749057e-06,
      "loss": 3.1559,
      "step": 92500
    },
    {
      "epoch": 9.961252825314821,
      "grad_norm": 0.8657666444778442,
      "learning_rate": 2.566533778687641e-06,
      "loss": 3.1297,
      "step": 92550
    },
    {
      "epoch": 9.96663437735443,
      "grad_norm": 0.8371118307113647,
      "learning_rate": 2.243292748626225e-06,
      "loss": 3.1296,
      "step": 92600
    },
    {
      "epoch": 9.972015929394038,
      "grad_norm": 0.8734837174415588,
      "learning_rate": 1.9200517185648096e-06,
      "loss": 3.1226,
      "step": 92650
    },
    {
      "epoch": 9.977397481433645,
      "grad_norm": 0.7993928790092468,
      "learning_rate": 1.596810688503394e-06,
      "loss": 3.1258,
      "step": 92700
    },
    {
      "epoch": 9.982779033473253,
      "grad_norm": 0.8332442045211792,
      "learning_rate": 1.2735696584419782e-06,
      "loss": 3.1128,
      "step": 92750
    },
    {
      "epoch": 9.988160585512862,
      "grad_norm": 0.9027200937271118,
      "learning_rate": 9.503286283805623e-07,
      "loss": 3.1115,
      "step": 92800
    },
    {
      "epoch": 9.99354213755247,
      "grad_norm": 0.8293177485466003,
      "learning_rate": 6.270875983191466e-07,
      "loss": 3.1236,
      "step": 92850
    },
    {
      "epoch": 9.998923689592079,
      "grad_norm": 0.8855078220367432,
      "learning_rate": 3.038465682577308e-07,
      "loss": 3.1244,
      "step": 92900
    },
    {
      "epoch": 10.0,
      "step": 92910,
      "total_flos": 7.76821211136e+17,
      "train_loss": 3.4522010074032865,
      "train_runtime": 81258.1448,
      "train_samples_per_second": 36.587,
      "train_steps_per_second": 1.143
    }
  ],
  "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": 7.76821211136e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}