{
  "best_metric": 3.3028476238250732,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_495/checkpoint-90000",
  "epoch": 10.0,
  "eval_steps": 1000,
  "global_step": 92910,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005381552039608223,
      "grad_norm": 1.6117655038833618,
      "learning_rate": 0.00027,
      "loss": 8.7843,
      "step": 50
    },
    {
      "epoch": 0.010763104079216447,
      "grad_norm": 1.1765116453170776,
      "learning_rate": 0.00057,
      "loss": 6.9711,
      "step": 100
    },
    {
      "epoch": 0.01614465611882467,
      "grad_norm": 1.6633868217468262,
      "learning_rate": 0.0005997090830729447,
      "loss": 6.5373,
      "step": 150
    },
    {
      "epoch": 0.021526208158432893,
      "grad_norm": 1.8836766481399536,
      "learning_rate": 0.0005993858420428833,
      "loss": 6.2429,
      "step": 200
    },
    {
      "epoch": 0.026907760198041114,
      "grad_norm": 1.5037431716918945,
      "learning_rate": 0.0005990626010128218,
      "loss": 6.0966,
      "step": 250
    },
    {
      "epoch": 0.03228931223764934,
      "grad_norm": 2.2753920555114746,
      "learning_rate": 0.0005987393599827604,
      "loss": 5.953,
      "step": 300
    },
    {
      "epoch": 0.03767086427725756,
      "grad_norm": 3.0884439945220947,
      "learning_rate": 0.000598416118952699,
      "loss": 5.869,
      "step": 350
    },
    {
      "epoch": 0.04305241631686579,
      "grad_norm": 1.4899078607559204,
      "learning_rate": 0.0005980928779226376,
      "loss": 5.8032,
      "step": 400
    },
    {
      "epoch": 0.048433968356474004,
      "grad_norm": 1.125209093093872,
      "learning_rate": 0.0005977696368925762,
      "loss": 5.7355,
      "step": 450
    },
    {
      "epoch": 0.05381552039608223,
      "grad_norm": 1.960442066192627,
      "learning_rate": 0.0005974463958625148,
      "loss": 5.6583,
      "step": 500
    },
    {
      "epoch": 0.05919707243569045,
      "grad_norm": 1.044176697731018,
      "learning_rate": 0.0005971231548324534,
      "loss": 5.5711,
      "step": 550
    },
    {
      "epoch": 0.06457862447529868,
      "grad_norm": 1.6373082399368286,
      "learning_rate": 0.0005967999138023919,
      "loss": 5.5099,
      "step": 600
    },
    {
      "epoch": 0.0699601765149069,
      "grad_norm": 1.8287220001220703,
      "learning_rate": 0.0005964766727723304,
      "loss": 5.4242,
      "step": 650
    },
    {
      "epoch": 0.07534172855451512,
      "grad_norm": 1.3936927318572998,
      "learning_rate": 0.0005961534317422691,
      "loss": 5.3571,
      "step": 700
    },
    {
      "epoch": 0.08072328059412334,
      "grad_norm": 0.9875015020370483,
      "learning_rate": 0.0005958301907122077,
      "loss": 5.3186,
      "step": 750
    },
    {
      "epoch": 0.08610483263373157,
      "grad_norm": 1.0579822063446045,
      "learning_rate": 0.0005955069496821463,
      "loss": 5.2725,
      "step": 800
    },
    {
      "epoch": 0.09148638467333979,
      "grad_norm": 1.3142889738082886,
      "learning_rate": 0.0005951837086520848,
      "loss": 5.2145,
      "step": 850
    },
    {
      "epoch": 0.09686793671294801,
      "grad_norm": 1.207129716873169,
      "learning_rate": 0.0005948604676220235,
      "loss": 5.1809,
      "step": 900
    },
    {
      "epoch": 0.10224948875255624,
      "grad_norm": 1.3603378534317017,
      "learning_rate": 0.000594537226591962,
      "loss": 5.1166,
      "step": 950
    },
    {
      "epoch": 0.10763104079216446,
      "grad_norm": 1.0100822448730469,
      "learning_rate": 0.0005942139855619007,
      "loss": 5.0872,
      "step": 1000
    },
    {
      "epoch": 0.10763104079216446,
      "eval_accuracy": 0.22757434275053265,
      "eval_loss": 5.0165934562683105,
      "eval_runtime": 186.2978,
      "eval_samples_per_second": 96.679,
      "eval_steps_per_second": 6.044,
      "step": 1000
    },
    {
      "epoch": 0.11301259283177269,
      "grad_norm": 0.9639580249786377,
      "learning_rate": 0.0005938907445318392,
      "loss": 5.0357,
      "step": 1050
    },
    {
      "epoch": 0.1183941448713809,
      "grad_norm": 0.8208778500556946,
      "learning_rate": 0.0005935675035017777,
      "loss": 5.0074,
      "step": 1100
    },
    {
      "epoch": 0.12377569691098914,
      "grad_norm": 0.9588396549224854,
      "learning_rate": 0.0005932442624717164,
      "loss": 4.9928,
      "step": 1150
    },
    {
      "epoch": 0.12915724895059735,
      "grad_norm": 0.8537307977676392,
      "learning_rate": 0.0005929210214416549,
      "loss": 4.9625,
      "step": 1200
    },
    {
      "epoch": 0.13453880099020557,
      "grad_norm": 1.1293931007385254,
      "learning_rate": 0.0005925977804115936,
      "loss": 4.9296,
      "step": 1250
    },
    {
      "epoch": 0.1399203530298138,
      "grad_norm": 1.039612054824829,
      "learning_rate": 0.0005922745393815321,
      "loss": 4.904,
      "step": 1300
    },
    {
      "epoch": 0.14530190506942203,
      "grad_norm": 0.986505925655365,
      "learning_rate": 0.0005919512983514707,
      "loss": 4.8739,
      "step": 1350
    },
    {
      "epoch": 0.15068345710903025,
      "grad_norm": 0.9047093391418457,
      "learning_rate": 0.0005916280573214093,
      "loss": 4.8174,
      "step": 1400
    },
    {
      "epoch": 0.15606500914863847,
      "grad_norm": 0.770999014377594,
      "learning_rate": 0.0005913048162913478,
      "loss": 4.8181,
      "step": 1450
    },
    {
      "epoch": 0.16144656118824668,
      "grad_norm": 0.9082644581794739,
      "learning_rate": 0.0005909815752612864,
      "loss": 4.8036,
      "step": 1500
    },
    {
      "epoch": 0.1668281132278549,
      "grad_norm": 0.8358648419380188,
      "learning_rate": 0.000590658334231225,
      "loss": 4.7798,
      "step": 1550
    },
    {
      "epoch": 0.17220966526746315,
      "grad_norm": 1.0487140417099,
      "learning_rate": 0.0005903350932011637,
      "loss": 4.7364,
      "step": 1600
    },
    {
      "epoch": 0.17759121730707136,
      "grad_norm": 1.1815866231918335,
      "learning_rate": 0.0005900118521711022,
      "loss": 4.7392,
      "step": 1650
    },
    {
      "epoch": 0.18297276934667958,
      "grad_norm": 0.9304591417312622,
      "learning_rate": 0.0005896886111410408,
      "loss": 4.7118,
      "step": 1700
    },
    {
      "epoch": 0.1883543213862878,
      "grad_norm": 1.0020699501037598,
      "learning_rate": 0.0005893653701109793,
      "loss": 4.6978,
      "step": 1750
    },
    {
      "epoch": 0.19373587342589602,
      "grad_norm": 0.8822704553604126,
      "learning_rate": 0.000589042129080918,
      "loss": 4.6769,
      "step": 1800
    },
    {
      "epoch": 0.19911742546550426,
      "grad_norm": 0.944671630859375,
      "learning_rate": 0.0005887188880508566,
      "loss": 4.6617,
      "step": 1850
    },
    {
      "epoch": 0.20449897750511248,
      "grad_norm": 1.0103189945220947,
      "learning_rate": 0.0005883956470207951,
      "loss": 4.6347,
      "step": 1900
    },
    {
      "epoch": 0.2098805295447207,
      "grad_norm": 0.9612853527069092,
      "learning_rate": 0.0005880724059907337,
      "loss": 4.5904,
      "step": 1950
    },
    {
      "epoch": 0.2152620815843289,
      "grad_norm": 0.9868743419647217,
      "learning_rate": 0.0005877491649606723,
      "loss": 4.58,
      "step": 2000
    },
    {
      "epoch": 0.2152620815843289,
      "eval_accuracy": 0.27043877621644785,
      "eval_loss": 4.506998538970947,
      "eval_runtime": 184.7948,
      "eval_samples_per_second": 97.465,
      "eval_steps_per_second": 6.093,
      "step": 2000
    },
    {
      "epoch": 0.22064363362393713,
      "grad_norm": 0.8025125861167908,
      "learning_rate": 0.0005874259239306109,
      "loss": 4.5503,
      "step": 2050
    },
    {
      "epoch": 0.22602518566354537,
      "grad_norm": 1.0225203037261963,
      "learning_rate": 0.0005871026829005494,
      "loss": 4.5634,
      "step": 2100
    },
    {
      "epoch": 0.2314067377031536,
      "grad_norm": 0.9046829342842102,
      "learning_rate": 0.0005867794418704881,
      "loss": 4.5213,
      "step": 2150
    },
    {
      "epoch": 0.2367882897427618,
      "grad_norm": 0.7576098442077637,
      "learning_rate": 0.0005864562008404266,
      "loss": 4.5007,
      "step": 2200
    },
    {
      "epoch": 0.24216984178237003,
      "grad_norm": 0.9998608231544495,
      "learning_rate": 0.0005861329598103653,
      "loss": 4.4993,
      "step": 2250
    },
    {
      "epoch": 0.24755139382197827,
      "grad_norm": 1.168628215789795,
      "learning_rate": 0.0005858097187803038,
      "loss": 4.4896,
      "step": 2300
    },
    {
      "epoch": 0.2529329458615865,
      "grad_norm": 0.8030685782432556,
      "learning_rate": 0.0005854864777502423,
      "loss": 4.4601,
      "step": 2350
    },
    {
      "epoch": 0.2583144979011947,
      "grad_norm": 1.0477194786071777,
      "learning_rate": 0.000585163236720181,
      "loss": 4.4508,
      "step": 2400
    },
    {
      "epoch": 0.2636960499408029,
      "grad_norm": 0.86094731092453,
      "learning_rate": 0.0005848399956901195,
      "loss": 4.428,
      "step": 2450
    },
    {
      "epoch": 0.26907760198041114,
      "grad_norm": 0.9053926467895508,
      "learning_rate": 0.0005845167546600582,
      "loss": 4.4063,
      "step": 2500
    },
    {
      "epoch": 0.27445915402001936,
      "grad_norm": 0.7576990127563477,
      "learning_rate": 0.0005841935136299967,
      "loss": 4.4131,
      "step": 2550
    },
    {
      "epoch": 0.2798407060596276,
      "grad_norm": 0.7213003635406494,
      "learning_rate": 0.0005838702725999353,
      "loss": 4.3988,
      "step": 2600
    },
    {
      "epoch": 0.2852222580992358,
      "grad_norm": 0.9967924356460571,
      "learning_rate": 0.0005835470315698739,
      "loss": 4.3945,
      "step": 2650
    },
    {
      "epoch": 0.29060381013884407,
      "grad_norm": 0.7419942617416382,
      "learning_rate": 0.0005832237905398125,
      "loss": 4.3636,
      "step": 2700
    },
    {
      "epoch": 0.2959853621784523,
      "grad_norm": 0.7668341994285583,
      "learning_rate": 0.0005829005495097511,
      "loss": 4.3658,
      "step": 2750
    },
    {
      "epoch": 0.3013669142180605,
      "grad_norm": 0.7742299437522888,
      "learning_rate": 0.0005825773084796896,
      "loss": 4.334,
      "step": 2800
    },
    {
      "epoch": 0.3067484662576687,
      "grad_norm": 0.9014106392860413,
      "learning_rate": 0.0005822540674496282,
      "loss": 4.327,
      "step": 2850
    },
    {
      "epoch": 0.31213001829727693,
      "grad_norm": 0.6771566867828369,
      "learning_rate": 0.0005819308264195668,
      "loss": 4.3486,
      "step": 2900
    },
    {
      "epoch": 0.31751157033688515,
      "grad_norm": 0.712129533290863,
      "learning_rate": 0.0005816075853895054,
      "loss": 4.3111,
      "step": 2950
    },
    {
      "epoch": 0.32289312237649337,
      "grad_norm": 0.6686953902244568,
      "learning_rate": 0.000581284344359444,
      "loss": 4.3079,
      "step": 3000
    },
    {
      "epoch": 0.32289312237649337,
      "eval_accuracy": 0.2981868766651734,
      "eval_loss": 4.241795063018799,
      "eval_runtime": 184.6997,
      "eval_samples_per_second": 97.515,
      "eval_steps_per_second": 6.096,
      "step": 3000
    },
    {
      "epoch": 0.3282746744161016,
      "grad_norm": 0.6886458992958069,
      "learning_rate": 0.0005809611033293825,
      "loss": 4.3161,
      "step": 3050
    },
    {
      "epoch": 0.3336562264557098,
      "grad_norm": 0.7012606859207153,
      "learning_rate": 0.0005806378622993211,
      "loss": 4.2908,
      "step": 3100
    },
    {
      "epoch": 0.3390377784953181,
      "grad_norm": 0.6854538321495056,
      "learning_rate": 0.0005803146212692597,
      "loss": 4.2775,
      "step": 3150
    },
    {
      "epoch": 0.3444193305349263,
      "grad_norm": 0.8339408040046692,
      "learning_rate": 0.0005799913802391983,
      "loss": 4.27,
      "step": 3200
    },
    {
      "epoch": 0.3498008825745345,
      "grad_norm": 0.8958104848861694,
      "learning_rate": 0.0005796681392091369,
      "loss": 4.2535,
      "step": 3250
    },
    {
      "epoch": 0.35518243461414273,
      "grad_norm": 0.671660304069519,
      "learning_rate": 0.0005793448981790755,
      "loss": 4.2516,
      "step": 3300
    },
    {
      "epoch": 0.36056398665375095,
      "grad_norm": 0.7408727407455444,
      "learning_rate": 0.0005790216571490141,
      "loss": 4.2658,
      "step": 3350
    },
    {
      "epoch": 0.36594553869335916,
      "grad_norm": 0.6920570731163025,
      "learning_rate": 0.0005786984161189527,
      "loss": 4.2325,
      "step": 3400
    },
    {
      "epoch": 0.3713270907329674,
      "grad_norm": 0.8909915685653687,
      "learning_rate": 0.0005783751750888912,
      "loss": 4.2163,
      "step": 3450
    },
    {
      "epoch": 0.3767086427725756,
      "grad_norm": 0.8377681970596313,
      "learning_rate": 0.0005780519340588297,
      "loss": 4.2289,
      "step": 3500
    },
    {
      "epoch": 0.3820901948121838,
      "grad_norm": 0.6133842468261719,
      "learning_rate": 0.0005777286930287684,
      "loss": 4.216,
      "step": 3550
    },
    {
      "epoch": 0.38747174685179203,
      "grad_norm": 0.5939428806304932,
      "learning_rate": 0.000577405451998707,
      "loss": 4.2227,
      "step": 3600
    },
    {
      "epoch": 0.3928532988914003,
      "grad_norm": 0.6408166885375977,
      "learning_rate": 0.0005770822109686456,
      "loss": 4.2049,
      "step": 3650
    },
    {
      "epoch": 0.3982348509310085,
      "grad_norm": 0.819424033164978,
      "learning_rate": 0.0005767589699385841,
      "loss": 4.2005,
      "step": 3700
    },
    {
      "epoch": 0.40361640297061674,
      "grad_norm": 0.6583958864212036,
      "learning_rate": 0.0005764357289085228,
      "loss": 4.1926,
      "step": 3750
    },
    {
      "epoch": 0.40899795501022496,
      "grad_norm": 0.6472761631011963,
      "learning_rate": 0.0005761124878784613,
      "loss": 4.1791,
      "step": 3800
    },
    {
      "epoch": 0.4143795070498332,
      "grad_norm": 0.6922594308853149,
      "learning_rate": 0.0005757892468484,
      "loss": 4.1849,
      "step": 3850
    },
    {
      "epoch": 0.4197610590894414,
      "grad_norm": 0.6930834054946899,
      "learning_rate": 0.0005754660058183385,
      "loss": 4.1757,
      "step": 3900
    },
    {
      "epoch": 0.4251426111290496,
      "grad_norm": 0.6437042355537415,
      "learning_rate": 0.000575142764788277,
      "loss": 4.1714,
      "step": 3950
    },
    {
      "epoch": 0.4305241631686578,
      "grad_norm": 0.7084268927574158,
      "learning_rate": 0.0005748195237582157,
      "loss": 4.1579,
      "step": 4000
    },
    {
      "epoch": 0.4305241631686578,
      "eval_accuracy": 0.31272235134410686,
      "eval_loss": 4.0872697830200195,
      "eval_runtime": 184.7092,
      "eval_samples_per_second": 97.51,
      "eval_steps_per_second": 6.096,
      "step": 4000
    },
    {
      "epoch": 0.43590571520826604,
      "grad_norm": 0.6244553327560425,
      "learning_rate": 0.0005744962827281542,
      "loss": 4.1348,
      "step": 4050
    },
    {
      "epoch": 0.44128726724787426,
      "grad_norm": 0.7042115330696106,
      "learning_rate": 0.0005741730416980928,
      "loss": 4.1397,
      "step": 4100
    },
    {
      "epoch": 0.44666881928748253,
      "grad_norm": 0.7811130881309509,
      "learning_rate": 0.0005738498006680314,
      "loss": 4.1573,
      "step": 4150
    },
    {
      "epoch": 0.45205037132709075,
      "grad_norm": 0.6470036506652832,
      "learning_rate": 0.00057352655963797,
      "loss": 4.1357,
      "step": 4200
    },
    {
      "epoch": 0.45743192336669897,
      "grad_norm": 0.6716099977493286,
      "learning_rate": 0.0005732033186079086,
      "loss": 4.1327,
      "step": 4250
    },
    {
      "epoch": 0.4628134754063072,
      "grad_norm": 0.7920870184898376,
      "learning_rate": 0.0005728800775778471,
      "loss": 4.1263,
      "step": 4300
    },
    {
      "epoch": 0.4681950274459154,
      "grad_norm": 0.7230188846588135,
      "learning_rate": 0.0005725568365477857,
      "loss": 4.1104,
      "step": 4350
    },
    {
      "epoch": 0.4735765794855236,
      "grad_norm": 0.7376595735549927,
      "learning_rate": 0.0005722335955177243,
      "loss": 4.1103,
      "step": 4400
    },
    {
      "epoch": 0.47895813152513184,
      "grad_norm": 1.1194459199905396,
      "learning_rate": 0.000571910354487663,
      "loss": 4.1222,
      "step": 4450
    },
    {
      "epoch": 0.48433968356474005,
      "grad_norm": 0.6437053680419922,
      "learning_rate": 0.0005715871134576015,
      "loss": 4.1131,
      "step": 4500
    },
    {
      "epoch": 0.48972123560434827,
      "grad_norm": 0.7767078876495361,
      "learning_rate": 0.0005712638724275401,
      "loss": 4.1017,
      "step": 4550
    },
    {
      "epoch": 0.49510278764395654,
      "grad_norm": 0.6522454023361206,
      "learning_rate": 0.0005709406313974786,
      "loss": 4.0979,
      "step": 4600
    },
    {
      "epoch": 0.5004843396835648,
      "grad_norm": 0.6566357612609863,
      "learning_rate": 0.0005706173903674173,
      "loss": 4.0735,
      "step": 4650
    },
    {
      "epoch": 0.505865891723173,
      "grad_norm": 0.5966187715530396,
      "learning_rate": 0.0005702941493373559,
      "loss": 4.0895,
      "step": 4700
    },
    {
      "epoch": 0.5112474437627812,
      "grad_norm": 0.7359771132469177,
      "learning_rate": 0.0005699709083072944,
      "loss": 4.0868,
      "step": 4750
    },
    {
      "epoch": 0.5166289958023894,
      "grad_norm": 0.7156068086624146,
      "learning_rate": 0.000569647667277233,
      "loss": 4.0897,
      "step": 4800
    },
    {
      "epoch": 0.5220105478419976,
      "grad_norm": 0.6234465837478638,
      "learning_rate": 0.0005693244262471716,
      "loss": 4.0549,
      "step": 4850
    },
    {
      "epoch": 0.5273920998816058,
      "grad_norm": 0.7085710167884827,
      "learning_rate": 0.0005690011852171102,
      "loss": 4.0802,
      "step": 4900
    },
    {
      "epoch": 0.5327736519212141,
      "grad_norm": 0.5667790174484253,
      "learning_rate": 0.0005686779441870487,
      "loss": 4.0628,
      "step": 4950
    },
    {
      "epoch": 0.5381552039608223,
      "grad_norm": 0.6275297999382019,
      "learning_rate": 0.0005683547031569874,
      "loss": 4.0589,
      "step": 5000
    },
    {
      "epoch": 0.5381552039608223,
      "eval_accuracy": 0.3213402637940002,
      "eval_loss": 3.990799903869629,
      "eval_runtime": 184.6417,
      "eval_samples_per_second": 97.546,
      "eval_steps_per_second": 6.098,
      "step": 5000
    },
    {
      "epoch": 0.5435367560004305,
      "grad_norm": 0.6828199625015259,
      "learning_rate": 0.0005680314621269259,
      "loss": 4.056,
      "step": 5050
    },
    {
      "epoch": 0.5489183080400387,
      "grad_norm": 0.6843883991241455,
      "learning_rate": 0.0005677082210968646,
      "loss": 4.0418,
      "step": 5100
    },
    {
      "epoch": 0.5542998600796469,
      "grad_norm": 0.5970234274864197,
      "learning_rate": 0.0005673849800668031,
      "loss": 4.0339,
      "step": 5150
    },
    {
      "epoch": 0.5596814121192552,
      "grad_norm": 0.6066529154777527,
      "learning_rate": 0.0005670617390367416,
      "loss": 4.0554,
      "step": 5200
    },
    {
      "epoch": 0.5650629641588634,
      "grad_norm": 0.7181238532066345,
      "learning_rate": 0.0005667384980066803,
      "loss": 4.0289,
      "step": 5250
    },
    {
      "epoch": 0.5704445161984716,
      "grad_norm": 0.6789528131484985,
      "learning_rate": 0.0005664152569766188,
      "loss": 4.0306,
      "step": 5300
    },
    {
      "epoch": 0.5758260682380799,
      "grad_norm": 0.6828908920288086,
      "learning_rate": 0.0005660920159465575,
      "loss": 4.0383,
      "step": 5350
    },
    {
      "epoch": 0.5812076202776881,
      "grad_norm": 0.5343970656394958,
      "learning_rate": 0.000565768774916496,
      "loss": 4.0472,
      "step": 5400
    },
    {
      "epoch": 0.5865891723172963,
      "grad_norm": 0.6703925728797913,
      "learning_rate": 0.0005654455338864346,
      "loss": 4.0139,
      "step": 5450
    },
    {
      "epoch": 0.5919707243569046,
      "grad_norm": 0.5589005351066589,
      "learning_rate": 0.0005651222928563732,
      "loss": 4.0171,
      "step": 5500
    },
    {
      "epoch": 0.5973522763965128,
      "grad_norm": 0.6419985890388489,
      "learning_rate": 0.0005647990518263118,
      "loss": 3.994,
      "step": 5550
    },
    {
      "epoch": 0.602733828436121,
      "grad_norm": 0.5866629481315613,
      "learning_rate": 0.0005644758107962504,
      "loss": 4.0104,
      "step": 5600
    },
    {
      "epoch": 0.6081153804757292,
      "grad_norm": 0.9215855002403259,
      "learning_rate": 0.0005641525697661889,
      "loss": 4.0096,
      "step": 5650
    },
    {
      "epoch": 0.6134969325153374,
      "grad_norm": 0.5779905319213867,
      "learning_rate": 0.0005638293287361275,
      "loss": 3.9754,
      "step": 5700
    },
    {
      "epoch": 0.6188784845549457,
      "grad_norm": 0.6425080299377441,
      "learning_rate": 0.0005635060877060661,
      "loss": 3.9917,
      "step": 5750
    },
    {
      "epoch": 0.6242600365945539,
      "grad_norm": 0.6887379884719849,
      "learning_rate": 0.0005631828466760047,
      "loss": 3.9941,
      "step": 5800
    },
    {
      "epoch": 0.6296415886341621,
      "grad_norm": 0.6834486722946167,
      "learning_rate": 0.0005628596056459433,
      "loss": 3.9953,
      "step": 5850
    },
    {
      "epoch": 0.6350231406737703,
      "grad_norm": 0.7504608631134033,
      "learning_rate": 0.0005625363646158818,
      "loss": 3.9892,
      "step": 5900
    },
    {
      "epoch": 0.6404046927133785,
      "grad_norm": 0.6510689854621887,
      "learning_rate": 0.0005622131235858205,
      "loss": 3.9722,
      "step": 5950
    },
    {
      "epoch": 0.6457862447529867,
      "grad_norm": 0.5909599661827087,
      "learning_rate": 0.000561889882555759,
      "loss": 3.9673,
      "step": 6000
    },
    {
      "epoch": 0.6457862447529867,
      "eval_accuracy": 0.32834229049631664,
      "eval_loss": 3.9165666103363037,
      "eval_runtime": 184.5514,
      "eval_samples_per_second": 97.593,
      "eval_steps_per_second": 6.101,
      "step": 6000
    },
    {
      "epoch": 0.651167796792595,
      "grad_norm": 0.6396195292472839,
      "learning_rate": 0.0005615666415256976,
      "loss": 3.997,
      "step": 6050
    },
    {
      "epoch": 0.6565493488322032,
      "grad_norm": 0.5541490316390991,
      "learning_rate": 0.0005612434004956361,
      "loss": 3.981,
      "step": 6100
    },
    {
      "epoch": 0.6619309008718114,
      "grad_norm": 0.6580690145492554,
      "learning_rate": 0.0005609201594655748,
      "loss": 3.9881,
      "step": 6150
    },
    {
      "epoch": 0.6673124529114196,
      "grad_norm": 0.6880823373794556,
      "learning_rate": 0.0005605969184355134,
      "loss": 3.9573,
      "step": 6200
    },
    {
      "epoch": 0.6726940049510278,
      "grad_norm": 0.6356266140937805,
      "learning_rate": 0.000560273677405452,
      "loss": 3.9975,
      "step": 6250
    },
    {
      "epoch": 0.6780755569906362,
      "grad_norm": 0.5227646827697754,
      "learning_rate": 0.0005599504363753905,
      "loss": 3.9533,
      "step": 6300
    },
    {
      "epoch": 0.6834571090302444,
      "grad_norm": 0.6362395882606506,
      "learning_rate": 0.0005596271953453291,
      "loss": 3.9563,
      "step": 6350
    },
    {
      "epoch": 0.6888386610698526,
      "grad_norm": 0.5801821947097778,
      "learning_rate": 0.0005593039543152677,
      "loss": 3.9428,
      "step": 6400
    },
    {
      "epoch": 0.6942202131094608,
      "grad_norm": 0.6077560782432556,
      "learning_rate": 0.0005589807132852063,
      "loss": 3.9554,
      "step": 6450
    },
    {
      "epoch": 0.699601765149069,
      "grad_norm": 0.5769425630569458,
      "learning_rate": 0.0005586574722551449,
      "loss": 3.9447,
      "step": 6500
    },
    {
      "epoch": 0.7049833171886772,
      "grad_norm": 0.6278015971183777,
      "learning_rate": 0.0005583342312250834,
      "loss": 3.9678,
      "step": 6550
    },
    {
      "epoch": 0.7103648692282855,
      "grad_norm": 0.5199827551841736,
      "learning_rate": 0.0005580109901950221,
      "loss": 3.9618,
      "step": 6600
    },
    {
      "epoch": 0.7157464212678937,
      "grad_norm": 0.659668505191803,
      "learning_rate": 0.0005576877491649606,
      "loss": 3.9206,
      "step": 6650
    },
    {
      "epoch": 0.7211279733075019,
      "grad_norm": 0.6791085004806519,
      "learning_rate": 0.0005573645081348993,
      "loss": 3.9427,
      "step": 6700
    },
    {
      "epoch": 0.7265095253471101,
      "grad_norm": 0.6757652759552002,
      "learning_rate": 0.0005570412671048378,
      "loss": 3.9381,
      "step": 6750
    },
    {
      "epoch": 0.7318910773867183,
      "grad_norm": 0.5604714155197144,
      "learning_rate": 0.0005567180260747763,
      "loss": 3.9385,
      "step": 6800
    },
    {
      "epoch": 0.7372726294263265,
      "grad_norm": 0.6604406237602234,
      "learning_rate": 0.000556394785044715,
      "loss": 3.9272,
      "step": 6850
    },
    {
      "epoch": 0.7426541814659348,
      "grad_norm": 0.564445972442627,
      "learning_rate": 0.0005560715440146535,
      "loss": 3.9496,
      "step": 6900
    },
    {
      "epoch": 0.748035733505543,
      "grad_norm": 0.5737416744232178,
      "learning_rate": 0.0005557483029845921,
      "loss": 3.9387,
      "step": 6950
    },
    {
      "epoch": 0.7534172855451512,
      "grad_norm": 0.5588909983634949,
      "learning_rate": 0.0005554250619545307,
      "loss": 3.9074,
      "step": 7000
    },
    {
      "epoch": 0.7534172855451512,
      "eval_accuracy": 0.33360815270424543,
      "eval_loss": 3.859344720840454,
      "eval_runtime": 184.9069,
      "eval_samples_per_second": 97.406,
      "eval_steps_per_second": 6.09,
      "step": 7000
    },
    {
      "epoch": 0.7587988375847594,
      "grad_norm": 0.6285735964775085,
      "learning_rate": 0.0005551018209244694,
      "loss": 3.9333,
      "step": 7050
    },
    {
      "epoch": 0.7641803896243676,
      "grad_norm": 0.5597401261329651,
      "learning_rate": 0.0005547785798944079,
      "loss": 3.9209,
      "step": 7100
    },
    {
      "epoch": 0.7695619416639758,
      "grad_norm": 0.578063428401947,
      "learning_rate": 0.0005544553388643464,
      "loss": 3.9312,
      "step": 7150
    },
    {
      "epoch": 0.7749434937035841,
      "grad_norm": 0.5833026170730591,
      "learning_rate": 0.000554132097834285,
      "loss": 3.9328,
      "step": 7200
    },
    {
      "epoch": 0.7803250457431924,
      "grad_norm": 0.518353283405304,
      "learning_rate": 0.0005538088568042236,
      "loss": 3.9029,
      "step": 7250
    },
    {
      "epoch": 0.7857065977828006,
      "grad_norm": 0.6059916019439697,
      "learning_rate": 0.0005534856157741623,
      "loss": 3.9199,
      "step": 7300
    },
    {
      "epoch": 0.7910881498224088,
      "grad_norm": 0.5463325381278992,
      "learning_rate": 0.0005531623747441008,
      "loss": 3.9054,
      "step": 7350
    },
    {
      "epoch": 0.796469701862017,
      "grad_norm": 0.6463024616241455,
      "learning_rate": 0.0005528391337140394,
      "loss": 3.9191,
      "step": 7400
    },
    {
      "epoch": 0.8018512539016253,
      "grad_norm": 0.6024007797241211,
      "learning_rate": 0.0005525158926839779,
      "loss": 3.9284,
      "step": 7450
    },
    {
      "epoch": 0.8072328059412335,
      "grad_norm": 0.5554441213607788,
      "learning_rate": 0.0005521926516539166,
      "loss": 3.9015,
      "step": 7500
    },
    {
      "epoch": 0.8126143579808417,
      "grad_norm": 0.5887565612792969,
      "learning_rate": 0.0005518694106238552,
      "loss": 3.9024,
      "step": 7550
    },
    {
      "epoch": 0.8179959100204499,
      "grad_norm": 0.575228214263916,
      "learning_rate": 0.0005515461695937937,
      "loss": 3.8981,
      "step": 7600
    },
    {
      "epoch": 0.8233774620600581,
      "grad_norm": 0.6179103255271912,
      "learning_rate": 0.0005512229285637323,
      "loss": 3.9081,
      "step": 7650
    },
    {
      "epoch": 0.8287590140996663,
      "grad_norm": 0.6597625017166138,
      "learning_rate": 0.0005508996875336709,
      "loss": 3.8958,
      "step": 7700
    },
    {
      "epoch": 0.8341405661392746,
      "grad_norm": 0.5965509414672852,
      "learning_rate": 0.0005505764465036095,
      "loss": 3.8879,
      "step": 7750
    },
    {
      "epoch": 0.8395221181788828,
      "grad_norm": 0.512361466884613,
      "learning_rate": 0.000550253205473548,
      "loss": 3.8809,
      "step": 7800
    },
    {
      "epoch": 0.844903670218491,
      "grad_norm": 0.6104485988616943,
      "learning_rate": 0.0005499299644434867,
      "loss": 3.8852,
      "step": 7850
    },
    {
      "epoch": 0.8502852222580992,
      "grad_norm": 0.6456380486488342,
      "learning_rate": 0.0005496067234134252,
      "loss": 3.8783,
      "step": 7900
    },
    {
      "epoch": 0.8556667742977074,
      "grad_norm": 0.5954079627990723,
      "learning_rate": 0.0005492834823833639,
      "loss": 3.889,
      "step": 7950
    },
    {
      "epoch": 0.8610483263373157,
      "grad_norm": 0.6212036609649658,
      "learning_rate": 0.0005489602413533024,
      "loss": 3.8712,
      "step": 8000
    },
    {
      "epoch": 0.8610483263373157,
      "eval_accuracy": 0.33805890094779,
      "eval_loss": 3.8169474601745605,
      "eval_runtime": 184.4499,
      "eval_samples_per_second": 97.647,
      "eval_steps_per_second": 6.105,
      "step": 8000
    },
    {
      "epoch": 0.8664298783769239,
      "grad_norm": 0.5757076144218445,
      "learning_rate": 0.0005486370003232409,
      "loss": 3.8725,
      "step": 8050
    },
    {
      "epoch": 0.8718114304165321,
      "grad_norm": 0.6085401177406311,
      "learning_rate": 0.0005483137592931796,
      "loss": 3.8905,
      "step": 8100
    },
    {
      "epoch": 0.8771929824561403,
      "grad_norm": 0.5095037817955017,
      "learning_rate": 0.0005479905182631181,
      "loss": 3.8799,
      "step": 8150
    },
    {
      "epoch": 0.8825745344957485,
      "grad_norm": 0.5414556860923767,
      "learning_rate": 0.0005476672772330568,
      "loss": 3.8747,
      "step": 8200
    },
    {
      "epoch": 0.8879560865353568,
      "grad_norm": 0.5936631560325623,
      "learning_rate": 0.0005473440362029953,
      "loss": 3.863,
      "step": 8250
    },
    {
      "epoch": 0.8933376385749651,
      "grad_norm": 0.5667993426322937,
      "learning_rate": 0.0005470207951729339,
      "loss": 3.8607,
      "step": 8300
    },
    {
      "epoch": 0.8987191906145733,
      "grad_norm": 0.5451300144195557,
      "learning_rate": 0.0005466975541428725,
      "loss": 3.8857,
      "step": 8350
    },
    {
      "epoch": 0.9041007426541815,
      "grad_norm": 0.5745592713356018,
      "learning_rate": 0.0005463743131128111,
      "loss": 3.8554,
      "step": 8400
    },
    {
      "epoch": 0.9094822946937897,
      "grad_norm": 0.5426753759384155,
      "learning_rate": 0.0005460510720827497,
      "loss": 3.8752,
      "step": 8450
    },
    {
      "epoch": 0.9148638467333979,
      "grad_norm": 0.508082389831543,
      "learning_rate": 0.0005457278310526882,
      "loss": 3.8584,
      "step": 8500
    },
    {
      "epoch": 0.9202453987730062,
      "grad_norm": 0.532146155834198,
      "learning_rate": 0.0005454045900226268,
      "loss": 3.8476,
      "step": 8550
    },
    {
      "epoch": 0.9256269508126144,
      "grad_norm": 0.5778146386146545,
      "learning_rate": 0.0005450813489925654,
      "loss": 3.8612,
      "step": 8600
    },
    {
      "epoch": 0.9310085028522226,
      "grad_norm": 0.5984019637107849,
      "learning_rate": 0.000544758107962504,
      "loss": 3.8478,
      "step": 8650
    },
    {
      "epoch": 0.9363900548918308,
      "grad_norm": 0.609585165977478,
      "learning_rate": 0.0005444348669324426,
      "loss": 3.8541,
      "step": 8700
    },
    {
      "epoch": 0.941771606931439,
      "grad_norm": 0.5877447724342346,
      "learning_rate": 0.0005441116259023811,
      "loss": 3.8298,
      "step": 8750
    },
    {
      "epoch": 0.9471531589710472,
      "grad_norm": 0.6004450917243958,
      "learning_rate": 0.0005437883848723198,
      "loss": 3.8393,
      "step": 8800
    },
    {
      "epoch": 0.9525347110106555,
      "grad_norm": 0.5448794960975647,
      "learning_rate": 0.0005434651438422583,
      "loss": 3.8515,
      "step": 8850
    },
    {
      "epoch": 0.9579162630502637,
      "grad_norm": 0.5065131783485413,
      "learning_rate": 0.0005431419028121969,
      "loss": 3.8478,
      "step": 8900
    },
    {
      "epoch": 0.9632978150898719,
      "grad_norm": 0.6335159540176392,
      "learning_rate": 0.0005428186617821354,
      "loss": 3.8411,
      "step": 8950
    },
    {
      "epoch": 0.9686793671294801,
      "grad_norm": 0.5555748343467712,
      "learning_rate": 0.0005424954207520741,
      "loss": 3.8516,
      "step": 9000
    },
    {
      "epoch": 0.9686793671294801,
      "eval_accuracy": 0.34156469502601206,
      "eval_loss": 3.775557279586792,
      "eval_runtime": 184.197,
      "eval_samples_per_second": 97.781,
      "eval_steps_per_second": 6.113,
      "step": 9000
    },
    {
      "epoch": 0.9740609191690883,
      "grad_norm": 0.6176546812057495,
      "learning_rate": 0.0005421721797220127,
      "loss": 3.8444,
      "step": 9050
    },
    {
      "epoch": 0.9794424712086965,
      "grad_norm": 0.5457547307014465,
      "learning_rate": 0.0005418489386919513,
      "loss": 3.8305,
      "step": 9100
    },
    {
      "epoch": 0.9848240232483048,
      "grad_norm": 0.5831869840621948,
      "learning_rate": 0.0005415256976618898,
      "loss": 3.8196,
      "step": 9150
    },
    {
      "epoch": 0.9902055752879131,
      "grad_norm": 0.5737512111663818,
      "learning_rate": 0.0005412024566318284,
      "loss": 3.8124,
      "step": 9200
    },
    {
      "epoch": 0.9955871273275213,
      "grad_norm": 0.5983269810676575,
      "learning_rate": 0.000540879215601767,
      "loss": 3.8368,
      "step": 9250
    },
    {
      "epoch": 1.0009686793671295,
      "grad_norm": 0.5368288159370422,
      "learning_rate": 0.0005405559745717056,
      "loss": 3.8048,
      "step": 9300
    },
    {
      "epoch": 1.0063502314067376,
      "grad_norm": 0.5906519889831543,
      "learning_rate": 0.0005402327335416442,
      "loss": 3.7763,
      "step": 9350
    },
    {
      "epoch": 1.011731783446346,
      "grad_norm": 0.6060296893119812,
      "learning_rate": 0.0005399094925115827,
      "loss": 3.7644,
      "step": 9400
    },
    {
      "epoch": 1.017113335485954,
      "grad_norm": 0.5626549124717712,
      "learning_rate": 0.0005395862514815214,
      "loss": 3.7658,
      "step": 9450
    },
    {
      "epoch": 1.0224948875255624,
      "grad_norm": 0.521769642829895,
      "learning_rate": 0.0005392630104514599,
      "loss": 3.7667,
      "step": 9500
    },
    {
      "epoch": 1.0278764395651705,
      "grad_norm": 0.5677669048309326,
      "learning_rate": 0.0005389397694213986,
      "loss": 3.7803,
      "step": 9550
    },
    {
      "epoch": 1.0332579916047788,
      "grad_norm": 0.5613917708396912,
      "learning_rate": 0.0005386165283913371,
      "loss": 3.7582,
      "step": 9600
    },
    {
      "epoch": 1.0386395436443872,
      "grad_norm": 0.5857552289962769,
      "learning_rate": 0.0005382932873612756,
      "loss": 3.7806,
      "step": 9650
    },
    {
      "epoch": 1.0440210956839953,
      "grad_norm": 0.5657854080200195,
      "learning_rate": 0.0005379700463312143,
      "loss": 3.7687,
      "step": 9700
    },
    {
      "epoch": 1.0494026477236036,
      "grad_norm": 0.580058217048645,
      "learning_rate": 0.0005376468053011528,
      "loss": 3.7603,
      "step": 9750
    },
    {
      "epoch": 1.0547841997632117,
      "grad_norm": 0.5268446207046509,
      "learning_rate": 0.0005373235642710914,
      "loss": 3.7591,
      "step": 9800
    },
    {
      "epoch": 1.06016575180282,
      "grad_norm": 0.5513815879821777,
      "learning_rate": 0.00053700032324103,
      "loss": 3.7748,
      "step": 9850
    },
    {
      "epoch": 1.0655473038424281,
      "grad_norm": 0.5946319699287415,
      "learning_rate": 0.0005366770822109687,
      "loss": 3.7816,
      "step": 9900
    },
    {
      "epoch": 1.0709288558820365,
      "grad_norm": 0.5772368311882019,
      "learning_rate": 0.0005363538411809072,
      "loss": 3.754,
      "step": 9950
    },
    {
      "epoch": 1.0763104079216446,
      "grad_norm": 0.6033820509910583,
      "learning_rate": 0.0005360306001508457,
      "loss": 3.7639,
      "step": 10000
    },
    {
      "epoch": 1.0763104079216446,
      "eval_accuracy": 0.345006709859087,
      "eval_loss": 3.7440133094787598,
      "eval_runtime": 184.605,
      "eval_samples_per_second": 97.565,
      "eval_steps_per_second": 6.1,
      "step": 10000
    },
    {
      "epoch": 1.081691959961253,
      "grad_norm": 0.7267680764198303,
      "learning_rate": 0.0005357073591207843,
      "loss": 3.7537,
      "step": 10050
    },
    {
      "epoch": 1.087073512000861,
      "grad_norm": 0.5396670699119568,
      "learning_rate": 0.0005353841180907229,
      "loss": 3.7696,
      "step": 10100
    },
    {
      "epoch": 1.0924550640404693,
      "grad_norm": 0.5720505118370056,
      "learning_rate": 0.0005350608770606616,
      "loss": 3.7771,
      "step": 10150
    },
    {
      "epoch": 1.0978366160800774,
      "grad_norm": 0.5750023722648621,
      "learning_rate": 0.0005347376360306001,
      "loss": 3.7454,
      "step": 10200
    },
    {
      "epoch": 1.1032181681196858,
      "grad_norm": 0.6659409403800964,
      "learning_rate": 0.0005344143950005387,
      "loss": 3.7663,
      "step": 10250
    },
    {
      "epoch": 1.1085997201592939,
      "grad_norm": 0.6214229464530945,
      "learning_rate": 0.0005340911539704773,
      "loss": 3.7545,
      "step": 10300
    },
    {
      "epoch": 1.1139812721989022,
      "grad_norm": 0.5536146759986877,
      "learning_rate": 0.0005337679129404159,
      "loss": 3.7517,
      "step": 10350
    },
    {
      "epoch": 1.1193628242385103,
      "grad_norm": 0.6057895421981812,
      "learning_rate": 0.0005334446719103545,
      "loss": 3.7618,
      "step": 10400
    },
    {
      "epoch": 1.1247443762781186,
      "grad_norm": 0.5203273296356201,
      "learning_rate": 0.000533121430880293,
      "loss": 3.7583,
      "step": 10450
    },
    {
      "epoch": 1.1301259283177267,
      "grad_norm": 0.4910421371459961,
      "learning_rate": 0.0005327981898502316,
      "loss": 3.7713,
      "step": 10500
    },
    {
      "epoch": 1.135507480357335,
      "grad_norm": 0.5717787742614746,
      "learning_rate": 0.0005324749488201702,
      "loss": 3.7681,
      "step": 10550
    },
    {
      "epoch": 1.1408890323969434,
      "grad_norm": 0.5272240042686462,
      "learning_rate": 0.0005321517077901088,
      "loss": 3.7644,
      "step": 10600
    },
    {
      "epoch": 1.1462705844365515,
      "grad_norm": 0.6392913460731506,
      "learning_rate": 0.0005318284667600473,
      "loss": 3.7694,
      "step": 10650
    },
    {
      "epoch": 1.1516521364761596,
      "grad_norm": 0.5354857444763184,
      "learning_rate": 0.000531505225729986,
      "loss": 3.742,
      "step": 10700
    },
    {
      "epoch": 1.157033688515768,
      "grad_norm": 0.7330885529518127,
      "learning_rate": 0.0005311819846999245,
      "loss": 3.7568,
      "step": 10750
    },
    {
      "epoch": 1.1624152405553763,
      "grad_norm": 0.5588583946228027,
      "learning_rate": 0.0005308587436698631,
      "loss": 3.7659,
      "step": 10800
    },
    {
      "epoch": 1.1677967925949844,
      "grad_norm": 0.534525990486145,
      "learning_rate": 0.0005305355026398017,
      "loss": 3.7399,
      "step": 10850
    },
    {
      "epoch": 1.1731783446345927,
      "grad_norm": 0.541637659072876,
      "learning_rate": 0.0005302122616097402,
      "loss": 3.7469,
      "step": 10900
    },
    {
      "epoch": 1.1785598966742008,
      "grad_norm": 0.5589224100112915,
      "learning_rate": 0.0005298890205796789,
      "loss": 3.7605,
      "step": 10950
    },
    {
      "epoch": 1.1839414487138091,
      "grad_norm": 0.5767397880554199,
      "learning_rate": 0.0005295657795496175,
      "loss": 3.7414,
      "step": 11000
    },
    {
      "epoch": 1.1839414487138091,
      "eval_accuracy": 0.34713293822072855,
      "eval_loss": 3.7198851108551025,
      "eval_runtime": 184.6204,
      "eval_samples_per_second": 97.557,
      "eval_steps_per_second": 6.099,
      "step": 11000
    },
    {
      "epoch": 1.1893230007534172,
      "grad_norm": 0.5567173957824707,
      "learning_rate": 0.0005292425385195561,
      "loss": 3.7439,
      "step": 11050
    },
    {
      "epoch": 1.1947045527930256,
      "grad_norm": 0.5579226016998291,
      "learning_rate": 0.0005289192974894946,
      "loss": 3.7435,
      "step": 11100
    },
    {
      "epoch": 1.2000861048326337,
      "grad_norm": 0.61114102602005,
      "learning_rate": 0.0005285960564594331,
      "loss": 3.7653,
      "step": 11150
    },
    {
      "epoch": 1.205467656872242,
      "grad_norm": 0.6080530881881714,
      "learning_rate": 0.0005282728154293718,
      "loss": 3.7311,
      "step": 11200
    },
    {
      "epoch": 1.21084920891185,
      "grad_norm": 0.5481367111206055,
      "learning_rate": 0.0005279495743993104,
      "loss": 3.7429,
      "step": 11250
    },
    {
      "epoch": 1.2162307609514584,
      "grad_norm": 0.6586524248123169,
      "learning_rate": 0.000527626333369249,
      "loss": 3.7476,
      "step": 11300
    },
    {
      "epoch": 1.2216123129910665,
      "grad_norm": 0.5334315299987793,
      "learning_rate": 0.0005273030923391875,
      "loss": 3.7436,
      "step": 11350
    },
    {
      "epoch": 1.2269938650306749,
      "grad_norm": 0.5702027082443237,
      "learning_rate": 0.0005269798513091261,
      "loss": 3.7232,
      "step": 11400
    },
    {
      "epoch": 1.232375417070283,
      "grad_norm": 0.5503053665161133,
      "learning_rate": 0.0005266566102790647,
      "loss": 3.7325,
      "step": 11450
    },
    {
      "epoch": 1.2377569691098913,
      "grad_norm": 0.6727051138877869,
      "learning_rate": 0.0005263333692490033,
      "loss": 3.7457,
      "step": 11500
    },
    {
      "epoch": 1.2431385211494996,
      "grad_norm": 0.5632249116897583,
      "learning_rate": 0.0005260101282189419,
      "loss": 3.7434,
      "step": 11550
    },
    {
      "epoch": 1.2485200731891077,
      "grad_norm": 0.6183726191520691,
      "learning_rate": 0.0005256868871888804,
      "loss": 3.7454,
      "step": 11600
    },
    {
      "epoch": 1.2539016252287158,
      "grad_norm": 0.6172442436218262,
      "learning_rate": 0.0005253636461588191,
      "loss": 3.7201,
      "step": 11650
    },
    {
      "epoch": 1.2592831772683242,
      "grad_norm": 0.5206655859947205,
      "learning_rate": 0.0005250404051287576,
      "loss": 3.7412,
      "step": 11700
    },
    {
      "epoch": 1.2646647293079325,
      "grad_norm": 0.6109122633934021,
      "learning_rate": 0.0005247171640986962,
      "loss": 3.7149,
      "step": 11750
    },
    {
      "epoch": 1.2700462813475406,
      "grad_norm": 0.5555276870727539,
      "learning_rate": 0.0005243939230686347,
      "loss": 3.7154,
      "step": 11800
    },
    {
      "epoch": 1.275427833387149,
      "grad_norm": 0.5312607884407043,
      "learning_rate": 0.0005240706820385734,
      "loss": 3.7459,
      "step": 11850
    },
    {
      "epoch": 1.280809385426757,
      "grad_norm": 0.552590548992157,
      "learning_rate": 0.000523747441008512,
      "loss": 3.713,
      "step": 11900
    },
    {
      "epoch": 1.2861909374663654,
      "grad_norm": 0.5426647663116455,
      "learning_rate": 0.0005234241999784506,
      "loss": 3.7425,
      "step": 11950
    },
    {
      "epoch": 1.2915724895059735,
      "grad_norm": 0.5763972997665405,
      "learning_rate": 0.0005231009589483891,
      "loss": 3.7364,
      "step": 12000
    },
    {
      "epoch": 1.2915724895059735,
      "eval_accuracy": 0.3500625460348704,
      "eval_loss": 3.695915460586548,
      "eval_runtime": 184.4233,
      "eval_samples_per_second": 97.661,
      "eval_steps_per_second": 6.106,
      "step": 12000
    },
    {
      "epoch": 1.2969540415455818,
      "grad_norm": 0.6403552889823914,
      "learning_rate": 0.0005227777179183277,
      "loss": 3.7176,
      "step": 12050
    },
    {
      "epoch": 1.30233559358519,
      "grad_norm": 0.5373594164848328,
      "learning_rate": 0.0005224544768882663,
      "loss": 3.7174,
      "step": 12100
    },
    {
      "epoch": 1.3077171456247982,
      "grad_norm": 0.5955021977424622,
      "learning_rate": 0.0005221312358582049,
      "loss": 3.72,
      "step": 12150
    },
    {
      "epoch": 1.3130986976644063,
      "grad_norm": 0.6005784869194031,
      "learning_rate": 0.0005218079948281435,
      "loss": 3.7168,
      "step": 12200
    },
    {
      "epoch": 1.3184802497040147,
      "grad_norm": 0.5965345501899719,
      "learning_rate": 0.000521484753798082,
      "loss": 3.7286,
      "step": 12250
    },
    {
      "epoch": 1.3238618017436228,
      "grad_norm": 0.4936142861843109,
      "learning_rate": 0.0005211615127680207,
      "loss": 3.7257,
      "step": 12300
    },
    {
      "epoch": 1.329243353783231,
      "grad_norm": 0.5646725296974182,
      "learning_rate": 0.0005208382717379592,
      "loss": 3.7049,
      "step": 12350
    },
    {
      "epoch": 1.3346249058228392,
      "grad_norm": 0.5552993416786194,
      "learning_rate": 0.0005205150307078979,
      "loss": 3.7232,
      "step": 12400
    },
    {
      "epoch": 1.3400064578624475,
      "grad_norm": 0.5514169931411743,
      "learning_rate": 0.0005201917896778364,
      "loss": 3.7065,
      "step": 12450
    },
    {
      "epoch": 1.3453880099020559,
      "grad_norm": 0.5545753240585327,
      "learning_rate": 0.0005198685486477749,
      "loss": 3.7162,
      "step": 12500
    },
    {
      "epoch": 1.350769561941664,
      "grad_norm": 0.5531725287437439,
      "learning_rate": 0.0005195453076177136,
      "loss": 3.7265,
      "step": 12550
    },
    {
      "epoch": 1.356151113981272,
      "grad_norm": 0.504911482334137,
      "learning_rate": 0.0005192220665876521,
      "loss": 3.7103,
      "step": 12600
    },
    {
      "epoch": 1.3615326660208804,
      "grad_norm": 0.5750048160552979,
      "learning_rate": 0.0005188988255575907,
      "loss": 3.715,
      "step": 12650
    },
    {
      "epoch": 1.3669142180604887,
      "grad_norm": 0.5439777970314026,
      "learning_rate": 0.0005185755845275293,
      "loss": 3.7145,
      "step": 12700
    },
    {
      "epoch": 1.3722957701000968,
      "grad_norm": 0.5391664505004883,
      "learning_rate": 0.000518252343497468,
      "loss": 3.7083,
      "step": 12750
    },
    {
      "epoch": 1.3776773221397052,
      "grad_norm": 0.5579096078872681,
      "learning_rate": 0.0005179291024674065,
      "loss": 3.7188,
      "step": 12800
    },
    {
      "epoch": 1.3830588741793133,
      "grad_norm": 0.563324511051178,
      "learning_rate": 0.000517605861437345,
      "loss": 3.7051,
      "step": 12850
    },
    {
      "epoch": 1.3884404262189216,
      "grad_norm": 0.5446837544441223,
      "learning_rate": 0.0005172826204072836,
      "loss": 3.7164,
      "step": 12900
    },
    {
      "epoch": 1.3938219782585297,
      "grad_norm": 0.5987839102745056,
      "learning_rate": 0.0005169593793772222,
      "loss": 3.7137,
      "step": 12950
    },
    {
      "epoch": 1.399203530298138,
      "grad_norm": 0.5936101078987122,
      "learning_rate": 0.0005166361383471609,
      "loss": 3.7086,
      "step": 13000
    },
    {
      "epoch": 1.399203530298138,
      "eval_accuracy": 0.3522383201133554,
      "eval_loss": 3.673328161239624,
      "eval_runtime": 184.7267,
      "eval_samples_per_second": 97.501,
      "eval_steps_per_second": 6.095,
      "step": 13000
    },
    {
      "epoch": 1.4045850823377461,
      "grad_norm": 0.6273950934410095,
      "learning_rate": 0.0005163128973170994,
      "loss": 3.7156,
      "step": 13050
    },
    {
      "epoch": 1.4099666343773545,
      "grad_norm": 0.5783877968788147,
      "learning_rate": 0.000515989656287038,
      "loss": 3.702,
      "step": 13100
    },
    {
      "epoch": 1.4153481864169626,
      "grad_norm": 0.5784569382667542,
      "learning_rate": 0.0005156664152569766,
      "loss": 3.7103,
      "step": 13150
    },
    {
      "epoch": 1.420729738456571,
      "grad_norm": 0.5583294034004211,
      "learning_rate": 0.0005153431742269152,
      "loss": 3.7069,
      "step": 13200
    },
    {
      "epoch": 1.426111290496179,
      "grad_norm": 0.5216031670570374,
      "learning_rate": 0.0005150199331968538,
      "loss": 3.7013,
      "step": 13250
    },
    {
      "epoch": 1.4314928425357873,
      "grad_norm": 0.6091950535774231,
      "learning_rate": 0.0005146966921667923,
      "loss": 3.7041,
      "step": 13300
    },
    {
      "epoch": 1.4368743945753955,
      "grad_norm": 0.5405470728874207,
      "learning_rate": 0.0005143734511367309,
      "loss": 3.7233,
      "step": 13350
    },
    {
      "epoch": 1.4422559466150038,
      "grad_norm": 0.5653330683708191,
      "learning_rate": 0.0005140502101066695,
      "loss": 3.7113,
      "step": 13400
    },
    {
      "epoch": 1.447637498654612,
      "grad_norm": 0.5083488821983337,
      "learning_rate": 0.0005137269690766081,
      "loss": 3.6946,
      "step": 13450
    },
    {
      "epoch": 1.4530190506942202,
      "grad_norm": 0.5432257652282715,
      "learning_rate": 0.0005134037280465466,
      "loss": 3.6909,
      "step": 13500
    },
    {
      "epoch": 1.4584006027338283,
      "grad_norm": 0.5383495092391968,
      "learning_rate": 0.0005130804870164853,
      "loss": 3.6824,
      "step": 13550
    },
    {
      "epoch": 1.4637821547734367,
      "grad_norm": 0.5217844247817993,
      "learning_rate": 0.0005127572459864238,
      "loss": 3.6961,
      "step": 13600
    },
    {
      "epoch": 1.469163706813045,
      "grad_norm": 0.5690218210220337,
      "learning_rate": 0.0005124340049563624,
      "loss": 3.6898,
      "step": 13650
    },
    {
      "epoch": 1.474545258852653,
      "grad_norm": 0.536909282207489,
      "learning_rate": 0.000512110763926301,
      "loss": 3.6847,
      "step": 13700
    },
    {
      "epoch": 1.4799268108922612,
      "grad_norm": 0.5096193552017212,
      "learning_rate": 0.0005117875228962395,
      "loss": 3.6682,
      "step": 13750
    },
    {
      "epoch": 1.4853083629318695,
      "grad_norm": 0.5519959330558777,
      "learning_rate": 0.0005114642818661782,
      "loss": 3.7007,
      "step": 13800
    },
    {
      "epoch": 1.4906899149714778,
      "grad_norm": 0.6047645807266235,
      "learning_rate": 0.0005111410408361168,
      "loss": 3.6784,
      "step": 13850
    },
    {
      "epoch": 1.496071467011086,
      "grad_norm": 0.5431356430053711,
      "learning_rate": 0.0005108177998060554,
      "loss": 3.6863,
      "step": 13900
    },
    {
      "epoch": 1.501453019050694,
      "grad_norm": 0.5571900606155396,
      "learning_rate": 0.0005104945587759939,
      "loss": 3.7115,
      "step": 13950
    },
    {
      "epoch": 1.5068345710903024,
      "grad_norm": 0.5458298921585083,
      "learning_rate": 0.0005101713177459324,
      "loss": 3.6937,
      "step": 14000
    },
    {
      "epoch": 1.5068345710903024,
      "eval_accuracy": 0.353437413383276,
      "eval_loss": 3.6556448936462402,
      "eval_runtime": 184.6362,
      "eval_samples_per_second": 97.549,
      "eval_steps_per_second": 6.098,
      "step": 14000
    },
    {
      "epoch": 1.5122161231299107,
      "grad_norm": 0.5262465476989746,
      "learning_rate": 0.0005098480767158711,
      "loss": 3.7033,
      "step": 14050
    },
    {
      "epoch": 1.5175976751695188,
      "grad_norm": 0.5345038175582886,
      "learning_rate": 0.0005095248356858097,
      "loss": 3.6971,
      "step": 14100
    },
    {
      "epoch": 1.5229792272091272,
      "grad_norm": 0.5168197751045227,
      "learning_rate": 0.0005092015946557483,
      "loss": 3.6759,
      "step": 14150
    },
    {
      "epoch": 1.5283607792487355,
      "grad_norm": 0.5598461627960205,
      "learning_rate": 0.0005088783536256868,
      "loss": 3.6936,
      "step": 14200
    },
    {
      "epoch": 1.5337423312883436,
      "grad_norm": 0.616462230682373,
      "learning_rate": 0.0005085551125956255,
      "loss": 3.6958,
      "step": 14250
    },
    {
      "epoch": 1.5391238833279517,
      "grad_norm": 0.5687780380249023,
      "learning_rate": 0.000508231871565564,
      "loss": 3.6718,
      "step": 14300
    },
    {
      "epoch": 1.54450543536756,
      "grad_norm": 0.6172040104866028,
      "learning_rate": 0.0005079086305355026,
      "loss": 3.6722,
      "step": 14350
    },
    {
      "epoch": 1.5498869874071683,
      "grad_norm": 0.5774914622306824,
      "learning_rate": 0.0005075853895054412,
      "loss": 3.7067,
      "step": 14400
    },
    {
      "epoch": 1.5552685394467765,
      "grad_norm": 0.5606103539466858,
      "learning_rate": 0.0005072621484753797,
      "loss": 3.6953,
      "step": 14450
    },
    {
      "epoch": 1.5606500914863846,
      "grad_norm": 0.6137360334396362,
      "learning_rate": 0.0005069389074453184,
      "loss": 3.6851,
      "step": 14500
    },
    {
      "epoch": 1.566031643525993,
      "grad_norm": 0.5306119322776794,
      "learning_rate": 0.0005066156664152569,
      "loss": 3.6717,
      "step": 14550
    },
    {
      "epoch": 1.5714131955656012,
      "grad_norm": 0.5361549258232117,
      "learning_rate": 0.0005062924253851955,
      "loss": 3.6858,
      "step": 14600
    },
    {
      "epoch": 1.5767947476052093,
      "grad_norm": 0.5695285201072693,
      "learning_rate": 0.0005059691843551341,
      "loss": 3.6793,
      "step": 14650
    },
    {
      "epoch": 1.5821762996448174,
      "grad_norm": 0.6082088351249695,
      "learning_rate": 0.0005056459433250727,
      "loss": 3.6823,
      "step": 14700
    },
    {
      "epoch": 1.5875578516844258,
      "grad_norm": 0.5494985580444336,
      "learning_rate": 0.0005053227022950113,
      "loss": 3.6773,
      "step": 14750
    },
    {
      "epoch": 1.592939403724034,
      "grad_norm": 0.5614803433418274,
      "learning_rate": 0.0005049994612649499,
      "loss": 3.6898,
      "step": 14800
    },
    {
      "epoch": 1.5983209557636422,
      "grad_norm": 0.5661324858665466,
      "learning_rate": 0.0005046762202348884,
      "loss": 3.6868,
      "step": 14850
    },
    {
      "epoch": 1.6037025078032503,
      "grad_norm": 0.5200116634368896,
      "learning_rate": 0.000504352979204827,
      "loss": 3.6836,
      "step": 14900
    },
    {
      "epoch": 1.6090840598428586,
      "grad_norm": 0.5705900192260742,
      "learning_rate": 0.0005040297381747656,
      "loss": 3.6905,
      "step": 14950
    },
    {
      "epoch": 1.614465611882467,
      "grad_norm": 0.5745868682861328,
      "learning_rate": 0.0005037064971447042,
      "loss": 3.6986,
      "step": 15000
    },
    {
      "epoch": 1.614465611882467,
      "eval_accuracy": 0.35545531481576653,
      "eval_loss": 3.6349782943725586,
      "eval_runtime": 184.312,
      "eval_samples_per_second": 97.72,
      "eval_steps_per_second": 6.109,
      "step": 15000
    },
    {
      "epoch": 1.619847163922075,
      "grad_norm": 0.5304009318351746,
      "learning_rate": 0.0005033832561146428,
      "loss": 3.6643,
      "step": 15050
    },
    {
      "epoch": 1.6252287159616834,
      "grad_norm": 0.5392767190933228,
      "learning_rate": 0.0005030600150845813,
      "loss": 3.6664,
      "step": 15100
    },
    {
      "epoch": 1.6306102680012917,
      "grad_norm": 0.5564727187156677,
      "learning_rate": 0.00050273677405452,
      "loss": 3.6673,
      "step": 15150
    },
    {
      "epoch": 1.6359918200408998,
      "grad_norm": 0.5869589447975159,
      "learning_rate": 0.0005024135330244585,
      "loss": 3.6861,
      "step": 15200
    },
    {
      "epoch": 1.641373372080508,
      "grad_norm": 0.6014476418495178,
      "learning_rate": 0.0005020902919943972,
      "loss": 3.6682,
      "step": 15250
    },
    {
      "epoch": 1.6467549241201163,
      "grad_norm": 0.5469924211502075,
      "learning_rate": 0.0005017670509643357,
      "loss": 3.6743,
      "step": 15300
    },
    {
      "epoch": 1.6521364761597246,
      "grad_norm": 0.5648680925369263,
      "learning_rate": 0.0005014438099342743,
      "loss": 3.6877,
      "step": 15350
    },
    {
      "epoch": 1.6575180281993327,
      "grad_norm": 0.6184208989143372,
      "learning_rate": 0.0005011205689042129,
      "loss": 3.6477,
      "step": 15400
    },
    {
      "epoch": 1.6628995802389408,
      "grad_norm": 0.575512707233429,
      "learning_rate": 0.0005007973278741514,
      "loss": 3.6665,
      "step": 15450
    },
    {
      "epoch": 1.6682811322785491,
      "grad_norm": 0.5355293154716492,
      "learning_rate": 0.00050047408684409,
      "loss": 3.68,
      "step": 15500
    },
    {
      "epoch": 1.6736626843181575,
      "grad_norm": 0.5851496458053589,
      "learning_rate": 0.0005001508458140286,
      "loss": 3.6668,
      "step": 15550
    },
    {
      "epoch": 1.6790442363577656,
      "grad_norm": 0.5289725661277771,
      "learning_rate": 0.0004998276047839673,
      "loss": 3.6951,
      "step": 15600
    },
    {
      "epoch": 1.6844257883973737,
      "grad_norm": 0.6031738519668579,
      "learning_rate": 0.0004995043637539058,
      "loss": 3.6694,
      "step": 15650
    },
    {
      "epoch": 1.689807340436982,
      "grad_norm": 0.567891001701355,
      "learning_rate": 0.0004991811227238443,
      "loss": 3.6925,
      "step": 15700
    },
    {
      "epoch": 1.6951888924765903,
      "grad_norm": 0.524456262588501,
      "learning_rate": 0.0004988578816937829,
      "loss": 3.6733,
      "step": 15750
    },
    {
      "epoch": 1.7005704445161984,
      "grad_norm": 0.6047706604003906,
      "learning_rate": 0.0004985346406637215,
      "loss": 3.6716,
      "step": 15800
    },
    {
      "epoch": 1.7059519965558065,
      "grad_norm": 0.6086586117744446,
      "learning_rate": 0.0004982113996336602,
      "loss": 3.6652,
      "step": 15850
    },
    {
      "epoch": 1.7113335485954149,
      "grad_norm": 0.5751634836196899,
      "learning_rate": 0.0004978881586035987,
      "loss": 3.6706,
      "step": 15900
    },
    {
      "epoch": 1.7167151006350232,
      "grad_norm": 0.5148400664329529,
      "learning_rate": 0.0004975649175735373,
      "loss": 3.6692,
      "step": 15950
    },
    {
      "epoch": 1.7220966526746313,
      "grad_norm": 0.5664333701133728,
      "learning_rate": 0.0004972416765434759,
      "loss": 3.6646,
      "step": 16000
    },
    {
      "epoch": 1.7220966526746313,
      "eval_accuracy": 0.3574298637460191,
      "eval_loss": 3.618809461593628,
      "eval_runtime": 184.6411,
      "eval_samples_per_second": 97.546,
      "eval_steps_per_second": 6.098,
      "step": 16000
    },
    {
      "epoch": 1.7274782047142396,
      "grad_norm": 0.5310845971107483,
      "learning_rate": 0.0004969184355134145,
      "loss": 3.6796,
      "step": 16050
    },
    {
      "epoch": 1.732859756753848,
      "grad_norm": 0.5434026122093201,
      "learning_rate": 0.0004966016593039543,
      "loss": 3.6685,
      "step": 16100
    },
    {
      "epoch": 1.738241308793456,
      "grad_norm": 0.5860040783882141,
      "learning_rate": 0.0004962784182738928,
      "loss": 3.6586,
      "step": 16150
    },
    {
      "epoch": 1.7436228608330642,
      "grad_norm": 0.5336900949478149,
      "learning_rate": 0.0004959551772438314,
      "loss": 3.675,
      "step": 16200
    },
    {
      "epoch": 1.7490044128726725,
      "grad_norm": 0.5451406836509705,
      "learning_rate": 0.0004956319362137701,
      "loss": 3.6633,
      "step": 16250
    },
    {
      "epoch": 1.7543859649122808,
      "grad_norm": 0.5304573178291321,
      "learning_rate": 0.0004953086951837086,
      "loss": 3.6377,
      "step": 16300
    },
    {
      "epoch": 1.759767516951889,
      "grad_norm": 0.521602988243103,
      "learning_rate": 0.0004949854541536472,
      "loss": 3.6885,
      "step": 16350
    },
    {
      "epoch": 1.765149068991497,
      "grad_norm": 0.5319018363952637,
      "learning_rate": 0.0004946622131235857,
      "loss": 3.6523,
      "step": 16400
    },
    {
      "epoch": 1.7705306210311054,
      "grad_norm": 0.6209896802902222,
      "learning_rate": 0.0004943389720935244,
      "loss": 3.6639,
      "step": 16450
    },
    {
      "epoch": 1.7759121730707137,
      "grad_norm": 0.5356787443161011,
      "learning_rate": 0.000494015731063463,
      "loss": 3.6537,
      "step": 16500
    },
    {
      "epoch": 1.7812937251103218,
      "grad_norm": 0.5571577548980713,
      "learning_rate": 0.0004936924900334016,
      "loss": 3.6549,
      "step": 16550
    },
    {
      "epoch": 1.78667527714993,
      "grad_norm": 0.5454883575439453,
      "learning_rate": 0.0004933692490033401,
      "loss": 3.6586,
      "step": 16600
    },
    {
      "epoch": 1.7920568291895382,
      "grad_norm": 0.5444821715354919,
      "learning_rate": 0.0004930460079732786,
      "loss": 3.6702,
      "step": 16650
    },
    {
      "epoch": 1.7974383812291466,
      "grad_norm": 0.5578252673149109,
      "learning_rate": 0.0004927227669432173,
      "loss": 3.6416,
      "step": 16700
    },
    {
      "epoch": 1.8028199332687547,
      "grad_norm": 0.5404045581817627,
      "learning_rate": 0.0004923995259131558,
      "loss": 3.6455,
      "step": 16750
    },
    {
      "epoch": 1.8082014853083628,
      "grad_norm": 0.5403972268104553,
      "learning_rate": 0.0004920762848830945,
      "loss": 3.6536,
      "step": 16800
    },
    {
      "epoch": 1.813583037347971,
      "grad_norm": 0.6036391258239746,
      "learning_rate": 0.000491753043853033,
      "loss": 3.6643,
      "step": 16850
    },
    {
      "epoch": 1.8189645893875794,
      "grad_norm": 0.5716856122016907,
      "learning_rate": 0.0004914298028229717,
      "loss": 3.6373,
      "step": 16900
    },
    {
      "epoch": 1.8243461414271875,
      "grad_norm": 0.5589890480041504,
      "learning_rate": 0.0004911065617929102,
      "loss": 3.6553,
      "step": 16950
    },
    {
      "epoch": 1.8297276934667959,
      "grad_norm": 0.5614646673202515,
      "learning_rate": 0.0004907833207628487,
      "loss": 3.6434,
      "step": 17000
    },
    {
      "epoch": 1.8297276934667959,
      "eval_accuracy": 0.3586415607509262,
      "eval_loss": 3.6042091846466064,
      "eval_runtime": 184.6067,
      "eval_samples_per_second": 97.564,
      "eval_steps_per_second": 6.099,
      "step": 17000
    },
    {
      "epoch": 1.8351092455064042,
      "grad_norm": 0.5553935170173645,
      "learning_rate": 0.0004904600797327874,
      "loss": 3.6565,
      "step": 17050
    },
    {
      "epoch": 1.8404907975460123,
      "grad_norm": 0.5318084359169006,
      "learning_rate": 0.0004901368387027259,
      "loss": 3.6334,
      "step": 17100
    },
    {
      "epoch": 1.8458723495856204,
      "grad_norm": 0.5458089113235474,
      "learning_rate": 0.0004898135976726646,
      "loss": 3.6462,
      "step": 17150
    },
    {
      "epoch": 1.8512539016252287,
      "grad_norm": 0.5222024321556091,
      "learning_rate": 0.0004894903566426031,
      "loss": 3.6486,
      "step": 17200
    },
    {
      "epoch": 1.856635453664837,
      "grad_norm": 0.576395571231842,
      "learning_rate": 0.0004891671156125417,
      "loss": 3.6422,
      "step": 17250
    },
    {
      "epoch": 1.8620170057044452,
      "grad_norm": 0.5254101157188416,
      "learning_rate": 0.0004888438745824803,
      "loss": 3.6532,
      "step": 17300
    },
    {
      "epoch": 1.8673985577440533,
      "grad_norm": 0.580644965171814,
      "learning_rate": 0.0004885206335524189,
      "loss": 3.6338,
      "step": 17350
    },
    {
      "epoch": 1.8727801097836616,
      "grad_norm": 0.5833187699317932,
      "learning_rate": 0.0004881973925223575,
      "loss": 3.6468,
      "step": 17400
    },
    {
      "epoch": 1.87816166182327,
      "grad_norm": 0.5698775053024292,
      "learning_rate": 0.00048787415149229604,
      "loss": 3.6259,
      "step": 17450
    },
    {
      "epoch": 1.883543213862878,
      "grad_norm": 0.5299211740493774,
      "learning_rate": 0.00048755091046223464,
      "loss": 3.6593,
      "step": 17500
    },
    {
      "epoch": 1.8889247659024861,
      "grad_norm": 0.5283573865890503,
      "learning_rate": 0.0004872276694321732,
      "loss": 3.6396,
      "step": 17550
    },
    {
      "epoch": 1.8943063179420945,
      "grad_norm": 0.5362534523010254,
      "learning_rate": 0.00048690442840211177,
      "loss": 3.6368,
      "step": 17600
    },
    {
      "epoch": 1.8996878699817028,
      "grad_norm": 0.6080977916717529,
      "learning_rate": 0.0004865811873720504,
      "loss": 3.6426,
      "step": 17650
    },
    {
      "epoch": 1.905069422021311,
      "grad_norm": 0.5486099123954773,
      "learning_rate": 0.00048625794634198896,
      "loss": 3.6298,
      "step": 17700
    },
    {
      "epoch": 1.910450974060919,
      "grad_norm": 0.556928813457489,
      "learning_rate": 0.00048593470531192756,
      "loss": 3.6388,
      "step": 17750
    },
    {
      "epoch": 1.9158325261005273,
      "grad_norm": 0.5231663584709167,
      "learning_rate": 0.00048561146428186615,
      "loss": 3.6384,
      "step": 17800
    },
    {
      "epoch": 1.9212140781401357,
      "grad_norm": 0.5433329343795776,
      "learning_rate": 0.0004852882232518047,
      "loss": 3.6283,
      "step": 17850
    },
    {
      "epoch": 1.9265956301797438,
      "grad_norm": 0.5465906262397766,
      "learning_rate": 0.00048496498222174334,
      "loss": 3.6332,
      "step": 17900
    },
    {
      "epoch": 1.931977182219352,
      "grad_norm": 0.5724872946739197,
      "learning_rate": 0.00048464174119168193,
      "loss": 3.6462,
      "step": 17950
    },
    {
      "epoch": 1.9373587342589604,
      "grad_norm": 0.6602355241775513,
      "learning_rate": 0.0004843185001616205,
      "loss": 3.6473,
      "step": 18000
    },
    {
      "epoch": 1.9373587342589604,
      "eval_accuracy": 0.3605781898233315,
      "eval_loss": 3.587989091873169,
      "eval_runtime": 184.6487,
      "eval_samples_per_second": 97.542,
      "eval_steps_per_second": 6.098,
      "step": 18000
    },
    {
      "epoch": 1.9427402862985685,
      "grad_norm": 0.5606858134269714,
      "learning_rate": 0.00048399525913155907,
      "loss": 3.6358,
      "step": 18050
    },
    {
      "epoch": 1.9481218383381766,
      "grad_norm": 0.739033579826355,
      "learning_rate": 0.0004836720181014976,
      "loss": 3.6298,
      "step": 18100
    },
    {
      "epoch": 1.953503390377785,
      "grad_norm": 0.5382635593414307,
      "learning_rate": 0.0004833487770714362,
      "loss": 3.626,
      "step": 18150
    },
    {
      "epoch": 1.9588849424173933,
      "grad_norm": 0.5662149786949158,
      "learning_rate": 0.00048302553604137485,
      "loss": 3.6201,
      "step": 18200
    },
    {
      "epoch": 1.9642664944570014,
      "grad_norm": 0.5744447708129883,
      "learning_rate": 0.0004827087598319146,
      "loss": 3.6442,
      "step": 18250
    },
    {
      "epoch": 1.9696480464966095,
      "grad_norm": 0.5152103304862976,
      "learning_rate": 0.00048238551880185325,
      "loss": 3.64,
      "step": 18300
    },
    {
      "epoch": 1.9750295985362178,
      "grad_norm": 0.5561021566390991,
      "learning_rate": 0.0004820622777717918,
      "loss": 3.6278,
      "step": 18350
    },
    {
      "epoch": 1.9804111505758262,
      "grad_norm": 0.5693721771240234,
      "learning_rate": 0.0004817390367417304,
      "loss": 3.6472,
      "step": 18400
    },
    {
      "epoch": 1.9857927026154343,
      "grad_norm": 0.5853719115257263,
      "learning_rate": 0.00048141579571166893,
      "loss": 3.6207,
      "step": 18450
    },
    {
      "epoch": 1.9911742546550424,
      "grad_norm": 0.597014844417572,
      "learning_rate": 0.0004810925546816075,
      "loss": 3.6292,
      "step": 18500
    },
    {
      "epoch": 1.9965558066946507,
      "grad_norm": 0.5561902523040771,
      "learning_rate": 0.0004807693136515461,
      "loss": 3.648,
      "step": 18550
    },
    {
      "epoch": 2.001937358734259,
      "grad_norm": 0.5626040101051331,
      "learning_rate": 0.0004804460726214847,
      "loss": 3.5959,
      "step": 18600
    },
    {
      "epoch": 2.007318910773867,
      "grad_norm": 0.5984902381896973,
      "learning_rate": 0.0004801228315914233,
      "loss": 3.539,
      "step": 18650
    },
    {
      "epoch": 2.0127004628134753,
      "grad_norm": 0.5392407178878784,
      "learning_rate": 0.0004797995905613619,
      "loss": 3.531,
      "step": 18700
    },
    {
      "epoch": 2.018082014853084,
      "grad_norm": 0.6213756799697876,
      "learning_rate": 0.00047947634953130044,
      "loss": 3.5341,
      "step": 18750
    },
    {
      "epoch": 2.023463566892692,
      "grad_norm": 0.5914905667304993,
      "learning_rate": 0.00047915310850123904,
      "loss": 3.5401,
      "step": 18800
    },
    {
      "epoch": 2.0288451189323,
      "grad_norm": 0.5456693172454834,
      "learning_rate": 0.0004788298674711777,
      "loss": 3.5331,
      "step": 18850
    },
    {
      "epoch": 2.034226670971908,
      "grad_norm": 0.5378903150558472,
      "learning_rate": 0.00047850662644111623,
      "loss": 3.5555,
      "step": 18900
    },
    {
      "epoch": 2.0396082230115167,
      "grad_norm": 0.5830462574958801,
      "learning_rate": 0.0004781833854110548,
      "loss": 3.532,
      "step": 18950
    },
    {
      "epoch": 2.044989775051125,
      "grad_norm": 0.5927799940109253,
      "learning_rate": 0.00047786014438099336,
      "loss": 3.5583,
      "step": 19000
    },
    {
      "epoch": 2.044989775051125,
      "eval_accuracy": 0.3620152329175658,
      "eval_loss": 3.5778863430023193,
      "eval_runtime": 184.4238,
      "eval_samples_per_second": 97.661,
      "eval_steps_per_second": 6.106,
      "step": 19000
    },
    {
      "epoch": 2.050371327090733,
      "grad_norm": 0.6135937571525574,
      "learning_rate": 0.00047753690335093196,
      "loss": 3.5692,
      "step": 19050
    },
    {
      "epoch": 2.055752879130341,
      "grad_norm": 0.5517251491546631,
      "learning_rate": 0.00047721366232087055,
      "loss": 3.5511,
      "step": 19100
    },
    {
      "epoch": 2.0611344311699495,
      "grad_norm": 0.5605584383010864,
      "learning_rate": 0.00047689042129080915,
      "loss": 3.5399,
      "step": 19150
    },
    {
      "epoch": 2.0665159832095576,
      "grad_norm": 0.54792720079422,
      "learning_rate": 0.00047656718026074774,
      "loss": 3.5455,
      "step": 19200
    },
    {
      "epoch": 2.0718975352491658,
      "grad_norm": 0.594642162322998,
      "learning_rate": 0.00047624393923068634,
      "loss": 3.5548,
      "step": 19250
    },
    {
      "epoch": 2.0772790872887743,
      "grad_norm": 0.5518810153007507,
      "learning_rate": 0.0004759206982006249,
      "loss": 3.5538,
      "step": 19300
    },
    {
      "epoch": 2.0826606393283824,
      "grad_norm": 0.5460236072540283,
      "learning_rate": 0.00047559745717056347,
      "loss": 3.5546,
      "step": 19350
    },
    {
      "epoch": 2.0880421913679905,
      "grad_norm": 0.5818831920623779,
      "learning_rate": 0.000475274216140502,
      "loss": 3.5441,
      "step": 19400
    },
    {
      "epoch": 2.0934237434075986,
      "grad_norm": 0.5418691635131836,
      "learning_rate": 0.00047495097511044066,
      "loss": 3.552,
      "step": 19450
    },
    {
      "epoch": 2.098805295447207,
      "grad_norm": 0.6234689354896545,
      "learning_rate": 0.00047462773408037925,
      "loss": 3.5575,
      "step": 19500
    },
    {
      "epoch": 2.1041868474868153,
      "grad_norm": 0.5577658414840698,
      "learning_rate": 0.0004743044930503178,
      "loss": 3.5691,
      "step": 19550
    },
    {
      "epoch": 2.1095683995264234,
      "grad_norm": 0.5367577075958252,
      "learning_rate": 0.0004739812520202564,
      "loss": 3.5516,
      "step": 19600
    },
    {
      "epoch": 2.1149499515660315,
      "grad_norm": 0.5605822205543518,
      "learning_rate": 0.000473658010990195,
      "loss": 3.5661,
      "step": 19650
    },
    {
      "epoch": 2.12033150360564,
      "grad_norm": 0.5611740946769714,
      "learning_rate": 0.0004733347699601336,
      "loss": 3.5445,
      "step": 19700
    },
    {
      "epoch": 2.125713055645248,
      "grad_norm": 0.5580589771270752,
      "learning_rate": 0.0004730115289300722,
      "loss": 3.5638,
      "step": 19750
    },
    {
      "epoch": 2.1310946076848563,
      "grad_norm": 0.5551881790161133,
      "learning_rate": 0.00047268828790001077,
      "loss": 3.5791,
      "step": 19800
    },
    {
      "epoch": 2.1364761597244644,
      "grad_norm": 0.5823877453804016,
      "learning_rate": 0.0004723650468699493,
      "loss": 3.5641,
      "step": 19850
    },
    {
      "epoch": 2.141857711764073,
      "grad_norm": 0.5327476263046265,
      "learning_rate": 0.0004720418058398879,
      "loss": 3.5788,
      "step": 19900
    },
    {
      "epoch": 2.147239263803681,
      "grad_norm": 0.5645259618759155,
      "learning_rate": 0.00047171856480982644,
      "loss": 3.5569,
      "step": 19950
    },
    {
      "epoch": 2.152620815843289,
      "grad_norm": 0.6214436292648315,
      "learning_rate": 0.0004713953237797651,
      "loss": 3.5672,
      "step": 20000
    },
    {
      "epoch": 2.152620815843289,
      "eval_accuracy": 0.3629375872822229,
      "eval_loss": 3.5702061653137207,
      "eval_runtime": 184.4838,
      "eval_samples_per_second": 97.629,
      "eval_steps_per_second": 6.104,
      "step": 20000
    },
    {
      "epoch": 2.1580023678828972,
      "grad_norm": 0.6375493407249451,
      "learning_rate": 0.0004710720827497037,
      "loss": 3.5479,
      "step": 20050
    },
    {
      "epoch": 2.163383919922506,
      "grad_norm": 0.5471000671386719,
      "learning_rate": 0.00047074884171964223,
      "loss": 3.5683,
      "step": 20100
    },
    {
      "epoch": 2.168765471962114,
      "grad_norm": 0.6272624135017395,
      "learning_rate": 0.0004704256006895808,
      "loss": 3.5552,
      "step": 20150
    },
    {
      "epoch": 2.174147024001722,
      "grad_norm": 0.5281191468238831,
      "learning_rate": 0.00047010235965951936,
      "loss": 3.5606,
      "step": 20200
    },
    {
      "epoch": 2.1795285760413305,
      "grad_norm": 0.5559777021408081,
      "learning_rate": 0.00046977911862945796,
      "loss": 3.5424,
      "step": 20250
    },
    {
      "epoch": 2.1849101280809387,
      "grad_norm": 0.5731501579284668,
      "learning_rate": 0.0004694558775993966,
      "loss": 3.5593,
      "step": 20300
    },
    {
      "epoch": 2.1902916801205468,
      "grad_norm": 0.5684100389480591,
      "learning_rate": 0.00046913263656933515,
      "loss": 3.5506,
      "step": 20350
    },
    {
      "epoch": 2.195673232160155,
      "grad_norm": 0.627325177192688,
      "learning_rate": 0.00046880939553927374,
      "loss": 3.5475,
      "step": 20400
    },
    {
      "epoch": 2.2010547841997634,
      "grad_norm": 0.5675097703933716,
      "learning_rate": 0.00046848615450921234,
      "loss": 3.5461,
      "step": 20450
    },
    {
      "epoch": 2.2064363362393715,
      "grad_norm": 0.5912321209907532,
      "learning_rate": 0.0004681629134791509,
      "loss": 3.5582,
      "step": 20500
    },
    {
      "epoch": 2.2118178882789796,
      "grad_norm": 0.5270475149154663,
      "learning_rate": 0.00046783967244908947,
      "loss": 3.575,
      "step": 20550
    },
    {
      "epoch": 2.2171994403185877,
      "grad_norm": 0.5744220614433289,
      "learning_rate": 0.0004675164314190281,
      "loss": 3.559,
      "step": 20600
    },
    {
      "epoch": 2.2225809923581963,
      "grad_norm": 0.5762168765068054,
      "learning_rate": 0.00046719319038896666,
      "loss": 3.5585,
      "step": 20650
    },
    {
      "epoch": 2.2279625443978044,
      "grad_norm": 0.6091347932815552,
      "learning_rate": 0.00046686994935890526,
      "loss": 3.5631,
      "step": 20700
    },
    {
      "epoch": 2.2333440964374125,
      "grad_norm": 0.5728710889816284,
      "learning_rate": 0.0004665467083288438,
      "loss": 3.5612,
      "step": 20750
    },
    {
      "epoch": 2.2387256484770206,
      "grad_norm": 0.5855892896652222,
      "learning_rate": 0.0004662234672987824,
      "loss": 3.5564,
      "step": 20800
    },
    {
      "epoch": 2.244107200516629,
      "grad_norm": 0.5865923166275024,
      "learning_rate": 0.0004659066910893222,
      "loss": 3.5502,
      "step": 20850
    },
    {
      "epoch": 2.2494887525562373,
      "grad_norm": 0.5471289753913879,
      "learning_rate": 0.0004655834500592608,
      "loss": 3.5732,
      "step": 20900
    },
    {
      "epoch": 2.2548703045958454,
      "grad_norm": 0.5742563009262085,
      "learning_rate": 0.00046526020902919944,
      "loss": 3.5591,
      "step": 20950
    },
    {
      "epoch": 2.2602518566354535,
      "grad_norm": 0.5915543437004089,
      "learning_rate": 0.000464936967999138,
      "loss": 3.5534,
      "step": 21000
    },
    {
      "epoch": 2.2602518566354535,
      "eval_accuracy": 0.3639646830307332,
      "eval_loss": 3.5602052211761475,
      "eval_runtime": 184.9594,
      "eval_samples_per_second": 97.378,
      "eval_steps_per_second": 6.088,
      "step": 21000
    },
    {
      "epoch": 2.265633408675062,
      "grad_norm": 0.6229991316795349,
      "learning_rate": 0.0004646137269690766,
      "loss": 3.5646,
      "step": 21050
    },
    {
      "epoch": 2.27101496071467,
      "grad_norm": 0.5470733642578125,
      "learning_rate": 0.00046429048593901517,
      "loss": 3.5405,
      "step": 21100
    },
    {
      "epoch": 2.2763965127542782,
      "grad_norm": 0.5559133291244507,
      "learning_rate": 0.0004639672449089537,
      "loss": 3.5681,
      "step": 21150
    },
    {
      "epoch": 2.281778064793887,
      "grad_norm": 0.586371123790741,
      "learning_rate": 0.0004636440038788923,
      "loss": 3.5481,
      "step": 21200
    },
    {
      "epoch": 2.287159616833495,
      "grad_norm": 0.5637801885604858,
      "learning_rate": 0.00046332076284883095,
      "loss": 3.5569,
      "step": 21250
    },
    {
      "epoch": 2.292541168873103,
      "grad_norm": 0.6041868925094604,
      "learning_rate": 0.0004629975218187695,
      "loss": 3.5539,
      "step": 21300
    },
    {
      "epoch": 2.297922720912711,
      "grad_norm": 0.5667281746864319,
      "learning_rate": 0.0004626742807887081,
      "loss": 3.5323,
      "step": 21350
    },
    {
      "epoch": 2.303304272952319,
      "grad_norm": 0.5741965770721436,
      "learning_rate": 0.00046235103975864663,
      "loss": 3.5524,
      "step": 21400
    },
    {
      "epoch": 2.3086858249919278,
      "grad_norm": 0.5632966160774231,
      "learning_rate": 0.0004620277987285852,
      "loss": 3.5519,
      "step": 21450
    },
    {
      "epoch": 2.314067377031536,
      "grad_norm": 0.5545016527175903,
      "learning_rate": 0.0004617045576985239,
      "loss": 3.5487,
      "step": 21500
    },
    {
      "epoch": 2.319448929071144,
      "grad_norm": 0.5945175290107727,
      "learning_rate": 0.0004613813166684624,
      "loss": 3.5649,
      "step": 21550
    },
    {
      "epoch": 2.3248304811107525,
      "grad_norm": 0.5715595483779907,
      "learning_rate": 0.000461058075638401,
      "loss": 3.5639,
      "step": 21600
    },
    {
      "epoch": 2.3302120331503606,
      "grad_norm": 0.6180649399757385,
      "learning_rate": 0.00046073483460833955,
      "loss": 3.5566,
      "step": 21650
    },
    {
      "epoch": 2.3355935851899687,
      "grad_norm": 0.5411754250526428,
      "learning_rate": 0.00046041159357827814,
      "loss": 3.5633,
      "step": 21700
    },
    {
      "epoch": 2.340975137229577,
      "grad_norm": 0.5503979325294495,
      "learning_rate": 0.00046008835254821674,
      "loss": 3.5429,
      "step": 21750
    },
    {
      "epoch": 2.3463566892691854,
      "grad_norm": 0.5738011002540588,
      "learning_rate": 0.0004597651115181554,
      "loss": 3.5402,
      "step": 21800
    },
    {
      "epoch": 2.3517382413087935,
      "grad_norm": 0.5915544033050537,
      "learning_rate": 0.00045944187048809393,
      "loss": 3.5577,
      "step": 21850
    },
    {
      "epoch": 2.3571197933484016,
      "grad_norm": 0.6151456236839294,
      "learning_rate": 0.0004591186294580325,
      "loss": 3.5619,
      "step": 21900
    },
    {
      "epoch": 2.3625013453880097,
      "grad_norm": 0.5948041081428528,
      "learning_rate": 0.00045879538842797106,
      "loss": 3.5498,
      "step": 21950
    },
    {
      "epoch": 2.3678828974276183,
      "grad_norm": 0.597334086894989,
      "learning_rate": 0.00045847214739790966,
      "loss": 3.5551,
      "step": 22000
    },
    {
      "epoch": 2.3678828974276183,
      "eval_accuracy": 0.3649270216581061,
      "eval_loss": 3.5502870082855225,
      "eval_runtime": 186.0038,
      "eval_samples_per_second": 96.831,
      "eval_steps_per_second": 6.054,
      "step": 22000
    },
    {
      "epoch": 2.3732644494672264,
      "grad_norm": 0.556607186794281,
      "learning_rate": 0.0004581489063678482,
      "loss": 3.5555,
      "step": 22050
    },
    {
      "epoch": 2.3786460015068345,
      "grad_norm": 0.6205542087554932,
      "learning_rate": 0.00045782566533778685,
      "loss": 3.5576,
      "step": 22100
    },
    {
      "epoch": 2.384027553546443,
      "grad_norm": 0.5172244906425476,
      "learning_rate": 0.00045750242430772544,
      "loss": 3.5658,
      "step": 22150
    },
    {
      "epoch": 2.389409105586051,
      "grad_norm": 0.5946823358535767,
      "learning_rate": 0.000457179183277664,
      "loss": 3.5521,
      "step": 22200
    },
    {
      "epoch": 2.3947906576256592,
      "grad_norm": 0.600932240486145,
      "learning_rate": 0.0004568559422476026,
      "loss": 3.5612,
      "step": 22250
    },
    {
      "epoch": 2.4001722096652673,
      "grad_norm": 0.6086694598197937,
      "learning_rate": 0.00045653270121754117,
      "loss": 3.5581,
      "step": 22300
    },
    {
      "epoch": 2.4055537617048754,
      "grad_norm": 0.6031166911125183,
      "learning_rate": 0.0004562094601874797,
      "loss": 3.5667,
      "step": 22350
    },
    {
      "epoch": 2.410935313744484,
      "grad_norm": 0.5419747233390808,
      "learning_rate": 0.00045589268397801957,
      "loss": 3.5316,
      "step": 22400
    },
    {
      "epoch": 2.416316865784092,
      "grad_norm": 0.5568996667861938,
      "learning_rate": 0.00045556944294795817,
      "loss": 3.5659,
      "step": 22450
    },
    {
      "epoch": 2.4216984178237,
      "grad_norm": 0.5609857439994812,
      "learning_rate": 0.00045524620191789676,
      "loss": 3.5513,
      "step": 22500
    },
    {
      "epoch": 2.4270799698633088,
      "grad_norm": 0.5434156656265259,
      "learning_rate": 0.00045492296088783536,
      "loss": 3.5403,
      "step": 22550
    },
    {
      "epoch": 2.432461521902917,
      "grad_norm": 0.5359591841697693,
      "learning_rate": 0.0004545997198577739,
      "loss": 3.5464,
      "step": 22600
    },
    {
      "epoch": 2.437843073942525,
      "grad_norm": 0.6035705208778381,
      "learning_rate": 0.0004542764788277125,
      "loss": 3.5432,
      "step": 22650
    },
    {
      "epoch": 2.443224625982133,
      "grad_norm": 0.5715034008026123,
      "learning_rate": 0.00045395323779765103,
      "loss": 3.5508,
      "step": 22700
    },
    {
      "epoch": 2.4486061780217416,
      "grad_norm": 0.5985683798789978,
      "learning_rate": 0.0004536299967675897,
      "loss": 3.5508,
      "step": 22750
    },
    {
      "epoch": 2.4539877300613497,
      "grad_norm": 0.5577922463417053,
      "learning_rate": 0.0004533067557375283,
      "loss": 3.5565,
      "step": 22800
    },
    {
      "epoch": 2.459369282100958,
      "grad_norm": 0.6181262731552124,
      "learning_rate": 0.0004529835147074668,
      "loss": 3.5475,
      "step": 22850
    },
    {
      "epoch": 2.464750834140566,
      "grad_norm": 0.5955373644828796,
      "learning_rate": 0.0004526602736774054,
      "loss": 3.5359,
      "step": 22900
    },
    {
      "epoch": 2.4701323861801745,
      "grad_norm": 0.5582594275474548,
      "learning_rate": 0.00045233703264734395,
      "loss": 3.5376,
      "step": 22950
    },
    {
      "epoch": 2.4755139382197826,
      "grad_norm": 0.5990350246429443,
      "learning_rate": 0.00045201379161728255,
      "loss": 3.5504,
      "step": 23000
    },
    {
      "epoch": 2.4755139382197826,
      "eval_accuracy": 0.36597867295926245,
      "eval_loss": 3.541978597640991,
      "eval_runtime": 184.7351,
      "eval_samples_per_second": 97.496,
      "eval_steps_per_second": 6.095,
      "step": 23000
    },
    {
      "epoch": 2.4808954902593907,
      "grad_norm": 0.5420169234275818,
      "learning_rate": 0.0004516905505872212,
      "loss": 3.5606,
      "step": 23050
    },
    {
      "epoch": 2.4862770422989993,
      "grad_norm": 0.5671257972717285,
      "learning_rate": 0.00045136730955715973,
      "loss": 3.548,
      "step": 23100
    },
    {
      "epoch": 2.4916585943386074,
      "grad_norm": 0.6008259057998657,
      "learning_rate": 0.00045104406852709833,
      "loss": 3.5426,
      "step": 23150
    },
    {
      "epoch": 2.4970401463782155,
      "grad_norm": 0.5630355477333069,
      "learning_rate": 0.0004507208274970369,
      "loss": 3.5474,
      "step": 23200
    },
    {
      "epoch": 2.5024216984178236,
      "grad_norm": 0.6546486020088196,
      "learning_rate": 0.00045039758646697546,
      "loss": 3.5549,
      "step": 23250
    },
    {
      "epoch": 2.5078032504574317,
      "grad_norm": 0.5610313415527344,
      "learning_rate": 0.0004500743454369141,
      "loss": 3.5606,
      "step": 23300
    },
    {
      "epoch": 2.5131848024970402,
      "grad_norm": 0.6008824706077576,
      "learning_rate": 0.0004497511044068527,
      "loss": 3.554,
      "step": 23350
    },
    {
      "epoch": 2.5185663545366483,
      "grad_norm": 0.5663953423500061,
      "learning_rate": 0.00044942786337679125,
      "loss": 3.5343,
      "step": 23400
    },
    {
      "epoch": 2.5239479065762565,
      "grad_norm": 0.5768746733665466,
      "learning_rate": 0.00044910462234672984,
      "loss": 3.5481,
      "step": 23450
    },
    {
      "epoch": 2.529329458615865,
      "grad_norm": 0.5716328620910645,
      "learning_rate": 0.0004487813813166684,
      "loss": 3.5617,
      "step": 23500
    },
    {
      "epoch": 2.534711010655473,
      "grad_norm": 0.5698726773262024,
      "learning_rate": 0.000448458140286607,
      "loss": 3.5434,
      "step": 23550
    },
    {
      "epoch": 2.540092562695081,
      "grad_norm": 0.6069528460502625,
      "learning_rate": 0.00044813489925654563,
      "loss": 3.55,
      "step": 23600
    },
    {
      "epoch": 2.5454741147346893,
      "grad_norm": 0.5537393689155579,
      "learning_rate": 0.00044781165822648417,
      "loss": 3.5412,
      "step": 23650
    },
    {
      "epoch": 2.550855666774298,
      "grad_norm": 0.6079427003860474,
      "learning_rate": 0.00044748841719642276,
      "loss": 3.55,
      "step": 23700
    },
    {
      "epoch": 2.556237218813906,
      "grad_norm": 0.5689899921417236,
      "learning_rate": 0.00044716517616636136,
      "loss": 3.5378,
      "step": 23750
    },
    {
      "epoch": 2.561618770853514,
      "grad_norm": 0.5739214420318604,
      "learning_rate": 0.0004468419351362999,
      "loss": 3.5636,
      "step": 23800
    },
    {
      "epoch": 2.567000322893122,
      "grad_norm": 0.581843912601471,
      "learning_rate": 0.0004465186941062385,
      "loss": 3.5358,
      "step": 23850
    },
    {
      "epoch": 2.5723818749327307,
      "grad_norm": 0.5732669234275818,
      "learning_rate": 0.00044619545307617714,
      "loss": 3.5419,
      "step": 23900
    },
    {
      "epoch": 2.577763426972339,
      "grad_norm": 0.572491466999054,
      "learning_rate": 0.0004458722120461157,
      "loss": 3.5357,
      "step": 23950
    },
    {
      "epoch": 2.583144979011947,
      "grad_norm": 0.5753400325775146,
      "learning_rate": 0.0004455489710160543,
      "loss": 3.5292,
      "step": 24000
    },
    {
      "epoch": 2.583144979011947,
      "eval_accuracy": 0.36699979279894296,
      "eval_loss": 3.5284852981567383,
      "eval_runtime": 184.7189,
      "eval_samples_per_second": 97.505,
      "eval_steps_per_second": 6.096,
      "step": 24000
    },
    {
      "epoch": 2.5885265310515555,
      "grad_norm": 0.5776293873786926,
      "learning_rate": 0.0004452257299859928,
      "loss": 3.5573,
      "step": 24050
    },
    {
      "epoch": 2.5939080830911636,
      "grad_norm": 0.5396984815597534,
      "learning_rate": 0.0004449024889559314,
      "loss": 3.5348,
      "step": 24100
    },
    {
      "epoch": 2.5992896351307717,
      "grad_norm": 0.5629790425300598,
      "learning_rate": 0.00044457924792587,
      "loss": 3.5311,
      "step": 24150
    },
    {
      "epoch": 2.60467118717038,
      "grad_norm": 0.6055911779403687,
      "learning_rate": 0.0004442560068958086,
      "loss": 3.5521,
      "step": 24200
    },
    {
      "epoch": 2.610052739209988,
      "grad_norm": 0.6158533692359924,
      "learning_rate": 0.0004439327658657472,
      "loss": 3.5443,
      "step": 24250
    },
    {
      "epoch": 2.6154342912495965,
      "grad_norm": 0.5898953080177307,
      "learning_rate": 0.0004436095248356858,
      "loss": 3.5279,
      "step": 24300
    },
    {
      "epoch": 2.6208158432892046,
      "grad_norm": 0.6070363521575928,
      "learning_rate": 0.00044328628380562433,
      "loss": 3.5328,
      "step": 24350
    },
    {
      "epoch": 2.6261973953288127,
      "grad_norm": 0.6255342364311218,
      "learning_rate": 0.0004429630427755629,
      "loss": 3.5308,
      "step": 24400
    },
    {
      "epoch": 2.6315789473684212,
      "grad_norm": 0.5635362863540649,
      "learning_rate": 0.00044264626656610273,
      "loss": 3.5503,
      "step": 24450
    },
    {
      "epoch": 2.6369604994080293,
      "grad_norm": 0.5545345544815063,
      "learning_rate": 0.0004423230255360413,
      "loss": 3.5512,
      "step": 24500
    },
    {
      "epoch": 2.6423420514476375,
      "grad_norm": 0.5683971047401428,
      "learning_rate": 0.00044199978450598,
      "loss": 3.5455,
      "step": 24550
    },
    {
      "epoch": 2.6477236034872456,
      "grad_norm": 0.5963163375854492,
      "learning_rate": 0.0004416765434759185,
      "loss": 3.5487,
      "step": 24600
    },
    {
      "epoch": 2.653105155526854,
      "grad_norm": 0.5661613941192627,
      "learning_rate": 0.0004413533024458571,
      "loss": 3.5521,
      "step": 24650
    },
    {
      "epoch": 2.658486707566462,
      "grad_norm": 0.5891050100326538,
      "learning_rate": 0.00044103006141579565,
      "loss": 3.5376,
      "step": 24700
    },
    {
      "epoch": 2.6638682596060703,
      "grad_norm": 0.6227092146873474,
      "learning_rate": 0.00044070682038573425,
      "loss": 3.5448,
      "step": 24750
    },
    {
      "epoch": 2.6692498116456784,
      "grad_norm": 0.602344810962677,
      "learning_rate": 0.0004403835793556728,
      "loss": 3.5597,
      "step": 24800
    },
    {
      "epoch": 2.674631363685287,
      "grad_norm": 0.5948516726493835,
      "learning_rate": 0.00044006033832561143,
      "loss": 3.5432,
      "step": 24850
    },
    {
      "epoch": 2.680012915724895,
      "grad_norm": 0.562466561794281,
      "learning_rate": 0.00043973709729555003,
      "loss": 3.5447,
      "step": 24900
    },
    {
      "epoch": 2.685394467764503,
      "grad_norm": 0.5580047369003296,
      "learning_rate": 0.00043941385626548857,
      "loss": 3.5319,
      "step": 24950
    },
    {
      "epoch": 2.6907760198041117,
      "grad_norm": 0.6161678433418274,
      "learning_rate": 0.00043909061523542716,
      "loss": 3.548,
      "step": 25000
    },
    {
      "epoch": 2.6907760198041117,
      "eval_accuracy": 0.3677566688154586,
      "eval_loss": 3.519853353500366,
      "eval_runtime": 184.6274,
      "eval_samples_per_second": 97.553,
      "eval_steps_per_second": 6.099,
      "step": 25000
    },
    {
      "epoch": 2.69615757184372,
      "grad_norm": 0.5613181591033936,
      "learning_rate": 0.00043876737420536576,
      "loss": 3.5378,
      "step": 25050
    },
    {
      "epoch": 2.701539123883328,
      "grad_norm": 0.5858383178710938,
      "learning_rate": 0.00043844413317530435,
      "loss": 3.543,
      "step": 25100
    },
    {
      "epoch": 2.706920675922936,
      "grad_norm": 0.584028959274292,
      "learning_rate": 0.00043812089214524295,
      "loss": 3.5279,
      "step": 25150
    },
    {
      "epoch": 2.712302227962544,
      "grad_norm": 0.6211503744125366,
      "learning_rate": 0.00043779765111518154,
      "loss": 3.542,
      "step": 25200
    },
    {
      "epoch": 2.7176837800021527,
      "grad_norm": 0.6113668084144592,
      "learning_rate": 0.0004374744100851201,
      "loss": 3.5454,
      "step": 25250
    },
    {
      "epoch": 2.723065332041761,
      "grad_norm": 0.6018054485321045,
      "learning_rate": 0.0004371511690550587,
      "loss": 3.5472,
      "step": 25300
    },
    {
      "epoch": 2.728446884081369,
      "grad_norm": 0.581635057926178,
      "learning_rate": 0.0004368279280249972,
      "loss": 3.5521,
      "step": 25350
    },
    {
      "epoch": 2.7338284361209775,
      "grad_norm": 0.6109009385108948,
      "learning_rate": 0.00043650468699493587,
      "loss": 3.5456,
      "step": 25400
    },
    {
      "epoch": 2.7392099881605856,
      "grad_norm": 0.6068916320800781,
      "learning_rate": 0.00043618144596487446,
      "loss": 3.5174,
      "step": 25450
    },
    {
      "epoch": 2.7445915402001937,
      "grad_norm": 0.5705057978630066,
      "learning_rate": 0.000435858204934813,
      "loss": 3.5223,
      "step": 25500
    },
    {
      "epoch": 2.749973092239802,
      "grad_norm": 0.5605146288871765,
      "learning_rate": 0.0004355349639047516,
      "loss": 3.5414,
      "step": 25550
    },
    {
      "epoch": 2.7553546442794103,
      "grad_norm": 0.5750584602355957,
      "learning_rate": 0.0004352117228746902,
      "loss": 3.5424,
      "step": 25600
    },
    {
      "epoch": 2.7607361963190185,
      "grad_norm": 0.58180171251297,
      "learning_rate": 0.00043488848184462873,
      "loss": 3.5355,
      "step": 25650
    },
    {
      "epoch": 2.7661177483586266,
      "grad_norm": 0.6068442463874817,
      "learning_rate": 0.0004345652408145674,
      "loss": 3.5318,
      "step": 25700
    },
    {
      "epoch": 2.7714993003982347,
      "grad_norm": 0.6122245192527771,
      "learning_rate": 0.000434241999784506,
      "loss": 3.5256,
      "step": 25750
    },
    {
      "epoch": 2.776880852437843,
      "grad_norm": 0.5642405152320862,
      "learning_rate": 0.0004339187587544445,
      "loss": 3.5298,
      "step": 25800
    },
    {
      "epoch": 2.7822624044774513,
      "grad_norm": 0.604735791683197,
      "learning_rate": 0.0004335955177243831,
      "loss": 3.5435,
      "step": 25850
    },
    {
      "epoch": 2.7876439565170594,
      "grad_norm": 0.5870072841644287,
      "learning_rate": 0.00043327227669432165,
      "loss": 3.5304,
      "step": 25900
    },
    {
      "epoch": 2.793025508556668,
      "grad_norm": 0.5905956029891968,
      "learning_rate": 0.00043294903566426025,
      "loss": 3.5234,
      "step": 25950
    },
    {
      "epoch": 2.798407060596276,
      "grad_norm": 0.5809804201126099,
      "learning_rate": 0.0004326257946341989,
      "loss": 3.5277,
      "step": 26000
    },
    {
      "epoch": 2.798407060596276,
      "eval_accuracy": 0.368913713417795,
      "eval_loss": 3.5110578536987305,
      "eval_runtime": 184.5749,
      "eval_samples_per_second": 97.581,
      "eval_steps_per_second": 6.101,
      "step": 26000
    },
    {
      "epoch": 2.803788612635884,
      "grad_norm": 0.5987585186958313,
      "learning_rate": 0.00043230255360413744,
      "loss": 3.5359,
      "step": 26050
    },
    {
      "epoch": 2.8091701646754923,
      "grad_norm": 0.6006620526313782,
      "learning_rate": 0.00043197931257407603,
      "loss": 3.534,
      "step": 26100
    },
    {
      "epoch": 2.8145517167151004,
      "grad_norm": 0.5650146007537842,
      "learning_rate": 0.0004316560715440146,
      "loss": 3.5355,
      "step": 26150
    },
    {
      "epoch": 2.819933268754709,
      "grad_norm": 0.6045055985450745,
      "learning_rate": 0.00043133283051395317,
      "loss": 3.5055,
      "step": 26200
    },
    {
      "epoch": 2.825314820794317,
      "grad_norm": 0.5808995962142944,
      "learning_rate": 0.0004310095894838918,
      "loss": 3.5506,
      "step": 26250
    },
    {
      "epoch": 2.830696372833925,
      "grad_norm": 0.6251314878463745,
      "learning_rate": 0.0004306863484538304,
      "loss": 3.5199,
      "step": 26300
    },
    {
      "epoch": 2.8360779248735337,
      "grad_norm": 0.5796584486961365,
      "learning_rate": 0.00043036310742376895,
      "loss": 3.5251,
      "step": 26350
    },
    {
      "epoch": 2.841459476913142,
      "grad_norm": 0.5469850301742554,
      "learning_rate": 0.00043003986639370754,
      "loss": 3.524,
      "step": 26400
    },
    {
      "epoch": 2.84684102895275,
      "grad_norm": 0.5464411377906799,
      "learning_rate": 0.0004297166253636461,
      "loss": 3.5052,
      "step": 26450
    },
    {
      "epoch": 2.852222580992358,
      "grad_norm": 0.6001213788986206,
      "learning_rate": 0.0004293933843335847,
      "loss": 3.5216,
      "step": 26500
    },
    {
      "epoch": 2.857604133031966,
      "grad_norm": 0.6024080514907837,
      "learning_rate": 0.00042907014330352333,
      "loss": 3.5236,
      "step": 26550
    },
    {
      "epoch": 2.8629856850715747,
      "grad_norm": 0.5663939118385315,
      "learning_rate": 0.00042874690227346187,
      "loss": 3.5461,
      "step": 26600
    },
    {
      "epoch": 2.868367237111183,
      "grad_norm": 0.5653455853462219,
      "learning_rate": 0.00042842366124340046,
      "loss": 3.5291,
      "step": 26650
    },
    {
      "epoch": 2.873748789150791,
      "grad_norm": 0.5584545135498047,
      "learning_rate": 0.00042810042021333906,
      "loss": 3.5283,
      "step": 26700
    },
    {
      "epoch": 2.8791303411903995,
      "grad_norm": 0.5881627798080444,
      "learning_rate": 0.00042778364400387886,
      "loss": 3.5353,
      "step": 26750
    },
    {
      "epoch": 2.8845118932300076,
      "grad_norm": 0.5636775493621826,
      "learning_rate": 0.0004274604029738174,
      "loss": 3.5406,
      "step": 26800
    },
    {
      "epoch": 2.8898934452696157,
      "grad_norm": 0.6478109359741211,
      "learning_rate": 0.000427137161943756,
      "loss": 3.5313,
      "step": 26850
    },
    {
      "epoch": 2.895274997309224,
      "grad_norm": 0.6400372385978699,
      "learning_rate": 0.00042681392091369465,
      "loss": 3.5297,
      "step": 26900
    },
    {
      "epoch": 2.9006565493488323,
      "grad_norm": 0.6088976263999939,
      "learning_rate": 0.0004264906798836332,
      "loss": 3.5215,
      "step": 26950
    },
    {
      "epoch": 2.9060381013884404,
      "grad_norm": 0.6216518878936768,
      "learning_rate": 0.0004261674388535718,
      "loss": 3.5187,
      "step": 27000
    },
    {
      "epoch": 2.9060381013884404,
      "eval_accuracy": 0.3694821853268404,
      "eval_loss": 3.503814697265625,
      "eval_runtime": 184.8292,
      "eval_samples_per_second": 97.447,
      "eval_steps_per_second": 6.092,
      "step": 27000
    },
    {
      "epoch": 2.9114196534280485,
      "grad_norm": 0.5948754549026489,
      "learning_rate": 0.0004258441978235104,
      "loss": 3.5347,
      "step": 27050
    },
    {
      "epoch": 2.9168012054676566,
      "grad_norm": 0.5998795032501221,
      "learning_rate": 0.0004255209567934489,
      "loss": 3.5336,
      "step": 27100
    },
    {
      "epoch": 2.922182757507265,
      "grad_norm": 0.6143434643745422,
      "learning_rate": 0.0004251977157633875,
      "loss": 3.5265,
      "step": 27150
    },
    {
      "epoch": 2.9275643095468733,
      "grad_norm": 0.590839147567749,
      "learning_rate": 0.00042487447473332616,
      "loss": 3.5134,
      "step": 27200
    },
    {
      "epoch": 2.9329458615864814,
      "grad_norm": 0.5512373447418213,
      "learning_rate": 0.0004245512337032647,
      "loss": 3.5192,
      "step": 27250
    },
    {
      "epoch": 2.93832741362609,
      "grad_norm": 0.636211633682251,
      "learning_rate": 0.0004242279926732033,
      "loss": 3.527,
      "step": 27300
    },
    {
      "epoch": 2.943708965665698,
      "grad_norm": 0.5445491671562195,
      "learning_rate": 0.00042390475164314184,
      "loss": 3.5407,
      "step": 27350
    },
    {
      "epoch": 2.949090517705306,
      "grad_norm": 0.5808195471763611,
      "learning_rate": 0.00042358151061308043,
      "loss": 3.5187,
      "step": 27400
    },
    {
      "epoch": 2.9544720697449143,
      "grad_norm": 0.5917678475379944,
      "learning_rate": 0.000423258269583019,
      "loss": 3.5374,
      "step": 27450
    },
    {
      "epoch": 2.9598536217845224,
      "grad_norm": 0.6088602542877197,
      "learning_rate": 0.0004229350285529576,
      "loss": 3.5279,
      "step": 27500
    },
    {
      "epoch": 2.965235173824131,
      "grad_norm": 0.6586233377456665,
      "learning_rate": 0.0004226117875228962,
      "loss": 3.5174,
      "step": 27550
    },
    {
      "epoch": 2.970616725863739,
      "grad_norm": 0.5706504583358765,
      "learning_rate": 0.0004222885464928348,
      "loss": 3.5181,
      "step": 27600
    },
    {
      "epoch": 2.975998277903347,
      "grad_norm": 0.5670300722122192,
      "learning_rate": 0.00042196530546277335,
      "loss": 3.539,
      "step": 27650
    },
    {
      "epoch": 2.9813798299429557,
      "grad_norm": 0.5491836667060852,
      "learning_rate": 0.00042164206443271195,
      "loss": 3.5378,
      "step": 27700
    },
    {
      "epoch": 2.986761381982564,
      "grad_norm": 0.607046902179718,
      "learning_rate": 0.0004213188234026505,
      "loss": 3.5157,
      "step": 27750
    },
    {
      "epoch": 2.992142934022172,
      "grad_norm": 0.5921494364738464,
      "learning_rate": 0.00042099558237258914,
      "loss": 3.5087,
      "step": 27800
    },
    {
      "epoch": 2.9975244860617805,
      "grad_norm": 0.5658209323883057,
      "learning_rate": 0.00042067234134252773,
      "loss": 3.5393,
      "step": 27850
    },
    {
      "epoch": 3.0029060381013886,
      "grad_norm": 0.5574458837509155,
      "learning_rate": 0.00042034910031246627,
      "loss": 3.4678,
      "step": 27900
    },
    {
      "epoch": 3.0082875901409967,
      "grad_norm": 0.6002983450889587,
      "learning_rate": 0.00042002585928240486,
      "loss": 3.4162,
      "step": 27950
    },
    {
      "epoch": 3.0136691421806048,
      "grad_norm": 0.5765467286109924,
      "learning_rate": 0.00041970261825234346,
      "loss": 3.4414,
      "step": 28000
    },
    {
      "epoch": 3.0136691421806048,
      "eval_accuracy": 0.37054198559458285,
      "eval_loss": 3.4985103607177734,
      "eval_runtime": 184.6194,
      "eval_samples_per_second": 97.557,
      "eval_steps_per_second": 6.099,
      "step": 28000
    },
    {
      "epoch": 3.0190506942202133,
      "grad_norm": 0.6787931323051453,
      "learning_rate": 0.00041937937722228205,
      "loss": 3.4392,
      "step": 28050
    },
    {
      "epoch": 3.0244322462598214,
      "grad_norm": 0.5360280871391296,
      "learning_rate": 0.00041905613619222065,
      "loss": 3.4451,
      "step": 28100
    },
    {
      "epoch": 3.0298137982994295,
      "grad_norm": 0.6370947957038879,
      "learning_rate": 0.00041873289516215924,
      "loss": 3.4504,
      "step": 28150
    },
    {
      "epoch": 3.0351953503390376,
      "grad_norm": 0.6381440162658691,
      "learning_rate": 0.0004184096541320978,
      "loss": 3.4473,
      "step": 28200
    },
    {
      "epoch": 3.040576902378646,
      "grad_norm": 0.5641018152236938,
      "learning_rate": 0.0004180864131020364,
      "loss": 3.4285,
      "step": 28250
    },
    {
      "epoch": 3.0459584544182543,
      "grad_norm": 0.573603630065918,
      "learning_rate": 0.0004177631720719749,
      "loss": 3.4438,
      "step": 28300
    },
    {
      "epoch": 3.0513400064578624,
      "grad_norm": 0.635581374168396,
      "learning_rate": 0.00041743993104191357,
      "loss": 3.4441,
      "step": 28350
    },
    {
      "epoch": 3.0567215584974705,
      "grad_norm": 0.6056490540504456,
      "learning_rate": 0.00041711669001185216,
      "loss": 3.4447,
      "step": 28400
    },
    {
      "epoch": 3.062103110537079,
      "grad_norm": 0.6397773623466492,
      "learning_rate": 0.0004167934489817907,
      "loss": 3.4359,
      "step": 28450
    },
    {
      "epoch": 3.067484662576687,
      "grad_norm": 0.570362389087677,
      "learning_rate": 0.0004164702079517293,
      "loss": 3.4428,
      "step": 28500
    },
    {
      "epoch": 3.0728662146162953,
      "grad_norm": 0.6196599006652832,
      "learning_rate": 0.0004161469669216679,
      "loss": 3.4492,
      "step": 28550
    },
    {
      "epoch": 3.0782477666559034,
      "grad_norm": 0.5860263705253601,
      "learning_rate": 0.00041582372589160643,
      "loss": 3.4482,
      "step": 28600
    },
    {
      "epoch": 3.083629318695512,
      "grad_norm": 0.6199707388877869,
      "learning_rate": 0.0004155004848615451,
      "loss": 3.451,
      "step": 28650
    },
    {
      "epoch": 3.08901087073512,
      "grad_norm": 0.6375261545181274,
      "learning_rate": 0.0004151772438314837,
      "loss": 3.4516,
      "step": 28700
    },
    {
      "epoch": 3.094392422774728,
      "grad_norm": 0.5758550763130188,
      "learning_rate": 0.0004148540028014222,
      "loss": 3.4535,
      "step": 28750
    },
    {
      "epoch": 3.0997739748143363,
      "grad_norm": 0.6565601825714111,
      "learning_rate": 0.0004145307617713608,
      "loss": 3.4571,
      "step": 28800
    },
    {
      "epoch": 3.105155526853945,
      "grad_norm": 0.5880964994430542,
      "learning_rate": 0.00041420752074129935,
      "loss": 3.432,
      "step": 28850
    },
    {
      "epoch": 3.110537078893553,
      "grad_norm": 0.585135281085968,
      "learning_rate": 0.000413884279711238,
      "loss": 3.4393,
      "step": 28900
    },
    {
      "epoch": 3.115918630933161,
      "grad_norm": 0.5878077745437622,
      "learning_rate": 0.0004135610386811766,
      "loss": 3.457,
      "step": 28950
    },
    {
      "epoch": 3.121300182972769,
      "grad_norm": 0.5695430636405945,
      "learning_rate": 0.0004132442624717164,
      "loss": 3.4285,
      "step": 29000
    },
    {
      "epoch": 3.121300182972769,
      "eval_accuracy": 0.3709822470960071,
      "eval_loss": 3.4938888549804688,
      "eval_runtime": 184.3826,
      "eval_samples_per_second": 97.683,
      "eval_steps_per_second": 6.107,
      "step": 29000
    },
    {
      "epoch": 3.1266817350123777,
      "grad_norm": 0.5889701843261719,
      "learning_rate": 0.000412921021441655,
      "loss": 3.4346,
      "step": 29050
    },
    {
      "epoch": 3.132063287051986,
      "grad_norm": 0.5934194326400757,
      "learning_rate": 0.00041259778041159354,
      "loss": 3.4426,
      "step": 29100
    },
    {
      "epoch": 3.137444839091594,
      "grad_norm": 0.5477867126464844,
      "learning_rate": 0.00041227453938153213,
      "loss": 3.4467,
      "step": 29150
    },
    {
      "epoch": 3.1428263911312024,
      "grad_norm": 0.5779726505279541,
      "learning_rate": 0.00041195129835147067,
      "loss": 3.4411,
      "step": 29200
    },
    {
      "epoch": 3.1482079431708105,
      "grad_norm": 0.5577905178070068,
      "learning_rate": 0.00041162805732140927,
      "loss": 3.4518,
      "step": 29250
    },
    {
      "epoch": 3.1535894952104186,
      "grad_norm": 0.5991485714912415,
      "learning_rate": 0.0004113048162913479,
      "loss": 3.4528,
      "step": 29300
    },
    {
      "epoch": 3.1589710472500268,
      "grad_norm": 0.5515627264976501,
      "learning_rate": 0.00041098157526128646,
      "loss": 3.4553,
      "step": 29350
    },
    {
      "epoch": 3.1643525992896353,
      "grad_norm": 0.577649712562561,
      "learning_rate": 0.00041065833423122505,
      "loss": 3.4708,
      "step": 29400
    },
    {
      "epoch": 3.1697341513292434,
      "grad_norm": 0.609976589679718,
      "learning_rate": 0.00041033509320116365,
      "loss": 3.4677,
      "step": 29450
    },
    {
      "epoch": 3.1751157033688515,
      "grad_norm": 0.5896977186203003,
      "learning_rate": 0.0004100118521711022,
      "loss": 3.4464,
      "step": 29500
    },
    {
      "epoch": 3.1804972554084596,
      "grad_norm": 0.5586709976196289,
      "learning_rate": 0.0004096886111410408,
      "loss": 3.4503,
      "step": 29550
    },
    {
      "epoch": 3.185878807448068,
      "grad_norm": 0.6799648404121399,
      "learning_rate": 0.00040936537011097943,
      "loss": 3.4442,
      "step": 29600
    },
    {
      "epoch": 3.1912603594876763,
      "grad_norm": 0.5831512808799744,
      "learning_rate": 0.00040904212908091797,
      "loss": 3.4351,
      "step": 29650
    },
    {
      "epoch": 3.1966419115272844,
      "grad_norm": 0.6388946771621704,
      "learning_rate": 0.00040871888805085656,
      "loss": 3.4333,
      "step": 29700
    },
    {
      "epoch": 3.2020234635668925,
      "grad_norm": 0.6149418950080872,
      "learning_rate": 0.0004083956470207951,
      "loss": 3.4522,
      "step": 29750
    },
    {
      "epoch": 3.207405015606501,
      "grad_norm": 0.5800194144248962,
      "learning_rate": 0.0004080724059907337,
      "loss": 3.4605,
      "step": 29800
    },
    {
      "epoch": 3.212786567646109,
      "grad_norm": 0.665179967880249,
      "learning_rate": 0.00040774916496067235,
      "loss": 3.4525,
      "step": 29850
    },
    {
      "epoch": 3.2181681196857173,
      "grad_norm": 0.6393986344337463,
      "learning_rate": 0.0004074259239306109,
      "loss": 3.4625,
      "step": 29900
    },
    {
      "epoch": 3.2235496717253254,
      "grad_norm": 0.5729492902755737,
      "learning_rate": 0.0004071026829005495,
      "loss": 3.448,
      "step": 29950
    },
    {
      "epoch": 3.228931223764934,
      "grad_norm": 0.6174540519714355,
      "learning_rate": 0.0004067794418704881,
      "loss": 3.4394,
      "step": 30000
    },
    {
      "epoch": 3.228931223764934,
      "eval_accuracy": 0.3717323866334783,
      "eval_loss": 3.488069772720337,
      "eval_runtime": 184.6952,
      "eval_samples_per_second": 97.517,
      "eval_steps_per_second": 6.097,
      "step": 30000
    },
    {
      "epoch": 3.234312775804542,
      "grad_norm": 0.5699408650398254,
      "learning_rate": 0.0004064562008404266,
      "loss": 3.4601,
      "step": 30050
    },
    {
      "epoch": 3.23969432784415,
      "grad_norm": 0.5962291955947876,
      "learning_rate": 0.0004061329598103652,
      "loss": 3.4724,
      "step": 30100
    },
    {
      "epoch": 3.2450758798837587,
      "grad_norm": 0.6102081537246704,
      "learning_rate": 0.00040580971878030386,
      "loss": 3.4567,
      "step": 30150
    },
    {
      "epoch": 3.250457431923367,
      "grad_norm": 0.6508827805519104,
      "learning_rate": 0.0004054864777502424,
      "loss": 3.4537,
      "step": 30200
    },
    {
      "epoch": 3.255838983962975,
      "grad_norm": 0.6434182524681091,
      "learning_rate": 0.000405163236720181,
      "loss": 3.4557,
      "step": 30250
    },
    {
      "epoch": 3.261220536002583,
      "grad_norm": 0.5960790514945984,
      "learning_rate": 0.00040483999569011954,
      "loss": 3.4562,
      "step": 30300
    },
    {
      "epoch": 3.2666020880421915,
      "grad_norm": 0.5768137574195862,
      "learning_rate": 0.00040451675466005813,
      "loss": 3.4538,
      "step": 30350
    },
    {
      "epoch": 3.2719836400817996,
      "grad_norm": 0.6380401849746704,
      "learning_rate": 0.0004041935136299967,
      "loss": 3.4476,
      "step": 30400
    },
    {
      "epoch": 3.2773651921214078,
      "grad_norm": 0.6057765483856201,
      "learning_rate": 0.0004038702725999353,
      "loss": 3.4356,
      "step": 30450
    },
    {
      "epoch": 3.282746744161016,
      "grad_norm": 0.5699383020401001,
      "learning_rate": 0.0004035470315698739,
      "loss": 3.452,
      "step": 30500
    },
    {
      "epoch": 3.2881282962006244,
      "grad_norm": 0.5979689359664917,
      "learning_rate": 0.0004032237905398125,
      "loss": 3.4618,
      "step": 30550
    },
    {
      "epoch": 3.2935098482402325,
      "grad_norm": 0.6368375420570374,
      "learning_rate": 0.00040290054950975105,
      "loss": 3.4628,
      "step": 30600
    },
    {
      "epoch": 3.2988914002798406,
      "grad_norm": 0.6144343614578247,
      "learning_rate": 0.00040257730847968965,
      "loss": 3.4395,
      "step": 30650
    },
    {
      "epoch": 3.304272952319449,
      "grad_norm": 0.5661136507987976,
      "learning_rate": 0.0004022540674496283,
      "loss": 3.469,
      "step": 30700
    },
    {
      "epoch": 3.3096545043590573,
      "grad_norm": 0.6017364263534546,
      "learning_rate": 0.00040193082641956684,
      "loss": 3.4608,
      "step": 30750
    },
    {
      "epoch": 3.3150360563986654,
      "grad_norm": 0.5940915942192078,
      "learning_rate": 0.00040160758538950543,
      "loss": 3.4437,
      "step": 30800
    },
    {
      "epoch": 3.3204176084382735,
      "grad_norm": 0.6294538974761963,
      "learning_rate": 0.00040128434435944397,
      "loss": 3.4343,
      "step": 30850
    },
    {
      "epoch": 3.3257991604778816,
      "grad_norm": 0.650331437587738,
      "learning_rate": 0.00040096110332938257,
      "loss": 3.4636,
      "step": 30900
    },
    {
      "epoch": 3.33118071251749,
      "grad_norm": 0.6346154808998108,
      "learning_rate": 0.00040063786229932116,
      "loss": 3.4786,
      "step": 30950
    },
    {
      "epoch": 3.3365622645570983,
      "grad_norm": 0.5942608714103699,
      "learning_rate": 0.00040031462126925975,
      "loss": 3.4575,
      "step": 31000
    },
    {
      "epoch": 3.3365622645570983,
      "eval_accuracy": 0.3724254834048469,
      "eval_loss": 3.4845597743988037,
      "eval_runtime": 184.6678,
      "eval_samples_per_second": 97.532,
      "eval_steps_per_second": 6.097,
      "step": 31000
    },
    {
      "epoch": 3.3419438165967064,
      "grad_norm": 0.5669004321098328,
      "learning_rate": 0.00039999138023919835,
      "loss": 3.4639,
      "step": 31050
    },
    {
      "epoch": 3.347325368636315,
      "grad_norm": 0.6496381759643555,
      "learning_rate": 0.00039966813920913694,
      "loss": 3.457,
      "step": 31100
    },
    {
      "epoch": 3.352706920675923,
      "grad_norm": 0.5670893788337708,
      "learning_rate": 0.0003993448981790755,
      "loss": 3.4657,
      "step": 31150
    },
    {
      "epoch": 3.358088472715531,
      "grad_norm": 0.608622670173645,
      "learning_rate": 0.0003990216571490141,
      "loss": 3.437,
      "step": 31200
    },
    {
      "epoch": 3.3634700247551392,
      "grad_norm": 0.5956699848175049,
      "learning_rate": 0.0003986984161189526,
      "loss": 3.463,
      "step": 31250
    },
    {
      "epoch": 3.368851576794748,
      "grad_norm": 0.6540201306343079,
      "learning_rate": 0.00039837517508889127,
      "loss": 3.4603,
      "step": 31300
    },
    {
      "epoch": 3.374233128834356,
      "grad_norm": 0.5587558150291443,
      "learning_rate": 0.000398058398879431,
      "loss": 3.4485,
      "step": 31350
    },
    {
      "epoch": 3.379614680873964,
      "grad_norm": 0.5995318293571472,
      "learning_rate": 0.00039773515784936967,
      "loss": 3.4373,
      "step": 31400
    },
    {
      "epoch": 3.384996232913572,
      "grad_norm": 0.60701984167099,
      "learning_rate": 0.00039741191681930826,
      "loss": 3.4547,
      "step": 31450
    },
    {
      "epoch": 3.3903777849531807,
      "grad_norm": 0.5675340890884399,
      "learning_rate": 0.0003970886757892468,
      "loss": 3.4381,
      "step": 31500
    },
    {
      "epoch": 3.3957593369927888,
      "grad_norm": 0.5904266238212585,
      "learning_rate": 0.0003967654347591854,
      "loss": 3.4733,
      "step": 31550
    },
    {
      "epoch": 3.401140889032397,
      "grad_norm": 0.5990027189254761,
      "learning_rate": 0.00039644219372912394,
      "loss": 3.4597,
      "step": 31600
    },
    {
      "epoch": 3.4065224410720054,
      "grad_norm": 0.6386702656745911,
      "learning_rate": 0.0003961189526990626,
      "loss": 3.4413,
      "step": 31650
    },
    {
      "epoch": 3.4119039931116135,
      "grad_norm": 0.6855025887489319,
      "learning_rate": 0.0003957957116690012,
      "loss": 3.4663,
      "step": 31700
    },
    {
      "epoch": 3.4172855451512216,
      "grad_norm": 0.6087505221366882,
      "learning_rate": 0.0003954724706389397,
      "loss": 3.4613,
      "step": 31750
    },
    {
      "epoch": 3.4226670971908297,
      "grad_norm": 0.6036788821220398,
      "learning_rate": 0.0003951492296088783,
      "loss": 3.45,
      "step": 31800
    },
    {
      "epoch": 3.428048649230438,
      "grad_norm": 0.6118823289871216,
      "learning_rate": 0.0003948259885788169,
      "loss": 3.4542,
      "step": 31850
    },
    {
      "epoch": 3.4334302012700464,
      "grad_norm": 0.6175610423088074,
      "learning_rate": 0.00039450274754875545,
      "loss": 3.4705,
      "step": 31900
    },
    {
      "epoch": 3.4388117533096545,
      "grad_norm": 0.5950233340263367,
      "learning_rate": 0.0003941795065186941,
      "loss": 3.4619,
      "step": 31950
    },
    {
      "epoch": 3.4441933053492626,
      "grad_norm": 0.6266686320304871,
      "learning_rate": 0.0003938562654886327,
      "loss": 3.4513,
      "step": 32000
    },
    {
      "epoch": 3.4441933053492626,
      "eval_accuracy": 0.37286737469958836,
      "eval_loss": 3.4757885932922363,
      "eval_runtime": 184.4894,
      "eval_samples_per_second": 97.626,
      "eval_steps_per_second": 6.103,
      "step": 32000
    },
    {
      "epoch": 3.449574857388871,
      "grad_norm": 0.5964539647102356,
      "learning_rate": 0.00039353302445857124,
      "loss": 3.4683,
      "step": 32050
    },
    {
      "epoch": 3.4549564094284793,
      "grad_norm": 0.66469806432724,
      "learning_rate": 0.00039320978342850983,
      "loss": 3.4752,
      "step": 32100
    },
    {
      "epoch": 3.4603379614680874,
      "grad_norm": 0.5947219729423523,
      "learning_rate": 0.00039288654239844837,
      "loss": 3.4626,
      "step": 32150
    },
    {
      "epoch": 3.4657195135076955,
      "grad_norm": 0.6041039824485779,
      "learning_rate": 0.00039256330136838697,
      "loss": 3.4448,
      "step": 32200
    },
    {
      "epoch": 3.471101065547304,
      "grad_norm": 0.6173006296157837,
      "learning_rate": 0.0003922400603383256,
      "loss": 3.4376,
      "step": 32250
    },
    {
      "epoch": 3.476482617586912,
      "grad_norm": 0.6498045921325684,
      "learning_rate": 0.00039191681930826416,
      "loss": 3.4638,
      "step": 32300
    },
    {
      "epoch": 3.4818641696265202,
      "grad_norm": 0.6071352958679199,
      "learning_rate": 0.00039159357827820275,
      "loss": 3.4619,
      "step": 32350
    },
    {
      "epoch": 3.4872457216661283,
      "grad_norm": 0.6316366195678711,
      "learning_rate": 0.00039127033724814135,
      "loss": 3.448,
      "step": 32400
    },
    {
      "epoch": 3.492627273705737,
      "grad_norm": 0.6110300421714783,
      "learning_rate": 0.0003909470962180799,
      "loss": 3.4594,
      "step": 32450
    },
    {
      "epoch": 3.498008825745345,
      "grad_norm": 0.5853701829910278,
      "learning_rate": 0.00039062385518801854,
      "loss": 3.4595,
      "step": 32500
    },
    {
      "epoch": 3.503390377784953,
      "grad_norm": 0.7893177270889282,
      "learning_rate": 0.00039030061415795713,
      "loss": 3.4567,
      "step": 32550
    },
    {
      "epoch": 3.5087719298245617,
      "grad_norm": 0.5927507877349854,
      "learning_rate": 0.00038997737312789567,
      "loss": 3.4584,
      "step": 32600
    },
    {
      "epoch": 3.5141534818641698,
      "grad_norm": 0.5924676060676575,
      "learning_rate": 0.00038965413209783426,
      "loss": 3.4384,
      "step": 32650
    },
    {
      "epoch": 3.519535033903778,
      "grad_norm": 0.5994710922241211,
      "learning_rate": 0.0003893308910677728,
      "loss": 3.4538,
      "step": 32700
    },
    {
      "epoch": 3.524916585943386,
      "grad_norm": 0.6486222743988037,
      "learning_rate": 0.0003890076500377114,
      "loss": 3.4661,
      "step": 32750
    },
    {
      "epoch": 3.530298137982994,
      "grad_norm": 0.636140763759613,
      "learning_rate": 0.00038868440900765005,
      "loss": 3.4545,
      "step": 32800
    },
    {
      "epoch": 3.5356796900226026,
      "grad_norm": 0.6097226142883301,
      "learning_rate": 0.0003883611679775886,
      "loss": 3.4571,
      "step": 32850
    },
    {
      "epoch": 3.5410612420622107,
      "grad_norm": 0.5975202322006226,
      "learning_rate": 0.0003880379269475272,
      "loss": 3.45,
      "step": 32900
    },
    {
      "epoch": 3.546442794101819,
      "grad_norm": 0.6173543334007263,
      "learning_rate": 0.0003877146859174657,
      "loss": 3.4763,
      "step": 32950
    },
    {
      "epoch": 3.5518243461414274,
      "grad_norm": 0.5743898153305054,
      "learning_rate": 0.0003873914448874043,
      "loss": 3.466,
      "step": 33000
    },
    {
      "epoch": 3.5518243461414274,
      "eval_accuracy": 0.3738963175471915,
      "eval_loss": 3.4691905975341797,
      "eval_runtime": 184.6686,
      "eval_samples_per_second": 97.531,
      "eval_steps_per_second": 6.097,
      "step": 33000
    },
    {
      "epoch": 3.5572058981810355,
      "grad_norm": 0.5753702521324158,
      "learning_rate": 0.0003870682038573429,
      "loss": 3.447,
      "step": 33050
    },
    {
      "epoch": 3.5625874502206436,
      "grad_norm": 0.57344651222229,
      "learning_rate": 0.0003867449628272815,
      "loss": 3.4531,
      "step": 33100
    },
    {
      "epoch": 3.5679690022602517,
      "grad_norm": 0.596196711063385,
      "learning_rate": 0.0003864217217972201,
      "loss": 3.466,
      "step": 33150
    },
    {
      "epoch": 3.57335055429986,
      "grad_norm": 0.5894469618797302,
      "learning_rate": 0.0003860984807671587,
      "loss": 3.4502,
      "step": 33200
    },
    {
      "epoch": 3.5787321063394684,
      "grad_norm": 0.5605970621109009,
      "learning_rate": 0.00038577523973709724,
      "loss": 3.4534,
      "step": 33250
    },
    {
      "epoch": 3.5841136583790765,
      "grad_norm": 0.6303330659866333,
      "learning_rate": 0.00038545199870703583,
      "loss": 3.4537,
      "step": 33300
    },
    {
      "epoch": 3.5894952104186846,
      "grad_norm": 0.6184437274932861,
      "learning_rate": 0.0003851287576769744,
      "loss": 3.4538,
      "step": 33350
    },
    {
      "epoch": 3.594876762458293,
      "grad_norm": 0.5937089323997498,
      "learning_rate": 0.000384805516646913,
      "loss": 3.4745,
      "step": 33400
    },
    {
      "epoch": 3.6002583144979012,
      "grad_norm": 0.650280773639679,
      "learning_rate": 0.0003844822756168516,
      "loss": 3.4664,
      "step": 33450
    },
    {
      "epoch": 3.6056398665375093,
      "grad_norm": 0.6183555722236633,
      "learning_rate": 0.00038415903458679016,
      "loss": 3.4595,
      "step": 33500
    },
    {
      "epoch": 3.611021418577118,
      "grad_norm": 0.6294432878494263,
      "learning_rate": 0.00038383579355672875,
      "loss": 3.4687,
      "step": 33550
    },
    {
      "epoch": 3.616402970616726,
      "grad_norm": 0.5563703775405884,
      "learning_rate": 0.00038351255252666735,
      "loss": 3.4539,
      "step": 33600
    },
    {
      "epoch": 3.621784522656334,
      "grad_norm": 0.5843515992164612,
      "learning_rate": 0.00038318931149660594,
      "loss": 3.4602,
      "step": 33650
    },
    {
      "epoch": 3.627166074695942,
      "grad_norm": 0.6656185388565063,
      "learning_rate": 0.00038286607046654454,
      "loss": 3.4504,
      "step": 33700
    },
    {
      "epoch": 3.6325476267355503,
      "grad_norm": 0.5866097211837769,
      "learning_rate": 0.00038254282943648313,
      "loss": 3.4492,
      "step": 33750
    },
    {
      "epoch": 3.637929178775159,
      "grad_norm": 0.6325012445449829,
      "learning_rate": 0.00038221958840642167,
      "loss": 3.4465,
      "step": 33800
    },
    {
      "epoch": 3.643310730814767,
      "grad_norm": 0.5878468751907349,
      "learning_rate": 0.00038189634737636027,
      "loss": 3.4465,
      "step": 33850
    },
    {
      "epoch": 3.648692282854375,
      "grad_norm": 0.6447891592979431,
      "learning_rate": 0.0003815731063462988,
      "loss": 3.4554,
      "step": 33900
    },
    {
      "epoch": 3.6540738348939836,
      "grad_norm": 0.6783122420310974,
      "learning_rate": 0.00038124986531623745,
      "loss": 3.4506,
      "step": 33950
    },
    {
      "epoch": 3.6594553869335917,
      "grad_norm": 0.5637625455856323,
      "learning_rate": 0.00038092662428617605,
      "loss": 3.4664,
      "step": 34000
    },
    {
      "epoch": 3.6594553869335917,
      "eval_accuracy": 0.3743114802315306,
      "eval_loss": 3.4632701873779297,
      "eval_runtime": 184.7686,
      "eval_samples_per_second": 97.479,
      "eval_steps_per_second": 6.094,
      "step": 34000
    },
    {
      "epoch": 3.6648369389732,
      "grad_norm": 0.6158130764961243,
      "learning_rate": 0.0003806033832561146,
      "loss": 3.4482,
      "step": 34050
    },
    {
      "epoch": 3.670218491012808,
      "grad_norm": 0.6033890843391418,
      "learning_rate": 0.0003802801422260532,
      "loss": 3.4567,
      "step": 34100
    },
    {
      "epoch": 3.675600043052416,
      "grad_norm": 0.6929822564125061,
      "learning_rate": 0.0003799569011959918,
      "loss": 3.4441,
      "step": 34150
    },
    {
      "epoch": 3.6809815950920246,
      "grad_norm": 0.6433200240135193,
      "learning_rate": 0.0003796336601659303,
      "loss": 3.4583,
      "step": 34200
    },
    {
      "epoch": 3.6863631471316327,
      "grad_norm": 0.6610816717147827,
      "learning_rate": 0.00037931041913586897,
      "loss": 3.4545,
      "step": 34250
    },
    {
      "epoch": 3.691744699171241,
      "grad_norm": 0.6418637037277222,
      "learning_rate": 0.00037898717810580756,
      "loss": 3.467,
      "step": 34300
    },
    {
      "epoch": 3.6971262512108494,
      "grad_norm": 0.587744414806366,
      "learning_rate": 0.0003786639370757461,
      "loss": 3.446,
      "step": 34350
    },
    {
      "epoch": 3.7025078032504575,
      "grad_norm": 0.5943315029144287,
      "learning_rate": 0.0003783406960456847,
      "loss": 3.453,
      "step": 34400
    },
    {
      "epoch": 3.7078893552900656,
      "grad_norm": 0.6527094841003418,
      "learning_rate": 0.00037801745501562324,
      "loss": 3.4449,
      "step": 34450
    },
    {
      "epoch": 3.713270907329674,
      "grad_norm": 0.5982084274291992,
      "learning_rate": 0.0003776942139855619,
      "loss": 3.4604,
      "step": 34500
    },
    {
      "epoch": 3.7186524593692822,
      "grad_norm": 0.6434234380722046,
      "learning_rate": 0.0003773709729555005,
      "loss": 3.462,
      "step": 34550
    },
    {
      "epoch": 3.7240340114088903,
      "grad_norm": 0.6087839603424072,
      "learning_rate": 0.000377047731925439,
      "loss": 3.4495,
      "step": 34600
    },
    {
      "epoch": 3.7294155634484984,
      "grad_norm": 0.6338579058647156,
      "learning_rate": 0.0003767244908953776,
      "loss": 3.4548,
      "step": 34650
    },
    {
      "epoch": 3.7347971154881066,
      "grad_norm": 0.6043111085891724,
      "learning_rate": 0.0003764012498653162,
      "loss": 3.4783,
      "step": 34700
    },
    {
      "epoch": 3.740178667527715,
      "grad_norm": 0.6296453475952148,
      "learning_rate": 0.00037607800883525475,
      "loss": 3.4469,
      "step": 34750
    },
    {
      "epoch": 3.745560219567323,
      "grad_norm": 0.5895748734474182,
      "learning_rate": 0.0003757547678051934,
      "loss": 3.4689,
      "step": 34800
    },
    {
      "epoch": 3.7509417716069313,
      "grad_norm": 0.6216940879821777,
      "learning_rate": 0.000375431526775132,
      "loss": 3.4502,
      "step": 34850
    },
    {
      "epoch": 3.75632332364654,
      "grad_norm": 0.6479626893997192,
      "learning_rate": 0.00037510828574507054,
      "loss": 3.4286,
      "step": 34900
    },
    {
      "epoch": 3.761704875686148,
      "grad_norm": 0.626582682132721,
      "learning_rate": 0.00037478504471500913,
      "loss": 3.4464,
      "step": 34950
    },
    {
      "epoch": 3.767086427725756,
      "grad_norm": 0.6230822205543518,
      "learning_rate": 0.00037446180368494767,
      "loss": 3.4389,
      "step": 35000
    },
    {
      "epoch": 3.767086427725756,
      "eval_accuracy": 0.37504858142246406,
      "eval_loss": 3.4572198390960693,
      "eval_runtime": 184.5056,
      "eval_samples_per_second": 97.618,
      "eval_steps_per_second": 6.103,
      "step": 35000
    },
    {
      "epoch": 3.772467979765364,
      "grad_norm": 0.6130627393722534,
      "learning_rate": 0.00037413856265488627,
      "loss": 3.4359,
      "step": 35050
    },
    {
      "epoch": 3.7778495318049723,
      "grad_norm": 0.5983439683914185,
      "learning_rate": 0.0003738153216248249,
      "loss": 3.4625,
      "step": 35100
    },
    {
      "epoch": 3.783231083844581,
      "grad_norm": 0.6246692538261414,
      "learning_rate": 0.00037349208059476346,
      "loss": 3.449,
      "step": 35150
    },
    {
      "epoch": 3.788612635884189,
      "grad_norm": 0.6247369647026062,
      "learning_rate": 0.00037316883956470205,
      "loss": 3.4509,
      "step": 35200
    },
    {
      "epoch": 3.793994187923797,
      "grad_norm": 0.5881553888320923,
      "learning_rate": 0.00037284559853464064,
      "loss": 3.4625,
      "step": 35250
    },
    {
      "epoch": 3.7993757399634056,
      "grad_norm": 0.621890127658844,
      "learning_rate": 0.0003725223575045792,
      "loss": 3.4556,
      "step": 35300
    },
    {
      "epoch": 3.8047572920030137,
      "grad_norm": 0.6440849304199219,
      "learning_rate": 0.000372205581295119,
      "loss": 3.4437,
      "step": 35350
    },
    {
      "epoch": 3.810138844042622,
      "grad_norm": 0.6532849669456482,
      "learning_rate": 0.0003718823402650576,
      "loss": 3.4493,
      "step": 35400
    },
    {
      "epoch": 3.8155203960822304,
      "grad_norm": 0.704378068447113,
      "learning_rate": 0.00037155909923499624,
      "loss": 3.4462,
      "step": 35450
    },
    {
      "epoch": 3.8209019481218385,
      "grad_norm": 0.6049141883850098,
      "learning_rate": 0.0003712358582049348,
      "loss": 3.4795,
      "step": 35500
    },
    {
      "epoch": 3.8262835001614466,
      "grad_norm": 0.6370956897735596,
      "learning_rate": 0.00037091261717487337,
      "loss": 3.4381,
      "step": 35550
    },
    {
      "epoch": 3.8316650522010547,
      "grad_norm": 0.6178144216537476,
      "learning_rate": 0.00037058937614481197,
      "loss": 3.4739,
      "step": 35600
    },
    {
      "epoch": 3.837046604240663,
      "grad_norm": 0.6254431009292603,
      "learning_rate": 0.0003702661351147505,
      "loss": 3.4459,
      "step": 35650
    },
    {
      "epoch": 3.8424281562802713,
      "grad_norm": 0.6104562878608704,
      "learning_rate": 0.0003699428940846891,
      "loss": 3.4574,
      "step": 35700
    },
    {
      "epoch": 3.8478097083198795,
      "grad_norm": 0.6314630508422852,
      "learning_rate": 0.00036961965305462775,
      "loss": 3.4543,
      "step": 35750
    },
    {
      "epoch": 3.8531912603594876,
      "grad_norm": 0.6032555103302002,
      "learning_rate": 0.0003692964120245663,
      "loss": 3.4393,
      "step": 35800
    },
    {
      "epoch": 3.858572812399096,
      "grad_norm": 0.5797463655471802,
      "learning_rate": 0.0003689731709945049,
      "loss": 3.4418,
      "step": 35850
    },
    {
      "epoch": 3.863954364438704,
      "grad_norm": 0.6190691590309143,
      "learning_rate": 0.0003686499299644434,
      "loss": 3.4475,
      "step": 35900
    },
    {
      "epoch": 3.8693359164783123,
      "grad_norm": 0.6184253692626953,
      "learning_rate": 0.000368326688934382,
      "loss": 3.4371,
      "step": 35950
    },
    {
      "epoch": 3.8747174685179204,
      "grad_norm": 0.6029491424560547,
      "learning_rate": 0.0003680034479043206,
      "loss": 3.4297,
      "step": 36000
    },
    {
      "epoch": 3.8747174685179204,
      "eval_accuracy": 0.3752638227932245,
      "eval_loss": 3.45257568359375,
      "eval_runtime": 184.636,
      "eval_samples_per_second": 97.549,
      "eval_steps_per_second": 6.098,
      "step": 36000
    },
    {
      "epoch": 3.8800990205575285,
      "grad_norm": 0.6468108296394348,
      "learning_rate": 0.0003676802068742592,
      "loss": 3.4645,
      "step": 36050
    },
    {
      "epoch": 3.885480572597137,
      "grad_norm": 0.6638432741165161,
      "learning_rate": 0.0003673569658441978,
      "loss": 3.4543,
      "step": 36100
    },
    {
      "epoch": 3.890862124636745,
      "grad_norm": 0.5957279205322266,
      "learning_rate": 0.0003670337248141364,
      "loss": 3.4544,
      "step": 36150
    },
    {
      "epoch": 3.8962436766763533,
      "grad_norm": 0.6390054821968079,
      "learning_rate": 0.00036671048378407494,
      "loss": 3.4491,
      "step": 36200
    },
    {
      "epoch": 3.901625228715962,
      "grad_norm": 0.6385354399681091,
      "learning_rate": 0.00036638724275401353,
      "loss": 3.4407,
      "step": 36250
    },
    {
      "epoch": 3.90700678075557,
      "grad_norm": 0.6172071099281311,
      "learning_rate": 0.0003660640017239522,
      "loss": 3.4678,
      "step": 36300
    },
    {
      "epoch": 3.912388332795178,
      "grad_norm": 0.6054326891899109,
      "learning_rate": 0.0003657407606938907,
      "loss": 3.4324,
      "step": 36350
    },
    {
      "epoch": 3.9177698848347866,
      "grad_norm": 0.617918074131012,
      "learning_rate": 0.0003654175196638293,
      "loss": 3.4563,
      "step": 36400
    },
    {
      "epoch": 3.9231514368743947,
      "grad_norm": 0.6514351963996887,
      "learning_rate": 0.00036509427863376786,
      "loss": 3.4482,
      "step": 36450
    },
    {
      "epoch": 3.928532988914003,
      "grad_norm": 0.6361662149429321,
      "learning_rate": 0.00036477103760370645,
      "loss": 3.4561,
      "step": 36500
    },
    {
      "epoch": 3.933914540953611,
      "grad_norm": 0.5907837152481079,
      "learning_rate": 0.00036444779657364505,
      "loss": 3.451,
      "step": 36550
    },
    {
      "epoch": 3.939296092993219,
      "grad_norm": 0.6405068635940552,
      "learning_rate": 0.00036412455554358364,
      "loss": 3.455,
      "step": 36600
    },
    {
      "epoch": 3.9446776450328276,
      "grad_norm": 0.5964227914810181,
      "learning_rate": 0.00036380131451352224,
      "loss": 3.4496,
      "step": 36650
    },
    {
      "epoch": 3.9500591970724357,
      "grad_norm": 0.5983424782752991,
      "learning_rate": 0.00036347807348346083,
      "loss": 3.4556,
      "step": 36700
    },
    {
      "epoch": 3.955440749112044,
      "grad_norm": 0.6061781048774719,
      "learning_rate": 0.00036315483245339937,
      "loss": 3.4668,
      "step": 36750
    },
    {
      "epoch": 3.9608223011516523,
      "grad_norm": 0.619894802570343,
      "learning_rate": 0.00036283159142333797,
      "loss": 3.4392,
      "step": 36800
    },
    {
      "epoch": 3.9662038531912605,
      "grad_norm": 0.6279895901679993,
      "learning_rate": 0.0003625083503932765,
      "loss": 3.4514,
      "step": 36850
    },
    {
      "epoch": 3.9715854052308686,
      "grad_norm": 0.5655210018157959,
      "learning_rate": 0.00036218510936321516,
      "loss": 3.4708,
      "step": 36900
    },
    {
      "epoch": 3.9769669572704767,
      "grad_norm": 0.5848953127861023,
      "learning_rate": 0.00036186186833315375,
      "loss": 3.465,
      "step": 36950
    },
    {
      "epoch": 3.9823485093100848,
      "grad_norm": 0.6551256775856018,
      "learning_rate": 0.0003615386273030923,
      "loss": 3.445,
      "step": 37000
    },
    {
      "epoch": 3.9823485093100848,
      "eval_accuracy": 0.37610197116982547,
      "eval_loss": 3.4445886611938477,
      "eval_runtime": 187.3097,
      "eval_samples_per_second": 96.156,
      "eval_steps_per_second": 6.011,
      "step": 37000
    },
    {
      "epoch": 3.9877300613496933,
      "grad_norm": 0.6091813445091248,
      "learning_rate": 0.0003612153862730309,
      "loss": 3.451,
      "step": 37050
    },
    {
      "epoch": 3.9931116133893014,
      "grad_norm": 0.6060930490493774,
      "learning_rate": 0.0003608921452429695,
      "loss": 3.4552,
      "step": 37100
    },
    {
      "epoch": 3.9984931654289095,
      "grad_norm": 0.6171631813049316,
      "learning_rate": 0.000360568904212908,
      "loss": 3.4441,
      "step": 37150
    },
    {
      "epoch": 4.003874717468518,
      "grad_norm": 0.6074433326721191,
      "learning_rate": 0.00036024566318284667,
      "loss": 3.3806,
      "step": 37200
    },
    {
      "epoch": 4.009256269508126,
      "grad_norm": 0.5943333506584167,
      "learning_rate": 0.00035992242215278526,
      "loss": 3.3569,
      "step": 37250
    },
    {
      "epoch": 4.014637821547734,
      "grad_norm": 0.637510359287262,
      "learning_rate": 0.0003595991811227238,
      "loss": 3.3698,
      "step": 37300
    },
    {
      "epoch": 4.020019373587343,
      "grad_norm": 0.6431918740272522,
      "learning_rate": 0.0003592759400926624,
      "loss": 3.3724,
      "step": 37350
    },
    {
      "epoch": 4.0254009256269505,
      "grad_norm": 0.7091398239135742,
      "learning_rate": 0.00035895269906260094,
      "loss": 3.3644,
      "step": 37400
    },
    {
      "epoch": 4.030782477666559,
      "grad_norm": 0.6210770010948181,
      "learning_rate": 0.0003586359228531408,
      "loss": 3.3404,
      "step": 37450
    },
    {
      "epoch": 4.036164029706168,
      "grad_norm": 0.6259356141090393,
      "learning_rate": 0.00035831268182307934,
      "loss": 3.3691,
      "step": 37500
    },
    {
      "epoch": 4.041545581745775,
      "grad_norm": 0.6482320427894592,
      "learning_rate": 0.000357989440793018,
      "loss": 3.3458,
      "step": 37550
    },
    {
      "epoch": 4.046927133785384,
      "grad_norm": 0.6331608891487122,
      "learning_rate": 0.0003576661997629566,
      "loss": 3.3768,
      "step": 37600
    },
    {
      "epoch": 4.0523086858249915,
      "grad_norm": 0.6756312847137451,
      "learning_rate": 0.0003573429587328951,
      "loss": 3.3593,
      "step": 37650
    },
    {
      "epoch": 4.0576902378646,
      "grad_norm": 0.5940220355987549,
      "learning_rate": 0.0003570197177028337,
      "loss": 3.3723,
      "step": 37700
    },
    {
      "epoch": 4.063071789904209,
      "grad_norm": 0.604108989238739,
      "learning_rate": 0.00035669647667277226,
      "loss": 3.3676,
      "step": 37750
    },
    {
      "epoch": 4.068453341943816,
      "grad_norm": 0.6163511872291565,
      "learning_rate": 0.00035637323564271085,
      "loss": 3.3558,
      "step": 37800
    },
    {
      "epoch": 4.073834893983425,
      "grad_norm": 0.6365783214569092,
      "learning_rate": 0.0003560499946126495,
      "loss": 3.3668,
      "step": 37850
    },
    {
      "epoch": 4.079216446023033,
      "grad_norm": 0.6292728185653687,
      "learning_rate": 0.00035572675358258804,
      "loss": 3.3648,
      "step": 37900
    },
    {
      "epoch": 4.084597998062641,
      "grad_norm": 0.6514345407485962,
      "learning_rate": 0.00035540351255252664,
      "loss": 3.3643,
      "step": 37950
    },
    {
      "epoch": 4.08997955010225,
      "grad_norm": 0.6355460286140442,
      "learning_rate": 0.00035508027152246523,
      "loss": 3.354,
      "step": 38000
    },
    {
      "epoch": 4.08997955010225,
      "eval_accuracy": 0.3767296589027297,
      "eval_loss": 3.4461004734039307,
      "eval_runtime": 186.8586,
      "eval_samples_per_second": 96.388,
      "eval_steps_per_second": 6.026,
      "step": 38000
    },
    {
      "epoch": 4.095361102141858,
      "grad_norm": 0.6700498461723328,
      "learning_rate": 0.0003547570304924038,
      "loss": 3.3655,
      "step": 38050
    },
    {
      "epoch": 4.100742654181466,
      "grad_norm": 0.663998544216156,
      "learning_rate": 0.0003544337894623424,
      "loss": 3.3594,
      "step": 38100
    },
    {
      "epoch": 4.106124206221074,
      "grad_norm": 0.6124699711799622,
      "learning_rate": 0.000354110548432281,
      "loss": 3.3701,
      "step": 38150
    },
    {
      "epoch": 4.111505758260682,
      "grad_norm": 0.6116810441017151,
      "learning_rate": 0.00035378730740221956,
      "loss": 3.3662,
      "step": 38200
    },
    {
      "epoch": 4.1168873103002905,
      "grad_norm": 0.6466190218925476,
      "learning_rate": 0.00035346406637215815,
      "loss": 3.3641,
      "step": 38250
    },
    {
      "epoch": 4.122268862339899,
      "grad_norm": 0.5948997139930725,
      "learning_rate": 0.0003531408253420967,
      "loss": 3.3756,
      "step": 38300
    },
    {
      "epoch": 4.127650414379507,
      "grad_norm": 0.6248114705085754,
      "learning_rate": 0.0003528175843120353,
      "loss": 3.3627,
      "step": 38350
    },
    {
      "epoch": 4.133031966419115,
      "grad_norm": 0.6819987297058105,
      "learning_rate": 0.00035249434328197394,
      "loss": 3.3872,
      "step": 38400
    },
    {
      "epoch": 4.138413518458724,
      "grad_norm": 0.6541435122489929,
      "learning_rate": 0.0003521711022519125,
      "loss": 3.3846,
      "step": 38450
    },
    {
      "epoch": 4.1437950704983315,
      "grad_norm": 0.6173901557922363,
      "learning_rate": 0.00035184786122185107,
      "loss": 3.3829,
      "step": 38500
    },
    {
      "epoch": 4.14917662253794,
      "grad_norm": 0.6310703158378601,
      "learning_rate": 0.00035152462019178967,
      "loss": 3.3788,
      "step": 38550
    },
    {
      "epoch": 4.154558174577549,
      "grad_norm": 0.6219739317893982,
      "learning_rate": 0.0003512013791617282,
      "loss": 3.3957,
      "step": 38600
    },
    {
      "epoch": 4.159939726617156,
      "grad_norm": 0.6413523554801941,
      "learning_rate": 0.0003508781381316668,
      "loss": 3.3756,
      "step": 38650
    },
    {
      "epoch": 4.165321278656765,
      "grad_norm": 0.6500468254089355,
      "learning_rate": 0.00035055489710160545,
      "loss": 3.3924,
      "step": 38700
    },
    {
      "epoch": 4.1707028306963725,
      "grad_norm": 0.6107872724533081,
      "learning_rate": 0.000350231656071544,
      "loss": 3.3748,
      "step": 38750
    },
    {
      "epoch": 4.176084382735981,
      "grad_norm": 0.6339595317840576,
      "learning_rate": 0.0003499084150414826,
      "loss": 3.3773,
      "step": 38800
    },
    {
      "epoch": 4.18146593477559,
      "grad_norm": 0.6154005527496338,
      "learning_rate": 0.0003495851740114211,
      "loss": 3.3888,
      "step": 38850
    },
    {
      "epoch": 4.186847486815197,
      "grad_norm": 0.5948972702026367,
      "learning_rate": 0.0003492619329813597,
      "loss": 3.3617,
      "step": 38900
    },
    {
      "epoch": 4.192229038854806,
      "grad_norm": 0.6268037557601929,
      "learning_rate": 0.0003489386919512983,
      "loss": 3.3808,
      "step": 38950
    },
    {
      "epoch": 4.197610590894414,
      "grad_norm": 0.6886900067329407,
      "learning_rate": 0.0003486154509212369,
      "loss": 3.3701,
      "step": 39000
    },
    {
      "epoch": 4.197610590894414,
      "eval_accuracy": 0.3767333531009154,
      "eval_loss": 3.4467148780822754,
      "eval_runtime": 186.912,
      "eval_samples_per_second": 96.361,
      "eval_steps_per_second": 6.024,
      "step": 39000
    },
    {
      "epoch": 4.202992142934022,
      "grad_norm": 0.6147679686546326,
      "learning_rate": 0.0003482922098911755,
      "loss": 3.3867,
      "step": 39050
    },
    {
      "epoch": 4.208373694973631,
      "grad_norm": 0.6290424466133118,
      "learning_rate": 0.0003479689688611141,
      "loss": 3.375,
      "step": 39100
    },
    {
      "epoch": 4.213755247013238,
      "grad_norm": 0.6457464098930359,
      "learning_rate": 0.00034764572783105264,
      "loss": 3.3837,
      "step": 39150
    },
    {
      "epoch": 4.219136799052847,
      "grad_norm": 0.646575391292572,
      "learning_rate": 0.00034732248680099123,
      "loss": 3.3858,
      "step": 39200
    },
    {
      "epoch": 4.224518351092455,
      "grad_norm": 0.6269925236701965,
      "learning_rate": 0.0003469992457709299,
      "loss": 3.3909,
      "step": 39250
    },
    {
      "epoch": 4.229899903132063,
      "grad_norm": 0.6284508109092712,
      "learning_rate": 0.0003466760047408684,
      "loss": 3.3784,
      "step": 39300
    },
    {
      "epoch": 4.2352814551716715,
      "grad_norm": 0.6222100257873535,
      "learning_rate": 0.000346352763710807,
      "loss": 3.3784,
      "step": 39350
    },
    {
      "epoch": 4.24066300721128,
      "grad_norm": 0.708451509475708,
      "learning_rate": 0.00034602952268074556,
      "loss": 3.3835,
      "step": 39400
    },
    {
      "epoch": 4.246044559250888,
      "grad_norm": 0.6076235175132751,
      "learning_rate": 0.00034570628165068415,
      "loss": 3.3924,
      "step": 39450
    },
    {
      "epoch": 4.251426111290496,
      "grad_norm": 0.6589393615722656,
      "learning_rate": 0.0003453830406206227,
      "loss": 3.3961,
      "step": 39500
    },
    {
      "epoch": 4.256807663330104,
      "grad_norm": 0.6510894298553467,
      "learning_rate": 0.00034505979959056134,
      "loss": 3.3551,
      "step": 39550
    },
    {
      "epoch": 4.2621892153697125,
      "grad_norm": 0.8012067675590515,
      "learning_rate": 0.00034473655856049994,
      "loss": 3.3709,
      "step": 39600
    },
    {
      "epoch": 4.267570767409321,
      "grad_norm": 0.6146559119224548,
      "learning_rate": 0.0003444133175304385,
      "loss": 3.4107,
      "step": 39650
    },
    {
      "epoch": 4.272952319448929,
      "grad_norm": 0.6020640134811401,
      "learning_rate": 0.00034409007650037707,
      "loss": 3.3753,
      "step": 39700
    },
    {
      "epoch": 4.278333871488537,
      "grad_norm": 0.6155546307563782,
      "learning_rate": 0.00034376683547031567,
      "loss": 3.3717,
      "step": 39750
    },
    {
      "epoch": 4.283715423528146,
      "grad_norm": 0.6239294409751892,
      "learning_rate": 0.0003434435944402542,
      "loss": 3.3639,
      "step": 39800
    },
    {
      "epoch": 4.2890969755677535,
      "grad_norm": 0.6434022188186646,
      "learning_rate": 0.00034312035341019286,
      "loss": 3.3988,
      "step": 39850
    },
    {
      "epoch": 4.294478527607362,
      "grad_norm": 0.6435459852218628,
      "learning_rate": 0.00034279711238013145,
      "loss": 3.3989,
      "step": 39900
    },
    {
      "epoch": 4.299860079646971,
      "grad_norm": 0.6500222086906433,
      "learning_rate": 0.00034248033617067126,
      "loss": 3.397,
      "step": 39950
    },
    {
      "epoch": 4.305241631686578,
      "grad_norm": 0.622305691242218,
      "learning_rate": 0.00034215709514060985,
      "loss": 3.3786,
      "step": 40000
    },
    {
      "epoch": 4.305241631686578,
      "eval_accuracy": 0.377357346635634,
      "eval_loss": 3.438131093978882,
      "eval_runtime": 186.8572,
      "eval_samples_per_second": 96.389,
      "eval_steps_per_second": 6.026,
      "step": 40000
    },
    {
      "epoch": 4.310623183726187,
      "grad_norm": 0.6466324329376221,
      "learning_rate": 0.0003418338541105484,
      "loss": 3.3885,
      "step": 40050
    },
    {
      "epoch": 4.3160047357657945,
      "grad_norm": 0.632236897945404,
      "learning_rate": 0.000341510613080487,
      "loss": 3.3859,
      "step": 40100
    },
    {
      "epoch": 4.321386287805403,
      "grad_norm": 0.6783816814422607,
      "learning_rate": 0.0003411873720504255,
      "loss": 3.37,
      "step": 40150
    },
    {
      "epoch": 4.326767839845012,
      "grad_norm": 0.632381796836853,
      "learning_rate": 0.0003408641310203642,
      "loss": 3.3859,
      "step": 40200
    },
    {
      "epoch": 4.332149391884619,
      "grad_norm": 0.6269997358322144,
      "learning_rate": 0.00034054088999030277,
      "loss": 3.3862,
      "step": 40250
    },
    {
      "epoch": 4.337530943924228,
      "grad_norm": 0.6661253571510315,
      "learning_rate": 0.0003402176489602413,
      "loss": 3.3827,
      "step": 40300
    },
    {
      "epoch": 4.342912495963836,
      "grad_norm": 0.6136159896850586,
      "learning_rate": 0.0003398944079301799,
      "loss": 3.3986,
      "step": 40350
    },
    {
      "epoch": 4.348294048003444,
      "grad_norm": 0.6653974056243896,
      "learning_rate": 0.0003395711669001185,
      "loss": 3.3938,
      "step": 40400
    },
    {
      "epoch": 4.3536756000430525,
      "grad_norm": 0.6155456900596619,
      "learning_rate": 0.00033924792587005704,
      "loss": 3.393,
      "step": 40450
    },
    {
      "epoch": 4.359057152082661,
      "grad_norm": 0.690609872341156,
      "learning_rate": 0.0003389246848399957,
      "loss": 3.3828,
      "step": 40500
    },
    {
      "epoch": 4.364438704122269,
      "grad_norm": 0.6541188955307007,
      "learning_rate": 0.0003386014438099343,
      "loss": 3.3799,
      "step": 40550
    },
    {
      "epoch": 4.369820256161877,
      "grad_norm": 0.6513378620147705,
      "learning_rate": 0.0003382782027798728,
      "loss": 3.391,
      "step": 40600
    },
    {
      "epoch": 4.375201808201485,
      "grad_norm": 0.6599308252334595,
      "learning_rate": 0.0003379549617498114,
      "loss": 3.3743,
      "step": 40650
    },
    {
      "epoch": 4.3805833602410935,
      "grad_norm": 0.6568775773048401,
      "learning_rate": 0.00033763172071974996,
      "loss": 3.3868,
      "step": 40700
    },
    {
      "epoch": 4.385964912280702,
      "grad_norm": 0.6643113493919373,
      "learning_rate": 0.00033730847968968855,
      "loss": 3.3817,
      "step": 40750
    },
    {
      "epoch": 4.39134646432031,
      "grad_norm": 0.6479328870773315,
      "learning_rate": 0.0003369852386596272,
      "loss": 3.3893,
      "step": 40800
    },
    {
      "epoch": 4.396728016359918,
      "grad_norm": 0.6810503602027893,
      "learning_rate": 0.00033666199762956574,
      "loss": 3.3667,
      "step": 40850
    },
    {
      "epoch": 4.402109568399527,
      "grad_norm": 0.652211606502533,
      "learning_rate": 0.00033633875659950434,
      "loss": 3.3764,
      "step": 40900
    },
    {
      "epoch": 4.4074911204391345,
      "grad_norm": 0.6235228180885315,
      "learning_rate": 0.0003360155155694429,
      "loss": 3.3965,
      "step": 40950
    },
    {
      "epoch": 4.412872672478743,
      "grad_norm": 0.6516085863113403,
      "learning_rate": 0.0003356922745393815,
      "loss": 3.3923,
      "step": 41000
    },
    {
      "epoch": 4.412872672478743,
      "eval_accuracy": 0.3779175609252054,
      "eval_loss": 3.4350180625915527,
      "eval_runtime": 185.3058,
      "eval_samples_per_second": 97.196,
      "eval_steps_per_second": 6.076,
      "step": 41000
    },
    {
      "epoch": 4.418254224518351,
      "grad_norm": 0.653671383857727,
      "learning_rate": 0.0003353690335093201,
      "loss": 3.3929,
      "step": 41050
    },
    {
      "epoch": 4.423635776557959,
      "grad_norm": 0.6062602996826172,
      "learning_rate": 0.0003350457924792587,
      "loss": 3.3809,
      "step": 41100
    },
    {
      "epoch": 4.429017328597568,
      "grad_norm": 0.6373082995414734,
      "learning_rate": 0.00033472255144919726,
      "loss": 3.3912,
      "step": 41150
    },
    {
      "epoch": 4.4343988806371755,
      "grad_norm": 0.6125858426094055,
      "learning_rate": 0.00033439931041913585,
      "loss": 3.3907,
      "step": 41200
    },
    {
      "epoch": 4.439780432676784,
      "grad_norm": 0.6293478608131409,
      "learning_rate": 0.0003340760693890744,
      "loss": 3.3564,
      "step": 41250
    },
    {
      "epoch": 4.445161984716393,
      "grad_norm": 0.6728382706642151,
      "learning_rate": 0.000333752828359013,
      "loss": 3.4139,
      "step": 41300
    },
    {
      "epoch": 4.450543536756,
      "grad_norm": 0.6386088728904724,
      "learning_rate": 0.00033342958732895164,
      "loss": 3.4263,
      "step": 41350
    },
    {
      "epoch": 4.455925088795609,
      "grad_norm": 0.6368845105171204,
      "learning_rate": 0.0003331063462988902,
      "loss": 3.3761,
      "step": 41400
    },
    {
      "epoch": 4.461306640835216,
      "grad_norm": 0.6361759305000305,
      "learning_rate": 0.00033278310526882877,
      "loss": 3.3773,
      "step": 41450
    },
    {
      "epoch": 4.466688192874825,
      "grad_norm": 0.6400294899940491,
      "learning_rate": 0.0003324598642387673,
      "loss": 3.3874,
      "step": 41500
    },
    {
      "epoch": 4.4720697449144335,
      "grad_norm": 0.6286135911941528,
      "learning_rate": 0.0003321366232087059,
      "loss": 3.3928,
      "step": 41550
    },
    {
      "epoch": 4.477451296954041,
      "grad_norm": 0.6225272417068481,
      "learning_rate": 0.0003318133821786445,
      "loss": 3.3863,
      "step": 41600
    },
    {
      "epoch": 4.48283284899365,
      "grad_norm": 0.6586177349090576,
      "learning_rate": 0.0003314901411485831,
      "loss": 3.3986,
      "step": 41650
    },
    {
      "epoch": 4.488214401033258,
      "grad_norm": 0.6606486439704895,
      "learning_rate": 0.0003311669001185217,
      "loss": 3.3888,
      "step": 41700
    },
    {
      "epoch": 4.493595953072866,
      "grad_norm": 0.657923698425293,
      "learning_rate": 0.0003308436590884603,
      "loss": 3.3853,
      "step": 41750
    },
    {
      "epoch": 4.4989775051124745,
      "grad_norm": 0.6433077454566956,
      "learning_rate": 0.0003305204180583988,
      "loss": 3.388,
      "step": 41800
    },
    {
      "epoch": 4.504359057152083,
      "grad_norm": 0.6714504361152649,
      "learning_rate": 0.0003301971770283374,
      "loss": 3.3915,
      "step": 41850
    },
    {
      "epoch": 4.509740609191691,
      "grad_norm": 0.703129231929779,
      "learning_rate": 0.00032987393599827607,
      "loss": 3.3991,
      "step": 41900
    },
    {
      "epoch": 4.515122161231299,
      "grad_norm": 0.6598409414291382,
      "learning_rate": 0.0003295506949682146,
      "loss": 3.4052,
      "step": 41950
    },
    {
      "epoch": 4.520503713270907,
      "grad_norm": 0.6891419291496277,
      "learning_rate": 0.0003292274539381532,
      "loss": 3.389,
      "step": 42000
    },
    {
      "epoch": 4.520503713270907,
      "eval_accuracy": 0.37835901760839563,
      "eval_loss": 3.4275496006011963,
      "eval_runtime": 184.4745,
      "eval_samples_per_second": 97.634,
      "eval_steps_per_second": 6.104,
      "step": 42000
    },
    {
      "epoch": 4.5258852653105155,
      "grad_norm": 0.6308417916297913,
      "learning_rate": 0.00032890421290809174,
      "loss": 3.3936,
      "step": 42050
    },
    {
      "epoch": 4.531266817350124,
      "grad_norm": 0.640397310256958,
      "learning_rate": 0.00032858097187803034,
      "loss": 3.4059,
      "step": 42100
    },
    {
      "epoch": 4.536648369389732,
      "grad_norm": 0.6476185917854309,
      "learning_rate": 0.00032825773084796893,
      "loss": 3.3851,
      "step": 42150
    },
    {
      "epoch": 4.54202992142934,
      "grad_norm": 0.6753180027008057,
      "learning_rate": 0.00032793448981790753,
      "loss": 3.4025,
      "step": 42200
    },
    {
      "epoch": 4.547411473468949,
      "grad_norm": 0.664780855178833,
      "learning_rate": 0.0003276112487878461,
      "loss": 3.3951,
      "step": 42250
    },
    {
      "epoch": 4.5527930255085565,
      "grad_norm": 0.6564720869064331,
      "learning_rate": 0.0003272880077577847,
      "loss": 3.3786,
      "step": 42300
    },
    {
      "epoch": 4.558174577548165,
      "grad_norm": 0.6598086357116699,
      "learning_rate": 0.00032696476672772326,
      "loss": 3.402,
      "step": 42350
    },
    {
      "epoch": 4.563556129587774,
      "grad_norm": 0.6517024040222168,
      "learning_rate": 0.00032664152569766185,
      "loss": 3.3968,
      "step": 42400
    },
    {
      "epoch": 4.568937681627381,
      "grad_norm": 0.6901940703392029,
      "learning_rate": 0.0003263182846676004,
      "loss": 3.3889,
      "step": 42450
    },
    {
      "epoch": 4.57431923366699,
      "grad_norm": 0.6727946400642395,
      "learning_rate": 0.00032599504363753904,
      "loss": 3.4006,
      "step": 42500
    },
    {
      "epoch": 4.579700785706597,
      "grad_norm": 0.6285117864608765,
      "learning_rate": 0.00032567180260747764,
      "loss": 3.3855,
      "step": 42550
    },
    {
      "epoch": 4.585082337746206,
      "grad_norm": 0.6336430907249451,
      "learning_rate": 0.0003253485615774162,
      "loss": 3.399,
      "step": 42600
    },
    {
      "epoch": 4.5904638897858145,
      "grad_norm": 0.6821553111076355,
      "learning_rate": 0.00032502532054735477,
      "loss": 3.3826,
      "step": 42650
    },
    {
      "epoch": 4.595845441825422,
      "grad_norm": 0.6781034469604492,
      "learning_rate": 0.00032470207951729337,
      "loss": 3.3837,
      "step": 42700
    },
    {
      "epoch": 4.601226993865031,
      "grad_norm": 0.6807826161384583,
      "learning_rate": 0.0003243788384872319,
      "loss": 3.3979,
      "step": 42750
    },
    {
      "epoch": 4.606608545904638,
      "grad_norm": 0.6595771908760071,
      "learning_rate": 0.0003240620622777717,
      "loss": 3.3888,
      "step": 42800
    },
    {
      "epoch": 4.611990097944247,
      "grad_norm": 0.6954243779182434,
      "learning_rate": 0.00032373882124771036,
      "loss": 3.3703,
      "step": 42850
    },
    {
      "epoch": 4.6173716499838555,
      "grad_norm": 0.6635729670524597,
      "learning_rate": 0.00032341558021764896,
      "loss": 3.3973,
      "step": 42900
    },
    {
      "epoch": 4.622753202023463,
      "grad_norm": 0.7169069051742554,
      "learning_rate": 0.0003230923391875875,
      "loss": 3.4055,
      "step": 42950
    },
    {
      "epoch": 4.628134754063072,
      "grad_norm": 0.643351137638092,
      "learning_rate": 0.0003227690981575261,
      "loss": 3.3754,
      "step": 43000
    },
    {
      "epoch": 4.628134754063072,
      "eval_accuracy": 0.3789961581425398,
      "eval_loss": 3.423278331756592,
      "eval_runtime": 184.6106,
      "eval_samples_per_second": 97.562,
      "eval_steps_per_second": 6.099,
      "step": 43000
    },
    {
      "epoch": 4.63351630610268,
      "grad_norm": 0.6247761845588684,
      "learning_rate": 0.0003224458571274647,
      "loss": 3.3895,
      "step": 43050
    },
    {
      "epoch": 4.638897858142288,
      "grad_norm": 0.6495500802993774,
      "learning_rate": 0.00032212261609740323,
      "loss": 3.3633,
      "step": 43100
    },
    {
      "epoch": 4.6442794101818965,
      "grad_norm": 0.6213272213935852,
      "learning_rate": 0.0003217993750673419,
      "loss": 3.3898,
      "step": 43150
    },
    {
      "epoch": 4.649660962221505,
      "grad_norm": 0.6968953609466553,
      "learning_rate": 0.00032147613403728047,
      "loss": 3.4009,
      "step": 43200
    },
    {
      "epoch": 4.655042514261113,
      "grad_norm": 0.6720950603485107,
      "learning_rate": 0.000321152893007219,
      "loss": 3.3992,
      "step": 43250
    },
    {
      "epoch": 4.660424066300721,
      "grad_norm": 0.648715078830719,
      "learning_rate": 0.0003208296519771576,
      "loss": 3.3778,
      "step": 43300
    },
    {
      "epoch": 4.665805618340329,
      "grad_norm": 0.6420325636863708,
      "learning_rate": 0.00032050641094709615,
      "loss": 3.397,
      "step": 43350
    },
    {
      "epoch": 4.6711871703799375,
      "grad_norm": 0.6525468230247498,
      "learning_rate": 0.00032018316991703474,
      "loss": 3.3925,
      "step": 43400
    },
    {
      "epoch": 4.676568722419546,
      "grad_norm": 0.6904116272926331,
      "learning_rate": 0.0003198599288869734,
      "loss": 3.3921,
      "step": 43450
    },
    {
      "epoch": 4.681950274459154,
      "grad_norm": 0.6783482432365417,
      "learning_rate": 0.00031953668785691193,
      "loss": 3.4008,
      "step": 43500
    },
    {
      "epoch": 4.687331826498762,
      "grad_norm": 0.6485925316810608,
      "learning_rate": 0.0003192134468268505,
      "loss": 3.3889,
      "step": 43550
    },
    {
      "epoch": 4.692713378538371,
      "grad_norm": 0.6691424250602722,
      "learning_rate": 0.0003188902057967891,
      "loss": 3.3749,
      "step": 43600
    },
    {
      "epoch": 4.6980949305779784,
      "grad_norm": 0.6782438158988953,
      "learning_rate": 0.00031856696476672766,
      "loss": 3.3979,
      "step": 43650
    },
    {
      "epoch": 4.703476482617587,
      "grad_norm": 0.7157396674156189,
      "learning_rate": 0.0003182437237366663,
      "loss": 3.3879,
      "step": 43700
    },
    {
      "epoch": 4.7088580346571955,
      "grad_norm": 0.6502918004989624,
      "learning_rate": 0.0003179204827066049,
      "loss": 3.3739,
      "step": 43750
    },
    {
      "epoch": 4.714239586696803,
      "grad_norm": 0.6235714554786682,
      "learning_rate": 0.00031759724167654344,
      "loss": 3.3888,
      "step": 43800
    },
    {
      "epoch": 4.719621138736412,
      "grad_norm": 0.6773161292076111,
      "learning_rate": 0.00031727400064648204,
      "loss": 3.3858,
      "step": 43850
    },
    {
      "epoch": 4.725002690776019,
      "grad_norm": 0.6831490397453308,
      "learning_rate": 0.0003169507596164206,
      "loss": 3.3861,
      "step": 43900
    },
    {
      "epoch": 4.730384242815628,
      "grad_norm": 0.644305944442749,
      "learning_rate": 0.0003166275185863592,
      "loss": 3.3901,
      "step": 43950
    },
    {
      "epoch": 4.7357657948552365,
      "grad_norm": 0.7171474695205688,
      "learning_rate": 0.0003163042775562978,
      "loss": 3.3811,
      "step": 44000
    },
    {
      "epoch": 4.7357657948552365,
      "eval_accuracy": 0.3795620223822776,
      "eval_loss": 3.4181690216064453,
      "eval_runtime": 184.9367,
      "eval_samples_per_second": 97.39,
      "eval_steps_per_second": 6.089,
      "step": 44000
    },
    {
      "epoch": 4.741147346894844,
      "grad_norm": 0.7104071974754333,
      "learning_rate": 0.00031598103652623636,
      "loss": 3.3849,
      "step": 44050
    },
    {
      "epoch": 4.746528898934453,
      "grad_norm": 0.6331132650375366,
      "learning_rate": 0.00031565779549617496,
      "loss": 3.382,
      "step": 44100
    },
    {
      "epoch": 4.751910450974061,
      "grad_norm": 0.669050395488739,
      "learning_rate": 0.00031533455446611355,
      "loss": 3.3709,
      "step": 44150
    },
    {
      "epoch": 4.757292003013669,
      "grad_norm": 0.6390455365180969,
      "learning_rate": 0.0003150113134360521,
      "loss": 3.403,
      "step": 44200
    },
    {
      "epoch": 4.7626735550532775,
      "grad_norm": 0.6818135380744934,
      "learning_rate": 0.0003146880724059907,
      "loss": 3.3829,
      "step": 44250
    },
    {
      "epoch": 4.768055107092886,
      "grad_norm": 0.6637402772903442,
      "learning_rate": 0.00031436483137592934,
      "loss": 3.4032,
      "step": 44300
    },
    {
      "epoch": 4.773436659132494,
      "grad_norm": 0.674086332321167,
      "learning_rate": 0.0003140415903458679,
      "loss": 3.3966,
      "step": 44350
    },
    {
      "epoch": 4.778818211172102,
      "grad_norm": 0.6924407482147217,
      "learning_rate": 0.00031371834931580647,
      "loss": 3.3863,
      "step": 44400
    },
    {
      "epoch": 4.78419976321171,
      "grad_norm": 0.6802992224693298,
      "learning_rate": 0.000313395108285745,
      "loss": 3.4059,
      "step": 44450
    },
    {
      "epoch": 4.7895813152513185,
      "grad_norm": 0.6734175682067871,
      "learning_rate": 0.0003130718672556836,
      "loss": 3.3743,
      "step": 44500
    },
    {
      "epoch": 4.794962867290927,
      "grad_norm": 0.6906740069389343,
      "learning_rate": 0.0003127486262256222,
      "loss": 3.3817,
      "step": 44550
    },
    {
      "epoch": 4.800344419330535,
      "grad_norm": 0.6398283839225769,
      "learning_rate": 0.0003124253851955608,
      "loss": 3.3937,
      "step": 44600
    },
    {
      "epoch": 4.805725971370143,
      "grad_norm": 0.6402740478515625,
      "learning_rate": 0.0003121021441654994,
      "loss": 3.4015,
      "step": 44650
    },
    {
      "epoch": 4.811107523409751,
      "grad_norm": 0.6882855296134949,
      "learning_rate": 0.000311778903135438,
      "loss": 3.3789,
      "step": 44700
    },
    {
      "epoch": 4.8164890754493594,
      "grad_norm": 0.6392636895179749,
      "learning_rate": 0.0003114556621053765,
      "loss": 3.3849,
      "step": 44750
    },
    {
      "epoch": 4.821870627488968,
      "grad_norm": 0.6478724479675293,
      "learning_rate": 0.0003111324210753151,
      "loss": 3.4083,
      "step": 44800
    },
    {
      "epoch": 4.827252179528576,
      "grad_norm": 0.6727092862129211,
      "learning_rate": 0.00031080918004525377,
      "loss": 3.4042,
      "step": 44850
    },
    {
      "epoch": 4.832633731568184,
      "grad_norm": 0.681175947189331,
      "learning_rate": 0.0003104859390151923,
      "loss": 3.3854,
      "step": 44900
    },
    {
      "epoch": 4.838015283607793,
      "grad_norm": 0.6760862469673157,
      "learning_rate": 0.0003101626979851309,
      "loss": 3.3831,
      "step": 44950
    },
    {
      "epoch": 4.8433968356474,
      "grad_norm": 0.6785559058189392,
      "learning_rate": 0.00030983945695506945,
      "loss": 3.3894,
      "step": 45000
    },
    {
      "epoch": 4.8433968356474,
      "eval_accuracy": 0.38030520813492863,
      "eval_loss": 3.4138073921203613,
      "eval_runtime": 184.496,
      "eval_samples_per_second": 97.623,
      "eval_steps_per_second": 6.103,
      "step": 45000
    },
    {
      "epoch": 4.848778387687009,
      "grad_norm": 0.6968111991882324,
      "learning_rate": 0.00030951621592500804,
      "loss": 3.3986,
      "step": 45050
    },
    {
      "epoch": 4.8541599397266175,
      "grad_norm": 0.6569320559501648,
      "learning_rate": 0.00030919297489494663,
      "loss": 3.3936,
      "step": 45100
    },
    {
      "epoch": 4.859541491766225,
      "grad_norm": 0.6438539624214172,
      "learning_rate": 0.00030886973386488523,
      "loss": 3.3937,
      "step": 45150
    },
    {
      "epoch": 4.864923043805834,
      "grad_norm": 0.6866356730461121,
      "learning_rate": 0.0003085464928348238,
      "loss": 3.3841,
      "step": 45200
    },
    {
      "epoch": 4.870304595845441,
      "grad_norm": 0.6964440941810608,
      "learning_rate": 0.0003082232518047624,
      "loss": 3.404,
      "step": 45250
    },
    {
      "epoch": 4.87568614788505,
      "grad_norm": 0.6882241368293762,
      "learning_rate": 0.00030790001077470096,
      "loss": 3.3983,
      "step": 45300
    },
    {
      "epoch": 4.8810676999246585,
      "grad_norm": 0.6609480977058411,
      "learning_rate": 0.00030757676974463955,
      "loss": 3.4132,
      "step": 45350
    },
    {
      "epoch": 4.886449251964266,
      "grad_norm": 0.664958655834198,
      "learning_rate": 0.0003072535287145781,
      "loss": 3.3773,
      "step": 45400
    },
    {
      "epoch": 4.891830804003875,
      "grad_norm": 0.715764582157135,
      "learning_rate": 0.00030693028768451674,
      "loss": 3.3921,
      "step": 45450
    },
    {
      "epoch": 4.897212356043483,
      "grad_norm": 0.6146078109741211,
      "learning_rate": 0.00030660704665445534,
      "loss": 3.3802,
      "step": 45500
    },
    {
      "epoch": 4.902593908083091,
      "grad_norm": 0.6503034830093384,
      "learning_rate": 0.0003062838056243939,
      "loss": 3.3987,
      "step": 45550
    },
    {
      "epoch": 4.9079754601226995,
      "grad_norm": 0.6508064270019531,
      "learning_rate": 0.00030596056459433247,
      "loss": 3.3721,
      "step": 45600
    },
    {
      "epoch": 4.913357012162308,
      "grad_norm": 0.6286031603813171,
      "learning_rate": 0.00030563732356427107,
      "loss": 3.3893,
      "step": 45650
    },
    {
      "epoch": 4.918738564201916,
      "grad_norm": 0.6623638868331909,
      "learning_rate": 0.00030531408253420966,
      "loss": 3.379,
      "step": 45700
    },
    {
      "epoch": 4.924120116241524,
      "grad_norm": 0.668628454208374,
      "learning_rate": 0.00030499084150414826,
      "loss": 3.3861,
      "step": 45750
    },
    {
      "epoch": 4.929501668281132,
      "grad_norm": 0.6662799119949341,
      "learning_rate": 0.00030466760047408685,
      "loss": 3.3934,
      "step": 45800
    },
    {
      "epoch": 4.9348832203207404,
      "grad_norm": 0.6624265313148499,
      "learning_rate": 0.0003043443594440254,
      "loss": 3.3864,
      "step": 45850
    },
    {
      "epoch": 4.940264772360349,
      "grad_norm": 0.6628495454788208,
      "learning_rate": 0.0003040275832345652,
      "loss": 3.404,
      "step": 45900
    },
    {
      "epoch": 4.945646324399957,
      "grad_norm": 0.6551867127418518,
      "learning_rate": 0.0003037043422045038,
      "loss": 3.3789,
      "step": 45950
    },
    {
      "epoch": 4.951027876439565,
      "grad_norm": 0.7384966015815735,
      "learning_rate": 0.0003033811011744424,
      "loss": 3.3842,
      "step": 46000
    },
    {
      "epoch": 4.951027876439565,
      "eval_accuracy": 0.3805596731981902,
      "eval_loss": 3.4086427688598633,
      "eval_runtime": 184.6207,
      "eval_samples_per_second": 97.557,
      "eval_steps_per_second": 6.099,
      "step": 46000
    },
    {
      "epoch": 4.956409428479174,
      "grad_norm": 0.680334746837616,
      "learning_rate": 0.00030305786014438093,
      "loss": 3.3907,
      "step": 46050
    },
    {
      "epoch": 4.961790980518781,
      "grad_norm": 0.6668956875801086,
      "learning_rate": 0.0003027346191143196,
      "loss": 3.3966,
      "step": 46100
    },
    {
      "epoch": 4.96717253255839,
      "grad_norm": 0.6901334524154663,
      "learning_rate": 0.00030241137808425817,
      "loss": 3.3896,
      "step": 46150
    },
    {
      "epoch": 4.9725540845979985,
      "grad_norm": 0.6751418709754944,
      "learning_rate": 0.0003020881370541967,
      "loss": 3.3745,
      "step": 46200
    },
    {
      "epoch": 4.977935636637606,
      "grad_norm": 0.7121784090995789,
      "learning_rate": 0.0003017648960241353,
      "loss": 3.385,
      "step": 46250
    },
    {
      "epoch": 4.983317188677215,
      "grad_norm": 0.6521340012550354,
      "learning_rate": 0.00030144165499407385,
      "loss": 3.4098,
      "step": 46300
    },
    {
      "epoch": 4.988698740716822,
      "grad_norm": 0.8021095395088196,
      "learning_rate": 0.00030111841396401244,
      "loss": 3.4035,
      "step": 46350
    },
    {
      "epoch": 4.994080292756431,
      "grad_norm": 0.6512081027030945,
      "learning_rate": 0.0003007951729339511,
      "loss": 3.3843,
      "step": 46400
    },
    {
      "epoch": 4.9994618447960395,
      "grad_norm": 0.7062845230102539,
      "learning_rate": 0.00030047193190388963,
      "loss": 3.396,
      "step": 46450
    },
    {
      "epoch": 5.004843396835647,
      "grad_norm": 0.6628386974334717,
      "learning_rate": 0.0003001486908738282,
      "loss": 3.3034,
      "step": 46500
    },
    {
      "epoch": 5.010224948875256,
      "grad_norm": 0.7442699670791626,
      "learning_rate": 0.0002998254498437668,
      "loss": 3.2796,
      "step": 46550
    },
    {
      "epoch": 5.015606500914864,
      "grad_norm": 0.7247018218040466,
      "learning_rate": 0.0002995022088137054,
      "loss": 3.2988,
      "step": 46600
    },
    {
      "epoch": 5.020988052954472,
      "grad_norm": 0.6431019902229309,
      "learning_rate": 0.00029917896778364396,
      "loss": 3.3149,
      "step": 46650
    },
    {
      "epoch": 5.0263696049940805,
      "grad_norm": 0.6791779398918152,
      "learning_rate": 0.00029885572675358255,
      "loss": 3.2962,
      "step": 46700
    },
    {
      "epoch": 5.031751157033688,
      "grad_norm": 0.6947548985481262,
      "learning_rate": 0.00029853248572352114,
      "loss": 3.3008,
      "step": 46750
    },
    {
      "epoch": 5.037132709073297,
      "grad_norm": 0.6451291441917419,
      "learning_rate": 0.00029820924469345974,
      "loss": 3.3082,
      "step": 46800
    },
    {
      "epoch": 5.042514261112905,
      "grad_norm": 0.698814332485199,
      "learning_rate": 0.0002978860036633983,
      "loss": 3.2975,
      "step": 46850
    },
    {
      "epoch": 5.047895813152513,
      "grad_norm": 0.702712893486023,
      "learning_rate": 0.00029756276263333693,
      "loss": 3.3188,
      "step": 46900
    },
    {
      "epoch": 5.0532773651921215,
      "grad_norm": 0.6753806471824646,
      "learning_rate": 0.00029723952160327547,
      "loss": 3.3139,
      "step": 46950
    },
    {
      "epoch": 5.05865891723173,
      "grad_norm": 0.7441586256027222,
      "learning_rate": 0.00029691628057321406,
      "loss": 3.3241,
      "step": 47000
    },
    {
      "epoch": 5.05865891723173,
      "eval_accuracy": 0.3806568088798963,
      "eval_loss": 3.4130375385284424,
      "eval_runtime": 184.7285,
      "eval_samples_per_second": 97.5,
      "eval_steps_per_second": 6.095,
      "step": 47000
    },
    {
      "epoch": 5.064040469271338,
      "grad_norm": 0.7186902761459351,
      "learning_rate": 0.00029659303954315266,
      "loss": 3.3039,
      "step": 47050
    },
    {
      "epoch": 5.069422021310946,
      "grad_norm": 0.6443047523498535,
      "learning_rate": 0.00029626979851309125,
      "loss": 3.3242,
      "step": 47100
    },
    {
      "epoch": 5.074803573350554,
      "grad_norm": 0.6580514907836914,
      "learning_rate": 0.00029594655748302985,
      "loss": 3.2998,
      "step": 47150
    },
    {
      "epoch": 5.080185125390162,
      "grad_norm": 0.6719970107078552,
      "learning_rate": 0.0002956233164529684,
      "loss": 3.3033,
      "step": 47200
    },
    {
      "epoch": 5.085566677429771,
      "grad_norm": 0.7296539545059204,
      "learning_rate": 0.000295300075422907,
      "loss": 3.299,
      "step": 47250
    },
    {
      "epoch": 5.090948229469379,
      "grad_norm": 0.7392880320549011,
      "learning_rate": 0.0002949768343928456,
      "loss": 3.3106,
      "step": 47300
    },
    {
      "epoch": 5.096329781508987,
      "grad_norm": 0.6548142433166504,
      "learning_rate": 0.00029465359336278417,
      "loss": 3.2949,
      "step": 47350
    },
    {
      "epoch": 5.101711333548596,
      "grad_norm": 0.6920472383499146,
      "learning_rate": 0.0002943303523327227,
      "loss": 3.302,
      "step": 47400
    },
    {
      "epoch": 5.107092885588203,
      "grad_norm": 0.6915966272354126,
      "learning_rate": 0.00029400711130266136,
      "loss": 3.315,
      "step": 47450
    },
    {
      "epoch": 5.112474437627812,
      "grad_norm": 0.7536485195159912,
      "learning_rate": 0.0002936838702725999,
      "loss": 3.3202,
      "step": 47500
    },
    {
      "epoch": 5.1178559896674205,
      "grad_norm": 0.6746701002120972,
      "learning_rate": 0.0002933606292425385,
      "loss": 3.3131,
      "step": 47550
    },
    {
      "epoch": 5.123237541707028,
      "grad_norm": 0.6896171569824219,
      "learning_rate": 0.0002930373882124771,
      "loss": 3.3146,
      "step": 47600
    },
    {
      "epoch": 5.128619093746637,
      "grad_norm": 0.6922616362571716,
      "learning_rate": 0.0002927141471824157,
      "loss": 3.3254,
      "step": 47650
    },
    {
      "epoch": 5.134000645786244,
      "grad_norm": 0.6527367234230042,
      "learning_rate": 0.0002923909061523542,
      "loss": 3.3171,
      "step": 47700
    },
    {
      "epoch": 5.139382197825853,
      "grad_norm": 0.7178485989570618,
      "learning_rate": 0.0002920676651222928,
      "loss": 3.322,
      "step": 47750
    },
    {
      "epoch": 5.1447637498654615,
      "grad_norm": 0.6550273895263672,
      "learning_rate": 0.0002917444240922314,
      "loss": 3.3252,
      "step": 47800
    },
    {
      "epoch": 5.150145301905069,
      "grad_norm": 0.6551401615142822,
      "learning_rate": 0.00029142118306216996,
      "loss": 3.3017,
      "step": 47850
    },
    {
      "epoch": 5.155526853944678,
      "grad_norm": 0.7181161046028137,
      "learning_rate": 0.0002910979420321086,
      "loss": 3.3123,
      "step": 47900
    },
    {
      "epoch": 5.160908405984286,
      "grad_norm": 0.7358765602111816,
      "learning_rate": 0.00029077470100204715,
      "loss": 3.3117,
      "step": 47950
    },
    {
      "epoch": 5.166289958023894,
      "grad_norm": 0.6698744297027588,
      "learning_rate": 0.00029045145997198574,
      "loss": 3.312,
      "step": 48000
    },
    {
      "epoch": 5.166289958023894,
      "eval_accuracy": 0.3811957272034561,
      "eval_loss": 3.4106051921844482,
      "eval_runtime": 184.3793,
      "eval_samples_per_second": 97.685,
      "eval_steps_per_second": 6.107,
      "step": 48000
    },
    {
      "epoch": 5.1716715100635025,
      "grad_norm": 0.7061935067176819,
      "learning_rate": 0.00029012821894192433,
      "loss": 3.3312,
      "step": 48050
    },
    {
      "epoch": 5.17705306210311,
      "grad_norm": 0.6918174028396606,
      "learning_rate": 0.00028980497791186293,
      "loss": 3.3123,
      "step": 48100
    },
    {
      "epoch": 5.182434614142719,
      "grad_norm": 0.6964541673660278,
      "learning_rate": 0.00028948820170240274,
      "loss": 3.3137,
      "step": 48150
    },
    {
      "epoch": 5.187816166182327,
      "grad_norm": 0.6781343221664429,
      "learning_rate": 0.00028916496067234133,
      "loss": 3.3102,
      "step": 48200
    },
    {
      "epoch": 5.193197718221935,
      "grad_norm": 0.6664396524429321,
      "learning_rate": 0.0002888417196422799,
      "loss": 3.3151,
      "step": 48250
    },
    {
      "epoch": 5.198579270261543,
      "grad_norm": 0.6614742875099182,
      "learning_rate": 0.00028851847861221847,
      "loss": 3.3069,
      "step": 48300
    },
    {
      "epoch": 5.203960822301152,
      "grad_norm": 0.6851457953453064,
      "learning_rate": 0.00028819523758215706,
      "loss": 3.3248,
      "step": 48350
    },
    {
      "epoch": 5.20934237434076,
      "grad_norm": 0.7182928323745728,
      "learning_rate": 0.00028787199655209566,
      "loss": 3.3229,
      "step": 48400
    },
    {
      "epoch": 5.214723926380368,
      "grad_norm": 0.7236764430999756,
      "learning_rate": 0.00028754875552203425,
      "loss": 3.337,
      "step": 48450
    },
    {
      "epoch": 5.220105478419977,
      "grad_norm": 0.6826680302619934,
      "learning_rate": 0.0002872255144919728,
      "loss": 3.3206,
      "step": 48500
    },
    {
      "epoch": 5.225487030459584,
      "grad_norm": 0.7440321445465088,
      "learning_rate": 0.00028690227346191144,
      "loss": 3.3193,
      "step": 48550
    },
    {
      "epoch": 5.230868582499193,
      "grad_norm": 0.698179304599762,
      "learning_rate": 0.00028657903243185,
      "loss": 3.3166,
      "step": 48600
    },
    {
      "epoch": 5.236250134538801,
      "grad_norm": 0.6869778037071228,
      "learning_rate": 0.0002862557914017886,
      "loss": 3.3437,
      "step": 48650
    },
    {
      "epoch": 5.241631686578409,
      "grad_norm": 0.6472514867782593,
      "learning_rate": 0.00028593255037172717,
      "loss": 3.3263,
      "step": 48700
    },
    {
      "epoch": 5.247013238618018,
      "grad_norm": 0.6622984409332275,
      "learning_rate": 0.00028560930934166576,
      "loss": 3.3033,
      "step": 48750
    },
    {
      "epoch": 5.252394790657625,
      "grad_norm": 0.687536358833313,
      "learning_rate": 0.00028528606831160436,
      "loss": 3.3054,
      "step": 48800
    },
    {
      "epoch": 5.257776342697234,
      "grad_norm": 0.6665416359901428,
      "learning_rate": 0.0002849628272815429,
      "loss": 3.3273,
      "step": 48850
    },
    {
      "epoch": 5.2631578947368425,
      "grad_norm": 0.7293050289154053,
      "learning_rate": 0.0002846395862514815,
      "loss": 3.3237,
      "step": 48900
    },
    {
      "epoch": 5.26853944677645,
      "grad_norm": 0.6864955425262451,
      "learning_rate": 0.0002843163452214201,
      "loss": 3.3194,
      "step": 48950
    },
    {
      "epoch": 5.273920998816059,
      "grad_norm": 0.6750340461730957,
      "learning_rate": 0.0002839931041913587,
      "loss": 3.3303,
      "step": 49000
    },
    {
      "epoch": 5.273920998816059,
      "eval_accuracy": 0.3815918756324277,
      "eval_loss": 3.406046152114868,
      "eval_runtime": 184.5273,
      "eval_samples_per_second": 97.606,
      "eval_steps_per_second": 6.102,
      "step": 49000
    },
    {
      "epoch": 5.279302550855666,
      "grad_norm": 0.6839402914047241,
      "learning_rate": 0.0002836698631612972,
      "loss": 3.3091,
      "step": 49050
    },
    {
      "epoch": 5.284684102895275,
      "grad_norm": 0.6985600590705872,
      "learning_rate": 0.00028334662213123587,
      "loss": 3.3188,
      "step": 49100
    },
    {
      "epoch": 5.2900656549348835,
      "grad_norm": 0.6883201599121094,
      "learning_rate": 0.0002830233811011744,
      "loss": 3.3188,
      "step": 49150
    },
    {
      "epoch": 5.295447206974491,
      "grad_norm": 0.7022517323493958,
      "learning_rate": 0.000282700140071113,
      "loss": 3.3139,
      "step": 49200
    },
    {
      "epoch": 5.3008287590141,
      "grad_norm": 0.7535263299942017,
      "learning_rate": 0.0002823768990410516,
      "loss": 3.3594,
      "step": 49250
    },
    {
      "epoch": 5.306210311053708,
      "grad_norm": 0.7504818439483643,
      "learning_rate": 0.00028205365801099014,
      "loss": 3.3368,
      "step": 49300
    },
    {
      "epoch": 5.311591863093316,
      "grad_norm": 0.6709638237953186,
      "learning_rate": 0.00028173041698092874,
      "loss": 3.3417,
      "step": 49350
    },
    {
      "epoch": 5.316973415132924,
      "grad_norm": 0.7753866910934448,
      "learning_rate": 0.00028140717595086733,
      "loss": 3.3401,
      "step": 49400
    },
    {
      "epoch": 5.322354967172533,
      "grad_norm": 0.753419041633606,
      "learning_rate": 0.0002810839349208059,
      "loss": 3.3209,
      "step": 49450
    },
    {
      "epoch": 5.327736519212141,
      "grad_norm": 0.7451731562614441,
      "learning_rate": 0.00028076069389074447,
      "loss": 3.3269,
      "step": 49500
    },
    {
      "epoch": 5.333118071251749,
      "grad_norm": 0.6875990033149719,
      "learning_rate": 0.0002804374528606831,
      "loss": 3.3183,
      "step": 49550
    },
    {
      "epoch": 5.338499623291357,
      "grad_norm": 0.7078640460968018,
      "learning_rate": 0.00028011421183062166,
      "loss": 3.3154,
      "step": 49600
    },
    {
      "epoch": 5.343881175330965,
      "grad_norm": 0.7015400528907776,
      "learning_rate": 0.00027979097080056025,
      "loss": 3.3199,
      "step": 49650
    },
    {
      "epoch": 5.349262727370574,
      "grad_norm": 0.6855900883674622,
      "learning_rate": 0.00027946772977049885,
      "loss": 3.3365,
      "step": 49700
    },
    {
      "epoch": 5.354644279410182,
      "grad_norm": 0.6965287923812866,
      "learning_rate": 0.00027914448874043744,
      "loss": 3.3461,
      "step": 49750
    },
    {
      "epoch": 5.36002583144979,
      "grad_norm": 0.706678569316864,
      "learning_rate": 0.00027882124771037603,
      "loss": 3.3384,
      "step": 49800
    },
    {
      "epoch": 5.365407383489399,
      "grad_norm": 0.7243395447731018,
      "learning_rate": 0.0002784980066803146,
      "loss": 3.3406,
      "step": 49850
    },
    {
      "epoch": 5.370788935529006,
      "grad_norm": 0.7101507782936096,
      "learning_rate": 0.00027817476565025317,
      "loss": 3.3327,
      "step": 49900
    },
    {
      "epoch": 5.376170487568615,
      "grad_norm": 0.7034843564033508,
      "learning_rate": 0.00027785152462019176,
      "loss": 3.3268,
      "step": 49950
    },
    {
      "epoch": 5.3815520396082235,
      "grad_norm": 0.6907040476799011,
      "learning_rate": 0.00027752828359013036,
      "loss": 3.3415,
      "step": 50000
    },
    {
      "epoch": 5.3815520396082235,
      "eval_accuracy": 0.38182634856433134,
      "eval_loss": 3.402235507965088,
      "eval_runtime": 184.8851,
      "eval_samples_per_second": 97.417,
      "eval_steps_per_second": 6.09,
      "step": 50000
    },
    {
      "epoch": 5.386933591647831,
      "grad_norm": 0.6705796122550964,
      "learning_rate": 0.0002772050425600689,
      "loss": 3.356,
      "step": 50050
    },
    {
      "epoch": 5.39231514368744,
      "grad_norm": 0.7291210889816284,
      "learning_rate": 0.00027688180153000755,
      "loss": 3.3134,
      "step": 50100
    },
    {
      "epoch": 5.397696695727047,
      "grad_norm": 0.7164068222045898,
      "learning_rate": 0.0002765585604999461,
      "loss": 3.3312,
      "step": 50150
    },
    {
      "epoch": 5.403078247766656,
      "grad_norm": 0.7845756411552429,
      "learning_rate": 0.0002762353194698847,
      "loss": 3.3374,
      "step": 50200
    },
    {
      "epoch": 5.4084597998062645,
      "grad_norm": 0.7239874601364136,
      "learning_rate": 0.0002759120784398233,
      "loss": 3.3221,
      "step": 50250
    },
    {
      "epoch": 5.413841351845872,
      "grad_norm": 0.6949788331985474,
      "learning_rate": 0.00027558883740976187,
      "loss": 3.341,
      "step": 50300
    },
    {
      "epoch": 5.419222903885481,
      "grad_norm": 0.680122971534729,
      "learning_rate": 0.0002752655963797004,
      "loss": 3.3322,
      "step": 50350
    },
    {
      "epoch": 5.424604455925088,
      "grad_norm": 0.679383397102356,
      "learning_rate": 0.000274942355349639,
      "loss": 3.3406,
      "step": 50400
    },
    {
      "epoch": 5.429986007964697,
      "grad_norm": 0.7332752346992493,
      "learning_rate": 0.0002746191143195776,
      "loss": 3.3444,
      "step": 50450
    },
    {
      "epoch": 5.435367560004305,
      "grad_norm": 0.7282567620277405,
      "learning_rate": 0.0002742958732895162,
      "loss": 3.3267,
      "step": 50500
    },
    {
      "epoch": 5.440749112043913,
      "grad_norm": 0.7126004695892334,
      "learning_rate": 0.000273979097080056,
      "loss": 3.3275,
      "step": 50550
    },
    {
      "epoch": 5.446130664083522,
      "grad_norm": 0.7799147963523865,
      "learning_rate": 0.0002736558560499946,
      "loss": 3.3277,
      "step": 50600
    },
    {
      "epoch": 5.45151221612313,
      "grad_norm": 0.6828324794769287,
      "learning_rate": 0.0002733390798405344,
      "loss": 3.3377,
      "step": 50650
    },
    {
      "epoch": 5.456893768162738,
      "grad_norm": 0.7099668383598328,
      "learning_rate": 0.000273015838810473,
      "loss": 3.325,
      "step": 50700
    },
    {
      "epoch": 5.462275320202346,
      "grad_norm": 0.7239611148834229,
      "learning_rate": 0.0002726925977804116,
      "loss": 3.3314,
      "step": 50750
    },
    {
      "epoch": 5.467656872241955,
      "grad_norm": 0.6668616533279419,
      "learning_rate": 0.00027236935675035013,
      "loss": 3.3309,
      "step": 50800
    },
    {
      "epoch": 5.473038424281563,
      "grad_norm": 0.6595761179924011,
      "learning_rate": 0.00027204611572028873,
      "loss": 3.3113,
      "step": 50850
    },
    {
      "epoch": 5.478419976321171,
      "grad_norm": 0.6890519261360168,
      "learning_rate": 0.0002717228746902273,
      "loss": 3.324,
      "step": 50900
    },
    {
      "epoch": 5.483801528360779,
      "grad_norm": 0.7110984325408936,
      "learning_rate": 0.0002713996336601659,
      "loss": 3.3335,
      "step": 50950
    },
    {
      "epoch": 5.489183080400387,
      "grad_norm": 0.6996448040008545,
      "learning_rate": 0.0002710763926301045,
      "loss": 3.3321,
      "step": 51000
    },
    {
      "epoch": 5.489183080400387,
      "eval_accuracy": 0.38222977673678654,
      "eval_loss": 3.397602081298828,
      "eval_runtime": 184.4703,
      "eval_samples_per_second": 97.636,
      "eval_steps_per_second": 6.104,
      "step": 51000
    },
    {
      "epoch": 5.494564632439996,
      "grad_norm": 0.6930586099624634,
      "learning_rate": 0.00027075315160004305,
      "loss": 3.3259,
      "step": 51050
    },
    {
      "epoch": 5.499946184479604,
      "grad_norm": 0.6693201661109924,
      "learning_rate": 0.00027042991056998165,
      "loss": 3.3238,
      "step": 51100
    },
    {
      "epoch": 5.505327736519212,
      "grad_norm": 0.7076327204704285,
      "learning_rate": 0.00027010666953992024,
      "loss": 3.3132,
      "step": 51150
    },
    {
      "epoch": 5.510709288558821,
      "grad_norm": 0.7098067998886108,
      "learning_rate": 0.00026978342850985884,
      "loss": 3.3202,
      "step": 51200
    },
    {
      "epoch": 5.516090840598428,
      "grad_norm": 0.7085769176483154,
      "learning_rate": 0.0002694601874797974,
      "loss": 3.3231,
      "step": 51250
    },
    {
      "epoch": 5.521472392638037,
      "grad_norm": 0.7210639715194702,
      "learning_rate": 0.000269136946449736,
      "loss": 3.3344,
      "step": 51300
    },
    {
      "epoch": 5.5268539446776455,
      "grad_norm": 0.7563997507095337,
      "learning_rate": 0.00026881370541967457,
      "loss": 3.3448,
      "step": 51350
    },
    {
      "epoch": 5.532235496717253,
      "grad_norm": 0.6973581314086914,
      "learning_rate": 0.00026849046438961316,
      "loss": 3.3427,
      "step": 51400
    },
    {
      "epoch": 5.537617048756862,
      "grad_norm": 0.6693292856216431,
      "learning_rate": 0.00026816722335955176,
      "loss": 3.3359,
      "step": 51450
    },
    {
      "epoch": 5.542998600796469,
      "grad_norm": 0.7080926299095154,
      "learning_rate": 0.00026784398232949035,
      "loss": 3.3283,
      "step": 51500
    },
    {
      "epoch": 5.548380152836078,
      "grad_norm": 0.6850876808166504,
      "learning_rate": 0.00026752074129942895,
      "loss": 3.3366,
      "step": 51550
    },
    {
      "epoch": 5.553761704875686,
      "grad_norm": 0.7193228602409363,
      "learning_rate": 0.0002671975002693675,
      "loss": 3.3334,
      "step": 51600
    },
    {
      "epoch": 5.559143256915294,
      "grad_norm": 0.7010207176208496,
      "learning_rate": 0.0002668742592393061,
      "loss": 3.3395,
      "step": 51650
    },
    {
      "epoch": 5.564524808954903,
      "grad_norm": 0.6764205694198608,
      "learning_rate": 0.0002665510182092447,
      "loss": 3.3341,
      "step": 51700
    },
    {
      "epoch": 5.569906360994511,
      "grad_norm": 0.6543567180633545,
      "learning_rate": 0.00026622777717918327,
      "loss": 3.3362,
      "step": 51750
    },
    {
      "epoch": 5.575287913034119,
      "grad_norm": 0.7548144459724426,
      "learning_rate": 0.0002659045361491218,
      "loss": 3.3307,
      "step": 51800
    },
    {
      "epoch": 5.580669465073727,
      "grad_norm": 0.7587510943412781,
      "learning_rate": 0.00026558129511906046,
      "loss": 3.3517,
      "step": 51850
    },
    {
      "epoch": 5.586051017113336,
      "grad_norm": 0.6862114667892456,
      "learning_rate": 0.000265258054088999,
      "loss": 3.3552,
      "step": 51900
    },
    {
      "epoch": 5.591432569152944,
      "grad_norm": 0.8210598230361938,
      "learning_rate": 0.0002649348130589376,
      "loss": 3.3333,
      "step": 51950
    },
    {
      "epoch": 5.596814121192552,
      "grad_norm": 0.7292066216468811,
      "learning_rate": 0.0002646115720288762,
      "loss": 3.345,
      "step": 52000
    },
    {
      "epoch": 5.596814121192552,
      "eval_accuracy": 0.3826878573118124,
      "eval_loss": 3.3923404216766357,
      "eval_runtime": 184.6516,
      "eval_samples_per_second": 97.54,
      "eval_steps_per_second": 6.098,
      "step": 52000
    },
    {
      "epoch": 5.60219567323216,
      "grad_norm": 0.7065765857696533,
      "learning_rate": 0.00026428833099881473,
      "loss": 3.3593,
      "step": 52050
    },
    {
      "epoch": 5.607577225271768,
      "grad_norm": 0.7100358009338379,
      "learning_rate": 0.0002639650899687533,
      "loss": 3.3294,
      "step": 52100
    },
    {
      "epoch": 5.612958777311377,
      "grad_norm": 0.7037196755409241,
      "learning_rate": 0.0002636418489386919,
      "loss": 3.3131,
      "step": 52150
    },
    {
      "epoch": 5.618340329350985,
      "grad_norm": 0.7100152373313904,
      "learning_rate": 0.0002633186079086305,
      "loss": 3.3275,
      "step": 52200
    },
    {
      "epoch": 5.623721881390593,
      "grad_norm": 0.6936700344085693,
      "learning_rate": 0.00026299536687856905,
      "loss": 3.3193,
      "step": 52250
    },
    {
      "epoch": 5.629103433430201,
      "grad_norm": 0.7310315370559692,
      "learning_rate": 0.0002626721258485077,
      "loss": 3.3352,
      "step": 52300
    },
    {
      "epoch": 5.634484985469809,
      "grad_norm": 0.7135825753211975,
      "learning_rate": 0.00026234888481844624,
      "loss": 3.33,
      "step": 52350
    },
    {
      "epoch": 5.639866537509418,
      "grad_norm": 0.6975454688072205,
      "learning_rate": 0.00026202564378838484,
      "loss": 3.3285,
      "step": 52400
    },
    {
      "epoch": 5.645248089549026,
      "grad_norm": 0.6941152811050415,
      "learning_rate": 0.00026170240275832343,
      "loss": 3.3472,
      "step": 52450
    },
    {
      "epoch": 5.650629641588634,
      "grad_norm": 0.7469999194145203,
      "learning_rate": 0.00026137916172826203,
      "loss": 3.3204,
      "step": 52500
    },
    {
      "epoch": 5.656011193628243,
      "grad_norm": 0.7071282863616943,
      "learning_rate": 0.0002610559206982006,
      "loss": 3.3354,
      "step": 52550
    },
    {
      "epoch": 5.66139274566785,
      "grad_norm": 0.705010175704956,
      "learning_rate": 0.00026073267966813916,
      "loss": 3.3401,
      "step": 52600
    },
    {
      "epoch": 5.666774297707459,
      "grad_norm": 0.769747257232666,
      "learning_rate": 0.00026040943863807776,
      "loss": 3.3283,
      "step": 52650
    },
    {
      "epoch": 5.672155849747067,
      "grad_norm": 0.7160783410072327,
      "learning_rate": 0.00026008619760801635,
      "loss": 3.3407,
      "step": 52700
    },
    {
      "epoch": 5.677537401786675,
      "grad_norm": 0.7041793465614319,
      "learning_rate": 0.00025976295657795495,
      "loss": 3.3489,
      "step": 52750
    },
    {
      "epoch": 5.682918953826284,
      "grad_norm": 0.7076753377914429,
      "learning_rate": 0.0002594397155478935,
      "loss": 3.343,
      "step": 52800
    },
    {
      "epoch": 5.688300505865891,
      "grad_norm": 0.7296324968338013,
      "learning_rate": 0.00025911647451783214,
      "loss": 3.3402,
      "step": 52850
    },
    {
      "epoch": 5.6936820579055,
      "grad_norm": 0.70832759141922,
      "learning_rate": 0.0002587932334877707,
      "loss": 3.3191,
      "step": 52900
    },
    {
      "epoch": 5.699063609945108,
      "grad_norm": 0.6894454956054688,
      "learning_rate": 0.00025846999245770927,
      "loss": 3.3372,
      "step": 52950
    },
    {
      "epoch": 5.704445161984716,
      "grad_norm": 0.7215243577957153,
      "learning_rate": 0.00025814675142764787,
      "loss": 3.3433,
      "step": 53000
    },
    {
      "epoch": 5.704445161984716,
      "eval_accuracy": 0.38319276728148627,
      "eval_loss": 3.3882508277893066,
      "eval_runtime": 186.9278,
      "eval_samples_per_second": 96.353,
      "eval_steps_per_second": 6.024,
      "step": 53000
    },
    {
      "epoch": 5.709826714024325,
      "grad_norm": 0.7630006074905396,
      "learning_rate": 0.00025782351039758646,
      "loss": 3.3234,
      "step": 53050
    },
    {
      "epoch": 5.715208266063933,
      "grad_norm": 0.6765692234039307,
      "learning_rate": 0.000257500269367525,
      "loss": 3.3324,
      "step": 53100
    },
    {
      "epoch": 5.720589818103541,
      "grad_norm": 0.7011041641235352,
      "learning_rate": 0.0002571770283374636,
      "loss": 3.3426,
      "step": 53150
    },
    {
      "epoch": 5.725971370143149,
      "grad_norm": 0.6970390677452087,
      "learning_rate": 0.0002568537873074022,
      "loss": 3.3182,
      "step": 53200
    },
    {
      "epoch": 5.731352922182758,
      "grad_norm": 0.7535760402679443,
      "learning_rate": 0.0002565305462773408,
      "loss": 3.3382,
      "step": 53250
    },
    {
      "epoch": 5.736734474222366,
      "grad_norm": 0.7167165279388428,
      "learning_rate": 0.0002562073052472794,
      "loss": 3.3495,
      "step": 53300
    },
    {
      "epoch": 5.742116026261974,
      "grad_norm": 0.7161158323287964,
      "learning_rate": 0.0002558840642172179,
      "loss": 3.3477,
      "step": 53350
    },
    {
      "epoch": 5.747497578301582,
      "grad_norm": 0.7228608727455139,
      "learning_rate": 0.00025556082318715657,
      "loss": 3.3316,
      "step": 53400
    },
    {
      "epoch": 5.75287913034119,
      "grad_norm": 0.7354365587234497,
      "learning_rate": 0.0002552375821570951,
      "loss": 3.35,
      "step": 53450
    },
    {
      "epoch": 5.758260682380799,
      "grad_norm": 0.7172362208366394,
      "learning_rate": 0.0002549143411270337,
      "loss": 3.3494,
      "step": 53500
    },
    {
      "epoch": 5.763642234420407,
      "grad_norm": 0.7296724319458008,
      "learning_rate": 0.0002545911000969723,
      "loss": 3.3451,
      "step": 53550
    },
    {
      "epoch": 5.769023786460015,
      "grad_norm": 0.7641502022743225,
      "learning_rate": 0.0002542678590669109,
      "loss": 3.3496,
      "step": 53600
    },
    {
      "epoch": 5.774405338499624,
      "grad_norm": 0.6677383184432983,
      "learning_rate": 0.00025394461803684943,
      "loss": 3.3406,
      "step": 53650
    },
    {
      "epoch": 5.779786890539231,
      "grad_norm": 0.7079816460609436,
      "learning_rate": 0.00025362137700678803,
      "loss": 3.3472,
      "step": 53700
    },
    {
      "epoch": 5.78516844257884,
      "grad_norm": 0.8173105120658875,
      "learning_rate": 0.0002532981359767266,
      "loss": 3.3269,
      "step": 53750
    },
    {
      "epoch": 5.790549994618448,
      "grad_norm": 0.740807294845581,
      "learning_rate": 0.0002529748949466652,
      "loss": 3.3255,
      "step": 53800
    },
    {
      "epoch": 5.795931546658056,
      "grad_norm": 0.7093555927276611,
      "learning_rate": 0.0002526516539166038,
      "loss": 3.3337,
      "step": 53850
    },
    {
      "epoch": 5.801313098697665,
      "grad_norm": 0.7256129384040833,
      "learning_rate": 0.00025232841288654235,
      "loss": 3.3258,
      "step": 53900
    },
    {
      "epoch": 5.806694650737272,
      "grad_norm": 0.6999818682670593,
      "learning_rate": 0.00025200517185648095,
      "loss": 3.3464,
      "step": 53950
    },
    {
      "epoch": 5.812076202776881,
      "grad_norm": 0.7375802397727966,
      "learning_rate": 0.00025168193082641954,
      "loss": 3.3351,
      "step": 54000
    },
    {
      "epoch": 5.812076202776881,
      "eval_accuracy": 0.3836312903367055,
      "eval_loss": 3.3833701610565186,
      "eval_runtime": 185.1377,
      "eval_samples_per_second": 97.284,
      "eval_steps_per_second": 6.082,
      "step": 54000
    },
    {
      "epoch": 5.817457754816489,
      "grad_norm": 0.714192807674408,
      "learning_rate": 0.00025135868979635814,
      "loss": 3.3304,
      "step": 54050
    },
    {
      "epoch": 5.822839306856097,
      "grad_norm": 0.7494600415229797,
      "learning_rate": 0.0002510354487662967,
      "loss": 3.3435,
      "step": 54100
    },
    {
      "epoch": 5.828220858895706,
      "grad_norm": 0.730343759059906,
      "learning_rate": 0.0002507122077362353,
      "loss": 3.3399,
      "step": 54150
    },
    {
      "epoch": 5.833602410935313,
      "grad_norm": 0.7611250877380371,
      "learning_rate": 0.00025038896670617387,
      "loss": 3.349,
      "step": 54200
    },
    {
      "epoch": 5.838983962974922,
      "grad_norm": 0.7368343472480774,
      "learning_rate": 0.00025006572567611246,
      "loss": 3.3408,
      "step": 54250
    },
    {
      "epoch": 5.84436551501453,
      "grad_norm": 0.711065948009491,
      "learning_rate": 0.00024974248464605106,
      "loss": 3.3482,
      "step": 54300
    },
    {
      "epoch": 5.849747067054138,
      "grad_norm": 0.7511163353919983,
      "learning_rate": 0.00024941924361598965,
      "loss": 3.3501,
      "step": 54350
    },
    {
      "epoch": 5.855128619093747,
      "grad_norm": 0.682986319065094,
      "learning_rate": 0.00024909600258592825,
      "loss": 3.3234,
      "step": 54400
    },
    {
      "epoch": 5.860510171133355,
      "grad_norm": 0.6947693824768066,
      "learning_rate": 0.0002487727615558668,
      "loss": 3.3509,
      "step": 54450
    },
    {
      "epoch": 5.865891723172963,
      "grad_norm": 0.6829459071159363,
      "learning_rate": 0.0002484495205258054,
      "loss": 3.3442,
      "step": 54500
    },
    {
      "epoch": 5.871273275212571,
      "grad_norm": 0.7875406742095947,
      "learning_rate": 0.000248126279495744,
      "loss": 3.3377,
      "step": 54550
    },
    {
      "epoch": 5.87665482725218,
      "grad_norm": 0.7647419571876526,
      "learning_rate": 0.00024780303846568257,
      "loss": 3.3455,
      "step": 54600
    },
    {
      "epoch": 5.882036379291788,
      "grad_norm": 0.687256395816803,
      "learning_rate": 0.0002474797974356211,
      "loss": 3.3324,
      "step": 54650
    },
    {
      "epoch": 5.887417931331396,
      "grad_norm": 0.7644538283348083,
      "learning_rate": 0.00024715655640555976,
      "loss": 3.3397,
      "step": 54700
    },
    {
      "epoch": 5.892799483371004,
      "grad_norm": 0.7015940546989441,
      "learning_rate": 0.0002468333153754983,
      "loss": 3.334,
      "step": 54750
    },
    {
      "epoch": 5.898181035410612,
      "grad_norm": 0.701823353767395,
      "learning_rate": 0.0002465100743454369,
      "loss": 3.3449,
      "step": 54800
    },
    {
      "epoch": 5.903562587450221,
      "grad_norm": 0.684441328048706,
      "learning_rate": 0.0002461868333153755,
      "loss": 3.3369,
      "step": 54850
    },
    {
      "epoch": 5.9089441394898286,
      "grad_norm": 0.7343729734420776,
      "learning_rate": 0.0002458635922853141,
      "loss": 3.3461,
      "step": 54900
    },
    {
      "epoch": 5.914325691529437,
      "grad_norm": 0.7203894257545471,
      "learning_rate": 0.0002455403512552526,
      "loss": 3.3287,
      "step": 54950
    },
    {
      "epoch": 5.919707243569046,
      "grad_norm": 0.7090471982955933,
      "learning_rate": 0.0002452171102251912,
      "loss": 3.3374,
      "step": 55000
    },
    {
      "epoch": 5.919707243569046,
      "eval_accuracy": 0.38405818753292864,
      "eval_loss": 3.379855155944824,
      "eval_runtime": 184.4932,
      "eval_samples_per_second": 97.624,
      "eval_steps_per_second": 6.103,
      "step": 55000
    },
    {
      "epoch": 5.925088795608653,
      "grad_norm": 0.6990830898284912,
      "learning_rate": 0.0002448938691951298,
      "loss": 3.3292,
      "step": 55050
    },
    {
      "epoch": 5.930470347648262,
      "grad_norm": 0.7348403334617615,
      "learning_rate": 0.0002445706281650684,
      "loss": 3.3253,
      "step": 55100
    },
    {
      "epoch": 5.93585189968787,
      "grad_norm": 0.7289194464683533,
      "learning_rate": 0.000244247387135007,
      "loss": 3.3327,
      "step": 55150
    },
    {
      "epoch": 5.941233451727478,
      "grad_norm": 0.74270099401474,
      "learning_rate": 0.00024392414610494557,
      "loss": 3.3126,
      "step": 55200
    },
    {
      "epoch": 5.946615003767087,
      "grad_norm": 0.6974494457244873,
      "learning_rate": 0.00024360090507488414,
      "loss": 3.3297,
      "step": 55250
    },
    {
      "epoch": 5.951996555806694,
      "grad_norm": 0.7457931637763977,
      "learning_rate": 0.00024327766404482273,
      "loss": 3.3188,
      "step": 55300
    },
    {
      "epoch": 5.957378107846303,
      "grad_norm": 0.7141943573951721,
      "learning_rate": 0.0002429544230147613,
      "loss": 3.3607,
      "step": 55350
    },
    {
      "epoch": 5.962759659885911,
      "grad_norm": 0.7391859889030457,
      "learning_rate": 0.00024263118198469992,
      "loss": 3.3317,
      "step": 55400
    },
    {
      "epoch": 5.968141211925519,
      "grad_norm": 0.7539886832237244,
      "learning_rate": 0.0002423079409546385,
      "loss": 3.3524,
      "step": 55450
    },
    {
      "epoch": 5.973522763965128,
      "grad_norm": 0.764616072177887,
      "learning_rate": 0.00024198469992457706,
      "loss": 3.3379,
      "step": 55500
    },
    {
      "epoch": 5.978904316004736,
      "grad_norm": 0.7487912178039551,
      "learning_rate": 0.00024166145889451568,
      "loss": 3.3334,
      "step": 55550
    },
    {
      "epoch": 5.984285868044344,
      "grad_norm": 0.7643945217132568,
      "learning_rate": 0.00024133821786445425,
      "loss": 3.3554,
      "step": 55600
    },
    {
      "epoch": 5.989667420083952,
      "grad_norm": 0.726444661617279,
      "learning_rate": 0.0002410149768343928,
      "loss": 3.3291,
      "step": 55650
    },
    {
      "epoch": 5.995048972123561,
      "grad_norm": 0.7337467670440674,
      "learning_rate": 0.0002406917358043314,
      "loss": 3.3395,
      "step": 55700
    },
    {
      "epoch": 6.000430524163169,
      "grad_norm": 0.7354254722595215,
      "learning_rate": 0.00024036849477427,
      "loss": 3.3185,
      "step": 55750
    },
    {
      "epoch": 6.005812076202777,
      "grad_norm": 0.7437455058097839,
      "learning_rate": 0.00024004525374420857,
      "loss": 3.251,
      "step": 55800
    },
    {
      "epoch": 6.011193628242385,
      "grad_norm": 0.7114704251289368,
      "learning_rate": 0.00023972201271414716,
      "loss": 3.2531,
      "step": 55850
    },
    {
      "epoch": 6.016575180281993,
      "grad_norm": 0.7447876930236816,
      "learning_rate": 0.00023939877168408573,
      "loss": 3.2347,
      "step": 55900
    },
    {
      "epoch": 6.021956732321602,
      "grad_norm": 0.7433951497077942,
      "learning_rate": 0.00023907553065402433,
      "loss": 3.2466,
      "step": 55950
    },
    {
      "epoch": 6.0273382843612096,
      "grad_norm": 0.7271364331245422,
      "learning_rate": 0.00023875875444456413,
      "loss": 3.2475,
      "step": 56000
    },
    {
      "epoch": 6.0273382843612096,
      "eval_accuracy": 0.3842132352038399,
      "eval_loss": 3.383183002471924,
      "eval_runtime": 184.6903,
      "eval_samples_per_second": 97.52,
      "eval_steps_per_second": 6.097,
      "step": 56000
    },
    {
      "epoch": 6.032719836400818,
      "grad_norm": 0.672038733959198,
      "learning_rate": 0.0002384355134145027,
      "loss": 3.2535,
      "step": 56050
    },
    {
      "epoch": 6.038101388440427,
      "grad_norm": 0.7392109036445618,
      "learning_rate": 0.00023811227238444132,
      "loss": 3.23,
      "step": 56100
    },
    {
      "epoch": 6.043482940480034,
      "grad_norm": 0.7268337607383728,
      "learning_rate": 0.0002377890313543799,
      "loss": 3.2515,
      "step": 56150
    },
    {
      "epoch": 6.048864492519643,
      "grad_norm": 0.7550994753837585,
      "learning_rate": 0.00023746579032431849,
      "loss": 3.2548,
      "step": 56200
    },
    {
      "epoch": 6.0542460445592505,
      "grad_norm": 0.7194846868515015,
      "learning_rate": 0.00023714254929425708,
      "loss": 3.2536,
      "step": 56250
    },
    {
      "epoch": 6.059627596598859,
      "grad_norm": 0.7446109652519226,
      "learning_rate": 0.00023681930826419565,
      "loss": 3.2548,
      "step": 56300
    },
    {
      "epoch": 6.065009148638468,
      "grad_norm": 0.7766132354736328,
      "learning_rate": 0.00023649606723413424,
      "loss": 3.2696,
      "step": 56350
    },
    {
      "epoch": 6.070390700678075,
      "grad_norm": 0.7171474695205688,
      "learning_rate": 0.0002361728262040728,
      "loss": 3.2452,
      "step": 56400
    },
    {
      "epoch": 6.075772252717684,
      "grad_norm": 0.7210790514945984,
      "learning_rate": 0.00023584958517401138,
      "loss": 3.2505,
      "step": 56450
    },
    {
      "epoch": 6.081153804757292,
      "grad_norm": 0.743676483631134,
      "learning_rate": 0.00023552634414395,
      "loss": 3.2716,
      "step": 56500
    },
    {
      "epoch": 6.0865353567969,
      "grad_norm": 0.7467906475067139,
      "learning_rate": 0.00023520310311388857,
      "loss": 3.2724,
      "step": 56550
    },
    {
      "epoch": 6.091916908836509,
      "grad_norm": 0.8049376010894775,
      "learning_rate": 0.00023487986208382713,
      "loss": 3.2701,
      "step": 56600
    },
    {
      "epoch": 6.097298460876116,
      "grad_norm": 0.7321614623069763,
      "learning_rate": 0.00023455662105376576,
      "loss": 3.2599,
      "step": 56650
    },
    {
      "epoch": 6.102680012915725,
      "grad_norm": 0.779611349105835,
      "learning_rate": 0.00023423338002370432,
      "loss": 3.2633,
      "step": 56700
    },
    {
      "epoch": 6.108061564955333,
      "grad_norm": 0.7684844136238098,
      "learning_rate": 0.0002339101389936429,
      "loss": 3.2466,
      "step": 56750
    },
    {
      "epoch": 6.113443116994941,
      "grad_norm": 0.7690160870552063,
      "learning_rate": 0.00023358689796358149,
      "loss": 3.2413,
      "step": 56800
    },
    {
      "epoch": 6.11882466903455,
      "grad_norm": 0.7911422252655029,
      "learning_rate": 0.00023326365693352008,
      "loss": 3.2716,
      "step": 56850
    },
    {
      "epoch": 6.124206221074158,
      "grad_norm": 0.7434123754501343,
      "learning_rate": 0.00023294041590345865,
      "loss": 3.2745,
      "step": 56900
    },
    {
      "epoch": 6.129587773113766,
      "grad_norm": 0.7318158149719238,
      "learning_rate": 0.00023261717487339724,
      "loss": 3.2717,
      "step": 56950
    },
    {
      "epoch": 6.134969325153374,
      "grad_norm": 0.748771071434021,
      "learning_rate": 0.0002322939338433358,
      "loss": 3.2754,
      "step": 57000
    },
    {
      "epoch": 6.134969325153374,
      "eval_accuracy": 0.3840938256801318,
      "eval_loss": 3.383234739303589,
      "eval_runtime": 184.5035,
      "eval_samples_per_second": 97.619,
      "eval_steps_per_second": 6.103,
      "step": 57000
    },
    {
      "epoch": 6.140350877192983,
      "grad_norm": 0.7642286419868469,
      "learning_rate": 0.0002319706928132744,
      "loss": 3.2569,
      "step": 57050
    },
    {
      "epoch": 6.1457324292325906,
      "grad_norm": 0.7967156767845154,
      "learning_rate": 0.000231647451783213,
      "loss": 3.272,
      "step": 57100
    },
    {
      "epoch": 6.151113981272199,
      "grad_norm": 0.7421084046363831,
      "learning_rate": 0.00023132421075315157,
      "loss": 3.2694,
      "step": 57150
    },
    {
      "epoch": 6.156495533311807,
      "grad_norm": 0.7525152564048767,
      "learning_rate": 0.0002310009697230902,
      "loss": 3.2596,
      "step": 57200
    },
    {
      "epoch": 6.161877085351415,
      "grad_norm": 0.6975091695785522,
      "learning_rate": 0.00023067772869302876,
      "loss": 3.2706,
      "step": 57250
    },
    {
      "epoch": 6.167258637391024,
      "grad_norm": 0.7624300122261047,
      "learning_rate": 0.00023035448766296732,
      "loss": 3.2747,
      "step": 57300
    },
    {
      "epoch": 6.1726401894306315,
      "grad_norm": 0.7489650845527649,
      "learning_rate": 0.00023003124663290592,
      "loss": 3.2598,
      "step": 57350
    },
    {
      "epoch": 6.17802174147024,
      "grad_norm": 0.747357189655304,
      "learning_rate": 0.0002297080056028445,
      "loss": 3.2593,
      "step": 57400
    },
    {
      "epoch": 6.183403293509849,
      "grad_norm": 0.7370163202285767,
      "learning_rate": 0.00022938476457278308,
      "loss": 3.2656,
      "step": 57450
    },
    {
      "epoch": 6.188784845549456,
      "grad_norm": 0.7547042369842529,
      "learning_rate": 0.00022906152354272168,
      "loss": 3.2699,
      "step": 57500
    },
    {
      "epoch": 6.194166397589065,
      "grad_norm": 0.7671395540237427,
      "learning_rate": 0.00022873828251266024,
      "loss": 3.2792,
      "step": 57550
    },
    {
      "epoch": 6.1995479496286725,
      "grad_norm": 0.7778907418251038,
      "learning_rate": 0.00022841504148259884,
      "loss": 3.2836,
      "step": 57600
    },
    {
      "epoch": 6.204929501668281,
      "grad_norm": 0.7617507576942444,
      "learning_rate": 0.00022809180045253743,
      "loss": 3.2632,
      "step": 57650
    },
    {
      "epoch": 6.21031105370789,
      "grad_norm": 0.8062065839767456,
      "learning_rate": 0.000227768559422476,
      "loss": 3.2826,
      "step": 57700
    },
    {
      "epoch": 6.215692605747497,
      "grad_norm": 0.7325953841209412,
      "learning_rate": 0.00022744531839241457,
      "loss": 3.2799,
      "step": 57750
    },
    {
      "epoch": 6.221074157787106,
      "grad_norm": 0.7609289288520813,
      "learning_rate": 0.0002271220773623532,
      "loss": 3.2695,
      "step": 57800
    },
    {
      "epoch": 6.226455709826714,
      "grad_norm": 0.8026849031448364,
      "learning_rate": 0.00022679883633229176,
      "loss": 3.27,
      "step": 57850
    },
    {
      "epoch": 6.231837261866322,
      "grad_norm": 0.7824159860610962,
      "learning_rate": 0.00022647559530223032,
      "loss": 3.2942,
      "step": 57900
    },
    {
      "epoch": 6.237218813905931,
      "grad_norm": 0.7842664122581482,
      "learning_rate": 0.00022615235427216895,
      "loss": 3.2698,
      "step": 57950
    },
    {
      "epoch": 6.242600365945538,
      "grad_norm": 0.8548306226730347,
      "learning_rate": 0.0002258291132421075,
      "loss": 3.2745,
      "step": 58000
    },
    {
      "epoch": 6.242600365945538,
      "eval_accuracy": 0.384586240567707,
      "eval_loss": 3.378992795944214,
      "eval_runtime": 184.88,
      "eval_samples_per_second": 97.42,
      "eval_steps_per_second": 6.09,
      "step": 58000
    },
    {
      "epoch": 6.247981917985147,
      "grad_norm": 0.744491696357727,
      "learning_rate": 0.0002255058722120461,
      "loss": 3.2725,
      "step": 58050
    },
    {
      "epoch": 6.253363470024755,
      "grad_norm": 0.7470669150352478,
      "learning_rate": 0.00022518263118198468,
      "loss": 3.2718,
      "step": 58100
    },
    {
      "epoch": 6.258745022064363,
      "grad_norm": 0.7122950553894043,
      "learning_rate": 0.00022485939015192327,
      "loss": 3.2667,
      "step": 58150
    },
    {
      "epoch": 6.264126574103972,
      "grad_norm": 0.7258158922195435,
      "learning_rate": 0.00022453614912186186,
      "loss": 3.265,
      "step": 58200
    },
    {
      "epoch": 6.26950812614358,
      "grad_norm": 0.7712364196777344,
      "learning_rate": 0.00022421290809180043,
      "loss": 3.2725,
      "step": 58250
    },
    {
      "epoch": 6.274889678183188,
      "grad_norm": 0.7405200600624084,
      "learning_rate": 0.000223889667061739,
      "loss": 3.2729,
      "step": 58300
    },
    {
      "epoch": 6.280271230222796,
      "grad_norm": 0.7869605422019958,
      "learning_rate": 0.00022356642603167762,
      "loss": 3.2867,
      "step": 58350
    },
    {
      "epoch": 6.285652782262405,
      "grad_norm": 0.7676852345466614,
      "learning_rate": 0.0002232431850016162,
      "loss": 3.2751,
      "step": 58400
    },
    {
      "epoch": 6.2910343343020125,
      "grad_norm": 0.8062360286712646,
      "learning_rate": 0.00022291994397155476,
      "loss": 3.2947,
      "step": 58450
    },
    {
      "epoch": 6.296415886341621,
      "grad_norm": 0.7730180621147156,
      "learning_rate": 0.00022259670294149338,
      "loss": 3.2726,
      "step": 58500
    },
    {
      "epoch": 6.301797438381229,
      "grad_norm": 0.7443013787269592,
      "learning_rate": 0.00022227346191143195,
      "loss": 3.2728,
      "step": 58550
    },
    {
      "epoch": 6.307178990420837,
      "grad_norm": 0.7263332605361938,
      "learning_rate": 0.00022195022088137051,
      "loss": 3.2891,
      "step": 58600
    },
    {
      "epoch": 6.312560542460446,
      "grad_norm": 0.730728268623352,
      "learning_rate": 0.0002216269798513091,
      "loss": 3.2826,
      "step": 58650
    },
    {
      "epoch": 6.3179420945000535,
      "grad_norm": 0.7482548952102661,
      "learning_rate": 0.00022130373882124768,
      "loss": 3.276,
      "step": 58700
    },
    {
      "epoch": 6.323323646539662,
      "grad_norm": 0.7771552205085754,
      "learning_rate": 0.00022098049779118627,
      "loss": 3.2764,
      "step": 58750
    },
    {
      "epoch": 6.328705198579271,
      "grad_norm": 0.7530174851417542,
      "learning_rate": 0.00022065725676112487,
      "loss": 3.2929,
      "step": 58800
    },
    {
      "epoch": 6.334086750618878,
      "grad_norm": 0.7332282066345215,
      "learning_rate": 0.00022033401573106343,
      "loss": 3.2768,
      "step": 58850
    },
    {
      "epoch": 6.339468302658487,
      "grad_norm": 0.8075442314147949,
      "learning_rate": 0.000220010774701002,
      "loss": 3.2745,
      "step": 58900
    },
    {
      "epoch": 6.344849854698095,
      "grad_norm": 0.7218873500823975,
      "learning_rate": 0.00021968753367094062,
      "loss": 3.2761,
      "step": 58950
    },
    {
      "epoch": 6.350231406737703,
      "grad_norm": 0.733670711517334,
      "learning_rate": 0.0002193642926408792,
      "loss": 3.2611,
      "step": 59000
    },
    {
      "epoch": 6.350231406737703,
      "eval_accuracy": 0.3853270359568261,
      "eval_loss": 3.375654935836792,
      "eval_runtime": 184.5756,
      "eval_samples_per_second": 97.581,
      "eval_steps_per_second": 6.1,
      "step": 59000
    },
    {
      "epoch": 6.355612958777312,
      "grad_norm": 0.7387909293174744,
      "learning_rate": 0.00021904105161081778,
      "loss": 3.258,
      "step": 59050
    },
    {
      "epoch": 6.360994510816919,
      "grad_norm": 0.869426965713501,
      "learning_rate": 0.00021871781058075638,
      "loss": 3.2898,
      "step": 59100
    },
    {
      "epoch": 6.366376062856528,
      "grad_norm": 0.7325130701065063,
      "learning_rate": 0.00021839456955069495,
      "loss": 3.2768,
      "step": 59150
    },
    {
      "epoch": 6.371757614896136,
      "grad_norm": 0.7691362500190735,
      "learning_rate": 0.00021807132852063354,
      "loss": 3.2778,
      "step": 59200
    },
    {
      "epoch": 6.377139166935744,
      "grad_norm": 0.7573683857917786,
      "learning_rate": 0.0002177480874905721,
      "loss": 3.2984,
      "step": 59250
    },
    {
      "epoch": 6.382520718975353,
      "grad_norm": 0.7838876843452454,
      "learning_rate": 0.0002174248464605107,
      "loss": 3.2733,
      "step": 59300
    },
    {
      "epoch": 6.387902271014961,
      "grad_norm": 0.793933093547821,
      "learning_rate": 0.0002171016054304493,
      "loss": 3.2801,
      "step": 59350
    },
    {
      "epoch": 6.393283823054569,
      "grad_norm": 0.7980042099952698,
      "learning_rate": 0.00021677836440038787,
      "loss": 3.2682,
      "step": 59400
    },
    {
      "epoch": 6.398665375094177,
      "grad_norm": 0.7875555753707886,
      "learning_rate": 0.0002164615881909277,
      "loss": 3.2946,
      "step": 59450
    },
    {
      "epoch": 6.404046927133785,
      "grad_norm": 0.76567542552948,
      "learning_rate": 0.00021613834716086627,
      "loss": 3.2778,
      "step": 59500
    },
    {
      "epoch": 6.4094284791733935,
      "grad_norm": 0.7699328660964966,
      "learning_rate": 0.00021581510613080483,
      "loss": 3.3058,
      "step": 59550
    },
    {
      "epoch": 6.414810031213002,
      "grad_norm": 0.7370243668556213,
      "learning_rate": 0.00021549186510074346,
      "loss": 3.275,
      "step": 59600
    },
    {
      "epoch": 6.42019158325261,
      "grad_norm": 0.7352474927902222,
      "learning_rate": 0.00021516862407068202,
      "loss": 3.2637,
      "step": 59650
    },
    {
      "epoch": 6.425573135292218,
      "grad_norm": 0.7487669587135315,
      "learning_rate": 0.0002148453830406206,
      "loss": 3.263,
      "step": 59700
    },
    {
      "epoch": 6.430954687331827,
      "grad_norm": 0.8222890496253967,
      "learning_rate": 0.00021452214201055919,
      "loss": 3.2685,
      "step": 59750
    },
    {
      "epoch": 6.4363362393714345,
      "grad_norm": 0.7670249342918396,
      "learning_rate": 0.00021419890098049778,
      "loss": 3.2656,
      "step": 59800
    },
    {
      "epoch": 6.441717791411043,
      "grad_norm": 0.7992569208145142,
      "learning_rate": 0.00021387565995043638,
      "loss": 3.2902,
      "step": 59850
    },
    {
      "epoch": 6.447099343450651,
      "grad_norm": 0.7975773215293884,
      "learning_rate": 0.00021355241892037494,
      "loss": 3.2832,
      "step": 59900
    },
    {
      "epoch": 6.452480895490259,
      "grad_norm": 0.7684232592582703,
      "learning_rate": 0.0002132291778903135,
      "loss": 3.2945,
      "step": 59950
    },
    {
      "epoch": 6.457862447529868,
      "grad_norm": 0.7654765844345093,
      "learning_rate": 0.00021290593686025213,
      "loss": 3.2576,
      "step": 60000
    },
    {
      "epoch": 6.457862447529868,
      "eval_accuracy": 0.3856219198943546,
      "eval_loss": 3.3727223873138428,
      "eval_runtime": 184.4472,
      "eval_samples_per_second": 97.649,
      "eval_steps_per_second": 6.105,
      "step": 60000
    },
    {
      "epoch": 6.4632439995694755,
      "grad_norm": 0.7334644794464111,
      "learning_rate": 0.0002125826958301907,
      "loss": 3.2746,
      "step": 60050
    },
    {
      "epoch": 6.468625551609084,
      "grad_norm": 0.7792539000511169,
      "learning_rate": 0.00021225945480012927,
      "loss": 3.2693,
      "step": 60100
    },
    {
      "epoch": 6.474007103648693,
      "grad_norm": 0.7655577659606934,
      "learning_rate": 0.00021193621377006786,
      "loss": 3.2839,
      "step": 60150
    },
    {
      "epoch": 6.4793886556883,
      "grad_norm": 0.7592105865478516,
      "learning_rate": 0.00021161297274000646,
      "loss": 3.2704,
      "step": 60200
    },
    {
      "epoch": 6.484770207727909,
      "grad_norm": 0.815912127494812,
      "learning_rate": 0.00021128973170994502,
      "loss": 3.2894,
      "step": 60250
    },
    {
      "epoch": 6.490151759767517,
      "grad_norm": 0.742038905620575,
      "learning_rate": 0.00021096649067988362,
      "loss": 3.2778,
      "step": 60300
    },
    {
      "epoch": 6.495533311807125,
      "grad_norm": 0.7956986427307129,
      "learning_rate": 0.00021064324964982219,
      "loss": 3.2668,
      "step": 60350
    },
    {
      "epoch": 6.500914863846734,
      "grad_norm": 0.7016414403915405,
      "learning_rate": 0.00021032000861976078,
      "loss": 3.2708,
      "step": 60400
    },
    {
      "epoch": 6.506296415886341,
      "grad_norm": 0.7408983707427979,
      "learning_rate": 0.00020999676758969938,
      "loss": 3.2812,
      "step": 60450
    },
    {
      "epoch": 6.51167796792595,
      "grad_norm": 0.7210450768470764,
      "learning_rate": 0.00020967352655963794,
      "loss": 3.2736,
      "step": 60500
    },
    {
      "epoch": 6.517059519965558,
      "grad_norm": 0.7317191958427429,
      "learning_rate": 0.0002093502855295765,
      "loss": 3.2999,
      "step": 60550
    },
    {
      "epoch": 6.522441072005166,
      "grad_norm": 0.7658736705780029,
      "learning_rate": 0.00020902704449951513,
      "loss": 3.3033,
      "step": 60600
    },
    {
      "epoch": 6.5278226240447745,
      "grad_norm": 0.7588192820549011,
      "learning_rate": 0.0002087038034694537,
      "loss": 3.2845,
      "step": 60650
    },
    {
      "epoch": 6.533204176084383,
      "grad_norm": 0.7971400022506714,
      "learning_rate": 0.00020838056243939227,
      "loss": 3.284,
      "step": 60700
    },
    {
      "epoch": 6.538585728123991,
      "grad_norm": 0.7562872171401978,
      "learning_rate": 0.0002080573214093309,
      "loss": 3.2741,
      "step": 60750
    },
    {
      "epoch": 6.543967280163599,
      "grad_norm": 0.7392501831054688,
      "learning_rate": 0.00020773408037926946,
      "loss": 3.2865,
      "step": 60800
    },
    {
      "epoch": 6.549348832203208,
      "grad_norm": 0.7754622101783752,
      "learning_rate": 0.00020741083934920805,
      "loss": 3.2653,
      "step": 60850
    },
    {
      "epoch": 6.5547303842428155,
      "grad_norm": 0.7140413522720337,
      "learning_rate": 0.00020708759831914662,
      "loss": 3.2833,
      "step": 60900
    },
    {
      "epoch": 6.560111936282424,
      "grad_norm": 0.7713199257850647,
      "learning_rate": 0.00020676435728908521,
      "loss": 3.2903,
      "step": 60950
    },
    {
      "epoch": 6.565493488322032,
      "grad_norm": 0.7456794381141663,
      "learning_rate": 0.0002064411162590238,
      "loss": 3.2939,
      "step": 61000
    },
    {
      "epoch": 6.565493488322032,
      "eval_accuracy": 0.38606848326327214,
      "eval_loss": 3.368579626083374,
      "eval_runtime": 184.7639,
      "eval_samples_per_second": 97.481,
      "eval_steps_per_second": 6.094,
      "step": 61000
    },
    {
      "epoch": 6.57087504036164,
      "grad_norm": 0.7683389186859131,
      "learning_rate": 0.00020611787522896238,
      "loss": 3.2849,
      "step": 61050
    },
    {
      "epoch": 6.576256592401249,
      "grad_norm": 0.8046767711639404,
      "learning_rate": 0.00020579463419890094,
      "loss": 3.3024,
      "step": 61100
    },
    {
      "epoch": 6.5816381444408565,
      "grad_norm": 0.7584671378135681,
      "learning_rate": 0.00020547139316883957,
      "loss": 3.2972,
      "step": 61150
    },
    {
      "epoch": 6.587019696480465,
      "grad_norm": 0.7123998999595642,
      "learning_rate": 0.00020514815213877813,
      "loss": 3.2849,
      "step": 61200
    },
    {
      "epoch": 6.592401248520073,
      "grad_norm": 0.7630634307861328,
      "learning_rate": 0.0002048249111087167,
      "loss": 3.2829,
      "step": 61250
    },
    {
      "epoch": 6.597782800559681,
      "grad_norm": 0.7825208902359009,
      "learning_rate": 0.00020450167007865532,
      "loss": 3.2779,
      "step": 61300
    },
    {
      "epoch": 6.60316435259929,
      "grad_norm": 0.822171688079834,
      "learning_rate": 0.0002041784290485939,
      "loss": 3.2678,
      "step": 61350
    },
    {
      "epoch": 6.608545904638898,
      "grad_norm": 0.745606005191803,
      "learning_rate": 0.00020385518801853246,
      "loss": 3.2759,
      "step": 61400
    },
    {
      "epoch": 6.613927456678506,
      "grad_norm": 0.7718546986579895,
      "learning_rate": 0.00020353194698847105,
      "loss": 3.2836,
      "step": 61450
    },
    {
      "epoch": 6.619309008718115,
      "grad_norm": 0.7141976952552795,
      "learning_rate": 0.00020320870595840965,
      "loss": 3.2647,
      "step": 61500
    },
    {
      "epoch": 6.624690560757722,
      "grad_norm": 0.7380606532096863,
      "learning_rate": 0.00020288546492834821,
      "loss": 3.2891,
      "step": 61550
    },
    {
      "epoch": 6.630072112797331,
      "grad_norm": 0.7463494539260864,
      "learning_rate": 0.0002025622238982868,
      "loss": 3.2688,
      "step": 61600
    },
    {
      "epoch": 6.635453664836939,
      "grad_norm": 0.76692795753479,
      "learning_rate": 0.00020223898286822538,
      "loss": 3.2883,
      "step": 61650
    },
    {
      "epoch": 6.640835216876547,
      "grad_norm": 0.8029922246932983,
      "learning_rate": 0.00020191574183816397,
      "loss": 3.2836,
      "step": 61700
    },
    {
      "epoch": 6.6462167689161555,
      "grad_norm": 0.728897213935852,
      "learning_rate": 0.00020159250080810257,
      "loss": 3.2701,
      "step": 61750
    },
    {
      "epoch": 6.651598320955763,
      "grad_norm": 0.7733128070831299,
      "learning_rate": 0.00020126925977804113,
      "loss": 3.2767,
      "step": 61800
    },
    {
      "epoch": 6.656979872995372,
      "grad_norm": 0.7983481884002686,
      "learning_rate": 0.00020095248356858097,
      "loss": 3.2963,
      "step": 61850
    },
    {
      "epoch": 6.66236142503498,
      "grad_norm": 0.8184331059455872,
      "learning_rate": 0.00020062924253851953,
      "loss": 3.2907,
      "step": 61900
    },
    {
      "epoch": 6.667742977074588,
      "grad_norm": 0.7854340076446533,
      "learning_rate": 0.00020030600150845813,
      "loss": 3.2837,
      "step": 61950
    },
    {
      "epoch": 6.6731245291141965,
      "grad_norm": 0.843238353729248,
      "learning_rate": 0.0001999827604783967,
      "loss": 3.2781,
      "step": 62000
    },
    {
      "epoch": 6.6731245291141965,
      "eval_accuracy": 0.3863260992602803,
      "eval_loss": 3.364689350128174,
      "eval_runtime": 184.8366,
      "eval_samples_per_second": 97.443,
      "eval_steps_per_second": 6.092,
      "step": 62000
    },
    {
      "epoch": 6.678506081153805,
      "grad_norm": 0.7781000733375549,
      "learning_rate": 0.0001996595194483353,
      "loss": 3.2809,
      "step": 62050
    },
    {
      "epoch": 6.683887633193413,
      "grad_norm": 0.8087568879127502,
      "learning_rate": 0.00019933627841827389,
      "loss": 3.2962,
      "step": 62100
    },
    {
      "epoch": 6.689269185233021,
      "grad_norm": 0.7818818092346191,
      "learning_rate": 0.00019901303738821245,
      "loss": 3.3061,
      "step": 62150
    },
    {
      "epoch": 6.69465073727263,
      "grad_norm": 0.7550804615020752,
      "learning_rate": 0.00019868979635815102,
      "loss": 3.3017,
      "step": 62200
    },
    {
      "epoch": 6.7000322893122375,
      "grad_norm": 0.736332356929779,
      "learning_rate": 0.00019836655532808964,
      "loss": 3.2831,
      "step": 62250
    },
    {
      "epoch": 6.705413841351846,
      "grad_norm": 0.7578887343406677,
      "learning_rate": 0.0001980433142980282,
      "loss": 3.2868,
      "step": 62300
    },
    {
      "epoch": 6.710795393391454,
      "grad_norm": 0.751354992389679,
      "learning_rate": 0.00019772007326796678,
      "loss": 3.2909,
      "step": 62350
    },
    {
      "epoch": 6.716176945431062,
      "grad_norm": 0.7875543236732483,
      "learning_rate": 0.0001973968322379054,
      "loss": 3.2875,
      "step": 62400
    },
    {
      "epoch": 6.721558497470671,
      "grad_norm": 0.7486498355865479,
      "learning_rate": 0.00019707359120784397,
      "loss": 3.2804,
      "step": 62450
    },
    {
      "epoch": 6.7269400495102785,
      "grad_norm": 0.790752112865448,
      "learning_rate": 0.00019675035017778253,
      "loss": 3.3072,
      "step": 62500
    },
    {
      "epoch": 6.732321601549887,
      "grad_norm": 0.7708467841148376,
      "learning_rate": 0.00019643357396832237,
      "loss": 3.2902,
      "step": 62550
    },
    {
      "epoch": 6.737703153589496,
      "grad_norm": 0.7900062799453735,
      "learning_rate": 0.00019611033293826096,
      "loss": 3.281,
      "step": 62600
    },
    {
      "epoch": 6.743084705629103,
      "grad_norm": 0.7324670553207397,
      "learning_rate": 0.00019578709190819953,
      "loss": 3.2743,
      "step": 62650
    },
    {
      "epoch": 6.748466257668712,
      "grad_norm": 0.7415837645530701,
      "learning_rate": 0.0001954638508781381,
      "loss": 3.3112,
      "step": 62700
    },
    {
      "epoch": 6.75384780970832,
      "grad_norm": 0.7605777382850647,
      "learning_rate": 0.00019514060984807672,
      "loss": 3.2669,
      "step": 62750
    },
    {
      "epoch": 6.759229361747928,
      "grad_norm": 0.7561804056167603,
      "learning_rate": 0.0001948173688180153,
      "loss": 3.2779,
      "step": 62800
    },
    {
      "epoch": 6.7646109137875365,
      "grad_norm": 0.799517810344696,
      "learning_rate": 0.00019449412778795386,
      "loss": 3.2907,
      "step": 62850
    },
    {
      "epoch": 6.769992465827144,
      "grad_norm": 0.9140427112579346,
      "learning_rate": 0.00019417088675789245,
      "loss": 3.2925,
      "step": 62900
    },
    {
      "epoch": 6.775374017866753,
      "grad_norm": 0.7958894968032837,
      "learning_rate": 0.00019384764572783104,
      "loss": 3.2867,
      "step": 62950
    },
    {
      "epoch": 6.780755569906361,
      "grad_norm": 0.7784082293510437,
      "learning_rate": 0.0001935244046977696,
      "loss": 3.2786,
      "step": 63000
    },
    {
      "epoch": 6.780755569906361,
      "eval_accuracy": 0.38668899990558064,
      "eval_loss": 3.360412836074829,
      "eval_runtime": 185.3645,
      "eval_samples_per_second": 97.165,
      "eval_steps_per_second": 6.075,
      "step": 63000
    },
    {
      "epoch": 6.786137121945969,
      "grad_norm": 0.7796688675880432,
      "learning_rate": 0.0001932011636677082,
      "loss": 3.3014,
      "step": 63050
    },
    {
      "epoch": 6.7915186739855775,
      "grad_norm": 0.8077431917190552,
      "learning_rate": 0.00019287792263764677,
      "loss": 3.2886,
      "step": 63100
    },
    {
      "epoch": 6.796900226025185,
      "grad_norm": 0.7771782875061035,
      "learning_rate": 0.00019255468160758537,
      "loss": 3.2901,
      "step": 63150
    },
    {
      "epoch": 6.802281778064794,
      "grad_norm": 0.8028762340545654,
      "learning_rate": 0.00019223144057752396,
      "loss": 3.2833,
      "step": 63200
    },
    {
      "epoch": 6.807663330104402,
      "grad_norm": 0.7487192749977112,
      "learning_rate": 0.00019190819954746253,
      "loss": 3.2754,
      "step": 63250
    },
    {
      "epoch": 6.813044882144011,
      "grad_norm": 0.7907421588897705,
      "learning_rate": 0.0001915849585174011,
      "loss": 3.2896,
      "step": 63300
    },
    {
      "epoch": 6.8184264341836185,
      "grad_norm": 0.7865395545959473,
      "learning_rate": 0.00019126171748733972,
      "loss": 3.2825,
      "step": 63350
    },
    {
      "epoch": 6.823807986223227,
      "grad_norm": 0.8095396757125854,
      "learning_rate": 0.0001909384764572783,
      "loss": 3.2916,
      "step": 63400
    },
    {
      "epoch": 6.829189538262835,
      "grad_norm": 0.812186598777771,
      "learning_rate": 0.00019061523542721688,
      "loss": 3.2914,
      "step": 63450
    },
    {
      "epoch": 6.834571090302443,
      "grad_norm": 0.767548680305481,
      "learning_rate": 0.00019029199439715548,
      "loss": 3.2989,
      "step": 63500
    },
    {
      "epoch": 6.839952642342052,
      "grad_norm": 0.9850395917892456,
      "learning_rate": 0.00018996875336709404,
      "loss": 3.2761,
      "step": 63550
    },
    {
      "epoch": 6.8453341943816595,
      "grad_norm": 0.8007294535636902,
      "learning_rate": 0.00018964551233703264,
      "loss": 3.2923,
      "step": 63600
    },
    {
      "epoch": 6.850715746421268,
      "grad_norm": 0.7697404026985168,
      "learning_rate": 0.0001893222713069712,
      "loss": 3.2658,
      "step": 63650
    },
    {
      "epoch": 6.856097298460876,
      "grad_norm": 0.775759220123291,
      "learning_rate": 0.0001889990302769098,
      "loss": 3.2943,
      "step": 63700
    },
    {
      "epoch": 6.861478850500484,
      "grad_norm": 0.7596869468688965,
      "learning_rate": 0.0001886757892468484,
      "loss": 3.2866,
      "step": 63750
    },
    {
      "epoch": 6.866860402540093,
      "grad_norm": 0.7999332547187805,
      "learning_rate": 0.00018835254821678696,
      "loss": 3.2685,
      "step": 63800
    },
    {
      "epoch": 6.8722419545797,
      "grad_norm": 0.8248339891433716,
      "learning_rate": 0.00018802930718672553,
      "loss": 3.298,
      "step": 63850
    },
    {
      "epoch": 6.877623506619309,
      "grad_norm": 0.7721416354179382,
      "learning_rate": 0.00018770606615666415,
      "loss": 3.2776,
      "step": 63900
    },
    {
      "epoch": 6.8830050586589175,
      "grad_norm": 0.8377927541732788,
      "learning_rate": 0.00018738282512660272,
      "loss": 3.2831,
      "step": 63950
    },
    {
      "epoch": 6.888386610698525,
      "grad_norm": 0.8438067436218262,
      "learning_rate": 0.0001870595840965413,
      "loss": 3.2927,
      "step": 64000
    },
    {
      "epoch": 6.888386610698525,
      "eval_accuracy": 0.3870464679064903,
      "eval_loss": 3.357008457183838,
      "eval_runtime": 185.2485,
      "eval_samples_per_second": 97.226,
      "eval_steps_per_second": 6.078,
      "step": 64000
    },
    {
      "epoch": 6.893768162738134,
      "grad_norm": 0.7458718419075012,
      "learning_rate": 0.0001867363430664799,
      "loss": 3.289,
      "step": 64050
    },
    {
      "epoch": 6.899149714777742,
      "grad_norm": 0.8297790288925171,
      "learning_rate": 0.00018641310203641848,
      "loss": 3.2955,
      "step": 64100
    },
    {
      "epoch": 6.90453126681735,
      "grad_norm": 0.7854870557785034,
      "learning_rate": 0.00018608986100635705,
      "loss": 3.2603,
      "step": 64150
    },
    {
      "epoch": 6.9099128188569585,
      "grad_norm": 0.7834721207618713,
      "learning_rate": 0.00018576661997629564,
      "loss": 3.2736,
      "step": 64200
    },
    {
      "epoch": 6.915294370896566,
      "grad_norm": 0.7471182346343994,
      "learning_rate": 0.00018544337894623423,
      "loss": 3.2952,
      "step": 64250
    },
    {
      "epoch": 6.920675922936175,
      "grad_norm": 0.7817671895027161,
      "learning_rate": 0.0001851201379161728,
      "loss": 3.2912,
      "step": 64300
    },
    {
      "epoch": 6.926057474975783,
      "grad_norm": 0.7854493856430054,
      "learning_rate": 0.0001847968968861114,
      "loss": 3.2829,
      "step": 64350
    },
    {
      "epoch": 6.931439027015391,
      "grad_norm": 0.7644644379615784,
      "learning_rate": 0.00018447365585604996,
      "loss": 3.283,
      "step": 64400
    },
    {
      "epoch": 6.9368205790549995,
      "grad_norm": 0.79556804895401,
      "learning_rate": 0.00018415041482598859,
      "loss": 3.2803,
      "step": 64450
    },
    {
      "epoch": 6.942202131094608,
      "grad_norm": 0.8365612030029297,
      "learning_rate": 0.00018382717379592715,
      "loss": 3.279,
      "step": 64500
    },
    {
      "epoch": 6.947583683134216,
      "grad_norm": 0.8554062843322754,
      "learning_rate": 0.00018350393276586572,
      "loss": 3.2721,
      "step": 64550
    },
    {
      "epoch": 6.952965235173824,
      "grad_norm": 0.8001241683959961,
      "learning_rate": 0.00018318069173580434,
      "loss": 3.2768,
      "step": 64600
    },
    {
      "epoch": 6.958346787213433,
      "grad_norm": 0.7490769624710083,
      "learning_rate": 0.0001828574507057429,
      "loss": 3.2979,
      "step": 64650
    },
    {
      "epoch": 6.9637283392530405,
      "grad_norm": 0.809732973575592,
      "learning_rate": 0.00018253420967568148,
      "loss": 3.2905,
      "step": 64700
    },
    {
      "epoch": 6.969109891292649,
      "grad_norm": 0.7997918725013733,
      "learning_rate": 0.00018221096864562007,
      "loss": 3.288,
      "step": 64750
    },
    {
      "epoch": 6.974491443332257,
      "grad_norm": 0.7825332880020142,
      "learning_rate": 0.00018188772761555867,
      "loss": 3.2976,
      "step": 64800
    },
    {
      "epoch": 6.979872995371865,
      "grad_norm": 0.7849071621894836,
      "learning_rate": 0.00018156448658549723,
      "loss": 3.2743,
      "step": 64850
    },
    {
      "epoch": 6.985254547411474,
      "grad_norm": 0.7942416667938232,
      "learning_rate": 0.00018124124555543583,
      "loss": 3.279,
      "step": 64900
    },
    {
      "epoch": 6.990636099451081,
      "grad_norm": 0.8016338348388672,
      "learning_rate": 0.0001809180045253744,
      "loss": 3.3058,
      "step": 64950
    },
    {
      "epoch": 6.99601765149069,
      "grad_norm": 1.0467689037322998,
      "learning_rate": 0.00018059476349531296,
      "loss": 3.2897,
      "step": 65000
    },
    {
      "epoch": 6.99601765149069,
      "eval_accuracy": 0.3873104944238795,
      "eval_loss": 3.3523385524749756,
      "eval_runtime": 185.1152,
      "eval_samples_per_second": 97.296,
      "eval_steps_per_second": 6.083,
      "step": 65000
    },
    {
      "epoch": 7.0013992035302985,
      "grad_norm": 0.8019529581069946,
      "learning_rate": 0.00018027152246525159,
      "loss": 3.2621,
      "step": 65050
    },
    {
      "epoch": 7.006780755569906,
      "grad_norm": 0.7568375468254089,
      "learning_rate": 0.00017994828143519015,
      "loss": 3.1794,
      "step": 65100
    },
    {
      "epoch": 7.012162307609515,
      "grad_norm": 0.7430813908576965,
      "learning_rate": 0.00017962504040512872,
      "loss": 3.1624,
      "step": 65150
    },
    {
      "epoch": 7.017543859649122,
      "grad_norm": 0.772469699382782,
      "learning_rate": 0.00017930179937506734,
      "loss": 3.2228,
      "step": 65200
    },
    {
      "epoch": 7.022925411688731,
      "grad_norm": 0.8518636226654053,
      "learning_rate": 0.0001789785583450059,
      "loss": 3.1921,
      "step": 65250
    },
    {
      "epoch": 7.0283069637283395,
      "grad_norm": 0.791405200958252,
      "learning_rate": 0.00017865531731494448,
      "loss": 3.1906,
      "step": 65300
    },
    {
      "epoch": 7.033688515767947,
      "grad_norm": 0.8231589198112488,
      "learning_rate": 0.00017833207628488307,
      "loss": 3.1997,
      "step": 65350
    },
    {
      "epoch": 7.039070067807556,
      "grad_norm": 0.8060066103935242,
      "learning_rate": 0.00017800883525482167,
      "loss": 3.2046,
      "step": 65400
    },
    {
      "epoch": 7.044451619847164,
      "grad_norm": 0.7832460403442383,
      "learning_rate": 0.00017768559422476026,
      "loss": 3.2033,
      "step": 65450
    },
    {
      "epoch": 7.049833171886772,
      "grad_norm": 0.8058651089668274,
      "learning_rate": 0.00017736235319469883,
      "loss": 3.2072,
      "step": 65500
    },
    {
      "epoch": 7.0552147239263805,
      "grad_norm": 0.7742359638214111,
      "learning_rate": 0.0001770391121646374,
      "loss": 3.1953,
      "step": 65550
    },
    {
      "epoch": 7.060596275965988,
      "grad_norm": 0.803683340549469,
      "learning_rate": 0.00017671587113457602,
      "loss": 3.2056,
      "step": 65600
    },
    {
      "epoch": 7.065977828005597,
      "grad_norm": 0.7680373787879944,
      "learning_rate": 0.0001763926301045146,
      "loss": 3.1984,
      "step": 65650
    },
    {
      "epoch": 7.071359380045205,
      "grad_norm": 0.7999513149261475,
      "learning_rate": 0.00017606938907445315,
      "loss": 3.2236,
      "step": 65700
    },
    {
      "epoch": 7.076740932084813,
      "grad_norm": 0.7916874885559082,
      "learning_rate": 0.00017574614804439178,
      "loss": 3.1951,
      "step": 65750
    },
    {
      "epoch": 7.0821224841244215,
      "grad_norm": 0.7580865025520325,
      "learning_rate": 0.00017542290701433034,
      "loss": 3.2117,
      "step": 65800
    },
    {
      "epoch": 7.08750403616403,
      "grad_norm": 0.8287043571472168,
      "learning_rate": 0.0001750996659842689,
      "loss": 3.2259,
      "step": 65850
    },
    {
      "epoch": 7.092885588203638,
      "grad_norm": 0.8034733533859253,
      "learning_rate": 0.0001747764249542075,
      "loss": 3.2206,
      "step": 65900
    },
    {
      "epoch": 7.098267140243246,
      "grad_norm": 0.7903028726577759,
      "learning_rate": 0.0001744531839241461,
      "loss": 3.2228,
      "step": 65950
    },
    {
      "epoch": 7.103648692282855,
      "grad_norm": 0.8349509835243225,
      "learning_rate": 0.00017412994289408467,
      "loss": 3.2322,
      "step": 66000
    },
    {
      "epoch": 7.103648692282855,
      "eval_accuracy": 0.3872474757489471,
      "eval_loss": 3.3593993186950684,
      "eval_runtime": 184.6002,
      "eval_samples_per_second": 97.568,
      "eval_steps_per_second": 6.1,
      "step": 66000
    },
    {
      "epoch": 7.109030244322462,
      "grad_norm": 0.8231300115585327,
      "learning_rate": 0.00017380670186402326,
      "loss": 3.2055,
      "step": 66050
    },
    {
      "epoch": 7.114411796362071,
      "grad_norm": 0.8903416395187378,
      "learning_rate": 0.00017348346083396183,
      "loss": 3.2181,
      "step": 66100
    },
    {
      "epoch": 7.119793348401679,
      "grad_norm": 0.8198266625404358,
      "learning_rate": 0.00017316021980390042,
      "loss": 3.2099,
      "step": 66150
    },
    {
      "epoch": 7.125174900441287,
      "grad_norm": 0.783141553401947,
      "learning_rate": 0.00017283697877383902,
      "loss": 3.2097,
      "step": 66200
    },
    {
      "epoch": 7.130556452480896,
      "grad_norm": 0.7667449116706848,
      "learning_rate": 0.0001725137377437776,
      "loss": 3.1992,
      "step": 66250
    },
    {
      "epoch": 7.135938004520503,
      "grad_norm": 0.8115949630737305,
      "learning_rate": 0.00017219049671371615,
      "loss": 3.2247,
      "step": 66300
    },
    {
      "epoch": 7.141319556560112,
      "grad_norm": 0.7727253437042236,
      "learning_rate": 0.00017186725568365478,
      "loss": 3.2203,
      "step": 66350
    },
    {
      "epoch": 7.1467011085997205,
      "grad_norm": 0.7740350961685181,
      "learning_rate": 0.00017154401465359334,
      "loss": 3.2199,
      "step": 66400
    },
    {
      "epoch": 7.152082660639328,
      "grad_norm": 0.7830681204795837,
      "learning_rate": 0.00017122077362353194,
      "loss": 3.2105,
      "step": 66450
    },
    {
      "epoch": 7.157464212678937,
      "grad_norm": 0.810469388961792,
      "learning_rate": 0.00017089753259347053,
      "loss": 3.2302,
      "step": 66500
    },
    {
      "epoch": 7.162845764718545,
      "grad_norm": 0.8127453327178955,
      "learning_rate": 0.0001705742915634091,
      "loss": 3.2225,
      "step": 66550
    },
    {
      "epoch": 7.168227316758153,
      "grad_norm": 0.8133081197738647,
      "learning_rate": 0.0001702510505333477,
      "loss": 3.2229,
      "step": 66600
    },
    {
      "epoch": 7.1736088687977615,
      "grad_norm": 0.8117032647132874,
      "learning_rate": 0.00016992780950328626,
      "loss": 3.2174,
      "step": 66650
    },
    {
      "epoch": 7.178990420837369,
      "grad_norm": 0.7892330288887024,
      "learning_rate": 0.00016960456847322486,
      "loss": 3.2199,
      "step": 66700
    },
    {
      "epoch": 7.184371972876978,
      "grad_norm": 0.7563703656196594,
      "learning_rate": 0.00016928132744316345,
      "loss": 3.2345,
      "step": 66750
    },
    {
      "epoch": 7.189753524916586,
      "grad_norm": 0.7536913752555847,
      "learning_rate": 0.00016895808641310202,
      "loss": 3.2084,
      "step": 66800
    },
    {
      "epoch": 7.195135076956194,
      "grad_norm": 0.8196943402290344,
      "learning_rate": 0.0001686348453830406,
      "loss": 3.2209,
      "step": 66850
    },
    {
      "epoch": 7.2005166289958025,
      "grad_norm": 0.7916746139526367,
      "learning_rate": 0.0001683116043529792,
      "loss": 3.2216,
      "step": 66900
    },
    {
      "epoch": 7.205898181035411,
      "grad_norm": 0.7951803803443909,
      "learning_rate": 0.00016798836332291778,
      "loss": 3.2184,
      "step": 66950
    },
    {
      "epoch": 7.211279733075019,
      "grad_norm": 0.8097406029701233,
      "learning_rate": 0.00016766512229285634,
      "loss": 3.2149,
      "step": 67000
    },
    {
      "epoch": 7.211279733075019,
      "eval_accuracy": 0.38753573186031887,
      "eval_loss": 3.3569376468658447,
      "eval_runtime": 184.6912,
      "eval_samples_per_second": 97.52,
      "eval_steps_per_second": 6.097,
      "step": 67000
    },
    {
      "epoch": 7.216661285114627,
      "grad_norm": 0.8015643954277039,
      "learning_rate": 0.00016734188126279494,
      "loss": 3.2318,
      "step": 67050
    },
    {
      "epoch": 7.222042837154235,
      "grad_norm": 0.7905345559120178,
      "learning_rate": 0.00016701864023273353,
      "loss": 3.2268,
      "step": 67100
    },
    {
      "epoch": 7.2274243891938434,
      "grad_norm": 0.8032287359237671,
      "learning_rate": 0.00016670186402327334,
      "loss": 3.2127,
      "step": 67150
    },
    {
      "epoch": 7.232805941233452,
      "grad_norm": 0.8226504325866699,
      "learning_rate": 0.0001663786229932119,
      "loss": 3.2253,
      "step": 67200
    },
    {
      "epoch": 7.23818749327306,
      "grad_norm": 0.8013386130332947,
      "learning_rate": 0.00016605538196315053,
      "loss": 3.2136,
      "step": 67250
    },
    {
      "epoch": 7.243569045312668,
      "grad_norm": 0.7741095423698425,
      "learning_rate": 0.0001657321409330891,
      "loss": 3.2263,
      "step": 67300
    },
    {
      "epoch": 7.248950597352277,
      "grad_norm": 0.7725183963775635,
      "learning_rate": 0.00016540889990302766,
      "loss": 3.2461,
      "step": 67350
    },
    {
      "epoch": 7.254332149391884,
      "grad_norm": 0.845971941947937,
      "learning_rate": 0.00016508565887296629,
      "loss": 3.2246,
      "step": 67400
    },
    {
      "epoch": 7.259713701431493,
      "grad_norm": 0.8192383646965027,
      "learning_rate": 0.00016476241784290485,
      "loss": 3.2089,
      "step": 67450
    },
    {
      "epoch": 7.265095253471101,
      "grad_norm": 0.7881582975387573,
      "learning_rate": 0.00016443917681284342,
      "loss": 3.2185,
      "step": 67500
    },
    {
      "epoch": 7.270476805510709,
      "grad_norm": 0.7972434163093567,
      "learning_rate": 0.00016411593578278202,
      "loss": 3.2279,
      "step": 67550
    },
    {
      "epoch": 7.275858357550318,
      "grad_norm": 0.8054504990577698,
      "learning_rate": 0.0001637926947527206,
      "loss": 3.2168,
      "step": 67600
    },
    {
      "epoch": 7.281239909589925,
      "grad_norm": 0.9870213270187378,
      "learning_rate": 0.00016346945372265918,
      "loss": 3.2263,
      "step": 67650
    },
    {
      "epoch": 7.286621461629534,
      "grad_norm": 0.7929873466491699,
      "learning_rate": 0.00016314621269259777,
      "loss": 3.2217,
      "step": 67700
    },
    {
      "epoch": 7.2920030136691425,
      "grad_norm": 0.8502267003059387,
      "learning_rate": 0.00016282297166253634,
      "loss": 3.2171,
      "step": 67750
    },
    {
      "epoch": 7.29738456570875,
      "grad_norm": 0.7992681264877319,
      "learning_rate": 0.00016249973063247494,
      "loss": 3.233,
      "step": 67800
    },
    {
      "epoch": 7.302766117748359,
      "grad_norm": 0.7842960953712463,
      "learning_rate": 0.00016217648960241353,
      "loss": 3.2266,
      "step": 67850
    },
    {
      "epoch": 7.308147669787967,
      "grad_norm": 0.7908195853233337,
      "learning_rate": 0.0001618532485723521,
      "loss": 3.2199,
      "step": 67900
    },
    {
      "epoch": 7.313529221827575,
      "grad_norm": 0.9430187940597534,
      "learning_rate": 0.00016153000754229067,
      "loss": 3.2218,
      "step": 67950
    },
    {
      "epoch": 7.3189107738671835,
      "grad_norm": 0.7958492040634155,
      "learning_rate": 0.0001612067665122293,
      "loss": 3.2237,
      "step": 68000
    },
    {
      "epoch": 7.3189107738671835,
      "eval_accuracy": 0.38775401551193817,
      "eval_loss": 3.354123830795288,
      "eval_runtime": 184.4291,
      "eval_samples_per_second": 97.658,
      "eval_steps_per_second": 6.105,
      "step": 68000
    },
    {
      "epoch": 7.324292325906791,
      "grad_norm": 0.784818708896637,
      "learning_rate": 0.00016088352548216785,
      "loss": 3.2084,
      "step": 68050
    },
    {
      "epoch": 7.3296738779464,
      "grad_norm": 0.7515696287155151,
      "learning_rate": 0.00016056028445210642,
      "loss": 3.214,
      "step": 68100
    },
    {
      "epoch": 7.335055429986008,
      "grad_norm": 0.8103103637695312,
      "learning_rate": 0.00016023704342204504,
      "loss": 3.2215,
      "step": 68150
    },
    {
      "epoch": 7.340436982025616,
      "grad_norm": 0.7764951586723328,
      "learning_rate": 0.0001599138023919836,
      "loss": 3.2311,
      "step": 68200
    },
    {
      "epoch": 7.3458185340652244,
      "grad_norm": 0.8081309795379639,
      "learning_rate": 0.0001595905613619222,
      "loss": 3.226,
      "step": 68250
    },
    {
      "epoch": 7.351200086104833,
      "grad_norm": 0.791309118270874,
      "learning_rate": 0.00015926732033186077,
      "loss": 3.2143,
      "step": 68300
    },
    {
      "epoch": 7.356581638144441,
      "grad_norm": 0.8047209978103638,
      "learning_rate": 0.00015894407930179934,
      "loss": 3.2546,
      "step": 68350
    },
    {
      "epoch": 7.361963190184049,
      "grad_norm": 0.8465035557746887,
      "learning_rate": 0.00015862083827173796,
      "loss": 3.2317,
      "step": 68400
    },
    {
      "epoch": 7.367344742223658,
      "grad_norm": 0.8064917922019958,
      "learning_rate": 0.00015829759724167653,
      "loss": 3.2442,
      "step": 68450
    },
    {
      "epoch": 7.372726294263265,
      "grad_norm": 0.8181912302970886,
      "learning_rate": 0.0001579743562116151,
      "loss": 3.2323,
      "step": 68500
    },
    {
      "epoch": 7.378107846302874,
      "grad_norm": 0.7603694796562195,
      "learning_rate": 0.00015765111518155372,
      "loss": 3.2263,
      "step": 68550
    },
    {
      "epoch": 7.383489398342482,
      "grad_norm": 0.7845638990402222,
      "learning_rate": 0.0001573278741514923,
      "loss": 3.2383,
      "step": 68600
    },
    {
      "epoch": 7.38887095038209,
      "grad_norm": 0.8256173729896545,
      "learning_rate": 0.00015700463312143085,
      "loss": 3.2371,
      "step": 68650
    },
    {
      "epoch": 7.394252502421699,
      "grad_norm": 0.7949895262718201,
      "learning_rate": 0.00015668139209136945,
      "loss": 3.2313,
      "step": 68700
    },
    {
      "epoch": 7.399634054461306,
      "grad_norm": 0.8290249705314636,
      "learning_rate": 0.00015635815106130804,
      "loss": 3.2535,
      "step": 68750
    },
    {
      "epoch": 7.405015606500915,
      "grad_norm": 0.8266456723213196,
      "learning_rate": 0.0001560349100312466,
      "loss": 3.2336,
      "step": 68800
    },
    {
      "epoch": 7.4103971585405235,
      "grad_norm": 0.7951944470405579,
      "learning_rate": 0.0001557116690011852,
      "loss": 3.2226,
      "step": 68850
    },
    {
      "epoch": 7.415778710580131,
      "grad_norm": 0.8083871603012085,
      "learning_rate": 0.00015538842797112377,
      "loss": 3.2418,
      "step": 68900
    },
    {
      "epoch": 7.42116026261974,
      "grad_norm": 0.789167046546936,
      "learning_rate": 0.00015506518694106237,
      "loss": 3.2354,
      "step": 68950
    },
    {
      "epoch": 7.426541814659347,
      "grad_norm": 0.8217098712921143,
      "learning_rate": 0.00015474194591100096,
      "loss": 3.2238,
      "step": 69000
    },
    {
      "epoch": 7.426541814659347,
      "eval_accuracy": 0.38835714769219637,
      "eval_loss": 3.349827289581299,
      "eval_runtime": 184.7119,
      "eval_samples_per_second": 97.509,
      "eval_steps_per_second": 6.096,
      "step": 69000
    },
    {
      "epoch": 7.431923366698956,
      "grad_norm": 0.8126811981201172,
      "learning_rate": 0.00015441870488093953,
      "loss": 3.2442,
      "step": 69050
    },
    {
      "epoch": 7.4373049187385645,
      "grad_norm": 0.9607537984848022,
      "learning_rate": 0.0001540954638508781,
      "loss": 3.216,
      "step": 69100
    },
    {
      "epoch": 7.442686470778172,
      "grad_norm": 0.8086814880371094,
      "learning_rate": 0.00015377222282081672,
      "loss": 3.2337,
      "step": 69150
    },
    {
      "epoch": 7.448068022817781,
      "grad_norm": 0.8346220850944519,
      "learning_rate": 0.0001534489817907553,
      "loss": 3.2318,
      "step": 69200
    },
    {
      "epoch": 7.453449574857389,
      "grad_norm": 0.8226746916770935,
      "learning_rate": 0.00015312574076069388,
      "loss": 3.2374,
      "step": 69250
    },
    {
      "epoch": 7.458831126896997,
      "grad_norm": 0.8161358833312988,
      "learning_rate": 0.00015280249973063248,
      "loss": 3.2381,
      "step": 69300
    },
    {
      "epoch": 7.4642126789366054,
      "grad_norm": 0.8100681304931641,
      "learning_rate": 0.00015248572352117228,
      "loss": 3.2397,
      "step": 69350
    },
    {
      "epoch": 7.469594230976213,
      "grad_norm": 0.8125149011611938,
      "learning_rate": 0.00015216248249111085,
      "loss": 3.2401,
      "step": 69400
    },
    {
      "epoch": 7.474975783015822,
      "grad_norm": 0.8136345744132996,
      "learning_rate": 0.00015183924146104945,
      "loss": 3.2329,
      "step": 69450
    },
    {
      "epoch": 7.48035733505543,
      "grad_norm": 0.8273459076881409,
      "learning_rate": 0.00015151600043098804,
      "loss": 3.2415,
      "step": 69500
    },
    {
      "epoch": 7.485738887095038,
      "grad_norm": 0.823276698589325,
      "learning_rate": 0.0001511927594009266,
      "loss": 3.2372,
      "step": 69550
    },
    {
      "epoch": 7.491120439134646,
      "grad_norm": 0.8145060539245605,
      "learning_rate": 0.00015086951837086518,
      "loss": 3.2337,
      "step": 69600
    },
    {
      "epoch": 7.496501991174255,
      "grad_norm": 0.8343734741210938,
      "learning_rate": 0.0001505462773408038,
      "loss": 3.2174,
      "step": 69650
    },
    {
      "epoch": 7.501883543213863,
      "grad_norm": 0.8493902683258057,
      "learning_rate": 0.00015022303631074236,
      "loss": 3.2238,
      "step": 69700
    },
    {
      "epoch": 7.507265095253471,
      "grad_norm": 0.8339319825172424,
      "learning_rate": 0.00014989979528068096,
      "loss": 3.2439,
      "step": 69750
    },
    {
      "epoch": 7.51264664729308,
      "grad_norm": 0.7970526218414307,
      "learning_rate": 0.00014957655425061953,
      "loss": 3.2329,
      "step": 69800
    },
    {
      "epoch": 7.518028199332687,
      "grad_norm": 0.8021500706672668,
      "learning_rate": 0.00014925331322055812,
      "loss": 3.2463,
      "step": 69850
    },
    {
      "epoch": 7.523409751372296,
      "grad_norm": 0.8772468566894531,
      "learning_rate": 0.0001489300721904967,
      "loss": 3.2297,
      "step": 69900
    },
    {
      "epoch": 7.528791303411904,
      "grad_norm": 0.8051742911338806,
      "learning_rate": 0.00014860683116043528,
      "loss": 3.2178,
      "step": 69950
    },
    {
      "epoch": 7.534172855451512,
      "grad_norm": 0.7842574119567871,
      "learning_rate": 0.00014828359013037385,
      "loss": 3.2394,
      "step": 70000
    },
    {
      "epoch": 7.534172855451512,
      "eval_accuracy": 0.3887887169625955,
      "eval_loss": 3.3451998233795166,
      "eval_runtime": 184.5808,
      "eval_samples_per_second": 97.578,
      "eval_steps_per_second": 6.1,
      "step": 70000
    },
    {
      "epoch": 7.539554407491121,
      "grad_norm": 0.8179084062576294,
      "learning_rate": 0.00014796034910031245,
      "loss": 3.2395,
      "step": 70050
    },
    {
      "epoch": 7.544935959530728,
      "grad_norm": 0.872739851474762,
      "learning_rate": 0.00014763710807025104,
      "loss": 3.214,
      "step": 70100
    },
    {
      "epoch": 7.550317511570337,
      "grad_norm": 0.8286665678024292,
      "learning_rate": 0.0001473138670401896,
      "loss": 3.2258,
      "step": 70150
    },
    {
      "epoch": 7.5556990636099455,
      "grad_norm": 0.8176371455192566,
      "learning_rate": 0.0001469906260101282,
      "loss": 3.2416,
      "step": 70200
    },
    {
      "epoch": 7.561080615649553,
      "grad_norm": 0.8359342813491821,
      "learning_rate": 0.0001466673849800668,
      "loss": 3.2232,
      "step": 70250
    },
    {
      "epoch": 7.566462167689162,
      "grad_norm": 0.8256863355636597,
      "learning_rate": 0.0001463441439500054,
      "loss": 3.2405,
      "step": 70300
    },
    {
      "epoch": 7.57184371972877,
      "grad_norm": 0.832430899143219,
      "learning_rate": 0.00014602090291994396,
      "loss": 3.2448,
      "step": 70350
    },
    {
      "epoch": 7.577225271768378,
      "grad_norm": 0.812171995639801,
      "learning_rate": 0.00014569766188988255,
      "loss": 3.2402,
      "step": 70400
    },
    {
      "epoch": 7.5826068238079865,
      "grad_norm": 0.849791944026947,
      "learning_rate": 0.00014537442085982112,
      "loss": 3.2329,
      "step": 70450
    },
    {
      "epoch": 7.587988375847594,
      "grad_norm": 0.8218454122543335,
      "learning_rate": 0.00014505117982975972,
      "loss": 3.2569,
      "step": 70500
    },
    {
      "epoch": 7.593369927887203,
      "grad_norm": 0.8892902135848999,
      "learning_rate": 0.00014472793879969828,
      "loss": 3.2399,
      "step": 70550
    },
    {
      "epoch": 7.598751479926811,
      "grad_norm": 0.8445575833320618,
      "learning_rate": 0.00014440469776963688,
      "loss": 3.2096,
      "step": 70600
    },
    {
      "epoch": 7.604133031966419,
      "grad_norm": 0.8453192114830017,
      "learning_rate": 0.00014408145673957545,
      "loss": 3.2261,
      "step": 70650
    },
    {
      "epoch": 7.609514584006027,
      "grad_norm": 0.8189914226531982,
      "learning_rate": 0.00014375821570951404,
      "loss": 3.236,
      "step": 70700
    },
    {
      "epoch": 7.614896136045635,
      "grad_norm": 0.8180625438690186,
      "learning_rate": 0.00014343497467945264,
      "loss": 3.2269,
      "step": 70750
    },
    {
      "epoch": 7.620277688085244,
      "grad_norm": 0.8000085353851318,
      "learning_rate": 0.00014311173364939123,
      "loss": 3.2445,
      "step": 70800
    },
    {
      "epoch": 7.625659240124852,
      "grad_norm": 0.8189267516136169,
      "learning_rate": 0.0001427884926193298,
      "loss": 3.221,
      "step": 70850
    },
    {
      "epoch": 7.63104079216446,
      "grad_norm": 0.8770104646682739,
      "learning_rate": 0.0001424652515892684,
      "loss": 3.2435,
      "step": 70900
    },
    {
      "epoch": 7.636422344204068,
      "grad_norm": 0.8498367667198181,
      "learning_rate": 0.000142142010559207,
      "loss": 3.2299,
      "step": 70950
    },
    {
      "epoch": 7.641803896243677,
      "grad_norm": 0.8652117848396301,
      "learning_rate": 0.00014181876952914555,
      "loss": 3.2487,
      "step": 71000
    },
    {
      "epoch": 7.641803896243677,
      "eval_accuracy": 0.3894372660499601,
      "eval_loss": 3.341649293899536,
      "eval_runtime": 184.9462,
      "eval_samples_per_second": 97.385,
      "eval_steps_per_second": 6.088,
      "step": 71000
    },
    {
      "epoch": 7.647185448283285,
      "grad_norm": 0.8142329454421997,
      "learning_rate": 0.00014149552849908415,
      "loss": 3.2602,
      "step": 71050
    },
    {
      "epoch": 7.652567000322893,
      "grad_norm": 0.869361400604248,
      "learning_rate": 0.00014117228746902272,
      "loss": 3.2542,
      "step": 71100
    },
    {
      "epoch": 7.657948552362502,
      "grad_norm": 0.792498767375946,
      "learning_rate": 0.0001408490464389613,
      "loss": 3.2124,
      "step": 71150
    },
    {
      "epoch": 7.663330104402109,
      "grad_norm": 0.7901838421821594,
      "learning_rate": 0.00014052580540889988,
      "loss": 3.2253,
      "step": 71200
    },
    {
      "epoch": 7.668711656441718,
      "grad_norm": 0.8150041103363037,
      "learning_rate": 0.00014020256437883847,
      "loss": 3.2392,
      "step": 71250
    },
    {
      "epoch": 7.674093208481326,
      "grad_norm": 0.8320419788360596,
      "learning_rate": 0.00013987932334877707,
      "loss": 3.2272,
      "step": 71300
    },
    {
      "epoch": 7.679474760520934,
      "grad_norm": 0.8377978801727295,
      "learning_rate": 0.00013956254713931687,
      "loss": 3.2441,
      "step": 71350
    },
    {
      "epoch": 7.684856312560543,
      "grad_norm": 0.7959548234939575,
      "learning_rate": 0.00013923930610925547,
      "loss": 3.2421,
      "step": 71400
    },
    {
      "epoch": 7.69023786460015,
      "grad_norm": 0.8453052043914795,
      "learning_rate": 0.00013891606507919404,
      "loss": 3.2083,
      "step": 71450
    },
    {
      "epoch": 7.695619416639759,
      "grad_norm": 0.8451975584030151,
      "learning_rate": 0.00013859282404913263,
      "loss": 3.2368,
      "step": 71500
    },
    {
      "epoch": 7.7010009686793675,
      "grad_norm": 0.8180558681488037,
      "learning_rate": 0.0001382695830190712,
      "loss": 3.2399,
      "step": 71550
    },
    {
      "epoch": 7.706382520718975,
      "grad_norm": 0.8149081468582153,
      "learning_rate": 0.0001379463419890098,
      "loss": 3.2282,
      "step": 71600
    },
    {
      "epoch": 7.711764072758584,
      "grad_norm": 0.8298981785774231,
      "learning_rate": 0.00013762310095894836,
      "loss": 3.2366,
      "step": 71650
    },
    {
      "epoch": 7.717145624798192,
      "grad_norm": 0.8108612298965454,
      "learning_rate": 0.00013729985992888696,
      "loss": 3.2308,
      "step": 71700
    },
    {
      "epoch": 7.7225271768378,
      "grad_norm": 0.8104018568992615,
      "learning_rate": 0.00013697661889882555,
      "loss": 3.2337,
      "step": 71750
    },
    {
      "epoch": 7.727908728877408,
      "grad_norm": 0.8229024410247803,
      "learning_rate": 0.00013665337786876412,
      "loss": 3.2546,
      "step": 71800
    },
    {
      "epoch": 7.733290280917016,
      "grad_norm": 0.8073616027832031,
      "learning_rate": 0.0001363301368387027,
      "loss": 3.2341,
      "step": 71850
    },
    {
      "epoch": 7.738671832956625,
      "grad_norm": 0.8207933306694031,
      "learning_rate": 0.0001360068958086413,
      "loss": 3.2404,
      "step": 71900
    },
    {
      "epoch": 7.744053384996233,
      "grad_norm": 0.7952953577041626,
      "learning_rate": 0.00013568365477857988,
      "loss": 3.2419,
      "step": 71950
    },
    {
      "epoch": 7.749434937035841,
      "grad_norm": 0.8869734406471252,
      "learning_rate": 0.00013536041374851847,
      "loss": 3.2442,
      "step": 72000
    },
    {
      "epoch": 7.749434937035841,
      "eval_accuracy": 0.3894115153155481,
      "eval_loss": 3.33847975730896,
      "eval_runtime": 184.598,
      "eval_samples_per_second": 97.569,
      "eval_steps_per_second": 6.1,
      "step": 72000
    },
    {
      "epoch": 7.754816489075449,
      "grad_norm": 0.880556583404541,
      "learning_rate": 0.00013503717271845706,
      "loss": 3.2394,
      "step": 72050
    },
    {
      "epoch": 7.760198041115058,
      "grad_norm": 0.7544320821762085,
      "learning_rate": 0.00013471393168839563,
      "loss": 3.2569,
      "step": 72100
    },
    {
      "epoch": 7.765579593154666,
      "grad_norm": 0.8141136169433594,
      "learning_rate": 0.00013439069065833423,
      "loss": 3.2304,
      "step": 72150
    },
    {
      "epoch": 7.770961145194274,
      "grad_norm": 0.8603343963623047,
      "learning_rate": 0.0001340674496282728,
      "loss": 3.2456,
      "step": 72200
    },
    {
      "epoch": 7.776342697233883,
      "grad_norm": 0.8688738942146301,
      "learning_rate": 0.0001337442085982114,
      "loss": 3.2221,
      "step": 72250
    },
    {
      "epoch": 7.78172424927349,
      "grad_norm": 0.8236707448959351,
      "learning_rate": 0.00013342096756814996,
      "loss": 3.247,
      "step": 72300
    },
    {
      "epoch": 7.787105801313099,
      "grad_norm": 0.8244578838348389,
      "learning_rate": 0.00013309772653808855,
      "loss": 3.2482,
      "step": 72350
    },
    {
      "epoch": 7.792487353352707,
      "grad_norm": 0.867263913154602,
      "learning_rate": 0.00013277448550802715,
      "loss": 3.2227,
      "step": 72400
    },
    {
      "epoch": 7.797868905392315,
      "grad_norm": 0.8356115818023682,
      "learning_rate": 0.0001324512444779657,
      "loss": 3.2356,
      "step": 72450
    },
    {
      "epoch": 7.803250457431924,
      "grad_norm": 0.8518567681312561,
      "learning_rate": 0.0001321280034479043,
      "loss": 3.2419,
      "step": 72500
    },
    {
      "epoch": 7.808632009471531,
      "grad_norm": 0.8349553942680359,
      "learning_rate": 0.0001318047624178429,
      "loss": 3.2326,
      "step": 72550
    },
    {
      "epoch": 7.81401356151114,
      "grad_norm": 0.8264884352684021,
      "learning_rate": 0.0001314815213877815,
      "loss": 3.239,
      "step": 72600
    },
    {
      "epoch": 7.819395113550748,
      "grad_norm": 0.8145789504051208,
      "learning_rate": 0.00013115828035772007,
      "loss": 3.2395,
      "step": 72650
    },
    {
      "epoch": 7.824776665590356,
      "grad_norm": 0.8533801436424255,
      "learning_rate": 0.00013083503932765866,
      "loss": 3.2414,
      "step": 72700
    },
    {
      "epoch": 7.830158217629965,
      "grad_norm": 0.853441596031189,
      "learning_rate": 0.00013051179829759723,
      "loss": 3.2316,
      "step": 72750
    },
    {
      "epoch": 7.835539769669572,
      "grad_norm": 0.8426158428192139,
      "learning_rate": 0.00013018855726753582,
      "loss": 3.2577,
      "step": 72800
    },
    {
      "epoch": 7.840921321709181,
      "grad_norm": 0.8060492873191833,
      "learning_rate": 0.0001298653162374744,
      "loss": 3.2267,
      "step": 72850
    },
    {
      "epoch": 7.846302873748789,
      "grad_norm": 0.8100488185882568,
      "learning_rate": 0.00012954207520741298,
      "loss": 3.2407,
      "step": 72900
    },
    {
      "epoch": 7.851684425788397,
      "grad_norm": 0.8481305837631226,
      "learning_rate": 0.00012921883417735155,
      "loss": 3.2192,
      "step": 72950
    },
    {
      "epoch": 7.857065977828006,
      "grad_norm": 0.8777086138725281,
      "learning_rate": 0.00012889559314729015,
      "loss": 3.2569,
      "step": 73000
    },
    {
      "epoch": 7.857065977828006,
      "eval_accuracy": 0.3896842340639624,
      "eval_loss": 3.3339617252349854,
      "eval_runtime": 184.6586,
      "eval_samples_per_second": 97.537,
      "eval_steps_per_second": 6.098,
      "step": 73000
    },
    {
      "epoch": 7.862447529867614,
      "grad_norm": 0.8324296474456787,
      "learning_rate": 0.00012857235211722874,
      "loss": 3.2141,
      "step": 73050
    },
    {
      "epoch": 7.867829081907222,
      "grad_norm": 0.8081611394882202,
      "learning_rate": 0.00012824911108716734,
      "loss": 3.2168,
      "step": 73100
    },
    {
      "epoch": 7.87321063394683,
      "grad_norm": 0.8032556176185608,
      "learning_rate": 0.0001279258700571059,
      "loss": 3.2353,
      "step": 73150
    },
    {
      "epoch": 7.878592185986438,
      "grad_norm": 0.8371847867965698,
      "learning_rate": 0.0001276026290270445,
      "loss": 3.2337,
      "step": 73200
    },
    {
      "epoch": 7.883973738026047,
      "grad_norm": 0.8007276654243469,
      "learning_rate": 0.00012727938799698307,
      "loss": 3.2279,
      "step": 73250
    },
    {
      "epoch": 7.889355290065655,
      "grad_norm": 0.8758023381233215,
      "learning_rate": 0.00012695614696692166,
      "loss": 3.2199,
      "step": 73300
    },
    {
      "epoch": 7.894736842105263,
      "grad_norm": 0.9168129563331604,
      "learning_rate": 0.00012663290593686023,
      "loss": 3.2366,
      "step": 73350
    },
    {
      "epoch": 7.900118394144871,
      "grad_norm": 0.8541741967201233,
      "learning_rate": 0.00012630966490679882,
      "loss": 3.2348,
      "step": 73400
    },
    {
      "epoch": 7.90549994618448,
      "grad_norm": 0.8171018362045288,
      "learning_rate": 0.0001259864238767374,
      "loss": 3.2464,
      "step": 73450
    },
    {
      "epoch": 7.910881498224088,
      "grad_norm": 0.8532667756080627,
      "learning_rate": 0.00012566318284667598,
      "loss": 3.2119,
      "step": 73500
    },
    {
      "epoch": 7.916263050263696,
      "grad_norm": 0.8246223330497742,
      "learning_rate": 0.00012533994181661458,
      "loss": 3.2272,
      "step": 73550
    },
    {
      "epoch": 7.921644602303305,
      "grad_norm": 0.8351603746414185,
      "learning_rate": 0.00012501670078655317,
      "loss": 3.2407,
      "step": 73600
    },
    {
      "epoch": 7.927026154342912,
      "grad_norm": 0.8015763163566589,
      "learning_rate": 0.00012469345975649174,
      "loss": 3.2179,
      "step": 73650
    },
    {
      "epoch": 7.932407706382521,
      "grad_norm": 0.8650892376899719,
      "learning_rate": 0.00012437021872643034,
      "loss": 3.2296,
      "step": 73700
    },
    {
      "epoch": 7.937789258422129,
      "grad_norm": 0.8532193303108215,
      "learning_rate": 0.00012405344251697014,
      "loss": 3.2463,
      "step": 73750
    },
    {
      "epoch": 7.943170810461737,
      "grad_norm": 0.810485303401947,
      "learning_rate": 0.00012373020148690874,
      "loss": 3.2286,
      "step": 73800
    },
    {
      "epoch": 7.948552362501346,
      "grad_norm": 0.7970120906829834,
      "learning_rate": 0.0001234069604568473,
      "loss": 3.2275,
      "step": 73850
    },
    {
      "epoch": 7.953933914540953,
      "grad_norm": 0.8421259522438049,
      "learning_rate": 0.0001230837194267859,
      "loss": 3.246,
      "step": 73900
    },
    {
      "epoch": 7.959315466580562,
      "grad_norm": 0.8532693982124329,
      "learning_rate": 0.00012276047839672447,
      "loss": 3.2488,
      "step": 73950
    },
    {
      "epoch": 7.96469701862017,
      "grad_norm": 0.8387293219566345,
      "learning_rate": 0.00012243723736666306,
      "loss": 3.2501,
      "step": 74000
    },
    {
      "epoch": 7.96469701862017,
      "eval_accuracy": 0.3901934901491489,
      "eval_loss": 3.3306832313537598,
      "eval_runtime": 184.7101,
      "eval_samples_per_second": 97.51,
      "eval_steps_per_second": 6.096,
      "step": 74000
    },
    {
      "epoch": 7.970078570659778,
      "grad_norm": 0.8431631326675415,
      "learning_rate": 0.00012211399633660166,
      "loss": 3.2543,
      "step": 74050
    },
    {
      "epoch": 7.975460122699387,
      "grad_norm": 0.8307387232780457,
      "learning_rate": 0.00012179075530654022,
      "loss": 3.2306,
      "step": 74100
    },
    {
      "epoch": 7.980841674738995,
      "grad_norm": 0.8382935523986816,
      "learning_rate": 0.00012146751427647882,
      "loss": 3.2373,
      "step": 74150
    },
    {
      "epoch": 7.986223226778603,
      "grad_norm": 0.8308893442153931,
      "learning_rate": 0.0001211442732464174,
      "loss": 3.2195,
      "step": 74200
    },
    {
      "epoch": 7.991604778818211,
      "grad_norm": 0.8867136240005493,
      "learning_rate": 0.00012082103221635598,
      "loss": 3.2518,
      "step": 74250
    },
    {
      "epoch": 7.996986330857819,
      "grad_norm": 0.8336354494094849,
      "learning_rate": 0.00012049779118629456,
      "loss": 3.2431,
      "step": 74300
    },
    {
      "epoch": 8.002367882897428,
      "grad_norm": 0.8303877711296082,
      "learning_rate": 0.00012017455015623316,
      "loss": 3.1969,
      "step": 74350
    },
    {
      "epoch": 8.007749434937036,
      "grad_norm": 0.836917519569397,
      "learning_rate": 0.00011985130912617175,
      "loss": 3.1684,
      "step": 74400
    },
    {
      "epoch": 8.013130986976645,
      "grad_norm": 0.8200033903121948,
      "learning_rate": 0.00011952806809611032,
      "loss": 3.1589,
      "step": 74450
    },
    {
      "epoch": 8.018512539016251,
      "grad_norm": 0.831483006477356,
      "learning_rate": 0.00011920482706604891,
      "loss": 3.1496,
      "step": 74500
    },
    {
      "epoch": 8.02389409105586,
      "grad_norm": 0.8704330921173096,
      "learning_rate": 0.0001188815860359875,
      "loss": 3.1754,
      "step": 74550
    },
    {
      "epoch": 8.029275643095469,
      "grad_norm": 0.8458114266395569,
      "learning_rate": 0.00011855834500592608,
      "loss": 3.1645,
      "step": 74600
    },
    {
      "epoch": 8.034657195135077,
      "grad_norm": 0.8807541728019714,
      "learning_rate": 0.00011823510397586466,
      "loss": 3.1698,
      "step": 74650
    },
    {
      "epoch": 8.040038747174686,
      "grad_norm": 0.8740618824958801,
      "learning_rate": 0.00011791186294580325,
      "loss": 3.1632,
      "step": 74700
    },
    {
      "epoch": 8.045420299214294,
      "grad_norm": 0.8318957686424255,
      "learning_rate": 0.00011758862191574182,
      "loss": 3.1803,
      "step": 74750
    },
    {
      "epoch": 8.050801851253901,
      "grad_norm": 0.9062291979789734,
      "learning_rate": 0.00011726538088568041,
      "loss": 3.1795,
      "step": 74800
    },
    {
      "epoch": 8.05618340329351,
      "grad_norm": 0.8924227952957153,
      "learning_rate": 0.000116942139855619,
      "loss": 3.155,
      "step": 74850
    },
    {
      "epoch": 8.061564955333118,
      "grad_norm": 0.8725975155830383,
      "learning_rate": 0.00011661889882555759,
      "loss": 3.1632,
      "step": 74900
    },
    {
      "epoch": 8.066946507372727,
      "grad_norm": 0.8697208762168884,
      "learning_rate": 0.00011629565779549616,
      "loss": 3.1696,
      "step": 74950
    },
    {
      "epoch": 8.072328059412335,
      "grad_norm": 0.8555173873901367,
      "learning_rate": 0.00011597241676543475,
      "loss": 3.1614,
      "step": 75000
    },
    {
      "epoch": 8.072328059412335,
      "eval_accuracy": 0.39014285790342734,
      "eval_loss": 3.3369150161743164,
      "eval_runtime": 184.9651,
      "eval_samples_per_second": 97.375,
      "eval_steps_per_second": 6.088,
      "step": 75000
    },
    {
      "epoch": 8.077709611451942,
      "grad_norm": 0.8447503447532654,
      "learning_rate": 0.00011564917573537335,
      "loss": 3.1697,
      "step": 75050
    },
    {
      "epoch": 8.08309116349155,
      "grad_norm": 0.8746142983436584,
      "learning_rate": 0.00011532593470531191,
      "loss": 3.1678,
      "step": 75100
    },
    {
      "epoch": 8.088472715531159,
      "grad_norm": 0.8564710021018982,
      "learning_rate": 0.0001150026936752505,
      "loss": 3.1848,
      "step": 75150
    },
    {
      "epoch": 8.093854267570768,
      "grad_norm": 0.87080979347229,
      "learning_rate": 0.00011467945264518909,
      "loss": 3.1768,
      "step": 75200
    },
    {
      "epoch": 8.099235819610376,
      "grad_norm": 0.8317235708236694,
      "learning_rate": 0.00011435621161512766,
      "loss": 3.1678,
      "step": 75250
    },
    {
      "epoch": 8.104617371649983,
      "grad_norm": 0.8491494059562683,
      "learning_rate": 0.00011403297058506625,
      "loss": 3.1711,
      "step": 75300
    },
    {
      "epoch": 8.109998923689592,
      "grad_norm": 0.8416190147399902,
      "learning_rate": 0.00011370972955500485,
      "loss": 3.1771,
      "step": 75350
    },
    {
      "epoch": 8.1153804757292,
      "grad_norm": 0.846297562122345,
      "learning_rate": 0.00011338648852494343,
      "loss": 3.1605,
      "step": 75400
    },
    {
      "epoch": 8.120762027768809,
      "grad_norm": 0.8915745615959167,
      "learning_rate": 0.00011306324749488201,
      "loss": 3.1726,
      "step": 75450
    },
    {
      "epoch": 8.126143579808417,
      "grad_norm": 0.8499758243560791,
      "learning_rate": 0.00011274000646482059,
      "loss": 3.1699,
      "step": 75500
    },
    {
      "epoch": 8.131525131848026,
      "grad_norm": 0.8181933164596558,
      "learning_rate": 0.00011241676543475918,
      "loss": 3.1845,
      "step": 75550
    },
    {
      "epoch": 8.136906683887632,
      "grad_norm": 0.8639881014823914,
      "learning_rate": 0.00011209352440469775,
      "loss": 3.168,
      "step": 75600
    },
    {
      "epoch": 8.142288235927241,
      "grad_norm": 0.8862914443016052,
      "learning_rate": 0.00011177028337463635,
      "loss": 3.1765,
      "step": 75650
    },
    {
      "epoch": 8.14766978796685,
      "grad_norm": 0.8618167042732239,
      "learning_rate": 0.00011144704234457493,
      "loss": 3.1776,
      "step": 75700
    },
    {
      "epoch": 8.153051340006458,
      "grad_norm": 0.8700433969497681,
      "learning_rate": 0.00011112380131451351,
      "loss": 3.1728,
      "step": 75750
    },
    {
      "epoch": 8.158432892046067,
      "grad_norm": 0.8872320652008057,
      "learning_rate": 0.00011080056028445209,
      "loss": 3.1696,
      "step": 75800
    },
    {
      "epoch": 8.163814444085673,
      "grad_norm": 0.8248878121376038,
      "learning_rate": 0.00011047731925439068,
      "loss": 3.1714,
      "step": 75850
    },
    {
      "epoch": 8.169195996125282,
      "grad_norm": 0.8656007051467896,
      "learning_rate": 0.00011015407822432928,
      "loss": 3.1678,
      "step": 75900
    },
    {
      "epoch": 8.17457754816489,
      "grad_norm": 0.886645495891571,
      "learning_rate": 0.00010983083719426785,
      "loss": 3.1808,
      "step": 75950
    },
    {
      "epoch": 8.1799591002045,
      "grad_norm": 0.917984127998352,
      "learning_rate": 0.00010950759616420644,
      "loss": 3.1899,
      "step": 76000
    },
    {
      "epoch": 8.1799591002045,
      "eval_accuracy": 0.39039047783475656,
      "eval_loss": 3.333770513534546,
      "eval_runtime": 184.4579,
      "eval_samples_per_second": 97.643,
      "eval_steps_per_second": 6.104,
      "step": 76000
    },
    {
      "epoch": 8.185340652244108,
      "grad_norm": 0.8363960385322571,
      "learning_rate": 0.00010918435513414502,
      "loss": 3.1928,
      "step": 76050
    },
    {
      "epoch": 8.190722204283716,
      "grad_norm": 0.8816800713539124,
      "learning_rate": 0.00010886111410408359,
      "loss": 3.1695,
      "step": 76100
    },
    {
      "epoch": 8.196103756323323,
      "grad_norm": 0.8266114592552185,
      "learning_rate": 0.00010853787307402218,
      "loss": 3.1846,
      "step": 76150
    },
    {
      "epoch": 8.201485308362932,
      "grad_norm": 0.8366254568099976,
      "learning_rate": 0.00010821463204396078,
      "loss": 3.1775,
      "step": 76200
    },
    {
      "epoch": 8.20686686040254,
      "grad_norm": 0.8426965475082397,
      "learning_rate": 0.00010789139101389935,
      "loss": 3.1647,
      "step": 76250
    },
    {
      "epoch": 8.212248412442149,
      "grad_norm": 0.799103319644928,
      "learning_rate": 0.00010756814998383794,
      "loss": 3.1751,
      "step": 76300
    },
    {
      "epoch": 8.217629964481757,
      "grad_norm": 0.8718476295471191,
      "learning_rate": 0.00010724490895377652,
      "loss": 3.1747,
      "step": 76350
    },
    {
      "epoch": 8.223011516521364,
      "grad_norm": 0.8915057182312012,
      "learning_rate": 0.00010692166792371512,
      "loss": 3.1874,
      "step": 76400
    },
    {
      "epoch": 8.228393068560973,
      "grad_norm": 0.8778514266014099,
      "learning_rate": 0.00010659842689365368,
      "loss": 3.1709,
      "step": 76450
    },
    {
      "epoch": 8.233774620600581,
      "grad_norm": 0.8106761574745178,
      "learning_rate": 0.00010627518586359228,
      "loss": 3.1766,
      "step": 76500
    },
    {
      "epoch": 8.23915617264019,
      "grad_norm": 0.8877562284469604,
      "learning_rate": 0.00010595194483353086,
      "loss": 3.178,
      "step": 76550
    },
    {
      "epoch": 8.244537724679798,
      "grad_norm": 0.874200165271759,
      "learning_rate": 0.00010562870380346944,
      "loss": 3.1837,
      "step": 76600
    },
    {
      "epoch": 8.249919276719407,
      "grad_norm": 0.8712376952171326,
      "learning_rate": 0.00010530546277340802,
      "loss": 3.1791,
      "step": 76650
    },
    {
      "epoch": 8.255300828759013,
      "grad_norm": 0.842028021812439,
      "learning_rate": 0.00010498222174334662,
      "loss": 3.188,
      "step": 76700
    },
    {
      "epoch": 8.260682380798622,
      "grad_norm": 0.8732202053070068,
      "learning_rate": 0.00010465898071328519,
      "loss": 3.1822,
      "step": 76750
    },
    {
      "epoch": 8.26606393283823,
      "grad_norm": 0.8106330037117004,
      "learning_rate": 0.00010433573968322378,
      "loss": 3.1808,
      "step": 76800
    },
    {
      "epoch": 8.27144548487784,
      "grad_norm": 0.8642439246177673,
      "learning_rate": 0.00010401249865316237,
      "loss": 3.1608,
      "step": 76850
    },
    {
      "epoch": 8.276827036917448,
      "grad_norm": 0.8715514540672302,
      "learning_rate": 0.00010368925762310096,
      "loss": 3.1802,
      "step": 76900
    },
    {
      "epoch": 8.282208588957054,
      "grad_norm": 0.8595199584960938,
      "learning_rate": 0.00010336601659303954,
      "loss": 3.1766,
      "step": 76950
    },
    {
      "epoch": 8.287590140996663,
      "grad_norm": 0.8247595429420471,
      "learning_rate": 0.00010304277556297812,
      "loss": 3.1874,
      "step": 77000
    },
    {
      "epoch": 8.287590140996663,
      "eval_accuracy": 0.3907075269613992,
      "eval_loss": 3.330806016921997,
      "eval_runtime": 184.729,
      "eval_samples_per_second": 97.5,
      "eval_steps_per_second": 6.095,
      "step": 77000
    },
    {
      "epoch": 8.292971693036272,
      "grad_norm": 0.9381096959114075,
      "learning_rate": 0.00010271953453291671,
      "loss": 3.166,
      "step": 77050
    },
    {
      "epoch": 8.29835324507588,
      "grad_norm": 0.8634301424026489,
      "learning_rate": 0.00010239629350285528,
      "loss": 3.1607,
      "step": 77100
    },
    {
      "epoch": 8.303734797115489,
      "grad_norm": 0.8498184680938721,
      "learning_rate": 0.00010207305247279387,
      "loss": 3.176,
      "step": 77150
    },
    {
      "epoch": 8.309116349155097,
      "grad_norm": 0.8834637999534607,
      "learning_rate": 0.00010174981144273246,
      "loss": 3.1865,
      "step": 77200
    },
    {
      "epoch": 8.314497901194704,
      "grad_norm": 0.877652108669281,
      "learning_rate": 0.00010142657041267104,
      "loss": 3.1713,
      "step": 77250
    },
    {
      "epoch": 8.319879453234313,
      "grad_norm": 0.8718097805976868,
      "learning_rate": 0.00010110332938260962,
      "loss": 3.17,
      "step": 77300
    },
    {
      "epoch": 8.325261005273921,
      "grad_norm": 0.8588361740112305,
      "learning_rate": 0.00010078655317314944,
      "loss": 3.1928,
      "step": 77350
    },
    {
      "epoch": 8.33064255731353,
      "grad_norm": 0.8447306752204895,
      "learning_rate": 0.00010046331214308802,
      "loss": 3.1892,
      "step": 77400
    },
    {
      "epoch": 8.336024109353138,
      "grad_norm": 0.8749763369560242,
      "learning_rate": 0.0001001400711130266,
      "loss": 3.1865,
      "step": 77450
    },
    {
      "epoch": 8.341405661392745,
      "grad_norm": 0.8686197400093079,
      "learning_rate": 9.98168300829652e-05,
      "loss": 3.1846,
      "step": 77500
    },
    {
      "epoch": 8.346787213432354,
      "grad_norm": 0.8535328507423401,
      "learning_rate": 9.949358905290376e-05,
      "loss": 3.1957,
      "step": 77550
    },
    {
      "epoch": 8.352168765471962,
      "grad_norm": 0.8441202044487,
      "learning_rate": 9.917034802284236e-05,
      "loss": 3.1839,
      "step": 77600
    },
    {
      "epoch": 8.35755031751157,
      "grad_norm": 0.8363760113716125,
      "learning_rate": 9.884710699278094e-05,
      "loss": 3.193,
      "step": 77650
    },
    {
      "epoch": 8.36293186955118,
      "grad_norm": 0.8740014433860779,
      "learning_rate": 9.852386596271953e-05,
      "loss": 3.1935,
      "step": 77700
    },
    {
      "epoch": 8.368313421590786,
      "grad_norm": 0.8385241627693176,
      "learning_rate": 9.82006249326581e-05,
      "loss": 3.1776,
      "step": 77750
    },
    {
      "epoch": 8.373694973630395,
      "grad_norm": 0.8790156841278076,
      "learning_rate": 9.78773839025967e-05,
      "loss": 3.1687,
      "step": 77800
    },
    {
      "epoch": 8.379076525670003,
      "grad_norm": 0.9463781714439392,
      "learning_rate": 9.755414287253529e-05,
      "loss": 3.1798,
      "step": 77850
    },
    {
      "epoch": 8.384458077709612,
      "grad_norm": 0.8797891736030579,
      "learning_rate": 9.723090184247386e-05,
      "loss": 3.1966,
      "step": 77900
    },
    {
      "epoch": 8.38983962974922,
      "grad_norm": 0.8536663055419922,
      "learning_rate": 9.690766081241245e-05,
      "loss": 3.1673,
      "step": 77950
    },
    {
      "epoch": 8.395221181788829,
      "grad_norm": 0.880811870098114,
      "learning_rate": 9.658441978235103e-05,
      "loss": 3.1641,
      "step": 78000
    },
    {
      "epoch": 8.395221181788829,
      "eval_accuracy": 0.3911067176712296,
      "eval_loss": 3.3282108306884766,
      "eval_runtime": 185.041,
      "eval_samples_per_second": 97.335,
      "eval_steps_per_second": 6.085,
      "step": 78000
    },
    {
      "epoch": 8.400602733828435,
      "grad_norm": 0.8960733413696289,
      "learning_rate": 9.626117875228961e-05,
      "loss": 3.1936,
      "step": 78050
    },
    {
      "epoch": 8.405984285868044,
      "grad_norm": 0.893669605255127,
      "learning_rate": 9.59379377222282e-05,
      "loss": 3.1858,
      "step": 78100
    },
    {
      "epoch": 8.411365837907653,
      "grad_norm": 0.8383551836013794,
      "learning_rate": 9.561469669216679e-05,
      "loss": 3.1571,
      "step": 78150
    },
    {
      "epoch": 8.416747389947261,
      "grad_norm": 0.8875780701637268,
      "learning_rate": 9.529145566210537e-05,
      "loss": 3.1831,
      "step": 78200
    },
    {
      "epoch": 8.42212894198687,
      "grad_norm": 0.9209667444229126,
      "learning_rate": 9.496821463204395e-05,
      "loss": 3.1833,
      "step": 78250
    },
    {
      "epoch": 8.427510494026476,
      "grad_norm": 0.8284728527069092,
      "learning_rate": 9.464497360198253e-05,
      "loss": 3.1816,
      "step": 78300
    },
    {
      "epoch": 8.432892046066085,
      "grad_norm": 0.8703120946884155,
      "learning_rate": 9.432173257192113e-05,
      "loss": 3.1751,
      "step": 78350
    },
    {
      "epoch": 8.438273598105694,
      "grad_norm": 0.8745947480201721,
      "learning_rate": 9.39984915418597e-05,
      "loss": 3.1674,
      "step": 78400
    },
    {
      "epoch": 8.443655150145302,
      "grad_norm": 0.8946985602378845,
      "learning_rate": 9.367525051179829e-05,
      "loss": 3.1602,
      "step": 78450
    },
    {
      "epoch": 8.44903670218491,
      "grad_norm": 0.867800772190094,
      "learning_rate": 9.335200948173688e-05,
      "loss": 3.1897,
      "step": 78500
    },
    {
      "epoch": 8.45441825422452,
      "grad_norm": 0.8727012872695923,
      "learning_rate": 9.302876845167545e-05,
      "loss": 3.1946,
      "step": 78550
    },
    {
      "epoch": 8.459799806264126,
      "grad_norm": 0.8764132261276245,
      "learning_rate": 9.270552742161403e-05,
      "loss": 3.1935,
      "step": 78600
    },
    {
      "epoch": 8.465181358303735,
      "grad_norm": 0.8671554923057556,
      "learning_rate": 9.238228639155263e-05,
      "loss": 3.1577,
      "step": 78650
    },
    {
      "epoch": 8.470562910343343,
      "grad_norm": 0.8665030002593994,
      "learning_rate": 9.205904536149122e-05,
      "loss": 3.1851,
      "step": 78700
    },
    {
      "epoch": 8.475944462382952,
      "grad_norm": 0.8668508529663086,
      "learning_rate": 9.173580433142979e-05,
      "loss": 3.2038,
      "step": 78750
    },
    {
      "epoch": 8.48132601442256,
      "grad_norm": 0.8960902094841003,
      "learning_rate": 9.141256330136838e-05,
      "loss": 3.1813,
      "step": 78800
    },
    {
      "epoch": 8.486707566462167,
      "grad_norm": 0.8812357783317566,
      "learning_rate": 9.108932227130697e-05,
      "loss": 3.1816,
      "step": 78850
    },
    {
      "epoch": 8.492089118501776,
      "grad_norm": 0.9052111506462097,
      "learning_rate": 9.076608124124555e-05,
      "loss": 3.1729,
      "step": 78900
    },
    {
      "epoch": 8.497470670541384,
      "grad_norm": 0.9190753102302551,
      "learning_rate": 9.044284021118413e-05,
      "loss": 3.1959,
      "step": 78950
    },
    {
      "epoch": 8.502852222580993,
      "grad_norm": 0.8276004195213318,
      "learning_rate": 9.011959918112272e-05,
      "loss": 3.185,
      "step": 79000
    },
    {
      "epoch": 8.502852222580993,
      "eval_accuracy": 0.3911831006513632,
      "eval_loss": 3.32541561126709,
      "eval_runtime": 184.5689,
      "eval_samples_per_second": 97.584,
      "eval_steps_per_second": 6.101,
      "step": 79000
    },
    {
      "epoch": 8.508233774620601,
      "grad_norm": 0.9947301745414734,
      "learning_rate": 8.979635815106129e-05,
      "loss": 3.1801,
      "step": 79050
    },
    {
      "epoch": 8.513615326660208,
      "grad_norm": 0.8838504552841187,
      "learning_rate": 8.947311712099989e-05,
      "loss": 3.196,
      "step": 79100
    },
    {
      "epoch": 8.518996878699816,
      "grad_norm": 0.8844854235649109,
      "learning_rate": 8.914987609093847e-05,
      "loss": 3.1692,
      "step": 79150
    },
    {
      "epoch": 8.524378430739425,
      "grad_norm": 0.9206638932228088,
      "learning_rate": 8.882663506087706e-05,
      "loss": 3.1824,
      "step": 79200
    },
    {
      "epoch": 8.529759982779034,
      "grad_norm": 0.8852978944778442,
      "learning_rate": 8.850339403081563e-05,
      "loss": 3.2002,
      "step": 79250
    },
    {
      "epoch": 8.535141534818642,
      "grad_norm": 0.9327110052108765,
      "learning_rate": 8.818015300075422e-05,
      "loss": 3.2021,
      "step": 79300
    },
    {
      "epoch": 8.54052308685825,
      "grad_norm": 0.8577570915222168,
      "learning_rate": 8.785691197069282e-05,
      "loss": 3.1762,
      "step": 79350
    },
    {
      "epoch": 8.545904638897857,
      "grad_norm": 0.8478599190711975,
      "learning_rate": 8.753367094063139e-05,
      "loss": 3.1831,
      "step": 79400
    },
    {
      "epoch": 8.551286190937466,
      "grad_norm": 0.909991443157196,
      "learning_rate": 8.721042991056998e-05,
      "loss": 3.1794,
      "step": 79450
    },
    {
      "epoch": 8.556667742977075,
      "grad_norm": 0.9163396954536438,
      "learning_rate": 8.688718888050856e-05,
      "loss": 3.1727,
      "step": 79500
    },
    {
      "epoch": 8.562049295016683,
      "grad_norm": 0.862952470779419,
      "learning_rate": 8.656394785044713e-05,
      "loss": 3.1933,
      "step": 79550
    },
    {
      "epoch": 8.567430847056292,
      "grad_norm": 0.8733549118041992,
      "learning_rate": 8.624070682038572e-05,
      "loss": 3.1773,
      "step": 79600
    },
    {
      "epoch": 8.572812399095898,
      "grad_norm": 0.8446792364120483,
      "learning_rate": 8.591746579032432e-05,
      "loss": 3.1822,
      "step": 79650
    },
    {
      "epoch": 8.578193951135507,
      "grad_norm": 0.8886731863021851,
      "learning_rate": 8.55942247602629e-05,
      "loss": 3.1945,
      "step": 79700
    },
    {
      "epoch": 8.583575503175116,
      "grad_norm": 0.8565948009490967,
      "learning_rate": 8.52774485508027e-05,
      "loss": 3.1875,
      "step": 79750
    },
    {
      "epoch": 8.588957055214724,
      "grad_norm": 0.8837953209877014,
      "learning_rate": 8.49542075207413e-05,
      "loss": 3.1851,
      "step": 79800
    },
    {
      "epoch": 8.594338607254333,
      "grad_norm": 0.9210212826728821,
      "learning_rate": 8.463096649067987e-05,
      "loss": 3.1748,
      "step": 79850
    },
    {
      "epoch": 8.599720159293941,
      "grad_norm": 0.9159842729568481,
      "learning_rate": 8.430772546061846e-05,
      "loss": 3.188,
      "step": 79900
    },
    {
      "epoch": 8.605101711333548,
      "grad_norm": 0.8564115166664124,
      "learning_rate": 8.398448443055704e-05,
      "loss": 3.1862,
      "step": 79950
    },
    {
      "epoch": 8.610483263373157,
      "grad_norm": 0.8882423639297485,
      "learning_rate": 8.366124340049564e-05,
      "loss": 3.1817,
      "step": 80000
    },
    {
      "epoch": 8.610483263373157,
      "eval_accuracy": 0.3914083380878026,
      "eval_loss": 3.3225831985473633,
      "eval_runtime": 184.6455,
      "eval_samples_per_second": 97.544,
      "eval_steps_per_second": 6.098,
      "step": 80000
    },
    {
      "epoch": 8.615864815412765,
      "grad_norm": 0.9362099766731262,
      "learning_rate": 8.33380023704342e-05,
      "loss": 3.186,
      "step": 80050
    },
    {
      "epoch": 8.621246367452374,
      "grad_norm": 0.8790286779403687,
      "learning_rate": 8.30147613403728e-05,
      "loss": 3.1654,
      "step": 80100
    },
    {
      "epoch": 8.626627919491982,
      "grad_norm": 0.8834596872329712,
      "learning_rate": 8.269152031031138e-05,
      "loss": 3.1895,
      "step": 80150
    },
    {
      "epoch": 8.632009471531589,
      "grad_norm": 0.8374781012535095,
      "learning_rate": 8.236827928024996e-05,
      "loss": 3.181,
      "step": 80200
    },
    {
      "epoch": 8.637391023571197,
      "grad_norm": 0.9129075407981873,
      "learning_rate": 8.204503825018854e-05,
      "loss": 3.1877,
      "step": 80250
    },
    {
      "epoch": 8.642772575610806,
      "grad_norm": 0.9645001292228699,
      "learning_rate": 8.172179722012714e-05,
      "loss": 3.1876,
      "step": 80300
    },
    {
      "epoch": 8.648154127650415,
      "grad_norm": 0.8795619010925293,
      "learning_rate": 8.13985561900657e-05,
      "loss": 3.1889,
      "step": 80350
    },
    {
      "epoch": 8.653535679690023,
      "grad_norm": 0.8914580941200256,
      "learning_rate": 8.10753151600043e-05,
      "loss": 3.1922,
      "step": 80400
    },
    {
      "epoch": 8.658917231729632,
      "grad_norm": 0.8828052878379822,
      "learning_rate": 8.07520741299429e-05,
      "loss": 3.178,
      "step": 80450
    },
    {
      "epoch": 8.664298783769238,
      "grad_norm": 0.8768082857131958,
      "learning_rate": 8.042883309988148e-05,
      "loss": 3.1827,
      "step": 80500
    },
    {
      "epoch": 8.669680335808847,
      "grad_norm": 0.8295359015464783,
      "learning_rate": 8.010559206982006e-05,
      "loss": 3.183,
      "step": 80550
    },
    {
      "epoch": 8.675061887848456,
      "grad_norm": 0.8298890590667725,
      "learning_rate": 7.978235103975864e-05,
      "loss": 3.1791,
      "step": 80600
    },
    {
      "epoch": 8.680443439888064,
      "grad_norm": 0.8362737894058228,
      "learning_rate": 7.945911000969723e-05,
      "loss": 3.1816,
      "step": 80650
    },
    {
      "epoch": 8.685824991927673,
      "grad_norm": 0.8717547059059143,
      "learning_rate": 7.91358689796358e-05,
      "loss": 3.1963,
      "step": 80700
    },
    {
      "epoch": 8.69120654396728,
      "grad_norm": 0.8629084825515747,
      "learning_rate": 7.88126279495744e-05,
      "loss": 3.2005,
      "step": 80750
    },
    {
      "epoch": 8.696588096006888,
      "grad_norm": 0.891313374042511,
      "learning_rate": 7.848938691951298e-05,
      "loss": 3.1822,
      "step": 80800
    },
    {
      "epoch": 8.701969648046497,
      "grad_norm": 0.9416958093643188,
      "learning_rate": 7.816614588945156e-05,
      "loss": 3.1739,
      "step": 80850
    },
    {
      "epoch": 8.707351200086105,
      "grad_norm": 0.9029965400695801,
      "learning_rate": 7.784290485939014e-05,
      "loss": 3.1881,
      "step": 80900
    },
    {
      "epoch": 8.712732752125714,
      "grad_norm": 0.8759965300559998,
      "learning_rate": 7.751966382932873e-05,
      "loss": 3.1905,
      "step": 80950
    },
    {
      "epoch": 8.718114304165322,
      "grad_norm": 0.8635278344154358,
      "learning_rate": 7.719642279926731e-05,
      "loss": 3.1891,
      "step": 81000
    },
    {
      "epoch": 8.718114304165322,
      "eval_accuracy": 0.3921397893285697,
      "eval_loss": 3.3176803588867188,
      "eval_runtime": 184.7567,
      "eval_samples_per_second": 97.485,
      "eval_steps_per_second": 6.095,
      "step": 81000
    },
    {
      "epoch": 8.723495856204929,
      "grad_norm": 0.9018222093582153,
      "learning_rate": 7.68731817692059e-05,
      "loss": 3.1802,
      "step": 81050
    },
    {
      "epoch": 8.728877408244538,
      "grad_norm": 0.8515042066574097,
      "learning_rate": 7.654994073914448e-05,
      "loss": 3.1897,
      "step": 81100
    },
    {
      "epoch": 8.734258960284146,
      "grad_norm": 0.8561316132545471,
      "learning_rate": 7.622669970908307e-05,
      "loss": 3.1992,
      "step": 81150
    },
    {
      "epoch": 8.739640512323755,
      "grad_norm": 0.8872911930084229,
      "learning_rate": 7.590345867902164e-05,
      "loss": 3.1834,
      "step": 81200
    },
    {
      "epoch": 8.745022064363363,
      "grad_norm": 0.8783745765686035,
      "learning_rate": 7.558021764896023e-05,
      "loss": 3.1589,
      "step": 81250
    },
    {
      "epoch": 8.75040361640297,
      "grad_norm": 0.8908545970916748,
      "learning_rate": 7.525697661889883e-05,
      "loss": 3.1957,
      "step": 81300
    },
    {
      "epoch": 8.755785168442578,
      "grad_norm": 0.8940176963806152,
      "learning_rate": 7.493373558883741e-05,
      "loss": 3.1899,
      "step": 81350
    },
    {
      "epoch": 8.761166720482187,
      "grad_norm": 0.8686371445655823,
      "learning_rate": 7.461049455877599e-05,
      "loss": 3.2028,
      "step": 81400
    },
    {
      "epoch": 8.766548272521796,
      "grad_norm": 0.8972199559211731,
      "learning_rate": 7.428725352871457e-05,
      "loss": 3.1772,
      "step": 81450
    },
    {
      "epoch": 8.771929824561404,
      "grad_norm": 0.9102570414543152,
      "learning_rate": 7.396401249865315e-05,
      "loss": 3.181,
      "step": 81500
    },
    {
      "epoch": 8.777311376601011,
      "grad_norm": 0.8299551606178284,
      "learning_rate": 7.364077146859173e-05,
      "loss": 3.1861,
      "step": 81550
    },
    {
      "epoch": 8.78269292864062,
      "grad_norm": 0.852982223033905,
      "learning_rate": 7.331753043853033e-05,
      "loss": 3.1815,
      "step": 81600
    },
    {
      "epoch": 8.788074480680228,
      "grad_norm": 0.8657204508781433,
      "learning_rate": 7.299428940846891e-05,
      "loss": 3.1924,
      "step": 81650
    },
    {
      "epoch": 8.793456032719837,
      "grad_norm": 0.8692888021469116,
      "learning_rate": 7.267104837840749e-05,
      "loss": 3.1805,
      "step": 81700
    },
    {
      "epoch": 8.798837584759445,
      "grad_norm": 0.8677384853363037,
      "learning_rate": 7.234780734834607e-05,
      "loss": 3.2016,
      "step": 81750
    },
    {
      "epoch": 8.804219136799054,
      "grad_norm": 0.9234066605567932,
      "learning_rate": 7.202456631828465e-05,
      "loss": 3.1823,
      "step": 81800
    },
    {
      "epoch": 8.80960068883866,
      "grad_norm": 0.8415629267692566,
      "learning_rate": 7.170132528822325e-05,
      "loss": 3.16,
      "step": 81850
    },
    {
      "epoch": 8.814982240878269,
      "grad_norm": 0.8653513193130493,
      "learning_rate": 7.137808425816183e-05,
      "loss": 3.1854,
      "step": 81900
    },
    {
      "epoch": 8.820363792917878,
      "grad_norm": 0.9204857349395752,
      "learning_rate": 7.105484322810041e-05,
      "loss": 3.1788,
      "step": 81950
    },
    {
      "epoch": 8.825745344957486,
      "grad_norm": 0.9475948810577393,
      "learning_rate": 7.073160219803899e-05,
      "loss": 3.1655,
      "step": 82000
    },
    {
      "epoch": 8.825745344957486,
      "eval_accuracy": 0.3923214569569955,
      "eval_loss": 3.316098928451538,
      "eval_runtime": 184.8571,
      "eval_samples_per_second": 97.432,
      "eval_steps_per_second": 6.091,
      "step": 82000
    },
    {
      "epoch": 8.831126896997095,
      "grad_norm": 0.9237799048423767,
      "learning_rate": 7.040836116797757e-05,
      "loss": 3.1975,
      "step": 82050
    },
    {
      "epoch": 8.836508449036701,
      "grad_norm": 0.8860271573066711,
      "learning_rate": 7.008512013791617e-05,
      "loss": 3.1737,
      "step": 82100
    },
    {
      "epoch": 8.84189000107631,
      "grad_norm": 0.8764588832855225,
      "learning_rate": 6.976187910785475e-05,
      "loss": 3.1814,
      "step": 82150
    },
    {
      "epoch": 8.847271553115919,
      "grad_norm": 0.8662134408950806,
      "learning_rate": 6.943863807779334e-05,
      "loss": 3.1879,
      "step": 82200
    },
    {
      "epoch": 8.852653105155527,
      "grad_norm": 0.8557435274124146,
      "learning_rate": 6.911539704773192e-05,
      "loss": 3.1715,
      "step": 82250
    },
    {
      "epoch": 8.858034657195136,
      "grad_norm": 0.8877943158149719,
      "learning_rate": 6.87921560176705e-05,
      "loss": 3.1928,
      "step": 82300
    },
    {
      "epoch": 8.863416209234742,
      "grad_norm": 0.9014151692390442,
      "learning_rate": 6.846891498760909e-05,
      "loss": 3.1795,
      "step": 82350
    },
    {
      "epoch": 8.868797761274351,
      "grad_norm": 0.9253289699554443,
      "learning_rate": 6.814567395754767e-05,
      "loss": 3.1607,
      "step": 82400
    },
    {
      "epoch": 8.87417931331396,
      "grad_norm": 0.8912492394447327,
      "learning_rate": 6.782243292748626e-05,
      "loss": 3.1784,
      "step": 82450
    },
    {
      "epoch": 8.879560865353568,
      "grad_norm": 0.9154777526855469,
      "learning_rate": 6.750565671802607e-05,
      "loss": 3.192,
      "step": 82500
    },
    {
      "epoch": 8.884942417393177,
      "grad_norm": 0.9307251572608948,
      "learning_rate": 6.718241568796465e-05,
      "loss": 3.1775,
      "step": 82550
    },
    {
      "epoch": 8.890323969432785,
      "grad_norm": 0.8818643689155579,
      "learning_rate": 6.685917465790323e-05,
      "loss": 3.166,
      "step": 82600
    },
    {
      "epoch": 8.895705521472392,
      "grad_norm": 0.8821066617965698,
      "learning_rate": 6.653593362784182e-05,
      "loss": 3.1702,
      "step": 82650
    },
    {
      "epoch": 8.901087073512,
      "grad_norm": 0.9354448318481445,
      "learning_rate": 6.62126925977804e-05,
      "loss": 3.171,
      "step": 82700
    },
    {
      "epoch": 8.906468625551609,
      "grad_norm": 0.8969210982322693,
      "learning_rate": 6.588945156771899e-05,
      "loss": 3.1851,
      "step": 82750
    },
    {
      "epoch": 8.911850177591218,
      "grad_norm": 0.8801003098487854,
      "learning_rate": 6.556621053765757e-05,
      "loss": 3.1559,
      "step": 82800
    },
    {
      "epoch": 8.917231729630826,
      "grad_norm": 0.914297878742218,
      "learning_rate": 6.524296950759615e-05,
      "loss": 3.1676,
      "step": 82850
    },
    {
      "epoch": 8.922613281670433,
      "grad_norm": 0.8998667001724243,
      "learning_rate": 6.491972847753474e-05,
      "loss": 3.1729,
      "step": 82900
    },
    {
      "epoch": 8.927994833710041,
      "grad_norm": 0.911956250667572,
      "learning_rate": 6.459648744747333e-05,
      "loss": 3.1796,
      "step": 82950
    },
    {
      "epoch": 8.93337638574965,
      "grad_norm": 0.9175410270690918,
      "learning_rate": 6.427324641741192e-05,
      "loss": 3.1707,
      "step": 83000
    },
    {
      "epoch": 8.93337638574965,
      "eval_accuracy": 0.3926996776594777,
      "eval_loss": 3.3129827976226807,
      "eval_runtime": 184.4576,
      "eval_samples_per_second": 97.643,
      "eval_steps_per_second": 6.104,
      "step": 83000
    },
    {
      "epoch": 8.938757937789259,
      "grad_norm": 0.8617079257965088,
      "learning_rate": 6.39500053873505e-05,
      "loss": 3.1874,
      "step": 83050
    },
    {
      "epoch": 8.944139489828867,
      "grad_norm": 0.8827839493751526,
      "learning_rate": 6.362676435728908e-05,
      "loss": 3.174,
      "step": 83100
    },
    {
      "epoch": 8.949521041868476,
      "grad_norm": 0.8599488139152527,
      "learning_rate": 6.330352332722766e-05,
      "loss": 3.2049,
      "step": 83150
    },
    {
      "epoch": 8.954902593908082,
      "grad_norm": 0.905812680721283,
      "learning_rate": 6.298028229716624e-05,
      "loss": 3.1845,
      "step": 83200
    },
    {
      "epoch": 8.960284145947691,
      "grad_norm": 0.8685833215713501,
      "learning_rate": 6.265704126710484e-05,
      "loss": 3.1687,
      "step": 83250
    },
    {
      "epoch": 8.9656656979873,
      "grad_norm": 0.90411376953125,
      "learning_rate": 6.233380023704342e-05,
      "loss": 3.1906,
      "step": 83300
    },
    {
      "epoch": 8.971047250026908,
      "grad_norm": 0.8720734119415283,
      "learning_rate": 6.2010559206982e-05,
      "loss": 3.1915,
      "step": 83350
    },
    {
      "epoch": 8.976428802066517,
      "grad_norm": 0.8964160680770874,
      "learning_rate": 6.168731817692058e-05,
      "loss": 3.1848,
      "step": 83400
    },
    {
      "epoch": 8.981810354106123,
      "grad_norm": 0.9113012552261353,
      "learning_rate": 6.136407714685916e-05,
      "loss": 3.1753,
      "step": 83450
    },
    {
      "epoch": 8.987191906145732,
      "grad_norm": 0.9036979079246521,
      "learning_rate": 6.104083611679776e-05,
      "loss": 3.1912,
      "step": 83500
    },
    {
      "epoch": 8.99257345818534,
      "grad_norm": 0.9076685905456543,
      "learning_rate": 6.071759508673634e-05,
      "loss": 3.1701,
      "step": 83550
    },
    {
      "epoch": 8.997955010224949,
      "grad_norm": 0.9387393593788147,
      "learning_rate": 6.039435405667492e-05,
      "loss": 3.1792,
      "step": 83600
    },
    {
      "epoch": 9.003336562264558,
      "grad_norm": 0.8677722215652466,
      "learning_rate": 6.007111302661351e-05,
      "loss": 3.1325,
      "step": 83650
    },
    {
      "epoch": 9.008718114304166,
      "grad_norm": 0.898324191570282,
      "learning_rate": 5.974787199655209e-05,
      "loss": 3.1276,
      "step": 83700
    },
    {
      "epoch": 9.014099666343773,
      "grad_norm": 0.8768854737281799,
      "learning_rate": 5.942463096649068e-05,
      "loss": 3.1129,
      "step": 83750
    },
    {
      "epoch": 9.019481218383381,
      "grad_norm": 0.8867700099945068,
      "learning_rate": 5.910138993642926e-05,
      "loss": 3.1221,
      "step": 83800
    },
    {
      "epoch": 9.02486277042299,
      "grad_norm": 0.8973957300186157,
      "learning_rate": 5.877814890636784e-05,
      "loss": 3.1166,
      "step": 83850
    },
    {
      "epoch": 9.030244322462599,
      "grad_norm": 0.8828250765800476,
      "learning_rate": 5.845490787630643e-05,
      "loss": 3.1444,
      "step": 83900
    },
    {
      "epoch": 9.035625874502207,
      "grad_norm": 0.8597937226295471,
      "learning_rate": 5.813166684624501e-05,
      "loss": 3.1197,
      "step": 83950
    },
    {
      "epoch": 9.041007426541814,
      "grad_norm": 0.8845257759094238,
      "learning_rate": 5.7808425816183596e-05,
      "loss": 3.1265,
      "step": 84000
    },
    {
      "epoch": 9.041007426541814,
      "eval_accuracy": 0.3927993123576036,
      "eval_loss": 3.3138818740844727,
      "eval_runtime": 184.539,
      "eval_samples_per_second": 97.6,
      "eval_steps_per_second": 6.102,
      "step": 84000
    },
    {
      "epoch": 9.046388978581422,
      "grad_norm": 0.8949583172798157,
      "learning_rate": 5.748518478612218e-05,
      "loss": 3.1287,
      "step": 84050
    },
    {
      "epoch": 9.051770530621031,
      "grad_norm": 0.8685891628265381,
      "learning_rate": 5.716194375606076e-05,
      "loss": 3.1128,
      "step": 84100
    },
    {
      "epoch": 9.05715208266064,
      "grad_norm": 0.863949179649353,
      "learning_rate": 5.6838702725999346e-05,
      "loss": 3.1296,
      "step": 84150
    },
    {
      "epoch": 9.062533634700248,
      "grad_norm": 0.9483818411827087,
      "learning_rate": 5.651546169593793e-05,
      "loss": 3.1382,
      "step": 84200
    },
    {
      "epoch": 9.067915186739857,
      "grad_norm": 0.898429811000824,
      "learning_rate": 5.619222066587652e-05,
      "loss": 3.1151,
      "step": 84250
    },
    {
      "epoch": 9.073296738779463,
      "grad_norm": 0.9140533208847046,
      "learning_rate": 5.58689796358151e-05,
      "loss": 3.1487,
      "step": 84300
    },
    {
      "epoch": 9.078678290819072,
      "grad_norm": 0.9005162119865417,
      "learning_rate": 5.554573860575369e-05,
      "loss": 3.1276,
      "step": 84350
    },
    {
      "epoch": 9.08405984285868,
      "grad_norm": 0.9215816855430603,
      "learning_rate": 5.522249757569227e-05,
      "loss": 3.1322,
      "step": 84400
    },
    {
      "epoch": 9.089441394898289,
      "grad_norm": 0.9173760414123535,
      "learning_rate": 5.489925654563085e-05,
      "loss": 3.1239,
      "step": 84450
    },
    {
      "epoch": 9.094822946937898,
      "grad_norm": 0.8493995070457458,
      "learning_rate": 5.457601551556944e-05,
      "loss": 3.1223,
      "step": 84500
    },
    {
      "epoch": 9.100204498977504,
      "grad_norm": 0.8942146897315979,
      "learning_rate": 5.425277448550802e-05,
      "loss": 3.144,
      "step": 84550
    },
    {
      "epoch": 9.105586051017113,
      "grad_norm": 0.9167282581329346,
      "learning_rate": 5.392953345544661e-05,
      "loss": 3.1309,
      "step": 84600
    },
    {
      "epoch": 9.110967603056721,
      "grad_norm": 0.9377874732017517,
      "learning_rate": 5.360629242538519e-05,
      "loss": 3.1163,
      "step": 84650
    },
    {
      "epoch": 9.11634915509633,
      "grad_norm": 0.8942631483078003,
      "learning_rate": 5.328305139532377e-05,
      "loss": 3.112,
      "step": 84700
    },
    {
      "epoch": 9.121730707135939,
      "grad_norm": 0.8881064653396606,
      "learning_rate": 5.295981036526236e-05,
      "loss": 3.121,
      "step": 84750
    },
    {
      "epoch": 9.127112259175545,
      "grad_norm": 0.9246423244476318,
      "learning_rate": 5.263656933520094e-05,
      "loss": 3.1307,
      "step": 84800
    },
    {
      "epoch": 9.132493811215154,
      "grad_norm": 0.8789636492729187,
      "learning_rate": 5.231332830513953e-05,
      "loss": 3.1251,
      "step": 84850
    },
    {
      "epoch": 9.137875363254762,
      "grad_norm": 0.9263179302215576,
      "learning_rate": 5.199008727507811e-05,
      "loss": 3.1385,
      "step": 84900
    },
    {
      "epoch": 9.143256915294371,
      "grad_norm": 0.9350448846817017,
      "learning_rate": 5.166684624501669e-05,
      "loss": 3.1264,
      "step": 84950
    },
    {
      "epoch": 9.14863846733398,
      "grad_norm": 0.9025295376777649,
      "learning_rate": 5.1343605214955286e-05,
      "loss": 3.1258,
      "step": 85000
    },
    {
      "epoch": 9.14863846733398,
      "eval_accuracy": 0.39295207831787077,
      "eval_loss": 3.3131446838378906,
      "eval_runtime": 184.5599,
      "eval_samples_per_second": 97.589,
      "eval_steps_per_second": 6.101,
      "step": 85000
    },
    {
      "epoch": 9.154020019373588,
      "grad_norm": 0.9414235949516296,
      "learning_rate": 5.102036418489387e-05,
      "loss": 3.1233,
      "step": 85050
    },
    {
      "epoch": 9.159401571413195,
      "grad_norm": 0.8915159106254578,
      "learning_rate": 5.0697123154832455e-05,
      "loss": 3.1198,
      "step": 85100
    },
    {
      "epoch": 9.164783123452803,
      "grad_norm": 0.9033054709434509,
      "learning_rate": 5.0373882124771036e-05,
      "loss": 3.1274,
      "step": 85150
    },
    {
      "epoch": 9.170164675492412,
      "grad_norm": 0.8779189586639404,
      "learning_rate": 5.005064109470962e-05,
      "loss": 3.1229,
      "step": 85200
    },
    {
      "epoch": 9.17554622753202,
      "grad_norm": 0.8858757019042969,
      "learning_rate": 4.9727400064648205e-05,
      "loss": 3.1166,
      "step": 85250
    },
    {
      "epoch": 9.180927779571629,
      "grad_norm": 0.8739831447601318,
      "learning_rate": 4.9404159034586786e-05,
      "loss": 3.1356,
      "step": 85300
    },
    {
      "epoch": 9.186309331611236,
      "grad_norm": 0.9114834666252136,
      "learning_rate": 4.9080918004525374e-05,
      "loss": 3.1189,
      "step": 85350
    },
    {
      "epoch": 9.191690883650844,
      "grad_norm": 0.9239445924758911,
      "learning_rate": 4.876414179506518e-05,
      "loss": 3.145,
      "step": 85400
    },
    {
      "epoch": 9.197072435690453,
      "grad_norm": 0.9356056451797485,
      "learning_rate": 4.844090076500377e-05,
      "loss": 3.1225,
      "step": 85450
    },
    {
      "epoch": 9.202453987730062,
      "grad_norm": 0.9145978689193726,
      "learning_rate": 4.811765973494235e-05,
      "loss": 3.1418,
      "step": 85500
    },
    {
      "epoch": 9.20783553976967,
      "grad_norm": 0.9663805961608887,
      "learning_rate": 4.779441870488094e-05,
      "loss": 3.1302,
      "step": 85550
    },
    {
      "epoch": 9.213217091809279,
      "grad_norm": 0.8993372321128845,
      "learning_rate": 4.747117767481952e-05,
      "loss": 3.1292,
      "step": 85600
    },
    {
      "epoch": 9.218598643848885,
      "grad_norm": 0.8745630979537964,
      "learning_rate": 4.71479366447581e-05,
      "loss": 3.1227,
      "step": 85650
    },
    {
      "epoch": 9.223980195888494,
      "grad_norm": 0.86039137840271,
      "learning_rate": 4.682469561469669e-05,
      "loss": 3.1389,
      "step": 85700
    },
    {
      "epoch": 9.229361747928102,
      "grad_norm": 0.9199390411376953,
      "learning_rate": 4.650145458463527e-05,
      "loss": 3.1364,
      "step": 85750
    },
    {
      "epoch": 9.234743299967711,
      "grad_norm": 0.9464226365089417,
      "learning_rate": 4.6178213554573856e-05,
      "loss": 3.1197,
      "step": 85800
    },
    {
      "epoch": 9.24012485200732,
      "grad_norm": 0.90545254945755,
      "learning_rate": 4.585497252451244e-05,
      "loss": 3.1347,
      "step": 85850
    },
    {
      "epoch": 9.245506404046926,
      "grad_norm": 0.943743109703064,
      "learning_rate": 4.553173149445102e-05,
      "loss": 3.1428,
      "step": 85900
    },
    {
      "epoch": 9.250887956086535,
      "grad_norm": 0.9059910774230957,
      "learning_rate": 4.520849046438961e-05,
      "loss": 3.1222,
      "step": 85950
    },
    {
      "epoch": 9.256269508126143,
      "grad_norm": 0.9065704941749573,
      "learning_rate": 4.488524943432819e-05,
      "loss": 3.1525,
      "step": 86000
    },
    {
      "epoch": 9.256269508126143,
      "eval_accuracy": 0.39329553009625234,
      "eval_loss": 3.3110220432281494,
      "eval_runtime": 184.9835,
      "eval_samples_per_second": 97.365,
      "eval_steps_per_second": 6.087,
      "step": 86000
    },
    {
      "epoch": 9.261651060165752,
      "grad_norm": 0.9051395654678345,
      "learning_rate": 4.456200840426678e-05,
      "loss": 3.1265,
      "step": 86050
    },
    {
      "epoch": 9.26703261220536,
      "grad_norm": 0.8704345226287842,
      "learning_rate": 4.4238767374205363e-05,
      "loss": 3.1381,
      "step": 86100
    },
    {
      "epoch": 9.272414164244967,
      "grad_norm": 0.9406769871711731,
      "learning_rate": 4.3915526344143945e-05,
      "loss": 3.1278,
      "step": 86150
    },
    {
      "epoch": 9.277795716284576,
      "grad_norm": 0.858677864074707,
      "learning_rate": 4.359228531408253e-05,
      "loss": 3.1111,
      "step": 86200
    },
    {
      "epoch": 9.283177268324184,
      "grad_norm": 0.9546279311180115,
      "learning_rate": 4.3269044284021114e-05,
      "loss": 3.1253,
      "step": 86250
    },
    {
      "epoch": 9.288558820363793,
      "grad_norm": 0.8848311901092529,
      "learning_rate": 4.29458032539597e-05,
      "loss": 3.1237,
      "step": 86300
    },
    {
      "epoch": 9.293940372403402,
      "grad_norm": 0.9586207270622253,
      "learning_rate": 4.262256222389828e-05,
      "loss": 3.1308,
      "step": 86350
    },
    {
      "epoch": 9.29932192444301,
      "grad_norm": 0.876517653465271,
      "learning_rate": 4.229932119383687e-05,
      "loss": 3.1307,
      "step": 86400
    },
    {
      "epoch": 9.304703476482617,
      "grad_norm": 0.8859730958938599,
      "learning_rate": 4.197608016377545e-05,
      "loss": 3.1464,
      "step": 86450
    },
    {
      "epoch": 9.310085028522225,
      "grad_norm": 0.8714919090270996,
      "learning_rate": 4.165283913371403e-05,
      "loss": 3.1382,
      "step": 86500
    },
    {
      "epoch": 9.315466580561834,
      "grad_norm": 0.936630368232727,
      "learning_rate": 4.132959810365262e-05,
      "loss": 3.1208,
      "step": 86550
    },
    {
      "epoch": 9.320848132601443,
      "grad_norm": 0.8881075382232666,
      "learning_rate": 4.10063570735912e-05,
      "loss": 3.1199,
      "step": 86600
    },
    {
      "epoch": 9.326229684641051,
      "grad_norm": 0.9490821361541748,
      "learning_rate": 4.068311604352979e-05,
      "loss": 3.1226,
      "step": 86650
    },
    {
      "epoch": 9.331611236680658,
      "grad_norm": 0.9327830672264099,
      "learning_rate": 4.035987501346837e-05,
      "loss": 3.1466,
      "step": 86700
    },
    {
      "epoch": 9.336992788720266,
      "grad_norm": 0.9075290560722351,
      "learning_rate": 4.003663398340695e-05,
      "loss": 3.1365,
      "step": 86750
    },
    {
      "epoch": 9.342374340759875,
      "grad_norm": 0.8921899199485779,
      "learning_rate": 3.9713392953345546e-05,
      "loss": 3.1491,
      "step": 86800
    },
    {
      "epoch": 9.347755892799483,
      "grad_norm": 0.9279853105545044,
      "learning_rate": 3.939015192328413e-05,
      "loss": 3.1321,
      "step": 86850
    },
    {
      "epoch": 9.353137444839092,
      "grad_norm": 0.9148728847503662,
      "learning_rate": 3.9066910893222715e-05,
      "loss": 3.1337,
      "step": 86900
    },
    {
      "epoch": 9.3585189968787,
      "grad_norm": 0.907360315322876,
      "learning_rate": 3.8743669863161296e-05,
      "loss": 3.1314,
      "step": 86950
    },
    {
      "epoch": 9.363900548918307,
      "grad_norm": 0.9280900359153748,
      "learning_rate": 3.842042883309988e-05,
      "loss": 3.1264,
      "step": 87000
    },
    {
      "epoch": 9.363900548918307,
      "eval_accuracy": 0.39339614267036854,
      "eval_loss": 3.3104796409606934,
      "eval_runtime": 184.6531,
      "eval_samples_per_second": 97.54,
      "eval_steps_per_second": 6.098,
      "step": 87000
    },
    {
      "epoch": 9.369282100957916,
      "grad_norm": 0.8657009601593018,
      "learning_rate": 3.8097187803038465e-05,
      "loss": 3.1423,
      "step": 87050
    },
    {
      "epoch": 9.374663652997524,
      "grad_norm": 0.8895671367645264,
      "learning_rate": 3.7773946772977047e-05,
      "loss": 3.11,
      "step": 87100
    },
    {
      "epoch": 9.380045205037133,
      "grad_norm": 0.8989640474319458,
      "learning_rate": 3.745070574291563e-05,
      "loss": 3.1225,
      "step": 87150
    },
    {
      "epoch": 9.385426757076742,
      "grad_norm": 0.9100672006607056,
      "learning_rate": 3.7127464712854216e-05,
      "loss": 3.1423,
      "step": 87200
    },
    {
      "epoch": 9.390808309116348,
      "grad_norm": 0.883725106716156,
      "learning_rate": 3.6804223682792803e-05,
      "loss": 3.1252,
      "step": 87250
    },
    {
      "epoch": 9.396189861155957,
      "grad_norm": 0.9332802891731262,
      "learning_rate": 3.6480982652731385e-05,
      "loss": 3.1588,
      "step": 87300
    },
    {
      "epoch": 9.401571413195565,
      "grad_norm": 0.878340482711792,
      "learning_rate": 3.6157741622669966e-05,
      "loss": 3.1199,
      "step": 87350
    },
    {
      "epoch": 9.406952965235174,
      "grad_norm": 0.8933216333389282,
      "learning_rate": 3.5834500592608554e-05,
      "loss": 3.1312,
      "step": 87400
    },
    {
      "epoch": 9.412334517274783,
      "grad_norm": 0.9225236177444458,
      "learning_rate": 3.5511259562547135e-05,
      "loss": 3.1305,
      "step": 87450
    },
    {
      "epoch": 9.417716069314391,
      "grad_norm": 0.8944328427314758,
      "learning_rate": 3.519448335308695e-05,
      "loss": 3.1197,
      "step": 87500
    },
    {
      "epoch": 9.423097621353998,
      "grad_norm": 0.8815526366233826,
      "learning_rate": 3.4871242323025536e-05,
      "loss": 3.1282,
      "step": 87550
    },
    {
      "epoch": 9.428479173393606,
      "grad_norm": 0.947081446647644,
      "learning_rate": 3.454800129296412e-05,
      "loss": 3.1075,
      "step": 87600
    },
    {
      "epoch": 9.433860725433215,
      "grad_norm": 0.9011362791061401,
      "learning_rate": 3.42247602629027e-05,
      "loss": 3.1441,
      "step": 87650
    },
    {
      "epoch": 9.439242277472824,
      "grad_norm": 0.8931843042373657,
      "learning_rate": 3.3901519232841286e-05,
      "loss": 3.1562,
      "step": 87700
    },
    {
      "epoch": 9.444623829512432,
      "grad_norm": 0.8832892179489136,
      "learning_rate": 3.3578278202779874e-05,
      "loss": 3.1324,
      "step": 87750
    },
    {
      "epoch": 9.450005381552039,
      "grad_norm": 0.9089193344116211,
      "learning_rate": 3.3255037172718455e-05,
      "loss": 3.1315,
      "step": 87800
    },
    {
      "epoch": 9.455386933591647,
      "grad_norm": 1.5505813360214233,
      "learning_rate": 3.2931796142657036e-05,
      "loss": 3.1245,
      "step": 87850
    },
    {
      "epoch": 9.460768485631256,
      "grad_norm": 0.8742733001708984,
      "learning_rate": 3.2608555112595624e-05,
      "loss": 3.1338,
      "step": 87900
    },
    {
      "epoch": 9.466150037670864,
      "grad_norm": 0.9426006078720093,
      "learning_rate": 3.2285314082534205e-05,
      "loss": 3.1198,
      "step": 87950
    },
    {
      "epoch": 9.471531589710473,
      "grad_norm": 0.9033601880073547,
      "learning_rate": 3.196207305247279e-05,
      "loss": 3.1522,
      "step": 88000
    },
    {
      "epoch": 9.471531589710473,
      "eval_accuracy": 0.39373003299462245,
      "eval_loss": 3.30660080909729,
      "eval_runtime": 184.6201,
      "eval_samples_per_second": 97.557,
      "eval_steps_per_second": 6.099,
      "step": 88000
    },
    {
      "epoch": 9.476913141750082,
      "grad_norm": 0.8803892731666565,
      "learning_rate": 3.1638832022411374e-05,
      "loss": 3.1347,
      "step": 88050
    },
    {
      "epoch": 9.482294693789688,
      "grad_norm": 0.9048923254013062,
      "learning_rate": 3.1315590992349955e-05,
      "loss": 3.1199,
      "step": 88100
    },
    {
      "epoch": 9.487676245829297,
      "grad_norm": 0.9061964154243469,
      "learning_rate": 3.099234996228854e-05,
      "loss": 3.1378,
      "step": 88150
    },
    {
      "epoch": 9.493057797868905,
      "grad_norm": 0.9181330800056458,
      "learning_rate": 3.066910893222713e-05,
      "loss": 3.131,
      "step": 88200
    },
    {
      "epoch": 9.498439349908514,
      "grad_norm": 0.9095847010612488,
      "learning_rate": 3.0345867902165712e-05,
      "loss": 3.1291,
      "step": 88250
    },
    {
      "epoch": 9.503820901948123,
      "grad_norm": 0.9039878249168396,
      "learning_rate": 3.00226268721043e-05,
      "loss": 3.1375,
      "step": 88300
    },
    {
      "epoch": 9.50920245398773,
      "grad_norm": 0.9264708757400513,
      "learning_rate": 2.9699385842042878e-05,
      "loss": 3.1186,
      "step": 88350
    },
    {
      "epoch": 9.514584006027338,
      "grad_norm": 0.9030917882919312,
      "learning_rate": 2.9376144811981465e-05,
      "loss": 3.1325,
      "step": 88400
    },
    {
      "epoch": 9.519965558066946,
      "grad_norm": 0.9015549421310425,
      "learning_rate": 2.9059368602521275e-05,
      "loss": 3.1507,
      "step": 88450
    },
    {
      "epoch": 9.525347110106555,
      "grad_norm": 0.9642078280448914,
      "learning_rate": 2.8736127572459863e-05,
      "loss": 3.1393,
      "step": 88500
    },
    {
      "epoch": 9.530728662146164,
      "grad_norm": 0.8946633338928223,
      "learning_rate": 2.8412886542398448e-05,
      "loss": 3.1269,
      "step": 88550
    },
    {
      "epoch": 9.536110214185772,
      "grad_norm": 0.9017453789710999,
      "learning_rate": 2.8089645512337032e-05,
      "loss": 3.1265,
      "step": 88600
    },
    {
      "epoch": 9.541491766225379,
      "grad_norm": 0.8839432001113892,
      "learning_rate": 2.7766404482275613e-05,
      "loss": 3.1254,
      "step": 88650
    },
    {
      "epoch": 9.546873318264987,
      "grad_norm": 0.9194327592849731,
      "learning_rate": 2.7443163452214198e-05,
      "loss": 3.1311,
      "step": 88700
    },
    {
      "epoch": 9.552254870304596,
      "grad_norm": 0.9012552499771118,
      "learning_rate": 2.7119922422152782e-05,
      "loss": 3.1302,
      "step": 88750
    },
    {
      "epoch": 9.557636422344205,
      "grad_norm": 0.894688606262207,
      "learning_rate": 2.6796681392091367e-05,
      "loss": 3.1381,
      "step": 88800
    },
    {
      "epoch": 9.563017974383813,
      "grad_norm": 0.8723885416984558,
      "learning_rate": 2.6473440362029955e-05,
      "loss": 3.1406,
      "step": 88850
    },
    {
      "epoch": 9.56839952642342,
      "grad_norm": 0.914779782295227,
      "learning_rate": 2.6150199331968536e-05,
      "loss": 3.1384,
      "step": 88900
    },
    {
      "epoch": 9.573781078463028,
      "grad_norm": 0.9142804145812988,
      "learning_rate": 2.582695830190712e-05,
      "loss": 3.1213,
      "step": 88950
    },
    {
      "epoch": 9.579162630502637,
      "grad_norm": 0.9292569756507874,
      "learning_rate": 2.5503717271845705e-05,
      "loss": 3.1493,
      "step": 89000
    },
    {
      "epoch": 9.579162630502637,
      "eval_accuracy": 0.39403013227076605,
      "eval_loss": 3.303948163986206,
      "eval_runtime": 184.7384,
      "eval_samples_per_second": 97.495,
      "eval_steps_per_second": 6.095,
      "step": 89000
    },
    {
      "epoch": 9.584544182542245,
      "grad_norm": 0.8625149130821228,
      "learning_rate": 2.518047624178429e-05,
      "loss": 3.1315,
      "step": 89050
    },
    {
      "epoch": 9.589925734581854,
      "grad_norm": 0.9564598798751831,
      "learning_rate": 2.4857235211722874e-05,
      "loss": 3.1362,
      "step": 89100
    },
    {
      "epoch": 9.59530728662146,
      "grad_norm": 0.9190493226051331,
      "learning_rate": 2.4533994181661455e-05,
      "loss": 3.12,
      "step": 89150
    },
    {
      "epoch": 9.60068883866107,
      "grad_norm": 0.8848850131034851,
      "learning_rate": 2.421075315160004e-05,
      "loss": 3.1241,
      "step": 89200
    },
    {
      "epoch": 9.606070390700678,
      "grad_norm": 1.0095571279525757,
      "learning_rate": 2.3887512121538624e-05,
      "loss": 3.145,
      "step": 89250
    },
    {
      "epoch": 9.611451942740286,
      "grad_norm": 0.8862415552139282,
      "learning_rate": 2.3564271091477212e-05,
      "loss": 3.1304,
      "step": 89300
    },
    {
      "epoch": 9.616833494779895,
      "grad_norm": 0.8857293725013733,
      "learning_rate": 2.3241030061415796e-05,
      "loss": 3.14,
      "step": 89350
    },
    {
      "epoch": 9.622215046819504,
      "grad_norm": 0.887195885181427,
      "learning_rate": 2.2917789031354377e-05,
      "loss": 3.1439,
      "step": 89400
    },
    {
      "epoch": 9.62759659885911,
      "grad_norm": 0.8865780234336853,
      "learning_rate": 2.2594548001292962e-05,
      "loss": 3.1397,
      "step": 89450
    },
    {
      "epoch": 9.632978150898719,
      "grad_norm": 0.9336400628089905,
      "learning_rate": 2.2271306971231546e-05,
      "loss": 3.1413,
      "step": 89500
    },
    {
      "epoch": 9.638359702938327,
      "grad_norm": 0.9491809606552124,
      "learning_rate": 2.194806594117013e-05,
      "loss": 3.1541,
      "step": 89550
    },
    {
      "epoch": 9.643741254977936,
      "grad_norm": 0.9103695154190063,
      "learning_rate": 2.1624824911108715e-05,
      "loss": 3.1467,
      "step": 89600
    },
    {
      "epoch": 9.649122807017545,
      "grad_norm": 0.9001594185829163,
      "learning_rate": 2.1301583881047296e-05,
      "loss": 3.1282,
      "step": 89650
    },
    {
      "epoch": 9.654504359057151,
      "grad_norm": 0.9728744029998779,
      "learning_rate": 2.097834285098588e-05,
      "loss": 3.1277,
      "step": 89700
    },
    {
      "epoch": 9.65988591109676,
      "grad_norm": 0.9372629523277283,
      "learning_rate": 2.065510182092447e-05,
      "loss": 3.1457,
      "step": 89750
    },
    {
      "epoch": 9.665267463136368,
      "grad_norm": 0.9052270650863647,
      "learning_rate": 2.0331860790863053e-05,
      "loss": 3.1316,
      "step": 89800
    },
    {
      "epoch": 9.670649015175977,
      "grad_norm": 0.9259636402130127,
      "learning_rate": 2.0008619760801638e-05,
      "loss": 3.1415,
      "step": 89850
    },
    {
      "epoch": 9.676030567215586,
      "grad_norm": 0.886457622051239,
      "learning_rate": 1.968537873074022e-05,
      "loss": 3.1369,
      "step": 89900
    },
    {
      "epoch": 9.681412119255192,
      "grad_norm": 0.9304108023643494,
      "learning_rate": 1.9362137700678803e-05,
      "loss": 3.1268,
      "step": 89950
    },
    {
      "epoch": 9.6867936712948,
      "grad_norm": 1.5604965686798096,
      "learning_rate": 1.9038896670617388e-05,
      "loss": 3.1222,
      "step": 90000
    },
    {
      "epoch": 9.6867936712948,
      "eval_accuracy": 0.3941440004971956,
      "eval_loss": 3.3028476238250732,
      "eval_runtime": 184.8637,
      "eval_samples_per_second": 97.429,
      "eval_steps_per_second": 6.091,
      "step": 90000
    },
    {
      "epoch": 9.69217522333441,
      "grad_norm": 0.9006509184837341,
      "learning_rate": 1.8715655640555972e-05,
      "loss": 3.1474,
      "step": 90050
    },
    {
      "epoch": 9.697556775374018,
      "grad_norm": 0.9110347032546997,
      "learning_rate": 1.8392414610494557e-05,
      "loss": 3.1244,
      "step": 90100
    },
    {
      "epoch": 9.702938327413626,
      "grad_norm": 0.9055619239807129,
      "learning_rate": 1.806917358043314e-05,
      "loss": 3.1425,
      "step": 90150
    },
    {
      "epoch": 9.708319879453235,
      "grad_norm": 0.864754319190979,
      "learning_rate": 1.7745932550371726e-05,
      "loss": 3.1258,
      "step": 90200
    },
    {
      "epoch": 9.713701431492842,
      "grad_norm": 0.9538236260414124,
      "learning_rate": 1.742269152031031e-05,
      "loss": 3.1427,
      "step": 90250
    },
    {
      "epoch": 9.71908298353245,
      "grad_norm": 0.8747223615646362,
      "learning_rate": 1.7099450490248895e-05,
      "loss": 3.1304,
      "step": 90300
    },
    {
      "epoch": 9.724464535572059,
      "grad_norm": 0.9114173650741577,
      "learning_rate": 1.677620946018748e-05,
      "loss": 3.1236,
      "step": 90350
    },
    {
      "epoch": 9.729846087611667,
      "grad_norm": 0.8518925905227661,
      "learning_rate": 1.6452968430126064e-05,
      "loss": 3.1251,
      "step": 90400
    },
    {
      "epoch": 9.735227639651276,
      "grad_norm": 0.9126296043395996,
      "learning_rate": 1.6129727400064645e-05,
      "loss": 3.1233,
      "step": 90450
    },
    {
      "epoch": 9.740609191690883,
      "grad_norm": 0.8861443996429443,
      "learning_rate": 1.580648637000323e-05,
      "loss": 3.1365,
      "step": 90500
    },
    {
      "epoch": 9.745990743730491,
      "grad_norm": 0.8633650541305542,
      "learning_rate": 1.5483245339941817e-05,
      "loss": 3.1337,
      "step": 90550
    },
    {
      "epoch": 9.7513722957701,
      "grad_norm": 0.951790452003479,
      "learning_rate": 1.5160004309880398e-05,
      "loss": 3.1405,
      "step": 90600
    },
    {
      "epoch": 9.756753847809708,
      "grad_norm": 0.8869327306747437,
      "learning_rate": 1.4836763279818985e-05,
      "loss": 3.1319,
      "step": 90650
    },
    {
      "epoch": 9.762135399849317,
      "grad_norm": 0.8825336694717407,
      "learning_rate": 1.4513522249757567e-05,
      "loss": 3.13,
      "step": 90700
    },
    {
      "epoch": 9.767516951888926,
      "grad_norm": 0.906427264213562,
      "learning_rate": 1.4190281219696152e-05,
      "loss": 3.1501,
      "step": 90750
    },
    {
      "epoch": 9.772898503928532,
      "grad_norm": 0.9699810147285461,
      "learning_rate": 1.3867040189634736e-05,
      "loss": 3.134,
      "step": 90800
    },
    {
      "epoch": 9.77828005596814,
      "grad_norm": 0.8849841952323914,
      "learning_rate": 1.3543799159573321e-05,
      "loss": 3.1346,
      "step": 90850
    },
    {
      "epoch": 9.78366160800775,
      "grad_norm": 0.9202008843421936,
      "learning_rate": 1.3220558129511905e-05,
      "loss": 3.1392,
      "step": 90900
    },
    {
      "epoch": 9.789043160047358,
      "grad_norm": 0.9104524254798889,
      "learning_rate": 1.2897317099450488e-05,
      "loss": 3.1384,
      "step": 90950
    },
    {
      "epoch": 9.794424712086967,
      "grad_norm": 0.9032725095748901,
      "learning_rate": 1.2574076069389073e-05,
      "loss": 3.1375,
      "step": 91000
    },
    {
      "epoch": 9.794424712086967,
      "eval_accuracy": 0.39456611696635485,
      "eval_loss": 3.3005428314208984,
      "eval_runtime": 184.4783,
      "eval_samples_per_second": 97.632,
      "eval_steps_per_second": 6.104,
      "step": 91000
    },
    {
      "epoch": 9.799806264126573,
      "grad_norm": 0.9103838205337524,
      "learning_rate": 1.2250835039327659e-05,
      "loss": 3.1205,
      "step": 91050
    },
    {
      "epoch": 9.805187816166182,
      "grad_norm": 0.8900352120399475,
      "learning_rate": 1.1927594009266242e-05,
      "loss": 3.1186,
      "step": 91100
    },
    {
      "epoch": 9.81056936820579,
      "grad_norm": 0.9095427393913269,
      "learning_rate": 1.1604352979204826e-05,
      "loss": 3.1354,
      "step": 91150
    },
    {
      "epoch": 9.815950920245399,
      "grad_norm": 0.910943329334259,
      "learning_rate": 1.1281111949143409e-05,
      "loss": 3.1302,
      "step": 91200
    },
    {
      "epoch": 9.821332472285007,
      "grad_norm": 0.899513304233551,
      "learning_rate": 1.0957870919081995e-05,
      "loss": 3.1446,
      "step": 91250
    },
    {
      "epoch": 9.826714024324616,
      "grad_norm": 0.9401593804359436,
      "learning_rate": 1.063462988902058e-05,
      "loss": 3.1492,
      "step": 91300
    },
    {
      "epoch": 9.832095576364223,
      "grad_norm": 0.9215407967567444,
      "learning_rate": 1.0311388858959162e-05,
      "loss": 3.1322,
      "step": 91350
    },
    {
      "epoch": 9.837477128403831,
      "grad_norm": 0.8816438317298889,
      "learning_rate": 9.988147828897747e-06,
      "loss": 3.1333,
      "step": 91400
    },
    {
      "epoch": 9.84285868044344,
      "grad_norm": 0.9174292683601379,
      "learning_rate": 9.66490679883633e-06,
      "loss": 3.1206,
      "step": 91450
    },
    {
      "epoch": 9.848240232483048,
      "grad_norm": 0.8972243666648865,
      "learning_rate": 9.341665768774916e-06,
      "loss": 3.1213,
      "step": 91500
    },
    {
      "epoch": 9.853621784522657,
      "grad_norm": 0.8909099102020264,
      "learning_rate": 9.018424738713499e-06,
      "loss": 3.1341,
      "step": 91550
    },
    {
      "epoch": 9.859003336562264,
      "grad_norm": 0.8665739893913269,
      "learning_rate": 8.695183708652085e-06,
      "loss": 3.1364,
      "step": 91600
    },
    {
      "epoch": 9.864384888601872,
      "grad_norm": 0.8592131733894348,
      "learning_rate": 8.37194267859067e-06,
      "loss": 3.1238,
      "step": 91650
    },
    {
      "epoch": 9.869766440641481,
      "grad_norm": 0.9557643532752991,
      "learning_rate": 8.048701648529252e-06,
      "loss": 3.1298,
      "step": 91700
    },
    {
      "epoch": 9.87514799268109,
      "grad_norm": 0.8807770609855652,
      "learning_rate": 7.725460618467837e-06,
      "loss": 3.1426,
      "step": 91750
    },
    {
      "epoch": 9.880529544720698,
      "grad_norm": 0.9055199027061462,
      "learning_rate": 7.40221958840642e-06,
      "loss": 3.1585,
      "step": 91800
    },
    {
      "epoch": 9.885911096760307,
      "grad_norm": 0.9519686102867126,
      "learning_rate": 7.078978558345006e-06,
      "loss": 3.1366,
      "step": 91850
    },
    {
      "epoch": 9.891292648799913,
      "grad_norm": 0.8609046936035156,
      "learning_rate": 6.75573752828359e-06,
      "loss": 3.1341,
      "step": 91900
    },
    {
      "epoch": 9.896674200839522,
      "grad_norm": 0.8886732459068298,
      "learning_rate": 6.432496498222174e-06,
      "loss": 3.1366,
      "step": 91950
    },
    {
      "epoch": 9.90205575287913,
      "grad_norm": 0.8871758580207825,
      "learning_rate": 6.1092554681607575e-06,
      "loss": 3.1266,
      "step": 92000
    },
    {
      "epoch": 9.90205575287913,
      "eval_accuracy": 0.3946833534323067,
      "eval_loss": 3.2989983558654785,
      "eval_runtime": 184.7106,
      "eval_samples_per_second": 97.509,
      "eval_steps_per_second": 6.096,
      "step": 92000
    },
    {
      "epoch": 9.907437304918739,
      "grad_norm": 0.9213749766349792,
      "learning_rate": 5.786014438099342e-06,
      "loss": 3.1351,
      "step": 92050
    },
    {
      "epoch": 9.912818856958348,
      "grad_norm": 0.8700889945030212,
      "learning_rate": 5.4627734080379264e-06,
      "loss": 3.1267,
      "step": 92100
    },
    {
      "epoch": 9.918200408997954,
      "grad_norm": 0.901222825050354,
      "learning_rate": 5.139532377976511e-06,
      "loss": 3.1486,
      "step": 92150
    },
    {
      "epoch": 9.923581961037563,
      "grad_norm": 0.8922194838523865,
      "learning_rate": 4.816291347915095e-06,
      "loss": 3.1234,
      "step": 92200
    },
    {
      "epoch": 9.928963513077171,
      "grad_norm": 0.876013994216919,
      "learning_rate": 4.493050317853679e-06,
      "loss": 3.1366,
      "step": 92250
    },
    {
      "epoch": 9.93434506511678,
      "grad_norm": 0.8912842869758606,
      "learning_rate": 4.169809287792264e-06,
      "loss": 3.1165,
      "step": 92300
    },
    {
      "epoch": 9.939726617156388,
      "grad_norm": 0.967231035232544,
      "learning_rate": 3.846568257730847e-06,
      "loss": 3.1164,
      "step": 92350
    },
    {
      "epoch": 9.945108169195997,
      "grad_norm": 0.9233632683753967,
      "learning_rate": 3.523327227669432e-06,
      "loss": 3.1274,
      "step": 92400
    },
    {
      "epoch": 9.950489721235604,
      "grad_norm": 0.8899291157722473,
      "learning_rate": 3.2000861976080162e-06,
      "loss": 3.1119,
      "step": 92450
    },
    {
      "epoch": 9.955871273275212,
      "grad_norm": 0.8986747860908508,
      "learning_rate": 2.8768451675466007e-06,
      "loss": 3.1101,
      "step": 92500
    },
    {
      "epoch": 9.961252825314821,
      "grad_norm": 0.8942649364471436,
      "learning_rate": 2.5536041374851848e-06,
      "loss": 3.1362,
      "step": 92550
    },
    {
      "epoch": 9.96663437735443,
      "grad_norm": 0.8900682926177979,
      "learning_rate": 2.230363107423769e-06,
      "loss": 3.1236,
      "step": 92600
    },
    {
      "epoch": 9.972015929394038,
      "grad_norm": 0.9322834014892578,
      "learning_rate": 1.9071220773623531e-06,
      "loss": 3.1256,
      "step": 92650
    },
    {
      "epoch": 9.977397481433645,
      "grad_norm": 0.9146674275398254,
      "learning_rate": 1.5838810473009372e-06,
      "loss": 3.1287,
      "step": 92700
    },
    {
      "epoch": 9.982779033473253,
      "grad_norm": 0.9451167583465576,
      "learning_rate": 1.2606400172395215e-06,
      "loss": 3.1258,
      "step": 92750
    },
    {
      "epoch": 9.988160585512862,
      "grad_norm": 0.8404658436775208,
      "learning_rate": 9.373989871781058e-07,
      "loss": 3.1275,
      "step": 92800
    },
    {
      "epoch": 9.99354213755247,
      "grad_norm": 0.9016643166542053,
      "learning_rate": 6.1415795711669e-07,
      "loss": 3.1313,
      "step": 92850
    },
    {
      "epoch": 9.998923689592079,
      "grad_norm": 0.8875418901443481,
      "learning_rate": 2.909169270552742e-07,
      "loss": 3.1211,
      "step": 92900
    },
    {
      "epoch": 10.0,
      "step": 92910,
      "total_flos": 7.7682904989696e+17,
      "train_loss": 3.4555358953643265,
      "train_runtime": 79991.7543,
      "train_samples_per_second": 37.167,
      "train_steps_per_second": 1.161
    }
  ],
  "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": 7.7682904989696e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}