{
  "best_metric": 3.3030855655670166,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_0_6910/checkpoint-90000",
  "epoch": 10.0,
  "eval_steps": 1000,
  "global_step": 92750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005390835579514825,
      "grad_norm": 2.8118252754211426,
      "learning_rate": 0.000276,
      "loss": 9.0225,
      "step": 50
    },
    {
      "epoch": 0.01078167115902965,
      "grad_norm": 1.6856672763824463,
      "learning_rate": 0.0005759999999999999,
      "loss": 6.9773,
      "step": 100
    },
    {
      "epoch": 0.016172506738544475,
      "grad_norm": 0.9568011164665222,
      "learning_rate": 0.000599702104695089,
      "loss": 6.5177,
      "step": 150
    },
    {
      "epoch": 0.0215633423180593,
      "grad_norm": 1.2443560361862183,
      "learning_rate": 0.0005993783054506205,
      "loss": 6.2578,
      "step": 200
    },
    {
      "epoch": 0.026954177897574125,
      "grad_norm": 1.41825270652771,
      "learning_rate": 0.0005990545062061521,
      "loss": 6.0969,
      "step": 250
    },
    {
      "epoch": 0.03234501347708895,
      "grad_norm": 1.9072396755218506,
      "learning_rate": 0.0005987307069616836,
      "loss": 5.9635,
      "step": 300
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 1.5556931495666504,
      "learning_rate": 0.0005984069077172153,
      "loss": 5.8839,
      "step": 350
    },
    {
      "epoch": 0.0431266846361186,
      "grad_norm": 1.5125844478607178,
      "learning_rate": 0.0005980831084727469,
      "loss": 5.8238,
      "step": 400
    },
    {
      "epoch": 0.04851752021563342,
      "grad_norm": 1.5804698467254639,
      "learning_rate": 0.0005977593092282784,
      "loss": 5.7627,
      "step": 450
    },
    {
      "epoch": 0.05390835579514825,
      "grad_norm": 1.1904308795928955,
      "learning_rate": 0.00059743550998381,
      "loss": 5.6518,
      "step": 500
    },
    {
      "epoch": 0.05929919137466307,
      "grad_norm": 1.009339690208435,
      "learning_rate": 0.0005971117107393416,
      "loss": 5.6142,
      "step": 550
    },
    {
      "epoch": 0.0646900269541779,
      "grad_norm": 1.3357371091842651,
      "learning_rate": 0.0005967879114948732,
      "loss": 5.5406,
      "step": 600
    },
    {
      "epoch": 0.07008086253369272,
      "grad_norm": 1.434008240699768,
      "learning_rate": 0.0005964641122504047,
      "loss": 5.4783,
      "step": 650
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 0.8490063548088074,
      "learning_rate": 0.0005961403130059363,
      "loss": 5.3996,
      "step": 700
    },
    {
      "epoch": 0.08086253369272237,
      "grad_norm": 1.1594209671020508,
      "learning_rate": 0.0005958165137614678,
      "loss": 5.331,
      "step": 750
    },
    {
      "epoch": 0.0862533692722372,
      "grad_norm": 0.9333789348602295,
      "learning_rate": 0.0005954927145169995,
      "loss": 5.2596,
      "step": 800
    },
    {
      "epoch": 0.09164420485175202,
      "grad_norm": 1.0127500295639038,
      "learning_rate": 0.0005951689152725309,
      "loss": 5.2418,
      "step": 850
    },
    {
      "epoch": 0.09703504043126684,
      "grad_norm": 1.113052487373352,
      "learning_rate": 0.0005948451160280626,
      "loss": 5.1815,
      "step": 900
    },
    {
      "epoch": 0.10242587601078167,
      "grad_norm": 0.9113388061523438,
      "learning_rate": 0.0005945213167835941,
      "loss": 5.1162,
      "step": 950
    },
    {
      "epoch": 0.1078167115902965,
      "grad_norm": 0.9989785552024841,
      "learning_rate": 0.0005941975175391257,
      "loss": 5.1216,
      "step": 1000
    },
    {
      "epoch": 0.1078167115902965,
      "eval_accuracy": 0.22690449769715637,
      "eval_loss": 5.032634735107422,
      "eval_runtime": 184.9778,
      "eval_samples_per_second": 97.368,
      "eval_steps_per_second": 6.087,
      "step": 1000
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 1.0346425771713257,
      "learning_rate": 0.0005938737182946572,
      "loss": 5.0533,
      "step": 1050
    },
    {
      "epoch": 0.11859838274932614,
      "grad_norm": 1.3372762203216553,
      "learning_rate": 0.0005935499190501888,
      "loss": 5.014,
      "step": 1100
    },
    {
      "epoch": 0.12398921832884097,
      "grad_norm": 1.185377836227417,
      "learning_rate": 0.0005932261198057204,
      "loss": 5.0192,
      "step": 1150
    },
    {
      "epoch": 0.1293800539083558,
      "grad_norm": 1.010504961013794,
      "learning_rate": 0.000592902320561252,
      "loss": 4.9761,
      "step": 1200
    },
    {
      "epoch": 0.1347708894878706,
      "grad_norm": 0.8940131068229675,
      "learning_rate": 0.0005925785213167835,
      "loss": 4.9547,
      "step": 1250
    },
    {
      "epoch": 0.14016172506738545,
      "grad_norm": 1.0484291315078735,
      "learning_rate": 0.0005922547220723151,
      "loss": 4.9102,
      "step": 1300
    },
    {
      "epoch": 0.14555256064690028,
      "grad_norm": 0.882337212562561,
      "learning_rate": 0.0005919309228278468,
      "loss": 4.8568,
      "step": 1350
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 1.2623690366744995,
      "learning_rate": 0.0005916071235833783,
      "loss": 4.8624,
      "step": 1400
    },
    {
      "epoch": 0.15633423180592992,
      "grad_norm": 0.7710188627243042,
      "learning_rate": 0.0005912833243389097,
      "loss": 4.8192,
      "step": 1450
    },
    {
      "epoch": 0.16172506738544473,
      "grad_norm": 0.933644711971283,
      "learning_rate": 0.0005909595250944414,
      "loss": 4.8205,
      "step": 1500
    },
    {
      "epoch": 0.16711590296495957,
      "grad_norm": 1.3466095924377441,
      "learning_rate": 0.000590635725849973,
      "loss": 4.8013,
      "step": 1550
    },
    {
      "epoch": 0.1725067385444744,
      "grad_norm": 1.1399619579315186,
      "learning_rate": 0.0005903119266055045,
      "loss": 4.7772,
      "step": 1600
    },
    {
      "epoch": 0.1778975741239892,
      "grad_norm": 0.8201636075973511,
      "learning_rate": 0.0005899881273610361,
      "loss": 4.7511,
      "step": 1650
    },
    {
      "epoch": 0.18328840970350405,
      "grad_norm": 0.902688205242157,
      "learning_rate": 0.0005896643281165677,
      "loss": 4.7286,
      "step": 1700
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 0.8644108176231384,
      "learning_rate": 0.0005893405288720993,
      "loss": 4.6775,
      "step": 1750
    },
    {
      "epoch": 0.1940700808625337,
      "grad_norm": 0.7629256844520569,
      "learning_rate": 0.0005890167296276308,
      "loss": 4.704,
      "step": 1800
    },
    {
      "epoch": 0.19946091644204852,
      "grad_norm": 1.2781596183776855,
      "learning_rate": 0.0005886929303831624,
      "loss": 4.6549,
      "step": 1850
    },
    {
      "epoch": 0.20485175202156333,
      "grad_norm": 0.7812705636024475,
      "learning_rate": 0.0005883691311386939,
      "loss": 4.632,
      "step": 1900
    },
    {
      "epoch": 0.21024258760107817,
      "grad_norm": 0.7588018178939819,
      "learning_rate": 0.0005880453318942256,
      "loss": 4.608,
      "step": 1950
    },
    {
      "epoch": 0.215633423180593,
      "grad_norm": 0.7466468811035156,
      "learning_rate": 0.0005877215326497571,
      "loss": 4.5906,
      "step": 2000
    },
    {
      "epoch": 0.215633423180593,
      "eval_accuracy": 0.26987845327398857,
      "eval_loss": 4.5157904624938965,
      "eval_runtime": 183.7342,
      "eval_samples_per_second": 98.027,
      "eval_steps_per_second": 6.128,
      "step": 2000
    },
    {
      "epoch": 0.2210242587601078,
      "grad_norm": 0.826682448387146,
      "learning_rate": 0.0005873977334052887,
      "loss": 4.576,
      "step": 2050
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 0.9647506475448608,
      "learning_rate": 0.0005870739341608202,
      "loss": 4.5599,
      "step": 2100
    },
    {
      "epoch": 0.23180592991913745,
      "grad_norm": 1.005610466003418,
      "learning_rate": 0.0005867501349163519,
      "loss": 4.5335,
      "step": 2150
    },
    {
      "epoch": 0.2371967654986523,
      "grad_norm": 1.0172079801559448,
      "learning_rate": 0.0005864263356718833,
      "loss": 4.5304,
      "step": 2200
    },
    {
      "epoch": 0.24258760107816713,
      "grad_norm": 0.8978919982910156,
      "learning_rate": 0.000586102536427415,
      "loss": 4.5045,
      "step": 2250
    },
    {
      "epoch": 0.24797843665768193,
      "grad_norm": 0.9270089268684387,
      "learning_rate": 0.0005857787371829465,
      "loss": 4.5004,
      "step": 2300
    },
    {
      "epoch": 0.25336927223719674,
      "grad_norm": 1.124731183052063,
      "learning_rate": 0.0005854549379384781,
      "loss": 4.4774,
      "step": 2350
    },
    {
      "epoch": 0.2587601078167116,
      "grad_norm": 0.7949519157409668,
      "learning_rate": 0.0005851311386940096,
      "loss": 4.4615,
      "step": 2400
    },
    {
      "epoch": 0.2641509433962264,
      "grad_norm": 0.7883573770523071,
      "learning_rate": 0.0005848073394495412,
      "loss": 4.4503,
      "step": 2450
    },
    {
      "epoch": 0.2695417789757412,
      "grad_norm": 0.9283429384231567,
      "learning_rate": 0.0005844835402050728,
      "loss": 4.4294,
      "step": 2500
    },
    {
      "epoch": 0.2749326145552561,
      "grad_norm": 0.703683078289032,
      "learning_rate": 0.0005841597409606044,
      "loss": 4.3969,
      "step": 2550
    },
    {
      "epoch": 0.2803234501347709,
      "grad_norm": 0.812880277633667,
      "learning_rate": 0.000583835941716136,
      "loss": 4.4119,
      "step": 2600
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.9256618618965149,
      "learning_rate": 0.0005835121424716675,
      "loss": 4.3917,
      "step": 2650
    },
    {
      "epoch": 0.29110512129380056,
      "grad_norm": 0.7537206411361694,
      "learning_rate": 0.0005831883432271992,
      "loss": 4.3791,
      "step": 2700
    },
    {
      "epoch": 0.29649595687331537,
      "grad_norm": 0.9869926571846008,
      "learning_rate": 0.0005828645439827307,
      "loss": 4.3759,
      "step": 2750
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 0.8514196276664734,
      "learning_rate": 0.0005825407447382622,
      "loss": 4.3314,
      "step": 2800
    },
    {
      "epoch": 0.30727762803234504,
      "grad_norm": 0.7071971893310547,
      "learning_rate": 0.0005822169454937938,
      "loss": 4.3577,
      "step": 2850
    },
    {
      "epoch": 0.31266846361185985,
      "grad_norm": 0.8087393045425415,
      "learning_rate": 0.0005818931462493254,
      "loss": 4.326,
      "step": 2900
    },
    {
      "epoch": 0.31805929919137466,
      "grad_norm": 0.8092687129974365,
      "learning_rate": 0.0005815693470048569,
      "loss": 4.3304,
      "step": 2950
    },
    {
      "epoch": 0.32345013477088946,
      "grad_norm": 0.7119380235671997,
      "learning_rate": 0.0005812455477603885,
      "loss": 4.3198,
      "step": 3000
    },
    {
      "epoch": 0.32345013477088946,
      "eval_accuracy": 0.29812342337868974,
      "eval_loss": 4.240254878997803,
      "eval_runtime": 183.7893,
      "eval_samples_per_second": 97.998,
      "eval_steps_per_second": 6.127,
      "step": 3000
    },
    {
      "epoch": 0.3288409703504043,
      "grad_norm": 0.7455107569694519,
      "learning_rate": 0.0005809217485159201,
      "loss": 4.3188,
      "step": 3050
    },
    {
      "epoch": 0.33423180592991913,
      "grad_norm": 0.8890029191970825,
      "learning_rate": 0.0005805979492714517,
      "loss": 4.2983,
      "step": 3100
    },
    {
      "epoch": 0.33962264150943394,
      "grad_norm": 0.8936235308647156,
      "learning_rate": 0.0005802741500269832,
      "loss": 4.2817,
      "step": 3150
    },
    {
      "epoch": 0.3450134770889488,
      "grad_norm": 0.6431916356086731,
      "learning_rate": 0.0005799503507825148,
      "loss": 4.2802,
      "step": 3200
    },
    {
      "epoch": 0.3504043126684636,
      "grad_norm": 0.7159081697463989,
      "learning_rate": 0.0005796265515380463,
      "loss": 4.2707,
      "step": 3250
    },
    {
      "epoch": 0.3557951482479784,
      "grad_norm": 0.9267504215240479,
      "learning_rate": 0.000579302752293578,
      "loss": 4.277,
      "step": 3300
    },
    {
      "epoch": 0.3611859838274933,
      "grad_norm": 0.8296390771865845,
      "learning_rate": 0.0005789789530491095,
      "loss": 4.2509,
      "step": 3350
    },
    {
      "epoch": 0.3665768194070081,
      "grad_norm": 0.7630621790885925,
      "learning_rate": 0.0005786551538046411,
      "loss": 4.2406,
      "step": 3400
    },
    {
      "epoch": 0.3719676549865229,
      "grad_norm": 0.6547260284423828,
      "learning_rate": 0.0005783313545601726,
      "loss": 4.2344,
      "step": 3450
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 0.8587298393249512,
      "learning_rate": 0.0005780075553157043,
      "loss": 4.2341,
      "step": 3500
    },
    {
      "epoch": 0.38274932614555257,
      "grad_norm": 0.7555488348007202,
      "learning_rate": 0.0005776837560712357,
      "loss": 4.2299,
      "step": 3550
    },
    {
      "epoch": 0.3881401617250674,
      "grad_norm": 0.8421213626861572,
      "learning_rate": 0.0005773599568267673,
      "loss": 4.2202,
      "step": 3600
    },
    {
      "epoch": 0.3935309973045822,
      "grad_norm": 0.7566924095153809,
      "learning_rate": 0.0005770361575822989,
      "loss": 4.2103,
      "step": 3650
    },
    {
      "epoch": 0.39892183288409705,
      "grad_norm": 0.7638437747955322,
      "learning_rate": 0.0005767123583378305,
      "loss": 4.1973,
      "step": 3700
    },
    {
      "epoch": 0.40431266846361186,
      "grad_norm": 0.6439513564109802,
      "learning_rate": 0.000576388559093362,
      "loss": 4.199,
      "step": 3750
    },
    {
      "epoch": 0.40970350404312667,
      "grad_norm": 0.7719266414642334,
      "learning_rate": 0.0005760647598488936,
      "loss": 4.1635,
      "step": 3800
    },
    {
      "epoch": 0.41509433962264153,
      "grad_norm": 0.6647982597351074,
      "learning_rate": 0.0005757409606044253,
      "loss": 4.1739,
      "step": 3850
    },
    {
      "epoch": 0.42048517520215634,
      "grad_norm": 0.7858614325523376,
      "learning_rate": 0.0005754171613599568,
      "loss": 4.1932,
      "step": 3900
    },
    {
      "epoch": 0.42587601078167114,
      "grad_norm": 1.070395588874817,
      "learning_rate": 0.0005750933621154884,
      "loss": 4.1588,
      "step": 3950
    },
    {
      "epoch": 0.431266846361186,
      "grad_norm": 0.6882054805755615,
      "learning_rate": 0.0005747695628710199,
      "loss": 4.1634,
      "step": 4000
    },
    {
      "epoch": 0.431266846361186,
      "eval_accuracy": 0.3118610599024015,
      "eval_loss": 4.099079608917236,
      "eval_runtime": 183.6446,
      "eval_samples_per_second": 98.075,
      "eval_steps_per_second": 6.131,
      "step": 4000
    },
    {
      "epoch": 0.4366576819407008,
      "grad_norm": 0.8163891434669495,
      "learning_rate": 0.0005744457636265515,
      "loss": 4.1703,
      "step": 4050
    },
    {
      "epoch": 0.4420485175202156,
      "grad_norm": 0.7172017097473145,
      "learning_rate": 0.0005741219643820831,
      "loss": 4.1633,
      "step": 4100
    },
    {
      "epoch": 0.4474393530997305,
      "grad_norm": 0.7089101672172546,
      "learning_rate": 0.0005737981651376146,
      "loss": 4.1486,
      "step": 4150
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 0.6500125527381897,
      "learning_rate": 0.0005734743658931462,
      "loss": 4.1541,
      "step": 4200
    },
    {
      "epoch": 0.4582210242587601,
      "grad_norm": 0.6067988276481628,
      "learning_rate": 0.0005731505666486778,
      "loss": 4.1386,
      "step": 4250
    },
    {
      "epoch": 0.4636118598382749,
      "grad_norm": 0.8405300974845886,
      "learning_rate": 0.0005728267674042093,
      "loss": 4.1407,
      "step": 4300
    },
    {
      "epoch": 0.46900269541778977,
      "grad_norm": 0.65191650390625,
      "learning_rate": 0.0005725029681597409,
      "loss": 4.1283,
      "step": 4350
    },
    {
      "epoch": 0.4743935309973046,
      "grad_norm": 0.674238920211792,
      "learning_rate": 0.0005721791689152725,
      "loss": 4.1114,
      "step": 4400
    },
    {
      "epoch": 0.4797843665768194,
      "grad_norm": 0.660973072052002,
      "learning_rate": 0.0005718553696708041,
      "loss": 4.1208,
      "step": 4450
    },
    {
      "epoch": 0.48517520215633425,
      "grad_norm": 0.6465425491333008,
      "learning_rate": 0.0005715315704263356,
      "loss": 4.1158,
      "step": 4500
    },
    {
      "epoch": 0.49056603773584906,
      "grad_norm": 0.7483091950416565,
      "learning_rate": 0.0005712077711818672,
      "loss": 4.1276,
      "step": 4550
    },
    {
      "epoch": 0.49595687331536387,
      "grad_norm": 0.845150351524353,
      "learning_rate": 0.0005708839719373987,
      "loss": 4.1192,
      "step": 4600
    },
    {
      "epoch": 0.5013477088948787,
      "grad_norm": 0.634871244430542,
      "learning_rate": 0.0005705601726929304,
      "loss": 4.0755,
      "step": 4650
    },
    {
      "epoch": 0.5067385444743935,
      "grad_norm": 0.6169816851615906,
      "learning_rate": 0.0005702363734484619,
      "loss": 4.078,
      "step": 4700
    },
    {
      "epoch": 0.5121293800539084,
      "grad_norm": 0.8197508454322815,
      "learning_rate": 0.0005699125742039935,
      "loss": 4.084,
      "step": 4750
    },
    {
      "epoch": 0.5175202156334232,
      "grad_norm": 0.733070969581604,
      "learning_rate": 0.000569588774959525,
      "loss": 4.0813,
      "step": 4800
    },
    {
      "epoch": 0.522911051212938,
      "grad_norm": 0.6208024024963379,
      "learning_rate": 0.0005692649757150567,
      "loss": 4.0748,
      "step": 4850
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 0.7824249863624573,
      "learning_rate": 0.0005689411764705881,
      "loss": 4.0771,
      "step": 4900
    },
    {
      "epoch": 0.5336927223719676,
      "grad_norm": 0.6890459656715393,
      "learning_rate": 0.0005686173772261197,
      "loss": 4.0671,
      "step": 4950
    },
    {
      "epoch": 0.5390835579514824,
      "grad_norm": 0.7617940902709961,
      "learning_rate": 0.0005682935779816514,
      "loss": 4.0659,
      "step": 5000
    },
    {
      "epoch": 0.5390835579514824,
      "eval_accuracy": 0.32073376337421977,
      "eval_loss": 3.998711347579956,
      "eval_runtime": 183.5304,
      "eval_samples_per_second": 98.136,
      "eval_steps_per_second": 6.135,
      "step": 5000
    },
    {
      "epoch": 0.5444743935309974,
      "grad_norm": 0.6309065222740173,
      "learning_rate": 0.0005679697787371829,
      "loss": 4.0746,
      "step": 5050
    },
    {
      "epoch": 0.5498652291105122,
      "grad_norm": 0.5925028920173645,
      "learning_rate": 0.0005676459794927145,
      "loss": 4.0574,
      "step": 5100
    },
    {
      "epoch": 0.555256064690027,
      "grad_norm": 0.6035439968109131,
      "learning_rate": 0.000567322180248246,
      "loss": 4.0516,
      "step": 5150
    },
    {
      "epoch": 0.5606469002695418,
      "grad_norm": 0.7275799512863159,
      "learning_rate": 0.0005669983810037777,
      "loss": 4.0651,
      "step": 5200
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.6090968251228333,
      "learning_rate": 0.0005666745817593092,
      "loss": 4.0379,
      "step": 5250
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.632185161113739,
      "learning_rate": 0.0005663507825148408,
      "loss": 4.0381,
      "step": 5300
    },
    {
      "epoch": 0.5768194070080862,
      "grad_norm": 0.6599447131156921,
      "learning_rate": 0.0005660269832703723,
      "loss": 4.0278,
      "step": 5350
    },
    {
      "epoch": 0.5822102425876011,
      "grad_norm": 0.648209810256958,
      "learning_rate": 0.0005657031840259039,
      "loss": 4.0327,
      "step": 5400
    },
    {
      "epoch": 0.5876010781671159,
      "grad_norm": 0.6686100363731384,
      "learning_rate": 0.0005653793847814355,
      "loss": 4.0357,
      "step": 5450
    },
    {
      "epoch": 0.5929919137466307,
      "grad_norm": 0.7332231998443604,
      "learning_rate": 0.000565055585536967,
      "loss": 4.0131,
      "step": 5500
    },
    {
      "epoch": 0.5983827493261455,
      "grad_norm": 0.6814959645271301,
      "learning_rate": 0.0005647317862924986,
      "loss": 4.033,
      "step": 5550
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.6917067766189575,
      "learning_rate": 0.0005644079870480302,
      "loss": 3.9815,
      "step": 5600
    },
    {
      "epoch": 0.6091644204851752,
      "grad_norm": 0.6626110672950745,
      "learning_rate": 0.0005640841878035617,
      "loss": 4.0186,
      "step": 5650
    },
    {
      "epoch": 0.6145552560646901,
      "grad_norm": 0.7377511262893677,
      "learning_rate": 0.0005637603885590933,
      "loss": 4.0184,
      "step": 5700
    },
    {
      "epoch": 0.6199460916442049,
      "grad_norm": 0.6328345537185669,
      "learning_rate": 0.0005634365893146248,
      "loss": 4.019,
      "step": 5750
    },
    {
      "epoch": 0.6253369272237197,
      "grad_norm": 0.6522849798202515,
      "learning_rate": 0.0005631127900701565,
      "loss": 4.01,
      "step": 5800
    },
    {
      "epoch": 0.6307277628032345,
      "grad_norm": 0.6383638978004456,
      "learning_rate": 0.000562788990825688,
      "loss": 3.9816,
      "step": 5850
    },
    {
      "epoch": 0.6361185983827493,
      "grad_norm": 0.593140721321106,
      "learning_rate": 0.0005624651915812196,
      "loss": 3.9949,
      "step": 5900
    },
    {
      "epoch": 0.6415094339622641,
      "grad_norm": 0.7360444068908691,
      "learning_rate": 0.0005621413923367511,
      "loss": 4.0016,
      "step": 5950
    },
    {
      "epoch": 0.6469002695417789,
      "grad_norm": 0.6608056426048279,
      "learning_rate": 0.0005618175930922828,
      "loss": 4.0076,
      "step": 6000
    },
    {
      "epoch": 0.6469002695417789,
      "eval_accuracy": 0.32758476256247404,
      "eval_loss": 3.921957492828369,
      "eval_runtime": 183.5766,
      "eval_samples_per_second": 98.112,
      "eval_steps_per_second": 6.134,
      "step": 6000
    },
    {
      "epoch": 0.6522911051212938,
      "grad_norm": 0.6179393529891968,
      "learning_rate": 0.0005614937938478143,
      "loss": 3.9939,
      "step": 6050
    },
    {
      "epoch": 0.6576819407008087,
      "grad_norm": 0.7146060466766357,
      "learning_rate": 0.0005611699946033459,
      "loss": 3.9929,
      "step": 6100
    },
    {
      "epoch": 0.6630727762803235,
      "grad_norm": 0.601253867149353,
      "learning_rate": 0.0005608461953588774,
      "loss": 3.9838,
      "step": 6150
    },
    {
      "epoch": 0.6684636118598383,
      "grad_norm": 0.6216392517089844,
      "learning_rate": 0.000560522396114409,
      "loss": 3.9788,
      "step": 6200
    },
    {
      "epoch": 0.6738544474393531,
      "grad_norm": 0.6294983625411987,
      "learning_rate": 0.0005601985968699405,
      "loss": 3.9608,
      "step": 6250
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 0.7225786447525024,
      "learning_rate": 0.0005598747976254721,
      "loss": 3.9794,
      "step": 6300
    },
    {
      "epoch": 0.6846361185983828,
      "grad_norm": 0.6607632637023926,
      "learning_rate": 0.0005595509983810038,
      "loss": 3.9496,
      "step": 6350
    },
    {
      "epoch": 0.6900269541778976,
      "grad_norm": 0.5790310502052307,
      "learning_rate": 0.0005592271991365353,
      "loss": 3.9592,
      "step": 6400
    },
    {
      "epoch": 0.6954177897574124,
      "grad_norm": 0.6292189955711365,
      "learning_rate": 0.0005589033998920669,
      "loss": 3.9773,
      "step": 6450
    },
    {
      "epoch": 0.7008086253369272,
      "grad_norm": 0.6256137490272522,
      "learning_rate": 0.0005585796006475984,
      "loss": 3.9487,
      "step": 6500
    },
    {
      "epoch": 0.706199460916442,
      "grad_norm": 0.6231578588485718,
      "learning_rate": 0.0005582558014031301,
      "loss": 3.9727,
      "step": 6550
    },
    {
      "epoch": 0.7115902964959568,
      "grad_norm": 0.6470305323600769,
      "learning_rate": 0.0005579320021586616,
      "loss": 3.9563,
      "step": 6600
    },
    {
      "epoch": 0.7169811320754716,
      "grad_norm": 0.5552076697349548,
      "learning_rate": 0.0005576082029141932,
      "loss": 3.951,
      "step": 6650
    },
    {
      "epoch": 0.7223719676549866,
      "grad_norm": 0.5381990671157837,
      "learning_rate": 0.0005572844036697247,
      "loss": 3.9356,
      "step": 6700
    },
    {
      "epoch": 0.7277628032345014,
      "grad_norm": 0.6558448076248169,
      "learning_rate": 0.0005569606044252563,
      "loss": 3.9426,
      "step": 6750
    },
    {
      "epoch": 0.7331536388140162,
      "grad_norm": 0.8135426640510559,
      "learning_rate": 0.0005566368051807879,
      "loss": 3.9613,
      "step": 6800
    },
    {
      "epoch": 0.738544474393531,
      "grad_norm": 0.6013303995132446,
      "learning_rate": 0.0005563130059363194,
      "loss": 3.9451,
      "step": 6850
    },
    {
      "epoch": 0.7439353099730458,
      "grad_norm": 0.5324015617370605,
      "learning_rate": 0.000555989206691851,
      "loss": 3.9444,
      "step": 6900
    },
    {
      "epoch": 0.7493261455525606,
      "grad_norm": 0.6945801377296448,
      "learning_rate": 0.0005556654074473826,
      "loss": 3.9473,
      "step": 6950
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.7069705128669739,
      "learning_rate": 0.0005553416082029141,
      "loss": 3.9328,
      "step": 7000
    },
    {
      "epoch": 0.7547169811320755,
      "eval_accuracy": 0.3323933047655917,
      "eval_loss": 3.8696444034576416,
      "eval_runtime": 183.451,
      "eval_samples_per_second": 98.179,
      "eval_steps_per_second": 6.138,
      "step": 7000
    },
    {
      "epoch": 0.7601078167115903,
      "grad_norm": 0.6576606631278992,
      "learning_rate": 0.0005550178089584457,
      "loss": 3.9266,
      "step": 7050
    },
    {
      "epoch": 0.7654986522911051,
      "grad_norm": 0.5154832005500793,
      "learning_rate": 0.0005546940097139772,
      "loss": 3.9252,
      "step": 7100
    },
    {
      "epoch": 0.77088948787062,
      "grad_norm": 0.6892321109771729,
      "learning_rate": 0.0005543702104695089,
      "loss": 3.9271,
      "step": 7150
    },
    {
      "epoch": 0.7762803234501348,
      "grad_norm": 0.6380577087402344,
      "learning_rate": 0.0005540464112250404,
      "loss": 3.9261,
      "step": 7200
    },
    {
      "epoch": 0.7816711590296496,
      "grad_norm": 0.652199923992157,
      "learning_rate": 0.000553722611980572,
      "loss": 3.9387,
      "step": 7250
    },
    {
      "epoch": 0.7870619946091644,
      "grad_norm": 0.5706573724746704,
      "learning_rate": 0.0005533988127361035,
      "loss": 3.9201,
      "step": 7300
    },
    {
      "epoch": 0.7924528301886793,
      "grad_norm": 0.5596190690994263,
      "learning_rate": 0.0005530750134916352,
      "loss": 3.9361,
      "step": 7350
    },
    {
      "epoch": 0.7978436657681941,
      "grad_norm": 0.6239616274833679,
      "learning_rate": 0.0005527512142471668,
      "loss": 3.9104,
      "step": 7400
    },
    {
      "epoch": 0.8032345013477089,
      "grad_norm": 0.5858375430107117,
      "learning_rate": 0.0005524274150026982,
      "loss": 3.9105,
      "step": 7450
    },
    {
      "epoch": 0.8086253369272237,
      "grad_norm": 0.5788413286209106,
      "learning_rate": 0.0005521036157582299,
      "loss": 3.904,
      "step": 7500
    },
    {
      "epoch": 0.8140161725067385,
      "grad_norm": 0.6172971725463867,
      "learning_rate": 0.0005517798165137614,
      "loss": 3.9068,
      "step": 7550
    },
    {
      "epoch": 0.8194070080862533,
      "grad_norm": 0.6352159976959229,
      "learning_rate": 0.000551456017269293,
      "loss": 3.8812,
      "step": 7600
    },
    {
      "epoch": 0.8247978436657682,
      "grad_norm": 0.6148518323898315,
      "learning_rate": 0.0005511322180248245,
      "loss": 3.8997,
      "step": 7650
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 0.6033445000648499,
      "learning_rate": 0.0005508084187803562,
      "loss": 3.9111,
      "step": 7700
    },
    {
      "epoch": 0.8355795148247979,
      "grad_norm": 0.5412169694900513,
      "learning_rate": 0.0005504846195358877,
      "loss": 3.9042,
      "step": 7750
    },
    {
      "epoch": 0.8409703504043127,
      "grad_norm": 0.5904088616371155,
      "learning_rate": 0.0005501608202914193,
      "loss": 3.892,
      "step": 7800
    },
    {
      "epoch": 0.8463611859838275,
      "grad_norm": 0.6405267715454102,
      "learning_rate": 0.0005498370210469508,
      "loss": 3.8977,
      "step": 7850
    },
    {
      "epoch": 0.8517520215633423,
      "grad_norm": 0.6236185431480408,
      "learning_rate": 0.0005495132218024824,
      "loss": 3.8806,
      "step": 7900
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.6019570231437683,
      "learning_rate": 0.000549189422558014,
      "loss": 3.8888,
      "step": 7950
    },
    {
      "epoch": 0.862533692722372,
      "grad_norm": 0.5633127093315125,
      "learning_rate": 0.0005488656233135456,
      "loss": 3.8875,
      "step": 8000
    },
    {
      "epoch": 0.862533692722372,
      "eval_accuracy": 0.33722933614932643,
      "eval_loss": 3.821709632873535,
      "eval_runtime": 183.4138,
      "eval_samples_per_second": 98.199,
      "eval_steps_per_second": 6.139,
      "step": 8000
    },
    {
      "epoch": 0.8679245283018868,
      "grad_norm": 0.5992864966392517,
      "learning_rate": 0.0005485418240690771,
      "loss": 3.8707,
      "step": 8050
    },
    {
      "epoch": 0.8733153638814016,
      "grad_norm": 0.6274523735046387,
      "learning_rate": 0.0005482180248246087,
      "loss": 3.8864,
      "step": 8100
    },
    {
      "epoch": 0.8787061994609164,
      "grad_norm": 0.6176576614379883,
      "learning_rate": 0.0005478942255801403,
      "loss": 3.8807,
      "step": 8150
    },
    {
      "epoch": 0.8840970350404312,
      "grad_norm": 0.5266938805580139,
      "learning_rate": 0.0005475704263356718,
      "loss": 3.8706,
      "step": 8200
    },
    {
      "epoch": 0.889487870619946,
      "grad_norm": 0.5737940073013306,
      "learning_rate": 0.0005472466270912034,
      "loss": 3.8805,
      "step": 8250
    },
    {
      "epoch": 0.894878706199461,
      "grad_norm": 0.6148428320884705,
      "learning_rate": 0.000546922827846735,
      "loss": 3.8693,
      "step": 8300
    },
    {
      "epoch": 0.9002695417789758,
      "grad_norm": 0.5487964749336243,
      "learning_rate": 0.0005465990286022665,
      "loss": 3.869,
      "step": 8350
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 0.6526573896408081,
      "learning_rate": 0.0005462752293577981,
      "loss": 3.8841,
      "step": 8400
    },
    {
      "epoch": 0.9110512129380054,
      "grad_norm": 0.601149320602417,
      "learning_rate": 0.0005459514301133296,
      "loss": 3.8793,
      "step": 8450
    },
    {
      "epoch": 0.9164420485175202,
      "grad_norm": 0.5263657569885254,
      "learning_rate": 0.0005456276308688613,
      "loss": 3.8761,
      "step": 8500
    },
    {
      "epoch": 0.921832884097035,
      "grad_norm": 0.5656020045280457,
      "learning_rate": 0.0005453038316243929,
      "loss": 3.8672,
      "step": 8550
    },
    {
      "epoch": 0.9272237196765498,
      "grad_norm": 0.565776526927948,
      "learning_rate": 0.0005449800323799244,
      "loss": 3.8777,
      "step": 8600
    },
    {
      "epoch": 0.9326145552560647,
      "grad_norm": 0.5656868815422058,
      "learning_rate": 0.000544656233135456,
      "loss": 3.8577,
      "step": 8650
    },
    {
      "epoch": 0.9380053908355795,
      "grad_norm": 0.5700314044952393,
      "learning_rate": 0.0005443324338909875,
      "loss": 3.8624,
      "step": 8700
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.5940127968788147,
      "learning_rate": 0.0005440086346465192,
      "loss": 3.8726,
      "step": 8750
    },
    {
      "epoch": 0.9487870619946092,
      "grad_norm": 0.5483199954032898,
      "learning_rate": 0.0005436848354020506,
      "loss": 3.8541,
      "step": 8800
    },
    {
      "epoch": 0.954177897574124,
      "grad_norm": 0.6202383041381836,
      "learning_rate": 0.0005433610361575823,
      "loss": 3.8599,
      "step": 8850
    },
    {
      "epoch": 0.9595687331536388,
      "grad_norm": 0.5427079200744629,
      "learning_rate": 0.0005430372369131138,
      "loss": 3.8437,
      "step": 8900
    },
    {
      "epoch": 0.9649595687331537,
      "grad_norm": 0.5505421757698059,
      "learning_rate": 0.0005427134376686454,
      "loss": 3.858,
      "step": 8950
    },
    {
      "epoch": 0.9703504043126685,
      "grad_norm": 0.6305214166641235,
      "learning_rate": 0.0005423896384241769,
      "loss": 3.8437,
      "step": 9000
    },
    {
      "epoch": 0.9703504043126685,
      "eval_accuracy": 0.3406345176534323,
      "eval_loss": 3.7832400798797607,
      "eval_runtime": 183.7219,
      "eval_samples_per_second": 98.034,
      "eval_steps_per_second": 6.129,
      "step": 9000
    },
    {
      "epoch": 0.9757412398921833,
      "grad_norm": 0.7392898797988892,
      "learning_rate": 0.0005420658391797086,
      "loss": 3.8372,
      "step": 9050
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 0.587247908115387,
      "learning_rate": 0.0005417420399352401,
      "loss": 3.8504,
      "step": 9100
    },
    {
      "epoch": 0.9865229110512129,
      "grad_norm": 0.5904769897460938,
      "learning_rate": 0.0005414182406907717,
      "loss": 3.844,
      "step": 9150
    },
    {
      "epoch": 0.9919137466307277,
      "grad_norm": 0.632688581943512,
      "learning_rate": 0.0005410944414463032,
      "loss": 3.8564,
      "step": 9200
    },
    {
      "epoch": 0.9973045822102425,
      "grad_norm": 0.5667609572410583,
      "learning_rate": 0.0005407706422018348,
      "loss": 3.8552,
      "step": 9250
    },
    {
      "epoch": 1.0026954177897573,
      "grad_norm": 0.6239280700683594,
      "learning_rate": 0.0005404468429573664,
      "loss": 3.8025,
      "step": 9300
    },
    {
      "epoch": 1.0080862533692723,
      "grad_norm": 0.6430540680885315,
      "learning_rate": 0.000540123043712898,
      "loss": 3.7743,
      "step": 9350
    },
    {
      "epoch": 1.013477088948787,
      "grad_norm": 0.5992752909660339,
      "learning_rate": 0.0005397992444684295,
      "loss": 3.7575,
      "step": 9400
    },
    {
      "epoch": 1.0188679245283019,
      "grad_norm": 0.6434339284896851,
      "learning_rate": 0.0005394754452239611,
      "loss": 3.7703,
      "step": 9450
    },
    {
      "epoch": 1.0242587601078168,
      "grad_norm": 0.5548680424690247,
      "learning_rate": 0.0005391516459794927,
      "loss": 3.7811,
      "step": 9500
    },
    {
      "epoch": 1.0296495956873315,
      "grad_norm": 0.5591529011726379,
      "learning_rate": 0.0005388278467350242,
      "loss": 3.7885,
      "step": 9550
    },
    {
      "epoch": 1.0350404312668464,
      "grad_norm": 0.5492196083068848,
      "learning_rate": 0.0005385040474905557,
      "loss": 3.7816,
      "step": 9600
    },
    {
      "epoch": 1.0404312668463611,
      "grad_norm": 0.5632776618003845,
      "learning_rate": 0.0005381802482460874,
      "loss": 3.7914,
      "step": 9650
    },
    {
      "epoch": 1.045822102425876,
      "grad_norm": 0.5463435053825378,
      "learning_rate": 0.000537856449001619,
      "loss": 3.7925,
      "step": 9700
    },
    {
      "epoch": 1.0512129380053907,
      "grad_norm": 0.5662521719932556,
      "learning_rate": 0.0005375326497571505,
      "loss": 3.7623,
      "step": 9750
    },
    {
      "epoch": 1.0566037735849056,
      "grad_norm": 0.6173110008239746,
      "learning_rate": 0.000537208850512682,
      "loss": 3.7692,
      "step": 9800
    },
    {
      "epoch": 1.0619946091644206,
      "grad_norm": 0.5675989389419556,
      "learning_rate": 0.0005368850512682137,
      "loss": 3.7665,
      "step": 9850
    },
    {
      "epoch": 1.0673854447439353,
      "grad_norm": 0.5368490815162659,
      "learning_rate": 0.0005365612520237453,
      "loss": 3.7797,
      "step": 9900
    },
    {
      "epoch": 1.0727762803234502,
      "grad_norm": 0.5896486639976501,
      "learning_rate": 0.0005362374527792768,
      "loss": 3.7825,
      "step": 9950
    },
    {
      "epoch": 1.0781671159029649,
      "grad_norm": 0.5884218215942383,
      "learning_rate": 0.0005359136535348084,
      "loss": 3.7795,
      "step": 10000
    },
    {
      "epoch": 1.0781671159029649,
      "eval_accuracy": 0.3442962286256681,
      "eval_loss": 3.7569897174835205,
      "eval_runtime": 183.6339,
      "eval_samples_per_second": 98.081,
      "eval_steps_per_second": 6.132,
      "step": 10000
    },
    {
      "epoch": 1.0835579514824798,
      "grad_norm": 0.5574485063552856,
      "learning_rate": 0.0005355898542903399,
      "loss": 3.7575,
      "step": 10050
    },
    {
      "epoch": 1.0889487870619945,
      "grad_norm": 0.5631133317947388,
      "learning_rate": 0.0005352660550458716,
      "loss": 3.7605,
      "step": 10100
    },
    {
      "epoch": 1.0943396226415094,
      "grad_norm": 0.6444439888000488,
      "learning_rate": 0.000534942255801403,
      "loss": 3.7508,
      "step": 10150
    },
    {
      "epoch": 1.0997304582210243,
      "grad_norm": 0.6213650107383728,
      "learning_rate": 0.0005346184565569347,
      "loss": 3.7672,
      "step": 10200
    },
    {
      "epoch": 1.105121293800539,
      "grad_norm": 0.5763867497444153,
      "learning_rate": 0.0005342946573124662,
      "loss": 3.7855,
      "step": 10250
    },
    {
      "epoch": 1.110512129380054,
      "grad_norm": 0.6422098278999329,
      "learning_rate": 0.0005339708580679978,
      "loss": 3.7876,
      "step": 10300
    },
    {
      "epoch": 1.1159029649595686,
      "grad_norm": 0.564333975315094,
      "learning_rate": 0.0005336470588235293,
      "loss": 3.7619,
      "step": 10350
    },
    {
      "epoch": 1.1212938005390836,
      "grad_norm": 0.573645830154419,
      "learning_rate": 0.000533323259579061,
      "loss": 3.7662,
      "step": 10400
    },
    {
      "epoch": 1.1266846361185983,
      "grad_norm": 0.5722724199295044,
      "learning_rate": 0.0005329994603345925,
      "loss": 3.7615,
      "step": 10450
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.5559325814247131,
      "learning_rate": 0.0005326756610901241,
      "loss": 3.7594,
      "step": 10500
    },
    {
      "epoch": 1.137466307277628,
      "grad_norm": 0.553900957107544,
      "learning_rate": 0.0005323518618456556,
      "loss": 3.76,
      "step": 10550
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.6239808201789856,
      "learning_rate": 0.0005320280626011872,
      "loss": 3.7736,
      "step": 10600
    },
    {
      "epoch": 1.1482479784366577,
      "grad_norm": 0.6128347516059875,
      "learning_rate": 0.0005317042633567188,
      "loss": 3.7596,
      "step": 10650
    },
    {
      "epoch": 1.1536388140161726,
      "grad_norm": 0.6108589172363281,
      "learning_rate": 0.0005313804641122504,
      "loss": 3.7629,
      "step": 10700
    },
    {
      "epoch": 1.1590296495956873,
      "grad_norm": 0.5402905941009521,
      "learning_rate": 0.0005310566648677819,
      "loss": 3.7593,
      "step": 10750
    },
    {
      "epoch": 1.1644204851752022,
      "grad_norm": 0.6511978507041931,
      "learning_rate": 0.0005307328656233135,
      "loss": 3.7501,
      "step": 10800
    },
    {
      "epoch": 1.169811320754717,
      "grad_norm": 0.5607616901397705,
      "learning_rate": 0.000530409066378845,
      "loss": 3.7458,
      "step": 10850
    },
    {
      "epoch": 1.1752021563342319,
      "grad_norm": 1.1150946617126465,
      "learning_rate": 0.0005300852671343766,
      "loss": 3.7631,
      "step": 10900
    },
    {
      "epoch": 1.1805929919137466,
      "grad_norm": 0.6214979290962219,
      "learning_rate": 0.0005297614678899081,
      "loss": 3.751,
      "step": 10950
    },
    {
      "epoch": 1.1859838274932615,
      "grad_norm": 0.542398989200592,
      "learning_rate": 0.0005294376686454398,
      "loss": 3.7415,
      "step": 11000
    },
    {
      "epoch": 1.1859838274932615,
      "eval_accuracy": 0.3474025060353963,
      "eval_loss": 3.724672555923462,
      "eval_runtime": 183.3754,
      "eval_samples_per_second": 98.219,
      "eval_steps_per_second": 6.14,
      "step": 11000
    },
    {
      "epoch": 1.1913746630727764,
      "grad_norm": 0.5348007082939148,
      "learning_rate": 0.0005291138694009714,
      "loss": 3.7517,
      "step": 11050
    },
    {
      "epoch": 1.196765498652291,
      "grad_norm": 0.6700088977813721,
      "learning_rate": 0.0005287900701565029,
      "loss": 3.7463,
      "step": 11100
    },
    {
      "epoch": 1.202156334231806,
      "grad_norm": 0.5958192348480225,
      "learning_rate": 0.0005284662709120345,
      "loss": 3.7478,
      "step": 11150
    },
    {
      "epoch": 1.2075471698113207,
      "grad_norm": 0.6137239336967468,
      "learning_rate": 0.0005281424716675661,
      "loss": 3.7436,
      "step": 11200
    },
    {
      "epoch": 1.2129380053908356,
      "grad_norm": 0.5924616456031799,
      "learning_rate": 0.0005278186724230977,
      "loss": 3.7357,
      "step": 11250
    },
    {
      "epoch": 1.2183288409703503,
      "grad_norm": 0.5333642959594727,
      "learning_rate": 0.0005274948731786292,
      "loss": 3.7353,
      "step": 11300
    },
    {
      "epoch": 1.2237196765498652,
      "grad_norm": 0.6624649167060852,
      "learning_rate": 0.0005271710739341608,
      "loss": 3.7523,
      "step": 11350
    },
    {
      "epoch": 1.2291105121293802,
      "grad_norm": 0.5522056221961975,
      "learning_rate": 0.0005268472746896923,
      "loss": 3.7463,
      "step": 11400
    },
    {
      "epoch": 1.2345013477088949,
      "grad_norm": 0.6152771711349487,
      "learning_rate": 0.000526523475445224,
      "loss": 3.7354,
      "step": 11450
    },
    {
      "epoch": 1.2398921832884098,
      "grad_norm": 0.5271309614181519,
      "learning_rate": 0.0005261996762007554,
      "loss": 3.7252,
      "step": 11500
    },
    {
      "epoch": 1.2452830188679245,
      "grad_norm": 0.5856218338012695,
      "learning_rate": 0.0005258758769562871,
      "loss": 3.7444,
      "step": 11550
    },
    {
      "epoch": 1.2506738544474394,
      "grad_norm": 0.6060943007469177,
      "learning_rate": 0.0005255520777118186,
      "loss": 3.7456,
      "step": 11600
    },
    {
      "epoch": 1.256064690026954,
      "grad_norm": 0.5646070837974548,
      "learning_rate": 0.0005252282784673502,
      "loss": 3.7347,
      "step": 11650
    },
    {
      "epoch": 1.261455525606469,
      "grad_norm": 0.5944088697433472,
      "learning_rate": 0.0005249044792228817,
      "loss": 3.7372,
      "step": 11700
    },
    {
      "epoch": 1.266846361185984,
      "grad_norm": 0.5187632441520691,
      "learning_rate": 0.0005245806799784133,
      "loss": 3.7333,
      "step": 11750
    },
    {
      "epoch": 1.2722371967654986,
      "grad_norm": 0.6018880009651184,
      "learning_rate": 0.0005242568807339449,
      "loss": 3.7307,
      "step": 11800
    },
    {
      "epoch": 1.2776280323450135,
      "grad_norm": 0.584701418876648,
      "learning_rate": 0.0005239330814894765,
      "loss": 3.7259,
      "step": 11850
    },
    {
      "epoch": 1.2830188679245282,
      "grad_norm": 0.5245819687843323,
      "learning_rate": 0.000523609282245008,
      "loss": 3.7477,
      "step": 11900
    },
    {
      "epoch": 1.2884097035040432,
      "grad_norm": 0.5737572908401489,
      "learning_rate": 0.0005232854830005396,
      "loss": 3.741,
      "step": 11950
    },
    {
      "epoch": 1.2938005390835579,
      "grad_norm": 0.5504003763198853,
      "learning_rate": 0.0005229616837560712,
      "loss": 3.7386,
      "step": 12000
    },
    {
      "epoch": 1.2938005390835579,
      "eval_accuracy": 0.34930121524995433,
      "eval_loss": 3.7041478157043457,
      "eval_runtime": 183.6844,
      "eval_samples_per_second": 98.054,
      "eval_steps_per_second": 6.13,
      "step": 12000
    },
    {
      "epoch": 1.2991913746630728,
      "grad_norm": 0.5650700926780701,
      "learning_rate": 0.0005226378845116028,
      "loss": 3.7186,
      "step": 12050
    },
    {
      "epoch": 1.3045822102425877,
      "grad_norm": 0.5875822305679321,
      "learning_rate": 0.0005223140852671344,
      "loss": 3.732,
      "step": 12100
    },
    {
      "epoch": 1.3099730458221024,
      "grad_norm": 0.6313338279724121,
      "learning_rate": 0.0005219902860226659,
      "loss": 3.7404,
      "step": 12150
    },
    {
      "epoch": 1.3153638814016173,
      "grad_norm": 0.553345799446106,
      "learning_rate": 0.0005216664867781975,
      "loss": 3.7401,
      "step": 12200
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.6565985679626465,
      "learning_rate": 0.000521342687533729,
      "loss": 3.7227,
      "step": 12250
    },
    {
      "epoch": 1.326145552560647,
      "grad_norm": 0.5814613103866577,
      "learning_rate": 0.0005210188882892606,
      "loss": 3.7334,
      "step": 12300
    },
    {
      "epoch": 1.3315363881401616,
      "grad_norm": 0.5827574729919434,
      "learning_rate": 0.0005206950890447922,
      "loss": 3.7207,
      "step": 12350
    },
    {
      "epoch": 1.3369272237196765,
      "grad_norm": 0.5803568363189697,
      "learning_rate": 0.0005203712898003238,
      "loss": 3.7345,
      "step": 12400
    },
    {
      "epoch": 1.3423180592991915,
      "grad_norm": 0.6482006311416626,
      "learning_rate": 0.0005200474905558553,
      "loss": 3.7252,
      "step": 12450
    },
    {
      "epoch": 1.3477088948787062,
      "grad_norm": 0.699213445186615,
      "learning_rate": 0.0005197236913113869,
      "loss": 3.7266,
      "step": 12500
    },
    {
      "epoch": 1.353099730458221,
      "grad_norm": 0.5527870059013367,
      "learning_rate": 0.0005193998920669184,
      "loss": 3.7349,
      "step": 12550
    },
    {
      "epoch": 1.3584905660377358,
      "grad_norm": 0.5830569267272949,
      "learning_rate": 0.0005190760928224501,
      "loss": 3.7173,
      "step": 12600
    },
    {
      "epoch": 1.3638814016172507,
      "grad_norm": 0.5488144159317017,
      "learning_rate": 0.0005187522935779816,
      "loss": 3.7258,
      "step": 12650
    },
    {
      "epoch": 1.3692722371967654,
      "grad_norm": 0.5564178824424744,
      "learning_rate": 0.0005184284943335132,
      "loss": 3.7399,
      "step": 12700
    },
    {
      "epoch": 1.3746630727762803,
      "grad_norm": 0.5656182169914246,
      "learning_rate": 0.0005181046950890447,
      "loss": 3.7232,
      "step": 12750
    },
    {
      "epoch": 1.3800539083557952,
      "grad_norm": 0.5447649359703064,
      "learning_rate": 0.0005177808958445764,
      "loss": 3.7155,
      "step": 12800
    },
    {
      "epoch": 1.38544474393531,
      "grad_norm": 0.5839430689811707,
      "learning_rate": 0.0005174570966001078,
      "loss": 3.711,
      "step": 12850
    },
    {
      "epoch": 1.3908355795148248,
      "grad_norm": 0.5413840413093567,
      "learning_rate": 0.0005171332973556395,
      "loss": 3.7186,
      "step": 12900
    },
    {
      "epoch": 1.3962264150943398,
      "grad_norm": 0.5594467520713806,
      "learning_rate": 0.000516809498111171,
      "loss": 3.7149,
      "step": 12950
    },
    {
      "epoch": 1.4016172506738545,
      "grad_norm": 0.5479294657707214,
      "learning_rate": 0.0005164856988667026,
      "loss": 3.7126,
      "step": 13000
    },
    {
      "epoch": 1.4016172506738545,
      "eval_accuracy": 0.3515265350452827,
      "eval_loss": 3.68241024017334,
      "eval_runtime": 183.4286,
      "eval_samples_per_second": 98.191,
      "eval_steps_per_second": 6.139,
      "step": 13000
    },
    {
      "epoch": 1.4070080862533692,
      "grad_norm": 0.6145225167274475,
      "learning_rate": 0.0005161618996222341,
      "loss": 3.7236,
      "step": 13050
    },
    {
      "epoch": 1.412398921832884,
      "grad_norm": 0.5675894021987915,
      "learning_rate": 0.0005158381003777657,
      "loss": 3.7337,
      "step": 13100
    },
    {
      "epoch": 1.417789757412399,
      "grad_norm": 0.5551406741142273,
      "learning_rate": 0.0005155143011332973,
      "loss": 3.7143,
      "step": 13150
    },
    {
      "epoch": 1.4231805929919137,
      "grad_norm": 0.5070226192474365,
      "learning_rate": 0.0005151905018888289,
      "loss": 3.7167,
      "step": 13200
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.5413742661476135,
      "learning_rate": 0.0005148731786292498,
      "loss": 3.7111,
      "step": 13250
    },
    {
      "epoch": 1.4339622641509435,
      "grad_norm": 0.5501816272735596,
      "learning_rate": 0.0005145493793847814,
      "loss": 3.7103,
      "step": 13300
    },
    {
      "epoch": 1.4393530997304582,
      "grad_norm": 0.593174934387207,
      "learning_rate": 0.000514225580140313,
      "loss": 3.6969,
      "step": 13350
    },
    {
      "epoch": 1.444743935309973,
      "grad_norm": 0.577828586101532,
      "learning_rate": 0.0005139017808958445,
      "loss": 3.7057,
      "step": 13400
    },
    {
      "epoch": 1.4501347708894878,
      "grad_norm": 0.6046187877655029,
      "learning_rate": 0.0005135779816513762,
      "loss": 3.7161,
      "step": 13450
    },
    {
      "epoch": 1.4555256064690028,
      "grad_norm": 0.5544584393501282,
      "learning_rate": 0.0005132541824069076,
      "loss": 3.7009,
      "step": 13500
    },
    {
      "epoch": 1.4609164420485174,
      "grad_norm": 0.5399063229560852,
      "learning_rate": 0.0005129303831624393,
      "loss": 3.7132,
      "step": 13550
    },
    {
      "epoch": 1.4663072776280324,
      "grad_norm": 0.5806102156639099,
      "learning_rate": 0.0005126065839179708,
      "loss": 3.7158,
      "step": 13600
    },
    {
      "epoch": 1.4716981132075473,
      "grad_norm": 0.5742630362510681,
      "learning_rate": 0.0005122827846735024,
      "loss": 3.6862,
      "step": 13650
    },
    {
      "epoch": 1.477088948787062,
      "grad_norm": 0.5323203206062317,
      "learning_rate": 0.0005119589854290339,
      "loss": 3.6859,
      "step": 13700
    },
    {
      "epoch": 1.482479784366577,
      "grad_norm": 0.5530096888542175,
      "learning_rate": 0.0005116351861845655,
      "loss": 3.7044,
      "step": 13750
    },
    {
      "epoch": 1.4878706199460916,
      "grad_norm": 0.6719633340835571,
      "learning_rate": 0.0005113113869400971,
      "loss": 3.7217,
      "step": 13800
    },
    {
      "epoch": 1.4932614555256065,
      "grad_norm": 0.6222640872001648,
      "learning_rate": 0.0005109875876956287,
      "loss": 3.7135,
      "step": 13850
    },
    {
      "epoch": 1.4986522911051212,
      "grad_norm": 0.631252110004425,
      "learning_rate": 0.0005106637884511602,
      "loss": 3.7266,
      "step": 13900
    },
    {
      "epoch": 1.5040431266846361,
      "grad_norm": 0.5431990027427673,
      "learning_rate": 0.0005103399892066918,
      "loss": 3.7016,
      "step": 13950
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.5950245261192322,
      "learning_rate": 0.0005100161899622234,
      "loss": 3.703,
      "step": 14000
    },
    {
      "epoch": 1.509433962264151,
      "eval_accuracy": 0.35352118475978095,
      "eval_loss": 3.6599960327148438,
      "eval_runtime": 183.3906,
      "eval_samples_per_second": 98.211,
      "eval_steps_per_second": 6.14,
      "step": 14000
    },
    {
      "epoch": 1.5148247978436657,
      "grad_norm": 0.5107572674751282,
      "learning_rate": 0.000509692390717755,
      "loss": 3.6925,
      "step": 14050
    },
    {
      "epoch": 1.5202156334231804,
      "grad_norm": 0.6692053079605103,
      "learning_rate": 0.0005093685914732865,
      "loss": 3.6872,
      "step": 14100
    },
    {
      "epoch": 1.5256064690026954,
      "grad_norm": 0.5478536486625671,
      "learning_rate": 0.0005090447922288181,
      "loss": 3.6865,
      "step": 14150
    },
    {
      "epoch": 1.5309973045822103,
      "grad_norm": 0.5649746060371399,
      "learning_rate": 0.0005087209929843496,
      "loss": 3.7203,
      "step": 14200
    },
    {
      "epoch": 1.536388140161725,
      "grad_norm": 0.6324547529220581,
      "learning_rate": 0.0005083971937398812,
      "loss": 3.695,
      "step": 14250
    },
    {
      "epoch": 1.54177897574124,
      "grad_norm": 0.6280053853988647,
      "learning_rate": 0.0005080733944954127,
      "loss": 3.6973,
      "step": 14300
    },
    {
      "epoch": 1.5471698113207548,
      "grad_norm": 0.5381866693496704,
      "learning_rate": 0.0005077495952509444,
      "loss": 3.6772,
      "step": 14350
    },
    {
      "epoch": 1.5525606469002695,
      "grad_norm": 0.5399860739707947,
      "learning_rate": 0.0005074257960064759,
      "loss": 3.6922,
      "step": 14400
    },
    {
      "epoch": 1.5579514824797842,
      "grad_norm": 0.5678397417068481,
      "learning_rate": 0.0005071019967620075,
      "loss": 3.6985,
      "step": 14450
    },
    {
      "epoch": 1.5633423180592994,
      "grad_norm": 0.5648748874664307,
      "learning_rate": 0.000506778197517539,
      "loss": 3.69,
      "step": 14500
    },
    {
      "epoch": 1.568733153638814,
      "grad_norm": 0.5564712285995483,
      "learning_rate": 0.0005064543982730707,
      "loss": 3.6994,
      "step": 14550
    },
    {
      "epoch": 1.5741239892183287,
      "grad_norm": 0.6412194967269897,
      "learning_rate": 0.0005061305990286023,
      "loss": 3.6939,
      "step": 14600
    },
    {
      "epoch": 1.5795148247978437,
      "grad_norm": 0.6159881353378296,
      "learning_rate": 0.0005058067997841338,
      "loss": 3.6991,
      "step": 14650
    },
    {
      "epoch": 1.5849056603773586,
      "grad_norm": 0.583437979221344,
      "learning_rate": 0.0005054830005396654,
      "loss": 3.6886,
      "step": 14700
    },
    {
      "epoch": 1.5902964959568733,
      "grad_norm": 0.5995129942893982,
      "learning_rate": 0.0005051592012951969,
      "loss": 3.6783,
      "step": 14750
    },
    {
      "epoch": 1.595687331536388,
      "grad_norm": 0.5731451511383057,
      "learning_rate": 0.0005048354020507286,
      "loss": 3.701,
      "step": 14800
    },
    {
      "epoch": 1.6010781671159031,
      "grad_norm": 0.6119737029075623,
      "learning_rate": 0.00050451160280626,
      "loss": 3.6983,
      "step": 14850
    },
    {
      "epoch": 1.6064690026954178,
      "grad_norm": 0.5756235718727112,
      "learning_rate": 0.0005041878035617917,
      "loss": 3.6896,
      "step": 14900
    },
    {
      "epoch": 1.6118598382749325,
      "grad_norm": 0.5833989977836609,
      "learning_rate": 0.0005038640043173232,
      "loss": 3.6857,
      "step": 14950
    },
    {
      "epoch": 1.6172506738544474,
      "grad_norm": 0.6236714124679565,
      "learning_rate": 0.0005035402050728548,
      "loss": 3.6784,
      "step": 15000
    },
    {
      "epoch": 1.6172506738544474,
      "eval_accuracy": 0.3555121402760935,
      "eval_loss": 3.6398637294769287,
      "eval_runtime": 183.4963,
      "eval_samples_per_second": 98.155,
      "eval_steps_per_second": 6.136,
      "step": 15000
    },
    {
      "epoch": 1.6226415094339623,
      "grad_norm": 0.5502594709396362,
      "learning_rate": 0.0005032164058283863,
      "loss": 3.6833,
      "step": 15050
    },
    {
      "epoch": 1.628032345013477,
      "grad_norm": 0.5723167061805725,
      "learning_rate": 0.0005028926065839179,
      "loss": 3.6792,
      "step": 15100
    },
    {
      "epoch": 1.633423180592992,
      "grad_norm": 0.5737243294715881,
      "learning_rate": 0.0005025688073394495,
      "loss": 3.6794,
      "step": 15150
    },
    {
      "epoch": 1.6388140161725069,
      "grad_norm": 0.548410952091217,
      "learning_rate": 0.0005022450080949811,
      "loss": 3.6866,
      "step": 15200
    },
    {
      "epoch": 1.6442048517520216,
      "grad_norm": 0.654476523399353,
      "learning_rate": 0.0005019212088505126,
      "loss": 3.6874,
      "step": 15250
    },
    {
      "epoch": 1.6495956873315363,
      "grad_norm": 0.511240541934967,
      "learning_rate": 0.0005015974096060442,
      "loss": 3.6893,
      "step": 15300
    },
    {
      "epoch": 1.6549865229110512,
      "grad_norm": 0.5759029984474182,
      "learning_rate": 0.0005012736103615758,
      "loss": 3.6728,
      "step": 15350
    },
    {
      "epoch": 1.6603773584905661,
      "grad_norm": 0.5841777920722961,
      "learning_rate": 0.0005009498111171074,
      "loss": 3.6776,
      "step": 15400
    },
    {
      "epoch": 1.6657681940700808,
      "grad_norm": 0.5825933814048767,
      "learning_rate": 0.0005006260118726389,
      "loss": 3.6777,
      "step": 15450
    },
    {
      "epoch": 1.6711590296495957,
      "grad_norm": 0.5646806359291077,
      "learning_rate": 0.0005003022126281705,
      "loss": 3.6786,
      "step": 15500
    },
    {
      "epoch": 1.6765498652291106,
      "grad_norm": 0.627292275428772,
      "learning_rate": 0.000499978413383702,
      "loss": 3.6701,
      "step": 15550
    },
    {
      "epoch": 1.6819407008086253,
      "grad_norm": 0.5698474645614624,
      "learning_rate": 0.0004996546141392336,
      "loss": 3.6674,
      "step": 15600
    },
    {
      "epoch": 1.68733153638814,
      "grad_norm": 0.5848374366760254,
      "learning_rate": 0.0004993308148947651,
      "loss": 3.677,
      "step": 15650
    },
    {
      "epoch": 1.692722371967655,
      "grad_norm": 0.5110187530517578,
      "learning_rate": 0.0004990070156502968,
      "loss": 3.6661,
      "step": 15700
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.5932977199554443,
      "learning_rate": 0.0004986832164058284,
      "loss": 3.6621,
      "step": 15750
    },
    {
      "epoch": 1.7035040431266846,
      "grad_norm": 0.5683602094650269,
      "learning_rate": 0.0004983594171613599,
      "loss": 3.6748,
      "step": 15800
    },
    {
      "epoch": 1.7088948787061995,
      "grad_norm": 0.6081506013870239,
      "learning_rate": 0.0004980356179168915,
      "loss": 3.6814,
      "step": 15850
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5934674143791199,
      "learning_rate": 0.000497711818672423,
      "loss": 3.704,
      "step": 15900
    },
    {
      "epoch": 1.719676549865229,
      "grad_norm": 0.5152272582054138,
      "learning_rate": 0.0004973880194279547,
      "loss": 3.6704,
      "step": 15950
    },
    {
      "epoch": 1.7250673854447438,
      "grad_norm": 0.5506434440612793,
      "learning_rate": 0.0004970642201834862,
      "loss": 3.6787,
      "step": 16000
    },
    {
      "epoch": 1.7250673854447438,
      "eval_accuracy": 0.3569788456087012,
      "eval_loss": 3.6205105781555176,
      "eval_runtime": 183.4216,
      "eval_samples_per_second": 98.195,
      "eval_steps_per_second": 6.139,
      "step": 16000
    },
    {
      "epoch": 1.7304582210242587,
      "grad_norm": 0.5687834024429321,
      "learning_rate": 0.0004967404209390178,
      "loss": 3.6465,
      "step": 16050
    },
    {
      "epoch": 1.7358490566037736,
      "grad_norm": 0.5210959911346436,
      "learning_rate": 0.0004964166216945493,
      "loss": 3.6549,
      "step": 16100
    },
    {
      "epoch": 1.7412398921832883,
      "grad_norm": 0.567616879940033,
      "learning_rate": 0.000496092822450081,
      "loss": 3.6694,
      "step": 16150
    },
    {
      "epoch": 1.7466307277628033,
      "grad_norm": 0.5949267745018005,
      "learning_rate": 0.0004957690232056125,
      "loss": 3.6572,
      "step": 16200
    },
    {
      "epoch": 1.7520215633423182,
      "grad_norm": 0.5664389729499817,
      "learning_rate": 0.0004954452239611441,
      "loss": 3.6673,
      "step": 16250
    },
    {
      "epoch": 1.7574123989218329,
      "grad_norm": 0.5412474870681763,
      "learning_rate": 0.0004951214247166756,
      "loss": 3.658,
      "step": 16300
    },
    {
      "epoch": 1.7628032345013476,
      "grad_norm": 0.5369322896003723,
      "learning_rate": 0.0004947976254722072,
      "loss": 3.6511,
      "step": 16350
    },
    {
      "epoch": 1.7681940700808625,
      "grad_norm": 0.5792037844657898,
      "learning_rate": 0.0004944738262277387,
      "loss": 3.667,
      "step": 16400
    },
    {
      "epoch": 1.7735849056603774,
      "grad_norm": 0.6002668142318726,
      "learning_rate": 0.0004941500269832703,
      "loss": 3.6646,
      "step": 16450
    },
    {
      "epoch": 1.778975741239892,
      "grad_norm": 0.5983853340148926,
      "learning_rate": 0.0004938262277388019,
      "loss": 3.6523,
      "step": 16500
    },
    {
      "epoch": 1.784366576819407,
      "grad_norm": 0.5764635801315308,
      "learning_rate": 0.0004935024284943335,
      "loss": 3.6473,
      "step": 16550
    },
    {
      "epoch": 1.789757412398922,
      "grad_norm": 0.5904407501220703,
      "learning_rate": 0.000493178629249865,
      "loss": 3.6545,
      "step": 16600
    },
    {
      "epoch": 1.7951482479784366,
      "grad_norm": 0.5274083018302917,
      "learning_rate": 0.0004928548300053966,
      "loss": 3.6631,
      "step": 16650
    },
    {
      "epoch": 1.8005390835579513,
      "grad_norm": 0.5829530954360962,
      "learning_rate": 0.0004925310307609282,
      "loss": 3.654,
      "step": 16700
    },
    {
      "epoch": 1.8059299191374663,
      "grad_norm": 0.522125780582428,
      "learning_rate": 0.0004922072315164598,
      "loss": 3.6488,
      "step": 16750
    },
    {
      "epoch": 1.8113207547169812,
      "grad_norm": 0.5361061096191406,
      "learning_rate": 0.0004918834322719913,
      "loss": 3.6535,
      "step": 16800
    },
    {
      "epoch": 1.8167115902964959,
      "grad_norm": 0.5397604703903198,
      "learning_rate": 0.0004915596330275229,
      "loss": 3.6442,
      "step": 16850
    },
    {
      "epoch": 1.8221024258760108,
      "grad_norm": 0.5718424320220947,
      "learning_rate": 0.0004912358337830544,
      "loss": 3.6489,
      "step": 16900
    },
    {
      "epoch": 1.8274932614555257,
      "grad_norm": 0.5481303930282593,
      "learning_rate": 0.000490912034538586,
      "loss": 3.655,
      "step": 16950
    },
    {
      "epoch": 1.8328840970350404,
      "grad_norm": 0.5255277156829834,
      "learning_rate": 0.0004905882352941175,
      "loss": 3.6622,
      "step": 17000
    },
    {
      "epoch": 1.8328840970350404,
      "eval_accuracy": 0.35840252439773435,
      "eval_loss": 3.60727596282959,
      "eval_runtime": 183.9803,
      "eval_samples_per_second": 97.896,
      "eval_steps_per_second": 6.12,
      "step": 17000
    },
    {
      "epoch": 1.838274932614555,
      "grad_norm": 0.5432764887809753,
      "learning_rate": 0.0004902644360496492,
      "loss": 3.6562,
      "step": 17050
    },
    {
      "epoch": 1.8436657681940702,
      "grad_norm": 0.5756969451904297,
      "learning_rate": 0.0004899406368051808,
      "loss": 3.6436,
      "step": 17100
    },
    {
      "epoch": 1.849056603773585,
      "grad_norm": 0.5970697402954102,
      "learning_rate": 0.0004896168375607123,
      "loss": 3.6439,
      "step": 17150
    },
    {
      "epoch": 1.8544474393530996,
      "grad_norm": 0.5418063402175903,
      "learning_rate": 0.0004892930383162439,
      "loss": 3.6429,
      "step": 17200
    },
    {
      "epoch": 1.8598382749326146,
      "grad_norm": 0.5667785406112671,
      "learning_rate": 0.0004889757150566648,
      "loss": 3.6382,
      "step": 17250
    },
    {
      "epoch": 1.8652291105121295,
      "grad_norm": 0.5172768235206604,
      "learning_rate": 0.0004886519158121964,
      "loss": 3.6628,
      "step": 17300
    },
    {
      "epoch": 1.8706199460916442,
      "grad_norm": 0.5945035219192505,
      "learning_rate": 0.000488328116567728,
      "loss": 3.6666,
      "step": 17350
    },
    {
      "epoch": 1.8760107816711589,
      "grad_norm": 0.5649627447128296,
      "learning_rate": 0.0004880043173232595,
      "loss": 3.6471,
      "step": 17400
    },
    {
      "epoch": 1.881401617250674,
      "grad_norm": 0.538306474685669,
      "learning_rate": 0.0004876805180787911,
      "loss": 3.6564,
      "step": 17450
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.5613325834274292,
      "learning_rate": 0.000487363194819212,
      "loss": 3.6392,
      "step": 17500
    },
    {
      "epoch": 1.8921832884097034,
      "grad_norm": 0.5915970206260681,
      "learning_rate": 0.0004870393955747436,
      "loss": 3.6539,
      "step": 17550
    },
    {
      "epoch": 1.8975741239892183,
      "grad_norm": 0.5147404670715332,
      "learning_rate": 0.0004867155963302752,
      "loss": 3.6422,
      "step": 17600
    },
    {
      "epoch": 1.9029649595687332,
      "grad_norm": 0.5603023767471313,
      "learning_rate": 0.0004863917970858068,
      "loss": 3.6386,
      "step": 17650
    },
    {
      "epoch": 1.908355795148248,
      "grad_norm": 0.5458604693412781,
      "learning_rate": 0.00048606799784133833,
      "loss": 3.6524,
      "step": 17700
    },
    {
      "epoch": 1.9137466307277629,
      "grad_norm": 0.606473445892334,
      "learning_rate": 0.00048574419859686994,
      "loss": 3.6401,
      "step": 17750
    },
    {
      "epoch": 1.9191374663072778,
      "grad_norm": 0.5725317001342773,
      "learning_rate": 0.0004854203993524015,
      "loss": 3.6557,
      "step": 17800
    },
    {
      "epoch": 1.9245283018867925,
      "grad_norm": 0.6147105097770691,
      "learning_rate": 0.000485096600107933,
      "loss": 3.6629,
      "step": 17850
    },
    {
      "epoch": 1.9299191374663072,
      "grad_norm": 0.5641502141952515,
      "learning_rate": 0.00048477280086346464,
      "loss": 3.6383,
      "step": 17900
    },
    {
      "epoch": 1.935309973045822,
      "grad_norm": 0.5484232902526855,
      "learning_rate": 0.00048444900161899614,
      "loss": 3.6478,
      "step": 17950
    },
    {
      "epoch": 1.940700808625337,
      "grad_norm": 0.5765843391418457,
      "learning_rate": 0.00048412520237452774,
      "loss": 3.6447,
      "step": 18000
    },
    {
      "epoch": 1.940700808625337,
      "eval_accuracy": 0.3603276362640313,
      "eval_loss": 3.5914807319641113,
      "eval_runtime": 183.476,
      "eval_samples_per_second": 98.165,
      "eval_steps_per_second": 6.137,
      "step": 18000
    },
    {
      "epoch": 1.9460916442048517,
      "grad_norm": 0.5953685641288757,
      "learning_rate": 0.0004838014031300593,
      "loss": 3.6288,
      "step": 18050
    },
    {
      "epoch": 1.9514824797843666,
      "grad_norm": 0.5497512817382812,
      "learning_rate": 0.0004834776038855909,
      "loss": 3.6586,
      "step": 18100
    },
    {
      "epoch": 1.9568733153638815,
      "grad_norm": 0.5568259954452515,
      "learning_rate": 0.00048315380464112245,
      "loss": 3.6575,
      "step": 18150
    },
    {
      "epoch": 1.9622641509433962,
      "grad_norm": 0.6135269999504089,
      "learning_rate": 0.00048283000539665405,
      "loss": 3.6467,
      "step": 18200
    },
    {
      "epoch": 1.967654986522911,
      "grad_norm": 0.5931421518325806,
      "learning_rate": 0.0004825062061521856,
      "loss": 3.648,
      "step": 18250
    },
    {
      "epoch": 1.9730458221024259,
      "grad_norm": 0.558696448802948,
      "learning_rate": 0.00048218240690771716,
      "loss": 3.6273,
      "step": 18300
    },
    {
      "epoch": 1.9784366576819408,
      "grad_norm": 0.6715732216835022,
      "learning_rate": 0.00048185860766324876,
      "loss": 3.6528,
      "step": 18350
    },
    {
      "epoch": 1.9838274932614555,
      "grad_norm": 0.5224708914756775,
      "learning_rate": 0.0004815348084187803,
      "loss": 3.6293,
      "step": 18400
    },
    {
      "epoch": 1.9892183288409704,
      "grad_norm": 0.5655230283737183,
      "learning_rate": 0.0004812110091743119,
      "loss": 3.6291,
      "step": 18450
    },
    {
      "epoch": 1.9946091644204853,
      "grad_norm": 0.5798264741897583,
      "learning_rate": 0.00048088720992984347,
      "loss": 3.6214,
      "step": 18500
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0706647634506226,
      "learning_rate": 0.00048056341068537507,
      "loss": 3.6367,
      "step": 18550
    },
    {
      "epoch": 2.0053908355795147,
      "grad_norm": 0.6311784982681274,
      "learning_rate": 0.00048023961144090657,
      "loss": 3.5528,
      "step": 18600
    },
    {
      "epoch": 2.01078167115903,
      "grad_norm": 0.5586987137794495,
      "learning_rate": 0.00047991581219643817,
      "loss": 3.5434,
      "step": 18650
    },
    {
      "epoch": 2.0161725067385445,
      "grad_norm": 0.6267822980880737,
      "learning_rate": 0.0004795920129519697,
      "loss": 3.552,
      "step": 18700
    },
    {
      "epoch": 2.0215633423180592,
      "grad_norm": 0.5614279508590698,
      "learning_rate": 0.0004792682137075013,
      "loss": 3.5445,
      "step": 18750
    },
    {
      "epoch": 2.026954177897574,
      "grad_norm": 0.5612114071846008,
      "learning_rate": 0.0004789444144630329,
      "loss": 3.5517,
      "step": 18800
    },
    {
      "epoch": 2.032345013477089,
      "grad_norm": 0.5366635322570801,
      "learning_rate": 0.00047862061521856443,
      "loss": 3.5508,
      "step": 18850
    },
    {
      "epoch": 2.0377358490566038,
      "grad_norm": 0.5916035175323486,
      "learning_rate": 0.00047829681597409603,
      "loss": 3.5641,
      "step": 18900
    },
    {
      "epoch": 2.0431266846361185,
      "grad_norm": 0.5509299635887146,
      "learning_rate": 0.0004779730167296276,
      "loss": 3.5591,
      "step": 18950
    },
    {
      "epoch": 2.0485175202156336,
      "grad_norm": 0.6209540963172913,
      "learning_rate": 0.0004776492174851592,
      "loss": 3.5598,
      "step": 19000
    },
    {
      "epoch": 2.0485175202156336,
      "eval_accuracy": 0.3615285766330448,
      "eval_loss": 3.5827741622924805,
      "eval_runtime": 183.6104,
      "eval_samples_per_second": 98.094,
      "eval_steps_per_second": 6.133,
      "step": 19000
    },
    {
      "epoch": 2.0539083557951483,
      "grad_norm": 0.5647927522659302,
      "learning_rate": 0.00047732541824069074,
      "loss": 3.5683,
      "step": 19050
    },
    {
      "epoch": 2.059299191374663,
      "grad_norm": 0.5451118350028992,
      "learning_rate": 0.0004770016189962223,
      "loss": 3.5618,
      "step": 19100
    },
    {
      "epoch": 2.0646900269541777,
      "grad_norm": 0.561872661113739,
      "learning_rate": 0.0004766778197517539,
      "loss": 3.5564,
      "step": 19150
    },
    {
      "epoch": 2.070080862533693,
      "grad_norm": 0.5675294995307922,
      "learning_rate": 0.0004763540205072854,
      "loss": 3.5752,
      "step": 19200
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.5630126595497131,
      "learning_rate": 0.00047603022126281705,
      "loss": 3.5646,
      "step": 19250
    },
    {
      "epoch": 2.0808625336927222,
      "grad_norm": 0.603242814540863,
      "learning_rate": 0.00047570642201834855,
      "loss": 3.5555,
      "step": 19300
    },
    {
      "epoch": 2.0862533692722374,
      "grad_norm": 0.5813462138175964,
      "learning_rate": 0.00047538262277388015,
      "loss": 3.5722,
      "step": 19350
    },
    {
      "epoch": 2.091644204851752,
      "grad_norm": 0.5469586253166199,
      "learning_rate": 0.0004750588235294117,
      "loss": 3.5634,
      "step": 19400
    },
    {
      "epoch": 2.0970350404312668,
      "grad_norm": 0.6438509225845337,
      "learning_rate": 0.0004747350242849433,
      "loss": 3.5481,
      "step": 19450
    },
    {
      "epoch": 2.1024258760107815,
      "grad_norm": 0.5822634696960449,
      "learning_rate": 0.00047441122504047486,
      "loss": 3.5723,
      "step": 19500
    },
    {
      "epoch": 2.1078167115902966,
      "grad_norm": 0.5717223882675171,
      "learning_rate": 0.0004740874257960064,
      "loss": 3.5625,
      "step": 19550
    },
    {
      "epoch": 2.1132075471698113,
      "grad_norm": 0.5862929821014404,
      "learning_rate": 0.000473763626551538,
      "loss": 3.5588,
      "step": 19600
    },
    {
      "epoch": 2.118598382749326,
      "grad_norm": 0.5764575600624084,
      "learning_rate": 0.00047343982730706956,
      "loss": 3.5623,
      "step": 19650
    },
    {
      "epoch": 2.123989218328841,
      "grad_norm": 0.5479846596717834,
      "learning_rate": 0.00047311602806260117,
      "loss": 3.5536,
      "step": 19700
    },
    {
      "epoch": 2.129380053908356,
      "grad_norm": 0.6235895752906799,
      "learning_rate": 0.0004727922288181327,
      "loss": 3.5448,
      "step": 19750
    },
    {
      "epoch": 2.1347708894878705,
      "grad_norm": 0.5840886831283569,
      "learning_rate": 0.0004724684295736643,
      "loss": 3.5682,
      "step": 19800
    },
    {
      "epoch": 2.1401617250673857,
      "grad_norm": 0.5456245541572571,
      "learning_rate": 0.0004721446303291959,
      "loss": 3.5595,
      "step": 19850
    },
    {
      "epoch": 2.1455525606469004,
      "grad_norm": 0.5832547545433044,
      "learning_rate": 0.0004718208310847275,
      "loss": 3.5621,
      "step": 19900
    },
    {
      "epoch": 2.150943396226415,
      "grad_norm": 0.5474064946174622,
      "learning_rate": 0.000471497031840259,
      "loss": 3.5543,
      "step": 19950
    },
    {
      "epoch": 2.1563342318059298,
      "grad_norm": 0.5883390307426453,
      "learning_rate": 0.0004711732325957905,
      "loss": 3.5586,
      "step": 20000
    },
    {
      "epoch": 2.1563342318059298,
      "eval_accuracy": 0.36281231050257284,
      "eval_loss": 3.5723373889923096,
      "eval_runtime": 183.7122,
      "eval_samples_per_second": 98.039,
      "eval_steps_per_second": 6.129,
      "step": 20000
    },
    {
      "epoch": 2.161725067385445,
      "grad_norm": 0.5845204591751099,
      "learning_rate": 0.00047084943335132213,
      "loss": 3.5599,
      "step": 20050
    },
    {
      "epoch": 2.1671159029649596,
      "grad_norm": 0.5878739953041077,
      "learning_rate": 0.0004705256341068537,
      "loss": 3.5556,
      "step": 20100
    },
    {
      "epoch": 2.1725067385444743,
      "grad_norm": 0.5857297778129578,
      "learning_rate": 0.0004702018348623853,
      "loss": 3.5674,
      "step": 20150
    },
    {
      "epoch": 2.177897574123989,
      "grad_norm": 0.5607808828353882,
      "learning_rate": 0.00046987803561791684,
      "loss": 3.5675,
      "step": 20200
    },
    {
      "epoch": 2.183288409703504,
      "grad_norm": 0.5514035224914551,
      "learning_rate": 0.00046955423637344844,
      "loss": 3.5755,
      "step": 20250
    },
    {
      "epoch": 2.188679245283019,
      "grad_norm": 0.607312023639679,
      "learning_rate": 0.00046923043712898,
      "loss": 3.5548,
      "step": 20300
    },
    {
      "epoch": 2.1940700808625335,
      "grad_norm": 0.5242741703987122,
      "learning_rate": 0.0004689066378845116,
      "loss": 3.5481,
      "step": 20350
    },
    {
      "epoch": 2.1994609164420487,
      "grad_norm": 0.6142726540565491,
      "learning_rate": 0.00046858283864004315,
      "loss": 3.5619,
      "step": 20400
    },
    {
      "epoch": 2.2048517520215634,
      "grad_norm": 0.7217954397201538,
      "learning_rate": 0.0004682590393955747,
      "loss": 3.5574,
      "step": 20450
    },
    {
      "epoch": 2.210242587601078,
      "grad_norm": 0.5498636364936829,
      "learning_rate": 0.0004679352401511063,
      "loss": 3.572,
      "step": 20500
    },
    {
      "epoch": 2.215633423180593,
      "grad_norm": 0.5720429420471191,
      "learning_rate": 0.0004676114409066378,
      "loss": 3.5675,
      "step": 20550
    },
    {
      "epoch": 2.221024258760108,
      "grad_norm": 0.6373676061630249,
      "learning_rate": 0.00046728764166216946,
      "loss": 3.5666,
      "step": 20600
    },
    {
      "epoch": 2.2264150943396226,
      "grad_norm": 0.5550791025161743,
      "learning_rate": 0.00046696384241770095,
      "loss": 3.5731,
      "step": 20650
    },
    {
      "epoch": 2.2318059299191373,
      "grad_norm": 0.6740472912788391,
      "learning_rate": 0.00046664004317323256,
      "loss": 3.5562,
      "step": 20700
    },
    {
      "epoch": 2.2371967654986524,
      "grad_norm": 0.5599523186683655,
      "learning_rate": 0.0004663162439287641,
      "loss": 3.5742,
      "step": 20750
    },
    {
      "epoch": 2.242587601078167,
      "grad_norm": 0.5968847870826721,
      "learning_rate": 0.00046599244468429566,
      "loss": 3.5682,
      "step": 20800
    },
    {
      "epoch": 2.247978436657682,
      "grad_norm": 0.577918291091919,
      "learning_rate": 0.00046566864543982726,
      "loss": 3.5715,
      "step": 20850
    },
    {
      "epoch": 2.2533692722371965,
      "grad_norm": 0.544826865196228,
      "learning_rate": 0.0004653448461953588,
      "loss": 3.5616,
      "step": 20900
    },
    {
      "epoch": 2.2587601078167117,
      "grad_norm": 0.5795401930809021,
      "learning_rate": 0.0004650210469508904,
      "loss": 3.5613,
      "step": 20950
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.5422238707542419,
      "learning_rate": 0.00046469724770642197,
      "loss": 3.5585,
      "step": 21000
    },
    {
      "epoch": 2.2641509433962264,
      "eval_accuracy": 0.3636982661498121,
      "eval_loss": 3.5617871284484863,
      "eval_runtime": 183.3857,
      "eval_samples_per_second": 98.214,
      "eval_steps_per_second": 6.14,
      "step": 21000
    },
    {
      "epoch": 2.269541778975741,
      "grad_norm": 0.6669674515724182,
      "learning_rate": 0.0004643734484619536,
      "loss": 3.556,
      "step": 21050
    },
    {
      "epoch": 2.274932614555256,
      "grad_norm": 0.6207566261291504,
      "learning_rate": 0.0004640496492174851,
      "loss": 3.546,
      "step": 21100
    },
    {
      "epoch": 2.280323450134771,
      "grad_norm": 0.595628559589386,
      "learning_rate": 0.00046372584997301673,
      "loss": 3.5727,
      "step": 21150
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.5632020235061646,
      "learning_rate": 0.0004634020507285483,
      "loss": 3.5522,
      "step": 21200
    },
    {
      "epoch": 2.2911051212938007,
      "grad_norm": 0.5731921195983887,
      "learning_rate": 0.0004630782514840798,
      "loss": 3.5504,
      "step": 21250
    },
    {
      "epoch": 2.2964959568733154,
      "grad_norm": 0.5752100348472595,
      "learning_rate": 0.0004627544522396114,
      "loss": 3.5611,
      "step": 21300
    },
    {
      "epoch": 2.30188679245283,
      "grad_norm": 0.5865209102630615,
      "learning_rate": 0.00046243065299514293,
      "loss": 3.5384,
      "step": 21350
    },
    {
      "epoch": 2.3072776280323453,
      "grad_norm": 0.5781216621398926,
      "learning_rate": 0.00046210685375067454,
      "loss": 3.5746,
      "step": 21400
    },
    {
      "epoch": 2.31266846361186,
      "grad_norm": 0.5815436244010925,
      "learning_rate": 0.0004617830545062061,
      "loss": 3.5338,
      "step": 21450
    },
    {
      "epoch": 2.3180592991913747,
      "grad_norm": 0.5783815979957581,
      "learning_rate": 0.0004614592552617377,
      "loss": 3.5657,
      "step": 21500
    },
    {
      "epoch": 2.3234501347708894,
      "grad_norm": 0.6043853163719177,
      "learning_rate": 0.00046114193200215864,
      "loss": 3.5665,
      "step": 21550
    },
    {
      "epoch": 2.3288409703504045,
      "grad_norm": 0.5591253042221069,
      "learning_rate": 0.0004608181327576902,
      "loss": 3.5399,
      "step": 21600
    },
    {
      "epoch": 2.334231805929919,
      "grad_norm": 0.6128972172737122,
      "learning_rate": 0.00046049433351322175,
      "loss": 3.5586,
      "step": 21650
    },
    {
      "epoch": 2.339622641509434,
      "grad_norm": 0.5784398317337036,
      "learning_rate": 0.00046017053426875335,
      "loss": 3.5736,
      "step": 21700
    },
    {
      "epoch": 2.3450134770889486,
      "grad_norm": 0.5711215734481812,
      "learning_rate": 0.0004598467350242849,
      "loss": 3.5537,
      "step": 21750
    },
    {
      "epoch": 2.3504043126684637,
      "grad_norm": 0.5763333439826965,
      "learning_rate": 0.0004595229357798165,
      "loss": 3.5654,
      "step": 21800
    },
    {
      "epoch": 2.3557951482479784,
      "grad_norm": 0.6388970613479614,
      "learning_rate": 0.00045919913653534806,
      "loss": 3.5649,
      "step": 21850
    },
    {
      "epoch": 2.361185983827493,
      "grad_norm": 0.5283222794532776,
      "learning_rate": 0.00045887533729087966,
      "loss": 3.558,
      "step": 21900
    },
    {
      "epoch": 2.3665768194070083,
      "grad_norm": 0.5484132170677185,
      "learning_rate": 0.00045855153804641116,
      "loss": 3.5504,
      "step": 21950
    },
    {
      "epoch": 2.371967654986523,
      "grad_norm": 0.5693395733833313,
      "learning_rate": 0.0004582277388019427,
      "loss": 3.546,
      "step": 22000
    },
    {
      "epoch": 2.371967654986523,
      "eval_accuracy": 0.36476686730146757,
      "eval_loss": 3.5517172813415527,
      "eval_runtime": 183.5403,
      "eval_samples_per_second": 98.131,
      "eval_steps_per_second": 6.135,
      "step": 22000
    },
    {
      "epoch": 2.3773584905660377,
      "grad_norm": 0.5790075659751892,
      "learning_rate": 0.0004579039395574743,
      "loss": 3.5475,
      "step": 22050
    },
    {
      "epoch": 2.382749326145553,
      "grad_norm": 0.5600558519363403,
      "learning_rate": 0.00045758014031300586,
      "loss": 3.5632,
      "step": 22100
    },
    {
      "epoch": 2.3881401617250675,
      "grad_norm": 0.5839028358459473,
      "learning_rate": 0.00045725634106853747,
      "loss": 3.557,
      "step": 22150
    },
    {
      "epoch": 2.393530997304582,
      "grad_norm": 0.582438051700592,
      "learning_rate": 0.000456932541824069,
      "loss": 3.5437,
      "step": 22200
    },
    {
      "epoch": 2.398921832884097,
      "grad_norm": 0.8435466885566711,
      "learning_rate": 0.0004566087425796006,
      "loss": 3.5393,
      "step": 22250
    },
    {
      "epoch": 2.404312668463612,
      "grad_norm": 0.6102995872497559,
      "learning_rate": 0.0004562849433351322,
      "loss": 3.5668,
      "step": 22300
    },
    {
      "epoch": 2.4097035040431267,
      "grad_norm": 0.5426021814346313,
      "learning_rate": 0.0004559611440906638,
      "loss": 3.5759,
      "step": 22350
    },
    {
      "epoch": 2.4150943396226414,
      "grad_norm": 0.6063995957374573,
      "learning_rate": 0.00045563734484619533,
      "loss": 3.5585,
      "step": 22400
    },
    {
      "epoch": 2.420485175202156,
      "grad_norm": 0.5520073771476746,
      "learning_rate": 0.0004553135456017269,
      "loss": 3.564,
      "step": 22450
    },
    {
      "epoch": 2.4258760107816713,
      "grad_norm": 0.6225586533546448,
      "learning_rate": 0.0004549897463572585,
      "loss": 3.5633,
      "step": 22500
    },
    {
      "epoch": 2.431266846361186,
      "grad_norm": 0.5612086653709412,
      "learning_rate": 0.00045466594711279,
      "loss": 3.543,
      "step": 22550
    },
    {
      "epoch": 2.4366576819407006,
      "grad_norm": 0.563895046710968,
      "learning_rate": 0.00045434214786832164,
      "loss": 3.5606,
      "step": 22600
    },
    {
      "epoch": 2.442048517520216,
      "grad_norm": 0.5550323128700256,
      "learning_rate": 0.00045401834862385314,
      "loss": 3.5558,
      "step": 22650
    },
    {
      "epoch": 2.4474393530997305,
      "grad_norm": 0.610876202583313,
      "learning_rate": 0.00045369454937938474,
      "loss": 3.563,
      "step": 22700
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.5941126346588135,
      "learning_rate": 0.0004533707501349163,
      "loss": 3.5505,
      "step": 22750
    },
    {
      "epoch": 2.4582210242587603,
      "grad_norm": 0.5647153258323669,
      "learning_rate": 0.0004530469508904479,
      "loss": 3.5483,
      "step": 22800
    },
    {
      "epoch": 2.463611859838275,
      "grad_norm": 0.6184137463569641,
      "learning_rate": 0.00045272315164597945,
      "loss": 3.5388,
      "step": 22850
    },
    {
      "epoch": 2.4690026954177897,
      "grad_norm": 0.5737881064414978,
      "learning_rate": 0.000452399352401511,
      "loss": 3.5513,
      "step": 22900
    },
    {
      "epoch": 2.4743935309973044,
      "grad_norm": 0.5692729353904724,
      "learning_rate": 0.0004520755531570426,
      "loss": 3.5391,
      "step": 22950
    },
    {
      "epoch": 2.4797843665768196,
      "grad_norm": 0.5915802121162415,
      "learning_rate": 0.00045175175391257415,
      "loss": 3.5531,
      "step": 23000
    },
    {
      "epoch": 2.4797843665768196,
      "eval_accuracy": 0.3655561218785519,
      "eval_loss": 3.5417704582214355,
      "eval_runtime": 183.6152,
      "eval_samples_per_second": 98.091,
      "eval_steps_per_second": 6.132,
      "step": 23000
    },
    {
      "epoch": 2.4851752021563343,
      "grad_norm": 0.59952312707901,
      "learning_rate": 0.00045142795466810576,
      "loss": 3.5615,
      "step": 23050
    },
    {
      "epoch": 2.490566037735849,
      "grad_norm": 0.6111100316047668,
      "learning_rate": 0.0004511041554236373,
      "loss": 3.5572,
      "step": 23100
    },
    {
      "epoch": 2.4959568733153636,
      "grad_norm": 0.6241294145584106,
      "learning_rate": 0.0004507803561791689,
      "loss": 3.5702,
      "step": 23150
    },
    {
      "epoch": 2.501347708894879,
      "grad_norm": 0.5904486179351807,
      "learning_rate": 0.00045045655693470046,
      "loss": 3.5615,
      "step": 23200
    },
    {
      "epoch": 2.5067385444743935,
      "grad_norm": 0.5772064328193665,
      "learning_rate": 0.00045013275769023207,
      "loss": 3.5602,
      "step": 23250
    },
    {
      "epoch": 2.512129380053908,
      "grad_norm": 0.5506559014320374,
      "learning_rate": 0.00044980895844576356,
      "loss": 3.5598,
      "step": 23300
    },
    {
      "epoch": 2.5175202156334233,
      "grad_norm": 0.583018958568573,
      "learning_rate": 0.0004494851592012951,
      "loss": 3.5572,
      "step": 23350
    },
    {
      "epoch": 2.522911051212938,
      "grad_norm": 0.573809802532196,
      "learning_rate": 0.0004491613599568267,
      "loss": 3.535,
      "step": 23400
    },
    {
      "epoch": 2.5283018867924527,
      "grad_norm": 0.6045467257499695,
      "learning_rate": 0.00044883756071235827,
      "loss": 3.5499,
      "step": 23450
    },
    {
      "epoch": 2.533692722371968,
      "grad_norm": 0.5793042182922363,
      "learning_rate": 0.0004485137614678899,
      "loss": 3.5448,
      "step": 23500
    },
    {
      "epoch": 2.5390835579514826,
      "grad_norm": 0.5313010215759277,
      "learning_rate": 0.00044819643820831083,
      "loss": 3.5483,
      "step": 23550
    },
    {
      "epoch": 2.5444743935309972,
      "grad_norm": 0.6156229376792908,
      "learning_rate": 0.0004478726389638424,
      "loss": 3.5537,
      "step": 23600
    },
    {
      "epoch": 2.5498652291105124,
      "grad_norm": 0.5773780345916748,
      "learning_rate": 0.00044754883971937393,
      "loss": 3.5403,
      "step": 23650
    },
    {
      "epoch": 2.555256064690027,
      "grad_norm": 0.5939944386482239,
      "learning_rate": 0.00044722504047490553,
      "loss": 3.5289,
      "step": 23700
    },
    {
      "epoch": 2.560646900269542,
      "grad_norm": 0.6025189161300659,
      "learning_rate": 0.0004469012412304371,
      "loss": 3.5384,
      "step": 23750
    },
    {
      "epoch": 2.5660377358490565,
      "grad_norm": 0.5510246157646179,
      "learning_rate": 0.0004465774419859687,
      "loss": 3.5567,
      "step": 23800
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.5493279695510864,
      "learning_rate": 0.00044625364274150024,
      "loss": 3.5405,
      "step": 23850
    },
    {
      "epoch": 2.5768194070080863,
      "grad_norm": 0.5634421706199646,
      "learning_rate": 0.00044592984349703184,
      "loss": 3.5443,
      "step": 23900
    },
    {
      "epoch": 2.582210242587601,
      "grad_norm": 0.6100355982780457,
      "learning_rate": 0.00044560604425256334,
      "loss": 3.5354,
      "step": 23950
    },
    {
      "epoch": 2.5876010781671157,
      "grad_norm": 0.5959258079528809,
      "learning_rate": 0.000445282245008095,
      "loss": 3.5406,
      "step": 24000
    },
    {
      "epoch": 2.5876010781671157,
      "eval_accuracy": 0.3667099068942539,
      "eval_loss": 3.5328526496887207,
      "eval_runtime": 183.3949,
      "eval_samples_per_second": 98.209,
      "eval_steps_per_second": 6.14,
      "step": 24000
    },
    {
      "epoch": 2.592991913746631,
      "grad_norm": 0.5708218812942505,
      "learning_rate": 0.0004449584457636265,
      "loss": 3.5664,
      "step": 24050
    },
    {
      "epoch": 2.5983827493261455,
      "grad_norm": 0.5921903252601624,
      "learning_rate": 0.00044463464651915805,
      "loss": 3.5394,
      "step": 24100
    },
    {
      "epoch": 2.6037735849056602,
      "grad_norm": 0.5770680904388428,
      "learning_rate": 0.00044431084727468965,
      "loss": 3.5536,
      "step": 24150
    },
    {
      "epoch": 2.6091644204851754,
      "grad_norm": 0.5639709234237671,
      "learning_rate": 0.0004439870480302212,
      "loss": 3.5388,
      "step": 24200
    },
    {
      "epoch": 2.61455525606469,
      "grad_norm": 0.5982383489608765,
      "learning_rate": 0.0004436632487857528,
      "loss": 3.5558,
      "step": 24250
    },
    {
      "epoch": 2.6199460916442048,
      "grad_norm": 0.6448655724525452,
      "learning_rate": 0.00044333944954128436,
      "loss": 3.5496,
      "step": 24300
    },
    {
      "epoch": 2.62533692722372,
      "grad_norm": 0.5641359686851501,
      "learning_rate": 0.00044301565029681596,
      "loss": 3.543,
      "step": 24350
    },
    {
      "epoch": 2.6307277628032346,
      "grad_norm": 0.5653684139251709,
      "learning_rate": 0.0004426918510523475,
      "loss": 3.545,
      "step": 24400
    },
    {
      "epoch": 2.6361185983827493,
      "grad_norm": 0.5824022889137268,
      "learning_rate": 0.00044237452779276846,
      "loss": 3.5449,
      "step": 24450
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.560650646686554,
      "learning_rate": 0.0004420507285483,
      "loss": 3.5525,
      "step": 24500
    },
    {
      "epoch": 2.6469002695417787,
      "grad_norm": 0.5790745615959167,
      "learning_rate": 0.0004417269293038316,
      "loss": 3.537,
      "step": 24550
    },
    {
      "epoch": 2.652291105121294,
      "grad_norm": 0.6426572799682617,
      "learning_rate": 0.0004414031300593631,
      "loss": 3.5435,
      "step": 24600
    },
    {
      "epoch": 2.6576819407008085,
      "grad_norm": 0.5720652341842651,
      "learning_rate": 0.0004410793308148948,
      "loss": 3.5407,
      "step": 24650
    },
    {
      "epoch": 2.6630727762803232,
      "grad_norm": 0.5447050929069519,
      "learning_rate": 0.00044075553157042627,
      "loss": 3.5486,
      "step": 24700
    },
    {
      "epoch": 2.6684636118598384,
      "grad_norm": 0.602033793926239,
      "learning_rate": 0.0004404317323259579,
      "loss": 3.5622,
      "step": 24750
    },
    {
      "epoch": 2.673854447439353,
      "grad_norm": 0.5450906157493591,
      "learning_rate": 0.0004401079330814894,
      "loss": 3.5564,
      "step": 24800
    },
    {
      "epoch": 2.6792452830188678,
      "grad_norm": 0.5530940890312195,
      "learning_rate": 0.000439784133837021,
      "loss": 3.5576,
      "step": 24850
    },
    {
      "epoch": 2.684636118598383,
      "grad_norm": 0.5873911380767822,
      "learning_rate": 0.0004394603345925526,
      "loss": 3.5276,
      "step": 24900
    },
    {
      "epoch": 2.6900269541778976,
      "grad_norm": 0.5609381198883057,
      "learning_rate": 0.00043913653534808413,
      "loss": 3.5476,
      "step": 24950
    },
    {
      "epoch": 2.6954177897574123,
      "grad_norm": 0.5607409477233887,
      "learning_rate": 0.00043881273610361574,
      "loss": 3.5264,
      "step": 25000
    },
    {
      "epoch": 2.6954177897574123,
      "eval_accuracy": 0.36771798838739667,
      "eval_loss": 3.5224156379699707,
      "eval_runtime": 183.5829,
      "eval_samples_per_second": 98.108,
      "eval_steps_per_second": 6.133,
      "step": 25000
    },
    {
      "epoch": 2.7008086253369274,
      "grad_norm": 0.6033768057823181,
      "learning_rate": 0.0004384889368591473,
      "loss": 3.5403,
      "step": 25050
    },
    {
      "epoch": 2.706199460916442,
      "grad_norm": 0.5833696722984314,
      "learning_rate": 0.0004381651376146789,
      "loss": 3.5425,
      "step": 25100
    },
    {
      "epoch": 2.711590296495957,
      "grad_norm": 0.5619771480560303,
      "learning_rate": 0.00043784133837021044,
      "loss": 3.5439,
      "step": 25150
    },
    {
      "epoch": 2.7169811320754715,
      "grad_norm": 0.6554853320121765,
      "learning_rate": 0.00043751753912574205,
      "loss": 3.5431,
      "step": 25200
    },
    {
      "epoch": 2.7223719676549867,
      "grad_norm": 0.627469003200531,
      "learning_rate": 0.0004371937398812736,
      "loss": 3.5358,
      "step": 25250
    },
    {
      "epoch": 2.7277628032345014,
      "grad_norm": 0.6013321280479431,
      "learning_rate": 0.0004368699406368051,
      "loss": 3.5432,
      "step": 25300
    },
    {
      "epoch": 2.733153638814016,
      "grad_norm": 0.6281169056892395,
      "learning_rate": 0.0004365461413923367,
      "loss": 3.5394,
      "step": 25350
    },
    {
      "epoch": 2.7385444743935308,
      "grad_norm": 0.562379002571106,
      "learning_rate": 0.00043622234214786825,
      "loss": 3.5336,
      "step": 25400
    },
    {
      "epoch": 2.743935309973046,
      "grad_norm": 0.5520660281181335,
      "learning_rate": 0.00043589854290339985,
      "loss": 3.5501,
      "step": 25450
    },
    {
      "epoch": 2.7493261455525606,
      "grad_norm": 0.5592831969261169,
      "learning_rate": 0.0004355747436589314,
      "loss": 3.5281,
      "step": 25500
    },
    {
      "epoch": 2.7547169811320753,
      "grad_norm": 0.6207462549209595,
      "learning_rate": 0.000435250944414463,
      "loss": 3.519,
      "step": 25550
    },
    {
      "epoch": 2.7601078167115904,
      "grad_norm": 0.5609697103500366,
      "learning_rate": 0.00043492714516999456,
      "loss": 3.5249,
      "step": 25600
    },
    {
      "epoch": 2.765498652291105,
      "grad_norm": 0.5504409074783325,
      "learning_rate": 0.0004346033459255261,
      "loss": 3.5315,
      "step": 25650
    },
    {
      "epoch": 2.77088948787062,
      "grad_norm": 0.5725831985473633,
      "learning_rate": 0.0004342795466810577,
      "loss": 3.5187,
      "step": 25700
    },
    {
      "epoch": 2.776280323450135,
      "grad_norm": 0.5668317675590515,
      "learning_rate": 0.00043395574743658927,
      "loss": 3.5219,
      "step": 25750
    },
    {
      "epoch": 2.7816711590296497,
      "grad_norm": 0.583371639251709,
      "learning_rate": 0.00043363194819212087,
      "loss": 3.5371,
      "step": 25800
    },
    {
      "epoch": 2.7870619946091644,
      "grad_norm": 0.5703887343406677,
      "learning_rate": 0.0004333081489476524,
      "loss": 3.522,
      "step": 25850
    },
    {
      "epoch": 2.7924528301886795,
      "grad_norm": 0.5865967869758606,
      "learning_rate": 0.000432984349703184,
      "loss": 3.5239,
      "step": 25900
    },
    {
      "epoch": 2.797843665768194,
      "grad_norm": 0.5793086886405945,
      "learning_rate": 0.0004326605504587155,
      "loss": 3.5336,
      "step": 25950
    },
    {
      "epoch": 2.803234501347709,
      "grad_norm": 0.6235855221748352,
      "learning_rate": 0.0004323367512142472,
      "loss": 3.5396,
      "step": 26000
    },
    {
      "epoch": 2.803234501347709,
      "eval_accuracy": 0.3685781933002239,
      "eval_loss": 3.5141384601593018,
      "eval_runtime": 183.8236,
      "eval_samples_per_second": 97.98,
      "eval_steps_per_second": 6.125,
      "step": 26000
    },
    {
      "epoch": 2.8086253369272236,
      "grad_norm": 0.6220427751541138,
      "learning_rate": 0.0004320129519697787,
      "loss": 3.5379,
      "step": 26050
    },
    {
      "epoch": 2.8140161725067383,
      "grad_norm": 0.5779548287391663,
      "learning_rate": 0.00043168915272531023,
      "loss": 3.5207,
      "step": 26100
    },
    {
      "epoch": 2.8194070080862534,
      "grad_norm": 0.573950469493866,
      "learning_rate": 0.00043136535348084183,
      "loss": 3.5309,
      "step": 26150
    },
    {
      "epoch": 2.824797843665768,
      "grad_norm": 0.6422613859176636,
      "learning_rate": 0.0004310415542363734,
      "loss": 3.5288,
      "step": 26200
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.6020054817199707,
      "learning_rate": 0.000430717754991905,
      "loss": 3.5225,
      "step": 26250
    },
    {
      "epoch": 2.835579514824798,
      "grad_norm": 0.5860601663589478,
      "learning_rate": 0.00043039395574743654,
      "loss": 3.5358,
      "step": 26300
    },
    {
      "epoch": 2.8409703504043127,
      "grad_norm": 0.5909083485603333,
      "learning_rate": 0.00043007015650296814,
      "loss": 3.5416,
      "step": 26350
    },
    {
      "epoch": 2.8463611859838274,
      "grad_norm": 0.5843289494514465,
      "learning_rate": 0.0004297463572584997,
      "loss": 3.5282,
      "step": 26400
    },
    {
      "epoch": 2.8517520215633425,
      "grad_norm": 0.5413671731948853,
      "learning_rate": 0.0004294225580140313,
      "loss": 3.5301,
      "step": 26450
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6106083393096924,
      "learning_rate": 0.00042909875876956285,
      "loss": 3.5269,
      "step": 26500
    },
    {
      "epoch": 2.862533692722372,
      "grad_norm": 0.5917198061943054,
      "learning_rate": 0.0004287749595250944,
      "loss": 3.5182,
      "step": 26550
    },
    {
      "epoch": 2.867924528301887,
      "grad_norm": 0.5803675651550293,
      "learning_rate": 0.000428451160280626,
      "loss": 3.528,
      "step": 26600
    },
    {
      "epoch": 2.8733153638814017,
      "grad_norm": 0.6025941371917725,
      "learning_rate": 0.0004281273610361575,
      "loss": 3.524,
      "step": 26650
    },
    {
      "epoch": 2.8787061994609164,
      "grad_norm": 0.5806294679641724,
      "learning_rate": 0.0004278035617916891,
      "loss": 3.5223,
      "step": 26700
    },
    {
      "epoch": 2.884097035040431,
      "grad_norm": 0.5862604975700378,
      "learning_rate": 0.00042747976254722066,
      "loss": 3.5418,
      "step": 26750
    },
    {
      "epoch": 2.889487870619946,
      "grad_norm": 0.6699230670928955,
      "learning_rate": 0.00042715596330275226,
      "loss": 3.5272,
      "step": 26800
    },
    {
      "epoch": 2.894878706199461,
      "grad_norm": 0.5533138513565063,
      "learning_rate": 0.0004268321640582838,
      "loss": 3.5215,
      "step": 26850
    },
    {
      "epoch": 2.9002695417789757,
      "grad_norm": 0.6020789742469788,
      "learning_rate": 0.0004265083648138154,
      "loss": 3.5449,
      "step": 26900
    },
    {
      "epoch": 2.9056603773584904,
      "grad_norm": 0.5816729664802551,
      "learning_rate": 0.00042618456556934697,
      "loss": 3.5387,
      "step": 26950
    },
    {
      "epoch": 2.9110512129380055,
      "grad_norm": 0.5571572184562683,
      "learning_rate": 0.0004258607663248785,
      "loss": 3.5222,
      "step": 27000
    },
    {
      "epoch": 2.9110512129380055,
      "eval_accuracy": 0.36945111060092545,
      "eval_loss": 3.506294012069702,
      "eval_runtime": 183.5095,
      "eval_samples_per_second": 98.148,
      "eval_steps_per_second": 6.136,
      "step": 27000
    },
    {
      "epoch": 2.91644204851752,
      "grad_norm": 0.6122097969055176,
      "learning_rate": 0.0004255369670804101,
      "loss": 3.5398,
      "step": 27050
    },
    {
      "epoch": 2.921832884097035,
      "grad_norm": 0.5753620862960815,
      "learning_rate": 0.0004252131678359417,
      "loss": 3.5455,
      "step": 27100
    },
    {
      "epoch": 2.92722371967655,
      "grad_norm": 0.5587215423583984,
      "learning_rate": 0.0004248893685914733,
      "loss": 3.529,
      "step": 27150
    },
    {
      "epoch": 2.9326145552560647,
      "grad_norm": 0.5809001326560974,
      "learning_rate": 0.00042456556934700483,
      "loss": 3.5189,
      "step": 27200
    },
    {
      "epoch": 2.9380053908355794,
      "grad_norm": 0.5886865258216858,
      "learning_rate": 0.00042424177010253643,
      "loss": 3.5316,
      "step": 27250
    },
    {
      "epoch": 2.9433962264150946,
      "grad_norm": 0.6303690075874329,
      "learning_rate": 0.00042391797085806793,
      "loss": 3.5273,
      "step": 27300
    },
    {
      "epoch": 2.9487870619946093,
      "grad_norm": 0.6268852353096008,
      "learning_rate": 0.0004235941716135995,
      "loss": 3.521,
      "step": 27350
    },
    {
      "epoch": 2.954177897574124,
      "grad_norm": 0.5647582411766052,
      "learning_rate": 0.0004232703723691311,
      "loss": 3.5257,
      "step": 27400
    },
    {
      "epoch": 2.9595687331536387,
      "grad_norm": 0.5563523173332214,
      "learning_rate": 0.00042295304910955204,
      "loss": 3.5211,
      "step": 27450
    },
    {
      "epoch": 2.964959568733154,
      "grad_norm": 0.5889139771461487,
      "learning_rate": 0.0004226292498650836,
      "loss": 3.5278,
      "step": 27500
    },
    {
      "epoch": 2.9703504043126685,
      "grad_norm": 0.5559007525444031,
      "learning_rate": 0.0004223054506206152,
      "loss": 3.5315,
      "step": 27550
    },
    {
      "epoch": 2.975741239892183,
      "grad_norm": 0.5638795495033264,
      "learning_rate": 0.00042198165137614674,
      "loss": 3.5319,
      "step": 27600
    },
    {
      "epoch": 2.981132075471698,
      "grad_norm": 0.5552601218223572,
      "learning_rate": 0.00042165785213167835,
      "loss": 3.5331,
      "step": 27650
    },
    {
      "epoch": 2.986522911051213,
      "grad_norm": 0.5989307165145874,
      "learning_rate": 0.0004213340528872099,
      "loss": 3.5241,
      "step": 27700
    },
    {
      "epoch": 2.9919137466307277,
      "grad_norm": 0.5570740103721619,
      "learning_rate": 0.00042101025364274145,
      "loss": 3.5344,
      "step": 27750
    },
    {
      "epoch": 2.9973045822102424,
      "grad_norm": 0.6432395577430725,
      "learning_rate": 0.00042068645439827305,
      "loss": 3.531,
      "step": 27800
    },
    {
      "epoch": 3.0026954177897576,
      "grad_norm": 0.5959632396697998,
      "learning_rate": 0.0004203626551538046,
      "loss": 3.4774,
      "step": 27850
    },
    {
      "epoch": 3.0080862533692723,
      "grad_norm": 0.5988036394119263,
      "learning_rate": 0.0004200388559093362,
      "loss": 3.4268,
      "step": 27900
    },
    {
      "epoch": 3.013477088948787,
      "grad_norm": 0.6331761479377747,
      "learning_rate": 0.0004197150566648677,
      "loss": 3.4278,
      "step": 27950
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.5820286273956299,
      "learning_rate": 0.00041939125742039936,
      "loss": 3.4471,
      "step": 28000
    },
    {
      "epoch": 3.018867924528302,
      "eval_accuracy": 0.37059087939409935,
      "eval_loss": 3.5022330284118652,
      "eval_runtime": 183.4286,
      "eval_samples_per_second": 98.191,
      "eval_steps_per_second": 6.139,
      "step": 28000
    },
    {
      "epoch": 3.024258760107817,
      "grad_norm": 0.5634085536003113,
      "learning_rate": 0.00041906745817593086,
      "loss": 3.441,
      "step": 28050
    },
    {
      "epoch": 3.0296495956873315,
      "grad_norm": 0.5626387000083923,
      "learning_rate": 0.00041874365893146247,
      "loss": 3.4477,
      "step": 28100
    },
    {
      "epoch": 3.035040431266846,
      "grad_norm": 0.6222658157348633,
      "learning_rate": 0.000418419859686994,
      "loss": 3.4389,
      "step": 28150
    },
    {
      "epoch": 3.0404312668463613,
      "grad_norm": 0.6200679540634155,
      "learning_rate": 0.00041809606044252557,
      "loss": 3.4408,
      "step": 28200
    },
    {
      "epoch": 3.045822102425876,
      "grad_norm": 0.6300202012062073,
      "learning_rate": 0.00041777226119805717,
      "loss": 3.4246,
      "step": 28250
    },
    {
      "epoch": 3.0512129380053907,
      "grad_norm": 0.5943900942802429,
      "learning_rate": 0.0004174484619535887,
      "loss": 3.4441,
      "step": 28300
    },
    {
      "epoch": 3.056603773584906,
      "grad_norm": 0.5677528977394104,
      "learning_rate": 0.00041712466270912033,
      "loss": 3.4205,
      "step": 28350
    },
    {
      "epoch": 3.0619946091644206,
      "grad_norm": 0.6124181151390076,
      "learning_rate": 0.0004168008634646519,
      "loss": 3.449,
      "step": 28400
    },
    {
      "epoch": 3.0673854447439353,
      "grad_norm": 0.6310901641845703,
      "learning_rate": 0.0004164770642201835,
      "loss": 3.4178,
      "step": 28450
    },
    {
      "epoch": 3.07277628032345,
      "grad_norm": 0.6272411942481995,
      "learning_rate": 0.00041615326497571503,
      "loss": 3.4442,
      "step": 28500
    },
    {
      "epoch": 3.078167115902965,
      "grad_norm": 0.6128734350204468,
      "learning_rate": 0.0004158294657312466,
      "loss": 3.4544,
      "step": 28550
    },
    {
      "epoch": 3.08355795148248,
      "grad_norm": 0.6625658273696899,
      "learning_rate": 0.0004155056664867782,
      "loss": 3.439,
      "step": 28600
    },
    {
      "epoch": 3.0889487870619945,
      "grad_norm": 0.6313899755477905,
      "learning_rate": 0.0004151818672423097,
      "loss": 3.4497,
      "step": 28650
    },
    {
      "epoch": 3.0943396226415096,
      "grad_norm": 0.5638771653175354,
      "learning_rate": 0.0004148580679978413,
      "loss": 3.4477,
      "step": 28700
    },
    {
      "epoch": 3.0997304582210243,
      "grad_norm": 0.595225989818573,
      "learning_rate": 0.00041453426875337284,
      "loss": 3.4742,
      "step": 28750
    },
    {
      "epoch": 3.105121293800539,
      "grad_norm": 0.6290755867958069,
      "learning_rate": 0.00041421046950890445,
      "loss": 3.4346,
      "step": 28800
    },
    {
      "epoch": 3.1105121293800537,
      "grad_norm": 0.6257621645927429,
      "learning_rate": 0.000413886670264436,
      "loss": 3.4416,
      "step": 28850
    },
    {
      "epoch": 3.115902964959569,
      "grad_norm": 0.6242688894271851,
      "learning_rate": 0.0004135628710199676,
      "loss": 3.4387,
      "step": 28900
    },
    {
      "epoch": 3.1212938005390836,
      "grad_norm": 0.6004605293273926,
      "learning_rate": 0.00041323907177549915,
      "loss": 3.4676,
      "step": 28950
    },
    {
      "epoch": 3.1266846361185983,
      "grad_norm": 0.6270654797554016,
      "learning_rate": 0.0004129152725310307,
      "loss": 3.4396,
      "step": 29000
    },
    {
      "epoch": 3.1266846361185983,
      "eval_accuracy": 0.37067845362167784,
      "eval_loss": 3.495734453201294,
      "eval_runtime": 183.7423,
      "eval_samples_per_second": 98.023,
      "eval_steps_per_second": 6.128,
      "step": 29000
    },
    {
      "epoch": 3.1320754716981134,
      "grad_norm": 0.6041761040687561,
      "learning_rate": 0.0004125914732865623,
      "loss": 3.4579,
      "step": 29050
    },
    {
      "epoch": 3.137466307277628,
      "grad_norm": 0.5562618970870972,
      "learning_rate": 0.00041226767404209386,
      "loss": 3.4598,
      "step": 29100
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.6546818017959595,
      "learning_rate": 0.00041194387479762546,
      "loss": 3.4509,
      "step": 29150
    },
    {
      "epoch": 3.1482479784366575,
      "grad_norm": 0.6424098610877991,
      "learning_rate": 0.000411620075553157,
      "loss": 3.4595,
      "step": 29200
    },
    {
      "epoch": 3.1536388140161726,
      "grad_norm": 0.593454897403717,
      "learning_rate": 0.0004112962763086886,
      "loss": 3.4614,
      "step": 29250
    },
    {
      "epoch": 3.1590296495956873,
      "grad_norm": 0.5967779159545898,
      "learning_rate": 0.0004109724770642201,
      "loss": 3.4656,
      "step": 29300
    },
    {
      "epoch": 3.164420485175202,
      "grad_norm": 0.6650448441505432,
      "learning_rate": 0.00041064867781975177,
      "loss": 3.4551,
      "step": 29350
    },
    {
      "epoch": 3.169811320754717,
      "grad_norm": 0.5960310697555542,
      "learning_rate": 0.00041032487857528327,
      "loss": 3.4487,
      "step": 29400
    },
    {
      "epoch": 3.175202156334232,
      "grad_norm": 0.6126105189323425,
      "learning_rate": 0.0004100010793308148,
      "loss": 3.4341,
      "step": 29450
    },
    {
      "epoch": 3.1805929919137466,
      "grad_norm": 0.5907238125801086,
      "learning_rate": 0.0004096772800863464,
      "loss": 3.4591,
      "step": 29500
    },
    {
      "epoch": 3.1859838274932613,
      "grad_norm": 0.5988099575042725,
      "learning_rate": 0.000409353480841878,
      "loss": 3.4454,
      "step": 29550
    },
    {
      "epoch": 3.1913746630727764,
      "grad_norm": 0.587455689907074,
      "learning_rate": 0.0004090296815974096,
      "loss": 3.4524,
      "step": 29600
    },
    {
      "epoch": 3.196765498652291,
      "grad_norm": 0.5905243754386902,
      "learning_rate": 0.00040870588235294113,
      "loss": 3.467,
      "step": 29650
    },
    {
      "epoch": 3.202156334231806,
      "grad_norm": 0.6206773519515991,
      "learning_rate": 0.00040838208310847273,
      "loss": 3.4518,
      "step": 29700
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.6231102347373962,
      "learning_rate": 0.0004080582838640043,
      "loss": 3.4605,
      "step": 29750
    },
    {
      "epoch": 3.2129380053908356,
      "grad_norm": 0.641909658908844,
      "learning_rate": 0.0004077344846195359,
      "loss": 3.4698,
      "step": 29800
    },
    {
      "epoch": 3.2183288409703503,
      "grad_norm": 0.6328471302986145,
      "learning_rate": 0.00040741068537506744,
      "loss": 3.4754,
      "step": 29850
    },
    {
      "epoch": 3.223719676549865,
      "grad_norm": 0.5950383543968201,
      "learning_rate": 0.000407086886130599,
      "loss": 3.4404,
      "step": 29900
    },
    {
      "epoch": 3.22911051212938,
      "grad_norm": 0.5742519497871399,
      "learning_rate": 0.0004067630868861306,
      "loss": 3.4444,
      "step": 29950
    },
    {
      "epoch": 3.234501347708895,
      "grad_norm": 0.5945967435836792,
      "learning_rate": 0.0004064392876416621,
      "loss": 3.4687,
      "step": 30000
    },
    {
      "epoch": 3.234501347708895,
      "eval_accuracy": 0.3716783861482345,
      "eval_loss": 3.490046501159668,
      "eval_runtime": 183.4166,
      "eval_samples_per_second": 98.197,
      "eval_steps_per_second": 6.139,
      "step": 30000
    },
    {
      "epoch": 3.2398921832884096,
      "grad_norm": 0.5931757092475891,
      "learning_rate": 0.0004061154883971937,
      "loss": 3.456,
      "step": 30050
    },
    {
      "epoch": 3.2452830188679247,
      "grad_norm": 0.5897646546363831,
      "learning_rate": 0.00040579168915272525,
      "loss": 3.4629,
      "step": 30100
    },
    {
      "epoch": 3.2506738544474394,
      "grad_norm": 0.6198738217353821,
      "learning_rate": 0.00040546788990825685,
      "loss": 3.4665,
      "step": 30150
    },
    {
      "epoch": 3.256064690026954,
      "grad_norm": 0.5888984799385071,
      "learning_rate": 0.0004051440906637884,
      "loss": 3.4521,
      "step": 30200
    },
    {
      "epoch": 3.2614555256064692,
      "grad_norm": 0.6324925422668457,
      "learning_rate": 0.00040482029141931995,
      "loss": 3.4564,
      "step": 30250
    },
    {
      "epoch": 3.266846361185984,
      "grad_norm": 0.6119275093078613,
      "learning_rate": 0.00040449649217485156,
      "loss": 3.4414,
      "step": 30300
    },
    {
      "epoch": 3.2722371967654986,
      "grad_norm": 0.5899844169616699,
      "learning_rate": 0.0004041726929303831,
      "loss": 3.4562,
      "step": 30350
    },
    {
      "epoch": 3.2776280323450133,
      "grad_norm": 0.6019811630249023,
      "learning_rate": 0.00040385536967080406,
      "loss": 3.4674,
      "step": 30400
    },
    {
      "epoch": 3.2830188679245285,
      "grad_norm": 0.6066054701805115,
      "learning_rate": 0.00040353157042633567,
      "loss": 3.4593,
      "step": 30450
    },
    {
      "epoch": 3.288409703504043,
      "grad_norm": 0.6244698762893677,
      "learning_rate": 0.0004032077711818672,
      "loss": 3.4782,
      "step": 30500
    },
    {
      "epoch": 3.293800539083558,
      "grad_norm": 0.6058152318000793,
      "learning_rate": 0.0004028839719373988,
      "loss": 3.4692,
      "step": 30550
    },
    {
      "epoch": 3.2991913746630726,
      "grad_norm": 0.5960109233856201,
      "learning_rate": 0.00040256017269293037,
      "loss": 3.4614,
      "step": 30600
    },
    {
      "epoch": 3.3045822102425877,
      "grad_norm": 0.6146271824836731,
      "learning_rate": 0.00040223637344846187,
      "loss": 3.462,
      "step": 30650
    },
    {
      "epoch": 3.3099730458221024,
      "grad_norm": 0.6514822244644165,
      "learning_rate": 0.00040191257420399347,
      "loss": 3.4523,
      "step": 30700
    },
    {
      "epoch": 3.315363881401617,
      "grad_norm": 0.6561552882194519,
      "learning_rate": 0.000401588774959525,
      "loss": 3.4581,
      "step": 30750
    },
    {
      "epoch": 3.3207547169811322,
      "grad_norm": 0.6325382590293884,
      "learning_rate": 0.00040126497571505663,
      "loss": 3.4593,
      "step": 30800
    },
    {
      "epoch": 3.326145552560647,
      "grad_norm": 0.5675181746482849,
      "learning_rate": 0.0004009411764705882,
      "loss": 3.4515,
      "step": 30850
    },
    {
      "epoch": 3.3315363881401616,
      "grad_norm": 0.5973735451698303,
      "learning_rate": 0.0004006173772261198,
      "loss": 3.4729,
      "step": 30900
    },
    {
      "epoch": 3.3369272237196768,
      "grad_norm": 0.5759937763214111,
      "learning_rate": 0.00040029357798165133,
      "loss": 3.4586,
      "step": 30950
    },
    {
      "epoch": 3.3423180592991915,
      "grad_norm": 0.6003343462944031,
      "learning_rate": 0.00039996977873718294,
      "loss": 3.4757,
      "step": 31000
    },
    {
      "epoch": 3.3423180592991915,
      "eval_accuracy": 0.37271232702867707,
      "eval_loss": 3.483132839202881,
      "eval_runtime": 183.5871,
      "eval_samples_per_second": 98.106,
      "eval_steps_per_second": 6.133,
      "step": 31000
    },
    {
      "epoch": 3.347708894878706,
      "grad_norm": 0.6348472833633423,
      "learning_rate": 0.0003996459794927145,
      "loss": 3.4504,
      "step": 31050
    },
    {
      "epoch": 3.353099730458221,
      "grad_norm": 0.6023390889167786,
      "learning_rate": 0.00039932218024824604,
      "loss": 3.4698,
      "step": 31100
    },
    {
      "epoch": 3.358490566037736,
      "grad_norm": 0.6116836071014404,
      "learning_rate": 0.00039899838100377764,
      "loss": 3.4493,
      "step": 31150
    },
    {
      "epoch": 3.3638814016172507,
      "grad_norm": 0.5540566444396973,
      "learning_rate": 0.0003986745817593092,
      "loss": 3.4486,
      "step": 31200
    },
    {
      "epoch": 3.3692722371967654,
      "grad_norm": 0.6478095054626465,
      "learning_rate": 0.0003983507825148408,
      "loss": 3.4682,
      "step": 31250
    },
    {
      "epoch": 3.37466307277628,
      "grad_norm": 0.5724659562110901,
      "learning_rate": 0.00039802698327037235,
      "loss": 3.4532,
      "step": 31300
    },
    {
      "epoch": 3.3800539083557952,
      "grad_norm": 0.5920699238777161,
      "learning_rate": 0.00039770318402590396,
      "loss": 3.4404,
      "step": 31350
    },
    {
      "epoch": 3.38544474393531,
      "grad_norm": 0.6260170936584473,
      "learning_rate": 0.00039737938478143545,
      "loss": 3.449,
      "step": 31400
    },
    {
      "epoch": 3.3908355795148246,
      "grad_norm": 0.5793566703796387,
      "learning_rate": 0.000397055585536967,
      "loss": 3.4676,
      "step": 31450
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.5834464430809021,
      "learning_rate": 0.0003967317862924986,
      "loss": 3.4691,
      "step": 31500
    },
    {
      "epoch": 3.4016172506738545,
      "grad_norm": 0.6603100895881653,
      "learning_rate": 0.00039640798704803016,
      "loss": 3.4556,
      "step": 31550
    },
    {
      "epoch": 3.407008086253369,
      "grad_norm": 0.5993121266365051,
      "learning_rate": 0.00039608418780356176,
      "loss": 3.4406,
      "step": 31600
    },
    {
      "epoch": 3.4123989218328843,
      "grad_norm": 0.661611020565033,
      "learning_rate": 0.0003957603885590933,
      "loss": 3.4628,
      "step": 31650
    },
    {
      "epoch": 3.417789757412399,
      "grad_norm": 0.6264667510986328,
      "learning_rate": 0.0003954365893146249,
      "loss": 3.4674,
      "step": 31700
    },
    {
      "epoch": 3.4231805929919137,
      "grad_norm": 0.5887061357498169,
      "learning_rate": 0.00039511279007015647,
      "loss": 3.4641,
      "step": 31750
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.6657390594482422,
      "learning_rate": 0.00039478899082568807,
      "loss": 3.453,
      "step": 31800
    },
    {
      "epoch": 3.4339622641509435,
      "grad_norm": 0.5978161692619324,
      "learning_rate": 0.0003944651915812196,
      "loss": 3.4537,
      "step": 31850
    },
    {
      "epoch": 3.439353099730458,
      "grad_norm": 0.5931206345558167,
      "learning_rate": 0.0003941413923367512,
      "loss": 3.4645,
      "step": 31900
    },
    {
      "epoch": 3.444743935309973,
      "grad_norm": 0.5882127285003662,
      "learning_rate": 0.0003938175930922828,
      "loss": 3.4478,
      "step": 31950
    },
    {
      "epoch": 3.450134770889488,
      "grad_norm": 0.5751291513442993,
      "learning_rate": 0.0003934937938478143,
      "loss": 3.4616,
      "step": 32000
    },
    {
      "epoch": 3.450134770889488,
      "eval_accuracy": 0.3731474818443741,
      "eval_loss": 3.476580858230591,
      "eval_runtime": 183.6874,
      "eval_samples_per_second": 98.052,
      "eval_steps_per_second": 6.13,
      "step": 32000
    },
    {
      "epoch": 3.4555256064690028,
      "grad_norm": 0.5731863975524902,
      "learning_rate": 0.0003931699946033459,
      "loss": 3.4844,
      "step": 32050
    },
    {
      "epoch": 3.4609164420485174,
      "grad_norm": 0.6018608212471008,
      "learning_rate": 0.00039284619535887743,
      "loss": 3.4735,
      "step": 32100
    },
    {
      "epoch": 3.466307277628032,
      "grad_norm": 0.5918667316436768,
      "learning_rate": 0.00039252239611440904,
      "loss": 3.4458,
      "step": 32150
    },
    {
      "epoch": 3.4716981132075473,
      "grad_norm": 0.5810291171073914,
      "learning_rate": 0.0003921985968699406,
      "loss": 3.4621,
      "step": 32200
    },
    {
      "epoch": 3.477088948787062,
      "grad_norm": 0.5926442742347717,
      "learning_rate": 0.0003918747976254722,
      "loss": 3.4497,
      "step": 32250
    },
    {
      "epoch": 3.4824797843665767,
      "grad_norm": 0.6344408392906189,
      "learning_rate": 0.00039155099838100374,
      "loss": 3.4549,
      "step": 32300
    },
    {
      "epoch": 3.487870619946092,
      "grad_norm": 0.5833978652954102,
      "learning_rate": 0.0003912271991365353,
      "loss": 3.4461,
      "step": 32350
    },
    {
      "epoch": 3.4932614555256065,
      "grad_norm": 0.6158148646354675,
      "learning_rate": 0.0003909033998920669,
      "loss": 3.4513,
      "step": 32400
    },
    {
      "epoch": 3.498652291105121,
      "grad_norm": 0.6152658462524414,
      "learning_rate": 0.00039057960064759845,
      "loss": 3.4622,
      "step": 32450
    },
    {
      "epoch": 3.5040431266846364,
      "grad_norm": 0.618352472782135,
      "learning_rate": 0.00039025580140313005,
      "loss": 3.4679,
      "step": 32500
    },
    {
      "epoch": 3.509433962264151,
      "grad_norm": 0.6015332341194153,
      "learning_rate": 0.0003899320021586616,
      "loss": 3.4678,
      "step": 32550
    },
    {
      "epoch": 3.5148247978436657,
      "grad_norm": 0.6236855387687683,
      "learning_rate": 0.0003896082029141932,
      "loss": 3.4557,
      "step": 32600
    },
    {
      "epoch": 3.5202156334231804,
      "grad_norm": 0.6326050758361816,
      "learning_rate": 0.00038928440366972476,
      "loss": 3.4591,
      "step": 32650
    },
    {
      "epoch": 3.525606469002695,
      "grad_norm": 0.5875770449638367,
      "learning_rate": 0.00038896060442525636,
      "loss": 3.4755,
      "step": 32700
    },
    {
      "epoch": 3.5309973045822103,
      "grad_norm": 0.625442624092102,
      "learning_rate": 0.00038863680518078786,
      "loss": 3.4483,
      "step": 32750
    },
    {
      "epoch": 3.536388140161725,
      "grad_norm": 0.6161897778511047,
      "learning_rate": 0.0003883130059363194,
      "loss": 3.4769,
      "step": 32800
    },
    {
      "epoch": 3.5417789757412397,
      "grad_norm": 0.543728768825531,
      "learning_rate": 0.000387989206691851,
      "loss": 3.4769,
      "step": 32850
    },
    {
      "epoch": 3.547169811320755,
      "grad_norm": 0.5994245409965515,
      "learning_rate": 0.00038766540744738256,
      "loss": 3.4708,
      "step": 32900
    },
    {
      "epoch": 3.5525606469002695,
      "grad_norm": 0.6172114014625549,
      "learning_rate": 0.00038734160820291417,
      "loss": 3.4589,
      "step": 32950
    },
    {
      "epoch": 3.557951482479784,
      "grad_norm": 0.5996062755584717,
      "learning_rate": 0.0003870178089584457,
      "loss": 3.4542,
      "step": 33000
    },
    {
      "epoch": 3.557951482479784,
      "eval_accuracy": 0.3733641357026762,
      "eval_loss": 3.4711246490478516,
      "eval_runtime": 183.9565,
      "eval_samples_per_second": 97.909,
      "eval_steps_per_second": 6.121,
      "step": 33000
    },
    {
      "epoch": 3.5633423180592994,
      "grad_norm": 0.6201596260070801,
      "learning_rate": 0.0003866940097139773,
      "loss": 3.4653,
      "step": 33050
    },
    {
      "epoch": 3.568733153638814,
      "grad_norm": 0.6010172963142395,
      "learning_rate": 0.0003863702104695089,
      "loss": 3.4675,
      "step": 33100
    },
    {
      "epoch": 3.5741239892183287,
      "grad_norm": 0.6184124946594238,
      "learning_rate": 0.0003860464112250404,
      "loss": 3.47,
      "step": 33150
    },
    {
      "epoch": 3.579514824797844,
      "grad_norm": 0.6181142926216125,
      "learning_rate": 0.00038572261198057203,
      "loss": 3.4644,
      "step": 33200
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.596049964427948,
      "learning_rate": 0.0003853988127361036,
      "loss": 3.4499,
      "step": 33250
    },
    {
      "epoch": 3.5902964959568733,
      "grad_norm": 0.6390442848205566,
      "learning_rate": 0.0003850750134916352,
      "loss": 3.4757,
      "step": 33300
    },
    {
      "epoch": 3.595687331536388,
      "grad_norm": 0.6242977976799011,
      "learning_rate": 0.0003847512142471667,
      "loss": 3.4634,
      "step": 33350
    },
    {
      "epoch": 3.601078167115903,
      "grad_norm": 0.5911014080047607,
      "learning_rate": 0.0003844274150026983,
      "loss": 3.4587,
      "step": 33400
    },
    {
      "epoch": 3.606469002695418,
      "grad_norm": 0.6290136575698853,
      "learning_rate": 0.00038410361575822984,
      "loss": 3.4752,
      "step": 33450
    },
    {
      "epoch": 3.6118598382749325,
      "grad_norm": 0.5740649700164795,
      "learning_rate": 0.00038377981651376144,
      "loss": 3.453,
      "step": 33500
    },
    {
      "epoch": 3.617250673854447,
      "grad_norm": 0.6023151874542236,
      "learning_rate": 0.000383456017269293,
      "loss": 3.4519,
      "step": 33550
    },
    {
      "epoch": 3.6226415094339623,
      "grad_norm": 0.6273173093795776,
      "learning_rate": 0.00038313221802482454,
      "loss": 3.4473,
      "step": 33600
    },
    {
      "epoch": 3.628032345013477,
      "grad_norm": 0.6236704587936401,
      "learning_rate": 0.00038280841878035615,
      "loss": 3.4628,
      "step": 33650
    },
    {
      "epoch": 3.6334231805929917,
      "grad_norm": 0.6301255226135254,
      "learning_rate": 0.0003824846195358877,
      "loss": 3.4684,
      "step": 33700
    },
    {
      "epoch": 3.638814016172507,
      "grad_norm": 0.6301966905593872,
      "learning_rate": 0.0003821608202914193,
      "loss": 3.4473,
      "step": 33750
    },
    {
      "epoch": 3.6442048517520216,
      "grad_norm": 0.6245051026344299,
      "learning_rate": 0.00038183702104695085,
      "loss": 3.4498,
      "step": 33800
    },
    {
      "epoch": 3.6495956873315363,
      "grad_norm": 0.6175593733787537,
      "learning_rate": 0.00038151322180248246,
      "loss": 3.4379,
      "step": 33850
    },
    {
      "epoch": 3.6549865229110514,
      "grad_norm": 0.6254450082778931,
      "learning_rate": 0.000381189422558014,
      "loss": 3.4292,
      "step": 33900
    },
    {
      "epoch": 3.660377358490566,
      "grad_norm": 0.6150121092796326,
      "learning_rate": 0.0003808656233135456,
      "loss": 3.4558,
      "step": 33950
    },
    {
      "epoch": 3.665768194070081,
      "grad_norm": 0.6018447875976562,
      "learning_rate": 0.00038054182406907716,
      "loss": 3.4533,
      "step": 34000
    },
    {
      "epoch": 3.665768194070081,
      "eval_accuracy": 0.374140243280335,
      "eval_loss": 3.465543270111084,
      "eval_runtime": 183.4349,
      "eval_samples_per_second": 98.187,
      "eval_steps_per_second": 6.138,
      "step": 34000
    },
    {
      "epoch": 3.671159029649596,
      "grad_norm": 0.6516909599304199,
      "learning_rate": 0.00038021802482460866,
      "loss": 3.4546,
      "step": 34050
    },
    {
      "epoch": 3.6765498652291106,
      "grad_norm": 0.6079384088516235,
      "learning_rate": 0.00037989422558014027,
      "loss": 3.4524,
      "step": 34100
    },
    {
      "epoch": 3.6819407008086253,
      "grad_norm": 0.6471598148345947,
      "learning_rate": 0.0003795704263356718,
      "loss": 3.448,
      "step": 34150
    },
    {
      "epoch": 3.68733153638814,
      "grad_norm": 0.6753168702125549,
      "learning_rate": 0.0003792466270912034,
      "loss": 3.4608,
      "step": 34200
    },
    {
      "epoch": 3.6927223719676547,
      "grad_norm": 0.6510220766067505,
      "learning_rate": 0.00037892282784673497,
      "loss": 3.4502,
      "step": 34250
    },
    {
      "epoch": 3.69811320754717,
      "grad_norm": 0.6369727253913879,
      "learning_rate": 0.0003785990286022666,
      "loss": 3.454,
      "step": 34300
    },
    {
      "epoch": 3.7035040431266846,
      "grad_norm": 0.6588965654373169,
      "learning_rate": 0.00037827522935779813,
      "loss": 3.4524,
      "step": 34350
    },
    {
      "epoch": 3.7088948787061993,
      "grad_norm": 0.6100491881370544,
      "learning_rate": 0.0003779579060982191,
      "loss": 3.4605,
      "step": 34400
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.6216384768486023,
      "learning_rate": 0.00037763410685375063,
      "loss": 3.4318,
      "step": 34450
    },
    {
      "epoch": 3.719676549865229,
      "grad_norm": 0.645187258720398,
      "learning_rate": 0.00037731030760928223,
      "loss": 3.4734,
      "step": 34500
    },
    {
      "epoch": 3.725067385444744,
      "grad_norm": 0.6211811304092407,
      "learning_rate": 0.0003769865083648138,
      "loss": 3.4501,
      "step": 34550
    },
    {
      "epoch": 3.730458221024259,
      "grad_norm": 0.5801553726196289,
      "learning_rate": 0.0003766627091203454,
      "loss": 3.4566,
      "step": 34600
    },
    {
      "epoch": 3.7358490566037736,
      "grad_norm": 0.6076372265815735,
      "learning_rate": 0.00037633890987587694,
      "loss": 3.4587,
      "step": 34650
    },
    {
      "epoch": 3.7412398921832883,
      "grad_norm": 0.5914555788040161,
      "learning_rate": 0.00037601511063140855,
      "loss": 3.4643,
      "step": 34700
    },
    {
      "epoch": 3.7466307277628035,
      "grad_norm": 0.6262001991271973,
      "learning_rate": 0.00037569131138694004,
      "loss": 3.4698,
      "step": 34750
    },
    {
      "epoch": 3.752021563342318,
      "grad_norm": 0.5983145236968994,
      "learning_rate": 0.0003753675121424716,
      "loss": 3.4542,
      "step": 34800
    },
    {
      "epoch": 3.757412398921833,
      "grad_norm": 0.6043577790260315,
      "learning_rate": 0.0003750437128980032,
      "loss": 3.4627,
      "step": 34850
    },
    {
      "epoch": 3.7628032345013476,
      "grad_norm": 0.6562215089797974,
      "learning_rate": 0.00037471991365353475,
      "loss": 3.455,
      "step": 34900
    },
    {
      "epoch": 3.7681940700808623,
      "grad_norm": 0.6480857133865356,
      "learning_rate": 0.00037439611440906635,
      "loss": 3.446,
      "step": 34950
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.6734146475791931,
      "learning_rate": 0.0003740723151645979,
      "loss": 3.4704,
      "step": 35000
    },
    {
      "epoch": 3.7735849056603774,
      "eval_accuracy": 0.37506574857873876,
      "eval_loss": 3.4583606719970703,
      "eval_runtime": 183.5302,
      "eval_samples_per_second": 98.136,
      "eval_steps_per_second": 6.135,
      "step": 35000
    },
    {
      "epoch": 3.778975741239892,
      "grad_norm": 0.5972731113433838,
      "learning_rate": 0.0003737485159201295,
      "loss": 3.4544,
      "step": 35050
    },
    {
      "epoch": 3.784366576819407,
      "grad_norm": 0.6210933923721313,
      "learning_rate": 0.00037342471667566106,
      "loss": 3.4579,
      "step": 35100
    },
    {
      "epoch": 3.789757412398922,
      "grad_norm": 0.6391624212265015,
      "learning_rate": 0.00037310091743119266,
      "loss": 3.4427,
      "step": 35150
    },
    {
      "epoch": 3.7951482479784366,
      "grad_norm": 0.5807982683181763,
      "learning_rate": 0.0003727771181867242,
      "loss": 3.4476,
      "step": 35200
    },
    {
      "epoch": 3.8005390835579513,
      "grad_norm": 0.6649582982063293,
      "learning_rate": 0.00037245331894225576,
      "loss": 3.4544,
      "step": 35250
    },
    {
      "epoch": 3.8059299191374665,
      "grad_norm": 0.6210810542106628,
      "learning_rate": 0.00037212951969778737,
      "loss": 3.4477,
      "step": 35300
    },
    {
      "epoch": 3.811320754716981,
      "grad_norm": 0.6078605651855469,
      "learning_rate": 0.00037180572045331887,
      "loss": 3.4687,
      "step": 35350
    },
    {
      "epoch": 3.816711590296496,
      "grad_norm": 0.5738853216171265,
      "learning_rate": 0.00037148192120885047,
      "loss": 3.4499,
      "step": 35400
    },
    {
      "epoch": 3.822102425876011,
      "grad_norm": 0.6362777352333069,
      "learning_rate": 0.000371158121964382,
      "loss": 3.4559,
      "step": 35450
    },
    {
      "epoch": 3.8274932614555257,
      "grad_norm": 0.6148720979690552,
      "learning_rate": 0.0003708343227199136,
      "loss": 3.4472,
      "step": 35500
    },
    {
      "epoch": 3.8328840970350404,
      "grad_norm": 0.5940957069396973,
      "learning_rate": 0.0003705105234754452,
      "loss": 3.4512,
      "step": 35550
    },
    {
      "epoch": 3.838274932614555,
      "grad_norm": 0.6278628706932068,
      "learning_rate": 0.0003701867242309768,
      "loss": 3.4608,
      "step": 35600
    },
    {
      "epoch": 3.8436657681940702,
      "grad_norm": 0.6533029079437256,
      "learning_rate": 0.00036986292498650833,
      "loss": 3.4467,
      "step": 35650
    },
    {
      "epoch": 3.849056603773585,
      "grad_norm": 0.6200972199440002,
      "learning_rate": 0.0003695391257420399,
      "loss": 3.4462,
      "step": 35700
    },
    {
      "epoch": 3.8544474393530996,
      "grad_norm": 0.6362287998199463,
      "learning_rate": 0.0003692153264975715,
      "loss": 3.4568,
      "step": 35750
    },
    {
      "epoch": 3.8598382749326143,
      "grad_norm": 0.6161172986030579,
      "learning_rate": 0.00036889152725310304,
      "loss": 3.4427,
      "step": 35800
    },
    {
      "epoch": 3.8652291105121295,
      "grad_norm": 0.6286973357200623,
      "learning_rate": 0.00036856772800863464,
      "loss": 3.4532,
      "step": 35850
    },
    {
      "epoch": 3.870619946091644,
      "grad_norm": 0.6777767539024353,
      "learning_rate": 0.0003682439287641662,
      "loss": 3.4637,
      "step": 35900
    },
    {
      "epoch": 3.876010781671159,
      "grad_norm": 0.6047759652137756,
      "learning_rate": 0.0003679201295196978,
      "loss": 3.4663,
      "step": 35950
    },
    {
      "epoch": 3.881401617250674,
      "grad_norm": 0.5936576724052429,
      "learning_rate": 0.00036759633027522935,
      "loss": 3.4368,
      "step": 36000
    },
    {
      "epoch": 3.881401617250674,
      "eval_accuracy": 0.37563161281847657,
      "eval_loss": 3.452744960784912,
      "eval_runtime": 183.8136,
      "eval_samples_per_second": 97.985,
      "eval_steps_per_second": 6.126,
      "step": 36000
    },
    {
      "epoch": 3.8867924528301887,
      "grad_norm": 0.6451914310455322,
      "learning_rate": 0.00036727253103076084,
      "loss": 3.4734,
      "step": 36050
    },
    {
      "epoch": 3.8921832884097034,
      "grad_norm": 0.6473202109336853,
      "learning_rate": 0.00036694873178629245,
      "loss": 3.4377,
      "step": 36100
    },
    {
      "epoch": 3.8975741239892185,
      "grad_norm": 0.6400440335273743,
      "learning_rate": 0.000366624932541824,
      "loss": 3.4438,
      "step": 36150
    },
    {
      "epoch": 3.9029649595687332,
      "grad_norm": 0.6485514044761658,
      "learning_rate": 0.0003663011332973556,
      "loss": 3.4499,
      "step": 36200
    },
    {
      "epoch": 3.908355795148248,
      "grad_norm": 0.5860769748687744,
      "learning_rate": 0.00036597733405288715,
      "loss": 3.4564,
      "step": 36250
    },
    {
      "epoch": 3.913746630727763,
      "grad_norm": 0.5832682847976685,
      "learning_rate": 0.00036565353480841876,
      "loss": 3.4587,
      "step": 36300
    },
    {
      "epoch": 3.9191374663072778,
      "grad_norm": 0.5989060401916504,
      "learning_rate": 0.0003653297355639503,
      "loss": 3.4729,
      "step": 36350
    },
    {
      "epoch": 3.9245283018867925,
      "grad_norm": 0.6010338664054871,
      "learning_rate": 0.00036501241230437126,
      "loss": 3.4403,
      "step": 36400
    },
    {
      "epoch": 3.929919137466307,
      "grad_norm": 0.6393582224845886,
      "learning_rate": 0.0003646886130599028,
      "loss": 3.4452,
      "step": 36450
    },
    {
      "epoch": 3.935309973045822,
      "grad_norm": 0.6078340411186218,
      "learning_rate": 0.0003643648138154344,
      "loss": 3.4593,
      "step": 36500
    },
    {
      "epoch": 3.940700808625337,
      "grad_norm": 0.5605495572090149,
      "learning_rate": 0.00036404101457096597,
      "loss": 3.4739,
      "step": 36550
    },
    {
      "epoch": 3.9460916442048517,
      "grad_norm": 0.6415327787399292,
      "learning_rate": 0.00036371721532649757,
      "loss": 3.4257,
      "step": 36600
    },
    {
      "epoch": 3.9514824797843664,
      "grad_norm": 0.6295135617256165,
      "learning_rate": 0.0003633934160820291,
      "loss": 3.4429,
      "step": 36650
    },
    {
      "epoch": 3.9568733153638815,
      "grad_norm": 0.6907050013542175,
      "learning_rate": 0.00036306961683756073,
      "loss": 3.4498,
      "step": 36700
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.6789411902427673,
      "learning_rate": 0.0003627458175930922,
      "loss": 3.4295,
      "step": 36750
    },
    {
      "epoch": 3.967654986522911,
      "grad_norm": 0.5623884201049805,
      "learning_rate": 0.0003624220183486238,
      "loss": 3.4439,
      "step": 36800
    },
    {
      "epoch": 3.973045822102426,
      "grad_norm": 0.6243301630020142,
      "learning_rate": 0.0003620982191041554,
      "loss": 3.4516,
      "step": 36850
    },
    {
      "epoch": 3.9784366576819408,
      "grad_norm": 0.6289325952529907,
      "learning_rate": 0.00036177441985968693,
      "loss": 3.4506,
      "step": 36900
    },
    {
      "epoch": 3.9838274932614555,
      "grad_norm": 0.6316044926643372,
      "learning_rate": 0.00036145062061521854,
      "loss": 3.4304,
      "step": 36950
    },
    {
      "epoch": 3.9892183288409706,
      "grad_norm": 0.6350688934326172,
      "learning_rate": 0.0003611268213707501,
      "loss": 3.4169,
      "step": 37000
    },
    {
      "epoch": 3.9892183288409706,
      "eval_accuracy": 0.3760328679331754,
      "eval_loss": 3.445186138153076,
      "eval_runtime": 183.4662,
      "eval_samples_per_second": 98.171,
      "eval_steps_per_second": 6.137,
      "step": 37000
    },
    {
      "epoch": 3.9946091644204853,
      "grad_norm": 0.633863627910614,
      "learning_rate": 0.0003608030221262817,
      "loss": 3.47,
      "step": 37050
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3224987983703613,
      "learning_rate": 0.00036047922288181324,
      "loss": 3.4343,
      "step": 37100
    },
    {
      "epoch": 4.005390835579515,
      "grad_norm": 0.6018860340118408,
      "learning_rate": 0.00036015542363734485,
      "loss": 3.3632,
      "step": 37150
    },
    {
      "epoch": 4.010781671159029,
      "grad_norm": 0.6127363443374634,
      "learning_rate": 0.0003598316243928764,
      "loss": 3.3584,
      "step": 37200
    },
    {
      "epoch": 4.0161725067385445,
      "grad_norm": 0.6423741579055786,
      "learning_rate": 0.00035950782514840795,
      "loss": 3.3555,
      "step": 37250
    },
    {
      "epoch": 4.02156334231806,
      "grad_norm": 0.5862413644790649,
      "learning_rate": 0.00035918402590393955,
      "loss": 3.3588,
      "step": 37300
    },
    {
      "epoch": 4.026954177897574,
      "grad_norm": 0.6085898876190186,
      "learning_rate": 0.00035886022665947105,
      "loss": 3.3414,
      "step": 37350
    },
    {
      "epoch": 4.032345013477089,
      "grad_norm": 0.5962231755256653,
      "learning_rate": 0.0003585364274150027,
      "loss": 3.3457,
      "step": 37400
    },
    {
      "epoch": 4.037735849056604,
      "grad_norm": 0.6367636919021606,
      "learning_rate": 0.0003582126281705342,
      "loss": 3.3553,
      "step": 37450
    },
    {
      "epoch": 4.0431266846361185,
      "grad_norm": 0.6896623373031616,
      "learning_rate": 0.0003578888289260658,
      "loss": 3.3537,
      "step": 37500
    },
    {
      "epoch": 4.048517520215634,
      "grad_norm": 0.6605024933815002,
      "learning_rate": 0.00035756502968159736,
      "loss": 3.3657,
      "step": 37550
    },
    {
      "epoch": 4.053908355795148,
      "grad_norm": 0.6501374244689941,
      "learning_rate": 0.00035724123043712896,
      "loss": 3.3591,
      "step": 37600
    },
    {
      "epoch": 4.059299191374663,
      "grad_norm": 0.6687803864479065,
      "learning_rate": 0.0003569174311926605,
      "loss": 3.3656,
      "step": 37650
    },
    {
      "epoch": 4.064690026954178,
      "grad_norm": 0.6082162261009216,
      "learning_rate": 0.00035659363194819206,
      "loss": 3.3675,
      "step": 37700
    },
    {
      "epoch": 4.070080862533692,
      "grad_norm": 0.5966090559959412,
      "learning_rate": 0.00035626983270372367,
      "loss": 3.3737,
      "step": 37750
    },
    {
      "epoch": 4.0754716981132075,
      "grad_norm": 0.6330704689025879,
      "learning_rate": 0.0003559460334592552,
      "loss": 3.3645,
      "step": 37800
    },
    {
      "epoch": 4.080862533692723,
      "grad_norm": 0.6244950294494629,
      "learning_rate": 0.0003556222342147868,
      "loss": 3.3694,
      "step": 37850
    },
    {
      "epoch": 4.086253369272237,
      "grad_norm": 0.6000611186027527,
      "learning_rate": 0.0003552984349703184,
      "loss": 3.3807,
      "step": 37900
    },
    {
      "epoch": 4.091644204851752,
      "grad_norm": 0.6247694492340088,
      "learning_rate": 0.00035497463572585,
      "loss": 3.3803,
      "step": 37950
    },
    {
      "epoch": 4.097035040431267,
      "grad_norm": 0.6111884713172913,
      "learning_rate": 0.00035465083648138153,
      "loss": 3.36,
      "step": 38000
    },
    {
      "epoch": 4.097035040431267,
      "eval_accuracy": 0.37655614024088996,
      "eval_loss": 3.4490854740142822,
      "eval_runtime": 183.4823,
      "eval_samples_per_second": 98.162,
      "eval_steps_per_second": 6.137,
      "step": 38000
    },
    {
      "epoch": 4.1024258760107815,
      "grad_norm": 0.6444174647331238,
      "learning_rate": 0.00035432703723691314,
      "loss": 3.3765,
      "step": 38050
    },
    {
      "epoch": 4.107816711590297,
      "grad_norm": 0.6280059814453125,
      "learning_rate": 0.00035400323799244463,
      "loss": 3.3616,
      "step": 38100
    },
    {
      "epoch": 4.113207547169812,
      "grad_norm": 0.6255881786346436,
      "learning_rate": 0.0003536794387479762,
      "loss": 3.377,
      "step": 38150
    },
    {
      "epoch": 4.118598382749326,
      "grad_norm": 0.65887850522995,
      "learning_rate": 0.0003533556395035078,
      "loss": 3.3807,
      "step": 38200
    },
    {
      "epoch": 4.123989218328841,
      "grad_norm": 0.6200643181800842,
      "learning_rate": 0.00035303184025903934,
      "loss": 3.3683,
      "step": 38250
    },
    {
      "epoch": 4.129380053908355,
      "grad_norm": 0.6016085147857666,
      "learning_rate": 0.00035270804101457094,
      "loss": 3.3729,
      "step": 38300
    },
    {
      "epoch": 4.1347708894878705,
      "grad_norm": 0.6084494590759277,
      "learning_rate": 0.0003523842417701025,
      "loss": 3.3562,
      "step": 38350
    },
    {
      "epoch": 4.140161725067386,
      "grad_norm": 0.6303635835647583,
      "learning_rate": 0.0003520604425256341,
      "loss": 3.378,
      "step": 38400
    },
    {
      "epoch": 4.1455525606469,
      "grad_norm": 0.6024056077003479,
      "learning_rate": 0.000351743119266055,
      "loss": 3.3748,
      "step": 38450
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.6303093433380127,
      "learning_rate": 0.0003514193200215866,
      "loss": 3.3691,
      "step": 38500
    },
    {
      "epoch": 4.15633423180593,
      "grad_norm": 0.6160764694213867,
      "learning_rate": 0.00035109552077711815,
      "loss": 3.3674,
      "step": 38550
    },
    {
      "epoch": 4.1617250673854445,
      "grad_norm": 0.6600135564804077,
      "learning_rate": 0.00035077172153264976,
      "loss": 3.3676,
      "step": 38600
    },
    {
      "epoch": 4.16711590296496,
      "grad_norm": 0.6453423500061035,
      "learning_rate": 0.0003504479222881813,
      "loss": 3.3716,
      "step": 38650
    },
    {
      "epoch": 4.172506738544475,
      "grad_norm": 0.6940420866012573,
      "learning_rate": 0.0003501241230437129,
      "loss": 3.3884,
      "step": 38700
    },
    {
      "epoch": 4.177897574123989,
      "grad_norm": 0.615378737449646,
      "learning_rate": 0.0003498003237992444,
      "loss": 3.3635,
      "step": 38750
    },
    {
      "epoch": 4.183288409703504,
      "grad_norm": 0.6390400528907776,
      "learning_rate": 0.000349476524554776,
      "loss": 3.3801,
      "step": 38800
    },
    {
      "epoch": 4.188679245283019,
      "grad_norm": 0.6343681812286377,
      "learning_rate": 0.00034915272531030756,
      "loss": 3.3872,
      "step": 38850
    },
    {
      "epoch": 4.1940700808625335,
      "grad_norm": 0.5857776403427124,
      "learning_rate": 0.0003488289260658391,
      "loss": 3.4035,
      "step": 38900
    },
    {
      "epoch": 4.199460916442049,
      "grad_norm": 0.6787095665931702,
      "learning_rate": 0.0003485051268213707,
      "loss": 3.3685,
      "step": 38950
    },
    {
      "epoch": 4.204851752021563,
      "grad_norm": 0.6941482424736023,
      "learning_rate": 0.00034818132757690227,
      "loss": 3.3837,
      "step": 39000
    },
    {
      "epoch": 4.204851752021563,
      "eval_accuracy": 0.3769635885701943,
      "eval_loss": 3.4448821544647217,
      "eval_runtime": 183.6433,
      "eval_samples_per_second": 98.076,
      "eval_steps_per_second": 6.131,
      "step": 39000
    },
    {
      "epoch": 4.210242587601078,
      "grad_norm": 0.6923218965530396,
      "learning_rate": 0.0003478575283324339,
      "loss": 3.3915,
      "step": 39050
    },
    {
      "epoch": 4.215633423180593,
      "grad_norm": 0.6843828558921814,
      "learning_rate": 0.0003475337290879654,
      "loss": 3.392,
      "step": 39100
    },
    {
      "epoch": 4.2210242587601075,
      "grad_norm": 0.6378276348114014,
      "learning_rate": 0.00034720992984349703,
      "loss": 3.381,
      "step": 39150
    },
    {
      "epoch": 4.226415094339623,
      "grad_norm": 0.6532413363456726,
      "learning_rate": 0.0003468861305990286,
      "loss": 3.3718,
      "step": 39200
    },
    {
      "epoch": 4.231805929919138,
      "grad_norm": 0.6036219000816345,
      "learning_rate": 0.0003465623313545602,
      "loss": 3.3749,
      "step": 39250
    },
    {
      "epoch": 4.237196765498652,
      "grad_norm": 0.6405287384986877,
      "learning_rate": 0.00034623853211009173,
      "loss": 3.3757,
      "step": 39300
    },
    {
      "epoch": 4.242587601078167,
      "grad_norm": 0.5965821146965027,
      "learning_rate": 0.0003459212088505127,
      "loss": 3.3881,
      "step": 39350
    },
    {
      "epoch": 4.247978436657682,
      "grad_norm": 0.6920449137687683,
      "learning_rate": 0.0003455974096060442,
      "loss": 3.3822,
      "step": 39400
    },
    {
      "epoch": 4.2533692722371965,
      "grad_norm": 0.6426447629928589,
      "learning_rate": 0.0003452736103615758,
      "loss": 3.392,
      "step": 39450
    },
    {
      "epoch": 4.258760107816712,
      "grad_norm": 0.6216702461242676,
      "learning_rate": 0.00034494981111710734,
      "loss": 3.3791,
      "step": 39500
    },
    {
      "epoch": 4.264150943396227,
      "grad_norm": 0.6139161586761475,
      "learning_rate": 0.00034462601187263894,
      "loss": 3.3677,
      "step": 39550
    },
    {
      "epoch": 4.269541778975741,
      "grad_norm": 0.6620310544967651,
      "learning_rate": 0.0003443022126281705,
      "loss": 3.3826,
      "step": 39600
    },
    {
      "epoch": 4.274932614555256,
      "grad_norm": 0.6156401038169861,
      "learning_rate": 0.00034397841338370204,
      "loss": 3.3912,
      "step": 39650
    },
    {
      "epoch": 4.280323450134771,
      "grad_norm": 0.6847094297409058,
      "learning_rate": 0.00034365461413923365,
      "loss": 3.3743,
      "step": 39700
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.6415292620658875,
      "learning_rate": 0.0003433308148947652,
      "loss": 3.3973,
      "step": 39750
    },
    {
      "epoch": 4.291105121293801,
      "grad_norm": 0.6539881229400635,
      "learning_rate": 0.0003430070156502968,
      "loss": 3.3823,
      "step": 39800
    },
    {
      "epoch": 4.296495956873315,
      "grad_norm": 0.6364943981170654,
      "learning_rate": 0.00034268321640582835,
      "loss": 3.3856,
      "step": 39850
    },
    {
      "epoch": 4.30188679245283,
      "grad_norm": 0.6464160084724426,
      "learning_rate": 0.00034235941716135996,
      "loss": 3.3868,
      "step": 39900
    },
    {
      "epoch": 4.307277628032345,
      "grad_norm": 0.7435222864151001,
      "learning_rate": 0.0003420356179168915,
      "loss": 3.3939,
      "step": 39950
    },
    {
      "epoch": 4.3126684636118595,
      "grad_norm": 0.6538057327270508,
      "learning_rate": 0.0003417118186724231,
      "loss": 3.4004,
      "step": 40000
    },
    {
      "epoch": 4.3126684636118595,
      "eval_accuracy": 0.37684906842643784,
      "eval_loss": 3.4412429332733154,
      "eval_runtime": 183.3977,
      "eval_samples_per_second": 98.207,
      "eval_steps_per_second": 6.14,
      "step": 40000
    },
    {
      "epoch": 4.318059299191375,
      "grad_norm": 0.6281383037567139,
      "learning_rate": 0.00034138801942795467,
      "loss": 3.389,
      "step": 40050
    },
    {
      "epoch": 4.32345013477089,
      "grad_norm": 0.6336797475814819,
      "learning_rate": 0.00034106422018348616,
      "loss": 3.3922,
      "step": 40100
    },
    {
      "epoch": 4.328840970350404,
      "grad_norm": 0.638562798500061,
      "learning_rate": 0.00034074042093901777,
      "loss": 3.3743,
      "step": 40150
    },
    {
      "epoch": 4.334231805929919,
      "grad_norm": 0.6539092063903809,
      "learning_rate": 0.0003404166216945493,
      "loss": 3.4048,
      "step": 40200
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.680752694606781,
      "learning_rate": 0.0003400928224500809,
      "loss": 3.4125,
      "step": 40250
    },
    {
      "epoch": 4.345013477088949,
      "grad_norm": 0.6655305624008179,
      "learning_rate": 0.00033976902320561247,
      "loss": 3.3806,
      "step": 40300
    },
    {
      "epoch": 4.350404312668464,
      "grad_norm": 0.6368870735168457,
      "learning_rate": 0.0003394452239611441,
      "loss": 3.3936,
      "step": 40350
    },
    {
      "epoch": 4.355795148247978,
      "grad_norm": 0.6510953307151794,
      "learning_rate": 0.00033912142471667563,
      "loss": 3.3909,
      "step": 40400
    },
    {
      "epoch": 4.361185983827493,
      "grad_norm": 0.6622096300125122,
      "learning_rate": 0.00033879762547220723,
      "loss": 3.3909,
      "step": 40450
    },
    {
      "epoch": 4.366576819407008,
      "grad_norm": 0.6338421702384949,
      "learning_rate": 0.0003384738262277388,
      "loss": 3.3849,
      "step": 40500
    },
    {
      "epoch": 4.3719676549865225,
      "grad_norm": 0.6532634496688843,
      "learning_rate": 0.00033815002698327033,
      "loss": 3.3902,
      "step": 40550
    },
    {
      "epoch": 4.377358490566038,
      "grad_norm": 0.6484360098838806,
      "learning_rate": 0.00033782622773880194,
      "loss": 3.3972,
      "step": 40600
    },
    {
      "epoch": 4.382749326145553,
      "grad_norm": 0.6758602857589722,
      "learning_rate": 0.0003375024284943335,
      "loss": 3.3879,
      "step": 40650
    },
    {
      "epoch": 4.388140161725067,
      "grad_norm": 0.6421069502830505,
      "learning_rate": 0.0003371786292498651,
      "loss": 3.4004,
      "step": 40700
    },
    {
      "epoch": 4.393530997304582,
      "grad_norm": 0.6300466060638428,
      "learning_rate": 0.0003368548300053966,
      "loss": 3.4042,
      "step": 40750
    },
    {
      "epoch": 4.398921832884097,
      "grad_norm": 0.6359042525291443,
      "learning_rate": 0.0003365310307609282,
      "loss": 3.3919,
      "step": 40800
    },
    {
      "epoch": 4.404312668463612,
      "grad_norm": 0.6414315104484558,
      "learning_rate": 0.00033620723151645975,
      "loss": 3.3817,
      "step": 40850
    },
    {
      "epoch": 4.409703504043127,
      "grad_norm": 0.6350900530815125,
      "learning_rate": 0.0003358834322719913,
      "loss": 3.385,
      "step": 40900
    },
    {
      "epoch": 4.415094339622642,
      "grad_norm": 1.0420467853546143,
      "learning_rate": 0.0003355596330275229,
      "loss": 3.3806,
      "step": 40950
    },
    {
      "epoch": 4.420485175202156,
      "grad_norm": 0.6752620935440063,
      "learning_rate": 0.00033523583378305445,
      "loss": 3.4147,
      "step": 41000
    },
    {
      "epoch": 4.420485175202156,
      "eval_accuracy": 0.3782003843921865,
      "eval_loss": 3.4338560104370117,
      "eval_runtime": 183.5528,
      "eval_samples_per_second": 98.124,
      "eval_steps_per_second": 6.134,
      "step": 41000
    },
    {
      "epoch": 4.425876010781671,
      "grad_norm": 0.6579931378364563,
      "learning_rate": 0.00033491203453858606,
      "loss": 3.3897,
      "step": 41050
    },
    {
      "epoch": 4.431266846361186,
      "grad_norm": 0.6752576231956482,
      "learning_rate": 0.0003345882352941176,
      "loss": 3.3981,
      "step": 41100
    },
    {
      "epoch": 4.436657681940701,
      "grad_norm": 0.6758126020431519,
      "learning_rate": 0.0003342644360496492,
      "loss": 3.4058,
      "step": 41150
    },
    {
      "epoch": 4.442048517520216,
      "grad_norm": 0.635006308555603,
      "learning_rate": 0.00033394063680518076,
      "loss": 3.3896,
      "step": 41200
    },
    {
      "epoch": 4.44743935309973,
      "grad_norm": 0.6431044340133667,
      "learning_rate": 0.00033361683756071237,
      "loss": 3.4063,
      "step": 41250
    },
    {
      "epoch": 4.452830188679245,
      "grad_norm": 0.618403971195221,
      "learning_rate": 0.0003332930383162439,
      "loss": 3.3787,
      "step": 41300
    },
    {
      "epoch": 4.45822102425876,
      "grad_norm": 0.7186667323112488,
      "learning_rate": 0.0003329692390717754,
      "loss": 3.3842,
      "step": 41350
    },
    {
      "epoch": 4.463611859838275,
      "grad_norm": 0.6048040390014648,
      "learning_rate": 0.0003326454398273071,
      "loss": 3.4078,
      "step": 41400
    },
    {
      "epoch": 4.46900269541779,
      "grad_norm": 0.6665133833885193,
      "learning_rate": 0.00033232164058283857,
      "loss": 3.3829,
      "step": 41450
    },
    {
      "epoch": 4.474393530997305,
      "grad_norm": 0.8045024275779724,
      "learning_rate": 0.0003319978413383702,
      "loss": 3.3956,
      "step": 41500
    },
    {
      "epoch": 4.479784366576819,
      "grad_norm": 0.667209267616272,
      "learning_rate": 0.0003316740420939017,
      "loss": 3.3875,
      "step": 41550
    },
    {
      "epoch": 4.485175202156334,
      "grad_norm": 0.6306861639022827,
      "learning_rate": 0.00033135024284943333,
      "loss": 3.4,
      "step": 41600
    },
    {
      "epoch": 4.490566037735849,
      "grad_norm": 0.6058953404426575,
      "learning_rate": 0.0003310264436049649,
      "loss": 3.3834,
      "step": 41650
    },
    {
      "epoch": 4.495956873315364,
      "grad_norm": 0.5988507270812988,
      "learning_rate": 0.00033070912034538583,
      "loss": 3.4056,
      "step": 41700
    },
    {
      "epoch": 4.501347708894879,
      "grad_norm": 0.6373119354248047,
      "learning_rate": 0.0003303853211009174,
      "loss": 3.39,
      "step": 41750
    },
    {
      "epoch": 4.506738544474393,
      "grad_norm": 0.6669942140579224,
      "learning_rate": 0.000330061521856449,
      "loss": 3.3935,
      "step": 41800
    },
    {
      "epoch": 4.512129380053908,
      "grad_norm": 0.6317675113677979,
      "learning_rate": 0.00032973772261198054,
      "loss": 3.381,
      "step": 41850
    },
    {
      "epoch": 4.517520215633423,
      "grad_norm": 0.6255390644073486,
      "learning_rate": 0.00032941392336751214,
      "loss": 3.3942,
      "step": 41900
    },
    {
      "epoch": 4.5229110512129385,
      "grad_norm": 0.6369220018386841,
      "learning_rate": 0.0003290901241230437,
      "loss": 3.3759,
      "step": 41950
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.6291524171829224,
      "learning_rate": 0.0003287663248785753,
      "loss": 3.3877,
      "step": 42000
    },
    {
      "epoch": 4.528301886792453,
      "eval_accuracy": 0.37846582339711726,
      "eval_loss": 3.429187059402466,
      "eval_runtime": 183.4827,
      "eval_samples_per_second": 98.162,
      "eval_steps_per_second": 6.137,
      "step": 42000
    },
    {
      "epoch": 4.533692722371968,
      "grad_norm": 0.6402625441551208,
      "learning_rate": 0.00032844252563410685,
      "loss": 3.3852,
      "step": 42050
    },
    {
      "epoch": 4.539083557951482,
      "grad_norm": 0.667153537273407,
      "learning_rate": 0.00032811872638963834,
      "loss": 3.3935,
      "step": 42100
    },
    {
      "epoch": 4.544474393530997,
      "grad_norm": 0.6692747473716736,
      "learning_rate": 0.00032779492714516995,
      "loss": 3.4201,
      "step": 42150
    },
    {
      "epoch": 4.549865229110512,
      "grad_norm": 0.6011302471160889,
      "learning_rate": 0.0003274711279007015,
      "loss": 3.3878,
      "step": 42200
    },
    {
      "epoch": 4.555256064690027,
      "grad_norm": 0.6687884330749512,
      "learning_rate": 0.0003271473286562331,
      "loss": 3.3848,
      "step": 42250
    },
    {
      "epoch": 4.560646900269542,
      "grad_norm": 0.6640747785568237,
      "learning_rate": 0.00032682352941176466,
      "loss": 3.3882,
      "step": 42300
    },
    {
      "epoch": 4.566037735849057,
      "grad_norm": 0.6145426034927368,
      "learning_rate": 0.00032649973016729626,
      "loss": 3.3902,
      "step": 42350
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.6300755143165588,
      "learning_rate": 0.0003261759309228278,
      "loss": 3.3816,
      "step": 42400
    },
    {
      "epoch": 4.576819407008086,
      "grad_norm": 0.6550257205963135,
      "learning_rate": 0.0003258521316783594,
      "loss": 3.3922,
      "step": 42450
    },
    {
      "epoch": 4.5822102425876015,
      "grad_norm": 0.6228023767471313,
      "learning_rate": 0.00032552833243389097,
      "loss": 3.3895,
      "step": 42500
    },
    {
      "epoch": 4.587601078167116,
      "grad_norm": 0.6689121723175049,
      "learning_rate": 0.0003252045331894225,
      "loss": 3.3885,
      "step": 42550
    },
    {
      "epoch": 4.592991913746631,
      "grad_norm": 0.6270008683204651,
      "learning_rate": 0.0003248807339449541,
      "loss": 3.3753,
      "step": 42600
    },
    {
      "epoch": 4.598382749326145,
      "grad_norm": 0.6733295917510986,
      "learning_rate": 0.00032455693470048567,
      "loss": 3.3757,
      "step": 42650
    },
    {
      "epoch": 4.60377358490566,
      "grad_norm": 0.6679153442382812,
      "learning_rate": 0.0003242331354560173,
      "loss": 3.3833,
      "step": 42700
    },
    {
      "epoch": 4.609164420485175,
      "grad_norm": 0.6631647348403931,
      "learning_rate": 0.0003239093362115488,
      "loss": 3.4027,
      "step": 42750
    },
    {
      "epoch": 4.6145552560646905,
      "grad_norm": 0.6664791107177734,
      "learning_rate": 0.00032358553696708043,
      "loss": 3.3741,
      "step": 42800
    },
    {
      "epoch": 4.619946091644205,
      "grad_norm": 0.6315823793411255,
      "learning_rate": 0.00032326173772261193,
      "loss": 3.3814,
      "step": 42850
    },
    {
      "epoch": 4.62533692722372,
      "grad_norm": 0.6814761161804199,
      "learning_rate": 0.00032293793847814353,
      "loss": 3.3802,
      "step": 42900
    },
    {
      "epoch": 4.630727762803234,
      "grad_norm": 0.6890589594841003,
      "learning_rate": 0.0003226141392336751,
      "loss": 3.4014,
      "step": 42950
    },
    {
      "epoch": 4.636118598382749,
      "grad_norm": 0.6870799660682678,
      "learning_rate": 0.00032229033998920663,
      "loss": 3.4082,
      "step": 43000
    },
    {
      "epoch": 4.636118598382749,
      "eval_accuracy": 0.3788883744778278,
      "eval_loss": 3.4247143268585205,
      "eval_runtime": 183.8361,
      "eval_samples_per_second": 97.973,
      "eval_steps_per_second": 6.125,
      "step": 43000
    },
    {
      "epoch": 4.6415094339622645,
      "grad_norm": 0.642059862613678,
      "learning_rate": 0.00032196654074473824,
      "loss": 3.3919,
      "step": 43050
    },
    {
      "epoch": 4.646900269541779,
      "grad_norm": 0.6234888434410095,
      "learning_rate": 0.0003216427415002698,
      "loss": 3.3981,
      "step": 43100
    },
    {
      "epoch": 4.652291105121294,
      "grad_norm": 0.6437119245529175,
      "learning_rate": 0.0003213189422558014,
      "loss": 3.4103,
      "step": 43150
    },
    {
      "epoch": 4.657681940700809,
      "grad_norm": 0.6424395442008972,
      "learning_rate": 0.00032099514301133295,
      "loss": 3.382,
      "step": 43200
    },
    {
      "epoch": 4.663072776280323,
      "grad_norm": 0.6444264650344849,
      "learning_rate": 0.00032067134376686455,
      "loss": 3.3805,
      "step": 43250
    },
    {
      "epoch": 4.668463611859838,
      "grad_norm": 0.6374757289886475,
      "learning_rate": 0.0003203475445223961,
      "loss": 3.3922,
      "step": 43300
    },
    {
      "epoch": 4.6738544474393535,
      "grad_norm": 0.709404468536377,
      "learning_rate": 0.0003200237452779277,
      "loss": 3.3917,
      "step": 43350
    },
    {
      "epoch": 4.679245283018868,
      "grad_norm": 0.6628051996231079,
      "learning_rate": 0.00031969994603345926,
      "loss": 3.402,
      "step": 43400
    },
    {
      "epoch": 4.684636118598383,
      "grad_norm": 0.624970555305481,
      "learning_rate": 0.00031937614678899075,
      "loss": 3.3909,
      "step": 43450
    },
    {
      "epoch": 4.690026954177897,
      "grad_norm": 0.6710526347160339,
      "learning_rate": 0.00031905234754452236,
      "loss": 3.4032,
      "step": 43500
    },
    {
      "epoch": 4.695417789757412,
      "grad_norm": 0.667590320110321,
      "learning_rate": 0.0003187285483000539,
      "loss": 3.3876,
      "step": 43550
    },
    {
      "epoch": 4.7008086253369274,
      "grad_norm": 0.65190190076828,
      "learning_rate": 0.0003184047490555855,
      "loss": 3.378,
      "step": 43600
    },
    {
      "epoch": 4.706199460916442,
      "grad_norm": 0.6367104649543762,
      "learning_rate": 0.00031808094981111706,
      "loss": 3.393,
      "step": 43650
    },
    {
      "epoch": 4.711590296495957,
      "grad_norm": 0.6436135172843933,
      "learning_rate": 0.00031775715056664867,
      "loss": 3.3811,
      "step": 43700
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.626654326915741,
      "learning_rate": 0.0003174333513221802,
      "loss": 3.3992,
      "step": 43750
    },
    {
      "epoch": 4.722371967654986,
      "grad_norm": 0.6450473666191101,
      "learning_rate": 0.00031710955207771177,
      "loss": 3.3913,
      "step": 43800
    },
    {
      "epoch": 4.727762803234501,
      "grad_norm": 0.6737494468688965,
      "learning_rate": 0.0003167857528332434,
      "loss": 3.3945,
      "step": 43850
    },
    {
      "epoch": 4.7331536388140165,
      "grad_norm": 0.5905812382698059,
      "learning_rate": 0.0003164619535887749,
      "loss": 3.3977,
      "step": 43900
    },
    {
      "epoch": 4.738544474393531,
      "grad_norm": 0.649468183517456,
      "learning_rate": 0.00031613815434430653,
      "loss": 3.3887,
      "step": 43950
    },
    {
      "epoch": 4.743935309973046,
      "grad_norm": 0.63776034116745,
      "learning_rate": 0.0003158143550998381,
      "loss": 3.3935,
      "step": 44000
    },
    {
      "epoch": 4.743935309973046,
      "eval_accuracy": 0.3795477888539739,
      "eval_loss": 3.4194493293762207,
      "eval_runtime": 183.3803,
      "eval_samples_per_second": 98.217,
      "eval_steps_per_second": 6.14,
      "step": 44000
    },
    {
      "epoch": 4.74932614555256,
      "grad_norm": 0.6686685085296631,
      "learning_rate": 0.0003154905558553697,
      "loss": 3.4027,
      "step": 44050
    },
    {
      "epoch": 4.754716981132075,
      "grad_norm": 0.6547311544418335,
      "learning_rate": 0.0003151667566109012,
      "loss": 3.3956,
      "step": 44100
    },
    {
      "epoch": 4.7601078167115904,
      "grad_norm": 0.6947376728057861,
      "learning_rate": 0.00031484295736643284,
      "loss": 3.4059,
      "step": 44150
    },
    {
      "epoch": 4.765498652291106,
      "grad_norm": 0.6252668499946594,
      "learning_rate": 0.00031451915812196434,
      "loss": 3.4025,
      "step": 44200
    },
    {
      "epoch": 4.77088948787062,
      "grad_norm": 0.6244868040084839,
      "learning_rate": 0.0003141953588774959,
      "loss": 3.3631,
      "step": 44250
    },
    {
      "epoch": 4.776280323450135,
      "grad_norm": 0.6843665838241577,
      "learning_rate": 0.0003138715596330275,
      "loss": 3.3909,
      "step": 44300
    },
    {
      "epoch": 4.781671159029649,
      "grad_norm": 0.6329135298728943,
      "learning_rate": 0.00031354776038855904,
      "loss": 3.3944,
      "step": 44350
    },
    {
      "epoch": 4.787061994609164,
      "grad_norm": 0.6198347806930542,
      "learning_rate": 0.00031322396114409065,
      "loss": 3.3779,
      "step": 44400
    },
    {
      "epoch": 4.7924528301886795,
      "grad_norm": 0.5985150337219238,
      "learning_rate": 0.0003129001618996222,
      "loss": 3.4032,
      "step": 44450
    },
    {
      "epoch": 4.797843665768194,
      "grad_norm": 0.6926071047782898,
      "learning_rate": 0.0003125763626551538,
      "loss": 3.3996,
      "step": 44500
    },
    {
      "epoch": 4.803234501347709,
      "grad_norm": 0.6708675026893616,
      "learning_rate": 0.00031225256341068535,
      "loss": 3.3724,
      "step": 44550
    },
    {
      "epoch": 4.808625336927224,
      "grad_norm": 0.6399439573287964,
      "learning_rate": 0.00031192876416621696,
      "loss": 3.3872,
      "step": 44600
    },
    {
      "epoch": 4.814016172506738,
      "grad_norm": 0.6167232394218445,
      "learning_rate": 0.0003116049649217485,
      "loss": 3.3877,
      "step": 44650
    },
    {
      "epoch": 4.819407008086253,
      "grad_norm": 0.6150270104408264,
      "learning_rate": 0.00031128116567728,
      "loss": 3.3872,
      "step": 44700
    },
    {
      "epoch": 4.824797843665769,
      "grad_norm": 0.623081386089325,
      "learning_rate": 0.00031095736643281166,
      "loss": 3.401,
      "step": 44750
    },
    {
      "epoch": 4.830188679245283,
      "grad_norm": 0.7186699509620667,
      "learning_rate": 0.00031063356718834316,
      "loss": 3.3718,
      "step": 44800
    },
    {
      "epoch": 4.835579514824798,
      "grad_norm": 0.702882707118988,
      "learning_rate": 0.00031030976794387476,
      "loss": 3.386,
      "step": 44850
    },
    {
      "epoch": 4.840970350404312,
      "grad_norm": 0.6542748808860779,
      "learning_rate": 0.0003099859686994063,
      "loss": 3.3638,
      "step": 44900
    },
    {
      "epoch": 4.846361185983827,
      "grad_norm": 0.6316834688186646,
      "learning_rate": 0.0003096621694549379,
      "loss": 3.4019,
      "step": 44950
    },
    {
      "epoch": 4.8517520215633425,
      "grad_norm": 0.6602932810783386,
      "learning_rate": 0.00030933837021046947,
      "loss": 3.3906,
      "step": 45000
    },
    {
      "epoch": 4.8517520215633425,
      "eval_accuracy": 0.38019666390000195,
      "eval_loss": 3.412461042404175,
      "eval_runtime": 183.6637,
      "eval_samples_per_second": 98.065,
      "eval_steps_per_second": 6.131,
      "step": 45000
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.6776353120803833,
      "learning_rate": 0.0003090145709660011,
      "loss": 3.3985,
      "step": 45050
    },
    {
      "epoch": 4.862533692722372,
      "grad_norm": 0.6363914608955383,
      "learning_rate": 0.0003086907717215326,
      "loss": 3.4018,
      "step": 45100
    },
    {
      "epoch": 4.867924528301887,
      "grad_norm": 0.6715949177742004,
      "learning_rate": 0.0003083669724770642,
      "loss": 3.3743,
      "step": 45150
    },
    {
      "epoch": 4.873315363881401,
      "grad_norm": 0.6310675740242004,
      "learning_rate": 0.0003080431732325958,
      "loss": 3.3816,
      "step": 45200
    },
    {
      "epoch": 4.878706199460916,
      "grad_norm": 0.7090080976486206,
      "learning_rate": 0.00030772584997301673,
      "loss": 3.3979,
      "step": 45250
    },
    {
      "epoch": 4.884097035040432,
      "grad_norm": 0.6891008019447327,
      "learning_rate": 0.0003074020507285483,
      "loss": 3.4082,
      "step": 45300
    },
    {
      "epoch": 4.889487870619946,
      "grad_norm": 0.6853170990943909,
      "learning_rate": 0.0003070782514840799,
      "loss": 3.3852,
      "step": 45350
    },
    {
      "epoch": 4.894878706199461,
      "grad_norm": 0.6550562381744385,
      "learning_rate": 0.00030675445223961144,
      "loss": 3.3967,
      "step": 45400
    },
    {
      "epoch": 4.900269541778976,
      "grad_norm": 0.6445778608322144,
      "learning_rate": 0.00030643065299514294,
      "loss": 3.3929,
      "step": 45450
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.6672085523605347,
      "learning_rate": 0.00030610685375067454,
      "loss": 3.3912,
      "step": 45500
    },
    {
      "epoch": 4.9110512129380055,
      "grad_norm": 0.6518294811248779,
      "learning_rate": 0.0003057830545062061,
      "loss": 3.3933,
      "step": 45550
    },
    {
      "epoch": 4.916442048517521,
      "grad_norm": 0.6551598310470581,
      "learning_rate": 0.0003054592552617377,
      "loss": 3.3893,
      "step": 45600
    },
    {
      "epoch": 4.921832884097035,
      "grad_norm": 0.6479905843734741,
      "learning_rate": 0.00030513545601726925,
      "loss": 3.3906,
      "step": 45650
    },
    {
      "epoch": 4.92722371967655,
      "grad_norm": 0.6924375891685486,
      "learning_rate": 0.00030481165677280085,
      "loss": 3.3681,
      "step": 45700
    },
    {
      "epoch": 4.932614555256064,
      "grad_norm": 0.6434122920036316,
      "learning_rate": 0.0003044878575283324,
      "loss": 3.3833,
      "step": 45750
    },
    {
      "epoch": 4.938005390835579,
      "grad_norm": 0.7251989245414734,
      "learning_rate": 0.000304164058283864,
      "loss": 3.4028,
      "step": 45800
    },
    {
      "epoch": 4.943396226415095,
      "grad_norm": 0.704974353313446,
      "learning_rate": 0.00030384025903939556,
      "loss": 3.3812,
      "step": 45850
    },
    {
      "epoch": 4.948787061994609,
      "grad_norm": 0.6566643118858337,
      "learning_rate": 0.0003035164597949271,
      "loss": 3.3898,
      "step": 45900
    },
    {
      "epoch": 4.954177897574124,
      "grad_norm": 0.619677722454071,
      "learning_rate": 0.0003031926605504587,
      "loss": 3.3809,
      "step": 45950
    },
    {
      "epoch": 4.959568733153639,
      "grad_norm": 0.6274846196174622,
      "learning_rate": 0.00030286886130599026,
      "loss": 3.3874,
      "step": 46000
    },
    {
      "epoch": 4.959568733153639,
      "eval_accuracy": 0.38035877400862117,
      "eval_loss": 3.4135515689849854,
      "eval_runtime": 183.7281,
      "eval_samples_per_second": 98.031,
      "eval_steps_per_second": 6.129,
      "step": 46000
    },
    {
      "epoch": 4.964959568733153,
      "grad_norm": 0.6403006911277771,
      "learning_rate": 0.00030254506206152187,
      "loss": 3.3988,
      "step": 46050
    },
    {
      "epoch": 4.9703504043126685,
      "grad_norm": 0.6312180757522583,
      "learning_rate": 0.00030222126281705336,
      "loss": 3.3685,
      "step": 46100
    },
    {
      "epoch": 4.975741239892184,
      "grad_norm": 0.6532524228096008,
      "learning_rate": 0.000301897463572585,
      "loss": 3.3742,
      "step": 46150
    },
    {
      "epoch": 4.981132075471698,
      "grad_norm": 0.6965556144714355,
      "learning_rate": 0.0003015736643281165,
      "loss": 3.3847,
      "step": 46200
    },
    {
      "epoch": 4.986522911051213,
      "grad_norm": 0.6514734625816345,
      "learning_rate": 0.00030124986508364807,
      "loss": 3.4039,
      "step": 46250
    },
    {
      "epoch": 4.991913746630727,
      "grad_norm": 0.6422972679138184,
      "learning_rate": 0.0003009260658391797,
      "loss": 3.4007,
      "step": 46300
    },
    {
      "epoch": 4.997304582210242,
      "grad_norm": 0.6854938268661499,
      "learning_rate": 0.0003006022665947112,
      "loss": 3.3792,
      "step": 46350
    },
    {
      "epoch": 5.002695417789758,
      "grad_norm": 0.6698465943336487,
      "learning_rate": 0.00030027846735024283,
      "loss": 3.3286,
      "step": 46400
    },
    {
      "epoch": 5.008086253369272,
      "grad_norm": 0.7109942436218262,
      "learning_rate": 0.00029995466810577443,
      "loss": 3.3071,
      "step": 46450
    },
    {
      "epoch": 5.013477088948787,
      "grad_norm": 0.6470203995704651,
      "learning_rate": 0.00029963086886130593,
      "loss": 3.3059,
      "step": 46500
    },
    {
      "epoch": 5.018867924528302,
      "grad_norm": 0.6870930790901184,
      "learning_rate": 0.00029930706961683754,
      "loss": 3.3071,
      "step": 46550
    },
    {
      "epoch": 5.024258760107816,
      "grad_norm": 0.6598406434059143,
      "learning_rate": 0.0002989832703723691,
      "loss": 3.2899,
      "step": 46600
    },
    {
      "epoch": 5.0296495956873315,
      "grad_norm": 0.7045952081680298,
      "learning_rate": 0.0002986594711279007,
      "loss": 3.2978,
      "step": 46650
    },
    {
      "epoch": 5.035040431266847,
      "grad_norm": 0.643715500831604,
      "learning_rate": 0.00029833567188343224,
      "loss": 3.2985,
      "step": 46700
    },
    {
      "epoch": 5.040431266846361,
      "grad_norm": 0.688113272190094,
      "learning_rate": 0.00029801187263896385,
      "loss": 3.3033,
      "step": 46750
    },
    {
      "epoch": 5.045822102425876,
      "grad_norm": 0.6977607607841492,
      "learning_rate": 0.0002976880733944954,
      "loss": 3.3141,
      "step": 46800
    },
    {
      "epoch": 5.051212938005391,
      "grad_norm": 0.6813511252403259,
      "learning_rate": 0.00029736427415002695,
      "loss": 3.3115,
      "step": 46850
    },
    {
      "epoch": 5.056603773584905,
      "grad_norm": 0.6466038823127747,
      "learning_rate": 0.0002970404749055585,
      "loss": 3.3061,
      "step": 46900
    },
    {
      "epoch": 5.061994609164421,
      "grad_norm": 0.6962968111038208,
      "learning_rate": 0.0002967166756610901,
      "loss": 3.304,
      "step": 46950
    },
    {
      "epoch": 5.067385444743936,
      "grad_norm": 0.6491953134536743,
      "learning_rate": 0.00029639287641662165,
      "loss": 3.314,
      "step": 47000
    },
    {
      "epoch": 5.067385444743936,
      "eval_accuracy": 0.38063018892238176,
      "eval_loss": 3.413198947906494,
      "eval_runtime": 183.7299,
      "eval_samples_per_second": 98.03,
      "eval_steps_per_second": 6.129,
      "step": 47000
    },
    {
      "epoch": 5.07277628032345,
      "grad_norm": 0.6980647444725037,
      "learning_rate": 0.00029606907717215326,
      "loss": 3.3073,
      "step": 47050
    },
    {
      "epoch": 5.078167115902965,
      "grad_norm": 0.6380411386489868,
      "learning_rate": 0.0002957452779276848,
      "loss": 3.3019,
      "step": 47100
    },
    {
      "epoch": 5.083557951482479,
      "grad_norm": 0.643720805644989,
      "learning_rate": 0.00029542147868321636,
      "loss": 3.3155,
      "step": 47150
    },
    {
      "epoch": 5.0889487870619945,
      "grad_norm": 0.6720905303955078,
      "learning_rate": 0.00029509767943874796,
      "loss": 3.314,
      "step": 47200
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.6560536026954651,
      "learning_rate": 0.0002947738801942795,
      "loss": 3.2982,
      "step": 47250
    },
    {
      "epoch": 5.099730458221024,
      "grad_norm": 0.68781578540802,
      "learning_rate": 0.0002944500809498111,
      "loss": 3.3222,
      "step": 47300
    },
    {
      "epoch": 5.105121293800539,
      "grad_norm": 0.6361771821975708,
      "learning_rate": 0.00029412628170534267,
      "loss": 3.3223,
      "step": 47350
    },
    {
      "epoch": 5.110512129380054,
      "grad_norm": 0.6629987955093384,
      "learning_rate": 0.0002938024824608742,
      "loss": 3.3037,
      "step": 47400
    },
    {
      "epoch": 5.115902964959568,
      "grad_norm": 0.7126314640045166,
      "learning_rate": 0.00029347868321640577,
      "loss": 3.3074,
      "step": 47450
    },
    {
      "epoch": 5.121293800539084,
      "grad_norm": 0.7076786160469055,
      "learning_rate": 0.0002931548839719374,
      "loss": 3.2979,
      "step": 47500
    },
    {
      "epoch": 5.126684636118599,
      "grad_norm": 0.647883415222168,
      "learning_rate": 0.0002928310847274689,
      "loss": 3.3051,
      "step": 47550
    },
    {
      "epoch": 5.132075471698113,
      "grad_norm": 0.7320659756660461,
      "learning_rate": 0.00029250728548300053,
      "loss": 3.2956,
      "step": 47600
    },
    {
      "epoch": 5.137466307277628,
      "grad_norm": 0.7425462603569031,
      "learning_rate": 0.0002921834862385321,
      "loss": 3.3094,
      "step": 47650
    },
    {
      "epoch": 5.142857142857143,
      "grad_norm": 0.726519763469696,
      "learning_rate": 0.0002918596869940637,
      "loss": 3.3247,
      "step": 47700
    },
    {
      "epoch": 5.1482479784366575,
      "grad_norm": 0.6692786812782288,
      "learning_rate": 0.0002915358877495952,
      "loss": 3.3143,
      "step": 47750
    },
    {
      "epoch": 5.153638814016173,
      "grad_norm": 0.6244843602180481,
      "learning_rate": 0.0002912120885051268,
      "loss": 3.2996,
      "step": 47800
    },
    {
      "epoch": 5.159029649595688,
      "grad_norm": 0.6885166168212891,
      "learning_rate": 0.00029089476524554774,
      "loss": 3.3091,
      "step": 47850
    },
    {
      "epoch": 5.164420485175202,
      "grad_norm": 0.695482611656189,
      "learning_rate": 0.0002905709660010793,
      "loss": 3.333,
      "step": 47900
    },
    {
      "epoch": 5.169811320754717,
      "grad_norm": 0.6709295511245728,
      "learning_rate": 0.0002902471667566109,
      "loss": 3.3144,
      "step": 47950
    },
    {
      "epoch": 5.175202156334231,
      "grad_norm": 0.6911921501159668,
      "learning_rate": 0.00028992336751214245,
      "loss": 3.3085,
      "step": 48000
    },
    {
      "epoch": 5.175202156334231,
      "eval_accuracy": 0.3809645138581869,
      "eval_loss": 3.4129929542541504,
      "eval_runtime": 183.2606,
      "eval_samples_per_second": 98.281,
      "eval_steps_per_second": 6.144,
      "step": 48000
    },
    {
      "epoch": 5.180592991913747,
      "grad_norm": 0.6802172064781189,
      "learning_rate": 0.00028959956826767405,
      "loss": 3.3148,
      "step": 48050
    },
    {
      "epoch": 5.185983827493262,
      "grad_norm": 0.6930907964706421,
      "learning_rate": 0.00028927576902320555,
      "loss": 3.304,
      "step": 48100
    },
    {
      "epoch": 5.191374663072776,
      "grad_norm": 0.6901477575302124,
      "learning_rate": 0.00028895196977873715,
      "loss": 3.3125,
      "step": 48150
    },
    {
      "epoch": 5.196765498652291,
      "grad_norm": 0.7224954962730408,
      "learning_rate": 0.0002886281705342687,
      "loss": 3.3272,
      "step": 48200
    },
    {
      "epoch": 5.202156334231806,
      "grad_norm": 0.665001392364502,
      "learning_rate": 0.0002883043712898003,
      "loss": 3.3197,
      "step": 48250
    },
    {
      "epoch": 5.2075471698113205,
      "grad_norm": 0.6910035610198975,
      "learning_rate": 0.00028798057204533186,
      "loss": 3.3222,
      "step": 48300
    },
    {
      "epoch": 5.212938005390836,
      "grad_norm": 0.6930015087127686,
      "learning_rate": 0.00028765677280086346,
      "loss": 3.3286,
      "step": 48350
    },
    {
      "epoch": 5.218328840970351,
      "grad_norm": 0.7036418914794922,
      "learning_rate": 0.000287332973556395,
      "loss": 3.3236,
      "step": 48400
    },
    {
      "epoch": 5.223719676549865,
      "grad_norm": 0.671301007270813,
      "learning_rate": 0.0002870091743119266,
      "loss": 3.3137,
      "step": 48450
    },
    {
      "epoch": 5.22911051212938,
      "grad_norm": 0.7079598903656006,
      "learning_rate": 0.00028668537506745817,
      "loss": 3.3494,
      "step": 48500
    },
    {
      "epoch": 5.234501347708895,
      "grad_norm": 0.6326192617416382,
      "learning_rate": 0.0002863615758229897,
      "loss": 3.3194,
      "step": 48550
    },
    {
      "epoch": 5.2398921832884096,
      "grad_norm": 0.6443894505500793,
      "learning_rate": 0.00028603777657852127,
      "loss": 3.3204,
      "step": 48600
    },
    {
      "epoch": 5.245283018867925,
      "grad_norm": 0.6314318776130676,
      "learning_rate": 0.0002857139773340529,
      "loss": 3.3184,
      "step": 48650
    },
    {
      "epoch": 5.250673854447439,
      "grad_norm": 1.178070068359375,
      "learning_rate": 0.0002853901780895844,
      "loss": 3.3221,
      "step": 48700
    },
    {
      "epoch": 5.256064690026954,
      "grad_norm": 0.6846694350242615,
      "learning_rate": 0.00028506637884511603,
      "loss": 3.3428,
      "step": 48750
    },
    {
      "epoch": 5.261455525606469,
      "grad_norm": 0.6815780401229858,
      "learning_rate": 0.0002847425796006476,
      "loss": 3.3422,
      "step": 48800
    },
    {
      "epoch": 5.2668463611859835,
      "grad_norm": 0.6925288438796997,
      "learning_rate": 0.00028441878035617913,
      "loss": 3.3171,
      "step": 48850
    },
    {
      "epoch": 5.272237196765499,
      "grad_norm": 0.7188766598701477,
      "learning_rate": 0.00028409498111171073,
      "loss": 3.3183,
      "step": 48900
    },
    {
      "epoch": 5.277628032345014,
      "grad_norm": 0.7365155220031738,
      "learning_rate": 0.0002837711818672423,
      "loss": 3.3391,
      "step": 48950
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.668428897857666,
      "learning_rate": 0.00028344738262277384,
      "loss": 3.3291,
      "step": 49000
    },
    {
      "epoch": 5.283018867924528,
      "eval_accuracy": 0.38120333290560315,
      "eval_loss": 3.4074130058288574,
      "eval_runtime": 183.7218,
      "eval_samples_per_second": 98.034,
      "eval_steps_per_second": 6.129,
      "step": 49000
    },
    {
      "epoch": 5.288409703504043,
      "grad_norm": 0.6253606677055359,
      "learning_rate": 0.00028312358337830544,
      "loss": 3.3379,
      "step": 49050
    },
    {
      "epoch": 5.293800539083558,
      "grad_norm": 0.6850340962409973,
      "learning_rate": 0.000282799784133837,
      "loss": 3.3321,
      "step": 49100
    },
    {
      "epoch": 5.2991913746630726,
      "grad_norm": 0.6750385165214539,
      "learning_rate": 0.00028247598488936854,
      "loss": 3.3324,
      "step": 49150
    },
    {
      "epoch": 5.304582210242588,
      "grad_norm": 0.714887261390686,
      "learning_rate": 0.00028215218564490015,
      "loss": 3.3403,
      "step": 49200
    },
    {
      "epoch": 5.309973045822103,
      "grad_norm": 0.6349644660949707,
      "learning_rate": 0.0002818283864004317,
      "loss": 3.3279,
      "step": 49250
    },
    {
      "epoch": 5.315363881401617,
      "grad_norm": 0.7085480093955994,
      "learning_rate": 0.0002815045871559633,
      "loss": 3.3278,
      "step": 49300
    },
    {
      "epoch": 5.320754716981132,
      "grad_norm": 0.6977726817131042,
      "learning_rate": 0.00028118078791149485,
      "loss": 3.338,
      "step": 49350
    },
    {
      "epoch": 5.3261455525606465,
      "grad_norm": 0.7166506052017212,
      "learning_rate": 0.0002808569886670264,
      "loss": 3.3244,
      "step": 49400
    },
    {
      "epoch": 5.331536388140162,
      "grad_norm": 0.7002288103103638,
      "learning_rate": 0.00028053318942255795,
      "loss": 3.3195,
      "step": 49450
    },
    {
      "epoch": 5.336927223719677,
      "grad_norm": 0.7169665098190308,
      "learning_rate": 0.00028020939017808956,
      "loss": 3.3464,
      "step": 49500
    },
    {
      "epoch": 5.342318059299191,
      "grad_norm": 0.6901816129684448,
      "learning_rate": 0.0002798855909336211,
      "loss": 3.317,
      "step": 49550
    },
    {
      "epoch": 5.347708894878706,
      "grad_norm": 0.7446085214614868,
      "learning_rate": 0.0002795617916891527,
      "loss": 3.3362,
      "step": 49600
    },
    {
      "epoch": 5.353099730458221,
      "grad_norm": 0.6764737963676453,
      "learning_rate": 0.00027923799244468426,
      "loss": 3.331,
      "step": 49650
    },
    {
      "epoch": 5.3584905660377355,
      "grad_norm": 0.6573972105979919,
      "learning_rate": 0.00027891419320021587,
      "loss": 3.3411,
      "step": 49700
    },
    {
      "epoch": 5.363881401617251,
      "grad_norm": 0.7208666205406189,
      "learning_rate": 0.0002785903939557474,
      "loss": 3.3244,
      "step": 49750
    },
    {
      "epoch": 5.369272237196766,
      "grad_norm": 0.741628110408783,
      "learning_rate": 0.00027826659471127897,
      "loss": 3.3276,
      "step": 49800
    },
    {
      "epoch": 5.37466307277628,
      "grad_norm": 0.6378738880157471,
      "learning_rate": 0.0002779427954668105,
      "loss": 3.3416,
      "step": 49850
    },
    {
      "epoch": 5.380053908355795,
      "grad_norm": 0.6997308135032654,
      "learning_rate": 0.0002776189962223421,
      "loss": 3.3354,
      "step": 49900
    },
    {
      "epoch": 5.38544474393531,
      "grad_norm": 0.7298374176025391,
      "learning_rate": 0.0002772951969778737,
      "loss": 3.3216,
      "step": 49950
    },
    {
      "epoch": 5.390835579514825,
      "grad_norm": 0.6540275812149048,
      "learning_rate": 0.0002769713977334053,
      "loss": 3.3413,
      "step": 50000
    },
    {
      "epoch": 5.390835579514825,
      "eval_accuracy": 0.3816847738515091,
      "eval_loss": 3.401179075241089,
      "eval_runtime": 183.4607,
      "eval_samples_per_second": 98.174,
      "eval_steps_per_second": 6.138,
      "step": 50000
    },
    {
      "epoch": 5.39622641509434,
      "grad_norm": 0.6782923936843872,
      "learning_rate": 0.00027665407447382623,
      "loss": 3.3526,
      "step": 50050
    },
    {
      "epoch": 5.401617250673855,
      "grad_norm": 0.6657195091247559,
      "learning_rate": 0.0002763302752293578,
      "loss": 3.3432,
      "step": 50100
    },
    {
      "epoch": 5.407008086253369,
      "grad_norm": 0.6890636086463928,
      "learning_rate": 0.00027600647598488933,
      "loss": 3.3339,
      "step": 50150
    },
    {
      "epoch": 5.412398921832884,
      "grad_norm": 0.7247375845909119,
      "learning_rate": 0.0002756826767404209,
      "loss": 3.3174,
      "step": 50200
    },
    {
      "epoch": 5.4177897574123985,
      "grad_norm": 0.6520726680755615,
      "learning_rate": 0.0002753588774959525,
      "loss": 3.3458,
      "step": 50250
    },
    {
      "epoch": 5.423180592991914,
      "grad_norm": 0.6332509517669678,
      "learning_rate": 0.00027503507825148404,
      "loss": 3.3392,
      "step": 50300
    },
    {
      "epoch": 5.428571428571429,
      "grad_norm": 0.7123890519142151,
      "learning_rate": 0.00027471127900701564,
      "loss": 3.3293,
      "step": 50350
    },
    {
      "epoch": 5.433962264150943,
      "grad_norm": 0.7349735498428345,
      "learning_rate": 0.0002743874797625472,
      "loss": 3.3496,
      "step": 50400
    },
    {
      "epoch": 5.439353099730458,
      "grad_norm": 0.6921404004096985,
      "learning_rate": 0.0002740636805180788,
      "loss": 3.3469,
      "step": 50450
    },
    {
      "epoch": 5.444743935309973,
      "grad_norm": 0.6628040075302124,
      "learning_rate": 0.00027373988127361035,
      "loss": 3.3299,
      "step": 50500
    },
    {
      "epoch": 5.450134770889488,
      "grad_norm": 0.6844006776809692,
      "learning_rate": 0.0002734160820291419,
      "loss": 3.3262,
      "step": 50550
    },
    {
      "epoch": 5.455525606469003,
      "grad_norm": 0.714565098285675,
      "learning_rate": 0.00027309228278467345,
      "loss": 3.3431,
      "step": 50600
    },
    {
      "epoch": 5.460916442048518,
      "grad_norm": 0.8186658620834351,
      "learning_rate": 0.00027276848354020506,
      "loss": 3.3354,
      "step": 50650
    },
    {
      "epoch": 5.466307277628032,
      "grad_norm": 0.6907452344894409,
      "learning_rate": 0.0002724446842957366,
      "loss": 3.3278,
      "step": 50700
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.6899653077125549,
      "learning_rate": 0.0002721208850512682,
      "loss": 3.3141,
      "step": 50750
    },
    {
      "epoch": 5.4770889487870615,
      "grad_norm": 0.7394895553588867,
      "learning_rate": 0.00027179708580679976,
      "loss": 3.3432,
      "step": 50800
    },
    {
      "epoch": 5.482479784366577,
      "grad_norm": 0.7098657488822937,
      "learning_rate": 0.0002714732865623313,
      "loss": 3.3277,
      "step": 50850
    },
    {
      "epoch": 5.487870619946092,
      "grad_norm": 0.6476383805274963,
      "learning_rate": 0.0002711494873178629,
      "loss": 3.3377,
      "step": 50900
    },
    {
      "epoch": 5.493261455525606,
      "grad_norm": 0.7089292407035828,
      "learning_rate": 0.00027082568807339447,
      "loss": 3.3536,
      "step": 50950
    },
    {
      "epoch": 5.498652291105121,
      "grad_norm": 0.6908971071243286,
      "learning_rate": 0.000270501888828926,
      "loss": 3.3168,
      "step": 51000
    },
    {
      "epoch": 5.498652291105121,
      "eval_accuracy": 0.3820141007544748,
      "eval_loss": 3.397711753845215,
      "eval_runtime": 183.6128,
      "eval_samples_per_second": 98.092,
      "eval_steps_per_second": 6.132,
      "step": 51000
    },
    {
      "epoch": 5.504043126684636,
      "grad_norm": 0.6817182898521423,
      "learning_rate": 0.0002701780895844576,
      "loss": 3.3255,
      "step": 51050
    },
    {
      "epoch": 5.509433962264151,
      "grad_norm": 0.707398533821106,
      "learning_rate": 0.0002698542903399892,
      "loss": 3.3347,
      "step": 51100
    },
    {
      "epoch": 5.514824797843666,
      "grad_norm": 0.6728003621101379,
      "learning_rate": 0.0002695304910955207,
      "loss": 3.3509,
      "step": 51150
    },
    {
      "epoch": 5.520215633423181,
      "grad_norm": 0.700847864151001,
      "learning_rate": 0.00026920669185105233,
      "loss": 3.3274,
      "step": 51200
    },
    {
      "epoch": 5.525606469002695,
      "grad_norm": 0.7568596005439758,
      "learning_rate": 0.0002688828926065839,
      "loss": 3.3324,
      "step": 51250
    },
    {
      "epoch": 5.53099730458221,
      "grad_norm": 0.6875473856925964,
      "learning_rate": 0.0002685590933621155,
      "loss": 3.3451,
      "step": 51300
    },
    {
      "epoch": 5.536388140161725,
      "grad_norm": 0.6690535545349121,
      "learning_rate": 0.00026823529411764704,
      "loss": 3.324,
      "step": 51350
    },
    {
      "epoch": 5.54177897574124,
      "grad_norm": 0.7146397829055786,
      "learning_rate": 0.0002679114948731786,
      "loss": 3.3327,
      "step": 51400
    },
    {
      "epoch": 5.547169811320755,
      "grad_norm": 0.7005475759506226,
      "learning_rate": 0.0002675876956287102,
      "loss": 3.3507,
      "step": 51450
    },
    {
      "epoch": 5.55256064690027,
      "grad_norm": 0.6918277144432068,
      "learning_rate": 0.00026726389638424174,
      "loss": 3.3357,
      "step": 51500
    },
    {
      "epoch": 5.557951482479784,
      "grad_norm": 0.701012134552002,
      "learning_rate": 0.0002669400971397733,
      "loss": 3.3272,
      "step": 51550
    },
    {
      "epoch": 5.563342318059299,
      "grad_norm": 0.7084556221961975,
      "learning_rate": 0.0002666162978953049,
      "loss": 3.3545,
      "step": 51600
    },
    {
      "epoch": 5.568733153638814,
      "grad_norm": 0.6553547978401184,
      "learning_rate": 0.00026629249865083645,
      "loss": 3.3508,
      "step": 51650
    },
    {
      "epoch": 5.574123989218329,
      "grad_norm": 0.7182124853134155,
      "learning_rate": 0.00026596869940636805,
      "loss": 3.3333,
      "step": 51700
    },
    {
      "epoch": 5.579514824797844,
      "grad_norm": 0.700818657875061,
      "learning_rate": 0.0002656449001618996,
      "loss": 3.3294,
      "step": 51750
    },
    {
      "epoch": 5.584905660377358,
      "grad_norm": 0.7259700894355774,
      "learning_rate": 0.0002653211009174312,
      "loss": 3.3407,
      "step": 51800
    },
    {
      "epoch": 5.590296495956873,
      "grad_norm": 0.7825680375099182,
      "learning_rate": 0.0002649973016729627,
      "loss": 3.3395,
      "step": 51850
    },
    {
      "epoch": 5.595687331536388,
      "grad_norm": 0.6643126606941223,
      "learning_rate": 0.0002646735024284943,
      "loss": 3.3443,
      "step": 51900
    },
    {
      "epoch": 5.601078167115903,
      "grad_norm": 0.6935634613037109,
      "learning_rate": 0.00026434970318402586,
      "loss": 3.3319,
      "step": 51950
    },
    {
      "epoch": 5.606469002695418,
      "grad_norm": 0.7245262265205383,
      "learning_rate": 0.00026402590393955746,
      "loss": 3.3291,
      "step": 52000
    },
    {
      "epoch": 5.606469002695418,
      "eval_accuracy": 0.3823364738726204,
      "eval_loss": 3.394861936569214,
      "eval_runtime": 183.6987,
      "eval_samples_per_second": 98.046,
      "eval_steps_per_second": 6.13,
      "step": 52000
    },
    {
      "epoch": 5.611859838274933,
      "grad_norm": 0.6866488456726074,
      "learning_rate": 0.000263702104695089,
      "loss": 3.3532,
      "step": 52050
    },
    {
      "epoch": 5.617250673854447,
      "grad_norm": 0.7162625193595886,
      "learning_rate": 0.0002633783054506206,
      "loss": 3.3288,
      "step": 52100
    },
    {
      "epoch": 5.622641509433962,
      "grad_norm": 0.7691003084182739,
      "learning_rate": 0.00026305450620615217,
      "loss": 3.3342,
      "step": 52150
    },
    {
      "epoch": 5.628032345013477,
      "grad_norm": 0.7199847102165222,
      "learning_rate": 0.0002627307069616837,
      "loss": 3.3293,
      "step": 52200
    },
    {
      "epoch": 5.633423180592992,
      "grad_norm": 0.6812527179718018,
      "learning_rate": 0.0002624069077172153,
      "loss": 3.3322,
      "step": 52250
    },
    {
      "epoch": 5.638814016172507,
      "grad_norm": 0.6690404415130615,
      "learning_rate": 0.0002620831084727469,
      "loss": 3.3338,
      "step": 52300
    },
    {
      "epoch": 5.644204851752022,
      "grad_norm": 0.6687739491462708,
      "learning_rate": 0.0002617593092282784,
      "loss": 3.3402,
      "step": 52350
    },
    {
      "epoch": 5.649595687331536,
      "grad_norm": 0.6985411047935486,
      "learning_rate": 0.00026143550998381003,
      "loss": 3.3269,
      "step": 52400
    },
    {
      "epoch": 5.654986522911051,
      "grad_norm": 0.6869599223136902,
      "learning_rate": 0.0002611117107393416,
      "loss": 3.3572,
      "step": 52450
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.7344062328338623,
      "learning_rate": 0.00026078791149487313,
      "loss": 3.3319,
      "step": 52500
    },
    {
      "epoch": 5.665768194070081,
      "grad_norm": 0.7084963321685791,
      "learning_rate": 0.00026046411225040474,
      "loss": 3.3483,
      "step": 52550
    },
    {
      "epoch": 5.671159029649596,
      "grad_norm": 0.6755357980728149,
      "learning_rate": 0.0002601403130059363,
      "loss": 3.3606,
      "step": 52600
    },
    {
      "epoch": 5.67654986522911,
      "grad_norm": 0.7375144362449646,
      "learning_rate": 0.0002598165137614679,
      "loss": 3.3331,
      "step": 52650
    },
    {
      "epoch": 5.681940700808625,
      "grad_norm": 0.7138906121253967,
      "learning_rate": 0.00025949271451699944,
      "loss": 3.3237,
      "step": 52700
    },
    {
      "epoch": 5.6873315363881405,
      "grad_norm": 0.6811950206756592,
      "learning_rate": 0.000259168915272531,
      "loss": 3.3397,
      "step": 52750
    },
    {
      "epoch": 5.692722371967655,
      "grad_norm": 0.6969444751739502,
      "learning_rate": 0.0002588451160280626,
      "loss": 3.3228,
      "step": 52800
    },
    {
      "epoch": 5.69811320754717,
      "grad_norm": 0.6877330541610718,
      "learning_rate": 0.00025852131678359415,
      "loss": 3.3201,
      "step": 52850
    },
    {
      "epoch": 5.703504043126685,
      "grad_norm": 0.6857284903526306,
      "learning_rate": 0.0002581975175391257,
      "loss": 3.3244,
      "step": 52900
    },
    {
      "epoch": 5.708894878706199,
      "grad_norm": 0.7053078413009644,
      "learning_rate": 0.00025788019427954665,
      "loss": 3.3277,
      "step": 52950
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.7215891480445862,
      "learning_rate": 0.00025755639503507826,
      "loss": 3.3407,
      "step": 53000
    },
    {
      "epoch": 5.714285714285714,
      "eval_accuracy": 0.38302468126403727,
      "eval_loss": 3.3901915550231934,
      "eval_runtime": 183.3315,
      "eval_samples_per_second": 98.243,
      "eval_steps_per_second": 6.142,
      "step": 53000
    },
    {
      "epoch": 5.719676549865229,
      "grad_norm": 0.7056070566177368,
      "learning_rate": 0.00025723907177549915,
      "loss": 3.3326,
      "step": 53050
    },
    {
      "epoch": 5.725067385444744,
      "grad_norm": 0.7360427975654602,
      "learning_rate": 0.00025691527253103076,
      "loss": 3.3417,
      "step": 53100
    },
    {
      "epoch": 5.730458221024259,
      "grad_norm": 0.6608803868293762,
      "learning_rate": 0.0002565914732865623,
      "loss": 3.3341,
      "step": 53150
    },
    {
      "epoch": 5.735849056603773,
      "grad_norm": 0.7453342080116272,
      "learning_rate": 0.00025626767404209386,
      "loss": 3.3453,
      "step": 53200
    },
    {
      "epoch": 5.741239892183288,
      "grad_norm": 0.6823360919952393,
      "learning_rate": 0.00025594387479762546,
      "loss": 3.317,
      "step": 53250
    },
    {
      "epoch": 5.7466307277628035,
      "grad_norm": 0.7249504923820496,
      "learning_rate": 0.000255620075553157,
      "loss": 3.3596,
      "step": 53300
    },
    {
      "epoch": 5.752021563342318,
      "grad_norm": 0.6726418137550354,
      "learning_rate": 0.00025529627630868857,
      "loss": 3.342,
      "step": 53350
    },
    {
      "epoch": 5.757412398921833,
      "grad_norm": 0.7085649371147156,
      "learning_rate": 0.00025497247706422017,
      "loss": 3.3468,
      "step": 53400
    },
    {
      "epoch": 5.762803234501348,
      "grad_norm": 0.6718982458114624,
      "learning_rate": 0.0002546486778197517,
      "loss": 3.3439,
      "step": 53450
    },
    {
      "epoch": 5.768194070080862,
      "grad_norm": 0.6999180316925049,
      "learning_rate": 0.00025432487857528327,
      "loss": 3.326,
      "step": 53500
    },
    {
      "epoch": 5.773584905660377,
      "grad_norm": 0.7056639790534973,
      "learning_rate": 0.0002540010793308149,
      "loss": 3.325,
      "step": 53550
    },
    {
      "epoch": 5.7789757412398925,
      "grad_norm": 0.7867892384529114,
      "learning_rate": 0.0002536772800863464,
      "loss": 3.3445,
      "step": 53600
    },
    {
      "epoch": 5.784366576819407,
      "grad_norm": 0.673121988773346,
      "learning_rate": 0.00025335348084187803,
      "loss": 3.3291,
      "step": 53650
    },
    {
      "epoch": 5.789757412398922,
      "grad_norm": 0.6664978265762329,
      "learning_rate": 0.0002530296815974096,
      "loss": 3.349,
      "step": 53700
    },
    {
      "epoch": 5.795148247978437,
      "grad_norm": 0.6702294945716858,
      "learning_rate": 0.0002527058823529412,
      "loss": 3.3366,
      "step": 53750
    },
    {
      "epoch": 5.800539083557951,
      "grad_norm": 0.6929095387458801,
      "learning_rate": 0.0002523820831084727,
      "loss": 3.3253,
      "step": 53800
    },
    {
      "epoch": 5.8059299191374665,
      "grad_norm": 0.7197151184082031,
      "learning_rate": 0.0002520582838640043,
      "loss": 3.3309,
      "step": 53850
    },
    {
      "epoch": 5.811320754716981,
      "grad_norm": 0.7062203884124756,
      "learning_rate": 0.00025173448461953584,
      "loss": 3.3312,
      "step": 53900
    },
    {
      "epoch": 5.816711590296496,
      "grad_norm": 0.7010804414749146,
      "learning_rate": 0.00025141068537506744,
      "loss": 3.3279,
      "step": 53950
    },
    {
      "epoch": 5.822102425876011,
      "grad_norm": 0.6943368315696716,
      "learning_rate": 0.000251086886130599,
      "loss": 3.3243,
      "step": 54000
    },
    {
      "epoch": 5.822102425876011,
      "eval_accuracy": 0.3835389353820632,
      "eval_loss": 3.3847732543945312,
      "eval_runtime": 184.1098,
      "eval_samples_per_second": 97.827,
      "eval_steps_per_second": 6.116,
      "step": 54000
    },
    {
      "epoch": 5.827493261455525,
      "grad_norm": 0.7037324905395508,
      "learning_rate": 0.0002507630868861306,
      "loss": 3.3363,
      "step": 54050
    },
    {
      "epoch": 5.83288409703504,
      "grad_norm": 0.688883900642395,
      "learning_rate": 0.00025043928764166215,
      "loss": 3.3239,
      "step": 54100
    },
    {
      "epoch": 5.8382749326145555,
      "grad_norm": 0.7349210977554321,
      "learning_rate": 0.00025011548839719375,
      "loss": 3.3382,
      "step": 54150
    },
    {
      "epoch": 5.84366576819407,
      "grad_norm": 0.6746708154678345,
      "learning_rate": 0.0002497916891527253,
      "loss": 3.3284,
      "step": 54200
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.6808030605316162,
      "learning_rate": 0.00024946788990825686,
      "loss": 3.3277,
      "step": 54250
    },
    {
      "epoch": 5.8544474393531,
      "grad_norm": 0.7390064001083374,
      "learning_rate": 0.0002491440906637884,
      "loss": 3.3295,
      "step": 54300
    },
    {
      "epoch": 5.859838274932614,
      "grad_norm": 0.743797779083252,
      "learning_rate": 0.00024882029141932,
      "loss": 3.3404,
      "step": 54350
    },
    {
      "epoch": 5.8652291105121295,
      "grad_norm": 0.6696168184280396,
      "learning_rate": 0.00024849649217485156,
      "loss": 3.3267,
      "step": 54400
    },
    {
      "epoch": 5.870619946091644,
      "grad_norm": 0.7122966051101685,
      "learning_rate": 0.00024817269293038317,
      "loss": 3.345,
      "step": 54450
    },
    {
      "epoch": 5.876010781671159,
      "grad_norm": 0.6853727698326111,
      "learning_rate": 0.0002478488936859147,
      "loss": 3.345,
      "step": 54500
    },
    {
      "epoch": 5.881401617250674,
      "grad_norm": 0.7064388394355774,
      "learning_rate": 0.00024752509444144627,
      "loss": 3.3174,
      "step": 54550
    },
    {
      "epoch": 5.886792452830189,
      "grad_norm": 0.6728473901748657,
      "learning_rate": 0.00024720129519697787,
      "loss": 3.3434,
      "step": 54600
    },
    {
      "epoch": 5.892183288409703,
      "grad_norm": 0.6776952743530273,
      "learning_rate": 0.0002468774959525094,
      "loss": 3.3255,
      "step": 54650
    },
    {
      "epoch": 5.8975741239892185,
      "grad_norm": 0.7131171226501465,
      "learning_rate": 0.00024655369670804097,
      "loss": 3.332,
      "step": 54700
    },
    {
      "epoch": 5.902964959568733,
      "grad_norm": 0.665278434753418,
      "learning_rate": 0.0002462298974635726,
      "loss": 3.3409,
      "step": 54750
    },
    {
      "epoch": 5.908355795148248,
      "grad_norm": 0.6969947814941406,
      "learning_rate": 0.00024590609821910413,
      "loss": 3.3351,
      "step": 54800
    },
    {
      "epoch": 5.913746630727763,
      "grad_norm": 0.7671633362770081,
      "learning_rate": 0.0002455822989746357,
      "loss": 3.3422,
      "step": 54850
    },
    {
      "epoch": 5.919137466307277,
      "grad_norm": 0.707665741443634,
      "learning_rate": 0.0002452584997301673,
      "loss": 3.3422,
      "step": 54900
    },
    {
      "epoch": 5.9245283018867925,
      "grad_norm": 0.7261333465576172,
      "learning_rate": 0.00024493470048569883,
      "loss": 3.3448,
      "step": 54950
    },
    {
      "epoch": 5.929919137466308,
      "grad_norm": 0.7175887823104858,
      "learning_rate": 0.00024461090124123044,
      "loss": 3.3262,
      "step": 55000
    },
    {
      "epoch": 5.929919137466308,
      "eval_accuracy": 0.3838151310228876,
      "eval_loss": 3.3818440437316895,
      "eval_runtime": 183.5588,
      "eval_samples_per_second": 98.121,
      "eval_steps_per_second": 6.134,
      "step": 55000
    },
    {
      "epoch": 5.935309973045822,
      "grad_norm": 0.7148836255073547,
      "learning_rate": 0.000244287101996762,
      "loss": 3.3569,
      "step": 55050
    },
    {
      "epoch": 5.940700808625337,
      "grad_norm": 0.6707710027694702,
      "learning_rate": 0.00024396330275229354,
      "loss": 3.3205,
      "step": 55100
    },
    {
      "epoch": 5.946091644204852,
      "grad_norm": 0.6846052408218384,
      "learning_rate": 0.00024363950350782512,
      "loss": 3.3286,
      "step": 55150
    },
    {
      "epoch": 5.951482479784366,
      "grad_norm": 0.7205194234848022,
      "learning_rate": 0.0002433157042633567,
      "loss": 3.325,
      "step": 55200
    },
    {
      "epoch": 5.9568733153638815,
      "grad_norm": 0.683093786239624,
      "learning_rate": 0.00024299190501888827,
      "loss": 3.3315,
      "step": 55250
    },
    {
      "epoch": 5.962264150943396,
      "grad_norm": 0.6996801495552063,
      "learning_rate": 0.00024266810577441985,
      "loss": 3.3234,
      "step": 55300
    },
    {
      "epoch": 5.967654986522911,
      "grad_norm": 0.7320035099983215,
      "learning_rate": 0.00024234430652995143,
      "loss": 3.3252,
      "step": 55350
    },
    {
      "epoch": 5.973045822102426,
      "grad_norm": 0.6926111578941345,
      "learning_rate": 0.00024202050728548298,
      "loss": 3.3272,
      "step": 55400
    },
    {
      "epoch": 5.97843665768194,
      "grad_norm": 0.702363133430481,
      "learning_rate": 0.00024169670804101456,
      "loss": 3.3475,
      "step": 55450
    },
    {
      "epoch": 5.9838274932614555,
      "grad_norm": 0.7452805042266846,
      "learning_rate": 0.0002413729087965461,
      "loss": 3.3308,
      "step": 55500
    },
    {
      "epoch": 5.989218328840971,
      "grad_norm": 0.6895799040794373,
      "learning_rate": 0.00024104910955207768,
      "loss": 3.3347,
      "step": 55550
    },
    {
      "epoch": 5.994609164420485,
      "grad_norm": 0.7022313475608826,
      "learning_rate": 0.00024072531030760926,
      "loss": 3.3497,
      "step": 55600
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.475501298904419,
      "learning_rate": 0.00024040151106314084,
      "loss": 3.3485,
      "step": 55650
    },
    {
      "epoch": 6.005390835579515,
      "grad_norm": 0.7008572816848755,
      "learning_rate": 0.0002400777118186724,
      "loss": 3.2428,
      "step": 55700
    },
    {
      "epoch": 6.010781671159029,
      "grad_norm": 0.7365021705627441,
      "learning_rate": 0.00023975391257420397,
      "loss": 3.2469,
      "step": 55750
    },
    {
      "epoch": 6.0161725067385445,
      "grad_norm": 0.7550241947174072,
      "learning_rate": 0.00023943011332973555,
      "loss": 3.239,
      "step": 55800
    },
    {
      "epoch": 6.02156334231806,
      "grad_norm": 0.6840744018554688,
      "learning_rate": 0.00023910631408526712,
      "loss": 3.2463,
      "step": 55850
    },
    {
      "epoch": 6.026954177897574,
      "grad_norm": 0.6900550723075867,
      "learning_rate": 0.0002387825148407987,
      "loss": 3.2467,
      "step": 55900
    },
    {
      "epoch": 6.032345013477089,
      "grad_norm": 0.7249858975410461,
      "learning_rate": 0.00023845871559633025,
      "loss": 3.2418,
      "step": 55950
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.6880434155464172,
      "learning_rate": 0.0002381349163518618,
      "loss": 3.2386,
      "step": 56000
    },
    {
      "epoch": 6.037735849056604,
      "eval_accuracy": 0.38375395944704804,
      "eval_loss": 3.3848366737365723,
      "eval_runtime": 183.6771,
      "eval_samples_per_second": 98.058,
      "eval_steps_per_second": 6.13,
      "step": 56000
    },
    {
      "epoch": 6.0431266846361185,
      "grad_norm": 0.7010095119476318,
      "learning_rate": 0.00023781111710739338,
      "loss": 3.2457,
      "step": 56050
    },
    {
      "epoch": 6.048517520215634,
      "grad_norm": 0.7619755864143372,
      "learning_rate": 0.00023748731786292496,
      "loss": 3.2441,
      "step": 56100
    },
    {
      "epoch": 6.053908355795148,
      "grad_norm": 0.756662130355835,
      "learning_rate": 0.00023716351861845654,
      "loss": 3.2619,
      "step": 56150
    },
    {
      "epoch": 6.059299191374663,
      "grad_norm": 0.7468305826187134,
      "learning_rate": 0.0002368397193739881,
      "loss": 3.2425,
      "step": 56200
    },
    {
      "epoch": 6.064690026954178,
      "grad_norm": 0.7461797595024109,
      "learning_rate": 0.0002365159201295197,
      "loss": 3.2516,
      "step": 56250
    },
    {
      "epoch": 6.070080862533692,
      "grad_norm": 0.7048510313034058,
      "learning_rate": 0.00023619212088505127,
      "loss": 3.2592,
      "step": 56300
    },
    {
      "epoch": 6.0754716981132075,
      "grad_norm": 0.7342618703842163,
      "learning_rate": 0.0002358683216405828,
      "loss": 3.2552,
      "step": 56350
    },
    {
      "epoch": 6.080862533692723,
      "grad_norm": 0.7538067698478699,
      "learning_rate": 0.00023554452239611437,
      "loss": 3.2766,
      "step": 56400
    },
    {
      "epoch": 6.086253369272237,
      "grad_norm": 0.7238980531692505,
      "learning_rate": 0.00023522072315164595,
      "loss": 3.264,
      "step": 56450
    },
    {
      "epoch": 6.091644204851752,
      "grad_norm": 0.7239121794700623,
      "learning_rate": 0.00023489692390717752,
      "loss": 3.2454,
      "step": 56500
    },
    {
      "epoch": 6.097035040431267,
      "grad_norm": 0.6743500828742981,
      "learning_rate": 0.0002345731246627091,
      "loss": 3.2766,
      "step": 56550
    },
    {
      "epoch": 6.1024258760107815,
      "grad_norm": 0.740608811378479,
      "learning_rate": 0.00023424932541824068,
      "loss": 3.2583,
      "step": 56600
    },
    {
      "epoch": 6.107816711590297,
      "grad_norm": 0.729874849319458,
      "learning_rate": 0.00023392552617377226,
      "loss": 3.2427,
      "step": 56650
    },
    {
      "epoch": 6.113207547169812,
      "grad_norm": 0.758840799331665,
      "learning_rate": 0.00023360172692930384,
      "loss": 3.2734,
      "step": 56700
    },
    {
      "epoch": 6.118598382749326,
      "grad_norm": 0.7450424432754517,
      "learning_rate": 0.00023327792768483539,
      "loss": 3.2821,
      "step": 56750
    },
    {
      "epoch": 6.123989218328841,
      "grad_norm": 0.7052544355392456,
      "learning_rate": 0.00023295412844036694,
      "loss": 3.2703,
      "step": 56800
    },
    {
      "epoch": 6.129380053908355,
      "grad_norm": 0.7544761300086975,
      "learning_rate": 0.00023263032919589851,
      "loss": 3.253,
      "step": 56850
    },
    {
      "epoch": 6.1347708894878705,
      "grad_norm": 0.7224695086479187,
      "learning_rate": 0.0002323065299514301,
      "loss": 3.2691,
      "step": 56900
    },
    {
      "epoch": 6.140161725067386,
      "grad_norm": 0.7088389992713928,
      "learning_rate": 0.00023198273070696167,
      "loss": 3.2525,
      "step": 56950
    },
    {
      "epoch": 6.1455525606469,
      "grad_norm": 0.7072451114654541,
      "learning_rate": 0.00023165893146249325,
      "loss": 3.2689,
      "step": 57000
    },
    {
      "epoch": 6.1455525606469,
      "eval_accuracy": 0.3841858546761106,
      "eval_loss": 3.3838181495666504,
      "eval_runtime": 183.7277,
      "eval_samples_per_second": 98.031,
      "eval_steps_per_second": 6.129,
      "step": 57000
    },
    {
      "epoch": 6.150943396226415,
      "grad_norm": 0.6568054556846619,
      "learning_rate": 0.0002313351322180248,
      "loss": 3.2615,
      "step": 57050
    },
    {
      "epoch": 6.15633423180593,
      "grad_norm": 0.7381845116615295,
      "learning_rate": 0.00023101133297355638,
      "loss": 3.2692,
      "step": 57100
    },
    {
      "epoch": 6.1617250673854445,
      "grad_norm": 0.6702138185501099,
      "learning_rate": 0.00023068753372908795,
      "loss": 3.2844,
      "step": 57150
    },
    {
      "epoch": 6.16711590296496,
      "grad_norm": 0.7349806427955627,
      "learning_rate": 0.0002303637344846195,
      "loss": 3.2594,
      "step": 57200
    },
    {
      "epoch": 6.172506738544475,
      "grad_norm": 0.694933295249939,
      "learning_rate": 0.00023003993524015108,
      "loss": 3.2597,
      "step": 57250
    },
    {
      "epoch": 6.177897574123989,
      "grad_norm": 0.7540972232818604,
      "learning_rate": 0.00022971613599568266,
      "loss": 3.275,
      "step": 57300
    },
    {
      "epoch": 6.183288409703504,
      "grad_norm": 0.6981645226478577,
      "learning_rate": 0.0002293923367512142,
      "loss": 3.2605,
      "step": 57350
    },
    {
      "epoch": 6.188679245283019,
      "grad_norm": 0.7014349699020386,
      "learning_rate": 0.0002290685375067458,
      "loss": 3.267,
      "step": 57400
    },
    {
      "epoch": 6.1940700808625335,
      "grad_norm": 0.758674144744873,
      "learning_rate": 0.00022874473826227736,
      "loss": 3.2639,
      "step": 57450
    },
    {
      "epoch": 6.199460916442049,
      "grad_norm": 0.7108901739120483,
      "learning_rate": 0.00022842093901780894,
      "loss": 3.2551,
      "step": 57500
    },
    {
      "epoch": 6.204851752021563,
      "grad_norm": 0.7790617942810059,
      "learning_rate": 0.00022810361575822987,
      "loss": 3.2789,
      "step": 57550
    },
    {
      "epoch": 6.210242587601078,
      "grad_norm": 0.7338255643844604,
      "learning_rate": 0.00022777981651376145,
      "loss": 3.2685,
      "step": 57600
    },
    {
      "epoch": 6.215633423180593,
      "grad_norm": 0.7300395965576172,
      "learning_rate": 0.00022745601726929302,
      "loss": 3.2808,
      "step": 57650
    },
    {
      "epoch": 6.2210242587601075,
      "grad_norm": 0.755389928817749,
      "learning_rate": 0.00022713221802482457,
      "loss": 3.2727,
      "step": 57700
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.6979179382324219,
      "learning_rate": 0.00022680841878035615,
      "loss": 3.2646,
      "step": 57750
    },
    {
      "epoch": 6.231805929919138,
      "grad_norm": 0.7372843027114868,
      "learning_rate": 0.00022648461953588773,
      "loss": 3.2586,
      "step": 57800
    },
    {
      "epoch": 6.237196765498652,
      "grad_norm": 0.6861436367034912,
      "learning_rate": 0.0002261608202914193,
      "loss": 3.2727,
      "step": 57850
    },
    {
      "epoch": 6.242587601078167,
      "grad_norm": 0.7225656509399414,
      "learning_rate": 0.00022583702104695088,
      "loss": 3.2816,
      "step": 57900
    },
    {
      "epoch": 6.247978436657682,
      "grad_norm": 0.7209280729293823,
      "learning_rate": 0.00022551322180248246,
      "loss": 3.2733,
      "step": 57950
    },
    {
      "epoch": 6.2533692722371965,
      "grad_norm": 0.7402727603912354,
      "learning_rate": 0.00022518942255801399,
      "loss": 3.2754,
      "step": 58000
    },
    {
      "epoch": 6.2533692722371965,
      "eval_accuracy": 0.3843185198521321,
      "eval_loss": 3.3821375370025635,
      "eval_runtime": 183.8391,
      "eval_samples_per_second": 97.972,
      "eval_steps_per_second": 6.125,
      "step": 58000
    },
    {
      "epoch": 6.258760107816712,
      "grad_norm": 0.7029194831848145,
      "learning_rate": 0.00022486562331354556,
      "loss": 3.2612,
      "step": 58050
    },
    {
      "epoch": 6.264150943396227,
      "grad_norm": 0.717560350894928,
      "learning_rate": 0.00022454182406907714,
      "loss": 3.2752,
      "step": 58100
    },
    {
      "epoch": 6.269541778975741,
      "grad_norm": 0.7272930145263672,
      "learning_rate": 0.00022421802482460872,
      "loss": 3.2697,
      "step": 58150
    },
    {
      "epoch": 6.274932614555256,
      "grad_norm": 0.7455615997314453,
      "learning_rate": 0.0002238942255801403,
      "loss": 3.292,
      "step": 58200
    },
    {
      "epoch": 6.280323450134771,
      "grad_norm": 0.7354786992073059,
      "learning_rate": 0.00022357042633567187,
      "loss": 3.2735,
      "step": 58250
    },
    {
      "epoch": 6.285714285714286,
      "grad_norm": 0.7349162697792053,
      "learning_rate": 0.00022324662709120345,
      "loss": 3.2632,
      "step": 58300
    },
    {
      "epoch": 6.291105121293801,
      "grad_norm": 0.7539188861846924,
      "learning_rate": 0.00022292282784673503,
      "loss": 3.2846,
      "step": 58350
    },
    {
      "epoch": 6.296495956873315,
      "grad_norm": 0.719907283782959,
      "learning_rate": 0.00022259902860226655,
      "loss": 3.2689,
      "step": 58400
    },
    {
      "epoch": 6.30188679245283,
      "grad_norm": 0.7047970294952393,
      "learning_rate": 0.00022227522935779813,
      "loss": 3.2944,
      "step": 58450
    },
    {
      "epoch": 6.307277628032345,
      "grad_norm": 0.7011443972587585,
      "learning_rate": 0.0002219514301133297,
      "loss": 3.2935,
      "step": 58500
    },
    {
      "epoch": 6.3126684636118595,
      "grad_norm": 0.766208827495575,
      "learning_rate": 0.00022162763086886129,
      "loss": 3.2833,
      "step": 58550
    },
    {
      "epoch": 6.318059299191375,
      "grad_norm": 0.7611593008041382,
      "learning_rate": 0.00022130383162439286,
      "loss": 3.2873,
      "step": 58600
    },
    {
      "epoch": 6.32345013477089,
      "grad_norm": 0.7359244227409363,
      "learning_rate": 0.00022098003237992444,
      "loss": 3.2847,
      "step": 58650
    },
    {
      "epoch": 6.328840970350404,
      "grad_norm": 0.7401198148727417,
      "learning_rate": 0.00022065623313545602,
      "loss": 3.284,
      "step": 58700
    },
    {
      "epoch": 6.334231805929919,
      "grad_norm": 0.7132663726806641,
      "learning_rate": 0.00022033243389098757,
      "loss": 3.2731,
      "step": 58750
    },
    {
      "epoch": 6.339622641509434,
      "grad_norm": 0.7397975325584412,
      "learning_rate": 0.00022000863464651915,
      "loss": 3.2809,
      "step": 58800
    },
    {
      "epoch": 6.345013477088949,
      "grad_norm": 0.6927112340927124,
      "learning_rate": 0.0002196848354020507,
      "loss": 3.2931,
      "step": 58850
    },
    {
      "epoch": 6.350404312668464,
      "grad_norm": 0.7106407880783081,
      "learning_rate": 0.00021936103615758227,
      "loss": 3.2802,
      "step": 58900
    },
    {
      "epoch": 6.355795148247978,
      "grad_norm": 0.7471071481704712,
      "learning_rate": 0.00021903723691311385,
      "loss": 3.2769,
      "step": 58950
    },
    {
      "epoch": 6.361185983827493,
      "grad_norm": 0.7540058493614197,
      "learning_rate": 0.00021871343766864543,
      "loss": 3.2938,
      "step": 59000
    },
    {
      "epoch": 6.361185983827493,
      "eval_accuracy": 0.38488079854657203,
      "eval_loss": 3.3768014907836914,
      "eval_runtime": 183.5526,
      "eval_samples_per_second": 98.124,
      "eval_steps_per_second": 6.134,
      "step": 59000
    },
    {
      "epoch": 6.366576819407008,
      "grad_norm": 0.732081413269043,
      "learning_rate": 0.00021838963842417698,
      "loss": 3.2813,
      "step": 59050
    },
    {
      "epoch": 6.3719676549865225,
      "grad_norm": 0.7750065326690674,
      "learning_rate": 0.00021806583917970856,
      "loss": 3.2791,
      "step": 59100
    },
    {
      "epoch": 6.377358490566038,
      "grad_norm": 0.7145469188690186,
      "learning_rate": 0.00021774203993524014,
      "loss": 3.2749,
      "step": 59150
    },
    {
      "epoch": 6.382749326145553,
      "grad_norm": 0.7660088539123535,
      "learning_rate": 0.00021741824069077171,
      "loss": 3.2785,
      "step": 59200
    },
    {
      "epoch": 6.388140161725067,
      "grad_norm": 0.7172083854675293,
      "learning_rate": 0.00021709444144630326,
      "loss": 3.2812,
      "step": 59250
    },
    {
      "epoch": 6.393530997304582,
      "grad_norm": 0.6981208920478821,
      "learning_rate": 0.00021677064220183484,
      "loss": 3.2828,
      "step": 59300
    },
    {
      "epoch": 6.398921832884097,
      "grad_norm": 0.6857271790504456,
      "learning_rate": 0.00021644684295736642,
      "loss": 3.2783,
      "step": 59350
    },
    {
      "epoch": 6.404312668463612,
      "grad_norm": 0.7612074613571167,
      "learning_rate": 0.00021612304371289797,
      "loss": 3.3057,
      "step": 59400
    },
    {
      "epoch": 6.409703504043127,
      "grad_norm": 0.8293981552124023,
      "learning_rate": 0.00021579924446842955,
      "loss": 3.2896,
      "step": 59450
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.7098532319068909,
      "learning_rate": 0.00021547544522396113,
      "loss": 3.2764,
      "step": 59500
    },
    {
      "epoch": 6.420485175202156,
      "grad_norm": 0.7373313903808594,
      "learning_rate": 0.0002151516459794927,
      "loss": 3.2737,
      "step": 59550
    },
    {
      "epoch": 6.425876010781671,
      "grad_norm": 0.7213019132614136,
      "learning_rate": 0.00021482784673502428,
      "loss": 3.2958,
      "step": 59600
    },
    {
      "epoch": 6.431266846361186,
      "grad_norm": 0.7404071092605591,
      "learning_rate": 0.00021450404749055586,
      "loss": 3.2866,
      "step": 59650
    },
    {
      "epoch": 6.436657681940701,
      "grad_norm": 0.7444236278533936,
      "learning_rate": 0.00021418024824608738,
      "loss": 3.2949,
      "step": 59700
    },
    {
      "epoch": 6.442048517520216,
      "grad_norm": 0.7594828605651855,
      "learning_rate": 0.00021385644900161896,
      "loss": 3.2932,
      "step": 59750
    },
    {
      "epoch": 6.44743935309973,
      "grad_norm": 0.7362959980964661,
      "learning_rate": 0.00021353264975715054,
      "loss": 3.2844,
      "step": 59800
    },
    {
      "epoch": 6.452830188679245,
      "grad_norm": 0.7293346524238586,
      "learning_rate": 0.00021320885051268211,
      "loss": 3.2673,
      "step": 59850
    },
    {
      "epoch": 6.45822102425876,
      "grad_norm": 0.774370551109314,
      "learning_rate": 0.0002128850512682137,
      "loss": 3.2825,
      "step": 59900
    },
    {
      "epoch": 6.463611859838275,
      "grad_norm": 0.7451016902923584,
      "learning_rate": 0.00021256125202374527,
      "loss": 3.2725,
      "step": 59950
    },
    {
      "epoch": 6.46900269541779,
      "grad_norm": 0.7309680581092834,
      "learning_rate": 0.0002122439287641662,
      "loss": 3.2827,
      "step": 60000
    },
    {
      "epoch": 6.46900269541779,
      "eval_accuracy": 0.38543851381972377,
      "eval_loss": 3.3723459243774414,
      "eval_runtime": 183.9807,
      "eval_samples_per_second": 97.896,
      "eval_steps_per_second": 6.12,
      "step": 60000
    },
    {
      "epoch": 6.474393530997305,
      "grad_norm": 0.7711241841316223,
      "learning_rate": 0.00021192012951969775,
      "loss": 3.2794,
      "step": 60050
    },
    {
      "epoch": 6.479784366576819,
      "grad_norm": 0.7258850336074829,
      "learning_rate": 0.00021159633027522932,
      "loss": 3.2932,
      "step": 60100
    },
    {
      "epoch": 6.485175202156334,
      "grad_norm": 0.7217243909835815,
      "learning_rate": 0.0002112725310307609,
      "loss": 3.2709,
      "step": 60150
    },
    {
      "epoch": 6.490566037735849,
      "grad_norm": 0.714972734451294,
      "learning_rate": 0.00021094873178629248,
      "loss": 3.284,
      "step": 60200
    },
    {
      "epoch": 6.495956873315364,
      "grad_norm": 0.7145763635635376,
      "learning_rate": 0.00021062493254182406,
      "loss": 3.2794,
      "step": 60250
    },
    {
      "epoch": 6.501347708894879,
      "grad_norm": 0.719743013381958,
      "learning_rate": 0.00021030113329735563,
      "loss": 3.276,
      "step": 60300
    },
    {
      "epoch": 6.506738544474393,
      "grad_norm": 0.8085545897483826,
      "learning_rate": 0.0002099773340528872,
      "loss": 3.2608,
      "step": 60350
    },
    {
      "epoch": 6.512129380053908,
      "grad_norm": 0.7587844133377075,
      "learning_rate": 0.0002096535348084188,
      "loss": 3.2921,
      "step": 60400
    },
    {
      "epoch": 6.517520215633423,
      "grad_norm": 0.8297287821769714,
      "learning_rate": 0.0002093297355639503,
      "loss": 3.2801,
      "step": 60450
    },
    {
      "epoch": 6.5229110512129385,
      "grad_norm": 0.7191682457923889,
      "learning_rate": 0.0002090059363194819,
      "loss": 3.2757,
      "step": 60500
    },
    {
      "epoch": 6.528301886792453,
      "grad_norm": 0.701370358467102,
      "learning_rate": 0.00020868213707501347,
      "loss": 3.2991,
      "step": 60550
    },
    {
      "epoch": 6.533692722371968,
      "grad_norm": 0.7378591895103455,
      "learning_rate": 0.00020835833783054505,
      "loss": 3.2726,
      "step": 60600
    },
    {
      "epoch": 6.539083557951482,
      "grad_norm": 0.7391147613525391,
      "learning_rate": 0.00020803453858607662,
      "loss": 3.2876,
      "step": 60650
    },
    {
      "epoch": 6.544474393530997,
      "grad_norm": 0.6959494948387146,
      "learning_rate": 0.0002077107393416082,
      "loss": 3.2869,
      "step": 60700
    },
    {
      "epoch": 6.549865229110512,
      "grad_norm": 0.6961156725883484,
      "learning_rate": 0.00020738694009713975,
      "loss": 3.2903,
      "step": 60750
    },
    {
      "epoch": 6.555256064690027,
      "grad_norm": 0.819736123085022,
      "learning_rate": 0.00020706314085267133,
      "loss": 3.2861,
      "step": 60800
    },
    {
      "epoch": 6.560646900269542,
      "grad_norm": 0.7416958808898926,
      "learning_rate": 0.00020673934160820288,
      "loss": 3.2745,
      "step": 60850
    },
    {
      "epoch": 6.566037735849057,
      "grad_norm": 0.7390998005867004,
      "learning_rate": 0.00020641554236373446,
      "loss": 3.271,
      "step": 60900
    },
    {
      "epoch": 6.571428571428571,
      "grad_norm": 0.823673665523529,
      "learning_rate": 0.00020609174311926604,
      "loss": 3.2896,
      "step": 60950
    },
    {
      "epoch": 6.576819407008086,
      "grad_norm": 0.7763093113899231,
      "learning_rate": 0.0002057679438747976,
      "loss": 3.2757,
      "step": 61000
    },
    {
      "epoch": 6.576819407008086,
      "eval_accuracy": 0.38573676599677453,
      "eval_loss": 3.3671045303344727,
      "eval_runtime": 183.4205,
      "eval_samples_per_second": 98.195,
      "eval_steps_per_second": 6.139,
      "step": 61000
    },
    {
      "epoch": 6.5822102425876015,
      "grad_norm": 0.7161189913749695,
      "learning_rate": 0.00020544414463032916,
      "loss": 3.2724,
      "step": 61050
    },
    {
      "epoch": 6.587601078167116,
      "grad_norm": 0.7647314071655273,
      "learning_rate": 0.00020512034538586074,
      "loss": 3.2795,
      "step": 61100
    },
    {
      "epoch": 6.592991913746631,
      "grad_norm": 0.7199732065200806,
      "learning_rate": 0.00020479654614139232,
      "loss": 3.2953,
      "step": 61150
    },
    {
      "epoch": 6.598382749326145,
      "grad_norm": 0.707043468952179,
      "learning_rate": 0.0002044727468969239,
      "loss": 3.2936,
      "step": 61200
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.7686954140663147,
      "learning_rate": 0.00020414894765245547,
      "loss": 3.2855,
      "step": 61250
    },
    {
      "epoch": 6.609164420485175,
      "grad_norm": 0.7518014907836914,
      "learning_rate": 0.00020382514840798702,
      "loss": 3.2775,
      "step": 61300
    },
    {
      "epoch": 6.6145552560646905,
      "grad_norm": 0.8239266872406006,
      "learning_rate": 0.0002035013491635186,
      "loss": 3.2874,
      "step": 61350
    },
    {
      "epoch": 6.619946091644205,
      "grad_norm": 0.7640479207038879,
      "learning_rate": 0.00020317754991905015,
      "loss": 3.2888,
      "step": 61400
    },
    {
      "epoch": 6.62533692722372,
      "grad_norm": 0.764625072479248,
      "learning_rate": 0.00020285375067458173,
      "loss": 3.2869,
      "step": 61450
    },
    {
      "epoch": 6.630727762803234,
      "grad_norm": 0.7239562273025513,
      "learning_rate": 0.0002025299514301133,
      "loss": 3.2814,
      "step": 61500
    },
    {
      "epoch": 6.636118598382749,
      "grad_norm": 0.7507439255714417,
      "learning_rate": 0.00020220615218564489,
      "loss": 3.2802,
      "step": 61550
    },
    {
      "epoch": 6.6415094339622645,
      "grad_norm": 0.7298441529273987,
      "learning_rate": 0.00020188235294117646,
      "loss": 3.3071,
      "step": 61600
    },
    {
      "epoch": 6.646900269541779,
      "grad_norm": 0.719257116317749,
      "learning_rate": 0.00020155855369670804,
      "loss": 3.2638,
      "step": 61650
    },
    {
      "epoch": 6.652291105121294,
      "grad_norm": 0.7430747747421265,
      "learning_rate": 0.00020123475445223956,
      "loss": 3.2825,
      "step": 61700
    },
    {
      "epoch": 6.657681940700809,
      "grad_norm": 0.7874649167060852,
      "learning_rate": 0.00020091095520777114,
      "loss": 3.2985,
      "step": 61750
    },
    {
      "epoch": 6.663072776280323,
      "grad_norm": 0.7608209252357483,
      "learning_rate": 0.00020058715596330272,
      "loss": 3.2834,
      "step": 61800
    },
    {
      "epoch": 6.668463611859838,
      "grad_norm": 0.7211445569992065,
      "learning_rate": 0.0002002633567188343,
      "loss": 3.2955,
      "step": 61850
    },
    {
      "epoch": 6.6738544474393535,
      "grad_norm": 0.7505002617835999,
      "learning_rate": 0.00019993955747436588,
      "loss": 3.2848,
      "step": 61900
    },
    {
      "epoch": 6.679245283018868,
      "grad_norm": 0.7617312669754028,
      "learning_rate": 0.00019961575822989745,
      "loss": 3.2796,
      "step": 61950
    },
    {
      "epoch": 6.684636118598383,
      "grad_norm": 0.7387272119522095,
      "learning_rate": 0.00019929195898542903,
      "loss": 3.278,
      "step": 62000
    },
    {
      "epoch": 6.684636118598383,
      "eval_accuracy": 0.3858140182000106,
      "eval_loss": 3.3650400638580322,
      "eval_runtime": 183.6699,
      "eval_samples_per_second": 98.062,
      "eval_steps_per_second": 6.131,
      "step": 62000
    },
    {
      "epoch": 6.690026954177897,
      "grad_norm": 0.7572900652885437,
      "learning_rate": 0.0001989681597409606,
      "loss": 3.2769,
      "step": 62050
    },
    {
      "epoch": 6.695417789757412,
      "grad_norm": 0.7592546343803406,
      "learning_rate": 0.0001986508364813815,
      "loss": 3.2829,
      "step": 62100
    },
    {
      "epoch": 6.7008086253369274,
      "grad_norm": 0.7549013495445251,
      "learning_rate": 0.00019832703723691308,
      "loss": 3.2989,
      "step": 62150
    },
    {
      "epoch": 6.706199460916442,
      "grad_norm": 0.7903435230255127,
      "learning_rate": 0.00019800323799244466,
      "loss": 3.3063,
      "step": 62200
    },
    {
      "epoch": 6.711590296495957,
      "grad_norm": 0.791134774684906,
      "learning_rate": 0.00019767943874797624,
      "loss": 3.2809,
      "step": 62250
    },
    {
      "epoch": 6.716981132075472,
      "grad_norm": 0.7578250169754028,
      "learning_rate": 0.00019735563950350782,
      "loss": 3.2862,
      "step": 62300
    },
    {
      "epoch": 6.722371967654986,
      "grad_norm": 0.7541927695274353,
      "learning_rate": 0.0001970318402590394,
      "loss": 3.2864,
      "step": 62350
    },
    {
      "epoch": 6.727762803234501,
      "grad_norm": 0.7560279369354248,
      "learning_rate": 0.00019670804101457097,
      "loss": 3.2722,
      "step": 62400
    },
    {
      "epoch": 6.7331536388140165,
      "grad_norm": 0.7128151059150696,
      "learning_rate": 0.00019638424177010252,
      "loss": 3.2774,
      "step": 62450
    },
    {
      "epoch": 6.738544474393531,
      "grad_norm": 0.8017320036888123,
      "learning_rate": 0.00019606044252563407,
      "loss": 3.2861,
      "step": 62500
    },
    {
      "epoch": 6.743935309973046,
      "grad_norm": 0.7617440819740295,
      "learning_rate": 0.00019573664328116565,
      "loss": 3.2845,
      "step": 62550
    },
    {
      "epoch": 6.74932614555256,
      "grad_norm": 0.7816303968429565,
      "learning_rate": 0.00019541284403669723,
      "loss": 3.2797,
      "step": 62600
    },
    {
      "epoch": 6.754716981132075,
      "grad_norm": 0.7712652087211609,
      "learning_rate": 0.0001950890447922288,
      "loss": 3.2916,
      "step": 62650
    },
    {
      "epoch": 6.7601078167115904,
      "grad_norm": 0.7972909808158875,
      "learning_rate": 0.00019476524554776038,
      "loss": 3.2795,
      "step": 62700
    },
    {
      "epoch": 6.765498652291106,
      "grad_norm": 0.7326833605766296,
      "learning_rate": 0.00019444144630329193,
      "loss": 3.2975,
      "step": 62750
    },
    {
      "epoch": 6.77088948787062,
      "grad_norm": 0.7460118532180786,
      "learning_rate": 0.0001941176470588235,
      "loss": 3.2945,
      "step": 62800
    },
    {
      "epoch": 6.776280323450135,
      "grad_norm": 0.820135235786438,
      "learning_rate": 0.0001937938478143551,
      "loss": 3.2876,
      "step": 62850
    },
    {
      "epoch": 6.781671159029649,
      "grad_norm": 0.7615317702293396,
      "learning_rate": 0.00019347004856988664,
      "loss": 3.2822,
      "step": 62900
    },
    {
      "epoch": 6.787061994609164,
      "grad_norm": 0.7746335864067078,
      "learning_rate": 0.00019314624932541822,
      "loss": 3.283,
      "step": 62950
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.745581865310669,
      "learning_rate": 0.0001928224500809498,
      "loss": 3.2879,
      "step": 63000
    },
    {
      "epoch": 6.7924528301886795,
      "eval_accuracy": 0.38680895269372784,
      "eval_loss": 3.35954213142395,
      "eval_runtime": 183.6719,
      "eval_samples_per_second": 98.061,
      "eval_steps_per_second": 6.13,
      "step": 63000
    },
    {
      "epoch": 6.797843665768194,
      "grad_norm": 0.7740397453308105,
      "learning_rate": 0.00019249865083648137,
      "loss": 3.2726,
      "step": 63050
    },
    {
      "epoch": 6.803234501347709,
      "grad_norm": 0.8130731582641602,
      "learning_rate": 0.00019217485159201292,
      "loss": 3.2999,
      "step": 63100
    },
    {
      "epoch": 6.808625336927224,
      "grad_norm": 0.7525258660316467,
      "learning_rate": 0.0001918510523475445,
      "loss": 3.2857,
      "step": 63150
    },
    {
      "epoch": 6.814016172506738,
      "grad_norm": 0.7616598606109619,
      "learning_rate": 0.00019152725310307608,
      "loss": 3.2845,
      "step": 63200
    },
    {
      "epoch": 6.819407008086253,
      "grad_norm": 0.7608476281166077,
      "learning_rate": 0.00019120345385860766,
      "loss": 3.293,
      "step": 63250
    },
    {
      "epoch": 6.824797843665769,
      "grad_norm": 0.7839943766593933,
      "learning_rate": 0.00019087965461413923,
      "loss": 3.2917,
      "step": 63300
    },
    {
      "epoch": 6.830188679245283,
      "grad_norm": 0.7487995028495789,
      "learning_rate": 0.00019055585536967079,
      "loss": 3.2826,
      "step": 63350
    },
    {
      "epoch": 6.835579514824798,
      "grad_norm": 0.7642489671707153,
      "learning_rate": 0.00019023205612520234,
      "loss": 3.2767,
      "step": 63400
    },
    {
      "epoch": 6.840970350404312,
      "grad_norm": 0.7495631575584412,
      "learning_rate": 0.00018990825688073391,
      "loss": 3.2824,
      "step": 63450
    },
    {
      "epoch": 6.846361185983827,
      "grad_norm": 0.7581333518028259,
      "learning_rate": 0.0001895844576362655,
      "loss": 3.2842,
      "step": 63500
    },
    {
      "epoch": 6.8517520215633425,
      "grad_norm": 0.7547215819358826,
      "learning_rate": 0.00018926065839179707,
      "loss": 3.3006,
      "step": 63550
    },
    {
      "epoch": 6.857142857142857,
      "grad_norm": 0.821328341960907,
      "learning_rate": 0.00018893685914732865,
      "loss": 3.286,
      "step": 63600
    },
    {
      "epoch": 6.862533692722372,
      "grad_norm": 0.7756742238998413,
      "learning_rate": 0.00018861305990286022,
      "loss": 3.2891,
      "step": 63650
    },
    {
      "epoch": 6.867924528301887,
      "grad_norm": 0.7616881728172302,
      "learning_rate": 0.0001882892606583918,
      "loss": 3.2645,
      "step": 63700
    },
    {
      "epoch": 6.873315363881401,
      "grad_norm": 0.7382721900939941,
      "learning_rate": 0.00018796546141392333,
      "loss": 3.275,
      "step": 63750
    },
    {
      "epoch": 6.878706199460916,
      "grad_norm": 0.7397435307502747,
      "learning_rate": 0.0001876416621694549,
      "loss": 3.2884,
      "step": 63800
    },
    {
      "epoch": 6.884097035040432,
      "grad_norm": 0.7460421323776245,
      "learning_rate": 0.00018731786292498648,
      "loss": 3.2613,
      "step": 63850
    },
    {
      "epoch": 6.889487870619946,
      "grad_norm": 0.7353686690330505,
      "learning_rate": 0.00018699406368051806,
      "loss": 3.276,
      "step": 63900
    },
    {
      "epoch": 6.894878706199461,
      "grad_norm": 0.7506459355354309,
      "learning_rate": 0.00018667026443604964,
      "loss": 3.293,
      "step": 63950
    },
    {
      "epoch": 6.900269541778976,
      "grad_norm": 0.7485863566398621,
      "learning_rate": 0.00018634646519158121,
      "loss": 3.3019,
      "step": 64000
    },
    {
      "epoch": 6.900269541778976,
      "eval_accuracy": 0.3870750436159855,
      "eval_loss": 3.3544468879699707,
      "eval_runtime": 183.6704,
      "eval_samples_per_second": 98.062,
      "eval_steps_per_second": 6.131,
      "step": 64000
    },
    {
      "epoch": 6.90566037735849,
      "grad_norm": 0.7352957129478455,
      "learning_rate": 0.0001860226659471128,
      "loss": 3.2824,
      "step": 64050
    },
    {
      "epoch": 6.9110512129380055,
      "grad_norm": 0.7454565763473511,
      "learning_rate": 0.00018569886670264434,
      "loss": 3.2811,
      "step": 64100
    },
    {
      "epoch": 6.916442048517521,
      "grad_norm": 0.7280657291412354,
      "learning_rate": 0.00018537506745817592,
      "loss": 3.2829,
      "step": 64150
    },
    {
      "epoch": 6.921832884097035,
      "grad_norm": 0.7745518088340759,
      "learning_rate": 0.00018505774419859684,
      "loss": 3.3071,
      "step": 64200
    },
    {
      "epoch": 6.92722371967655,
      "grad_norm": 0.79837566614151,
      "learning_rate": 0.00018473394495412842,
      "loss": 3.2772,
      "step": 64250
    },
    {
      "epoch": 6.932614555256064,
      "grad_norm": 0.8125883936882019,
      "learning_rate": 0.00018441014570966,
      "loss": 3.2835,
      "step": 64300
    },
    {
      "epoch": 6.938005390835579,
      "grad_norm": 0.7624317407608032,
      "learning_rate": 0.00018408634646519158,
      "loss": 3.2743,
      "step": 64350
    },
    {
      "epoch": 6.943396226415095,
      "grad_norm": 0.7544397711753845,
      "learning_rate": 0.00018376254722072316,
      "loss": 3.2835,
      "step": 64400
    },
    {
      "epoch": 6.948787061994609,
      "grad_norm": 0.7494552731513977,
      "learning_rate": 0.0001834387479762547,
      "loss": 3.2593,
      "step": 64450
    },
    {
      "epoch": 6.954177897574124,
      "grad_norm": 0.7906284332275391,
      "learning_rate": 0.00018311494873178628,
      "loss": 3.3027,
      "step": 64500
    },
    {
      "epoch": 6.959568733153639,
      "grad_norm": 0.76076340675354,
      "learning_rate": 0.00018279114948731783,
      "loss": 3.2956,
      "step": 64550
    },
    {
      "epoch": 6.964959568733153,
      "grad_norm": 0.7786262035369873,
      "learning_rate": 0.0001824673502428494,
      "loss": 3.296,
      "step": 64600
    },
    {
      "epoch": 6.9703504043126685,
      "grad_norm": 0.8318013548851013,
      "learning_rate": 0.000182143550998381,
      "loss": 3.3069,
      "step": 64650
    },
    {
      "epoch": 6.975741239892184,
      "grad_norm": 0.7505510449409485,
      "learning_rate": 0.00018181975175391257,
      "loss": 3.3001,
      "step": 64700
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.7878348231315613,
      "learning_rate": 0.00018149595250944414,
      "loss": 3.2874,
      "step": 64750
    },
    {
      "epoch": 6.986522911051213,
      "grad_norm": 0.8053597807884216,
      "learning_rate": 0.0001811721532649757,
      "loss": 3.2644,
      "step": 64800
    },
    {
      "epoch": 6.991913746630727,
      "grad_norm": 0.7475190162658691,
      "learning_rate": 0.00018084835402050727,
      "loss": 3.2769,
      "step": 64850
    },
    {
      "epoch": 6.997304582210242,
      "grad_norm": 0.7410888671875,
      "learning_rate": 0.00018052455477603885,
      "loss": 3.2803,
      "step": 64900
    },
    {
      "epoch": 7.002695417789758,
      "grad_norm": 0.7842205762863159,
      "learning_rate": 0.0001802007555315704,
      "loss": 3.2315,
      "step": 64950
    },
    {
      "epoch": 7.008086253369272,
      "grad_norm": 0.8470344543457031,
      "learning_rate": 0.00017987695628710198,
      "loss": 3.2106,
      "step": 65000
    },
    {
      "epoch": 7.008086253369272,
      "eval_accuracy": 0.38683339959348606,
      "eval_loss": 3.3595986366271973,
      "eval_runtime": 183.7302,
      "eval_samples_per_second": 98.03,
      "eval_steps_per_second": 6.129,
      "step": 65000
    },
    {
      "epoch": 7.013477088948787,
      "grad_norm": 0.7936300039291382,
      "learning_rate": 0.00017955315704263356,
      "loss": 3.1962,
      "step": 65050
    },
    {
      "epoch": 7.018867924528302,
      "grad_norm": 0.8864548802375793,
      "learning_rate": 0.0001792293577981651,
      "loss": 3.2104,
      "step": 65100
    },
    {
      "epoch": 7.024258760107816,
      "grad_norm": 0.7770150303840637,
      "learning_rate": 0.00017890555855369668,
      "loss": 3.2211,
      "step": 65150
    },
    {
      "epoch": 7.0296495956873315,
      "grad_norm": 0.7711372375488281,
      "learning_rate": 0.00017858175930922826,
      "loss": 3.1993,
      "step": 65200
    },
    {
      "epoch": 7.035040431266847,
      "grad_norm": 0.7707984447479248,
      "learning_rate": 0.00017825796006475984,
      "loss": 3.1926,
      "step": 65250
    },
    {
      "epoch": 7.040431266846361,
      "grad_norm": 0.7653559446334839,
      "learning_rate": 0.00017793416082029142,
      "loss": 3.2114,
      "step": 65300
    },
    {
      "epoch": 7.045822102425876,
      "grad_norm": 0.7938640713691711,
      "learning_rate": 0.000177610361575823,
      "loss": 3.1917,
      "step": 65350
    },
    {
      "epoch": 7.051212938005391,
      "grad_norm": 0.7933249473571777,
      "learning_rate": 0.00017728656233135452,
      "loss": 3.2102,
      "step": 65400
    },
    {
      "epoch": 7.056603773584905,
      "grad_norm": 0.7714968323707581,
      "learning_rate": 0.0001769627630868861,
      "loss": 3.2091,
      "step": 65450
    },
    {
      "epoch": 7.061994609164421,
      "grad_norm": 0.7854500412940979,
      "learning_rate": 0.00017663896384241767,
      "loss": 3.2047,
      "step": 65500
    },
    {
      "epoch": 7.067385444743936,
      "grad_norm": 0.8066527247428894,
      "learning_rate": 0.00017631516459794925,
      "loss": 3.2168,
      "step": 65550
    },
    {
      "epoch": 7.07277628032345,
      "grad_norm": 0.7777254581451416,
      "learning_rate": 0.00017599136535348083,
      "loss": 3.2176,
      "step": 65600
    },
    {
      "epoch": 7.078167115902965,
      "grad_norm": 0.8117446899414062,
      "learning_rate": 0.0001756675661090124,
      "loss": 3.2035,
      "step": 65650
    },
    {
      "epoch": 7.083557951482479,
      "grad_norm": 0.7938959002494812,
      "learning_rate": 0.00017534376686454398,
      "loss": 3.2154,
      "step": 65700
    },
    {
      "epoch": 7.0889487870619945,
      "grad_norm": 0.7688341736793518,
      "learning_rate": 0.00017501996762007556,
      "loss": 3.2104,
      "step": 65750
    },
    {
      "epoch": 7.09433962264151,
      "grad_norm": 0.8066598176956177,
      "learning_rate": 0.00017469616837560709,
      "loss": 3.2143,
      "step": 65800
    },
    {
      "epoch": 7.099730458221024,
      "grad_norm": 0.7198970317840576,
      "learning_rate": 0.00017437236913113866,
      "loss": 3.1953,
      "step": 65850
    },
    {
      "epoch": 7.105121293800539,
      "grad_norm": 0.7910106778144836,
      "learning_rate": 0.00017404856988667024,
      "loss": 3.206,
      "step": 65900
    },
    {
      "epoch": 7.110512129380054,
      "grad_norm": 0.8067068457603455,
      "learning_rate": 0.00017372477064220182,
      "loss": 3.194,
      "step": 65950
    },
    {
      "epoch": 7.115902964959568,
      "grad_norm": 0.7943997383117676,
      "learning_rate": 0.0001734009713977334,
      "loss": 3.2226,
      "step": 66000
    },
    {
      "epoch": 7.115902964959568,
      "eval_accuracy": 0.38742360208009435,
      "eval_loss": 3.3600785732269287,
      "eval_runtime": 183.7791,
      "eval_samples_per_second": 98.004,
      "eval_steps_per_second": 6.127,
      "step": 66000
    },
    {
      "epoch": 7.121293800539084,
      "grad_norm": 0.7855022549629211,
      "learning_rate": 0.00017307717215326497,
      "loss": 3.211,
      "step": 66050
    },
    {
      "epoch": 7.126684636118599,
      "grad_norm": 0.8030198216438293,
      "learning_rate": 0.00017275337290879655,
      "loss": 3.2127,
      "step": 66100
    },
    {
      "epoch": 7.132075471698113,
      "grad_norm": 0.7985265851020813,
      "learning_rate": 0.0001724295736643281,
      "loss": 3.2214,
      "step": 66150
    },
    {
      "epoch": 7.137466307277628,
      "grad_norm": 0.779312014579773,
      "learning_rate": 0.00017210577441985968,
      "loss": 3.2065,
      "step": 66200
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 0.8055819272994995,
      "learning_rate": 0.00017178197517539123,
      "loss": 3.2262,
      "step": 66250
    },
    {
      "epoch": 7.1482479784366575,
      "grad_norm": 0.7920130491256714,
      "learning_rate": 0.0001714581759309228,
      "loss": 3.2101,
      "step": 66300
    },
    {
      "epoch": 7.153638814016173,
      "grad_norm": 0.786949098110199,
      "learning_rate": 0.00017113437668645439,
      "loss": 3.2134,
      "step": 66350
    },
    {
      "epoch": 7.159029649595688,
      "grad_norm": 0.7556127309799194,
      "learning_rate": 0.00017081057744198596,
      "loss": 3.2158,
      "step": 66400
    },
    {
      "epoch": 7.164420485175202,
      "grad_norm": 0.7686795592308044,
      "learning_rate": 0.00017048677819751751,
      "loss": 3.223,
      "step": 66450
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.7993310689926147,
      "learning_rate": 0.0001701629789530491,
      "loss": 3.2131,
      "step": 66500
    },
    {
      "epoch": 7.175202156334231,
      "grad_norm": 0.8089261651039124,
      "learning_rate": 0.00016983917970858067,
      "loss": 3.228,
      "step": 66550
    },
    {
      "epoch": 7.180592991913747,
      "grad_norm": 0.7769164443016052,
      "learning_rate": 0.00016951538046411225,
      "loss": 3.2136,
      "step": 66600
    },
    {
      "epoch": 7.185983827493262,
      "grad_norm": 0.7841304540634155,
      "learning_rate": 0.0001691915812196438,
      "loss": 3.2292,
      "step": 66650
    },
    {
      "epoch": 7.191374663072776,
      "grad_norm": 0.7953046560287476,
      "learning_rate": 0.00016886778197517538,
      "loss": 3.2204,
      "step": 66700
    },
    {
      "epoch": 7.196765498652291,
      "grad_norm": 0.7785905599594116,
      "learning_rate": 0.00016854398273070693,
      "loss": 3.2299,
      "step": 66750
    },
    {
      "epoch": 7.202156334231806,
      "grad_norm": 0.7831030488014221,
      "learning_rate": 0.0001682201834862385,
      "loss": 3.2033,
      "step": 66800
    },
    {
      "epoch": 7.2075471698113205,
      "grad_norm": 0.7967961430549622,
      "learning_rate": 0.00016789638424177008,
      "loss": 3.2196,
      "step": 66850
    },
    {
      "epoch": 7.212938005390836,
      "grad_norm": 0.786827564239502,
      "learning_rate": 0.00016757258499730166,
      "loss": 3.2386,
      "step": 66900
    },
    {
      "epoch": 7.218328840970351,
      "grad_norm": 0.8236501216888428,
      "learning_rate": 0.00016724878575283324,
      "loss": 3.2287,
      "step": 66950
    },
    {
      "epoch": 7.223719676549865,
      "grad_norm": 0.7607845067977905,
      "learning_rate": 0.00016692498650836481,
      "loss": 3.2306,
      "step": 67000
    },
    {
      "epoch": 7.223719676549865,
      "eval_accuracy": 0.38756930560265357,
      "eval_loss": 3.355984926223755,
      "eval_runtime": 183.5772,
      "eval_samples_per_second": 98.111,
      "eval_steps_per_second": 6.134,
      "step": 67000
    },
    {
      "epoch": 7.22911051212938,
      "grad_norm": 0.7747772932052612,
      "learning_rate": 0.0001666011872638964,
      "loss": 3.2265,
      "step": 67050
    },
    {
      "epoch": 7.234501347708895,
      "grad_norm": 0.8035948872566223,
      "learning_rate": 0.00016627738801942792,
      "loss": 3.2282,
      "step": 67100
    },
    {
      "epoch": 7.2398921832884096,
      "grad_norm": 0.771018922328949,
      "learning_rate": 0.0001659535887749595,
      "loss": 3.2286,
      "step": 67150
    },
    {
      "epoch": 7.245283018867925,
      "grad_norm": 0.8212960362434387,
      "learning_rate": 0.00016562978953049107,
      "loss": 3.2199,
      "step": 67200
    },
    {
      "epoch": 7.250673854447439,
      "grad_norm": 0.8093283772468567,
      "learning_rate": 0.00016530599028602265,
      "loss": 3.2192,
      "step": 67250
    },
    {
      "epoch": 7.256064690026954,
      "grad_norm": 0.8153406977653503,
      "learning_rate": 0.00016498219104155423,
      "loss": 3.2356,
      "step": 67300
    },
    {
      "epoch": 7.261455525606469,
      "grad_norm": 0.7982839345932007,
      "learning_rate": 0.0001646583917970858,
      "loss": 3.224,
      "step": 67350
    },
    {
      "epoch": 7.2668463611859835,
      "grad_norm": 0.8163326978683472,
      "learning_rate": 0.00016433459255261738,
      "loss": 3.2309,
      "step": 67400
    },
    {
      "epoch": 7.272237196765499,
      "grad_norm": 0.8001326322555542,
      "learning_rate": 0.00016401079330814896,
      "loss": 3.2408,
      "step": 67450
    },
    {
      "epoch": 7.277628032345014,
      "grad_norm": 0.8096315860748291,
      "learning_rate": 0.00016368699406368048,
      "loss": 3.2359,
      "step": 67500
    },
    {
      "epoch": 7.283018867924528,
      "grad_norm": 0.9220945239067078,
      "learning_rate": 0.00016336319481921206,
      "loss": 3.2263,
      "step": 67550
    },
    {
      "epoch": 7.288409703504043,
      "grad_norm": 0.7517732977867126,
      "learning_rate": 0.00016303939557474364,
      "loss": 3.2471,
      "step": 67600
    },
    {
      "epoch": 7.293800539083558,
      "grad_norm": 0.8864023089408875,
      "learning_rate": 0.00016271559633027522,
      "loss": 3.2311,
      "step": 67650
    },
    {
      "epoch": 7.2991913746630726,
      "grad_norm": 0.7982566952705383,
      "learning_rate": 0.0001623917970858068,
      "loss": 3.23,
      "step": 67700
    },
    {
      "epoch": 7.304582210242588,
      "grad_norm": 0.7961156964302063,
      "learning_rate": 0.00016206799784133837,
      "loss": 3.2334,
      "step": 67750
    },
    {
      "epoch": 7.309973045822103,
      "grad_norm": 0.7799623012542725,
      "learning_rate": 0.00016174419859686992,
      "loss": 3.2151,
      "step": 67800
    },
    {
      "epoch": 7.315363881401617,
      "grad_norm": 0.7993525266647339,
      "learning_rate": 0.0001614203993524015,
      "loss": 3.2525,
      "step": 67850
    },
    {
      "epoch": 7.320754716981132,
      "grad_norm": 0.7678369879722595,
      "learning_rate": 0.00016110307609282242,
      "loss": 3.2317,
      "step": 67900
    },
    {
      "epoch": 7.3261455525606465,
      "grad_norm": 0.8402950763702393,
      "learning_rate": 0.000160779276848354,
      "loss": 3.2364,
      "step": 67950
    },
    {
      "epoch": 7.331536388140162,
      "grad_norm": 0.8143036365509033,
      "learning_rate": 0.00016045547760388558,
      "loss": 3.2122,
      "step": 68000
    },
    {
      "epoch": 7.331536388140162,
      "eval_accuracy": 0.38808019148115724,
      "eval_loss": 3.353743553161621,
      "eval_runtime": 183.5805,
      "eval_samples_per_second": 98.11,
      "eval_steps_per_second": 6.134,
      "step": 68000
    },
    {
      "epoch": 7.336927223719677,
      "grad_norm": 0.7968747019767761,
      "learning_rate": 0.00016013167835941716,
      "loss": 3.2338,
      "step": 68050
    },
    {
      "epoch": 7.342318059299191,
      "grad_norm": 0.8042038679122925,
      "learning_rate": 0.00015980787911494873,
      "loss": 3.2394,
      "step": 68100
    },
    {
      "epoch": 7.347708894878706,
      "grad_norm": 0.8089545369148254,
      "learning_rate": 0.00015948407987048029,
      "loss": 3.2243,
      "step": 68150
    },
    {
      "epoch": 7.353099730458221,
      "grad_norm": 0.839443564414978,
      "learning_rate": 0.00015916028062601186,
      "loss": 3.2456,
      "step": 68200
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.8253589868545532,
      "learning_rate": 0.00015883648138154344,
      "loss": 3.2482,
      "step": 68250
    },
    {
      "epoch": 7.363881401617251,
      "grad_norm": 0.805784285068512,
      "learning_rate": 0.000158512682137075,
      "loss": 3.2355,
      "step": 68300
    },
    {
      "epoch": 7.369272237196766,
      "grad_norm": 0.8239293098449707,
      "learning_rate": 0.00015818888289260657,
      "loss": 3.2337,
      "step": 68350
    },
    {
      "epoch": 7.37466307277628,
      "grad_norm": 0.8073683381080627,
      "learning_rate": 0.00015786508364813815,
      "loss": 3.2261,
      "step": 68400
    },
    {
      "epoch": 7.380053908355795,
      "grad_norm": 0.871188759803772,
      "learning_rate": 0.0001575412844036697,
      "loss": 3.2112,
      "step": 68450
    },
    {
      "epoch": 7.38544474393531,
      "grad_norm": 0.8160022497177124,
      "learning_rate": 0.00015721748515920127,
      "loss": 3.2173,
      "step": 68500
    },
    {
      "epoch": 7.390835579514825,
      "grad_norm": 0.7948161959648132,
      "learning_rate": 0.00015689368591473285,
      "loss": 3.2271,
      "step": 68550
    },
    {
      "epoch": 7.39622641509434,
      "grad_norm": 0.7891157269477844,
      "learning_rate": 0.00015656988667026443,
      "loss": 3.2428,
      "step": 68600
    },
    {
      "epoch": 7.401617250673855,
      "grad_norm": 0.824519157409668,
      "learning_rate": 0.000156246087425796,
      "loss": 3.2153,
      "step": 68650
    },
    {
      "epoch": 7.407008086253369,
      "grad_norm": 0.8005858659744263,
      "learning_rate": 0.00015592228818132756,
      "loss": 3.2209,
      "step": 68700
    },
    {
      "epoch": 7.412398921832884,
      "grad_norm": 0.8065062761306763,
      "learning_rate": 0.0001555984889368591,
      "loss": 3.2241,
      "step": 68750
    },
    {
      "epoch": 7.4177897574123985,
      "grad_norm": 0.8123283386230469,
      "learning_rate": 0.0001552746896923907,
      "loss": 3.235,
      "step": 68800
    },
    {
      "epoch": 7.423180592991914,
      "grad_norm": 0.7728703022003174,
      "learning_rate": 0.00015495089044792226,
      "loss": 3.2198,
      "step": 68850
    },
    {
      "epoch": 7.428571428571429,
      "grad_norm": 0.7691994905471802,
      "learning_rate": 0.00015462709120345384,
      "loss": 3.2346,
      "step": 68900
    },
    {
      "epoch": 7.433962264150943,
      "grad_norm": 0.835544228553772,
      "learning_rate": 0.00015430329195898542,
      "loss": 3.2275,
      "step": 68950
    },
    {
      "epoch": 7.439353099730458,
      "grad_norm": 0.7981140613555908,
      "learning_rate": 0.000153979492714517,
      "loss": 3.2476,
      "step": 69000
    },
    {
      "epoch": 7.439353099730458,
      "eval_accuracy": 0.3881780877330781,
      "eval_loss": 3.350919723510742,
      "eval_runtime": 183.8391,
      "eval_samples_per_second": 97.972,
      "eval_steps_per_second": 6.125,
      "step": 69000
    },
    {
      "epoch": 7.444743935309973,
      "grad_norm": 0.804359495639801,
      "learning_rate": 0.00015365569347004858,
      "loss": 3.2264,
      "step": 69050
    },
    {
      "epoch": 7.450134770889488,
      "grad_norm": 0.7978847026824951,
      "learning_rate": 0.00015333189422558015,
      "loss": 3.2213,
      "step": 69100
    },
    {
      "epoch": 7.455525606469003,
      "grad_norm": 0.8682163953781128,
      "learning_rate": 0.00015300809498111168,
      "loss": 3.2043,
      "step": 69150
    },
    {
      "epoch": 7.460916442048518,
      "grad_norm": 0.7981349229812622,
      "learning_rate": 0.00015268429573664325,
      "loss": 3.2463,
      "step": 69200
    },
    {
      "epoch": 7.466307277628032,
      "grad_norm": 0.8066042065620422,
      "learning_rate": 0.00015236049649217483,
      "loss": 3.2327,
      "step": 69250
    },
    {
      "epoch": 7.471698113207547,
      "grad_norm": 0.8053681254386902,
      "learning_rate": 0.0001520366972477064,
      "loss": 3.2211,
      "step": 69300
    },
    {
      "epoch": 7.4770889487870615,
      "grad_norm": 0.8309054374694824,
      "learning_rate": 0.000151712898003238,
      "loss": 3.2419,
      "step": 69350
    },
    {
      "epoch": 7.482479784366577,
      "grad_norm": 0.8017661571502686,
      "learning_rate": 0.00015138909875876956,
      "loss": 3.243,
      "step": 69400
    },
    {
      "epoch": 7.487870619946092,
      "grad_norm": 0.7868233919143677,
      "learning_rate": 0.00015106529951430114,
      "loss": 3.2369,
      "step": 69450
    },
    {
      "epoch": 7.493261455525606,
      "grad_norm": 0.8187854290008545,
      "learning_rate": 0.0001507415002698327,
      "loss": 3.2252,
      "step": 69500
    },
    {
      "epoch": 7.498652291105121,
      "grad_norm": 0.841691255569458,
      "learning_rate": 0.00015041770102536424,
      "loss": 3.2313,
      "step": 69550
    },
    {
      "epoch": 7.504043126684636,
      "grad_norm": 0.8016337752342224,
      "learning_rate": 0.00015009390178089582,
      "loss": 3.2255,
      "step": 69600
    },
    {
      "epoch": 7.509433962264151,
      "grad_norm": 0.7767087817192078,
      "learning_rate": 0.0001497701025364274,
      "loss": 3.2288,
      "step": 69650
    },
    {
      "epoch": 7.514824797843666,
      "grad_norm": 0.8359373211860657,
      "learning_rate": 0.00014944630329195898,
      "loss": 3.2401,
      "step": 69700
    },
    {
      "epoch": 7.520215633423181,
      "grad_norm": 0.818621039390564,
      "learning_rate": 0.00014912250404749055,
      "loss": 3.2298,
      "step": 69750
    },
    {
      "epoch": 7.525606469002695,
      "grad_norm": 0.8097342848777771,
      "learning_rate": 0.0001487987048030221,
      "loss": 3.2457,
      "step": 69800
    },
    {
      "epoch": 7.53099730458221,
      "grad_norm": 0.7905571460723877,
      "learning_rate": 0.00014847490555855368,
      "loss": 3.2211,
      "step": 69850
    },
    {
      "epoch": 7.536388140161725,
      "grad_norm": 0.7882384061813354,
      "learning_rate": 0.00014815110631408526,
      "loss": 3.2294,
      "step": 69900
    },
    {
      "epoch": 7.54177897574124,
      "grad_norm": 0.8214282393455505,
      "learning_rate": 0.0001478273070696168,
      "loss": 3.2369,
      "step": 69950
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.8215126395225525,
      "learning_rate": 0.0001475035078251484,
      "loss": 3.2342,
      "step": 70000
    },
    {
      "epoch": 7.547169811320755,
      "eval_accuracy": 0.3886583334972181,
      "eval_loss": 3.3478920459747314,
      "eval_runtime": 183.3704,
      "eval_samples_per_second": 98.222,
      "eval_steps_per_second": 6.141,
      "step": 70000
    },
    {
      "epoch": 7.55256064690027,
      "grad_norm": 0.8579515218734741,
      "learning_rate": 0.00014717970858067997,
      "loss": 3.215,
      "step": 70050
    },
    {
      "epoch": 7.557951482479784,
      "grad_norm": 0.7795184850692749,
      "learning_rate": 0.00014685590933621154,
      "loss": 3.2367,
      "step": 70100
    },
    {
      "epoch": 7.563342318059299,
      "grad_norm": 0.8258717060089111,
      "learning_rate": 0.0001465321100917431,
      "loss": 3.2346,
      "step": 70150
    },
    {
      "epoch": 7.568733153638814,
      "grad_norm": 0.778454065322876,
      "learning_rate": 0.00014620831084727467,
      "loss": 3.2334,
      "step": 70200
    },
    {
      "epoch": 7.574123989218329,
      "grad_norm": 0.7907705307006836,
      "learning_rate": 0.00014588451160280625,
      "loss": 3.2301,
      "step": 70250
    },
    {
      "epoch": 7.579514824797844,
      "grad_norm": 0.8008063435554504,
      "learning_rate": 0.0001455607123583378,
      "loss": 3.2434,
      "step": 70300
    },
    {
      "epoch": 7.584905660377358,
      "grad_norm": 0.7994003891944885,
      "learning_rate": 0.00014523691311386938,
      "loss": 3.2157,
      "step": 70350
    },
    {
      "epoch": 7.590296495956873,
      "grad_norm": 0.8054956197738647,
      "learning_rate": 0.00014491311386940096,
      "loss": 3.2219,
      "step": 70400
    },
    {
      "epoch": 7.595687331536388,
      "grad_norm": 0.7711500525474548,
      "learning_rate": 0.00014458931462493253,
      "loss": 3.2291,
      "step": 70450
    },
    {
      "epoch": 7.601078167115903,
      "grad_norm": 0.7859018445014954,
      "learning_rate": 0.00014426551538046408,
      "loss": 3.2422,
      "step": 70500
    },
    {
      "epoch": 7.606469002695418,
      "grad_norm": 0.8233794569969177,
      "learning_rate": 0.00014394171613599566,
      "loss": 3.2292,
      "step": 70550
    },
    {
      "epoch": 7.611859838274933,
      "grad_norm": 0.8168668746948242,
      "learning_rate": 0.0001436243928764166,
      "loss": 3.254,
      "step": 70600
    },
    {
      "epoch": 7.617250673854447,
      "grad_norm": 0.8281172513961792,
      "learning_rate": 0.00014330059363194816,
      "loss": 3.2292,
      "step": 70650
    },
    {
      "epoch": 7.622641509433962,
      "grad_norm": 0.824066698551178,
      "learning_rate": 0.00014297679438747974,
      "loss": 3.2311,
      "step": 70700
    },
    {
      "epoch": 7.628032345013477,
      "grad_norm": 0.8089962005615234,
      "learning_rate": 0.00014265299514301132,
      "loss": 3.2236,
      "step": 70750
    },
    {
      "epoch": 7.633423180592992,
      "grad_norm": 0.7989951372146606,
      "learning_rate": 0.0001423291958985429,
      "loss": 3.227,
      "step": 70800
    },
    {
      "epoch": 7.638814016172507,
      "grad_norm": 0.8355391621589661,
      "learning_rate": 0.00014200539665407445,
      "loss": 3.2313,
      "step": 70850
    },
    {
      "epoch": 7.644204851752022,
      "grad_norm": 0.8052586317062378,
      "learning_rate": 0.00014168159740960602,
      "loss": 3.2264,
      "step": 70900
    },
    {
      "epoch": 7.649595687331536,
      "grad_norm": 0.8242946863174438,
      "learning_rate": 0.0001413577981651376,
      "loss": 3.2342,
      "step": 70950
    },
    {
      "epoch": 7.654986522911051,
      "grad_norm": 0.829434335231781,
      "learning_rate": 0.00014103399892066918,
      "loss": 3.2355,
      "step": 71000
    },
    {
      "epoch": 7.654986522911051,
      "eval_accuracy": 0.3890025458458144,
      "eval_loss": 3.3414785861968994,
      "eval_runtime": 183.6585,
      "eval_samples_per_second": 98.068,
      "eval_steps_per_second": 6.131,
      "step": 71000
    },
    {
      "epoch": 7.660377358490566,
      "grad_norm": 0.7990546226501465,
      "learning_rate": 0.00014071019967620073,
      "loss": 3.2366,
      "step": 71050
    },
    {
      "epoch": 7.665768194070081,
      "grad_norm": 0.807453453540802,
      "learning_rate": 0.0001403864004317323,
      "loss": 3.2209,
      "step": 71100
    },
    {
      "epoch": 7.671159029649596,
      "grad_norm": 0.8239540457725525,
      "learning_rate": 0.00014006260118726389,
      "loss": 3.2114,
      "step": 71150
    },
    {
      "epoch": 7.67654986522911,
      "grad_norm": 0.7842625975608826,
      "learning_rate": 0.00013973880194279546,
      "loss": 3.22,
      "step": 71200
    },
    {
      "epoch": 7.681940700808625,
      "grad_norm": 0.7766878008842468,
      "learning_rate": 0.00013941500269832704,
      "loss": 3.2296,
      "step": 71250
    },
    {
      "epoch": 7.6873315363881405,
      "grad_norm": 0.7896228432655334,
      "learning_rate": 0.0001390912034538586,
      "loss": 3.2524,
      "step": 71300
    },
    {
      "epoch": 7.692722371967655,
      "grad_norm": 0.7728307247161865,
      "learning_rate": 0.00013876740420939017,
      "loss": 3.2405,
      "step": 71350
    },
    {
      "epoch": 7.69811320754717,
      "grad_norm": 0.8142659664154053,
      "learning_rate": 0.00013844360496492175,
      "loss": 3.2361,
      "step": 71400
    },
    {
      "epoch": 7.703504043126685,
      "grad_norm": 0.8092532753944397,
      "learning_rate": 0.00013811980572045333,
      "loss": 3.224,
      "step": 71450
    },
    {
      "epoch": 7.708894878706199,
      "grad_norm": 0.8074396252632141,
      "learning_rate": 0.00013779600647598488,
      "loss": 3.227,
      "step": 71500
    },
    {
      "epoch": 7.714285714285714,
      "grad_norm": 0.8047695159912109,
      "learning_rate": 0.00013747220723151645,
      "loss": 3.2548,
      "step": 71550
    },
    {
      "epoch": 7.719676549865229,
      "grad_norm": 0.8205470442771912,
      "learning_rate": 0.00013714840798704803,
      "loss": 3.2417,
      "step": 71600
    },
    {
      "epoch": 7.725067385444744,
      "grad_norm": 0.8592079281806946,
      "learning_rate": 0.00013682460874257958,
      "loss": 3.2491,
      "step": 71650
    },
    {
      "epoch": 7.730458221024259,
      "grad_norm": 0.8482884168624878,
      "learning_rate": 0.00013650080949811116,
      "loss": 3.2452,
      "step": 71700
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.820476233959198,
      "learning_rate": 0.00013617701025364274,
      "loss": 3.2502,
      "step": 71750
    },
    {
      "epoch": 7.741239892183288,
      "grad_norm": 0.8236547112464905,
      "learning_rate": 0.0001358532110091743,
      "loss": 3.2414,
      "step": 71800
    },
    {
      "epoch": 7.7466307277628035,
      "grad_norm": 0.8307732343673706,
      "learning_rate": 0.00013552941176470587,
      "loss": 3.2554,
      "step": 71850
    },
    {
      "epoch": 7.752021563342318,
      "grad_norm": 0.7903879284858704,
      "learning_rate": 0.00013520561252023744,
      "loss": 3.2328,
      "step": 71900
    },
    {
      "epoch": 7.757412398921833,
      "grad_norm": 0.8310343027114868,
      "learning_rate": 0.00013488181327576902,
      "loss": 3.2614,
      "step": 71950
    },
    {
      "epoch": 7.762803234501348,
      "grad_norm": 0.8106740117073059,
      "learning_rate": 0.00013455801403130057,
      "loss": 3.2304,
      "step": 72000
    },
    {
      "epoch": 7.762803234501348,
      "eval_accuracy": 0.3895956819603936,
      "eval_loss": 3.340195417404175,
      "eval_runtime": 183.5574,
      "eval_samples_per_second": 98.122,
      "eval_steps_per_second": 6.134,
      "step": 72000
    },
    {
      "epoch": 7.768194070080862,
      "grad_norm": 0.749928891658783,
      "learning_rate": 0.00013423421478683215,
      "loss": 3.2317,
      "step": 72050
    },
    {
      "epoch": 7.773584905660377,
      "grad_norm": 0.8262786269187927,
      "learning_rate": 0.00013391041554236373,
      "loss": 3.22,
      "step": 72100
    },
    {
      "epoch": 7.7789757412398925,
      "grad_norm": 0.79350745677948,
      "learning_rate": 0.00013358661629789528,
      "loss": 3.234,
      "step": 72150
    },
    {
      "epoch": 7.784366576819407,
      "grad_norm": 0.8237420320510864,
      "learning_rate": 0.00013326281705342685,
      "loss": 3.2357,
      "step": 72200
    },
    {
      "epoch": 7.789757412398922,
      "grad_norm": 0.806442141532898,
      "learning_rate": 0.00013293901780895843,
      "loss": 3.2334,
      "step": 72250
    },
    {
      "epoch": 7.795148247978437,
      "grad_norm": 0.8202387690544128,
      "learning_rate": 0.00013261521856449,
      "loss": 3.2303,
      "step": 72300
    },
    {
      "epoch": 7.800539083557951,
      "grad_norm": 0.8479371070861816,
      "learning_rate": 0.00013229141932002156,
      "loss": 3.2328,
      "step": 72350
    },
    {
      "epoch": 7.8059299191374665,
      "grad_norm": 0.8242699503898621,
      "learning_rate": 0.00013196762007555314,
      "loss": 3.2365,
      "step": 72400
    },
    {
      "epoch": 7.811320754716981,
      "grad_norm": 0.882240355014801,
      "learning_rate": 0.00013164382083108472,
      "loss": 3.2313,
      "step": 72450
    },
    {
      "epoch": 7.816711590296496,
      "grad_norm": 0.8102225661277771,
      "learning_rate": 0.0001313200215866163,
      "loss": 3.239,
      "step": 72500
    },
    {
      "epoch": 7.822102425876011,
      "grad_norm": 0.8708174824714661,
      "learning_rate": 0.00013099622234214784,
      "loss": 3.2317,
      "step": 72550
    },
    {
      "epoch": 7.827493261455525,
      "grad_norm": 0.8136692643165588,
      "learning_rate": 0.00013067242309767942,
      "loss": 3.2421,
      "step": 72600
    },
    {
      "epoch": 7.83288409703504,
      "grad_norm": 0.7785462737083435,
      "learning_rate": 0.000130348623853211,
      "loss": 3.2495,
      "step": 72650
    },
    {
      "epoch": 7.8382749326145555,
      "grad_norm": 0.8305777907371521,
      "learning_rate": 0.00013002482460874258,
      "loss": 3.2455,
      "step": 72700
    },
    {
      "epoch": 7.84366576819407,
      "grad_norm": Infinity,
      "learning_rate": 0.0001297075013491635,
      "loss": 3.2382,
      "step": 72750
    },
    {
      "epoch": 7.849056603773585,
      "grad_norm": 0.8630931377410889,
      "learning_rate": 0.00012938370210469508,
      "loss": 3.2622,
      "step": 72800
    },
    {
      "epoch": 7.8544474393531,
      "grad_norm": 0.817644476890564,
      "learning_rate": 0.00012905990286022666,
      "loss": 3.2318,
      "step": 72850
    },
    {
      "epoch": 7.859838274932614,
      "grad_norm": 0.8427354693412781,
      "learning_rate": 0.0001287361036157582,
      "loss": 3.2369,
      "step": 72900
    },
    {
      "epoch": 7.8652291105121295,
      "grad_norm": 0.8348943591117859,
      "learning_rate": 0.00012841230437128979,
      "loss": 3.2545,
      "step": 72950
    },
    {
      "epoch": 7.870619946091644,
      "grad_norm": 0.8035464882850647,
      "learning_rate": 0.00012808850512682136,
      "loss": 3.2316,
      "step": 73000
    },
    {
      "epoch": 7.870619946091644,
      "eval_accuracy": 0.38966630633747307,
      "eval_loss": 3.3341665267944336,
      "eval_runtime": 183.9628,
      "eval_samples_per_second": 97.906,
      "eval_steps_per_second": 6.121,
      "step": 73000
    },
    {
      "epoch": 7.876010781671159,
      "grad_norm": 0.8106042146682739,
      "learning_rate": 0.00012776470588235294,
      "loss": 3.2295,
      "step": 73050
    },
    {
      "epoch": 7.881401617250674,
      "grad_norm": 0.8646166920661926,
      "learning_rate": 0.0001274409066378845,
      "loss": 3.2397,
      "step": 73100
    },
    {
      "epoch": 7.886792452830189,
      "grad_norm": 0.8176259398460388,
      "learning_rate": 0.00012711710739341607,
      "loss": 3.2437,
      "step": 73150
    },
    {
      "epoch": 7.892183288409703,
      "grad_norm": 0.7888458967208862,
      "learning_rate": 0.00012679330814894765,
      "loss": 3.2228,
      "step": 73200
    },
    {
      "epoch": 7.8975741239892185,
      "grad_norm": 0.8205003142356873,
      "learning_rate": 0.00012646950890447922,
      "loss": 3.2399,
      "step": 73250
    },
    {
      "epoch": 7.902964959568733,
      "grad_norm": 0.8603438138961792,
      "learning_rate": 0.00012614570966001077,
      "loss": 3.24,
      "step": 73300
    },
    {
      "epoch": 7.908355795148248,
      "grad_norm": 0.8492730259895325,
      "learning_rate": 0.00012582191041554235,
      "loss": 3.2435,
      "step": 73350
    },
    {
      "epoch": 7.913746630727763,
      "grad_norm": 0.8674182295799255,
      "learning_rate": 0.00012549811117107393,
      "loss": 3.2483,
      "step": 73400
    },
    {
      "epoch": 7.919137466307277,
      "grad_norm": 0.8385972380638123,
      "learning_rate": 0.0001251743119266055,
      "loss": 3.2268,
      "step": 73450
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.8631293177604675,
      "learning_rate": 0.00012485051268213706,
      "loss": 3.2479,
      "step": 73500
    },
    {
      "epoch": 7.929919137466308,
      "grad_norm": 0.8175431489944458,
      "learning_rate": 0.00012452671343766864,
      "loss": 3.2322,
      "step": 73550
    },
    {
      "epoch": 7.935309973045822,
      "grad_norm": 0.8723227381706238,
      "learning_rate": 0.00012420291419320021,
      "loss": 3.2427,
      "step": 73600
    },
    {
      "epoch": 7.940700808625337,
      "grad_norm": 0.8406923413276672,
      "learning_rate": 0.0001238791149487318,
      "loss": 3.2237,
      "step": 73650
    },
    {
      "epoch": 7.946091644204852,
      "grad_norm": 0.8450952172279358,
      "learning_rate": 0.00012355531570426334,
      "loss": 3.228,
      "step": 73700
    },
    {
      "epoch": 7.951482479784366,
      "grad_norm": 0.7840113043785095,
      "learning_rate": 0.00012323151645979492,
      "loss": 3.221,
      "step": 73750
    },
    {
      "epoch": 7.9568733153638815,
      "grad_norm": 0.8030226230621338,
      "learning_rate": 0.0001229077172153265,
      "loss": 3.2,
      "step": 73800
    },
    {
      "epoch": 7.962264150943396,
      "grad_norm": 0.8302611112594604,
      "learning_rate": 0.00012258391797085805,
      "loss": 3.2252,
      "step": 73850
    },
    {
      "epoch": 7.967654986522911,
      "grad_norm": 0.8094905614852905,
      "learning_rate": 0.00012226011872638963,
      "loss": 3.2324,
      "step": 73900
    },
    {
      "epoch": 7.973045822102426,
      "grad_norm": 0.8156700730323792,
      "learning_rate": 0.00012193631948192119,
      "loss": 3.2545,
      "step": 73950
    },
    {
      "epoch": 7.97843665768194,
      "grad_norm": 0.810791552066803,
      "learning_rate": 0.00012161252023745277,
      "loss": 3.2465,
      "step": 74000
    },
    {
      "epoch": 7.97843665768194,
      "eval_accuracy": 0.3901558962499651,
      "eval_loss": 3.331369400024414,
      "eval_runtime": 183.6514,
      "eval_samples_per_second": 98.072,
      "eval_steps_per_second": 6.131,
      "step": 74000
    },
    {
      "epoch": 7.9838274932614555,
      "grad_norm": 0.8125215172767639,
      "learning_rate": 0.00012128872099298435,
      "loss": 3.239,
      "step": 74050
    },
    {
      "epoch": 7.989218328840971,
      "grad_norm": 0.8124715685844421,
      "learning_rate": 0.00012096492174851591,
      "loss": 3.2293,
      "step": 74100
    },
    {
      "epoch": 7.994609164420485,
      "grad_norm": 0.9727672934532166,
      "learning_rate": 0.00012064112250404749,
      "loss": 3.2225,
      "step": 74150
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.8616753816604614,
      "learning_rate": 0.00012031732325957905,
      "loss": 3.2173,
      "step": 74200
    },
    {
      "epoch": 8.005390835579515,
      "grad_norm": 0.8144136667251587,
      "learning_rate": 0.00011999352401511062,
      "loss": 3.1608,
      "step": 74250
    },
    {
      "epoch": 8.01078167115903,
      "grad_norm": 0.8414373397827148,
      "learning_rate": 0.00011966972477064219,
      "loss": 3.1587,
      "step": 74300
    },
    {
      "epoch": 8.016172506738544,
      "grad_norm": 0.8705826997756958,
      "learning_rate": 0.00011934592552617377,
      "loss": 3.1476,
      "step": 74350
    },
    {
      "epoch": 8.021563342318059,
      "grad_norm": 0.7987647652626038,
      "learning_rate": 0.00011902212628170532,
      "loss": 3.1579,
      "step": 74400
    },
    {
      "epoch": 8.026954177897574,
      "grad_norm": 0.8605558276176453,
      "learning_rate": 0.0001186983270372369,
      "loss": 3.1628,
      "step": 74450
    },
    {
      "epoch": 8.032345013477089,
      "grad_norm": 0.8623753786087036,
      "learning_rate": 0.00011837452779276848,
      "loss": 3.1631,
      "step": 74500
    },
    {
      "epoch": 8.037735849056604,
      "grad_norm": 0.831736147403717,
      "learning_rate": 0.00011805072854830005,
      "loss": 3.1766,
      "step": 74550
    },
    {
      "epoch": 8.04312668463612,
      "grad_norm": 0.7906613945960999,
      "learning_rate": 0.0001177269293038316,
      "loss": 3.1654,
      "step": 74600
    },
    {
      "epoch": 8.048517520215633,
      "grad_norm": 0.8411964178085327,
      "learning_rate": 0.00011740313005936318,
      "loss": 3.1646,
      "step": 74650
    },
    {
      "epoch": 8.053908355795148,
      "grad_norm": 0.8155192732810974,
      "learning_rate": 0.00011707933081489476,
      "loss": 3.1527,
      "step": 74700
    },
    {
      "epoch": 8.059299191374663,
      "grad_norm": 0.830322802066803,
      "learning_rate": 0.00011675553157042632,
      "loss": 3.1701,
      "step": 74750
    },
    {
      "epoch": 8.064690026954178,
      "grad_norm": 0.8275706171989441,
      "learning_rate": 0.00011643173232595789,
      "loss": 3.1642,
      "step": 74800
    },
    {
      "epoch": 8.070080862533693,
      "grad_norm": 0.884319007396698,
      "learning_rate": 0.00011610793308148947,
      "loss": 3.1643,
      "step": 74850
    },
    {
      "epoch": 8.075471698113208,
      "grad_norm": 0.8118297457695007,
      "learning_rate": 0.00011578413383702104,
      "loss": 3.1777,
      "step": 74900
    },
    {
      "epoch": 8.080862533692722,
      "grad_norm": 0.8348281979560852,
      "learning_rate": 0.00011546033459255261,
      "loss": 3.1559,
      "step": 74950
    },
    {
      "epoch": 8.086253369272237,
      "grad_norm": 0.8485154509544373,
      "learning_rate": 0.00011513653534808419,
      "loss": 3.1592,
      "step": 75000
    },
    {
      "epoch": 8.086253369272237,
      "eval_accuracy": 0.38975431517660275,
      "eval_loss": 3.3387372493743896,
      "eval_runtime": 183.7567,
      "eval_samples_per_second": 98.015,
      "eval_steps_per_second": 6.128,
      "step": 75000
    },
    {
      "epoch": 8.091644204851752,
      "grad_norm": 0.8829059600830078,
      "learning_rate": 0.00011481273610361575,
      "loss": 3.1685,
      "step": 75050
    },
    {
      "epoch": 8.097035040431267,
      "grad_norm": 0.8124864101409912,
      "learning_rate": 0.00011448893685914731,
      "loss": 3.1565,
      "step": 75100
    },
    {
      "epoch": 8.102425876010782,
      "grad_norm": 0.8549532294273376,
      "learning_rate": 0.00011416513761467889,
      "loss": 3.1708,
      "step": 75150
    },
    {
      "epoch": 8.107816711590296,
      "grad_norm": 0.8692648410797119,
      "learning_rate": 0.00011384133837021047,
      "loss": 3.1843,
      "step": 75200
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.8498212695121765,
      "learning_rate": 0.00011351753912574202,
      "loss": 3.1675,
      "step": 75250
    },
    {
      "epoch": 8.118598382749326,
      "grad_norm": 0.7981224656105042,
      "learning_rate": 0.0001131937398812736,
      "loss": 3.1784,
      "step": 75300
    },
    {
      "epoch": 8.123989218328841,
      "grad_norm": 0.8637334704399109,
      "learning_rate": 0.00011286994063680517,
      "loss": 3.1798,
      "step": 75350
    },
    {
      "epoch": 8.129380053908356,
      "grad_norm": 0.8062846660614014,
      "learning_rate": 0.00011254614139233675,
      "loss": 3.1836,
      "step": 75400
    },
    {
      "epoch": 8.134770889487871,
      "grad_norm": 0.8146725296974182,
      "learning_rate": 0.0001122223421478683,
      "loss": 3.172,
      "step": 75450
    },
    {
      "epoch": 8.140161725067385,
      "grad_norm": 0.8586001992225647,
      "learning_rate": 0.00011189854290339988,
      "loss": 3.1706,
      "step": 75500
    },
    {
      "epoch": 8.1455525606469,
      "grad_norm": 0.8062219619750977,
      "learning_rate": 0.00011157474365893146,
      "loss": 3.1751,
      "step": 75550
    },
    {
      "epoch": 8.150943396226415,
      "grad_norm": 0.8286907076835632,
      "learning_rate": 0.00011125094441446302,
      "loss": 3.1779,
      "step": 75600
    },
    {
      "epoch": 8.15633423180593,
      "grad_norm": 0.8133478164672852,
      "learning_rate": 0.00011092714516999459,
      "loss": 3.1643,
      "step": 75650
    },
    {
      "epoch": 8.161725067385445,
      "grad_norm": 0.8967101573944092,
      "learning_rate": 0.00011060334592552616,
      "loss": 3.176,
      "step": 75700
    },
    {
      "epoch": 8.167115902964959,
      "grad_norm": 0.8152874112129211,
      "learning_rate": 0.00011027954668105773,
      "loss": 3.1753,
      "step": 75750
    },
    {
      "epoch": 8.172506738544474,
      "grad_norm": 0.8211420774459839,
      "learning_rate": 0.0001099557474365893,
      "loss": 3.194,
      "step": 75800
    },
    {
      "epoch": 8.177897574123989,
      "grad_norm": 0.8731687068939209,
      "learning_rate": 0.00010963194819212088,
      "loss": 3.199,
      "step": 75850
    },
    {
      "epoch": 8.183288409703504,
      "grad_norm": 0.8784967660903931,
      "learning_rate": 0.00010930814894765243,
      "loss": 3.1609,
      "step": 75900
    },
    {
      "epoch": 8.18867924528302,
      "grad_norm": 0.8387054204940796,
      "learning_rate": 0.00010898434970318401,
      "loss": 3.1826,
      "step": 75950
    },
    {
      "epoch": 8.194070080862534,
      "grad_norm": 0.8362334370613098,
      "learning_rate": 0.00010866055045871559,
      "loss": 3.185,
      "step": 76000
    },
    {
      "epoch": 8.194070080862534,
      "eval_accuracy": 0.3901588298779361,
      "eval_loss": 3.3355491161346436,
      "eval_runtime": 183.8342,
      "eval_samples_per_second": 97.974,
      "eval_steps_per_second": 6.125,
      "step": 76000
    },
    {
      "epoch": 8.199460916442048,
      "grad_norm": 0.8048451542854309,
      "learning_rate": 0.00010833675121424717,
      "loss": 3.1804,
      "step": 76050
    },
    {
      "epoch": 8.204851752021563,
      "grad_norm": 0.8230847120285034,
      "learning_rate": 0.00010801295196977872,
      "loss": 3.1555,
      "step": 76100
    },
    {
      "epoch": 8.210242587601078,
      "grad_norm": 0.82015061378479,
      "learning_rate": 0.0001076891527253103,
      "loss": 3.1869,
      "step": 76150
    },
    {
      "epoch": 8.215633423180593,
      "grad_norm": 0.8503551483154297,
      "learning_rate": 0.00010737182946573123,
      "loss": 3.1691,
      "step": 76200
    },
    {
      "epoch": 8.221024258760108,
      "grad_norm": 0.8172155022621155,
      "learning_rate": 0.0001070480302212628,
      "loss": 3.1655,
      "step": 76250
    },
    {
      "epoch": 8.226415094339623,
      "grad_norm": 0.7921382188796997,
      "learning_rate": 0.00010672423097679438,
      "loss": 3.1808,
      "step": 76300
    },
    {
      "epoch": 8.231805929919137,
      "grad_norm": 0.8233168125152588,
      "learning_rate": 0.00010640043173232595,
      "loss": 3.1727,
      "step": 76350
    },
    {
      "epoch": 8.237196765498652,
      "grad_norm": 0.8275365233421326,
      "learning_rate": 0.00010607663248785753,
      "loss": 3.166,
      "step": 76400
    },
    {
      "epoch": 8.242587601078167,
      "grad_norm": 0.8599715232849121,
      "learning_rate": 0.00010575283324338908,
      "loss": 3.1878,
      "step": 76450
    },
    {
      "epoch": 8.247978436657682,
      "grad_norm": 0.8927331566810608,
      "learning_rate": 0.00010542903399892066,
      "loss": 3.1882,
      "step": 76500
    },
    {
      "epoch": 8.253369272237197,
      "grad_norm": 0.8399921655654907,
      "learning_rate": 0.00010510523475445224,
      "loss": 3.1921,
      "step": 76550
    },
    {
      "epoch": 8.25876010781671,
      "grad_norm": 0.8249356150627136,
      "learning_rate": 0.0001047814355099838,
      "loss": 3.1809,
      "step": 76600
    },
    {
      "epoch": 8.264150943396226,
      "grad_norm": 0.8486528396606445,
      "learning_rate": 0.00010445763626551537,
      "loss": 3.1722,
      "step": 76650
    },
    {
      "epoch": 8.269541778975741,
      "grad_norm": 0.8354886174201965,
      "learning_rate": 0.00010413383702104694,
      "loss": 3.1934,
      "step": 76700
    },
    {
      "epoch": 8.274932614555256,
      "grad_norm": 0.8391222357749939,
      "learning_rate": 0.00010381003777657852,
      "loss": 3.1813,
      "step": 76750
    },
    {
      "epoch": 8.280323450134771,
      "grad_norm": 0.8095042705535889,
      "learning_rate": 0.00010348623853211008,
      "loss": 3.1764,
      "step": 76800
    },
    {
      "epoch": 8.285714285714286,
      "grad_norm": 0.8261136412620544,
      "learning_rate": 0.00010316243928764165,
      "loss": 3.1795,
      "step": 76850
    },
    {
      "epoch": 8.2911051212938,
      "grad_norm": 0.8423893451690674,
      "learning_rate": 0.00010283864004317323,
      "loss": 3.1935,
      "step": 76900
    },
    {
      "epoch": 8.296495956873315,
      "grad_norm": 0.8349312543869019,
      "learning_rate": 0.00010251484079870479,
      "loss": 3.1764,
      "step": 76950
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.864117443561554,
      "learning_rate": 0.00010219104155423637,
      "loss": 3.1758,
      "step": 77000
    },
    {
      "epoch": 8.30188679245283,
      "eval_accuracy": 0.3905072796891571,
      "eval_loss": 3.3320937156677246,
      "eval_runtime": 183.9498,
      "eval_samples_per_second": 97.913,
      "eval_steps_per_second": 6.121,
      "step": 77000
    },
    {
      "epoch": 8.307277628032345,
      "grad_norm": 0.8279615044593811,
      "learning_rate": 0.00010186724230976793,
      "loss": 3.1842,
      "step": 77050
    },
    {
      "epoch": 8.31266846361186,
      "grad_norm": 0.9203056693077087,
      "learning_rate": 0.0001015434430652995,
      "loss": 3.1882,
      "step": 77100
    },
    {
      "epoch": 8.318059299191376,
      "grad_norm": 0.9166232943534851,
      "learning_rate": 0.00010121964382083107,
      "loss": 3.1777,
      "step": 77150
    },
    {
      "epoch": 8.323450134770889,
      "grad_norm": 0.8042053580284119,
      "learning_rate": 0.00010089584457636265,
      "loss": 3.171,
      "step": 77200
    },
    {
      "epoch": 8.328840970350404,
      "grad_norm": 0.7936014533042908,
      "learning_rate": 0.00010057204533189423,
      "loss": 3.1757,
      "step": 77250
    },
    {
      "epoch": 8.33423180592992,
      "grad_norm": 0.8509289026260376,
      "learning_rate": 0.00010024824608742578,
      "loss": 3.1787,
      "step": 77300
    },
    {
      "epoch": 8.339622641509434,
      "grad_norm": 0.8344526290893555,
      "learning_rate": 9.992444684295736e-05,
      "loss": 3.1696,
      "step": 77350
    },
    {
      "epoch": 8.34501347708895,
      "grad_norm": 0.8416390419006348,
      "learning_rate": 9.960064759848894e-05,
      "loss": 3.2041,
      "step": 77400
    },
    {
      "epoch": 8.350404312668463,
      "grad_norm": 0.8684095740318298,
      "learning_rate": 9.92768483540205e-05,
      "loss": 3.1784,
      "step": 77450
    },
    {
      "epoch": 8.355795148247978,
      "grad_norm": 0.8323090672492981,
      "learning_rate": 9.895304910955206e-05,
      "loss": 3.1704,
      "step": 77500
    },
    {
      "epoch": 8.361185983827493,
      "grad_norm": 0.8483664989471436,
      "learning_rate": 9.862924986508364e-05,
      "loss": 3.1927,
      "step": 77550
    },
    {
      "epoch": 8.366576819407008,
      "grad_norm": 0.818683922290802,
      "learning_rate": 9.83054506206152e-05,
      "loss": 3.186,
      "step": 77600
    },
    {
      "epoch": 8.371967654986523,
      "grad_norm": 0.8603020906448364,
      "learning_rate": 9.798165137614678e-05,
      "loss": 3.1818,
      "step": 77650
    },
    {
      "epoch": 8.377358490566039,
      "grad_norm": 0.8788433074951172,
      "learning_rate": 9.765785213167835e-05,
      "loss": 3.1842,
      "step": 77700
    },
    {
      "epoch": 8.382749326145552,
      "grad_norm": 0.8362876772880554,
      "learning_rate": 9.733405288720992e-05,
      "loss": 3.1768,
      "step": 77750
    },
    {
      "epoch": 8.388140161725067,
      "grad_norm": 0.8131971955299377,
      "learning_rate": 9.701025364274149e-05,
      "loss": 3.1838,
      "step": 77800
    },
    {
      "epoch": 8.393530997304582,
      "grad_norm": 0.8119576573371887,
      "learning_rate": 9.668645439827307e-05,
      "loss": 3.1686,
      "step": 77850
    },
    {
      "epoch": 8.398921832884097,
      "grad_norm": 0.8404786586761475,
      "learning_rate": 9.636265515380464e-05,
      "loss": 3.1771,
      "step": 77900
    },
    {
      "epoch": 8.404312668463612,
      "grad_norm": 0.8290375471115112,
      "learning_rate": 9.60388559093362e-05,
      "loss": 3.1822,
      "step": 77950
    },
    {
      "epoch": 8.409703504043126,
      "grad_norm": 0.8503365516662598,
      "learning_rate": 9.571505666486777e-05,
      "loss": 3.1928,
      "step": 78000
    },
    {
      "epoch": 8.409703504043126,
      "eval_accuracy": 0.390971770784564,
      "eval_loss": 3.3296592235565186,
      "eval_runtime": 183.6471,
      "eval_samples_per_second": 98.074,
      "eval_steps_per_second": 6.131,
      "step": 78000
    },
    {
      "epoch": 8.415094339622641,
      "grad_norm": 0.8500294089317322,
      "learning_rate": 9.539125742039935e-05,
      "loss": 3.1717,
      "step": 78050
    },
    {
      "epoch": 8.420485175202156,
      "grad_norm": 0.8462092280387878,
      "learning_rate": 9.506745817593093e-05,
      "loss": 3.1679,
      "step": 78100
    },
    {
      "epoch": 8.425876010781671,
      "grad_norm": 0.8475029468536377,
      "learning_rate": 9.474365893146248e-05,
      "loss": 3.1917,
      "step": 78150
    },
    {
      "epoch": 8.431266846361186,
      "grad_norm": 0.887370765209198,
      "learning_rate": 9.441985968699406e-05,
      "loss": 3.2017,
      "step": 78200
    },
    {
      "epoch": 8.436657681940702,
      "grad_norm": 0.8378453254699707,
      "learning_rate": 9.409606044252563e-05,
      "loss": 3.1763,
      "step": 78250
    },
    {
      "epoch": 8.442048517520215,
      "grad_norm": 0.8680948615074158,
      "learning_rate": 9.37722611980572e-05,
      "loss": 3.1699,
      "step": 78300
    },
    {
      "epoch": 8.44743935309973,
      "grad_norm": 0.8394586443901062,
      "learning_rate": 9.344846195358876e-05,
      "loss": 3.177,
      "step": 78350
    },
    {
      "epoch": 8.452830188679245,
      "grad_norm": 0.850395679473877,
      "learning_rate": 9.312466270912034e-05,
      "loss": 3.1694,
      "step": 78400
    },
    {
      "epoch": 8.45822102425876,
      "grad_norm": 0.8615196943283081,
      "learning_rate": 9.28008634646519e-05,
      "loss": 3.1778,
      "step": 78450
    },
    {
      "epoch": 8.463611859838275,
      "grad_norm": 0.8771566152572632,
      "learning_rate": 9.247706422018348e-05,
      "loss": 3.1854,
      "step": 78500
    },
    {
      "epoch": 8.46900269541779,
      "grad_norm": 0.8475139737129211,
      "learning_rate": 9.215326497571505e-05,
      "loss": 3.1787,
      "step": 78550
    },
    {
      "epoch": 8.474393530997304,
      "grad_norm": 0.9131626486778259,
      "learning_rate": 9.182946573124661e-05,
      "loss": 3.1901,
      "step": 78600
    },
    {
      "epoch": 8.479784366576819,
      "grad_norm": 0.8300061821937561,
      "learning_rate": 9.151214247166756e-05,
      "loss": 3.1735,
      "step": 78650
    },
    {
      "epoch": 8.485175202156334,
      "grad_norm": 0.8313758373260498,
      "learning_rate": 9.118834322719913e-05,
      "loss": 3.1882,
      "step": 78700
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.8368052244186401,
      "learning_rate": 9.08645439827307e-05,
      "loss": 3.1876,
      "step": 78750
    },
    {
      "epoch": 8.495956873315365,
      "grad_norm": 0.9024884700775146,
      "learning_rate": 9.054074473826227e-05,
      "loss": 3.1979,
      "step": 78800
    },
    {
      "epoch": 8.501347708894878,
      "grad_norm": 0.8912298679351807,
      "learning_rate": 9.021694549379385e-05,
      "loss": 3.1707,
      "step": 78850
    },
    {
      "epoch": 8.506738544474393,
      "grad_norm": 0.838214099407196,
      "learning_rate": 8.989314624932541e-05,
      "loss": 3.1788,
      "step": 78900
    },
    {
      "epoch": 8.512129380053908,
      "grad_norm": 0.7878483533859253,
      "learning_rate": 8.956934700485697e-05,
      "loss": 3.1794,
      "step": 78950
    },
    {
      "epoch": 8.517520215633423,
      "grad_norm": 0.8773841857910156,
      "learning_rate": 8.924554776038855e-05,
      "loss": 3.1793,
      "step": 79000
    },
    {
      "epoch": 8.517520215633423,
      "eval_accuracy": 0.3911975514854425,
      "eval_loss": 3.3264498710632324,
      "eval_runtime": 184.0527,
      "eval_samples_per_second": 97.858,
      "eval_steps_per_second": 6.118,
      "step": 79000
    },
    {
      "epoch": 8.522911051212938,
      "grad_norm": 0.885176956653595,
      "learning_rate": 8.892174851592013e-05,
      "loss": 3.1912,
      "step": 79050
    },
    {
      "epoch": 8.528301886792454,
      "grad_norm": 0.8777654767036438,
      "learning_rate": 8.859794927145168e-05,
      "loss": 3.187,
      "step": 79100
    },
    {
      "epoch": 8.533692722371967,
      "grad_norm": 0.8870535492897034,
      "learning_rate": 8.827415002698326e-05,
      "loss": 3.1758,
      "step": 79150
    },
    {
      "epoch": 8.539083557951482,
      "grad_norm": 0.8505139946937561,
      "learning_rate": 8.795035078251483e-05,
      "loss": 3.1889,
      "step": 79200
    },
    {
      "epoch": 8.544474393530997,
      "grad_norm": 0.8751934766769409,
      "learning_rate": 8.762655153804641e-05,
      "loss": 3.1847,
      "step": 79250
    },
    {
      "epoch": 8.549865229110512,
      "grad_norm": 0.9024791717529297,
      "learning_rate": 8.730275229357798e-05,
      "loss": 3.1797,
      "step": 79300
    },
    {
      "epoch": 8.555256064690028,
      "grad_norm": 0.9003559947013855,
      "learning_rate": 8.697895304910954e-05,
      "loss": 3.2043,
      "step": 79350
    },
    {
      "epoch": 8.560646900269543,
      "grad_norm": 0.8512437343597412,
      "learning_rate": 8.665515380464112e-05,
      "loss": 3.1781,
      "step": 79400
    },
    {
      "epoch": 8.566037735849056,
      "grad_norm": 0.8742110729217529,
      "learning_rate": 8.633135456017268e-05,
      "loss": 3.1958,
      "step": 79450
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 0.8985169529914856,
      "learning_rate": 8.600755531570426e-05,
      "loss": 3.1726,
      "step": 79500
    },
    {
      "epoch": 8.576819407008086,
      "grad_norm": 0.8324291110038757,
      "learning_rate": 8.568375607123582e-05,
      "loss": 3.1744,
      "step": 79550
    },
    {
      "epoch": 8.582210242587601,
      "grad_norm": 0.8357748985290527,
      "learning_rate": 8.53599568267674e-05,
      "loss": 3.1933,
      "step": 79600
    },
    {
      "epoch": 8.587601078167117,
      "grad_norm": 0.8282589912414551,
      "learning_rate": 8.503615758229897e-05,
      "loss": 3.1728,
      "step": 79650
    },
    {
      "epoch": 8.59299191374663,
      "grad_norm": 0.9156341552734375,
      "learning_rate": 8.471235833783054e-05,
      "loss": 3.1786,
      "step": 79700
    },
    {
      "epoch": 8.598382749326145,
      "grad_norm": 0.8525185585021973,
      "learning_rate": 8.438855909336211e-05,
      "loss": 3.1799,
      "step": 79750
    },
    {
      "epoch": 8.60377358490566,
      "grad_norm": 0.8564472794532776,
      "learning_rate": 8.406475984889367e-05,
      "loss": 3.1775,
      "step": 79800
    },
    {
      "epoch": 8.609164420485175,
      "grad_norm": 0.8167390823364258,
      "learning_rate": 8.374096060442525e-05,
      "loss": 3.176,
      "step": 79850
    },
    {
      "epoch": 8.61455525606469,
      "grad_norm": 0.8467491269111633,
      "learning_rate": 8.341716135995683e-05,
      "loss": 3.1902,
      "step": 79900
    },
    {
      "epoch": 8.619946091644206,
      "grad_norm": 0.8552255034446716,
      "learning_rate": 8.309336211548838e-05,
      "loss": 3.1948,
      "step": 79950
    },
    {
      "epoch": 8.625336927223719,
      "grad_norm": 0.8199063539505005,
      "learning_rate": 8.276956287101996e-05,
      "loss": 3.1757,
      "step": 80000
    },
    {
      "epoch": 8.625336927223719,
      "eval_accuracy": 0.3916971374636135,
      "eval_loss": 3.32309889793396,
      "eval_runtime": 183.6042,
      "eval_samples_per_second": 98.097,
      "eval_steps_per_second": 6.133,
      "step": 80000
    },
    {
      "epoch": 8.630727762803234,
      "grad_norm": 0.8239213228225708,
      "learning_rate": 8.244576362655153e-05,
      "loss": 3.173,
      "step": 80050
    },
    {
      "epoch": 8.63611859838275,
      "grad_norm": 0.8665240406990051,
      "learning_rate": 8.212196438208311e-05,
      "loss": 3.174,
      "step": 80100
    },
    {
      "epoch": 8.641509433962264,
      "grad_norm": 0.8977245092391968,
      "learning_rate": 8.179816513761467e-05,
      "loss": 3.1986,
      "step": 80150
    },
    {
      "epoch": 8.64690026954178,
      "grad_norm": 0.821174144744873,
      "learning_rate": 8.147436589314624e-05,
      "loss": 3.1776,
      "step": 80200
    },
    {
      "epoch": 8.652291105121293,
      "grad_norm": 0.8627246618270874,
      "learning_rate": 8.115056664867782e-05,
      "loss": 3.1993,
      "step": 80250
    },
    {
      "epoch": 8.657681940700808,
      "grad_norm": 0.8275755643844604,
      "learning_rate": 8.082676740420938e-05,
      "loss": 3.1825,
      "step": 80300
    },
    {
      "epoch": 8.663072776280323,
      "grad_norm": 0.8466624617576599,
      "learning_rate": 8.050296815974096e-05,
      "loss": 3.186,
      "step": 80350
    },
    {
      "epoch": 8.668463611859838,
      "grad_norm": 0.8501802086830139,
      "learning_rate": 8.017916891527252e-05,
      "loss": 3.1807,
      "step": 80400
    },
    {
      "epoch": 8.673854447439354,
      "grad_norm": 0.8679507970809937,
      "learning_rate": 7.985536967080409e-05,
      "loss": 3.1781,
      "step": 80450
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.8288442492485046,
      "learning_rate": 7.953157042633566e-05,
      "loss": 3.1756,
      "step": 80500
    },
    {
      "epoch": 8.684636118598382,
      "grad_norm": 0.9615182280540466,
      "learning_rate": 7.920777118186724e-05,
      "loss": 3.1839,
      "step": 80550
    },
    {
      "epoch": 8.690026954177897,
      "grad_norm": 0.8873838186264038,
      "learning_rate": 7.888397193739879e-05,
      "loss": 3.1769,
      "step": 80600
    },
    {
      "epoch": 8.695417789757412,
      "grad_norm": 0.8448495268821716,
      "learning_rate": 7.856017269293037e-05,
      "loss": 3.1886,
      "step": 80650
    },
    {
      "epoch": 8.700808625336927,
      "grad_norm": 0.880897045135498,
      "learning_rate": 7.823637344846195e-05,
      "loss": 3.1907,
      "step": 80700
    },
    {
      "epoch": 8.706199460916443,
      "grad_norm": 0.858959436416626,
      "learning_rate": 7.791257420399353e-05,
      "loss": 3.1827,
      "step": 80750
    },
    {
      "epoch": 8.711590296495956,
      "grad_norm": 0.8667418360710144,
      "learning_rate": 7.758877495952508e-05,
      "loss": 3.1533,
      "step": 80800
    },
    {
      "epoch": 8.716981132075471,
      "grad_norm": 0.8277961611747742,
      "learning_rate": 7.726497571505665e-05,
      "loss": 3.1854,
      "step": 80850
    },
    {
      "epoch": 8.722371967654986,
      "grad_norm": 0.8159651160240173,
      "learning_rate": 7.694117647058823e-05,
      "loss": 3.1746,
      "step": 80900
    },
    {
      "epoch": 8.727762803234501,
      "grad_norm": 0.8652306795120239,
      "learning_rate": 7.661737722611981e-05,
      "loss": 3.1753,
      "step": 80950
    },
    {
      "epoch": 8.733153638814017,
      "grad_norm": 0.8533473014831543,
      "learning_rate": 7.629357798165137e-05,
      "loss": 3.1885,
      "step": 81000
    },
    {
      "epoch": 8.733153638814017,
      "eval_accuracy": 0.3921426143036529,
      "eval_loss": 3.3177406787872314,
      "eval_runtime": 183.8875,
      "eval_samples_per_second": 97.946,
      "eval_steps_per_second": 6.123,
      "step": 81000
    },
    {
      "epoch": 8.738544474393532,
      "grad_norm": 0.881313145160675,
      "learning_rate": 7.596977873718294e-05,
      "loss": 3.1966,
      "step": 81050
    },
    {
      "epoch": 8.743935309973045,
      "grad_norm": 0.8376140594482422,
      "learning_rate": 7.564597949271451e-05,
      "loss": 3.1998,
      "step": 81100
    },
    {
      "epoch": 8.74932614555256,
      "grad_norm": 0.8664501309394836,
      "learning_rate": 7.532218024824608e-05,
      "loss": 3.1735,
      "step": 81150
    },
    {
      "epoch": 8.754716981132075,
      "grad_norm": 0.8317979574203491,
      "learning_rate": 7.499838100377764e-05,
      "loss": 3.1951,
      "step": 81200
    },
    {
      "epoch": 8.76010781671159,
      "grad_norm": 0.8491113185882568,
      "learning_rate": 7.467458175930922e-05,
      "loss": 3.1775,
      "step": 81250
    },
    {
      "epoch": 8.765498652291106,
      "grad_norm": 0.8893978595733643,
      "learning_rate": 7.435078251484078e-05,
      "loss": 3.1669,
      "step": 81300
    },
    {
      "epoch": 8.77088948787062,
      "grad_norm": 0.8895040154457092,
      "learning_rate": 7.402698327037236e-05,
      "loss": 3.1894,
      "step": 81350
    },
    {
      "epoch": 8.776280323450134,
      "grad_norm": 0.9167323708534241,
      "learning_rate": 7.370318402590393e-05,
      "loss": 3.2023,
      "step": 81400
    },
    {
      "epoch": 8.78167115902965,
      "grad_norm": 0.8975420594215393,
      "learning_rate": 7.338586076632488e-05,
      "loss": 3.1803,
      "step": 81450
    },
    {
      "epoch": 8.787061994609164,
      "grad_norm": 0.8334147334098816,
      "learning_rate": 7.306206152185644e-05,
      "loss": 3.1854,
      "step": 81500
    },
    {
      "epoch": 8.79245283018868,
      "grad_norm": 0.8818899393081665,
      "learning_rate": 7.273826227738801e-05,
      "loss": 3.1842,
      "step": 81550
    },
    {
      "epoch": 8.797843665768195,
      "grad_norm": 0.8167200088500977,
      "learning_rate": 7.241446303291958e-05,
      "loss": 3.1631,
      "step": 81600
    },
    {
      "epoch": 8.80323450134771,
      "grad_norm": 0.8305187821388245,
      "learning_rate": 7.209066378845115e-05,
      "loss": 3.1923,
      "step": 81650
    },
    {
      "epoch": 8.808625336927223,
      "grad_norm": 0.8849042057991028,
      "learning_rate": 7.176686454398273e-05,
      "loss": 3.1956,
      "step": 81700
    },
    {
      "epoch": 8.814016172506738,
      "grad_norm": 0.8665215969085693,
      "learning_rate": 7.144306529951429e-05,
      "loss": 3.1857,
      "step": 81750
    },
    {
      "epoch": 8.819407008086253,
      "grad_norm": 0.8836647868156433,
      "learning_rate": 7.111926605504587e-05,
      "loss": 3.1706,
      "step": 81800
    },
    {
      "epoch": 8.824797843665769,
      "grad_norm": 0.8913038372993469,
      "learning_rate": 7.079546681057743e-05,
      "loss": 3.1907,
      "step": 81850
    },
    {
      "epoch": 8.830188679245284,
      "grad_norm": 0.8813517093658447,
      "learning_rate": 7.047166756610901e-05,
      "loss": 3.1911,
      "step": 81900
    },
    {
      "epoch": 8.835579514824797,
      "grad_norm": 0.8769224286079407,
      "learning_rate": 7.014786832164057e-05,
      "loss": 3.1783,
      "step": 81950
    },
    {
      "epoch": 8.840970350404312,
      "grad_norm": 0.8626630306243896,
      "learning_rate": 6.982406907717215e-05,
      "loss": 3.1939,
      "step": 82000
    },
    {
      "epoch": 8.840970350404312,
      "eval_accuracy": 0.39245173176948506,
      "eval_loss": 3.3154091835021973,
      "eval_runtime": 183.9266,
      "eval_samples_per_second": 97.925,
      "eval_steps_per_second": 6.122,
      "step": 82000
    },
    {
      "epoch": 8.846361185983827,
      "grad_norm": 0.8761478662490845,
      "learning_rate": 6.950026983270372e-05,
      "loss": 3.1862,
      "step": 82050
    },
    {
      "epoch": 8.851752021563343,
      "grad_norm": 0.8595954179763794,
      "learning_rate": 6.91764705882353e-05,
      "loss": 3.1753,
      "step": 82100
    },
    {
      "epoch": 8.857142857142858,
      "grad_norm": 0.8653388619422913,
      "learning_rate": 6.885267134376686e-05,
      "loss": 3.1944,
      "step": 82150
    },
    {
      "epoch": 8.862533692722373,
      "grad_norm": 0.9042940735816956,
      "learning_rate": 6.852887209929844e-05,
      "loss": 3.1794,
      "step": 82200
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.8679096698760986,
      "learning_rate": 6.820507285483e-05,
      "loss": 3.1972,
      "step": 82250
    },
    {
      "epoch": 8.873315363881401,
      "grad_norm": 0.8591122627258301,
      "learning_rate": 6.788127361036156e-05,
      "loss": 3.1823,
      "step": 82300
    },
    {
      "epoch": 8.878706199460916,
      "grad_norm": 0.829720675945282,
      "learning_rate": 6.755747436589314e-05,
      "loss": 3.1901,
      "step": 82350
    },
    {
      "epoch": 8.884097035040432,
      "grad_norm": 0.8924640417098999,
      "learning_rate": 6.72336751214247e-05,
      "loss": 3.1803,
      "step": 82400
    },
    {
      "epoch": 8.889487870619947,
      "grad_norm": 0.8685306906700134,
      "learning_rate": 6.690987587695628e-05,
      "loss": 3.1928,
      "step": 82450
    },
    {
      "epoch": 8.89487870619946,
      "grad_norm": 0.8657023906707764,
      "learning_rate": 6.658607663248785e-05,
      "loss": 3.1798,
      "step": 82500
    },
    {
      "epoch": 8.900269541778975,
      "grad_norm": 0.8230668902397156,
      "learning_rate": 6.626227738801942e-05,
      "loss": 3.1695,
      "step": 82550
    },
    {
      "epoch": 8.90566037735849,
      "grad_norm": 0.8264906406402588,
      "learning_rate": 6.593847814355099e-05,
      "loss": 3.1872,
      "step": 82600
    },
    {
      "epoch": 8.911051212938006,
      "grad_norm": 0.872795820236206,
      "learning_rate": 6.561467889908257e-05,
      "loss": 3.1811,
      "step": 82650
    },
    {
      "epoch": 8.91644204851752,
      "grad_norm": 0.8508303761482239,
      "learning_rate": 6.529087965461413e-05,
      "loss": 3.1977,
      "step": 82700
    },
    {
      "epoch": 8.921832884097036,
      "grad_norm": 0.8353860974311829,
      "learning_rate": 6.496708041014571e-05,
      "loss": 3.1709,
      "step": 82750
    },
    {
      "epoch": 8.92722371967655,
      "grad_norm": 0.8931666016578674,
      "learning_rate": 6.464328116567727e-05,
      "loss": 3.1748,
      "step": 82800
    },
    {
      "epoch": 8.932614555256064,
      "grad_norm": 0.8589481115341187,
      "learning_rate": 6.431948192120885e-05,
      "loss": 3.1785,
      "step": 82850
    },
    {
      "epoch": 8.93800539083558,
      "grad_norm": 0.8512519598007202,
      "learning_rate": 6.399568267674041e-05,
      "loss": 3.1763,
      "step": 82900
    },
    {
      "epoch": 8.943396226415095,
      "grad_norm": 0.8806709051132202,
      "learning_rate": 6.367188343227199e-05,
      "loss": 3.1768,
      "step": 82950
    },
    {
      "epoch": 8.94878706199461,
      "grad_norm": 0.8603725433349609,
      "learning_rate": 6.334808418780356e-05,
      "loss": 3.172,
      "step": 83000
    },
    {
      "epoch": 8.94878706199461,
      "eval_accuracy": 0.3926380714720869,
      "eval_loss": 3.3128795623779297,
      "eval_runtime": 184.4695,
      "eval_samples_per_second": 97.637,
      "eval_steps_per_second": 6.104,
      "step": 83000
    },
    {
      "epoch": 8.954177897574123,
      "grad_norm": 0.8853550553321838,
      "learning_rate": 6.302428494333512e-05,
      "loss": 3.169,
      "step": 83050
    },
    {
      "epoch": 8.959568733153638,
      "grad_norm": 0.8855602741241455,
      "learning_rate": 6.27004856988667e-05,
      "loss": 3.1791,
      "step": 83100
    },
    {
      "epoch": 8.964959568733153,
      "grad_norm": 0.8563063740730286,
      "learning_rate": 6.237668645439826e-05,
      "loss": 3.2049,
      "step": 83150
    },
    {
      "epoch": 8.970350404312669,
      "grad_norm": 0.8717796206474304,
      "learning_rate": 6.205288720992984e-05,
      "loss": 3.182,
      "step": 83200
    },
    {
      "epoch": 8.975741239892184,
      "grad_norm": 0.8790916204452515,
      "learning_rate": 6.17290879654614e-05,
      "loss": 3.1686,
      "step": 83250
    },
    {
      "epoch": 8.981132075471699,
      "grad_norm": 0.8493698239326477,
      "learning_rate": 6.140528872099298e-05,
      "loss": 3.178,
      "step": 83300
    },
    {
      "epoch": 8.986522911051212,
      "grad_norm": 0.8447183966636658,
      "learning_rate": 6.108148947652455e-05,
      "loss": 3.182,
      "step": 83350
    },
    {
      "epoch": 8.991913746630727,
      "grad_norm": 0.8738790154457092,
      "learning_rate": 6.075769023205612e-05,
      "loss": 3.1719,
      "step": 83400
    },
    {
      "epoch": 8.997304582210242,
      "grad_norm": 0.9319518804550171,
      "learning_rate": 6.043389098758769e-05,
      "loss": 3.1974,
      "step": 83450
    },
    {
      "epoch": 9.002695417789758,
      "grad_norm": 0.8708575367927551,
      "learning_rate": 6.0110091743119265e-05,
      "loss": 3.1247,
      "step": 83500
    },
    {
      "epoch": 9.008086253369273,
      "grad_norm": 0.8379282355308533,
      "learning_rate": 5.978629249865083e-05,
      "loss": 3.1218,
      "step": 83550
    },
    {
      "epoch": 9.013477088948788,
      "grad_norm": 0.8564487099647522,
      "learning_rate": 5.94624932541824e-05,
      "loss": 3.1391,
      "step": 83600
    },
    {
      "epoch": 9.018867924528301,
      "grad_norm": 0.8536072969436646,
      "learning_rate": 5.9145169994603345e-05,
      "loss": 3.1322,
      "step": 83650
    },
    {
      "epoch": 9.024258760107816,
      "grad_norm": 0.8448429703712463,
      "learning_rate": 5.882137075013491e-05,
      "loss": 3.1157,
      "step": 83700
    },
    {
      "epoch": 9.029649595687331,
      "grad_norm": 0.8528944849967957,
      "learning_rate": 5.849757150566649e-05,
      "loss": 3.1142,
      "step": 83750
    },
    {
      "epoch": 9.035040431266847,
      "grad_norm": 0.8785792589187622,
      "learning_rate": 5.817377226119805e-05,
      "loss": 3.1166,
      "step": 83800
    },
    {
      "epoch": 9.040431266846362,
      "grad_norm": 0.872730553150177,
      "learning_rate": 5.784997301672963e-05,
      "loss": 3.1049,
      "step": 83850
    },
    {
      "epoch": 9.045822102425875,
      "grad_norm": 0.8977234363555908,
      "learning_rate": 5.752617377226119e-05,
      "loss": 3.1209,
      "step": 83900
    },
    {
      "epoch": 9.05121293800539,
      "grad_norm": 0.8592477440834045,
      "learning_rate": 5.7202374527792764e-05,
      "loss": 3.125,
      "step": 83950
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.8540244698524475,
      "learning_rate": 5.6878575283324335e-05,
      "loss": 3.1069,
      "step": 84000
    },
    {
      "epoch": 9.056603773584905,
      "eval_accuracy": 0.39254386941835173,
      "eval_loss": 3.316199541091919,
      "eval_runtime": 183.6321,
      "eval_samples_per_second": 98.082,
      "eval_steps_per_second": 6.132,
      "step": 84000
    },
    {
      "epoch": 9.06199460916442,
      "grad_norm": 0.8707414865493774,
      "learning_rate": 5.6554776038855905e-05,
      "loss": 3.1212,
      "step": 84050
    },
    {
      "epoch": 9.067385444743936,
      "grad_norm": 0.8910436630249023,
      "learning_rate": 5.623097679438747e-05,
      "loss": 3.1472,
      "step": 84100
    },
    {
      "epoch": 9.07277628032345,
      "grad_norm": 0.9022126197814941,
      "learning_rate": 5.590717754991905e-05,
      "loss": 3.1256,
      "step": 84150
    },
    {
      "epoch": 9.078167115902964,
      "grad_norm": 0.8683561682701111,
      "learning_rate": 5.558337830545061e-05,
      "loss": 3.1405,
      "step": 84200
    },
    {
      "epoch": 9.08355795148248,
      "grad_norm": 0.8293735384941101,
      "learning_rate": 5.525957906098219e-05,
      "loss": 3.1366,
      "step": 84250
    },
    {
      "epoch": 9.088948787061994,
      "grad_norm": 0.8638619780540466,
      "learning_rate": 5.493577981651375e-05,
      "loss": 3.1273,
      "step": 84300
    },
    {
      "epoch": 9.09433962264151,
      "grad_norm": 0.8600364327430725,
      "learning_rate": 5.4611980572045324e-05,
      "loss": 3.1203,
      "step": 84350
    },
    {
      "epoch": 9.099730458221025,
      "grad_norm": 0.9247821569442749,
      "learning_rate": 5.4288181327576895e-05,
      "loss": 3.1105,
      "step": 84400
    },
    {
      "epoch": 9.10512129380054,
      "grad_norm": 0.8678122758865356,
      "learning_rate": 5.3964382083108466e-05,
      "loss": 3.1373,
      "step": 84450
    },
    {
      "epoch": 9.110512129380053,
      "grad_norm": 0.8575831651687622,
      "learning_rate": 5.3640582838640043e-05,
      "loss": 3.1233,
      "step": 84500
    },
    {
      "epoch": 9.115902964959568,
      "grad_norm": 0.8633356690406799,
      "learning_rate": 5.331678359417161e-05,
      "loss": 3.133,
      "step": 84550
    },
    {
      "epoch": 9.121293800539084,
      "grad_norm": 0.8657795190811157,
      "learning_rate": 5.2992984349703185e-05,
      "loss": 3.1199,
      "step": 84600
    },
    {
      "epoch": 9.126684636118599,
      "grad_norm": 0.8592625856399536,
      "learning_rate": 5.266918510523475e-05,
      "loss": 3.108,
      "step": 84650
    },
    {
      "epoch": 9.132075471698114,
      "grad_norm": 0.8899664282798767,
      "learning_rate": 5.234538586076632e-05,
      "loss": 3.1206,
      "step": 84700
    },
    {
      "epoch": 9.137466307277627,
      "grad_norm": 0.9175904393196106,
      "learning_rate": 5.202158661629789e-05,
      "loss": 3.1242,
      "step": 84750
    },
    {
      "epoch": 9.142857142857142,
      "grad_norm": 0.8841559886932373,
      "learning_rate": 5.169778737182946e-05,
      "loss": 3.1228,
      "step": 84800
    },
    {
      "epoch": 9.148247978436657,
      "grad_norm": 0.8705527186393738,
      "learning_rate": 5.1373988127361026e-05,
      "loss": 3.1341,
      "step": 84850
    },
    {
      "epoch": 9.153638814016173,
      "grad_norm": 0.8766990303993225,
      "learning_rate": 5.1050188882892604e-05,
      "loss": 3.1433,
      "step": 84900
    },
    {
      "epoch": 9.159029649595688,
      "grad_norm": 0.817348062992096,
      "learning_rate": 5.072638963842417e-05,
      "loss": 3.1296,
      "step": 84950
    },
    {
      "epoch": 9.164420485175203,
      "grad_norm": 0.9388226270675659,
      "learning_rate": 5.0402590393955746e-05,
      "loss": 3.1433,
      "step": 85000
    },
    {
      "epoch": 9.164420485175203,
      "eval_accuracy": 0.39282951786041603,
      "eval_loss": 3.314669370651245,
      "eval_runtime": 183.9047,
      "eval_samples_per_second": 97.937,
      "eval_steps_per_second": 6.123,
      "step": 85000
    },
    {
      "epoch": 9.169811320754716,
      "grad_norm": 0.8864387273788452,
      "learning_rate": 5.007879114948731e-05,
      "loss": 3.1343,
      "step": 85050
    },
    {
      "epoch": 9.175202156334231,
      "grad_norm": 0.8899182677268982,
      "learning_rate": 4.975499190501889e-05,
      "loss": 3.1324,
      "step": 85100
    },
    {
      "epoch": 9.180592991913747,
      "grad_norm": 0.8436346054077148,
      "learning_rate": 4.943119266055045e-05,
      "loss": 3.1394,
      "step": 85150
    },
    {
      "epoch": 9.185983827493262,
      "grad_norm": 0.8905600905418396,
      "learning_rate": 4.910739341608202e-05,
      "loss": 3.142,
      "step": 85200
    },
    {
      "epoch": 9.191374663072777,
      "grad_norm": 0.8945043087005615,
      "learning_rate": 4.87835941716136e-05,
      "loss": 3.1179,
      "step": 85250
    },
    {
      "epoch": 9.19676549865229,
      "grad_norm": 0.8980481624603271,
      "learning_rate": 4.8459794927145164e-05,
      "loss": 3.1269,
      "step": 85300
    },
    {
      "epoch": 9.202156334231805,
      "grad_norm": 0.8833536505699158,
      "learning_rate": 4.813599568267674e-05,
      "loss": 3.1221,
      "step": 85350
    },
    {
      "epoch": 9.20754716981132,
      "grad_norm": 0.8827757835388184,
      "learning_rate": 4.7812196438208306e-05,
      "loss": 3.1308,
      "step": 85400
    },
    {
      "epoch": 9.212938005390836,
      "grad_norm": 0.8767707347869873,
      "learning_rate": 4.748839719373988e-05,
      "loss": 3.1349,
      "step": 85450
    },
    {
      "epoch": 9.21832884097035,
      "grad_norm": 0.8896255493164062,
      "learning_rate": 4.716459794927145e-05,
      "loss": 3.1217,
      "step": 85500
    },
    {
      "epoch": 9.223719676549866,
      "grad_norm": 0.9134429693222046,
      "learning_rate": 4.684079870480302e-05,
      "loss": 3.1118,
      "step": 85550
    },
    {
      "epoch": 9.22911051212938,
      "grad_norm": 0.8544310927391052,
      "learning_rate": 4.651699946033459e-05,
      "loss": 3.1185,
      "step": 85600
    },
    {
      "epoch": 9.234501347708894,
      "grad_norm": 0.8505896925926208,
      "learning_rate": 4.619320021586616e-05,
      "loss": 3.1347,
      "step": 85650
    },
    {
      "epoch": 9.23989218328841,
      "grad_norm": 0.8800784349441528,
      "learning_rate": 4.587587695628709e-05,
      "loss": 3.1258,
      "step": 85700
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.8774470686912537,
      "learning_rate": 4.555207771181867e-05,
      "loss": 3.1123,
      "step": 85750
    },
    {
      "epoch": 9.25067385444744,
      "grad_norm": 0.87869793176651,
      "learning_rate": 4.522827846735024e-05,
      "loss": 3.1254,
      "step": 85800
    },
    {
      "epoch": 9.256064690026955,
      "grad_norm": 0.8597740530967712,
      "learning_rate": 4.490447922288181e-05,
      "loss": 3.1507,
      "step": 85850
    },
    {
      "epoch": 9.261455525606468,
      "grad_norm": 0.849094033241272,
      "learning_rate": 4.458067997841338e-05,
      "loss": 3.1339,
      "step": 85900
    },
    {
      "epoch": 9.266846361185983,
      "grad_norm": 0.945573627948761,
      "learning_rate": 4.4256880733944947e-05,
      "loss": 3.1356,
      "step": 85950
    },
    {
      "epoch": 9.272237196765499,
      "grad_norm": 0.8902300000190735,
      "learning_rate": 4.3933081489476524e-05,
      "loss": 3.1322,
      "step": 86000
    },
    {
      "epoch": 9.272237196765499,
      "eval_accuracy": 0.39319730788566803,
      "eval_loss": 3.3121745586395264,
      "eval_runtime": 183.5939,
      "eval_samples_per_second": 98.102,
      "eval_steps_per_second": 6.133,
      "step": 86000
    },
    {
      "epoch": 9.277628032345014,
      "grad_norm": 0.8416682481765747,
      "learning_rate": 4.360928224500809e-05,
      "loss": 3.1422,
      "step": 86050
    },
    {
      "epoch": 9.283018867924529,
      "grad_norm": 0.8767873644828796,
      "learning_rate": 4.3285483000539666e-05,
      "loss": 3.1349,
      "step": 86100
    },
    {
      "epoch": 9.288409703504042,
      "grad_norm": 0.8905349969863892,
      "learning_rate": 4.296168375607123e-05,
      "loss": 3.1462,
      "step": 86150
    },
    {
      "epoch": 9.293800539083557,
      "grad_norm": 0.9114990234375,
      "learning_rate": 4.263788451160281e-05,
      "loss": 3.1402,
      "step": 86200
    },
    {
      "epoch": 9.299191374663073,
      "grad_norm": 0.9382705092430115,
      "learning_rate": 4.231408526713437e-05,
      "loss": 3.1439,
      "step": 86250
    },
    {
      "epoch": 9.304582210242588,
      "grad_norm": 0.9107972979545593,
      "learning_rate": 4.199028602266594e-05,
      "loss": 3.1434,
      "step": 86300
    },
    {
      "epoch": 9.309973045822103,
      "grad_norm": 0.8746132850646973,
      "learning_rate": 4.1666486778197514e-05,
      "loss": 3.1147,
      "step": 86350
    },
    {
      "epoch": 9.315363881401618,
      "grad_norm": 0.8748690485954285,
      "learning_rate": 4.1342687533729085e-05,
      "loss": 3.1236,
      "step": 86400
    },
    {
      "epoch": 9.320754716981131,
      "grad_norm": 0.8657193183898926,
      "learning_rate": 4.101888828926065e-05,
      "loss": 3.1341,
      "step": 86450
    },
    {
      "epoch": 9.326145552560646,
      "grad_norm": 0.8812388181686401,
      "learning_rate": 4.0695089044792226e-05,
      "loss": 3.1317,
      "step": 86500
    },
    {
      "epoch": 9.331536388140162,
      "grad_norm": 0.8681718111038208,
      "learning_rate": 4.037128980032379e-05,
      "loss": 3.1414,
      "step": 86550
    },
    {
      "epoch": 9.336927223719677,
      "grad_norm": 0.8916813135147095,
      "learning_rate": 4.004749055585537e-05,
      "loss": 3.1533,
      "step": 86600
    },
    {
      "epoch": 9.342318059299192,
      "grad_norm": 0.8998400568962097,
      "learning_rate": 3.972369131138694e-05,
      "loss": 3.1206,
      "step": 86650
    },
    {
      "epoch": 9.347708894878707,
      "grad_norm": 0.8771690726280212,
      "learning_rate": 3.93998920669185e-05,
      "loss": 3.1332,
      "step": 86700
    },
    {
      "epoch": 9.35309973045822,
      "grad_norm": 0.8819947242736816,
      "learning_rate": 3.907609282245008e-05,
      "loss": 3.1126,
      "step": 86750
    },
    {
      "epoch": 9.358490566037736,
      "grad_norm": 0.9157130718231201,
      "learning_rate": 3.8752293577981645e-05,
      "loss": 3.1176,
      "step": 86800
    },
    {
      "epoch": 9.36388140161725,
      "grad_norm": 0.8591070771217346,
      "learning_rate": 3.842849433351322e-05,
      "loss": 3.1351,
      "step": 86850
    },
    {
      "epoch": 9.369272237196766,
      "grad_norm": 0.8667553663253784,
      "learning_rate": 3.810469508904479e-05,
      "loss": 3.1406,
      "step": 86900
    },
    {
      "epoch": 9.374663072776281,
      "grad_norm": 0.9189602136611938,
      "learning_rate": 3.7780895844576364e-05,
      "loss": 3.1398,
      "step": 86950
    },
    {
      "epoch": 9.380053908355794,
      "grad_norm": 0.9026527404785156,
      "learning_rate": 3.745709660010793e-05,
      "loss": 3.1188,
      "step": 87000
    },
    {
      "epoch": 9.380053908355794,
      "eval_accuracy": 0.3933680015724246,
      "eval_loss": 3.3105313777923584,
      "eval_runtime": 183.4007,
      "eval_samples_per_second": 98.206,
      "eval_steps_per_second": 6.14,
      "step": 87000
    },
    {
      "epoch": 9.38544474393531,
      "grad_norm": 0.8697655200958252,
      "learning_rate": 3.71332973556395e-05,
      "loss": 3.1387,
      "step": 87050
    },
    {
      "epoch": 9.390835579514825,
      "grad_norm": 0.8661813735961914,
      "learning_rate": 3.680949811117107e-05,
      "loss": 3.1282,
      "step": 87100
    },
    {
      "epoch": 9.39622641509434,
      "grad_norm": 0.8661714196205139,
      "learning_rate": 3.648569886670264e-05,
      "loss": 3.1331,
      "step": 87150
    },
    {
      "epoch": 9.401617250673855,
      "grad_norm": 0.8739043474197388,
      "learning_rate": 3.616189962223421e-05,
      "loss": 3.1303,
      "step": 87200
    },
    {
      "epoch": 9.40700808625337,
      "grad_norm": 0.8703292608261108,
      "learning_rate": 3.583810037776578e-05,
      "loss": 3.1426,
      "step": 87250
    },
    {
      "epoch": 9.412398921832883,
      "grad_norm": 0.9095264077186584,
      "learning_rate": 3.5514301133297354e-05,
      "loss": 3.1497,
      "step": 87300
    },
    {
      "epoch": 9.417789757412399,
      "grad_norm": 0.8935539126396179,
      "learning_rate": 3.5190501888828925e-05,
      "loss": 3.1295,
      "step": 87350
    },
    {
      "epoch": 9.423180592991914,
      "grad_norm": 0.9008485674858093,
      "learning_rate": 3.4866702644360496e-05,
      "loss": 3.1277,
      "step": 87400
    },
    {
      "epoch": 9.428571428571429,
      "grad_norm": 0.9170891642570496,
      "learning_rate": 3.4542903399892067e-05,
      "loss": 3.1333,
      "step": 87450
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.9074438214302063,
      "learning_rate": 3.421910415542363e-05,
      "loss": 3.1358,
      "step": 87500
    },
    {
      "epoch": 9.439353099730457,
      "grad_norm": 0.887983500957489,
      "learning_rate": 3.38953049109552e-05,
      "loss": 3.1457,
      "step": 87550
    },
    {
      "epoch": 9.444743935309972,
      "grad_norm": 0.9073708653450012,
      "learning_rate": 3.357150566648677e-05,
      "loss": 3.1329,
      "step": 87600
    },
    {
      "epoch": 9.450134770889488,
      "grad_norm": 0.8670662641525269,
      "learning_rate": 3.324770642201834e-05,
      "loss": 3.1483,
      "step": 87650
    },
    {
      "epoch": 9.455525606469003,
      "grad_norm": 0.9071794748306274,
      "learning_rate": 3.2923907177549914e-05,
      "loss": 3.1072,
      "step": 87700
    },
    {
      "epoch": 9.460916442048518,
      "grad_norm": 0.8758538961410522,
      "learning_rate": 3.2600107933081485e-05,
      "loss": 3.1269,
      "step": 87750
    },
    {
      "epoch": 9.466307277628033,
      "grad_norm": 0.8275009989738464,
      "learning_rate": 3.2276308688613056e-05,
      "loss": 3.1231,
      "step": 87800
    },
    {
      "epoch": 9.471698113207546,
      "grad_norm": 0.8741101026535034,
      "learning_rate": 3.195250944414463e-05,
      "loss": 3.1334,
      "step": 87850
    },
    {
      "epoch": 9.477088948787062,
      "grad_norm": 0.9081462025642395,
      "learning_rate": 3.16287101996762e-05,
      "loss": 3.1442,
      "step": 87900
    },
    {
      "epoch": 9.482479784366577,
      "grad_norm": 0.8778511881828308,
      "learning_rate": 3.130491095520777e-05,
      "loss": 3.1457,
      "step": 87950
    },
    {
      "epoch": 9.487870619946092,
      "grad_norm": 0.8574414849281311,
      "learning_rate": 3.098111171073934e-05,
      "loss": 3.1361,
      "step": 88000
    },
    {
      "epoch": 9.487870619946092,
      "eval_accuracy": 0.39378316425676374,
      "eval_loss": 3.3077938556671143,
      "eval_runtime": 183.7065,
      "eval_samples_per_second": 98.042,
      "eval_steps_per_second": 6.129,
      "step": 88000
    },
    {
      "epoch": 9.493261455525607,
      "grad_norm": 0.8746349811553955,
      "learning_rate": 3.065731246627091e-05,
      "loss": 3.1334,
      "step": 88050
    },
    {
      "epoch": 9.498652291105122,
      "grad_norm": 0.8603171706199646,
      "learning_rate": 3.033351322180248e-05,
      "loss": 3.1328,
      "step": 88100
    },
    {
      "epoch": 9.504043126684635,
      "grad_norm": 0.8541721105575562,
      "learning_rate": 3.000971397733405e-05,
      "loss": 3.1322,
      "step": 88150
    },
    {
      "epoch": 9.50943396226415,
      "grad_norm": 0.8555578589439392,
      "learning_rate": 2.968591473286562e-05,
      "loss": 3.1273,
      "step": 88200
    },
    {
      "epoch": 9.514824797843666,
      "grad_norm": 0.8623125553131104,
      "learning_rate": 2.936211548839719e-05,
      "loss": 3.1337,
      "step": 88250
    },
    {
      "epoch": 9.520215633423181,
      "grad_norm": 0.9109192490577698,
      "learning_rate": 2.903831624392876e-05,
      "loss": 3.1421,
      "step": 88300
    },
    {
      "epoch": 9.525606469002696,
      "grad_norm": 0.8681851625442505,
      "learning_rate": 2.8714516999460332e-05,
      "loss": 3.1416,
      "step": 88350
    },
    {
      "epoch": 9.530997304582211,
      "grad_norm": 0.8718369603157043,
      "learning_rate": 2.83907177549919e-05,
      "loss": 3.1426,
      "step": 88400
    },
    {
      "epoch": 9.536388140161725,
      "grad_norm": 0.8970053791999817,
      "learning_rate": 2.806691851052347e-05,
      "loss": 3.1343,
      "step": 88450
    },
    {
      "epoch": 9.54177897574124,
      "grad_norm": 0.9087472558021545,
      "learning_rate": 2.774311926605504e-05,
      "loss": 3.1285,
      "step": 88500
    },
    {
      "epoch": 9.547169811320755,
      "grad_norm": 0.8881940245628357,
      "learning_rate": 2.7419320021586613e-05,
      "loss": 3.1207,
      "step": 88550
    },
    {
      "epoch": 9.55256064690027,
      "grad_norm": 0.8824938535690308,
      "learning_rate": 2.7095520777118187e-05,
      "loss": 3.1361,
      "step": 88600
    },
    {
      "epoch": 9.557951482479785,
      "grad_norm": 0.878555178642273,
      "learning_rate": 2.6771721532649758e-05,
      "loss": 3.107,
      "step": 88650
    },
    {
      "epoch": 9.563342318059298,
      "grad_norm": 0.8404247760772705,
      "learning_rate": 2.6447922288181325e-05,
      "loss": 3.1298,
      "step": 88700
    },
    {
      "epoch": 9.568733153638814,
      "grad_norm": 0.8609907627105713,
      "learning_rate": 2.6124123043712896e-05,
      "loss": 3.1371,
      "step": 88750
    },
    {
      "epoch": 9.574123989218329,
      "grad_norm": 0.9168536067008972,
      "learning_rate": 2.5800323799244467e-05,
      "loss": 3.1142,
      "step": 88800
    },
    {
      "epoch": 9.579514824797844,
      "grad_norm": 0.9115109443664551,
      "learning_rate": 2.5476524554776038e-05,
      "loss": 3.1325,
      "step": 88850
    },
    {
      "epoch": 9.584905660377359,
      "grad_norm": 0.8815240263938904,
      "learning_rate": 2.5159201295196976e-05,
      "loss": 3.1531,
      "step": 88900
    },
    {
      "epoch": 9.590296495956874,
      "grad_norm": 0.8526226282119751,
      "learning_rate": 2.4835402050728544e-05,
      "loss": 3.1387,
      "step": 88950
    },
    {
      "epoch": 9.595687331536388,
      "grad_norm": 0.8618078231811523,
      "learning_rate": 2.4511602806260115e-05,
      "loss": 3.1304,
      "step": 89000
    },
    {
      "epoch": 9.595687331536388,
      "eval_accuracy": 0.3937702345631138,
      "eval_loss": 3.305745840072632,
      "eval_runtime": 183.6824,
      "eval_samples_per_second": 98.055,
      "eval_steps_per_second": 6.13,
      "step": 89000
    },
    {
      "epoch": 9.601078167115903,
      "grad_norm": 0.8251466155052185,
      "learning_rate": 2.4187803561791686e-05,
      "loss": 3.127,
      "step": 89050
    },
    {
      "epoch": 9.606469002695418,
      "grad_norm": 0.849942684173584,
      "learning_rate": 2.3864004317323257e-05,
      "loss": 3.1325,
      "step": 89100
    },
    {
      "epoch": 9.611859838274933,
      "grad_norm": 0.9657033681869507,
      "learning_rate": 2.354020507285483e-05,
      "loss": 3.1566,
      "step": 89150
    },
    {
      "epoch": 9.617250673854448,
      "grad_norm": 0.8909383416175842,
      "learning_rate": 2.3216405828386402e-05,
      "loss": 3.1215,
      "step": 89200
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.9422598481178284,
      "learning_rate": 2.289260658391797e-05,
      "loss": 3.1439,
      "step": 89250
    },
    {
      "epoch": 9.628032345013477,
      "grad_norm": 0.8946119546890259,
      "learning_rate": 2.256880733944954e-05,
      "loss": 3.1319,
      "step": 89300
    },
    {
      "epoch": 9.633423180592992,
      "grad_norm": 0.8867332935333252,
      "learning_rate": 2.224500809498111e-05,
      "loss": 3.1286,
      "step": 89350
    },
    {
      "epoch": 9.638814016172507,
      "grad_norm": 0.8961672782897949,
      "learning_rate": 2.1921208850512682e-05,
      "loss": 3.1327,
      "step": 89400
    },
    {
      "epoch": 9.644204851752022,
      "grad_norm": 0.8897280097007751,
      "learning_rate": 2.1597409606044253e-05,
      "loss": 3.1312,
      "step": 89450
    },
    {
      "epoch": 9.649595687331537,
      "grad_norm": 0.8461384773254395,
      "learning_rate": 2.127361036157582e-05,
      "loss": 3.1449,
      "step": 89500
    },
    {
      "epoch": 9.65498652291105,
      "grad_norm": 0.8975701332092285,
      "learning_rate": 2.094981111710739e-05,
      "loss": 3.1314,
      "step": 89550
    },
    {
      "epoch": 9.660377358490566,
      "grad_norm": 0.8743809461593628,
      "learning_rate": 2.0626011872638962e-05,
      "loss": 3.1226,
      "step": 89600
    },
    {
      "epoch": 9.66576819407008,
      "grad_norm": 0.8966118693351746,
      "learning_rate": 2.0302212628170533e-05,
      "loss": 3.1243,
      "step": 89650
    },
    {
      "epoch": 9.671159029649596,
      "grad_norm": 0.9084198474884033,
      "learning_rate": 1.99784133837021e-05,
      "loss": 3.126,
      "step": 89700
    },
    {
      "epoch": 9.676549865229111,
      "grad_norm": 0.8550707697868347,
      "learning_rate": 1.965461413923367e-05,
      "loss": 3.1412,
      "step": 89750
    },
    {
      "epoch": 9.681940700808624,
      "grad_norm": 0.883452832698822,
      "learning_rate": 1.9330814894765242e-05,
      "loss": 3.1228,
      "step": 89800
    },
    {
      "epoch": 9.68733153638814,
      "grad_norm": 0.8623883724212646,
      "learning_rate": 1.9007015650296813e-05,
      "loss": 3.1269,
      "step": 89850
    },
    {
      "epoch": 9.692722371967655,
      "grad_norm": 0.8669401407241821,
      "learning_rate": 1.8683216405828384e-05,
      "loss": 3.1452,
      "step": 89900
    },
    {
      "epoch": 9.69811320754717,
      "grad_norm": 0.8647443652153015,
      "learning_rate": 1.8359417161359955e-05,
      "loss": 3.1487,
      "step": 89950
    },
    {
      "epoch": 9.703504043126685,
      "grad_norm": 0.8886281847953796,
      "learning_rate": 1.8035617916891526e-05,
      "loss": 3.1329,
      "step": 90000
    },
    {
      "epoch": 9.703504043126685,
      "eval_accuracy": 0.39417724828086687,
      "eval_loss": 3.3030855655670166,
      "eval_runtime": 183.5836,
      "eval_samples_per_second": 98.108,
      "eval_steps_per_second": 6.133,
      "step": 90000
    },
    {
      "epoch": 9.7088948787062,
      "grad_norm": 0.8624467849731445,
      "learning_rate": 1.7711818672423097e-05,
      "loss": 3.1197,
      "step": 90050
    },
    {
      "epoch": 9.714285714285714,
      "grad_norm": 0.9024360179901123,
      "learning_rate": 1.7388019427954664e-05,
      "loss": 3.1357,
      "step": 90100
    },
    {
      "epoch": 9.719676549865229,
      "grad_norm": 0.8829125761985779,
      "learning_rate": 1.706422018348624e-05,
      "loss": 3.1514,
      "step": 90150
    },
    {
      "epoch": 9.725067385444744,
      "grad_norm": 0.8832122683525085,
      "learning_rate": 1.674042093901781e-05,
      "loss": 3.1308,
      "step": 90200
    },
    {
      "epoch": 9.730458221024259,
      "grad_norm": 0.8811697959899902,
      "learning_rate": 1.6416621694549377e-05,
      "loss": 3.1369,
      "step": 90250
    },
    {
      "epoch": 9.735849056603774,
      "grad_norm": 0.9028815031051636,
      "learning_rate": 1.6092822450080948e-05,
      "loss": 3.1302,
      "step": 90300
    },
    {
      "epoch": 9.74123989218329,
      "grad_norm": 0.8779157996177673,
      "learning_rate": 1.576902320561252e-05,
      "loss": 3.1376,
      "step": 90350
    },
    {
      "epoch": 9.746630727762803,
      "grad_norm": 0.8744044899940491,
      "learning_rate": 1.544522396114409e-05,
      "loss": 3.1471,
      "step": 90400
    },
    {
      "epoch": 9.752021563342318,
      "grad_norm": 0.8857881426811218,
      "learning_rate": 1.5121424716675659e-05,
      "loss": 3.1234,
      "step": 90450
    },
    {
      "epoch": 9.757412398921833,
      "grad_norm": 0.8945528864860535,
      "learning_rate": 1.479762547220723e-05,
      "loss": 3.1355,
      "step": 90500
    },
    {
      "epoch": 9.762803234501348,
      "grad_norm": 0.9267755746841431,
      "learning_rate": 1.4473826227738802e-05,
      "loss": 3.152,
      "step": 90550
    },
    {
      "epoch": 9.768194070080863,
      "grad_norm": 0.8681300282478333,
      "learning_rate": 1.4150026983270371e-05,
      "loss": 3.1292,
      "step": 90600
    },
    {
      "epoch": 9.773584905660378,
      "grad_norm": 0.8631065487861633,
      "learning_rate": 1.3826227738801942e-05,
      "loss": 3.1229,
      "step": 90650
    },
    {
      "epoch": 9.778975741239892,
      "grad_norm": 0.8668704032897949,
      "learning_rate": 1.3502428494333512e-05,
      "loss": 3.1076,
      "step": 90700
    },
    {
      "epoch": 9.784366576819407,
      "grad_norm": 0.8805832862854004,
      "learning_rate": 1.3178629249865082e-05,
      "loss": 3.1499,
      "step": 90750
    },
    {
      "epoch": 9.789757412398922,
      "grad_norm": 0.9205412864685059,
      "learning_rate": 1.2854830005396653e-05,
      "loss": 3.1072,
      "step": 90800
    },
    {
      "epoch": 9.795148247978437,
      "grad_norm": 0.8832406997680664,
      "learning_rate": 1.2531030760928222e-05,
      "loss": 3.1221,
      "step": 90850
    },
    {
      "epoch": 9.800539083557952,
      "grad_norm": 0.8675433397293091,
      "learning_rate": 1.2207231516459793e-05,
      "loss": 3.1423,
      "step": 90900
    },
    {
      "epoch": 9.805929919137466,
      "grad_norm": 0.9155154824256897,
      "learning_rate": 1.1883432271991366e-05,
      "loss": 3.1176,
      "step": 90950
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.8952852487564087,
      "learning_rate": 1.1559633027522935e-05,
      "loss": 3.133,
      "step": 91000
    },
    {
      "epoch": 9.81132075471698,
      "eval_accuracy": 0.3943811897512946,
      "eval_loss": 3.3014745712280273,
      "eval_runtime": 183.7431,
      "eval_samples_per_second": 98.023,
      "eval_steps_per_second": 6.128,
      "step": 91000
    },
    {
      "epoch": 9.816711590296496,
      "grad_norm": 0.8473813533782959,
      "learning_rate": 1.1235833783054506e-05,
      "loss": 3.1286,
      "step": 91050
    },
    {
      "epoch": 9.822102425876011,
      "grad_norm": 0.8618873357772827,
      "learning_rate": 1.0912034538586075e-05,
      "loss": 3.1257,
      "step": 91100
    },
    {
      "epoch": 9.827493261455526,
      "grad_norm": 0.8717328906059265,
      "learning_rate": 1.0588235294117646e-05,
      "loss": 3.1372,
      "step": 91150
    },
    {
      "epoch": 9.832884097035041,
      "grad_norm": 0.8867208957672119,
      "learning_rate": 1.0264436049649217e-05,
      "loss": 3.1372,
      "step": 91200
    },
    {
      "epoch": 9.838274932614555,
      "grad_norm": 0.8444369435310364,
      "learning_rate": 9.940636805180786e-06,
      "loss": 3.1292,
      "step": 91250
    },
    {
      "epoch": 9.84366576819407,
      "grad_norm": 0.8623387813568115,
      "learning_rate": 9.616837560712357e-06,
      "loss": 3.1277,
      "step": 91300
    },
    {
      "epoch": 9.849056603773585,
      "grad_norm": 0.9045534729957581,
      "learning_rate": 9.293038316243928e-06,
      "loss": 3.1321,
      "step": 91350
    },
    {
      "epoch": 9.8544474393531,
      "grad_norm": 0.8703294992446899,
      "learning_rate": 8.969239071775497e-06,
      "loss": 3.1132,
      "step": 91400
    },
    {
      "epoch": 9.859838274932615,
      "grad_norm": 0.8696820139884949,
      "learning_rate": 8.64543982730707e-06,
      "loss": 3.1257,
      "step": 91450
    },
    {
      "epoch": 9.865229110512129,
      "grad_norm": 0.8719615340232849,
      "learning_rate": 8.321640582838639e-06,
      "loss": 3.1223,
      "step": 91500
    },
    {
      "epoch": 9.870619946091644,
      "grad_norm": 0.8584866523742676,
      "learning_rate": 7.99784133837021e-06,
      "loss": 3.1242,
      "step": 91550
    },
    {
      "epoch": 9.876010781671159,
      "grad_norm": 0.855100691318512,
      "learning_rate": 7.674042093901779e-06,
      "loss": 3.13,
      "step": 91600
    },
    {
      "epoch": 9.881401617250674,
      "grad_norm": 0.8618912696838379,
      "learning_rate": 7.350242849433351e-06,
      "loss": 3.1491,
      "step": 91650
    },
    {
      "epoch": 9.88679245283019,
      "grad_norm": 0.9182708263397217,
      "learning_rate": 7.026443604964921e-06,
      "loss": 3.1439,
      "step": 91700
    },
    {
      "epoch": 9.892183288409704,
      "grad_norm": 0.911716103553772,
      "learning_rate": 6.702644360496492e-06,
      "loss": 3.1325,
      "step": 91750
    },
    {
      "epoch": 9.897574123989218,
      "grad_norm": 0.888717770576477,
      "learning_rate": 6.378845116028062e-06,
      "loss": 3.1463,
      "step": 91800
    },
    {
      "epoch": 9.902964959568733,
      "grad_norm": 0.8845711946487427,
      "learning_rate": 6.055045871559633e-06,
      "loss": 3.1092,
      "step": 91850
    },
    {
      "epoch": 9.908355795148248,
      "grad_norm": 0.8652920722961426,
      "learning_rate": 5.731246627091203e-06,
      "loss": 3.1351,
      "step": 91900
    },
    {
      "epoch": 9.913746630727763,
      "grad_norm": 0.901452362537384,
      "learning_rate": 5.407447382622774e-06,
      "loss": 3.1131,
      "step": 91950
    },
    {
      "epoch": 9.919137466307278,
      "grad_norm": 0.8996922373771667,
      "learning_rate": 5.083648138154344e-06,
      "loss": 3.1425,
      "step": 92000
    },
    {
      "epoch": 9.919137466307278,
      "eval_accuracy": 0.3946086002454903,
      "eval_loss": 3.2996339797973633,
      "eval_runtime": 183.757,
      "eval_samples_per_second": 98.015,
      "eval_steps_per_second": 6.128,
      "step": 92000
    },
    {
      "epoch": 9.924528301886792,
      "grad_norm": 0.8938733339309692,
      "learning_rate": 4.7598488936859145e-06,
      "loss": 3.1449,
      "step": 92050
    },
    {
      "epoch": 9.929919137466307,
      "grad_norm": 0.855878472328186,
      "learning_rate": 4.4360496492174846e-06,
      "loss": 3.127,
      "step": 92100
    },
    {
      "epoch": 9.935309973045822,
      "grad_norm": 0.9077852964401245,
      "learning_rate": 4.112250404749055e-06,
      "loss": 3.1375,
      "step": 92150
    },
    {
      "epoch": 9.940700808625337,
      "grad_norm": 0.8704142570495605,
      "learning_rate": 3.788451160280626e-06,
      "loss": 3.1189,
      "step": 92200
    },
    {
      "epoch": 9.946091644204852,
      "grad_norm": 0.8859520554542542,
      "learning_rate": 3.464651915812196e-06,
      "loss": 3.1297,
      "step": 92250
    },
    {
      "epoch": 9.951482479784367,
      "grad_norm": 0.8482578992843628,
      "learning_rate": 3.1408526713437664e-06,
      "loss": 3.1209,
      "step": 92300
    },
    {
      "epoch": 9.95687331536388,
      "grad_norm": 0.8768433332443237,
      "learning_rate": 2.817053426875337e-06,
      "loss": 3.1263,
      "step": 92350
    },
    {
      "epoch": 9.962264150943396,
      "grad_norm": 0.8390297889709473,
      "learning_rate": 2.4932541824069074e-06,
      "loss": 3.1259,
      "step": 92400
    },
    {
      "epoch": 9.967654986522911,
      "grad_norm": 0.9365578889846802,
      "learning_rate": 2.169454937938478e-06,
      "loss": 3.128,
      "step": 92450
    },
    {
      "epoch": 9.973045822102426,
      "grad_norm": 0.8617887496948242,
      "learning_rate": 1.8456556934700485e-06,
      "loss": 3.1347,
      "step": 92500
    },
    {
      "epoch": 9.978436657681941,
      "grad_norm": 0.9396256804466248,
      "learning_rate": 1.521856449001619e-06,
      "loss": 3.1253,
      "step": 92550
    },
    {
      "epoch": 9.983827493261456,
      "grad_norm": 0.9035614728927612,
      "learning_rate": 1.1980572045331894e-06,
      "loss": 3.1229,
      "step": 92600
    },
    {
      "epoch": 9.98921832884097,
      "grad_norm": 0.840107798576355,
      "learning_rate": 8.742579600647598e-07,
      "loss": 3.1141,
      "step": 92650
    },
    {
      "epoch": 9.994609164420485,
      "grad_norm": 0.9073900580406189,
      "learning_rate": 5.504587155963303e-07,
      "loss": 3.1322,
      "step": 92700
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.8777025938034058,
      "learning_rate": 2.2665947112790068e-07,
      "loss": 3.1244,
      "step": 92750
    },
    {
      "epoch": 10.0,
      "step": 92750,
      "total_flos": 7.75449427968e+17,
      "train_loss": 3.4578574852133697,
      "train_runtime": 79538.6235,
      "train_samples_per_second": 37.312,
      "train_steps_per_second": 1.166
    }
  ],
  "logging_steps": 50,
  "max_steps": 92750,
  "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": 7.75449427968e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}