{
  "best_metric": 3.299715518951416,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_2000_634/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": 1.3841798305511475,
      "learning_rate": 0.0003,
      "loss": 8.6258,
      "step": 50
    },
    {
      "epoch": 0.01078167115902965,
      "grad_norm": 2.7142674922943115,
      "learning_rate": 0.0006,
      "loss": 6.9518,
      "step": 100
    },
    {
      "epoch": 0.016172506738544475,
      "grad_norm": 1.7501500844955444,
      "learning_rate": 0.0005996762007555315,
      "loss": 6.4766,
      "step": 150
    },
    {
      "epoch": 0.0215633423180593,
      "grad_norm": 2.089492082595825,
      "learning_rate": 0.000599352401511063,
      "loss": 6.2242,
      "step": 200
    },
    {
      "epoch": 0.026954177897574125,
      "grad_norm": 1.0795507431030273,
      "learning_rate": 0.0005990286022665946,
      "loss": 6.0814,
      "step": 250
    },
    {
      "epoch": 0.03234501347708895,
      "grad_norm": 2.468144655227661,
      "learning_rate": 0.0005987048030221263,
      "loss": 5.9861,
      "step": 300
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 1.4130438566207886,
      "learning_rate": 0.0005983810037776578,
      "loss": 5.8758,
      "step": 350
    },
    {
      "epoch": 0.0431266846361186,
      "grad_norm": 1.5332931280136108,
      "learning_rate": 0.0005980572045331894,
      "loss": 5.7848,
      "step": 400
    },
    {
      "epoch": 0.04851752021563342,
      "grad_norm": 1.451593279838562,
      "learning_rate": 0.0005977334052887209,
      "loss": 5.713,
      "step": 450
    },
    {
      "epoch": 0.05390835579514825,
      "grad_norm": 1.3739091157913208,
      "learning_rate": 0.0005974096060442526,
      "loss": 5.6508,
      "step": 500
    },
    {
      "epoch": 0.05929919137466307,
      "grad_norm": 1.2970751523971558,
      "learning_rate": 0.0005970858067997841,
      "loss": 5.5723,
      "step": 550
    },
    {
      "epoch": 0.0646900269541779,
      "grad_norm": 1.2038410902023315,
      "learning_rate": 0.0005967620075553157,
      "loss": 5.5093,
      "step": 600
    },
    {
      "epoch": 0.07008086253369272,
      "grad_norm": 1.2364015579223633,
      "learning_rate": 0.0005964382083108472,
      "loss": 5.4167,
      "step": 650
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 1.1776434183120728,
      "learning_rate": 0.0005961144090663788,
      "loss": 5.3687,
      "step": 700
    },
    {
      "epoch": 0.08086253369272237,
      "grad_norm": 1.2404123544692993,
      "learning_rate": 0.0005957906098219104,
      "loss": 5.2965,
      "step": 750
    },
    {
      "epoch": 0.0862533692722372,
      "grad_norm": 0.8565037250518799,
      "learning_rate": 0.0005954668105774419,
      "loss": 5.2536,
      "step": 800
    },
    {
      "epoch": 0.09164420485175202,
      "grad_norm": 0.8625679612159729,
      "learning_rate": 0.0005951430113329735,
      "loss": 5.1917,
      "step": 850
    },
    {
      "epoch": 0.09703504043126684,
      "grad_norm": 1.4676177501678467,
      "learning_rate": 0.0005948192120885051,
      "loss": 5.1705,
      "step": 900
    },
    {
      "epoch": 0.10242587601078167,
      "grad_norm": 1.3620243072509766,
      "learning_rate": 0.0005944954128440366,
      "loss": 5.129,
      "step": 950
    },
    {
      "epoch": 0.1078167115902965,
      "grad_norm": 1.802571177482605,
      "learning_rate": 0.0005941716135995682,
      "loss": 5.0781,
      "step": 1000
    },
    {
      "epoch": 0.1078167115902965,
      "eval_accuracy": 0.22607515020446844,
      "eval_loss": 5.030185699462891,
      "eval_runtime": 184.9577,
      "eval_samples_per_second": 97.379,
      "eval_steps_per_second": 6.088,
      "step": 1000
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 1.0640990734100342,
      "learning_rate": 0.0005938478143550997,
      "loss": 5.0384,
      "step": 1050
    },
    {
      "epoch": 0.11859838274932614,
      "grad_norm": 0.9512579441070557,
      "learning_rate": 0.0005935240151106314,
      "loss": 5.01,
      "step": 1100
    },
    {
      "epoch": 0.12398921832884097,
      "grad_norm": 1.4336494207382202,
      "learning_rate": 0.0005932002158661629,
      "loss": 4.986,
      "step": 1150
    },
    {
      "epoch": 0.1293800539083558,
      "grad_norm": 0.986595094203949,
      "learning_rate": 0.0005928764166216945,
      "loss": 4.9237,
      "step": 1200
    },
    {
      "epoch": 0.1347708894878706,
      "grad_norm": 0.9924365878105164,
      "learning_rate": 0.000592552617377226,
      "loss": 4.8993,
      "step": 1250
    },
    {
      "epoch": 0.14016172506738545,
      "grad_norm": 1.405576229095459,
      "learning_rate": 0.0005922288181327577,
      "loss": 4.8778,
      "step": 1300
    },
    {
      "epoch": 0.14555256064690028,
      "grad_norm": 0.7532362341880798,
      "learning_rate": 0.0005919050188882893,
      "loss": 4.8412,
      "step": 1350
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 0.9290658235549927,
      "learning_rate": 0.0005915812196438207,
      "loss": 4.8361,
      "step": 1400
    },
    {
      "epoch": 0.15633423180592992,
      "grad_norm": 1.0212785005569458,
      "learning_rate": 0.0005912574203993524,
      "loss": 4.8431,
      "step": 1450
    },
    {
      "epoch": 0.16172506738544473,
      "grad_norm": 1.3435090780258179,
      "learning_rate": 0.0005909336211548839,
      "loss": 4.801,
      "step": 1500
    },
    {
      "epoch": 0.16711590296495957,
      "grad_norm": 1.0386897325515747,
      "learning_rate": 0.0005906098219104155,
      "loss": 4.7558,
      "step": 1550
    },
    {
      "epoch": 0.1725067385444744,
      "grad_norm": 0.8441452383995056,
      "learning_rate": 0.000590286022665947,
      "loss": 4.749,
      "step": 1600
    },
    {
      "epoch": 0.1778975741239892,
      "grad_norm": 0.9853036999702454,
      "learning_rate": 0.0005899622234214787,
      "loss": 4.7201,
      "step": 1650
    },
    {
      "epoch": 0.18328840970350405,
      "grad_norm": 0.9956970810890198,
      "learning_rate": 0.0005896384241770102,
      "loss": 4.6933,
      "step": 1700
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 0.8019234538078308,
      "learning_rate": 0.0005893146249325418,
      "loss": 4.7021,
      "step": 1750
    },
    {
      "epoch": 0.1940700808625337,
      "grad_norm": 0.8837598562240601,
      "learning_rate": 0.0005889908256880733,
      "loss": 4.6704,
      "step": 1800
    },
    {
      "epoch": 0.19946091644204852,
      "grad_norm": 1.046995997428894,
      "learning_rate": 0.0005886670264436049,
      "loss": 4.6374,
      "step": 1850
    },
    {
      "epoch": 0.20485175202156333,
      "grad_norm": 1.0166348218917847,
      "learning_rate": 0.0005883432271991365,
      "loss": 4.6279,
      "step": 1900
    },
    {
      "epoch": 0.21024258760107817,
      "grad_norm": 0.8063094615936279,
      "learning_rate": 0.0005880194279546681,
      "loss": 4.6256,
      "step": 1950
    },
    {
      "epoch": 0.215633423180593,
      "grad_norm": 0.9398548603057861,
      "learning_rate": 0.0005876956287101996,
      "loss": 4.5763,
      "step": 2000
    },
    {
      "epoch": 0.215633423180593,
      "eval_accuracy": 0.2714252357849155,
      "eval_loss": 4.502427577972412,
      "eval_runtime": 183.5351,
      "eval_samples_per_second": 98.134,
      "eval_steps_per_second": 6.135,
      "step": 2000
    },
    {
      "epoch": 0.2210242587601078,
      "grad_norm": 0.980785608291626,
      "learning_rate": 0.0005873718294657312,
      "loss": 4.5575,
      "step": 2050
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 0.8288772106170654,
      "learning_rate": 0.0005870480302212628,
      "loss": 4.5462,
      "step": 2100
    },
    {
      "epoch": 0.23180592991913745,
      "grad_norm": 0.9006871581077576,
      "learning_rate": 0.0005867242309767943,
      "loss": 4.5289,
      "step": 2150
    },
    {
      "epoch": 0.2371967654986523,
      "grad_norm": 0.7568821310997009,
      "learning_rate": 0.0005864004317323259,
      "loss": 4.4877,
      "step": 2200
    },
    {
      "epoch": 0.24258760107816713,
      "grad_norm": 0.8699207305908203,
      "learning_rate": 0.0005860766324878575,
      "loss": 4.5022,
      "step": 2250
    },
    {
      "epoch": 0.24797843665768193,
      "grad_norm": 0.9681375026702881,
      "learning_rate": 0.000585752833243389,
      "loss": 4.4781,
      "step": 2300
    },
    {
      "epoch": 0.25336927223719674,
      "grad_norm": 1.0356428623199463,
      "learning_rate": 0.0005854290339989206,
      "loss": 4.4547,
      "step": 2350
    },
    {
      "epoch": 0.2587601078167116,
      "grad_norm": 1.0217968225479126,
      "learning_rate": 0.0005851052347544521,
      "loss": 4.4496,
      "step": 2400
    },
    {
      "epoch": 0.2641509433962264,
      "grad_norm": 0.9390143752098083,
      "learning_rate": 0.0005847814355099838,
      "loss": 4.4324,
      "step": 2450
    },
    {
      "epoch": 0.2695417789757412,
      "grad_norm": 0.8502791523933411,
      "learning_rate": 0.0005844576362655154,
      "loss": 4.4238,
      "step": 2500
    },
    {
      "epoch": 0.2749326145552561,
      "grad_norm": 0.9871562719345093,
      "learning_rate": 0.0005841338370210469,
      "loss": 4.3975,
      "step": 2550
    },
    {
      "epoch": 0.2803234501347709,
      "grad_norm": 0.9953110814094543,
      "learning_rate": 0.0005838100377765785,
      "loss": 4.3725,
      "step": 2600
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.8370280861854553,
      "learning_rate": 0.0005834862385321101,
      "loss": 4.3714,
      "step": 2650
    },
    {
      "epoch": 0.29110512129380056,
      "grad_norm": 0.8906695246696472,
      "learning_rate": 0.0005831624392876417,
      "loss": 4.3636,
      "step": 2700
    },
    {
      "epoch": 0.29649595687331537,
      "grad_norm": 0.8926903009414673,
      "learning_rate": 0.0005828386400431731,
      "loss": 4.3712,
      "step": 2750
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 0.6380210518836975,
      "learning_rate": 0.0005825148407987048,
      "loss": 4.3652,
      "step": 2800
    },
    {
      "epoch": 0.30727762803234504,
      "grad_norm": 0.9012649655342102,
      "learning_rate": 0.0005821910415542363,
      "loss": 4.3264,
      "step": 2850
    },
    {
      "epoch": 0.31266846361185985,
      "grad_norm": 0.8436964750289917,
      "learning_rate": 0.0005818672423097679,
      "loss": 4.3127,
      "step": 2900
    },
    {
      "epoch": 0.31805929919137466,
      "grad_norm": 0.8248408436775208,
      "learning_rate": 0.0005815434430652994,
      "loss": 4.3424,
      "step": 2950
    },
    {
      "epoch": 0.32345013477088946,
      "grad_norm": 0.7463982105255127,
      "learning_rate": 0.0005812196438208311,
      "loss": 4.3107,
      "step": 3000
    },
    {
      "epoch": 0.32345013477088946,
      "eval_accuracy": 0.2991547565898248,
      "eval_loss": 4.231179237365723,
      "eval_runtime": 183.5046,
      "eval_samples_per_second": 98.15,
      "eval_steps_per_second": 6.136,
      "step": 3000
    },
    {
      "epoch": 0.3288409703504043,
      "grad_norm": 0.7356308102607727,
      "learning_rate": 0.0005808958445763626,
      "loss": 4.3028,
      "step": 3050
    },
    {
      "epoch": 0.33423180592991913,
      "grad_norm": 0.7332613468170166,
      "learning_rate": 0.0005805720453318942,
      "loss": 4.281,
      "step": 3100
    },
    {
      "epoch": 0.33962264150943394,
      "grad_norm": 0.9393066763877869,
      "learning_rate": 0.0005802482460874257,
      "loss": 4.2707,
      "step": 3150
    },
    {
      "epoch": 0.3450134770889488,
      "grad_norm": 0.7228361964225769,
      "learning_rate": 0.0005799244468429573,
      "loss": 4.2769,
      "step": 3200
    },
    {
      "epoch": 0.3504043126684636,
      "grad_norm": 0.7841880917549133,
      "learning_rate": 0.0005796006475984889,
      "loss": 4.2499,
      "step": 3250
    },
    {
      "epoch": 0.3557951482479784,
      "grad_norm": 0.7926186919212341,
      "learning_rate": 0.0005792768483540205,
      "loss": 4.2548,
      "step": 3300
    },
    {
      "epoch": 0.3611859838274933,
      "grad_norm": 0.8690990805625916,
      "learning_rate": 0.000578953049109552,
      "loss": 4.254,
      "step": 3350
    },
    {
      "epoch": 0.3665768194070081,
      "grad_norm": 0.7845708727836609,
      "learning_rate": 0.0005786292498650836,
      "loss": 4.2325,
      "step": 3400
    },
    {
      "epoch": 0.3719676549865229,
      "grad_norm": 0.8615543246269226,
      "learning_rate": 0.0005783054506206152,
      "loss": 4.2247,
      "step": 3450
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 0.7429792284965515,
      "learning_rate": 0.0005779816513761467,
      "loss": 4.216,
      "step": 3500
    },
    {
      "epoch": 0.38274932614555257,
      "grad_norm": 0.7726993560791016,
      "learning_rate": 0.0005776578521316782,
      "loss": 4.1962,
      "step": 3550
    },
    {
      "epoch": 0.3881401617250674,
      "grad_norm": 0.7264295220375061,
      "learning_rate": 0.0005773340528872099,
      "loss": 4.2044,
      "step": 3600
    },
    {
      "epoch": 0.3935309973045822,
      "grad_norm": 0.7918074131011963,
      "learning_rate": 0.0005770102536427414,
      "loss": 4.219,
      "step": 3650
    },
    {
      "epoch": 0.39892183288409705,
      "grad_norm": 0.7423751950263977,
      "learning_rate": 0.000576686454398273,
      "loss": 4.2045,
      "step": 3700
    },
    {
      "epoch": 0.40431266846361186,
      "grad_norm": 0.6395682096481323,
      "learning_rate": 0.0005763626551538045,
      "loss": 4.1897,
      "step": 3750
    },
    {
      "epoch": 0.40970350404312667,
      "grad_norm": 0.6632921695709229,
      "learning_rate": 0.0005760388559093362,
      "loss": 4.182,
      "step": 3800
    },
    {
      "epoch": 0.41509433962264153,
      "grad_norm": 0.6251309514045715,
      "learning_rate": 0.0005757150566648678,
      "loss": 4.1772,
      "step": 3850
    },
    {
      "epoch": 0.42048517520215634,
      "grad_norm": 0.8533205986022949,
      "learning_rate": 0.0005753912574203993,
      "loss": 4.1699,
      "step": 3900
    },
    {
      "epoch": 0.42587601078167114,
      "grad_norm": 0.707554817199707,
      "learning_rate": 0.0005750674581759309,
      "loss": 4.1668,
      "step": 3950
    },
    {
      "epoch": 0.431266846361186,
      "grad_norm": 0.7947620153427124,
      "learning_rate": 0.0005747436589314624,
      "loss": 4.1524,
      "step": 4000
    },
    {
      "epoch": 0.431266846361186,
      "eval_accuracy": 0.3127696153503061,
      "eval_loss": 4.084148406982422,
      "eval_runtime": 183.63,
      "eval_samples_per_second": 98.083,
      "eval_steps_per_second": 6.132,
      "step": 4000
    },
    {
      "epoch": 0.4366576819407008,
      "grad_norm": 0.5997743606567383,
      "learning_rate": 0.0005744198596869941,
      "loss": 4.1636,
      "step": 4050
    },
    {
      "epoch": 0.4420485175202156,
      "grad_norm": 0.7036544680595398,
      "learning_rate": 0.0005740960604425255,
      "loss": 4.156,
      "step": 4100
    },
    {
      "epoch": 0.4474393530997305,
      "grad_norm": 0.7143358588218689,
      "learning_rate": 0.0005737722611980572,
      "loss": 4.1384,
      "step": 4150
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 0.6587009429931641,
      "learning_rate": 0.0005734484619535887,
      "loss": 4.113,
      "step": 4200
    },
    {
      "epoch": 0.4582210242587601,
      "grad_norm": 0.704142153263092,
      "learning_rate": 0.0005731246627091203,
      "loss": 4.1487,
      "step": 4250
    },
    {
      "epoch": 0.4636118598382749,
      "grad_norm": 0.5820618867874146,
      "learning_rate": 0.0005728008634646518,
      "loss": 4.1219,
      "step": 4300
    },
    {
      "epoch": 0.46900269541778977,
      "grad_norm": 0.6929563879966736,
      "learning_rate": 0.0005724770642201835,
      "loss": 4.1263,
      "step": 4350
    },
    {
      "epoch": 0.4743935309973046,
      "grad_norm": 0.7341722846031189,
      "learning_rate": 0.000572153264975715,
      "loss": 4.1214,
      "step": 4400
    },
    {
      "epoch": 0.4797843665768194,
      "grad_norm": 0.5978211164474487,
      "learning_rate": 0.0005718294657312466,
      "loss": 4.0946,
      "step": 4450
    },
    {
      "epoch": 0.48517520215633425,
      "grad_norm": 0.6326689124107361,
      "learning_rate": 0.0005715056664867781,
      "loss": 4.0929,
      "step": 4500
    },
    {
      "epoch": 0.49056603773584906,
      "grad_norm": 0.6336254477500916,
      "learning_rate": 0.0005711818672423097,
      "loss": 4.0867,
      "step": 4550
    },
    {
      "epoch": 0.49595687331536387,
      "grad_norm": 0.6374770998954773,
      "learning_rate": 0.0005708580679978413,
      "loss": 4.0936,
      "step": 4600
    },
    {
      "epoch": 0.5013477088948787,
      "grad_norm": 0.7158097624778748,
      "learning_rate": 0.0005705342687533729,
      "loss": 4.0868,
      "step": 4650
    },
    {
      "epoch": 0.5067385444743935,
      "grad_norm": 0.6559752225875854,
      "learning_rate": 0.0005702104695089044,
      "loss": 4.0694,
      "step": 4700
    },
    {
      "epoch": 0.5121293800539084,
      "grad_norm": 0.6569370627403259,
      "learning_rate": 0.000569886670264436,
      "loss": 4.0812,
      "step": 4750
    },
    {
      "epoch": 0.5175202156334232,
      "grad_norm": 0.6718046069145203,
      "learning_rate": 0.0005695628710199675,
      "loss": 4.0682,
      "step": 4800
    },
    {
      "epoch": 0.522911051212938,
      "grad_norm": 0.6529590487480164,
      "learning_rate": 0.0005692390717754991,
      "loss": 4.0635,
      "step": 4850
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 0.6065393686294556,
      "learning_rate": 0.0005689152725310306,
      "loss": 4.0566,
      "step": 4900
    },
    {
      "epoch": 0.5336927223719676,
      "grad_norm": 0.711345911026001,
      "learning_rate": 0.0005685914732865623,
      "loss": 4.0593,
      "step": 4950
    },
    {
      "epoch": 0.5390835579514824,
      "grad_norm": 0.6914499998092651,
      "learning_rate": 0.0005682676740420939,
      "loss": 4.0724,
      "step": 5000
    },
    {
      "epoch": 0.5390835579514824,
      "eval_accuracy": 0.32181616344262765,
      "eval_loss": 3.98882794380188,
      "eval_runtime": 183.5537,
      "eval_samples_per_second": 98.124,
      "eval_steps_per_second": 6.134,
      "step": 5000
    },
    {
      "epoch": 0.5444743935309974,
      "grad_norm": 0.6853660345077515,
      "learning_rate": 0.0005679438747976254,
      "loss": 4.0452,
      "step": 5050
    },
    {
      "epoch": 0.5498652291105122,
      "grad_norm": 0.6142929792404175,
      "learning_rate": 0.000567620075553157,
      "loss": 4.0442,
      "step": 5100
    },
    {
      "epoch": 0.555256064690027,
      "grad_norm": 0.6437095999717712,
      "learning_rate": 0.0005672962763086886,
      "loss": 4.0475,
      "step": 5150
    },
    {
      "epoch": 0.5606469002695418,
      "grad_norm": 0.7355647683143616,
      "learning_rate": 0.0005669724770642202,
      "loss": 4.0336,
      "step": 5200
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.6523656249046326,
      "learning_rate": 0.0005666486778197517,
      "loss": 4.0297,
      "step": 5250
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.6674039959907532,
      "learning_rate": 0.0005663248785752833,
      "loss": 4.0411,
      "step": 5300
    },
    {
      "epoch": 0.5768194070080862,
      "grad_norm": 0.599259078502655,
      "learning_rate": 0.0005660010793308148,
      "loss": 4.0249,
      "step": 5350
    },
    {
      "epoch": 0.5822102425876011,
      "grad_norm": 0.6059934496879578,
      "learning_rate": 0.0005656772800863465,
      "loss": 4.0416,
      "step": 5400
    },
    {
      "epoch": 0.5876010781671159,
      "grad_norm": 0.5647556185722351,
      "learning_rate": 0.0005653534808418779,
      "loss": 4.0184,
      "step": 5450
    },
    {
      "epoch": 0.5929919137466307,
      "grad_norm": 0.6619639992713928,
      "learning_rate": 0.0005650296815974096,
      "loss": 4.0231,
      "step": 5500
    },
    {
      "epoch": 0.5983827493261455,
      "grad_norm": 0.6255214810371399,
      "learning_rate": 0.0005647058823529411,
      "loss": 4.0157,
      "step": 5550
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.5285488963127136,
      "learning_rate": 0.0005643820831084727,
      "loss": 4.005,
      "step": 5600
    },
    {
      "epoch": 0.6091644204851752,
      "grad_norm": 0.8316670656204224,
      "learning_rate": 0.0005640582838640042,
      "loss": 4.0181,
      "step": 5650
    },
    {
      "epoch": 0.6145552560646901,
      "grad_norm": 0.7264182567596436,
      "learning_rate": 0.0005637344846195358,
      "loss": 4.0123,
      "step": 5700
    },
    {
      "epoch": 0.6199460916442049,
      "grad_norm": 0.6089359521865845,
      "learning_rate": 0.0005634106853750674,
      "loss": 4.0165,
      "step": 5750
    },
    {
      "epoch": 0.6253369272237197,
      "grad_norm": 0.769210696220398,
      "learning_rate": 0.000563086886130599,
      "loss": 3.9894,
      "step": 5800
    },
    {
      "epoch": 0.6307277628032345,
      "grad_norm": 0.7284867763519287,
      "learning_rate": 0.0005627630868861305,
      "loss": 3.9899,
      "step": 5850
    },
    {
      "epoch": 0.6361185983827493,
      "grad_norm": 0.604616105556488,
      "learning_rate": 0.0005624392876416621,
      "loss": 3.9758,
      "step": 5900
    },
    {
      "epoch": 0.6415094339622641,
      "grad_norm": 0.685613751411438,
      "learning_rate": 0.0005621154883971937,
      "loss": 3.9787,
      "step": 5950
    },
    {
      "epoch": 0.6469002695417789,
      "grad_norm": 0.685235321521759,
      "learning_rate": 0.0005617916891527253,
      "loss": 3.9609,
      "step": 6000
    },
    {
      "epoch": 0.6469002695417789,
      "eval_accuracy": 0.32830632639045004,
      "eval_loss": 3.917071580886841,
      "eval_runtime": 183.57,
      "eval_samples_per_second": 98.115,
      "eval_steps_per_second": 6.134,
      "step": 6000
    },
    {
      "epoch": 0.6522911051212938,
      "grad_norm": 0.6348211169242859,
      "learning_rate": 0.0005614678899082568,
      "loss": 3.9973,
      "step": 6050
    },
    {
      "epoch": 0.6576819407008087,
      "grad_norm": 0.5842711925506592,
      "learning_rate": 0.0005611440906637884,
      "loss": 3.9818,
      "step": 6100
    },
    {
      "epoch": 0.6630727762803235,
      "grad_norm": 0.6440193057060242,
      "learning_rate": 0.00056082029141932,
      "loss": 3.9752,
      "step": 6150
    },
    {
      "epoch": 0.6684636118598383,
      "grad_norm": 0.6132529377937317,
      "learning_rate": 0.0005604964921748515,
      "loss": 3.9854,
      "step": 6200
    },
    {
      "epoch": 0.6738544474393531,
      "grad_norm": 0.6459117531776428,
      "learning_rate": 0.000560172692930383,
      "loss": 3.9462,
      "step": 6250
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 0.6339119672775269,
      "learning_rate": 0.0005598488936859147,
      "loss": 3.9565,
      "step": 6300
    },
    {
      "epoch": 0.6846361185983828,
      "grad_norm": 0.6577672958374023,
      "learning_rate": 0.0005595250944414463,
      "loss": 3.9497,
      "step": 6350
    },
    {
      "epoch": 0.6900269541778976,
      "grad_norm": 0.6435871720314026,
      "learning_rate": 0.0005592012951969778,
      "loss": 3.9424,
      "step": 6400
    },
    {
      "epoch": 0.6954177897574124,
      "grad_norm": 0.7063788771629333,
      "learning_rate": 0.0005588774959525094,
      "loss": 3.9694,
      "step": 6450
    },
    {
      "epoch": 0.7008086253369272,
      "grad_norm": 0.5613532662391663,
      "learning_rate": 0.000558553696708041,
      "loss": 3.9543,
      "step": 6500
    },
    {
      "epoch": 0.706199460916442,
      "grad_norm": 0.559969425201416,
      "learning_rate": 0.0005582298974635726,
      "loss": 3.9512,
      "step": 6550
    },
    {
      "epoch": 0.7115902964959568,
      "grad_norm": 0.58429354429245,
      "learning_rate": 0.0005579060982191041,
      "loss": 3.9474,
      "step": 6600
    },
    {
      "epoch": 0.7169811320754716,
      "grad_norm": Infinity,
      "learning_rate": 0.0005575887749595251,
      "loss": 3.9414,
      "step": 6650
    },
    {
      "epoch": 0.7223719676549866,
      "grad_norm": 0.589440107345581,
      "learning_rate": 0.0005572649757150566,
      "loss": 3.9341,
      "step": 6700
    },
    {
      "epoch": 0.7277628032345014,
      "grad_norm": 0.7343171834945679,
      "learning_rate": 0.0005569411764705882,
      "loss": 3.9383,
      "step": 6750
    },
    {
      "epoch": 0.7331536388140162,
      "grad_norm": 0.6029019355773926,
      "learning_rate": 0.0005566173772261198,
      "loss": 3.9469,
      "step": 6800
    },
    {
      "epoch": 0.738544474393531,
      "grad_norm": 0.5020064115524292,
      "learning_rate": 0.0005562935779816513,
      "loss": 3.931,
      "step": 6850
    },
    {
      "epoch": 0.7439353099730458,
      "grad_norm": 0.6130984425544739,
      "learning_rate": 0.0005559697787371828,
      "loss": 3.9264,
      "step": 6900
    },
    {
      "epoch": 0.7493261455525606,
      "grad_norm": 0.6600217819213867,
      "learning_rate": 0.0005556459794927145,
      "loss": 3.9137,
      "step": 6950
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.5343778133392334,
      "learning_rate": 0.000555322180248246,
      "loss": 3.9274,
      "step": 7000
    },
    {
      "epoch": 0.7547169811320755,
      "eval_accuracy": 0.33360293736563035,
      "eval_loss": 3.858804941177368,
      "eval_runtime": 183.535,
      "eval_samples_per_second": 98.134,
      "eval_steps_per_second": 6.135,
      "step": 7000
    },
    {
      "epoch": 0.7601078167115903,
      "grad_norm": 0.6115443110466003,
      "learning_rate": 0.0005549983810037776,
      "loss": 3.931,
      "step": 7050
    },
    {
      "epoch": 0.7654986522911051,
      "grad_norm": 0.5609748959541321,
      "learning_rate": 0.0005546745817593091,
      "loss": 3.9311,
      "step": 7100
    },
    {
      "epoch": 0.77088948787062,
      "grad_norm": 0.6355463266372681,
      "learning_rate": 0.0005543507825148408,
      "loss": 3.9063,
      "step": 7150
    },
    {
      "epoch": 0.7762803234501348,
      "grad_norm": 0.6557473540306091,
      "learning_rate": 0.0005540269832703723,
      "loss": 3.9159,
      "step": 7200
    },
    {
      "epoch": 0.7816711590296496,
      "grad_norm": 0.6337311863899231,
      "learning_rate": 0.0005537031840259039,
      "loss": 3.9184,
      "step": 7250
    },
    {
      "epoch": 0.7870619946091644,
      "grad_norm": 0.5891764760017395,
      "learning_rate": 0.0005533793847814354,
      "loss": 3.9032,
      "step": 7300
    },
    {
      "epoch": 0.7924528301886793,
      "grad_norm": 0.6272363066673279,
      "learning_rate": 0.000553055585536967,
      "loss": 3.9148,
      "step": 7350
    },
    {
      "epoch": 0.7978436657681941,
      "grad_norm": 0.6207209229469299,
      "learning_rate": 0.0005527317862924987,
      "loss": 3.8796,
      "step": 7400
    },
    {
      "epoch": 0.8032345013477089,
      "grad_norm": 0.5580083131790161,
      "learning_rate": 0.0005524079870480301,
      "loss": 3.8865,
      "step": 7450
    },
    {
      "epoch": 0.8086253369272237,
      "grad_norm": 0.7376517057418823,
      "learning_rate": 0.0005520841878035618,
      "loss": 3.9018,
      "step": 7500
    },
    {
      "epoch": 0.8140161725067385,
      "grad_norm": 0.6539864540100098,
      "learning_rate": 0.0005517603885590933,
      "loss": 3.8971,
      "step": 7550
    },
    {
      "epoch": 0.8194070080862533,
      "grad_norm": 0.640784502029419,
      "learning_rate": 0.0005514365893146249,
      "loss": 3.9161,
      "step": 7600
    },
    {
      "epoch": 0.8247978436657682,
      "grad_norm": 0.5456549525260925,
      "learning_rate": 0.0005511127900701564,
      "loss": 3.8959,
      "step": 7650
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 0.5484740734100342,
      "learning_rate": 0.000550788990825688,
      "loss": 3.871,
      "step": 7700
    },
    {
      "epoch": 0.8355795148247979,
      "grad_norm": 0.605842649936676,
      "learning_rate": 0.0005504651915812196,
      "loss": 3.8834,
      "step": 7750
    },
    {
      "epoch": 0.8409703504043127,
      "grad_norm": 0.5898163914680481,
      "learning_rate": 0.0005501413923367512,
      "loss": 3.8977,
      "step": 7800
    },
    {
      "epoch": 0.8463611859838275,
      "grad_norm": 0.7656612992286682,
      "learning_rate": 0.0005498175930922827,
      "loss": 3.8864,
      "step": 7850
    },
    {
      "epoch": 0.8517520215633423,
      "grad_norm": 0.8152493834495544,
      "learning_rate": 0.0005494937938478143,
      "loss": 3.8847,
      "step": 7900
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.5602722764015198,
      "learning_rate": 0.0005491699946033459,
      "loss": 3.8864,
      "step": 7950
    },
    {
      "epoch": 0.862533692722372,
      "grad_norm": 0.5723937153816223,
      "learning_rate": 0.0005488461953588775,
      "loss": 3.8618,
      "step": 8000
    },
    {
      "epoch": 0.862533692722372,
      "eval_accuracy": 0.33786419497282644,
      "eval_loss": 3.8131983280181885,
      "eval_runtime": 183.2819,
      "eval_samples_per_second": 98.269,
      "eval_steps_per_second": 6.144,
      "step": 8000
    },
    {
      "epoch": 0.8679245283018868,
      "grad_norm": 0.5539727210998535,
      "learning_rate": 0.000548522396114409,
      "loss": 3.879,
      "step": 8050
    },
    {
      "epoch": 0.8733153638814016,
      "grad_norm": 0.7839546203613281,
      "learning_rate": 0.0005481985968699406,
      "loss": 3.8589,
      "step": 8100
    },
    {
      "epoch": 0.8787061994609164,
      "grad_norm": 0.5483460426330566,
      "learning_rate": 0.0005478747976254721,
      "loss": 3.8715,
      "step": 8150
    },
    {
      "epoch": 0.8840970350404312,
      "grad_norm": 0.5753620266914368,
      "learning_rate": 0.0005475509983810037,
      "loss": 3.8617,
      "step": 8200
    },
    {
      "epoch": 0.889487870619946,
      "grad_norm": 0.5939356088638306,
      "learning_rate": 0.0005472271991365352,
      "loss": 3.8748,
      "step": 8250
    },
    {
      "epoch": 0.894878706199461,
      "grad_norm": 0.5494759678840637,
      "learning_rate": 0.0005469033998920669,
      "loss": 3.8485,
      "step": 8300
    },
    {
      "epoch": 0.9002695417789758,
      "grad_norm": 0.5566021800041199,
      "learning_rate": 0.0005465796006475984,
      "loss": 3.8652,
      "step": 8350
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 0.5274083018302917,
      "learning_rate": 0.00054625580140313,
      "loss": 3.866,
      "step": 8400
    },
    {
      "epoch": 0.9110512129380054,
      "grad_norm": 0.5682411193847656,
      "learning_rate": 0.0005459320021586615,
      "loss": 3.8363,
      "step": 8450
    },
    {
      "epoch": 0.9164420485175202,
      "grad_norm": 0.5468075275421143,
      "learning_rate": 0.0005456082029141932,
      "loss": 3.8486,
      "step": 8500
    },
    {
      "epoch": 0.921832884097035,
      "grad_norm": 0.5886979699134827,
      "learning_rate": 0.0005452844036697248,
      "loss": 3.8496,
      "step": 8550
    },
    {
      "epoch": 0.9272237196765498,
      "grad_norm": 0.566277027130127,
      "learning_rate": 0.0005449606044252563,
      "loss": 3.8782,
      "step": 8600
    },
    {
      "epoch": 0.9326145552560647,
      "grad_norm": 0.6201555728912354,
      "learning_rate": 0.0005446368051807879,
      "loss": 3.8507,
      "step": 8650
    },
    {
      "epoch": 0.9380053908355795,
      "grad_norm": 0.5942217111587524,
      "learning_rate": 0.0005443130059363194,
      "loss": 3.8613,
      "step": 8700
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.5227022767066956,
      "learning_rate": 0.0005439892066918511,
      "loss": 3.8428,
      "step": 8750
    },
    {
      "epoch": 0.9487870619946092,
      "grad_norm": 0.6055588126182556,
      "learning_rate": 0.0005436654074473825,
      "loss": 3.8415,
      "step": 8800
    },
    {
      "epoch": 0.954177897574124,
      "grad_norm": 0.5056185722351074,
      "learning_rate": 0.0005433416082029142,
      "loss": 3.8431,
      "step": 8850
    },
    {
      "epoch": 0.9595687331536388,
      "grad_norm": 0.6033360958099365,
      "learning_rate": 0.0005430178089584457,
      "loss": 3.8452,
      "step": 8900
    },
    {
      "epoch": 0.9649595687331537,
      "grad_norm": 0.6905035972595215,
      "learning_rate": 0.0005426940097139773,
      "loss": 3.8455,
      "step": 8950
    },
    {
      "epoch": 0.9703504043126685,
      "grad_norm": 0.6233043670654297,
      "learning_rate": 0.0005423766864543982,
      "loss": 3.8625,
      "step": 9000
    },
    {
      "epoch": 0.9703504043126685,
      "eval_accuracy": 0.34123297775951444,
      "eval_loss": 3.775101900100708,
      "eval_runtime": 183.702,
      "eval_samples_per_second": 98.045,
      "eval_steps_per_second": 6.129,
      "step": 9000
    },
    {
      "epoch": 0.9757412398921833,
      "grad_norm": 0.6275179982185364,
      "learning_rate": 0.0005420528872099298,
      "loss": 3.839,
      "step": 9050
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 0.5922442674636841,
      "learning_rate": 0.0005417290879654613,
      "loss": 3.8288,
      "step": 9100
    },
    {
      "epoch": 0.9865229110512129,
      "grad_norm": 0.5687377452850342,
      "learning_rate": 0.000541405288720993,
      "loss": 3.8416,
      "step": 9150
    },
    {
      "epoch": 0.9919137466307277,
      "grad_norm": 0.5579700469970703,
      "learning_rate": 0.0005410814894765245,
      "loss": 3.8436,
      "step": 9200
    },
    {
      "epoch": 0.9973045822102425,
      "grad_norm": 0.5312648415565491,
      "learning_rate": 0.0005407576902320561,
      "loss": 3.8355,
      "step": 9250
    },
    {
      "epoch": 1.0026954177897573,
      "grad_norm": 0.5382205247879028,
      "learning_rate": 0.0005404338909875876,
      "loss": 3.8133,
      "step": 9300
    },
    {
      "epoch": 1.0080862533692723,
      "grad_norm": 0.6140266060829163,
      "learning_rate": 0.0005401100917431192,
      "loss": 3.7783,
      "step": 9350
    },
    {
      "epoch": 1.013477088948787,
      "grad_norm": 0.5851580500602722,
      "learning_rate": 0.0005397862924986508,
      "loss": 3.7885,
      "step": 9400
    },
    {
      "epoch": 1.0188679245283019,
      "grad_norm": 0.5980308651924133,
      "learning_rate": 0.0005394624932541824,
      "loss": 3.7783,
      "step": 9450
    },
    {
      "epoch": 1.0242587601078168,
      "grad_norm": 0.5626499652862549,
      "learning_rate": 0.0005391386940097139,
      "loss": 3.7758,
      "step": 9500
    },
    {
      "epoch": 1.0296495956873315,
      "grad_norm": 0.5909863114356995,
      "learning_rate": 0.0005388148947652455,
      "loss": 3.7869,
      "step": 9550
    },
    {
      "epoch": 1.0350404312668464,
      "grad_norm": 0.5526650547981262,
      "learning_rate": 0.000538491095520777,
      "loss": 3.7655,
      "step": 9600
    },
    {
      "epoch": 1.0404312668463611,
      "grad_norm": 0.5704436898231506,
      "learning_rate": 0.0005381672962763086,
      "loss": 3.7576,
      "step": 9650
    },
    {
      "epoch": 1.045822102425876,
      "grad_norm": 0.5485429167747498,
      "learning_rate": 0.0005378434970318403,
      "loss": 3.7695,
      "step": 9700
    },
    {
      "epoch": 1.0512129380053907,
      "grad_norm": 0.5683411955833435,
      "learning_rate": 0.0005375196977873718,
      "loss": 3.7622,
      "step": 9750
    },
    {
      "epoch": 1.0566037735849056,
      "grad_norm": 0.5670942068099976,
      "learning_rate": 0.0005371958985429034,
      "loss": 3.7672,
      "step": 9800
    },
    {
      "epoch": 1.0619946091644206,
      "grad_norm": 0.6690695881843567,
      "learning_rate": 0.0005368720992984349,
      "loss": 3.7521,
      "step": 9850
    },
    {
      "epoch": 1.0673854447439353,
      "grad_norm": 0.6253776550292969,
      "learning_rate": 0.0005365483000539665,
      "loss": 3.762,
      "step": 9900
    },
    {
      "epoch": 1.0727762803234502,
      "grad_norm": 0.5629079937934875,
      "learning_rate": 0.0005362245008094981,
      "loss": 3.7594,
      "step": 9950
    },
    {
      "epoch": 1.0781671159029649,
      "grad_norm": 0.5760037302970886,
      "learning_rate": 0.0005359007015650297,
      "loss": 3.7656,
      "step": 10000
    },
    {
      "epoch": 1.0781671159029649,
      "eval_accuracy": 0.34546098758303934,
      "eval_loss": 3.7440903186798096,
      "eval_runtime": 183.3277,
      "eval_samples_per_second": 98.245,
      "eval_steps_per_second": 6.142,
      "step": 10000
    },
    {
      "epoch": 1.0835579514824798,
      "grad_norm": 0.5451662540435791,
      "learning_rate": 0.0005355769023205612,
      "loss": 3.755,
      "step": 10050
    },
    {
      "epoch": 1.0889487870619945,
      "grad_norm": 0.5988211035728455,
      "learning_rate": 0.0005352531030760928,
      "loss": 3.7454,
      "step": 10100
    },
    {
      "epoch": 1.0943396226415094,
      "grad_norm": 0.6496660709381104,
      "learning_rate": 0.0005349293038316244,
      "loss": 3.777,
      "step": 10150
    },
    {
      "epoch": 1.0997304582210243,
      "grad_norm": 0.5746974945068359,
      "learning_rate": 0.0005346055045871559,
      "loss": 3.7792,
      "step": 10200
    },
    {
      "epoch": 1.105121293800539,
      "grad_norm": 0.5537252426147461,
      "learning_rate": 0.0005342817053426874,
      "loss": 3.7429,
      "step": 10250
    },
    {
      "epoch": 1.110512129380054,
      "grad_norm": 0.6647258400917053,
      "learning_rate": 0.0005339579060982191,
      "loss": 3.7704,
      "step": 10300
    },
    {
      "epoch": 1.1159029649595686,
      "grad_norm": 0.5465642809867859,
      "learning_rate": 0.0005336341068537506,
      "loss": 3.7703,
      "step": 10350
    },
    {
      "epoch": 1.1212938005390836,
      "grad_norm": 0.5680330991744995,
      "learning_rate": 0.0005333103076092822,
      "loss": 3.7573,
      "step": 10400
    },
    {
      "epoch": 1.1266846361185983,
      "grad_norm": 0.5591064095497131,
      "learning_rate": 0.0005329865083648137,
      "loss": 3.7225,
      "step": 10450
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.5781257152557373,
      "learning_rate": 0.0005326627091203454,
      "loss": 3.7439,
      "step": 10500
    },
    {
      "epoch": 1.137466307277628,
      "grad_norm": 0.5286954045295715,
      "learning_rate": 0.0005323389098758769,
      "loss": 3.7463,
      "step": 10550
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.5636879205703735,
      "learning_rate": 0.0005320151106314085,
      "loss": 3.7536,
      "step": 10600
    },
    {
      "epoch": 1.1482479784366577,
      "grad_norm": 0.5693058371543884,
      "learning_rate": 0.00053169131138694,
      "loss": 3.7615,
      "step": 10650
    },
    {
      "epoch": 1.1536388140161726,
      "grad_norm": 0.62501060962677,
      "learning_rate": 0.0005313675121424716,
      "loss": 3.7439,
      "step": 10700
    },
    {
      "epoch": 1.1590296495956873,
      "grad_norm": 0.5845157504081726,
      "learning_rate": 0.0005310437128980032,
      "loss": 3.7449,
      "step": 10750
    },
    {
      "epoch": 1.1644204851752022,
      "grad_norm": 0.5758997201919556,
      "learning_rate": 0.0005307199136535348,
      "loss": 3.7219,
      "step": 10800
    },
    {
      "epoch": 1.169811320754717,
      "grad_norm": 0.6027551889419556,
      "learning_rate": 0.0005303961144090663,
      "loss": 3.7569,
      "step": 10850
    },
    {
      "epoch": 1.1752021563342319,
      "grad_norm": 0.5498311519622803,
      "learning_rate": 0.0005300723151645979,
      "loss": 3.7575,
      "step": 10900
    },
    {
      "epoch": 1.1805929919137466,
      "grad_norm": 0.5848629474639893,
      "learning_rate": 0.0005297485159201295,
      "loss": 3.7171,
      "step": 10950
    },
    {
      "epoch": 1.1859838274932615,
      "grad_norm": 0.5120974779129028,
      "learning_rate": 0.000529424716675661,
      "loss": 3.7343,
      "step": 11000
    },
    {
      "epoch": 1.1859838274932615,
      "eval_accuracy": 0.3476425202645785,
      "eval_loss": 3.7190380096435547,
      "eval_runtime": 183.3061,
      "eval_samples_per_second": 98.256,
      "eval_steps_per_second": 6.143,
      "step": 11000
    },
    {
      "epoch": 1.1913746630727764,
      "grad_norm": 0.5583203434944153,
      "learning_rate": 0.000529107393416082,
      "loss": 3.7451,
      "step": 11050
    },
    {
      "epoch": 1.196765498652291,
      "grad_norm": 0.5904495716094971,
      "learning_rate": 0.0005287835941716135,
      "loss": 3.74,
      "step": 11100
    },
    {
      "epoch": 1.202156334231806,
      "grad_norm": 0.5562983751296997,
      "learning_rate": 0.0005284597949271452,
      "loss": 3.7345,
      "step": 11150
    },
    {
      "epoch": 1.2075471698113207,
      "grad_norm": 0.676979660987854,
      "learning_rate": 0.0005281359956826767,
      "loss": 3.7461,
      "step": 11200
    },
    {
      "epoch": 1.2129380053908356,
      "grad_norm": 0.507462739944458,
      "learning_rate": 0.0005278121964382083,
      "loss": 3.756,
      "step": 11250
    },
    {
      "epoch": 1.2183288409703503,
      "grad_norm": 0.57256019115448,
      "learning_rate": 0.0005274883971937398,
      "loss": 3.7256,
      "step": 11300
    },
    {
      "epoch": 1.2237196765498652,
      "grad_norm": 0.5681268572807312,
      "learning_rate": 0.0005271645979492714,
      "loss": 3.747,
      "step": 11350
    },
    {
      "epoch": 1.2291105121293802,
      "grad_norm": 0.5520728230476379,
      "learning_rate": 0.000526840798704803,
      "loss": 3.7469,
      "step": 11400
    },
    {
      "epoch": 1.2345013477088949,
      "grad_norm": 0.5298144817352295,
      "learning_rate": 0.0005265169994603346,
      "loss": 3.7509,
      "step": 11450
    },
    {
      "epoch": 1.2398921832884098,
      "grad_norm": 0.5538373589515686,
      "learning_rate": 0.0005261932002158661,
      "loss": 3.7152,
      "step": 11500
    },
    {
      "epoch": 1.2452830188679245,
      "grad_norm": 0.601010262966156,
      "learning_rate": 0.0005258694009713977,
      "loss": 3.7366,
      "step": 11550
    },
    {
      "epoch": 1.2506738544474394,
      "grad_norm": 0.5493463277816772,
      "learning_rate": 0.0005255456017269292,
      "loss": 3.7331,
      "step": 11600
    },
    {
      "epoch": 1.256064690026954,
      "grad_norm": 0.5921666622161865,
      "learning_rate": 0.0005252218024824608,
      "loss": 3.732,
      "step": 11650
    },
    {
      "epoch": 1.261455525606469,
      "grad_norm": 0.5424702763557434,
      "learning_rate": 0.0005248980032379924,
      "loss": 3.7297,
      "step": 11700
    },
    {
      "epoch": 1.266846361185984,
      "grad_norm": 0.6088834404945374,
      "learning_rate": 0.000524574203993524,
      "loss": 3.7384,
      "step": 11750
    },
    {
      "epoch": 1.2722371967654986,
      "grad_norm": 0.5311514139175415,
      "learning_rate": 0.0005242504047490555,
      "loss": 3.7268,
      "step": 11800
    },
    {
      "epoch": 1.2776280323450135,
      "grad_norm": 0.5391415953636169,
      "learning_rate": 0.0005239266055045871,
      "loss": 3.7088,
      "step": 11850
    },
    {
      "epoch": 1.2830188679245282,
      "grad_norm": 0.5253583788871765,
      "learning_rate": 0.0005236028062601186,
      "loss": 3.7158,
      "step": 11900
    },
    {
      "epoch": 1.2884097035040432,
      "grad_norm": 0.5387663841247559,
      "learning_rate": 0.0005232790070156503,
      "loss": 3.7243,
      "step": 11950
    },
    {
      "epoch": 1.2938005390835579,
      "grad_norm": 0.5523077845573425,
      "learning_rate": 0.0005229552077711818,
      "loss": 3.7195,
      "step": 12000
    },
    {
      "epoch": 1.2938005390835579,
      "eval_accuracy": 0.34966281206060096,
      "eval_loss": 3.695082664489746,
      "eval_runtime": 183.4346,
      "eval_samples_per_second": 98.188,
      "eval_steps_per_second": 6.138,
      "step": 12000
    },
    {
      "epoch": 1.2991913746630728,
      "grad_norm": 0.5675397515296936,
      "learning_rate": 0.0005226314085267134,
      "loss": 3.701,
      "step": 12050
    },
    {
      "epoch": 1.3045822102425877,
      "grad_norm": 0.6067949533462524,
      "learning_rate": 0.000522307609282245,
      "loss": 3.7387,
      "step": 12100
    },
    {
      "epoch": 1.3099730458221024,
      "grad_norm": 0.5414419174194336,
      "learning_rate": 0.0005219838100377766,
      "loss": 3.7205,
      "step": 12150
    },
    {
      "epoch": 1.3153638814016173,
      "grad_norm": 0.6006377339363098,
      "learning_rate": 0.000521660010793308,
      "loss": 3.7045,
      "step": 12200
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.6085687875747681,
      "learning_rate": 0.0005213362115488396,
      "loss": 3.7319,
      "step": 12250
    },
    {
      "epoch": 1.326145552560647,
      "grad_norm": 0.5864773392677307,
      "learning_rate": 0.0005210124123043713,
      "loss": 3.6989,
      "step": 12300
    },
    {
      "epoch": 1.3315363881401616,
      "grad_norm": 0.5509727001190186,
      "learning_rate": 0.0005206886130599028,
      "loss": 3.725,
      "step": 12350
    },
    {
      "epoch": 1.3369272237196765,
      "grad_norm": 0.5826424360275269,
      "learning_rate": 0.0005203648138154344,
      "loss": 3.6955,
      "step": 12400
    },
    {
      "epoch": 1.3423180592991915,
      "grad_norm": 0.5667881369590759,
      "learning_rate": 0.0005200410145709659,
      "loss": 3.7183,
      "step": 12450
    },
    {
      "epoch": 1.3477088948787062,
      "grad_norm": 0.6350487470626831,
      "learning_rate": 0.0005197172153264976,
      "loss": 3.7223,
      "step": 12500
    },
    {
      "epoch": 1.353099730458221,
      "grad_norm": 0.565371572971344,
      "learning_rate": 0.0005193934160820291,
      "loss": 3.7318,
      "step": 12550
    },
    {
      "epoch": 1.3584905660377358,
      "grad_norm": 0.5364895462989807,
      "learning_rate": 0.0005190696168375607,
      "loss": 3.7087,
      "step": 12600
    },
    {
      "epoch": 1.3638814016172507,
      "grad_norm": 0.6211722493171692,
      "learning_rate": 0.0005187458175930922,
      "loss": 3.7054,
      "step": 12650
    },
    {
      "epoch": 1.3692722371967654,
      "grad_norm": 0.5237928628921509,
      "learning_rate": 0.0005184220183486238,
      "loss": 3.7025,
      "step": 12700
    },
    {
      "epoch": 1.3746630727762803,
      "grad_norm": 0.5414772033691406,
      "learning_rate": 0.0005180982191041554,
      "loss": 3.7171,
      "step": 12750
    },
    {
      "epoch": 1.3800539083557952,
      "grad_norm": 0.5802706480026245,
      "learning_rate": 0.000517774419859687,
      "loss": 3.7097,
      "step": 12800
    },
    {
      "epoch": 1.38544474393531,
      "grad_norm": 0.5995144248008728,
      "learning_rate": 0.0005174506206152185,
      "loss": 3.7307,
      "step": 12850
    },
    {
      "epoch": 1.3908355795148248,
      "grad_norm": 0.5822244882583618,
      "learning_rate": 0.0005171268213707501,
      "loss": 3.7303,
      "step": 12900
    },
    {
      "epoch": 1.3962264150943398,
      "grad_norm": 0.5466347336769104,
      "learning_rate": 0.0005168030221262816,
      "loss": 3.7282,
      "step": 12950
    },
    {
      "epoch": 1.4016172506738545,
      "grad_norm": 0.5338788032531738,
      "learning_rate": 0.0005164792228818132,
      "loss": 3.7112,
      "step": 13000
    },
    {
      "epoch": 1.4016172506738545,
      "eval_accuracy": 0.35203926802287927,
      "eval_loss": 3.6746132373809814,
      "eval_runtime": 183.6347,
      "eval_samples_per_second": 98.081,
      "eval_steps_per_second": 6.132,
      "step": 13000
    },
    {
      "epoch": 1.4070080862533692,
      "grad_norm": 0.6075963973999023,
      "learning_rate": 0.0005161554236373448,
      "loss": 3.692,
      "step": 13050
    },
    {
      "epoch": 1.412398921832884,
      "grad_norm": 0.5783334374427795,
      "learning_rate": 0.0005158381003777657,
      "loss": 3.7088,
      "step": 13100
    },
    {
      "epoch": 1.417789757412399,
      "grad_norm": 0.4899292588233948,
      "learning_rate": 0.0005155143011332973,
      "loss": 3.7093,
      "step": 13150
    },
    {
      "epoch": 1.4231805929919137,
      "grad_norm": 0.5416874885559082,
      "learning_rate": 0.0005151905018888289,
      "loss": 3.7333,
      "step": 13200
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.5604920983314514,
      "learning_rate": 0.0005148667026443604,
      "loss": 3.6912,
      "step": 13250
    },
    {
      "epoch": 1.4339622641509435,
      "grad_norm": 0.5507873892784119,
      "learning_rate": 0.000514542903399892,
      "loss": 3.698,
      "step": 13300
    },
    {
      "epoch": 1.4393530997304582,
      "grad_norm": 0.5818184614181519,
      "learning_rate": 0.0005142191041554237,
      "loss": 3.6776,
      "step": 13350
    },
    {
      "epoch": 1.444743935309973,
      "grad_norm": 0.6013184785842896,
      "learning_rate": 0.0005138953049109552,
      "loss": 3.7165,
      "step": 13400
    },
    {
      "epoch": 1.4501347708894878,
      "grad_norm": 0.6403388381004333,
      "learning_rate": 0.0005135715056664868,
      "loss": 3.7128,
      "step": 13450
    },
    {
      "epoch": 1.4555256064690028,
      "grad_norm": 0.595710813999176,
      "learning_rate": 0.0005132477064220183,
      "loss": 3.6982,
      "step": 13500
    },
    {
      "epoch": 1.4609164420485174,
      "grad_norm": 0.6105093955993652,
      "learning_rate": 0.0005129239071775499,
      "loss": 3.7149,
      "step": 13550
    },
    {
      "epoch": 1.4663072776280324,
      "grad_norm": 0.566348135471344,
      "learning_rate": 0.0005126001079330814,
      "loss": 3.6997,
      "step": 13600
    },
    {
      "epoch": 1.4716981132075473,
      "grad_norm": 0.5651235580444336,
      "learning_rate": 0.000512276308688613,
      "loss": 3.7034,
      "step": 13650
    },
    {
      "epoch": 1.477088948787062,
      "grad_norm": 0.5697197318077087,
      "learning_rate": 0.0005119525094441446,
      "loss": 3.7153,
      "step": 13700
    },
    {
      "epoch": 1.482479784366577,
      "grad_norm": 0.6158642172813416,
      "learning_rate": 0.0005116287101996762,
      "loss": 3.6978,
      "step": 13750
    },
    {
      "epoch": 1.4878706199460916,
      "grad_norm": 0.5030224323272705,
      "learning_rate": 0.0005113049109552077,
      "loss": 3.6963,
      "step": 13800
    },
    {
      "epoch": 1.4932614555256065,
      "grad_norm": 0.5537750720977783,
      "learning_rate": 0.0005109811117107393,
      "loss": 3.6907,
      "step": 13850
    },
    {
      "epoch": 1.4986522911051212,
      "grad_norm": 0.6205897927284241,
      "learning_rate": 0.0005106573124662708,
      "loss": 3.6805,
      "step": 13900
    },
    {
      "epoch": 1.5040431266846361,
      "grad_norm": 0.5059426426887512,
      "learning_rate": 0.0005103335132218025,
      "loss": 3.7003,
      "step": 13950
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.5683820843696594,
      "learning_rate": 0.000510009713977334,
      "loss": 3.6911,
      "step": 14000
    },
    {
      "epoch": 1.509433962264151,
      "eval_accuracy": 0.35447570037923115,
      "eval_loss": 3.6494083404541016,
      "eval_runtime": 183.437,
      "eval_samples_per_second": 98.186,
      "eval_steps_per_second": 6.138,
      "step": 14000
    },
    {
      "epoch": 1.5148247978436657,
      "grad_norm": 0.5536700487136841,
      "learning_rate": 0.0005096859147328656,
      "loss": 3.6698,
      "step": 14050
    },
    {
      "epoch": 1.5202156334231804,
      "grad_norm": 0.5463404655456543,
      "learning_rate": 0.0005093621154883971,
      "loss": 3.7006,
      "step": 14100
    },
    {
      "epoch": 1.5256064690026954,
      "grad_norm": 0.5179774761199951,
      "learning_rate": 0.0005090383162439288,
      "loss": 3.6828,
      "step": 14150
    },
    {
      "epoch": 1.5309973045822103,
      "grad_norm": 0.5397458076477051,
      "learning_rate": 0.0005087145169994602,
      "loss": 3.6819,
      "step": 14200
    },
    {
      "epoch": 1.536388140161725,
      "grad_norm": 0.5047475695610046,
      "learning_rate": 0.0005083907177549918,
      "loss": 3.6879,
      "step": 14250
    },
    {
      "epoch": 1.54177897574124,
      "grad_norm": 0.5295481085777283,
      "learning_rate": 0.0005080669185105234,
      "loss": 3.6971,
      "step": 14300
    },
    {
      "epoch": 1.5471698113207548,
      "grad_norm": 0.49893248081207275,
      "learning_rate": 0.000507743119266055,
      "loss": 3.7041,
      "step": 14350
    },
    {
      "epoch": 1.5525606469002695,
      "grad_norm": 0.6531094312667847,
      "learning_rate": 0.0005074193200215865,
      "loss": 3.6874,
      "step": 14400
    },
    {
      "epoch": 1.5579514824797842,
      "grad_norm": 0.5354477763175964,
      "learning_rate": 0.0005070955207771181,
      "loss": 3.7074,
      "step": 14450
    },
    {
      "epoch": 1.5633423180592994,
      "grad_norm": 0.7448453903198242,
      "learning_rate": 0.0005067717215326498,
      "loss": 3.6686,
      "step": 14500
    },
    {
      "epoch": 1.568733153638814,
      "grad_norm": 0.5273525714874268,
      "learning_rate": 0.0005064479222881813,
      "loss": 3.6823,
      "step": 14550
    },
    {
      "epoch": 1.5741239892183287,
      "grad_norm": 0.5093368887901306,
      "learning_rate": 0.0005061241230437129,
      "loss": 3.6727,
      "step": 14600
    },
    {
      "epoch": 1.5795148247978437,
      "grad_norm": 0.5493596196174622,
      "learning_rate": 0.0005058003237992444,
      "loss": 3.6886,
      "step": 14650
    },
    {
      "epoch": 1.5849056603773586,
      "grad_norm": 0.5913286209106445,
      "learning_rate": 0.000505476524554776,
      "loss": 3.6968,
      "step": 14700
    },
    {
      "epoch": 1.5902964959568733,
      "grad_norm": 0.5922813415527344,
      "learning_rate": 0.0005051527253103076,
      "loss": 3.6575,
      "step": 14750
    },
    {
      "epoch": 1.595687331536388,
      "grad_norm": 0.562714695930481,
      "learning_rate": 0.0005048289260658392,
      "loss": 3.6826,
      "step": 14800
    },
    {
      "epoch": 1.6010781671159031,
      "grad_norm": 0.5367226600646973,
      "learning_rate": 0.0005045051268213707,
      "loss": 3.6719,
      "step": 14850
    },
    {
      "epoch": 1.6064690026954178,
      "grad_norm": 0.5268148183822632,
      "learning_rate": 0.0005041813275769023,
      "loss": 3.6715,
      "step": 14900
    },
    {
      "epoch": 1.6118598382749325,
      "grad_norm": 0.5168265104293823,
      "learning_rate": 0.0005038575283324338,
      "loss": 3.6692,
      "step": 14950
    },
    {
      "epoch": 1.6172506738544474,
      "grad_norm": 0.5893277525901794,
      "learning_rate": 0.0005035337290879654,
      "loss": 3.6767,
      "step": 15000
    },
    {
      "epoch": 1.6172506738544474,
      "eval_accuracy": 0.356096910118311,
      "eval_loss": 3.634113311767578,
      "eval_runtime": 183.2068,
      "eval_samples_per_second": 98.31,
      "eval_steps_per_second": 6.146,
      "step": 15000
    },
    {
      "epoch": 1.6226415094339623,
      "grad_norm": 0.5801236033439636,
      "learning_rate": 0.000503209929843497,
      "loss": 3.6864,
      "step": 15050
    },
    {
      "epoch": 1.628032345013477,
      "grad_norm": 0.5008922815322876,
      "learning_rate": 0.0005028926065839179,
      "loss": 3.6614,
      "step": 15100
    },
    {
      "epoch": 1.633423180592992,
      "grad_norm": 0.5347914099693298,
      "learning_rate": 0.0005025688073394495,
      "loss": 3.6811,
      "step": 15150
    },
    {
      "epoch": 1.6388140161725069,
      "grad_norm": 0.6196523904800415,
      "learning_rate": 0.0005022450080949811,
      "loss": 3.6564,
      "step": 15200
    },
    {
      "epoch": 1.6442048517520216,
      "grad_norm": 0.5722506046295166,
      "learning_rate": 0.0005019212088505126,
      "loss": 3.6983,
      "step": 15250
    },
    {
      "epoch": 1.6495956873315363,
      "grad_norm": 0.5670409202575684,
      "learning_rate": 0.0005015974096060442,
      "loss": 3.6822,
      "step": 15300
    },
    {
      "epoch": 1.6549865229110512,
      "grad_norm": 0.5775005221366882,
      "learning_rate": 0.0005012736103615758,
      "loss": 3.676,
      "step": 15350
    },
    {
      "epoch": 1.6603773584905661,
      "grad_norm": 0.5482152700424194,
      "learning_rate": 0.0005009498111171074,
      "loss": 3.6664,
      "step": 15400
    },
    {
      "epoch": 1.6657681940700808,
      "grad_norm": 0.5493103861808777,
      "learning_rate": 0.0005006260118726389,
      "loss": 3.6771,
      "step": 15450
    },
    {
      "epoch": 1.6711590296495957,
      "grad_norm": 0.5443813800811768,
      "learning_rate": 0.0005003022126281705,
      "loss": 3.6884,
      "step": 15500
    },
    {
      "epoch": 1.6765498652291106,
      "grad_norm": 0.5360202193260193,
      "learning_rate": 0.000499978413383702,
      "loss": 3.6633,
      "step": 15550
    },
    {
      "epoch": 1.6819407008086253,
      "grad_norm": 0.5757080316543579,
      "learning_rate": 0.0004996546141392336,
      "loss": 3.6626,
      "step": 15600
    },
    {
      "epoch": 1.68733153638814,
      "grad_norm": 0.5827114582061768,
      "learning_rate": 0.0004993308148947651,
      "loss": 3.657,
      "step": 15650
    },
    {
      "epoch": 1.692722371967655,
      "grad_norm": 0.5391814708709717,
      "learning_rate": 0.0004990070156502968,
      "loss": 3.673,
      "step": 15700
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.6123207211494446,
      "learning_rate": 0.0004986832164058284,
      "loss": 3.6493,
      "step": 15750
    },
    {
      "epoch": 1.7035040431266846,
      "grad_norm": 0.5068091154098511,
      "learning_rate": 0.0004983594171613599,
      "loss": 3.6679,
      "step": 15800
    },
    {
      "epoch": 1.7088948787061995,
      "grad_norm": 0.5883404016494751,
      "learning_rate": 0.0004980356179168915,
      "loss": 3.6597,
      "step": 15850
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5995203256607056,
      "learning_rate": 0.000497711818672423,
      "loss": 3.6626,
      "step": 15900
    },
    {
      "epoch": 1.719676549865229,
      "grad_norm": 0.5240509510040283,
      "learning_rate": 0.0004973880194279547,
      "loss": 3.6495,
      "step": 15950
    },
    {
      "epoch": 1.7250673854447438,
      "grad_norm": 0.5879561305046082,
      "learning_rate": 0.0004970642201834862,
      "loss": 3.6497,
      "step": 16000
    },
    {
      "epoch": 1.7250673854447438,
      "eval_accuracy": 0.3578040642916522,
      "eval_loss": 3.616135597229004,
      "eval_runtime": 183.5062,
      "eval_samples_per_second": 98.149,
      "eval_steps_per_second": 6.136,
      "step": 16000
    },
    {
      "epoch": 1.7304582210242587,
      "grad_norm": 0.5466377139091492,
      "learning_rate": 0.0004967404209390178,
      "loss": 3.6711,
      "step": 16050
    },
    {
      "epoch": 1.7358490566037736,
      "grad_norm": 0.5741949081420898,
      "learning_rate": 0.0004964166216945493,
      "loss": 3.6722,
      "step": 16100
    },
    {
      "epoch": 1.7412398921832883,
      "grad_norm": 0.6283635497093201,
      "learning_rate": 0.000496092822450081,
      "loss": 3.6545,
      "step": 16150
    },
    {
      "epoch": 1.7466307277628033,
      "grad_norm": 0.5305262207984924,
      "learning_rate": 0.0004957690232056125,
      "loss": 3.6599,
      "step": 16200
    },
    {
      "epoch": 1.7520215633423182,
      "grad_norm": 0.5391592979431152,
      "learning_rate": 0.0004954452239611441,
      "loss": 3.6603,
      "step": 16250
    },
    {
      "epoch": 1.7574123989218329,
      "grad_norm": 0.5375102162361145,
      "learning_rate": 0.0004951214247166756,
      "loss": 3.6588,
      "step": 16300
    },
    {
      "epoch": 1.7628032345013476,
      "grad_norm": 0.5385131239891052,
      "learning_rate": 0.0004947976254722072,
      "loss": 3.6598,
      "step": 16350
    },
    {
      "epoch": 1.7681940700808625,
      "grad_norm": 0.581318199634552,
      "learning_rate": 0.0004944738262277387,
      "loss": 3.664,
      "step": 16400
    },
    {
      "epoch": 1.7735849056603774,
      "grad_norm": 0.5098201632499695,
      "learning_rate": 0.0004941500269832703,
      "loss": 3.666,
      "step": 16450
    },
    {
      "epoch": 1.778975741239892,
      "grad_norm": 0.5255228877067566,
      "learning_rate": 0.0004938262277388019,
      "loss": 3.6614,
      "step": 16500
    },
    {
      "epoch": 1.784366576819407,
      "grad_norm": 0.5504864454269409,
      "learning_rate": 0.0004935024284943335,
      "loss": 3.641,
      "step": 16550
    },
    {
      "epoch": 1.789757412398922,
      "grad_norm": 0.510482132434845,
      "learning_rate": 0.000493178629249865,
      "loss": 3.6548,
      "step": 16600
    },
    {
      "epoch": 1.7951482479784366,
      "grad_norm": 0.5756171941757202,
      "learning_rate": 0.0004928548300053966,
      "loss": 3.6613,
      "step": 16650
    },
    {
      "epoch": 1.8005390835579513,
      "grad_norm": 0.5339798927307129,
      "learning_rate": 0.0004925310307609282,
      "loss": 3.653,
      "step": 16700
    },
    {
      "epoch": 1.8059299191374663,
      "grad_norm": 0.588362991809845,
      "learning_rate": 0.0004922072315164598,
      "loss": 3.6417,
      "step": 16750
    },
    {
      "epoch": 1.8113207547169812,
      "grad_norm": 0.5696936249732971,
      "learning_rate": 0.0004918834322719913,
      "loss": 3.6488,
      "step": 16800
    },
    {
      "epoch": 1.8167115902964959,
      "grad_norm": 0.5738952159881592,
      "learning_rate": 0.0004915596330275229,
      "loss": 3.6448,
      "step": 16850
    },
    {
      "epoch": 1.8221024258760108,
      "grad_norm": 0.5368397235870361,
      "learning_rate": 0.0004912358337830544,
      "loss": 3.6527,
      "step": 16900
    },
    {
      "epoch": 1.8274932614555257,
      "grad_norm": 0.5771840214729309,
      "learning_rate": 0.000490912034538586,
      "loss": 3.6372,
      "step": 16950
    },
    {
      "epoch": 1.8328840970350404,
      "grad_norm": 0.5940222144126892,
      "learning_rate": 0.0004905882352941175,
      "loss": 3.6291,
      "step": 17000
    },
    {
      "epoch": 1.8328840970350404,
      "eval_accuracy": 0.359461020830823,
      "eval_loss": 3.600602626800537,
      "eval_runtime": 183.2755,
      "eval_samples_per_second": 98.273,
      "eval_steps_per_second": 6.144,
      "step": 17000
    },
    {
      "epoch": 1.838274932614555,
      "grad_norm": 0.49475905299186707,
      "learning_rate": 0.0004902644360496492,
      "loss": 3.6373,
      "step": 17050
    },
    {
      "epoch": 1.8436657681940702,
      "grad_norm": 0.5278207063674927,
      "learning_rate": 0.0004899406368051808,
      "loss": 3.6397,
      "step": 17100
    },
    {
      "epoch": 1.849056603773585,
      "grad_norm": 0.5330104827880859,
      "learning_rate": 0.0004896233135456017,
      "loss": 3.6291,
      "step": 17150
    },
    {
      "epoch": 1.8544474393530996,
      "grad_norm": 0.5108185410499573,
      "learning_rate": 0.0004892995143011333,
      "loss": 3.6578,
      "step": 17200
    },
    {
      "epoch": 1.8598382749326146,
      "grad_norm": 0.5817229747772217,
      "learning_rate": 0.0004889757150566648,
      "loss": 3.6642,
      "step": 17250
    },
    {
      "epoch": 1.8652291105121295,
      "grad_norm": 0.5322571992874146,
      "learning_rate": 0.0004886519158121964,
      "loss": 3.648,
      "step": 17300
    },
    {
      "epoch": 1.8706199460916442,
      "grad_norm": 0.5755774974822998,
      "learning_rate": 0.000488328116567728,
      "loss": 3.636,
      "step": 17350
    },
    {
      "epoch": 1.8760107816711589,
      "grad_norm": 0.5633023977279663,
      "learning_rate": 0.0004880043173232595,
      "loss": 3.6531,
      "step": 17400
    },
    {
      "epoch": 1.881401617250674,
      "grad_norm": 0.5242737531661987,
      "learning_rate": 0.0004876805180787911,
      "loss": 3.6244,
      "step": 17450
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.5220695734024048,
      "learning_rate": 0.0004873567188343227,
      "loss": 3.6473,
      "step": 17500
    },
    {
      "epoch": 1.8921832884097034,
      "grad_norm": 0.4793733060359955,
      "learning_rate": 0.0004870329195898542,
      "loss": 3.6514,
      "step": 17550
    },
    {
      "epoch": 1.8975741239892183,
      "grad_norm": 0.6032518148422241,
      "learning_rate": 0.00048670912034538583,
      "loss": 3.6398,
      "step": 17600
    },
    {
      "epoch": 1.9029649595687332,
      "grad_norm": 0.562566876411438,
      "learning_rate": 0.0004863853211009174,
      "loss": 3.6368,
      "step": 17650
    },
    {
      "epoch": 1.908355795148248,
      "grad_norm": 0.5620681643486023,
      "learning_rate": 0.000486061521856449,
      "loss": 3.6535,
      "step": 17700
    },
    {
      "epoch": 1.9137466307277629,
      "grad_norm": 0.548251748085022,
      "learning_rate": 0.00048573772261198054,
      "loss": 3.6267,
      "step": 17750
    },
    {
      "epoch": 1.9191374663072778,
      "grad_norm": 0.5563006401062012,
      "learning_rate": 0.00048541392336751214,
      "loss": 3.632,
      "step": 17800
    },
    {
      "epoch": 1.9245283018867925,
      "grad_norm": 0.5012133121490479,
      "learning_rate": 0.0004850901241230437,
      "loss": 3.6225,
      "step": 17850
    },
    {
      "epoch": 1.9299191374663072,
      "grad_norm": 0.5342919826507568,
      "learning_rate": 0.0004847663248785753,
      "loss": 3.6266,
      "step": 17900
    },
    {
      "epoch": 1.935309973045822,
      "grad_norm": 0.5178313851356506,
      "learning_rate": 0.0004844425256341068,
      "loss": 3.6355,
      "step": 17950
    },
    {
      "epoch": 1.940700808625337,
      "grad_norm": 0.5467429161071777,
      "learning_rate": 0.00048411872638963834,
      "loss": 3.643,
      "step": 18000
    },
    {
      "epoch": 1.940700808625337,
      "eval_accuracy": 0.36094337217927597,
      "eval_loss": 3.584843873977661,
      "eval_runtime": 183.5851,
      "eval_samples_per_second": 98.107,
      "eval_steps_per_second": 6.133,
      "step": 18000
    },
    {
      "epoch": 1.9460916442048517,
      "grad_norm": 0.5631021857261658,
      "learning_rate": 0.00048379492714516995,
      "loss": 3.6179,
      "step": 18050
    },
    {
      "epoch": 1.9514824797843666,
      "grad_norm": 0.541903018951416,
      "learning_rate": 0.0004834711279007015,
      "loss": 3.6427,
      "step": 18100
    },
    {
      "epoch": 1.9568733153638815,
      "grad_norm": 0.5587621927261353,
      "learning_rate": 0.0004831473286562331,
      "loss": 3.6403,
      "step": 18150
    },
    {
      "epoch": 1.9622641509433962,
      "grad_norm": 1.227616548538208,
      "learning_rate": 0.00048282352941176465,
      "loss": 3.6295,
      "step": 18200
    },
    {
      "epoch": 1.967654986522911,
      "grad_norm": 0.5264652371406555,
      "learning_rate": 0.00048249973016729626,
      "loss": 3.645,
      "step": 18250
    },
    {
      "epoch": 1.9730458221024259,
      "grad_norm": 0.5398224592208862,
      "learning_rate": 0.0004821759309228278,
      "loss": 3.6412,
      "step": 18300
    },
    {
      "epoch": 1.9784366576819408,
      "grad_norm": 0.5698915123939514,
      "learning_rate": 0.00048185213167835936,
      "loss": 3.6253,
      "step": 18350
    },
    {
      "epoch": 1.9838274932614555,
      "grad_norm": 0.5805723071098328,
      "learning_rate": 0.00048152833243389096,
      "loss": 3.6555,
      "step": 18400
    },
    {
      "epoch": 1.9892183288409704,
      "grad_norm": 0.5575518012046814,
      "learning_rate": 0.0004812045331894225,
      "loss": 3.611,
      "step": 18450
    },
    {
      "epoch": 1.9946091644204853,
      "grad_norm": 0.5759623050689697,
      "learning_rate": 0.0004808807339449541,
      "loss": 3.617,
      "step": 18500
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1932631731033325,
      "learning_rate": 0.0004805569347004856,
      "loss": 3.6358,
      "step": 18550
    },
    {
      "epoch": 2.0053908355795147,
      "grad_norm": 0.5486853122711182,
      "learning_rate": 0.0004802331354560173,
      "loss": 3.5442,
      "step": 18600
    },
    {
      "epoch": 2.01078167115903,
      "grad_norm": 0.5466881394386292,
      "learning_rate": 0.00047990933621154877,
      "loss": 3.5451,
      "step": 18650
    },
    {
      "epoch": 2.0161725067385445,
      "grad_norm": 0.5424332618713379,
      "learning_rate": 0.0004795855369670804,
      "loss": 3.5382,
      "step": 18700
    },
    {
      "epoch": 2.0215633423180592,
      "grad_norm": 0.5714851021766663,
      "learning_rate": 0.0004792617377226119,
      "loss": 3.5577,
      "step": 18750
    },
    {
      "epoch": 2.026954177897574,
      "grad_norm": 0.5695775151252747,
      "learning_rate": 0.0004789379384781435,
      "loss": 3.5428,
      "step": 18800
    },
    {
      "epoch": 2.032345013477089,
      "grad_norm": 0.530637264251709,
      "learning_rate": 0.0004786141392336751,
      "loss": 3.5231,
      "step": 18850
    },
    {
      "epoch": 2.0377358490566038,
      "grad_norm": 0.5261335968971252,
      "learning_rate": 0.00047829033998920663,
      "loss": 3.5522,
      "step": 18900
    },
    {
      "epoch": 2.0431266846361185,
      "grad_norm": 0.5697975158691406,
      "learning_rate": 0.00047796654074473824,
      "loss": 3.5444,
      "step": 18950
    },
    {
      "epoch": 2.0485175202156336,
      "grad_norm": 0.6406385898590088,
      "learning_rate": 0.0004776427415002698,
      "loss": 3.5593,
      "step": 19000
    },
    {
      "epoch": 2.0485175202156336,
      "eval_accuracy": 0.36186920343634316,
      "eval_loss": 3.57981014251709,
      "eval_runtime": 183.1277,
      "eval_samples_per_second": 98.352,
      "eval_steps_per_second": 6.149,
      "step": 19000
    },
    {
      "epoch": 2.0539083557951483,
      "grad_norm": 0.5842219591140747,
      "learning_rate": 0.0004773189422558014,
      "loss": 3.5202,
      "step": 19050
    },
    {
      "epoch": 2.059299191374663,
      "grad_norm": 0.5753844380378723,
      "learning_rate": 0.00047699514301133294,
      "loss": 3.5584,
      "step": 19100
    },
    {
      "epoch": 2.0646900269541777,
      "grad_norm": 0.5149918794631958,
      "learning_rate": 0.00047667134376686455,
      "loss": 3.5434,
      "step": 19150
    },
    {
      "epoch": 2.070080862533693,
      "grad_norm": 0.5646060705184937,
      "learning_rate": 0.0004763540205072854,
      "loss": 3.5652,
      "step": 19200
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.6100060343742371,
      "learning_rate": 0.00047603022126281705,
      "loss": 3.563,
      "step": 19250
    },
    {
      "epoch": 2.0808625336927222,
      "grad_norm": 0.5924269556999207,
      "learning_rate": 0.00047570642201834855,
      "loss": 3.5591,
      "step": 19300
    },
    {
      "epoch": 2.0862533692722374,
      "grad_norm": 0.5271193385124207,
      "learning_rate": 0.00047538262277388015,
      "loss": 3.5502,
      "step": 19350
    },
    {
      "epoch": 2.091644204851752,
      "grad_norm": 0.559655487537384,
      "learning_rate": 0.0004750588235294117,
      "loss": 3.5551,
      "step": 19400
    },
    {
      "epoch": 2.0970350404312668,
      "grad_norm": 0.5192553997039795,
      "learning_rate": 0.0004747350242849433,
      "loss": 3.5419,
      "step": 19450
    },
    {
      "epoch": 2.1024258760107815,
      "grad_norm": 0.5999012589454651,
      "learning_rate": 0.00047441122504047486,
      "loss": 3.5376,
      "step": 19500
    },
    {
      "epoch": 2.1078167115902966,
      "grad_norm": 0.601438581943512,
      "learning_rate": 0.0004740874257960064,
      "loss": 3.5554,
      "step": 19550
    },
    {
      "epoch": 2.1132075471698113,
      "grad_norm": 0.5252161622047424,
      "learning_rate": 0.000473763626551538,
      "loss": 3.5518,
      "step": 19600
    },
    {
      "epoch": 2.118598382749326,
      "grad_norm": 0.568138062953949,
      "learning_rate": 0.00047343982730706956,
      "loss": 3.5517,
      "step": 19650
    },
    {
      "epoch": 2.123989218328841,
      "grad_norm": 0.5670929551124573,
      "learning_rate": 0.00047311602806260117,
      "loss": 3.5404,
      "step": 19700
    },
    {
      "epoch": 2.129380053908356,
      "grad_norm": 0.6419331431388855,
      "learning_rate": 0.0004727922288181327,
      "loss": 3.5383,
      "step": 19750
    },
    {
      "epoch": 2.1347708894878705,
      "grad_norm": 0.5455891489982605,
      "learning_rate": 0.0004724684295736643,
      "loss": 3.5673,
      "step": 19800
    },
    {
      "epoch": 2.1401617250673857,
      "grad_norm": 0.5633212327957153,
      "learning_rate": 0.0004721446303291959,
      "loss": 3.5679,
      "step": 19850
    },
    {
      "epoch": 2.1455525606469004,
      "grad_norm": Infinity,
      "learning_rate": 0.0004718273070696168,
      "loss": 3.5567,
      "step": 19900
    },
    {
      "epoch": 2.150943396226415,
      "grad_norm": 0.5624653697013855,
      "learning_rate": 0.0004715035078251483,
      "loss": 3.5539,
      "step": 19950
    },
    {
      "epoch": 2.1563342318059298,
      "grad_norm": 0.5584042072296143,
      "learning_rate": 0.0004711797085806799,
      "loss": 3.5552,
      "step": 20000
    },
    {
      "epoch": 2.1563342318059298,
      "eval_accuracy": 0.3635111658769956,
      "eval_loss": 3.5678915977478027,
      "eval_runtime": 183.3667,
      "eval_samples_per_second": 98.224,
      "eval_steps_per_second": 6.141,
      "step": 20000
    },
    {
      "epoch": 2.161725067385445,
      "grad_norm": 0.5540956258773804,
      "learning_rate": 0.0004708559093362115,
      "loss": 3.5543,
      "step": 20050
    },
    {
      "epoch": 2.1671159029649596,
      "grad_norm": 0.5896965265274048,
      "learning_rate": 0.0004705321100917431,
      "loss": 3.5654,
      "step": 20100
    },
    {
      "epoch": 2.1725067385444743,
      "grad_norm": 0.5863565802574158,
      "learning_rate": 0.00047020831084727463,
      "loss": 3.5427,
      "step": 20150
    },
    {
      "epoch": 2.177897574123989,
      "grad_norm": 0.6331449747085571,
      "learning_rate": 0.00046988451160280624,
      "loss": 3.5658,
      "step": 20200
    },
    {
      "epoch": 2.183288409703504,
      "grad_norm": 0.5432201027870178,
      "learning_rate": 0.0004695607123583378,
      "loss": 3.5648,
      "step": 20250
    },
    {
      "epoch": 2.188679245283019,
      "grad_norm": 0.5611268877983093,
      "learning_rate": 0.00046923691311386934,
      "loss": 3.5401,
      "step": 20300
    },
    {
      "epoch": 2.1940700808625335,
      "grad_norm": 0.5305395722389221,
      "learning_rate": 0.00046891311386940094,
      "loss": 3.5583,
      "step": 20350
    },
    {
      "epoch": 2.1994609164420487,
      "grad_norm": 0.5022453665733337,
      "learning_rate": 0.0004685893146249325,
      "loss": 3.5571,
      "step": 20400
    },
    {
      "epoch": 2.2048517520215634,
      "grad_norm": 0.5850080847740173,
      "learning_rate": 0.0004682655153804641,
      "loss": 3.5559,
      "step": 20450
    },
    {
      "epoch": 2.210242587601078,
      "grad_norm": 0.5080869793891907,
      "learning_rate": 0.00046794171613599565,
      "loss": 3.5601,
      "step": 20500
    },
    {
      "epoch": 2.215633423180593,
      "grad_norm": 0.5989983081817627,
      "learning_rate": 0.00046761791689152725,
      "loss": 3.563,
      "step": 20550
    },
    {
      "epoch": 2.221024258760108,
      "grad_norm": 0.6006690859794617,
      "learning_rate": 0.00046729411764705875,
      "loss": 3.5674,
      "step": 20600
    },
    {
      "epoch": 2.2264150943396226,
      "grad_norm": 0.5296274423599243,
      "learning_rate": 0.0004669703184025904,
      "loss": 3.5525,
      "step": 20650
    },
    {
      "epoch": 2.2318059299191373,
      "grad_norm": 0.5273208618164062,
      "learning_rate": 0.0004666465191581219,
      "loss": 3.5551,
      "step": 20700
    },
    {
      "epoch": 2.2371967654986524,
      "grad_norm": 0.6000422239303589,
      "learning_rate": 0.00046632271991365346,
      "loss": 3.556,
      "step": 20750
    },
    {
      "epoch": 2.242587601078167,
      "grad_norm": 0.542732298374176,
      "learning_rate": 0.00046599892066918506,
      "loss": 3.5593,
      "step": 20800
    },
    {
      "epoch": 2.247978436657682,
      "grad_norm": 0.5846306085586548,
      "learning_rate": 0.0004656751214247166,
      "loss": 3.5424,
      "step": 20850
    },
    {
      "epoch": 2.2533692722371965,
      "grad_norm": 0.6560302972793579,
      "learning_rate": 0.0004653513221802482,
      "loss": 3.554,
      "step": 20900
    },
    {
      "epoch": 2.2587601078167117,
      "grad_norm": 0.5398501753807068,
      "learning_rate": 0.00046502752293577977,
      "loss": 3.5597,
      "step": 20950
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.5567417144775391,
      "learning_rate": 0.00046470372369131137,
      "loss": 3.5423,
      "step": 21000
    },
    {
      "epoch": 2.2641509433962264,
      "eval_accuracy": 0.3644758948679004,
      "eval_loss": 3.5552797317504883,
      "eval_runtime": 183.5672,
      "eval_samples_per_second": 98.117,
      "eval_steps_per_second": 6.134,
      "step": 21000
    },
    {
      "epoch": 2.269541778975741,
      "grad_norm": 0.5651289224624634,
      "learning_rate": 0.0004643799244468429,
      "loss": 3.5383,
      "step": 21050
    },
    {
      "epoch": 2.274932614555256,
      "grad_norm": 0.5148941278457642,
      "learning_rate": 0.0004640561252023745,
      "loss": 3.5681,
      "step": 21100
    },
    {
      "epoch": 2.280323450134771,
      "grad_norm": 0.5875728726387024,
      "learning_rate": 0.0004637323259579061,
      "loss": 3.5605,
      "step": 21150
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.5935449004173279,
      "learning_rate": 0.0004634085267134376,
      "loss": 3.5523,
      "step": 21200
    },
    {
      "epoch": 2.2911051212938007,
      "grad_norm": 0.5729395747184753,
      "learning_rate": 0.00046308472746896923,
      "loss": 3.5244,
      "step": 21250
    },
    {
      "epoch": 2.2964959568733154,
      "grad_norm": 0.6214632391929626,
      "learning_rate": 0.00046276092822450073,
      "loss": 3.5476,
      "step": 21300
    },
    {
      "epoch": 2.30188679245283,
      "grad_norm": 0.5737814903259277,
      "learning_rate": 0.00046243712898003233,
      "loss": 3.5626,
      "step": 21350
    },
    {
      "epoch": 2.3072776280323453,
      "grad_norm": 0.5795217156410217,
      "learning_rate": 0.0004621133297355639,
      "loss": 3.54,
      "step": 21400
    },
    {
      "epoch": 2.31266846361186,
      "grad_norm": 0.5368177890777588,
      "learning_rate": 0.0004617895304910955,
      "loss": 3.552,
      "step": 21450
    },
    {
      "epoch": 2.3180592991913747,
      "grad_norm": 0.6149646639823914,
      "learning_rate": 0.00046146573124662704,
      "loss": 3.557,
      "step": 21500
    },
    {
      "epoch": 2.3234501347708894,
      "grad_norm": 0.5619512796401978,
      "learning_rate": 0.00046114193200215864,
      "loss": 3.5438,
      "step": 21550
    },
    {
      "epoch": 2.3288409703504045,
      "grad_norm": 0.5701526403427124,
      "learning_rate": 0.0004608181327576902,
      "loss": 3.5537,
      "step": 21600
    },
    {
      "epoch": 2.334231805929919,
      "grad_norm": 0.5332589149475098,
      "learning_rate": 0.00046049433351322175,
      "loss": 3.5537,
      "step": 21650
    },
    {
      "epoch": 2.339622641509434,
      "grad_norm": 0.583706259727478,
      "learning_rate": 0.00046017053426875335,
      "loss": 3.5544,
      "step": 21700
    },
    {
      "epoch": 2.3450134770889486,
      "grad_norm": 0.5967761278152466,
      "learning_rate": 0.0004598467350242849,
      "loss": 3.5528,
      "step": 21750
    },
    {
      "epoch": 2.3504043126684637,
      "grad_norm": 0.5579556822776794,
      "learning_rate": 0.0004595229357798165,
      "loss": 3.5632,
      "step": 21800
    },
    {
      "epoch": 2.3557951482479784,
      "grad_norm": 0.5610327124595642,
      "learning_rate": 0.00045919913653534806,
      "loss": 3.5557,
      "step": 21850
    },
    {
      "epoch": 2.361185983827493,
      "grad_norm": 0.5787166953086853,
      "learning_rate": 0.00045887533729087966,
      "loss": 3.5439,
      "step": 21900
    },
    {
      "epoch": 2.3665768194070083,
      "grad_norm": 0.540556013584137,
      "learning_rate": 0.00045855153804641116,
      "loss": 3.558,
      "step": 21950
    },
    {
      "epoch": 2.371967654986523,
      "grad_norm": 0.5531706809997559,
      "learning_rate": 0.0004582277388019427,
      "loss": 3.5562,
      "step": 22000
    },
    {
      "epoch": 2.371967654986523,
      "eval_accuracy": 0.3654050943644898,
      "eval_loss": 3.5464489459991455,
      "eval_runtime": 183.2178,
      "eval_samples_per_second": 98.304,
      "eval_steps_per_second": 6.146,
      "step": 22000
    },
    {
      "epoch": 2.3773584905660377,
      "grad_norm": 0.5978881120681763,
      "learning_rate": 0.0004579039395574743,
      "loss": 3.5342,
      "step": 22050
    },
    {
      "epoch": 2.382749326145553,
      "grad_norm": 0.5623920559883118,
      "learning_rate": 0.00045758014031300586,
      "loss": 3.5555,
      "step": 22100
    },
    {
      "epoch": 2.3881401617250675,
      "grad_norm": 0.5382367372512817,
      "learning_rate": 0.00045725634106853747,
      "loss": 3.5515,
      "step": 22150
    },
    {
      "epoch": 2.393530997304582,
      "grad_norm": 0.6158401370048523,
      "learning_rate": 0.000456932541824069,
      "loss": 3.5763,
      "step": 22200
    },
    {
      "epoch": 2.398921832884097,
      "grad_norm": 0.5851870775222778,
      "learning_rate": 0.0004566087425796006,
      "loss": 3.546,
      "step": 22250
    },
    {
      "epoch": 2.404312668463612,
      "grad_norm": 0.6621217131614685,
      "learning_rate": 0.0004562849433351322,
      "loss": 3.527,
      "step": 22300
    },
    {
      "epoch": 2.4097035040431267,
      "grad_norm": 0.5817117094993591,
      "learning_rate": 0.0004559611440906638,
      "loss": 3.55,
      "step": 22350
    },
    {
      "epoch": 2.4150943396226414,
      "grad_norm": 0.5394627451896667,
      "learning_rate": 0.00045563734484619533,
      "loss": 3.5395,
      "step": 22400
    },
    {
      "epoch": 2.420485175202156,
      "grad_norm": 0.6029559969902039,
      "learning_rate": 0.0004553135456017269,
      "loss": 3.569,
      "step": 22450
    },
    {
      "epoch": 2.4258760107816713,
      "grad_norm": 0.6211172342300415,
      "learning_rate": 0.0004549897463572585,
      "loss": 3.5407,
      "step": 22500
    },
    {
      "epoch": 2.431266846361186,
      "grad_norm": 0.5414947867393494,
      "learning_rate": 0.00045466594711279,
      "loss": 3.5477,
      "step": 22550
    },
    {
      "epoch": 2.4366576819407006,
      "grad_norm": 0.5086063742637634,
      "learning_rate": 0.00045434214786832164,
      "loss": 3.5477,
      "step": 22600
    },
    {
      "epoch": 2.442048517520216,
      "grad_norm": 0.5853644609451294,
      "learning_rate": 0.00045401834862385314,
      "loss": 3.5445,
      "step": 22650
    },
    {
      "epoch": 2.4474393530997305,
      "grad_norm": 0.5894696116447449,
      "learning_rate": 0.00045369454937938474,
      "loss": 3.5327,
      "step": 22700
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.5951809883117676,
      "learning_rate": 0.0004533707501349163,
      "loss": 3.5572,
      "step": 22750
    },
    {
      "epoch": 2.4582210242587603,
      "grad_norm": 0.5352622270584106,
      "learning_rate": 0.0004530469508904479,
      "loss": 3.5671,
      "step": 22800
    },
    {
      "epoch": 2.463611859838275,
      "grad_norm": 0.6040215492248535,
      "learning_rate": 0.00045272315164597945,
      "loss": 3.5349,
      "step": 22850
    },
    {
      "epoch": 2.4690026954177897,
      "grad_norm": 0.573526918888092,
      "learning_rate": 0.000452399352401511,
      "loss": 3.5441,
      "step": 22900
    },
    {
      "epoch": 2.4743935309973044,
      "grad_norm": 0.5855453610420227,
      "learning_rate": 0.0004520755531570426,
      "loss": 3.5475,
      "step": 22950
    },
    {
      "epoch": 2.4797843665768196,
      "grad_norm": 0.5997734665870667,
      "learning_rate": 0.00045175175391257415,
      "loss": 3.5392,
      "step": 23000
    },
    {
      "epoch": 2.4797843665768196,
      "eval_accuracy": 0.36637493004112187,
      "eval_loss": 3.5344290733337402,
      "eval_runtime": 183.3961,
      "eval_samples_per_second": 98.208,
      "eval_steps_per_second": 6.14,
      "step": 23000
    },
    {
      "epoch": 2.4851752021563343,
      "grad_norm": 0.5770864486694336,
      "learning_rate": 0.00045142795466810576,
      "loss": 3.5379,
      "step": 23050
    },
    {
      "epoch": 2.490566037735849,
      "grad_norm": 0.5228535532951355,
      "learning_rate": 0.0004511041554236373,
      "loss": 3.5355,
      "step": 23100
    },
    {
      "epoch": 2.4959568733153636,
      "grad_norm": 0.6147720217704773,
      "learning_rate": 0.0004507803561791689,
      "loss": 3.5329,
      "step": 23150
    },
    {
      "epoch": 2.501347708894879,
      "grad_norm": 0.5598917007446289,
      "learning_rate": 0.00045045655693470046,
      "loss": 3.5474,
      "step": 23200
    },
    {
      "epoch": 2.5067385444743935,
      "grad_norm": 0.545137345790863,
      "learning_rate": 0.00045013275769023207,
      "loss": 3.5582,
      "step": 23250
    },
    {
      "epoch": 2.512129380053908,
      "grad_norm": 0.5641994476318359,
      "learning_rate": 0.00044980895844576356,
      "loss": 3.5553,
      "step": 23300
    },
    {
      "epoch": 2.5175202156334233,
      "grad_norm": 0.5789488554000854,
      "learning_rate": 0.0004494851592012951,
      "loss": 3.5306,
      "step": 23350
    },
    {
      "epoch": 2.522911051212938,
      "grad_norm": 0.5591729283332825,
      "learning_rate": 0.0004491613599568267,
      "loss": 3.5349,
      "step": 23400
    },
    {
      "epoch": 2.5283018867924527,
      "grad_norm": 0.5694953799247742,
      "learning_rate": 0.00044883756071235827,
      "loss": 3.539,
      "step": 23450
    },
    {
      "epoch": 2.533692722371968,
      "grad_norm": 0.5788413286209106,
      "learning_rate": 0.0004485137614678899,
      "loss": 3.5186,
      "step": 23500
    },
    {
      "epoch": 2.5390835579514826,
      "grad_norm": 0.5994324088096619,
      "learning_rate": 0.0004481899622234214,
      "loss": 3.5555,
      "step": 23550
    },
    {
      "epoch": 2.5444743935309972,
      "grad_norm": 0.6380375623703003,
      "learning_rate": 0.00044786616297895303,
      "loss": 3.5488,
      "step": 23600
    },
    {
      "epoch": 2.5498652291105124,
      "grad_norm": 0.5977026224136353,
      "learning_rate": 0.0004475423637344846,
      "loss": 3.5424,
      "step": 23650
    },
    {
      "epoch": 2.555256064690027,
      "grad_norm": 0.5507897734642029,
      "learning_rate": 0.00044721856449001613,
      "loss": 3.5557,
      "step": 23700
    },
    {
      "epoch": 2.560646900269542,
      "grad_norm": 0.5803244709968567,
      "learning_rate": 0.00044689476524554774,
      "loss": 3.5641,
      "step": 23750
    },
    {
      "epoch": 2.5660377358490565,
      "grad_norm": 0.609472930431366,
      "learning_rate": 0.0004465709660010793,
      "loss": 3.5369,
      "step": 23800
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.5822898745536804,
      "learning_rate": 0.0004462471667566109,
      "loss": 3.5363,
      "step": 23850
    },
    {
      "epoch": 2.5768194070080863,
      "grad_norm": 0.5829198360443115,
      "learning_rate": 0.00044592336751214244,
      "loss": 3.534,
      "step": 23900
    },
    {
      "epoch": 2.582210242587601,
      "grad_norm": 0.5264182686805725,
      "learning_rate": 0.00044560604425256334,
      "loss": 3.5611,
      "step": 23950
    },
    {
      "epoch": 2.5876010781671157,
      "grad_norm": 0.5896156430244446,
      "learning_rate": 0.000445282245008095,
      "loss": 3.541,
      "step": 24000
    },
    {
      "epoch": 2.5876010781671157,
      "eval_accuracy": 0.3671398463713358,
      "eval_loss": 3.5273725986480713,
      "eval_runtime": 183.5998,
      "eval_samples_per_second": 98.099,
      "eval_steps_per_second": 6.133,
      "step": 24000
    },
    {
      "epoch": 2.592991913746631,
      "grad_norm": 0.6344112157821655,
      "learning_rate": 0.0004449584457636265,
      "loss": 3.5453,
      "step": 24050
    },
    {
      "epoch": 2.5983827493261455,
      "grad_norm": 0.6354188919067383,
      "learning_rate": 0.00044463464651915805,
      "loss": 3.5216,
      "step": 24100
    },
    {
      "epoch": 2.6037735849056602,
      "grad_norm": 0.5827237963676453,
      "learning_rate": 0.00044431084727468965,
      "loss": 3.5458,
      "step": 24150
    },
    {
      "epoch": 2.6091644204851754,
      "grad_norm": 0.5500607490539551,
      "learning_rate": 0.0004439870480302212,
      "loss": 3.5379,
      "step": 24200
    },
    {
      "epoch": 2.61455525606469,
      "grad_norm": 0.5530035495758057,
      "learning_rate": 0.0004436632487857528,
      "loss": 3.5422,
      "step": 24250
    },
    {
      "epoch": 2.6199460916442048,
      "grad_norm": 0.585949182510376,
      "learning_rate": 0.00044333944954128436,
      "loss": 3.5475,
      "step": 24300
    },
    {
      "epoch": 2.62533692722372,
      "grad_norm": 0.6067348718643188,
      "learning_rate": 0.00044301565029681596,
      "loss": 3.5488,
      "step": 24350
    },
    {
      "epoch": 2.6307277628032346,
      "grad_norm": 0.5916098356246948,
      "learning_rate": 0.0004426918510523475,
      "loss": 3.5251,
      "step": 24400
    },
    {
      "epoch": 2.6361185983827493,
      "grad_norm": 0.5653256177902222,
      "learning_rate": 0.0004423680518078791,
      "loss": 3.5518,
      "step": 24450
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.6101019978523254,
      "learning_rate": 0.00044204425256341067,
      "loss": 3.5162,
      "step": 24500
    },
    {
      "epoch": 2.6469002695417787,
      "grad_norm": 0.5783448815345764,
      "learning_rate": 0.0004417204533189422,
      "loss": 3.5196,
      "step": 24550
    },
    {
      "epoch": 2.652291105121294,
      "grad_norm": 0.5608351230621338,
      "learning_rate": 0.0004413966540744738,
      "loss": 3.5275,
      "step": 24600
    },
    {
      "epoch": 2.6576819407008085,
      "grad_norm": 0.533299446105957,
      "learning_rate": 0.0004410728548300053,
      "loss": 3.5358,
      "step": 24650
    },
    {
      "epoch": 2.6630727762803232,
      "grad_norm": 0.5338042378425598,
      "learning_rate": 0.0004407490555855369,
      "loss": 3.5365,
      "step": 24700
    },
    {
      "epoch": 2.6684636118598384,
      "grad_norm": 0.5940648913383484,
      "learning_rate": 0.0004404252563410685,
      "loss": 3.5506,
      "step": 24750
    },
    {
      "epoch": 2.673854447439353,
      "grad_norm": 0.5929331183433533,
      "learning_rate": 0.0004401014570966001,
      "loss": 3.5268,
      "step": 24800
    },
    {
      "epoch": 2.6792452830188678,
      "grad_norm": 0.5341362953186035,
      "learning_rate": 0.00043977765785213163,
      "loss": 3.5281,
      "step": 24850
    },
    {
      "epoch": 2.684636118598383,
      "grad_norm": 0.5370336174964905,
      "learning_rate": 0.0004394538586076632,
      "loss": 3.5383,
      "step": 24900
    },
    {
      "epoch": 2.6900269541778976,
      "grad_norm": 0.5800198316574097,
      "learning_rate": 0.0004391300593631948,
      "loss": 3.5321,
      "step": 24950
    },
    {
      "epoch": 2.6954177897574123,
      "grad_norm": 0.5767235159873962,
      "learning_rate": 0.00043880626011872634,
      "loss": 3.5477,
      "step": 25000
    },
    {
      "epoch": 2.6954177897574123,
      "eval_accuracy": 0.36847790668477115,
      "eval_loss": 3.51676607131958,
      "eval_runtime": 183.255,
      "eval_samples_per_second": 98.284,
      "eval_steps_per_second": 6.144,
      "step": 25000
    },
    {
      "epoch": 2.7008086253369274,
      "grad_norm": 0.6077183485031128,
      "learning_rate": 0.00043848246087425794,
      "loss": 3.5476,
      "step": 25050
    },
    {
      "epoch": 2.706199460916442,
      "grad_norm": 0.5808961391448975,
      "learning_rate": 0.0004381586616297895,
      "loss": 3.5578,
      "step": 25100
    },
    {
      "epoch": 2.711590296495957,
      "grad_norm": 0.6259551048278809,
      "learning_rate": 0.0004378348623853211,
      "loss": 3.5405,
      "step": 25150
    },
    {
      "epoch": 2.7169811320754715,
      "grad_norm": 0.6004087328910828,
      "learning_rate": 0.00043751106314085265,
      "loss": 3.5239,
      "step": 25200
    },
    {
      "epoch": 2.7223719676549867,
      "grad_norm": 0.5496945381164551,
      "learning_rate": 0.00043718726389638425,
      "loss": 3.5258,
      "step": 25250
    },
    {
      "epoch": 2.7277628032345014,
      "grad_norm": 0.6094425916671753,
      "learning_rate": 0.00043686346465191575,
      "loss": 3.536,
      "step": 25300
    },
    {
      "epoch": 2.733153638814016,
      "grad_norm": 0.5401579737663269,
      "learning_rate": 0.0004365396654074473,
      "loss": 3.5275,
      "step": 25350
    },
    {
      "epoch": 2.7385444743935308,
      "grad_norm": 0.5293307304382324,
      "learning_rate": 0.00043622234214786825,
      "loss": 3.5194,
      "step": 25400
    },
    {
      "epoch": 2.743935309973046,
      "grad_norm": 0.5775792598724365,
      "learning_rate": 0.00043589854290339985,
      "loss": 3.535,
      "step": 25450
    },
    {
      "epoch": 2.7493261455525606,
      "grad_norm": 0.5909881591796875,
      "learning_rate": 0.0004355747436589314,
      "loss": 3.5507,
      "step": 25500
    },
    {
      "epoch": 2.7547169811320753,
      "grad_norm": 0.5832464098930359,
      "learning_rate": 0.000435250944414463,
      "loss": 3.5444,
      "step": 25550
    },
    {
      "epoch": 2.7601078167115904,
      "grad_norm": 0.5698075294494629,
      "learning_rate": 0.00043492714516999456,
      "loss": 3.535,
      "step": 25600
    },
    {
      "epoch": 2.765498652291105,
      "grad_norm": 0.5500668883323669,
      "learning_rate": 0.0004346033459255261,
      "loss": 3.5232,
      "step": 25650
    },
    {
      "epoch": 2.77088948787062,
      "grad_norm": 0.5673616528511047,
      "learning_rate": 0.0004342795466810577,
      "loss": 3.5455,
      "step": 25700
    },
    {
      "epoch": 2.776280323450135,
      "grad_norm": 0.5733495354652405,
      "learning_rate": 0.00043395574743658927,
      "loss": 3.5295,
      "step": 25750
    },
    {
      "epoch": 2.7816711590296497,
      "grad_norm": 0.5701019763946533,
      "learning_rate": 0.00043363194819212087,
      "loss": 3.5307,
      "step": 25800
    },
    {
      "epoch": 2.7870619946091644,
      "grad_norm": 0.5702261924743652,
      "learning_rate": 0.0004333081489476524,
      "loss": 3.516,
      "step": 25850
    },
    {
      "epoch": 2.7924528301886795,
      "grad_norm": 0.5822378396987915,
      "learning_rate": 0.000432984349703184,
      "loss": 3.5372,
      "step": 25900
    },
    {
      "epoch": 2.797843665768194,
      "grad_norm": 0.550873875617981,
      "learning_rate": 0.0004326605504587155,
      "loss": 3.5085,
      "step": 25950
    },
    {
      "epoch": 2.803234501347709,
      "grad_norm": 0.5707508325576782,
      "learning_rate": 0.0004323367512142472,
      "loss": 3.5275,
      "step": 26000
    },
    {
      "epoch": 2.803234501347709,
      "eval_accuracy": 0.36893414016070414,
      "eval_loss": 3.510256290435791,
      "eval_runtime": 183.5418,
      "eval_samples_per_second": 98.13,
      "eval_steps_per_second": 6.135,
      "step": 26000
    },
    {
      "epoch": 2.8086253369272236,
      "grad_norm": 0.5530628561973572,
      "learning_rate": 0.0004320129519697787,
      "loss": 3.5324,
      "step": 26050
    },
    {
      "epoch": 2.8140161725067383,
      "grad_norm": 0.536308228969574,
      "learning_rate": 0.00043168915272531023,
      "loss": 3.5192,
      "step": 26100
    },
    {
      "epoch": 2.8194070080862534,
      "grad_norm": 0.5508840680122375,
      "learning_rate": 0.00043136535348084183,
      "loss": 3.5308,
      "step": 26150
    },
    {
      "epoch": 2.824797843665768,
      "grad_norm": 0.661864697933197,
      "learning_rate": 0.0004310415542363734,
      "loss": 3.5415,
      "step": 26200
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.5606989860534668,
      "learning_rate": 0.000430717754991905,
      "loss": 3.5281,
      "step": 26250
    },
    {
      "epoch": 2.835579514824798,
      "grad_norm": 0.5549330711364746,
      "learning_rate": 0.00043039395574743654,
      "loss": 3.5263,
      "step": 26300
    },
    {
      "epoch": 2.8409703504043127,
      "grad_norm": 0.5238580107688904,
      "learning_rate": 0.00043007015650296814,
      "loss": 3.5394,
      "step": 26350
    },
    {
      "epoch": 2.8463611859838274,
      "grad_norm": 0.5641552805900574,
      "learning_rate": 0.0004297463572584997,
      "loss": 3.5218,
      "step": 26400
    },
    {
      "epoch": 2.8517520215633425,
      "grad_norm": 0.5834696888923645,
      "learning_rate": 0.0004294225580140313,
      "loss": 3.5309,
      "step": 26450
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6112784147262573,
      "learning_rate": 0.00042909875876956285,
      "loss": 3.52,
      "step": 26500
    },
    {
      "epoch": 2.862533692722372,
      "grad_norm": 0.5841647386550903,
      "learning_rate": 0.0004287749595250944,
      "loss": 3.5032,
      "step": 26550
    },
    {
      "epoch": 2.867924528301887,
      "grad_norm": 0.5762807726860046,
      "learning_rate": 0.000428451160280626,
      "loss": 3.5271,
      "step": 26600
    },
    {
      "epoch": 2.8733153638814017,
      "grad_norm": 0.5195743441581726,
      "learning_rate": 0.0004281273610361575,
      "loss": 3.5221,
      "step": 26650
    },
    {
      "epoch": 2.8787061994609164,
      "grad_norm": 0.5398359298706055,
      "learning_rate": 0.0004278035617916891,
      "loss": 3.5163,
      "step": 26700
    },
    {
      "epoch": 2.884097035040431,
      "grad_norm": 0.5633315443992615,
      "learning_rate": 0.00042747976254722066,
      "loss": 3.5195,
      "step": 26750
    },
    {
      "epoch": 2.889487870619946,
      "grad_norm": 0.5828991532325745,
      "learning_rate": 0.00042715596330275226,
      "loss": 3.5012,
      "step": 26800
    },
    {
      "epoch": 2.894878706199461,
      "grad_norm": 0.6273923516273499,
      "learning_rate": 0.0004268321640582838,
      "loss": 3.5202,
      "step": 26850
    },
    {
      "epoch": 2.9002695417789757,
      "grad_norm": 0.5381441712379456,
      "learning_rate": 0.0004265083648138154,
      "loss": 3.5087,
      "step": 26900
    },
    {
      "epoch": 2.9056603773584904,
      "grad_norm": 0.596377432346344,
      "learning_rate": 0.00042618456556934697,
      "loss": 3.5073,
      "step": 26950
    },
    {
      "epoch": 2.9110512129380055,
      "grad_norm": 0.5934976935386658,
      "learning_rate": 0.0004258607663248785,
      "loss": 3.5341,
      "step": 27000
    },
    {
      "epoch": 2.9110512129380055,
      "eval_accuracy": 0.3700002422959398,
      "eval_loss": 3.500957727432251,
      "eval_runtime": 183.1714,
      "eval_samples_per_second": 98.329,
      "eval_steps_per_second": 6.147,
      "step": 27000
    },
    {
      "epoch": 2.91644204851752,
      "grad_norm": 0.6017852425575256,
      "learning_rate": 0.0004255369670804101,
      "loss": 3.5455,
      "step": 27050
    },
    {
      "epoch": 2.921832884097035,
      "grad_norm": 0.5310950875282288,
      "learning_rate": 0.0004252131678359417,
      "loss": 3.5272,
      "step": 27100
    },
    {
      "epoch": 2.92722371967655,
      "grad_norm": 0.5664175152778625,
      "learning_rate": 0.0004248893685914733,
      "loss": 3.5418,
      "step": 27150
    },
    {
      "epoch": 2.9326145552560647,
      "grad_norm": 0.6421070694923401,
      "learning_rate": 0.00042456556934700483,
      "loss": 3.5271,
      "step": 27200
    },
    {
      "epoch": 2.9380053908355794,
      "grad_norm": 0.5426092743873596,
      "learning_rate": 0.00042424177010253643,
      "loss": 3.542,
      "step": 27250
    },
    {
      "epoch": 2.9433962264150946,
      "grad_norm": 0.597486138343811,
      "learning_rate": 0.00042391797085806793,
      "loss": 3.5224,
      "step": 27300
    },
    {
      "epoch": 2.9487870619946093,
      "grad_norm": 0.6039035320281982,
      "learning_rate": 0.0004235941716135995,
      "loss": 3.5114,
      "step": 27350
    },
    {
      "epoch": 2.954177897574124,
      "grad_norm": 0.5943632125854492,
      "learning_rate": 0.0004232703723691311,
      "loss": 3.5101,
      "step": 27400
    },
    {
      "epoch": 2.9595687331536387,
      "grad_norm": 0.5611184239387512,
      "learning_rate": 0.00042294657312466264,
      "loss": 3.527,
      "step": 27450
    },
    {
      "epoch": 2.964959568733154,
      "grad_norm": 0.5667484402656555,
      "learning_rate": 0.00042262277388019424,
      "loss": 3.5303,
      "step": 27500
    },
    {
      "epoch": 2.9703504043126685,
      "grad_norm": 0.5914982557296753,
      "learning_rate": 0.0004222989746357258,
      "loss": 3.5148,
      "step": 27550
    },
    {
      "epoch": 2.975741239892183,
      "grad_norm": 0.5547810792922974,
      "learning_rate": 0.0004219751753912574,
      "loss": 3.5162,
      "step": 27600
    },
    {
      "epoch": 2.981132075471698,
      "grad_norm": 0.5683316588401794,
      "learning_rate": 0.00042165137614678895,
      "loss": 3.5384,
      "step": 27650
    },
    {
      "epoch": 2.986522911051213,
      "grad_norm": 0.5850192904472351,
      "learning_rate": 0.00042132757690232055,
      "loss": 3.5048,
      "step": 27700
    },
    {
      "epoch": 2.9919137466307277,
      "grad_norm": 0.5472140908241272,
      "learning_rate": 0.0004210037776578521,
      "loss": 3.4964,
      "step": 27750
    },
    {
      "epoch": 2.9973045822102424,
      "grad_norm": 0.6173052787780762,
      "learning_rate": 0.00042067997841338365,
      "loss": 3.5167,
      "step": 27800
    },
    {
      "epoch": 3.0026954177897576,
      "grad_norm": 0.5687317848205566,
      "learning_rate": 0.00042035617916891526,
      "loss": 3.4765,
      "step": 27850
    },
    {
      "epoch": 3.0080862533692723,
      "grad_norm": 0.5316717624664307,
      "learning_rate": 0.0004200323799244468,
      "loss": 3.4207,
      "step": 27900
    },
    {
      "epoch": 3.013477088948787,
      "grad_norm": 0.6342532634735107,
      "learning_rate": 0.0004197085806799784,
      "loss": 3.3998,
      "step": 27950
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.5721489787101746,
      "learning_rate": 0.0004193847814355099,
      "loss": 3.4218,
      "step": 28000
    },
    {
      "epoch": 3.018867924528302,
      "eval_accuracy": 0.37088456814986187,
      "eval_loss": 3.4964616298675537,
      "eval_runtime": 183.4604,
      "eval_samples_per_second": 98.174,
      "eval_steps_per_second": 6.138,
      "step": 28000
    },
    {
      "epoch": 3.024258760107817,
      "grad_norm": 0.5991702079772949,
      "learning_rate": 0.0004190609821910415,
      "loss": 3.4134,
      "step": 28050
    },
    {
      "epoch": 3.0296495956873315,
      "grad_norm": 0.5457400679588318,
      "learning_rate": 0.00041873718294657306,
      "loss": 3.4297,
      "step": 28100
    },
    {
      "epoch": 3.035040431266846,
      "grad_norm": 0.5663076043128967,
      "learning_rate": 0.00041841338370210467,
      "loss": 3.431,
      "step": 28150
    },
    {
      "epoch": 3.0404312668463613,
      "grad_norm": 0.6150392889976501,
      "learning_rate": 0.0004180895844576362,
      "loss": 3.4296,
      "step": 28200
    },
    {
      "epoch": 3.045822102425876,
      "grad_norm": 0.5990293622016907,
      "learning_rate": 0.00041776578521316777,
      "loss": 3.435,
      "step": 28250
    },
    {
      "epoch": 3.0512129380053907,
      "grad_norm": 0.5907416939735413,
      "learning_rate": 0.0004174419859686994,
      "loss": 3.4342,
      "step": 28300
    },
    {
      "epoch": 3.056603773584906,
      "grad_norm": 0.5918059945106506,
      "learning_rate": 0.0004171181867242309,
      "loss": 3.4428,
      "step": 28350
    },
    {
      "epoch": 3.0619946091644206,
      "grad_norm": 0.6297253966331482,
      "learning_rate": 0.00041679438747976253,
      "loss": 3.4174,
      "step": 28400
    },
    {
      "epoch": 3.0673854447439353,
      "grad_norm": 0.6004771590232849,
      "learning_rate": 0.0004164705882352941,
      "loss": 3.4398,
      "step": 28450
    },
    {
      "epoch": 3.07277628032345,
      "grad_norm": 0.6342185139656067,
      "learning_rate": 0.0004161467889908257,
      "loss": 3.4255,
      "step": 28500
    },
    {
      "epoch": 3.078167115902965,
      "grad_norm": 0.5759187340736389,
      "learning_rate": 0.00041582298974635724,
      "loss": 3.4223,
      "step": 28550
    },
    {
      "epoch": 3.08355795148248,
      "grad_norm": 0.5629438757896423,
      "learning_rate": 0.00041549919050188884,
      "loss": 3.4434,
      "step": 28600
    },
    {
      "epoch": 3.0889487870619945,
      "grad_norm": 0.6492778658866882,
      "learning_rate": 0.00041517539125742034,
      "loss": 3.4449,
      "step": 28650
    },
    {
      "epoch": 3.0943396226415096,
      "grad_norm": 0.5692011713981628,
      "learning_rate": 0.0004148515920129519,
      "loss": 3.447,
      "step": 28700
    },
    {
      "epoch": 3.0997304582210243,
      "grad_norm": 0.9911279678344727,
      "learning_rate": 0.0004145277927684835,
      "loss": 3.443,
      "step": 28750
    },
    {
      "epoch": 3.105121293800539,
      "grad_norm": 0.6052446365356445,
      "learning_rate": 0.00041420399352401504,
      "loss": 3.4396,
      "step": 28800
    },
    {
      "epoch": 3.1105121293800537,
      "grad_norm": 0.5958963632583618,
      "learning_rate": 0.00041388019427954665,
      "loss": 3.4439,
      "step": 28850
    },
    {
      "epoch": 3.115902964959569,
      "grad_norm": 0.5762249231338501,
      "learning_rate": 0.0004135563950350782,
      "loss": 3.4425,
      "step": 28900
    },
    {
      "epoch": 3.1212938005390836,
      "grad_norm": 0.5552219152450562,
      "learning_rate": 0.0004132325957906098,
      "loss": 3.4359,
      "step": 28950
    },
    {
      "epoch": 3.1266846361185983,
      "grad_norm": 0.59885573387146,
      "learning_rate": 0.00041290879654614135,
      "loss": 3.4461,
      "step": 29000
    },
    {
      "epoch": 3.1266846361185983,
      "eval_accuracy": 0.37161156462222855,
      "eval_loss": 3.489034414291382,
      "eval_runtime": 183.4689,
      "eval_samples_per_second": 98.169,
      "eval_steps_per_second": 6.137,
      "step": 29000
    },
    {
      "epoch": 3.1320754716981134,
      "grad_norm": 0.5445829033851624,
      "learning_rate": 0.00041258499730167296,
      "loss": 3.4258,
      "step": 29050
    },
    {
      "epoch": 3.137466307277628,
      "grad_norm": 0.5851534008979797,
      "learning_rate": 0.0004122611980572045,
      "loss": 3.4514,
      "step": 29100
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.5788666605949402,
      "learning_rate": 0.00041193739881273606,
      "loss": 3.4466,
      "step": 29150
    },
    {
      "epoch": 3.1482479784366575,
      "grad_norm": 0.6072415113449097,
      "learning_rate": 0.00041161359956826766,
      "loss": 3.4602,
      "step": 29200
    },
    {
      "epoch": 3.1536388140161726,
      "grad_norm": 0.6163262724876404,
      "learning_rate": 0.0004112898003237992,
      "loss": 3.4614,
      "step": 29250
    },
    {
      "epoch": 3.1590296495956873,
      "grad_norm": 0.612508237361908,
      "learning_rate": 0.0004109660010793308,
      "loss": 3.4541,
      "step": 29300
    },
    {
      "epoch": 3.164420485175202,
      "grad_norm": 0.5615676045417786,
      "learning_rate": 0.0004106422018348623,
      "loss": 3.4416,
      "step": 29350
    },
    {
      "epoch": 3.169811320754717,
      "grad_norm": 0.6066969633102417,
      "learning_rate": 0.0004103184025903939,
      "loss": 3.4516,
      "step": 29400
    },
    {
      "epoch": 3.175202156334232,
      "grad_norm": 0.603179395198822,
      "learning_rate": 0.00040999460334592547,
      "loss": 3.4643,
      "step": 29450
    },
    {
      "epoch": 3.1805929919137466,
      "grad_norm": 0.5823822021484375,
      "learning_rate": 0.000409670804101457,
      "loss": 3.4381,
      "step": 29500
    },
    {
      "epoch": 3.1859838274932613,
      "grad_norm": 0.5800785422325134,
      "learning_rate": 0.000409353480841878,
      "loss": 3.4505,
      "step": 29550
    },
    {
      "epoch": 3.1913746630727764,
      "grad_norm": 0.5989530086517334,
      "learning_rate": 0.0004090296815974096,
      "loss": 3.4613,
      "step": 29600
    },
    {
      "epoch": 3.196765498652291,
      "grad_norm": 0.5782071352005005,
      "learning_rate": 0.00040870588235294113,
      "loss": 3.4469,
      "step": 29650
    },
    {
      "epoch": 3.202156334231806,
      "grad_norm": 0.6211130023002625,
      "learning_rate": 0.00040838208310847273,
      "loss": 3.4652,
      "step": 29700
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.6505089998245239,
      "learning_rate": 0.0004080582838640043,
      "loss": 3.4618,
      "step": 29750
    },
    {
      "epoch": 3.2129380053908356,
      "grad_norm": 0.676662027835846,
      "learning_rate": 0.0004077344846195359,
      "loss": 3.437,
      "step": 29800
    },
    {
      "epoch": 3.2183288409703503,
      "grad_norm": 0.6046519875526428,
      "learning_rate": 0.00040741068537506744,
      "loss": 3.4592,
      "step": 29850
    },
    {
      "epoch": 3.223719676549865,
      "grad_norm": 0.6096304655075073,
      "learning_rate": 0.000407086886130599,
      "loss": 3.4508,
      "step": 29900
    },
    {
      "epoch": 3.22911051212938,
      "grad_norm": 0.5660011172294617,
      "learning_rate": 0.0004067630868861306,
      "loss": 3.4471,
      "step": 29950
    },
    {
      "epoch": 3.234501347708895,
      "grad_norm": 0.6080735921859741,
      "learning_rate": 0.0004064392876416621,
      "loss": 3.4554,
      "step": 30000
    },
    {
      "epoch": 3.234501347708895,
      "eval_accuracy": 0.3721109332946239,
      "eval_loss": 3.484154224395752,
      "eval_runtime": 183.5489,
      "eval_samples_per_second": 98.126,
      "eval_steps_per_second": 6.135,
      "step": 30000
    },
    {
      "epoch": 3.2398921832884096,
      "grad_norm": 0.5689899325370789,
      "learning_rate": 0.0004061154883971937,
      "loss": 3.4345,
      "step": 30050
    },
    {
      "epoch": 3.2452830188679247,
      "grad_norm": 0.6104323267936707,
      "learning_rate": 0.00040579168915272525,
      "loss": 3.451,
      "step": 30100
    },
    {
      "epoch": 3.2506738544474394,
      "grad_norm": 0.5824128985404968,
      "learning_rate": 0.00040546788990825685,
      "loss": 3.4518,
      "step": 30150
    },
    {
      "epoch": 3.256064690026954,
      "grad_norm": 0.5787423849105835,
      "learning_rate": 0.0004051440906637884,
      "loss": 3.4673,
      "step": 30200
    },
    {
      "epoch": 3.2614555256064692,
      "grad_norm": 0.5913103222846985,
      "learning_rate": 0.00040482029141931995,
      "loss": 3.4537,
      "step": 30250
    },
    {
      "epoch": 3.266846361185984,
      "grad_norm": 0.6089011430740356,
      "learning_rate": 0.00040449649217485156,
      "loss": 3.4319,
      "step": 30300
    },
    {
      "epoch": 3.2722371967654986,
      "grad_norm": 0.5752812623977661,
      "learning_rate": 0.0004041726929303831,
      "loss": 3.4487,
      "step": 30350
    },
    {
      "epoch": 3.2776280323450133,
      "grad_norm": 0.6006338596343994,
      "learning_rate": 0.0004038488936859147,
      "loss": 3.4607,
      "step": 30400
    },
    {
      "epoch": 3.2830188679245285,
      "grad_norm": 0.7376881241798401,
      "learning_rate": 0.00040352509444144626,
      "loss": 3.45,
      "step": 30450
    },
    {
      "epoch": 3.288409703504043,
      "grad_norm": 0.6478651165962219,
      "learning_rate": 0.00040320129519697787,
      "loss": 3.4595,
      "step": 30500
    },
    {
      "epoch": 3.293800539083558,
      "grad_norm": 0.6603450775146484,
      "learning_rate": 0.0004028774959525094,
      "loss": 3.4558,
      "step": 30550
    },
    {
      "epoch": 3.2991913746630726,
      "grad_norm": 0.5964590311050415,
      "learning_rate": 0.000402553696708041,
      "loss": 3.4487,
      "step": 30600
    },
    {
      "epoch": 3.3045822102425877,
      "grad_norm": 0.5935302376747131,
      "learning_rate": 0.0004022298974635726,
      "loss": 3.4468,
      "step": 30650
    },
    {
      "epoch": 3.3099730458221024,
      "grad_norm": 0.6866962313652039,
      "learning_rate": 0.00040190609821910407,
      "loss": 3.4606,
      "step": 30700
    },
    {
      "epoch": 3.315363881401617,
      "grad_norm": 0.5697274804115295,
      "learning_rate": 0.0004015822989746357,
      "loss": 3.4381,
      "step": 30750
    },
    {
      "epoch": 3.3207547169811322,
      "grad_norm": 0.5480004549026489,
      "learning_rate": 0.0004012584997301672,
      "loss": 3.4342,
      "step": 30800
    },
    {
      "epoch": 3.326145552560647,
      "grad_norm": 0.591159999370575,
      "learning_rate": 0.00040093470048569883,
      "loss": 3.4336,
      "step": 30850
    },
    {
      "epoch": 3.3315363881401616,
      "grad_norm": 0.564811646938324,
      "learning_rate": 0.0004006109012412304,
      "loss": 3.4658,
      "step": 30900
    },
    {
      "epoch": 3.3369272237196768,
      "grad_norm": 0.6144384145736694,
      "learning_rate": 0.000400287101996762,
      "loss": 3.4619,
      "step": 30950
    },
    {
      "epoch": 3.3423180592991915,
      "grad_norm": 0.5722789168357849,
      "learning_rate": 0.00039996330275229354,
      "loss": 3.4404,
      "step": 31000
    },
    {
      "epoch": 3.3423180592991915,
      "eval_accuracy": 0.37290725030941624,
      "eval_loss": 3.476534605026245,
      "eval_runtime": 183.314,
      "eval_samples_per_second": 98.252,
      "eval_steps_per_second": 6.142,
      "step": 31000
    },
    {
      "epoch": 3.347708894878706,
      "grad_norm": 0.5752816200256348,
      "learning_rate": 0.00039963950350782514,
      "loss": 3.4498,
      "step": 31050
    },
    {
      "epoch": 3.353099730458221,
      "grad_norm": 0.6340233087539673,
      "learning_rate": 0.0003993157042633567,
      "loss": 3.4681,
      "step": 31100
    },
    {
      "epoch": 3.358490566037736,
      "grad_norm": 0.6339342594146729,
      "learning_rate": 0.00039899190501888824,
      "loss": 3.4499,
      "step": 31150
    },
    {
      "epoch": 3.3638814016172507,
      "grad_norm": 0.5568013191223145,
      "learning_rate": 0.00039866810577441985,
      "loss": 3.4723,
      "step": 31200
    },
    {
      "epoch": 3.3692722371967654,
      "grad_norm": 0.5924887657165527,
      "learning_rate": 0.0003983443065299514,
      "loss": 3.4685,
      "step": 31250
    },
    {
      "epoch": 3.37466307277628,
      "grad_norm": 0.6027880907058716,
      "learning_rate": 0.000398020507285483,
      "loss": 3.4603,
      "step": 31300
    },
    {
      "epoch": 3.3800539083557952,
      "grad_norm": 0.5847180485725403,
      "learning_rate": 0.0003976967080410145,
      "loss": 3.4554,
      "step": 31350
    },
    {
      "epoch": 3.38544474393531,
      "grad_norm": 0.6057071089744568,
      "learning_rate": 0.0003973729087965461,
      "loss": 3.4646,
      "step": 31400
    },
    {
      "epoch": 3.3908355795148246,
      "grad_norm": 0.6064717769622803,
      "learning_rate": 0.00039704910955207765,
      "loss": 3.4559,
      "step": 31450
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.7252898216247559,
      "learning_rate": 0.00039672531030760926,
      "loss": 3.4556,
      "step": 31500
    },
    {
      "epoch": 3.4016172506738545,
      "grad_norm": 0.5483014583587646,
      "learning_rate": 0.00039640798704803016,
      "loss": 3.4635,
      "step": 31550
    },
    {
      "epoch": 3.407008086253369,
      "grad_norm": 0.5969088673591614,
      "learning_rate": 0.00039608418780356176,
      "loss": 3.4526,
      "step": 31600
    },
    {
      "epoch": 3.4123989218328843,
      "grad_norm": 0.6174740791320801,
      "learning_rate": 0.0003957603885590933,
      "loss": 3.4417,
      "step": 31650
    },
    {
      "epoch": 3.417789757412399,
      "grad_norm": 0.5688740611076355,
      "learning_rate": 0.0003954365893146249,
      "loss": 3.4588,
      "step": 31700
    },
    {
      "epoch": 3.4231805929919137,
      "grad_norm": 0.6028859615325928,
      "learning_rate": 0.00039511279007015647,
      "loss": 3.4622,
      "step": 31750
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.5561661124229431,
      "learning_rate": 0.00039478899082568807,
      "loss": 3.4419,
      "step": 31800
    },
    {
      "epoch": 3.4339622641509435,
      "grad_norm": 0.6191286444664001,
      "learning_rate": 0.0003944651915812196,
      "loss": 3.4556,
      "step": 31850
    },
    {
      "epoch": 3.439353099730458,
      "grad_norm": 0.6074486374855042,
      "learning_rate": 0.0003941413923367512,
      "loss": 3.468,
      "step": 31900
    },
    {
      "epoch": 3.444743935309973,
      "grad_norm": 0.5516985058784485,
      "learning_rate": 0.0003938175930922828,
      "loss": 3.4587,
      "step": 31950
    },
    {
      "epoch": 3.450134770889488,
      "grad_norm": 0.5780602097511292,
      "learning_rate": 0.0003934937938478143,
      "loss": 3.4539,
      "step": 32000
    },
    {
      "epoch": 3.450134770889488,
      "eval_accuracy": 0.3738237374181314,
      "eval_loss": 3.4704768657684326,
      "eval_runtime": 183.3826,
      "eval_samples_per_second": 98.215,
      "eval_steps_per_second": 6.14,
      "step": 32000
    },
    {
      "epoch": 3.4555256064690028,
      "grad_norm": 0.5890681147575378,
      "learning_rate": 0.0003931699946033459,
      "loss": 3.4567,
      "step": 32050
    },
    {
      "epoch": 3.4609164420485174,
      "grad_norm": 0.6089186668395996,
      "learning_rate": 0.00039284619535887743,
      "loss": 3.4616,
      "step": 32100
    },
    {
      "epoch": 3.466307277628032,
      "grad_norm": 0.6278788447380066,
      "learning_rate": 0.00039252239611440904,
      "loss": 3.4596,
      "step": 32150
    },
    {
      "epoch": 3.4716981132075473,
      "grad_norm": 0.6116218566894531,
      "learning_rate": 0.0003921985968699406,
      "loss": 3.439,
      "step": 32200
    },
    {
      "epoch": 3.477088948787062,
      "grad_norm": 0.5771948099136353,
      "learning_rate": 0.0003918747976254722,
      "loss": 3.4533,
      "step": 32250
    },
    {
      "epoch": 3.4824797843665767,
      "grad_norm": 0.5892335176467896,
      "learning_rate": 0.00039155099838100374,
      "loss": 3.4757,
      "step": 32300
    },
    {
      "epoch": 3.487870619946092,
      "grad_norm": 0.6450402140617371,
      "learning_rate": 0.0003912271991365353,
      "loss": 3.4477,
      "step": 32350
    },
    {
      "epoch": 3.4932614555256065,
      "grad_norm": 0.585907518863678,
      "learning_rate": 0.0003909033998920669,
      "loss": 3.4573,
      "step": 32400
    },
    {
      "epoch": 3.498652291105121,
      "grad_norm": 0.6161054372787476,
      "learning_rate": 0.00039057960064759845,
      "loss": 3.4465,
      "step": 32450
    },
    {
      "epoch": 3.5040431266846364,
      "grad_norm": 0.5852574110031128,
      "learning_rate": 0.00039025580140313005,
      "loss": 3.4551,
      "step": 32500
    },
    {
      "epoch": 3.509433962264151,
      "grad_norm": 0.6066722273826599,
      "learning_rate": 0.0003899320021586616,
      "loss": 3.4501,
      "step": 32550
    },
    {
      "epoch": 3.5148247978436657,
      "grad_norm": 0.5780394077301025,
      "learning_rate": 0.0003896082029141932,
      "loss": 3.4549,
      "step": 32600
    },
    {
      "epoch": 3.5202156334231804,
      "grad_norm": 0.6035788059234619,
      "learning_rate": 0.00038928440366972476,
      "loss": 3.448,
      "step": 32650
    },
    {
      "epoch": 3.525606469002695,
      "grad_norm": 0.6049900650978088,
      "learning_rate": 0.00038896060442525636,
      "loss": 3.4605,
      "step": 32700
    },
    {
      "epoch": 3.5309973045822103,
      "grad_norm": 0.6193360686302185,
      "learning_rate": 0.00038863680518078786,
      "loss": 3.4481,
      "step": 32750
    },
    {
      "epoch": 3.536388140161725,
      "grad_norm": 0.6224262714385986,
      "learning_rate": 0.0003883130059363194,
      "loss": 3.4517,
      "step": 32800
    },
    {
      "epoch": 3.5417789757412397,
      "grad_norm": 0.6505341529846191,
      "learning_rate": 0.000387989206691851,
      "loss": 3.4337,
      "step": 32850
    },
    {
      "epoch": 3.547169811320755,
      "grad_norm": 0.5855464935302734,
      "learning_rate": 0.00038766540744738256,
      "loss": 3.4484,
      "step": 32900
    },
    {
      "epoch": 3.5525606469002695,
      "grad_norm": 0.5817187428474426,
      "learning_rate": 0.00038734160820291417,
      "loss": 3.4388,
      "step": 32950
    },
    {
      "epoch": 3.557951482479784,
      "grad_norm": 0.6249986290931702,
      "learning_rate": 0.0003870178089584457,
      "loss": 3.4407,
      "step": 33000
    },
    {
      "epoch": 3.557951482479784,
      "eval_accuracy": 0.3741863121047683,
      "eval_loss": 3.4653964042663574,
      "eval_runtime": 183.4614,
      "eval_samples_per_second": 98.173,
      "eval_steps_per_second": 6.138,
      "step": 33000
    },
    {
      "epoch": 3.5633423180592994,
      "grad_norm": 0.6825558543205261,
      "learning_rate": 0.0003866940097139773,
      "loss": 3.4488,
      "step": 33050
    },
    {
      "epoch": 3.568733153638814,
      "grad_norm": 0.6727454662322998,
      "learning_rate": 0.0003863702104695089,
      "loss": 3.4594,
      "step": 33100
    },
    {
      "epoch": 3.5741239892183287,
      "grad_norm": 0.5715111494064331,
      "learning_rate": 0.0003860464112250404,
      "loss": 3.4315,
      "step": 33150
    },
    {
      "epoch": 3.579514824797844,
      "grad_norm": 0.5947990417480469,
      "learning_rate": 0.00038572261198057203,
      "loss": 3.4395,
      "step": 33200
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.6582793593406677,
      "learning_rate": 0.0003853988127361036,
      "loss": 3.4704,
      "step": 33250
    },
    {
      "epoch": 3.5902964959568733,
      "grad_norm": 0.6481404900550842,
      "learning_rate": 0.0003850750134916352,
      "loss": 3.4585,
      "step": 33300
    },
    {
      "epoch": 3.595687331536388,
      "grad_norm": 0.5785871148109436,
      "learning_rate": 0.0003847512142471667,
      "loss": 3.446,
      "step": 33350
    },
    {
      "epoch": 3.601078167115903,
      "grad_norm": 0.6385973691940308,
      "learning_rate": 0.0003844274150026983,
      "loss": 3.4475,
      "step": 33400
    },
    {
      "epoch": 3.606469002695418,
      "grad_norm": 0.6048685908317566,
      "learning_rate": 0.00038410361575822984,
      "loss": 3.4571,
      "step": 33450
    },
    {
      "epoch": 3.6118598382749325,
      "grad_norm": 0.5991024971008301,
      "learning_rate": 0.00038377981651376144,
      "loss": 3.4378,
      "step": 33500
    },
    {
      "epoch": 3.617250673854447,
      "grad_norm": 0.6056102514266968,
      "learning_rate": 0.00038346249325418234,
      "loss": 3.4473,
      "step": 33550
    },
    {
      "epoch": 3.6226415094339623,
      "grad_norm": 0.6247057914733887,
      "learning_rate": 0.00038313869400971395,
      "loss": 3.4526,
      "step": 33600
    },
    {
      "epoch": 3.628032345013477,
      "grad_norm": 0.6171483397483826,
      "learning_rate": 0.0003828148947652455,
      "loss": 3.4612,
      "step": 33650
    },
    {
      "epoch": 3.6334231805929917,
      "grad_norm": 0.5629798769950867,
      "learning_rate": 0.0003824910955207771,
      "loss": 3.4562,
      "step": 33700
    },
    {
      "epoch": 3.638814016172507,
      "grad_norm": 0.5999249815940857,
      "learning_rate": 0.00038216729627630865,
      "loss": 3.4415,
      "step": 33750
    },
    {
      "epoch": 3.6442048517520216,
      "grad_norm": 0.5683233737945557,
      "learning_rate": 0.00038184349703184026,
      "loss": 3.4568,
      "step": 33800
    },
    {
      "epoch": 3.6495956873315363,
      "grad_norm": 0.6162096858024597,
      "learning_rate": 0.0003815196977873718,
      "loss": 3.4557,
      "step": 33850
    },
    {
      "epoch": 3.6549865229110514,
      "grad_norm": 0.6410558223724365,
      "learning_rate": 0.0003811958985429034,
      "loss": 3.4458,
      "step": 33900
    },
    {
      "epoch": 3.660377358490566,
      "grad_norm": 0.6777019500732422,
      "learning_rate": 0.00038087209929843496,
      "loss": 3.4405,
      "step": 33950
    },
    {
      "epoch": 3.665768194070081,
      "grad_norm": 0.6722299456596375,
      "learning_rate": 0.00038054830005396646,
      "loss": 3.4573,
      "step": 34000
    },
    {
      "epoch": 3.665768194070081,
      "eval_accuracy": 0.37471262669334165,
      "eval_loss": 3.458420991897583,
      "eval_runtime": 183.3402,
      "eval_samples_per_second": 98.238,
      "eval_steps_per_second": 6.142,
      "step": 34000
    },
    {
      "epoch": 3.671159029649596,
      "grad_norm": 0.6154153943061829,
      "learning_rate": 0.00038022450080949806,
      "loss": 3.4406,
      "step": 34050
    },
    {
      "epoch": 3.6765498652291106,
      "grad_norm": 0.5900728702545166,
      "learning_rate": 0.0003799007015650296,
      "loss": 3.4567,
      "step": 34100
    },
    {
      "epoch": 3.6819407008086253,
      "grad_norm": 0.6302698850631714,
      "learning_rate": 0.0003795769023205612,
      "loss": 3.4545,
      "step": 34150
    },
    {
      "epoch": 3.68733153638814,
      "grad_norm": 0.5984343886375427,
      "learning_rate": 0.00037925310307609277,
      "loss": 3.4368,
      "step": 34200
    },
    {
      "epoch": 3.6927223719676547,
      "grad_norm": 0.6989060044288635,
      "learning_rate": 0.0003789293038316244,
      "loss": 3.4501,
      "step": 34250
    },
    {
      "epoch": 3.69811320754717,
      "grad_norm": 0.6396805644035339,
      "learning_rate": 0.0003786055045871559,
      "loss": 3.4404,
      "step": 34300
    },
    {
      "epoch": 3.7035040431266846,
      "grad_norm": 0.560369074344635,
      "learning_rate": 0.0003782817053426875,
      "loss": 3.4632,
      "step": 34350
    },
    {
      "epoch": 3.7088948787061993,
      "grad_norm": 0.6261559128761292,
      "learning_rate": 0.0003779579060982191,
      "loss": 3.4469,
      "step": 34400
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.6551036238670349,
      "learning_rate": 0.00037763410685375063,
      "loss": 3.433,
      "step": 34450
    },
    {
      "epoch": 3.719676549865229,
      "grad_norm": 0.7240360975265503,
      "learning_rate": 0.00037731030760928223,
      "loss": 3.4508,
      "step": 34500
    },
    {
      "epoch": 3.725067385444744,
      "grad_norm": 0.6055688858032227,
      "learning_rate": 0.0003769865083648138,
      "loss": 3.4537,
      "step": 34550
    },
    {
      "epoch": 3.730458221024259,
      "grad_norm": 0.6346741318702698,
      "learning_rate": 0.0003766627091203454,
      "loss": 3.4566,
      "step": 34600
    },
    {
      "epoch": 3.7358490566037736,
      "grad_norm": 0.5869565010070801,
      "learning_rate": 0.00037633890987587694,
      "loss": 3.4455,
      "step": 34650
    },
    {
      "epoch": 3.7412398921832883,
      "grad_norm": 0.606854259967804,
      "learning_rate": 0.00037601511063140855,
      "loss": 3.4511,
      "step": 34700
    },
    {
      "epoch": 3.7466307277628035,
      "grad_norm": 0.6614265441894531,
      "learning_rate": 0.00037569131138694004,
      "loss": 3.4331,
      "step": 34750
    },
    {
      "epoch": 3.752021563342318,
      "grad_norm": 0.6091951727867126,
      "learning_rate": 0.0003753675121424716,
      "loss": 3.4687,
      "step": 34800
    },
    {
      "epoch": 3.757412398921833,
      "grad_norm": 0.6363885402679443,
      "learning_rate": 0.0003750437128980032,
      "loss": 3.454,
      "step": 34850
    },
    {
      "epoch": 3.7628032345013476,
      "grad_norm": 0.6237848997116089,
      "learning_rate": 0.00037471991365353475,
      "loss": 3.4517,
      "step": 34900
    },
    {
      "epoch": 3.7681940700808623,
      "grad_norm": 0.6054222583770752,
      "learning_rate": 0.00037439611440906635,
      "loss": 3.4515,
      "step": 34950
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.6497372984886169,
      "learning_rate": 0.0003740723151645979,
      "loss": 3.4392,
      "step": 35000
    },
    {
      "epoch": 3.7735849056603774,
      "eval_accuracy": 0.3750332613652822,
      "eval_loss": 3.4538044929504395,
      "eval_runtime": 183.7049,
      "eval_samples_per_second": 98.043,
      "eval_steps_per_second": 6.129,
      "step": 35000
    },
    {
      "epoch": 3.778975741239892,
      "grad_norm": 0.7446255683898926,
      "learning_rate": 0.0003737485159201295,
      "loss": 3.4523,
      "step": 35050
    },
    {
      "epoch": 3.784366576819407,
      "grad_norm": 0.5870745778083801,
      "learning_rate": 0.00037342471667566106,
      "loss": 3.4448,
      "step": 35100
    },
    {
      "epoch": 3.789757412398922,
      "grad_norm": 0.6146149635314941,
      "learning_rate": 0.00037310091743119266,
      "loss": 3.4419,
      "step": 35150
    },
    {
      "epoch": 3.7951482479784366,
      "grad_norm": 0.7114392518997192,
      "learning_rate": 0.0003727771181867242,
      "loss": 3.4663,
      "step": 35200
    },
    {
      "epoch": 3.8005390835579513,
      "grad_norm": 0.6365261077880859,
      "learning_rate": 0.00037245331894225576,
      "loss": 3.4394,
      "step": 35250
    },
    {
      "epoch": 3.8059299191374665,
      "grad_norm": 0.6266224980354309,
      "learning_rate": 0.00037212951969778737,
      "loss": 3.4338,
      "step": 35300
    },
    {
      "epoch": 3.811320754716981,
      "grad_norm": 0.5985236167907715,
      "learning_rate": 0.0003718121964382083,
      "loss": 3.4362,
      "step": 35350
    },
    {
      "epoch": 3.816711590296496,
      "grad_norm": 0.6070831418037415,
      "learning_rate": 0.0003714883971937398,
      "loss": 3.4484,
      "step": 35400
    },
    {
      "epoch": 3.822102425876011,
      "grad_norm": 0.556121826171875,
      "learning_rate": 0.0003711645979492714,
      "loss": 3.4579,
      "step": 35450
    },
    {
      "epoch": 3.8274932614555257,
      "grad_norm": 0.5937209129333496,
      "learning_rate": 0.00037084079870480297,
      "loss": 3.4385,
      "step": 35500
    },
    {
      "epoch": 3.8328840970350404,
      "grad_norm": 0.5834438800811768,
      "learning_rate": 0.0003705169994603345,
      "loss": 3.4426,
      "step": 35550
    },
    {
      "epoch": 3.838274932614555,
      "grad_norm": 0.6240514516830444,
      "learning_rate": 0.00037019320021586613,
      "loss": 3.4437,
      "step": 35600
    },
    {
      "epoch": 3.8436657681940702,
      "grad_norm": 0.6120893955230713,
      "learning_rate": 0.0003698694009713977,
      "loss": 3.4429,
      "step": 35650
    },
    {
      "epoch": 3.849056603773585,
      "grad_norm": 0.6248950958251953,
      "learning_rate": 0.0003695456017269293,
      "loss": 3.4407,
      "step": 35700
    },
    {
      "epoch": 3.8544474393530996,
      "grad_norm": 0.5935930013656616,
      "learning_rate": 0.00036922180248246083,
      "loss": 3.45,
      "step": 35750
    },
    {
      "epoch": 3.8598382749326143,
      "grad_norm": 0.643417239189148,
      "learning_rate": 0.00036889800323799244,
      "loss": 3.4408,
      "step": 35800
    },
    {
      "epoch": 3.8652291105121295,
      "grad_norm": 0.6277262568473816,
      "learning_rate": 0.000368574203993524,
      "loss": 3.4397,
      "step": 35850
    },
    {
      "epoch": 3.870619946091644,
      "grad_norm": 0.6452029347419739,
      "learning_rate": 0.0003682504047490556,
      "loss": 3.4507,
      "step": 35900
    },
    {
      "epoch": 3.876010781671159,
      "grad_norm": 0.645775556564331,
      "learning_rate": 0.00036792660550458714,
      "loss": 3.4488,
      "step": 35950
    },
    {
      "epoch": 3.881401617250674,
      "grad_norm": 0.6617048382759094,
      "learning_rate": 0.00036760280626011864,
      "loss": 3.4464,
      "step": 36000
    },
    {
      "epoch": 3.881401617250674,
      "eval_accuracy": 0.37615086496934197,
      "eval_loss": 3.445986032485962,
      "eval_runtime": 183.1416,
      "eval_samples_per_second": 98.345,
      "eval_steps_per_second": 6.148,
      "step": 36000
    },
    {
      "epoch": 3.8867924528301887,
      "grad_norm": 0.6022095084190369,
      "learning_rate": 0.0003672790070156503,
      "loss": 3.4428,
      "step": 36050
    },
    {
      "epoch": 3.8921832884097034,
      "grad_norm": 0.5766339302062988,
      "learning_rate": 0.0003669552077711818,
      "loss": 3.4499,
      "step": 36100
    },
    {
      "epoch": 3.8975741239892185,
      "grad_norm": 0.6311817765235901,
      "learning_rate": 0.0003666314085267134,
      "loss": 3.4279,
      "step": 36150
    },
    {
      "epoch": 3.9029649595687332,
      "grad_norm": 0.6049012541770935,
      "learning_rate": 0.00036630760928224495,
      "loss": 3.4384,
      "step": 36200
    },
    {
      "epoch": 3.908355795148248,
      "grad_norm": 0.6929218769073486,
      "learning_rate": 0.00036598381003777656,
      "loss": 3.4577,
      "step": 36250
    },
    {
      "epoch": 3.913746630727763,
      "grad_norm": 0.6345332264900208,
      "learning_rate": 0.0003656600107933081,
      "loss": 3.4422,
      "step": 36300
    },
    {
      "epoch": 3.9191374663072778,
      "grad_norm": 0.6367397308349609,
      "learning_rate": 0.0003653362115488397,
      "loss": 3.4492,
      "step": 36350
    },
    {
      "epoch": 3.9245283018867925,
      "grad_norm": 0.6182951331138611,
      "learning_rate": 0.00036501241230437126,
      "loss": 3.4355,
      "step": 36400
    },
    {
      "epoch": 3.929919137466307,
      "grad_norm": 0.5866326093673706,
      "learning_rate": 0.0003646886130599028,
      "loss": 3.458,
      "step": 36450
    },
    {
      "epoch": 3.935309973045822,
      "grad_norm": 0.6007795333862305,
      "learning_rate": 0.0003643648138154344,
      "loss": 3.4518,
      "step": 36500
    },
    {
      "epoch": 3.940700808625337,
      "grad_norm": 0.575502336025238,
      "learning_rate": 0.00036404101457096597,
      "loss": 3.4513,
      "step": 36550
    },
    {
      "epoch": 3.9460916442048517,
      "grad_norm": 0.5908992290496826,
      "learning_rate": 0.00036371721532649757,
      "loss": 3.452,
      "step": 36600
    },
    {
      "epoch": 3.9514824797843664,
      "grad_norm": 0.6322728395462036,
      "learning_rate": 0.0003633934160820291,
      "loss": 3.4509,
      "step": 36650
    },
    {
      "epoch": 3.9568733153638815,
      "grad_norm": 0.7075260877609253,
      "learning_rate": 0.00036306961683756073,
      "loss": 3.4427,
      "step": 36700
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.6351099610328674,
      "learning_rate": 0.0003627458175930922,
      "loss": 3.474,
      "step": 36750
    },
    {
      "epoch": 3.967654986522911,
      "grad_norm": 0.6212638020515442,
      "learning_rate": 0.0003624220183486238,
      "loss": 3.4598,
      "step": 36800
    },
    {
      "epoch": 3.973045822102426,
      "grad_norm": 0.6004298329353333,
      "learning_rate": 0.0003620982191041554,
      "loss": 3.4321,
      "step": 36850
    },
    {
      "epoch": 3.9784366576819408,
      "grad_norm": 0.6405156254768372,
      "learning_rate": 0.00036177441985968693,
      "loss": 3.4454,
      "step": 36900
    },
    {
      "epoch": 3.9838274932614555,
      "grad_norm": 0.6019781231880188,
      "learning_rate": 0.00036145062061521854,
      "loss": 3.424,
      "step": 36950
    },
    {
      "epoch": 3.9892183288409706,
      "grad_norm": 0.6448103785514832,
      "learning_rate": 0.0003611268213707501,
      "loss": 3.4344,
      "step": 37000
    },
    {
      "epoch": 3.9892183288409706,
      "eval_accuracy": 0.37672487817566586,
      "eval_loss": 3.4396376609802246,
      "eval_runtime": 183.5533,
      "eval_samples_per_second": 98.124,
      "eval_steps_per_second": 6.134,
      "step": 37000
    },
    {
      "epoch": 3.9946091644204853,
      "grad_norm": 0.6120318174362183,
      "learning_rate": 0.0003608030221262817,
      "loss": 3.4419,
      "step": 37050
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1697500944137573,
      "learning_rate": 0.00036047922288181324,
      "loss": 3.4236,
      "step": 37100
    },
    {
      "epoch": 4.005390835579515,
      "grad_norm": 0.6020103693008423,
      "learning_rate": 0.00036015542363734485,
      "loss": 3.3492,
      "step": 37150
    },
    {
      "epoch": 4.010781671159029,
      "grad_norm": 0.5951403975486755,
      "learning_rate": 0.0003598316243928764,
      "loss": 3.3509,
      "step": 37200
    },
    {
      "epoch": 4.0161725067385445,
      "grad_norm": 0.57025545835495,
      "learning_rate": 0.00035950782514840795,
      "loss": 3.3444,
      "step": 37250
    },
    {
      "epoch": 4.02156334231806,
      "grad_norm": 0.6555331349372864,
      "learning_rate": 0.00035918402590393955,
      "loss": 3.3651,
      "step": 37300
    },
    {
      "epoch": 4.026954177897574,
      "grad_norm": 0.5830883979797363,
      "learning_rate": 0.00035886022665947105,
      "loss": 3.3525,
      "step": 37350
    },
    {
      "epoch": 4.032345013477089,
      "grad_norm": 0.6118924021720886,
      "learning_rate": 0.0003585364274150027,
      "loss": 3.3481,
      "step": 37400
    },
    {
      "epoch": 4.037735849056604,
      "grad_norm": 0.6575861573219299,
      "learning_rate": 0.0003582126281705342,
      "loss": 3.3609,
      "step": 37450
    },
    {
      "epoch": 4.0431266846361185,
      "grad_norm": 0.664242684841156,
      "learning_rate": 0.0003578888289260658,
      "loss": 3.3625,
      "step": 37500
    },
    {
      "epoch": 4.048517520215634,
      "grad_norm": 0.5893779993057251,
      "learning_rate": 0.00035756502968159736,
      "loss": 3.3602,
      "step": 37550
    },
    {
      "epoch": 4.053908355795148,
      "grad_norm": 0.662519633769989,
      "learning_rate": 0.00035724123043712896,
      "loss": 3.3475,
      "step": 37600
    },
    {
      "epoch": 4.059299191374663,
      "grad_norm": 0.609507143497467,
      "learning_rate": 0.0003569174311926605,
      "loss": 3.3682,
      "step": 37650
    },
    {
      "epoch": 4.064690026954178,
      "grad_norm": 0.6340688467025757,
      "learning_rate": 0.00035659363194819206,
      "loss": 3.3747,
      "step": 37700
    },
    {
      "epoch": 4.070080862533692,
      "grad_norm": 0.6548017263412476,
      "learning_rate": 0.00035626983270372367,
      "loss": 3.3722,
      "step": 37750
    },
    {
      "epoch": 4.0754716981132075,
      "grad_norm": 0.6166157126426697,
      "learning_rate": 0.0003559460334592552,
      "loss": 3.3431,
      "step": 37800
    },
    {
      "epoch": 4.080862533692723,
      "grad_norm": 0.6283932328224182,
      "learning_rate": 0.0003556222342147868,
      "loss": 3.3594,
      "step": 37850
    },
    {
      "epoch": 4.086253369272237,
      "grad_norm": 0.6251431703567505,
      "learning_rate": 0.0003552984349703184,
      "loss": 3.3586,
      "step": 37900
    },
    {
      "epoch": 4.091644204851752,
      "grad_norm": 0.6172202825546265,
      "learning_rate": 0.00035497463572585,
      "loss": 3.3612,
      "step": 37950
    },
    {
      "epoch": 4.097035040431267,
      "grad_norm": 0.655680775642395,
      "learning_rate": 0.00035465083648138153,
      "loss": 3.3725,
      "step": 38000
    },
    {
      "epoch": 4.097035040431267,
      "eval_accuracy": 0.37726716473874794,
      "eval_loss": 3.4422571659088135,
      "eval_runtime": 183.6362,
      "eval_samples_per_second": 98.08,
      "eval_steps_per_second": 6.132,
      "step": 38000
    },
    {
      "epoch": 4.1024258760107815,
      "grad_norm": 0.6230634450912476,
      "learning_rate": 0.00035432703723691314,
      "loss": 3.3661,
      "step": 38050
    },
    {
      "epoch": 4.107816711590297,
      "grad_norm": 0.6373627781867981,
      "learning_rate": 0.00035400323799244463,
      "loss": 3.3758,
      "step": 38100
    },
    {
      "epoch": 4.113207547169812,
      "grad_norm": 0.6033377051353455,
      "learning_rate": 0.0003536794387479762,
      "loss": 3.3743,
      "step": 38150
    },
    {
      "epoch": 4.118598382749326,
      "grad_norm": 0.6158263683319092,
      "learning_rate": 0.0003533556395035078,
      "loss": 3.3557,
      "step": 38200
    },
    {
      "epoch": 4.123989218328841,
      "grad_norm": 0.688728928565979,
      "learning_rate": 0.00035303184025903934,
      "loss": 3.3649,
      "step": 38250
    },
    {
      "epoch": 4.129380053908355,
      "grad_norm": 0.6338489055633545,
      "learning_rate": 0.00035270804101457094,
      "loss": 3.3688,
      "step": 38300
    },
    {
      "epoch": 4.1347708894878705,
      "grad_norm": 0.6235848069190979,
      "learning_rate": 0.0003523842417701025,
      "loss": 3.3523,
      "step": 38350
    },
    {
      "epoch": 4.140161725067386,
      "grad_norm": 0.6039944887161255,
      "learning_rate": 0.0003520604425256341,
      "loss": 3.3666,
      "step": 38400
    },
    {
      "epoch": 4.1455525606469,
      "grad_norm": 0.6759158372879028,
      "learning_rate": 0.00035173664328116565,
      "loss": 3.3516,
      "step": 38450
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.6314166188240051,
      "learning_rate": 0.00035141284403669725,
      "loss": 3.3835,
      "step": 38500
    },
    {
      "epoch": 4.15633423180593,
      "grad_norm": 0.6417436003684998,
      "learning_rate": 0.0003510890447922288,
      "loss": 3.3594,
      "step": 38550
    },
    {
      "epoch": 4.1617250673854445,
      "grad_norm": 0.6634682416915894,
      "learning_rate": 0.00035076524554776035,
      "loss": 3.368,
      "step": 38600
    },
    {
      "epoch": 4.16711590296496,
      "grad_norm": 0.6313776969909668,
      "learning_rate": 0.00035044144630329196,
      "loss": 3.3669,
      "step": 38650
    },
    {
      "epoch": 4.172506738544475,
      "grad_norm": 0.616958737373352,
      "learning_rate": 0.00035011764705882346,
      "loss": 3.3885,
      "step": 38700
    },
    {
      "epoch": 4.177897574123989,
      "grad_norm": 0.7168732285499573,
      "learning_rate": 0.0003497938478143551,
      "loss": 3.3589,
      "step": 38750
    },
    {
      "epoch": 4.183288409703504,
      "grad_norm": 0.7030311226844788,
      "learning_rate": 0.0003494700485698866,
      "loss": 3.3839,
      "step": 38800
    },
    {
      "epoch": 4.188679245283019,
      "grad_norm": 0.6451276540756226,
      "learning_rate": 0.0003491462493254182,
      "loss": 3.374,
      "step": 38850
    },
    {
      "epoch": 4.1940700808625335,
      "grad_norm": 0.6581581234931946,
      "learning_rate": 0.00034882245008094977,
      "loss": 3.3815,
      "step": 38900
    },
    {
      "epoch": 4.199460916442049,
      "grad_norm": 0.6221701502799988,
      "learning_rate": 0.0003484986508364813,
      "loss": 3.3756,
      "step": 38950
    },
    {
      "epoch": 4.204851752021563,
      "grad_norm": 0.7239381074905396,
      "learning_rate": 0.0003481748515920129,
      "loss": 3.371,
      "step": 39000
    },
    {
      "epoch": 4.204851752021563,
      "eval_accuracy": 0.37740939136889706,
      "eval_loss": 3.438514232635498,
      "eval_runtime": 183.1338,
      "eval_samples_per_second": 98.349,
      "eval_steps_per_second": 6.149,
      "step": 39000
    },
    {
      "epoch": 4.210242587601078,
      "grad_norm": 0.6811919212341309,
      "learning_rate": 0.00034785105234754447,
      "loss": 3.3755,
      "step": 39050
    },
    {
      "epoch": 4.215633423180593,
      "grad_norm": 0.6211300492286682,
      "learning_rate": 0.0003475272531030761,
      "loss": 3.3774,
      "step": 39100
    },
    {
      "epoch": 4.2210242587601075,
      "grad_norm": 0.6302964091300964,
      "learning_rate": 0.00034720345385860763,
      "loss": 3.3873,
      "step": 39150
    },
    {
      "epoch": 4.226415094339623,
      "grad_norm": 0.6861122250556946,
      "learning_rate": 0.00034687965461413923,
      "loss": 3.3823,
      "step": 39200
    },
    {
      "epoch": 4.231805929919138,
      "grad_norm": 0.6081944108009338,
      "learning_rate": 0.0003465558553696708,
      "loss": 3.38,
      "step": 39250
    },
    {
      "epoch": 4.237196765498652,
      "grad_norm": 0.6504591703414917,
      "learning_rate": 0.0003462320561252024,
      "loss": 3.375,
      "step": 39300
    },
    {
      "epoch": 4.242587601078167,
      "grad_norm": 0.6743220090866089,
      "learning_rate": 0.00034590825688073394,
      "loss": 3.3905,
      "step": 39350
    },
    {
      "epoch": 4.247978436657682,
      "grad_norm": 0.6235594749450684,
      "learning_rate": 0.0003455909336211549,
      "loss": 3.3754,
      "step": 39400
    },
    {
      "epoch": 4.2533692722371965,
      "grad_norm": 0.6357383728027344,
      "learning_rate": 0.0003452671343766864,
      "loss": 3.3785,
      "step": 39450
    },
    {
      "epoch": 4.258760107816712,
      "grad_norm": 0.6747174859046936,
      "learning_rate": 0.000344943335132218,
      "loss": 3.377,
      "step": 39500
    },
    {
      "epoch": 4.264150943396227,
      "grad_norm": 0.6351191997528076,
      "learning_rate": 0.00034461953588774954,
      "loss": 3.3805,
      "step": 39550
    },
    {
      "epoch": 4.269541778975741,
      "grad_norm": 0.5940690636634827,
      "learning_rate": 0.00034429573664328115,
      "loss": 3.3745,
      "step": 39600
    },
    {
      "epoch": 4.274932614555256,
      "grad_norm": 0.5747135281562805,
      "learning_rate": 0.0003439719373988127,
      "loss": 3.3735,
      "step": 39650
    },
    {
      "epoch": 4.280323450134771,
      "grad_norm": 0.6357155442237854,
      "learning_rate": 0.00034364813815434425,
      "loss": 3.3891,
      "step": 39700
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.6541160345077515,
      "learning_rate": 0.00034332433890987585,
      "loss": 3.3779,
      "step": 39750
    },
    {
      "epoch": 4.291105121293801,
      "grad_norm": 0.6745419502258301,
      "learning_rate": 0.0003430005396654074,
      "loss": 3.384,
      "step": 39800
    },
    {
      "epoch": 4.296495956873315,
      "grad_norm": 0.641961395740509,
      "learning_rate": 0.000342676740420939,
      "loss": 3.3535,
      "step": 39850
    },
    {
      "epoch": 4.30188679245283,
      "grad_norm": 0.6052359342575073,
      "learning_rate": 0.00034235294117647056,
      "loss": 3.3919,
      "step": 39900
    },
    {
      "epoch": 4.307277628032345,
      "grad_norm": 0.6415327787399292,
      "learning_rate": 0.00034202914193200216,
      "loss": 3.3692,
      "step": 39950
    },
    {
      "epoch": 4.3126684636118595,
      "grad_norm": 0.6692045331001282,
      "learning_rate": 0.0003417053426875337,
      "loss": 3.3901,
      "step": 40000
    },
    {
      "epoch": 4.3126684636118595,
      "eval_accuracy": 0.3779461366347006,
      "eval_loss": 3.4338619709014893,
      "eval_runtime": 183.59,
      "eval_samples_per_second": 98.104,
      "eval_steps_per_second": 6.133,
      "step": 40000
    },
    {
      "epoch": 4.318059299191375,
      "grad_norm": 0.659393846988678,
      "learning_rate": 0.0003413815434430653,
      "loss": 3.4026,
      "step": 40050
    },
    {
      "epoch": 4.32345013477089,
      "grad_norm": 0.6064386963844299,
      "learning_rate": 0.0003410577441985968,
      "loss": 3.3698,
      "step": 40100
    },
    {
      "epoch": 4.328840970350404,
      "grad_norm": 0.6234272718429565,
      "learning_rate": 0.00034073394495412837,
      "loss": 3.3865,
      "step": 40150
    },
    {
      "epoch": 4.334231805929919,
      "grad_norm": 0.6467458009719849,
      "learning_rate": 0.00034041014570965997,
      "loss": 3.4039,
      "step": 40200
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.5988481640815735,
      "learning_rate": 0.0003400863464651915,
      "loss": 3.3805,
      "step": 40250
    },
    {
      "epoch": 4.345013477088949,
      "grad_norm": 0.660705029964447,
      "learning_rate": 0.0003397625472207231,
      "loss": 3.3948,
      "step": 40300
    },
    {
      "epoch": 4.350404312668464,
      "grad_norm": 0.6406025886535645,
      "learning_rate": 0.0003394387479762547,
      "loss": 3.401,
      "step": 40350
    },
    {
      "epoch": 4.355795148247978,
      "grad_norm": 0.6672471761703491,
      "learning_rate": 0.0003391149487317863,
      "loss": 3.3882,
      "step": 40400
    },
    {
      "epoch": 4.361185983827493,
      "grad_norm": 0.6221401691436768,
      "learning_rate": 0.00033879114948731783,
      "loss": 3.3945,
      "step": 40450
    },
    {
      "epoch": 4.366576819407008,
      "grad_norm": 0.6297544240951538,
      "learning_rate": 0.00033846735024284944,
      "loss": 3.3631,
      "step": 40500
    },
    {
      "epoch": 4.3719676549865225,
      "grad_norm": 0.6644108295440674,
      "learning_rate": 0.000338143550998381,
      "loss": 3.3732,
      "step": 40550
    },
    {
      "epoch": 4.377358490566038,
      "grad_norm": 0.633611798286438,
      "learning_rate": 0.00033781975175391254,
      "loss": 3.3804,
      "step": 40600
    },
    {
      "epoch": 4.382749326145553,
      "grad_norm": 0.641505241394043,
      "learning_rate": 0.00033749595250944414,
      "loss": 3.3789,
      "step": 40650
    },
    {
      "epoch": 4.388140161725067,
      "grad_norm": 0.6398807764053345,
      "learning_rate": 0.00033717215326497564,
      "loss": 3.3847,
      "step": 40700
    },
    {
      "epoch": 4.393530997304582,
      "grad_norm": 0.6694504022598267,
      "learning_rate": 0.0003368483540205073,
      "loss": 3.3853,
      "step": 40750
    },
    {
      "epoch": 4.398921832884097,
      "grad_norm": 0.6023000478744507,
      "learning_rate": 0.0003365245547760388,
      "loss": 3.3843,
      "step": 40800
    },
    {
      "epoch": 4.404312668463612,
      "grad_norm": 0.6796262860298157,
      "learning_rate": 0.0003362007555315704,
      "loss": 3.3754,
      "step": 40850
    },
    {
      "epoch": 4.409703504043127,
      "grad_norm": 0.6454067230224609,
      "learning_rate": 0.00033587695628710195,
      "loss": 3.3825,
      "step": 40900
    },
    {
      "epoch": 4.415094339622642,
      "grad_norm": 0.6159910559654236,
      "learning_rate": 0.00033555315704263355,
      "loss": 3.3901,
      "step": 40950
    },
    {
      "epoch": 4.420485175202156,
      "grad_norm": 0.6237287521362305,
      "learning_rate": 0.0003352293577981651,
      "loss": 3.3831,
      "step": 41000
    },
    {
      "epoch": 4.420485175202156,
      "eval_accuracy": 0.378417364209152,
      "eval_loss": 3.4264814853668213,
      "eval_runtime": 183.3491,
      "eval_samples_per_second": 98.233,
      "eval_steps_per_second": 6.141,
      "step": 41000
    },
    {
      "epoch": 4.425876010781671,
      "grad_norm": 0.6706305742263794,
      "learning_rate": 0.00033490555855369665,
      "loss": 3.3701,
      "step": 41050
    },
    {
      "epoch": 4.431266846361186,
      "grad_norm": 0.6600586175918579,
      "learning_rate": 0.00033458175930922826,
      "loss": 3.3875,
      "step": 41100
    },
    {
      "epoch": 4.436657681940701,
      "grad_norm": 0.5768634080886841,
      "learning_rate": 0.0003342579600647598,
      "loss": 3.383,
      "step": 41150
    },
    {
      "epoch": 4.442048517520216,
      "grad_norm": 0.6009304523468018,
      "learning_rate": 0.0003339341608202914,
      "loss": 3.3686,
      "step": 41200
    },
    {
      "epoch": 4.44743935309973,
      "grad_norm": 0.6819839477539062,
      "learning_rate": 0.00033361036157582297,
      "loss": 3.3808,
      "step": 41250
    },
    {
      "epoch": 4.452830188679245,
      "grad_norm": 0.63370281457901,
      "learning_rate": 0.00033328656233135457,
      "loss": 3.3742,
      "step": 41300
    },
    {
      "epoch": 4.45822102425876,
      "grad_norm": 0.6805122494697571,
      "learning_rate": 0.0003329627630868861,
      "loss": 3.3999,
      "step": 41350
    },
    {
      "epoch": 4.463611859838275,
      "grad_norm": 0.6374489068984985,
      "learning_rate": 0.0003326454398273071,
      "loss": 3.3923,
      "step": 41400
    },
    {
      "epoch": 4.46900269541779,
      "grad_norm": 0.632504403591156,
      "learning_rate": 0.00033232164058283857,
      "loss": 3.3802,
      "step": 41450
    },
    {
      "epoch": 4.474393530997305,
      "grad_norm": 0.6276558637619019,
      "learning_rate": 0.0003319978413383702,
      "loss": 3.371,
      "step": 41500
    },
    {
      "epoch": 4.479784366576819,
      "grad_norm": 0.6790854334831238,
      "learning_rate": 0.0003316740420939017,
      "loss": 3.3857,
      "step": 41550
    },
    {
      "epoch": 4.485175202156334,
      "grad_norm": 0.6570220589637756,
      "learning_rate": 0.00033135024284943333,
      "loss": 3.3908,
      "step": 41600
    },
    {
      "epoch": 4.490566037735849,
      "grad_norm": 0.6306306719779968,
      "learning_rate": 0.0003310264436049649,
      "loss": 3.3758,
      "step": 41650
    },
    {
      "epoch": 4.495956873315364,
      "grad_norm": 0.6574743986129761,
      "learning_rate": 0.0003307026443604965,
      "loss": 3.3988,
      "step": 41700
    },
    {
      "epoch": 4.501347708894879,
      "grad_norm": 0.6541933417320251,
      "learning_rate": 0.00033037884511602804,
      "loss": 3.379,
      "step": 41750
    },
    {
      "epoch": 4.506738544474393,
      "grad_norm": 0.6046768426895142,
      "learning_rate": 0.0003300550458715596,
      "loss": 3.3775,
      "step": 41800
    },
    {
      "epoch": 4.512129380053908,
      "grad_norm": 0.6767274737358093,
      "learning_rate": 0.0003297312466270912,
      "loss": 3.3878,
      "step": 41850
    },
    {
      "epoch": 4.517520215633423,
      "grad_norm": 0.6689441204071045,
      "learning_rate": 0.00032940744738262274,
      "loss": 3.3872,
      "step": 41900
    },
    {
      "epoch": 4.5229110512129385,
      "grad_norm": 0.6398838758468628,
      "learning_rate": 0.00032908364813815435,
      "loss": 3.3898,
      "step": 41950
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.6058347225189209,
      "learning_rate": 0.0003287598488936859,
      "loss": 3.3797,
      "step": 42000
    },
    {
      "epoch": 4.528301886792453,
      "eval_accuracy": 0.37927583067577425,
      "eval_loss": 3.4238638877868652,
      "eval_runtime": 183.4895,
      "eval_samples_per_second": 98.158,
      "eval_steps_per_second": 6.137,
      "step": 42000
    },
    {
      "epoch": 4.533692722371968,
      "grad_norm": 0.5938351154327393,
      "learning_rate": 0.0003284360496492175,
      "loss": 3.3765,
      "step": 42050
    },
    {
      "epoch": 4.539083557951482,
      "grad_norm": 0.6558327674865723,
      "learning_rate": 0.000328112250404749,
      "loss": 3.4059,
      "step": 42100
    },
    {
      "epoch": 4.544474393530997,
      "grad_norm": 0.64605313539505,
      "learning_rate": 0.00032778845116028066,
      "loss": 3.3917,
      "step": 42150
    },
    {
      "epoch": 4.549865229110512,
      "grad_norm": 0.5833234190940857,
      "learning_rate": 0.00032746465191581215,
      "loss": 3.3712,
      "step": 42200
    },
    {
      "epoch": 4.555256064690027,
      "grad_norm": 0.822323203086853,
      "learning_rate": 0.0003271408526713437,
      "loss": 3.3953,
      "step": 42250
    },
    {
      "epoch": 4.560646900269542,
      "grad_norm": 0.6433358192443848,
      "learning_rate": 0.0003268170534268753,
      "loss": 3.3939,
      "step": 42300
    },
    {
      "epoch": 4.566037735849057,
      "grad_norm": 0.7177082300186157,
      "learning_rate": 0.00032649325418240686,
      "loss": 3.3757,
      "step": 42350
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.6373773217201233,
      "learning_rate": 0.00032616945493793846,
      "loss": 3.4017,
      "step": 42400
    },
    {
      "epoch": 4.576819407008086,
      "grad_norm": 0.6122283935546875,
      "learning_rate": 0.00032584565569347,
      "loss": 3.3703,
      "step": 42450
    },
    {
      "epoch": 4.5822102425876015,
      "grad_norm": 0.6813517212867737,
      "learning_rate": 0.0003255218564490016,
      "loss": 3.3709,
      "step": 42500
    },
    {
      "epoch": 4.587601078167116,
      "grad_norm": 0.6224530935287476,
      "learning_rate": 0.00032519805720453317,
      "loss": 3.3845,
      "step": 42550
    },
    {
      "epoch": 4.592991913746631,
      "grad_norm": 0.6742523312568665,
      "learning_rate": 0.0003248742579600647,
      "loss": 3.3823,
      "step": 42600
    },
    {
      "epoch": 4.598382749326145,
      "grad_norm": 0.7158631682395935,
      "learning_rate": 0.0003245504587155963,
      "loss": 3.3956,
      "step": 42650
    },
    {
      "epoch": 4.60377358490566,
      "grad_norm": 0.6356601715087891,
      "learning_rate": 0.0003242266594711278,
      "loss": 3.3887,
      "step": 42700
    },
    {
      "epoch": 4.609164420485175,
      "grad_norm": 0.6447049975395203,
      "learning_rate": 0.0003239028602266595,
      "loss": 3.3758,
      "step": 42750
    },
    {
      "epoch": 4.6145552560646905,
      "grad_norm": 0.6369867324829102,
      "learning_rate": 0.000323579060982191,
      "loss": 3.3985,
      "step": 42800
    },
    {
      "epoch": 4.619946091644205,
      "grad_norm": 0.6150829195976257,
      "learning_rate": 0.0003232552617377226,
      "loss": 3.3909,
      "step": 42850
    },
    {
      "epoch": 4.62533692722372,
      "grad_norm": 0.5987628102302551,
      "learning_rate": 0.00032293146249325413,
      "loss": 3.378,
      "step": 42900
    },
    {
      "epoch": 4.630727762803234,
      "grad_norm": 0.6027454733848572,
      "learning_rate": 0.00032260766324878574,
      "loss": 3.382,
      "step": 42950
    },
    {
      "epoch": 4.636118598382749,
      "grad_norm": 0.6479946374893188,
      "learning_rate": 0.0003222838640043173,
      "loss": 3.3968,
      "step": 43000
    },
    {
      "epoch": 4.636118598382749,
      "eval_accuracy": 0.3795472455895348,
      "eval_loss": 3.419617176055908,
      "eval_runtime": 183.1797,
      "eval_samples_per_second": 98.324,
      "eval_steps_per_second": 6.147,
      "step": 43000
    },
    {
      "epoch": 4.6415094339622645,
      "grad_norm": 0.6356445550918579,
      "learning_rate": 0.00032196006475984884,
      "loss": 3.3766,
      "step": 43050
    },
    {
      "epoch": 4.646900269541779,
      "grad_norm": 0.6246060132980347,
      "learning_rate": 0.00032163626551538044,
      "loss": 3.3818,
      "step": 43100
    },
    {
      "epoch": 4.652291105121294,
      "grad_norm": 0.6289001107215881,
      "learning_rate": 0.000321312466270912,
      "loss": 3.3773,
      "step": 43150
    },
    {
      "epoch": 4.657681940700809,
      "grad_norm": 0.6760928630828857,
      "learning_rate": 0.0003209886670264436,
      "loss": 3.3847,
      "step": 43200
    },
    {
      "epoch": 4.663072776280323,
      "grad_norm": 0.6528955698013306,
      "learning_rate": 0.00032066486778197515,
      "loss": 3.3837,
      "step": 43250
    },
    {
      "epoch": 4.668463611859838,
      "grad_norm": 0.6557238101959229,
      "learning_rate": 0.00032034106853750675,
      "loss": 3.3919,
      "step": 43300
    },
    {
      "epoch": 4.6738544474393535,
      "grad_norm": 0.6291058659553528,
      "learning_rate": 0.0003200172692930383,
      "loss": 3.3674,
      "step": 43350
    },
    {
      "epoch": 4.679245283018868,
      "grad_norm": 0.7018590569496155,
      "learning_rate": 0.0003196934700485699,
      "loss": 3.377,
      "step": 43400
    },
    {
      "epoch": 4.684636118598383,
      "grad_norm": 0.6758108139038086,
      "learning_rate": 0.00031937614678899075,
      "loss": 3.3833,
      "step": 43450
    },
    {
      "epoch": 4.690026954177897,
      "grad_norm": 0.6351593136787415,
      "learning_rate": 0.00031905234754452236,
      "loss": 3.383,
      "step": 43500
    },
    {
      "epoch": 4.695417789757412,
      "grad_norm": 0.639167308807373,
      "learning_rate": 0.0003187285483000539,
      "loss": 3.3697,
      "step": 43550
    },
    {
      "epoch": 4.7008086253369274,
      "grad_norm": 0.6248657703399658,
      "learning_rate": 0.0003184047490555855,
      "loss": 3.3745,
      "step": 43600
    },
    {
      "epoch": 4.706199460916442,
      "grad_norm": 0.633437991142273,
      "learning_rate": 0.00031808094981111706,
      "loss": 3.3931,
      "step": 43650
    },
    {
      "epoch": 4.711590296495957,
      "grad_norm": 0.627947211265564,
      "learning_rate": 0.00031775715056664867,
      "loss": 3.384,
      "step": 43700
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.6561819911003113,
      "learning_rate": 0.0003174333513221802,
      "loss": 3.3817,
      "step": 43750
    },
    {
      "epoch": 4.722371967654986,
      "grad_norm": 0.6273093819618225,
      "learning_rate": 0.00031710955207771177,
      "loss": 3.3727,
      "step": 43800
    },
    {
      "epoch": 4.727762803234501,
      "grad_norm": 0.6125487089157104,
      "learning_rate": 0.0003167857528332434,
      "loss": 3.3952,
      "step": 43850
    },
    {
      "epoch": 4.7331536388140165,
      "grad_norm": 0.6168906092643738,
      "learning_rate": 0.0003164619535887749,
      "loss": 3.4017,
      "step": 43900
    },
    {
      "epoch": 4.738544474393531,
      "grad_norm": 0.631703794002533,
      "learning_rate": 0.00031613815434430653,
      "loss": 3.4001,
      "step": 43950
    },
    {
      "epoch": 4.743935309973046,
      "grad_norm": 0.6847707629203796,
      "learning_rate": 0.0003158143550998381,
      "loss": 3.3652,
      "step": 44000
    },
    {
      "epoch": 4.743935309973046,
      "eval_accuracy": 0.38033150213377975,
      "eval_loss": 3.4119884967803955,
      "eval_runtime": 183.7579,
      "eval_samples_per_second": 98.015,
      "eval_steps_per_second": 6.128,
      "step": 44000
    },
    {
      "epoch": 4.74932614555256,
      "grad_norm": 0.6523143649101257,
      "learning_rate": 0.0003154905558553697,
      "loss": 3.3999,
      "step": 44050
    },
    {
      "epoch": 4.754716981132075,
      "grad_norm": 0.6526023149490356,
      "learning_rate": 0.0003151667566109012,
      "loss": 3.3786,
      "step": 44100
    },
    {
      "epoch": 4.7601078167115904,
      "grad_norm": 0.6396304368972778,
      "learning_rate": 0.00031484295736643284,
      "loss": 3.3723,
      "step": 44150
    },
    {
      "epoch": 4.765498652291106,
      "grad_norm": 0.6429187059402466,
      "learning_rate": 0.00031451915812196434,
      "loss": 3.3932,
      "step": 44200
    },
    {
      "epoch": 4.77088948787062,
      "grad_norm": 0.6210240721702576,
      "learning_rate": 0.0003141953588774959,
      "loss": 3.3862,
      "step": 44250
    },
    {
      "epoch": 4.776280323450135,
      "grad_norm": 0.6462406516075134,
      "learning_rate": 0.0003138715596330275,
      "loss": 3.3807,
      "step": 44300
    },
    {
      "epoch": 4.781671159029649,
      "grad_norm": 0.6584225296974182,
      "learning_rate": 0.00031354776038855904,
      "loss": 3.3916,
      "step": 44350
    },
    {
      "epoch": 4.787061994609164,
      "grad_norm": 0.6552765369415283,
      "learning_rate": 0.00031322396114409065,
      "loss": 3.3723,
      "step": 44400
    },
    {
      "epoch": 4.7924528301886795,
      "grad_norm": 0.6333467364311218,
      "learning_rate": 0.0003129001618996222,
      "loss": 3.39,
      "step": 44450
    },
    {
      "epoch": 4.797843665768194,
      "grad_norm": 0.6195804476737976,
      "learning_rate": 0.0003125763626551538,
      "loss": 3.3721,
      "step": 44500
    },
    {
      "epoch": 4.803234501347709,
      "grad_norm": 0.6451464891433716,
      "learning_rate": 0.00031225256341068535,
      "loss": 3.3719,
      "step": 44550
    },
    {
      "epoch": 4.808625336927224,
      "grad_norm": 0.6498815417289734,
      "learning_rate": 0.00031192876416621696,
      "loss": 3.3853,
      "step": 44600
    },
    {
      "epoch": 4.814016172506738,
      "grad_norm": 0.6606132984161377,
      "learning_rate": 0.0003116049649217485,
      "loss": 3.375,
      "step": 44650
    },
    {
      "epoch": 4.819407008086253,
      "grad_norm": 0.6065442562103271,
      "learning_rate": 0.00031128116567728,
      "loss": 3.3695,
      "step": 44700
    },
    {
      "epoch": 4.824797843665769,
      "grad_norm": 0.6994802355766296,
      "learning_rate": 0.00031095736643281166,
      "loss": 3.3718,
      "step": 44750
    },
    {
      "epoch": 4.830188679245283,
      "grad_norm": 0.6376444697380066,
      "learning_rate": 0.00031063356718834316,
      "loss": 3.3855,
      "step": 44800
    },
    {
      "epoch": 4.835579514824798,
      "grad_norm": 0.6514849662780762,
      "learning_rate": 0.00031030976794387476,
      "loss": 3.3803,
      "step": 44850
    },
    {
      "epoch": 4.840970350404312,
      "grad_norm": 0.6474654078483582,
      "learning_rate": 0.0003099859686994063,
      "loss": 3.3833,
      "step": 44900
    },
    {
      "epoch": 4.846361185983827,
      "grad_norm": 0.6419128179550171,
      "learning_rate": 0.0003096621694549379,
      "loss": 3.3786,
      "step": 44950
    },
    {
      "epoch": 4.8517520215633425,
      "grad_norm": 0.6200140714645386,
      "learning_rate": 0.00030933837021046947,
      "loss": 3.3886,
      "step": 45000
    },
    {
      "epoch": 4.8517520215633425,
      "eval_accuracy": 0.38061834575760994,
      "eval_loss": 3.4101431369781494,
      "eval_runtime": 183.1471,
      "eval_samples_per_second": 98.342,
      "eval_steps_per_second": 6.148,
      "step": 45000
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.6373332738876343,
      "learning_rate": 0.0003090145709660011,
      "loss": 3.3643,
      "step": 45050
    },
    {
      "epoch": 4.862533692722372,
      "grad_norm": 0.6662492752075195,
      "learning_rate": 0.0003086907717215326,
      "loss": 3.3895,
      "step": 45100
    },
    {
      "epoch": 4.867924528301887,
      "grad_norm": 0.7503253817558289,
      "learning_rate": 0.0003083669724770642,
      "loss": 3.3651,
      "step": 45150
    },
    {
      "epoch": 4.873315363881401,
      "grad_norm": 0.6341877579689026,
      "learning_rate": 0.0003080431732325958,
      "loss": 3.3864,
      "step": 45200
    },
    {
      "epoch": 4.878706199460916,
      "grad_norm": 0.6354518532752991,
      "learning_rate": 0.00030771937398812733,
      "loss": 3.3839,
      "step": 45250
    },
    {
      "epoch": 4.884097035040432,
      "grad_norm": 0.6411907076835632,
      "learning_rate": 0.00030739557474365894,
      "loss": 3.389,
      "step": 45300
    },
    {
      "epoch": 4.889487870619946,
      "grad_norm": 0.6452317833900452,
      "learning_rate": 0.0003070717754991905,
      "loss": 3.3803,
      "step": 45350
    },
    {
      "epoch": 4.894878706199461,
      "grad_norm": 0.61771160364151,
      "learning_rate": 0.0003067479762547221,
      "loss": 3.3768,
      "step": 45400
    },
    {
      "epoch": 4.900269541778976,
      "grad_norm": 0.6439537405967712,
      "learning_rate": 0.00030643065299514294,
      "loss": 3.3857,
      "step": 45450
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.6568528413772583,
      "learning_rate": 0.00030610685375067454,
      "loss": 3.3843,
      "step": 45500
    },
    {
      "epoch": 4.9110512129380055,
      "grad_norm": 0.6745019555091858,
      "learning_rate": 0.0003057830545062061,
      "loss": 3.3737,
      "step": 45550
    },
    {
      "epoch": 4.916442048517521,
      "grad_norm": 0.6615657806396484,
      "learning_rate": 0.0003054592552617377,
      "loss": 3.3996,
      "step": 45600
    },
    {
      "epoch": 4.921832884097035,
      "grad_norm": 0.6774879097938538,
      "learning_rate": 0.00030513545601726925,
      "loss": 3.3726,
      "step": 45650
    },
    {
      "epoch": 4.92722371967655,
      "grad_norm": 0.6349773406982422,
      "learning_rate": 0.00030481165677280085,
      "loss": 3.3848,
      "step": 45700
    },
    {
      "epoch": 4.932614555256064,
      "grad_norm": 0.6601877808570862,
      "learning_rate": 0.0003044878575283324,
      "loss": 3.3864,
      "step": 45750
    },
    {
      "epoch": 4.938005390835579,
      "grad_norm": 0.7116173505783081,
      "learning_rate": 0.000304164058283864,
      "loss": 3.3814,
      "step": 45800
    },
    {
      "epoch": 4.943396226415095,
      "grad_norm": 0.7644145488739014,
      "learning_rate": 0.00030384025903939556,
      "loss": 3.382,
      "step": 45850
    },
    {
      "epoch": 4.948787061994609,
      "grad_norm": 0.6855260729789734,
      "learning_rate": 0.0003035164597949271,
      "loss": 3.3844,
      "step": 45900
    },
    {
      "epoch": 4.954177897574124,
      "grad_norm": 0.6151649951934814,
      "learning_rate": 0.0003031926605504587,
      "loss": 3.3674,
      "step": 45950
    },
    {
      "epoch": 4.959568733153639,
      "grad_norm": 0.6668580174446106,
      "learning_rate": 0.00030286886130599026,
      "loss": 3.3804,
      "step": 46000
    },
    {
      "epoch": 4.959568733153639,
      "eval_accuracy": 0.3810672994900594,
      "eval_loss": 3.4056036472320557,
      "eval_runtime": 183.5372,
      "eval_samples_per_second": 98.133,
      "eval_steps_per_second": 6.135,
      "step": 46000
    },
    {
      "epoch": 4.964959568733153,
      "grad_norm": 0.6548580527305603,
      "learning_rate": 0.00030254506206152187,
      "loss": 3.387,
      "step": 46050
    },
    {
      "epoch": 4.9703504043126685,
      "grad_norm": 0.6375557780265808,
      "learning_rate": 0.00030222126281705336,
      "loss": 3.3795,
      "step": 46100
    },
    {
      "epoch": 4.975741239892184,
      "grad_norm": 0.6740067601203918,
      "learning_rate": 0.000301897463572585,
      "loss": 3.3812,
      "step": 46150
    },
    {
      "epoch": 4.981132075471698,
      "grad_norm": 0.6744655966758728,
      "learning_rate": 0.0003015736643281165,
      "loss": 3.3782,
      "step": 46200
    },
    {
      "epoch": 4.986522911051213,
      "grad_norm": 0.6689214110374451,
      "learning_rate": 0.00030124986508364807,
      "loss": 3.4003,
      "step": 46250
    },
    {
      "epoch": 4.991913746630727,
      "grad_norm": 0.6952337026596069,
      "learning_rate": 0.0003009260658391797,
      "loss": 3.3866,
      "step": 46300
    },
    {
      "epoch": 4.997304582210242,
      "grad_norm": 0.6591714024543762,
      "learning_rate": 0.0003006022665947112,
      "loss": 3.3759,
      "step": 46350
    },
    {
      "epoch": 5.002695417789758,
      "grad_norm": 0.6524887084960938,
      "learning_rate": 0.00030027846735024283,
      "loss": 3.3332,
      "step": 46400
    },
    {
      "epoch": 5.008086253369272,
      "grad_norm": 0.6842234134674072,
      "learning_rate": 0.00029995466810577443,
      "loss": 3.2811,
      "step": 46450
    },
    {
      "epoch": 5.013477088948787,
      "grad_norm": 0.6671633720397949,
      "learning_rate": 0.00029963086886130593,
      "loss": 3.2851,
      "step": 46500
    },
    {
      "epoch": 5.018867924528302,
      "grad_norm": 0.6657042503356934,
      "learning_rate": 0.00029930706961683754,
      "loss": 3.2927,
      "step": 46550
    },
    {
      "epoch": 5.024258760107816,
      "grad_norm": 0.6508234739303589,
      "learning_rate": 0.0002989832703723691,
      "loss": 3.2926,
      "step": 46600
    },
    {
      "epoch": 5.0296495956873315,
      "grad_norm": 0.7433764934539795,
      "learning_rate": 0.0002986594711279007,
      "loss": 3.2969,
      "step": 46650
    },
    {
      "epoch": 5.035040431266847,
      "grad_norm": 0.6221592426300049,
      "learning_rate": 0.00029833567188343224,
      "loss": 3.2891,
      "step": 46700
    },
    {
      "epoch": 5.040431266846361,
      "grad_norm": 0.6607252955436707,
      "learning_rate": 0.00029801187263896385,
      "loss": 3.3061,
      "step": 46750
    },
    {
      "epoch": 5.045822102425876,
      "grad_norm": 0.7137832045555115,
      "learning_rate": 0.0002976880733944954,
      "loss": 3.2899,
      "step": 46800
    },
    {
      "epoch": 5.051212938005391,
      "grad_norm": 0.6454117894172668,
      "learning_rate": 0.00029736427415002695,
      "loss": 3.308,
      "step": 46850
    },
    {
      "epoch": 5.056603773584905,
      "grad_norm": 0.6954532265663147,
      "learning_rate": 0.0002970404749055585,
      "loss": 3.296,
      "step": 46900
    },
    {
      "epoch": 5.061994609164421,
      "grad_norm": 0.6978551149368286,
      "learning_rate": 0.0002967166756610901,
      "loss": 3.2855,
      "step": 46950
    },
    {
      "epoch": 5.067385444743936,
      "grad_norm": 0.7152319550514221,
      "learning_rate": 0.00029639287641662165,
      "loss": 3.3041,
      "step": 47000
    },
    {
      "epoch": 5.067385444743936,
      "eval_accuracy": 0.3810717542584598,
      "eval_loss": 3.4087512493133545,
      "eval_runtime": 183.5405,
      "eval_samples_per_second": 98.131,
      "eval_steps_per_second": 6.135,
      "step": 47000
    },
    {
      "epoch": 5.07277628032345,
      "grad_norm": 0.6855849623680115,
      "learning_rate": 0.00029606907717215326,
      "loss": 3.3057,
      "step": 47050
    },
    {
      "epoch": 5.078167115902965,
      "grad_norm": 0.6449735164642334,
      "learning_rate": 0.0002957452779276848,
      "loss": 3.3095,
      "step": 47100
    },
    {
      "epoch": 5.083557951482479,
      "grad_norm": 0.668123722076416,
      "learning_rate": 0.00029542147868321636,
      "loss": 3.304,
      "step": 47150
    },
    {
      "epoch": 5.0889487870619945,
      "grad_norm": 0.6669653058052063,
      "learning_rate": 0.00029509767943874796,
      "loss": 3.2993,
      "step": 47200
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.6730850338935852,
      "learning_rate": 0.0002947738801942795,
      "loss": 3.3039,
      "step": 47250
    },
    {
      "epoch": 5.099730458221024,
      "grad_norm": 0.6958271265029907,
      "learning_rate": 0.0002944500809498111,
      "loss": 3.2981,
      "step": 47300
    },
    {
      "epoch": 5.105121293800539,
      "grad_norm": 0.6824030876159668,
      "learning_rate": 0.00029412628170534267,
      "loss": 3.2891,
      "step": 47350
    },
    {
      "epoch": 5.110512129380054,
      "grad_norm": 0.6391562223434448,
      "learning_rate": 0.0002938024824608742,
      "loss": 3.3064,
      "step": 47400
    },
    {
      "epoch": 5.115902964959568,
      "grad_norm": 0.6703659296035767,
      "learning_rate": 0.00029347868321640577,
      "loss": 3.3213,
      "step": 47450
    },
    {
      "epoch": 5.121293800539084,
      "grad_norm": 0.6760079860687256,
      "learning_rate": 0.0002931548839719374,
      "loss": 3.3149,
      "step": 47500
    },
    {
      "epoch": 5.126684636118599,
      "grad_norm": 0.6816515326499939,
      "learning_rate": 0.00029283756071235833,
      "loss": 3.3225,
      "step": 47550
    },
    {
      "epoch": 5.132075471698113,
      "grad_norm": 0.6747937202453613,
      "learning_rate": 0.0002925137614678899,
      "loss": 3.314,
      "step": 47600
    },
    {
      "epoch": 5.137466307277628,
      "grad_norm": 0.6805335879325867,
      "learning_rate": 0.00029218996222342143,
      "loss": 3.3041,
      "step": 47650
    },
    {
      "epoch": 5.142857142857143,
      "grad_norm": 0.6687912940979004,
      "learning_rate": 0.00029186616297895303,
      "loss": 3.3185,
      "step": 47700
    },
    {
      "epoch": 5.1482479784366575,
      "grad_norm": 0.6776725649833679,
      "learning_rate": 0.0002915423637344846,
      "loss": 3.311,
      "step": 47750
    },
    {
      "epoch": 5.153638814016173,
      "grad_norm": 0.6747115850448608,
      "learning_rate": 0.00029121856449001613,
      "loss": 3.3006,
      "step": 47800
    },
    {
      "epoch": 5.159029649595688,
      "grad_norm": 0.6835746169090271,
      "learning_rate": 0.00029089476524554774,
      "loss": 3.3243,
      "step": 47850
    },
    {
      "epoch": 5.164420485175202,
      "grad_norm": 0.6365933418273926,
      "learning_rate": 0.0002905709660010793,
      "loss": 3.3159,
      "step": 47900
    },
    {
      "epoch": 5.169811320754717,
      "grad_norm": 0.8081470727920532,
      "learning_rate": 0.0002902471667566109,
      "loss": 3.3033,
      "step": 47950
    },
    {
      "epoch": 5.175202156334231,
      "grad_norm": 0.6865705251693726,
      "learning_rate": 0.00028992336751214245,
      "loss": 3.3075,
      "step": 48000
    },
    {
      "epoch": 5.175202156334231,
      "eval_accuracy": 0.3815551509563464,
      "eval_loss": 3.405240297317505,
      "eval_runtime": 183.5473,
      "eval_samples_per_second": 98.127,
      "eval_steps_per_second": 6.135,
      "step": 48000
    },
    {
      "epoch": 5.180592991913747,
      "grad_norm": 0.6026649475097656,
      "learning_rate": 0.00028959956826767405,
      "loss": 3.3132,
      "step": 48050
    },
    {
      "epoch": 5.185983827493262,
      "grad_norm": 0.7102674245834351,
      "learning_rate": 0.00028927576902320555,
      "loss": 3.3234,
      "step": 48100
    },
    {
      "epoch": 5.191374663072776,
      "grad_norm": 0.6844705939292908,
      "learning_rate": 0.00028895196977873715,
      "loss": 3.319,
      "step": 48150
    },
    {
      "epoch": 5.196765498652291,
      "grad_norm": 0.733917236328125,
      "learning_rate": 0.0002886281705342687,
      "loss": 3.329,
      "step": 48200
    },
    {
      "epoch": 5.202156334231806,
      "grad_norm": 0.6499011516571045,
      "learning_rate": 0.0002883043712898003,
      "loss": 3.3221,
      "step": 48250
    },
    {
      "epoch": 5.2075471698113205,
      "grad_norm": 0.6283177733421326,
      "learning_rate": 0.00028798057204533186,
      "loss": 3.2958,
      "step": 48300
    },
    {
      "epoch": 5.212938005390836,
      "grad_norm": 0.6744676232337952,
      "learning_rate": 0.00028765677280086346,
      "loss": 3.3312,
      "step": 48350
    },
    {
      "epoch": 5.218328840970351,
      "grad_norm": 0.6724770665168762,
      "learning_rate": 0.000287332973556395,
      "loss": 3.3213,
      "step": 48400
    },
    {
      "epoch": 5.223719676549865,
      "grad_norm": 0.7276390790939331,
      "learning_rate": 0.0002870091743119266,
      "loss": 3.3264,
      "step": 48450
    },
    {
      "epoch": 5.22911051212938,
      "grad_norm": 0.6782853007316589,
      "learning_rate": 0.00028668537506745817,
      "loss": 3.3324,
      "step": 48500
    },
    {
      "epoch": 5.234501347708895,
      "grad_norm": 0.7115477323532104,
      "learning_rate": 0.0002863615758229897,
      "loss": 3.3095,
      "step": 48550
    },
    {
      "epoch": 5.2398921832884096,
      "grad_norm": 0.6378598213195801,
      "learning_rate": 0.00028603777657852127,
      "loss": 3.3119,
      "step": 48600
    },
    {
      "epoch": 5.245283018867925,
      "grad_norm": 0.678990364074707,
      "learning_rate": 0.0002857139773340529,
      "loss": 3.3342,
      "step": 48650
    },
    {
      "epoch": 5.250673854447439,
      "grad_norm": 0.7339763045310974,
      "learning_rate": 0.0002853901780895844,
      "loss": 3.3229,
      "step": 48700
    },
    {
      "epoch": 5.256064690026954,
      "grad_norm": 0.7002367973327637,
      "learning_rate": 0.00028506637884511603,
      "loss": 3.3221,
      "step": 48750
    },
    {
      "epoch": 5.261455525606469,
      "grad_norm": 0.6659910082817078,
      "learning_rate": 0.0002847425796006476,
      "loss": 3.3332,
      "step": 48800
    },
    {
      "epoch": 5.2668463611859835,
      "grad_norm": 0.6622328758239746,
      "learning_rate": 0.00028441878035617913,
      "loss": 3.3246,
      "step": 48850
    },
    {
      "epoch": 5.272237196765499,
      "grad_norm": 0.6887483596801758,
      "learning_rate": 0.00028409498111171073,
      "loss": 3.3156,
      "step": 48900
    },
    {
      "epoch": 5.277628032345014,
      "grad_norm": 0.6723185181617737,
      "learning_rate": 0.0002837711818672423,
      "loss": 3.326,
      "step": 48950
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.6477317214012146,
      "learning_rate": 0.00028344738262277384,
      "loss": 3.3062,
      "step": 49000
    },
    {
      "epoch": 5.283018867924528,
      "eval_accuracy": 0.3819458667409273,
      "eval_loss": 3.401156425476074,
      "eval_runtime": 183.2869,
      "eval_samples_per_second": 98.267,
      "eval_steps_per_second": 6.143,
      "step": 49000
    },
    {
      "epoch": 5.288409703504043,
      "grad_norm": 0.7285588979721069,
      "learning_rate": 0.00028312358337830544,
      "loss": 3.3276,
      "step": 49050
    },
    {
      "epoch": 5.293800539083558,
      "grad_norm": 0.6508597731590271,
      "learning_rate": 0.000282799784133837,
      "loss": 3.3288,
      "step": 49100
    },
    {
      "epoch": 5.2991913746630726,
      "grad_norm": 0.6885204911231995,
      "learning_rate": 0.00028247598488936854,
      "loss": 3.3244,
      "step": 49150
    },
    {
      "epoch": 5.304582210242588,
      "grad_norm": 0.6657274961471558,
      "learning_rate": 0.00028215218564490015,
      "loss": 3.3197,
      "step": 49200
    },
    {
      "epoch": 5.309973045822103,
      "grad_norm": 0.6937828063964844,
      "learning_rate": 0.0002818283864004317,
      "loss": 3.2995,
      "step": 49250
    },
    {
      "epoch": 5.315363881401617,
      "grad_norm": 0.6877214312553406,
      "learning_rate": 0.0002815045871559633,
      "loss": 3.3058,
      "step": 49300
    },
    {
      "epoch": 5.320754716981132,
      "grad_norm": 0.7068207859992981,
      "learning_rate": 0.00028118078791149485,
      "loss": 3.3351,
      "step": 49350
    },
    {
      "epoch": 5.3261455525606465,
      "grad_norm": 0.6211855411529541,
      "learning_rate": 0.0002808569886670264,
      "loss": 3.3069,
      "step": 49400
    },
    {
      "epoch": 5.331536388140162,
      "grad_norm": 0.7042295932769775,
      "learning_rate": 0.00028053318942255795,
      "loss": 3.322,
      "step": 49450
    },
    {
      "epoch": 5.336927223719677,
      "grad_norm": 0.6894974708557129,
      "learning_rate": 0.00028020939017808956,
      "loss": 3.3321,
      "step": 49500
    },
    {
      "epoch": 5.342318059299191,
      "grad_norm": 0.645094633102417,
      "learning_rate": 0.0002798920669185105,
      "loss": 3.3203,
      "step": 49550
    },
    {
      "epoch": 5.347708894878706,
      "grad_norm": 0.6860967874526978,
      "learning_rate": 0.00027956826767404206,
      "loss": 3.3143,
      "step": 49600
    },
    {
      "epoch": 5.353099730458221,
      "grad_norm": 0.7084428668022156,
      "learning_rate": 0.00027924446842957367,
      "loss": 3.3198,
      "step": 49650
    },
    {
      "epoch": 5.3584905660377355,
      "grad_norm": 0.6508020162582397,
      "learning_rate": 0.0002789206691851052,
      "loss": 3.2959,
      "step": 49700
    },
    {
      "epoch": 5.363881401617251,
      "grad_norm": 0.651019811630249,
      "learning_rate": 0.00027859686994063677,
      "loss": 3.3288,
      "step": 49750
    },
    {
      "epoch": 5.369272237196766,
      "grad_norm": 0.6705125570297241,
      "learning_rate": 0.0002782730706961683,
      "loss": 3.3182,
      "step": 49800
    },
    {
      "epoch": 5.37466307277628,
      "grad_norm": 0.6464545130729675,
      "learning_rate": 0.0002779492714516999,
      "loss": 3.3219,
      "step": 49850
    },
    {
      "epoch": 5.380053908355795,
      "grad_norm": 0.6712095737457275,
      "learning_rate": 0.00027762547220723147,
      "loss": 3.3218,
      "step": 49900
    },
    {
      "epoch": 5.38544474393531,
      "grad_norm": 0.7088808417320251,
      "learning_rate": 0.0002773016729627631,
      "loss": 3.32,
      "step": 49950
    },
    {
      "epoch": 5.390835579514825,
      "grad_norm": 0.6843079924583435,
      "learning_rate": 0.00027697787371829463,
      "loss": 3.331,
      "step": 50000
    },
    {
      "epoch": 5.390835579514825,
      "eval_accuracy": 0.3825648622428064,
      "eval_loss": 3.397858142852783,
      "eval_runtime": 183.2184,
      "eval_samples_per_second": 98.303,
      "eval_steps_per_second": 6.146,
      "step": 50000
    },
    {
      "epoch": 5.39622641509434,
      "grad_norm": 0.6787253618240356,
      "learning_rate": 0.00027665407447382623,
      "loss": 3.324,
      "step": 50050
    },
    {
      "epoch": 5.401617250673855,
      "grad_norm": 0.7045350074768066,
      "learning_rate": 0.0002763302752293578,
      "loss": 3.3343,
      "step": 50100
    },
    {
      "epoch": 5.407008086253369,
      "grad_norm": 0.6545983552932739,
      "learning_rate": 0.00027600647598488933,
      "loss": 3.3073,
      "step": 50150
    },
    {
      "epoch": 5.412398921832884,
      "grad_norm": 0.7040044665336609,
      "learning_rate": 0.0002756826767404209,
      "loss": 3.3056,
      "step": 50200
    },
    {
      "epoch": 5.4177897574123985,
      "grad_norm": 0.682117760181427,
      "learning_rate": 0.0002753588774959525,
      "loss": 3.3306,
      "step": 50250
    },
    {
      "epoch": 5.423180592991914,
      "grad_norm": 0.6643507480621338,
      "learning_rate": 0.00027503507825148404,
      "loss": 3.3126,
      "step": 50300
    },
    {
      "epoch": 5.428571428571429,
      "grad_norm": 0.6264309287071228,
      "learning_rate": 0.00027471127900701564,
      "loss": 3.3265,
      "step": 50350
    },
    {
      "epoch": 5.433962264150943,
      "grad_norm": 0.6541180610656738,
      "learning_rate": 0.0002743874797625472,
      "loss": 3.3153,
      "step": 50400
    },
    {
      "epoch": 5.439353099730458,
      "grad_norm": 0.7172690629959106,
      "learning_rate": 0.0002740636805180788,
      "loss": 3.327,
      "step": 50450
    },
    {
      "epoch": 5.444743935309973,
      "grad_norm": 0.6622113585472107,
      "learning_rate": 0.00027373988127361035,
      "loss": 3.3246,
      "step": 50500
    },
    {
      "epoch": 5.450134770889488,
      "grad_norm": 0.6717416644096375,
      "learning_rate": 0.0002734160820291419,
      "loss": 3.3573,
      "step": 50550
    },
    {
      "epoch": 5.455525606469003,
      "grad_norm": 0.6707860231399536,
      "learning_rate": 0.00027309228278467345,
      "loss": 3.3132,
      "step": 50600
    },
    {
      "epoch": 5.460916442048518,
      "grad_norm": 0.6682204008102417,
      "learning_rate": 0.00027276848354020506,
      "loss": 3.3331,
      "step": 50650
    },
    {
      "epoch": 5.466307277628032,
      "grad_norm": 0.6391171813011169,
      "learning_rate": 0.0002724446842957366,
      "loss": 3.3307,
      "step": 50700
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.681674063205719,
      "learning_rate": 0.0002721208850512682,
      "loss": 3.3208,
      "step": 50750
    },
    {
      "epoch": 5.4770889487870615,
      "grad_norm": 0.6989498138427734,
      "learning_rate": 0.00027179708580679976,
      "loss": 3.3355,
      "step": 50800
    },
    {
      "epoch": 5.482479784366577,
      "grad_norm": 0.6760035753250122,
      "learning_rate": 0.0002714732865623313,
      "loss": 3.336,
      "step": 50850
    },
    {
      "epoch": 5.487870619946092,
      "grad_norm": 0.6494047045707703,
      "learning_rate": 0.0002711494873178629,
      "loss": 3.328,
      "step": 50900
    },
    {
      "epoch": 5.493261455525606,
      "grad_norm": 0.7357152104377747,
      "learning_rate": 0.00027082568807339447,
      "loss": 3.3348,
      "step": 50950
    },
    {
      "epoch": 5.498652291105121,
      "grad_norm": 0.6467098593711853,
      "learning_rate": 0.000270501888828926,
      "loss": 3.3158,
      "step": 51000
    },
    {
      "epoch": 5.498652291105121,
      "eval_accuracy": 0.3829662260103931,
      "eval_loss": 3.393737316131592,
      "eval_runtime": 183.6994,
      "eval_samples_per_second": 98.046,
      "eval_steps_per_second": 6.13,
      "step": 51000
    },
    {
      "epoch": 5.504043126684636,
      "grad_norm": 0.6844212412834167,
      "learning_rate": 0.0002701780895844576,
      "loss": 3.3327,
      "step": 51050
    },
    {
      "epoch": 5.509433962264151,
      "grad_norm": 0.6608554124832153,
      "learning_rate": 0.0002698542903399892,
      "loss": 3.3068,
      "step": 51100
    },
    {
      "epoch": 5.514824797843666,
      "grad_norm": 0.7180224061012268,
      "learning_rate": 0.0002695304910955207,
      "loss": 3.3203,
      "step": 51150
    },
    {
      "epoch": 5.520215633423181,
      "grad_norm": 0.7351436614990234,
      "learning_rate": 0.00026920669185105233,
      "loss": 3.3291,
      "step": 51200
    },
    {
      "epoch": 5.525606469002695,
      "grad_norm": 0.6572839021682739,
      "learning_rate": 0.0002688828926065839,
      "loss": 3.3288,
      "step": 51250
    },
    {
      "epoch": 5.53099730458221,
      "grad_norm": 0.7058694958686829,
      "learning_rate": 0.0002685590933621155,
      "loss": 3.3398,
      "step": 51300
    },
    {
      "epoch": 5.536388140161725,
      "grad_norm": 0.6541451811790466,
      "learning_rate": 0.00026823529411764704,
      "loss": 3.3093,
      "step": 51350
    },
    {
      "epoch": 5.54177897574124,
      "grad_norm": 0.7691888809204102,
      "learning_rate": 0.0002679114948731786,
      "loss": 3.3288,
      "step": 51400
    },
    {
      "epoch": 5.547169811320755,
      "grad_norm": 0.6912489533424377,
      "learning_rate": 0.0002675876956287102,
      "loss": 3.3276,
      "step": 51450
    },
    {
      "epoch": 5.55256064690027,
      "grad_norm": 0.6621717810630798,
      "learning_rate": 0.00026726389638424174,
      "loss": 3.3131,
      "step": 51500
    },
    {
      "epoch": 5.557951482479784,
      "grad_norm": 0.674435555934906,
      "learning_rate": 0.0002669400971397733,
      "loss": 3.3392,
      "step": 51550
    },
    {
      "epoch": 5.563342318059299,
      "grad_norm": 0.7033376097679138,
      "learning_rate": 0.00026662277388019424,
      "loss": 3.3262,
      "step": 51600
    },
    {
      "epoch": 5.568733153638814,
      "grad_norm": 0.6880776882171631,
      "learning_rate": 0.00026629897463572585,
      "loss": 3.3215,
      "step": 51650
    },
    {
      "epoch": 5.574123989218329,
      "grad_norm": 0.6662737727165222,
      "learning_rate": 0.0002659751753912574,
      "loss": 3.3392,
      "step": 51700
    },
    {
      "epoch": 5.579514824797844,
      "grad_norm": 0.6773297190666199,
      "learning_rate": 0.00026565137614678895,
      "loss": 3.328,
      "step": 51750
    },
    {
      "epoch": 5.584905660377358,
      "grad_norm": 0.7141923308372498,
      "learning_rate": 0.0002653275769023205,
      "loss": 3.3329,
      "step": 51800
    },
    {
      "epoch": 5.590296495956873,
      "grad_norm": 0.6370801329612732,
      "learning_rate": 0.0002650037776578521,
      "loss": 3.3387,
      "step": 51850
    },
    {
      "epoch": 5.595687331536388,
      "grad_norm": 0.7268838286399841,
      "learning_rate": 0.00026468645439827306,
      "loss": 3.3367,
      "step": 51900
    },
    {
      "epoch": 5.601078167115903,
      "grad_norm": 0.6624352335929871,
      "learning_rate": 0.0002643626551538046,
      "loss": 3.3439,
      "step": 51950
    },
    {
      "epoch": 5.606469002695418,
      "grad_norm": 0.6981469988822937,
      "learning_rate": 0.0002640388559093362,
      "loss": 3.3354,
      "step": 52000
    },
    {
      "epoch": 5.606469002695418,
      "eval_accuracy": 0.38319341919881317,
      "eval_loss": 3.390268087387085,
      "eval_runtime": 183.3396,
      "eval_samples_per_second": 98.238,
      "eval_steps_per_second": 6.142,
      "step": 52000
    },
    {
      "epoch": 5.611859838274933,
      "grad_norm": 0.7103446125984192,
      "learning_rate": 0.00026371505666486776,
      "loss": 3.3376,
      "step": 52050
    },
    {
      "epoch": 5.617250673854447,
      "grad_norm": 0.6647046804428101,
      "learning_rate": 0.0002633912574203993,
      "loss": 3.3324,
      "step": 52100
    },
    {
      "epoch": 5.622641509433962,
      "grad_norm": 0.6633039712905884,
      "learning_rate": 0.00026306745817593086,
      "loss": 3.3342,
      "step": 52150
    },
    {
      "epoch": 5.628032345013477,
      "grad_norm": 0.7168214917182922,
      "learning_rate": 0.00026274365893146247,
      "loss": 3.3202,
      "step": 52200
    },
    {
      "epoch": 5.633423180592992,
      "grad_norm": 0.6797349452972412,
      "learning_rate": 0.000262419859686994,
      "loss": 3.3205,
      "step": 52250
    },
    {
      "epoch": 5.638814016172507,
      "grad_norm": 0.7016252279281616,
      "learning_rate": 0.0002620960604425256,
      "loss": 3.341,
      "step": 52300
    },
    {
      "epoch": 5.644204851752022,
      "grad_norm": 0.6899751424789429,
      "learning_rate": 0.0002617722611980572,
      "loss": 3.3142,
      "step": 52350
    },
    {
      "epoch": 5.649595687331536,
      "grad_norm": 0.6652300953865051,
      "learning_rate": 0.0002614484619535888,
      "loss": 3.3168,
      "step": 52400
    },
    {
      "epoch": 5.654986522911051,
      "grad_norm": 0.6907793283462524,
      "learning_rate": 0.00026112466270912033,
      "loss": 3.3208,
      "step": 52450
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.700583279132843,
      "learning_rate": 0.0002608008634646519,
      "loss": 3.3349,
      "step": 52500
    },
    {
      "epoch": 5.665768194070081,
      "grad_norm": 0.7041873335838318,
      "learning_rate": 0.00026047706422018343,
      "loss": 3.3405,
      "step": 52550
    },
    {
      "epoch": 5.671159029649596,
      "grad_norm": 0.7494111657142639,
      "learning_rate": 0.00026015326497571504,
      "loss": 3.3385,
      "step": 52600
    },
    {
      "epoch": 5.67654986522911,
      "grad_norm": 0.645006000995636,
      "learning_rate": 0.0002598294657312466,
      "loss": 3.3303,
      "step": 52650
    },
    {
      "epoch": 5.681940700808625,
      "grad_norm": 0.6822366118431091,
      "learning_rate": 0.0002595056664867782,
      "loss": 3.3312,
      "step": 52700
    },
    {
      "epoch": 5.6873315363881405,
      "grad_norm": 0.6748678088188171,
      "learning_rate": 0.00025918186724230974,
      "loss": 3.3285,
      "step": 52750
    },
    {
      "epoch": 5.692722371967655,
      "grad_norm": 0.6696112155914307,
      "learning_rate": 0.00025885806799784135,
      "loss": 3.3249,
      "step": 52800
    },
    {
      "epoch": 5.69811320754717,
      "grad_norm": 0.6802454590797424,
      "learning_rate": 0.0002585342687533729,
      "loss": 3.3251,
      "step": 52850
    },
    {
      "epoch": 5.703504043126685,
      "grad_norm": 0.7200360894203186,
      "learning_rate": 0.00025821046950890445,
      "loss": 3.3194,
      "step": 52900
    },
    {
      "epoch": 5.708894878706199,
      "grad_norm": 0.6851761937141418,
      "learning_rate": 0.000257886670264436,
      "loss": 3.3145,
      "step": 52950
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.6522359251976013,
      "learning_rate": 0.0002575628710199676,
      "loss": 3.3422,
      "step": 53000
    },
    {
      "epoch": 5.714285714285714,
      "eval_accuracy": 0.38380991568427253,
      "eval_loss": 3.3851399421691895,
      "eval_runtime": 183.4753,
      "eval_samples_per_second": 98.166,
      "eval_steps_per_second": 6.137,
      "step": 53000
    },
    {
      "epoch": 5.719676549865229,
      "grad_norm": 0.6580410003662109,
      "learning_rate": 0.00025723907177549915,
      "loss": 3.3208,
      "step": 53050
    },
    {
      "epoch": 5.725067385444744,
      "grad_norm": 0.6901311278343201,
      "learning_rate": 0.00025691527253103076,
      "loss": 3.314,
      "step": 53100
    },
    {
      "epoch": 5.730458221024259,
      "grad_norm": 0.7142326235771179,
      "learning_rate": 0.0002565914732865623,
      "loss": 3.3408,
      "step": 53150
    },
    {
      "epoch": 5.735849056603773,
      "grad_norm": 0.6859725117683411,
      "learning_rate": 0.00025626767404209386,
      "loss": 3.3055,
      "step": 53200
    },
    {
      "epoch": 5.741239892183288,
      "grad_norm": 0.6842724680900574,
      "learning_rate": 0.00025594387479762546,
      "loss": 3.3248,
      "step": 53250
    },
    {
      "epoch": 5.7466307277628035,
      "grad_norm": 0.6995661854743958,
      "learning_rate": 0.000255620075553157,
      "loss": 3.3282,
      "step": 53300
    },
    {
      "epoch": 5.752021563342318,
      "grad_norm": 0.6869672536849976,
      "learning_rate": 0.00025529627630868857,
      "loss": 3.3208,
      "step": 53350
    },
    {
      "epoch": 5.757412398921833,
      "grad_norm": 0.6568567156791687,
      "learning_rate": 0.00025497247706422017,
      "loss": 3.3467,
      "step": 53400
    },
    {
      "epoch": 5.762803234501348,
      "grad_norm": 0.6627490520477295,
      "learning_rate": 0.0002546486778197517,
      "loss": 3.3176,
      "step": 53450
    },
    {
      "epoch": 5.768194070080862,
      "grad_norm": 0.6492809653282166,
      "learning_rate": 0.00025432487857528327,
      "loss": 3.3171,
      "step": 53500
    },
    {
      "epoch": 5.773584905660377,
      "grad_norm": 0.7379694581031799,
      "learning_rate": 0.0002540010793308149,
      "loss": 3.3445,
      "step": 53550
    },
    {
      "epoch": 5.7789757412398925,
      "grad_norm": 0.7134669423103333,
      "learning_rate": 0.0002536772800863464,
      "loss": 3.3258,
      "step": 53600
    },
    {
      "epoch": 5.784366576819407,
      "grad_norm": 0.7085931897163391,
      "learning_rate": 0.00025335348084187803,
      "loss": 3.3441,
      "step": 53650
    },
    {
      "epoch": 5.789757412398922,
      "grad_norm": 0.7263320088386536,
      "learning_rate": 0.0002530296815974096,
      "loss": 3.3446,
      "step": 53700
    },
    {
      "epoch": 5.795148247978437,
      "grad_norm": 0.7074159383773804,
      "learning_rate": 0.0002527058823529412,
      "loss": 3.3256,
      "step": 53750
    },
    {
      "epoch": 5.800539083557951,
      "grad_norm": 0.6964407563209534,
      "learning_rate": 0.0002523820831084727,
      "loss": 3.3323,
      "step": 53800
    },
    {
      "epoch": 5.8059299191374665,
      "grad_norm": 0.6850724220275879,
      "learning_rate": 0.0002520582838640043,
      "loss": 3.3159,
      "step": 53850
    },
    {
      "epoch": 5.811320754716981,
      "grad_norm": 0.6719462871551514,
      "learning_rate": 0.00025173448461953584,
      "loss": 3.3493,
      "step": 53900
    },
    {
      "epoch": 5.816711590296496,
      "grad_norm": 0.6726512908935547,
      "learning_rate": 0.00025141068537506744,
      "loss": 3.3207,
      "step": 53950
    },
    {
      "epoch": 5.822102425876011,
      "grad_norm": 0.6462234258651733,
      "learning_rate": 0.000251086886130599,
      "loss": 3.3243,
      "step": 54000
    },
    {
      "epoch": 5.822102425876011,
      "eval_accuracy": 0.38394768754602127,
      "eval_loss": 3.3802895545959473,
      "eval_runtime": 183.3426,
      "eval_samples_per_second": 98.237,
      "eval_steps_per_second": 6.142,
      "step": 54000
    },
    {
      "epoch": 5.827493261455525,
      "grad_norm": 0.7041391134262085,
      "learning_rate": 0.0002507630868861306,
      "loss": 3.319,
      "step": 54050
    },
    {
      "epoch": 5.83288409703504,
      "grad_norm": 0.671320915222168,
      "learning_rate": 0.00025043928764166215,
      "loss": 3.3248,
      "step": 54100
    },
    {
      "epoch": 5.8382749326145555,
      "grad_norm": 0.6599231958389282,
      "learning_rate": 0.00025011548839719375,
      "loss": 3.342,
      "step": 54150
    },
    {
      "epoch": 5.84366576819407,
      "grad_norm": 0.6870060563087463,
      "learning_rate": 0.0002497916891527253,
      "loss": 3.3316,
      "step": 54200
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.6556626558303833,
      "learning_rate": 0.00024946788990825686,
      "loss": 3.3394,
      "step": 54250
    },
    {
      "epoch": 5.8544474393531,
      "grad_norm": 0.7121956944465637,
      "learning_rate": 0.0002491440906637884,
      "loss": 3.3378,
      "step": 54300
    },
    {
      "epoch": 5.859838274932614,
      "grad_norm": 0.6935588717460632,
      "learning_rate": 0.00024882029141932,
      "loss": 3.3287,
      "step": 54350
    },
    {
      "epoch": 5.8652291105121295,
      "grad_norm": 0.6818873882293701,
      "learning_rate": 0.00024849649217485156,
      "loss": 3.3147,
      "step": 54400
    },
    {
      "epoch": 5.870619946091644,
      "grad_norm": 0.6630216836929321,
      "learning_rate": 0.00024817269293038317,
      "loss": 3.3099,
      "step": 54450
    },
    {
      "epoch": 5.876010781671159,
      "grad_norm": 0.7189406752586365,
      "learning_rate": 0.0002478488936859147,
      "loss": 3.3263,
      "step": 54500
    },
    {
      "epoch": 5.881401617250674,
      "grad_norm": 0.708378791809082,
      "learning_rate": 0.00024752509444144627,
      "loss": 3.3216,
      "step": 54550
    },
    {
      "epoch": 5.886792452830189,
      "grad_norm": 0.696700394153595,
      "learning_rate": 0.00024720129519697787,
      "loss": 3.3531,
      "step": 54600
    },
    {
      "epoch": 5.892183288409703,
      "grad_norm": 0.6870036125183105,
      "learning_rate": 0.0002468774959525094,
      "loss": 3.3366,
      "step": 54650
    },
    {
      "epoch": 5.8975741239892185,
      "grad_norm": 0.7101106643676758,
      "learning_rate": 0.00024655369670804097,
      "loss": 3.3365,
      "step": 54700
    },
    {
      "epoch": 5.902964959568733,
      "grad_norm": 0.7199245691299438,
      "learning_rate": 0.0002462298974635726,
      "loss": 3.3377,
      "step": 54750
    },
    {
      "epoch": 5.908355795148248,
      "grad_norm": 0.714850127696991,
      "learning_rate": 0.00024590609821910413,
      "loss": 3.3418,
      "step": 54800
    },
    {
      "epoch": 5.913746630727763,
      "grad_norm": 0.7096270322799683,
      "learning_rate": 0.0002455822989746357,
      "loss": 3.3353,
      "step": 54850
    },
    {
      "epoch": 5.919137466307277,
      "grad_norm": 0.691370964050293,
      "learning_rate": 0.0002452584997301673,
      "loss": 3.3158,
      "step": 54900
    },
    {
      "epoch": 5.9245283018867925,
      "grad_norm": 0.6646804213523865,
      "learning_rate": 0.00024493470048569883,
      "loss": 3.3219,
      "step": 54950
    },
    {
      "epoch": 5.929919137466308,
      "grad_norm": 0.6864519119262695,
      "learning_rate": 0.00024461090124123044,
      "loss": 3.3245,
      "step": 55000
    },
    {
      "epoch": 5.929919137466308,
      "eval_accuracy": 0.3845000788276701,
      "eval_loss": 3.375938892364502,
      "eval_runtime": 183.427,
      "eval_samples_per_second": 98.192,
      "eval_steps_per_second": 6.139,
      "step": 55000
    },
    {
      "epoch": 5.935309973045822,
      "grad_norm": 0.6417977809906006,
      "learning_rate": 0.000244287101996762,
      "loss": 3.3377,
      "step": 55050
    },
    {
      "epoch": 5.940700808625337,
      "grad_norm": 0.7369255423545837,
      "learning_rate": 0.00024396330275229354,
      "loss": 3.3254,
      "step": 55100
    },
    {
      "epoch": 5.946091644204852,
      "grad_norm": 0.6987436413764954,
      "learning_rate": 0.00024363950350782512,
      "loss": 3.3251,
      "step": 55150
    },
    {
      "epoch": 5.951482479784366,
      "grad_norm": 0.689603865146637,
      "learning_rate": 0.0002433157042633567,
      "loss": 3.3442,
      "step": 55200
    },
    {
      "epoch": 5.9568733153638815,
      "grad_norm": 0.7123790979385376,
      "learning_rate": 0.00024299190501888827,
      "loss": 3.3154,
      "step": 55250
    },
    {
      "epoch": 5.962264150943396,
      "grad_norm": 0.666307270526886,
      "learning_rate": 0.00024266810577441985,
      "loss": 3.3173,
      "step": 55300
    },
    {
      "epoch": 5.967654986522911,
      "grad_norm": 0.6958657503128052,
      "learning_rate": 0.00024234430652995143,
      "loss": 3.3256,
      "step": 55350
    },
    {
      "epoch": 5.973045822102426,
      "grad_norm": 0.7178813219070435,
      "learning_rate": 0.00024202050728548298,
      "loss": 3.3286,
      "step": 55400
    },
    {
      "epoch": 5.97843665768194,
      "grad_norm": 0.690979540348053,
      "learning_rate": 0.00024169670804101456,
      "loss": 3.3219,
      "step": 55450
    },
    {
      "epoch": 5.9838274932614555,
      "grad_norm": 0.6696831583976746,
      "learning_rate": 0.0002413729087965461,
      "loss": 3.3019,
      "step": 55500
    },
    {
      "epoch": 5.989218328840971,
      "grad_norm": 0.6762212514877319,
      "learning_rate": 0.00024104910955207768,
      "loss": 3.3312,
      "step": 55550
    },
    {
      "epoch": 5.994609164420485,
      "grad_norm": 0.6979408264160156,
      "learning_rate": 0.00024072531030760926,
      "loss": 3.3208,
      "step": 55600
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.4232317209243774,
      "learning_rate": 0.00024040151106314084,
      "loss": 3.3492,
      "step": 55650
    },
    {
      "epoch": 6.005390835579515,
      "grad_norm": 0.6714563965797424,
      "learning_rate": 0.0002400777118186724,
      "loss": 3.2286,
      "step": 55700
    },
    {
      "epoch": 6.010781671159029,
      "grad_norm": 0.6843574047088623,
      "learning_rate": 0.00023975391257420397,
      "loss": 3.2305,
      "step": 55750
    },
    {
      "epoch": 6.0161725067385445,
      "grad_norm": 0.7195513248443604,
      "learning_rate": 0.00023943011332973555,
      "loss": 3.2422,
      "step": 55800
    },
    {
      "epoch": 6.02156334231806,
      "grad_norm": 0.6866133809089661,
      "learning_rate": 0.00023910631408526712,
      "loss": 3.2445,
      "step": 55850
    },
    {
      "epoch": 6.026954177897574,
      "grad_norm": 0.7409161925315857,
      "learning_rate": 0.0002387825148407987,
      "loss": 3.2428,
      "step": 55900
    },
    {
      "epoch": 6.032345013477089,
      "grad_norm": 0.7081006169319153,
      "learning_rate": 0.00023845871559633025,
      "loss": 3.2432,
      "step": 55950
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.736912727355957,
      "learning_rate": 0.0002381413923367512,
      "loss": 3.2569,
      "step": 56000
    },
    {
      "epoch": 6.037735849056604,
      "eval_accuracy": 0.38479583198830114,
      "eval_loss": 3.3801846504211426,
      "eval_runtime": 183.4392,
      "eval_samples_per_second": 98.185,
      "eval_steps_per_second": 6.138,
      "step": 56000
    },
    {
      "epoch": 6.0431266846361185,
      "grad_norm": 0.7223060131072998,
      "learning_rate": 0.00023781759309228275,
      "loss": 3.2541,
      "step": 56050
    },
    {
      "epoch": 6.048517520215634,
      "grad_norm": 0.853127658367157,
      "learning_rate": 0.00023749379384781433,
      "loss": 3.2419,
      "step": 56100
    },
    {
      "epoch": 6.053908355795148,
      "grad_norm": 0.7139176726341248,
      "learning_rate": 0.0002371699946033459,
      "loss": 3.2564,
      "step": 56150
    },
    {
      "epoch": 6.059299191374663,
      "grad_norm": 0.7075513005256653,
      "learning_rate": 0.0002368461953588775,
      "loss": 3.2494,
      "step": 56200
    },
    {
      "epoch": 6.064690026954178,
      "grad_norm": 0.7035233974456787,
      "learning_rate": 0.00023652239611440904,
      "loss": 3.2336,
      "step": 56250
    },
    {
      "epoch": 6.070080862533692,
      "grad_norm": 0.7118614912033081,
      "learning_rate": 0.00023619859686994062,
      "loss": 3.2595,
      "step": 56300
    },
    {
      "epoch": 6.0754716981132075,
      "grad_norm": 0.7053843140602112,
      "learning_rate": 0.00023587479762547217,
      "loss": 3.2453,
      "step": 56350
    },
    {
      "epoch": 6.080862533692723,
      "grad_norm": 0.7055346369743347,
      "learning_rate": 0.00023555099838100374,
      "loss": 3.2494,
      "step": 56400
    },
    {
      "epoch": 6.086253369272237,
      "grad_norm": 0.739841878414154,
      "learning_rate": 0.00023522719913653532,
      "loss": 3.2442,
      "step": 56450
    },
    {
      "epoch": 6.091644204851752,
      "grad_norm": 0.7180924415588379,
      "learning_rate": 0.0002349033998920669,
      "loss": 3.2631,
      "step": 56500
    },
    {
      "epoch": 6.097035040431267,
      "grad_norm": 0.7231001853942871,
      "learning_rate": 0.00023457960064759848,
      "loss": 3.2418,
      "step": 56550
    },
    {
      "epoch": 6.1024258760107815,
      "grad_norm": 0.706662118434906,
      "learning_rate": 0.00023425580140313005,
      "loss": 3.2493,
      "step": 56600
    },
    {
      "epoch": 6.107816711590297,
      "grad_norm": 0.6997068524360657,
      "learning_rate": 0.00023393200215866163,
      "loss": 3.2453,
      "step": 56650
    },
    {
      "epoch": 6.113207547169812,
      "grad_norm": 0.6997630000114441,
      "learning_rate": 0.00023360820291419316,
      "loss": 3.2687,
      "step": 56700
    },
    {
      "epoch": 6.118598382749326,
      "grad_norm": 0.7305110692977905,
      "learning_rate": 0.00023328440366972473,
      "loss": 3.2583,
      "step": 56750
    },
    {
      "epoch": 6.123989218328841,
      "grad_norm": 0.7409155368804932,
      "learning_rate": 0.0002329606044252563,
      "loss": 3.2608,
      "step": 56800
    },
    {
      "epoch": 6.129380053908355,
      "grad_norm": 0.6931422352790833,
      "learning_rate": 0.0002326368051807879,
      "loss": 3.2628,
      "step": 56850
    },
    {
      "epoch": 6.1347708894878705,
      "grad_norm": 0.7668306827545166,
      "learning_rate": 0.00023231300593631947,
      "loss": 3.2696,
      "step": 56900
    },
    {
      "epoch": 6.140161725067386,
      "grad_norm": 0.7603601217269897,
      "learning_rate": 0.00023198920669185104,
      "loss": 3.2466,
      "step": 56950
    },
    {
      "epoch": 6.1455525606469,
      "grad_norm": 0.7167591452598572,
      "learning_rate": 0.00023166540744738262,
      "loss": 3.2452,
      "step": 57000
    },
    {
      "epoch": 6.1455525606469,
      "eval_accuracy": 0.38466523121714813,
      "eval_loss": 3.3798298835754395,
      "eval_runtime": 183.6772,
      "eval_samples_per_second": 98.058,
      "eval_steps_per_second": 6.13,
      "step": 57000
    },
    {
      "epoch": 6.150943396226415,
      "grad_norm": 0.6705722808837891,
      "learning_rate": 0.0002313416082029142,
      "loss": 3.2731,
      "step": 57050
    },
    {
      "epoch": 6.15633423180593,
      "grad_norm": 0.691725492477417,
      "learning_rate": 0.00023101780895844572,
      "loss": 3.2644,
      "step": 57100
    },
    {
      "epoch": 6.1617250673854445,
      "grad_norm": 0.6815258264541626,
      "learning_rate": 0.0002306940097139773,
      "loss": 3.247,
      "step": 57150
    },
    {
      "epoch": 6.16711590296496,
      "grad_norm": 0.6972390413284302,
      "learning_rate": 0.00023037021046950888,
      "loss": 3.2524,
      "step": 57200
    },
    {
      "epoch": 6.172506738544475,
      "grad_norm": 0.695257842540741,
      "learning_rate": 0.00023004641122504046,
      "loss": 3.256,
      "step": 57250
    },
    {
      "epoch": 6.177897574123989,
      "grad_norm": 0.8920659422874451,
      "learning_rate": 0.00022972261198057203,
      "loss": 3.255,
      "step": 57300
    },
    {
      "epoch": 6.183288409703504,
      "grad_norm": 0.7532164454460144,
      "learning_rate": 0.0002293988127361036,
      "loss": 3.271,
      "step": 57350
    },
    {
      "epoch": 6.188679245283019,
      "grad_norm": 0.6974930167198181,
      "learning_rate": 0.00022907501349163516,
      "loss": 3.2621,
      "step": 57400
    },
    {
      "epoch": 6.1940700808625335,
      "grad_norm": 0.7437677979469299,
      "learning_rate": 0.00022875121424716674,
      "loss": 3.2682,
      "step": 57450
    },
    {
      "epoch": 6.199460916442049,
      "grad_norm": 0.7247287631034851,
      "learning_rate": 0.00022842741500269832,
      "loss": 3.2651,
      "step": 57500
    },
    {
      "epoch": 6.204851752021563,
      "grad_norm": 0.7260898947715759,
      "learning_rate": 0.00022810361575822987,
      "loss": 3.2551,
      "step": 57550
    },
    {
      "epoch": 6.210242587601078,
      "grad_norm": 0.783756673336029,
      "learning_rate": 0.00022777981651376145,
      "loss": 3.2437,
      "step": 57600
    },
    {
      "epoch": 6.215633423180593,
      "grad_norm": 0.7135037779808044,
      "learning_rate": 0.00022745601726929302,
      "loss": 3.2631,
      "step": 57650
    },
    {
      "epoch": 6.2210242587601075,
      "grad_norm": 0.7173543572425842,
      "learning_rate": 0.00022713221802482457,
      "loss": 3.2685,
      "step": 57700
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.7847826480865479,
      "learning_rate": 0.00022680841878035615,
      "loss": 3.283,
      "step": 57750
    },
    {
      "epoch": 6.231805929919138,
      "grad_norm": 0.7096986770629883,
      "learning_rate": 0.00022648461953588773,
      "loss": 3.28,
      "step": 57800
    },
    {
      "epoch": 6.237196765498652,
      "grad_norm": 0.7000459432601929,
      "learning_rate": 0.0002261608202914193,
      "loss": 3.2509,
      "step": 57850
    },
    {
      "epoch": 6.242587601078167,
      "grad_norm": 0.6716017127037048,
      "learning_rate": 0.00022583702104695088,
      "loss": 3.2718,
      "step": 57900
    },
    {
      "epoch": 6.247978436657682,
      "grad_norm": 0.712790310382843,
      "learning_rate": 0.00022551322180248246,
      "loss": 3.2622,
      "step": 57950
    },
    {
      "epoch": 6.2533692722371965,
      "grad_norm": 0.7777376770973206,
      "learning_rate": 0.00022518942255801399,
      "loss": 3.2473,
      "step": 58000
    },
    {
      "epoch": 6.2533692722371965,
      "eval_accuracy": 0.3850733314637793,
      "eval_loss": 3.378077983856201,
      "eval_runtime": 183.3483,
      "eval_samples_per_second": 98.234,
      "eval_steps_per_second": 6.141,
      "step": 58000
    },
    {
      "epoch": 6.258760107816712,
      "grad_norm": 0.7233048677444458,
      "learning_rate": 0.00022486562331354556,
      "loss": 3.2652,
      "step": 58050
    },
    {
      "epoch": 6.264150943396227,
      "grad_norm": 0.7433217167854309,
      "learning_rate": 0.00022454182406907714,
      "loss": 3.2765,
      "step": 58100
    },
    {
      "epoch": 6.269541778975741,
      "grad_norm": 0.7410855889320374,
      "learning_rate": 0.00022421802482460872,
      "loss": 3.2542,
      "step": 58150
    },
    {
      "epoch": 6.274932614555256,
      "grad_norm": 0.7226539850234985,
      "learning_rate": 0.00022390070156502967,
      "loss": 3.2707,
      "step": 58200
    },
    {
      "epoch": 6.280323450134771,
      "grad_norm": 0.7170883417129517,
      "learning_rate": 0.00022357690232056125,
      "loss": 3.2763,
      "step": 58250
    },
    {
      "epoch": 6.285714285714286,
      "grad_norm": 0.7725226879119873,
      "learning_rate": 0.0002232531030760928,
      "loss": 3.2611,
      "step": 58300
    },
    {
      "epoch": 6.291105121293801,
      "grad_norm": 0.766638457775116,
      "learning_rate": 0.00022292930383162435,
      "loss": 3.2514,
      "step": 58350
    },
    {
      "epoch": 6.296495956873315,
      "grad_norm": 0.7184855341911316,
      "learning_rate": 0.00022260550458715593,
      "loss": 3.2592,
      "step": 58400
    },
    {
      "epoch": 6.30188679245283,
      "grad_norm": 0.7369925379753113,
      "learning_rate": 0.0002222817053426875,
      "loss": 3.2782,
      "step": 58450
    },
    {
      "epoch": 6.307277628032345,
      "grad_norm": 0.6875489354133606,
      "learning_rate": 0.00022195790609821908,
      "loss": 3.2501,
      "step": 58500
    },
    {
      "epoch": 6.3126684636118595,
      "grad_norm": 0.7300560474395752,
      "learning_rate": 0.00022163410685375066,
      "loss": 3.2723,
      "step": 58550
    },
    {
      "epoch": 6.318059299191375,
      "grad_norm": 0.7411653995513916,
      "learning_rate": 0.00022131030760928224,
      "loss": 3.2799,
      "step": 58600
    },
    {
      "epoch": 6.32345013477089,
      "grad_norm": 0.7001060843467712,
      "learning_rate": 0.00022098650836481382,
      "loss": 3.2727,
      "step": 58650
    },
    {
      "epoch": 6.328840970350404,
      "grad_norm": 0.7271462678909302,
      "learning_rate": 0.0002206627091203454,
      "loss": 3.2883,
      "step": 58700
    },
    {
      "epoch": 6.334231805929919,
      "grad_norm": 0.7320724129676819,
      "learning_rate": 0.00022033890987587692,
      "loss": 3.2852,
      "step": 58750
    },
    {
      "epoch": 6.339622641509434,
      "grad_norm": 0.7684124112129211,
      "learning_rate": 0.0002200151106314085,
      "loss": 3.2706,
      "step": 58800
    },
    {
      "epoch": 6.345013477088949,
      "grad_norm": 0.7218632102012634,
      "learning_rate": 0.00021969131138694007,
      "loss": 3.2767,
      "step": 58850
    },
    {
      "epoch": 6.350404312668464,
      "grad_norm": 0.7296848297119141,
      "learning_rate": 0.00021936751214247165,
      "loss": 3.2619,
      "step": 58900
    },
    {
      "epoch": 6.355795148247978,
      "grad_norm": 0.7030173540115356,
      "learning_rate": 0.00021904371289800323,
      "loss": 3.2828,
      "step": 58950
    },
    {
      "epoch": 6.361185983827493,
      "grad_norm": 0.6917963624000549,
      "learning_rate": 0.0002187199136535348,
      "loss": 3.2807,
      "step": 59000
    },
    {
      "epoch": 6.361185983827493,
      "eval_accuracy": 0.38561768243172984,
      "eval_loss": 3.3727059364318848,
      "eval_runtime": 183.1739,
      "eval_samples_per_second": 98.327,
      "eval_steps_per_second": 6.147,
      "step": 59000
    },
    {
      "epoch": 6.366576819407008,
      "grad_norm": 0.7581689357757568,
      "learning_rate": 0.00021839611440906638,
      "loss": 3.2838,
      "step": 59050
    },
    {
      "epoch": 6.3719676549865225,
      "grad_norm": 0.7609435319900513,
      "learning_rate": 0.00021807231516459793,
      "loss": 3.2745,
      "step": 59100
    },
    {
      "epoch": 6.377358490566038,
      "grad_norm": 0.747968316078186,
      "learning_rate": 0.00021774851592012948,
      "loss": 3.2765,
      "step": 59150
    },
    {
      "epoch": 6.382749326145553,
      "grad_norm": 0.733305037021637,
      "learning_rate": 0.00021742471667566106,
      "loss": 3.2944,
      "step": 59200
    },
    {
      "epoch": 6.388140161725067,
      "grad_norm": 0.7445916533470154,
      "learning_rate": 0.00021710091743119264,
      "loss": 3.2692,
      "step": 59250
    },
    {
      "epoch": 6.393530997304582,
      "grad_norm": 0.7506402730941772,
      "learning_rate": 0.00021677711818672422,
      "loss": 3.2604,
      "step": 59300
    },
    {
      "epoch": 6.398921832884097,
      "grad_norm": 0.7963356971740723,
      "learning_rate": 0.0002164533189422558,
      "loss": 3.2751,
      "step": 59350
    },
    {
      "epoch": 6.404312668463612,
      "grad_norm": 0.7167596817016602,
      "learning_rate": 0.00021612951969778734,
      "loss": 3.2961,
      "step": 59400
    },
    {
      "epoch": 6.409703504043127,
      "grad_norm": 0.7183796763420105,
      "learning_rate": 0.00021580572045331892,
      "loss": 3.2764,
      "step": 59450
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.6938195824623108,
      "learning_rate": 0.0002154819212088505,
      "loss": 3.2624,
      "step": 59500
    },
    {
      "epoch": 6.420485175202156,
      "grad_norm": 0.702420175075531,
      "learning_rate": 0.00021515812196438208,
      "loss": 3.2971,
      "step": 59550
    },
    {
      "epoch": 6.425876010781671,
      "grad_norm": 0.7254041433334351,
      "learning_rate": 0.00021483432271991363,
      "loss": 3.2668,
      "step": 59600
    },
    {
      "epoch": 6.431266846361186,
      "grad_norm": 0.7375361323356628,
      "learning_rate": 0.0002145105234754452,
      "loss": 3.2671,
      "step": 59650
    },
    {
      "epoch": 6.436657681940701,
      "grad_norm": 0.7460655570030212,
      "learning_rate": 0.00021418672423097676,
      "loss": 3.305,
      "step": 59700
    },
    {
      "epoch": 6.442048517520216,
      "grad_norm": 0.7111104130744934,
      "learning_rate": 0.00021386292498650833,
      "loss": 3.286,
      "step": 59750
    },
    {
      "epoch": 6.44743935309973,
      "grad_norm": 0.7258356809616089,
      "learning_rate": 0.0002135391257420399,
      "loss": 3.2734,
      "step": 59800
    },
    {
      "epoch": 6.452830188679245,
      "grad_norm": 0.7276678085327148,
      "learning_rate": 0.0002132153264975715,
      "loss": 3.2684,
      "step": 59850
    },
    {
      "epoch": 6.45822102425876,
      "grad_norm": 0.7468143701553345,
      "learning_rate": 0.00021289152725310307,
      "loss": 3.2648,
      "step": 59900
    },
    {
      "epoch": 6.463611859838275,
      "grad_norm": 0.7350443005561829,
      "learning_rate": 0.00021256772800863464,
      "loss": 3.2631,
      "step": 59950
    },
    {
      "epoch": 6.46900269541779,
      "grad_norm": 0.6993089318275452,
      "learning_rate": 0.0002122439287641662,
      "loss": 3.2817,
      "step": 60000
    },
    {
      "epoch": 6.46900269541779,
      "eval_accuracy": 0.38607826202317547,
      "eval_loss": 3.366489887237549,
      "eval_runtime": 183.7834,
      "eval_samples_per_second": 98.001,
      "eval_steps_per_second": 6.127,
      "step": 60000
    },
    {
      "epoch": 6.474393530997305,
      "grad_norm": 0.7127420902252197,
      "learning_rate": 0.00021192012951969775,
      "loss": 3.2576,
      "step": 60050
    },
    {
      "epoch": 6.479784366576819,
      "grad_norm": 0.7203495502471924,
      "learning_rate": 0.00021159633027522932,
      "loss": 3.2822,
      "step": 60100
    },
    {
      "epoch": 6.485175202156334,
      "grad_norm": 0.7426835298538208,
      "learning_rate": 0.0002112725310307609,
      "loss": 3.2841,
      "step": 60150
    },
    {
      "epoch": 6.490566037735849,
      "grad_norm": 0.7064744234085083,
      "learning_rate": 0.00021094873178629248,
      "loss": 3.2526,
      "step": 60200
    },
    {
      "epoch": 6.495956873315364,
      "grad_norm": 0.7300220131874084,
      "learning_rate": 0.00021063140852671343,
      "loss": 3.2619,
      "step": 60250
    },
    {
      "epoch": 6.501347708894879,
      "grad_norm": 0.7384023666381836,
      "learning_rate": 0.000210307609282245,
      "loss": 3.297,
      "step": 60300
    },
    {
      "epoch": 6.506738544474393,
      "grad_norm": 0.757164478302002,
      "learning_rate": 0.00020998381003777653,
      "loss": 3.2567,
      "step": 60350
    },
    {
      "epoch": 6.512129380053908,
      "grad_norm": 0.7513059377670288,
      "learning_rate": 0.0002096600107933081,
      "loss": 3.2715,
      "step": 60400
    },
    {
      "epoch": 6.517520215633423,
      "grad_norm": 0.7424511313438416,
      "learning_rate": 0.0002093362115488397,
      "loss": 3.275,
      "step": 60450
    },
    {
      "epoch": 6.5229110512129385,
      "grad_norm": 0.704623281955719,
      "learning_rate": 0.00020901888828926064,
      "loss": 3.28,
      "step": 60500
    },
    {
      "epoch": 6.528301886792453,
      "grad_norm": 0.7265028953552246,
      "learning_rate": 0.00020869508904479222,
      "loss": 3.2796,
      "step": 60550
    },
    {
      "epoch": 6.533692722371968,
      "grad_norm": 0.7257518172264099,
      "learning_rate": 0.0002083712898003238,
      "loss": 3.2759,
      "step": 60600
    },
    {
      "epoch": 6.539083557951482,
      "grad_norm": 0.7516768574714661,
      "learning_rate": 0.00020804749055585537,
      "loss": 3.2809,
      "step": 60650
    },
    {
      "epoch": 6.544474393530997,
      "grad_norm": 0.7546630501747131,
      "learning_rate": 0.0002077236913113869,
      "loss": 3.2721,
      "step": 60700
    },
    {
      "epoch": 6.549865229110512,
      "grad_norm": 0.8262805342674255,
      "learning_rate": 0.00020739989206691847,
      "loss": 3.2756,
      "step": 60750
    },
    {
      "epoch": 6.555256064690027,
      "grad_norm": 0.7627807259559631,
      "learning_rate": 0.00020707609282245005,
      "loss": 3.2692,
      "step": 60800
    },
    {
      "epoch": 6.560646900269542,
      "grad_norm": 0.7664798498153687,
      "learning_rate": 0.00020675229357798163,
      "loss": 3.2687,
      "step": 60850
    },
    {
      "epoch": 6.566037735849057,
      "grad_norm": 0.75236576795578,
      "learning_rate": 0.0002064284943335132,
      "loss": 3.266,
      "step": 60900
    },
    {
      "epoch": 6.571428571428571,
      "grad_norm": 0.7668713331222534,
      "learning_rate": 0.00020610469508904478,
      "loss": 3.2746,
      "step": 60950
    },
    {
      "epoch": 6.576819407008086,
      "grad_norm": 0.6870357990264893,
      "learning_rate": 0.00020578089584457636,
      "loss": 3.2753,
      "step": 61000
    },
    {
      "epoch": 6.576819407008086,
      "eval_accuracy": 0.38637314596070393,
      "eval_loss": 3.363420248031616,
      "eval_runtime": 183.2196,
      "eval_samples_per_second": 98.303,
      "eval_steps_per_second": 6.146,
      "step": 61000
    },
    {
      "epoch": 6.5822102425876015,
      "grad_norm": 0.7615857720375061,
      "learning_rate": 0.00020545709660010794,
      "loss": 3.2695,
      "step": 61050
    },
    {
      "epoch": 6.587601078167116,
      "grad_norm": 0.7549526691436768,
      "learning_rate": 0.00020513329735563946,
      "loss": 3.288,
      "step": 61100
    },
    {
      "epoch": 6.592991913746631,
      "grad_norm": 0.76313716173172,
      "learning_rate": 0.00020480949811117104,
      "loss": 3.2877,
      "step": 61150
    },
    {
      "epoch": 6.598382749326145,
      "grad_norm": 0.7764831185340881,
      "learning_rate": 0.00020448569886670262,
      "loss": 3.2824,
      "step": 61200
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.7454062104225159,
      "learning_rate": 0.0002041618996222342,
      "loss": 3.2825,
      "step": 61250
    },
    {
      "epoch": 6.609164420485175,
      "grad_norm": 0.7208166718482971,
      "learning_rate": 0.00020383810037776577,
      "loss": 3.2659,
      "step": 61300
    },
    {
      "epoch": 6.6145552560646905,
      "grad_norm": 0.7168710827827454,
      "learning_rate": 0.00020351430113329735,
      "loss": 3.2976,
      "step": 61350
    },
    {
      "epoch": 6.619946091644205,
      "grad_norm": 0.7853072285652161,
      "learning_rate": 0.00020319050188882893,
      "loss": 3.2847,
      "step": 61400
    },
    {
      "epoch": 6.62533692722372,
      "grad_norm": 0.7360592484474182,
      "learning_rate": 0.00020286670264436048,
      "loss": 3.2946,
      "step": 61450
    },
    {
      "epoch": 6.630727762803234,
      "grad_norm": 0.7565436363220215,
      "learning_rate": 0.00020254290339989206,
      "loss": 3.281,
      "step": 61500
    },
    {
      "epoch": 6.636118598382749,
      "grad_norm": 0.7261871099472046,
      "learning_rate": 0.0002022191041554236,
      "loss": 3.2848,
      "step": 61550
    },
    {
      "epoch": 6.6415094339622645,
      "grad_norm": 0.7736920714378357,
      "learning_rate": 0.00020189530491095519,
      "loss": 3.2826,
      "step": 61600
    },
    {
      "epoch": 6.646900269541779,
      "grad_norm": 0.74223393201828,
      "learning_rate": 0.00020157150566648676,
      "loss": 3.28,
      "step": 61650
    },
    {
      "epoch": 6.652291105121294,
      "grad_norm": 0.777579128742218,
      "learning_rate": 0.00020124770642201834,
      "loss": 3.2633,
      "step": 61700
    },
    {
      "epoch": 6.657681940700809,
      "grad_norm": 0.764687180519104,
      "learning_rate": 0.0002009239071775499,
      "loss": 3.2872,
      "step": 61750
    },
    {
      "epoch": 6.663072776280323,
      "grad_norm": 0.7332888245582581,
      "learning_rate": 0.00020060010793308147,
      "loss": 3.2934,
      "step": 61800
    },
    {
      "epoch": 6.668463611859838,
      "grad_norm": 0.8394750356674194,
      "learning_rate": 0.00020027630868861305,
      "loss": 3.2917,
      "step": 61850
    },
    {
      "epoch": 6.6738544474393535,
      "grad_norm": 0.7316187620162964,
      "learning_rate": 0.00019995250944414462,
      "loss": 3.2639,
      "step": 61900
    },
    {
      "epoch": 6.679245283018868,
      "grad_norm": 0.7434121370315552,
      "learning_rate": 0.00019962871019967617,
      "loss": 3.2656,
      "step": 61950
    },
    {
      "epoch": 6.684636118598383,
      "grad_norm": 0.8291236162185669,
      "learning_rate": 0.00019930491095520775,
      "loss": 3.2679,
      "step": 62000
    },
    {
      "epoch": 6.684636118598383,
      "eval_accuracy": 0.3871309998532099,
      "eval_loss": 3.358170986175537,
      "eval_runtime": 183.5019,
      "eval_samples_per_second": 98.152,
      "eval_steps_per_second": 6.136,
      "step": 62000
    },
    {
      "epoch": 6.690026954177897,
      "grad_norm": 0.7411676645278931,
      "learning_rate": 0.0001989811117107393,
      "loss": 3.2909,
      "step": 62050
    },
    {
      "epoch": 6.695417789757412,
      "grad_norm": 0.7574923038482666,
      "learning_rate": 0.00019865731246627088,
      "loss": 3.2834,
      "step": 62100
    },
    {
      "epoch": 6.7008086253369274,
      "grad_norm": 0.7495836615562439,
      "learning_rate": 0.00019833351322180246,
      "loss": 3.2719,
      "step": 62150
    },
    {
      "epoch": 6.706199460916442,
      "grad_norm": 0.7210440635681152,
      "learning_rate": 0.00019800971397733404,
      "loss": 3.2732,
      "step": 62200
    },
    {
      "epoch": 6.711590296495957,
      "grad_norm": 0.7707087397575378,
      "learning_rate": 0.00019768591473286561,
      "loss": 3.2867,
      "step": 62250
    },
    {
      "epoch": 6.716981132075472,
      "grad_norm": 0.7962491512298584,
      "learning_rate": 0.0001973621154883972,
      "loss": 3.2703,
      "step": 62300
    },
    {
      "epoch": 6.722371967654986,
      "grad_norm": 0.7390112280845642,
      "learning_rate": 0.00019703831624392877,
      "loss": 3.2883,
      "step": 62350
    },
    {
      "epoch": 6.727762803234501,
      "grad_norm": 0.8107832670211792,
      "learning_rate": 0.0001967145169994603,
      "loss": 3.271,
      "step": 62400
    },
    {
      "epoch": 6.7331536388140165,
      "grad_norm": 0.741752028465271,
      "learning_rate": 0.00019639071775499187,
      "loss": 3.2754,
      "step": 62450
    },
    {
      "epoch": 6.738544474393531,
      "grad_norm": 0.7602443695068359,
      "learning_rate": 0.00019606691851052345,
      "loss": 3.2785,
      "step": 62500
    },
    {
      "epoch": 6.743935309973046,
      "grad_norm": 0.716080904006958,
      "learning_rate": 0.00019574311926605503,
      "loss": 3.2864,
      "step": 62550
    },
    {
      "epoch": 6.74932614555256,
      "grad_norm": 0.7163423299789429,
      "learning_rate": 0.0001954193200215866,
      "loss": 3.269,
      "step": 62600
    },
    {
      "epoch": 6.754716981132075,
      "grad_norm": 0.7376225590705872,
      "learning_rate": 0.00019509552077711818,
      "loss": 3.274,
      "step": 62650
    },
    {
      "epoch": 6.7601078167115904,
      "grad_norm": 0.7619684338569641,
      "learning_rate": 0.00019477172153264976,
      "loss": 3.2638,
      "step": 62700
    },
    {
      "epoch": 6.765498652291106,
      "grad_norm": 0.7855158448219299,
      "learning_rate": 0.00019444792228818134,
      "loss": 3.2921,
      "step": 62750
    },
    {
      "epoch": 6.77088948787062,
      "grad_norm": 0.7887711524963379,
      "learning_rate": 0.00019412412304371286,
      "loss": 3.2783,
      "step": 62800
    },
    {
      "epoch": 6.776280323450135,
      "grad_norm": 0.7664616107940674,
      "learning_rate": 0.00019380032379924444,
      "loss": 3.2934,
      "step": 62850
    },
    {
      "epoch": 6.781671159029649,
      "grad_norm": 0.7111510634422302,
      "learning_rate": 0.00019347652455477602,
      "loss": 3.2909,
      "step": 62900
    },
    {
      "epoch": 6.787061994609164,
      "grad_norm": 0.7316784262657166,
      "learning_rate": 0.0001931527253103076,
      "loss": 3.2634,
      "step": 62950
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.7745811939239502,
      "learning_rate": 0.00019282892606583917,
      "loss": 3.2727,
      "step": 63000
    },
    {
      "epoch": 6.7924528301886795,
      "eval_accuracy": 0.3873692756361871,
      "eval_loss": 3.3546576499938965,
      "eval_runtime": 183.7912,
      "eval_samples_per_second": 97.997,
      "eval_steps_per_second": 6.127,
      "step": 63000
    },
    {
      "epoch": 6.797843665768194,
      "grad_norm": 0.7474355101585388,
      "learning_rate": 0.00019250512682137075,
      "loss": 3.2738,
      "step": 63050
    },
    {
      "epoch": 6.803234501347709,
      "grad_norm": 0.7750080823898315,
      "learning_rate": 0.0001921813275769023,
      "loss": 3.2733,
      "step": 63100
    },
    {
      "epoch": 6.808625336927224,
      "grad_norm": 0.7421109676361084,
      "learning_rate": 0.00019185752833243388,
      "loss": 3.2766,
      "step": 63150
    },
    {
      "epoch": 6.814016172506738,
      "grad_norm": 0.7296343445777893,
      "learning_rate": 0.00019153372908796545,
      "loss": 3.2777,
      "step": 63200
    },
    {
      "epoch": 6.819407008086253,
      "grad_norm": 0.7462671995162964,
      "learning_rate": 0.000191209929843497,
      "loss": 3.27,
      "step": 63250
    },
    {
      "epoch": 6.824797843665769,
      "grad_norm": 0.7307702898979187,
      "learning_rate": 0.00019088613059902858,
      "loss": 3.2594,
      "step": 63300
    },
    {
      "epoch": 6.830188679245283,
      "grad_norm": 0.751537561416626,
      "learning_rate": 0.00019056233135456016,
      "loss": 3.2779,
      "step": 63350
    },
    {
      "epoch": 6.835579514824798,
      "grad_norm": 0.7643833756446838,
      "learning_rate": 0.0001902385321100917,
      "loss": 3.2702,
      "step": 63400
    },
    {
      "epoch": 6.840970350404312,
      "grad_norm": 0.7389995455741882,
      "learning_rate": 0.0001899147328656233,
      "loss": 3.2886,
      "step": 63450
    },
    {
      "epoch": 6.846361185983827,
      "grad_norm": 0.7349757552146912,
      "learning_rate": 0.00018959093362115487,
      "loss": 3.2749,
      "step": 63500
    },
    {
      "epoch": 6.8517520215633425,
      "grad_norm": 0.7448405623435974,
      "learning_rate": 0.00018926713437668644,
      "loss": 3.2778,
      "step": 63550
    },
    {
      "epoch": 6.857142857142857,
      "grad_norm": 0.7943175435066223,
      "learning_rate": 0.00018894333513221802,
      "loss": 3.271,
      "step": 63600
    },
    {
      "epoch": 6.862533692722372,
      "grad_norm": 0.7844904661178589,
      "learning_rate": 0.00018861953588774957,
      "loss": 3.2804,
      "step": 63650
    },
    {
      "epoch": 6.867924528301887,
      "grad_norm": 0.7307280898094177,
      "learning_rate": 0.00018829573664328115,
      "loss": 3.2785,
      "step": 63700
    },
    {
      "epoch": 6.873315363881401,
      "grad_norm": 0.7532712817192078,
      "learning_rate": 0.0001879719373988127,
      "loss": 3.2801,
      "step": 63750
    },
    {
      "epoch": 6.878706199460916,
      "grad_norm": 0.7567185759544373,
      "learning_rate": 0.00018764813815434428,
      "loss": 3.2661,
      "step": 63800
    },
    {
      "epoch": 6.884097035040432,
      "grad_norm": 0.7488755583763123,
      "learning_rate": 0.00018732433890987586,
      "loss": 3.2661,
      "step": 63850
    },
    {
      "epoch": 6.889487870619946,
      "grad_norm": 0.7576258778572083,
      "learning_rate": 0.00018700053966540743,
      "loss": 3.2932,
      "step": 63900
    },
    {
      "epoch": 6.894878706199461,
      "grad_norm": 0.7660182118415833,
      "learning_rate": 0.000186676740420939,
      "loss": 3.2827,
      "step": 63950
    },
    {
      "epoch": 6.900269541778976,
      "grad_norm": 0.7318092584609985,
      "learning_rate": 0.0001863529411764706,
      "loss": 3.2896,
      "step": 64000
    },
    {
      "epoch": 6.900269541778976,
      "eval_accuracy": 0.3878326802027159,
      "eval_loss": 3.352012872695923,
      "eval_runtime": 183.2112,
      "eval_samples_per_second": 98.307,
      "eval_steps_per_second": 6.146,
      "step": 64000
    },
    {
      "epoch": 6.90566037735849,
      "grad_norm": 0.7451673150062561,
      "learning_rate": 0.00018603561791689149,
      "loss": 3.273,
      "step": 64050
    },
    {
      "epoch": 6.9110512129380055,
      "grad_norm": 0.7861278653144836,
      "learning_rate": 0.00018571181867242306,
      "loss": 3.2591,
      "step": 64100
    },
    {
      "epoch": 6.916442048517521,
      "grad_norm": 0.7036826610565186,
      "learning_rate": 0.00018538801942795464,
      "loss": 3.2688,
      "step": 64150
    },
    {
      "epoch": 6.921832884097035,
      "grad_norm": 0.7634555697441101,
      "learning_rate": 0.00018506422018348622,
      "loss": 3.2769,
      "step": 64200
    },
    {
      "epoch": 6.92722371967655,
      "grad_norm": 0.7458739280700684,
      "learning_rate": 0.0001847404209390178,
      "loss": 3.2709,
      "step": 64250
    },
    {
      "epoch": 6.932614555256064,
      "grad_norm": 0.8086719512939453,
      "learning_rate": 0.00018441662169454937,
      "loss": 3.2676,
      "step": 64300
    },
    {
      "epoch": 6.938005390835579,
      "grad_norm": 0.716335654258728,
      "learning_rate": 0.00018409282245008095,
      "loss": 3.2789,
      "step": 64350
    },
    {
      "epoch": 6.943396226415095,
      "grad_norm": 0.7624198794364929,
      "learning_rate": 0.00018376902320561253,
      "loss": 3.2732,
      "step": 64400
    },
    {
      "epoch": 6.948787061994609,
      "grad_norm": 0.77959144115448,
      "learning_rate": 0.00018344522396114405,
      "loss": 3.2667,
      "step": 64450
    },
    {
      "epoch": 6.954177897574124,
      "grad_norm": 0.766917884349823,
      "learning_rate": 0.00018312142471667563,
      "loss": 3.2833,
      "step": 64500
    },
    {
      "epoch": 6.959568733153639,
      "grad_norm": 0.7285516858100891,
      "learning_rate": 0.0001827976254722072,
      "loss": 3.275,
      "step": 64550
    },
    {
      "epoch": 6.964959568733153,
      "grad_norm": 0.7674565315246582,
      "learning_rate": 0.00018247382622773879,
      "loss": 3.2768,
      "step": 64600
    },
    {
      "epoch": 6.9703504043126685,
      "grad_norm": 0.7331922054290771,
      "learning_rate": 0.00018215002698327036,
      "loss": 3.2847,
      "step": 64650
    },
    {
      "epoch": 6.975741239892184,
      "grad_norm": 0.7738147377967834,
      "learning_rate": 0.00018182622773880194,
      "loss": 3.2892,
      "step": 64700
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.7373318672180176,
      "learning_rate": 0.00018150242849433352,
      "loss": 3.2808,
      "step": 64750
    },
    {
      "epoch": 6.986522911051213,
      "grad_norm": 0.7882824540138245,
      "learning_rate": 0.00018117862924986507,
      "loss": 3.2797,
      "step": 64800
    },
    {
      "epoch": 6.991913746630727,
      "grad_norm": 0.718296468257904,
      "learning_rate": 0.00018085483000539662,
      "loss": 3.2775,
      "step": 64850
    },
    {
      "epoch": 6.997304582210242,
      "grad_norm": 0.7915265560150146,
      "learning_rate": 0.0001805310307609282,
      "loss": 3.2826,
      "step": 64900
    },
    {
      "epoch": 7.002695417789758,
      "grad_norm": 0.7813780903816223,
      "learning_rate": 0.00018020723151645978,
      "loss": 3.228,
      "step": 64950
    },
    {
      "epoch": 7.008086253369272,
      "grad_norm": 0.8058916926383972,
      "learning_rate": 0.00017988343227199135,
      "loss": 3.1934,
      "step": 65000
    },
    {
      "epoch": 7.008086253369272,
      "eval_accuracy": 0.387832028285389,
      "eval_loss": 3.354559898376465,
      "eval_runtime": 183.3424,
      "eval_samples_per_second": 98.237,
      "eval_steps_per_second": 6.142,
      "step": 65000
    },
    {
      "epoch": 7.013477088948787,
      "grad_norm": 0.7725360989570618,
      "learning_rate": 0.00017955963302752293,
      "loss": 3.1876,
      "step": 65050
    },
    {
      "epoch": 7.018867924528302,
      "grad_norm": 0.7770508527755737,
      "learning_rate": 0.00017923583378305448,
      "loss": 3.1881,
      "step": 65100
    },
    {
      "epoch": 7.024258760107816,
      "grad_norm": 0.8097355961799622,
      "learning_rate": 0.00017891203453858606,
      "loss": 3.1979,
      "step": 65150
    },
    {
      "epoch": 7.0296495956873315,
      "grad_norm": 0.7231821417808533,
      "learning_rate": 0.00017858823529411764,
      "loss": 3.175,
      "step": 65200
    },
    {
      "epoch": 7.035040431266847,
      "grad_norm": 0.8374466300010681,
      "learning_rate": 0.00017826443604964921,
      "loss": 3.1885,
      "step": 65250
    },
    {
      "epoch": 7.040431266846361,
      "grad_norm": 0.8206698298454285,
      "learning_rate": 0.00017794063680518077,
      "loss": 3.1949,
      "step": 65300
    },
    {
      "epoch": 7.045822102425876,
      "grad_norm": 0.735885500907898,
      "learning_rate": 0.00017761683756071234,
      "loss": 3.1847,
      "step": 65350
    },
    {
      "epoch": 7.051212938005391,
      "grad_norm": 0.8094024062156677,
      "learning_rate": 0.00017729303831624392,
      "loss": 3.2144,
      "step": 65400
    },
    {
      "epoch": 7.056603773584905,
      "grad_norm": 0.7802730798721313,
      "learning_rate": 0.00017696923907177547,
      "loss": 3.2029,
      "step": 65450
    },
    {
      "epoch": 7.061994609164421,
      "grad_norm": 0.7770246863365173,
      "learning_rate": 0.00017664543982730705,
      "loss": 3.1913,
      "step": 65500
    },
    {
      "epoch": 7.067385444743936,
      "grad_norm": 0.7353746891021729,
      "learning_rate": 0.00017632164058283863,
      "loss": 3.2044,
      "step": 65550
    },
    {
      "epoch": 7.07277628032345,
      "grad_norm": 0.7481196522712708,
      "learning_rate": 0.0001759978413383702,
      "loss": 3.1822,
      "step": 65600
    },
    {
      "epoch": 7.078167115902965,
      "grad_norm": 0.7949692010879517,
      "learning_rate": 0.00017567404209390178,
      "loss": 3.2052,
      "step": 65650
    },
    {
      "epoch": 7.083557951482479,
      "grad_norm": 0.7633875012397766,
      "learning_rate": 0.00017535024284943333,
      "loss": 3.2092,
      "step": 65700
    },
    {
      "epoch": 7.0889487870619945,
      "grad_norm": 0.8302983641624451,
      "learning_rate": 0.00017502644360496488,
      "loss": 3.2087,
      "step": 65750
    },
    {
      "epoch": 7.09433962264151,
      "grad_norm": 0.794335126876831,
      "learning_rate": 0.00017470264436049646,
      "loss": 3.2061,
      "step": 65800
    },
    {
      "epoch": 7.099730458221024,
      "grad_norm": 0.7204697132110596,
      "learning_rate": 0.00017437884511602804,
      "loss": 3.1899,
      "step": 65850
    },
    {
      "epoch": 7.105121293800539,
      "grad_norm": 0.7534080743789673,
      "learning_rate": 0.00017405504587155962,
      "loss": 3.2178,
      "step": 65900
    },
    {
      "epoch": 7.110512129380054,
      "grad_norm": 0.7708962559700012,
      "learning_rate": 0.0001737312466270912,
      "loss": 3.1813,
      "step": 65950
    },
    {
      "epoch": 7.115902964959568,
      "grad_norm": 0.7770982384681702,
      "learning_rate": 0.00017340744738262277,
      "loss": 3.2125,
      "step": 66000
    },
    {
      "epoch": 7.115902964959568,
      "eval_accuracy": 0.3879341619999346,
      "eval_loss": 3.3549797534942627,
      "eval_runtime": 183.6028,
      "eval_samples_per_second": 98.098,
      "eval_steps_per_second": 6.133,
      "step": 66000
    },
    {
      "epoch": 7.121293800539084,
      "grad_norm": 0.8400378823280334,
      "learning_rate": 0.00017308364813815435,
      "loss": 3.204,
      "step": 66050
    },
    {
      "epoch": 7.126684636118599,
      "grad_norm": 0.783427357673645,
      "learning_rate": 0.00017275984889368593,
      "loss": 3.1943,
      "step": 66100
    },
    {
      "epoch": 7.132075471698113,
      "grad_norm": 0.7737600207328796,
      "learning_rate": 0.00017243604964921745,
      "loss": 3.2306,
      "step": 66150
    },
    {
      "epoch": 7.137466307277628,
      "grad_norm": 0.7946293354034424,
      "learning_rate": 0.00017211225040474903,
      "loss": 3.2256,
      "step": 66200
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 0.797230064868927,
      "learning_rate": 0.0001717884511602806,
      "loss": 3.2133,
      "step": 66250
    },
    {
      "epoch": 7.1482479784366575,
      "grad_norm": 0.817846953868866,
      "learning_rate": 0.00017146465191581218,
      "loss": 3.2025,
      "step": 66300
    },
    {
      "epoch": 7.153638814016173,
      "grad_norm": 0.786282479763031,
      "learning_rate": 0.00017114085267134376,
      "loss": 3.21,
      "step": 66350
    },
    {
      "epoch": 7.159029649595688,
      "grad_norm": 0.7998514771461487,
      "learning_rate": 0.00017081705342687534,
      "loss": 3.2079,
      "step": 66400
    },
    {
      "epoch": 7.164420485175202,
      "grad_norm": 0.7554528117179871,
      "learning_rate": 0.0001704932541824069,
      "loss": 3.2061,
      "step": 66450
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.7405669093132019,
      "learning_rate": 0.00017016945493793847,
      "loss": 3.2116,
      "step": 66500
    },
    {
      "epoch": 7.175202156334231,
      "grad_norm": 0.7673985958099365,
      "learning_rate": 0.00016984565569347002,
      "loss": 3.2082,
      "step": 66550
    },
    {
      "epoch": 7.180592991913747,
      "grad_norm": 0.7783784866333008,
      "learning_rate": 0.0001695218564490016,
      "loss": 3.2171,
      "step": 66600
    },
    {
      "epoch": 7.185983827493262,
      "grad_norm": 0.7714428901672363,
      "learning_rate": 0.00016919805720453317,
      "loss": 3.2155,
      "step": 66650
    },
    {
      "epoch": 7.191374663072776,
      "grad_norm": 0.7873608469963074,
      "learning_rate": 0.00016887425796006475,
      "loss": 3.2122,
      "step": 66700
    },
    {
      "epoch": 7.196765498652291,
      "grad_norm": 0.8219302892684937,
      "learning_rate": 0.00016855045871559633,
      "loss": 3.1983,
      "step": 66750
    },
    {
      "epoch": 7.202156334231806,
      "grad_norm": 0.802437961101532,
      "learning_rate": 0.00016822665947112788,
      "loss": 3.2066,
      "step": 66800
    },
    {
      "epoch": 7.2075471698113205,
      "grad_norm": 0.7578773498535156,
      "learning_rate": 0.00016790286022665946,
      "loss": 3.2035,
      "step": 66850
    },
    {
      "epoch": 7.212938005390836,
      "grad_norm": 0.7701559662818909,
      "learning_rate": 0.00016757906098219103,
      "loss": 3.2044,
      "step": 66900
    },
    {
      "epoch": 7.218328840970351,
      "grad_norm": 0.8019431829452515,
      "learning_rate": 0.0001672552617377226,
      "loss": 3.207,
      "step": 66950
    },
    {
      "epoch": 7.223719676549865,
      "grad_norm": 0.7901089787483215,
      "learning_rate": 0.00016693146249325416,
      "loss": 3.215,
      "step": 67000
    },
    {
      "epoch": 7.223719676549865,
      "eval_accuracy": 0.3883080365869042,
      "eval_loss": 3.351943016052246,
      "eval_runtime": 183.3094,
      "eval_samples_per_second": 98.255,
      "eval_steps_per_second": 6.143,
      "step": 67000
    },
    {
      "epoch": 7.22911051212938,
      "grad_norm": 0.7946869134902954,
      "learning_rate": 0.00016660766324878574,
      "loss": 3.2258,
      "step": 67050
    },
    {
      "epoch": 7.234501347708895,
      "grad_norm": 0.77195143699646,
      "learning_rate": 0.0001662838640043173,
      "loss": 3.2286,
      "step": 67100
    },
    {
      "epoch": 7.2398921832884096,
      "grad_norm": 0.782701849937439,
      "learning_rate": 0.00016596006475984887,
      "loss": 3.186,
      "step": 67150
    },
    {
      "epoch": 7.245283018867925,
      "grad_norm": 0.756524384021759,
      "learning_rate": 0.00016563626551538045,
      "loss": 3.2187,
      "step": 67200
    },
    {
      "epoch": 7.250673854447439,
      "grad_norm": 0.7699859142303467,
      "learning_rate": 0.00016531246627091202,
      "loss": 3.2101,
      "step": 67250
    },
    {
      "epoch": 7.256064690026954,
      "grad_norm": 0.7793636322021484,
      "learning_rate": 0.0001649886670264436,
      "loss": 3.2176,
      "step": 67300
    },
    {
      "epoch": 7.261455525606469,
      "grad_norm": 0.7925220727920532,
      "learning_rate": 0.00016466486778197518,
      "loss": 3.232,
      "step": 67350
    },
    {
      "epoch": 7.2668463611859835,
      "grad_norm": 0.7722257375717163,
      "learning_rate": 0.0001643410685375067,
      "loss": 3.2253,
      "step": 67400
    },
    {
      "epoch": 7.272237196765499,
      "grad_norm": 0.8224971294403076,
      "learning_rate": 0.00016401726929303828,
      "loss": 3.237,
      "step": 67450
    },
    {
      "epoch": 7.277628032345014,
      "grad_norm": 0.7844073176383972,
      "learning_rate": 0.00016369347004856986,
      "loss": 3.2074,
      "step": 67500
    },
    {
      "epoch": 7.283018867924528,
      "grad_norm": 0.7575783133506775,
      "learning_rate": 0.00016336967080410143,
      "loss": 3.1997,
      "step": 67550
    },
    {
      "epoch": 7.288409703504043,
      "grad_norm": 0.7861126661300659,
      "learning_rate": 0.000163045871559633,
      "loss": 3.2356,
      "step": 67600
    },
    {
      "epoch": 7.293800539083558,
      "grad_norm": 0.8623127341270447,
      "learning_rate": 0.0001627220723151646,
      "loss": 3.2232,
      "step": 67650
    },
    {
      "epoch": 7.2991913746630726,
      "grad_norm": 0.7551688551902771,
      "learning_rate": 0.00016239827307069617,
      "loss": 3.2158,
      "step": 67700
    },
    {
      "epoch": 7.304582210242588,
      "grad_norm": 0.7738710641860962,
      "learning_rate": 0.00016207447382622775,
      "loss": 3.223,
      "step": 67750
    },
    {
      "epoch": 7.309973045822103,
      "grad_norm": 0.7757263779640198,
      "learning_rate": 0.00016175067458175932,
      "loss": 3.2344,
      "step": 67800
    },
    {
      "epoch": 7.315363881401617,
      "grad_norm": 0.7811141014099121,
      "learning_rate": 0.00016142687533729085,
      "loss": 3.219,
      "step": 67850
    },
    {
      "epoch": 7.320754716981132,
      "grad_norm": 0.8079759478569031,
      "learning_rate": 0.00016110307609282242,
      "loss": 3.2148,
      "step": 67900
    },
    {
      "epoch": 7.3261455525606465,
      "grad_norm": 0.8070315718650818,
      "learning_rate": 0.000160779276848354,
      "loss": 3.2236,
      "step": 67950
    },
    {
      "epoch": 7.331536388140162,
      "grad_norm": 0.8166431784629822,
      "learning_rate": 0.00016045547760388558,
      "loss": 3.2193,
      "step": 68000
    },
    {
      "epoch": 7.331536388140162,
      "eval_accuracy": 0.3887239598414581,
      "eval_loss": 3.3503804206848145,
      "eval_runtime": 183.4566,
      "eval_samples_per_second": 98.176,
      "eval_steps_per_second": 6.138,
      "step": 68000
    },
    {
      "epoch": 7.336927223719677,
      "grad_norm": 0.7583974003791809,
      "learning_rate": 0.00016013815434430653,
      "loss": 3.2139,
      "step": 68050
    },
    {
      "epoch": 7.342318059299191,
      "grad_norm": 0.757961630821228,
      "learning_rate": 0.0001598143550998381,
      "loss": 3.2173,
      "step": 68100
    },
    {
      "epoch": 7.347708894878706,
      "grad_norm": 0.7688447833061218,
      "learning_rate": 0.00015949055585536966,
      "loss": 3.2238,
      "step": 68150
    },
    {
      "epoch": 7.353099730458221,
      "grad_norm": 0.7891886830329895,
      "learning_rate": 0.0001591667566109012,
      "loss": 3.2162,
      "step": 68200
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.8004134297370911,
      "learning_rate": 0.0001588429573664328,
      "loss": 3.2212,
      "step": 68250
    },
    {
      "epoch": 7.363881401617251,
      "grad_norm": 0.8025893568992615,
      "learning_rate": 0.00015851915812196437,
      "loss": 3.2367,
      "step": 68300
    },
    {
      "epoch": 7.369272237196766,
      "grad_norm": 0.8140208125114441,
      "learning_rate": 0.00015819535887749594,
      "loss": 3.2145,
      "step": 68350
    },
    {
      "epoch": 7.37466307277628,
      "grad_norm": 0.7988510727882385,
      "learning_rate": 0.00015787155963302752,
      "loss": 3.2117,
      "step": 68400
    },
    {
      "epoch": 7.380053908355795,
      "grad_norm": 0.8174982666969299,
      "learning_rate": 0.0001575477603885591,
      "loss": 3.2268,
      "step": 68450
    },
    {
      "epoch": 7.38544474393531,
      "grad_norm": 0.8240893483161926,
      "learning_rate": 0.00015722396114409065,
      "loss": 3.199,
      "step": 68500
    },
    {
      "epoch": 7.390835579514825,
      "grad_norm": 0.7737301588058472,
      "learning_rate": 0.00015690016189962223,
      "loss": 3.1935,
      "step": 68550
    },
    {
      "epoch": 7.39622641509434,
      "grad_norm": 0.7589561939239502,
      "learning_rate": 0.00015657636265515378,
      "loss": 3.2208,
      "step": 68600
    },
    {
      "epoch": 7.401617250673855,
      "grad_norm": 0.8396160006523132,
      "learning_rate": 0.00015625256341068536,
      "loss": 3.2331,
      "step": 68650
    },
    {
      "epoch": 7.407008086253369,
      "grad_norm": 0.7733933925628662,
      "learning_rate": 0.00015592876416621693,
      "loss": 3.2099,
      "step": 68700
    },
    {
      "epoch": 7.412398921832884,
      "grad_norm": 0.8380154967308044,
      "learning_rate": 0.0001556049649217485,
      "loss": 3.2235,
      "step": 68750
    },
    {
      "epoch": 7.4177897574123985,
      "grad_norm": 0.799769401550293,
      "learning_rate": 0.00015528116567728006,
      "loss": 3.2154,
      "step": 68800
    },
    {
      "epoch": 7.423180592991914,
      "grad_norm": 0.794215977191925,
      "learning_rate": 0.00015495736643281164,
      "loss": 3.2324,
      "step": 68850
    },
    {
      "epoch": 7.428571428571429,
      "grad_norm": 0.8496611714363098,
      "learning_rate": 0.00015463356718834322,
      "loss": 3.2261,
      "step": 68900
    },
    {
      "epoch": 7.433962264150943,
      "grad_norm": 0.7774752378463745,
      "learning_rate": 0.0001543097679438748,
      "loss": 3.2289,
      "step": 68950
    },
    {
      "epoch": 7.439353099730458,
      "grad_norm": 0.8549684286117554,
      "learning_rate": 0.00015398596869940637,
      "loss": 3.2403,
      "step": 69000
    },
    {
      "epoch": 7.439353099730458,
      "eval_accuracy": 0.38884412993538087,
      "eval_loss": 3.345785617828369,
      "eval_runtime": 183.4217,
      "eval_samples_per_second": 98.195,
      "eval_steps_per_second": 6.139,
      "step": 69000
    },
    {
      "epoch": 7.444743935309973,
      "grad_norm": 0.766400933265686,
      "learning_rate": 0.00015366216945493792,
      "loss": 3.2498,
      "step": 69050
    },
    {
      "epoch": 7.450134770889488,
      "grad_norm": 0.7826684713363647,
      "learning_rate": 0.00015333837021046947,
      "loss": 3.2286,
      "step": 69100
    },
    {
      "epoch": 7.455525606469003,
      "grad_norm": 0.7703137993812561,
      "learning_rate": 0.00015301457096600105,
      "loss": 3.2104,
      "step": 69150
    },
    {
      "epoch": 7.460916442048518,
      "grad_norm": 0.7931674718856812,
      "learning_rate": 0.00015269077172153263,
      "loss": 3.2337,
      "step": 69200
    },
    {
      "epoch": 7.466307277628032,
      "grad_norm": 0.8100343346595764,
      "learning_rate": 0.0001523669724770642,
      "loss": 3.2269,
      "step": 69250
    },
    {
      "epoch": 7.471698113207547,
      "grad_norm": 0.7942655682563782,
      "learning_rate": 0.00015204317323259578,
      "loss": 3.2206,
      "step": 69300
    },
    {
      "epoch": 7.4770889487870615,
      "grad_norm": 0.8322016000747681,
      "learning_rate": 0.00015171937398812736,
      "loss": 3.221,
      "step": 69350
    },
    {
      "epoch": 7.482479784366577,
      "grad_norm": 0.7898223400115967,
      "learning_rate": 0.00015139557474365894,
      "loss": 3.248,
      "step": 69400
    },
    {
      "epoch": 7.487870619946092,
      "grad_norm": 0.8208786845207214,
      "learning_rate": 0.00015107177549919046,
      "loss": 3.2311,
      "step": 69450
    },
    {
      "epoch": 7.493261455525606,
      "grad_norm": 0.7814972400665283,
      "learning_rate": 0.00015074797625472204,
      "loss": 3.2196,
      "step": 69500
    },
    {
      "epoch": 7.498652291105121,
      "grad_norm": 0.8326282501220703,
      "learning_rate": 0.00015042417701025362,
      "loss": 3.2347,
      "step": 69550
    },
    {
      "epoch": 7.504043126684636,
      "grad_norm": 0.8042081594467163,
      "learning_rate": 0.0001501003777657852,
      "loss": 3.2242,
      "step": 69600
    },
    {
      "epoch": 7.509433962264151,
      "grad_norm": 0.7709265947341919,
      "learning_rate": 0.00014977657852131677,
      "loss": 3.2092,
      "step": 69650
    },
    {
      "epoch": 7.514824797843666,
      "grad_norm": 0.7810226678848267,
      "learning_rate": 0.0001494592552617377,
      "loss": 3.2464,
      "step": 69700
    },
    {
      "epoch": 7.520215633423181,
      "grad_norm": 0.8230692744255066,
      "learning_rate": 0.00014913545601726928,
      "loss": 3.2242,
      "step": 69750
    },
    {
      "epoch": 7.525606469002695,
      "grad_norm": 0.7870705127716064,
      "learning_rate": 0.00014881165677280085,
      "loss": 3.2335,
      "step": 69800
    },
    {
      "epoch": 7.53099730458221,
      "grad_norm": 0.7646998763084412,
      "learning_rate": 0.00014848785752833243,
      "loss": 3.219,
      "step": 69850
    },
    {
      "epoch": 7.536388140161725,
      "grad_norm": 0.8329770565032959,
      "learning_rate": 0.00014816405828386398,
      "loss": 3.239,
      "step": 69900
    },
    {
      "epoch": 7.54177897574124,
      "grad_norm": 0.8003082275390625,
      "learning_rate": 0.00014784025903939556,
      "loss": 3.2322,
      "step": 69950
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.7998071312904358,
      "learning_rate": 0.00014751645979492714,
      "loss": 3.2254,
      "step": 70000
    },
    {
      "epoch": 7.547169811320755,
      "eval_accuracy": 0.3892735261480237,
      "eval_loss": 3.343639373779297,
      "eval_runtime": 183.6901,
      "eval_samples_per_second": 98.051,
      "eval_steps_per_second": 6.13,
      "step": 70000
    },
    {
      "epoch": 7.55256064690027,
      "grad_norm": 0.7884393930435181,
      "learning_rate": 0.00014719266055045871,
      "loss": 3.2285,
      "step": 70050
    },
    {
      "epoch": 7.557951482479784,
      "grad_norm": 0.8215367197990417,
      "learning_rate": 0.00014686886130599027,
      "loss": 3.238,
      "step": 70100
    },
    {
      "epoch": 7.563342318059299,
      "grad_norm": 0.794144332408905,
      "learning_rate": 0.00014654506206152184,
      "loss": 3.2335,
      "step": 70150
    },
    {
      "epoch": 7.568733153638814,
      "grad_norm": 0.8532745838165283,
      "learning_rate": 0.00014622126281705342,
      "loss": 3.23,
      "step": 70200
    },
    {
      "epoch": 7.574123989218329,
      "grad_norm": 0.8423836827278137,
      "learning_rate": 0.000145897463572585,
      "loss": 3.2318,
      "step": 70250
    },
    {
      "epoch": 7.579514824797844,
      "grad_norm": 0.8306622505187988,
      "learning_rate": 0.00014557366432811658,
      "loss": 3.2334,
      "step": 70300
    },
    {
      "epoch": 7.584905660377358,
      "grad_norm": 0.8016265630722046,
      "learning_rate": 0.00014524986508364813,
      "loss": 3.2495,
      "step": 70350
    },
    {
      "epoch": 7.590296495956873,
      "grad_norm": 0.7804158329963684,
      "learning_rate": 0.0001449260658391797,
      "loss": 3.22,
      "step": 70400
    },
    {
      "epoch": 7.595687331536388,
      "grad_norm": 0.8219479322433472,
      "learning_rate": 0.00014460226659471128,
      "loss": 3.2402,
      "step": 70450
    },
    {
      "epoch": 7.601078167115903,
      "grad_norm": 0.8083851933479309,
      "learning_rate": 0.00014427846735024283,
      "loss": 3.2458,
      "step": 70500
    },
    {
      "epoch": 7.606469002695418,
      "grad_norm": 0.7973158359527588,
      "learning_rate": 0.0001439546681057744,
      "loss": 3.2348,
      "step": 70550
    },
    {
      "epoch": 7.611859838274933,
      "grad_norm": 0.8355487585067749,
      "learning_rate": 0.000143630868861306,
      "loss": 3.2238,
      "step": 70600
    },
    {
      "epoch": 7.617250673854447,
      "grad_norm": 0.8130533695220947,
      "learning_rate": 0.00014330706961683754,
      "loss": 3.2209,
      "step": 70650
    },
    {
      "epoch": 7.622641509433962,
      "grad_norm": 0.824119508266449,
      "learning_rate": 0.00014298327037236912,
      "loss": 3.2475,
      "step": 70700
    },
    {
      "epoch": 7.628032345013477,
      "grad_norm": 0.8470996022224426,
      "learning_rate": 0.0001426594711279007,
      "loss": 3.2272,
      "step": 70750
    },
    {
      "epoch": 7.633423180592992,
      "grad_norm": 0.812836229801178,
      "learning_rate": 0.00014233567188343224,
      "loss": 3.2344,
      "step": 70800
    },
    {
      "epoch": 7.638814016172507,
      "grad_norm": 0.7536956667900085,
      "learning_rate": 0.00014201187263896382,
      "loss": 3.2221,
      "step": 70850
    },
    {
      "epoch": 7.644204851752022,
      "grad_norm": 0.8352275490760803,
      "learning_rate": 0.0001416880733944954,
      "loss": 3.2347,
      "step": 70900
    },
    {
      "epoch": 7.649595687331536,
      "grad_norm": 0.7672180533409119,
      "learning_rate": 0.00014136427415002695,
      "loss": 3.2151,
      "step": 70950
    },
    {
      "epoch": 7.654986522911051,
      "grad_norm": 0.8054909706115723,
      "learning_rate": 0.00014104047490555853,
      "loss": 3.2215,
      "step": 71000
    },
    {
      "epoch": 7.654986522911051,
      "eval_accuracy": 0.38935914462362153,
      "eval_loss": 3.3389949798583984,
      "eval_runtime": 183.3427,
      "eval_samples_per_second": 98.237,
      "eval_steps_per_second": 6.142,
      "step": 71000
    },
    {
      "epoch": 7.660377358490566,
      "grad_norm": 0.8233856558799744,
      "learning_rate": 0.0001407166756610901,
      "loss": 3.2494,
      "step": 71050
    },
    {
      "epoch": 7.665768194070081,
      "grad_norm": 0.8189040422439575,
      "learning_rate": 0.00014039287641662168,
      "loss": 3.2215,
      "step": 71100
    },
    {
      "epoch": 7.671159029649596,
      "grad_norm": 0.8202671408653259,
      "learning_rate": 0.00014006907717215326,
      "loss": 3.2173,
      "step": 71150
    },
    {
      "epoch": 7.67654986522911,
      "grad_norm": 0.8153406977653503,
      "learning_rate": 0.0001397452779276848,
      "loss": 3.223,
      "step": 71200
    },
    {
      "epoch": 7.681940700808625,
      "grad_norm": 0.8444876670837402,
      "learning_rate": 0.0001394214786832164,
      "loss": 3.2304,
      "step": 71250
    },
    {
      "epoch": 7.6873315363881405,
      "grad_norm": 0.836453378200531,
      "learning_rate": 0.00013909767943874797,
      "loss": 3.2242,
      "step": 71300
    },
    {
      "epoch": 7.692722371967655,
      "grad_norm": 0.7985456585884094,
      "learning_rate": 0.00013877388019427954,
      "loss": 3.2336,
      "step": 71350
    },
    {
      "epoch": 7.69811320754717,
      "grad_norm": 0.8078007102012634,
      "learning_rate": 0.0001384500809498111,
      "loss": 3.2358,
      "step": 71400
    },
    {
      "epoch": 7.703504043126685,
      "grad_norm": 0.9007571935653687,
      "learning_rate": 0.00013812628170534267,
      "loss": 3.2288,
      "step": 71450
    },
    {
      "epoch": 7.708894878706199,
      "grad_norm": 0.7986894249916077,
      "learning_rate": 0.00013780248246087425,
      "loss": 3.2364,
      "step": 71500
    },
    {
      "epoch": 7.714285714285714,
      "grad_norm": 0.8218685388565063,
      "learning_rate": 0.00013747868321640583,
      "loss": 3.2235,
      "step": 71550
    },
    {
      "epoch": 7.719676549865229,
      "grad_norm": 0.8080571293830872,
      "learning_rate": 0.00013715488397193738,
      "loss": 3.2249,
      "step": 71600
    },
    {
      "epoch": 7.725067385444744,
      "grad_norm": 0.854092001914978,
      "learning_rate": 0.00013683108472746896,
      "loss": 3.2193,
      "step": 71650
    },
    {
      "epoch": 7.730458221024259,
      "grad_norm": 0.7832936644554138,
      "learning_rate": 0.00013650728548300053,
      "loss": 3.2439,
      "step": 71700
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.779666006565094,
      "learning_rate": 0.0001361834862385321,
      "loss": 3.2301,
      "step": 71750
    },
    {
      "epoch": 7.741239892183288,
      "grad_norm": 0.8854478597640991,
      "learning_rate": 0.0001358596869940637,
      "loss": 3.2211,
      "step": 71800
    },
    {
      "epoch": 7.7466307277628035,
      "grad_norm": 0.7972925901412964,
      "learning_rate": 0.00013553588774959524,
      "loss": 3.2189,
      "step": 71850
    },
    {
      "epoch": 7.752021563342318,
      "grad_norm": 0.853445291519165,
      "learning_rate": 0.00013521208850512682,
      "loss": 3.2392,
      "step": 71900
    },
    {
      "epoch": 7.757412398921833,
      "grad_norm": 0.8733102083206177,
      "learning_rate": 0.0001348882892606584,
      "loss": 3.2315,
      "step": 71950
    },
    {
      "epoch": 7.762803234501348,
      "grad_norm": 0.7951270937919617,
      "learning_rate": 0.00013456449001618995,
      "loss": 3.235,
      "step": 72000
    },
    {
      "epoch": 7.762803234501348,
      "eval_accuracy": 0.39014046753989545,
      "eval_loss": 3.3319358825683594,
      "eval_runtime": 183.3781,
      "eval_samples_per_second": 98.218,
      "eval_steps_per_second": 6.14,
      "step": 72000
    },
    {
      "epoch": 7.768194070080862,
      "grad_norm": 0.7769864201545715,
      "learning_rate": 0.00013424069077172152,
      "loss": 3.2239,
      "step": 72050
    },
    {
      "epoch": 7.773584905660377,
      "grad_norm": 0.7999830842018127,
      "learning_rate": 0.0001339168915272531,
      "loss": 3.2342,
      "step": 72100
    },
    {
      "epoch": 7.7789757412398925,
      "grad_norm": 0.781497597694397,
      "learning_rate": 0.00013359309228278465,
      "loss": 3.2331,
      "step": 72150
    },
    {
      "epoch": 7.784366576819407,
      "grad_norm": 0.8018577098846436,
      "learning_rate": 0.00013326929303831623,
      "loss": 3.2561,
      "step": 72200
    },
    {
      "epoch": 7.789757412398922,
      "grad_norm": 0.8189263939857483,
      "learning_rate": 0.0001329454937938478,
      "loss": 3.211,
      "step": 72250
    },
    {
      "epoch": 7.795148247978437,
      "grad_norm": 0.8497806191444397,
      "learning_rate": 0.00013262169454937936,
      "loss": 3.2363,
      "step": 72300
    },
    {
      "epoch": 7.800539083557951,
      "grad_norm": 0.8176953792572021,
      "learning_rate": 0.00013229789530491093,
      "loss": 3.2281,
      "step": 72350
    },
    {
      "epoch": 7.8059299191374665,
      "grad_norm": 0.8766255378723145,
      "learning_rate": 0.0001319740960604425,
      "loss": 3.2288,
      "step": 72400
    },
    {
      "epoch": 7.811320754716981,
      "grad_norm": 0.8095853924751282,
      "learning_rate": 0.0001316502968159741,
      "loss": 3.2339,
      "step": 72450
    },
    {
      "epoch": 7.816711590296496,
      "grad_norm": 0.8431499600410461,
      "learning_rate": 0.00013132649757150564,
      "loss": 3.2225,
      "step": 72500
    },
    {
      "epoch": 7.822102425876011,
      "grad_norm": 0.8290701508522034,
      "learning_rate": 0.00013100269832703722,
      "loss": 3.2066,
      "step": 72550
    },
    {
      "epoch": 7.827493261455525,
      "grad_norm": 0.780984103679657,
      "learning_rate": 0.0001306788990825688,
      "loss": 3.2393,
      "step": 72600
    },
    {
      "epoch": 7.83288409703504,
      "grad_norm": 0.795058012008667,
      "learning_rate": 0.00013035509983810037,
      "loss": 3.2264,
      "step": 72650
    },
    {
      "epoch": 7.8382749326145555,
      "grad_norm": 0.833223819732666,
      "learning_rate": 0.00013003130059363192,
      "loss": 3.2348,
      "step": 72700
    },
    {
      "epoch": 7.84366576819407,
      "grad_norm": 0.8560569882392883,
      "learning_rate": 0.0001297075013491635,
      "loss": 3.229,
      "step": 72750
    },
    {
      "epoch": 7.849056603773585,
      "grad_norm": 0.8031648397445679,
      "learning_rate": 0.00012938370210469508,
      "loss": 3.2249,
      "step": 72800
    },
    {
      "epoch": 7.8544474393531,
      "grad_norm": 0.8022270202636719,
      "learning_rate": 0.00012905990286022666,
      "loss": 3.2204,
      "step": 72850
    },
    {
      "epoch": 7.859838274932614,
      "grad_norm": 0.7824501991271973,
      "learning_rate": 0.0001287361036157582,
      "loss": 3.2205,
      "step": 72900
    },
    {
      "epoch": 7.8652291105121295,
      "grad_norm": 0.8288241028785706,
      "learning_rate": 0.00012841230437128979,
      "loss": 3.2178,
      "step": 72950
    },
    {
      "epoch": 7.870619946091644,
      "grad_norm": 0.8087592124938965,
      "learning_rate": 0.00012808850512682136,
      "loss": 3.2186,
      "step": 73000
    },
    {
      "epoch": 7.870619946091644,
      "eval_accuracy": 0.3905230343578902,
      "eval_loss": 3.329652786254883,
      "eval_runtime": 183.3863,
      "eval_samples_per_second": 98.213,
      "eval_steps_per_second": 6.14,
      "step": 73000
    },
    {
      "epoch": 7.876010781671159,
      "grad_norm": 0.8783208727836609,
      "learning_rate": 0.00012776470588235294,
      "loss": 3.2303,
      "step": 73050
    },
    {
      "epoch": 7.881401617250674,
      "grad_norm": 0.850925862789154,
      "learning_rate": 0.0001274409066378845,
      "loss": 3.2095,
      "step": 73100
    },
    {
      "epoch": 7.886792452830189,
      "grad_norm": 0.7974444627761841,
      "learning_rate": 0.00012711710739341607,
      "loss": 3.2193,
      "step": 73150
    },
    {
      "epoch": 7.892183288409703,
      "grad_norm": 0.8161489367485046,
      "learning_rate": 0.00012679330814894765,
      "loss": 3.2342,
      "step": 73200
    },
    {
      "epoch": 7.8975741239892185,
      "grad_norm": 0.8146517872810364,
      "learning_rate": 0.00012646950890447922,
      "loss": 3.2266,
      "step": 73250
    },
    {
      "epoch": 7.902964959568733,
      "grad_norm": 0.8209285140037537,
      "learning_rate": 0.00012614570966001077,
      "loss": 3.2198,
      "step": 73300
    },
    {
      "epoch": 7.908355795148248,
      "grad_norm": 0.8260653614997864,
      "learning_rate": 0.00012582191041554235,
      "loss": 3.2057,
      "step": 73350
    },
    {
      "epoch": 7.913746630727763,
      "grad_norm": 0.8421151638031006,
      "learning_rate": 0.00012549811117107393,
      "loss": 3.2516,
      "step": 73400
    },
    {
      "epoch": 7.919137466307277,
      "grad_norm": 0.8532014489173889,
      "learning_rate": 0.0001251743119266055,
      "loss": 3.2325,
      "step": 73450
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.7685270309448242,
      "learning_rate": 0.00012485051268213706,
      "loss": 3.2406,
      "step": 73500
    },
    {
      "epoch": 7.929919137466308,
      "grad_norm": 0.859190821647644,
      "learning_rate": 0.00012452671343766864,
      "loss": 3.2347,
      "step": 73550
    },
    {
      "epoch": 7.935309973045822,
      "grad_norm": 0.8019168972969055,
      "learning_rate": 0.00012420291419320021,
      "loss": 3.2328,
      "step": 73600
    },
    {
      "epoch": 7.940700808625337,
      "grad_norm": 0.849294126033783,
      "learning_rate": 0.0001238791149487318,
      "loss": 3.2427,
      "step": 73650
    },
    {
      "epoch": 7.946091644204852,
      "grad_norm": 0.8092149496078491,
      "learning_rate": 0.00012356179168915272,
      "loss": 3.2481,
      "step": 73700
    },
    {
      "epoch": 7.951482479784366,
      "grad_norm": 0.8331597447395325,
      "learning_rate": 0.0001232379924446843,
      "loss": 3.2047,
      "step": 73750
    },
    {
      "epoch": 7.9568733153638815,
      "grad_norm": 0.8321600556373596,
      "learning_rate": 0.00012291419320021587,
      "loss": 3.2282,
      "step": 73800
    },
    {
      "epoch": 7.962264150943396,
      "grad_norm": 0.8063613772392273,
      "learning_rate": 0.00012259039395574742,
      "loss": 3.2137,
      "step": 73850
    },
    {
      "epoch": 7.967654986522911,
      "grad_norm": 0.853704035282135,
      "learning_rate": 0.000122266594711279,
      "loss": 3.2196,
      "step": 73900
    },
    {
      "epoch": 7.973045822102426,
      "grad_norm": 0.8520025014877319,
      "learning_rate": 0.00012194279546681056,
      "loss": 3.2307,
      "step": 73950
    },
    {
      "epoch": 7.97843665768194,
      "grad_norm": 0.8468704223632812,
      "learning_rate": 0.00012161899622234214,
      "loss": 3.2275,
      "step": 74000
    },
    {
      "epoch": 7.97843665768194,
      "eval_accuracy": 0.3908582285167979,
      "eval_loss": 3.326791286468506,
      "eval_runtime": 183.707,
      "eval_samples_per_second": 98.042,
      "eval_steps_per_second": 6.129,
      "step": 74000
    },
    {
      "epoch": 7.9838274932614555,
      "grad_norm": 0.8065071105957031,
      "learning_rate": 0.00012129519697787372,
      "loss": 3.2299,
      "step": 74050
    },
    {
      "epoch": 7.989218328840971,
      "grad_norm": 0.8211476802825928,
      "learning_rate": 0.00012097139773340527,
      "loss": 3.2332,
      "step": 74100
    },
    {
      "epoch": 7.994609164420485,
      "grad_norm": 0.7979119420051575,
      "learning_rate": 0.00012064759848893685,
      "loss": 3.2219,
      "step": 74150
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.7197800874710083,
      "learning_rate": 0.00012032379924446843,
      "loss": 3.2359,
      "step": 74200
    },
    {
      "epoch": 8.005390835579515,
      "grad_norm": 0.8147539496421814,
      "learning_rate": 0.00011999999999999999,
      "loss": 3.1524,
      "step": 74250
    },
    {
      "epoch": 8.01078167115903,
      "grad_norm": 0.8394148945808411,
      "learning_rate": 0.00011967620075553155,
      "loss": 3.1463,
      "step": 74300
    },
    {
      "epoch": 8.016172506738544,
      "grad_norm": 0.8568943738937378,
      "learning_rate": 0.00011935240151106313,
      "loss": 3.1327,
      "step": 74350
    },
    {
      "epoch": 8.021563342318059,
      "grad_norm": 0.8077348470687866,
      "learning_rate": 0.00011902860226659471,
      "loss": 3.1388,
      "step": 74400
    },
    {
      "epoch": 8.026954177897574,
      "grad_norm": 0.785599410533905,
      "learning_rate": 0.00011870480302212627,
      "loss": 3.1547,
      "step": 74450
    },
    {
      "epoch": 8.032345013477089,
      "grad_norm": 0.8443362712860107,
      "learning_rate": 0.00011838100377765784,
      "loss": 3.1424,
      "step": 74500
    },
    {
      "epoch": 8.037735849056604,
      "grad_norm": 0.8324596881866455,
      "learning_rate": 0.00011805720453318941,
      "loss": 3.1601,
      "step": 74550
    },
    {
      "epoch": 8.04312668463612,
      "grad_norm": 0.8503394722938538,
      "learning_rate": 0.00011773340528872098,
      "loss": 3.1528,
      "step": 74600
    },
    {
      "epoch": 8.048517520215633,
      "grad_norm": 0.8025884628295898,
      "learning_rate": 0.00011740960604425256,
      "loss": 3.1453,
      "step": 74650
    },
    {
      "epoch": 8.053908355795148,
      "grad_norm": 0.8412855267524719,
      "learning_rate": 0.00011708580679978412,
      "loss": 3.1539,
      "step": 74700
    },
    {
      "epoch": 8.059299191374663,
      "grad_norm": 0.8652845025062561,
      "learning_rate": 0.00011676200755531568,
      "loss": 3.1489,
      "step": 74750
    },
    {
      "epoch": 8.064690026954178,
      "grad_norm": 0.8231407999992371,
      "learning_rate": 0.00011643820831084726,
      "loss": 3.156,
      "step": 74800
    },
    {
      "epoch": 8.070080862533693,
      "grad_norm": 0.8152666687965393,
      "learning_rate": 0.00011611440906637884,
      "loss": 3.1646,
      "step": 74850
    },
    {
      "epoch": 8.075471698113208,
      "grad_norm": 0.8229835033416748,
      "learning_rate": 0.00011579060982191042,
      "loss": 3.141,
      "step": 74900
    },
    {
      "epoch": 8.080862533692722,
      "grad_norm": 0.8419477343559265,
      "learning_rate": 0.00011546681057744197,
      "loss": 3.1425,
      "step": 74950
    },
    {
      "epoch": 8.086253369272237,
      "grad_norm": 0.8185001015663147,
      "learning_rate": 0.00011514301133297355,
      "loss": 3.1539,
      "step": 75000
    },
    {
      "epoch": 8.086253369272237,
      "eval_accuracy": 0.3905573686704396,
      "eval_loss": 3.333664894104004,
      "eval_runtime": 183.2216,
      "eval_samples_per_second": 98.302,
      "eval_steps_per_second": 6.146,
      "step": 75000
    },
    {
      "epoch": 8.091644204851752,
      "grad_norm": 0.8968559503555298,
      "learning_rate": 0.00011481921208850512,
      "loss": 3.1622,
      "step": 75050
    },
    {
      "epoch": 8.097035040431267,
      "grad_norm": 0.7851213812828064,
      "learning_rate": 0.00011449541284403669,
      "loss": 3.1479,
      "step": 75100
    },
    {
      "epoch": 8.102425876010782,
      "grad_norm": 0.8505614995956421,
      "learning_rate": 0.00011417161359956825,
      "loss": 3.1686,
      "step": 75150
    },
    {
      "epoch": 8.107816711590296,
      "grad_norm": 0.8062684535980225,
      "learning_rate": 0.00011384781435509983,
      "loss": 3.1602,
      "step": 75200
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.9041953682899475,
      "learning_rate": 0.0001135240151106314,
      "loss": 3.1506,
      "step": 75250
    },
    {
      "epoch": 8.118598382749326,
      "grad_norm": 0.8515322208404541,
      "learning_rate": 0.00011320021586616297,
      "loss": 3.1608,
      "step": 75300
    },
    {
      "epoch": 8.123989218328841,
      "grad_norm": 0.8603762984275818,
      "learning_rate": 0.00011288289260658391,
      "loss": 3.1645,
      "step": 75350
    },
    {
      "epoch": 8.129380053908356,
      "grad_norm": 0.8565706014633179,
      "learning_rate": 0.00011255909336211549,
      "loss": 3.1632,
      "step": 75400
    },
    {
      "epoch": 8.134770889487871,
      "grad_norm": 0.932990550994873,
      "learning_rate": 0.00011223529411764705,
      "loss": 3.167,
      "step": 75450
    },
    {
      "epoch": 8.140161725067385,
      "grad_norm": 0.7992711663246155,
      "learning_rate": 0.00011191149487317862,
      "loss": 3.141,
      "step": 75500
    },
    {
      "epoch": 8.1455525606469,
      "grad_norm": 0.835195004940033,
      "learning_rate": 0.0001115876956287102,
      "loss": 3.172,
      "step": 75550
    },
    {
      "epoch": 8.150943396226415,
      "grad_norm": 0.8241792917251587,
      "learning_rate": 0.00011126389638424176,
      "loss": 3.1497,
      "step": 75600
    },
    {
      "epoch": 8.15633423180593,
      "grad_norm": 0.7964764833450317,
      "learning_rate": 0.00011094009713977334,
      "loss": 3.1637,
      "step": 75650
    },
    {
      "epoch": 8.161725067385445,
      "grad_norm": 0.8399732708930969,
      "learning_rate": 0.0001106162978953049,
      "loss": 3.1866,
      "step": 75700
    },
    {
      "epoch": 8.167115902964959,
      "grad_norm": 0.8350764513015747,
      "learning_rate": 0.00011029249865083646,
      "loss": 3.1631,
      "step": 75750
    },
    {
      "epoch": 8.172506738544474,
      "grad_norm": 0.8351109623908997,
      "learning_rate": 0.00010996869940636804,
      "loss": 3.159,
      "step": 75800
    },
    {
      "epoch": 8.177897574123989,
      "grad_norm": 0.8726935386657715,
      "learning_rate": 0.00010964490016189962,
      "loss": 3.1482,
      "step": 75850
    },
    {
      "epoch": 8.183288409703504,
      "grad_norm": 0.8468180894851685,
      "learning_rate": 0.00010932110091743117,
      "loss": 3.1839,
      "step": 75900
    },
    {
      "epoch": 8.18867924528302,
      "grad_norm": 0.8606348037719727,
      "learning_rate": 0.00010899730167296275,
      "loss": 3.1466,
      "step": 75950
    },
    {
      "epoch": 8.194070080862534,
      "grad_norm": 0.9001939296722412,
      "learning_rate": 0.00010867350242849432,
      "loss": 3.1617,
      "step": 76000
    },
    {
      "epoch": 8.194070080862534,
      "eval_accuracy": 0.3908328037410493,
      "eval_loss": 3.3323988914489746,
      "eval_runtime": 183.5683,
      "eval_samples_per_second": 98.116,
      "eval_steps_per_second": 6.134,
      "step": 76000
    },
    {
      "epoch": 8.199460916442048,
      "grad_norm": 0.9116581082344055,
      "learning_rate": 0.0001083497031840259,
      "loss": 3.171,
      "step": 76050
    },
    {
      "epoch": 8.204851752021563,
      "grad_norm": 0.8499873876571655,
      "learning_rate": 0.00010802590393955745,
      "loss": 3.1795,
      "step": 76100
    },
    {
      "epoch": 8.210242587601078,
      "grad_norm": 0.8768589496612549,
      "learning_rate": 0.00010770210469508903,
      "loss": 3.1671,
      "step": 76150
    },
    {
      "epoch": 8.215633423180593,
      "grad_norm": 0.8752418756484985,
      "learning_rate": 0.00010737830545062061,
      "loss": 3.1726,
      "step": 76200
    },
    {
      "epoch": 8.221024258760108,
      "grad_norm": 0.8474659323692322,
      "learning_rate": 0.00010705450620615219,
      "loss": 3.1591,
      "step": 76250
    },
    {
      "epoch": 8.226415094339623,
      "grad_norm": 0.8635431528091431,
      "learning_rate": 0.00010673070696168375,
      "loss": 3.1685,
      "step": 76300
    },
    {
      "epoch": 8.231805929919137,
      "grad_norm": 0.8334670662879944,
      "learning_rate": 0.00010640690771721531,
      "loss": 3.1738,
      "step": 76350
    },
    {
      "epoch": 8.237196765498652,
      "grad_norm": 0.8614861369132996,
      "learning_rate": 0.00010608310847274689,
      "loss": 3.1823,
      "step": 76400
    },
    {
      "epoch": 8.242587601078167,
      "grad_norm": 0.8414133191108704,
      "learning_rate": 0.00010575930922827846,
      "loss": 3.1497,
      "step": 76450
    },
    {
      "epoch": 8.247978436657682,
      "grad_norm": 0.8237685561180115,
      "learning_rate": 0.00010543550998381003,
      "loss": 3.158,
      "step": 76500
    },
    {
      "epoch": 8.253369272237197,
      "grad_norm": 0.8597197532653809,
      "learning_rate": 0.0001051117107393416,
      "loss": 3.1511,
      "step": 76550
    },
    {
      "epoch": 8.25876010781671,
      "grad_norm": 0.866402268409729,
      "learning_rate": 0.00010478791149487316,
      "loss": 3.1755,
      "step": 76600
    },
    {
      "epoch": 8.264150943396226,
      "grad_norm": 0.8706876039505005,
      "learning_rate": 0.00010446411225040474,
      "loss": 3.1563,
      "step": 76650
    },
    {
      "epoch": 8.269541778975741,
      "grad_norm": 0.8851219415664673,
      "learning_rate": 0.00010414031300593632,
      "loss": 3.1876,
      "step": 76700
    },
    {
      "epoch": 8.274932614555256,
      "grad_norm": 0.8189446330070496,
      "learning_rate": 0.00010381651376146787,
      "loss": 3.1711,
      "step": 76750
    },
    {
      "epoch": 8.280323450134771,
      "grad_norm": 0.8582973480224609,
      "learning_rate": 0.00010349271451699945,
      "loss": 3.1521,
      "step": 76800
    },
    {
      "epoch": 8.285714285714286,
      "grad_norm": 0.8564453721046448,
      "learning_rate": 0.00010316891527253102,
      "loss": 3.1961,
      "step": 76850
    },
    {
      "epoch": 8.2911051212938,
      "grad_norm": 0.837863028049469,
      "learning_rate": 0.0001028451160280626,
      "loss": 3.1797,
      "step": 76900
    },
    {
      "epoch": 8.296495956873315,
      "grad_norm": 0.8210705518722534,
      "learning_rate": 0.00010252779276848353,
      "loss": 3.1627,
      "step": 76950
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.8527716398239136,
      "learning_rate": 0.0001022039935240151,
      "loss": 3.1744,
      "step": 77000
    },
    {
      "epoch": 8.30188679245283,
      "eval_accuracy": 0.3909301567285311,
      "eval_loss": 3.3277838230133057,
      "eval_runtime": 183.3119,
      "eval_samples_per_second": 98.253,
      "eval_steps_per_second": 6.143,
      "step": 77000
    },
    {
      "epoch": 8.307277628032345,
      "grad_norm": 0.8103308081626892,
      "learning_rate": 0.00010188019427954668,
      "loss": 3.1687,
      "step": 77050
    },
    {
      "epoch": 8.31266846361186,
      "grad_norm": 0.8414570093154907,
      "learning_rate": 0.00010155639503507823,
      "loss": 3.171,
      "step": 77100
    },
    {
      "epoch": 8.318059299191376,
      "grad_norm": 0.8726758360862732,
      "learning_rate": 0.00010123259579060981,
      "loss": 3.1752,
      "step": 77150
    },
    {
      "epoch": 8.323450134770889,
      "grad_norm": 0.8334712982177734,
      "learning_rate": 0.00010090879654614139,
      "loss": 3.1825,
      "step": 77200
    },
    {
      "epoch": 8.328840970350404,
      "grad_norm": 0.8159593939781189,
      "learning_rate": 0.00010058499730167296,
      "loss": 3.185,
      "step": 77250
    },
    {
      "epoch": 8.33423180592992,
      "grad_norm": 0.8698613047599792,
      "learning_rate": 0.00010026119805720452,
      "loss": 3.1679,
      "step": 77300
    },
    {
      "epoch": 8.339622641509434,
      "grad_norm": 0.854110598564148,
      "learning_rate": 9.993739881273609e-05,
      "loss": 3.1685,
      "step": 77350
    },
    {
      "epoch": 8.34501347708895,
      "grad_norm": 0.8709824085235596,
      "learning_rate": 9.961359956826767e-05,
      "loss": 3.1878,
      "step": 77400
    },
    {
      "epoch": 8.350404312668463,
      "grad_norm": 0.8535655736923218,
      "learning_rate": 9.928980032379923e-05,
      "loss": 3.1643,
      "step": 77450
    },
    {
      "epoch": 8.355795148247978,
      "grad_norm": 0.9109821915626526,
      "learning_rate": 9.896600107933081e-05,
      "loss": 3.1706,
      "step": 77500
    },
    {
      "epoch": 8.361185983827493,
      "grad_norm": 0.8124513030052185,
      "learning_rate": 9.864220183486238e-05,
      "loss": 3.1827,
      "step": 77550
    },
    {
      "epoch": 8.366576819407008,
      "grad_norm": 0.8645600080490112,
      "learning_rate": 9.831840259039394e-05,
      "loss": 3.1798,
      "step": 77600
    },
    {
      "epoch": 8.371967654986523,
      "grad_norm": 0.8976259231567383,
      "learning_rate": 9.799460334592552e-05,
      "loss": 3.1675,
      "step": 77650
    },
    {
      "epoch": 8.377358490566039,
      "grad_norm": 0.8319971561431885,
      "learning_rate": 9.76708041014571e-05,
      "loss": 3.1847,
      "step": 77700
    },
    {
      "epoch": 8.382749326145552,
      "grad_norm": 0.8768755793571472,
      "learning_rate": 9.734700485698865e-05,
      "loss": 3.1646,
      "step": 77750
    },
    {
      "epoch": 8.388140161725067,
      "grad_norm": 0.8494071960449219,
      "learning_rate": 9.702320561252022e-05,
      "loss": 3.1844,
      "step": 77800
    },
    {
      "epoch": 8.393530997304582,
      "grad_norm": 0.8520126938819885,
      "learning_rate": 9.66994063680518e-05,
      "loss": 3.173,
      "step": 77850
    },
    {
      "epoch": 8.398921832884097,
      "grad_norm": 0.8483304381370544,
      "learning_rate": 9.637560712358338e-05,
      "loss": 3.176,
      "step": 77900
    },
    {
      "epoch": 8.404312668463612,
      "grad_norm": 0.8193878531455994,
      "learning_rate": 9.605180787911493e-05,
      "loss": 3.1709,
      "step": 77950
    },
    {
      "epoch": 8.409703504043126,
      "grad_norm": 0.8840237259864807,
      "learning_rate": 9.572800863464651e-05,
      "loss": 3.1638,
      "step": 78000
    },
    {
      "epoch": 8.409703504043126,
      "eval_accuracy": 0.3914478877389671,
      "eval_loss": 3.3253719806671143,
      "eval_runtime": 183.4973,
      "eval_samples_per_second": 98.154,
      "eval_steps_per_second": 6.136,
      "step": 78000
    },
    {
      "epoch": 8.415094339622641,
      "grad_norm": 0.881489098072052,
      "learning_rate": 9.540420939017809e-05,
      "loss": 3.1724,
      "step": 78050
    },
    {
      "epoch": 8.420485175202156,
      "grad_norm": 0.86197429895401,
      "learning_rate": 9.508041014570966e-05,
      "loss": 3.1732,
      "step": 78100
    },
    {
      "epoch": 8.425876010781671,
      "grad_norm": 0.9012764692306519,
      "learning_rate": 9.475661090124121e-05,
      "loss": 3.1727,
      "step": 78150
    },
    {
      "epoch": 8.431266846361186,
      "grad_norm": 0.8909586071968079,
      "learning_rate": 9.443281165677279e-05,
      "loss": 3.1593,
      "step": 78200
    },
    {
      "epoch": 8.436657681940702,
      "grad_norm": 0.848753035068512,
      "learning_rate": 9.410901241230437e-05,
      "loss": 3.1753,
      "step": 78250
    },
    {
      "epoch": 8.442048517520215,
      "grad_norm": 0.9008297324180603,
      "learning_rate": 9.378521316783593e-05,
      "loss": 3.1694,
      "step": 78300
    },
    {
      "epoch": 8.44743935309973,
      "grad_norm": 0.8660398125648499,
      "learning_rate": 9.346141392336751e-05,
      "loss": 3.1938,
      "step": 78350
    },
    {
      "epoch": 8.452830188679245,
      "grad_norm": 0.8556224703788757,
      "learning_rate": 9.313761467889907e-05,
      "loss": 3.1827,
      "step": 78400
    },
    {
      "epoch": 8.45822102425876,
      "grad_norm": 0.8545512557029724,
      "learning_rate": 9.281381543443064e-05,
      "loss": 3.1655,
      "step": 78450
    },
    {
      "epoch": 8.463611859838275,
      "grad_norm": 0.8092154264450073,
      "learning_rate": 9.249001618996222e-05,
      "loss": 3.1973,
      "step": 78500
    },
    {
      "epoch": 8.46900269541779,
      "grad_norm": 0.8078595399856567,
      "learning_rate": 9.21662169454938e-05,
      "loss": 3.1669,
      "step": 78550
    },
    {
      "epoch": 8.474393530997304,
      "grad_norm": 0.7993897199630737,
      "learning_rate": 9.184241770102534e-05,
      "loss": 3.1705,
      "step": 78600
    },
    {
      "epoch": 8.479784366576819,
      "grad_norm": 0.8611373901367188,
      "learning_rate": 9.151861845655692e-05,
      "loss": 3.1709,
      "step": 78650
    },
    {
      "epoch": 8.485175202156334,
      "grad_norm": 0.8634024262428284,
      "learning_rate": 9.11948192120885e-05,
      "loss": 3.1975,
      "step": 78700
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.8928923010826111,
      "learning_rate": 9.087101996762008e-05,
      "loss": 3.1808,
      "step": 78750
    },
    {
      "epoch": 8.495956873315365,
      "grad_norm": 0.8743278384208679,
      "learning_rate": 9.054722072315163e-05,
      "loss": 3.173,
      "step": 78800
    },
    {
      "epoch": 8.501347708894878,
      "grad_norm": 0.8760882019996643,
      "learning_rate": 9.02234214786832e-05,
      "loss": 3.1766,
      "step": 78850
    },
    {
      "epoch": 8.506738544474393,
      "grad_norm": 0.8380445241928101,
      "learning_rate": 8.989962223421478e-05,
      "loss": 3.1711,
      "step": 78900
    },
    {
      "epoch": 8.512129380053908,
      "grad_norm": 0.8691101670265198,
      "learning_rate": 8.958229897463571e-05,
      "loss": 3.1905,
      "step": 78950
    },
    {
      "epoch": 8.517520215633423,
      "grad_norm": 0.8391762971878052,
      "learning_rate": 8.925849973016729e-05,
      "loss": 3.1771,
      "step": 79000
    },
    {
      "epoch": 8.517520215633423,
      "eval_accuracy": 0.39183414875514755,
      "eval_loss": 3.3210854530334473,
      "eval_runtime": 183.4514,
      "eval_samples_per_second": 98.179,
      "eval_steps_per_second": 6.138,
      "step": 79000
    },
    {
      "epoch": 8.522911051212938,
      "grad_norm": 0.8594598174095154,
      "learning_rate": 8.893470048569886e-05,
      "loss": 3.1846,
      "step": 79050
    },
    {
      "epoch": 8.528301886792454,
      "grad_norm": 0.8403089642524719,
      "learning_rate": 8.861090124123044e-05,
      "loss": 3.1923,
      "step": 79100
    },
    {
      "epoch": 8.533692722371967,
      "grad_norm": 0.815895676612854,
      "learning_rate": 8.828710199676199e-05,
      "loss": 3.1661,
      "step": 79150
    },
    {
      "epoch": 8.539083557951482,
      "grad_norm": 0.8398244380950928,
      "learning_rate": 8.796330275229357e-05,
      "loss": 3.1722,
      "step": 79200
    },
    {
      "epoch": 8.544474393530997,
      "grad_norm": 0.8309531807899475,
      "learning_rate": 8.763950350782515e-05,
      "loss": 3.163,
      "step": 79250
    },
    {
      "epoch": 8.549865229110512,
      "grad_norm": 0.8906266093254089,
      "learning_rate": 8.731570426335671e-05,
      "loss": 3.1818,
      "step": 79300
    },
    {
      "epoch": 8.555256064690028,
      "grad_norm": 0.8349162340164185,
      "learning_rate": 8.699190501888828e-05,
      "loss": 3.1804,
      "step": 79350
    },
    {
      "epoch": 8.560646900269543,
      "grad_norm": 0.8070085048675537,
      "learning_rate": 8.666810577441985e-05,
      "loss": 3.1515,
      "step": 79400
    },
    {
      "epoch": 8.566037735849056,
      "grad_norm": 0.8301887512207031,
      "learning_rate": 8.634430652995142e-05,
      "loss": 3.1848,
      "step": 79450
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 0.8491299748420715,
      "learning_rate": 8.6020507285483e-05,
      "loss": 3.1771,
      "step": 79500
    },
    {
      "epoch": 8.576819407008086,
      "grad_norm": 0.8146814703941345,
      "learning_rate": 8.569670804101456e-05,
      "loss": 3.1755,
      "step": 79550
    },
    {
      "epoch": 8.582210242587601,
      "grad_norm": 0.8444635272026062,
      "learning_rate": 8.537290879654612e-05,
      "loss": 3.1763,
      "step": 79600
    },
    {
      "epoch": 8.587601078167117,
      "grad_norm": 0.8767122030258179,
      "learning_rate": 8.50491095520777e-05,
      "loss": 3.1571,
      "step": 79650
    },
    {
      "epoch": 8.59299191374663,
      "grad_norm": 0.8696253895759583,
      "learning_rate": 8.472531030760928e-05,
      "loss": 3.1813,
      "step": 79700
    },
    {
      "epoch": 8.598382749326145,
      "grad_norm": 0.8822606205940247,
      "learning_rate": 8.440151106314086e-05,
      "loss": 3.189,
      "step": 79750
    },
    {
      "epoch": 8.60377358490566,
      "grad_norm": 0.8302372694015503,
      "learning_rate": 8.407771181867241e-05,
      "loss": 3.1693,
      "step": 79800
    },
    {
      "epoch": 8.609164420485175,
      "grad_norm": 0.8339834213256836,
      "learning_rate": 8.375391257420398e-05,
      "loss": 3.1542,
      "step": 79850
    },
    {
      "epoch": 8.61455525606469,
      "grad_norm": 0.862395703792572,
      "learning_rate": 8.343011332973556e-05,
      "loss": 3.1626,
      "step": 79900
    },
    {
      "epoch": 8.619946091644206,
      "grad_norm": 0.846082329750061,
      "learning_rate": 8.310631408526714e-05,
      "loss": 3.1861,
      "step": 79950
    },
    {
      "epoch": 8.625336927223719,
      "grad_norm": 0.8844338059425354,
      "learning_rate": 8.278251484079869e-05,
      "loss": 3.1753,
      "step": 80000
    },
    {
      "epoch": 8.625336927223719,
      "eval_accuracy": 0.3923148291308388,
      "eval_loss": 3.3195295333862305,
      "eval_runtime": 183.5025,
      "eval_samples_per_second": 98.151,
      "eval_steps_per_second": 6.136,
      "step": 80000
    },
    {
      "epoch": 8.630727762803234,
      "grad_norm": 0.8043167591094971,
      "learning_rate": 8.245871559633027e-05,
      "loss": 3.1842,
      "step": 80050
    },
    {
      "epoch": 8.63611859838275,
      "grad_norm": 0.8519346714019775,
      "learning_rate": 8.213491635186185e-05,
      "loss": 3.1895,
      "step": 80100
    },
    {
      "epoch": 8.641509433962264,
      "grad_norm": 0.8410979509353638,
      "learning_rate": 8.181111710739341e-05,
      "loss": 3.1753,
      "step": 80150
    },
    {
      "epoch": 8.64690026954178,
      "grad_norm": 0.859257161617279,
      "learning_rate": 8.148731786292497e-05,
      "loss": 3.1692,
      "step": 80200
    },
    {
      "epoch": 8.652291105121293,
      "grad_norm": 0.8353541493415833,
      "learning_rate": 8.116351861845655e-05,
      "loss": 3.179,
      "step": 80250
    },
    {
      "epoch": 8.657681940700808,
      "grad_norm": 1.4765746593475342,
      "learning_rate": 8.083971937398812e-05,
      "loss": 3.1822,
      "step": 80300
    },
    {
      "epoch": 8.663072776280323,
      "grad_norm": 0.8867174386978149,
      "learning_rate": 8.05159201295197e-05,
      "loss": 3.1748,
      "step": 80350
    },
    {
      "epoch": 8.668463611859838,
      "grad_norm": 0.8695386052131653,
      "learning_rate": 8.019212088505126e-05,
      "loss": 3.1764,
      "step": 80400
    },
    {
      "epoch": 8.673854447439354,
      "grad_norm": 0.8609994649887085,
      "learning_rate": 7.986832164058282e-05,
      "loss": 3.1638,
      "step": 80450
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.8649408221244812,
      "learning_rate": 7.95445223961144e-05,
      "loss": 3.1631,
      "step": 80500
    },
    {
      "epoch": 8.684636118598382,
      "grad_norm": 0.8425276279449463,
      "learning_rate": 7.922072315164598e-05,
      "loss": 3.1902,
      "step": 80550
    },
    {
      "epoch": 8.690026954177897,
      "grad_norm": 0.8469082117080688,
      "learning_rate": 7.889692390717755e-05,
      "loss": 3.1604,
      "step": 80600
    },
    {
      "epoch": 8.695417789757412,
      "grad_norm": 0.9055061936378479,
      "learning_rate": 7.85731246627091e-05,
      "loss": 3.1877,
      "step": 80650
    },
    {
      "epoch": 8.700808625336927,
      "grad_norm": 0.8626482486724854,
      "learning_rate": 7.824932541824068e-05,
      "loss": 3.1765,
      "step": 80700
    },
    {
      "epoch": 8.706199460916443,
      "grad_norm": 0.913098931312561,
      "learning_rate": 7.792552617377226e-05,
      "loss": 3.1799,
      "step": 80750
    },
    {
      "epoch": 8.711590296495956,
      "grad_norm": 0.8522776961326599,
      "learning_rate": 7.760172692930382e-05,
      "loss": 3.1608,
      "step": 80800
    },
    {
      "epoch": 8.716981132075471,
      "grad_norm": 0.8847063183784485,
      "learning_rate": 7.727792768483539e-05,
      "loss": 3.1656,
      "step": 80850
    },
    {
      "epoch": 8.722371967654986,
      "grad_norm": 0.9037518501281738,
      "learning_rate": 7.695412844036697e-05,
      "loss": 3.1761,
      "step": 80900
    },
    {
      "epoch": 8.727762803234501,
      "grad_norm": 0.8497751951217651,
      "learning_rate": 7.663032919589854e-05,
      "loss": 3.1929,
      "step": 80950
    },
    {
      "epoch": 8.733153638814017,
      "grad_norm": 0.8641258478164673,
      "learning_rate": 7.630652995143011e-05,
      "loss": 3.1883,
      "step": 81000
    },
    {
      "epoch": 8.733153638814017,
      "eval_accuracy": 0.39242739352261463,
      "eval_loss": 3.3154139518737793,
      "eval_runtime": 183.473,
      "eval_samples_per_second": 98.167,
      "eval_steps_per_second": 6.137,
      "step": 81000
    },
    {
      "epoch": 8.738544474393532,
      "grad_norm": 0.8507005572319031,
      "learning_rate": 7.598273070696167e-05,
      "loss": 3.1759,
      "step": 81050
    },
    {
      "epoch": 8.743935309973045,
      "grad_norm": 0.8656426072120667,
      "learning_rate": 7.565893146249325e-05,
      "loss": 3.1855,
      "step": 81100
    },
    {
      "epoch": 8.74932614555256,
      "grad_norm": 0.8609006404876709,
      "learning_rate": 7.533513221802481e-05,
      "loss": 3.1698,
      "step": 81150
    },
    {
      "epoch": 8.754716981132075,
      "grad_norm": 0.973294198513031,
      "learning_rate": 7.501133297355639e-05,
      "loss": 3.1679,
      "step": 81200
    },
    {
      "epoch": 8.76010781671159,
      "grad_norm": 0.8869118094444275,
      "learning_rate": 7.468753372908796e-05,
      "loss": 3.1771,
      "step": 81250
    },
    {
      "epoch": 8.765498652291106,
      "grad_norm": 0.8876540064811707,
      "learning_rate": 7.436373448461953e-05,
      "loss": 3.1754,
      "step": 81300
    },
    {
      "epoch": 8.77088948787062,
      "grad_norm": 0.8496720790863037,
      "learning_rate": 7.40399352401511e-05,
      "loss": 3.1767,
      "step": 81350
    },
    {
      "epoch": 8.776280323450134,
      "grad_norm": 0.8450832366943359,
      "learning_rate": 7.371613599568268e-05,
      "loss": 3.1846,
      "step": 81400
    },
    {
      "epoch": 8.78167115902965,
      "grad_norm": 0.8679110407829285,
      "learning_rate": 7.339233675121424e-05,
      "loss": 3.1839,
      "step": 81450
    },
    {
      "epoch": 8.787061994609164,
      "grad_norm": 0.9450589418411255,
      "learning_rate": 7.306853750674582e-05,
      "loss": 3.2063,
      "step": 81500
    },
    {
      "epoch": 8.79245283018868,
      "grad_norm": 0.8842014670372009,
      "learning_rate": 7.274473826227738e-05,
      "loss": 3.1691,
      "step": 81550
    },
    {
      "epoch": 8.797843665768195,
      "grad_norm": 0.8416998386383057,
      "learning_rate": 7.242093901780896e-05,
      "loss": 3.1774,
      "step": 81600
    },
    {
      "epoch": 8.80323450134771,
      "grad_norm": 0.9002299904823303,
      "learning_rate": 7.209713977334052e-05,
      "loss": 3.1794,
      "step": 81650
    },
    {
      "epoch": 8.808625336927223,
      "grad_norm": 0.8381384015083313,
      "learning_rate": 7.17733405288721e-05,
      "loss": 3.1905,
      "step": 81700
    },
    {
      "epoch": 8.814016172506738,
      "grad_norm": 0.8705641031265259,
      "learning_rate": 7.144954128440366e-05,
      "loss": 3.1742,
      "step": 81750
    },
    {
      "epoch": 8.819407008086253,
      "grad_norm": 0.8659651279449463,
      "learning_rate": 7.112574203993523e-05,
      "loss": 3.1648,
      "step": 81800
    },
    {
      "epoch": 8.824797843665769,
      "grad_norm": 0.8624113202095032,
      "learning_rate": 7.08019427954668e-05,
      "loss": 3.181,
      "step": 81850
    },
    {
      "epoch": 8.830188679245284,
      "grad_norm": 0.8544336557388306,
      "learning_rate": 7.047814355099837e-05,
      "loss": 3.2008,
      "step": 81900
    },
    {
      "epoch": 8.835579514824797,
      "grad_norm": 0.8490225076675415,
      "learning_rate": 7.015434430652993e-05,
      "loss": 3.1846,
      "step": 81950
    },
    {
      "epoch": 8.840970350404312,
      "grad_norm": 0.8430733680725098,
      "learning_rate": 6.983054506206151e-05,
      "loss": 3.173,
      "step": 82000
    },
    {
      "epoch": 8.840970350404312,
      "eval_accuracy": 0.3929860866717567,
      "eval_loss": 3.311267137527466,
      "eval_runtime": 183.6755,
      "eval_samples_per_second": 98.059,
      "eval_steps_per_second": 6.13,
      "step": 82000
    },
    {
      "epoch": 8.846361185983827,
      "grad_norm": 0.8979526162147522,
      "learning_rate": 6.950674581759309e-05,
      "loss": 3.1733,
      "step": 82050
    },
    {
      "epoch": 8.851752021563343,
      "grad_norm": 0.8584031462669373,
      "learning_rate": 6.918294657312465e-05,
      "loss": 3.1726,
      "step": 82100
    },
    {
      "epoch": 8.857142857142858,
      "grad_norm": 0.8862038850784302,
      "learning_rate": 6.885914732865623e-05,
      "loss": 3.169,
      "step": 82150
    },
    {
      "epoch": 8.862533692722373,
      "grad_norm": 0.8636807799339294,
      "learning_rate": 6.85353480841878e-05,
      "loss": 3.1808,
      "step": 82200
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.8127702474594116,
      "learning_rate": 6.821154883971937e-05,
      "loss": 3.182,
      "step": 82250
    },
    {
      "epoch": 8.873315363881401,
      "grad_norm": 0.8639582991600037,
      "learning_rate": 6.788774959525094e-05,
      "loss": 3.1617,
      "step": 82300
    },
    {
      "epoch": 8.878706199460916,
      "grad_norm": 0.8684182167053223,
      "learning_rate": 6.756395035078252e-05,
      "loss": 3.1501,
      "step": 82350
    },
    {
      "epoch": 8.884097035040432,
      "grad_norm": 0.8748733401298523,
      "learning_rate": 6.724015110631408e-05,
      "loss": 3.1779,
      "step": 82400
    },
    {
      "epoch": 8.889487870619947,
      "grad_norm": 0.8723418116569519,
      "learning_rate": 6.691635186184566e-05,
      "loss": 3.1651,
      "step": 82450
    },
    {
      "epoch": 8.89487870619946,
      "grad_norm": 0.8358088135719299,
      "learning_rate": 6.659255261737722e-05,
      "loss": 3.1762,
      "step": 82500
    },
    {
      "epoch": 8.900269541778975,
      "grad_norm": 0.8295577168464661,
      "learning_rate": 6.626875337290879e-05,
      "loss": 3.1649,
      "step": 82550
    },
    {
      "epoch": 8.90566037735849,
      "grad_norm": 0.8575540781021118,
      "learning_rate": 6.594495412844036e-05,
      "loss": 3.1891,
      "step": 82600
    },
    {
      "epoch": 8.911051212938006,
      "grad_norm": 0.8286822438240051,
      "learning_rate": 6.562115488397193e-05,
      "loss": 3.1915,
      "step": 82650
    },
    {
      "epoch": 8.91644204851752,
      "grad_norm": 0.8888596296310425,
      "learning_rate": 6.52973556395035e-05,
      "loss": 3.1872,
      "step": 82700
    },
    {
      "epoch": 8.921832884097036,
      "grad_norm": 0.863433301448822,
      "learning_rate": 6.497355639503507e-05,
      "loss": 3.1827,
      "step": 82750
    },
    {
      "epoch": 8.92722371967655,
      "grad_norm": 0.8844254612922668,
      "learning_rate": 6.464975715056663e-05,
      "loss": 3.181,
      "step": 82800
    },
    {
      "epoch": 8.932614555256064,
      "grad_norm": 0.8509985208511353,
      "learning_rate": 6.432595790609821e-05,
      "loss": 3.1813,
      "step": 82850
    },
    {
      "epoch": 8.93800539083558,
      "grad_norm": 0.896156907081604,
      "learning_rate": 6.400215866162979e-05,
      "loss": 3.177,
      "step": 82900
    },
    {
      "epoch": 8.943396226415095,
      "grad_norm": 0.8738552927970886,
      "learning_rate": 6.367835941716135e-05,
      "loss": 3.1741,
      "step": 82950
    },
    {
      "epoch": 8.94878706199461,
      "grad_norm": 0.8401568531990051,
      "learning_rate": 6.335456017269293e-05,
      "loss": 3.1879,
      "step": 83000
    },
    {
      "epoch": 8.94878706199461,
      "eval_accuracy": 0.3931521082843372,
      "eval_loss": 3.3100833892822266,
      "eval_runtime": 183.3659,
      "eval_samples_per_second": 98.224,
      "eval_steps_per_second": 6.141,
      "step": 83000
    },
    {
      "epoch": 8.954177897574123,
      "grad_norm": 0.8782355785369873,
      "learning_rate": 6.30307609282245e-05,
      "loss": 3.1927,
      "step": 83050
    },
    {
      "epoch": 8.959568733153638,
      "grad_norm": 0.8759222030639648,
      "learning_rate": 6.270696168375607e-05,
      "loss": 3.1784,
      "step": 83100
    },
    {
      "epoch": 8.964959568733153,
      "grad_norm": 0.8785428404808044,
      "learning_rate": 6.238316243928764e-05,
      "loss": 3.1803,
      "step": 83150
    },
    {
      "epoch": 8.970350404312669,
      "grad_norm": 0.8433955311775208,
      "learning_rate": 6.205936319481921e-05,
      "loss": 3.188,
      "step": 83200
    },
    {
      "epoch": 8.975741239892184,
      "grad_norm": 0.877407968044281,
      "learning_rate": 6.173556395035078e-05,
      "loss": 3.1738,
      "step": 83250
    },
    {
      "epoch": 8.981132075471699,
      "grad_norm": 0.8445099592208862,
      "learning_rate": 6.141176470588236e-05,
      "loss": 3.1865,
      "step": 83300
    },
    {
      "epoch": 8.986522911051212,
      "grad_norm": 0.8456767797470093,
      "learning_rate": 6.108796546141392e-05,
      "loss": 3.1833,
      "step": 83350
    },
    {
      "epoch": 8.991913746630727,
      "grad_norm": 0.8464460968971252,
      "learning_rate": 6.076416621694549e-05,
      "loss": 3.1835,
      "step": 83400
    },
    {
      "epoch": 8.997304582210242,
      "grad_norm": 0.8834063410758972,
      "learning_rate": 6.0440366972477055e-05,
      "loss": 3.1796,
      "step": 83450
    },
    {
      "epoch": 9.002695417789758,
      "grad_norm": 0.8756682276725769,
      "learning_rate": 6.011656772800863e-05,
      "loss": 3.1574,
      "step": 83500
    },
    {
      "epoch": 9.008086253369273,
      "grad_norm": 0.8651514053344727,
      "learning_rate": 5.9792768483540197e-05,
      "loss": 3.1089,
      "step": 83550
    },
    {
      "epoch": 9.013477088948788,
      "grad_norm": 0.8796454668045044,
      "learning_rate": 5.946896923907177e-05,
      "loss": 3.1068,
      "step": 83600
    },
    {
      "epoch": 9.018867924528301,
      "grad_norm": 0.9122756123542786,
      "learning_rate": 5.9145169994603345e-05,
      "loss": 3.1101,
      "step": 83650
    },
    {
      "epoch": 9.024258760107816,
      "grad_norm": 0.9356156587600708,
      "learning_rate": 5.882137075013491e-05,
      "loss": 3.1069,
      "step": 83700
    },
    {
      "epoch": 9.029649595687331,
      "grad_norm": 0.8637686371803284,
      "learning_rate": 5.849757150566649e-05,
      "loss": 3.1241,
      "step": 83750
    },
    {
      "epoch": 9.035040431266847,
      "grad_norm": 0.830084502696991,
      "learning_rate": 5.817377226119805e-05,
      "loss": 3.1141,
      "step": 83800
    },
    {
      "epoch": 9.040431266846362,
      "grad_norm": 0.8634269833564758,
      "learning_rate": 5.784997301672963e-05,
      "loss": 3.1249,
      "step": 83850
    },
    {
      "epoch": 9.045822102425875,
      "grad_norm": 0.8700315356254578,
      "learning_rate": 5.752617377226119e-05,
      "loss": 3.1224,
      "step": 83900
    },
    {
      "epoch": 9.05121293800539,
      "grad_norm": 0.8730193376541138,
      "learning_rate": 5.7202374527792764e-05,
      "loss": 3.1196,
      "step": 83950
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.9231434464454651,
      "learning_rate": 5.6878575283324335e-05,
      "loss": 3.1161,
      "step": 84000
    },
    {
      "epoch": 9.056603773584905,
      "eval_accuracy": 0.39329487817892544,
      "eval_loss": 3.3126213550567627,
      "eval_runtime": 183.6807,
      "eval_samples_per_second": 98.056,
      "eval_steps_per_second": 6.13,
      "step": 84000
    },
    {
      "epoch": 9.06199460916442,
      "grad_norm": 0.8769927024841309,
      "learning_rate": 5.6554776038855905e-05,
      "loss": 3.1255,
      "step": 84050
    },
    {
      "epoch": 9.067385444743936,
      "grad_norm": 0.8235280513763428,
      "learning_rate": 5.623097679438747e-05,
      "loss": 3.1235,
      "step": 84100
    },
    {
      "epoch": 9.07277628032345,
      "grad_norm": 0.8591448664665222,
      "learning_rate": 5.590717754991905e-05,
      "loss": 3.1205,
      "step": 84150
    },
    {
      "epoch": 9.078167115902964,
      "grad_norm": 0.8690507411956787,
      "learning_rate": 5.558337830545061e-05,
      "loss": 3.1238,
      "step": 84200
    },
    {
      "epoch": 9.08355795148248,
      "grad_norm": 0.841168224811554,
      "learning_rate": 5.525957906098219e-05,
      "loss": 3.1343,
      "step": 84250
    },
    {
      "epoch": 9.088948787061994,
      "grad_norm": 0.8726950883865356,
      "learning_rate": 5.493577981651375e-05,
      "loss": 3.1138,
      "step": 84300
    },
    {
      "epoch": 9.09433962264151,
      "grad_norm": 0.898984968662262,
      "learning_rate": 5.461845655693469e-05,
      "loss": 3.131,
      "step": 84350
    },
    {
      "epoch": 9.099730458221025,
      "grad_norm": 0.871024489402771,
      "learning_rate": 5.429465731246627e-05,
      "loss": 3.1095,
      "step": 84400
    },
    {
      "epoch": 9.10512129380054,
      "grad_norm": 0.9210131764411926,
      "learning_rate": 5.3970858067997833e-05,
      "loss": 3.1129,
      "step": 84450
    },
    {
      "epoch": 9.110512129380053,
      "grad_norm": 0.8362647294998169,
      "learning_rate": 5.364705882352941e-05,
      "loss": 3.1118,
      "step": 84500
    },
    {
      "epoch": 9.115902964959568,
      "grad_norm": 0.9286665320396423,
      "learning_rate": 5.3323259579060975e-05,
      "loss": 3.1208,
      "step": 84550
    },
    {
      "epoch": 9.121293800539084,
      "grad_norm": 0.8982177972793579,
      "learning_rate": 5.299946033459255e-05,
      "loss": 3.1197,
      "step": 84600
    },
    {
      "epoch": 9.126684636118599,
      "grad_norm": 0.8783411979675293,
      "learning_rate": 5.267566109012412e-05,
      "loss": 3.1145,
      "step": 84650
    },
    {
      "epoch": 9.132075471698114,
      "grad_norm": 0.8987107276916504,
      "learning_rate": 5.235186184565569e-05,
      "loss": 3.1187,
      "step": 84700
    },
    {
      "epoch": 9.137466307277627,
      "grad_norm": 0.872234582901001,
      "learning_rate": 5.202806260118726e-05,
      "loss": 3.128,
      "step": 84750
    },
    {
      "epoch": 9.142857142857142,
      "grad_norm": 0.8869773745536804,
      "learning_rate": 5.170426335671883e-05,
      "loss": 3.132,
      "step": 84800
    },
    {
      "epoch": 9.148247978436657,
      "grad_norm": 0.8714069724082947,
      "learning_rate": 5.1380464112250394e-05,
      "loss": 3.1384,
      "step": 84850
    },
    {
      "epoch": 9.153638814016173,
      "grad_norm": 0.8658605217933655,
      "learning_rate": 5.105666486778197e-05,
      "loss": 3.1152,
      "step": 84900
    },
    {
      "epoch": 9.159029649595688,
      "grad_norm": 0.8711130619049072,
      "learning_rate": 5.0732865623313536e-05,
      "loss": 3.1266,
      "step": 84950
    },
    {
      "epoch": 9.164420485175203,
      "grad_norm": 0.8554506301879883,
      "learning_rate": 5.040906637884511e-05,
      "loss": 3.1182,
      "step": 85000
    },
    {
      "epoch": 9.164420485175203,
      "eval_accuracy": 0.3934888235836743,
      "eval_loss": 3.310671806335449,
      "eval_runtime": 183.8724,
      "eval_samples_per_second": 97.954,
      "eval_steps_per_second": 6.124,
      "step": 85000
    },
    {
      "epoch": 9.169811320754716,
      "grad_norm": 0.8684996962547302,
      "learning_rate": 5.0085267134376684e-05,
      "loss": 3.1199,
      "step": 85050
    },
    {
      "epoch": 9.175202156334231,
      "grad_norm": 0.8756259679794312,
      "learning_rate": 4.9761467889908255e-05,
      "loss": 3.1533,
      "step": 85100
    },
    {
      "epoch": 9.180592991913747,
      "grad_norm": 0.8757433295249939,
      "learning_rate": 4.9437668645439826e-05,
      "loss": 3.1191,
      "step": 85150
    },
    {
      "epoch": 9.185983827493262,
      "grad_norm": 0.857020378112793,
      "learning_rate": 4.911386940097139e-05,
      "loss": 3.1121,
      "step": 85200
    },
    {
      "epoch": 9.191374663072777,
      "grad_norm": 0.9065777659416199,
      "learning_rate": 4.879007015650297e-05,
      "loss": 3.1354,
      "step": 85250
    },
    {
      "epoch": 9.19676549865229,
      "grad_norm": 0.848362147808075,
      "learning_rate": 4.846627091203453e-05,
      "loss": 3.1066,
      "step": 85300
    },
    {
      "epoch": 9.202156334231805,
      "grad_norm": 0.8962082862854004,
      "learning_rate": 4.814247166756611e-05,
      "loss": 3.1158,
      "step": 85350
    },
    {
      "epoch": 9.20754716981132,
      "grad_norm": 0.8896428942680359,
      "learning_rate": 4.7818672423097674e-05,
      "loss": 3.1332,
      "step": 85400
    },
    {
      "epoch": 9.212938005390836,
      "grad_norm": 0.8223966360092163,
      "learning_rate": 4.7494873178629244e-05,
      "loss": 3.097,
      "step": 85450
    },
    {
      "epoch": 9.21832884097035,
      "grad_norm": 0.867064893245697,
      "learning_rate": 4.7171073934160815e-05,
      "loss": 3.1308,
      "step": 85500
    },
    {
      "epoch": 9.223719676549866,
      "grad_norm": 0.9328610301017761,
      "learning_rate": 4.6847274689692386e-05,
      "loss": 3.1116,
      "step": 85550
    },
    {
      "epoch": 9.22911051212938,
      "grad_norm": 0.8840168714523315,
      "learning_rate": 4.652347544522396e-05,
      "loss": 3.134,
      "step": 85600
    },
    {
      "epoch": 9.234501347708894,
      "grad_norm": 0.8884910345077515,
      "learning_rate": 4.619967620075553e-05,
      "loss": 3.1195,
      "step": 85650
    },
    {
      "epoch": 9.23989218328841,
      "grad_norm": 0.9012236595153809,
      "learning_rate": 4.587587695628709e-05,
      "loss": 3.1036,
      "step": 85700
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.8753321170806885,
      "learning_rate": 4.555207771181867e-05,
      "loss": 3.1286,
      "step": 85750
    },
    {
      "epoch": 9.25067385444744,
      "grad_norm": 0.8457467555999756,
      "learning_rate": 4.522827846735024e-05,
      "loss": 3.1347,
      "step": 85800
    },
    {
      "epoch": 9.256064690026955,
      "grad_norm": 0.8559665083885193,
      "learning_rate": 4.490447922288181e-05,
      "loss": 3.134,
      "step": 85850
    },
    {
      "epoch": 9.261455525606468,
      "grad_norm": 0.8733518123626709,
      "learning_rate": 4.458067997841338e-05,
      "loss": 3.0981,
      "step": 85900
    },
    {
      "epoch": 9.266846361185983,
      "grad_norm": 0.8558553457260132,
      "learning_rate": 4.4256880733944947e-05,
      "loss": 3.134,
      "step": 85950
    },
    {
      "epoch": 9.272237196765499,
      "grad_norm": 0.8802152872085571,
      "learning_rate": 4.3933081489476524e-05,
      "loss": 3.1283,
      "step": 86000
    },
    {
      "epoch": 9.272237196765499,
      "eval_accuracy": 0.39366853546011943,
      "eval_loss": 3.3087074756622314,
      "eval_runtime": 183.4292,
      "eval_samples_per_second": 98.19,
      "eval_steps_per_second": 6.139,
      "step": 86000
    },
    {
      "epoch": 9.277628032345014,
      "grad_norm": 0.8693817257881165,
      "learning_rate": 4.360928224500809e-05,
      "loss": 3.1253,
      "step": 86050
    },
    {
      "epoch": 9.283018867924529,
      "grad_norm": 0.929667592048645,
      "learning_rate": 4.3285483000539666e-05,
      "loss": 3.124,
      "step": 86100
    },
    {
      "epoch": 9.288409703504042,
      "grad_norm": 0.9104236960411072,
      "learning_rate": 4.296168375607123e-05,
      "loss": 3.1201,
      "step": 86150
    },
    {
      "epoch": 9.293800539083557,
      "grad_norm": 0.8924854397773743,
      "learning_rate": 4.263788451160281e-05,
      "loss": 3.134,
      "step": 86200
    },
    {
      "epoch": 9.299191374663073,
      "grad_norm": 0.8825605511665344,
      "learning_rate": 4.231408526713437e-05,
      "loss": 3.1373,
      "step": 86250
    },
    {
      "epoch": 9.304582210242588,
      "grad_norm": 0.8415103554725647,
      "learning_rate": 4.199028602266594e-05,
      "loss": 3.1284,
      "step": 86300
    },
    {
      "epoch": 9.309973045822103,
      "grad_norm": 0.8934675455093384,
      "learning_rate": 4.1666486778197514e-05,
      "loss": 3.0965,
      "step": 86350
    },
    {
      "epoch": 9.315363881401618,
      "grad_norm": 0.8981188535690308,
      "learning_rate": 4.1342687533729085e-05,
      "loss": 3.1221,
      "step": 86400
    },
    {
      "epoch": 9.320754716981131,
      "grad_norm": 0.8392598032951355,
      "learning_rate": 4.101888828926065e-05,
      "loss": 3.1186,
      "step": 86450
    },
    {
      "epoch": 9.326145552560646,
      "grad_norm": 0.8929069638252258,
      "learning_rate": 4.0695089044792226e-05,
      "loss": 3.1407,
      "step": 86500
    },
    {
      "epoch": 9.331536388140162,
      "grad_norm": 0.9197957515716553,
      "learning_rate": 4.037128980032379e-05,
      "loss": 3.1202,
      "step": 86550
    },
    {
      "epoch": 9.336927223719677,
      "grad_norm": 0.8734220266342163,
      "learning_rate": 4.004749055585537e-05,
      "loss": 3.1092,
      "step": 86600
    },
    {
      "epoch": 9.342318059299192,
      "grad_norm": 0.8806682229042053,
      "learning_rate": 3.972369131138694e-05,
      "loss": 3.1167,
      "step": 86650
    },
    {
      "epoch": 9.347708894878707,
      "grad_norm": 0.8593670129776001,
      "learning_rate": 3.93998920669185e-05,
      "loss": 3.1351,
      "step": 86700
    },
    {
      "epoch": 9.35309973045822,
      "grad_norm": 0.8641849160194397,
      "learning_rate": 3.907609282245008e-05,
      "loss": 3.1207,
      "step": 86750
    },
    {
      "epoch": 9.358490566037736,
      "grad_norm": 0.8967059254646301,
      "learning_rate": 3.8752293577981645e-05,
      "loss": 3.1447,
      "step": 86800
    },
    {
      "epoch": 9.36388140161725,
      "grad_norm": 0.851588249206543,
      "learning_rate": 3.842849433351322e-05,
      "loss": 3.1141,
      "step": 86850
    },
    {
      "epoch": 9.369272237196766,
      "grad_norm": 0.8949471116065979,
      "learning_rate": 3.810469508904479e-05,
      "loss": 3.1332,
      "step": 86900
    },
    {
      "epoch": 9.374663072776281,
      "grad_norm": 0.8716377019882202,
      "learning_rate": 3.7780895844576364e-05,
      "loss": 3.1285,
      "step": 86950
    },
    {
      "epoch": 9.380053908355794,
      "grad_norm": 0.8617969155311584,
      "learning_rate": 3.745709660010793e-05,
      "loss": 3.1145,
      "step": 87000
    },
    {
      "epoch": 9.380053908355794,
      "eval_accuracy": 0.39399134318981627,
      "eval_loss": 3.307685613632202,
      "eval_runtime": 183.5956,
      "eval_samples_per_second": 98.101,
      "eval_steps_per_second": 6.133,
      "step": 87000
    },
    {
      "epoch": 9.38544474393531,
      "grad_norm": 0.8978739976882935,
      "learning_rate": 3.71332973556395e-05,
      "loss": 3.1249,
      "step": 87050
    },
    {
      "epoch": 9.390835579514825,
      "grad_norm": 0.857925295829773,
      "learning_rate": 3.680949811117107e-05,
      "loss": 3.1099,
      "step": 87100
    },
    {
      "epoch": 9.39622641509434,
      "grad_norm": 0.8732088208198547,
      "learning_rate": 3.648569886670264e-05,
      "loss": 3.1194,
      "step": 87150
    },
    {
      "epoch": 9.401617250673855,
      "grad_norm": 0.9072591066360474,
      "learning_rate": 3.616189962223421e-05,
      "loss": 3.1287,
      "step": 87200
    },
    {
      "epoch": 9.40700808625337,
      "grad_norm": 0.8680362105369568,
      "learning_rate": 3.583810037776578e-05,
      "loss": 3.1274,
      "step": 87250
    },
    {
      "epoch": 9.412398921832883,
      "grad_norm": 0.8892818093299866,
      "learning_rate": 3.5514301133297354e-05,
      "loss": 3.1368,
      "step": 87300
    },
    {
      "epoch": 9.417789757412399,
      "grad_norm": 0.9270879626274109,
      "learning_rate": 3.519697787371829e-05,
      "loss": 3.1019,
      "step": 87350
    },
    {
      "epoch": 9.423180592991914,
      "grad_norm": 0.9087534546852112,
      "learning_rate": 3.487317862924986e-05,
      "loss": 3.1312,
      "step": 87400
    },
    {
      "epoch": 9.428571428571429,
      "grad_norm": 0.8477288484573364,
      "learning_rate": 3.4549379384781434e-05,
      "loss": 3.1269,
      "step": 87450
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.8833394646644592,
      "learning_rate": 3.4225580140313e-05,
      "loss": 3.1109,
      "step": 87500
    },
    {
      "epoch": 9.439353099730457,
      "grad_norm": 0.834549605846405,
      "learning_rate": 3.390178089584457e-05,
      "loss": 3.1338,
      "step": 87550
    },
    {
      "epoch": 9.444743935309972,
      "grad_norm": 0.8486028909683228,
      "learning_rate": 3.357798165137615e-05,
      "loss": 3.1179,
      "step": 87600
    },
    {
      "epoch": 9.450134770889488,
      "grad_norm": 0.8409470915794373,
      "learning_rate": 3.325418240690772e-05,
      "loss": 3.1084,
      "step": 87650
    },
    {
      "epoch": 9.455525606469003,
      "grad_norm": 0.9353410005569458,
      "learning_rate": 3.293038316243929e-05,
      "loss": 3.1225,
      "step": 87700
    },
    {
      "epoch": 9.460916442048518,
      "grad_norm": 0.886806309223175,
      "learning_rate": 3.260658391797086e-05,
      "loss": 3.1205,
      "step": 87750
    },
    {
      "epoch": 9.466307277628033,
      "grad_norm": 0.8259122371673584,
      "learning_rate": 3.2282784673502424e-05,
      "loss": 3.121,
      "step": 87800
    },
    {
      "epoch": 9.471698113207546,
      "grad_norm": 0.8698675036430359,
      "learning_rate": 3.1958985429033995e-05,
      "loss": 3.1229,
      "step": 87850
    },
    {
      "epoch": 9.477088948787062,
      "grad_norm": 0.8667991757392883,
      "learning_rate": 3.1635186184565565e-05,
      "loss": 3.1285,
      "step": 87900
    },
    {
      "epoch": 9.482479784366577,
      "grad_norm": 0.9097034931182861,
      "learning_rate": 3.1311386940097136e-05,
      "loss": 3.1274,
      "step": 87950
    },
    {
      "epoch": 9.487870619946092,
      "grad_norm": 0.9387816786766052,
      "learning_rate": 3.098758769562871e-05,
      "loss": 3.1251,
      "step": 88000
    },
    {
      "epoch": 9.487870619946092,
      "eval_accuracy": 0.3944075924030335,
      "eval_loss": 3.3036043643951416,
      "eval_runtime": 183.5422,
      "eval_samples_per_second": 98.13,
      "eval_steps_per_second": 6.135,
      "step": 88000
    },
    {
      "epoch": 9.493261455525607,
      "grad_norm": 0.8956674933433533,
      "learning_rate": 3.066378845116028e-05,
      "loss": 3.1346,
      "step": 88050
    },
    {
      "epoch": 9.498652291105122,
      "grad_norm": 0.9043859839439392,
      "learning_rate": 3.033998920669185e-05,
      "loss": 3.13,
      "step": 88100
    },
    {
      "epoch": 9.504043126684635,
      "grad_norm": 0.8860067129135132,
      "learning_rate": 3.0016189962223416e-05,
      "loss": 3.1347,
      "step": 88150
    },
    {
      "epoch": 9.50943396226415,
      "grad_norm": 0.8624042272567749,
      "learning_rate": 2.9692390717754987e-05,
      "loss": 3.1483,
      "step": 88200
    },
    {
      "epoch": 9.514824797843666,
      "grad_norm": 0.9003132581710815,
      "learning_rate": 2.9368591473286558e-05,
      "loss": 3.1226,
      "step": 88250
    },
    {
      "epoch": 9.520215633423181,
      "grad_norm": 0.884479284286499,
      "learning_rate": 2.904479222881813e-05,
      "loss": 3.1117,
      "step": 88300
    },
    {
      "epoch": 9.525606469002696,
      "grad_norm": 0.9293926954269409,
      "learning_rate": 2.87209929843497e-05,
      "loss": 3.1243,
      "step": 88350
    },
    {
      "epoch": 9.530997304582211,
      "grad_norm": 0.9077318906784058,
      "learning_rate": 2.8397193739881274e-05,
      "loss": 3.1164,
      "step": 88400
    },
    {
      "epoch": 9.536388140161725,
      "grad_norm": 0.8652298450469971,
      "learning_rate": 2.8073394495412842e-05,
      "loss": 3.1231,
      "step": 88450
    },
    {
      "epoch": 9.54177897574124,
      "grad_norm": 0.8628005385398865,
      "learning_rate": 2.7749595250944413e-05,
      "loss": 3.1198,
      "step": 88500
    },
    {
      "epoch": 9.547169811320755,
      "grad_norm": 0.8354774117469788,
      "learning_rate": 2.7425796006475984e-05,
      "loss": 3.1267,
      "step": 88550
    },
    {
      "epoch": 9.55256064690027,
      "grad_norm": 0.8727371692657471,
      "learning_rate": 2.7101996762007554e-05,
      "loss": 3.1216,
      "step": 88600
    },
    {
      "epoch": 9.557951482479785,
      "grad_norm": 0.8466863036155701,
      "learning_rate": 2.6778197517539125e-05,
      "loss": 3.1325,
      "step": 88650
    },
    {
      "epoch": 9.563342318059298,
      "grad_norm": 0.8633363842964172,
      "learning_rate": 2.6454398273070693e-05,
      "loss": 3.1185,
      "step": 88700
    },
    {
      "epoch": 9.568733153638814,
      "grad_norm": 0.8681420087814331,
      "learning_rate": 2.6130599028602264e-05,
      "loss": 3.1405,
      "step": 88750
    },
    {
      "epoch": 9.574123989218329,
      "grad_norm": 0.9068518877029419,
      "learning_rate": 2.5806799784133835e-05,
      "loss": 3.1347,
      "step": 88800
    },
    {
      "epoch": 9.579514824797844,
      "grad_norm": 0.8506572842597961,
      "learning_rate": 2.5483000539665406e-05,
      "loss": 3.1212,
      "step": 88850
    },
    {
      "epoch": 9.584905660377359,
      "grad_norm": 0.8176720142364502,
      "learning_rate": 2.5159201295196976e-05,
      "loss": 3.1273,
      "step": 88900
    },
    {
      "epoch": 9.590296495956874,
      "grad_norm": 0.9253360629081726,
      "learning_rate": 2.4835402050728544e-05,
      "loss": 3.1155,
      "step": 88950
    },
    {
      "epoch": 9.595687331536388,
      "grad_norm": 0.8876266479492188,
      "learning_rate": 2.4511602806260115e-05,
      "loss": 3.1404,
      "step": 89000
    },
    {
      "epoch": 9.595687331536388,
      "eval_accuracy": 0.39452298176989253,
      "eval_loss": 3.302065372467041,
      "eval_runtime": 183.5463,
      "eval_samples_per_second": 98.128,
      "eval_steps_per_second": 6.135,
      "step": 89000
    },
    {
      "epoch": 9.601078167115903,
      "grad_norm": 0.8948820233345032,
      "learning_rate": 2.4187803561791686e-05,
      "loss": 3.1014,
      "step": 89050
    },
    {
      "epoch": 9.606469002695418,
      "grad_norm": 0.8849186897277832,
      "learning_rate": 2.3864004317323257e-05,
      "loss": 3.1244,
      "step": 89100
    },
    {
      "epoch": 9.611859838274933,
      "grad_norm": 0.8476836085319519,
      "learning_rate": 2.354020507285483e-05,
      "loss": 3.1156,
      "step": 89150
    },
    {
      "epoch": 9.617250673854448,
      "grad_norm": 0.8894110321998596,
      "learning_rate": 2.3216405828386402e-05,
      "loss": 3.1006,
      "step": 89200
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.8986078500747681,
      "learning_rate": 2.289260658391797e-05,
      "loss": 3.1276,
      "step": 89250
    },
    {
      "epoch": 9.628032345013477,
      "grad_norm": 0.9119869470596313,
      "learning_rate": 2.256880733944954e-05,
      "loss": 3.1514,
      "step": 89300
    },
    {
      "epoch": 9.633423180592992,
      "grad_norm": 0.8732916712760925,
      "learning_rate": 2.224500809498111e-05,
      "loss": 3.1433,
      "step": 89350
    },
    {
      "epoch": 9.638814016172507,
      "grad_norm": 0.8498047590255737,
      "learning_rate": 2.1921208850512682e-05,
      "loss": 3.1164,
      "step": 89400
    },
    {
      "epoch": 9.644204851752022,
      "grad_norm": 0.8872281312942505,
      "learning_rate": 2.1597409606044253e-05,
      "loss": 3.1294,
      "step": 89450
    },
    {
      "epoch": 9.649595687331537,
      "grad_norm": 0.8698834776878357,
      "learning_rate": 2.127361036157582e-05,
      "loss": 3.1374,
      "step": 89500
    },
    {
      "epoch": 9.65498652291105,
      "grad_norm": 0.8387486338615417,
      "learning_rate": 2.094981111710739e-05,
      "loss": 3.1255,
      "step": 89550
    },
    {
      "epoch": 9.660377358490566,
      "grad_norm": 0.8434497714042664,
      "learning_rate": 2.0626011872638962e-05,
      "loss": 3.1261,
      "step": 89600
    },
    {
      "epoch": 9.66576819407008,
      "grad_norm": 0.8968543410301208,
      "learning_rate": 2.0302212628170533e-05,
      "loss": 3.1309,
      "step": 89650
    },
    {
      "epoch": 9.671159029649596,
      "grad_norm": 0.9121111035346985,
      "learning_rate": 1.99784133837021e-05,
      "loss": 3.1151,
      "step": 89700
    },
    {
      "epoch": 9.676549865229111,
      "grad_norm": 0.9186899065971375,
      "learning_rate": 1.965461413923367e-05,
      "loss": 3.1392,
      "step": 89750
    },
    {
      "epoch": 9.681940700808624,
      "grad_norm": 0.8917949199676514,
      "learning_rate": 1.9330814894765242e-05,
      "loss": 3.1306,
      "step": 89800
    },
    {
      "epoch": 9.68733153638814,
      "grad_norm": 0.8518571257591248,
      "learning_rate": 1.9007015650296813e-05,
      "loss": 3.1275,
      "step": 89850
    },
    {
      "epoch": 9.692722371967655,
      "grad_norm": 0.8881104588508606,
      "learning_rate": 1.8683216405828384e-05,
      "loss": 3.1199,
      "step": 89900
    },
    {
      "epoch": 9.69811320754717,
      "grad_norm": 0.8823761343955994,
      "learning_rate": 1.8359417161359955e-05,
      "loss": 3.1253,
      "step": 89950
    },
    {
      "epoch": 9.703504043126685,
      "grad_norm": 0.8819106817245483,
      "learning_rate": 1.8035617916891526e-05,
      "loss": 3.1261,
      "step": 90000
    },
    {
      "epoch": 9.703504043126685,
      "eval_accuracy": 0.3946354375087805,
      "eval_loss": 3.299715518951416,
      "eval_runtime": 183.4384,
      "eval_samples_per_second": 98.186,
      "eval_steps_per_second": 6.138,
      "step": 90000
    },
    {
      "epoch": 9.7088948787062,
      "grad_norm": 0.8763957023620605,
      "learning_rate": 1.7711818672423097e-05,
      "loss": 3.1182,
      "step": 90050
    },
    {
      "epoch": 9.714285714285714,
      "grad_norm": 0.8612681031227112,
      "learning_rate": 1.7388019427954664e-05,
      "loss": 3.1227,
      "step": 90100
    },
    {
      "epoch": 9.719676549865229,
      "grad_norm": 0.8781333565711975,
      "learning_rate": 1.706422018348624e-05,
      "loss": 3.1294,
      "step": 90150
    },
    {
      "epoch": 9.725067385444744,
      "grad_norm": 0.865515410900116,
      "learning_rate": 1.674042093901781e-05,
      "loss": 3.1297,
      "step": 90200
    },
    {
      "epoch": 9.730458221024259,
      "grad_norm": 0.8704222440719604,
      "learning_rate": 1.6423097679438745e-05,
      "loss": 3.1347,
      "step": 90250
    },
    {
      "epoch": 9.735849056603774,
      "grad_norm": 0.8565595746040344,
      "learning_rate": 1.6099298434970315e-05,
      "loss": 3.117,
      "step": 90300
    },
    {
      "epoch": 9.74123989218329,
      "grad_norm": 0.8548470735549927,
      "learning_rate": 1.577549919050189e-05,
      "loss": 3.1338,
      "step": 90350
    },
    {
      "epoch": 9.746630727762803,
      "grad_norm": 0.8468580842018127,
      "learning_rate": 1.5451699946033457e-05,
      "loss": 3.1301,
      "step": 90400
    },
    {
      "epoch": 9.752021563342318,
      "grad_norm": 0.9051481485366821,
      "learning_rate": 1.5127900701565028e-05,
      "loss": 3.1297,
      "step": 90450
    },
    {
      "epoch": 9.757412398921833,
      "grad_norm": 0.8959400057792664,
      "learning_rate": 1.4804101457096599e-05,
      "loss": 3.1377,
      "step": 90500
    },
    {
      "epoch": 9.762803234501348,
      "grad_norm": 0.9014003872871399,
      "learning_rate": 1.448030221262817e-05,
      "loss": 3.1243,
      "step": 90550
    },
    {
      "epoch": 9.768194070080863,
      "grad_norm": 0.8818705677986145,
      "learning_rate": 1.4156502968159739e-05,
      "loss": 3.1397,
      "step": 90600
    },
    {
      "epoch": 9.773584905660378,
      "grad_norm": 0.8922802209854126,
      "learning_rate": 1.383270372369131e-05,
      "loss": 3.1233,
      "step": 90650
    },
    {
      "epoch": 9.778975741239892,
      "grad_norm": 0.9080004096031189,
      "learning_rate": 1.3508904479222879e-05,
      "loss": 3.1246,
      "step": 90700
    },
    {
      "epoch": 9.784366576819407,
      "grad_norm": 0.9191465973854065,
      "learning_rate": 1.3185105234754452e-05,
      "loss": 3.1084,
      "step": 90750
    },
    {
      "epoch": 9.789757412398922,
      "grad_norm": 0.9019896984100342,
      "learning_rate": 1.2861305990286023e-05,
      "loss": 3.1193,
      "step": 90800
    },
    {
      "epoch": 9.795148247978437,
      "grad_norm": 0.8658702373504639,
      "learning_rate": 1.2537506745817592e-05,
      "loss": 3.1277,
      "step": 90850
    },
    {
      "epoch": 9.800539083557952,
      "grad_norm": 0.8594561219215393,
      "learning_rate": 1.2213707501349163e-05,
      "loss": 3.1363,
      "step": 90900
    },
    {
      "epoch": 9.805929919137466,
      "grad_norm": 0.8785389065742493,
      "learning_rate": 1.1889908256880732e-05,
      "loss": 3.1238,
      "step": 90950
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.8855112195014954,
      "learning_rate": 1.1566109012412303e-05,
      "loss": 3.1272,
      "step": 91000
    },
    {
      "epoch": 9.81132075471698,
      "eval_accuracy": 0.3948550249950536,
      "eval_loss": 3.297788381576538,
      "eval_runtime": 183.5523,
      "eval_samples_per_second": 98.125,
      "eval_steps_per_second": 6.134,
      "step": 91000
    },
    {
      "epoch": 9.816711590296496,
      "grad_norm": 0.868584394454956,
      "learning_rate": 1.1242309767943874e-05,
      "loss": 3.1144,
      "step": 91050
    },
    {
      "epoch": 9.822102425876011,
      "grad_norm": 0.8530396819114685,
      "learning_rate": 1.0918510523475443e-05,
      "loss": 3.1235,
      "step": 91100
    },
    {
      "epoch": 9.827493261455526,
      "grad_norm": 0.8684161901473999,
      "learning_rate": 1.0594711279007015e-05,
      "loss": 3.1208,
      "step": 91150
    },
    {
      "epoch": 9.832884097035041,
      "grad_norm": 0.8899792432785034,
      "learning_rate": 1.0270912034538586e-05,
      "loss": 3.1213,
      "step": 91200
    },
    {
      "epoch": 9.838274932614555,
      "grad_norm": 0.8613511919975281,
      "learning_rate": 9.947112790070156e-06,
      "loss": 3.1172,
      "step": 91250
    },
    {
      "epoch": 9.84366576819407,
      "grad_norm": 0.8709462881088257,
      "learning_rate": 9.623313545601726e-06,
      "loss": 3.1474,
      "step": 91300
    },
    {
      "epoch": 9.849056603773585,
      "grad_norm": 0.896946132183075,
      "learning_rate": 9.299514301133296e-06,
      "loss": 3.1258,
      "step": 91350
    },
    {
      "epoch": 9.8544474393531,
      "grad_norm": 0.8559094071388245,
      "learning_rate": 8.975715056664867e-06,
      "loss": 3.1215,
      "step": 91400
    },
    {
      "epoch": 9.859838274932615,
      "grad_norm": 0.8676149249076843,
      "learning_rate": 8.651915812196437e-06,
      "loss": 3.1358,
      "step": 91450
    },
    {
      "epoch": 9.865229110512129,
      "grad_norm": 0.8665869235992432,
      "learning_rate": 8.328116567728008e-06,
      "loss": 3.0967,
      "step": 91500
    },
    {
      "epoch": 9.870619946091644,
      "grad_norm": 0.8311969637870789,
      "learning_rate": 8.004317323259577e-06,
      "loss": 3.0991,
      "step": 91550
    },
    {
      "epoch": 9.876010781671159,
      "grad_norm": 0.8719154000282288,
      "learning_rate": 7.680518078791148e-06,
      "loss": 3.12,
      "step": 91600
    },
    {
      "epoch": 9.881401617250674,
      "grad_norm": 0.8440600633621216,
      "learning_rate": 7.356718834322719e-06,
      "loss": 3.1188,
      "step": 91650
    },
    {
      "epoch": 9.88679245283019,
      "grad_norm": 0.8644077777862549,
      "learning_rate": 7.03291958985429e-06,
      "loss": 3.1107,
      "step": 91700
    },
    {
      "epoch": 9.892183288409704,
      "grad_norm": 0.8675840497016907,
      "learning_rate": 6.70912034538586e-06,
      "loss": 3.1316,
      "step": 91750
    },
    {
      "epoch": 9.897574123989218,
      "grad_norm": 0.873511552810669,
      "learning_rate": 6.38532110091743e-06,
      "loss": 3.1241,
      "step": 91800
    },
    {
      "epoch": 9.902964959568733,
      "grad_norm": 0.9048805236816406,
      "learning_rate": 6.061521856449001e-06,
      "loss": 3.1268,
      "step": 91850
    },
    {
      "epoch": 9.908355795148248,
      "grad_norm": 0.8429412245750427,
      "learning_rate": 5.737722611980571e-06,
      "loss": 3.1232,
      "step": 91900
    },
    {
      "epoch": 9.913746630727763,
      "grad_norm": 0.8978517651557922,
      "learning_rate": 5.413923367512142e-06,
      "loss": 3.1066,
      "step": 91950
    },
    {
      "epoch": 9.919137466307278,
      "grad_norm": 0.8608512282371521,
      "learning_rate": 5.090124123043712e-06,
      "loss": 3.0993,
      "step": 92000
    },
    {
      "epoch": 9.919137466307278,
      "eval_accuracy": 0.39501604857479466,
      "eval_loss": 3.2966372966766357,
      "eval_runtime": 183.5216,
      "eval_samples_per_second": 98.141,
      "eval_steps_per_second": 6.136,
      "step": 92000
    },
    {
      "epoch": 9.924528301886792,
      "grad_norm": 0.8535104990005493,
      "learning_rate": 4.766324878575283e-06,
      "loss": 3.1249,
      "step": 92050
    },
    {
      "epoch": 9.929919137466307,
      "grad_norm": 0.9014983177185059,
      "learning_rate": 4.442525634106853e-06,
      "loss": 3.1135,
      "step": 92100
    },
    {
      "epoch": 9.935309973045822,
      "grad_norm": 0.8669118285179138,
      "learning_rate": 4.118726389638424e-06,
      "loss": 3.1419,
      "step": 92150
    },
    {
      "epoch": 9.940700808625337,
      "grad_norm": 0.8581469058990479,
      "learning_rate": 3.7949271451699944e-06,
      "loss": 3.1323,
      "step": 92200
    },
    {
      "epoch": 9.946091644204852,
      "grad_norm": 0.8778931498527527,
      "learning_rate": 3.4711279007015644e-06,
      "loss": 3.124,
      "step": 92250
    },
    {
      "epoch": 9.951482479784367,
      "grad_norm": 0.9205542206764221,
      "learning_rate": 3.1473286562331353e-06,
      "loss": 3.1097,
      "step": 92300
    },
    {
      "epoch": 9.95687331536388,
      "grad_norm": 0.9244022965431213,
      "learning_rate": 2.8235294117647054e-06,
      "loss": 3.1215,
      "step": 92350
    },
    {
      "epoch": 9.962264150943396,
      "grad_norm": 0.858595073223114,
      "learning_rate": 2.4997301672962763e-06,
      "loss": 3.1055,
      "step": 92400
    },
    {
      "epoch": 9.967654986522911,
      "grad_norm": 0.8582491874694824,
      "learning_rate": 2.1759309228278467e-06,
      "loss": 3.1251,
      "step": 92450
    },
    {
      "epoch": 9.973045822102426,
      "grad_norm": 0.8598997592926025,
      "learning_rate": 1.852131678359417e-06,
      "loss": 3.12,
      "step": 92500
    },
    {
      "epoch": 9.978436657681941,
      "grad_norm": 0.8376339077949524,
      "learning_rate": 1.5283324338909875e-06,
      "loss": 3.1245,
      "step": 92550
    },
    {
      "epoch": 9.983827493261456,
      "grad_norm": 0.8226578831672668,
      "learning_rate": 1.204533189422558e-06,
      "loss": 3.1198,
      "step": 92600
    },
    {
      "epoch": 9.98921832884097,
      "grad_norm": 0.8843859434127808,
      "learning_rate": 8.807339449541284e-07,
      "loss": 3.1088,
      "step": 92650
    },
    {
      "epoch": 9.994609164420485,
      "grad_norm": 0.886563777923584,
      "learning_rate": 5.569347004856989e-07,
      "loss": 3.131,
      "step": 92700
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.917934536933899,
      "learning_rate": 2.3313545601726927e-07,
      "loss": 3.1285,
      "step": 92750
    },
    {
      "epoch": 10.0,
      "step": 92750,
      "total_flos": 7.75449427968e+17,
      "train_loss": 3.4498943302123695,
      "train_runtime": 79544.4067,
      "train_samples_per_second": 37.309,
      "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
}