{
  "best_metric": 3.3063955307006836,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_500_495/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": 1.9115898609161377,
      "learning_rate": 0.0003,
      "loss": 8.518,
      "step": 50
    },
    {
      "epoch": 0.01078167115902965,
      "grad_norm": 5.626335144042969,
      "learning_rate": 0.0006,
      "loss": 6.8964,
      "step": 100
    },
    {
      "epoch": 0.016172506738544475,
      "grad_norm": 1.5297762155532837,
      "learning_rate": 0.0005996762007555315,
      "loss": 6.4903,
      "step": 150
    },
    {
      "epoch": 0.0215633423180593,
      "grad_norm": 1.232740044593811,
      "learning_rate": 0.000599352401511063,
      "loss": 6.2567,
      "step": 200
    },
    {
      "epoch": 0.026954177897574125,
      "grad_norm": 0.7754726409912109,
      "learning_rate": 0.0005990286022665946,
      "loss": 6.0728,
      "step": 250
    },
    {
      "epoch": 0.03234501347708895,
      "grad_norm": 2.1544833183288574,
      "learning_rate": 0.0005987048030221263,
      "loss": 5.9666,
      "step": 300
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 0.9812532663345337,
      "learning_rate": 0.0005983810037776578,
      "loss": 5.8668,
      "step": 350
    },
    {
      "epoch": 0.0431266846361186,
      "grad_norm": 1.3231236934661865,
      "learning_rate": 0.0005980572045331894,
      "loss": 5.809,
      "step": 400
    },
    {
      "epoch": 0.04851752021563342,
      "grad_norm": 1.8271440267562866,
      "learning_rate": 0.0005977334052887209,
      "loss": 5.7373,
      "step": 450
    },
    {
      "epoch": 0.05390835579514825,
      "grad_norm": 1.6826151609420776,
      "learning_rate": 0.0005974096060442526,
      "loss": 5.6784,
      "step": 500
    },
    {
      "epoch": 0.05929919137466307,
      "grad_norm": 2.017390012741089,
      "learning_rate": 0.0005970858067997841,
      "loss": 5.566,
      "step": 550
    },
    {
      "epoch": 0.0646900269541779,
      "grad_norm": 2.0407137870788574,
      "learning_rate": 0.0005967620075553157,
      "loss": 5.5348,
      "step": 600
    },
    {
      "epoch": 0.07008086253369272,
      "grad_norm": 0.9982050061225891,
      "learning_rate": 0.0005964382083108472,
      "loss": 5.4331,
      "step": 650
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 1.1897433996200562,
      "learning_rate": 0.0005961144090663788,
      "loss": 5.4115,
      "step": 700
    },
    {
      "epoch": 0.08086253369272237,
      "grad_norm": 1.7157807350158691,
      "learning_rate": 0.0005957906098219104,
      "loss": 5.3354,
      "step": 750
    },
    {
      "epoch": 0.0862533692722372,
      "grad_norm": 1.4302008152008057,
      "learning_rate": 0.0005954668105774419,
      "loss": 5.2837,
      "step": 800
    },
    {
      "epoch": 0.09164420485175202,
      "grad_norm": 1.2498868703842163,
      "learning_rate": 0.0005951430113329735,
      "loss": 5.2505,
      "step": 850
    },
    {
      "epoch": 0.09703504043126684,
      "grad_norm": 0.8495532274246216,
      "learning_rate": 0.0005948192120885051,
      "loss": 5.1733,
      "step": 900
    },
    {
      "epoch": 0.10242587601078167,
      "grad_norm": 0.9718948602676392,
      "learning_rate": 0.0005944954128440366,
      "loss": 5.161,
      "step": 950
    },
    {
      "epoch": 0.1078167115902965,
      "grad_norm": 1.1483657360076904,
      "learning_rate": 0.0005941716135995682,
      "loss": 5.1041,
      "step": 1000
    },
    {
      "epoch": 0.1078167115902965,
      "eval_accuracy": 0.22633993729207233,
      "eval_loss": 5.030312538146973,
      "eval_runtime": 188.3882,
      "eval_samples_per_second": 95.606,
      "eval_steps_per_second": 5.977,
      "step": 1000
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 1.1805005073547363,
      "learning_rate": 0.0005938478143550997,
      "loss": 5.0661,
      "step": 1050
    },
    {
      "epoch": 0.11859838274932614,
      "grad_norm": 1.1717469692230225,
      "learning_rate": 0.0005935240151106314,
      "loss": 5.0246,
      "step": 1100
    },
    {
      "epoch": 0.12398921832884097,
      "grad_norm": 0.9578151702880859,
      "learning_rate": 0.0005932002158661629,
      "loss": 5.0052,
      "step": 1150
    },
    {
      "epoch": 0.1293800539083558,
      "grad_norm": 1.1234848499298096,
      "learning_rate": 0.0005928764166216945,
      "loss": 4.9837,
      "step": 1200
    },
    {
      "epoch": 0.1347708894878706,
      "grad_norm": 1.1127243041992188,
      "learning_rate": 0.000592552617377226,
      "loss": 4.9561,
      "step": 1250
    },
    {
      "epoch": 0.14016172506738545,
      "grad_norm": 1.002482295036316,
      "learning_rate": 0.0005922288181327577,
      "loss": 4.902,
      "step": 1300
    },
    {
      "epoch": 0.14555256064690028,
      "grad_norm": 1.1132245063781738,
      "learning_rate": 0.0005919050188882893,
      "loss": 4.8863,
      "step": 1350
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 0.8769996762275696,
      "learning_rate": 0.0005915812196438207,
      "loss": 4.8603,
      "step": 1400
    },
    {
      "epoch": 0.15633423180592992,
      "grad_norm": 1.0863553285598755,
      "learning_rate": 0.0005912574203993524,
      "loss": 4.838,
      "step": 1450
    },
    {
      "epoch": 0.16172506738544473,
      "grad_norm": 0.9288261532783508,
      "learning_rate": 0.0005909336211548839,
      "loss": 4.8231,
      "step": 1500
    },
    {
      "epoch": 0.16711590296495957,
      "grad_norm": 1.0346804857254028,
      "learning_rate": 0.0005906098219104155,
      "loss": 4.7805,
      "step": 1550
    },
    {
      "epoch": 0.1725067385444744,
      "grad_norm": 0.972179651260376,
      "learning_rate": 0.000590286022665947,
      "loss": 4.7959,
      "step": 1600
    },
    {
      "epoch": 0.1778975741239892,
      "grad_norm": 1.0101567506790161,
      "learning_rate": 0.0005899622234214787,
      "loss": 4.7731,
      "step": 1650
    },
    {
      "epoch": 0.18328840970350405,
      "grad_norm": 0.979967474937439,
      "learning_rate": 0.0005896384241770102,
      "loss": 4.6941,
      "step": 1700
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 1.1450660228729248,
      "learning_rate": 0.0005893146249325418,
      "loss": 4.7073,
      "step": 1750
    },
    {
      "epoch": 0.1940700808625337,
      "grad_norm": 0.8868092894554138,
      "learning_rate": 0.0005889908256880733,
      "loss": 4.6979,
      "step": 1800
    },
    {
      "epoch": 0.19946091644204852,
      "grad_norm": 0.8870577216148376,
      "learning_rate": 0.0005886670264436049,
      "loss": 4.6546,
      "step": 1850
    },
    {
      "epoch": 0.20485175202156333,
      "grad_norm": 0.9839901924133301,
      "learning_rate": 0.0005883432271991365,
      "loss": 4.6524,
      "step": 1900
    },
    {
      "epoch": 0.21024258760107817,
      "grad_norm": 1.3535802364349365,
      "learning_rate": 0.0005880194279546681,
      "loss": 4.6183,
      "step": 1950
    },
    {
      "epoch": 0.215633423180593,
      "grad_norm": 1.1455411911010742,
      "learning_rate": 0.0005876956287101996,
      "loss": 4.5964,
      "step": 2000
    },
    {
      "epoch": 0.215633423180593,
      "eval_accuracy": 0.2674478881735786,
      "eval_loss": 4.539289474487305,
      "eval_runtime": 187.7214,
      "eval_samples_per_second": 95.945,
      "eval_steps_per_second": 5.998,
      "step": 2000
    },
    {
      "epoch": 0.2210242587601078,
      "grad_norm": 0.9703343510627747,
      "learning_rate": 0.0005873718294657312,
      "loss": 4.6083,
      "step": 2050
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 0.934937059879303,
      "learning_rate": 0.0005870480302212628,
      "loss": 4.575,
      "step": 2100
    },
    {
      "epoch": 0.23180592991913745,
      "grad_norm": 1.039948582649231,
      "learning_rate": 0.0005867242309767943,
      "loss": 4.5189,
      "step": 2150
    },
    {
      "epoch": 0.2371967654986523,
      "grad_norm": 0.926210343837738,
      "learning_rate": 0.0005864004317323259,
      "loss": 4.5622,
      "step": 2200
    },
    {
      "epoch": 0.24258760107816713,
      "grad_norm": 0.7271756529808044,
      "learning_rate": 0.0005860766324878575,
      "loss": 4.4915,
      "step": 2250
    },
    {
      "epoch": 0.24797843665768193,
      "grad_norm": 1.1152825355529785,
      "learning_rate": 0.000585752833243389,
      "loss": 4.4756,
      "step": 2300
    },
    {
      "epoch": 0.25336927223719674,
      "grad_norm": 0.8933312296867371,
      "learning_rate": 0.0005854290339989206,
      "loss": 4.4727,
      "step": 2350
    },
    {
      "epoch": 0.2587601078167116,
      "grad_norm": 1.0869262218475342,
      "learning_rate": 0.0005851052347544521,
      "loss": 4.4607,
      "step": 2400
    },
    {
      "epoch": 0.2641509433962264,
      "grad_norm": 0.9118139147758484,
      "learning_rate": 0.0005847814355099838,
      "loss": 4.4666,
      "step": 2450
    },
    {
      "epoch": 0.2695417789757412,
      "grad_norm": 0.7588400840759277,
      "learning_rate": 0.0005844576362655154,
      "loss": 4.4622,
      "step": 2500
    },
    {
      "epoch": 0.2749326145552561,
      "grad_norm": 0.8877196311950684,
      "learning_rate": 0.0005841338370210469,
      "loss": 4.4204,
      "step": 2550
    },
    {
      "epoch": 0.2803234501347709,
      "grad_norm": 0.8958740234375,
      "learning_rate": 0.0005838100377765785,
      "loss": 4.4112,
      "step": 2600
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 1.173532247543335,
      "learning_rate": 0.0005834862385321101,
      "loss": 4.4104,
      "step": 2650
    },
    {
      "epoch": 0.29110512129380056,
      "grad_norm": 0.862187385559082,
      "learning_rate": 0.0005831624392876417,
      "loss": 4.4036,
      "step": 2700
    },
    {
      "epoch": 0.29649595687331537,
      "grad_norm": 0.8075447082519531,
      "learning_rate": 0.0005828386400431731,
      "loss": 4.3633,
      "step": 2750
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 0.8274509906768799,
      "learning_rate": 0.0005825148407987048,
      "loss": 4.4045,
      "step": 2800
    },
    {
      "epoch": 0.30727762803234504,
      "grad_norm": 0.7691844701766968,
      "learning_rate": 0.0005821910415542363,
      "loss": 4.3696,
      "step": 2850
    },
    {
      "epoch": 0.31266846361185985,
      "grad_norm": 0.7903671860694885,
      "learning_rate": 0.0005818672423097679,
      "loss": 4.3474,
      "step": 2900
    },
    {
      "epoch": 0.31805929919137466,
      "grad_norm": 0.871162474155426,
      "learning_rate": 0.0005815434430652994,
      "loss": 4.3491,
      "step": 2950
    },
    {
      "epoch": 0.32345013477088946,
      "grad_norm": 0.6972927451133728,
      "learning_rate": 0.0005812196438208311,
      "loss": 4.3455,
      "step": 3000
    },
    {
      "epoch": 0.32345013477088946,
      "eval_accuracy": 0.29693041467048675,
      "eval_loss": 4.249656677246094,
      "eval_runtime": 187.5787,
      "eval_samples_per_second": 96.018,
      "eval_steps_per_second": 6.003,
      "step": 3000
    },
    {
      "epoch": 0.3288409703504043,
      "grad_norm": 0.8769593834877014,
      "learning_rate": 0.0005808958445763626,
      "loss": 4.325,
      "step": 3050
    },
    {
      "epoch": 0.33423180592991913,
      "grad_norm": 0.7991501092910767,
      "learning_rate": 0.0005805720453318942,
      "loss": 4.3205,
      "step": 3100
    },
    {
      "epoch": 0.33962264150943394,
      "grad_norm": 0.7523728609085083,
      "learning_rate": 0.0005802482460874257,
      "loss": 4.2834,
      "step": 3150
    },
    {
      "epoch": 0.3450134770889488,
      "grad_norm": 0.854878842830658,
      "learning_rate": 0.0005799244468429573,
      "loss": 4.2919,
      "step": 3200
    },
    {
      "epoch": 0.3504043126684636,
      "grad_norm": 0.7155695557594299,
      "learning_rate": 0.0005796006475984889,
      "loss": 4.2661,
      "step": 3250
    },
    {
      "epoch": 0.3557951482479784,
      "grad_norm": 0.7300028204917908,
      "learning_rate": 0.0005792768483540205,
      "loss": 4.2884,
      "step": 3300
    },
    {
      "epoch": 0.3611859838274933,
      "grad_norm": 0.6207231879234314,
      "learning_rate": 0.000578953049109552,
      "loss": 4.2695,
      "step": 3350
    },
    {
      "epoch": 0.3665768194070081,
      "grad_norm": 0.8405777215957642,
      "learning_rate": 0.0005786292498650836,
      "loss": 4.2635,
      "step": 3400
    },
    {
      "epoch": 0.3719676549865229,
      "grad_norm": 0.6905649304389954,
      "learning_rate": 0.0005783054506206152,
      "loss": 4.2467,
      "step": 3450
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 0.7021420001983643,
      "learning_rate": 0.0005779816513761467,
      "loss": 4.226,
      "step": 3500
    },
    {
      "epoch": 0.38274932614555257,
      "grad_norm": 0.7447681427001953,
      "learning_rate": 0.0005776578521316782,
      "loss": 4.2315,
      "step": 3550
    },
    {
      "epoch": 0.3881401617250674,
      "grad_norm": 0.8126360177993774,
      "learning_rate": 0.0005773340528872099,
      "loss": 4.2338,
      "step": 3600
    },
    {
      "epoch": 0.3935309973045822,
      "grad_norm": 0.821212649345398,
      "learning_rate": 0.0005770102536427414,
      "loss": 4.2432,
      "step": 3650
    },
    {
      "epoch": 0.39892183288409705,
      "grad_norm": 0.7117862105369568,
      "learning_rate": 0.000576686454398273,
      "loss": 4.2093,
      "step": 3700
    },
    {
      "epoch": 0.40431266846361186,
      "grad_norm": 0.7709187269210815,
      "learning_rate": 0.0005763626551538045,
      "loss": 4.2057,
      "step": 3750
    },
    {
      "epoch": 0.40970350404312667,
      "grad_norm": 0.648059606552124,
      "learning_rate": 0.0005760388559093362,
      "loss": 4.2246,
      "step": 3800
    },
    {
      "epoch": 0.41509433962264153,
      "grad_norm": 0.7581032514572144,
      "learning_rate": 0.0005757150566648678,
      "loss": 4.2061,
      "step": 3850
    },
    {
      "epoch": 0.42048517520215634,
      "grad_norm": 1.0342339277267456,
      "learning_rate": 0.0005753912574203993,
      "loss": 4.1752,
      "step": 3900
    },
    {
      "epoch": 0.42587601078167114,
      "grad_norm": 0.6775522828102112,
      "learning_rate": 0.0005750674581759309,
      "loss": 4.1718,
      "step": 3950
    },
    {
      "epoch": 0.431266846361186,
      "grad_norm": 0.7176401615142822,
      "learning_rate": 0.0005747436589314624,
      "loss": 4.1807,
      "step": 4000
    },
    {
      "epoch": 0.431266846361186,
      "eval_accuracy": 0.31078854724678473,
      "eval_loss": 4.110689163208008,
      "eval_runtime": 187.6754,
      "eval_samples_per_second": 95.969,
      "eval_steps_per_second": 6.0,
      "step": 4000
    },
    {
      "epoch": 0.4366576819407008,
      "grad_norm": 0.6973474621772766,
      "learning_rate": 0.0005744198596869941,
      "loss": 4.175,
      "step": 4050
    },
    {
      "epoch": 0.4420485175202156,
      "grad_norm": 0.6163977384567261,
      "learning_rate": 0.0005740960604425255,
      "loss": 4.1802,
      "step": 4100
    },
    {
      "epoch": 0.4474393530997305,
      "grad_norm": 0.701683759689331,
      "learning_rate": 0.0005737722611980572,
      "loss": 4.1702,
      "step": 4150
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 0.648784339427948,
      "learning_rate": 0.0005734484619535887,
      "loss": 4.1583,
      "step": 4200
    },
    {
      "epoch": 0.4582210242587601,
      "grad_norm": 0.6678009629249573,
      "learning_rate": 0.0005731246627091203,
      "loss": 4.1472,
      "step": 4250
    },
    {
      "epoch": 0.4636118598382749,
      "grad_norm": 0.7954819202423096,
      "learning_rate": 0.0005728008634646518,
      "loss": 4.1426,
      "step": 4300
    },
    {
      "epoch": 0.46900269541778977,
      "grad_norm": 0.6935424208641052,
      "learning_rate": 0.0005724770642201835,
      "loss": 4.1495,
      "step": 4350
    },
    {
      "epoch": 0.4743935309973046,
      "grad_norm": 0.7789283394813538,
      "learning_rate": 0.000572153264975715,
      "loss": 4.1333,
      "step": 4400
    },
    {
      "epoch": 0.4797843665768194,
      "grad_norm": 0.7601833939552307,
      "learning_rate": 0.0005718294657312466,
      "loss": 4.1398,
      "step": 4450
    },
    {
      "epoch": 0.48517520215633425,
      "grad_norm": 0.68370121717453,
      "learning_rate": 0.0005715056664867781,
      "loss": 4.1212,
      "step": 4500
    },
    {
      "epoch": 0.49056603773584906,
      "grad_norm": 0.6254522204399109,
      "learning_rate": 0.0005711818672423097,
      "loss": 4.1102,
      "step": 4550
    },
    {
      "epoch": 0.49595687331536387,
      "grad_norm": 0.6852529644966125,
      "learning_rate": 0.0005708580679978413,
      "loss": 4.1179,
      "step": 4600
    },
    {
      "epoch": 0.5013477088948787,
      "grad_norm": 0.7181583046913147,
      "learning_rate": 0.0005705342687533729,
      "loss": 4.0984,
      "step": 4650
    },
    {
      "epoch": 0.5067385444743935,
      "grad_norm": 0.7251303791999817,
      "learning_rate": 0.0005702104695089044,
      "loss": 4.1042,
      "step": 4700
    },
    {
      "epoch": 0.5121293800539084,
      "grad_norm": 0.6099450588226318,
      "learning_rate": 0.000569886670264436,
      "loss": 4.0865,
      "step": 4750
    },
    {
      "epoch": 0.5175202156334232,
      "grad_norm": 0.5896744132041931,
      "learning_rate": 0.0005695628710199675,
      "loss": 4.1034,
      "step": 4800
    },
    {
      "epoch": 0.522911051212938,
      "grad_norm": 0.6033713817596436,
      "learning_rate": 0.0005692390717754991,
      "loss": 4.095,
      "step": 4850
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 0.5704399347305298,
      "learning_rate": 0.0005689152725310306,
      "loss": 4.0999,
      "step": 4900
    },
    {
      "epoch": 0.5336927223719676,
      "grad_norm": 0.671288013458252,
      "learning_rate": 0.0005685914732865623,
      "loss": 4.0748,
      "step": 4950
    },
    {
      "epoch": 0.5390835579514824,
      "grad_norm": 0.6648839712142944,
      "learning_rate": 0.0005682676740420939,
      "loss": 4.0654,
      "step": 5000
    },
    {
      "epoch": 0.5390835579514824,
      "eval_accuracy": 0.3200598981639944,
      "eval_loss": 4.007607460021973,
      "eval_runtime": 187.3054,
      "eval_samples_per_second": 96.158,
      "eval_steps_per_second": 6.012,
      "step": 5000
    },
    {
      "epoch": 0.5444743935309974,
      "grad_norm": 0.5957515239715576,
      "learning_rate": 0.0005679438747976254,
      "loss": 4.0606,
      "step": 5050
    },
    {
      "epoch": 0.5498652291105122,
      "grad_norm": 0.7439665794372559,
      "learning_rate": 0.000567620075553157,
      "loss": 4.0673,
      "step": 5100
    },
    {
      "epoch": 0.555256064690027,
      "grad_norm": 0.7030623555183411,
      "learning_rate": 0.0005672962763086886,
      "loss": 4.0702,
      "step": 5150
    },
    {
      "epoch": 0.5606469002695418,
      "grad_norm": 0.6474356651306152,
      "learning_rate": 0.0005669724770642202,
      "loss": 4.0454,
      "step": 5200
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.6153950095176697,
      "learning_rate": 0.0005666486778197517,
      "loss": 4.0458,
      "step": 5250
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.6132007837295532,
      "learning_rate": 0.0005663248785752833,
      "loss": 4.0315,
      "step": 5300
    },
    {
      "epoch": 0.5768194070080862,
      "grad_norm": 0.6719494462013245,
      "learning_rate": 0.0005660010793308148,
      "loss": 4.0466,
      "step": 5350
    },
    {
      "epoch": 0.5822102425876011,
      "grad_norm": 0.6127015352249146,
      "learning_rate": 0.0005656772800863465,
      "loss": 4.0351,
      "step": 5400
    },
    {
      "epoch": 0.5876010781671159,
      "grad_norm": 0.6278625130653381,
      "learning_rate": 0.0005653534808418779,
      "loss": 4.0478,
      "step": 5450
    },
    {
      "epoch": 0.5929919137466307,
      "grad_norm": 0.6481402516365051,
      "learning_rate": 0.0005650296815974096,
      "loss": 4.0487,
      "step": 5500
    },
    {
      "epoch": 0.5983827493261455,
      "grad_norm": 0.6136019825935364,
      "learning_rate": 0.0005647058823529411,
      "loss": 4.0517,
      "step": 5550
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.6584049463272095,
      "learning_rate": 0.0005643820831084727,
      "loss": 4.0275,
      "step": 5600
    },
    {
      "epoch": 0.6091644204851752,
      "grad_norm": 0.6200391054153442,
      "learning_rate": 0.0005640582838640042,
      "loss": 4.0176,
      "step": 5650
    },
    {
      "epoch": 0.6145552560646901,
      "grad_norm": 0.6417002081871033,
      "learning_rate": 0.0005637344846195358,
      "loss": 4.0226,
      "step": 5700
    },
    {
      "epoch": 0.6199460916442049,
      "grad_norm": 0.6000820994377136,
      "learning_rate": 0.0005634106853750674,
      "loss": 4.0338,
      "step": 5750
    },
    {
      "epoch": 0.6253369272237197,
      "grad_norm": 0.6211546659469604,
      "learning_rate": 0.000563086886130599,
      "loss": 4.0222,
      "step": 5800
    },
    {
      "epoch": 0.6307277628032345,
      "grad_norm": 0.6056830286979675,
      "learning_rate": 0.0005627630868861305,
      "loss": 4.0358,
      "step": 5850
    },
    {
      "epoch": 0.6361185983827493,
      "grad_norm": 0.5934694409370422,
      "learning_rate": 0.0005624392876416621,
      "loss": 4.0042,
      "step": 5900
    },
    {
      "epoch": 0.6415094339622641,
      "grad_norm": 0.6150168180465698,
      "learning_rate": 0.0005621154883971937,
      "loss": 3.9998,
      "step": 5950
    },
    {
      "epoch": 0.6469002695417789,
      "grad_norm": 0.5978230237960815,
      "learning_rate": 0.0005617916891527253,
      "loss": 4.0111,
      "step": 6000
    },
    {
      "epoch": 0.6469002695417789,
      "eval_accuracy": 0.32617988072303283,
      "eval_loss": 3.932387351989746,
      "eval_runtime": 187.2479,
      "eval_samples_per_second": 96.188,
      "eval_steps_per_second": 6.013,
      "step": 6000
    },
    {
      "epoch": 0.6522911051212938,
      "grad_norm": 0.6879520416259766,
      "learning_rate": 0.0005614678899082568,
      "loss": 3.9816,
      "step": 6050
    },
    {
      "epoch": 0.6576819407008087,
      "grad_norm": 0.7578017711639404,
      "learning_rate": 0.0005611440906637884,
      "loss": 4.0126,
      "step": 6100
    },
    {
      "epoch": 0.6630727762803235,
      "grad_norm": 0.544524073600769,
      "learning_rate": 0.00056082029141932,
      "loss": 3.9986,
      "step": 6150
    },
    {
      "epoch": 0.6684636118598383,
      "grad_norm": 0.6755046248435974,
      "learning_rate": 0.0005604964921748515,
      "loss": 3.9759,
      "step": 6200
    },
    {
      "epoch": 0.6738544474393531,
      "grad_norm": 0.5876584053039551,
      "learning_rate": 0.000560172692930383,
      "loss": 3.991,
      "step": 6250
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 0.6401435732841492,
      "learning_rate": 0.0005598488936859147,
      "loss": 4.0086,
      "step": 6300
    },
    {
      "epoch": 0.6846361185983828,
      "grad_norm": 0.6608089208602905,
      "learning_rate": 0.0005595250944414463,
      "loss": 3.9701,
      "step": 6350
    },
    {
      "epoch": 0.6900269541778976,
      "grad_norm": 0.5429288148880005,
      "learning_rate": 0.0005592012951969778,
      "loss": 3.9757,
      "step": 6400
    },
    {
      "epoch": 0.6954177897574124,
      "grad_norm": 0.6896840929985046,
      "learning_rate": 0.0005588839719373988,
      "loss": 3.9599,
      "step": 6450
    },
    {
      "epoch": 0.7008086253369272,
      "grad_norm": 0.6430425643920898,
      "learning_rate": 0.0005585601726929303,
      "loss": 3.9727,
      "step": 6500
    },
    {
      "epoch": 0.706199460916442,
      "grad_norm": 0.6070827841758728,
      "learning_rate": 0.0005582363734484619,
      "loss": 3.9942,
      "step": 6550
    },
    {
      "epoch": 0.7115902964959568,
      "grad_norm": 0.6078307032585144,
      "learning_rate": 0.0005579125742039935,
      "loss": 3.9502,
      "step": 6600
    },
    {
      "epoch": 0.7169811320754716,
      "grad_norm": 0.7253237962722778,
      "learning_rate": 0.0005575887749595251,
      "loss": 3.9852,
      "step": 6650
    },
    {
      "epoch": 0.7223719676549866,
      "grad_norm": 0.6382951736450195,
      "learning_rate": 0.0005572649757150566,
      "loss": 3.9578,
      "step": 6700
    },
    {
      "epoch": 0.7277628032345014,
      "grad_norm": 0.5518078804016113,
      "learning_rate": 0.0005569411764705882,
      "loss": 3.9632,
      "step": 6750
    },
    {
      "epoch": 0.7331536388140162,
      "grad_norm": 0.6217591166496277,
      "learning_rate": 0.0005566173772261198,
      "loss": 3.9522,
      "step": 6800
    },
    {
      "epoch": 0.738544474393531,
      "grad_norm": 0.761849582195282,
      "learning_rate": 0.0005562935779816513,
      "loss": 3.9322,
      "step": 6850
    },
    {
      "epoch": 0.7439353099730458,
      "grad_norm": 0.7228758335113525,
      "learning_rate": 0.0005559697787371828,
      "loss": 3.9589,
      "step": 6900
    },
    {
      "epoch": 0.7493261455525606,
      "grad_norm": 0.6004955768585205,
      "learning_rate": 0.0005556459794927145,
      "loss": 3.9557,
      "step": 6950
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.6903455257415771,
      "learning_rate": 0.000555322180248246,
      "loss": 3.9427,
      "step": 7000
    },
    {
      "epoch": 0.7547169811320755,
      "eval_accuracy": 0.33205235200362987,
      "eval_loss": 3.8775031566619873,
      "eval_runtime": 187.876,
      "eval_samples_per_second": 95.866,
      "eval_steps_per_second": 5.993,
      "step": 7000
    },
    {
      "epoch": 0.7601078167115903,
      "grad_norm": 0.6758421063423157,
      "learning_rate": 0.0005549983810037776,
      "loss": 3.9325,
      "step": 7050
    },
    {
      "epoch": 0.7654986522911051,
      "grad_norm": 0.5868051648139954,
      "learning_rate": 0.0005546745817593091,
      "loss": 3.9466,
      "step": 7100
    },
    {
      "epoch": 0.77088948787062,
      "grad_norm": 0.5360812544822693,
      "learning_rate": 0.0005543507825148408,
      "loss": 3.9455,
      "step": 7150
    },
    {
      "epoch": 0.7762803234501348,
      "grad_norm": 0.6635931134223938,
      "learning_rate": 0.0005540269832703723,
      "loss": 3.9448,
      "step": 7200
    },
    {
      "epoch": 0.7816711590296496,
      "grad_norm": 0.5423426628112793,
      "learning_rate": 0.0005537031840259039,
      "loss": 3.9006,
      "step": 7250
    },
    {
      "epoch": 0.7870619946091644,
      "grad_norm": 0.6123334169387817,
      "learning_rate": 0.0005533793847814354,
      "loss": 3.9403,
      "step": 7300
    },
    {
      "epoch": 0.7924528301886793,
      "grad_norm": 0.540355920791626,
      "learning_rate": 0.000553055585536967,
      "loss": 3.9251,
      "step": 7350
    },
    {
      "epoch": 0.7978436657681941,
      "grad_norm": 0.711990475654602,
      "learning_rate": 0.0005527317862924987,
      "loss": 3.9266,
      "step": 7400
    },
    {
      "epoch": 0.8032345013477089,
      "grad_norm": 0.5533379912376404,
      "learning_rate": 0.0005524079870480301,
      "loss": 3.9171,
      "step": 7450
    },
    {
      "epoch": 0.8086253369272237,
      "grad_norm": 0.5620821118354797,
      "learning_rate": 0.0005520841878035618,
      "loss": 3.9021,
      "step": 7500
    },
    {
      "epoch": 0.8140161725067385,
      "grad_norm": 0.6481254696846008,
      "learning_rate": 0.0005517603885590933,
      "loss": 3.9298,
      "step": 7550
    },
    {
      "epoch": 0.8194070080862533,
      "grad_norm": 0.5652502775192261,
      "learning_rate": 0.0005514365893146249,
      "loss": 3.9313,
      "step": 7600
    },
    {
      "epoch": 0.8247978436657682,
      "grad_norm": 0.5965263843536377,
      "learning_rate": 0.0005511127900701564,
      "loss": 3.9063,
      "step": 7650
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 0.5765863656997681,
      "learning_rate": 0.000550788990825688,
      "loss": 3.9115,
      "step": 7700
    },
    {
      "epoch": 0.8355795148247979,
      "grad_norm": 0.5533828139305115,
      "learning_rate": 0.0005504651915812196,
      "loss": 3.8928,
      "step": 7750
    },
    {
      "epoch": 0.8409703504043127,
      "grad_norm": 0.6085563898086548,
      "learning_rate": 0.0005501413923367512,
      "loss": 3.9219,
      "step": 7800
    },
    {
      "epoch": 0.8463611859838275,
      "grad_norm": 0.6730923652648926,
      "learning_rate": 0.0005498175930922827,
      "loss": 3.8986,
      "step": 7850
    },
    {
      "epoch": 0.8517520215633423,
      "grad_norm": 0.5520169138908386,
      "learning_rate": 0.0005494937938478143,
      "loss": 3.9013,
      "step": 7900
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.5863463878631592,
      "learning_rate": 0.0005491699946033459,
      "loss": 3.8983,
      "step": 7950
    },
    {
      "epoch": 0.862533692722372,
      "grad_norm": 0.6581799387931824,
      "learning_rate": 0.0005488461953588775,
      "loss": 3.8863,
      "step": 8000
    },
    {
      "epoch": 0.862533692722372,
      "eval_accuracy": 0.3363742379222265,
      "eval_loss": 3.829810619354248,
      "eval_runtime": 187.6184,
      "eval_samples_per_second": 95.998,
      "eval_steps_per_second": 6.002,
      "step": 8000
    },
    {
      "epoch": 0.8679245283018868,
      "grad_norm": 0.5409788489341736,
      "learning_rate": 0.000548522396114409,
      "loss": 3.8947,
      "step": 8050
    },
    {
      "epoch": 0.8733153638814016,
      "grad_norm": 0.6086958050727844,
      "learning_rate": 0.0005481985968699406,
      "loss": 3.8799,
      "step": 8100
    },
    {
      "epoch": 0.8787061994609164,
      "grad_norm": 0.5412994623184204,
      "learning_rate": 0.0005478747976254721,
      "loss": 3.8916,
      "step": 8150
    },
    {
      "epoch": 0.8840970350404312,
      "grad_norm": 0.5382531881332397,
      "learning_rate": 0.0005475509983810037,
      "loss": 3.8804,
      "step": 8200
    },
    {
      "epoch": 0.889487870619946,
      "grad_norm": 0.6551288962364197,
      "learning_rate": 0.0005472271991365352,
      "loss": 3.8687,
      "step": 8250
    },
    {
      "epoch": 0.894878706199461,
      "grad_norm": 0.6076605319976807,
      "learning_rate": 0.0005469033998920669,
      "loss": 3.891,
      "step": 8300
    },
    {
      "epoch": 0.9002695417789758,
      "grad_norm": 0.6250050663948059,
      "learning_rate": 0.0005465796006475984,
      "loss": 3.8736,
      "step": 8350
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 0.6367249488830566,
      "learning_rate": 0.00054625580140313,
      "loss": 3.8892,
      "step": 8400
    },
    {
      "epoch": 0.9110512129380054,
      "grad_norm": 0.625838041305542,
      "learning_rate": 0.0005459320021586615,
      "loss": 3.8984,
      "step": 8450
    },
    {
      "epoch": 0.9164420485175202,
      "grad_norm": 0.6105564832687378,
      "learning_rate": 0.0005456082029141932,
      "loss": 3.8909,
      "step": 8500
    },
    {
      "epoch": 0.921832884097035,
      "grad_norm": 0.5936092138290405,
      "learning_rate": 0.0005452844036697248,
      "loss": 3.8709,
      "step": 8550
    },
    {
      "epoch": 0.9272237196765498,
      "grad_norm": 0.5503196716308594,
      "learning_rate": 0.0005449606044252563,
      "loss": 3.8656,
      "step": 8600
    },
    {
      "epoch": 0.9326145552560647,
      "grad_norm": 0.5261377096176147,
      "learning_rate": 0.0005446368051807879,
      "loss": 3.8574,
      "step": 8650
    },
    {
      "epoch": 0.9380053908355795,
      "grad_norm": 0.541580319404602,
      "learning_rate": 0.0005443130059363194,
      "loss": 3.8643,
      "step": 8700
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.559482216835022,
      "learning_rate": 0.0005439892066918511,
      "loss": 3.8616,
      "step": 8750
    },
    {
      "epoch": 0.9487870619946092,
      "grad_norm": 0.5329921245574951,
      "learning_rate": 0.0005436654074473825,
      "loss": 3.854,
      "step": 8800
    },
    {
      "epoch": 0.954177897574124,
      "grad_norm": 0.5369682312011719,
      "learning_rate": 0.0005433480841878035,
      "loss": 3.8458,
      "step": 8850
    },
    {
      "epoch": 0.9595687331536388,
      "grad_norm": 0.5372455716133118,
      "learning_rate": 0.000543024284943335,
      "loss": 3.8639,
      "step": 8900
    },
    {
      "epoch": 0.9649595687331537,
      "grad_norm": 0.6307722926139832,
      "learning_rate": 0.0005427004856988667,
      "loss": 3.8671,
      "step": 8950
    },
    {
      "epoch": 0.9703504043126685,
      "grad_norm": 0.5844284296035767,
      "learning_rate": 0.0005423766864543982,
      "loss": 3.8554,
      "step": 9000
    },
    {
      "epoch": 0.9703504043126685,
      "eval_accuracy": 0.34037483725155565,
      "eval_loss": 3.7922003269195557,
      "eval_runtime": 187.0501,
      "eval_samples_per_second": 96.29,
      "eval_steps_per_second": 6.02,
      "step": 9000
    },
    {
      "epoch": 0.9757412398921833,
      "grad_norm": 0.5957661271095276,
      "learning_rate": 0.0005420528872099298,
      "loss": 3.8489,
      "step": 9050
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 0.5926182270050049,
      "learning_rate": 0.0005417290879654613,
      "loss": 3.8633,
      "step": 9100
    },
    {
      "epoch": 0.9865229110512129,
      "grad_norm": 0.548740565776825,
      "learning_rate": 0.000541405288720993,
      "loss": 3.8363,
      "step": 9150
    },
    {
      "epoch": 0.9919137466307277,
      "grad_norm": 0.5834896564483643,
      "learning_rate": 0.0005410814894765245,
      "loss": 3.8322,
      "step": 9200
    },
    {
      "epoch": 0.9973045822102425,
      "grad_norm": 0.5723360180854797,
      "learning_rate": 0.0005407576902320561,
      "loss": 3.8482,
      "step": 9250
    },
    {
      "epoch": 1.0026954177897573,
      "grad_norm": 0.5614904761314392,
      "learning_rate": 0.0005404338909875876,
      "loss": 3.8201,
      "step": 9300
    },
    {
      "epoch": 1.0080862533692723,
      "grad_norm": 0.6050702333450317,
      "learning_rate": 0.0005401100917431192,
      "loss": 3.7859,
      "step": 9350
    },
    {
      "epoch": 1.013477088948787,
      "grad_norm": 0.5851100087165833,
      "learning_rate": 0.0005397862924986508,
      "loss": 3.7818,
      "step": 9400
    },
    {
      "epoch": 1.0188679245283019,
      "grad_norm": 0.5718414783477783,
      "learning_rate": 0.0005394624932541824,
      "loss": 3.7795,
      "step": 9450
    },
    {
      "epoch": 1.0242587601078168,
      "grad_norm": 0.6160867810249329,
      "learning_rate": 0.0005391386940097139,
      "loss": 3.778,
      "step": 9500
    },
    {
      "epoch": 1.0296495956873315,
      "grad_norm": 0.5421329140663147,
      "learning_rate": 0.0005388148947652455,
      "loss": 3.7779,
      "step": 9550
    },
    {
      "epoch": 1.0350404312668464,
      "grad_norm": 0.5929830074310303,
      "learning_rate": 0.000538491095520777,
      "loss": 3.7659,
      "step": 9600
    },
    {
      "epoch": 1.0404312668463611,
      "grad_norm": 0.5713817477226257,
      "learning_rate": 0.0005381672962763086,
      "loss": 3.8062,
      "step": 9650
    },
    {
      "epoch": 1.045822102425876,
      "grad_norm": 0.6254969239234924,
      "learning_rate": 0.0005378434970318403,
      "loss": 3.7924,
      "step": 9700
    },
    {
      "epoch": 1.0512129380053907,
      "grad_norm": 0.5617794394493103,
      "learning_rate": 0.0005375196977873718,
      "loss": 3.7907,
      "step": 9750
    },
    {
      "epoch": 1.0566037735849056,
      "grad_norm": 0.5884292721748352,
      "learning_rate": 0.0005371958985429034,
      "loss": 3.7984,
      "step": 9800
    },
    {
      "epoch": 1.0619946091644206,
      "grad_norm": 0.5575320720672607,
      "learning_rate": 0.0005368720992984349,
      "loss": 3.7939,
      "step": 9850
    },
    {
      "epoch": 1.0673854447439353,
      "grad_norm": 0.5543467402458191,
      "learning_rate": 0.0005365483000539665,
      "loss": 3.785,
      "step": 9900
    },
    {
      "epoch": 1.0727762803234502,
      "grad_norm": 0.607307493686676,
      "learning_rate": 0.0005362245008094981,
      "loss": 3.7797,
      "step": 9950
    },
    {
      "epoch": 1.0781671159029649,
      "grad_norm": 0.6108971834182739,
      "learning_rate": 0.0005359007015650297,
      "loss": 3.7746,
      "step": 10000
    },
    {
      "epoch": 1.0781671159029649,
      "eval_accuracy": 0.3434026672762818,
      "eval_loss": 3.7587594985961914,
      "eval_runtime": 187.3528,
      "eval_samples_per_second": 96.134,
      "eval_steps_per_second": 6.01,
      "step": 10000
    },
    {
      "epoch": 1.0835579514824798,
      "grad_norm": 0.5519567728042603,
      "learning_rate": 0.0005355769023205612,
      "loss": 3.7767,
      "step": 10050
    },
    {
      "epoch": 1.0889487870619945,
      "grad_norm": 0.5885294079780579,
      "learning_rate": 0.0005352531030760928,
      "loss": 3.7858,
      "step": 10100
    },
    {
      "epoch": 1.0943396226415094,
      "grad_norm": 0.6101541519165039,
      "learning_rate": 0.0005349293038316244,
      "loss": 3.7783,
      "step": 10150
    },
    {
      "epoch": 1.0997304582210243,
      "grad_norm": 0.5413455367088318,
      "learning_rate": 0.0005346055045871559,
      "loss": 3.777,
      "step": 10200
    },
    {
      "epoch": 1.105121293800539,
      "grad_norm": 0.5316413640975952,
      "learning_rate": 0.0005342817053426874,
      "loss": 3.7892,
      "step": 10250
    },
    {
      "epoch": 1.110512129380054,
      "grad_norm": 0.6017285585403442,
      "learning_rate": 0.0005339579060982191,
      "loss": 3.7736,
      "step": 10300
    },
    {
      "epoch": 1.1159029649595686,
      "grad_norm": 0.6177402138710022,
      "learning_rate": 0.0005336341068537506,
      "loss": 3.7604,
      "step": 10350
    },
    {
      "epoch": 1.1212938005390836,
      "grad_norm": 0.5573323965072632,
      "learning_rate": 0.0005333103076092822,
      "loss": 3.7756,
      "step": 10400
    },
    {
      "epoch": 1.1266846361185983,
      "grad_norm": 0.5655317306518555,
      "learning_rate": 0.0005329865083648137,
      "loss": 3.7654,
      "step": 10450
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.5832797884941101,
      "learning_rate": 0.0005326627091203454,
      "loss": 3.7555,
      "step": 10500
    },
    {
      "epoch": 1.137466307277628,
      "grad_norm": 0.5413033962249756,
      "learning_rate": 0.0005323389098758769,
      "loss": 3.7752,
      "step": 10550
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.6015472412109375,
      "learning_rate": 0.0005320151106314085,
      "loss": 3.7858,
      "step": 10600
    },
    {
      "epoch": 1.1482479784366577,
      "grad_norm": 0.6413048505783081,
      "learning_rate": 0.00053169131138694,
      "loss": 3.7643,
      "step": 10650
    },
    {
      "epoch": 1.1536388140161726,
      "grad_norm": 0.5632215738296509,
      "learning_rate": 0.0005313675121424716,
      "loss": 3.7595,
      "step": 10700
    },
    {
      "epoch": 1.1590296495956873,
      "grad_norm": 0.7078406810760498,
      "learning_rate": 0.0005310437128980032,
      "loss": 3.7473,
      "step": 10750
    },
    {
      "epoch": 1.1644204851752022,
      "grad_norm": 0.580906093120575,
      "learning_rate": 0.0005307199136535348,
      "loss": 3.7379,
      "step": 10800
    },
    {
      "epoch": 1.169811320754717,
      "grad_norm": 0.5415087938308716,
      "learning_rate": 0.0005303961144090663,
      "loss": 3.7567,
      "step": 10850
    },
    {
      "epoch": 1.1752021563342319,
      "grad_norm": 0.5719914436340332,
      "learning_rate": 0.0005300723151645979,
      "loss": 3.7588,
      "step": 10900
    },
    {
      "epoch": 1.1805929919137466,
      "grad_norm": 0.6258513331413269,
      "learning_rate": 0.0005297549919050189,
      "loss": 3.7707,
      "step": 10950
    },
    {
      "epoch": 1.1859838274932615,
      "grad_norm": 0.6074648499488831,
      "learning_rate": 0.0005294311926605504,
      "loss": 3.7845,
      "step": 11000
    },
    {
      "epoch": 1.1859838274932615,
      "eval_accuracy": 0.3456640598303646,
      "eval_loss": 3.7324743270874023,
      "eval_runtime": 187.399,
      "eval_samples_per_second": 96.11,
      "eval_steps_per_second": 6.009,
      "step": 11000
    },
    {
      "epoch": 1.1913746630727764,
      "grad_norm": 0.532958447933197,
      "learning_rate": 0.000529107393416082,
      "loss": 3.7742,
      "step": 11050
    },
    {
      "epoch": 1.196765498652291,
      "grad_norm": 0.5758987069129944,
      "learning_rate": 0.0005287835941716135,
      "loss": 3.7677,
      "step": 11100
    },
    {
      "epoch": 1.202156334231806,
      "grad_norm": 0.5204159021377563,
      "learning_rate": 0.0005284597949271452,
      "loss": 3.7653,
      "step": 11150
    },
    {
      "epoch": 1.2075471698113207,
      "grad_norm": 0.5608921051025391,
      "learning_rate": 0.0005281359956826767,
      "loss": 3.7539,
      "step": 11200
    },
    {
      "epoch": 1.2129380053908356,
      "grad_norm": 0.537016749382019,
      "learning_rate": 0.0005278121964382083,
      "loss": 3.7451,
      "step": 11250
    },
    {
      "epoch": 1.2183288409703503,
      "grad_norm": 0.5913570523262024,
      "learning_rate": 0.0005274883971937398,
      "loss": 3.7612,
      "step": 11300
    },
    {
      "epoch": 1.2237196765498652,
      "grad_norm": 0.6099703907966614,
      "learning_rate": 0.0005271645979492714,
      "loss": 3.7478,
      "step": 11350
    },
    {
      "epoch": 1.2291105121293802,
      "grad_norm": 0.5816707015037537,
      "learning_rate": 0.000526840798704803,
      "loss": 3.7651,
      "step": 11400
    },
    {
      "epoch": 1.2345013477088949,
      "grad_norm": 0.5798138380050659,
      "learning_rate": 0.0005265169994603346,
      "loss": 3.7535,
      "step": 11450
    },
    {
      "epoch": 1.2398921832884098,
      "grad_norm": 0.6095981001853943,
      "learning_rate": 0.0005261932002158661,
      "loss": 3.7464,
      "step": 11500
    },
    {
      "epoch": 1.2452830188679245,
      "grad_norm": 0.48052334785461426,
      "learning_rate": 0.0005258694009713977,
      "loss": 3.757,
      "step": 11550
    },
    {
      "epoch": 1.2506738544474394,
      "grad_norm": 0.5602318048477173,
      "learning_rate": 0.0005255456017269292,
      "loss": 3.7321,
      "step": 11600
    },
    {
      "epoch": 1.256064690026954,
      "grad_norm": 0.6862710118293762,
      "learning_rate": 0.0005252218024824608,
      "loss": 3.7624,
      "step": 11650
    },
    {
      "epoch": 1.261455525606469,
      "grad_norm": 0.680061399936676,
      "learning_rate": 0.0005248980032379924,
      "loss": 3.7615,
      "step": 11700
    },
    {
      "epoch": 1.266846361185984,
      "grad_norm": 0.5473019480705261,
      "learning_rate": 0.000524574203993524,
      "loss": 3.7522,
      "step": 11750
    },
    {
      "epoch": 1.2722371967654986,
      "grad_norm": 0.5845617055892944,
      "learning_rate": 0.0005242504047490555,
      "loss": 3.7444,
      "step": 11800
    },
    {
      "epoch": 1.2776280323450135,
      "grad_norm": 0.5994465351104736,
      "learning_rate": 0.0005239266055045871,
      "loss": 3.7533,
      "step": 11850
    },
    {
      "epoch": 1.2830188679245282,
      "grad_norm": 0.5354965329170227,
      "learning_rate": 0.0005236028062601186,
      "loss": 3.7394,
      "step": 11900
    },
    {
      "epoch": 1.2884097035040432,
      "grad_norm": 0.5937642455101013,
      "learning_rate": 0.0005232790070156503,
      "loss": 3.7542,
      "step": 11950
    },
    {
      "epoch": 1.2938005390835579,
      "grad_norm": 0.6336929202079773,
      "learning_rate": 0.0005229552077711818,
      "loss": 3.7478,
      "step": 12000
    },
    {
      "epoch": 1.2938005390835579,
      "eval_accuracy": 0.34786960480011075,
      "eval_loss": 3.708402633666992,
      "eval_runtime": 188.2113,
      "eval_samples_per_second": 95.696,
      "eval_steps_per_second": 5.983,
      "step": 12000
    },
    {
      "epoch": 1.2991913746630728,
      "grad_norm": 0.580945611000061,
      "learning_rate": 0.0005226314085267134,
      "loss": 3.7444,
      "step": 12050
    },
    {
      "epoch": 1.3045822102425877,
      "grad_norm": 0.5905938744544983,
      "learning_rate": 0.000522307609282245,
      "loss": 3.7524,
      "step": 12100
    },
    {
      "epoch": 1.3099730458221024,
      "grad_norm": 0.550325334072113,
      "learning_rate": 0.0005219838100377766,
      "loss": 3.7695,
      "step": 12150
    },
    {
      "epoch": 1.3153638814016173,
      "grad_norm": 0.5600820779800415,
      "learning_rate": 0.000521660010793308,
      "loss": 3.7399,
      "step": 12200
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.5471258759498596,
      "learning_rate": 0.0005213362115488396,
      "loss": 3.7442,
      "step": 12250
    },
    {
      "epoch": 1.326145552560647,
      "grad_norm": 0.5628847479820251,
      "learning_rate": 0.0005210124123043713,
      "loss": 3.7555,
      "step": 12300
    },
    {
      "epoch": 1.3315363881401616,
      "grad_norm": 0.592341959476471,
      "learning_rate": 0.0005206886130599028,
      "loss": 3.7435,
      "step": 12350
    },
    {
      "epoch": 1.3369272237196765,
      "grad_norm": 0.5691344738006592,
      "learning_rate": 0.0005203648138154344,
      "loss": 3.7308,
      "step": 12400
    },
    {
      "epoch": 1.3423180592991915,
      "grad_norm": 0.5798255801200867,
      "learning_rate": 0.0005200410145709659,
      "loss": 3.7462,
      "step": 12450
    },
    {
      "epoch": 1.3477088948787062,
      "grad_norm": 0.7156546711921692,
      "learning_rate": 0.0005197172153264976,
      "loss": 3.7264,
      "step": 12500
    },
    {
      "epoch": 1.353099730458221,
      "grad_norm": 0.5289100408554077,
      "learning_rate": 0.0005193934160820291,
      "loss": 3.7396,
      "step": 12550
    },
    {
      "epoch": 1.3584905660377358,
      "grad_norm": 0.5872498154640198,
      "learning_rate": 0.0005190696168375607,
      "loss": 3.7339,
      "step": 12600
    },
    {
      "epoch": 1.3638814016172507,
      "grad_norm": 0.6003397703170776,
      "learning_rate": 0.0005187458175930922,
      "loss": 3.7224,
      "step": 12650
    },
    {
      "epoch": 1.3692722371967654,
      "grad_norm": 0.5635182857513428,
      "learning_rate": 0.0005184220183486238,
      "loss": 3.7189,
      "step": 12700
    },
    {
      "epoch": 1.3746630727762803,
      "grad_norm": 0.5980514287948608,
      "learning_rate": 0.0005180982191041554,
      "loss": 3.7381,
      "step": 12750
    },
    {
      "epoch": 1.3800539083557952,
      "grad_norm": 0.5108292102813721,
      "learning_rate": 0.000517774419859687,
      "loss": 3.7194,
      "step": 12800
    },
    {
      "epoch": 1.38544474393531,
      "grad_norm": 0.6213290691375732,
      "learning_rate": 0.0005174506206152185,
      "loss": 3.7305,
      "step": 12850
    },
    {
      "epoch": 1.3908355795148248,
      "grad_norm": 0.512791633605957,
      "learning_rate": 0.0005171268213707501,
      "loss": 3.7317,
      "step": 12900
    },
    {
      "epoch": 1.3962264150943398,
      "grad_norm": 0.6446360349655151,
      "learning_rate": 0.0005168030221262816,
      "loss": 3.7234,
      "step": 12950
    },
    {
      "epoch": 1.4016172506738545,
      "grad_norm": 0.5144412517547607,
      "learning_rate": 0.0005164792228818132,
      "loss": 3.72,
      "step": 13000
    },
    {
      "epoch": 1.4016172506738545,
      "eval_accuracy": 0.3508476718022178,
      "eval_loss": 3.6830170154571533,
      "eval_runtime": 187.4709,
      "eval_samples_per_second": 96.074,
      "eval_steps_per_second": 6.006,
      "step": 13000
    },
    {
      "epoch": 1.4070080862533692,
      "grad_norm": 0.5337992310523987,
      "learning_rate": 0.0005161554236373448,
      "loss": 3.6965,
      "step": 13050
    },
    {
      "epoch": 1.412398921832884,
      "grad_norm": 0.5049533843994141,
      "learning_rate": 0.0005158316243928764,
      "loss": 3.719,
      "step": 13100
    },
    {
      "epoch": 1.417789757412399,
      "grad_norm": 0.563833475112915,
      "learning_rate": 0.0005155078251484079,
      "loss": 3.7227,
      "step": 13150
    },
    {
      "epoch": 1.4231805929919137,
      "grad_norm": 0.5263909101486206,
      "learning_rate": 0.0005151840259039395,
      "loss": 3.7043,
      "step": 13200
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.5385528206825256,
      "learning_rate": 0.000514860226659471,
      "loss": 3.7236,
      "step": 13250
    },
    {
      "epoch": 1.4339622641509435,
      "grad_norm": 0.5642369985580444,
      "learning_rate": 0.0005145364274150027,
      "loss": 3.718,
      "step": 13300
    },
    {
      "epoch": 1.4393530997304582,
      "grad_norm": 0.5917314887046814,
      "learning_rate": 0.0005142126281705343,
      "loss": 3.7219,
      "step": 13350
    },
    {
      "epoch": 1.444743935309973,
      "grad_norm": 0.5475001931190491,
      "learning_rate": 0.0005138888289260658,
      "loss": 3.712,
      "step": 13400
    },
    {
      "epoch": 1.4501347708894878,
      "grad_norm": 0.5014727711677551,
      "learning_rate": 0.0005135650296815974,
      "loss": 3.729,
      "step": 13450
    },
    {
      "epoch": 1.4555256064690028,
      "grad_norm": 0.6016579270362854,
      "learning_rate": 0.0005132412304371289,
      "loss": 3.7224,
      "step": 13500
    },
    {
      "epoch": 1.4609164420485174,
      "grad_norm": 0.5744902491569519,
      "learning_rate": 0.0005129174311926605,
      "loss": 3.6989,
      "step": 13550
    },
    {
      "epoch": 1.4663072776280324,
      "grad_norm": 0.5305960178375244,
      "learning_rate": 0.000512593631948192,
      "loss": 3.7266,
      "step": 13600
    },
    {
      "epoch": 1.4716981132075473,
      "grad_norm": 0.5261050462722778,
      "learning_rate": 0.0005122698327037237,
      "loss": 3.7084,
      "step": 13650
    },
    {
      "epoch": 1.477088948787062,
      "grad_norm": 0.5880641341209412,
      "learning_rate": 0.0005119460334592552,
      "loss": 3.7263,
      "step": 13700
    },
    {
      "epoch": 1.482479784366577,
      "grad_norm": 0.5516208410263062,
      "learning_rate": 0.0005116222342147868,
      "loss": 3.6993,
      "step": 13750
    },
    {
      "epoch": 1.4878706199460916,
      "grad_norm": 0.6383761167526245,
      "learning_rate": 0.0005112984349703183,
      "loss": 3.7225,
      "step": 13800
    },
    {
      "epoch": 1.4932614555256065,
      "grad_norm": 0.5817267894744873,
      "learning_rate": 0.00051097463572585,
      "loss": 3.7083,
      "step": 13850
    },
    {
      "epoch": 1.4986522911051212,
      "grad_norm": 0.5393953323364258,
      "learning_rate": 0.0005106573124662708,
      "loss": 3.7188,
      "step": 13900
    },
    {
      "epoch": 1.5040431266846361,
      "grad_norm": 0.5401408076286316,
      "learning_rate": 0.0005103335132218025,
      "loss": 3.7074,
      "step": 13950
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.5108155012130737,
      "learning_rate": 0.000510009713977334,
      "loss": 3.7062,
      "step": 14000
    },
    {
      "epoch": 1.509433962264151,
      "eval_accuracy": 0.3525950275440503,
      "eval_loss": 3.6650450229644775,
      "eval_runtime": 187.3211,
      "eval_samples_per_second": 96.15,
      "eval_steps_per_second": 6.011,
      "step": 14000
    },
    {
      "epoch": 1.5148247978436657,
      "grad_norm": 0.593525230884552,
      "learning_rate": 0.0005096859147328656,
      "loss": 3.7055,
      "step": 14050
    },
    {
      "epoch": 1.5202156334231804,
      "grad_norm": 0.5967386960983276,
      "learning_rate": 0.0005093621154883971,
      "loss": 3.7037,
      "step": 14100
    },
    {
      "epoch": 1.5256064690026954,
      "grad_norm": 0.5114768147468567,
      "learning_rate": 0.0005090383162439288,
      "loss": 3.6958,
      "step": 14150
    },
    {
      "epoch": 1.5309973045822103,
      "grad_norm": 0.6086651086807251,
      "learning_rate": 0.0005087145169994602,
      "loss": 3.6959,
      "step": 14200
    },
    {
      "epoch": 1.536388140161725,
      "grad_norm": 0.5664010643959045,
      "learning_rate": 0.0005083907177549918,
      "loss": 3.7123,
      "step": 14250
    },
    {
      "epoch": 1.54177897574124,
      "grad_norm": 0.5594226121902466,
      "learning_rate": 0.0005080669185105234,
      "loss": 3.7111,
      "step": 14300
    },
    {
      "epoch": 1.5471698113207548,
      "grad_norm": 0.5915853381156921,
      "learning_rate": 0.000507743119266055,
      "loss": 3.7037,
      "step": 14350
    },
    {
      "epoch": 1.5525606469002695,
      "grad_norm": 0.5531720519065857,
      "learning_rate": 0.0005074193200215865,
      "loss": 3.7185,
      "step": 14400
    },
    {
      "epoch": 1.5579514824797842,
      "grad_norm": 0.5341070890426636,
      "learning_rate": 0.0005070955207771181,
      "loss": 3.7107,
      "step": 14450
    },
    {
      "epoch": 1.5633423180592994,
      "grad_norm": 0.6099830269813538,
      "learning_rate": 0.0005067717215326498,
      "loss": 3.6937,
      "step": 14500
    },
    {
      "epoch": 1.568733153638814,
      "grad_norm": 0.5606914162635803,
      "learning_rate": 0.0005064479222881813,
      "loss": 3.7077,
      "step": 14550
    },
    {
      "epoch": 1.5741239892183287,
      "grad_norm": 0.5850932598114014,
      "learning_rate": 0.0005061241230437129,
      "loss": 3.6935,
      "step": 14600
    },
    {
      "epoch": 1.5795148247978437,
      "grad_norm": 0.570080041885376,
      "learning_rate": 0.0005058003237992444,
      "loss": 3.6894,
      "step": 14650
    },
    {
      "epoch": 1.5849056603773586,
      "grad_norm": 0.5481992363929749,
      "learning_rate": 0.000505476524554776,
      "loss": 3.7065,
      "step": 14700
    },
    {
      "epoch": 1.5902964959568733,
      "grad_norm": 0.5915714502334595,
      "learning_rate": 0.0005051527253103076,
      "loss": 3.708,
      "step": 14750
    },
    {
      "epoch": 1.595687331536388,
      "grad_norm": 0.5506225228309631,
      "learning_rate": 0.0005048289260658392,
      "loss": 3.6768,
      "step": 14800
    },
    {
      "epoch": 1.6010781671159031,
      "grad_norm": 0.5602514147758484,
      "learning_rate": 0.0005045051268213707,
      "loss": 3.6965,
      "step": 14850
    },
    {
      "epoch": 1.6064690026954178,
      "grad_norm": 0.5985303521156311,
      "learning_rate": 0.0005041813275769023,
      "loss": 3.7044,
      "step": 14900
    },
    {
      "epoch": 1.6118598382749325,
      "grad_norm": 0.6018468141555786,
      "learning_rate": 0.0005038575283324338,
      "loss": 3.6936,
      "step": 14950
    },
    {
      "epoch": 1.6172506738544474,
      "grad_norm": 0.6045594811439514,
      "learning_rate": 0.0005035337290879654,
      "loss": 3.6945,
      "step": 15000
    },
    {
      "epoch": 1.6172506738544474,
      "eval_accuracy": 0.3545141635015175,
      "eval_loss": 3.6467270851135254,
      "eval_runtime": 188.3924,
      "eval_samples_per_second": 95.604,
      "eval_steps_per_second": 5.977,
      "step": 15000
    },
    {
      "epoch": 1.6226415094339623,
      "grad_norm": 0.5498721599578857,
      "learning_rate": 0.000503209929843497,
      "loss": 3.6842,
      "step": 15050
    },
    {
      "epoch": 1.628032345013477,
      "grad_norm": 0.5301955342292786,
      "learning_rate": 0.0005028861305990286,
      "loss": 3.6793,
      "step": 15100
    },
    {
      "epoch": 1.633423180592992,
      "grad_norm": 0.5473493933677673,
      "learning_rate": 0.0005025623313545601,
      "loss": 3.6787,
      "step": 15150
    },
    {
      "epoch": 1.6388140161725069,
      "grad_norm": 0.5606446862220764,
      "learning_rate": 0.0005022385321100917,
      "loss": 3.6822,
      "step": 15200
    },
    {
      "epoch": 1.6442048517520216,
      "grad_norm": 0.5899948477745056,
      "learning_rate": 0.0005019147328656232,
      "loss": 3.6766,
      "step": 15250
    },
    {
      "epoch": 1.6495956873315363,
      "grad_norm": 0.5508028864860535,
      "learning_rate": 0.0005015909336211549,
      "loss": 3.6821,
      "step": 15300
    },
    {
      "epoch": 1.6549865229110512,
      "grad_norm": 0.540977954864502,
      "learning_rate": 0.0005012671343766864,
      "loss": 3.7041,
      "step": 15350
    },
    {
      "epoch": 1.6603773584905661,
      "grad_norm": 0.5751597881317139,
      "learning_rate": 0.000500943335132218,
      "loss": 3.695,
      "step": 15400
    },
    {
      "epoch": 1.6657681940700808,
      "grad_norm": 0.5815579891204834,
      "learning_rate": 0.0005006195358877495,
      "loss": 3.6627,
      "step": 15450
    },
    {
      "epoch": 1.6711590296495957,
      "grad_norm": 0.6284985542297363,
      "learning_rate": 0.0005002957366432812,
      "loss": 3.6786,
      "step": 15500
    },
    {
      "epoch": 1.6765498652291106,
      "grad_norm": 0.5556270480155945,
      "learning_rate": 0.0004999719373988127,
      "loss": 3.6775,
      "step": 15550
    },
    {
      "epoch": 1.6819407008086253,
      "grad_norm": 0.509715735912323,
      "learning_rate": 0.0004996481381543442,
      "loss": 3.6642,
      "step": 15600
    },
    {
      "epoch": 1.68733153638814,
      "grad_norm": 0.5490276217460632,
      "learning_rate": 0.0004993243389098758,
      "loss": 3.6823,
      "step": 15650
    },
    {
      "epoch": 1.692722371967655,
      "grad_norm": 0.5493997931480408,
      "learning_rate": 0.0004990005396654074,
      "loss": 3.6706,
      "step": 15700
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.5856523513793945,
      "learning_rate": 0.000498676740420939,
      "loss": 3.6974,
      "step": 15750
    },
    {
      "epoch": 1.7035040431266846,
      "grad_norm": 0.5182335376739502,
      "learning_rate": 0.0004983529411764705,
      "loss": 3.6908,
      "step": 15800
    },
    {
      "epoch": 1.7088948787061995,
      "grad_norm": 0.5529457330703735,
      "learning_rate": 0.0004980291419320022,
      "loss": 3.6858,
      "step": 15850
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5879248976707458,
      "learning_rate": 0.0004977053426875337,
      "loss": 3.6865,
      "step": 15900
    },
    {
      "epoch": 1.719676549865229,
      "grad_norm": 0.5359522700309753,
      "learning_rate": 0.0004973815434430653,
      "loss": 3.6724,
      "step": 15950
    },
    {
      "epoch": 1.7250673854447438,
      "grad_norm": 0.6166804432868958,
      "learning_rate": 0.0004970642201834862,
      "loss": 3.6696,
      "step": 16000
    },
    {
      "epoch": 1.7250673854447438,
      "eval_accuracy": 0.3563720192302573,
      "eval_loss": 3.6291186809539795,
      "eval_runtime": 187.9412,
      "eval_samples_per_second": 95.833,
      "eval_steps_per_second": 5.991,
      "step": 16000
    },
    {
      "epoch": 1.7304582210242587,
      "grad_norm": 0.5808438062667847,
      "learning_rate": 0.0004967404209390178,
      "loss": 3.6702,
      "step": 16050
    },
    {
      "epoch": 1.7358490566037736,
      "grad_norm": 0.5703431963920593,
      "learning_rate": 0.0004964166216945493,
      "loss": 3.658,
      "step": 16100
    },
    {
      "epoch": 1.7412398921832883,
      "grad_norm": 0.5545331239700317,
      "learning_rate": 0.000496092822450081,
      "loss": 3.68,
      "step": 16150
    },
    {
      "epoch": 1.7466307277628033,
      "grad_norm": 0.5666328072547913,
      "learning_rate": 0.0004957690232056125,
      "loss": 3.6912,
      "step": 16200
    },
    {
      "epoch": 1.7520215633423182,
      "grad_norm": 0.5478608012199402,
      "learning_rate": 0.0004954452239611441,
      "loss": 3.6673,
      "step": 16250
    },
    {
      "epoch": 1.7574123989218329,
      "grad_norm": 0.5847337245941162,
      "learning_rate": 0.0004951214247166756,
      "loss": 3.6887,
      "step": 16300
    },
    {
      "epoch": 1.7628032345013476,
      "grad_norm": 0.6067107915878296,
      "learning_rate": 0.0004947976254722072,
      "loss": 3.6728,
      "step": 16350
    },
    {
      "epoch": 1.7681940700808625,
      "grad_norm": 0.5245669484138489,
      "learning_rate": 0.0004944738262277387,
      "loss": 3.6636,
      "step": 16400
    },
    {
      "epoch": 1.7735849056603774,
      "grad_norm": 0.5606017708778381,
      "learning_rate": 0.0004941500269832703,
      "loss": 3.6698,
      "step": 16450
    },
    {
      "epoch": 1.778975741239892,
      "grad_norm": 0.5453688502311707,
      "learning_rate": 0.0004938262277388019,
      "loss": 3.6553,
      "step": 16500
    },
    {
      "epoch": 1.784366576819407,
      "grad_norm": 0.6251377463340759,
      "learning_rate": 0.0004935024284943335,
      "loss": 3.684,
      "step": 16550
    },
    {
      "epoch": 1.789757412398922,
      "grad_norm": 0.6245383620262146,
      "learning_rate": 0.000493178629249865,
      "loss": 3.6669,
      "step": 16600
    },
    {
      "epoch": 1.7951482479784366,
      "grad_norm": 0.5404795408248901,
      "learning_rate": 0.0004928548300053966,
      "loss": 3.6722,
      "step": 16650
    },
    {
      "epoch": 1.8005390835579513,
      "grad_norm": 0.5152893662452698,
      "learning_rate": 0.0004925310307609282,
      "loss": 3.6736,
      "step": 16700
    },
    {
      "epoch": 1.8059299191374663,
      "grad_norm": 0.5853434801101685,
      "learning_rate": 0.0004922072315164598,
      "loss": 3.6629,
      "step": 16750
    },
    {
      "epoch": 1.8113207547169812,
      "grad_norm": 0.53773033618927,
      "learning_rate": 0.0004918834322719913,
      "loss": 3.6618,
      "step": 16800
    },
    {
      "epoch": 1.8167115902964959,
      "grad_norm": 0.6675792336463928,
      "learning_rate": 0.0004915596330275229,
      "loss": 3.6682,
      "step": 16850
    },
    {
      "epoch": 1.8221024258760108,
      "grad_norm": 0.5869772434234619,
      "learning_rate": 0.0004912423097679439,
      "loss": 3.6543,
      "step": 16900
    },
    {
      "epoch": 1.8274932614555257,
      "grad_norm": 0.5314184427261353,
      "learning_rate": 0.0004909185105234754,
      "loss": 3.6832,
      "step": 16950
    },
    {
      "epoch": 1.8328840970350404,
      "grad_norm": 0.5484595894813538,
      "learning_rate": 0.000490594711279007,
      "loss": 3.6703,
      "step": 17000
    },
    {
      "epoch": 1.8328840970350404,
      "eval_accuracy": 0.35783242269537174,
      "eval_loss": 3.615173816680908,
      "eval_runtime": 187.71,
      "eval_samples_per_second": 95.951,
      "eval_steps_per_second": 5.999,
      "step": 17000
    },
    {
      "epoch": 1.838274932614555,
      "grad_norm": 0.5312723517417908,
      "learning_rate": 0.0004902709120345385,
      "loss": 3.6526,
      "step": 17050
    },
    {
      "epoch": 1.8436657681940702,
      "grad_norm": 0.5201370716094971,
      "learning_rate": 0.00048994711279007,
      "loss": 3.6655,
      "step": 17100
    },
    {
      "epoch": 1.849056603773585,
      "grad_norm": 0.541175127029419,
      "learning_rate": 0.0004896233135456017,
      "loss": 3.6618,
      "step": 17150
    },
    {
      "epoch": 1.8544474393530996,
      "grad_norm": 0.5954682230949402,
      "learning_rate": 0.0004892995143011333,
      "loss": 3.6487,
      "step": 17200
    },
    {
      "epoch": 1.8598382749326146,
      "grad_norm": 0.6293191909790039,
      "learning_rate": 0.0004889757150566648,
      "loss": 3.6673,
      "step": 17250
    },
    {
      "epoch": 1.8652291105121295,
      "grad_norm": 0.593071460723877,
      "learning_rate": 0.0004886519158121964,
      "loss": 3.6598,
      "step": 17300
    },
    {
      "epoch": 1.8706199460916442,
      "grad_norm": 0.6929914951324463,
      "learning_rate": 0.000488328116567728,
      "loss": 3.6619,
      "step": 17350
    },
    {
      "epoch": 1.8760107816711589,
      "grad_norm": 0.5749919414520264,
      "learning_rate": 0.0004880043173232595,
      "loss": 3.6562,
      "step": 17400
    },
    {
      "epoch": 1.881401617250674,
      "grad_norm": 0.5454095602035522,
      "learning_rate": 0.0004876805180787911,
      "loss": 3.6735,
      "step": 17450
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.5522634387016296,
      "learning_rate": 0.0004873567188343227,
      "loss": 3.6662,
      "step": 17500
    },
    {
      "epoch": 1.8921832884097034,
      "grad_norm": 0.5520308017730713,
      "learning_rate": 0.0004870329195898542,
      "loss": 3.6464,
      "step": 17550
    },
    {
      "epoch": 1.8975741239892183,
      "grad_norm": 0.596127986907959,
      "learning_rate": 0.00048670912034538583,
      "loss": 3.6552,
      "step": 17600
    },
    {
      "epoch": 1.9029649595687332,
      "grad_norm": 0.5734618306159973,
      "learning_rate": 0.0004863853211009174,
      "loss": 3.6327,
      "step": 17650
    },
    {
      "epoch": 1.908355795148248,
      "grad_norm": 0.5969244837760925,
      "learning_rate": 0.000486061521856449,
      "loss": 3.6419,
      "step": 17700
    },
    {
      "epoch": 1.9137466307277629,
      "grad_norm": 0.5883053541183472,
      "learning_rate": 0.00048573772261198054,
      "loss": 3.6462,
      "step": 17750
    },
    {
      "epoch": 1.9191374663072778,
      "grad_norm": 0.544793426990509,
      "learning_rate": 0.00048541392336751214,
      "loss": 3.664,
      "step": 17800
    },
    {
      "epoch": 1.9245283018867925,
      "grad_norm": 0.5943015813827515,
      "learning_rate": 0.0004850901241230437,
      "loss": 3.665,
      "step": 17850
    },
    {
      "epoch": 1.9299191374663072,
      "grad_norm": 0.5140420198440552,
      "learning_rate": 0.0004847663248785753,
      "loss": 3.6708,
      "step": 17900
    },
    {
      "epoch": 1.935309973045822,
      "grad_norm": 0.5902069807052612,
      "learning_rate": 0.0004844425256341068,
      "loss": 3.6597,
      "step": 17950
    },
    {
      "epoch": 1.940700808625337,
      "grad_norm": 0.6027345657348633,
      "learning_rate": 0.00048411872638963834,
      "loss": 3.6467,
      "step": 18000
    },
    {
      "epoch": 1.940700808625337,
      "eval_accuracy": 0.3591877588179696,
      "eval_loss": 3.598599910736084,
      "eval_runtime": 187.872,
      "eval_samples_per_second": 95.868,
      "eval_steps_per_second": 5.993,
      "step": 18000
    },
    {
      "epoch": 1.9460916442048517,
      "grad_norm": 0.5818517208099365,
      "learning_rate": 0.00048379492714516995,
      "loss": 3.6332,
      "step": 18050
    },
    {
      "epoch": 1.9514824797843666,
      "grad_norm": 0.6805805563926697,
      "learning_rate": 0.0004834711279007015,
      "loss": 3.6435,
      "step": 18100
    },
    {
      "epoch": 1.9568733153638815,
      "grad_norm": 0.5603196620941162,
      "learning_rate": 0.0004831473286562331,
      "loss": 3.6487,
      "step": 18150
    },
    {
      "epoch": 1.9622641509433962,
      "grad_norm": 0.5356538891792297,
      "learning_rate": 0.00048282352941176465,
      "loss": 3.6499,
      "step": 18200
    },
    {
      "epoch": 1.967654986522911,
      "grad_norm": 0.5439290404319763,
      "learning_rate": 0.00048249973016729626,
      "loss": 3.6414,
      "step": 18250
    },
    {
      "epoch": 1.9730458221024259,
      "grad_norm": 0.5327361226081848,
      "learning_rate": 0.0004821759309228278,
      "loss": 3.6649,
      "step": 18300
    },
    {
      "epoch": 1.9784366576819408,
      "grad_norm": 0.5530686378479004,
      "learning_rate": 0.00048185213167835936,
      "loss": 3.6452,
      "step": 18350
    },
    {
      "epoch": 1.9838274932614555,
      "grad_norm": 0.6126965284347534,
      "learning_rate": 0.00048152833243389096,
      "loss": 3.6502,
      "step": 18400
    },
    {
      "epoch": 1.9892183288409704,
      "grad_norm": 0.5579380989074707,
      "learning_rate": 0.0004812045331894225,
      "loss": 3.6557,
      "step": 18450
    },
    {
      "epoch": 1.9946091644204853,
      "grad_norm": 0.5861890912055969,
      "learning_rate": 0.0004808807339449541,
      "loss": 3.6357,
      "step": 18500
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1135197877883911,
      "learning_rate": 0.0004805569347004856,
      "loss": 3.6403,
      "step": 18550
    },
    {
      "epoch": 2.0053908355795147,
      "grad_norm": 0.5583805441856384,
      "learning_rate": 0.0004802331354560173,
      "loss": 3.536,
      "step": 18600
    },
    {
      "epoch": 2.01078167115903,
      "grad_norm": 0.5247920155525208,
      "learning_rate": 0.00047990933621154877,
      "loss": 3.5479,
      "step": 18650
    },
    {
      "epoch": 2.0161725067385445,
      "grad_norm": 0.5629153847694397,
      "learning_rate": 0.0004795855369670804,
      "loss": 3.5451,
      "step": 18700
    },
    {
      "epoch": 2.0215633423180592,
      "grad_norm": 0.6046398282051086,
      "learning_rate": 0.0004792617377226119,
      "loss": 3.5519,
      "step": 18750
    },
    {
      "epoch": 2.026954177897574,
      "grad_norm": 0.5622299313545227,
      "learning_rate": 0.0004789379384781435,
      "loss": 3.574,
      "step": 18800
    },
    {
      "epoch": 2.032345013477089,
      "grad_norm": 0.5278229117393494,
      "learning_rate": 0.0004786141392336751,
      "loss": 3.5721,
      "step": 18850
    },
    {
      "epoch": 2.0377358490566038,
      "grad_norm": 0.5192546248435974,
      "learning_rate": 0.00047829033998920663,
      "loss": 3.5541,
      "step": 18900
    },
    {
      "epoch": 2.0431266846361185,
      "grad_norm": 0.5911169052124023,
      "learning_rate": 0.00047796654074473824,
      "loss": 3.5595,
      "step": 18950
    },
    {
      "epoch": 2.0485175202156336,
      "grad_norm": 0.5658133625984192,
      "learning_rate": 0.0004776427415002698,
      "loss": 3.5905,
      "step": 19000
    },
    {
      "epoch": 2.0485175202156336,
      "eval_accuracy": 0.36072813080851546,
      "eval_loss": 3.588514566421509,
      "eval_runtime": 187.5876,
      "eval_samples_per_second": 96.014,
      "eval_steps_per_second": 6.003,
      "step": 19000
    },
    {
      "epoch": 2.0539083557951483,
      "grad_norm": 0.5861156582832336,
      "learning_rate": 0.0004773189422558014,
      "loss": 3.5654,
      "step": 19050
    },
    {
      "epoch": 2.059299191374663,
      "grad_norm": 0.5876423716545105,
      "learning_rate": 0.00047699514301133294,
      "loss": 3.559,
      "step": 19100
    },
    {
      "epoch": 2.0646900269541777,
      "grad_norm": 0.6080085635185242,
      "learning_rate": 0.00047667134376686455,
      "loss": 3.573,
      "step": 19150
    },
    {
      "epoch": 2.070080862533693,
      "grad_norm": 0.6167579293251038,
      "learning_rate": 0.0004763475445223961,
      "loss": 3.585,
      "step": 19200
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.5479595065116882,
      "learning_rate": 0.0004760237452779276,
      "loss": 3.568,
      "step": 19250
    },
    {
      "epoch": 2.0808625336927222,
      "grad_norm": 0.6446500420570374,
      "learning_rate": 0.0004756999460334592,
      "loss": 3.5579,
      "step": 19300
    },
    {
      "epoch": 2.0862533692722374,
      "grad_norm": 0.6173583269119263,
      "learning_rate": 0.00047537614678899075,
      "loss": 3.5745,
      "step": 19350
    },
    {
      "epoch": 2.091644204851752,
      "grad_norm": 0.5377396941184998,
      "learning_rate": 0.00047505234754452235,
      "loss": 3.5535,
      "step": 19400
    },
    {
      "epoch": 2.0970350404312668,
      "grad_norm": 0.5425545573234558,
      "learning_rate": 0.0004747285483000539,
      "loss": 3.5692,
      "step": 19450
    },
    {
      "epoch": 2.1024258760107815,
      "grad_norm": 0.6151023507118225,
      "learning_rate": 0.0004744047490555855,
      "loss": 3.5695,
      "step": 19500
    },
    {
      "epoch": 2.1078167115902966,
      "grad_norm": 0.6019276976585388,
      "learning_rate": 0.00047408094981111706,
      "loss": 3.5584,
      "step": 19550
    },
    {
      "epoch": 2.1132075471698113,
      "grad_norm": 0.5593970417976379,
      "learning_rate": 0.00047375715056664866,
      "loss": 3.5596,
      "step": 19600
    },
    {
      "epoch": 2.118598382749326,
      "grad_norm": 0.6009968519210815,
      "learning_rate": 0.0004734333513221802,
      "loss": 3.567,
      "step": 19650
    },
    {
      "epoch": 2.123989218328841,
      "grad_norm": 0.5649537444114685,
      "learning_rate": 0.00047310955207771177,
      "loss": 3.5717,
      "step": 19700
    },
    {
      "epoch": 2.129380053908356,
      "grad_norm": 0.5351967215538025,
      "learning_rate": 0.00047278575283324337,
      "loss": 3.5785,
      "step": 19750
    },
    {
      "epoch": 2.1347708894878705,
      "grad_norm": 0.5678059458732605,
      "learning_rate": 0.0004724619535887749,
      "loss": 3.5599,
      "step": 19800
    },
    {
      "epoch": 2.1401617250673857,
      "grad_norm": 0.5407218933105469,
      "learning_rate": 0.0004721381543443065,
      "loss": 3.5872,
      "step": 19850
    },
    {
      "epoch": 2.1455525606469004,
      "grad_norm": 0.5732417106628418,
      "learning_rate": 0.000471814355099838,
      "loss": 3.5882,
      "step": 19900
    },
    {
      "epoch": 2.150943396226415,
      "grad_norm": 0.6302648186683655,
      "learning_rate": 0.0004714905558553697,
      "loss": 3.5841,
      "step": 19950
    },
    {
      "epoch": 2.1563342318059298,
      "grad_norm": 0.5343143939971924,
      "learning_rate": 0.0004711667566109012,
      "loss": 3.5686,
      "step": 20000
    },
    {
      "epoch": 2.1563342318059298,
      "eval_accuracy": 0.36169535881584,
      "eval_loss": 3.582411050796509,
      "eval_runtime": 187.7427,
      "eval_samples_per_second": 95.934,
      "eval_steps_per_second": 5.998,
      "step": 20000
    },
    {
      "epoch": 2.161725067385445,
      "grad_norm": 0.5989320278167725,
      "learning_rate": 0.00047084295736643273,
      "loss": 3.5688,
      "step": 20050
    },
    {
      "epoch": 2.1671159029649596,
      "grad_norm": 0.5695555806159973,
      "learning_rate": 0.00047051915812196433,
      "loss": 3.5748,
      "step": 20100
    },
    {
      "epoch": 2.1725067385444743,
      "grad_norm": 0.5576919317245483,
      "learning_rate": 0.0004701953588774959,
      "loss": 3.5674,
      "step": 20150
    },
    {
      "epoch": 2.177897574123989,
      "grad_norm": 0.5610904097557068,
      "learning_rate": 0.0004698715596330275,
      "loss": 3.5857,
      "step": 20200
    },
    {
      "epoch": 2.183288409703504,
      "grad_norm": 0.5354993343353271,
      "learning_rate": 0.00046954776038855904,
      "loss": 3.5707,
      "step": 20250
    },
    {
      "epoch": 2.188679245283019,
      "grad_norm": 0.5753160715103149,
      "learning_rate": 0.00046922396114409064,
      "loss": 3.5748,
      "step": 20300
    },
    {
      "epoch": 2.1940700808625335,
      "grad_norm": 0.6492990851402283,
      "learning_rate": 0.0004689001618996222,
      "loss": 3.5586,
      "step": 20350
    },
    {
      "epoch": 2.1994609164420487,
      "grad_norm": 0.5679466128349304,
      "learning_rate": 0.0004685763626551538,
      "loss": 3.5652,
      "step": 20400
    },
    {
      "epoch": 2.2048517520215634,
      "grad_norm": 0.5694677233695984,
      "learning_rate": 0.00046825256341068535,
      "loss": 3.5668,
      "step": 20450
    },
    {
      "epoch": 2.210242587601078,
      "grad_norm": 0.5879806280136108,
      "learning_rate": 0.0004679287641662169,
      "loss": 3.5777,
      "step": 20500
    },
    {
      "epoch": 2.215633423180593,
      "grad_norm": 0.5413506627082825,
      "learning_rate": 0.0004676049649217485,
      "loss": 3.5756,
      "step": 20550
    },
    {
      "epoch": 2.221024258760108,
      "grad_norm": 0.5593597888946533,
      "learning_rate": 0.00046728116567728,
      "loss": 3.5686,
      "step": 20600
    },
    {
      "epoch": 2.2264150943396226,
      "grad_norm": 0.5434522032737732,
      "learning_rate": 0.0004669573664328116,
      "loss": 3.5702,
      "step": 20650
    },
    {
      "epoch": 2.2318059299191373,
      "grad_norm": 0.6712062954902649,
      "learning_rate": 0.00046663356718834316,
      "loss": 3.5654,
      "step": 20700
    },
    {
      "epoch": 2.2371967654986524,
      "grad_norm": 0.5769302248954773,
      "learning_rate": 0.00046630976794387476,
      "loss": 3.5653,
      "step": 20750
    },
    {
      "epoch": 2.242587601078167,
      "grad_norm": 0.599884569644928,
      "learning_rate": 0.0004659859686994063,
      "loss": 3.5716,
      "step": 20800
    },
    {
      "epoch": 2.247978436657682,
      "grad_norm": 0.6358910202980042,
      "learning_rate": 0.0004656621694549379,
      "loss": 3.567,
      "step": 20850
    },
    {
      "epoch": 2.2533692722371965,
      "grad_norm": 0.5754926800727844,
      "learning_rate": 0.0004653448461953588,
      "loss": 3.5622,
      "step": 20900
    },
    {
      "epoch": 2.2587601078167117,
      "grad_norm": 0.5771598219871521,
      "learning_rate": 0.0004650210469508904,
      "loss": 3.572,
      "step": 20950
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.6123298406600952,
      "learning_rate": 0.00046469724770642197,
      "loss": 3.5546,
      "step": 21000
    },
    {
      "epoch": 2.2641509433962264,
      "eval_accuracy": 0.3627602657693097,
      "eval_loss": 3.571512460708618,
      "eval_runtime": 187.3542,
      "eval_samples_per_second": 96.133,
      "eval_steps_per_second": 6.01,
      "step": 21000
    },
    {
      "epoch": 2.269541778975741,
      "grad_norm": 0.5739647746086121,
      "learning_rate": 0.0004643734484619536,
      "loss": 3.5696,
      "step": 21050
    },
    {
      "epoch": 2.274932614555256,
      "grad_norm": 0.5520042777061462,
      "learning_rate": 0.0004640496492174851,
      "loss": 3.5743,
      "step": 21100
    },
    {
      "epoch": 2.280323450134771,
      "grad_norm": 0.5906243324279785,
      "learning_rate": 0.00046372584997301673,
      "loss": 3.5845,
      "step": 21150
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.5810971856117249,
      "learning_rate": 0.0004634020507285483,
      "loss": 3.5379,
      "step": 21200
    },
    {
      "epoch": 2.2911051212938007,
      "grad_norm": 0.565323531627655,
      "learning_rate": 0.0004630782514840798,
      "loss": 3.565,
      "step": 21250
    },
    {
      "epoch": 2.2964959568733154,
      "grad_norm": 0.5445725321769714,
      "learning_rate": 0.0004627544522396114,
      "loss": 3.5676,
      "step": 21300
    },
    {
      "epoch": 2.30188679245283,
      "grad_norm": 0.5527558326721191,
      "learning_rate": 0.00046243065299514293,
      "loss": 3.5603,
      "step": 21350
    },
    {
      "epoch": 2.3072776280323453,
      "grad_norm": 0.560587465763092,
      "learning_rate": 0.00046210685375067454,
      "loss": 3.5752,
      "step": 21400
    },
    {
      "epoch": 2.31266846361186,
      "grad_norm": 0.5325582027435303,
      "learning_rate": 0.0004617830545062061,
      "loss": 3.5719,
      "step": 21450
    },
    {
      "epoch": 2.3180592991913747,
      "grad_norm": 0.5584622621536255,
      "learning_rate": 0.0004614592552617377,
      "loss": 3.5749,
      "step": 21500
    },
    {
      "epoch": 2.3234501347708894,
      "grad_norm": 0.5946611166000366,
      "learning_rate": 0.00046113545601726924,
      "loss": 3.5561,
      "step": 21550
    },
    {
      "epoch": 2.3288409703504045,
      "grad_norm": 0.5582345128059387,
      "learning_rate": 0.00046081165677280085,
      "loss": 3.5775,
      "step": 21600
    },
    {
      "epoch": 2.334231805929919,
      "grad_norm": 0.5604034066200256,
      "learning_rate": 0.0004604878575283324,
      "loss": 3.573,
      "step": 21650
    },
    {
      "epoch": 2.339622641509434,
      "grad_norm": 0.5571855902671814,
      "learning_rate": 0.00046016405828386395,
      "loss": 3.5513,
      "step": 21700
    },
    {
      "epoch": 2.3450134770889486,
      "grad_norm": 0.6498404145240784,
      "learning_rate": 0.00045984025903939555,
      "loss": 3.5687,
      "step": 21750
    },
    {
      "epoch": 2.3504043126684637,
      "grad_norm": 0.5699803829193115,
      "learning_rate": 0.0004595164597949271,
      "loss": 3.5887,
      "step": 21800
    },
    {
      "epoch": 2.3557951482479784,
      "grad_norm": 0.6102402806282043,
      "learning_rate": 0.0004591926605504587,
      "loss": 3.5741,
      "step": 21850
    },
    {
      "epoch": 2.361185983827493,
      "grad_norm": 0.6587836146354675,
      "learning_rate": 0.0004588688613059902,
      "loss": 3.5529,
      "step": 21900
    },
    {
      "epoch": 2.3665768194070083,
      "grad_norm": 0.6018949151039124,
      "learning_rate": 0.00045854506206152186,
      "loss": 3.5676,
      "step": 21950
    },
    {
      "epoch": 2.371967654986523,
      "grad_norm": 0.5756721496582031,
      "learning_rate": 0.00045822126281705336,
      "loss": 3.5576,
      "step": 22000
    },
    {
      "epoch": 2.371967654986523,
      "eval_accuracy": 0.363854726308265,
      "eval_loss": 3.5607616901397705,
      "eval_runtime": 187.5094,
      "eval_samples_per_second": 96.054,
      "eval_steps_per_second": 6.005,
      "step": 22000
    },
    {
      "epoch": 2.3773584905660377,
      "grad_norm": 0.5994178056716919,
      "learning_rate": 0.00045789746357258497,
      "loss": 3.5606,
      "step": 22050
    },
    {
      "epoch": 2.382749326145553,
      "grad_norm": 0.5385029911994934,
      "learning_rate": 0.0004575736643281165,
      "loss": 3.573,
      "step": 22100
    },
    {
      "epoch": 2.3881401617250675,
      "grad_norm": 0.5386703610420227,
      "learning_rate": 0.00045724986508364807,
      "loss": 3.576,
      "step": 22150
    },
    {
      "epoch": 2.393530997304582,
      "grad_norm": 0.5755857825279236,
      "learning_rate": 0.00045692606583917967,
      "loss": 3.5603,
      "step": 22200
    },
    {
      "epoch": 2.398921832884097,
      "grad_norm": 0.554608166217804,
      "learning_rate": 0.0004566022665947112,
      "loss": 3.5675,
      "step": 22250
    },
    {
      "epoch": 2.404312668463612,
      "grad_norm": 0.6008310317993164,
      "learning_rate": 0.0004562784673502428,
      "loss": 3.5556,
      "step": 22300
    },
    {
      "epoch": 2.4097035040431267,
      "grad_norm": 0.5571367144584656,
      "learning_rate": 0.0004559546681057744,
      "loss": 3.5784,
      "step": 22350
    },
    {
      "epoch": 2.4150943396226414,
      "grad_norm": 0.5885669589042664,
      "learning_rate": 0.000455630868861306,
      "loss": 3.5604,
      "step": 22400
    },
    {
      "epoch": 2.420485175202156,
      "grad_norm": 0.574574887752533,
      "learning_rate": 0.00045530706961683753,
      "loss": 3.563,
      "step": 22450
    },
    {
      "epoch": 2.4258760107816713,
      "grad_norm": 0.5961259603500366,
      "learning_rate": 0.00045498327037236914,
      "loss": 3.5716,
      "step": 22500
    },
    {
      "epoch": 2.431266846361186,
      "grad_norm": 0.5879604816436768,
      "learning_rate": 0.0004546594711279007,
      "loss": 3.5629,
      "step": 22550
    },
    {
      "epoch": 2.4366576819407006,
      "grad_norm": 0.6066700220108032,
      "learning_rate": 0.0004543356718834322,
      "loss": 3.5943,
      "step": 22600
    },
    {
      "epoch": 2.442048517520216,
      "grad_norm": 0.5698487758636475,
      "learning_rate": 0.0004540118726389638,
      "loss": 3.5546,
      "step": 22650
    },
    {
      "epoch": 2.4474393530997305,
      "grad_norm": 0.5691965818405151,
      "learning_rate": 0.00045368807339449534,
      "loss": 3.5457,
      "step": 22700
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.5681477785110474,
      "learning_rate": 0.00045336427415002694,
      "loss": 3.5509,
      "step": 22750
    },
    {
      "epoch": 2.4582210242587603,
      "grad_norm": 0.5589949488639832,
      "learning_rate": 0.0004530404749055585,
      "loss": 3.5605,
      "step": 22800
    },
    {
      "epoch": 2.463611859838275,
      "grad_norm": 0.5612940192222595,
      "learning_rate": 0.0004527166756610901,
      "loss": 3.5464,
      "step": 22850
    },
    {
      "epoch": 2.4690026954177897,
      "grad_norm": 0.5747497081756592,
      "learning_rate": 0.00045239287641662165,
      "loss": 3.5627,
      "step": 22900
    },
    {
      "epoch": 2.4743935309973044,
      "grad_norm": 0.5754920840263367,
      "learning_rate": 0.0004520690771721532,
      "loss": 3.5591,
      "step": 22950
    },
    {
      "epoch": 2.4797843665768196,
      "grad_norm": 0.5711870193481445,
      "learning_rate": 0.0004517452779276848,
      "loss": 3.5526,
      "step": 23000
    },
    {
      "epoch": 2.4797843665768196,
      "eval_accuracy": 0.3650906529071547,
      "eval_loss": 3.546846389770508,
      "eval_runtime": 188.1187,
      "eval_samples_per_second": 95.743,
      "eval_steps_per_second": 5.986,
      "step": 23000
    },
    {
      "epoch": 2.4851752021563343,
      "grad_norm": 0.5261489152908325,
      "learning_rate": 0.00045142147868321636,
      "loss": 3.5626,
      "step": 23050
    },
    {
      "epoch": 2.490566037735849,
      "grad_norm": 0.6213911771774292,
      "learning_rate": 0.0004511041554236373,
      "loss": 3.566,
      "step": 23100
    },
    {
      "epoch": 2.4959568733153636,
      "grad_norm": 0.5985024571418762,
      "learning_rate": 0.0004507803561791689,
      "loss": 3.5628,
      "step": 23150
    },
    {
      "epoch": 2.501347708894879,
      "grad_norm": 0.576748251914978,
      "learning_rate": 0.00045045655693470046,
      "loss": 3.5601,
      "step": 23200
    },
    {
      "epoch": 2.5067385444743935,
      "grad_norm": 0.6227774024009705,
      "learning_rate": 0.00045013275769023207,
      "loss": 3.5594,
      "step": 23250
    },
    {
      "epoch": 2.512129380053908,
      "grad_norm": 0.5366291403770447,
      "learning_rate": 0.00044980895844576356,
      "loss": 3.5443,
      "step": 23300
    },
    {
      "epoch": 2.5175202156334233,
      "grad_norm": 0.6307509541511536,
      "learning_rate": 0.0004494851592012951,
      "loss": 3.5499,
      "step": 23350
    },
    {
      "epoch": 2.522911051212938,
      "grad_norm": 0.5795412063598633,
      "learning_rate": 0.0004491613599568267,
      "loss": 3.5467,
      "step": 23400
    },
    {
      "epoch": 2.5283018867924527,
      "grad_norm": 0.5431268215179443,
      "learning_rate": 0.00044883756071235827,
      "loss": 3.5585,
      "step": 23450
    },
    {
      "epoch": 2.533692722371968,
      "grad_norm": 0.5200484395027161,
      "learning_rate": 0.0004485137614678899,
      "loss": 3.5626,
      "step": 23500
    },
    {
      "epoch": 2.5390835579514826,
      "grad_norm": 0.5646335482597351,
      "learning_rate": 0.00044819643820831083,
      "loss": 3.5698,
      "step": 23550
    },
    {
      "epoch": 2.5444743935309972,
      "grad_norm": 0.5931826829910278,
      "learning_rate": 0.0004478726389638424,
      "loss": 3.5731,
      "step": 23600
    },
    {
      "epoch": 2.5498652291105124,
      "grad_norm": 0.5869390964508057,
      "learning_rate": 0.00044754883971937393,
      "loss": 3.5585,
      "step": 23650
    },
    {
      "epoch": 2.555256064690027,
      "grad_norm": 0.5838721394538879,
      "learning_rate": 0.00044722504047490553,
      "loss": 3.5635,
      "step": 23700
    },
    {
      "epoch": 2.560646900269542,
      "grad_norm": 0.5717689394950867,
      "learning_rate": 0.0004469012412304371,
      "loss": 3.5562,
      "step": 23750
    },
    {
      "epoch": 2.5660377358490565,
      "grad_norm": 0.588871419429779,
      "learning_rate": 0.0004465774419859687,
      "loss": 3.5684,
      "step": 23800
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.6461533904075623,
      "learning_rate": 0.00044625364274150024,
      "loss": 3.5753,
      "step": 23850
    },
    {
      "epoch": 2.5768194070080863,
      "grad_norm": 0.5436891913414001,
      "learning_rate": 0.00044592984349703184,
      "loss": 3.5734,
      "step": 23900
    },
    {
      "epoch": 2.582210242587601,
      "grad_norm": 0.625299334526062,
      "learning_rate": 0.00044560604425256334,
      "loss": 3.5485,
      "step": 23950
    },
    {
      "epoch": 2.5876010781671157,
      "grad_norm": 0.5579541325569153,
      "learning_rate": 0.000445282245008095,
      "loss": 3.5451,
      "step": 24000
    },
    {
      "epoch": 2.5876010781671157,
      "eval_accuracy": 0.36589175064901086,
      "eval_loss": 3.5402700901031494,
      "eval_runtime": 187.5335,
      "eval_samples_per_second": 96.042,
      "eval_steps_per_second": 6.004,
      "step": 24000
    },
    {
      "epoch": 2.592991913746631,
      "grad_norm": 0.6185333132743835,
      "learning_rate": 0.0004449584457636265,
      "loss": 3.5467,
      "step": 24050
    },
    {
      "epoch": 2.5983827493261455,
      "grad_norm": 0.5775592923164368,
      "learning_rate": 0.00044463464651915805,
      "loss": 3.5458,
      "step": 24100
    },
    {
      "epoch": 2.6037735849056602,
      "grad_norm": 0.5688097476959229,
      "learning_rate": 0.00044431084727468965,
      "loss": 3.55,
      "step": 24150
    },
    {
      "epoch": 2.6091644204851754,
      "grad_norm": 0.6222105026245117,
      "learning_rate": 0.0004439870480302212,
      "loss": 3.565,
      "step": 24200
    },
    {
      "epoch": 2.61455525606469,
      "grad_norm": 0.5622762441635132,
      "learning_rate": 0.0004436632487857528,
      "loss": 3.5693,
      "step": 24250
    },
    {
      "epoch": 2.6199460916442048,
      "grad_norm": 0.6001819968223572,
      "learning_rate": 0.00044333944954128436,
      "loss": 3.5543,
      "step": 24300
    },
    {
      "epoch": 2.62533692722372,
      "grad_norm": 0.6119269728660583,
      "learning_rate": 0.00044301565029681596,
      "loss": 3.5469,
      "step": 24350
    },
    {
      "epoch": 2.6307277628032346,
      "grad_norm": 0.5447987914085388,
      "learning_rate": 0.0004426918510523475,
      "loss": 3.5599,
      "step": 24400
    },
    {
      "epoch": 2.6361185983827493,
      "grad_norm": 0.5981472134590149,
      "learning_rate": 0.0004423680518078791,
      "loss": 3.5324,
      "step": 24450
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.6249266266822815,
      "learning_rate": 0.00044204425256341067,
      "loss": 3.5395,
      "step": 24500
    },
    {
      "epoch": 2.6469002695417787,
      "grad_norm": 0.6205341815948486,
      "learning_rate": 0.0004417204533189422,
      "loss": 3.5569,
      "step": 24550
    },
    {
      "epoch": 2.652291105121294,
      "grad_norm": 0.609792172908783,
      "learning_rate": 0.0004413966540744738,
      "loss": 3.5346,
      "step": 24600
    },
    {
      "epoch": 2.6576819407008085,
      "grad_norm": 0.5270373821258545,
      "learning_rate": 0.0004410728548300053,
      "loss": 3.5518,
      "step": 24650
    },
    {
      "epoch": 2.6630727762803232,
      "grad_norm": 0.5638793706893921,
      "learning_rate": 0.0004407490555855369,
      "loss": 3.5521,
      "step": 24700
    },
    {
      "epoch": 2.6684636118598384,
      "grad_norm": 0.6222891211509705,
      "learning_rate": 0.0004404252563410685,
      "loss": 3.5552,
      "step": 24750
    },
    {
      "epoch": 2.673854447439353,
      "grad_norm": 0.5870048999786377,
      "learning_rate": 0.0004401014570966001,
      "loss": 3.5535,
      "step": 24800
    },
    {
      "epoch": 2.6792452830188678,
      "grad_norm": 0.5811405181884766,
      "learning_rate": 0.00043977765785213163,
      "loss": 3.5552,
      "step": 24850
    },
    {
      "epoch": 2.684636118598383,
      "grad_norm": 0.6036756038665771,
      "learning_rate": 0.0004394538586076632,
      "loss": 3.5722,
      "step": 24900
    },
    {
      "epoch": 2.6900269541778976,
      "grad_norm": 0.6059266328811646,
      "learning_rate": 0.0004391300593631948,
      "loss": 3.561,
      "step": 24950
    },
    {
      "epoch": 2.6954177897574123,
      "grad_norm": 0.559899091720581,
      "learning_rate": 0.00043880626011872634,
      "loss": 3.555,
      "step": 25000
    },
    {
      "epoch": 2.6954177897574123,
      "eval_accuracy": 0.36689939753060236,
      "eval_loss": 3.527742385864258,
      "eval_runtime": 187.8058,
      "eval_samples_per_second": 95.902,
      "eval_steps_per_second": 5.996,
      "step": 25000
    },
    {
      "epoch": 2.7008086253369274,
      "grad_norm": 0.5817413330078125,
      "learning_rate": 0.00043848246087425794,
      "loss": 3.5455,
      "step": 25050
    },
    {
      "epoch": 2.706199460916442,
      "grad_norm": 0.5311758518218994,
      "learning_rate": 0.0004381586616297895,
      "loss": 3.5554,
      "step": 25100
    },
    {
      "epoch": 2.711590296495957,
      "grad_norm": 0.6347346305847168,
      "learning_rate": 0.0004378348623853211,
      "loss": 3.5307,
      "step": 25150
    },
    {
      "epoch": 2.7169811320754715,
      "grad_norm": 0.6582555770874023,
      "learning_rate": 0.00043751106314085265,
      "loss": 3.552,
      "step": 25200
    },
    {
      "epoch": 2.7223719676549867,
      "grad_norm": 0.6206474304199219,
      "learning_rate": 0.00043718726389638425,
      "loss": 3.5568,
      "step": 25250
    },
    {
      "epoch": 2.7277628032345014,
      "grad_norm": 0.5440530776977539,
      "learning_rate": 0.00043686346465191575,
      "loss": 3.5593,
      "step": 25300
    },
    {
      "epoch": 2.733153638814016,
      "grad_norm": 0.5995965600013733,
      "learning_rate": 0.0004365396654074473,
      "loss": 3.5496,
      "step": 25350
    },
    {
      "epoch": 2.7385444743935308,
      "grad_norm": 0.7029544115066528,
      "learning_rate": 0.0004362158661629789,
      "loss": 3.5409,
      "step": 25400
    },
    {
      "epoch": 2.743935309973046,
      "grad_norm": 0.5936011075973511,
      "learning_rate": 0.00043589206691851045,
      "loss": 3.5428,
      "step": 25450
    },
    {
      "epoch": 2.7493261455525606,
      "grad_norm": 0.5993081331253052,
      "learning_rate": 0.00043556826767404206,
      "loss": 3.5572,
      "step": 25500
    },
    {
      "epoch": 2.7547169811320753,
      "grad_norm": 0.5754674077033997,
      "learning_rate": 0.0004352444684295736,
      "loss": 3.5503,
      "step": 25550
    },
    {
      "epoch": 2.7601078167115904,
      "grad_norm": 0.5702727437019348,
      "learning_rate": 0.0004349206691851052,
      "loss": 3.5517,
      "step": 25600
    },
    {
      "epoch": 2.765498652291105,
      "grad_norm": 0.597646951675415,
      "learning_rate": 0.00043459686994063676,
      "loss": 3.544,
      "step": 25650
    },
    {
      "epoch": 2.77088948787062,
      "grad_norm": 0.6668535470962524,
      "learning_rate": 0.00043427307069616837,
      "loss": 3.5493,
      "step": 25700
    },
    {
      "epoch": 2.776280323450135,
      "grad_norm": 0.564940333366394,
      "learning_rate": 0.0004339492714516999,
      "loss": 3.5378,
      "step": 25750
    },
    {
      "epoch": 2.7816711590296497,
      "grad_norm": 0.5661826729774475,
      "learning_rate": 0.00043362547220723147,
      "loss": 3.5621,
      "step": 25800
    },
    {
      "epoch": 2.7870619946091644,
      "grad_norm": 0.5790637135505676,
      "learning_rate": 0.0004333016729627631,
      "loss": 3.5346,
      "step": 25850
    },
    {
      "epoch": 2.7924528301886795,
      "grad_norm": 0.5834683775901794,
      "learning_rate": 0.0004329778737182946,
      "loss": 3.5441,
      "step": 25900
    },
    {
      "epoch": 2.797843665768194,
      "grad_norm": 0.6243922114372253,
      "learning_rate": 0.00043265407447382623,
      "loss": 3.5435,
      "step": 25950
    },
    {
      "epoch": 2.803234501347709,
      "grad_norm": 0.5399969816207886,
      "learning_rate": 0.0004323302752293577,
      "loss": 3.5566,
      "step": 26000
    },
    {
      "epoch": 2.803234501347709,
      "eval_accuracy": 0.3679905984829232,
      "eval_loss": 3.520369529724121,
      "eval_runtime": 187.7249,
      "eval_samples_per_second": 95.944,
      "eval_steps_per_second": 5.998,
      "step": 26000
    },
    {
      "epoch": 2.8086253369272236,
      "grad_norm": 0.6081315279006958,
      "learning_rate": 0.00043200647598488933,
      "loss": 3.5645,
      "step": 26050
    },
    {
      "epoch": 2.8140161725067383,
      "grad_norm": 0.6176601648330688,
      "learning_rate": 0.0004316826767404209,
      "loss": 3.5486,
      "step": 26100
    },
    {
      "epoch": 2.8194070080862534,
      "grad_norm": 0.6038433313369751,
      "learning_rate": 0.0004313588774959525,
      "loss": 3.542,
      "step": 26150
    },
    {
      "epoch": 2.824797843665768,
      "grad_norm": 0.5484240055084229,
      "learning_rate": 0.00043103507825148404,
      "loss": 3.5549,
      "step": 26200
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.5879635214805603,
      "learning_rate": 0.0004307112790070156,
      "loss": 3.5369,
      "step": 26250
    },
    {
      "epoch": 2.835579514824798,
      "grad_norm": 0.6630586981773376,
      "learning_rate": 0.0004303874797625472,
      "loss": 3.5354,
      "step": 26300
    },
    {
      "epoch": 2.8409703504043127,
      "grad_norm": 0.6347656846046448,
      "learning_rate": 0.00043006368051807874,
      "loss": 3.5574,
      "step": 26350
    },
    {
      "epoch": 2.8463611859838274,
      "grad_norm": 0.5696792006492615,
      "learning_rate": 0.00042973988127361035,
      "loss": 3.531,
      "step": 26400
    },
    {
      "epoch": 2.8517520215633425,
      "grad_norm": 0.5842821598052979,
      "learning_rate": 0.0004294160820291419,
      "loss": 3.5532,
      "step": 26450
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.5837120413780212,
      "learning_rate": 0.0004290922827846735,
      "loss": 3.5602,
      "step": 26500
    },
    {
      "epoch": 2.862533692722372,
      "grad_norm": 0.652936577796936,
      "learning_rate": 0.00042876848354020505,
      "loss": 3.5401,
      "step": 26550
    },
    {
      "epoch": 2.867924528301887,
      "grad_norm": 0.5488174557685852,
      "learning_rate": 0.00042844468429573655,
      "loss": 3.5421,
      "step": 26600
    },
    {
      "epoch": 2.8733153638814017,
      "grad_norm": 0.5535524487495422,
      "learning_rate": 0.00042812088505126815,
      "loss": 3.5271,
      "step": 26650
    },
    {
      "epoch": 2.8787061994609164,
      "grad_norm": 0.6259840130805969,
      "learning_rate": 0.0004277970858067997,
      "loss": 3.5667,
      "step": 26700
    },
    {
      "epoch": 2.884097035040431,
      "grad_norm": 0.626735270023346,
      "learning_rate": 0.0004274732865623313,
      "loss": 3.5335,
      "step": 26750
    },
    {
      "epoch": 2.889487870619946,
      "grad_norm": 0.565473198890686,
      "learning_rate": 0.00042714948731786286,
      "loss": 3.5181,
      "step": 26800
    },
    {
      "epoch": 2.894878706199461,
      "grad_norm": 0.6019836664199829,
      "learning_rate": 0.00042682568807339447,
      "loss": 3.5218,
      "step": 26850
    },
    {
      "epoch": 2.9002695417789757,
      "grad_norm": 0.5418981909751892,
      "learning_rate": 0.000426501888828926,
      "loss": 3.5398,
      "step": 26900
    },
    {
      "epoch": 2.9056603773584904,
      "grad_norm": 0.6377326250076294,
      "learning_rate": 0.0004261780895844576,
      "loss": 3.5597,
      "step": 26950
    },
    {
      "epoch": 2.9110512129380055,
      "grad_norm": 0.5947117805480957,
      "learning_rate": 0.00042585429033998917,
      "loss": 3.5438,
      "step": 27000
    },
    {
      "epoch": 2.9110512129380055,
      "eval_accuracy": 0.3689144739880097,
      "eval_loss": 3.5099987983703613,
      "eval_runtime": 187.8098,
      "eval_samples_per_second": 95.9,
      "eval_steps_per_second": 5.995,
      "step": 27000
    },
    {
      "epoch": 2.91644204851752,
      "grad_norm": 0.6234594583511353,
      "learning_rate": 0.0004255304910955207,
      "loss": 3.5303,
      "step": 27050
    },
    {
      "epoch": 2.921832884097035,
      "grad_norm": 0.5786582827568054,
      "learning_rate": 0.0004252066918510523,
      "loss": 3.5387,
      "step": 27100
    },
    {
      "epoch": 2.92722371967655,
      "grad_norm": 0.6039722561836243,
      "learning_rate": 0.0004248828926065839,
      "loss": 3.5334,
      "step": 27150
    },
    {
      "epoch": 2.9326145552560647,
      "grad_norm": 0.608992338180542,
      "learning_rate": 0.0004245590933621155,
      "loss": 3.544,
      "step": 27200
    },
    {
      "epoch": 2.9380053908355794,
      "grad_norm": 0.5596150755882263,
      "learning_rate": 0.00042423529411764703,
      "loss": 3.5481,
      "step": 27250
    },
    {
      "epoch": 2.9433962264150946,
      "grad_norm": 0.5880177617073059,
      "learning_rate": 0.00042391149487317864,
      "loss": 3.5136,
      "step": 27300
    },
    {
      "epoch": 2.9487870619946093,
      "grad_norm": 0.5926826596260071,
      "learning_rate": 0.00042358769562871013,
      "loss": 3.5327,
      "step": 27350
    },
    {
      "epoch": 2.954177897574124,
      "grad_norm": 0.5690999031066895,
      "learning_rate": 0.00042326389638424174,
      "loss": 3.5271,
      "step": 27400
    },
    {
      "epoch": 2.9595687331536387,
      "grad_norm": 0.5594161748886108,
      "learning_rate": 0.0004229400971397733,
      "loss": 3.5225,
      "step": 27450
    },
    {
      "epoch": 2.964959568733154,
      "grad_norm": 0.5708371996879578,
      "learning_rate": 0.00042261629789530484,
      "loss": 3.5475,
      "step": 27500
    },
    {
      "epoch": 2.9703504043126685,
      "grad_norm": 0.561190664768219,
      "learning_rate": 0.00042229249865083644,
      "loss": 3.5487,
      "step": 27550
    },
    {
      "epoch": 2.975741239892183,
      "grad_norm": 0.6065704226493835,
      "learning_rate": 0.0004219751753912574,
      "loss": 3.5301,
      "step": 27600
    },
    {
      "epoch": 2.981132075471698,
      "grad_norm": 0.6116462349891663,
      "learning_rate": 0.00042165137614678895,
      "loss": 3.5354,
      "step": 27650
    },
    {
      "epoch": 2.986522911051213,
      "grad_norm": 0.5894355177879333,
      "learning_rate": 0.00042132757690232055,
      "loss": 3.5255,
      "step": 27700
    },
    {
      "epoch": 2.9919137466307277,
      "grad_norm": 0.6621140837669373,
      "learning_rate": 0.0004210037776578521,
      "loss": 3.5259,
      "step": 27750
    },
    {
      "epoch": 2.9973045822102424,
      "grad_norm": 0.5993552207946777,
      "learning_rate": 0.00042067997841338365,
      "loss": 3.533,
      "step": 27800
    },
    {
      "epoch": 3.0026954177897576,
      "grad_norm": 0.5765702724456787,
      "learning_rate": 0.00042035617916891526,
      "loss": 3.4951,
      "step": 27850
    },
    {
      "epoch": 3.0080862533692723,
      "grad_norm": 0.6206012964248657,
      "learning_rate": 0.0004200323799244468,
      "loss": 3.4419,
      "step": 27900
    },
    {
      "epoch": 3.013477088948787,
      "grad_norm": 0.5946252942085266,
      "learning_rate": 0.0004197085806799784,
      "loss": 3.4422,
      "step": 27950
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.5602282881736755,
      "learning_rate": 0.0004193847814355099,
      "loss": 3.4421,
      "step": 28000
    },
    {
      "epoch": 3.018867924528302,
      "eval_accuracy": 0.36948424973170885,
      "eval_loss": 3.5056018829345703,
      "eval_runtime": 187.3106,
      "eval_samples_per_second": 96.156,
      "eval_steps_per_second": 6.011,
      "step": 28000
    },
    {
      "epoch": 3.024258760107817,
      "grad_norm": 0.565204381942749,
      "learning_rate": 0.0004190609821910415,
      "loss": 3.4405,
      "step": 28050
    },
    {
      "epoch": 3.0296495956873315,
      "grad_norm": 0.5876172184944153,
      "learning_rate": 0.00041873718294657306,
      "loss": 3.4473,
      "step": 28100
    },
    {
      "epoch": 3.035040431266846,
      "grad_norm": 0.6169458031654358,
      "learning_rate": 0.00041841338370210467,
      "loss": 3.4312,
      "step": 28150
    },
    {
      "epoch": 3.0404312668463613,
      "grad_norm": 0.5794179439544678,
      "learning_rate": 0.0004180895844576362,
      "loss": 3.4471,
      "step": 28200
    },
    {
      "epoch": 3.045822102425876,
      "grad_norm": 0.5454350709915161,
      "learning_rate": 0.00041776578521316777,
      "loss": 3.4392,
      "step": 28250
    },
    {
      "epoch": 3.0512129380053907,
      "grad_norm": 0.5963560938835144,
      "learning_rate": 0.0004174419859686994,
      "loss": 3.4641,
      "step": 28300
    },
    {
      "epoch": 3.056603773584906,
      "grad_norm": 0.6112507581710815,
      "learning_rate": 0.0004171181867242309,
      "loss": 3.4487,
      "step": 28350
    },
    {
      "epoch": 3.0619946091644206,
      "grad_norm": 0.5972319841384888,
      "learning_rate": 0.00041679438747976253,
      "loss": 3.4692,
      "step": 28400
    },
    {
      "epoch": 3.0673854447439353,
      "grad_norm": 0.5812078714370728,
      "learning_rate": 0.0004164705882352941,
      "loss": 3.4462,
      "step": 28450
    },
    {
      "epoch": 3.07277628032345,
      "grad_norm": 0.6155223250389099,
      "learning_rate": 0.0004161467889908257,
      "loss": 3.4396,
      "step": 28500
    },
    {
      "epoch": 3.078167115902965,
      "grad_norm": 0.6197593212127686,
      "learning_rate": 0.00041582298974635724,
      "loss": 3.4435,
      "step": 28550
    },
    {
      "epoch": 3.08355795148248,
      "grad_norm": 0.6040982604026794,
      "learning_rate": 0.00041549919050188884,
      "loss": 3.4621,
      "step": 28600
    },
    {
      "epoch": 3.0889487870619945,
      "grad_norm": 0.7128732204437256,
      "learning_rate": 0.00041517539125742034,
      "loss": 3.4482,
      "step": 28650
    },
    {
      "epoch": 3.0943396226415096,
      "grad_norm": 0.5997539162635803,
      "learning_rate": 0.0004148515920129519,
      "loss": 3.4704,
      "step": 28700
    },
    {
      "epoch": 3.0997304582210243,
      "grad_norm": 0.6413683891296387,
      "learning_rate": 0.0004145277927684835,
      "loss": 3.4344,
      "step": 28750
    },
    {
      "epoch": 3.105121293800539,
      "grad_norm": 0.5954497456550598,
      "learning_rate": 0.00041420399352401504,
      "loss": 3.465,
      "step": 28800
    },
    {
      "epoch": 3.1105121293800537,
      "grad_norm": 0.6719018220901489,
      "learning_rate": 0.00041388019427954665,
      "loss": 3.4504,
      "step": 28850
    },
    {
      "epoch": 3.115902964959569,
      "grad_norm": 0.5712775588035583,
      "learning_rate": 0.0004135563950350782,
      "loss": 3.4676,
      "step": 28900
    },
    {
      "epoch": 3.1212938005390836,
      "grad_norm": 0.5765361785888672,
      "learning_rate": 0.0004132325957906098,
      "loss": 3.4634,
      "step": 28950
    },
    {
      "epoch": 3.1266846361185983,
      "grad_norm": 0.6526620388031006,
      "learning_rate": 0.00041290879654614135,
      "loss": 3.4609,
      "step": 29000
    },
    {
      "epoch": 3.1266846361185983,
      "eval_accuracy": 0.37030783862134264,
      "eval_loss": 3.5025389194488525,
      "eval_runtime": 187.8717,
      "eval_samples_per_second": 95.869,
      "eval_steps_per_second": 5.993,
      "step": 29000
    },
    {
      "epoch": 3.1320754716981134,
      "grad_norm": 0.6089078187942505,
      "learning_rate": 0.00041258499730167296,
      "loss": 3.4653,
      "step": 29050
    },
    {
      "epoch": 3.137466307277628,
      "grad_norm": 0.5929785370826721,
      "learning_rate": 0.0004122611980572045,
      "loss": 3.4622,
      "step": 29100
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.6185963749885559,
      "learning_rate": 0.00041193739881273606,
      "loss": 3.4663,
      "step": 29150
    },
    {
      "epoch": 3.1482479784366575,
      "grad_norm": 0.5839598774909973,
      "learning_rate": 0.000411620075553157,
      "loss": 3.4628,
      "step": 29200
    },
    {
      "epoch": 3.1536388140161726,
      "grad_norm": 0.6440391540527344,
      "learning_rate": 0.0004112962763086886,
      "loss": 3.4815,
      "step": 29250
    },
    {
      "epoch": 3.1590296495956873,
      "grad_norm": 0.655301570892334,
      "learning_rate": 0.0004109724770642201,
      "loss": 3.4671,
      "step": 29300
    },
    {
      "epoch": 3.164420485175202,
      "grad_norm": 0.5958247780799866,
      "learning_rate": 0.00041064867781975177,
      "loss": 3.4627,
      "step": 29350
    },
    {
      "epoch": 3.169811320754717,
      "grad_norm": 0.581174910068512,
      "learning_rate": 0.00041032487857528327,
      "loss": 3.4552,
      "step": 29400
    },
    {
      "epoch": 3.175202156334232,
      "grad_norm": 0.6023386120796204,
      "learning_rate": 0.0004100010793308148,
      "loss": 3.4664,
      "step": 29450
    },
    {
      "epoch": 3.1805929919137466,
      "grad_norm": 0.6177906394004822,
      "learning_rate": 0.0004096772800863464,
      "loss": 3.4619,
      "step": 29500
    },
    {
      "epoch": 3.1859838274932613,
      "grad_norm": 0.6565192937850952,
      "learning_rate": 0.000409353480841878,
      "loss": 3.4737,
      "step": 29550
    },
    {
      "epoch": 3.1913746630727764,
      "grad_norm": 0.5951600670814514,
      "learning_rate": 0.0004090296815974096,
      "loss": 3.4712,
      "step": 29600
    },
    {
      "epoch": 3.196765498652291,
      "grad_norm": 0.636186957359314,
      "learning_rate": 0.00040870588235294113,
      "loss": 3.4576,
      "step": 29650
    },
    {
      "epoch": 3.202156334231806,
      "grad_norm": 0.6101711988449097,
      "learning_rate": 0.00040838208310847273,
      "loss": 3.4893,
      "step": 29700
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.634506344795227,
      "learning_rate": 0.0004080582838640043,
      "loss": 3.459,
      "step": 29750
    },
    {
      "epoch": 3.2129380053908356,
      "grad_norm": 0.6397202610969543,
      "learning_rate": 0.0004077344846195359,
      "loss": 3.4702,
      "step": 29800
    },
    {
      "epoch": 3.2183288409703503,
      "grad_norm": 0.6291985511779785,
      "learning_rate": 0.00040741068537506744,
      "loss": 3.4467,
      "step": 29850
    },
    {
      "epoch": 3.223719676549865,
      "grad_norm": 0.5848795771598816,
      "learning_rate": 0.000407086886130599,
      "loss": 3.4579,
      "step": 29900
    },
    {
      "epoch": 3.22911051212938,
      "grad_norm": 0.5925343036651611,
      "learning_rate": 0.0004067630868861306,
      "loss": 3.4721,
      "step": 29950
    },
    {
      "epoch": 3.234501347708895,
      "grad_norm": 0.5578032732009888,
      "learning_rate": 0.0004064392876416621,
      "loss": 3.4737,
      "step": 30000
    },
    {
      "epoch": 3.234501347708895,
      "eval_accuracy": 0.37095921268379045,
      "eval_loss": 3.4953773021698,
      "eval_runtime": 187.3567,
      "eval_samples_per_second": 96.132,
      "eval_steps_per_second": 6.01,
      "step": 30000
    },
    {
      "epoch": 3.2398921832884096,
      "grad_norm": 0.6164473295211792,
      "learning_rate": 0.0004061154883971937,
      "loss": 3.4777,
      "step": 30050
    },
    {
      "epoch": 3.2452830188679247,
      "grad_norm": 0.5678409337997437,
      "learning_rate": 0.00040579168915272525,
      "loss": 3.4606,
      "step": 30100
    },
    {
      "epoch": 3.2506738544474394,
      "grad_norm": 0.6322740912437439,
      "learning_rate": 0.00040546788990825685,
      "loss": 3.4631,
      "step": 30150
    },
    {
      "epoch": 3.256064690026954,
      "grad_norm": 0.5954024791717529,
      "learning_rate": 0.0004051440906637884,
      "loss": 3.4663,
      "step": 30200
    },
    {
      "epoch": 3.2614555256064692,
      "grad_norm": 0.5793306231498718,
      "learning_rate": 0.00040482029141931995,
      "loss": 3.4458,
      "step": 30250
    },
    {
      "epoch": 3.266846361185984,
      "grad_norm": 0.5884698629379272,
      "learning_rate": 0.00040449649217485156,
      "loss": 3.479,
      "step": 30300
    },
    {
      "epoch": 3.2722371967654986,
      "grad_norm": 0.5904477834701538,
      "learning_rate": 0.0004041726929303831,
      "loss": 3.4627,
      "step": 30350
    },
    {
      "epoch": 3.2776280323450133,
      "grad_norm": 0.5925611853599548,
      "learning_rate": 0.0004038488936859147,
      "loss": 3.465,
      "step": 30400
    },
    {
      "epoch": 3.2830188679245285,
      "grad_norm": 0.6218023300170898,
      "learning_rate": 0.00040352509444144626,
      "loss": 3.4659,
      "step": 30450
    },
    {
      "epoch": 3.288409703504043,
      "grad_norm": 0.5947563648223877,
      "learning_rate": 0.00040320129519697787,
      "loss": 3.4958,
      "step": 30500
    },
    {
      "epoch": 3.293800539083558,
      "grad_norm": 0.6579247713088989,
      "learning_rate": 0.0004028774959525094,
      "loss": 3.4644,
      "step": 30550
    },
    {
      "epoch": 3.2991913746630726,
      "grad_norm": 0.5691560506820679,
      "learning_rate": 0.000402553696708041,
      "loss": 3.4724,
      "step": 30600
    },
    {
      "epoch": 3.3045822102425877,
      "grad_norm": 0.5954251885414124,
      "learning_rate": 0.0004022298974635726,
      "loss": 3.4599,
      "step": 30650
    },
    {
      "epoch": 3.3099730458221024,
      "grad_norm": 0.6152036190032959,
      "learning_rate": 0.00040190609821910407,
      "loss": 3.4639,
      "step": 30700
    },
    {
      "epoch": 3.315363881401617,
      "grad_norm": 0.5858497023582458,
      "learning_rate": 0.0004015822989746357,
      "loss": 3.454,
      "step": 30750
    },
    {
      "epoch": 3.3207547169811322,
      "grad_norm": 0.5813567042350769,
      "learning_rate": 0.0004012584997301672,
      "loss": 3.4685,
      "step": 30800
    },
    {
      "epoch": 3.326145552560647,
      "grad_norm": 0.6456325054168701,
      "learning_rate": 0.00040093470048569883,
      "loss": 3.4794,
      "step": 30850
    },
    {
      "epoch": 3.3315363881401616,
      "grad_norm": 0.6200717687606812,
      "learning_rate": 0.0004006109012412304,
      "loss": 3.4679,
      "step": 30900
    },
    {
      "epoch": 3.3369272237196768,
      "grad_norm": 0.6692876815795898,
      "learning_rate": 0.000400287101996762,
      "loss": 3.4739,
      "step": 30950
    },
    {
      "epoch": 3.3423180592991915,
      "grad_norm": 0.5814414024353027,
      "learning_rate": 0.00039996330275229354,
      "loss": 3.4623,
      "step": 31000
    },
    {
      "epoch": 3.3423180592991915,
      "eval_accuracy": 0.3720433511984033,
      "eval_loss": 3.4902048110961914,
      "eval_runtime": 187.701,
      "eval_samples_per_second": 95.956,
      "eval_steps_per_second": 5.999,
      "step": 31000
    },
    {
      "epoch": 3.347708894878706,
      "grad_norm": 0.6048328876495361,
      "learning_rate": 0.00039963950350782514,
      "loss": 3.473,
      "step": 31050
    },
    {
      "epoch": 3.353099730458221,
      "grad_norm": 0.6048550009727478,
      "learning_rate": 0.0003993157042633567,
      "loss": 3.4611,
      "step": 31100
    },
    {
      "epoch": 3.358490566037736,
      "grad_norm": 0.6270948648452759,
      "learning_rate": 0.00039899190501888824,
      "loss": 3.4633,
      "step": 31150
    },
    {
      "epoch": 3.3638814016172507,
      "grad_norm": 0.6166711449623108,
      "learning_rate": 0.00039866810577441985,
      "loss": 3.473,
      "step": 31200
    },
    {
      "epoch": 3.3692722371967654,
      "grad_norm": 0.6020999550819397,
      "learning_rate": 0.0003983443065299514,
      "loss": 3.4617,
      "step": 31250
    },
    {
      "epoch": 3.37466307277628,
      "grad_norm": 0.6348170042037964,
      "learning_rate": 0.00039802698327037235,
      "loss": 3.4698,
      "step": 31300
    },
    {
      "epoch": 3.3800539083557952,
      "grad_norm": 0.6470772624015808,
      "learning_rate": 0.00039770318402590396,
      "loss": 3.4731,
      "step": 31350
    },
    {
      "epoch": 3.38544474393531,
      "grad_norm": 0.6075937747955322,
      "learning_rate": 0.00039737938478143545,
      "loss": 3.4858,
      "step": 31400
    },
    {
      "epoch": 3.3908355795148246,
      "grad_norm": 0.6555103063583374,
      "learning_rate": 0.000397055585536967,
      "loss": 3.4644,
      "step": 31450
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.5855467915534973,
      "learning_rate": 0.0003967317862924986,
      "loss": 3.4624,
      "step": 31500
    },
    {
      "epoch": 3.4016172506738545,
      "grad_norm": 0.6008960008621216,
      "learning_rate": 0.00039640798704803016,
      "loss": 3.449,
      "step": 31550
    },
    {
      "epoch": 3.407008086253369,
      "grad_norm": 0.6120670437812805,
      "learning_rate": 0.00039608418780356176,
      "loss": 3.4585,
      "step": 31600
    },
    {
      "epoch": 3.4123989218328843,
      "grad_norm": 0.6167988777160645,
      "learning_rate": 0.0003957603885590933,
      "loss": 3.4748,
      "step": 31650
    },
    {
      "epoch": 3.417789757412399,
      "grad_norm": 0.6113810539245605,
      "learning_rate": 0.0003954365893146249,
      "loss": 3.4616,
      "step": 31700
    },
    {
      "epoch": 3.4231805929919137,
      "grad_norm": 0.6468194723129272,
      "learning_rate": 0.00039511279007015647,
      "loss": 3.4704,
      "step": 31750
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.6060384511947632,
      "learning_rate": 0.00039478899082568807,
      "loss": 3.4692,
      "step": 31800
    },
    {
      "epoch": 3.4339622641509435,
      "grad_norm": 0.5738762021064758,
      "learning_rate": 0.0003944651915812196,
      "loss": 3.4586,
      "step": 31850
    },
    {
      "epoch": 3.439353099730458,
      "grad_norm": 0.596107542514801,
      "learning_rate": 0.0003941413923367512,
      "loss": 3.459,
      "step": 31900
    },
    {
      "epoch": 3.444743935309973,
      "grad_norm": 0.5924981832504272,
      "learning_rate": 0.0003938175930922828,
      "loss": 3.4679,
      "step": 31950
    },
    {
      "epoch": 3.450134770889488,
      "grad_norm": 0.6087164282798767,
      "learning_rate": 0.0003934937938478143,
      "loss": 3.4713,
      "step": 32000
    },
    {
      "epoch": 3.450134770889488,
      "eval_accuracy": 0.3723299775164579,
      "eval_loss": 3.4831252098083496,
      "eval_runtime": 187.6862,
      "eval_samples_per_second": 95.963,
      "eval_steps_per_second": 5.999,
      "step": 32000
    },
    {
      "epoch": 3.4555256064690028,
      "grad_norm": 0.6510775089263916,
      "learning_rate": 0.0003931699946033459,
      "loss": 3.4671,
      "step": 32050
    },
    {
      "epoch": 3.4609164420485174,
      "grad_norm": 0.5941176414489746,
      "learning_rate": 0.00039284619535887743,
      "loss": 3.4686,
      "step": 32100
    },
    {
      "epoch": 3.466307277628032,
      "grad_norm": 0.5818312764167786,
      "learning_rate": 0.00039252239611440904,
      "loss": 3.4657,
      "step": 32150
    },
    {
      "epoch": 3.4716981132075473,
      "grad_norm": 0.6189693808555603,
      "learning_rate": 0.0003921985968699406,
      "loss": 3.4633,
      "step": 32200
    },
    {
      "epoch": 3.477088948787062,
      "grad_norm": 0.6840120553970337,
      "learning_rate": 0.0003918747976254722,
      "loss": 3.4742,
      "step": 32250
    },
    {
      "epoch": 3.4824797843665767,
      "grad_norm": 0.6052665710449219,
      "learning_rate": 0.00039155099838100374,
      "loss": 3.4845,
      "step": 32300
    },
    {
      "epoch": 3.487870619946092,
      "grad_norm": 0.6465663909912109,
      "learning_rate": 0.0003912271991365353,
      "loss": 3.4689,
      "step": 32350
    },
    {
      "epoch": 3.4932614555256065,
      "grad_norm": 0.5961518883705139,
      "learning_rate": 0.0003909033998920669,
      "loss": 3.4694,
      "step": 32400
    },
    {
      "epoch": 3.498652291105121,
      "grad_norm": 0.6497889161109924,
      "learning_rate": 0.00039057960064759845,
      "loss": 3.4529,
      "step": 32450
    },
    {
      "epoch": 3.5040431266846364,
      "grad_norm": 0.5722944736480713,
      "learning_rate": 0.00039025580140313005,
      "loss": 3.4766,
      "step": 32500
    },
    {
      "epoch": 3.509433962264151,
      "grad_norm": 0.6770778894424438,
      "learning_rate": 0.0003899320021586616,
      "loss": 3.4527,
      "step": 32550
    },
    {
      "epoch": 3.5148247978436657,
      "grad_norm": 0.6125558018684387,
      "learning_rate": 0.0003896082029141932,
      "loss": 3.4736,
      "step": 32600
    },
    {
      "epoch": 3.5202156334231804,
      "grad_norm": 0.6196644306182861,
      "learning_rate": 0.00038928440366972476,
      "loss": 3.4602,
      "step": 32650
    },
    {
      "epoch": 3.525606469002695,
      "grad_norm": 0.6249521374702454,
      "learning_rate": 0.00038896060442525636,
      "loss": 3.4555,
      "step": 32700
    },
    {
      "epoch": 3.5309973045822103,
      "grad_norm": 0.6278482675552368,
      "learning_rate": 0.00038863680518078786,
      "loss": 3.4667,
      "step": 32750
    },
    {
      "epoch": 3.536388140161725,
      "grad_norm": 0.6120947003364563,
      "learning_rate": 0.0003883130059363194,
      "loss": 3.4728,
      "step": 32800
    },
    {
      "epoch": 3.5417789757412397,
      "grad_norm": 0.5920261740684509,
      "learning_rate": 0.000387989206691851,
      "loss": 3.458,
      "step": 32850
    },
    {
      "epoch": 3.547169811320755,
      "grad_norm": 0.576420247554779,
      "learning_rate": 0.00038766540744738256,
      "loss": 3.4604,
      "step": 32900
    },
    {
      "epoch": 3.5525606469002695,
      "grad_norm": 0.6335041522979736,
      "learning_rate": 0.00038734160820291417,
      "loss": 3.4826,
      "step": 32950
    },
    {
      "epoch": 3.557951482479784,
      "grad_norm": 0.6456249356269836,
      "learning_rate": 0.0003870178089584457,
      "loss": 3.4599,
      "step": 33000
    },
    {
      "epoch": 3.557951482479784,
      "eval_accuracy": 0.3727679573072381,
      "eval_loss": 3.4776175022125244,
      "eval_runtime": 187.4986,
      "eval_samples_per_second": 96.059,
      "eval_steps_per_second": 6.005,
      "step": 33000
    },
    {
      "epoch": 3.5633423180592994,
      "grad_norm": 0.5974683165550232,
      "learning_rate": 0.0003866940097139773,
      "loss": 3.4603,
      "step": 33050
    },
    {
      "epoch": 3.568733153638814,
      "grad_norm": 0.6262832283973694,
      "learning_rate": 0.0003863702104695089,
      "loss": 3.4828,
      "step": 33100
    },
    {
      "epoch": 3.5741239892183287,
      "grad_norm": 0.5947827100753784,
      "learning_rate": 0.0003860464112250404,
      "loss": 3.458,
      "step": 33150
    },
    {
      "epoch": 3.579514824797844,
      "grad_norm": 0.5985409617424011,
      "learning_rate": 0.00038572261198057203,
      "loss": 3.4769,
      "step": 33200
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.5783484578132629,
      "learning_rate": 0.0003853988127361036,
      "loss": 3.4725,
      "step": 33250
    },
    {
      "epoch": 3.5902964959568733,
      "grad_norm": 0.6530685424804688,
      "learning_rate": 0.0003850750134916352,
      "loss": 3.4968,
      "step": 33300
    },
    {
      "epoch": 3.595687331536388,
      "grad_norm": 0.6270087957382202,
      "learning_rate": 0.0003847512142471667,
      "loss": 3.4611,
      "step": 33350
    },
    {
      "epoch": 3.601078167115903,
      "grad_norm": 0.6156905889511108,
      "learning_rate": 0.0003844274150026983,
      "loss": 3.4747,
      "step": 33400
    },
    {
      "epoch": 3.606469002695418,
      "grad_norm": 0.5912847518920898,
      "learning_rate": 0.00038410361575822984,
      "loss": 3.4828,
      "step": 33450
    },
    {
      "epoch": 3.6118598382749325,
      "grad_norm": 0.5977141857147217,
      "learning_rate": 0.00038377981651376144,
      "loss": 3.4765,
      "step": 33500
    },
    {
      "epoch": 3.617250673854447,
      "grad_norm": 0.6528952121734619,
      "learning_rate": 0.000383456017269293,
      "loss": 3.4641,
      "step": 33550
    },
    {
      "epoch": 3.6226415094339623,
      "grad_norm": 0.6034547090530396,
      "learning_rate": 0.00038313221802482454,
      "loss": 3.4617,
      "step": 33600
    },
    {
      "epoch": 3.628032345013477,
      "grad_norm": 0.6089150309562683,
      "learning_rate": 0.00038280841878035615,
      "loss": 3.4533,
      "step": 33650
    },
    {
      "epoch": 3.6334231805929917,
      "grad_norm": 0.5918501615524292,
      "learning_rate": 0.0003824846195358877,
      "loss": 3.4497,
      "step": 33700
    },
    {
      "epoch": 3.638814016172507,
      "grad_norm": 0.594428539276123,
      "learning_rate": 0.0003821608202914193,
      "loss": 3.4731,
      "step": 33750
    },
    {
      "epoch": 3.6442048517520216,
      "grad_norm": 0.5788894891738892,
      "learning_rate": 0.00038183702104695085,
      "loss": 3.4525,
      "step": 33800
    },
    {
      "epoch": 3.6495956873315363,
      "grad_norm": 0.6797850728034973,
      "learning_rate": 0.00038151322180248246,
      "loss": 3.4659,
      "step": 33850
    },
    {
      "epoch": 3.6549865229110514,
      "grad_norm": 0.6085013747215271,
      "learning_rate": 0.000381189422558014,
      "loss": 3.4651,
      "step": 33900
    },
    {
      "epoch": 3.660377358490566,
      "grad_norm": 0.6374878287315369,
      "learning_rate": 0.0003808656233135456,
      "loss": 3.4603,
      "step": 33950
    },
    {
      "epoch": 3.665768194070081,
      "grad_norm": 0.6357040405273438,
      "learning_rate": 0.00038054182406907716,
      "loss": 3.4813,
      "step": 34000
    },
    {
      "epoch": 3.665768194070081,
      "eval_accuracy": 0.3736759694907037,
      "eval_loss": 3.4704673290252686,
      "eval_runtime": 187.7895,
      "eval_samples_per_second": 95.911,
      "eval_steps_per_second": 5.996,
      "step": 34000
    },
    {
      "epoch": 3.671159029649596,
      "grad_norm": 0.5881034731864929,
      "learning_rate": 0.00038021802482460866,
      "loss": 3.4617,
      "step": 34050
    },
    {
      "epoch": 3.6765498652291106,
      "grad_norm": 0.5796710848808289,
      "learning_rate": 0.00037989422558014027,
      "loss": 3.4691,
      "step": 34100
    },
    {
      "epoch": 3.6819407008086253,
      "grad_norm": 0.642318606376648,
      "learning_rate": 0.0003795704263356718,
      "loss": 3.4757,
      "step": 34150
    },
    {
      "epoch": 3.68733153638814,
      "grad_norm": 0.6038513779640198,
      "learning_rate": 0.0003792466270912034,
      "loss": 3.4749,
      "step": 34200
    },
    {
      "epoch": 3.6927223719676547,
      "grad_norm": 0.5752274990081787,
      "learning_rate": 0.00037892282784673497,
      "loss": 3.4663,
      "step": 34250
    },
    {
      "epoch": 3.69811320754717,
      "grad_norm": 0.6258960962295532,
      "learning_rate": 0.0003785990286022666,
      "loss": 3.4774,
      "step": 34300
    },
    {
      "epoch": 3.7035040431266846,
      "grad_norm": 0.6572497487068176,
      "learning_rate": 0.00037827522935779813,
      "loss": 3.4861,
      "step": 34350
    },
    {
      "epoch": 3.7088948787061993,
      "grad_norm": 0.6318001747131348,
      "learning_rate": 0.00037795143011332973,
      "loss": 3.4592,
      "step": 34400
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.6450494527816772,
      "learning_rate": 0.0003776276308688613,
      "loss": 3.4438,
      "step": 34450
    },
    {
      "epoch": 3.719676549865229,
      "grad_norm": 0.6240411400794983,
      "learning_rate": 0.00037730383162439283,
      "loss": 3.4636,
      "step": 34500
    },
    {
      "epoch": 3.725067385444744,
      "grad_norm": 0.6308651566505432,
      "learning_rate": 0.00037698003237992444,
      "loss": 3.4799,
      "step": 34550
    },
    {
      "epoch": 3.730458221024259,
      "grad_norm": 0.6322066187858582,
      "learning_rate": 0.000376656233135456,
      "loss": 3.4733,
      "step": 34600
    },
    {
      "epoch": 3.7358490566037736,
      "grad_norm": 0.6787533164024353,
      "learning_rate": 0.0003763324338909876,
      "loss": 3.4466,
      "step": 34650
    },
    {
      "epoch": 3.7412398921832883,
      "grad_norm": 0.6830219030380249,
      "learning_rate": 0.0003760086346465191,
      "loss": 3.4894,
      "step": 34700
    },
    {
      "epoch": 3.7466307277628035,
      "grad_norm": 0.6646257638931274,
      "learning_rate": 0.0003756848354020507,
      "loss": 3.4516,
      "step": 34750
    },
    {
      "epoch": 3.752021563342318,
      "grad_norm": 0.6488728523254395,
      "learning_rate": 0.00037536103615758224,
      "loss": 3.4828,
      "step": 34800
    },
    {
      "epoch": 3.757412398921833,
      "grad_norm": 0.6173174381256104,
      "learning_rate": 0.0003750372369131138,
      "loss": 3.4949,
      "step": 34850
    },
    {
      "epoch": 3.7628032345013476,
      "grad_norm": 0.6098762154579163,
      "learning_rate": 0.0003747134376686454,
      "loss": 3.4806,
      "step": 34900
    },
    {
      "epoch": 3.7681940700808623,
      "grad_norm": 0.5716989040374756,
      "learning_rate": 0.00037438963842417695,
      "loss": 3.4584,
      "step": 34950
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.6457527875900269,
      "learning_rate": 0.00037406583917970856,
      "loss": 3.4595,
      "step": 35000
    },
    {
      "epoch": 3.7735849056603774,
      "eval_accuracy": 0.37413861348701777,
      "eval_loss": 3.4643301963806152,
      "eval_runtime": 187.4007,
      "eval_samples_per_second": 96.11,
      "eval_steps_per_second": 6.009,
      "step": 35000
    },
    {
      "epoch": 3.778975741239892,
      "grad_norm": 0.6491796970367432,
      "learning_rate": 0.0003737420399352401,
      "loss": 3.4837,
      "step": 35050
    },
    {
      "epoch": 3.784366576819407,
      "grad_norm": 0.689035177230835,
      "learning_rate": 0.0003734182406907717,
      "loss": 3.4489,
      "step": 35100
    },
    {
      "epoch": 3.789757412398922,
      "grad_norm": 0.6304520964622498,
      "learning_rate": 0.00037309444144630326,
      "loss": 3.4715,
      "step": 35150
    },
    {
      "epoch": 3.7951482479784366,
      "grad_norm": 0.6018555760383606,
      "learning_rate": 0.00037277064220183487,
      "loss": 3.4543,
      "step": 35200
    },
    {
      "epoch": 3.8005390835579513,
      "grad_norm": 0.582009494304657,
      "learning_rate": 0.0003724468429573664,
      "loss": 3.4806,
      "step": 35250
    },
    {
      "epoch": 3.8059299191374665,
      "grad_norm": 0.6774134039878845,
      "learning_rate": 0.0003721230437128979,
      "loss": 3.469,
      "step": 35300
    },
    {
      "epoch": 3.811320754716981,
      "grad_norm": 0.5913166999816895,
      "learning_rate": 0.00037179924446842957,
      "loss": 3.47,
      "step": 35350
    },
    {
      "epoch": 3.816711590296496,
      "grad_norm": 0.6116369962692261,
      "learning_rate": 0.00037147544522396107,
      "loss": 3.4878,
      "step": 35400
    },
    {
      "epoch": 3.822102425876011,
      "grad_norm": 0.5959073305130005,
      "learning_rate": 0.000371158121964382,
      "loss": 3.4448,
      "step": 35450
    },
    {
      "epoch": 3.8274932614555257,
      "grad_norm": 0.611961841583252,
      "learning_rate": 0.0003708343227199136,
      "loss": 3.4809,
      "step": 35500
    },
    {
      "epoch": 3.8328840970350404,
      "grad_norm": 0.6396837830543518,
      "learning_rate": 0.0003705105234754452,
      "loss": 3.4641,
      "step": 35550
    },
    {
      "epoch": 3.838274932614555,
      "grad_norm": 0.5907486081123352,
      "learning_rate": 0.0003701867242309768,
      "loss": 3.4513,
      "step": 35600
    },
    {
      "epoch": 3.8436657681940702,
      "grad_norm": 0.6286391019821167,
      "learning_rate": 0.00036986292498650833,
      "loss": 3.4649,
      "step": 35650
    },
    {
      "epoch": 3.849056603773585,
      "grad_norm": 0.629386842250824,
      "learning_rate": 0.0003695391257420399,
      "loss": 3.4621,
      "step": 35700
    },
    {
      "epoch": 3.8544474393530996,
      "grad_norm": 0.6501761674880981,
      "learning_rate": 0.0003692153264975715,
      "loss": 3.4423,
      "step": 35750
    },
    {
      "epoch": 3.8598382749326143,
      "grad_norm": 0.6401487588882446,
      "learning_rate": 0.00036889152725310304,
      "loss": 3.4644,
      "step": 35800
    },
    {
      "epoch": 3.8652291105121295,
      "grad_norm": 0.6554639935493469,
      "learning_rate": 0.00036856772800863464,
      "loss": 3.4676,
      "step": 35850
    },
    {
      "epoch": 3.870619946091644,
      "grad_norm": 0.59914231300354,
      "learning_rate": 0.0003682439287641662,
      "loss": 3.4596,
      "step": 35900
    },
    {
      "epoch": 3.876010781671159,
      "grad_norm": 0.5814793705940247,
      "learning_rate": 0.0003679201295196978,
      "loss": 3.4636,
      "step": 35950
    },
    {
      "epoch": 3.881401617250674,
      "grad_norm": 0.6307700872421265,
      "learning_rate": 0.00036759633027522935,
      "loss": 3.4446,
      "step": 36000
    },
    {
      "epoch": 3.881401617250674,
      "eval_accuracy": 0.3749434054270596,
      "eval_loss": 3.458733320236206,
      "eval_runtime": 187.388,
      "eval_samples_per_second": 96.116,
      "eval_steps_per_second": 6.009,
      "step": 36000
    },
    {
      "epoch": 3.8867924528301887,
      "grad_norm": 0.6277196407318115,
      "learning_rate": 0.00036727253103076084,
      "loss": 3.4476,
      "step": 36050
    },
    {
      "epoch": 3.8921832884097034,
      "grad_norm": 0.6114755272865295,
      "learning_rate": 0.00036694873178629245,
      "loss": 3.4725,
      "step": 36100
    },
    {
      "epoch": 3.8975741239892185,
      "grad_norm": 0.6390984654426575,
      "learning_rate": 0.000366624932541824,
      "loss": 3.4624,
      "step": 36150
    },
    {
      "epoch": 3.9029649595687332,
      "grad_norm": 0.6099748611450195,
      "learning_rate": 0.0003663011332973556,
      "loss": 3.4424,
      "step": 36200
    },
    {
      "epoch": 3.908355795148248,
      "grad_norm": 0.6574003100395203,
      "learning_rate": 0.00036597733405288715,
      "loss": 3.4488,
      "step": 36250
    },
    {
      "epoch": 3.913746630727763,
      "grad_norm": 0.6434478163719177,
      "learning_rate": 0.00036565353480841876,
      "loss": 3.461,
      "step": 36300
    },
    {
      "epoch": 3.9191374663072778,
      "grad_norm": 0.5764980912208557,
      "learning_rate": 0.0003653297355639503,
      "loss": 3.4538,
      "step": 36350
    },
    {
      "epoch": 3.9245283018867925,
      "grad_norm": 0.6057455539703369,
      "learning_rate": 0.0003650059363194819,
      "loss": 3.4548,
      "step": 36400
    },
    {
      "epoch": 3.929919137466307,
      "grad_norm": 0.6364848613739014,
      "learning_rate": 0.00036468213707501347,
      "loss": 3.462,
      "step": 36450
    },
    {
      "epoch": 3.935309973045822,
      "grad_norm": 0.5842701196670532,
      "learning_rate": 0.000364358337830545,
      "loss": 3.4663,
      "step": 36500
    },
    {
      "epoch": 3.940700808625337,
      "grad_norm": 0.5927199125289917,
      "learning_rate": 0.0003640345385860766,
      "loss": 3.4543,
      "step": 36550
    },
    {
      "epoch": 3.9460916442048517,
      "grad_norm": 0.6089248657226562,
      "learning_rate": 0.00036371073934160817,
      "loss": 3.466,
      "step": 36600
    },
    {
      "epoch": 3.9514824797843664,
      "grad_norm": 0.6188242435455322,
      "learning_rate": 0.0003633869400971398,
      "loss": 3.4595,
      "step": 36650
    },
    {
      "epoch": 3.9568733153638815,
      "grad_norm": 0.6130086183547974,
      "learning_rate": 0.00036306314085267127,
      "loss": 3.4574,
      "step": 36700
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.6104381084442139,
      "learning_rate": 0.00036273934160820293,
      "loss": 3.4509,
      "step": 36750
    },
    {
      "epoch": 3.967654986522911,
      "grad_norm": 0.7232850790023804,
      "learning_rate": 0.00036241554236373443,
      "loss": 3.4471,
      "step": 36800
    },
    {
      "epoch": 3.973045822102426,
      "grad_norm": 0.6882140636444092,
      "learning_rate": 0.00036209174311926603,
      "loss": 3.4459,
      "step": 36850
    },
    {
      "epoch": 3.9784366576819408,
      "grad_norm": 0.6311055421829224,
      "learning_rate": 0.0003617679438747976,
      "loss": 3.4611,
      "step": 36900
    },
    {
      "epoch": 3.9838274932614555,
      "grad_norm": 0.6199382543563843,
      "learning_rate": 0.00036144414463032913,
      "loss": 3.4723,
      "step": 36950
    },
    {
      "epoch": 3.9892183288409706,
      "grad_norm": 0.6460267901420593,
      "learning_rate": 0.00036112034538586074,
      "loss": 3.467,
      "step": 37000
    },
    {
      "epoch": 3.9892183288409706,
      "eval_accuracy": 0.3754612450903834,
      "eval_loss": 3.453521251678467,
      "eval_runtime": 187.8709,
      "eval_samples_per_second": 95.869,
      "eval_steps_per_second": 5.993,
      "step": 37000
    },
    {
      "epoch": 3.9946091644204853,
      "grad_norm": 0.6030475497245789,
      "learning_rate": 0.0003607965461413923,
      "loss": 3.4608,
      "step": 37050
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.1987624168395996,
      "learning_rate": 0.0003604727468969239,
      "loss": 3.4534,
      "step": 37100
    },
    {
      "epoch": 4.005390835579515,
      "grad_norm": 0.6259908676147461,
      "learning_rate": 0.00036014894765245544,
      "loss": 3.3673,
      "step": 37150
    },
    {
      "epoch": 4.010781671159029,
      "grad_norm": 0.604569673538208,
      "learning_rate": 0.00035982514840798705,
      "loss": 3.3673,
      "step": 37200
    },
    {
      "epoch": 4.0161725067385445,
      "grad_norm": 0.6428099274635315,
      "learning_rate": 0.0003595013491635186,
      "loss": 3.3624,
      "step": 37250
    },
    {
      "epoch": 4.02156334231806,
      "grad_norm": 0.6633650064468384,
      "learning_rate": 0.0003591775499190502,
      "loss": 3.3465,
      "step": 37300
    },
    {
      "epoch": 4.026954177897574,
      "grad_norm": 0.6475902795791626,
      "learning_rate": 0.00035885375067458176,
      "loss": 3.3721,
      "step": 37350
    },
    {
      "epoch": 4.032345013477089,
      "grad_norm": 0.6463242173194885,
      "learning_rate": 0.00035852995143011325,
      "loss": 3.3786,
      "step": 37400
    },
    {
      "epoch": 4.037735849056604,
      "grad_norm": 0.6179232597351074,
      "learning_rate": 0.0003582126281705342,
      "loss": 3.3648,
      "step": 37450
    },
    {
      "epoch": 4.0431266846361185,
      "grad_norm": 0.6580100059509277,
      "learning_rate": 0.0003578888289260658,
      "loss": 3.3807,
      "step": 37500
    },
    {
      "epoch": 4.048517520215634,
      "grad_norm": 0.6432563662528992,
      "learning_rate": 0.00035756502968159736,
      "loss": 3.3707,
      "step": 37550
    },
    {
      "epoch": 4.053908355795148,
      "grad_norm": 0.6162436604499817,
      "learning_rate": 0.00035724123043712896,
      "loss": 3.366,
      "step": 37600
    },
    {
      "epoch": 4.059299191374663,
      "grad_norm": 0.6134903430938721,
      "learning_rate": 0.0003569174311926605,
      "loss": 3.3818,
      "step": 37650
    },
    {
      "epoch": 4.064690026954178,
      "grad_norm": 0.6512436270713806,
      "learning_rate": 0.00035659363194819206,
      "loss": 3.3842,
      "step": 37700
    },
    {
      "epoch": 4.070080862533692,
      "grad_norm": 0.6385947465896606,
      "learning_rate": 0.00035626983270372367,
      "loss": 3.3619,
      "step": 37750
    },
    {
      "epoch": 4.0754716981132075,
      "grad_norm": 0.6357753276824951,
      "learning_rate": 0.0003559460334592552,
      "loss": 3.3616,
      "step": 37800
    },
    {
      "epoch": 4.080862533692723,
      "grad_norm": 0.6429207921028137,
      "learning_rate": 0.0003556222342147868,
      "loss": 3.375,
      "step": 37850
    },
    {
      "epoch": 4.086253369272237,
      "grad_norm": 0.635795533657074,
      "learning_rate": 0.0003552984349703184,
      "loss": 3.3842,
      "step": 37900
    },
    {
      "epoch": 4.091644204851752,
      "grad_norm": 0.6620546579360962,
      "learning_rate": 0.00035497463572585,
      "loss": 3.3943,
      "step": 37950
    },
    {
      "epoch": 4.097035040431267,
      "grad_norm": 0.6397520899772644,
      "learning_rate": 0.00035465083648138153,
      "loss": 3.3801,
      "step": 38000
    },
    {
      "epoch": 4.097035040431267,
      "eval_accuracy": 0.3764028310161837,
      "eval_loss": 3.453479766845703,
      "eval_runtime": 187.2183,
      "eval_samples_per_second": 96.203,
      "eval_steps_per_second": 6.014,
      "step": 38000
    },
    {
      "epoch": 4.1024258760107815,
      "grad_norm": 0.6575223207473755,
      "learning_rate": 0.00035432703723691314,
      "loss": 3.3747,
      "step": 38050
    },
    {
      "epoch": 4.107816711590297,
      "grad_norm": 0.64016193151474,
      "learning_rate": 0.00035400323799244463,
      "loss": 3.3831,
      "step": 38100
    },
    {
      "epoch": 4.113207547169812,
      "grad_norm": 0.6374653577804565,
      "learning_rate": 0.0003536794387479762,
      "loss": 3.3751,
      "step": 38150
    },
    {
      "epoch": 4.118598382749326,
      "grad_norm": 0.6173773407936096,
      "learning_rate": 0.0003533556395035078,
      "loss": 3.3993,
      "step": 38200
    },
    {
      "epoch": 4.123989218328841,
      "grad_norm": 0.6391666531562805,
      "learning_rate": 0.00035303184025903934,
      "loss": 3.3912,
      "step": 38250
    },
    {
      "epoch": 4.129380053908355,
      "grad_norm": 0.6431846022605896,
      "learning_rate": 0.00035270804101457094,
      "loss": 3.3915,
      "step": 38300
    },
    {
      "epoch": 4.1347708894878705,
      "grad_norm": 0.5880658626556396,
      "learning_rate": 0.0003523842417701025,
      "loss": 3.382,
      "step": 38350
    },
    {
      "epoch": 4.140161725067386,
      "grad_norm": 0.6336498856544495,
      "learning_rate": 0.0003520604425256341,
      "loss": 3.3767,
      "step": 38400
    },
    {
      "epoch": 4.1455525606469,
      "grad_norm": 0.6082850098609924,
      "learning_rate": 0.00035173664328116565,
      "loss": 3.3863,
      "step": 38450
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.6710603833198547,
      "learning_rate": 0.00035141284403669725,
      "loss": 3.3936,
      "step": 38500
    },
    {
      "epoch": 4.15633423180593,
      "grad_norm": 0.6536480188369751,
      "learning_rate": 0.0003510890447922288,
      "loss": 3.3814,
      "step": 38550
    },
    {
      "epoch": 4.1617250673854445,
      "grad_norm": 0.6300691366195679,
      "learning_rate": 0.00035076524554776035,
      "loss": 3.3792,
      "step": 38600
    },
    {
      "epoch": 4.16711590296496,
      "grad_norm": 0.6134623289108276,
      "learning_rate": 0.00035044144630329196,
      "loss": 3.4067,
      "step": 38650
    },
    {
      "epoch": 4.172506738544475,
      "grad_norm": 0.6804368495941162,
      "learning_rate": 0.00035011764705882346,
      "loss": 3.384,
      "step": 38700
    },
    {
      "epoch": 4.177897574123989,
      "grad_norm": 0.6068801879882812,
      "learning_rate": 0.0003497938478143551,
      "loss": 3.3772,
      "step": 38750
    },
    {
      "epoch": 4.183288409703504,
      "grad_norm": 0.7131064534187317,
      "learning_rate": 0.0003494700485698866,
      "loss": 3.3989,
      "step": 38800
    },
    {
      "epoch": 4.188679245283019,
      "grad_norm": 0.6442277431488037,
      "learning_rate": 0.0003491462493254182,
      "loss": 3.4037,
      "step": 38850
    },
    {
      "epoch": 4.1940700808625335,
      "grad_norm": 0.6614530682563782,
      "learning_rate": 0.00034882245008094977,
      "loss": 3.4017,
      "step": 38900
    },
    {
      "epoch": 4.199460916442049,
      "grad_norm": 0.6263593435287476,
      "learning_rate": 0.0003484986508364813,
      "loss": 3.3943,
      "step": 38950
    },
    {
      "epoch": 4.204851752021563,
      "grad_norm": 0.6600920557975769,
      "learning_rate": 0.0003481748515920129,
      "loss": 3.3901,
      "step": 39000
    },
    {
      "epoch": 4.204851752021563,
      "eval_accuracy": 0.37629906750832093,
      "eval_loss": 3.448453664779663,
      "eval_runtime": 187.5563,
      "eval_samples_per_second": 96.03,
      "eval_steps_per_second": 6.004,
      "step": 39000
    },
    {
      "epoch": 4.210242587601078,
      "grad_norm": 0.627390444278717,
      "learning_rate": 0.00034785105234754447,
      "loss": 3.3867,
      "step": 39050
    },
    {
      "epoch": 4.215633423180593,
      "grad_norm": 0.6027140021324158,
      "learning_rate": 0.0003475272531030761,
      "loss": 3.3901,
      "step": 39100
    },
    {
      "epoch": 4.2210242587601075,
      "grad_norm": 0.6500639319419861,
      "learning_rate": 0.00034720345385860763,
      "loss": 3.3859,
      "step": 39150
    },
    {
      "epoch": 4.226415094339623,
      "grad_norm": 0.6324947476387024,
      "learning_rate": 0.00034687965461413923,
      "loss": 3.383,
      "step": 39200
    },
    {
      "epoch": 4.231805929919138,
      "grad_norm": 0.7030044794082642,
      "learning_rate": 0.0003465558553696708,
      "loss": 3.3804,
      "step": 39250
    },
    {
      "epoch": 4.237196765498652,
      "grad_norm": 0.656147301197052,
      "learning_rate": 0.0003462320561252024,
      "loss": 3.3941,
      "step": 39300
    },
    {
      "epoch": 4.242587601078167,
      "grad_norm": 0.6377809047698975,
      "learning_rate": 0.00034590825688073394,
      "loss": 3.3942,
      "step": 39350
    },
    {
      "epoch": 4.247978436657682,
      "grad_norm": 0.6354027986526489,
      "learning_rate": 0.00034558445763626543,
      "loss": 3.3987,
      "step": 39400
    },
    {
      "epoch": 4.2533692722371965,
      "grad_norm": 0.6445314884185791,
      "learning_rate": 0.00034526065839179704,
      "loss": 3.3887,
      "step": 39450
    },
    {
      "epoch": 4.258760107816712,
      "grad_norm": 0.6906786561012268,
      "learning_rate": 0.0003449368591473286,
      "loss": 3.3807,
      "step": 39500
    },
    {
      "epoch": 4.264150943396227,
      "grad_norm": 0.6557918787002563,
      "learning_rate": 0.00034461953588774954,
      "loss": 3.3852,
      "step": 39550
    },
    {
      "epoch": 4.269541778975741,
      "grad_norm": 0.6619495749473572,
      "learning_rate": 0.0003443022126281705,
      "loss": 3.3829,
      "step": 39600
    },
    {
      "epoch": 4.274932614555256,
      "grad_norm": 0.700074315071106,
      "learning_rate": 0.00034397841338370204,
      "loss": 3.4102,
      "step": 39650
    },
    {
      "epoch": 4.280323450134771,
      "grad_norm": 0.6916738748550415,
      "learning_rate": 0.00034365461413923365,
      "loss": 3.3953,
      "step": 39700
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.6402020454406738,
      "learning_rate": 0.0003433308148947652,
      "loss": 3.3878,
      "step": 39750
    },
    {
      "epoch": 4.291105121293801,
      "grad_norm": 0.5897455811500549,
      "learning_rate": 0.0003430070156502968,
      "loss": 3.4044,
      "step": 39800
    },
    {
      "epoch": 4.296495956873315,
      "grad_norm": 0.6505072712898254,
      "learning_rate": 0.00034268321640582835,
      "loss": 3.4052,
      "step": 39850
    },
    {
      "epoch": 4.30188679245283,
      "grad_norm": 0.6143913269042969,
      "learning_rate": 0.00034235941716135996,
      "loss": 3.3976,
      "step": 39900
    },
    {
      "epoch": 4.307277628032345,
      "grad_norm": 0.6959649920463562,
      "learning_rate": 0.0003420356179168915,
      "loss": 3.4007,
      "step": 39950
    },
    {
      "epoch": 4.3126684636118595,
      "grad_norm": 0.6691548228263855,
      "learning_rate": 0.0003417118186724231,
      "loss": 3.3924,
      "step": 40000
    },
    {
      "epoch": 4.3126684636118595,
      "eval_accuracy": 0.3769364253482407,
      "eval_loss": 3.444800853729248,
      "eval_runtime": 187.924,
      "eval_samples_per_second": 95.842,
      "eval_steps_per_second": 5.992,
      "step": 40000
    },
    {
      "epoch": 4.318059299191375,
      "grad_norm": 0.6325005888938904,
      "learning_rate": 0.00034138801942795467,
      "loss": 3.3868,
      "step": 40050
    },
    {
      "epoch": 4.32345013477089,
      "grad_norm": 0.6606913208961487,
      "learning_rate": 0.00034106422018348616,
      "loss": 3.4124,
      "step": 40100
    },
    {
      "epoch": 4.328840970350404,
      "grad_norm": 0.6430675983428955,
      "learning_rate": 0.00034074042093901777,
      "loss": 3.4218,
      "step": 40150
    },
    {
      "epoch": 4.334231805929919,
      "grad_norm": 0.653377115726471,
      "learning_rate": 0.0003404166216945493,
      "loss": 3.3903,
      "step": 40200
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.6559277176856995,
      "learning_rate": 0.0003400928224500809,
      "loss": 3.3963,
      "step": 40250
    },
    {
      "epoch": 4.345013477088949,
      "grad_norm": 0.6817251443862915,
      "learning_rate": 0.00033976902320561247,
      "loss": 3.4016,
      "step": 40300
    },
    {
      "epoch": 4.350404312668464,
      "grad_norm": 0.664612889289856,
      "learning_rate": 0.0003394452239611441,
      "loss": 3.3943,
      "step": 40350
    },
    {
      "epoch": 4.355795148247978,
      "grad_norm": 0.6640458703041077,
      "learning_rate": 0.00033912142471667563,
      "loss": 3.4109,
      "step": 40400
    },
    {
      "epoch": 4.361185983827493,
      "grad_norm": 0.6599859595298767,
      "learning_rate": 0.00033879762547220723,
      "loss": 3.38,
      "step": 40450
    },
    {
      "epoch": 4.366576819407008,
      "grad_norm": 0.6857253909111023,
      "learning_rate": 0.0003384738262277388,
      "loss": 3.3843,
      "step": 40500
    },
    {
      "epoch": 4.3719676549865225,
      "grad_norm": 0.6355632543563843,
      "learning_rate": 0.00033815002698327033,
      "loss": 3.407,
      "step": 40550
    },
    {
      "epoch": 4.377358490566038,
      "grad_norm": 0.6515060663223267,
      "learning_rate": 0.00033782622773880194,
      "loss": 3.3921,
      "step": 40600
    },
    {
      "epoch": 4.382749326145553,
      "grad_norm": 0.651151716709137,
      "learning_rate": 0.0003375024284943335,
      "loss": 3.3834,
      "step": 40650
    },
    {
      "epoch": 4.388140161725067,
      "grad_norm": 0.6231512427330017,
      "learning_rate": 0.0003371786292498651,
      "loss": 3.4116,
      "step": 40700
    },
    {
      "epoch": 4.393530997304582,
      "grad_norm": 0.7003295421600342,
      "learning_rate": 0.0003368548300053966,
      "loss": 3.4094,
      "step": 40750
    },
    {
      "epoch": 4.398921832884097,
      "grad_norm": 0.6507589817047119,
      "learning_rate": 0.0003365310307609282,
      "loss": 3.3824,
      "step": 40800
    },
    {
      "epoch": 4.404312668463612,
      "grad_norm": 0.6417136788368225,
      "learning_rate": 0.00033620723151645975,
      "loss": 3.4028,
      "step": 40850
    },
    {
      "epoch": 4.409703504043127,
      "grad_norm": 0.6464463472366333,
      "learning_rate": 0.0003358834322719913,
      "loss": 3.3901,
      "step": 40900
    },
    {
      "epoch": 4.415094339622642,
      "grad_norm": 0.672114372253418,
      "learning_rate": 0.0003355596330275229,
      "loss": 3.3962,
      "step": 40950
    },
    {
      "epoch": 4.420485175202156,
      "grad_norm": 0.6453450918197632,
      "learning_rate": 0.00033523583378305445,
      "loss": 3.4093,
      "step": 41000
    },
    {
      "epoch": 4.420485175202156,
      "eval_accuracy": 0.37747805999399586,
      "eval_loss": 3.4390106201171875,
      "eval_runtime": 187.6862,
      "eval_samples_per_second": 95.963,
      "eval_steps_per_second": 5.999,
      "step": 41000
    },
    {
      "epoch": 4.425876010781671,
      "grad_norm": 0.6747152209281921,
      "learning_rate": 0.00033491203453858606,
      "loss": 3.4099,
      "step": 41050
    },
    {
      "epoch": 4.431266846361186,
      "grad_norm": 0.6520073413848877,
      "learning_rate": 0.0003345882352941176,
      "loss": 3.3923,
      "step": 41100
    },
    {
      "epoch": 4.436657681940701,
      "grad_norm": 0.6549708843231201,
      "learning_rate": 0.0003342644360496492,
      "loss": 3.3711,
      "step": 41150
    },
    {
      "epoch": 4.442048517520216,
      "grad_norm": 0.6869304776191711,
      "learning_rate": 0.00033394063680518076,
      "loss": 3.4071,
      "step": 41200
    },
    {
      "epoch": 4.44743935309973,
      "grad_norm": 0.6499251127243042,
      "learning_rate": 0.00033361683756071237,
      "loss": 3.3953,
      "step": 41250
    },
    {
      "epoch": 4.452830188679245,
      "grad_norm": 0.6444708108901978,
      "learning_rate": 0.0003332930383162439,
      "loss": 3.3924,
      "step": 41300
    },
    {
      "epoch": 4.45822102425876,
      "grad_norm": 0.6329517364501953,
      "learning_rate": 0.0003329692390717754,
      "loss": 3.3836,
      "step": 41350
    },
    {
      "epoch": 4.463611859838275,
      "grad_norm": 0.6541862487792969,
      "learning_rate": 0.0003326454398273071,
      "loss": 3.395,
      "step": 41400
    },
    {
      "epoch": 4.46900269541779,
      "grad_norm": 0.6906949281692505,
      "learning_rate": 0.00033232164058283857,
      "loss": 3.3827,
      "step": 41450
    },
    {
      "epoch": 4.474393530997305,
      "grad_norm": 0.6574482321739197,
      "learning_rate": 0.0003319978413383702,
      "loss": 3.3845,
      "step": 41500
    },
    {
      "epoch": 4.479784366576819,
      "grad_norm": 0.6679044961929321,
      "learning_rate": 0.0003316740420939017,
      "loss": 3.4135,
      "step": 41550
    },
    {
      "epoch": 4.485175202156334,
      "grad_norm": 0.6667996644973755,
      "learning_rate": 0.00033135024284943333,
      "loss": 3.4046,
      "step": 41600
    },
    {
      "epoch": 4.490566037735849,
      "grad_norm": 0.6718025207519531,
      "learning_rate": 0.0003310264436049649,
      "loss": 3.41,
      "step": 41650
    },
    {
      "epoch": 4.495956873315364,
      "grad_norm": 0.6717631220817566,
      "learning_rate": 0.0003307026443604965,
      "loss": 3.4114,
      "step": 41700
    },
    {
      "epoch": 4.501347708894879,
      "grad_norm": 0.6933983564376831,
      "learning_rate": 0.00033037884511602804,
      "loss": 3.3954,
      "step": 41750
    },
    {
      "epoch": 4.506738544474393,
      "grad_norm": 0.6075118780136108,
      "learning_rate": 0.0003300550458715596,
      "loss": 3.4072,
      "step": 41800
    },
    {
      "epoch": 4.512129380053908,
      "grad_norm": 0.68716961145401,
      "learning_rate": 0.0003297312466270912,
      "loss": 3.3843,
      "step": 41850
    },
    {
      "epoch": 4.517520215633423,
      "grad_norm": 0.6730256676673889,
      "learning_rate": 0.00032940744738262274,
      "loss": 3.4068,
      "step": 41900
    },
    {
      "epoch": 4.5229110512129385,
      "grad_norm": 0.7278045415878296,
      "learning_rate": 0.00032908364813815435,
      "loss": 3.4097,
      "step": 41950
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.6512318849563599,
      "learning_rate": 0.0003287598488936859,
      "loss": 3.4104,
      "step": 42000
    },
    {
      "epoch": 4.528301886792453,
      "eval_accuracy": 0.378027300341898,
      "eval_loss": 3.4342195987701416,
      "eval_runtime": 187.6655,
      "eval_samples_per_second": 95.974,
      "eval_steps_per_second": 6.0,
      "step": 42000
    },
    {
      "epoch": 4.533692722371968,
      "grad_norm": 0.6605228185653687,
      "learning_rate": 0.0003284360496492175,
      "loss": 3.3834,
      "step": 42050
    },
    {
      "epoch": 4.539083557951482,
      "grad_norm": 0.7046045064926147,
      "learning_rate": 0.000328112250404749,
      "loss": 3.394,
      "step": 42100
    },
    {
      "epoch": 4.544474393530997,
      "grad_norm": 0.6316480040550232,
      "learning_rate": 0.00032778845116028066,
      "loss": 3.3978,
      "step": 42150
    },
    {
      "epoch": 4.549865229110512,
      "grad_norm": 0.6550145149230957,
      "learning_rate": 0.00032746465191581215,
      "loss": 3.4099,
      "step": 42200
    },
    {
      "epoch": 4.555256064690027,
      "grad_norm": 0.6940279006958008,
      "learning_rate": 0.0003271408526713437,
      "loss": 3.4054,
      "step": 42250
    },
    {
      "epoch": 4.560646900269542,
      "grad_norm": 0.6692448854446411,
      "learning_rate": 0.0003268170534268753,
      "loss": 3.3962,
      "step": 42300
    },
    {
      "epoch": 4.566037735849057,
      "grad_norm": 0.6523816585540771,
      "learning_rate": 0.00032649325418240686,
      "loss": 3.4179,
      "step": 42350
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.6672971844673157,
      "learning_rate": 0.00032616945493793846,
      "loss": 3.396,
      "step": 42400
    },
    {
      "epoch": 4.576819407008086,
      "grad_norm": 0.6267552971839905,
      "learning_rate": 0.00032584565569347,
      "loss": 3.4071,
      "step": 42450
    },
    {
      "epoch": 4.5822102425876015,
      "grad_norm": 0.6981790661811829,
      "learning_rate": 0.0003255218564490016,
      "loss": 3.3912,
      "step": 42500
    },
    {
      "epoch": 4.587601078167116,
      "grad_norm": 0.646481454372406,
      "learning_rate": 0.00032519805720453317,
      "loss": 3.4016,
      "step": 42550
    },
    {
      "epoch": 4.592991913746631,
      "grad_norm": 0.658094048500061,
      "learning_rate": 0.0003248742579600647,
      "loss": 3.3834,
      "step": 42600
    },
    {
      "epoch": 4.598382749326145,
      "grad_norm": 0.701236367225647,
      "learning_rate": 0.0003245504587155963,
      "loss": 3.4091,
      "step": 42650
    },
    {
      "epoch": 4.60377358490566,
      "grad_norm": 0.6524823904037476,
      "learning_rate": 0.0003242266594711278,
      "loss": 3.4031,
      "step": 42700
    },
    {
      "epoch": 4.609164420485175,
      "grad_norm": 0.6430251002311707,
      "learning_rate": 0.0003239028602266595,
      "loss": 3.3937,
      "step": 42750
    },
    {
      "epoch": 4.6145552560646905,
      "grad_norm": 0.6301037073135376,
      "learning_rate": 0.000323579060982191,
      "loss": 3.395,
      "step": 42800
    },
    {
      "epoch": 4.619946091644205,
      "grad_norm": 0.6488710641860962,
      "learning_rate": 0.0003232552617377226,
      "loss": 3.4024,
      "step": 42850
    },
    {
      "epoch": 4.62533692722372,
      "grad_norm": 0.664051353931427,
      "learning_rate": 0.00032293146249325413,
      "loss": 3.4154,
      "step": 42900
    },
    {
      "epoch": 4.630727762803234,
      "grad_norm": 0.6523007750511169,
      "learning_rate": 0.00032260766324878574,
      "loss": 3.424,
      "step": 42950
    },
    {
      "epoch": 4.636118598382749,
      "grad_norm": 0.6787187457084656,
      "learning_rate": 0.0003222838640043173,
      "loss": 3.3984,
      "step": 43000
    },
    {
      "epoch": 4.636118598382749,
      "eval_accuracy": 0.37819299599581513,
      "eval_loss": 3.4289870262145996,
      "eval_runtime": 187.3727,
      "eval_samples_per_second": 96.124,
      "eval_steps_per_second": 6.009,
      "step": 43000
    },
    {
      "epoch": 4.6415094339622645,
      "grad_norm": 0.6364569067955017,
      "learning_rate": 0.00032196006475984884,
      "loss": 3.4042,
      "step": 43050
    },
    {
      "epoch": 4.646900269541779,
      "grad_norm": 0.6747838854789734,
      "learning_rate": 0.00032163626551538044,
      "loss": 3.397,
      "step": 43100
    },
    {
      "epoch": 4.652291105121294,
      "grad_norm": 0.6599672436714172,
      "learning_rate": 0.000321312466270912,
      "loss": 3.4032,
      "step": 43150
    },
    {
      "epoch": 4.657681940700809,
      "grad_norm": 0.6930491328239441,
      "learning_rate": 0.0003209886670264436,
      "loss": 3.4026,
      "step": 43200
    },
    {
      "epoch": 4.663072776280323,
      "grad_norm": 0.6434908509254456,
      "learning_rate": 0.00032066486778197515,
      "loss": 3.3898,
      "step": 43250
    },
    {
      "epoch": 4.668463611859838,
      "grad_norm": 0.6763002276420593,
      "learning_rate": 0.00032034106853750675,
      "loss": 3.3911,
      "step": 43300
    },
    {
      "epoch": 4.6738544474393535,
      "grad_norm": 0.6328909993171692,
      "learning_rate": 0.0003200172692930383,
      "loss": 3.3884,
      "step": 43350
    },
    {
      "epoch": 4.679245283018868,
      "grad_norm": 0.6524956822395325,
      "learning_rate": 0.0003196934700485699,
      "loss": 3.4142,
      "step": 43400
    },
    {
      "epoch": 4.684636118598383,
      "grad_norm": 0.6790773272514343,
      "learning_rate": 0.0003193696708041014,
      "loss": 3.4084,
      "step": 43450
    },
    {
      "epoch": 4.690026954177897,
      "grad_norm": 0.6767542958259583,
      "learning_rate": 0.00031904587155963296,
      "loss": 3.4182,
      "step": 43500
    },
    {
      "epoch": 4.695417789757412,
      "grad_norm": 0.6552374362945557,
      "learning_rate": 0.00031872207231516456,
      "loss": 3.383,
      "step": 43550
    },
    {
      "epoch": 4.7008086253369274,
      "grad_norm": 0.6805891394615173,
      "learning_rate": 0.0003184047490555855,
      "loss": 3.3967,
      "step": 43600
    },
    {
      "epoch": 4.706199460916442,
      "grad_norm": 0.6635441780090332,
      "learning_rate": 0.00031808094981111706,
      "loss": 3.4145,
      "step": 43650
    },
    {
      "epoch": 4.711590296495957,
      "grad_norm": 0.6911014914512634,
      "learning_rate": 0.00031775715056664867,
      "loss": 3.4043,
      "step": 43700
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.6306861639022827,
      "learning_rate": 0.0003174333513221802,
      "loss": 3.3861,
      "step": 43750
    },
    {
      "epoch": 4.722371967654986,
      "grad_norm": 0.6720971465110779,
      "learning_rate": 0.00031710955207771177,
      "loss": 3.3955,
      "step": 43800
    },
    {
      "epoch": 4.727762803234501,
      "grad_norm": 0.681067168712616,
      "learning_rate": 0.0003167857528332434,
      "loss": 3.4104,
      "step": 43850
    },
    {
      "epoch": 4.7331536388140165,
      "grad_norm": 0.6770942807197571,
      "learning_rate": 0.0003164619535887749,
      "loss": 3.3891,
      "step": 43900
    },
    {
      "epoch": 4.738544474393531,
      "grad_norm": 0.6591625809669495,
      "learning_rate": 0.00031613815434430653,
      "loss": 3.4002,
      "step": 43950
    },
    {
      "epoch": 4.743935309973046,
      "grad_norm": 0.6355513334274292,
      "learning_rate": 0.0003158143550998381,
      "loss": 3.403,
      "step": 44000
    },
    {
      "epoch": 4.743935309973046,
      "eval_accuracy": 0.37904559520649533,
      "eval_loss": 3.4251604080200195,
      "eval_runtime": 187.4268,
      "eval_samples_per_second": 96.096,
      "eval_steps_per_second": 6.008,
      "step": 44000
    },
    {
      "epoch": 4.74932614555256,
      "grad_norm": 0.6216215491294861,
      "learning_rate": 0.0003154905558553697,
      "loss": 3.4181,
      "step": 44050
    },
    {
      "epoch": 4.754716981132075,
      "grad_norm": 0.7213981747627258,
      "learning_rate": 0.0003151667566109012,
      "loss": 3.3885,
      "step": 44100
    },
    {
      "epoch": 4.7601078167115904,
      "grad_norm": 0.6748440265655518,
      "learning_rate": 0.00031484295736643284,
      "loss": 3.3973,
      "step": 44150
    },
    {
      "epoch": 4.765498652291106,
      "grad_norm": 0.6230536103248596,
      "learning_rate": 0.00031451915812196434,
      "loss": 3.3889,
      "step": 44200
    },
    {
      "epoch": 4.77088948787062,
      "grad_norm": 0.6862159371376038,
      "learning_rate": 0.0003141953588774959,
      "loss": 3.4072,
      "step": 44250
    },
    {
      "epoch": 4.776280323450135,
      "grad_norm": 0.6609549522399902,
      "learning_rate": 0.0003138715596330275,
      "loss": 3.3927,
      "step": 44300
    },
    {
      "epoch": 4.781671159029649,
      "grad_norm": 0.7261171340942383,
      "learning_rate": 0.00031354776038855904,
      "loss": 3.3898,
      "step": 44350
    },
    {
      "epoch": 4.787061994609164,
      "grad_norm": 0.683654248714447,
      "learning_rate": 0.00031322396114409065,
      "loss": 3.3972,
      "step": 44400
    },
    {
      "epoch": 4.7924528301886795,
      "grad_norm": 0.6835164427757263,
      "learning_rate": 0.0003129001618996222,
      "loss": 3.4126,
      "step": 44450
    },
    {
      "epoch": 4.797843665768194,
      "grad_norm": 0.6999883055686951,
      "learning_rate": 0.0003125763626551538,
      "loss": 3.4072,
      "step": 44500
    },
    {
      "epoch": 4.803234501347709,
      "grad_norm": 0.660194456577301,
      "learning_rate": 0.00031225256341068535,
      "loss": 3.4001,
      "step": 44550
    },
    {
      "epoch": 4.808625336927224,
      "grad_norm": 0.662414014339447,
      "learning_rate": 0.00031192876416621696,
      "loss": 3.4085,
      "step": 44600
    },
    {
      "epoch": 4.814016172506738,
      "grad_norm": 0.6617624759674072,
      "learning_rate": 0.0003116049649217485,
      "loss": 3.4029,
      "step": 44650
    },
    {
      "epoch": 4.819407008086253,
      "grad_norm": 0.6734949350357056,
      "learning_rate": 0.00031128116567728,
      "loss": 3.3927,
      "step": 44700
    },
    {
      "epoch": 4.824797843665769,
      "grad_norm": 0.6460137367248535,
      "learning_rate": 0.00031095736643281166,
      "loss": 3.41,
      "step": 44750
    },
    {
      "epoch": 4.830188679245283,
      "grad_norm": 0.6399081945419312,
      "learning_rate": 0.00031063356718834316,
      "loss": 3.4198,
      "step": 44800
    },
    {
      "epoch": 4.835579514824798,
      "grad_norm": 0.671139121055603,
      "learning_rate": 0.00031030976794387476,
      "loss": 3.4087,
      "step": 44850
    },
    {
      "epoch": 4.840970350404312,
      "grad_norm": 0.6836479902267456,
      "learning_rate": 0.0003099859686994063,
      "loss": 3.4142,
      "step": 44900
    },
    {
      "epoch": 4.846361185983827,
      "grad_norm": 0.6696195006370544,
      "learning_rate": 0.0003096621694549379,
      "loss": 3.4148,
      "step": 44950
    },
    {
      "epoch": 4.8517520215633425,
      "grad_norm": 0.6819624304771423,
      "learning_rate": 0.00030933837021046947,
      "loss": 3.4082,
      "step": 45000
    },
    {
      "epoch": 4.8517520215633425,
      "eval_accuracy": 0.37946542996501054,
      "eval_loss": 3.4183225631713867,
      "eval_runtime": 187.3214,
      "eval_samples_per_second": 96.15,
      "eval_steps_per_second": 6.011,
      "step": 45000
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.7019557356834412,
      "learning_rate": 0.0003090145709660011,
      "loss": 3.3969,
      "step": 45050
    },
    {
      "epoch": 4.862533692722372,
      "grad_norm": 0.6515421271324158,
      "learning_rate": 0.0003086907717215326,
      "loss": 3.4006,
      "step": 45100
    },
    {
      "epoch": 4.867924528301887,
      "grad_norm": 0.6652536392211914,
      "learning_rate": 0.0003083669724770642,
      "loss": 3.407,
      "step": 45150
    },
    {
      "epoch": 4.873315363881401,
      "grad_norm": 0.712742269039154,
      "learning_rate": 0.0003080431732325958,
      "loss": 3.3875,
      "step": 45200
    },
    {
      "epoch": 4.878706199460916,
      "grad_norm": 0.6716951727867126,
      "learning_rate": 0.00030771937398812733,
      "loss": 3.3971,
      "step": 45250
    },
    {
      "epoch": 4.884097035040432,
      "grad_norm": 0.657441258430481,
      "learning_rate": 0.00030739557474365894,
      "loss": 3.3839,
      "step": 45300
    },
    {
      "epoch": 4.889487870619946,
      "grad_norm": 0.6646153926849365,
      "learning_rate": 0.0003070717754991905,
      "loss": 3.3969,
      "step": 45350
    },
    {
      "epoch": 4.894878706199461,
      "grad_norm": 0.7000280022621155,
      "learning_rate": 0.0003067479762547221,
      "loss": 3.4016,
      "step": 45400
    },
    {
      "epoch": 4.900269541778976,
      "grad_norm": 0.6623254418373108,
      "learning_rate": 0.0003064241770102536,
      "loss": 3.4024,
      "step": 45450
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.7038528919219971,
      "learning_rate": 0.00030610037776578514,
      "loss": 3.3923,
      "step": 45500
    },
    {
      "epoch": 4.9110512129380055,
      "grad_norm": 0.6824002265930176,
      "learning_rate": 0.00030577657852131674,
      "loss": 3.3806,
      "step": 45550
    },
    {
      "epoch": 4.916442048517521,
      "grad_norm": 0.6446487307548523,
      "learning_rate": 0.0003054592552617377,
      "loss": 3.3889,
      "step": 45600
    },
    {
      "epoch": 4.921832884097035,
      "grad_norm": 0.7261145710945129,
      "learning_rate": 0.00030513545601726925,
      "loss": 3.3935,
      "step": 45650
    },
    {
      "epoch": 4.92722371967655,
      "grad_norm": 0.651084303855896,
      "learning_rate": 0.00030481165677280085,
      "loss": 3.4054,
      "step": 45700
    },
    {
      "epoch": 4.932614555256064,
      "grad_norm": 0.6604491472244263,
      "learning_rate": 0.0003044878575283324,
      "loss": 3.4026,
      "step": 45750
    },
    {
      "epoch": 4.938005390835579,
      "grad_norm": 0.7495591044425964,
      "learning_rate": 0.000304164058283864,
      "loss": 3.3887,
      "step": 45800
    },
    {
      "epoch": 4.943396226415095,
      "grad_norm": 0.6877288818359375,
      "learning_rate": 0.00030384025903939556,
      "loss": 3.3952,
      "step": 45850
    },
    {
      "epoch": 4.948787061994609,
      "grad_norm": 0.6563786864280701,
      "learning_rate": 0.0003035164597949271,
      "loss": 3.3979,
      "step": 45900
    },
    {
      "epoch": 4.954177897574124,
      "grad_norm": 0.68060302734375,
      "learning_rate": 0.0003031926605504587,
      "loss": 3.3982,
      "step": 45950
    },
    {
      "epoch": 4.959568733153639,
      "grad_norm": 0.6740385293960571,
      "learning_rate": 0.00030286886130599026,
      "loss": 3.3902,
      "step": 46000
    },
    {
      "epoch": 4.959568733153639,
      "eval_accuracy": 0.3798636427988506,
      "eval_loss": 3.415046215057373,
      "eval_runtime": 188.2864,
      "eval_samples_per_second": 95.657,
      "eval_steps_per_second": 5.98,
      "step": 46000
    },
    {
      "epoch": 4.964959568733153,
      "grad_norm": 0.6443136930465698,
      "learning_rate": 0.00030254506206152187,
      "loss": 3.4066,
      "step": 46050
    },
    {
      "epoch": 4.9703504043126685,
      "grad_norm": 0.676317036151886,
      "learning_rate": 0.00030222126281705336,
      "loss": 3.3889,
      "step": 46100
    },
    {
      "epoch": 4.975741239892184,
      "grad_norm": 0.6700965762138367,
      "learning_rate": 0.000301897463572585,
      "loss": 3.4058,
      "step": 46150
    },
    {
      "epoch": 4.981132075471698,
      "grad_norm": 0.7073339223861694,
      "learning_rate": 0.0003015736643281165,
      "loss": 3.3953,
      "step": 46200
    },
    {
      "epoch": 4.986522911051213,
      "grad_norm": 0.6577731966972351,
      "learning_rate": 0.00030124986508364807,
      "loss": 3.4054,
      "step": 46250
    },
    {
      "epoch": 4.991913746630727,
      "grad_norm": 0.6806015968322754,
      "learning_rate": 0.0003009260658391797,
      "loss": 3.4067,
      "step": 46300
    },
    {
      "epoch": 4.997304582210242,
      "grad_norm": 0.6785401105880737,
      "learning_rate": 0.0003006022665947112,
      "loss": 3.3915,
      "step": 46350
    },
    {
      "epoch": 5.002695417789758,
      "grad_norm": 0.6428429484367371,
      "learning_rate": 0.00030027846735024283,
      "loss": 3.3601,
      "step": 46400
    },
    {
      "epoch": 5.008086253369272,
      "grad_norm": 0.6469174027442932,
      "learning_rate": 0.00029995466810577443,
      "loss": 3.2972,
      "step": 46450
    },
    {
      "epoch": 5.013477088948787,
      "grad_norm": 0.6720959544181824,
      "learning_rate": 0.00029963086886130593,
      "loss": 3.2985,
      "step": 46500
    },
    {
      "epoch": 5.018867924528302,
      "grad_norm": 0.714445173740387,
      "learning_rate": 0.00029930706961683754,
      "loss": 3.297,
      "step": 46550
    },
    {
      "epoch": 5.024258760107816,
      "grad_norm": 0.7507009506225586,
      "learning_rate": 0.0002989832703723691,
      "loss": 3.3071,
      "step": 46600
    },
    {
      "epoch": 5.0296495956873315,
      "grad_norm": 0.7144675254821777,
      "learning_rate": 0.0002986594711279007,
      "loss": 3.3031,
      "step": 46650
    },
    {
      "epoch": 5.035040431266847,
      "grad_norm": 0.6817595362663269,
      "learning_rate": 0.00029833567188343224,
      "loss": 3.3211,
      "step": 46700
    },
    {
      "epoch": 5.040431266846361,
      "grad_norm": 0.7118050456047058,
      "learning_rate": 0.00029801187263896385,
      "loss": 3.3085,
      "step": 46750
    },
    {
      "epoch": 5.045822102425876,
      "grad_norm": 0.6556680202484131,
      "learning_rate": 0.0002976880733944954,
      "loss": 3.3258,
      "step": 46800
    },
    {
      "epoch": 5.051212938005391,
      "grad_norm": 0.7107027769088745,
      "learning_rate": 0.00029736427415002695,
      "loss": 3.3078,
      "step": 46850
    },
    {
      "epoch": 5.056603773584905,
      "grad_norm": 0.6754008531570435,
      "learning_rate": 0.0002970404749055585,
      "loss": 3.3179,
      "step": 46900
    },
    {
      "epoch": 5.061994609164421,
      "grad_norm": 0.6981600522994995,
      "learning_rate": 0.0002967166756610901,
      "loss": 3.3276,
      "step": 46950
    },
    {
      "epoch": 5.067385444743936,
      "grad_norm": 0.7493944764137268,
      "learning_rate": 0.00029639287641662165,
      "loss": 3.315,
      "step": 47000
    },
    {
      "epoch": 5.067385444743936,
      "eval_accuracy": 0.38014038170411407,
      "eval_loss": 3.420238494873047,
      "eval_runtime": 187.6122,
      "eval_samples_per_second": 96.001,
      "eval_steps_per_second": 6.002,
      "step": 47000
    },
    {
      "epoch": 5.07277628032345,
      "grad_norm": 0.6746612787246704,
      "learning_rate": 0.00029606907717215326,
      "loss": 3.32,
      "step": 47050
    },
    {
      "epoch": 5.078167115902965,
      "grad_norm": 0.736753523349762,
      "learning_rate": 0.0002957452779276848,
      "loss": 3.3211,
      "step": 47100
    },
    {
      "epoch": 5.083557951482479,
      "grad_norm": 0.7640523314476013,
      "learning_rate": 0.00029542147868321636,
      "loss": 3.3097,
      "step": 47150
    },
    {
      "epoch": 5.0889487870619945,
      "grad_norm": 0.6825003623962402,
      "learning_rate": 0.00029509767943874796,
      "loss": 3.3206,
      "step": 47200
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.6508895754814148,
      "learning_rate": 0.0002947738801942795,
      "loss": 3.3091,
      "step": 47250
    },
    {
      "epoch": 5.099730458221024,
      "grad_norm": 0.6962198615074158,
      "learning_rate": 0.00029445655693470047,
      "loss": 3.3217,
      "step": 47300
    },
    {
      "epoch": 5.105121293800539,
      "grad_norm": 0.7021613121032715,
      "learning_rate": 0.000294132757690232,
      "loss": 3.3286,
      "step": 47350
    },
    {
      "epoch": 5.110512129380054,
      "grad_norm": 0.6529552936553955,
      "learning_rate": 0.0002938089584457636,
      "loss": 3.3382,
      "step": 47400
    },
    {
      "epoch": 5.115902964959568,
      "grad_norm": 0.7062910795211792,
      "learning_rate": 0.00029348515920129517,
      "loss": 3.3252,
      "step": 47450
    },
    {
      "epoch": 5.121293800539084,
      "grad_norm": 0.6736778616905212,
      "learning_rate": 0.0002931613599568267,
      "loss": 3.3304,
      "step": 47500
    },
    {
      "epoch": 5.126684636118599,
      "grad_norm": 0.6905795931816101,
      "learning_rate": 0.00029283756071235833,
      "loss": 3.3278,
      "step": 47550
    },
    {
      "epoch": 5.132075471698113,
      "grad_norm": 0.7116023898124695,
      "learning_rate": 0.0002925137614678899,
      "loss": 3.3291,
      "step": 47600
    },
    {
      "epoch": 5.137466307277628,
      "grad_norm": 0.7711155414581299,
      "learning_rate": 0.00029218996222342143,
      "loss": 3.3198,
      "step": 47650
    },
    {
      "epoch": 5.142857142857143,
      "grad_norm": 0.7277513742446899,
      "learning_rate": 0.00029186616297895303,
      "loss": 3.3177,
      "step": 47700
    },
    {
      "epoch": 5.1482479784366575,
      "grad_norm": 0.6594894528388977,
      "learning_rate": 0.000291548839719374,
      "loss": 3.3157,
      "step": 47750
    },
    {
      "epoch": 5.153638814016173,
      "grad_norm": 0.7083530426025391,
      "learning_rate": 0.00029122504047490554,
      "loss": 3.3343,
      "step": 47800
    },
    {
      "epoch": 5.159029649595688,
      "grad_norm": 0.7079737782478333,
      "learning_rate": 0.0002909012412304371,
      "loss": 3.3067,
      "step": 47850
    },
    {
      "epoch": 5.164420485175202,
      "grad_norm": 0.743047833442688,
      "learning_rate": 0.0002905774419859687,
      "loss": 3.3431,
      "step": 47900
    },
    {
      "epoch": 5.169811320754717,
      "grad_norm": 0.6552191972732544,
      "learning_rate": 0.00029025364274150024,
      "loss": 3.3267,
      "step": 47950
    },
    {
      "epoch": 5.175202156334231,
      "grad_norm": 0.6602107286453247,
      "learning_rate": 0.0002899298434970318,
      "loss": 3.3194,
      "step": 48000
    },
    {
      "epoch": 5.175202156334231,
      "eval_accuracy": 0.38059954880801805,
      "eval_loss": 3.4138906002044678,
      "eval_runtime": 187.9002,
      "eval_samples_per_second": 95.854,
      "eval_steps_per_second": 5.993,
      "step": 48000
    },
    {
      "epoch": 5.180592991913747,
      "grad_norm": 0.6958052515983582,
      "learning_rate": 0.0002896060442525634,
      "loss": 3.3332,
      "step": 48050
    },
    {
      "epoch": 5.185983827493262,
      "grad_norm": 0.6811221241950989,
      "learning_rate": 0.00028928224500809495,
      "loss": 3.3202,
      "step": 48100
    },
    {
      "epoch": 5.191374663072776,
      "grad_norm": 0.6962025165557861,
      "learning_rate": 0.0002889584457636265,
      "loss": 3.3349,
      "step": 48150
    },
    {
      "epoch": 5.196765498652291,
      "grad_norm": 0.6717342138290405,
      "learning_rate": 0.0002886346465191581,
      "loss": 3.3358,
      "step": 48200
    },
    {
      "epoch": 5.202156334231806,
      "grad_norm": 0.6936809420585632,
      "learning_rate": 0.00028831084727468965,
      "loss": 3.3275,
      "step": 48250
    },
    {
      "epoch": 5.2075471698113205,
      "grad_norm": 0.7409997582435608,
      "learning_rate": 0.00028798704803022126,
      "loss": 3.3286,
      "step": 48300
    },
    {
      "epoch": 5.212938005390836,
      "grad_norm": 0.7106682658195496,
      "learning_rate": 0.0002876632487857528,
      "loss": 3.332,
      "step": 48350
    },
    {
      "epoch": 5.218328840970351,
      "grad_norm": 0.65801602602005,
      "learning_rate": 0.0002873394495412844,
      "loss": 3.3419,
      "step": 48400
    },
    {
      "epoch": 5.223719676549865,
      "grad_norm": 0.7524127960205078,
      "learning_rate": 0.0002870156502968159,
      "loss": 3.3331,
      "step": 48450
    },
    {
      "epoch": 5.22911051212938,
      "grad_norm": 0.7451915144920349,
      "learning_rate": 0.0002866918510523475,
      "loss": 3.3196,
      "step": 48500
    },
    {
      "epoch": 5.234501347708895,
      "grad_norm": 0.7274192571640015,
      "learning_rate": 0.00028636805180787907,
      "loss": 3.3301,
      "step": 48550
    },
    {
      "epoch": 5.2398921832884096,
      "grad_norm": 0.6847207546234131,
      "learning_rate": 0.00028604425256341067,
      "loss": 3.3537,
      "step": 48600
    },
    {
      "epoch": 5.245283018867925,
      "grad_norm": 0.6851732730865479,
      "learning_rate": 0.0002857204533189422,
      "loss": 3.3343,
      "step": 48650
    },
    {
      "epoch": 5.250673854447439,
      "grad_norm": 0.7094375491142273,
      "learning_rate": 0.0002853966540744738,
      "loss": 3.3359,
      "step": 48700
    },
    {
      "epoch": 5.256064690026954,
      "grad_norm": 0.7639798521995544,
      "learning_rate": 0.0002850728548300054,
      "loss": 3.3449,
      "step": 48750
    },
    {
      "epoch": 5.261455525606469,
      "grad_norm": 0.6960095167160034,
      "learning_rate": 0.000284749055585537,
      "loss": 3.3278,
      "step": 48800
    },
    {
      "epoch": 5.2668463611859835,
      "grad_norm": 0.6844643950462341,
      "learning_rate": 0.0002844252563410685,
      "loss": 3.3229,
      "step": 48850
    },
    {
      "epoch": 5.272237196765499,
      "grad_norm": 0.7330331802368164,
      "learning_rate": 0.0002841014570966001,
      "loss": 3.338,
      "step": 48900
    },
    {
      "epoch": 5.277628032345014,
      "grad_norm": 0.7339238524436951,
      "learning_rate": 0.00028377765785213163,
      "loss": 3.3371,
      "step": 48950
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.7294909358024597,
      "learning_rate": 0.00028345385860766324,
      "loss": 3.3434,
      "step": 49000
    },
    {
      "epoch": 5.283018867924528,
      "eval_accuracy": 0.38087085506889085,
      "eval_loss": 3.410123586654663,
      "eval_runtime": 187.9534,
      "eval_samples_per_second": 95.827,
      "eval_steps_per_second": 5.991,
      "step": 49000
    },
    {
      "epoch": 5.288409703504043,
      "grad_norm": 0.7706004977226257,
      "learning_rate": 0.0002831300593631948,
      "loss": 3.3418,
      "step": 49050
    },
    {
      "epoch": 5.293800539083558,
      "grad_norm": 0.7382911443710327,
      "learning_rate": 0.0002828062601187264,
      "loss": 3.354,
      "step": 49100
    },
    {
      "epoch": 5.2991913746630726,
      "grad_norm": 0.7077482342720032,
      "learning_rate": 0.00028248246087425794,
      "loss": 3.3486,
      "step": 49150
    },
    {
      "epoch": 5.304582210242588,
      "grad_norm": 0.7349934577941895,
      "learning_rate": 0.0002821586616297895,
      "loss": 3.3568,
      "step": 49200
    },
    {
      "epoch": 5.309973045822103,
      "grad_norm": 0.6874622702598572,
      "learning_rate": 0.0002818348623853211,
      "loss": 3.3349,
      "step": 49250
    },
    {
      "epoch": 5.315363881401617,
      "grad_norm": 0.7308844327926636,
      "learning_rate": 0.00028151106314085265,
      "loss": 3.3386,
      "step": 49300
    },
    {
      "epoch": 5.320754716981132,
      "grad_norm": 0.7208496928215027,
      "learning_rate": 0.0002811872638963842,
      "loss": 3.3406,
      "step": 49350
    },
    {
      "epoch": 5.3261455525606465,
      "grad_norm": 0.6990295052528381,
      "learning_rate": 0.0002808634646519158,
      "loss": 3.3326,
      "step": 49400
    },
    {
      "epoch": 5.331536388140162,
      "grad_norm": 0.6737046241760254,
      "learning_rate": 0.00028053966540744736,
      "loss": 3.3159,
      "step": 49450
    },
    {
      "epoch": 5.336927223719677,
      "grad_norm": 0.725805938243866,
      "learning_rate": 0.0002802158661629789,
      "loss": 3.3448,
      "step": 49500
    },
    {
      "epoch": 5.342318059299191,
      "grad_norm": 0.6815341711044312,
      "learning_rate": 0.0002798920669185105,
      "loss": 3.3416,
      "step": 49550
    },
    {
      "epoch": 5.347708894878706,
      "grad_norm": 0.7110295295715332,
      "learning_rate": 0.00027956826767404206,
      "loss": 3.3343,
      "step": 49600
    },
    {
      "epoch": 5.353099730458221,
      "grad_norm": 0.6776632070541382,
      "learning_rate": 0.00027924446842957367,
      "loss": 3.3388,
      "step": 49650
    },
    {
      "epoch": 5.3584905660377355,
      "grad_norm": 0.7187811136245728,
      "learning_rate": 0.0002789206691851052,
      "loss": 3.3225,
      "step": 49700
    },
    {
      "epoch": 5.363881401617251,
      "grad_norm": 0.7040181756019592,
      "learning_rate": 0.00027859686994063677,
      "loss": 3.3609,
      "step": 49750
    },
    {
      "epoch": 5.369272237196766,
      "grad_norm": 0.7089135646820068,
      "learning_rate": 0.0002782730706961683,
      "loss": 3.3371,
      "step": 49800
    },
    {
      "epoch": 5.37466307277628,
      "grad_norm": 0.7521438002586365,
      "learning_rate": 0.0002779492714516999,
      "loss": 3.3469,
      "step": 49850
    },
    {
      "epoch": 5.380053908355795,
      "grad_norm": 0.7226114869117737,
      "learning_rate": 0.00027762547220723147,
      "loss": 3.3395,
      "step": 49900
    },
    {
      "epoch": 5.38544474393531,
      "grad_norm": 0.6697371602058411,
      "learning_rate": 0.0002773016729627631,
      "loss": 3.3519,
      "step": 49950
    },
    {
      "epoch": 5.390835579514825,
      "grad_norm": 0.7129895091056824,
      "learning_rate": 0.00027697787371829463,
      "loss": 3.3462,
      "step": 50000
    },
    {
      "epoch": 5.390835579514825,
      "eval_accuracy": 0.38111282505005367,
      "eval_loss": 3.410655975341797,
      "eval_runtime": 187.6619,
      "eval_samples_per_second": 95.976,
      "eval_steps_per_second": 6.0,
      "step": 50000
    },
    {
      "epoch": 5.39622641509434,
      "grad_norm": 0.7076137661933899,
      "learning_rate": 0.00027665407447382623,
      "loss": 3.3291,
      "step": 50050
    },
    {
      "epoch": 5.401617250673855,
      "grad_norm": 0.7562270760536194,
      "learning_rate": 0.0002763302752293578,
      "loss": 3.354,
      "step": 50100
    },
    {
      "epoch": 5.407008086253369,
      "grad_norm": 0.6961677074432373,
      "learning_rate": 0.00027600647598488933,
      "loss": 3.3319,
      "step": 50150
    },
    {
      "epoch": 5.412398921832884,
      "grad_norm": 0.677556574344635,
      "learning_rate": 0.0002756826767404209,
      "loss": 3.3385,
      "step": 50200
    },
    {
      "epoch": 5.4177897574123985,
      "grad_norm": 0.7727808356285095,
      "learning_rate": 0.0002753588774959525,
      "loss": 3.3574,
      "step": 50250
    },
    {
      "epoch": 5.423180592991914,
      "grad_norm": 0.7122824788093567,
      "learning_rate": 0.00027503507825148404,
      "loss": 3.322,
      "step": 50300
    },
    {
      "epoch": 5.428571428571429,
      "grad_norm": 0.7861427664756775,
      "learning_rate": 0.00027471127900701564,
      "loss": 3.3397,
      "step": 50350
    },
    {
      "epoch": 5.433962264150943,
      "grad_norm": 0.6940593123435974,
      "learning_rate": 0.0002743874797625472,
      "loss": 3.3523,
      "step": 50400
    },
    {
      "epoch": 5.439353099730458,
      "grad_norm": 0.7210487723350525,
      "learning_rate": 0.0002740636805180788,
      "loss": 3.3354,
      "step": 50450
    },
    {
      "epoch": 5.444743935309973,
      "grad_norm": 0.7028635740280151,
      "learning_rate": 0.00027373988127361035,
      "loss": 3.3486,
      "step": 50500
    },
    {
      "epoch": 5.450134770889488,
      "grad_norm": 0.7137354016304016,
      "learning_rate": 0.0002734160820291419,
      "loss": 3.346,
      "step": 50550
    },
    {
      "epoch": 5.455525606469003,
      "grad_norm": 0.6994820237159729,
      "learning_rate": 0.00027309228278467345,
      "loss": 3.3351,
      "step": 50600
    },
    {
      "epoch": 5.460916442048518,
      "grad_norm": 0.6950395703315735,
      "learning_rate": 0.00027276848354020506,
      "loss": 3.3469,
      "step": 50650
    },
    {
      "epoch": 5.466307277628032,
      "grad_norm": 0.7250814437866211,
      "learning_rate": 0.0002724446842957366,
      "loss": 3.3623,
      "step": 50700
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.7469569444656372,
      "learning_rate": 0.0002721208850512682,
      "loss": 3.3645,
      "step": 50750
    },
    {
      "epoch": 5.4770889487870615,
      "grad_norm": 0.7123376131057739,
      "learning_rate": 0.00027179708580679976,
      "loss": 3.3449,
      "step": 50800
    },
    {
      "epoch": 5.482479784366577,
      "grad_norm": 0.6993563771247864,
      "learning_rate": 0.0002714732865623313,
      "loss": 3.3459,
      "step": 50850
    },
    {
      "epoch": 5.487870619946092,
      "grad_norm": 0.6928504109382629,
      "learning_rate": 0.0002711494873178629,
      "loss": 3.3604,
      "step": 50900
    },
    {
      "epoch": 5.493261455525606,
      "grad_norm": 0.708638608455658,
      "learning_rate": 0.00027082568807339447,
      "loss": 3.3473,
      "step": 50950
    },
    {
      "epoch": 5.498652291105121,
      "grad_norm": 0.6820019483566284,
      "learning_rate": 0.000270501888828926,
      "loss": 3.3481,
      "step": 51000
    },
    {
      "epoch": 5.498652291105121,
      "eval_accuracy": 0.38180364011077816,
      "eval_loss": 3.4029934406280518,
      "eval_runtime": 187.9527,
      "eval_samples_per_second": 95.827,
      "eval_steps_per_second": 5.991,
      "step": 51000
    },
    {
      "epoch": 5.504043126684636,
      "grad_norm": 0.7395426630973816,
      "learning_rate": 0.0002701780895844576,
      "loss": 3.3385,
      "step": 51050
    },
    {
      "epoch": 5.509433962264151,
      "grad_norm": 0.7084470391273499,
      "learning_rate": 0.0002698542903399892,
      "loss": 3.3299,
      "step": 51100
    },
    {
      "epoch": 5.514824797843666,
      "grad_norm": 0.7226378917694092,
      "learning_rate": 0.0002695304910955207,
      "loss": 3.3308,
      "step": 51150
    },
    {
      "epoch": 5.520215633423181,
      "grad_norm": 0.7400267124176025,
      "learning_rate": 0.00026920669185105233,
      "loss": 3.3319,
      "step": 51200
    },
    {
      "epoch": 5.525606469002695,
      "grad_norm": 0.6931690573692322,
      "learning_rate": 0.0002688828926065839,
      "loss": 3.3606,
      "step": 51250
    },
    {
      "epoch": 5.53099730458221,
      "grad_norm": 0.7145960927009583,
      "learning_rate": 0.0002685590933621155,
      "loss": 3.3511,
      "step": 51300
    },
    {
      "epoch": 5.536388140161725,
      "grad_norm": 0.7469211220741272,
      "learning_rate": 0.00026823529411764704,
      "loss": 3.352,
      "step": 51350
    },
    {
      "epoch": 5.54177897574124,
      "grad_norm": 0.7150426506996155,
      "learning_rate": 0.0002679114948731786,
      "loss": 3.3304,
      "step": 51400
    },
    {
      "epoch": 5.547169811320755,
      "grad_norm": 0.6985902190208435,
      "learning_rate": 0.0002675876956287102,
      "loss": 3.3355,
      "step": 51450
    },
    {
      "epoch": 5.55256064690027,
      "grad_norm": 0.7522373795509338,
      "learning_rate": 0.00026726389638424174,
      "loss": 3.3749,
      "step": 51500
    },
    {
      "epoch": 5.557951482479784,
      "grad_norm": 0.7140241861343384,
      "learning_rate": 0.0002669400971397733,
      "loss": 3.3353,
      "step": 51550
    },
    {
      "epoch": 5.563342318059299,
      "grad_norm": 0.7518497705459595,
      "learning_rate": 0.0002666162978953049,
      "loss": 3.35,
      "step": 51600
    },
    {
      "epoch": 5.568733153638814,
      "grad_norm": 0.6793280839920044,
      "learning_rate": 0.00026629249865083645,
      "loss": 3.336,
      "step": 51650
    },
    {
      "epoch": 5.574123989218329,
      "grad_norm": 0.7399218678474426,
      "learning_rate": 0.00026596869940636805,
      "loss": 3.3401,
      "step": 51700
    },
    {
      "epoch": 5.579514824797844,
      "grad_norm": 0.6799671649932861,
      "learning_rate": 0.0002656449001618996,
      "loss": 3.3521,
      "step": 51750
    },
    {
      "epoch": 5.584905660377358,
      "grad_norm": 0.7114550471305847,
      "learning_rate": 0.0002653211009174312,
      "loss": 3.3258,
      "step": 51800
    },
    {
      "epoch": 5.590296495956873,
      "grad_norm": 0.6648984551429749,
      "learning_rate": 0.0002649973016729627,
      "loss": 3.3234,
      "step": 51850
    },
    {
      "epoch": 5.595687331536388,
      "grad_norm": 0.6796683073043823,
      "learning_rate": 0.0002646735024284943,
      "loss": 3.3763,
      "step": 51900
    },
    {
      "epoch": 5.601078167115903,
      "grad_norm": 0.7021110653877258,
      "learning_rate": 0.00026434970318402586,
      "loss": 3.3425,
      "step": 51950
    },
    {
      "epoch": 5.606469002695418,
      "grad_norm": 0.6809260249137878,
      "learning_rate": 0.00026402590393955746,
      "loss": 3.3563,
      "step": 52000
    },
    {
      "epoch": 5.606469002695418,
      "eval_accuracy": 0.382167409979181,
      "eval_loss": 3.3986570835113525,
      "eval_runtime": 187.5942,
      "eval_samples_per_second": 96.01,
      "eval_steps_per_second": 6.002,
      "step": 52000
    },
    {
      "epoch": 5.611859838274933,
      "grad_norm": 0.7304422855377197,
      "learning_rate": 0.000263702104695089,
      "loss": 3.3322,
      "step": 52050
    },
    {
      "epoch": 5.617250673854447,
      "grad_norm": 0.7089975476264954,
      "learning_rate": 0.0002633783054506206,
      "loss": 3.3526,
      "step": 52100
    },
    {
      "epoch": 5.622641509433962,
      "grad_norm": 0.7025209665298462,
      "learning_rate": 0.00026305450620615217,
      "loss": 3.326,
      "step": 52150
    },
    {
      "epoch": 5.628032345013477,
      "grad_norm": 0.7471228837966919,
      "learning_rate": 0.0002627307069616837,
      "loss": 3.3258,
      "step": 52200
    },
    {
      "epoch": 5.633423180592992,
      "grad_norm": 0.7315310835838318,
      "learning_rate": 0.0002624069077172153,
      "loss": 3.3439,
      "step": 52250
    },
    {
      "epoch": 5.638814016172507,
      "grad_norm": 0.7284122705459595,
      "learning_rate": 0.0002620831084727469,
      "loss": 3.3585,
      "step": 52300
    },
    {
      "epoch": 5.644204851752022,
      "grad_norm": 0.7069854736328125,
      "learning_rate": 0.0002617593092282784,
      "loss": 3.3298,
      "step": 52350
    },
    {
      "epoch": 5.649595687331536,
      "grad_norm": 0.7223303318023682,
      "learning_rate": 0.00026143550998381003,
      "loss": 3.3243,
      "step": 52400
    },
    {
      "epoch": 5.654986522911051,
      "grad_norm": 0.7994519472122192,
      "learning_rate": 0.0002611117107393416,
      "loss": 3.3368,
      "step": 52450
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.7556917667388916,
      "learning_rate": 0.00026078791149487313,
      "loss": 3.3491,
      "step": 52500
    },
    {
      "epoch": 5.665768194070081,
      "grad_norm": 0.7107183933258057,
      "learning_rate": 0.00026046411225040474,
      "loss": 3.3432,
      "step": 52550
    },
    {
      "epoch": 5.671159029649596,
      "grad_norm": 0.8043556809425354,
      "learning_rate": 0.0002601403130059363,
      "loss": 3.3448,
      "step": 52600
    },
    {
      "epoch": 5.67654986522911,
      "grad_norm": 0.7135907411575317,
      "learning_rate": 0.0002598165137614679,
      "loss": 3.3574,
      "step": 52650
    },
    {
      "epoch": 5.681940700808625,
      "grad_norm": 0.738898515701294,
      "learning_rate": 0.00025949271451699944,
      "loss": 3.3589,
      "step": 52700
    },
    {
      "epoch": 5.6873315363881405,
      "grad_norm": 0.7450215220451355,
      "learning_rate": 0.000259168915272531,
      "loss": 3.3521,
      "step": 52750
    },
    {
      "epoch": 5.692722371967655,
      "grad_norm": 0.7098891735076904,
      "learning_rate": 0.0002588451160280626,
      "loss": 3.332,
      "step": 52800
    },
    {
      "epoch": 5.69811320754717,
      "grad_norm": 0.7791381478309631,
      "learning_rate": 0.00025852131678359415,
      "loss": 3.3358,
      "step": 52850
    },
    {
      "epoch": 5.703504043126685,
      "grad_norm": 0.7195214629173279,
      "learning_rate": 0.0002581975175391257,
      "loss": 3.3575,
      "step": 52900
    },
    {
      "epoch": 5.708894878706199,
      "grad_norm": 0.7167583703994751,
      "learning_rate": 0.0002578737182946573,
      "loss": 3.3528,
      "step": 52950
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.6789204478263855,
      "learning_rate": 0.00025754991905018885,
      "loss": 3.3483,
      "step": 53000
    },
    {
      "epoch": 5.714285714285714,
      "eval_accuracy": 0.38240546845638257,
      "eval_loss": 3.3940911293029785,
      "eval_runtime": 188.0896,
      "eval_samples_per_second": 95.758,
      "eval_steps_per_second": 5.987,
      "step": 53000
    },
    {
      "epoch": 5.719676549865229,
      "grad_norm": 0.7385661602020264,
      "learning_rate": 0.00025722611980572046,
      "loss": 3.3309,
      "step": 53050
    },
    {
      "epoch": 5.725067385444744,
      "grad_norm": 0.7592161297798157,
      "learning_rate": 0.000256902320561252,
      "loss": 3.3538,
      "step": 53100
    },
    {
      "epoch": 5.730458221024259,
      "grad_norm": 0.7699127197265625,
      "learning_rate": 0.00025657852131678356,
      "loss": 3.3546,
      "step": 53150
    },
    {
      "epoch": 5.735849056603773,
      "grad_norm": 0.7535634636878967,
      "learning_rate": 0.0002562547220723151,
      "loss": 3.3488,
      "step": 53200
    },
    {
      "epoch": 5.741239892183288,
      "grad_norm": 0.7500416040420532,
      "learning_rate": 0.0002559309228278467,
      "loss": 3.3408,
      "step": 53250
    },
    {
      "epoch": 5.7466307277628035,
      "grad_norm": 0.70904141664505,
      "learning_rate": 0.00025560712358337827,
      "loss": 3.3645,
      "step": 53300
    },
    {
      "epoch": 5.752021563342318,
      "grad_norm": 0.739777147769928,
      "learning_rate": 0.00025528332433890987,
      "loss": 3.3288,
      "step": 53350
    },
    {
      "epoch": 5.757412398921833,
      "grad_norm": 0.7218491435050964,
      "learning_rate": 0.0002549595250944414,
      "loss": 3.3466,
      "step": 53400
    },
    {
      "epoch": 5.762803234501348,
      "grad_norm": 0.7896709442138672,
      "learning_rate": 0.000254635725849973,
      "loss": 3.3365,
      "step": 53450
    },
    {
      "epoch": 5.768194070080862,
      "grad_norm": 0.7408962845802307,
      "learning_rate": 0.0002543119266055046,
      "loss": 3.349,
      "step": 53500
    },
    {
      "epoch": 5.773584905660377,
      "grad_norm": 0.7032009959220886,
      "learning_rate": 0.00025398812736103613,
      "loss": 3.3334,
      "step": 53550
    },
    {
      "epoch": 5.7789757412398925,
      "grad_norm": 0.807627260684967,
      "learning_rate": 0.0002536643281165677,
      "loss": 3.3306,
      "step": 53600
    },
    {
      "epoch": 5.784366576819407,
      "grad_norm": 0.6804621815681458,
      "learning_rate": 0.0002533405288720993,
      "loss": 3.3277,
      "step": 53650
    },
    {
      "epoch": 5.789757412398922,
      "grad_norm": 0.7353481650352478,
      "learning_rate": 0.00025301672962763083,
      "loss": 3.338,
      "step": 53700
    },
    {
      "epoch": 5.795148247978437,
      "grad_norm": 0.7213873267173767,
      "learning_rate": 0.0002526994063680518,
      "loss": 3.3453,
      "step": 53750
    },
    {
      "epoch": 5.800539083557951,
      "grad_norm": 0.751306414604187,
      "learning_rate": 0.0002523756071235834,
      "loss": 3.3533,
      "step": 53800
    },
    {
      "epoch": 5.8059299191374665,
      "grad_norm": 0.6616135239601135,
      "learning_rate": 0.00025205180787911494,
      "loss": 3.3369,
      "step": 53850
    },
    {
      "epoch": 5.811320754716981,
      "grad_norm": 0.7167936563491821,
      "learning_rate": 0.0002517280086346465,
      "loss": 3.3414,
      "step": 53900
    },
    {
      "epoch": 5.816711590296496,
      "grad_norm": 0.7028331160545349,
      "learning_rate": 0.00025140420939017804,
      "loss": 3.3411,
      "step": 53950
    },
    {
      "epoch": 5.822102425876011,
      "grad_norm": 0.6635711193084717,
      "learning_rate": 0.00025108041014570965,
      "loss": 3.3368,
      "step": 54000
    },
    {
      "epoch": 5.822102425876011,
      "eval_accuracy": 0.38277771325003496,
      "eval_loss": 3.389596700668335,
      "eval_runtime": 187.815,
      "eval_samples_per_second": 95.898,
      "eval_steps_per_second": 5.995,
      "step": 54000
    },
    {
      "epoch": 5.827493261455525,
      "grad_norm": 0.7287641167640686,
      "learning_rate": 0.0002507566109012412,
      "loss": 3.3552,
      "step": 54050
    },
    {
      "epoch": 5.83288409703504,
      "grad_norm": 0.7763833999633789,
      "learning_rate": 0.0002504328116567728,
      "loss": 3.3402,
      "step": 54100
    },
    {
      "epoch": 5.8382749326145555,
      "grad_norm": 0.7085495591163635,
      "learning_rate": 0.00025010901241230435,
      "loss": 3.3601,
      "step": 54150
    },
    {
      "epoch": 5.84366576819407,
      "grad_norm": 0.724402904510498,
      "learning_rate": 0.0002497852131678359,
      "loss": 3.3646,
      "step": 54200
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.7037736773490906,
      "learning_rate": 0.0002494614139233675,
      "loss": 3.3403,
      "step": 54250
    },
    {
      "epoch": 5.8544474393531,
      "grad_norm": 0.751995325088501,
      "learning_rate": 0.00024913761467889906,
      "loss": 3.3494,
      "step": 54300
    },
    {
      "epoch": 5.859838274932614,
      "grad_norm": 0.749474287033081,
      "learning_rate": 0.0002488138154344306,
      "loss": 3.352,
      "step": 54350
    },
    {
      "epoch": 5.8652291105121295,
      "grad_norm": 0.7056989073753357,
      "learning_rate": 0.0002484900161899622,
      "loss": 3.3472,
      "step": 54400
    },
    {
      "epoch": 5.870619946091644,
      "grad_norm": 0.7020218372344971,
      "learning_rate": 0.00024816621694549376,
      "loss": 3.3353,
      "step": 54450
    },
    {
      "epoch": 5.876010781671159,
      "grad_norm": 0.7736152410507202,
      "learning_rate": 0.00024784241770102537,
      "loss": 3.3473,
      "step": 54500
    },
    {
      "epoch": 5.881401617250674,
      "grad_norm": 0.7925509214401245,
      "learning_rate": 0.0002475186184565569,
      "loss": 3.3463,
      "step": 54550
    },
    {
      "epoch": 5.886792452830189,
      "grad_norm": 0.7352725267410278,
      "learning_rate": 0.00024719481921208847,
      "loss": 3.346,
      "step": 54600
    },
    {
      "epoch": 5.892183288409703,
      "grad_norm": 0.7285502552986145,
      "learning_rate": 0.0002468710199676201,
      "loss": 3.3522,
      "step": 54650
    },
    {
      "epoch": 5.8975741239892185,
      "grad_norm": 0.7065455913543701,
      "learning_rate": 0.0002465472207231516,
      "loss": 3.3408,
      "step": 54700
    },
    {
      "epoch": 5.902964959568733,
      "grad_norm": 0.8250967264175415,
      "learning_rate": 0.0002462234214786832,
      "loss": 3.3413,
      "step": 54750
    },
    {
      "epoch": 5.908355795148248,
      "grad_norm": 0.7509498596191406,
      "learning_rate": 0.0002458996222342148,
      "loss": 3.339,
      "step": 54800
    },
    {
      "epoch": 5.913746630727763,
      "grad_norm": 0.6860405802726746,
      "learning_rate": 0.00024557582298974633,
      "loss": 3.3402,
      "step": 54850
    },
    {
      "epoch": 5.919137466307277,
      "grad_norm": 0.746910035610199,
      "learning_rate": 0.0002452520237452779,
      "loss": 3.3429,
      "step": 54900
    },
    {
      "epoch": 5.9245283018867925,
      "grad_norm": 0.8592144250869751,
      "learning_rate": 0.0002449282245008095,
      "loss": 3.3498,
      "step": 54950
    },
    {
      "epoch": 5.929919137466308,
      "grad_norm": 0.7327863574028015,
      "learning_rate": 0.00024460442525634104,
      "loss": 3.3578,
      "step": 55000
    },
    {
      "epoch": 5.929919137466308,
      "eval_accuracy": 0.38393095500129787,
      "eval_loss": 3.3858537673950195,
      "eval_runtime": 188.1736,
      "eval_samples_per_second": 95.715,
      "eval_steps_per_second": 5.984,
      "step": 55000
    },
    {
      "epoch": 5.935309973045822,
      "grad_norm": 0.7130724787712097,
      "learning_rate": 0.00024428062601187264,
      "loss": 3.3619,
      "step": 55050
    },
    {
      "epoch": 5.940700808625337,
      "grad_norm": 0.7520925998687744,
      "learning_rate": 0.0002439568267674042,
      "loss": 3.3506,
      "step": 55100
    },
    {
      "epoch": 5.946091644204852,
      "grad_norm": 0.7618809938430786,
      "learning_rate": 0.00024363302752293574,
      "loss": 3.3459,
      "step": 55150
    },
    {
      "epoch": 5.951482479784366,
      "grad_norm": 0.757305383682251,
      "learning_rate": 0.00024330922827846732,
      "loss": 3.3539,
      "step": 55200
    },
    {
      "epoch": 5.9568733153638815,
      "grad_norm": 0.7665929198265076,
      "learning_rate": 0.0002429854290339989,
      "loss": 3.3594,
      "step": 55250
    },
    {
      "epoch": 5.962264150943396,
      "grad_norm": 0.7202193737030029,
      "learning_rate": 0.00024266162978953048,
      "loss": 3.3256,
      "step": 55300
    },
    {
      "epoch": 5.967654986522911,
      "grad_norm": 0.7906439304351807,
      "learning_rate": 0.00024233783054506203,
      "loss": 3.3373,
      "step": 55350
    },
    {
      "epoch": 5.973045822102426,
      "grad_norm": 0.6932725310325623,
      "learning_rate": 0.0002420140313005936,
      "loss": 3.3375,
      "step": 55400
    },
    {
      "epoch": 5.97843665768194,
      "grad_norm": 0.7889389991760254,
      "learning_rate": 0.00024169023205612518,
      "loss": 3.3301,
      "step": 55450
    },
    {
      "epoch": 5.9838274932614555,
      "grad_norm": 0.7442350387573242,
      "learning_rate": 0.00024136643281165676,
      "loss": 3.351,
      "step": 55500
    },
    {
      "epoch": 5.989218328840971,
      "grad_norm": 0.7058565616607666,
      "learning_rate": 0.00024104263356718834,
      "loss": 3.3421,
      "step": 55550
    },
    {
      "epoch": 5.994609164420485,
      "grad_norm": 0.7452341914176941,
      "learning_rate": 0.0002407188343227199,
      "loss": 3.3257,
      "step": 55600
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.5780168771743774,
      "learning_rate": 0.00024039503507825147,
      "loss": 3.341,
      "step": 55650
    },
    {
      "epoch": 6.005390835579515,
      "grad_norm": 0.7813258767127991,
      "learning_rate": 0.00024007123583378302,
      "loss": 3.2441,
      "step": 55700
    },
    {
      "epoch": 6.010781671159029,
      "grad_norm": 0.7337445020675659,
      "learning_rate": 0.0002397474365893146,
      "loss": 3.2507,
      "step": 55750
    },
    {
      "epoch": 6.0161725067385445,
      "grad_norm": 0.7185764312744141,
      "learning_rate": 0.00023943011332973555,
      "loss": 3.2637,
      "step": 55800
    },
    {
      "epoch": 6.02156334231806,
      "grad_norm": 0.7511305212974548,
      "learning_rate": 0.00023910631408526712,
      "loss": 3.2531,
      "step": 55850
    },
    {
      "epoch": 6.026954177897574,
      "grad_norm": 0.7478423714637756,
      "learning_rate": 0.0002387825148407987,
      "loss": 3.2628,
      "step": 55900
    },
    {
      "epoch": 6.032345013477089,
      "grad_norm": 0.7264851927757263,
      "learning_rate": 0.00023845871559633025,
      "loss": 3.2552,
      "step": 55950
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.7317351698875427,
      "learning_rate": 0.0002381349163518618,
      "loss": 3.2562,
      "step": 56000
    },
    {
      "epoch": 6.037735849056604,
      "eval_accuracy": 0.3837084338870538,
      "eval_loss": 3.387885570526123,
      "eval_runtime": 187.6852,
      "eval_samples_per_second": 95.964,
      "eval_steps_per_second": 5.999,
      "step": 56000
    },
    {
      "epoch": 6.0431266846361185,
      "grad_norm": 0.8035656213760376,
      "learning_rate": 0.00023781111710739338,
      "loss": 3.2558,
      "step": 56050
    },
    {
      "epoch": 6.048517520215634,
      "grad_norm": 0.7192986011505127,
      "learning_rate": 0.00023748731786292496,
      "loss": 3.2742,
      "step": 56100
    },
    {
      "epoch": 6.053908355795148,
      "grad_norm": 0.6871985197067261,
      "learning_rate": 0.00023716351861845654,
      "loss": 3.2518,
      "step": 56150
    },
    {
      "epoch": 6.059299191374663,
      "grad_norm": 0.6761716604232788,
      "learning_rate": 0.0002368397193739881,
      "loss": 3.2607,
      "step": 56200
    },
    {
      "epoch": 6.064690026954178,
      "grad_norm": 0.7683160305023193,
      "learning_rate": 0.0002365159201295197,
      "loss": 3.2775,
      "step": 56250
    },
    {
      "epoch": 6.070080862533692,
      "grad_norm": 0.7152441143989563,
      "learning_rate": 0.00023619212088505127,
      "loss": 3.2585,
      "step": 56300
    },
    {
      "epoch": 6.0754716981132075,
      "grad_norm": 0.7848925590515137,
      "learning_rate": 0.0002358683216405828,
      "loss": 3.2735,
      "step": 56350
    },
    {
      "epoch": 6.080862533692723,
      "grad_norm": 0.7362651228904724,
      "learning_rate": 0.00023554452239611437,
      "loss": 3.273,
      "step": 56400
    },
    {
      "epoch": 6.086253369272237,
      "grad_norm": 0.7312922477722168,
      "learning_rate": 0.00023522072315164595,
      "loss": 3.2636,
      "step": 56450
    },
    {
      "epoch": 6.091644204851752,
      "grad_norm": 0.7509818077087402,
      "learning_rate": 0.0002349033998920669,
      "loss": 3.2747,
      "step": 56500
    },
    {
      "epoch": 6.097035040431267,
      "grad_norm": 0.7297114729881287,
      "learning_rate": 0.00023457960064759848,
      "loss": 3.2905,
      "step": 56550
    },
    {
      "epoch": 6.1024258760107815,
      "grad_norm": 0.7270781993865967,
      "learning_rate": 0.00023425580140313005,
      "loss": 3.2893,
      "step": 56600
    },
    {
      "epoch": 6.107816711590297,
      "grad_norm": 0.7570849657058716,
      "learning_rate": 0.00023393200215866163,
      "loss": 3.2655,
      "step": 56650
    },
    {
      "epoch": 6.113207547169812,
      "grad_norm": 0.7342410087585449,
      "learning_rate": 0.00023360820291419316,
      "loss": 3.2849,
      "step": 56700
    },
    {
      "epoch": 6.118598382749326,
      "grad_norm": 0.7107300758361816,
      "learning_rate": 0.00023328440366972473,
      "loss": 3.2713,
      "step": 56750
    },
    {
      "epoch": 6.123989218328841,
      "grad_norm": 0.73293536901474,
      "learning_rate": 0.0002329606044252563,
      "loss": 3.2649,
      "step": 56800
    },
    {
      "epoch": 6.129380053908355,
      "grad_norm": 0.7688845992088318,
      "learning_rate": 0.0002326368051807879,
      "loss": 3.2661,
      "step": 56850
    },
    {
      "epoch": 6.1347708894878705,
      "grad_norm": 0.7595694065093994,
      "learning_rate": 0.00023231300593631947,
      "loss": 3.2667,
      "step": 56900
    },
    {
      "epoch": 6.140161725067386,
      "grad_norm": 0.7559347152709961,
      "learning_rate": 0.00023198920669185104,
      "loss": 3.2798,
      "step": 56950
    },
    {
      "epoch": 6.1455525606469,
      "grad_norm": 0.7553902268409729,
      "learning_rate": 0.00023166540744738262,
      "loss": 3.2748,
      "step": 57000
    },
    {
      "epoch": 6.1455525606469,
      "eval_accuracy": 0.3838181733037464,
      "eval_loss": 3.3863282203674316,
      "eval_runtime": 187.9204,
      "eval_samples_per_second": 95.844,
      "eval_steps_per_second": 5.992,
      "step": 57000
    },
    {
      "epoch": 6.150943396226415,
      "grad_norm": 0.7465497851371765,
      "learning_rate": 0.0002313416082029142,
      "loss": 3.2845,
      "step": 57050
    },
    {
      "epoch": 6.15633423180593,
      "grad_norm": 0.7718679308891296,
      "learning_rate": 0.00023101780895844572,
      "loss": 3.2673,
      "step": 57100
    },
    {
      "epoch": 6.1617250673854445,
      "grad_norm": 0.8121668696403503,
      "learning_rate": 0.0002306940097139773,
      "loss": 3.276,
      "step": 57150
    },
    {
      "epoch": 6.16711590296496,
      "grad_norm": 0.7367902994155884,
      "learning_rate": 0.00023037021046950888,
      "loss": 3.2705,
      "step": 57200
    },
    {
      "epoch": 6.172506738544475,
      "grad_norm": 0.7376331686973572,
      "learning_rate": 0.00023004641122504046,
      "loss": 3.2723,
      "step": 57250
    },
    {
      "epoch": 6.177897574123989,
      "grad_norm": 0.8236753940582275,
      "learning_rate": 0.00022972261198057203,
      "loss": 3.2745,
      "step": 57300
    },
    {
      "epoch": 6.183288409703504,
      "grad_norm": 0.8473188877105713,
      "learning_rate": 0.0002293988127361036,
      "loss": 3.2814,
      "step": 57350
    },
    {
      "epoch": 6.188679245283019,
      "grad_norm": 0.7933148741722107,
      "learning_rate": 0.00022907501349163516,
      "loss": 3.2874,
      "step": 57400
    },
    {
      "epoch": 6.1940700808625335,
      "grad_norm": 0.7893574833869934,
      "learning_rate": 0.00022875121424716674,
      "loss": 3.3015,
      "step": 57450
    },
    {
      "epoch": 6.199460916442049,
      "grad_norm": 0.7557196021080017,
      "learning_rate": 0.00022842741500269832,
      "loss": 3.2545,
      "step": 57500
    },
    {
      "epoch": 6.204851752021563,
      "grad_norm": 0.7800120115280151,
      "learning_rate": 0.00022810361575822987,
      "loss": 3.3058,
      "step": 57550
    },
    {
      "epoch": 6.210242587601078,
      "grad_norm": 0.7379505038261414,
      "learning_rate": 0.00022777981651376145,
      "loss": 3.2753,
      "step": 57600
    },
    {
      "epoch": 6.215633423180593,
      "grad_norm": 0.7617130279541016,
      "learning_rate": 0.00022745601726929302,
      "loss": 3.2784,
      "step": 57650
    },
    {
      "epoch": 6.2210242587601075,
      "grad_norm": 0.7396572828292847,
      "learning_rate": 0.00022713221802482457,
      "loss": 3.2981,
      "step": 57700
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.7271841764450073,
      "learning_rate": 0.00022680841878035615,
      "loss": 3.2814,
      "step": 57750
    },
    {
      "epoch": 6.231805929919138,
      "grad_norm": 0.7972626686096191,
      "learning_rate": 0.00022648461953588773,
      "loss": 3.2878,
      "step": 57800
    },
    {
      "epoch": 6.237196765498652,
      "grad_norm": 0.7392578125,
      "learning_rate": 0.0002261608202914193,
      "loss": 3.2872,
      "step": 57850
    },
    {
      "epoch": 6.242587601078167,
      "grad_norm": 0.7566655874252319,
      "learning_rate": 0.00022583702104695088,
      "loss": 3.287,
      "step": 57900
    },
    {
      "epoch": 6.247978436657682,
      "grad_norm": 0.7734372615814209,
      "learning_rate": 0.00022551322180248246,
      "loss": 3.2726,
      "step": 57950
    },
    {
      "epoch": 6.2533692722371965,
      "grad_norm": 0.7498614192008972,
      "learning_rate": 0.00022518942255801399,
      "loss": 3.292,
      "step": 58000
    },
    {
      "epoch": 6.2533692722371965,
      "eval_accuracy": 0.3842432234008767,
      "eval_loss": 3.3843281269073486,
      "eval_runtime": 187.9288,
      "eval_samples_per_second": 95.84,
      "eval_steps_per_second": 5.992,
      "step": 58000
    },
    {
      "epoch": 6.258760107816712,
      "grad_norm": 0.7295949459075928,
      "learning_rate": 0.00022486562331354556,
      "loss": 3.282,
      "step": 58050
    },
    {
      "epoch": 6.264150943396227,
      "grad_norm": 0.7549505829811096,
      "learning_rate": 0.00022454182406907714,
      "loss": 3.2869,
      "step": 58100
    },
    {
      "epoch": 6.269541778975741,
      "grad_norm": 0.7437057495117188,
      "learning_rate": 0.00022421802482460872,
      "loss": 3.2805,
      "step": 58150
    },
    {
      "epoch": 6.274932614555256,
      "grad_norm": 0.7324735522270203,
      "learning_rate": 0.0002238942255801403,
      "loss": 3.279,
      "step": 58200
    },
    {
      "epoch": 6.280323450134771,
      "grad_norm": 0.8076735734939575,
      "learning_rate": 0.00022357042633567187,
      "loss": 3.2862,
      "step": 58250
    },
    {
      "epoch": 6.285714285714286,
      "grad_norm": 0.747568666934967,
      "learning_rate": 0.00022324662709120345,
      "loss": 3.2721,
      "step": 58300
    },
    {
      "epoch": 6.291105121293801,
      "grad_norm": 0.7982004284858704,
      "learning_rate": 0.00022292282784673503,
      "loss": 3.3009,
      "step": 58350
    },
    {
      "epoch": 6.296495956873315,
      "grad_norm": 0.8082500696182251,
      "learning_rate": 0.00022259902860226655,
      "loss": 3.2826,
      "step": 58400
    },
    {
      "epoch": 6.30188679245283,
      "grad_norm": 0.8358715176582336,
      "learning_rate": 0.00022227522935779813,
      "loss": 3.2843,
      "step": 58450
    },
    {
      "epoch": 6.307277628032345,
      "grad_norm": 0.8117343187332153,
      "learning_rate": 0.0002219514301133297,
      "loss": 3.275,
      "step": 58500
    },
    {
      "epoch": 6.3126684636118595,
      "grad_norm": 0.7375354170799255,
      "learning_rate": 0.00022162763086886129,
      "loss": 3.285,
      "step": 58550
    },
    {
      "epoch": 6.318059299191375,
      "grad_norm": 0.7584719061851501,
      "learning_rate": 0.00022130383162439286,
      "loss": 3.3029,
      "step": 58600
    },
    {
      "epoch": 6.32345013477089,
      "grad_norm": 0.7963485717773438,
      "learning_rate": 0.00022098003237992444,
      "loss": 3.279,
      "step": 58650
    },
    {
      "epoch": 6.328840970350404,
      "grad_norm": 0.7515376806259155,
      "learning_rate": 0.00022065623313545602,
      "loss": 3.292,
      "step": 58700
    },
    {
      "epoch": 6.334231805929919,
      "grad_norm": 0.7710037231445312,
      "learning_rate": 0.00022033243389098757,
      "loss": 3.2945,
      "step": 58750
    },
    {
      "epoch": 6.339622641509434,
      "grad_norm": 0.7963765859603882,
      "learning_rate": 0.00022000863464651915,
      "loss": 3.296,
      "step": 58800
    },
    {
      "epoch": 6.345013477088949,
      "grad_norm": 0.7372424006462097,
      "learning_rate": 0.0002196848354020507,
      "loss": 3.28,
      "step": 58850
    },
    {
      "epoch": 6.350404312668464,
      "grad_norm": 0.7944062352180481,
      "learning_rate": 0.00021936103615758227,
      "loss": 3.2815,
      "step": 58900
    },
    {
      "epoch": 6.355795148247978,
      "grad_norm": 0.7355282306671143,
      "learning_rate": 0.00021903723691311385,
      "loss": 3.2831,
      "step": 58950
    },
    {
      "epoch": 6.361185983827493,
      "grad_norm": 0.7471590042114258,
      "learning_rate": 0.00021871343766864543,
      "loss": 3.2756,
      "step": 59000
    },
    {
      "epoch": 6.361185983827493,
      "eval_accuracy": 0.3848307095652896,
      "eval_loss": 3.3796348571777344,
      "eval_runtime": 187.8514,
      "eval_samples_per_second": 95.879,
      "eval_steps_per_second": 5.994,
      "step": 59000
    },
    {
      "epoch": 6.366576819407008,
      "grad_norm": 0.7613081932067871,
      "learning_rate": 0.00021838963842417698,
      "loss": 3.2832,
      "step": 59050
    },
    {
      "epoch": 6.3719676549865225,
      "grad_norm": 0.7622696757316589,
      "learning_rate": 0.00021806583917970856,
      "loss": 3.2789,
      "step": 59100
    },
    {
      "epoch": 6.377358490566038,
      "grad_norm": 0.72235506772995,
      "learning_rate": 0.00021774203993524014,
      "loss": 3.2904,
      "step": 59150
    },
    {
      "epoch": 6.382749326145553,
      "grad_norm": 0.8166873455047607,
      "learning_rate": 0.00021741824069077171,
      "loss": 3.2955,
      "step": 59200
    },
    {
      "epoch": 6.388140161725067,
      "grad_norm": 0.7647126913070679,
      "learning_rate": 0.00021709444144630326,
      "loss": 3.2857,
      "step": 59250
    },
    {
      "epoch": 6.393530997304582,
      "grad_norm": 0.7704625129699707,
      "learning_rate": 0.00021677064220183484,
      "loss": 3.3068,
      "step": 59300
    },
    {
      "epoch": 6.398921832884097,
      "grad_norm": 0.7548606395721436,
      "learning_rate": 0.00021644684295736642,
      "loss": 3.2837,
      "step": 59350
    },
    {
      "epoch": 6.404312668463612,
      "grad_norm": 0.695271909236908,
      "learning_rate": 0.00021612304371289797,
      "loss": 3.2807,
      "step": 59400
    },
    {
      "epoch": 6.409703504043127,
      "grad_norm": 0.7588513493537903,
      "learning_rate": 0.00021579924446842955,
      "loss": 3.2976,
      "step": 59450
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.8067427277565002,
      "learning_rate": 0.00021547544522396113,
      "loss": 3.2828,
      "step": 59500
    },
    {
      "epoch": 6.420485175202156,
      "grad_norm": 0.777260959148407,
      "learning_rate": 0.0002151516459794927,
      "loss": 3.2736,
      "step": 59550
    },
    {
      "epoch": 6.425876010781671,
      "grad_norm": 0.7520415186882019,
      "learning_rate": 0.00021482784673502428,
      "loss": 3.2852,
      "step": 59600
    },
    {
      "epoch": 6.431266846361186,
      "grad_norm": 0.7825960516929626,
      "learning_rate": 0.00021450404749055586,
      "loss": 3.2913,
      "step": 59650
    },
    {
      "epoch": 6.436657681940701,
      "grad_norm": 0.7923210859298706,
      "learning_rate": 0.00021418024824608738,
      "loss": 3.2861,
      "step": 59700
    },
    {
      "epoch": 6.442048517520216,
      "grad_norm": 0.7896773815155029,
      "learning_rate": 0.00021385644900161896,
      "loss": 3.2976,
      "step": 59750
    },
    {
      "epoch": 6.44743935309973,
      "grad_norm": 0.7635443210601807,
      "learning_rate": 0.00021353264975715054,
      "loss": 3.2877,
      "step": 59800
    },
    {
      "epoch": 6.452830188679245,
      "grad_norm": 0.7932662963867188,
      "learning_rate": 0.00021320885051268211,
      "loss": 3.2946,
      "step": 59850
    },
    {
      "epoch": 6.45822102425876,
      "grad_norm": 0.7560425996780396,
      "learning_rate": 0.0002128850512682137,
      "loss": 3.2832,
      "step": 59900
    },
    {
      "epoch": 6.463611859838275,
      "grad_norm": 0.7501300573348999,
      "learning_rate": 0.00021256125202374527,
      "loss": 3.2952,
      "step": 59950
    },
    {
      "epoch": 6.46900269541779,
      "grad_norm": 0.7904677987098694,
      "learning_rate": 0.00021223745277927685,
      "loss": 3.2928,
      "step": 60000
    },
    {
      "epoch": 6.46900269541779,
      "eval_accuracy": 0.384727380668978,
      "eval_loss": 3.3773369789123535,
      "eval_runtime": 188.1838,
      "eval_samples_per_second": 95.71,
      "eval_steps_per_second": 5.984,
      "step": 60000
    },
    {
      "epoch": 6.474393530997305,
      "grad_norm": 0.7343600392341614,
      "learning_rate": 0.00021191365353480843,
      "loss": 3.3014,
      "step": 60050
    },
    {
      "epoch": 6.479784366576819,
      "grad_norm": 0.777730405330658,
      "learning_rate": 0.00021158985429033995,
      "loss": 3.2978,
      "step": 60100
    },
    {
      "epoch": 6.485175202156334,
      "grad_norm": 0.7713602185249329,
      "learning_rate": 0.00021126605504587153,
      "loss": 3.2898,
      "step": 60150
    },
    {
      "epoch": 6.490566037735849,
      "grad_norm": 0.8683859705924988,
      "learning_rate": 0.0002109422558014031,
      "loss": 3.3135,
      "step": 60200
    },
    {
      "epoch": 6.495956873315364,
      "grad_norm": 0.7600846290588379,
      "learning_rate": 0.00021061845655693468,
      "loss": 3.2873,
      "step": 60250
    },
    {
      "epoch": 6.501347708894879,
      "grad_norm": 0.7918633818626404,
      "learning_rate": 0.00021029465731246626,
      "loss": 3.2922,
      "step": 60300
    },
    {
      "epoch": 6.506738544474393,
      "grad_norm": 0.7769505381584167,
      "learning_rate": 0.0002099773340528872,
      "loss": 3.2941,
      "step": 60350
    },
    {
      "epoch": 6.512129380053908,
      "grad_norm": 0.8223928809165955,
      "learning_rate": 0.0002096535348084188,
      "loss": 3.2921,
      "step": 60400
    },
    {
      "epoch": 6.517520215633423,
      "grad_norm": 0.7309505939483643,
      "learning_rate": 0.0002093297355639503,
      "loss": 3.3065,
      "step": 60450
    },
    {
      "epoch": 6.5229110512129385,
      "grad_norm": 0.7776194214820862,
      "learning_rate": 0.0002090059363194819,
      "loss": 3.2934,
      "step": 60500
    },
    {
      "epoch": 6.528301886792453,
      "grad_norm": 0.7898682951927185,
      "learning_rate": 0.00020868213707501347,
      "loss": 3.3116,
      "step": 60550
    },
    {
      "epoch": 6.533692722371968,
      "grad_norm": 0.8079066872596741,
      "learning_rate": 0.00020835833783054505,
      "loss": 3.3073,
      "step": 60600
    },
    {
      "epoch": 6.539083557951482,
      "grad_norm": 0.8752650022506714,
      "learning_rate": 0.00020803453858607662,
      "loss": 3.2962,
      "step": 60650
    },
    {
      "epoch": 6.544474393530997,
      "grad_norm": 0.8422349691390991,
      "learning_rate": 0.0002077107393416082,
      "loss": 3.3121,
      "step": 60700
    },
    {
      "epoch": 6.549865229110512,
      "grad_norm": 0.7801697850227356,
      "learning_rate": 0.00020738694009713975,
      "loss": 3.2964,
      "step": 60750
    },
    {
      "epoch": 6.555256064690027,
      "grad_norm": 0.732362687587738,
      "learning_rate": 0.00020706314085267133,
      "loss": 3.2881,
      "step": 60800
    },
    {
      "epoch": 6.560646900269542,
      "grad_norm": 0.8021662831306458,
      "learning_rate": 0.00020673934160820288,
      "loss": 3.3067,
      "step": 60850
    },
    {
      "epoch": 6.566037735849057,
      "grad_norm": 0.8009557127952576,
      "learning_rate": 0.00020641554236373446,
      "loss": 3.2876,
      "step": 60900
    },
    {
      "epoch": 6.571428571428571,
      "grad_norm": 0.7953367233276367,
      "learning_rate": 0.00020609174311926604,
      "loss": 3.2895,
      "step": 60950
    },
    {
      "epoch": 6.576819407008086,
      "grad_norm": 0.7843896150588989,
      "learning_rate": 0.0002057679438747976,
      "loss": 3.2836,
      "step": 61000
    },
    {
      "epoch": 6.576819407008086,
      "eval_accuracy": 0.3854018977965303,
      "eval_loss": 3.373260498046875,
      "eval_runtime": 187.8909,
      "eval_samples_per_second": 95.859,
      "eval_steps_per_second": 5.993,
      "step": 61000
    },
    {
      "epoch": 6.5822102425876015,
      "grad_norm": 0.7954122424125671,
      "learning_rate": 0.00020544414463032916,
      "loss": 3.2847,
      "step": 61050
    },
    {
      "epoch": 6.587601078167116,
      "grad_norm": 0.7902364730834961,
      "learning_rate": 0.00020512034538586074,
      "loss": 3.2728,
      "step": 61100
    },
    {
      "epoch": 6.592991913746631,
      "grad_norm": 0.7611752152442932,
      "learning_rate": 0.00020479654614139232,
      "loss": 3.2865,
      "step": 61150
    },
    {
      "epoch": 6.598382749326145,
      "grad_norm": 0.8043197989463806,
      "learning_rate": 0.0002044727468969239,
      "loss": 3.2878,
      "step": 61200
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.7516340613365173,
      "learning_rate": 0.00020414894765245547,
      "loss": 3.2894,
      "step": 61250
    },
    {
      "epoch": 6.609164420485175,
      "grad_norm": 0.7979243397712708,
      "learning_rate": 0.00020382514840798702,
      "loss": 3.2967,
      "step": 61300
    },
    {
      "epoch": 6.6145552560646905,
      "grad_norm": 0.7496867775917053,
      "learning_rate": 0.0002035013491635186,
      "loss": 3.2942,
      "step": 61350
    },
    {
      "epoch": 6.619946091644205,
      "grad_norm": 0.8551615476608276,
      "learning_rate": 0.00020317754991905015,
      "loss": 3.2784,
      "step": 61400
    },
    {
      "epoch": 6.62533692722372,
      "grad_norm": 0.7860953211784363,
      "learning_rate": 0.00020285375067458173,
      "loss": 3.3065,
      "step": 61450
    },
    {
      "epoch": 6.630727762803234,
      "grad_norm": 0.8140352964401245,
      "learning_rate": 0.0002025299514301133,
      "loss": 3.2856,
      "step": 61500
    },
    {
      "epoch": 6.636118598382749,
      "grad_norm": 0.7997155785560608,
      "learning_rate": 0.00020220615218564489,
      "loss": 3.3001,
      "step": 61550
    },
    {
      "epoch": 6.6415094339622645,
      "grad_norm": 0.7644395232200623,
      "learning_rate": 0.00020188235294117646,
      "loss": 3.2859,
      "step": 61600
    },
    {
      "epoch": 6.646900269541779,
      "grad_norm": 0.7517040371894836,
      "learning_rate": 0.00020155855369670804,
      "loss": 3.2825,
      "step": 61650
    },
    {
      "epoch": 6.652291105121294,
      "grad_norm": 0.8023728132247925,
      "learning_rate": 0.00020123475445223956,
      "loss": 3.3109,
      "step": 61700
    },
    {
      "epoch": 6.657681940700809,
      "grad_norm": 0.7633260488510132,
      "learning_rate": 0.00020091095520777114,
      "loss": 3.2998,
      "step": 61750
    },
    {
      "epoch": 6.663072776280323,
      "grad_norm": 0.8267096281051636,
      "learning_rate": 0.00020058715596330272,
      "loss": 3.2899,
      "step": 61800
    },
    {
      "epoch": 6.668463611859838,
      "grad_norm": 0.8023743629455566,
      "learning_rate": 0.0002002633567188343,
      "loss": 3.2966,
      "step": 61850
    },
    {
      "epoch": 6.6738544474393535,
      "grad_norm": 0.8002311587333679,
      "learning_rate": 0.00019993955747436588,
      "loss": 3.2937,
      "step": 61900
    },
    {
      "epoch": 6.679245283018868,
      "grad_norm": 0.8148394823074341,
      "learning_rate": 0.00019961575822989745,
      "loss": 3.2936,
      "step": 61950
    },
    {
      "epoch": 6.684636118598383,
      "grad_norm": 0.80229651927948,
      "learning_rate": 0.00019929195898542903,
      "loss": 3.307,
      "step": 62000
    },
    {
      "epoch": 6.684636118598383,
      "eval_accuracy": 0.3859653716727362,
      "eval_loss": 3.3676891326904297,
      "eval_runtime": 188.1942,
      "eval_samples_per_second": 95.704,
      "eval_steps_per_second": 5.983,
      "step": 62000
    },
    {
      "epoch": 6.690026954177897,
      "grad_norm": 0.7992780208587646,
      "learning_rate": 0.0001989681597409606,
      "loss": 3.2901,
      "step": 62050
    },
    {
      "epoch": 6.695417789757412,
      "grad_norm": 0.7275341749191284,
      "learning_rate": 0.00019864436049649216,
      "loss": 3.2889,
      "step": 62100
    },
    {
      "epoch": 6.7008086253369274,
      "grad_norm": 0.8156207203865051,
      "learning_rate": 0.0001983205612520237,
      "loss": 3.2942,
      "step": 62150
    },
    {
      "epoch": 6.706199460916442,
      "grad_norm": 0.8169457912445068,
      "learning_rate": 0.0001979967620075553,
      "loss": 3.299,
      "step": 62200
    },
    {
      "epoch": 6.711590296495957,
      "grad_norm": 0.7171157002449036,
      "learning_rate": 0.00019767296276308686,
      "loss": 3.2987,
      "step": 62250
    },
    {
      "epoch": 6.716981132075472,
      "grad_norm": 0.7815842628479004,
      "learning_rate": 0.00019734916351861844,
      "loss": 3.2905,
      "step": 62300
    },
    {
      "epoch": 6.722371967654986,
      "grad_norm": 0.8042939305305481,
      "learning_rate": 0.00019702536427415002,
      "loss": 3.2956,
      "step": 62350
    },
    {
      "epoch": 6.727762803234501,
      "grad_norm": 0.8041614294052124,
      "learning_rate": 0.0001967015650296816,
      "loss": 3.2962,
      "step": 62400
    },
    {
      "epoch": 6.7331536388140165,
      "grad_norm": 0.7848486304283142,
      "learning_rate": 0.00019637776578521315,
      "loss": 3.2944,
      "step": 62450
    },
    {
      "epoch": 6.738544474393531,
      "grad_norm": 0.8133097290992737,
      "learning_rate": 0.00019605396654074473,
      "loss": 3.2887,
      "step": 62500
    },
    {
      "epoch": 6.743935309973046,
      "grad_norm": 0.7522289752960205,
      "learning_rate": 0.0001957301672962763,
      "loss": 3.3056,
      "step": 62550
    },
    {
      "epoch": 6.74932614555256,
      "grad_norm": 0.7786284685134888,
      "learning_rate": 0.00019540636805180785,
      "loss": 3.2838,
      "step": 62600
    },
    {
      "epoch": 6.754716981132075,
      "grad_norm": 0.7982814311981201,
      "learning_rate": 0.00019508256880733943,
      "loss": 3.2938,
      "step": 62650
    },
    {
      "epoch": 6.7601078167115904,
      "grad_norm": 0.8202456831932068,
      "learning_rate": 0.000194758769562871,
      "loss": 3.2877,
      "step": 62700
    },
    {
      "epoch": 6.765498652291106,
      "grad_norm": 0.7827909588813782,
      "learning_rate": 0.00019443497031840256,
      "loss": 3.2874,
      "step": 62750
    },
    {
      "epoch": 6.77088948787062,
      "grad_norm": 0.7439596056938171,
      "learning_rate": 0.00019411117107393414,
      "loss": 3.2878,
      "step": 62800
    },
    {
      "epoch": 6.776280323450135,
      "grad_norm": 0.8092724084854126,
      "learning_rate": 0.00019378737182946572,
      "loss": 3.2849,
      "step": 62850
    },
    {
      "epoch": 6.781671159029649,
      "grad_norm": 0.7997114658355713,
      "learning_rate": 0.0001934635725849973,
      "loss": 3.3103,
      "step": 62900
    },
    {
      "epoch": 6.787061994609164,
      "grad_norm": Infinity,
      "learning_rate": 0.00019314624932541822,
      "loss": 3.2921,
      "step": 62950
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.7939057946205139,
      "learning_rate": 0.0001928224500809498,
      "loss": 3.2797,
      "step": 63000
    },
    {
      "epoch": 6.7924528301886795,
      "eval_accuracy": 0.3862258126448275,
      "eval_loss": 3.364175796508789,
      "eval_runtime": 187.876,
      "eval_samples_per_second": 95.866,
      "eval_steps_per_second": 5.993,
      "step": 63000
    },
    {
      "epoch": 6.797843665768194,
      "grad_norm": 0.7783874869346619,
      "learning_rate": 0.00019249865083648137,
      "loss": 3.3023,
      "step": 63050
    },
    {
      "epoch": 6.803234501347709,
      "grad_norm": 0.8127438426017761,
      "learning_rate": 0.00019217485159201292,
      "loss": 3.2982,
      "step": 63100
    },
    {
      "epoch": 6.808625336927224,
      "grad_norm": 0.8036924600601196,
      "learning_rate": 0.0001918510523475445,
      "loss": 3.2927,
      "step": 63150
    },
    {
      "epoch": 6.814016172506738,
      "grad_norm": 0.7932625412940979,
      "learning_rate": 0.00019152725310307608,
      "loss": 3.3087,
      "step": 63200
    },
    {
      "epoch": 6.819407008086253,
      "grad_norm": 0.9488426446914673,
      "learning_rate": 0.00019120345385860766,
      "loss": 3.2917,
      "step": 63250
    },
    {
      "epoch": 6.824797843665769,
      "grad_norm": 0.8007382154464722,
      "learning_rate": 0.00019087965461413923,
      "loss": 3.2951,
      "step": 63300
    },
    {
      "epoch": 6.830188679245283,
      "grad_norm": 0.7953688502311707,
      "learning_rate": 0.00019055585536967079,
      "loss": 3.2765,
      "step": 63350
    },
    {
      "epoch": 6.835579514824798,
      "grad_norm": 0.7954296469688416,
      "learning_rate": 0.00019023205612520234,
      "loss": 3.2922,
      "step": 63400
    },
    {
      "epoch": 6.840970350404312,
      "grad_norm": 0.7817032337188721,
      "learning_rate": 0.00018990825688073391,
      "loss": 3.2839,
      "step": 63450
    },
    {
      "epoch": 6.846361185983827,
      "grad_norm": 0.7928920984268188,
      "learning_rate": 0.0001895844576362655,
      "loss": 3.3096,
      "step": 63500
    },
    {
      "epoch": 6.8517520215633425,
      "grad_norm": 0.824849009513855,
      "learning_rate": 0.00018926065839179707,
      "loss": 3.2859,
      "step": 63550
    },
    {
      "epoch": 6.857142857142857,
      "grad_norm": 0.8251289129257202,
      "learning_rate": 0.00018893685914732865,
      "loss": 3.3139,
      "step": 63600
    },
    {
      "epoch": 6.862533692722372,
      "grad_norm": 0.7916178107261658,
      "learning_rate": 0.00018861305990286022,
      "loss": 3.2982,
      "step": 63650
    },
    {
      "epoch": 6.867924528301887,
      "grad_norm": 0.8265122771263123,
      "learning_rate": 0.0001882892606583918,
      "loss": 3.298,
      "step": 63700
    },
    {
      "epoch": 6.873315363881401,
      "grad_norm": 0.7769911289215088,
      "learning_rate": 0.00018796546141392333,
      "loss": 3.3156,
      "step": 63750
    },
    {
      "epoch": 6.878706199460916,
      "grad_norm": 0.8272818326950073,
      "learning_rate": 0.0001876416621694549,
      "loss": 3.3094,
      "step": 63800
    },
    {
      "epoch": 6.884097035040432,
      "grad_norm": 0.8231576085090637,
      "learning_rate": 0.00018731786292498648,
      "loss": 3.2944,
      "step": 63850
    },
    {
      "epoch": 6.889487870619946,
      "grad_norm": 0.8012822866439819,
      "learning_rate": 0.00018699406368051806,
      "loss": 3.2897,
      "step": 63900
    },
    {
      "epoch": 6.894878706199461,
      "grad_norm": 0.8373528122901917,
      "learning_rate": 0.00018667026443604964,
      "loss": 3.2976,
      "step": 63950
    },
    {
      "epoch": 6.900269541778976,
      "grad_norm": 0.8032182455062866,
      "learning_rate": 0.00018634646519158121,
      "loss": 3.2845,
      "step": 64000
    },
    {
      "epoch": 6.900269541778976,
      "eval_accuracy": 0.386607401586832,
      "eval_loss": 3.3606295585632324,
      "eval_runtime": 187.9629,
      "eval_samples_per_second": 95.822,
      "eval_steps_per_second": 5.991,
      "step": 64000
    },
    {
      "epoch": 6.90566037735849,
      "grad_norm": 0.7420918345451355,
      "learning_rate": 0.0001860226659471128,
      "loss": 3.2917,
      "step": 64050
    },
    {
      "epoch": 6.9110512129380055,
      "grad_norm": 0.833095908164978,
      "learning_rate": 0.00018569886670264434,
      "loss": 3.294,
      "step": 64100
    },
    {
      "epoch": 6.916442048517521,
      "grad_norm": 0.8235822319984436,
      "learning_rate": 0.00018537506745817592,
      "loss": 3.3031,
      "step": 64150
    },
    {
      "epoch": 6.921832884097035,
      "grad_norm": 0.8346628546714783,
      "learning_rate": 0.00018505126821370747,
      "loss": 3.2886,
      "step": 64200
    },
    {
      "epoch": 6.92722371967655,
      "grad_norm": 0.7743715643882751,
      "learning_rate": 0.00018472746896923905,
      "loss": 3.3067,
      "step": 64250
    },
    {
      "epoch": 6.932614555256064,
      "grad_norm": 0.8426298499107361,
      "learning_rate": 0.00018440366972477063,
      "loss": 3.2944,
      "step": 64300
    },
    {
      "epoch": 6.938005390835579,
      "grad_norm": 0.772252082824707,
      "learning_rate": 0.0001840798704803022,
      "loss": 3.2894,
      "step": 64350
    },
    {
      "epoch": 6.943396226415095,
      "grad_norm": 0.8040913939476013,
      "learning_rate": 0.00018375607123583378,
      "loss": 3.3017,
      "step": 64400
    },
    {
      "epoch": 6.948787061994609,
      "grad_norm": 0.7839882373809814,
      "learning_rate": 0.00018343227199136533,
      "loss": 3.2766,
      "step": 64450
    },
    {
      "epoch": 6.954177897574124,
      "grad_norm": 0.7890709042549133,
      "learning_rate": 0.0001831084727468969,
      "loss": 3.3068,
      "step": 64500
    },
    {
      "epoch": 6.959568733153639,
      "grad_norm": 0.7926402688026428,
      "learning_rate": 0.0001827846735024285,
      "loss": 3.2977,
      "step": 64550
    },
    {
      "epoch": 6.964959568733153,
      "grad_norm": 0.7957824468612671,
      "learning_rate": 0.00018246087425796004,
      "loss": 3.2825,
      "step": 64600
    },
    {
      "epoch": 6.9703504043126685,
      "grad_norm": 0.8079566359519958,
      "learning_rate": 0.00018213707501349162,
      "loss": 3.2973,
      "step": 64650
    },
    {
      "epoch": 6.975741239892184,
      "grad_norm": 0.7776445746421814,
      "learning_rate": 0.0001818132757690232,
      "loss": 3.3011,
      "step": 64700
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.80714350938797,
      "learning_rate": 0.00018148947652455474,
      "loss": 3.2876,
      "step": 64750
    },
    {
      "epoch": 6.986522911051213,
      "grad_norm": 0.8431934118270874,
      "learning_rate": 0.00018116567728008632,
      "loss": 3.2964,
      "step": 64800
    },
    {
      "epoch": 6.991913746630727,
      "grad_norm": 0.8190756440162659,
      "learning_rate": 0.0001808418780356179,
      "loss": 3.2698,
      "step": 64850
    },
    {
      "epoch": 6.997304582210242,
      "grad_norm": 0.7817225456237793,
      "learning_rate": 0.00018051807879114948,
      "loss": 3.3162,
      "step": 64900
    },
    {
      "epoch": 7.002695417789758,
      "grad_norm": 0.7827392220497131,
      "learning_rate": 0.00018019427954668105,
      "loss": 3.2638,
      "step": 64950
    },
    {
      "epoch": 7.008086253369272,
      "grad_norm": 0.8535230159759521,
      "learning_rate": 0.00017987048030221263,
      "loss": 3.2128,
      "step": 65000
    },
    {
      "epoch": 7.008086253369272,
      "eval_accuracy": 0.3870300613204303,
      "eval_loss": 3.3620991706848145,
      "eval_runtime": 188.0973,
      "eval_samples_per_second": 95.754,
      "eval_steps_per_second": 5.986,
      "step": 65000
    },
    {
      "epoch": 7.013477088948787,
      "grad_norm": 0.8091498613357544,
      "learning_rate": 0.00017954668105774416,
      "loss": 3.2167,
      "step": 65050
    },
    {
      "epoch": 7.018867924528302,
      "grad_norm": 0.7643722891807556,
      "learning_rate": 0.00017922288181327573,
      "loss": 3.2169,
      "step": 65100
    },
    {
      "epoch": 7.024258760107816,
      "grad_norm": 0.8369283080101013,
      "learning_rate": 0.0001788990825688073,
      "loss": 3.2205,
      "step": 65150
    },
    {
      "epoch": 7.0296495956873315,
      "grad_norm": 0.7840744256973267,
      "learning_rate": 0.0001785752833243389,
      "loss": 3.2284,
      "step": 65200
    },
    {
      "epoch": 7.035040431266847,
      "grad_norm": 0.7983671426773071,
      "learning_rate": 0.00017825148407987047,
      "loss": 3.2039,
      "step": 65250
    },
    {
      "epoch": 7.040431266846361,
      "grad_norm": 0.8109228014945984,
      "learning_rate": 0.00017793416082029142,
      "loss": 3.2186,
      "step": 65300
    },
    {
      "epoch": 7.045822102425876,
      "grad_norm": 0.7853065133094788,
      "learning_rate": 0.000177610361575823,
      "loss": 3.2151,
      "step": 65350
    },
    {
      "epoch": 7.051212938005391,
      "grad_norm": 0.7788589596748352,
      "learning_rate": 0.00017728656233135452,
      "loss": 3.2145,
      "step": 65400
    },
    {
      "epoch": 7.056603773584905,
      "grad_norm": 0.7826195955276489,
      "learning_rate": 0.0001769627630868861,
      "loss": 3.2041,
      "step": 65450
    },
    {
      "epoch": 7.061994609164421,
      "grad_norm": 0.7999874949455261,
      "learning_rate": 0.00017663896384241767,
      "loss": 3.2177,
      "step": 65500
    },
    {
      "epoch": 7.067385444743936,
      "grad_norm": 0.834130048751831,
      "learning_rate": 0.00017631516459794925,
      "loss": 3.203,
      "step": 65550
    },
    {
      "epoch": 7.07277628032345,
      "grad_norm": 0.8355494737625122,
      "learning_rate": 0.00017599136535348083,
      "loss": 3.196,
      "step": 65600
    },
    {
      "epoch": 7.078167115902965,
      "grad_norm": 0.7943632006645203,
      "learning_rate": 0.0001756675661090124,
      "loss": 3.2143,
      "step": 65650
    },
    {
      "epoch": 7.083557951482479,
      "grad_norm": 0.7961003184318542,
      "learning_rate": 0.00017534376686454398,
      "loss": 3.2406,
      "step": 65700
    },
    {
      "epoch": 7.0889487870619945,
      "grad_norm": 0.8276134133338928,
      "learning_rate": 0.00017501996762007556,
      "loss": 3.2242,
      "step": 65750
    },
    {
      "epoch": 7.09433962264151,
      "grad_norm": 0.8373634815216064,
      "learning_rate": 0.00017469616837560709,
      "loss": 3.2149,
      "step": 65800
    },
    {
      "epoch": 7.099730458221024,
      "grad_norm": 0.7836037874221802,
      "learning_rate": 0.00017437236913113866,
      "loss": 3.2282,
      "step": 65850
    },
    {
      "epoch": 7.105121293800539,
      "grad_norm": 0.8280946016311646,
      "learning_rate": 0.00017404856988667024,
      "loss": 3.2394,
      "step": 65900
    },
    {
      "epoch": 7.110512129380054,
      "grad_norm": 0.8327875137329102,
      "learning_rate": 0.00017372477064220182,
      "loss": 3.2372,
      "step": 65950
    },
    {
      "epoch": 7.115902964959568,
      "grad_norm": 0.84897780418396,
      "learning_rate": 0.0001734009713977334,
      "loss": 3.2066,
      "step": 66000
    },
    {
      "epoch": 7.115902964959568,
      "eval_accuracy": 0.38692760164722123,
      "eval_loss": 3.3632514476776123,
      "eval_runtime": 188.3038,
      "eval_samples_per_second": 95.649,
      "eval_steps_per_second": 5.98,
      "step": 66000
    },
    {
      "epoch": 7.121293800539084,
      "grad_norm": 0.8668311238288879,
      "learning_rate": 0.00017307717215326497,
      "loss": 3.2588,
      "step": 66050
    },
    {
      "epoch": 7.126684636118599,
      "grad_norm": 0.8419384360313416,
      "learning_rate": 0.00017275337290879655,
      "loss": 3.2144,
      "step": 66100
    },
    {
      "epoch": 7.132075471698113,
      "grad_norm": 0.846095621585846,
      "learning_rate": 0.0001724295736643281,
      "loss": 3.2171,
      "step": 66150
    },
    {
      "epoch": 7.137466307277628,
      "grad_norm": 0.8568212389945984,
      "learning_rate": 0.00017210577441985968,
      "loss": 3.2174,
      "step": 66200
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 0.8674870133399963,
      "learning_rate": 0.00017178197517539123,
      "loss": 3.2317,
      "step": 66250
    },
    {
      "epoch": 7.1482479784366575,
      "grad_norm": 0.7653921246528625,
      "learning_rate": 0.0001714581759309228,
      "loss": 3.2348,
      "step": 66300
    },
    {
      "epoch": 7.153638814016173,
      "grad_norm": 0.8856856822967529,
      "learning_rate": 0.00017113437668645439,
      "loss": 3.2316,
      "step": 66350
    },
    {
      "epoch": 7.159029649595688,
      "grad_norm": 0.7898850440979004,
      "learning_rate": 0.00017081057744198596,
      "loss": 3.24,
      "step": 66400
    },
    {
      "epoch": 7.164420485175202,
      "grad_norm": 0.7871878147125244,
      "learning_rate": 0.00017048677819751751,
      "loss": 3.2378,
      "step": 66450
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.8268200159072876,
      "learning_rate": 0.0001701629789530491,
      "loss": 3.2243,
      "step": 66500
    },
    {
      "epoch": 7.175202156334231,
      "grad_norm": 0.829271674156189,
      "learning_rate": 0.00016983917970858067,
      "loss": 3.2261,
      "step": 66550
    },
    {
      "epoch": 7.180592991913747,
      "grad_norm": 0.8218183517456055,
      "learning_rate": 0.00016951538046411225,
      "loss": 3.2282,
      "step": 66600
    },
    {
      "epoch": 7.185983827493262,
      "grad_norm": 0.8347243666648865,
      "learning_rate": 0.0001691915812196438,
      "loss": 3.2301,
      "step": 66650
    },
    {
      "epoch": 7.191374663072776,
      "grad_norm": 0.8447914123535156,
      "learning_rate": 0.00016886778197517538,
      "loss": 3.2272,
      "step": 66700
    },
    {
      "epoch": 7.196765498652291,
      "grad_norm": 0.854339599609375,
      "learning_rate": 0.00016854398273070693,
      "loss": 3.2207,
      "step": 66750
    },
    {
      "epoch": 7.202156334231806,
      "grad_norm": 0.8469423651695251,
      "learning_rate": 0.0001682201834862385,
      "loss": 3.2379,
      "step": 66800
    },
    {
      "epoch": 7.2075471698113205,
      "grad_norm": 0.8873926401138306,
      "learning_rate": 0.00016789638424177008,
      "loss": 3.2445,
      "step": 66850
    },
    {
      "epoch": 7.212938005390836,
      "grad_norm": 0.8708266615867615,
      "learning_rate": 0.00016757258499730166,
      "loss": 3.2234,
      "step": 66900
    },
    {
      "epoch": 7.218328840970351,
      "grad_norm": 0.8092944622039795,
      "learning_rate": 0.00016724878575283324,
      "loss": 3.244,
      "step": 66950
    },
    {
      "epoch": 7.223719676549865,
      "grad_norm": 0.7969817519187927,
      "learning_rate": 0.00016692498650836481,
      "loss": 3.2329,
      "step": 67000
    },
    {
      "epoch": 7.223719676549865,
      "eval_accuracy": 0.38752866942261094,
      "eval_loss": 3.360783100128174,
      "eval_runtime": 187.837,
      "eval_samples_per_second": 95.886,
      "eval_steps_per_second": 5.995,
      "step": 67000
    },
    {
      "epoch": 7.22911051212938,
      "grad_norm": 0.872371256351471,
      "learning_rate": 0.0001666011872638964,
      "loss": 3.2582,
      "step": 67050
    },
    {
      "epoch": 7.234501347708895,
      "grad_norm": 0.7985860109329224,
      "learning_rate": 0.00016627738801942792,
      "loss": 3.2338,
      "step": 67100
    },
    {
      "epoch": 7.2398921832884096,
      "grad_norm": 0.8462668657302856,
      "learning_rate": 0.0001659535887749595,
      "loss": 3.2476,
      "step": 67150
    },
    {
      "epoch": 7.245283018867925,
      "grad_norm": 0.8263183832168579,
      "learning_rate": 0.00016562978953049107,
      "loss": 3.2315,
      "step": 67200
    },
    {
      "epoch": 7.250673854447439,
      "grad_norm": 0.7963122725486755,
      "learning_rate": 0.00016530599028602265,
      "loss": 3.2401,
      "step": 67250
    },
    {
      "epoch": 7.256064690026954,
      "grad_norm": 0.8212612867355347,
      "learning_rate": 0.00016498219104155423,
      "loss": 3.2396,
      "step": 67300
    },
    {
      "epoch": 7.261455525606469,
      "grad_norm": 0.8274388313293457,
      "learning_rate": 0.0001646583917970858,
      "loss": 3.2335,
      "step": 67350
    },
    {
      "epoch": 7.2668463611859835,
      "grad_norm": 0.8153033256530762,
      "learning_rate": 0.00016433459255261738,
      "loss": 3.2378,
      "step": 67400
    },
    {
      "epoch": 7.272237196765499,
      "grad_norm": 0.8221491575241089,
      "learning_rate": 0.00016401079330814896,
      "loss": 3.2338,
      "step": 67450
    },
    {
      "epoch": 7.277628032345014,
      "grad_norm": 0.8421692252159119,
      "learning_rate": 0.00016368699406368048,
      "loss": 3.2443,
      "step": 67500
    },
    {
      "epoch": 7.283018867924528,
      "grad_norm": 0.8362700343132019,
      "learning_rate": 0.00016336319481921206,
      "loss": 3.2484,
      "step": 67550
    },
    {
      "epoch": 7.288409703504043,
      "grad_norm": 0.7969926595687866,
      "learning_rate": 0.00016303939557474364,
      "loss": 3.2195,
      "step": 67600
    },
    {
      "epoch": 7.293800539083558,
      "grad_norm": 0.7820971012115479,
      "learning_rate": 0.00016271559633027522,
      "loss": 3.2425,
      "step": 67650
    },
    {
      "epoch": 7.2991913746630726,
      "grad_norm": 0.8669634461402893,
      "learning_rate": 0.0001623917970858068,
      "loss": 3.2312,
      "step": 67700
    },
    {
      "epoch": 7.304582210242588,
      "grad_norm": 0.8178690075874329,
      "learning_rate": 0.00016206799784133837,
      "loss": 3.2396,
      "step": 67750
    },
    {
      "epoch": 7.309973045822103,
      "grad_norm": 0.8719444870948792,
      "learning_rate": 0.00016174419859686992,
      "loss": 3.2307,
      "step": 67800
    },
    {
      "epoch": 7.315363881401617,
      "grad_norm": 0.7770711779594421,
      "learning_rate": 0.0001614203993524015,
      "loss": 3.24,
      "step": 67850
    },
    {
      "epoch": 7.320754716981132,
      "grad_norm": 0.8245651721954346,
      "learning_rate": 0.00016109660010793308,
      "loss": 3.2262,
      "step": 67900
    },
    {
      "epoch": 7.3261455525606465,
      "grad_norm": 0.8216030597686768,
      "learning_rate": 0.00016077280086346463,
      "loss": 3.2282,
      "step": 67950
    },
    {
      "epoch": 7.331536388140162,
      "grad_norm": 0.838242769241333,
      "learning_rate": 0.0001604490016189962,
      "loss": 3.2334,
      "step": 68000
    },
    {
      "epoch": 7.331536388140162,
      "eval_accuracy": 0.3875798449327716,
      "eval_loss": 3.357121706008911,
      "eval_runtime": 188.4467,
      "eval_samples_per_second": 95.576,
      "eval_steps_per_second": 5.975,
      "step": 68000
    },
    {
      "epoch": 7.336927223719677,
      "grad_norm": 0.8874527215957642,
      "learning_rate": 0.00016012520237452778,
      "loss": 3.2394,
      "step": 68050
    },
    {
      "epoch": 7.342318059299191,
      "grad_norm": 0.8234398365020752,
      "learning_rate": 0.00015980140313005933,
      "loss": 3.2681,
      "step": 68100
    },
    {
      "epoch": 7.347708894878706,
      "grad_norm": 0.9195815324783325,
      "learning_rate": 0.0001594776038855909,
      "loss": 3.2316,
      "step": 68150
    },
    {
      "epoch": 7.353099730458221,
      "grad_norm": 0.8223610520362854,
      "learning_rate": 0.0001591538046411225,
      "loss": 3.2292,
      "step": 68200
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.8694055676460266,
      "learning_rate": 0.00015883000539665407,
      "loss": 3.2589,
      "step": 68250
    },
    {
      "epoch": 7.363881401617251,
      "grad_norm": 0.8673613667488098,
      "learning_rate": 0.00015850620615218564,
      "loss": 3.2354,
      "step": 68300
    },
    {
      "epoch": 7.369272237196766,
      "grad_norm": 0.8228337168693542,
      "learning_rate": 0.0001581824069077172,
      "loss": 3.2454,
      "step": 68350
    },
    {
      "epoch": 7.37466307277628,
      "grad_norm": 0.8655876517295837,
      "learning_rate": 0.00015785860766324877,
      "loss": 3.2482,
      "step": 68400
    },
    {
      "epoch": 7.380053908355795,
      "grad_norm": 0.8178161978721619,
      "learning_rate": 0.00015753480841878032,
      "loss": 3.2314,
      "step": 68450
    },
    {
      "epoch": 7.38544474393531,
      "grad_norm": 0.8654986023902893,
      "learning_rate": 0.0001572110091743119,
      "loss": 3.2173,
      "step": 68500
    },
    {
      "epoch": 7.390835579514825,
      "grad_norm": 0.866359293460846,
      "learning_rate": 0.00015688720992984348,
      "loss": 3.2428,
      "step": 68550
    },
    {
      "epoch": 7.39622641509434,
      "grad_norm": 0.8526694178581238,
      "learning_rate": 0.00015656341068537506,
      "loss": 3.2247,
      "step": 68600
    },
    {
      "epoch": 7.401617250673855,
      "grad_norm": 0.8770955801010132,
      "learning_rate": 0.00015623961144090663,
      "loss": 3.2474,
      "step": 68650
    },
    {
      "epoch": 7.407008086253369,
      "grad_norm": 0.8251200318336487,
      "learning_rate": 0.0001559158121964382,
      "loss": 3.2239,
      "step": 68700
    },
    {
      "epoch": 7.412398921832884,
      "grad_norm": 0.886862576007843,
      "learning_rate": 0.0001555920129519698,
      "loss": 3.2334,
      "step": 68750
    },
    {
      "epoch": 7.4177897574123985,
      "grad_norm": 0.8815854787826538,
      "learning_rate": 0.0001552682137075013,
      "loss": 3.2502,
      "step": 68800
    },
    {
      "epoch": 7.423180592991914,
      "grad_norm": 0.8504962921142578,
      "learning_rate": 0.0001549444144630329,
      "loss": 3.2438,
      "step": 68850
    },
    {
      "epoch": 7.428571428571429,
      "grad_norm": 0.8565604090690613,
      "learning_rate": 0.00015462061521856447,
      "loss": 3.2513,
      "step": 68900
    },
    {
      "epoch": 7.433962264150943,
      "grad_norm": 0.8494173884391785,
      "learning_rate": 0.00015429681597409605,
      "loss": 3.2412,
      "step": 68950
    },
    {
      "epoch": 7.439353099730458,
      "grad_norm": 0.8011876344680786,
      "learning_rate": 0.000153979492714517,
      "loss": 3.2301,
      "step": 69000
    },
    {
      "epoch": 7.439353099730458,
      "eval_accuracy": 0.3875929919321971,
      "eval_loss": 3.354153871536255,
      "eval_runtime": 188.1144,
      "eval_samples_per_second": 95.745,
      "eval_steps_per_second": 5.986,
      "step": 69000
    },
    {
      "epoch": 7.444743935309973,
      "grad_norm": 0.8193632364273071,
      "learning_rate": 0.00015365569347004858,
      "loss": 3.2271,
      "step": 69050
    },
    {
      "epoch": 7.450134770889488,
      "grad_norm": 0.8151710629463196,
      "learning_rate": 0.00015333189422558015,
      "loss": 3.2332,
      "step": 69100
    },
    {
      "epoch": 7.455525606469003,
      "grad_norm": 0.8996269702911377,
      "learning_rate": 0.00015300809498111168,
      "loss": 3.2432,
      "step": 69150
    },
    {
      "epoch": 7.460916442048518,
      "grad_norm": 0.9261552691459656,
      "learning_rate": 0.00015268429573664325,
      "loss": 3.2588,
      "step": 69200
    },
    {
      "epoch": 7.466307277628032,
      "grad_norm": 0.8579587936401367,
      "learning_rate": 0.00015236049649217483,
      "loss": 3.2349,
      "step": 69250
    },
    {
      "epoch": 7.471698113207547,
      "grad_norm": 0.8887305855751038,
      "learning_rate": 0.0001520366972477064,
      "loss": 3.2629,
      "step": 69300
    },
    {
      "epoch": 7.4770889487870615,
      "grad_norm": 0.9070491194725037,
      "learning_rate": 0.000151712898003238,
      "loss": 3.2475,
      "step": 69350
    },
    {
      "epoch": 7.482479784366577,
      "grad_norm": 0.8597719073295593,
      "learning_rate": 0.00015138909875876956,
      "loss": 3.2416,
      "step": 69400
    },
    {
      "epoch": 7.487870619946092,
      "grad_norm": 0.8557640314102173,
      "learning_rate": 0.00015106529951430114,
      "loss": 3.2421,
      "step": 69450
    },
    {
      "epoch": 7.493261455525606,
      "grad_norm": 0.8627281188964844,
      "learning_rate": 0.0001507415002698327,
      "loss": 3.255,
      "step": 69500
    },
    {
      "epoch": 7.498652291105121,
      "grad_norm": 0.8331064581871033,
      "learning_rate": 0.00015041770102536424,
      "loss": 3.2324,
      "step": 69550
    },
    {
      "epoch": 7.504043126684636,
      "grad_norm": 0.9102504253387451,
      "learning_rate": 0.00015009390178089582,
      "loss": 3.2418,
      "step": 69600
    },
    {
      "epoch": 7.509433962264151,
      "grad_norm": 0.8683810234069824,
      "learning_rate": 0.0001497701025364274,
      "loss": 3.2566,
      "step": 69650
    },
    {
      "epoch": 7.514824797843666,
      "grad_norm": 0.8473634123802185,
      "learning_rate": 0.00014944630329195898,
      "loss": 3.2508,
      "step": 69700
    },
    {
      "epoch": 7.520215633423181,
      "grad_norm": 0.8372735381126404,
      "learning_rate": 0.00014912250404749055,
      "loss": 3.2712,
      "step": 69750
    },
    {
      "epoch": 7.525606469002695,
      "grad_norm": 0.8659421801567078,
      "learning_rate": 0.0001487987048030221,
      "loss": 3.2557,
      "step": 69800
    },
    {
      "epoch": 7.53099730458221,
      "grad_norm": 0.8479022979736328,
      "learning_rate": 0.00014847490555855368,
      "loss": 3.2343,
      "step": 69850
    },
    {
      "epoch": 7.536388140161725,
      "grad_norm": 0.8474248051643372,
      "learning_rate": 0.00014815110631408526,
      "loss": 3.2239,
      "step": 69900
    },
    {
      "epoch": 7.54177897574124,
      "grad_norm": 0.9101057648658752,
      "learning_rate": 0.0001478273070696168,
      "loss": 3.2595,
      "step": 69950
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.8913754224777222,
      "learning_rate": 0.0001475035078251484,
      "loss": 3.2397,
      "step": 70000
    },
    {
      "epoch": 7.547169811320755,
      "eval_accuracy": 0.3883309623462331,
      "eval_loss": 3.3493452072143555,
      "eval_runtime": 188.5729,
      "eval_samples_per_second": 95.512,
      "eval_steps_per_second": 5.971,
      "step": 70000
    },
    {
      "epoch": 7.55256064690027,
      "grad_norm": 0.8284294009208679,
      "learning_rate": 0.00014717970858067997,
      "loss": 3.2311,
      "step": 70050
    },
    {
      "epoch": 7.557951482479784,
      "grad_norm": 0.8397430777549744,
      "learning_rate": 0.00014685590933621154,
      "loss": 3.2495,
      "step": 70100
    },
    {
      "epoch": 7.563342318059299,
      "grad_norm": 0.8663442730903625,
      "learning_rate": 0.0001465321100917431,
      "loss": 3.2463,
      "step": 70150
    },
    {
      "epoch": 7.568733153638814,
      "grad_norm": 0.8732253313064575,
      "learning_rate": 0.00014620831084727467,
      "loss": 3.2396,
      "step": 70200
    },
    {
      "epoch": 7.574123989218329,
      "grad_norm": 0.8437363505363464,
      "learning_rate": 0.00014588451160280625,
      "loss": 3.2518,
      "step": 70250
    },
    {
      "epoch": 7.579514824797844,
      "grad_norm": 0.8276694416999817,
      "learning_rate": 0.0001455607123583378,
      "loss": 3.2439,
      "step": 70300
    },
    {
      "epoch": 7.584905660377358,
      "grad_norm": 0.8312061429023743,
      "learning_rate": 0.00014523691311386938,
      "loss": 3.2447,
      "step": 70350
    },
    {
      "epoch": 7.590296495956873,
      "grad_norm": 0.8033319711685181,
      "learning_rate": 0.00014491311386940096,
      "loss": 3.2543,
      "step": 70400
    },
    {
      "epoch": 7.595687331536388,
      "grad_norm": 0.8969256281852722,
      "learning_rate": 0.00014458931462493253,
      "loss": 3.2566,
      "step": 70450
    },
    {
      "epoch": 7.601078167115903,
      "grad_norm": 0.8511197566986084,
      "learning_rate": 0.00014426551538046408,
      "loss": 3.2485,
      "step": 70500
    },
    {
      "epoch": 7.606469002695418,
      "grad_norm": 0.8295609354972839,
      "learning_rate": 0.00014394171613599566,
      "loss": 3.2324,
      "step": 70550
    },
    {
      "epoch": 7.611859838274933,
      "grad_norm": 0.8153634071350098,
      "learning_rate": 0.00014361791689152724,
      "loss": 3.2491,
      "step": 70600
    },
    {
      "epoch": 7.617250673854447,
      "grad_norm": 0.8315316438674927,
      "learning_rate": 0.00014329411764705882,
      "loss": 3.229,
      "step": 70650
    },
    {
      "epoch": 7.622641509433962,
      "grad_norm": 0.857572615146637,
      "learning_rate": 0.0001429703184025904,
      "loss": 3.2302,
      "step": 70700
    },
    {
      "epoch": 7.628032345013477,
      "grad_norm": 0.8279692530632019,
      "learning_rate": 0.00014264651915812194,
      "loss": 3.2476,
      "step": 70750
    },
    {
      "epoch": 7.633423180592992,
      "grad_norm": 0.8981974124908447,
      "learning_rate": 0.00014232271991365352,
      "loss": 3.2547,
      "step": 70800
    },
    {
      "epoch": 7.638814016172507,
      "grad_norm": 0.8265762329101562,
      "learning_rate": 0.0001419989206691851,
      "loss": 3.2457,
      "step": 70850
    },
    {
      "epoch": 7.644204851752022,
      "grad_norm": 0.871878445148468,
      "learning_rate": 0.00014167512142471668,
      "loss": 3.2525,
      "step": 70900
    },
    {
      "epoch": 7.649595687331536,
      "grad_norm": 0.9873601198196411,
      "learning_rate": 0.00014135132218024823,
      "loss": 3.2456,
      "step": 70950
    },
    {
      "epoch": 7.654986522911051,
      "grad_norm": 0.8204500675201416,
      "learning_rate": 0.0001410275229357798,
      "loss": 3.2356,
      "step": 71000
    },
    {
      "epoch": 7.654986522911051,
      "eval_accuracy": 0.3886576815798912,
      "eval_loss": 3.3460803031921387,
      "eval_runtime": 187.9483,
      "eval_samples_per_second": 95.83,
      "eval_steps_per_second": 5.991,
      "step": 71000
    },
    {
      "epoch": 7.660377358490566,
      "grad_norm": 0.8447879552841187,
      "learning_rate": 0.00014070372369131138,
      "loss": 3.2571,
      "step": 71050
    },
    {
      "epoch": 7.665768194070081,
      "grad_norm": 0.8823720812797546,
      "learning_rate": 0.00014037992444684296,
      "loss": 3.251,
      "step": 71100
    },
    {
      "epoch": 7.671159029649596,
      "grad_norm": 0.8513026833534241,
      "learning_rate": 0.0001400561252023745,
      "loss": 3.2542,
      "step": 71150
    },
    {
      "epoch": 7.67654986522911,
      "grad_norm": 0.8223856687545776,
      "learning_rate": 0.0001397323259579061,
      "loss": 3.2377,
      "step": 71200
    },
    {
      "epoch": 7.681940700808625,
      "grad_norm": 0.8093695044517517,
      "learning_rate": 0.00013940852671343767,
      "loss": 3.2353,
      "step": 71250
    },
    {
      "epoch": 7.6873315363881405,
      "grad_norm": 0.855767011642456,
      "learning_rate": 0.00013908472746896924,
      "loss": 3.2416,
      "step": 71300
    },
    {
      "epoch": 7.692722371967655,
      "grad_norm": 0.861933708190918,
      "learning_rate": 0.0001387609282245008,
      "loss": 3.2258,
      "step": 71350
    },
    {
      "epoch": 7.69811320754717,
      "grad_norm": 0.8508812785148621,
      "learning_rate": 0.00013843712898003237,
      "loss": 3.2479,
      "step": 71400
    },
    {
      "epoch": 7.703504043126685,
      "grad_norm": 0.9000187516212463,
      "learning_rate": 0.00013811332973556395,
      "loss": 3.2494,
      "step": 71450
    },
    {
      "epoch": 7.708894878706199,
      "grad_norm": 0.8213217258453369,
      "learning_rate": 0.0001377895304910955,
      "loss": 3.234,
      "step": 71500
    },
    {
      "epoch": 7.714285714285714,
      "grad_norm": 0.8820887804031372,
      "learning_rate": 0.00013746573124662708,
      "loss": 3.263,
      "step": 71550
    },
    {
      "epoch": 7.719676549865229,
      "grad_norm": 0.8610178828239441,
      "learning_rate": 0.00013714193200215866,
      "loss": 3.2413,
      "step": 71600
    },
    {
      "epoch": 7.725067385444744,
      "grad_norm": 0.8472573161125183,
      "learning_rate": 0.0001368181327576902,
      "loss": 3.2256,
      "step": 71650
    },
    {
      "epoch": 7.730458221024259,
      "grad_norm": 0.9511776566505432,
      "learning_rate": 0.00013649433351322178,
      "loss": 3.2522,
      "step": 71700
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.8717077374458313,
      "learning_rate": 0.00013617701025364274,
      "loss": 3.2526,
      "step": 71750
    },
    {
      "epoch": 7.741239892183288,
      "grad_norm": 0.8659278750419617,
      "learning_rate": 0.0001358532110091743,
      "loss": 3.2549,
      "step": 71800
    },
    {
      "epoch": 7.7466307277628035,
      "grad_norm": 0.8259217739105225,
      "learning_rate": 0.00013552941176470587,
      "loss": 3.2458,
      "step": 71850
    },
    {
      "epoch": 7.752021563342318,
      "grad_norm": 0.906169593334198,
      "learning_rate": 0.00013520561252023744,
      "loss": 3.2155,
      "step": 71900
    },
    {
      "epoch": 7.757412398921833,
      "grad_norm": 0.8461496829986572,
      "learning_rate": 0.00013488181327576902,
      "loss": 3.2572,
      "step": 71950
    },
    {
      "epoch": 7.762803234501348,
      "grad_norm": 0.9183528423309326,
      "learning_rate": 0.00013455801403130057,
      "loss": 3.239,
      "step": 72000
    },
    {
      "epoch": 7.762803234501348,
      "eval_accuracy": 0.38920985555576443,
      "eval_loss": 3.3422865867614746,
      "eval_runtime": 188.2266,
      "eval_samples_per_second": 95.688,
      "eval_steps_per_second": 5.982,
      "step": 72000
    },
    {
      "epoch": 7.768194070080862,
      "grad_norm": 0.8566349148750305,
      "learning_rate": 0.00013423421478683215,
      "loss": 3.2618,
      "step": 72050
    },
    {
      "epoch": 7.773584905660377,
      "grad_norm": 0.8473686575889587,
      "learning_rate": 0.00013391041554236373,
      "loss": 3.2447,
      "step": 72100
    },
    {
      "epoch": 7.7789757412398925,
      "grad_norm": 0.8755962252616882,
      "learning_rate": 0.00013358661629789528,
      "loss": 3.2386,
      "step": 72150
    },
    {
      "epoch": 7.784366576819407,
      "grad_norm": 0.832993745803833,
      "learning_rate": 0.00013326281705342685,
      "loss": 3.2337,
      "step": 72200
    },
    {
      "epoch": 7.789757412398922,
      "grad_norm": 0.8647584915161133,
      "learning_rate": 0.00013293901780895843,
      "loss": 3.2494,
      "step": 72250
    },
    {
      "epoch": 7.795148247978437,
      "grad_norm": 0.877619206905365,
      "learning_rate": 0.00013261521856449,
      "loss": 3.245,
      "step": 72300
    },
    {
      "epoch": 7.800539083557951,
      "grad_norm": 0.920470118522644,
      "learning_rate": 0.00013229141932002156,
      "loss": 3.233,
      "step": 72350
    },
    {
      "epoch": 7.8059299191374665,
      "grad_norm": 0.9275642037391663,
      "learning_rate": 0.00013196762007555314,
      "loss": 3.2381,
      "step": 72400
    },
    {
      "epoch": 7.811320754716981,
      "grad_norm": 0.8567812442779541,
      "learning_rate": 0.00013164382083108472,
      "loss": 3.2538,
      "step": 72450
    },
    {
      "epoch": 7.816711590296496,
      "grad_norm": 0.8508996367454529,
      "learning_rate": 0.0001313200215866163,
      "loss": 3.2415,
      "step": 72500
    },
    {
      "epoch": 7.822102425876011,
      "grad_norm": 0.8709658980369568,
      "learning_rate": 0.00013099622234214784,
      "loss": 3.2361,
      "step": 72550
    },
    {
      "epoch": 7.827493261455525,
      "grad_norm": 0.8481534123420715,
      "learning_rate": 0.00013067242309767942,
      "loss": 3.2271,
      "step": 72600
    },
    {
      "epoch": 7.83288409703504,
      "grad_norm": 0.8258869647979736,
      "learning_rate": 0.000130348623853211,
      "loss": 3.2287,
      "step": 72650
    },
    {
      "epoch": 7.8382749326145555,
      "grad_norm": 0.8969262838363647,
      "learning_rate": 0.00013002482460874258,
      "loss": 3.2518,
      "step": 72700
    },
    {
      "epoch": 7.84366576819407,
      "grad_norm": 0.8625231981277466,
      "learning_rate": 0.00012970102536427413,
      "loss": 3.2441,
      "step": 72750
    },
    {
      "epoch": 7.849056603773585,
      "grad_norm": 0.8450524210929871,
      "learning_rate": 0.0001293772261198057,
      "loss": 3.2201,
      "step": 72800
    },
    {
      "epoch": 7.8544474393531,
      "grad_norm": 0.8817840814590454,
      "learning_rate": 0.00012905342687533728,
      "loss": 3.2418,
      "step": 72850
    },
    {
      "epoch": 7.859838274932614,
      "grad_norm": 0.8885464668273926,
      "learning_rate": 0.00012872962763086886,
      "loss": 3.241,
      "step": 72900
    },
    {
      "epoch": 7.8652291105121295,
      "grad_norm": 0.8640483021736145,
      "learning_rate": 0.00012840582838640044,
      "loss": 3.2513,
      "step": 72950
    },
    {
      "epoch": 7.870619946091644,
      "grad_norm": 0.8919004797935486,
      "learning_rate": 0.000128082029141932,
      "loss": 3.248,
      "step": 73000
    },
    {
      "epoch": 7.870619946091644,
      "eval_accuracy": 0.38931839979069105,
      "eval_loss": 3.3391408920288086,
      "eval_runtime": 187.7588,
      "eval_samples_per_second": 95.926,
      "eval_steps_per_second": 5.997,
      "step": 73000
    },
    {
      "epoch": 7.876010781671159,
      "grad_norm": 0.9472377300262451,
      "learning_rate": 0.00012775822989746357,
      "loss": 3.2351,
      "step": 73050
    },
    {
      "epoch": 7.881401617250674,
      "grad_norm": 0.8974602222442627,
      "learning_rate": 0.00012743443065299514,
      "loss": 3.2414,
      "step": 73100
    },
    {
      "epoch": 7.886792452830189,
      "grad_norm": 0.9496487379074097,
      "learning_rate": 0.00012711063140852672,
      "loss": 3.254,
      "step": 73150
    },
    {
      "epoch": 7.892183288409703,
      "grad_norm": 0.8661764860153198,
      "learning_rate": 0.00012678683216405827,
      "loss": 3.2198,
      "step": 73200
    },
    {
      "epoch": 7.8975741239892185,
      "grad_norm": 0.8774340748786926,
      "learning_rate": 0.00012646303291958985,
      "loss": 3.2429,
      "step": 73250
    },
    {
      "epoch": 7.902964959568733,
      "grad_norm": 0.8722697496414185,
      "learning_rate": 0.00012613923367512143,
      "loss": 3.2337,
      "step": 73300
    },
    {
      "epoch": 7.908355795148248,
      "grad_norm": 0.8887828588485718,
      "learning_rate": 0.00012581543443065298,
      "loss": 3.2708,
      "step": 73350
    },
    {
      "epoch": 7.913746630727763,
      "grad_norm": 0.8771435618400574,
      "learning_rate": 0.00012549163518618456,
      "loss": 3.2568,
      "step": 73400
    },
    {
      "epoch": 7.919137466307277,
      "grad_norm": 0.8421612977981567,
      "learning_rate": 0.00012516783594171613,
      "loss": 3.2556,
      "step": 73450
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.9187849760055542,
      "learning_rate": 0.00012484403669724768,
      "loss": 3.2609,
      "step": 73500
    },
    {
      "epoch": 7.929919137466308,
      "grad_norm": 0.8759550452232361,
      "learning_rate": 0.00012452023745277926,
      "loss": 3.2343,
      "step": 73550
    },
    {
      "epoch": 7.935309973045822,
      "grad_norm": 0.8831597566604614,
      "learning_rate": 0.00012419643820831084,
      "loss": 3.256,
      "step": 73600
    },
    {
      "epoch": 7.940700808625337,
      "grad_norm": 0.8703159093856812,
      "learning_rate": 0.0001238726389638424,
      "loss": 3.2427,
      "step": 73650
    },
    {
      "epoch": 7.946091644204852,
      "grad_norm": 0.8806517720222473,
      "learning_rate": 0.00012354883971937397,
      "loss": 3.2337,
      "step": 73700
    },
    {
      "epoch": 7.951482479784366,
      "grad_norm": 0.8828615546226501,
      "learning_rate": 0.00012322504047490555,
      "loss": 3.2424,
      "step": 73750
    },
    {
      "epoch": 7.9568733153638815,
      "grad_norm": 0.8587687611579895,
      "learning_rate": 0.0001229077172153265,
      "loss": 3.2401,
      "step": 73800
    },
    {
      "epoch": 7.962264150943396,
      "grad_norm": 0.8520889282226562,
      "learning_rate": 0.00012258391797085805,
      "loss": 3.2413,
      "step": 73850
    },
    {
      "epoch": 7.967654986522911,
      "grad_norm": 0.8795452117919922,
      "learning_rate": 0.00012226011872638963,
      "loss": 3.237,
      "step": 73900
    },
    {
      "epoch": 7.973045822102426,
      "grad_norm": 0.8479523658752441,
      "learning_rate": 0.00012193631948192119,
      "loss": 3.2574,
      "step": 73950
    },
    {
      "epoch": 7.97843665768194,
      "grad_norm": 0.8569572567939758,
      "learning_rate": 0.00012161252023745277,
      "loss": 3.2583,
      "step": 74000
    },
    {
      "epoch": 7.97843665768194,
      "eval_accuracy": 0.3900485471968044,
      "eval_loss": 3.3358373641967773,
      "eval_runtime": 188.1945,
      "eval_samples_per_second": 95.704,
      "eval_steps_per_second": 5.983,
      "step": 74000
    },
    {
      "epoch": 7.9838274932614555,
      "grad_norm": 0.8261906504631042,
      "learning_rate": 0.00012128872099298435,
      "loss": 3.2584,
      "step": 74050
    },
    {
      "epoch": 7.989218328840971,
      "grad_norm": 0.9081093072891235,
      "learning_rate": 0.00012096492174851591,
      "loss": 3.2528,
      "step": 74100
    },
    {
      "epoch": 7.994609164420485,
      "grad_norm": 0.8433132767677307,
      "learning_rate": 0.00012064112250404749,
      "loss": 3.2545,
      "step": 74150
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.7616924047470093,
      "learning_rate": 0.00012031732325957905,
      "loss": 3.2255,
      "step": 74200
    },
    {
      "epoch": 8.005390835579515,
      "grad_norm": 0.8508578538894653,
      "learning_rate": 0.00011999352401511062,
      "loss": 3.1774,
      "step": 74250
    },
    {
      "epoch": 8.01078167115903,
      "grad_norm": 0.87180495262146,
      "learning_rate": 0.00011966972477064219,
      "loss": 3.1611,
      "step": 74300
    },
    {
      "epoch": 8.016172506738544,
      "grad_norm": 0.8701367974281311,
      "learning_rate": 0.00011934592552617377,
      "loss": 3.1618,
      "step": 74350
    },
    {
      "epoch": 8.021563342318059,
      "grad_norm": 0.9527740478515625,
      "learning_rate": 0.00011902212628170532,
      "loss": 3.1657,
      "step": 74400
    },
    {
      "epoch": 8.026954177897574,
      "grad_norm": 0.8816781640052795,
      "learning_rate": 0.0001186983270372369,
      "loss": 3.165,
      "step": 74450
    },
    {
      "epoch": 8.032345013477089,
      "grad_norm": 0.8772432804107666,
      "learning_rate": 0.00011837452779276848,
      "loss": 3.169,
      "step": 74500
    },
    {
      "epoch": 8.037735849056604,
      "grad_norm": 0.9046964049339294,
      "learning_rate": 0.00011805072854830005,
      "loss": 3.1862,
      "step": 74550
    },
    {
      "epoch": 8.04312668463612,
      "grad_norm": 0.850584089756012,
      "learning_rate": 0.0001177269293038316,
      "loss": 3.1822,
      "step": 74600
    },
    {
      "epoch": 8.048517520215633,
      "grad_norm": 0.8665777444839478,
      "learning_rate": 0.00011740313005936318,
      "loss": 3.1756,
      "step": 74650
    },
    {
      "epoch": 8.053908355795148,
      "grad_norm": 0.8615607619285583,
      "learning_rate": 0.00011707933081489476,
      "loss": 3.1659,
      "step": 74700
    },
    {
      "epoch": 8.059299191374663,
      "grad_norm": 0.8437297940254211,
      "learning_rate": 0.00011675553157042632,
      "loss": 3.1962,
      "step": 74750
    },
    {
      "epoch": 8.064690026954178,
      "grad_norm": 0.876714825630188,
      "learning_rate": 0.00011643173232595789,
      "loss": 3.1722,
      "step": 74800
    },
    {
      "epoch": 8.070080862533693,
      "grad_norm": 0.8553579449653625,
      "learning_rate": 0.00011610793308148947,
      "loss": 3.1779,
      "step": 74850
    },
    {
      "epoch": 8.075471698113208,
      "grad_norm": 0.8584737777709961,
      "learning_rate": 0.00011578413383702104,
      "loss": 3.1723,
      "step": 74900
    },
    {
      "epoch": 8.080862533692722,
      "grad_norm": 0.9048159718513489,
      "learning_rate": 0.00011546033459255261,
      "loss": 3.1733,
      "step": 74950
    },
    {
      "epoch": 8.086253369272237,
      "grad_norm": 0.8661808371543884,
      "learning_rate": 0.00011513653534808419,
      "loss": 3.1717,
      "step": 75000
    },
    {
      "epoch": 8.086253369272237,
      "eval_accuracy": 0.3897854985554055,
      "eval_loss": 3.341247320175171,
      "eval_runtime": 187.5859,
      "eval_samples_per_second": 96.015,
      "eval_steps_per_second": 6.003,
      "step": 75000
    },
    {
      "epoch": 8.091644204851752,
      "grad_norm": 0.8615157008171082,
      "learning_rate": 0.00011481273610361575,
      "loss": 3.1795,
      "step": 75050
    },
    {
      "epoch": 8.097035040431267,
      "grad_norm": 0.8480276465415955,
      "learning_rate": 0.00011448893685914731,
      "loss": 3.1897,
      "step": 75100
    },
    {
      "epoch": 8.102425876010782,
      "grad_norm": 0.8777347207069397,
      "learning_rate": 0.00011416513761467889,
      "loss": 3.1793,
      "step": 75150
    },
    {
      "epoch": 8.107816711590296,
      "grad_norm": 0.8581756353378296,
      "learning_rate": 0.00011384133837021047,
      "loss": 3.1799,
      "step": 75200
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.897616446018219,
      "learning_rate": 0.00011351753912574202,
      "loss": 3.1801,
      "step": 75250
    },
    {
      "epoch": 8.118598382749326,
      "grad_norm": 0.8755094408988953,
      "learning_rate": 0.0001131937398812736,
      "loss": 3.1826,
      "step": 75300
    },
    {
      "epoch": 8.123989218328841,
      "grad_norm": 0.8945389986038208,
      "learning_rate": 0.00011286994063680517,
      "loss": 3.1673,
      "step": 75350
    },
    {
      "epoch": 8.129380053908356,
      "grad_norm": 0.9047781825065613,
      "learning_rate": 0.00011254614139233675,
      "loss": 3.1784,
      "step": 75400
    },
    {
      "epoch": 8.134770889487871,
      "grad_norm": 0.8923010230064392,
      "learning_rate": 0.0001122223421478683,
      "loss": 3.1798,
      "step": 75450
    },
    {
      "epoch": 8.140161725067385,
      "grad_norm": 0.9317930936813354,
      "learning_rate": 0.00011189854290339988,
      "loss": 3.181,
      "step": 75500
    },
    {
      "epoch": 8.1455525606469,
      "grad_norm": 0.8782373666763306,
      "learning_rate": 0.00011157474365893146,
      "loss": 3.1832,
      "step": 75550
    },
    {
      "epoch": 8.150943396226415,
      "grad_norm": 0.9229250550270081,
      "learning_rate": 0.00011125094441446302,
      "loss": 3.1852,
      "step": 75600
    },
    {
      "epoch": 8.15633423180593,
      "grad_norm": 0.9482505917549133,
      "learning_rate": 0.00011092714516999459,
      "loss": 3.1697,
      "step": 75650
    },
    {
      "epoch": 8.161725067385445,
      "grad_norm": 0.8892439007759094,
      "learning_rate": 0.00011060334592552616,
      "loss": 3.1789,
      "step": 75700
    },
    {
      "epoch": 8.167115902964959,
      "grad_norm": 0.8948121666908264,
      "learning_rate": 0.00011027954668105773,
      "loss": 3.1997,
      "step": 75750
    },
    {
      "epoch": 8.172506738544474,
      "grad_norm": 0.9063841104507446,
      "learning_rate": 0.0001099557474365893,
      "loss": 3.2041,
      "step": 75800
    },
    {
      "epoch": 8.177897574123989,
      "grad_norm": 0.8902472853660583,
      "learning_rate": 0.00010963194819212088,
      "loss": 3.1912,
      "step": 75850
    },
    {
      "epoch": 8.183288409703504,
      "grad_norm": 0.9260648488998413,
      "learning_rate": 0.00010930814894765243,
      "loss": 3.1748,
      "step": 75900
    },
    {
      "epoch": 8.18867924528302,
      "grad_norm": 0.9551469683647156,
      "learning_rate": 0.00010898434970318401,
      "loss": 3.1831,
      "step": 75950
    },
    {
      "epoch": 8.194070080862534,
      "grad_norm": 0.9048367738723755,
      "learning_rate": 0.00010866055045871559,
      "loss": 3.194,
      "step": 76000
    },
    {
      "epoch": 8.194070080862534,
      "eval_accuracy": 0.3899166425909976,
      "eval_loss": 3.3387255668640137,
      "eval_runtime": 187.7379,
      "eval_samples_per_second": 95.937,
      "eval_steps_per_second": 5.998,
      "step": 76000
    },
    {
      "epoch": 8.199460916442048,
      "grad_norm": 0.8931266665458679,
      "learning_rate": 0.00010833675121424717,
      "loss": 3.184,
      "step": 76050
    },
    {
      "epoch": 8.204851752021563,
      "grad_norm": 0.8630874752998352,
      "learning_rate": 0.00010801295196977872,
      "loss": 3.1901,
      "step": 76100
    },
    {
      "epoch": 8.210242587601078,
      "grad_norm": 0.8952473998069763,
      "learning_rate": 0.0001076891527253103,
      "loss": 3.2001,
      "step": 76150
    },
    {
      "epoch": 8.215633423180593,
      "grad_norm": 0.8969290256500244,
      "learning_rate": 0.00010737182946573123,
      "loss": 3.192,
      "step": 76200
    },
    {
      "epoch": 8.221024258760108,
      "grad_norm": 0.8658696413040161,
      "learning_rate": 0.0001070480302212628,
      "loss": 3.1861,
      "step": 76250
    },
    {
      "epoch": 8.226415094339623,
      "grad_norm": 0.8696344494819641,
      "learning_rate": 0.00010672423097679438,
      "loss": 3.1845,
      "step": 76300
    },
    {
      "epoch": 8.231805929919137,
      "grad_norm": 0.9221002459526062,
      "learning_rate": 0.00010640043173232595,
      "loss": 3.1888,
      "step": 76350
    },
    {
      "epoch": 8.237196765498652,
      "grad_norm": 0.9323232769966125,
      "learning_rate": 0.00010607663248785753,
      "loss": 3.1993,
      "step": 76400
    },
    {
      "epoch": 8.242587601078167,
      "grad_norm": 0.9110686182975769,
      "learning_rate": 0.00010575283324338908,
      "loss": 3.2035,
      "step": 76450
    },
    {
      "epoch": 8.247978436657682,
      "grad_norm": 0.9169686436653137,
      "learning_rate": 0.00010542903399892066,
      "loss": 3.1789,
      "step": 76500
    },
    {
      "epoch": 8.253369272237197,
      "grad_norm": Infinity,
      "learning_rate": 0.0001051117107393416,
      "loss": 3.1813,
      "step": 76550
    },
    {
      "epoch": 8.25876010781671,
      "grad_norm": 0.8762148022651672,
      "learning_rate": 0.00010478791149487316,
      "loss": 3.1851,
      "step": 76600
    },
    {
      "epoch": 8.264150943396226,
      "grad_norm": 0.8535869121551514,
      "learning_rate": 0.00010446411225040474,
      "loss": 3.1942,
      "step": 76650
    },
    {
      "epoch": 8.269541778975741,
      "grad_norm": 0.9197494983673096,
      "learning_rate": 0.00010414031300593632,
      "loss": 3.1874,
      "step": 76700
    },
    {
      "epoch": 8.274932614555256,
      "grad_norm": 0.8315194249153137,
      "learning_rate": 0.00010381651376146787,
      "loss": 3.1813,
      "step": 76750
    },
    {
      "epoch": 8.280323450134771,
      "grad_norm": 0.9296077489852905,
      "learning_rate": 0.00010349271451699945,
      "loss": 3.1992,
      "step": 76800
    },
    {
      "epoch": 8.285714285714286,
      "grad_norm": 0.9207568168640137,
      "learning_rate": 0.00010316891527253102,
      "loss": 3.1844,
      "step": 76850
    },
    {
      "epoch": 8.2911051212938,
      "grad_norm": 0.8942183256149292,
      "learning_rate": 0.0001028451160280626,
      "loss": 3.1912,
      "step": 76900
    },
    {
      "epoch": 8.296495956873315,
      "grad_norm": 0.8278528451919556,
      "learning_rate": 0.00010252131678359416,
      "loss": 3.1987,
      "step": 76950
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.8854562044143677,
      "learning_rate": 0.00010219751753912573,
      "loss": 3.1902,
      "step": 77000
    },
    {
      "epoch": 8.30188679245283,
      "eval_accuracy": 0.39015491837397476,
      "eval_loss": 3.336681365966797,
      "eval_runtime": 187.7838,
      "eval_samples_per_second": 95.914,
      "eval_steps_per_second": 5.996,
      "step": 77000
    },
    {
      "epoch": 8.307277628032345,
      "grad_norm": 0.8753238320350647,
      "learning_rate": 0.0001018737182946573,
      "loss": 3.2079,
      "step": 77050
    },
    {
      "epoch": 8.31266846361186,
      "grad_norm": 0.867482602596283,
      "learning_rate": 0.00010154991905018887,
      "loss": 3.1959,
      "step": 77100
    },
    {
      "epoch": 8.318059299191376,
      "grad_norm": 0.8500292301177979,
      "learning_rate": 0.00010122611980572045,
      "loss": 3.1798,
      "step": 77150
    },
    {
      "epoch": 8.323450134770889,
      "grad_norm": 0.8923583030700684,
      "learning_rate": 0.00010090232056125201,
      "loss": 3.1884,
      "step": 77200
    },
    {
      "epoch": 8.328840970350404,
      "grad_norm": 0.9104676246643066,
      "learning_rate": 0.00010057852131678359,
      "loss": 3.1968,
      "step": 77250
    },
    {
      "epoch": 8.33423180592992,
      "grad_norm": 0.9469248056411743,
      "learning_rate": 0.00010025472207231515,
      "loss": 3.1925,
      "step": 77300
    },
    {
      "epoch": 8.339622641509434,
      "grad_norm": 0.9284824728965759,
      "learning_rate": 9.993092282784673e-05,
      "loss": 3.1953,
      "step": 77350
    },
    {
      "epoch": 8.34501347708895,
      "grad_norm": 0.9235027432441711,
      "learning_rate": 9.96071235833783e-05,
      "loss": 3.1919,
      "step": 77400
    },
    {
      "epoch": 8.350404312668463,
      "grad_norm": 0.8894093036651611,
      "learning_rate": 9.928332433890986e-05,
      "loss": 3.1901,
      "step": 77450
    },
    {
      "epoch": 8.355795148247978,
      "grad_norm": 0.9287856817245483,
      "learning_rate": 9.895952509444144e-05,
      "loss": 3.208,
      "step": 77500
    },
    {
      "epoch": 8.361185983827493,
      "grad_norm": 0.8841333985328674,
      "learning_rate": 9.863572584997302e-05,
      "loss": 3.1928,
      "step": 77550
    },
    {
      "epoch": 8.366576819407008,
      "grad_norm": 0.8921537399291992,
      "learning_rate": 9.831192660550457e-05,
      "loss": 3.203,
      "step": 77600
    },
    {
      "epoch": 8.371967654986523,
      "grad_norm": 0.8815743327140808,
      "learning_rate": 9.798812736103614e-05,
      "loss": 3.1849,
      "step": 77650
    },
    {
      "epoch": 8.377358490566039,
      "grad_norm": 0.8852102756500244,
      "learning_rate": 9.766432811656772e-05,
      "loss": 3.1901,
      "step": 77700
    },
    {
      "epoch": 8.382749326145552,
      "grad_norm": 0.8962900042533875,
      "learning_rate": 9.73405288720993e-05,
      "loss": 3.1786,
      "step": 77750
    },
    {
      "epoch": 8.388140161725067,
      "grad_norm": 0.9113929271697998,
      "learning_rate": 9.701672962763086e-05,
      "loss": 3.1965,
      "step": 77800
    },
    {
      "epoch": 8.393530997304582,
      "grad_norm": 0.8877599239349365,
      "learning_rate": 9.669293038316243e-05,
      "loss": 3.1914,
      "step": 77850
    },
    {
      "epoch": 8.398921832884097,
      "grad_norm": 1.0134741067886353,
      "learning_rate": 9.6369131138694e-05,
      "loss": 3.1801,
      "step": 77900
    },
    {
      "epoch": 8.404312668463612,
      "grad_norm": 0.8980883955955505,
      "learning_rate": 9.604533189422557e-05,
      "loss": 3.1971,
      "step": 77950
    },
    {
      "epoch": 8.409703504043126,
      "grad_norm": 0.9101238250732422,
      "learning_rate": 9.572153264975715e-05,
      "loss": 3.1837,
      "step": 78000
    },
    {
      "epoch": 8.409703504043126,
      "eval_accuracy": 0.3904747924757006,
      "eval_loss": 3.333543539047241,
      "eval_runtime": 188.0712,
      "eval_samples_per_second": 95.767,
      "eval_steps_per_second": 5.987,
      "step": 78000
    },
    {
      "epoch": 8.415094339622641,
      "grad_norm": 0.8676778674125671,
      "learning_rate": 9.539773340528871e-05,
      "loss": 3.2043,
      "step": 78050
    },
    {
      "epoch": 8.420485175202156,
      "grad_norm": 0.9355192184448242,
      "learning_rate": 9.507393416082027e-05,
      "loss": 3.1883,
      "step": 78100
    },
    {
      "epoch": 8.425876010781671,
      "grad_norm": 0.9380497336387634,
      "learning_rate": 9.475013491635185e-05,
      "loss": 3.1893,
      "step": 78150
    },
    {
      "epoch": 8.431266846361186,
      "grad_norm": 0.8921291828155518,
      "learning_rate": 9.442633567188343e-05,
      "loss": 3.1987,
      "step": 78200
    },
    {
      "epoch": 8.436657681940702,
      "grad_norm": 0.952408492565155,
      "learning_rate": 9.410253642741498e-05,
      "loss": 3.2036,
      "step": 78250
    },
    {
      "epoch": 8.442048517520215,
      "grad_norm": 0.9249347448348999,
      "learning_rate": 9.377873718294656e-05,
      "loss": 3.1883,
      "step": 78300
    },
    {
      "epoch": 8.44743935309973,
      "grad_norm": 0.9118902683258057,
      "learning_rate": 9.345493793847814e-05,
      "loss": 3.2047,
      "step": 78350
    },
    {
      "epoch": 8.452830188679245,
      "grad_norm": 0.929966390132904,
      "learning_rate": 9.313113869400971e-05,
      "loss": 3.1827,
      "step": 78400
    },
    {
      "epoch": 8.45822102425876,
      "grad_norm": 0.9814649224281311,
      "learning_rate": 9.280733944954126e-05,
      "loss": 3.1744,
      "step": 78450
    },
    {
      "epoch": 8.463611859838275,
      "grad_norm": 0.9400845766067505,
      "learning_rate": 9.248354020507284e-05,
      "loss": 3.2012,
      "step": 78500
    },
    {
      "epoch": 8.46900269541779,
      "grad_norm": 0.9196180701255798,
      "learning_rate": 9.215974096060442e-05,
      "loss": 3.1851,
      "step": 78550
    },
    {
      "epoch": 8.474393530997304,
      "grad_norm": 0.8978266716003418,
      "learning_rate": 9.1835941716136e-05,
      "loss": 3.1822,
      "step": 78600
    },
    {
      "epoch": 8.479784366576819,
      "grad_norm": 0.8619605302810669,
      "learning_rate": 9.151214247166756e-05,
      "loss": 3.1947,
      "step": 78650
    },
    {
      "epoch": 8.485175202156334,
      "grad_norm": 0.9299235939979553,
      "learning_rate": 9.118834322719913e-05,
      "loss": 3.2065,
      "step": 78700
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.8792457580566406,
      "learning_rate": 9.08645439827307e-05,
      "loss": 3.2227,
      "step": 78750
    },
    {
      "epoch": 8.495956873315365,
      "grad_norm": 0.9111963510513306,
      "learning_rate": 9.054074473826227e-05,
      "loss": 3.1876,
      "step": 78800
    },
    {
      "epoch": 8.501347708894878,
      "grad_norm": 0.912173330783844,
      "learning_rate": 9.021694549379385e-05,
      "loss": 3.1958,
      "step": 78850
    },
    {
      "epoch": 8.506738544474393,
      "grad_norm": 0.8932207822799683,
      "learning_rate": 8.989314624932541e-05,
      "loss": 3.2019,
      "step": 78900
    },
    {
      "epoch": 8.512129380053908,
      "grad_norm": 0.8717414736747742,
      "learning_rate": 8.956934700485697e-05,
      "loss": 3.1856,
      "step": 78950
    },
    {
      "epoch": 8.517520215633423,
      "grad_norm": 0.9076499342918396,
      "learning_rate": 8.924554776038855e-05,
      "loss": 3.1819,
      "step": 79000
    },
    {
      "epoch": 8.517520215633423,
      "eval_accuracy": 0.3909775293876182,
      "eval_loss": 3.330061435699463,
      "eval_runtime": 187.9001,
      "eval_samples_per_second": 95.854,
      "eval_steps_per_second": 5.993,
      "step": 79000
    },
    {
      "epoch": 8.522911051212938,
      "grad_norm": 0.9383403062820435,
      "learning_rate": 8.892174851592013e-05,
      "loss": 3.1745,
      "step": 79050
    },
    {
      "epoch": 8.528301886792454,
      "grad_norm": 0.8300234079360962,
      "learning_rate": 8.859794927145168e-05,
      "loss": 3.1902,
      "step": 79100
    },
    {
      "epoch": 8.533692722371967,
      "grad_norm": 0.9023022055625916,
      "learning_rate": 8.827415002698326e-05,
      "loss": 3.1957,
      "step": 79150
    },
    {
      "epoch": 8.539083557951482,
      "grad_norm": 0.8733766078948975,
      "learning_rate": 8.795035078251483e-05,
      "loss": 3.1877,
      "step": 79200
    },
    {
      "epoch": 8.544474393530997,
      "grad_norm": 0.8842893242835999,
      "learning_rate": 8.762655153804641e-05,
      "loss": 3.1868,
      "step": 79250
    },
    {
      "epoch": 8.549865229110512,
      "grad_norm": 0.8878859281539917,
      "learning_rate": 8.730275229357798e-05,
      "loss": 3.1857,
      "step": 79300
    },
    {
      "epoch": 8.555256064690028,
      "grad_norm": 0.9339536428451538,
      "learning_rate": 8.697895304910954e-05,
      "loss": 3.1905,
      "step": 79350
    },
    {
      "epoch": 8.560646900269543,
      "grad_norm": 0.9216422438621521,
      "learning_rate": 8.665515380464112e-05,
      "loss": 3.1944,
      "step": 79400
    },
    {
      "epoch": 8.566037735849056,
      "grad_norm": 0.915511965751648,
      "learning_rate": 8.633135456017268e-05,
      "loss": 3.2111,
      "step": 79450
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 0.8784403204917908,
      "learning_rate": 8.600755531570426e-05,
      "loss": 3.1864,
      "step": 79500
    },
    {
      "epoch": 8.576819407008086,
      "grad_norm": 0.8907803893089294,
      "learning_rate": 8.568375607123582e-05,
      "loss": 3.1804,
      "step": 79550
    },
    {
      "epoch": 8.582210242587601,
      "grad_norm": 0.8826505541801453,
      "learning_rate": 8.53599568267674e-05,
      "loss": 3.1909,
      "step": 79600
    },
    {
      "epoch": 8.587601078167117,
      "grad_norm": 0.871987521648407,
      "learning_rate": 8.503615758229897e-05,
      "loss": 3.1889,
      "step": 79650
    },
    {
      "epoch": 8.59299191374663,
      "grad_norm": 0.9072272777557373,
      "learning_rate": 8.471235833783054e-05,
      "loss": 3.204,
      "step": 79700
    },
    {
      "epoch": 8.598382749326145,
      "grad_norm": 0.9233155846595764,
      "learning_rate": 8.438855909336211e-05,
      "loss": 3.2003,
      "step": 79750
    },
    {
      "epoch": 8.60377358490566,
      "grad_norm": 0.9233723878860474,
      "learning_rate": 8.406475984889367e-05,
      "loss": 3.2135,
      "step": 79800
    },
    {
      "epoch": 8.609164420485175,
      "grad_norm": 0.9055242538452148,
      "learning_rate": 8.374096060442525e-05,
      "loss": 3.1896,
      "step": 79850
    },
    {
      "epoch": 8.61455525606469,
      "grad_norm": 0.8707368969917297,
      "learning_rate": 8.341716135995683e-05,
      "loss": 3.1881,
      "step": 79900
    },
    {
      "epoch": 8.619946091644206,
      "grad_norm": 0.9200837016105652,
      "learning_rate": 8.309336211548838e-05,
      "loss": 3.2068,
      "step": 79950
    },
    {
      "epoch": 8.625336927223719,
      "grad_norm": 0.9371185898780823,
      "learning_rate": 8.276956287101996e-05,
      "loss": 3.217,
      "step": 80000
    },
    {
      "epoch": 8.625336927223719,
      "eval_accuracy": 0.39134336366088956,
      "eval_loss": 3.3252391815185547,
      "eval_runtime": 187.9935,
      "eval_samples_per_second": 95.807,
      "eval_steps_per_second": 5.99,
      "step": 80000
    },
    {
      "epoch": 8.630727762803234,
      "grad_norm": 0.9155548810958862,
      "learning_rate": 8.244576362655153e-05,
      "loss": 3.1843,
      "step": 80050
    },
    {
      "epoch": 8.63611859838275,
      "grad_norm": 0.9367262125015259,
      "learning_rate": 8.212196438208311e-05,
      "loss": 3.1933,
      "step": 80100
    },
    {
      "epoch": 8.641509433962264,
      "grad_norm": 0.9463507533073425,
      "learning_rate": 8.179816513761467e-05,
      "loss": 3.1984,
      "step": 80150
    },
    {
      "epoch": 8.64690026954178,
      "grad_norm": 0.978238046169281,
      "learning_rate": 8.147436589314624e-05,
      "loss": 3.2138,
      "step": 80200
    },
    {
      "epoch": 8.652291105121293,
      "grad_norm": 0.9424788355827332,
      "learning_rate": 8.115056664867782e-05,
      "loss": 3.1938,
      "step": 80250
    },
    {
      "epoch": 8.657681940700808,
      "grad_norm": 0.9349570870399475,
      "learning_rate": 8.082676740420938e-05,
      "loss": 3.2039,
      "step": 80300
    },
    {
      "epoch": 8.663072776280323,
      "grad_norm": 0.9339366555213928,
      "learning_rate": 8.050296815974096e-05,
      "loss": 3.1935,
      "step": 80350
    },
    {
      "epoch": 8.668463611859838,
      "grad_norm": 0.9126559495925903,
      "learning_rate": 8.017916891527252e-05,
      "loss": 3.2023,
      "step": 80400
    },
    {
      "epoch": 8.673854447439354,
      "grad_norm": 0.9444116353988647,
      "learning_rate": 7.985536967080409e-05,
      "loss": 3.1947,
      "step": 80450
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.9091424345970154,
      "learning_rate": 7.953157042633566e-05,
      "loss": 3.1969,
      "step": 80500
    },
    {
      "epoch": 8.684636118598382,
      "grad_norm": 0.8752775192260742,
      "learning_rate": 7.920777118186724e-05,
      "loss": 3.1877,
      "step": 80550
    },
    {
      "epoch": 8.690026954177897,
      "grad_norm": 0.8782562017440796,
      "learning_rate": 7.888397193739879e-05,
      "loss": 3.1763,
      "step": 80600
    },
    {
      "epoch": 8.695417789757412,
      "grad_norm": 0.8987087607383728,
      "learning_rate": 7.856017269293037e-05,
      "loss": 3.195,
      "step": 80650
    },
    {
      "epoch": 8.700808625336927,
      "grad_norm": 0.9534454345703125,
      "learning_rate": 7.824284943335132e-05,
      "loss": 3.1834,
      "step": 80700
    },
    {
      "epoch": 8.706199460916443,
      "grad_norm": 1.0178718566894531,
      "learning_rate": 7.791905018888289e-05,
      "loss": 3.1851,
      "step": 80750
    },
    {
      "epoch": 8.711590296495956,
      "grad_norm": 0.9170519709587097,
      "learning_rate": 7.759525094441445e-05,
      "loss": 3.1812,
      "step": 80800
    },
    {
      "epoch": 8.716981132075471,
      "grad_norm": 0.8797026872634888,
      "learning_rate": 7.727145169994603e-05,
      "loss": 3.1884,
      "step": 80850
    },
    {
      "epoch": 8.722371967654986,
      "grad_norm": 0.9378693103790283,
      "learning_rate": 7.69476524554776e-05,
      "loss": 3.2125,
      "step": 80900
    },
    {
      "epoch": 8.727762803234501,
      "grad_norm": 0.9277364015579224,
      "learning_rate": 7.662385321100916e-05,
      "loss": 3.1871,
      "step": 80950
    },
    {
      "epoch": 8.733153638814017,
      "grad_norm": 0.9785609841346741,
      "learning_rate": 7.630005396654073e-05,
      "loss": 3.1705,
      "step": 81000
    },
    {
      "epoch": 8.733153638814017,
      "eval_accuracy": 0.39164357158992097,
      "eval_loss": 3.32247257232666,
      "eval_runtime": 187.8542,
      "eval_samples_per_second": 95.878,
      "eval_steps_per_second": 5.994,
      "step": 81000
    },
    {
      "epoch": 8.738544474393532,
      "grad_norm": 0.973168671131134,
      "learning_rate": 7.597625472207231e-05,
      "loss": 3.1917,
      "step": 81050
    },
    {
      "epoch": 8.743935309973045,
      "grad_norm": 0.9081736207008362,
      "learning_rate": 7.565245547760389e-05,
      "loss": 3.1902,
      "step": 81100
    },
    {
      "epoch": 8.74932614555256,
      "grad_norm": 0.9135347008705139,
      "learning_rate": 7.532865623313544e-05,
      "loss": 3.1826,
      "step": 81150
    },
    {
      "epoch": 8.754716981132075,
      "grad_norm": 0.8944445252418518,
      "learning_rate": 7.500485698866702e-05,
      "loss": 3.2065,
      "step": 81200
    },
    {
      "epoch": 8.76010781671159,
      "grad_norm": 0.919346034526825,
      "learning_rate": 7.46810577441986e-05,
      "loss": 3.1938,
      "step": 81250
    },
    {
      "epoch": 8.765498652291106,
      "grad_norm": 0.8800582885742188,
      "learning_rate": 7.435725849973016e-05,
      "loss": 3.1885,
      "step": 81300
    },
    {
      "epoch": 8.77088948787062,
      "grad_norm": 0.9364388585090637,
      "learning_rate": 7.403345925526174e-05,
      "loss": 3.1847,
      "step": 81350
    },
    {
      "epoch": 8.776280323450134,
      "grad_norm": 0.9442126154899597,
      "learning_rate": 7.37096600107933e-05,
      "loss": 3.19,
      "step": 81400
    },
    {
      "epoch": 8.78167115902965,
      "grad_norm": 0.9176348447799683,
      "learning_rate": 7.338586076632488e-05,
      "loss": 3.1993,
      "step": 81450
    },
    {
      "epoch": 8.787061994609164,
      "grad_norm": 0.9438028931617737,
      "learning_rate": 7.306206152185644e-05,
      "loss": 3.2149,
      "step": 81500
    },
    {
      "epoch": 8.79245283018868,
      "grad_norm": 0.8861234188079834,
      "learning_rate": 7.273826227738801e-05,
      "loss": 3.1937,
      "step": 81550
    },
    {
      "epoch": 8.797843665768195,
      "grad_norm": 0.8872656226158142,
      "learning_rate": 7.241446303291958e-05,
      "loss": 3.1829,
      "step": 81600
    },
    {
      "epoch": 8.80323450134771,
      "grad_norm": 0.8715137839317322,
      "learning_rate": 7.209066378845115e-05,
      "loss": 3.1958,
      "step": 81650
    },
    {
      "epoch": 8.808625336927223,
      "grad_norm": 0.9722141623497009,
      "learning_rate": 7.176686454398273e-05,
      "loss": 3.1777,
      "step": 81700
    },
    {
      "epoch": 8.814016172506738,
      "grad_norm": 0.961341917514801,
      "learning_rate": 7.144306529951429e-05,
      "loss": 3.1831,
      "step": 81750
    },
    {
      "epoch": 8.819407008086253,
      "grad_norm": 0.8844859600067139,
      "learning_rate": 7.111926605504587e-05,
      "loss": 3.1845,
      "step": 81800
    },
    {
      "epoch": 8.824797843665769,
      "grad_norm": 0.9381985068321228,
      "learning_rate": 7.079546681057743e-05,
      "loss": 3.2038,
      "step": 81850
    },
    {
      "epoch": 8.830188679245284,
      "grad_norm": 0.9069252014160156,
      "learning_rate": 7.047166756610901e-05,
      "loss": 3.1925,
      "step": 81900
    },
    {
      "epoch": 8.835579514824797,
      "grad_norm": 0.9448679089546204,
      "learning_rate": 7.015434430652993e-05,
      "loss": 3.2069,
      "step": 81950
    },
    {
      "epoch": 8.840970350404312,
      "grad_norm": 0.9136323928833008,
      "learning_rate": 6.983054506206151e-05,
      "loss": 3.1868,
      "step": 82000
    },
    {
      "epoch": 8.840970350404312,
      "eval_accuracy": 0.3918973847358556,
      "eval_loss": 3.3201887607574463,
      "eval_runtime": 188.2231,
      "eval_samples_per_second": 95.69,
      "eval_steps_per_second": 5.982,
      "step": 82000
    },
    {
      "epoch": 8.846361185983827,
      "grad_norm": 0.893839418888092,
      "learning_rate": 6.950674581759309e-05,
      "loss": 3.1889,
      "step": 82050
    },
    {
      "epoch": 8.851752021563343,
      "grad_norm": 0.9275647401809692,
      "learning_rate": 6.918294657312465e-05,
      "loss": 3.1863,
      "step": 82100
    },
    {
      "epoch": 8.857142857142858,
      "grad_norm": 0.848946213722229,
      "learning_rate": 6.885914732865623e-05,
      "loss": 3.1965,
      "step": 82150
    },
    {
      "epoch": 8.862533692722373,
      "grad_norm": 0.9524142742156982,
      "learning_rate": 6.85353480841878e-05,
      "loss": 3.1925,
      "step": 82200
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.9237378835678101,
      "learning_rate": 6.821154883971937e-05,
      "loss": 3.2198,
      "step": 82250
    },
    {
      "epoch": 8.873315363881401,
      "grad_norm": 0.9178951978683472,
      "learning_rate": 6.788774959525094e-05,
      "loss": 3.2063,
      "step": 82300
    },
    {
      "epoch": 8.878706199460916,
      "grad_norm": 0.9111886024475098,
      "learning_rate": 6.756395035078252e-05,
      "loss": 3.1763,
      "step": 82350
    },
    {
      "epoch": 8.884097035040432,
      "grad_norm": 0.8813912868499756,
      "learning_rate": 6.724015110631408e-05,
      "loss": 3.1891,
      "step": 82400
    },
    {
      "epoch": 8.889487870619947,
      "grad_norm": 0.9190998077392578,
      "learning_rate": 6.691635186184566e-05,
      "loss": 3.1971,
      "step": 82450
    },
    {
      "epoch": 8.89487870619946,
      "grad_norm": 0.963971734046936,
      "learning_rate": 6.659255261737722e-05,
      "loss": 3.199,
      "step": 82500
    },
    {
      "epoch": 8.900269541778975,
      "grad_norm": 0.9044879078865051,
      "learning_rate": 6.626875337290879e-05,
      "loss": 3.2037,
      "step": 82550
    },
    {
      "epoch": 8.90566037735849,
      "grad_norm": 0.9314145445823669,
      "learning_rate": 6.594495412844036e-05,
      "loss": 3.1844,
      "step": 82600
    },
    {
      "epoch": 8.911051212938006,
      "grad_norm": 0.9088249206542969,
      "learning_rate": 6.562115488397193e-05,
      "loss": 3.1967,
      "step": 82650
    },
    {
      "epoch": 8.91644204851752,
      "grad_norm": 0.9422711730003357,
      "learning_rate": 6.52973556395035e-05,
      "loss": 3.1891,
      "step": 82700
    },
    {
      "epoch": 8.921832884097036,
      "grad_norm": 0.9077197909355164,
      "learning_rate": 6.497355639503507e-05,
      "loss": 3.2024,
      "step": 82750
    },
    {
      "epoch": 8.92722371967655,
      "grad_norm": 0.9061179757118225,
      "learning_rate": 6.464975715056663e-05,
      "loss": 3.1981,
      "step": 82800
    },
    {
      "epoch": 8.932614555256064,
      "grad_norm": 0.9592944979667664,
      "learning_rate": 6.432595790609821e-05,
      "loss": 3.204,
      "step": 82850
    },
    {
      "epoch": 8.93800539083558,
      "grad_norm": 0.8956493139266968,
      "learning_rate": 6.400215866162979e-05,
      "loss": 3.2006,
      "step": 82900
    },
    {
      "epoch": 8.943396226415095,
      "grad_norm": 0.8995712995529175,
      "learning_rate": 6.367835941716135e-05,
      "loss": 3.18,
      "step": 82950
    },
    {
      "epoch": 8.94878706199461,
      "grad_norm": 0.96676105260849,
      "learning_rate": 6.335456017269293e-05,
      "loss": 3.2008,
      "step": 83000
    },
    {
      "epoch": 8.94878706199461,
      "eval_accuracy": 0.39247530944614084,
      "eval_loss": 3.3157753944396973,
      "eval_runtime": 187.7013,
      "eval_samples_per_second": 95.956,
      "eval_steps_per_second": 5.999,
      "step": 83000
    },
    {
      "epoch": 8.954177897574123,
      "grad_norm": 0.9426843523979187,
      "learning_rate": 6.30307609282245e-05,
      "loss": 3.1781,
      "step": 83050
    },
    {
      "epoch": 8.959568733153638,
      "grad_norm": 0.8954929113388062,
      "learning_rate": 6.270696168375607e-05,
      "loss": 3.1773,
      "step": 83100
    },
    {
      "epoch": 8.964959568733153,
      "grad_norm": 0.9750449657440186,
      "learning_rate": 6.238316243928764e-05,
      "loss": 3.199,
      "step": 83150
    },
    {
      "epoch": 8.970350404312669,
      "grad_norm": 0.9364018440246582,
      "learning_rate": 6.205936319481921e-05,
      "loss": 3.2079,
      "step": 83200
    },
    {
      "epoch": 8.975741239892184,
      "grad_norm": 0.8858622908592224,
      "learning_rate": 6.173556395035078e-05,
      "loss": 3.1922,
      "step": 83250
    },
    {
      "epoch": 8.981132075471699,
      "grad_norm": 1.0074647665023804,
      "learning_rate": 6.141176470588236e-05,
      "loss": 3.1942,
      "step": 83300
    },
    {
      "epoch": 8.986522911051212,
      "grad_norm": 0.9643685817718506,
      "learning_rate": 6.108796546141392e-05,
      "loss": 3.1766,
      "step": 83350
    },
    {
      "epoch": 8.991913746630727,
      "grad_norm": 0.9266127943992615,
      "learning_rate": 6.076416621694549e-05,
      "loss": 3.1933,
      "step": 83400
    },
    {
      "epoch": 8.997304582210242,
      "grad_norm": 0.9136055707931519,
      "learning_rate": 6.0440366972477055e-05,
      "loss": 3.2052,
      "step": 83450
    },
    {
      "epoch": 9.002695417789758,
      "grad_norm": 0.9789246916770935,
      "learning_rate": 6.011656772800863e-05,
      "loss": 3.1685,
      "step": 83500
    },
    {
      "epoch": 9.008086253369273,
      "grad_norm": 0.9144122004508972,
      "learning_rate": 5.9792768483540197e-05,
      "loss": 3.1438,
      "step": 83550
    },
    {
      "epoch": 9.013477088948788,
      "grad_norm": 0.9275460243225098,
      "learning_rate": 5.946896923907177e-05,
      "loss": 3.1252,
      "step": 83600
    },
    {
      "epoch": 9.018867924528301,
      "grad_norm": 1.0540989637374878,
      "learning_rate": 5.9145169994603345e-05,
      "loss": 3.1482,
      "step": 83650
    },
    {
      "epoch": 9.024258760107816,
      "grad_norm": 0.8946913480758667,
      "learning_rate": 5.882137075013491e-05,
      "loss": 3.1316,
      "step": 83700
    },
    {
      "epoch": 9.029649595687331,
      "grad_norm": 0.9386449456214905,
      "learning_rate": 5.849757150566649e-05,
      "loss": 3.1166,
      "step": 83750
    },
    {
      "epoch": 9.035040431266847,
      "grad_norm": 0.916728675365448,
      "learning_rate": 5.817377226119805e-05,
      "loss": 3.1525,
      "step": 83800
    },
    {
      "epoch": 9.040431266846362,
      "grad_norm": 0.890503466129303,
      "learning_rate": 5.784997301672963e-05,
      "loss": 3.1461,
      "step": 83850
    },
    {
      "epoch": 9.045822102425875,
      "grad_norm": 0.8731235861778259,
      "learning_rate": 5.752617377226119e-05,
      "loss": 3.1499,
      "step": 83900
    },
    {
      "epoch": 9.05121293800539,
      "grad_norm": 0.9045177698135376,
      "learning_rate": 5.7202374527792764e-05,
      "loss": 3.1331,
      "step": 83950
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.9173175692558289,
      "learning_rate": 5.6878575283324335e-05,
      "loss": 3.1378,
      "step": 84000
    },
    {
      "epoch": 9.056603773584905,
      "eval_accuracy": 0.39228864378487555,
      "eval_loss": 3.3212382793426514,
      "eval_runtime": 188.0841,
      "eval_samples_per_second": 95.76,
      "eval_steps_per_second": 5.987,
      "step": 84000
    },
    {
      "epoch": 9.06199460916442,
      "grad_norm": 0.9891923069953918,
      "learning_rate": 5.6554776038855905e-05,
      "loss": 3.1346,
      "step": 84050
    },
    {
      "epoch": 9.067385444743936,
      "grad_norm": 0.9098547697067261,
      "learning_rate": 5.623097679438747e-05,
      "loss": 3.14,
      "step": 84100
    },
    {
      "epoch": 9.07277628032345,
      "grad_norm": 0.9232341647148132,
      "learning_rate": 5.590717754991905e-05,
      "loss": 3.1318,
      "step": 84150
    },
    {
      "epoch": 9.078167115902964,
      "grad_norm": 0.9636071920394897,
      "learning_rate": 5.558337830545061e-05,
      "loss": 3.1412,
      "step": 84200
    },
    {
      "epoch": 9.08355795148248,
      "grad_norm": 0.9031469821929932,
      "learning_rate": 5.525957906098219e-05,
      "loss": 3.1383,
      "step": 84250
    },
    {
      "epoch": 9.088948787061994,
      "grad_norm": 0.9007047414779663,
      "learning_rate": 5.493577981651375e-05,
      "loss": 3.1277,
      "step": 84300
    },
    {
      "epoch": 9.09433962264151,
      "grad_norm": 0.9313275814056396,
      "learning_rate": 5.4611980572045324e-05,
      "loss": 3.1248,
      "step": 84350
    },
    {
      "epoch": 9.099730458221025,
      "grad_norm": 0.8984010219573975,
      "learning_rate": 5.4288181327576895e-05,
      "loss": 3.1446,
      "step": 84400
    },
    {
      "epoch": 9.10512129380054,
      "grad_norm": 0.9281125664710999,
      "learning_rate": 5.3964382083108466e-05,
      "loss": 3.1514,
      "step": 84450
    },
    {
      "epoch": 9.110512129380053,
      "grad_norm": 0.921140730381012,
      "learning_rate": 5.3640582838640043e-05,
      "loss": 3.1314,
      "step": 84500
    },
    {
      "epoch": 9.115902964959568,
      "grad_norm": 0.9595467448234558,
      "learning_rate": 5.331678359417161e-05,
      "loss": 3.1323,
      "step": 84550
    },
    {
      "epoch": 9.121293800539084,
      "grad_norm": 0.9736076593399048,
      "learning_rate": 5.2992984349703185e-05,
      "loss": 3.1573,
      "step": 84600
    },
    {
      "epoch": 9.126684636118599,
      "grad_norm": 0.9062066078186035,
      "learning_rate": 5.266918510523475e-05,
      "loss": 3.1342,
      "step": 84650
    },
    {
      "epoch": 9.132075471698114,
      "grad_norm": 0.8909493684768677,
      "learning_rate": 5.234538586076632e-05,
      "loss": 3.1348,
      "step": 84700
    },
    {
      "epoch": 9.137466307277627,
      "grad_norm": 0.9599141478538513,
      "learning_rate": 5.202158661629789e-05,
      "loss": 3.1318,
      "step": 84750
    },
    {
      "epoch": 9.142857142857142,
      "grad_norm": 0.8932557106018066,
      "learning_rate": 5.169778737182946e-05,
      "loss": 3.132,
      "step": 84800
    },
    {
      "epoch": 9.148247978436657,
      "grad_norm": 0.9026200771331787,
      "learning_rate": 5.1373988127361026e-05,
      "loss": 3.1385,
      "step": 84850
    },
    {
      "epoch": 9.153638814016173,
      "grad_norm": 0.9215351939201355,
      "learning_rate": 5.1050188882892604e-05,
      "loss": 3.1445,
      "step": 84900
    },
    {
      "epoch": 9.159029649595688,
      "grad_norm": 0.9547321200370789,
      "learning_rate": 5.072638963842417e-05,
      "loss": 3.1533,
      "step": 84950
    },
    {
      "epoch": 9.164420485175203,
      "grad_norm": 0.9079762101173401,
      "learning_rate": 5.0402590393955746e-05,
      "loss": 3.1322,
      "step": 85000
    },
    {
      "epoch": 9.164420485175203,
      "eval_accuracy": 0.3925814633175356,
      "eval_loss": 3.317584276199341,
      "eval_runtime": 187.6367,
      "eval_samples_per_second": 95.989,
      "eval_steps_per_second": 6.001,
      "step": 85000
    },
    {
      "epoch": 9.169811320754716,
      "grad_norm": 0.8956434726715088,
      "learning_rate": 5.007879114948731e-05,
      "loss": 3.1324,
      "step": 85050
    },
    {
      "epoch": 9.175202156334231,
      "grad_norm": 0.9152560234069824,
      "learning_rate": 4.975499190501889e-05,
      "loss": 3.1568,
      "step": 85100
    },
    {
      "epoch": 9.180592991913747,
      "grad_norm": 0.915358304977417,
      "learning_rate": 4.943119266055045e-05,
      "loss": 3.1344,
      "step": 85150
    },
    {
      "epoch": 9.185983827493262,
      "grad_norm": 0.9076443910598755,
      "learning_rate": 4.910739341608202e-05,
      "loss": 3.1544,
      "step": 85200
    },
    {
      "epoch": 9.191374663072777,
      "grad_norm": 0.8821456432342529,
      "learning_rate": 4.87835941716136e-05,
      "loss": 3.1362,
      "step": 85250
    },
    {
      "epoch": 9.19676549865229,
      "grad_norm": 0.9051779508590698,
      "learning_rate": 4.8459794927145164e-05,
      "loss": 3.1339,
      "step": 85300
    },
    {
      "epoch": 9.202156334231805,
      "grad_norm": 0.9335005283355713,
      "learning_rate": 4.813599568267674e-05,
      "loss": 3.1469,
      "step": 85350
    },
    {
      "epoch": 9.20754716981132,
      "grad_norm": 0.9518271684646606,
      "learning_rate": 4.7812196438208306e-05,
      "loss": 3.1175,
      "step": 85400
    },
    {
      "epoch": 9.212938005390836,
      "grad_norm": 0.9518353343009949,
      "learning_rate": 4.748839719373988e-05,
      "loss": 3.1383,
      "step": 85450
    },
    {
      "epoch": 9.21832884097035,
      "grad_norm": 0.9310569763183594,
      "learning_rate": 4.716459794927145e-05,
      "loss": 3.16,
      "step": 85500
    },
    {
      "epoch": 9.223719676549866,
      "grad_norm": 0.98565274477005,
      "learning_rate": 4.684079870480302e-05,
      "loss": 3.1408,
      "step": 85550
    },
    {
      "epoch": 9.22911051212938,
      "grad_norm": 0.9924389123916626,
      "learning_rate": 4.651699946033459e-05,
      "loss": 3.153,
      "step": 85600
    },
    {
      "epoch": 9.234501347708894,
      "grad_norm": 0.9255833029747009,
      "learning_rate": 4.619320021586616e-05,
      "loss": 3.1516,
      "step": 85650
    },
    {
      "epoch": 9.23989218328841,
      "grad_norm": 0.9540524482727051,
      "learning_rate": 4.5869400971397725e-05,
      "loss": 3.1385,
      "step": 85700
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.9115455150604248,
      "learning_rate": 4.55456017269293e-05,
      "loss": 3.1441,
      "step": 85750
    },
    {
      "epoch": 9.25067385444744,
      "grad_norm": 0.8959429264068604,
      "learning_rate": 4.5221802482460866e-05,
      "loss": 3.1475,
      "step": 85800
    },
    {
      "epoch": 9.256064690026955,
      "grad_norm": 0.9074015021324158,
      "learning_rate": 4.4898003237992444e-05,
      "loss": 3.144,
      "step": 85850
    },
    {
      "epoch": 9.261455525606468,
      "grad_norm": 0.954766571521759,
      "learning_rate": 4.457420399352401e-05,
      "loss": 3.1538,
      "step": 85900
    },
    {
      "epoch": 9.266846361185983,
      "grad_norm": 0.9908513426780701,
      "learning_rate": 4.425040474905558e-05,
      "loss": 3.1401,
      "step": 85950
    },
    {
      "epoch": 9.272237196765499,
      "grad_norm": 0.9178355932235718,
      "learning_rate": 4.392660550458716e-05,
      "loss": 3.1334,
      "step": 86000
    },
    {
      "epoch": 9.272237196765499,
      "eval_accuracy": 0.3929567503920468,
      "eval_loss": 3.315417766571045,
      "eval_runtime": 187.4351,
      "eval_samples_per_second": 96.092,
      "eval_steps_per_second": 6.007,
      "step": 86000
    },
    {
      "epoch": 9.277628032345014,
      "grad_norm": 0.9122679233551025,
      "learning_rate": 4.360280626011872e-05,
      "loss": 3.1163,
      "step": 86050
    },
    {
      "epoch": 9.283018867924529,
      "grad_norm": 0.9035681486129761,
      "learning_rate": 4.32790070156503e-05,
      "loss": 3.1341,
      "step": 86100
    },
    {
      "epoch": 9.288409703504042,
      "grad_norm": 0.9279851913452148,
      "learning_rate": 4.295520777118186e-05,
      "loss": 3.153,
      "step": 86150
    },
    {
      "epoch": 9.293800539083557,
      "grad_norm": 0.9015566110610962,
      "learning_rate": 4.263140852671344e-05,
      "loss": 3.1309,
      "step": 86200
    },
    {
      "epoch": 9.299191374663073,
      "grad_norm": 0.8954885601997375,
      "learning_rate": 4.2307609282245004e-05,
      "loss": 3.1584,
      "step": 86250
    },
    {
      "epoch": 9.304582210242588,
      "grad_norm": 0.9186733961105347,
      "learning_rate": 4.1983810037776575e-05,
      "loss": 3.1386,
      "step": 86300
    },
    {
      "epoch": 9.309973045822103,
      "grad_norm": 0.9036773443222046,
      "learning_rate": 4.1660010793308146e-05,
      "loss": 3.1344,
      "step": 86350
    },
    {
      "epoch": 9.315363881401618,
      "grad_norm": 0.9160073399543762,
      "learning_rate": 4.133621154883972e-05,
      "loss": 3.1418,
      "step": 86400
    },
    {
      "epoch": 9.320754716981131,
      "grad_norm": 0.9248919486999512,
      "learning_rate": 4.101241230437128e-05,
      "loss": 3.1488,
      "step": 86450
    },
    {
      "epoch": 9.326145552560646,
      "grad_norm": 0.957439124584198,
      "learning_rate": 4.068861305990286e-05,
      "loss": 3.1424,
      "step": 86500
    },
    {
      "epoch": 9.331536388140162,
      "grad_norm": 0.9187906980514526,
      "learning_rate": 4.036481381543442e-05,
      "loss": 3.1554,
      "step": 86550
    },
    {
      "epoch": 9.336927223719677,
      "grad_norm": 0.9415502548217773,
      "learning_rate": 4.0041014570966e-05,
      "loss": 3.143,
      "step": 86600
    },
    {
      "epoch": 9.342318059299192,
      "grad_norm": 0.9612598419189453,
      "learning_rate": 3.9717215326497565e-05,
      "loss": 3.1304,
      "step": 86650
    },
    {
      "epoch": 9.347708894878707,
      "grad_norm": 0.8706883788108826,
      "learning_rate": 3.9393416082029136e-05,
      "loss": 3.134,
      "step": 86700
    },
    {
      "epoch": 9.35309973045822,
      "grad_norm": 0.9136846661567688,
      "learning_rate": 3.9069616837560706e-05,
      "loss": 3.1408,
      "step": 86750
    },
    {
      "epoch": 9.358490566037736,
      "grad_norm": 0.9092503786087036,
      "learning_rate": 3.874581759309228e-05,
      "loss": 3.153,
      "step": 86800
    },
    {
      "epoch": 9.36388140161725,
      "grad_norm": 0.9336559176445007,
      "learning_rate": 3.8422018348623855e-05,
      "loss": 3.1576,
      "step": 86850
    },
    {
      "epoch": 9.369272237196766,
      "grad_norm": 0.9446649551391602,
      "learning_rate": 3.809821910415542e-05,
      "loss": 3.1343,
      "step": 86900
    },
    {
      "epoch": 9.374663072776281,
      "grad_norm": 0.9409492611885071,
      "learning_rate": 3.7774419859687e-05,
      "loss": 3.1445,
      "step": 86950
    },
    {
      "epoch": 9.380053908355794,
      "grad_norm": 0.9114211201667786,
      "learning_rate": 3.745062061521856e-05,
      "loss": 3.1351,
      "step": 87000
    },
    {
      "epoch": 9.380053908355794,
      "eval_accuracy": 0.3931550419123082,
      "eval_loss": 3.313524007797241,
      "eval_runtime": 187.7613,
      "eval_samples_per_second": 95.925,
      "eval_steps_per_second": 5.997,
      "step": 87000
    },
    {
      "epoch": 9.38544474393531,
      "grad_norm": 0.9090732932090759,
      "learning_rate": 3.712682137075013e-05,
      "loss": 3.1364,
      "step": 87050
    },
    {
      "epoch": 9.390835579514825,
      "grad_norm": 0.9086277484893799,
      "learning_rate": 3.68030221262817e-05,
      "loss": 3.1433,
      "step": 87100
    },
    {
      "epoch": 9.39622641509434,
      "grad_norm": 0.957598090171814,
      "learning_rate": 3.648569886670264e-05,
      "loss": 3.1458,
      "step": 87150
    },
    {
      "epoch": 9.401617250673855,
      "grad_norm": 0.9030936360359192,
      "learning_rate": 3.616189962223421e-05,
      "loss": 3.1413,
      "step": 87200
    },
    {
      "epoch": 9.40700808625337,
      "grad_norm": 0.8740924000740051,
      "learning_rate": 3.583810037776578e-05,
      "loss": 3.1287,
      "step": 87250
    },
    {
      "epoch": 9.412398921832883,
      "grad_norm": 0.9102224707603455,
      "learning_rate": 3.5514301133297354e-05,
      "loss": 3.1572,
      "step": 87300
    },
    {
      "epoch": 9.417789757412399,
      "grad_norm": 0.9368909001350403,
      "learning_rate": 3.5190501888828925e-05,
      "loss": 3.1524,
      "step": 87350
    },
    {
      "epoch": 9.423180592991914,
      "grad_norm": 0.9046692848205566,
      "learning_rate": 3.4866702644360496e-05,
      "loss": 3.1581,
      "step": 87400
    },
    {
      "epoch": 9.428571428571429,
      "grad_norm": 0.9296504259109497,
      "learning_rate": 3.4542903399892067e-05,
      "loss": 3.148,
      "step": 87450
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.9260674118995667,
      "learning_rate": 3.421910415542363e-05,
      "loss": 3.1497,
      "step": 87500
    },
    {
      "epoch": 9.439353099730457,
      "grad_norm": 0.9342174530029297,
      "learning_rate": 3.38953049109552e-05,
      "loss": 3.1412,
      "step": 87550
    },
    {
      "epoch": 9.444743935309972,
      "grad_norm": 0.9455352425575256,
      "learning_rate": 3.357150566648677e-05,
      "loss": 3.126,
      "step": 87600
    },
    {
      "epoch": 9.450134770889488,
      "grad_norm": 0.9465023279190063,
      "learning_rate": 3.324770642201834e-05,
      "loss": 3.134,
      "step": 87650
    },
    {
      "epoch": 9.455525606469003,
      "grad_norm": 0.8920009136199951,
      "learning_rate": 3.2923907177549914e-05,
      "loss": 3.1474,
      "step": 87700
    },
    {
      "epoch": 9.460916442048518,
      "grad_norm": 0.9538445472717285,
      "learning_rate": 3.2600107933081485e-05,
      "loss": 3.1486,
      "step": 87750
    },
    {
      "epoch": 9.466307277628033,
      "grad_norm": 0.889605700969696,
      "learning_rate": 3.2276308688613056e-05,
      "loss": 3.1442,
      "step": 87800
    },
    {
      "epoch": 9.471698113207546,
      "grad_norm": 0.9327829480171204,
      "learning_rate": 3.195250944414463e-05,
      "loss": 3.1462,
      "step": 87850
    },
    {
      "epoch": 9.477088948787062,
      "grad_norm": 0.9338760375976562,
      "learning_rate": 3.16287101996762e-05,
      "loss": 3.1455,
      "step": 87900
    },
    {
      "epoch": 9.482479784366577,
      "grad_norm": 0.9253023862838745,
      "learning_rate": 3.130491095520777e-05,
      "loss": 3.141,
      "step": 87950
    },
    {
      "epoch": 9.487870619946092,
      "grad_norm": 0.922599196434021,
      "learning_rate": 3.098111171073934e-05,
      "loss": 3.1321,
      "step": 88000
    },
    {
      "epoch": 9.487870619946092,
      "eval_accuracy": 0.3934250443385272,
      "eval_loss": 3.311544895172119,
      "eval_runtime": 187.9489,
      "eval_samples_per_second": 95.829,
      "eval_steps_per_second": 5.991,
      "step": 88000
    },
    {
      "epoch": 9.493261455525607,
      "grad_norm": 0.9280849099159241,
      "learning_rate": 3.065731246627091e-05,
      "loss": 3.1346,
      "step": 88050
    },
    {
      "epoch": 9.498652291105122,
      "grad_norm": 0.9111308455467224,
      "learning_rate": 3.033351322180248e-05,
      "loss": 3.1326,
      "step": 88100
    },
    {
      "epoch": 9.504043126684635,
      "grad_norm": 0.878240704536438,
      "learning_rate": 3.000971397733405e-05,
      "loss": 3.1318,
      "step": 88150
    },
    {
      "epoch": 9.50943396226415,
      "grad_norm": 0.9397717118263245,
      "learning_rate": 2.968591473286562e-05,
      "loss": 3.1424,
      "step": 88200
    },
    {
      "epoch": 9.514824797843666,
      "grad_norm": 0.9439541697502136,
      "learning_rate": 2.936211548839719e-05,
      "loss": 3.1412,
      "step": 88250
    },
    {
      "epoch": 9.520215633423181,
      "grad_norm": 0.9720891714096069,
      "learning_rate": 2.903831624392876e-05,
      "loss": 3.1292,
      "step": 88300
    },
    {
      "epoch": 9.525606469002696,
      "grad_norm": 0.9237875938415527,
      "learning_rate": 2.8714516999460332e-05,
      "loss": 3.1326,
      "step": 88350
    },
    {
      "epoch": 9.530997304582211,
      "grad_norm": 0.9764357209205627,
      "learning_rate": 2.83907177549919e-05,
      "loss": 3.1372,
      "step": 88400
    },
    {
      "epoch": 9.536388140161725,
      "grad_norm": 0.942274272441864,
      "learning_rate": 2.806691851052347e-05,
      "loss": 3.1463,
      "step": 88450
    },
    {
      "epoch": 9.54177897574124,
      "grad_norm": 0.9024509787559509,
      "learning_rate": 2.774311926605504e-05,
      "loss": 3.1422,
      "step": 88500
    },
    {
      "epoch": 9.547169811320755,
      "grad_norm": 0.923398494720459,
      "learning_rate": 2.7419320021586613e-05,
      "loss": 3.1545,
      "step": 88550
    },
    {
      "epoch": 9.55256064690027,
      "grad_norm": 0.9231415390968323,
      "learning_rate": 2.7095520777118187e-05,
      "loss": 3.1488,
      "step": 88600
    },
    {
      "epoch": 9.557951482479785,
      "grad_norm": 0.9691388010978699,
      "learning_rate": 2.6771721532649758e-05,
      "loss": 3.1497,
      "step": 88650
    },
    {
      "epoch": 9.563342318059298,
      "grad_norm": 0.9140695929527283,
      "learning_rate": 2.6447922288181325e-05,
      "loss": 3.143,
      "step": 88700
    },
    {
      "epoch": 9.568733153638814,
      "grad_norm": 0.9069567322731018,
      "learning_rate": 2.6124123043712896e-05,
      "loss": 3.1387,
      "step": 88750
    },
    {
      "epoch": 9.574123989218329,
      "grad_norm": 0.9632625579833984,
      "learning_rate": 2.5800323799244467e-05,
      "loss": 3.13,
      "step": 88800
    },
    {
      "epoch": 9.579514824797844,
      "grad_norm": 0.8916159868240356,
      "learning_rate": 2.5476524554776038e-05,
      "loss": 3.15,
      "step": 88850
    },
    {
      "epoch": 9.584905660377359,
      "grad_norm": 0.9191704392433167,
      "learning_rate": 2.515272531030761e-05,
      "loss": 3.1395,
      "step": 88900
    },
    {
      "epoch": 9.590296495956874,
      "grad_norm": 0.8881397843360901,
      "learning_rate": 2.4828926065839176e-05,
      "loss": 3.123,
      "step": 88950
    },
    {
      "epoch": 9.595687331536388,
      "grad_norm": 0.9439510703086853,
      "learning_rate": 2.4505126821370747e-05,
      "loss": 3.1353,
      "step": 89000
    },
    {
      "epoch": 9.595687331536388,
      "eval_accuracy": 0.39363930783329737,
      "eval_loss": 3.308457136154175,
      "eval_runtime": 187.5014,
      "eval_samples_per_second": 96.058,
      "eval_steps_per_second": 6.005,
      "step": 89000
    },
    {
      "epoch": 9.601078167115903,
      "grad_norm": 0.9280842542648315,
      "learning_rate": 2.4181327576902318e-05,
      "loss": 3.1481,
      "step": 89050
    },
    {
      "epoch": 9.606469002695418,
      "grad_norm": 0.9251762628555298,
      "learning_rate": 2.385752833243389e-05,
      "loss": 3.1365,
      "step": 89100
    },
    {
      "epoch": 9.611859838274933,
      "grad_norm": 0.9192915558815002,
      "learning_rate": 2.353372908796546e-05,
      "loss": 3.1542,
      "step": 89150
    },
    {
      "epoch": 9.617250673854448,
      "grad_norm": 0.9379941821098328,
      "learning_rate": 2.3209929843497027e-05,
      "loss": 3.1426,
      "step": 89200
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.9859703779220581,
      "learning_rate": 2.289260658391797e-05,
      "loss": 3.1537,
      "step": 89250
    },
    {
      "epoch": 9.628032345013477,
      "grad_norm": 0.9222863912582397,
      "learning_rate": 2.256880733944954e-05,
      "loss": 3.1382,
      "step": 89300
    },
    {
      "epoch": 9.633423180592992,
      "grad_norm": 0.9080533981323242,
      "learning_rate": 2.224500809498111e-05,
      "loss": 3.1555,
      "step": 89350
    },
    {
      "epoch": 9.638814016172507,
      "grad_norm": 0.926685631275177,
      "learning_rate": 2.1921208850512682e-05,
      "loss": 3.1466,
      "step": 89400
    },
    {
      "epoch": 9.644204851752022,
      "grad_norm": 0.9160860180854797,
      "learning_rate": 2.1597409606044253e-05,
      "loss": 3.1422,
      "step": 89450
    },
    {
      "epoch": 9.649595687331537,
      "grad_norm": 0.9045087695121765,
      "learning_rate": 2.127361036157582e-05,
      "loss": 3.1423,
      "step": 89500
    },
    {
      "epoch": 9.65498652291105,
      "grad_norm": 0.9281613826751709,
      "learning_rate": 2.094981111710739e-05,
      "loss": 3.1433,
      "step": 89550
    },
    {
      "epoch": 9.660377358490566,
      "grad_norm": 0.9077588319778442,
      "learning_rate": 2.0626011872638962e-05,
      "loss": 3.1219,
      "step": 89600
    },
    {
      "epoch": 9.66576819407008,
      "grad_norm": 0.9211140871047974,
      "learning_rate": 2.0302212628170533e-05,
      "loss": 3.1379,
      "step": 89650
    },
    {
      "epoch": 9.671159029649596,
      "grad_norm": 0.9394590854644775,
      "learning_rate": 1.99784133837021e-05,
      "loss": 3.1425,
      "step": 89700
    },
    {
      "epoch": 9.676549865229111,
      "grad_norm": 0.9185537099838257,
      "learning_rate": 1.965461413923367e-05,
      "loss": 3.1399,
      "step": 89750
    },
    {
      "epoch": 9.681940700808624,
      "grad_norm": 0.8864555358886719,
      "learning_rate": 1.9330814894765242e-05,
      "loss": 3.1365,
      "step": 89800
    },
    {
      "epoch": 9.68733153638814,
      "grad_norm": 0.952918529510498,
      "learning_rate": 1.9007015650296813e-05,
      "loss": 3.1575,
      "step": 89850
    },
    {
      "epoch": 9.692722371967655,
      "grad_norm": 0.9361217617988586,
      "learning_rate": 1.8683216405828384e-05,
      "loss": 3.1383,
      "step": 89900
    },
    {
      "epoch": 9.69811320754717,
      "grad_norm": 0.9423272013664246,
      "learning_rate": 1.8359417161359955e-05,
      "loss": 3.1508,
      "step": 89950
    },
    {
      "epoch": 9.703504043126685,
      "grad_norm": 0.9519898295402527,
      "learning_rate": 1.8035617916891526e-05,
      "loss": 3.1414,
      "step": 90000
    },
    {
      "epoch": 9.703504043126685,
      "eval_accuracy": 0.3940164420069014,
      "eval_loss": 3.3063955307006836,
      "eval_runtime": 187.5748,
      "eval_samples_per_second": 96.02,
      "eval_steps_per_second": 6.003,
      "step": 90000
    },
    {
      "epoch": 9.7088948787062,
      "grad_norm": 0.9187700748443604,
      "learning_rate": 1.7711818672423097e-05,
      "loss": 3.1495,
      "step": 90050
    },
    {
      "epoch": 9.714285714285714,
      "grad_norm": 0.9302745461463928,
      "learning_rate": 1.7388019427954664e-05,
      "loss": 3.1235,
      "step": 90100
    },
    {
      "epoch": 9.719676549865229,
      "grad_norm": 0.9535933136940002,
      "learning_rate": 1.706422018348624e-05,
      "loss": 3.1415,
      "step": 90150
    },
    {
      "epoch": 9.725067385444744,
      "grad_norm": 0.8942263722419739,
      "learning_rate": 1.674042093901781e-05,
      "loss": 3.1423,
      "step": 90200
    },
    {
      "epoch": 9.730458221024259,
      "grad_norm": 0.9255458116531372,
      "learning_rate": 1.6416621694549377e-05,
      "loss": 3.1483,
      "step": 90250
    },
    {
      "epoch": 9.735849056603774,
      "grad_norm": 0.9040615558624268,
      "learning_rate": 1.6092822450080948e-05,
      "loss": 3.1325,
      "step": 90300
    },
    {
      "epoch": 9.74123989218329,
      "grad_norm": 0.9275330901145935,
      "learning_rate": 1.576902320561252e-05,
      "loss": 3.1358,
      "step": 90350
    },
    {
      "epoch": 9.746630727762803,
      "grad_norm": 0.9373171925544739,
      "learning_rate": 1.544522396114409e-05,
      "loss": 3.1712,
      "step": 90400
    },
    {
      "epoch": 9.752021563342318,
      "grad_norm": 0.9148626923561096,
      "learning_rate": 1.5121424716675659e-05,
      "loss": 3.1346,
      "step": 90450
    },
    {
      "epoch": 9.757412398921833,
      "grad_norm": 0.9298881888389587,
      "learning_rate": 1.479762547220723e-05,
      "loss": 3.1401,
      "step": 90500
    },
    {
      "epoch": 9.762803234501348,
      "grad_norm": 0.9553585052490234,
      "learning_rate": 1.4473826227738802e-05,
      "loss": 3.1409,
      "step": 90550
    },
    {
      "epoch": 9.768194070080863,
      "grad_norm": 0.8911868333816528,
      "learning_rate": 1.4150026983270371e-05,
      "loss": 3.1391,
      "step": 90600
    },
    {
      "epoch": 9.773584905660378,
      "grad_norm": 0.9505174160003662,
      "learning_rate": 1.3826227738801942e-05,
      "loss": 3.1435,
      "step": 90650
    },
    {
      "epoch": 9.778975741239892,
      "grad_norm": 0.8940786123275757,
      "learning_rate": 1.3502428494333512e-05,
      "loss": 3.1435,
      "step": 90700
    },
    {
      "epoch": 9.784366576819407,
      "grad_norm": 0.9654335379600525,
      "learning_rate": 1.3178629249865082e-05,
      "loss": 3.136,
      "step": 90750
    },
    {
      "epoch": 9.789757412398922,
      "grad_norm": 0.9177462458610535,
      "learning_rate": 1.2854830005396653e-05,
      "loss": 3.1599,
      "step": 90800
    },
    {
      "epoch": 9.795148247978437,
      "grad_norm": 0.8937206268310547,
      "learning_rate": 1.2531030760928222e-05,
      "loss": 3.1504,
      "step": 90850
    },
    {
      "epoch": 9.800539083557952,
      "grad_norm": 0.9319292902946472,
      "learning_rate": 1.2207231516459793e-05,
      "loss": 3.1356,
      "step": 90900
    },
    {
      "epoch": 9.805929919137466,
      "grad_norm": 0.9177371859550476,
      "learning_rate": 1.1883432271991366e-05,
      "loss": 3.1507,
      "step": 90950
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.9243341088294983,
      "learning_rate": 1.1559633027522935e-05,
      "loss": 3.1511,
      "step": 91000
    },
    {
      "epoch": 9.81132075471698,
      "eval_accuracy": 0.3941807251732769,
      "eval_loss": 3.304288148880005,
      "eval_runtime": 187.6044,
      "eval_samples_per_second": 96.005,
      "eval_steps_per_second": 6.002,
      "step": 91000
    },
    {
      "epoch": 9.816711590296496,
      "grad_norm": 0.9366071224212646,
      "learning_rate": 1.1235833783054506e-05,
      "loss": 3.1432,
      "step": 91050
    },
    {
      "epoch": 9.822102425876011,
      "grad_norm": 0.8996028900146484,
      "learning_rate": 1.0912034538586075e-05,
      "loss": 3.1392,
      "step": 91100
    },
    {
      "epoch": 9.827493261455526,
      "grad_norm": 0.9392869472503662,
      "learning_rate": 1.0588235294117646e-05,
      "loss": 3.1376,
      "step": 91150
    },
    {
      "epoch": 9.832884097035041,
      "grad_norm": 0.9204728007316589,
      "learning_rate": 1.0264436049649217e-05,
      "loss": 3.1242,
      "step": 91200
    },
    {
      "epoch": 9.838274932614555,
      "grad_norm": 0.9263073801994324,
      "learning_rate": 9.940636805180786e-06,
      "loss": 3.13,
      "step": 91250
    },
    {
      "epoch": 9.84366576819407,
      "grad_norm": 0.9406017661094666,
      "learning_rate": 9.616837560712357e-06,
      "loss": 3.125,
      "step": 91300
    },
    {
      "epoch": 9.849056603773585,
      "grad_norm": 0.9088371992111206,
      "learning_rate": 9.293038316243928e-06,
      "loss": 3.1387,
      "step": 91350
    },
    {
      "epoch": 9.8544474393531,
      "grad_norm": 0.9465605616569519,
      "learning_rate": 8.975715056664867e-06,
      "loss": 3.1628,
      "step": 91400
    },
    {
      "epoch": 9.859838274932615,
      "grad_norm": 0.9424158334732056,
      "learning_rate": 8.651915812196437e-06,
      "loss": 3.1449,
      "step": 91450
    },
    {
      "epoch": 9.865229110512129,
      "grad_norm": 0.9331150054931641,
      "learning_rate": 8.328116567728008e-06,
      "loss": 3.1452,
      "step": 91500
    },
    {
      "epoch": 9.870619946091644,
      "grad_norm": 0.8824761509895325,
      "learning_rate": 8.004317323259577e-06,
      "loss": 3.1565,
      "step": 91550
    },
    {
      "epoch": 9.876010781671159,
      "grad_norm": 0.8807875514030457,
      "learning_rate": 7.680518078791148e-06,
      "loss": 3.1295,
      "step": 91600
    },
    {
      "epoch": 9.881401617250674,
      "grad_norm": 0.9300699830055237,
      "learning_rate": 7.356718834322719e-06,
      "loss": 3.1535,
      "step": 91650
    },
    {
      "epoch": 9.88679245283019,
      "grad_norm": 0.9522417783737183,
      "learning_rate": 7.03291958985429e-06,
      "loss": 3.1476,
      "step": 91700
    },
    {
      "epoch": 9.892183288409704,
      "grad_norm": 0.9019232392311096,
      "learning_rate": 6.70912034538586e-06,
      "loss": 3.1431,
      "step": 91750
    },
    {
      "epoch": 9.897574123989218,
      "grad_norm": 0.91133052110672,
      "learning_rate": 6.38532110091743e-06,
      "loss": 3.1496,
      "step": 91800
    },
    {
      "epoch": 9.902964959568733,
      "grad_norm": 0.9428941011428833,
      "learning_rate": 6.061521856449001e-06,
      "loss": 3.1511,
      "step": 91850
    },
    {
      "epoch": 9.908355795148248,
      "grad_norm": 0.9570835828781128,
      "learning_rate": 5.737722611980571e-06,
      "loss": 3.145,
      "step": 91900
    },
    {
      "epoch": 9.913746630727763,
      "grad_norm": 0.907947838306427,
      "learning_rate": 5.413923367512142e-06,
      "loss": 3.1555,
      "step": 91950
    },
    {
      "epoch": 9.919137466307278,
      "grad_norm": 0.8969404101371765,
      "learning_rate": 5.090124123043712e-06,
      "loss": 3.1522,
      "step": 92000
    },
    {
      "epoch": 9.919137466307278,
      "eval_accuracy": 0.3943509842484822,
      "eval_loss": 3.303070068359375,
      "eval_runtime": 188.1629,
      "eval_samples_per_second": 95.72,
      "eval_steps_per_second": 5.984,
      "step": 92000
    },
    {
      "epoch": 9.924528301886792,
      "grad_norm": 0.9321413040161133,
      "learning_rate": 4.766324878575283e-06,
      "loss": 3.1493,
      "step": 92050
    },
    {
      "epoch": 9.929919137466307,
      "grad_norm": 0.8739159107208252,
      "learning_rate": 4.442525634106853e-06,
      "loss": 3.1489,
      "step": 92100
    },
    {
      "epoch": 9.935309973045822,
      "grad_norm": 0.8782985210418701,
      "learning_rate": 4.118726389638424e-06,
      "loss": 3.1392,
      "step": 92150
    },
    {
      "epoch": 9.940700808625337,
      "grad_norm": 0.905799925327301,
      "learning_rate": 3.7949271451699944e-06,
      "loss": 3.1412,
      "step": 92200
    },
    {
      "epoch": 9.946091644204852,
      "grad_norm": 0.933990478515625,
      "learning_rate": 3.4711279007015644e-06,
      "loss": 3.1318,
      "step": 92250
    },
    {
      "epoch": 9.951482479784367,
      "grad_norm": 0.951168417930603,
      "learning_rate": 3.1473286562331353e-06,
      "loss": 3.1352,
      "step": 92300
    },
    {
      "epoch": 9.95687331536388,
      "grad_norm": 0.935908854007721,
      "learning_rate": 2.8235294117647054e-06,
      "loss": 3.1434,
      "step": 92350
    },
    {
      "epoch": 9.962264150943396,
      "grad_norm": 0.8934540748596191,
      "learning_rate": 2.4997301672962763e-06,
      "loss": 3.1592,
      "step": 92400
    },
    {
      "epoch": 9.967654986522911,
      "grad_norm": 0.9462904930114746,
      "learning_rate": 2.1759309228278467e-06,
      "loss": 3.1313,
      "step": 92450
    },
    {
      "epoch": 9.973045822102426,
      "grad_norm": 0.9313557147979736,
      "learning_rate": 1.852131678359417e-06,
      "loss": 3.132,
      "step": 92500
    },
    {
      "epoch": 9.978436657681941,
      "grad_norm": 0.9732421636581421,
      "learning_rate": 1.5283324338909875e-06,
      "loss": 3.1402,
      "step": 92550
    },
    {
      "epoch": 9.983827493261456,
      "grad_norm": 0.8972058892250061,
      "learning_rate": 1.204533189422558e-06,
      "loss": 3.1391,
      "step": 92600
    },
    {
      "epoch": 9.98921832884097,
      "grad_norm": 0.9335411190986633,
      "learning_rate": 8.807339449541284e-07,
      "loss": 3.1346,
      "step": 92650
    },
    {
      "epoch": 9.994609164420485,
      "grad_norm": 0.9858599305152893,
      "learning_rate": 5.569347004856989e-07,
      "loss": 3.1294,
      "step": 92700
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.9937326908111572,
      "learning_rate": 2.3313545601726927e-07,
      "loss": 3.139,
      "step": 92750
    },
    {
      "epoch": 10.0,
      "step": 92750,
      "total_flos": 7.75449427968e+17,
      "train_loss": 3.467074710979616,
      "train_runtime": 80468.1749,
      "train_samples_per_second": 36.881,
      "train_steps_per_second": 1.153
    }
  ],
  "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
}