{
  "best_metric": 3.3018898963928223,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_2000_8397/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": 1.5813435316085815,
      "learning_rate": 0.0003,
      "loss": 8.6318,
      "step": 50
    },
    {
      "epoch": 0.01078167115902965,
      "grad_norm": 1.0082982778549194,
      "learning_rate": 0.0006,
      "loss": 6.8697,
      "step": 100
    },
    {
      "epoch": 0.016172506738544475,
      "grad_norm": 2.0049550533294678,
      "learning_rate": 0.0005996762007555315,
      "loss": 6.4573,
      "step": 150
    },
    {
      "epoch": 0.0215633423180593,
      "grad_norm": 2.044867515563965,
      "learning_rate": 0.000599352401511063,
      "loss": 6.239,
      "step": 200
    },
    {
      "epoch": 0.026954177897574125,
      "grad_norm": 1.7265796661376953,
      "learning_rate": 0.0005990286022665946,
      "loss": 6.0844,
      "step": 250
    },
    {
      "epoch": 0.03234501347708895,
      "grad_norm": 3.774665594100952,
      "learning_rate": 0.0005987048030221263,
      "loss": 5.9817,
      "step": 300
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 1.1093494892120361,
      "learning_rate": 0.0005983810037776578,
      "loss": 5.8816,
      "step": 350
    },
    {
      "epoch": 0.0431266846361186,
      "grad_norm": 1.401320457458496,
      "learning_rate": 0.0005980572045331894,
      "loss": 5.8369,
      "step": 400
    },
    {
      "epoch": 0.04851752021563342,
      "grad_norm": 1.4539614915847778,
      "learning_rate": 0.0005977334052887209,
      "loss": 5.756,
      "step": 450
    },
    {
      "epoch": 0.05390835579514825,
      "grad_norm": 1.237868309020996,
      "learning_rate": 0.0005974096060442526,
      "loss": 5.6613,
      "step": 500
    },
    {
      "epoch": 0.05929919137466307,
      "grad_norm": 1.7472739219665527,
      "learning_rate": 0.0005970858067997841,
      "loss": 5.5746,
      "step": 550
    },
    {
      "epoch": 0.0646900269541779,
      "grad_norm": 1.1309614181518555,
      "learning_rate": 0.0005967620075553157,
      "loss": 5.5044,
      "step": 600
    },
    {
      "epoch": 0.07008086253369272,
      "grad_norm": 1.2399049997329712,
      "learning_rate": 0.0005964382083108472,
      "loss": 5.4345,
      "step": 650
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 1.7309972047805786,
      "learning_rate": 0.0005961144090663788,
      "loss": 5.386,
      "step": 700
    },
    {
      "epoch": 0.08086253369272237,
      "grad_norm": 1.1939959526062012,
      "learning_rate": 0.0005957906098219104,
      "loss": 5.3088,
      "step": 750
    },
    {
      "epoch": 0.0862533692722372,
      "grad_norm": 1.6205377578735352,
      "learning_rate": 0.0005954668105774419,
      "loss": 5.2622,
      "step": 800
    },
    {
      "epoch": 0.09164420485175202,
      "grad_norm": 1.3945573568344116,
      "learning_rate": 0.0005951430113329735,
      "loss": 5.2251,
      "step": 850
    },
    {
      "epoch": 0.09703504043126684,
      "grad_norm": 1.3046308755874634,
      "learning_rate": 0.0005948192120885051,
      "loss": 5.1654,
      "step": 900
    },
    {
      "epoch": 0.10242587601078167,
      "grad_norm": 1.2131174802780151,
      "learning_rate": 0.0005944954128440366,
      "loss": 5.1448,
      "step": 950
    },
    {
      "epoch": 0.1078167115902965,
      "grad_norm": 0.9675126075744629,
      "learning_rate": 0.0005941716135995682,
      "loss": 5.1035,
      "step": 1000
    },
    {
      "epoch": 0.1078167115902965,
      "eval_accuracy": 0.22749035406825205,
      "eval_loss": 5.018069267272949,
      "eval_runtime": 184.6816,
      "eval_samples_per_second": 97.525,
      "eval_steps_per_second": 6.097,
      "step": 1000
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 1.0149263143539429,
      "learning_rate": 0.0005938478143550997,
      "loss": 5.0486,
      "step": 1050
    },
    {
      "epoch": 0.11859838274932614,
      "grad_norm": 1.3123433589935303,
      "learning_rate": 0.0005935240151106314,
      "loss": 5.0207,
      "step": 1100
    },
    {
      "epoch": 0.12398921832884097,
      "grad_norm": 0.9050843715667725,
      "learning_rate": 0.0005932002158661629,
      "loss": 4.9928,
      "step": 1150
    },
    {
      "epoch": 0.1293800539083558,
      "grad_norm": 1.274807095527649,
      "learning_rate": 0.0005928764166216945,
      "loss": 4.9535,
      "step": 1200
    },
    {
      "epoch": 0.1347708894878706,
      "grad_norm": 1.00777006149292,
      "learning_rate": 0.000592552617377226,
      "loss": 4.9137,
      "step": 1250
    },
    {
      "epoch": 0.14016172506738545,
      "grad_norm": 0.8887598514556885,
      "learning_rate": 0.0005922288181327577,
      "loss": 4.8818,
      "step": 1300
    },
    {
      "epoch": 0.14555256064690028,
      "grad_norm": 0.9829823970794678,
      "learning_rate": 0.0005919050188882893,
      "loss": 4.8648,
      "step": 1350
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 0.9974054098129272,
      "learning_rate": 0.0005915812196438207,
      "loss": 4.8498,
      "step": 1400
    },
    {
      "epoch": 0.15633423180592992,
      "grad_norm": 1.0419261455535889,
      "learning_rate": 0.0005912574203993524,
      "loss": 4.8005,
      "step": 1450
    },
    {
      "epoch": 0.16172506738544473,
      "grad_norm": 1.1737334728240967,
      "learning_rate": 0.0005909336211548839,
      "loss": 4.8138,
      "step": 1500
    },
    {
      "epoch": 0.16711590296495957,
      "grad_norm": 1.0961012840270996,
      "learning_rate": 0.0005906098219104155,
      "loss": 4.7659,
      "step": 1550
    },
    {
      "epoch": 0.1725067385444744,
      "grad_norm": 1.1323438882827759,
      "learning_rate": 0.000590286022665947,
      "loss": 4.7663,
      "step": 1600
    },
    {
      "epoch": 0.1778975741239892,
      "grad_norm": 0.8736843466758728,
      "learning_rate": 0.0005899622234214787,
      "loss": 4.7157,
      "step": 1650
    },
    {
      "epoch": 0.18328840970350405,
      "grad_norm": 0.7891756892204285,
      "learning_rate": 0.0005896384241770102,
      "loss": 4.7099,
      "step": 1700
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 1.0397777557373047,
      "learning_rate": 0.0005893146249325418,
      "loss": 4.6729,
      "step": 1750
    },
    {
      "epoch": 0.1940700808625337,
      "grad_norm": 1.151475429534912,
      "learning_rate": 0.0005889908256880733,
      "loss": 4.6804,
      "step": 1800
    },
    {
      "epoch": 0.19946091644204852,
      "grad_norm": 1.5584900379180908,
      "learning_rate": 0.0005886670264436049,
      "loss": 4.6534,
      "step": 1850
    },
    {
      "epoch": 0.20485175202156333,
      "grad_norm": 0.7592549920082092,
      "learning_rate": 0.0005883432271991365,
      "loss": 4.6368,
      "step": 1900
    },
    {
      "epoch": 0.21024258760107817,
      "grad_norm": 1.132956862449646,
      "learning_rate": 0.0005880194279546681,
      "loss": 4.6073,
      "step": 1950
    },
    {
      "epoch": 0.215633423180593,
      "grad_norm": 0.904139518737793,
      "learning_rate": 0.0005876956287101996,
      "loss": 4.5922,
      "step": 2000
    },
    {
      "epoch": 0.215633423180593,
      "eval_accuracy": 0.26994136329603313,
      "eval_loss": 4.514337062835693,
      "eval_runtime": 183.9245,
      "eval_samples_per_second": 97.926,
      "eval_steps_per_second": 6.122,
      "step": 2000
    },
    {
      "epoch": 0.2210242587601078,
      "grad_norm": 0.8630588054656982,
      "learning_rate": 0.0005873718294657312,
      "loss": 4.5736,
      "step": 2050
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 0.751631498336792,
      "learning_rate": 0.0005870480302212628,
      "loss": 4.5577,
      "step": 2100
    },
    {
      "epoch": 0.23180592991913745,
      "grad_norm": 0.9096190333366394,
      "learning_rate": 0.0005867242309767943,
      "loss": 4.5327,
      "step": 2150
    },
    {
      "epoch": 0.2371967654986523,
      "grad_norm": 0.9275792241096497,
      "learning_rate": 0.0005864004317323259,
      "loss": 4.513,
      "step": 2200
    },
    {
      "epoch": 0.24258760107816713,
      "grad_norm": 0.8501319885253906,
      "learning_rate": 0.0005860766324878575,
      "loss": 4.4957,
      "step": 2250
    },
    {
      "epoch": 0.24797843665768193,
      "grad_norm": 0.8335550427436829,
      "learning_rate": 0.000585752833243389,
      "loss": 4.505,
      "step": 2300
    },
    {
      "epoch": 0.25336927223719674,
      "grad_norm": 0.7996338605880737,
      "learning_rate": 0.0005854290339989206,
      "loss": 4.4554,
      "step": 2350
    },
    {
      "epoch": 0.2587601078167116,
      "grad_norm": 1.0249171257019043,
      "learning_rate": 0.0005851052347544521,
      "loss": 4.4629,
      "step": 2400
    },
    {
      "epoch": 0.2641509433962264,
      "grad_norm": 0.9536342620849609,
      "learning_rate": 0.0005847814355099838,
      "loss": 4.4412,
      "step": 2450
    },
    {
      "epoch": 0.2695417789757412,
      "grad_norm": 1.2684301137924194,
      "learning_rate": 0.0005844576362655154,
      "loss": 4.4274,
      "step": 2500
    },
    {
      "epoch": 0.2749326145552561,
      "grad_norm": 0.8013315796852112,
      "learning_rate": 0.0005841338370210469,
      "loss": 4.4099,
      "step": 2550
    },
    {
      "epoch": 0.2803234501347709,
      "grad_norm": 0.8502827882766724,
      "learning_rate": 0.0005838100377765785,
      "loss": 4.3802,
      "step": 2600
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.8461282849311829,
      "learning_rate": 0.0005834862385321101,
      "loss": 4.3947,
      "step": 2650
    },
    {
      "epoch": 0.29110512129380056,
      "grad_norm": 1.0242016315460205,
      "learning_rate": 0.0005831624392876417,
      "loss": 4.3976,
      "step": 2700
    },
    {
      "epoch": 0.29649595687331537,
      "grad_norm": 0.8631992936134338,
      "learning_rate": 0.0005828386400431731,
      "loss": 4.3521,
      "step": 2750
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 0.7447966933250427,
      "learning_rate": 0.0005825148407987048,
      "loss": 4.3323,
      "step": 2800
    },
    {
      "epoch": 0.30727762803234504,
      "grad_norm": 0.7234294414520264,
      "learning_rate": 0.0005821910415542363,
      "loss": 4.3364,
      "step": 2850
    },
    {
      "epoch": 0.31266846361185985,
      "grad_norm": 0.8805587887763977,
      "learning_rate": 0.0005818672423097679,
      "loss": 4.3307,
      "step": 2900
    },
    {
      "epoch": 0.31805929919137466,
      "grad_norm": 0.8215044140815735,
      "learning_rate": 0.0005815434430652994,
      "loss": 4.3141,
      "step": 2950
    },
    {
      "epoch": 0.32345013477088946,
      "grad_norm": 0.7359305620193481,
      "learning_rate": 0.0005812196438208311,
      "loss": 4.311,
      "step": 3000
    },
    {
      "epoch": 0.32345013477088946,
      "eval_accuracy": 0.298171882566655,
      "eval_loss": 4.2451090812683105,
      "eval_runtime": 183.7967,
      "eval_samples_per_second": 97.994,
      "eval_steps_per_second": 6.126,
      "step": 3000
    },
    {
      "epoch": 0.3288409703504043,
      "grad_norm": 0.770511269569397,
      "learning_rate": 0.0005808958445763626,
      "loss": 4.3027,
      "step": 3050
    },
    {
      "epoch": 0.33423180592991913,
      "grad_norm": 0.677880048751831,
      "learning_rate": 0.0005805720453318942,
      "loss": 4.3083,
      "step": 3100
    },
    {
      "epoch": 0.33962264150943394,
      "grad_norm": 0.6858367919921875,
      "learning_rate": 0.0005802482460874257,
      "loss": 4.2657,
      "step": 3150
    },
    {
      "epoch": 0.3450134770889488,
      "grad_norm": 0.6485952138900757,
      "learning_rate": 0.0005799244468429573,
      "loss": 4.261,
      "step": 3200
    },
    {
      "epoch": 0.3504043126684636,
      "grad_norm": 0.706744372844696,
      "learning_rate": 0.0005796006475984889,
      "loss": 4.2609,
      "step": 3250
    },
    {
      "epoch": 0.3557951482479784,
      "grad_norm": 0.7495291233062744,
      "learning_rate": 0.0005792768483540205,
      "loss": 4.2458,
      "step": 3300
    },
    {
      "epoch": 0.3611859838274933,
      "grad_norm": 0.6684705018997192,
      "learning_rate": 0.000578953049109552,
      "loss": 4.2625,
      "step": 3350
    },
    {
      "epoch": 0.3665768194070081,
      "grad_norm": 0.6015046834945679,
      "learning_rate": 0.0005786292498650836,
      "loss": 4.2344,
      "step": 3400
    },
    {
      "epoch": 0.3719676549865229,
      "grad_norm": 0.6746869087219238,
      "learning_rate": 0.0005783054506206152,
      "loss": 4.2115,
      "step": 3450
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 0.6800693273544312,
      "learning_rate": 0.0005779816513761467,
      "loss": 4.2238,
      "step": 3500
    },
    {
      "epoch": 0.38274932614555257,
      "grad_norm": 0.8763014674186707,
      "learning_rate": 0.0005776578521316782,
      "loss": 4.2176,
      "step": 3550
    },
    {
      "epoch": 0.3881401617250674,
      "grad_norm": 0.6071361899375916,
      "learning_rate": 0.0005773340528872099,
      "loss": 4.2172,
      "step": 3600
    },
    {
      "epoch": 0.3935309973045822,
      "grad_norm": 0.8424677848815918,
      "learning_rate": 0.0005770102536427414,
      "loss": 4.1909,
      "step": 3650
    },
    {
      "epoch": 0.39892183288409705,
      "grad_norm": 0.7024722099304199,
      "learning_rate": 0.000576686454398273,
      "loss": 4.193,
      "step": 3700
    },
    {
      "epoch": 0.40431266846361186,
      "grad_norm": 0.6678298711776733,
      "learning_rate": 0.0005763626551538045,
      "loss": 4.1954,
      "step": 3750
    },
    {
      "epoch": 0.40970350404312667,
      "grad_norm": 0.591792643070221,
      "learning_rate": 0.0005760388559093362,
      "loss": 4.1897,
      "step": 3800
    },
    {
      "epoch": 0.41509433962264153,
      "grad_norm": 0.7370854020118713,
      "learning_rate": 0.0005757150566648678,
      "loss": 4.1703,
      "step": 3850
    },
    {
      "epoch": 0.42048517520215634,
      "grad_norm": 0.8055949807167053,
      "learning_rate": 0.0005753912574203993,
      "loss": 4.1666,
      "step": 3900
    },
    {
      "epoch": 0.42587601078167114,
      "grad_norm": 0.6214135885238647,
      "learning_rate": 0.0005750674581759309,
      "loss": 4.1586,
      "step": 3950
    },
    {
      "epoch": 0.431266846361186,
      "grad_norm": 0.7452210783958435,
      "learning_rate": 0.0005747436589314624,
      "loss": 4.159,
      "step": 4000
    },
    {
      "epoch": 0.431266846361186,
      "eval_accuracy": 0.31298409615085193,
      "eval_loss": 4.08732271194458,
      "eval_runtime": 183.8347,
      "eval_samples_per_second": 97.974,
      "eval_steps_per_second": 6.125,
      "step": 4000
    },
    {
      "epoch": 0.4366576819407008,
      "grad_norm": 0.5454065203666687,
      "learning_rate": 0.0005744198596869941,
      "loss": 4.1509,
      "step": 4050
    },
    {
      "epoch": 0.4420485175202156,
      "grad_norm": 0.6480352282524109,
      "learning_rate": 0.0005740960604425255,
      "loss": 4.1515,
      "step": 4100
    },
    {
      "epoch": 0.4474393530997305,
      "grad_norm": 0.7491263151168823,
      "learning_rate": 0.0005737722611980572,
      "loss": 4.1543,
      "step": 4150
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 0.5414714217185974,
      "learning_rate": 0.0005734484619535887,
      "loss": 4.1397,
      "step": 4200
    },
    {
      "epoch": 0.4582210242587601,
      "grad_norm": 0.735364556312561,
      "learning_rate": 0.0005731246627091203,
      "loss": 4.1231,
      "step": 4250
    },
    {
      "epoch": 0.4636118598382749,
      "grad_norm": 0.7785477042198181,
      "learning_rate": 0.0005728008634646518,
      "loss": 4.1242,
      "step": 4300
    },
    {
      "epoch": 0.46900269541778977,
      "grad_norm": 0.6108700633049011,
      "learning_rate": 0.0005724770642201835,
      "loss": 4.118,
      "step": 4350
    },
    {
      "epoch": 0.4743935309973046,
      "grad_norm": 0.6415695548057556,
      "learning_rate": 0.000572153264975715,
      "loss": 4.1178,
      "step": 4400
    },
    {
      "epoch": 0.4797843665768194,
      "grad_norm": 0.8770371675491333,
      "learning_rate": 0.0005718294657312466,
      "loss": 4.1028,
      "step": 4450
    },
    {
      "epoch": 0.48517520215633425,
      "grad_norm": 0.7115625143051147,
      "learning_rate": 0.0005715056664867781,
      "loss": 4.1109,
      "step": 4500
    },
    {
      "epoch": 0.49056603773584906,
      "grad_norm": 0.660191535949707,
      "learning_rate": 0.0005711818672423097,
      "loss": 4.1011,
      "step": 4550
    },
    {
      "epoch": 0.49595687331536387,
      "grad_norm": 0.6192442178726196,
      "learning_rate": 0.0005708580679978413,
      "loss": 4.0857,
      "step": 4600
    },
    {
      "epoch": 0.5013477088948787,
      "grad_norm": 0.6783645153045654,
      "learning_rate": 0.0005705342687533729,
      "loss": 4.078,
      "step": 4650
    },
    {
      "epoch": 0.5067385444743935,
      "grad_norm": 0.6928367018699646,
      "learning_rate": 0.0005702104695089044,
      "loss": 4.0956,
      "step": 4700
    },
    {
      "epoch": 0.5121293800539084,
      "grad_norm": 0.6265180110931396,
      "learning_rate": 0.000569886670264436,
      "loss": 4.0822,
      "step": 4750
    },
    {
      "epoch": 0.5175202156334232,
      "grad_norm": 0.7510149478912354,
      "learning_rate": 0.0005695628710199675,
      "loss": 4.0755,
      "step": 4800
    },
    {
      "epoch": 0.522911051212938,
      "grad_norm": 0.7354061603546143,
      "learning_rate": 0.0005692390717754991,
      "loss": 4.0655,
      "step": 4850
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 0.6540889739990234,
      "learning_rate": 0.0005689152725310306,
      "loss": 4.055,
      "step": 4900
    },
    {
      "epoch": 0.5336927223719676,
      "grad_norm": 0.674555778503418,
      "learning_rate": 0.0005685914732865623,
      "loss": 4.0696,
      "step": 4950
    },
    {
      "epoch": 0.5390835579514824,
      "grad_norm": 0.4996623992919922,
      "learning_rate": 0.0005682676740420939,
      "loss": 4.0472,
      "step": 5000
    },
    {
      "epoch": 0.5390835579514824,
      "eval_accuracy": 0.3215982057496718,
      "eval_loss": 3.9894087314605713,
      "eval_runtime": 183.8499,
      "eval_samples_per_second": 97.966,
      "eval_steps_per_second": 6.125,
      "step": 5000
    },
    {
      "epoch": 0.5444743935309974,
      "grad_norm": 0.7127335071563721,
      "learning_rate": 0.0005679438747976254,
      "loss": 4.0568,
      "step": 5050
    },
    {
      "epoch": 0.5498652291105122,
      "grad_norm": 0.6820212602615356,
      "learning_rate": 0.000567620075553157,
      "loss": 4.0516,
      "step": 5100
    },
    {
      "epoch": 0.555256064690027,
      "grad_norm": 0.7278046011924744,
      "learning_rate": 0.0005672962763086886,
      "loss": 4.0498,
      "step": 5150
    },
    {
      "epoch": 0.5606469002695418,
      "grad_norm": 0.624863862991333,
      "learning_rate": 0.0005669724770642202,
      "loss": 4.0671,
      "step": 5200
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.6197010278701782,
      "learning_rate": 0.0005666486778197517,
      "loss": 4.041,
      "step": 5250
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.7419002056121826,
      "learning_rate": 0.0005663248785752833,
      "loss": 4.0326,
      "step": 5300
    },
    {
      "epoch": 0.5768194070080862,
      "grad_norm": 0.6294003129005432,
      "learning_rate": 0.0005660010793308148,
      "loss": 4.0263,
      "step": 5350
    },
    {
      "epoch": 0.5822102425876011,
      "grad_norm": 0.5749978423118591,
      "learning_rate": 0.0005656772800863465,
      "loss": 4.0359,
      "step": 5400
    },
    {
      "epoch": 0.5876010781671159,
      "grad_norm": 0.5415307879447937,
      "learning_rate": 0.0005653534808418779,
      "loss": 4.0382,
      "step": 5450
    },
    {
      "epoch": 0.5929919137466307,
      "grad_norm": 0.6727957725524902,
      "learning_rate": 0.0005650296815974096,
      "loss": 4.0181,
      "step": 5500
    },
    {
      "epoch": 0.5983827493261455,
      "grad_norm": 0.6418060660362244,
      "learning_rate": 0.0005647058823529411,
      "loss": 4.0069,
      "step": 5550
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.7099882960319519,
      "learning_rate": 0.0005643820831084727,
      "loss": 4.0226,
      "step": 5600
    },
    {
      "epoch": 0.6091644204851752,
      "grad_norm": 0.6498609185218811,
      "learning_rate": 0.0005640582838640042,
      "loss": 4.0046,
      "step": 5650
    },
    {
      "epoch": 0.6145552560646901,
      "grad_norm": 0.5713142156600952,
      "learning_rate": 0.0005637344846195358,
      "loss": 4.0043,
      "step": 5700
    },
    {
      "epoch": 0.6199460916442049,
      "grad_norm": 0.6617730259895325,
      "learning_rate": 0.0005634106853750674,
      "loss": 3.979,
      "step": 5750
    },
    {
      "epoch": 0.6253369272237197,
      "grad_norm": 0.6460627317428589,
      "learning_rate": 0.000563086886130599,
      "loss": 3.9846,
      "step": 5800
    },
    {
      "epoch": 0.6307277628032345,
      "grad_norm": 0.6782198548316956,
      "learning_rate": 0.0005627630868861305,
      "loss": 3.9947,
      "step": 5850
    },
    {
      "epoch": 0.6361185983827493,
      "grad_norm": 0.5800774097442627,
      "learning_rate": 0.0005624392876416621,
      "loss": 3.9922,
      "step": 5900
    },
    {
      "epoch": 0.6415094339622641,
      "grad_norm": 0.676412045955658,
      "learning_rate": 0.0005621154883971937,
      "loss": 3.9791,
      "step": 5950
    },
    {
      "epoch": 0.6469002695417789,
      "grad_norm": 0.583685040473938,
      "learning_rate": 0.0005617916891527253,
      "loss": 3.9931,
      "step": 6000
    },
    {
      "epoch": 0.6469002695417789,
      "eval_accuracy": 0.32843508006251015,
      "eval_loss": 3.9174036979675293,
      "eval_runtime": 183.9133,
      "eval_samples_per_second": 97.932,
      "eval_steps_per_second": 6.122,
      "step": 6000
    },
    {
      "epoch": 0.6522911051212938,
      "grad_norm": 0.5780828595161438,
      "learning_rate": 0.0005614678899082568,
      "loss": 3.9663,
      "step": 6050
    },
    {
      "epoch": 0.6576819407008087,
      "grad_norm": 0.5652017593383789,
      "learning_rate": 0.0005611440906637884,
      "loss": 3.9971,
      "step": 6100
    },
    {
      "epoch": 0.6630727762803235,
      "grad_norm": 0.8341758847236633,
      "learning_rate": 0.00056082029141932,
      "loss": 3.9793,
      "step": 6150
    },
    {
      "epoch": 0.6684636118598383,
      "grad_norm": 0.578648567199707,
      "learning_rate": 0.0005604964921748515,
      "loss": 3.9499,
      "step": 6200
    },
    {
      "epoch": 0.6738544474393531,
      "grad_norm": 0.59043949842453,
      "learning_rate": 0.000560172692930383,
      "loss": 3.9715,
      "step": 6250
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 0.6629766225814819,
      "learning_rate": 0.0005598488936859147,
      "loss": 3.9749,
      "step": 6300
    },
    {
      "epoch": 0.6846361185983828,
      "grad_norm": 0.5541216135025024,
      "learning_rate": 0.0005595250944414463,
      "loss": 3.9608,
      "step": 6350
    },
    {
      "epoch": 0.6900269541778976,
      "grad_norm": 0.6767361164093018,
      "learning_rate": 0.0005592012951969778,
      "loss": 3.9694,
      "step": 6400
    },
    {
      "epoch": 0.6954177897574124,
      "grad_norm": 0.6981496810913086,
      "learning_rate": 0.0005588774959525094,
      "loss": 3.9655,
      "step": 6450
    },
    {
      "epoch": 0.7008086253369272,
      "grad_norm": 0.5768890380859375,
      "learning_rate": 0.000558553696708041,
      "loss": 3.96,
      "step": 6500
    },
    {
      "epoch": 0.706199460916442,
      "grad_norm": 0.6919478178024292,
      "learning_rate": 0.0005582298974635726,
      "loss": 3.958,
      "step": 6550
    },
    {
      "epoch": 0.7115902964959568,
      "grad_norm": 0.5978683829307556,
      "learning_rate": 0.0005579060982191041,
      "loss": 3.9587,
      "step": 6600
    },
    {
      "epoch": 0.7169811320754716,
      "grad_norm": 0.6285462379455566,
      "learning_rate": 0.0005575822989746357,
      "loss": 3.9703,
      "step": 6650
    },
    {
      "epoch": 0.7223719676549866,
      "grad_norm": 0.7304993867874146,
      "learning_rate": 0.0005572584997301672,
      "loss": 3.9348,
      "step": 6700
    },
    {
      "epoch": 0.7277628032345014,
      "grad_norm": 0.5624628663063049,
      "learning_rate": 0.0005569347004856989,
      "loss": 3.9602,
      "step": 6750
    },
    {
      "epoch": 0.7331536388140162,
      "grad_norm": 0.6064887642860413,
      "learning_rate": 0.0005566109012412303,
      "loss": 3.9331,
      "step": 6800
    },
    {
      "epoch": 0.738544474393531,
      "grad_norm": 0.5876811146736145,
      "learning_rate": 0.000556287101996762,
      "loss": 3.9323,
      "step": 6850
    },
    {
      "epoch": 0.7439353099730458,
      "grad_norm": 0.5871797204017639,
      "learning_rate": 0.0005559633027522935,
      "loss": 3.9329,
      "step": 6900
    },
    {
      "epoch": 0.7493261455525606,
      "grad_norm": 0.6020776629447937,
      "learning_rate": 0.0005556395035078251,
      "loss": 3.9362,
      "step": 6950
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.6208156943321228,
      "learning_rate": 0.0005553157042633566,
      "loss": 3.9191,
      "step": 7000
    },
    {
      "epoch": 0.7547169811320755,
      "eval_accuracy": 0.33344278300899177,
      "eval_loss": 3.8618791103363037,
      "eval_runtime": 183.7498,
      "eval_samples_per_second": 98.019,
      "eval_steps_per_second": 6.128,
      "step": 7000
    },
    {
      "epoch": 0.7601078167115903,
      "grad_norm": 0.6798801422119141,
      "learning_rate": 0.0005549919050188882,
      "loss": 3.9331,
      "step": 7050
    },
    {
      "epoch": 0.7654986522911051,
      "grad_norm": 1.4628221988677979,
      "learning_rate": 0.0005546681057744198,
      "loss": 3.906,
      "step": 7100
    },
    {
      "epoch": 0.77088948787062,
      "grad_norm": 0.6019647121429443,
      "learning_rate": 0.0005543443065299514,
      "loss": 3.9351,
      "step": 7150
    },
    {
      "epoch": 0.7762803234501348,
      "grad_norm": 0.5555400252342224,
      "learning_rate": 0.000554020507285483,
      "loss": 3.9058,
      "step": 7200
    },
    {
      "epoch": 0.7816711590296496,
      "grad_norm": 0.6362798810005188,
      "learning_rate": 0.0005536967080410145,
      "loss": 3.8967,
      "step": 7250
    },
    {
      "epoch": 0.7870619946091644,
      "grad_norm": 0.5722786784172058,
      "learning_rate": 0.0005533729087965462,
      "loss": 3.9111,
      "step": 7300
    },
    {
      "epoch": 0.7924528301886793,
      "grad_norm": 0.580388069152832,
      "learning_rate": 0.0005530491095520777,
      "loss": 3.9099,
      "step": 7350
    },
    {
      "epoch": 0.7978436657681941,
      "grad_norm": 0.5334943532943726,
      "learning_rate": 0.0005527253103076093,
      "loss": 3.9072,
      "step": 7400
    },
    {
      "epoch": 0.8032345013477089,
      "grad_norm": 0.6169307827949524,
      "learning_rate": 0.0005524015110631408,
      "loss": 3.8832,
      "step": 7450
    },
    {
      "epoch": 0.8086253369272237,
      "grad_norm": 0.6007163524627686,
      "learning_rate": 0.0005520777118186724,
      "loss": 3.9024,
      "step": 7500
    },
    {
      "epoch": 0.8140161725067385,
      "grad_norm": 0.5735055804252625,
      "learning_rate": 0.0005517539125742039,
      "loss": 3.8979,
      "step": 7550
    },
    {
      "epoch": 0.8194070080862533,
      "grad_norm": 0.5585691928863525,
      "learning_rate": 0.0005514365893146249,
      "loss": 3.8902,
      "step": 7600
    },
    {
      "epoch": 0.8247978436657682,
      "grad_norm": 0.5613856315612793,
      "learning_rate": 0.0005511127900701564,
      "loss": 3.8952,
      "step": 7650
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 0.5988004803657532,
      "learning_rate": 0.000550788990825688,
      "loss": 3.8894,
      "step": 7700
    },
    {
      "epoch": 0.8355795148247979,
      "grad_norm": 0.5915306210517883,
      "learning_rate": 0.0005504651915812196,
      "loss": 3.8922,
      "step": 7750
    },
    {
      "epoch": 0.8409703504043127,
      "grad_norm": 0.7122664451599121,
      "learning_rate": 0.0005501413923367512,
      "loss": 3.8766,
      "step": 7800
    },
    {
      "epoch": 0.8463611859838275,
      "grad_norm": 0.5410688519477844,
      "learning_rate": 0.0005498175930922827,
      "loss": 3.8794,
      "step": 7850
    },
    {
      "epoch": 0.8517520215633423,
      "grad_norm": 0.5719017386436462,
      "learning_rate": 0.0005494937938478143,
      "loss": 3.8727,
      "step": 7900
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.570475697517395,
      "learning_rate": 0.0005491699946033459,
      "loss": 3.9063,
      "step": 7950
    },
    {
      "epoch": 0.862533692722372,
      "grad_norm": 0.5836682915687561,
      "learning_rate": 0.0005488461953588775,
      "loss": 3.8542,
      "step": 8000
    },
    {
      "epoch": 0.862533692722372,
      "eval_accuracy": 0.33794155582895036,
      "eval_loss": 3.816067695617676,
      "eval_runtime": 183.5803,
      "eval_samples_per_second": 98.11,
      "eval_steps_per_second": 6.134,
      "step": 8000
    },
    {
      "epoch": 0.8679245283018868,
      "grad_norm": 0.6498636603355408,
      "learning_rate": 0.000548522396114409,
      "loss": 3.895,
      "step": 8050
    },
    {
      "epoch": 0.8733153638814016,
      "grad_norm": 0.6250914931297302,
      "learning_rate": 0.0005481985968699406,
      "loss": 3.8735,
      "step": 8100
    },
    {
      "epoch": 0.8787061994609164,
      "grad_norm": 0.5511185526847839,
      "learning_rate": 0.0005478747976254721,
      "loss": 3.8758,
      "step": 8150
    },
    {
      "epoch": 0.8840970350404312,
      "grad_norm": 0.6440311074256897,
      "learning_rate": 0.0005475509983810037,
      "loss": 3.869,
      "step": 8200
    },
    {
      "epoch": 0.889487870619946,
      "grad_norm": 0.5398238301277161,
      "learning_rate": 0.0005472271991365352,
      "loss": 3.866,
      "step": 8250
    },
    {
      "epoch": 0.894878706199461,
      "grad_norm": 0.5822187066078186,
      "learning_rate": 0.0005469033998920669,
      "loss": 3.8864,
      "step": 8300
    },
    {
      "epoch": 0.9002695417789758,
      "grad_norm": 0.5405433773994446,
      "learning_rate": 0.0005465796006475984,
      "loss": 3.857,
      "step": 8350
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 0.5735152959823608,
      "learning_rate": 0.00054625580140313,
      "loss": 3.8628,
      "step": 8400
    },
    {
      "epoch": 0.9110512129380054,
      "grad_norm": 0.5623797178268433,
      "learning_rate": 0.0005459320021586615,
      "loss": 3.858,
      "step": 8450
    },
    {
      "epoch": 0.9164420485175202,
      "grad_norm": 0.5893810391426086,
      "learning_rate": 0.0005456082029141932,
      "loss": 3.8638,
      "step": 8500
    },
    {
      "epoch": 0.921832884097035,
      "grad_norm": 0.6306450366973877,
      "learning_rate": 0.0005452844036697248,
      "loss": 3.8565,
      "step": 8550
    },
    {
      "epoch": 0.9272237196765498,
      "grad_norm": 0.5499542951583862,
      "learning_rate": 0.0005449606044252563,
      "loss": 3.8639,
      "step": 8600
    },
    {
      "epoch": 0.9326145552560647,
      "grad_norm": 0.563671886920929,
      "learning_rate": 0.0005446368051807879,
      "loss": 3.844,
      "step": 8650
    },
    {
      "epoch": 0.9380053908355795,
      "grad_norm": 0.5789870619773865,
      "learning_rate": 0.0005443130059363194,
      "loss": 3.8492,
      "step": 8700
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.5480877757072449,
      "learning_rate": 0.0005439892066918511,
      "loss": 3.8458,
      "step": 8750
    },
    {
      "epoch": 0.9487870619946092,
      "grad_norm": 0.5352542996406555,
      "learning_rate": 0.0005436654074473825,
      "loss": 3.8487,
      "step": 8800
    },
    {
      "epoch": 0.954177897574124,
      "grad_norm": 0.5493354797363281,
      "learning_rate": 0.0005433416082029142,
      "loss": 3.8454,
      "step": 8850
    },
    {
      "epoch": 0.9595687331536388,
      "grad_norm": 0.5753820538520813,
      "learning_rate": 0.0005430178089584457,
      "loss": 3.8372,
      "step": 8900
    },
    {
      "epoch": 0.9649595687331537,
      "grad_norm": 0.5328227877616882,
      "learning_rate": 0.0005426940097139773,
      "loss": 3.8276,
      "step": 8950
    },
    {
      "epoch": 0.9703504043126685,
      "grad_norm": 0.5667588114738464,
      "learning_rate": 0.0005423702104695088,
      "loss": 3.8531,
      "step": 9000
    },
    {
      "epoch": 0.9703504043126685,
      "eval_accuracy": 0.34165379039401994,
      "eval_loss": 3.7784414291381836,
      "eval_runtime": 184.2031,
      "eval_samples_per_second": 97.778,
      "eval_steps_per_second": 6.113,
      "step": 9000
    },
    {
      "epoch": 0.9757412398921833,
      "grad_norm": 0.6010326743125916,
      "learning_rate": 0.0005420464112250404,
      "loss": 3.8678,
      "step": 9050
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 0.5293521285057068,
      "learning_rate": 0.000541722611980572,
      "loss": 3.8241,
      "step": 9100
    },
    {
      "epoch": 0.9865229110512129,
      "grad_norm": 0.5834252238273621,
      "learning_rate": 0.0005413988127361036,
      "loss": 3.8378,
      "step": 9150
    },
    {
      "epoch": 0.9919137466307277,
      "grad_norm": 0.5734151005744934,
      "learning_rate": 0.0005410750134916351,
      "loss": 3.8295,
      "step": 9200
    },
    {
      "epoch": 0.9973045822102425,
      "grad_norm": 0.5477326512336731,
      "learning_rate": 0.0005407512142471667,
      "loss": 3.8451,
      "step": 9250
    },
    {
      "epoch": 1.0026954177897573,
      "grad_norm": 0.5776128172874451,
      "learning_rate": 0.0005404274150026983,
      "loss": 3.7938,
      "step": 9300
    },
    {
      "epoch": 1.0080862533692723,
      "grad_norm": 0.5947125554084778,
      "learning_rate": 0.0005401036157582299,
      "loss": 3.7605,
      "step": 9350
    },
    {
      "epoch": 1.013477088948787,
      "grad_norm": 0.6069636940956116,
      "learning_rate": 0.0005397798165137614,
      "loss": 3.7525,
      "step": 9400
    },
    {
      "epoch": 1.0188679245283019,
      "grad_norm": 0.5255810618400574,
      "learning_rate": 0.000539456017269293,
      "loss": 3.7525,
      "step": 9450
    },
    {
      "epoch": 1.0242587601078168,
      "grad_norm": 0.546853244304657,
      "learning_rate": 0.0005391322180248245,
      "loss": 3.7576,
      "step": 9500
    },
    {
      "epoch": 1.0296495956873315,
      "grad_norm": 0.5615027546882629,
      "learning_rate": 0.0005388084187803561,
      "loss": 3.7679,
      "step": 9550
    },
    {
      "epoch": 1.0350404312668464,
      "grad_norm": 0.579430341720581,
      "learning_rate": 0.0005384846195358876,
      "loss": 3.7661,
      "step": 9600
    },
    {
      "epoch": 1.0404312668463611,
      "grad_norm": 0.6099395751953125,
      "learning_rate": 0.0005381608202914193,
      "loss": 3.7707,
      "step": 9650
    },
    {
      "epoch": 1.045822102425876,
      "grad_norm": 0.5670583248138428,
      "learning_rate": 0.0005378370210469509,
      "loss": 3.7544,
      "step": 9700
    },
    {
      "epoch": 1.0512129380053907,
      "grad_norm": 0.5511836409568787,
      "learning_rate": 0.0005375132218024824,
      "loss": 3.7525,
      "step": 9750
    },
    {
      "epoch": 1.0566037735849056,
      "grad_norm": 0.5288443565368652,
      "learning_rate": 0.000537189422558014,
      "loss": 3.7805,
      "step": 9800
    },
    {
      "epoch": 1.0619946091644206,
      "grad_norm": 0.5281191468238831,
      "learning_rate": 0.0005368656233135455,
      "loss": 3.7818,
      "step": 9850
    },
    {
      "epoch": 1.0673854447439353,
      "grad_norm": 0.5530230402946472,
      "learning_rate": 0.0005365483000539665,
      "loss": 3.7735,
      "step": 9900
    },
    {
      "epoch": 1.0727762803234502,
      "grad_norm": 0.6312490701675415,
      "learning_rate": 0.0005362245008094981,
      "loss": 3.7825,
      "step": 9950
    },
    {
      "epoch": 1.0781671159029649,
      "grad_norm": 0.5409021377563477,
      "learning_rate": 0.0005359007015650297,
      "loss": 3.7613,
      "step": 10000
    },
    {
      "epoch": 1.0781671159029649,
      "eval_accuracy": 0.3451733833889944,
      "eval_loss": 3.7442734241485596,
      "eval_runtime": 183.7504,
      "eval_samples_per_second": 98.019,
      "eval_steps_per_second": 6.128,
      "step": 10000
    },
    {
      "epoch": 1.0835579514824798,
      "grad_norm": 0.5455986857414246,
      "learning_rate": 0.0005355769023205612,
      "loss": 3.7552,
      "step": 10050
    },
    {
      "epoch": 1.0889487870619945,
      "grad_norm": 0.5736761093139648,
      "learning_rate": 0.0005352531030760928,
      "loss": 3.7591,
      "step": 10100
    },
    {
      "epoch": 1.0943396226415094,
      "grad_norm": 0.5683985948562622,
      "learning_rate": 0.0005349293038316244,
      "loss": 3.7521,
      "step": 10150
    },
    {
      "epoch": 1.0997304582210243,
      "grad_norm": 0.5910122394561768,
      "learning_rate": 0.0005346055045871559,
      "loss": 3.7653,
      "step": 10200
    },
    {
      "epoch": 1.105121293800539,
      "grad_norm": 0.6161583065986633,
      "learning_rate": 0.0005342817053426874,
      "loss": 3.753,
      "step": 10250
    },
    {
      "epoch": 1.110512129380054,
      "grad_norm": 0.5718348026275635,
      "learning_rate": 0.0005339579060982191,
      "loss": 3.7615,
      "step": 10300
    },
    {
      "epoch": 1.1159029649595686,
      "grad_norm": 0.576435387134552,
      "learning_rate": 0.0005336341068537506,
      "loss": 3.7567,
      "step": 10350
    },
    {
      "epoch": 1.1212938005390836,
      "grad_norm": 0.5745424628257751,
      "learning_rate": 0.0005333103076092822,
      "loss": 3.7612,
      "step": 10400
    },
    {
      "epoch": 1.1266846361185983,
      "grad_norm": 0.5627392530441284,
      "learning_rate": 0.0005329865083648137,
      "loss": 3.757,
      "step": 10450
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.6117731332778931,
      "learning_rate": 0.0005326627091203454,
      "loss": 3.7597,
      "step": 10500
    },
    {
      "epoch": 1.137466307277628,
      "grad_norm": 0.6170853972434998,
      "learning_rate": 0.0005323389098758769,
      "loss": 3.7656,
      "step": 10550
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.5776093006134033,
      "learning_rate": 0.0005320151106314085,
      "loss": 3.7554,
      "step": 10600
    },
    {
      "epoch": 1.1482479784366577,
      "grad_norm": 0.6169910430908203,
      "learning_rate": 0.00053169131138694,
      "loss": 3.7413,
      "step": 10650
    },
    {
      "epoch": 1.1536388140161726,
      "grad_norm": 0.5566491484642029,
      "learning_rate": 0.0005313675121424716,
      "loss": 3.7528,
      "step": 10700
    },
    {
      "epoch": 1.1590296495956873,
      "grad_norm": 0.5667610168457031,
      "learning_rate": 0.0005310437128980032,
      "loss": 3.7331,
      "step": 10750
    },
    {
      "epoch": 1.1644204851752022,
      "grad_norm": 0.5442368984222412,
      "learning_rate": 0.0005307199136535348,
      "loss": 3.7391,
      "step": 10800
    },
    {
      "epoch": 1.169811320754717,
      "grad_norm": 0.5643858313560486,
      "learning_rate": 0.0005303961144090663,
      "loss": 3.7309,
      "step": 10850
    },
    {
      "epoch": 1.1752021563342319,
      "grad_norm": 0.6005975604057312,
      "learning_rate": 0.0005300723151645979,
      "loss": 3.7448,
      "step": 10900
    },
    {
      "epoch": 1.1805929919137466,
      "grad_norm": 0.5089237093925476,
      "learning_rate": 0.0005297485159201295,
      "loss": 3.7585,
      "step": 10950
    },
    {
      "epoch": 1.1859838274932615,
      "grad_norm": 0.5960420370101929,
      "learning_rate": 0.000529424716675661,
      "loss": 3.7685,
      "step": 11000
    },
    {
      "epoch": 1.1859838274932615,
      "eval_accuracy": 0.34775128180528075,
      "eval_loss": 3.720940589904785,
      "eval_runtime": 183.5903,
      "eval_samples_per_second": 98.104,
      "eval_steps_per_second": 6.133,
      "step": 11000
    },
    {
      "epoch": 1.1913746630727764,
      "grad_norm": 0.575962245464325,
      "learning_rate": 0.0005291009174311926,
      "loss": 3.755,
      "step": 11050
    },
    {
      "epoch": 1.196765498652291,
      "grad_norm": 0.523404598236084,
      "learning_rate": 0.0005287771181867242,
      "loss": 3.7415,
      "step": 11100
    },
    {
      "epoch": 1.202156334231806,
      "grad_norm": 0.6159588694572449,
      "learning_rate": 0.0005284533189422558,
      "loss": 3.7365,
      "step": 11150
    },
    {
      "epoch": 1.2075471698113207,
      "grad_norm": 0.49879810214042664,
      "learning_rate": 0.0005281295196977873,
      "loss": 3.7591,
      "step": 11200
    },
    {
      "epoch": 1.2129380053908356,
      "grad_norm": 0.5774774551391602,
      "learning_rate": 0.0005278057204533189,
      "loss": 3.7483,
      "step": 11250
    },
    {
      "epoch": 1.2183288409703503,
      "grad_norm": 0.5057991743087769,
      "learning_rate": 0.0005274819212088505,
      "loss": 3.732,
      "step": 11300
    },
    {
      "epoch": 1.2237196765498652,
      "grad_norm": 0.5313456654548645,
      "learning_rate": 0.0005271581219643821,
      "loss": 3.7282,
      "step": 11350
    },
    {
      "epoch": 1.2291105121293802,
      "grad_norm": 0.7374264597892761,
      "learning_rate": 0.0005268343227199136,
      "loss": 3.7426,
      "step": 11400
    },
    {
      "epoch": 1.2345013477088949,
      "grad_norm": 0.5699669718742371,
      "learning_rate": 0.0005265105234754452,
      "loss": 3.7535,
      "step": 11450
    },
    {
      "epoch": 1.2398921832884098,
      "grad_norm": 0.567206084728241,
      "learning_rate": 0.0005261867242309767,
      "loss": 3.7431,
      "step": 11500
    },
    {
      "epoch": 1.2452830188679245,
      "grad_norm": 0.5753964781761169,
      "learning_rate": 0.0005258629249865083,
      "loss": 3.7371,
      "step": 11550
    },
    {
      "epoch": 1.2506738544474394,
      "grad_norm": 0.5284551978111267,
      "learning_rate": 0.0005255391257420398,
      "loss": 3.7509,
      "step": 11600
    },
    {
      "epoch": 1.256064690026954,
      "grad_norm": 0.5363852381706238,
      "learning_rate": 0.0005252153264975715,
      "loss": 3.7284,
      "step": 11650
    },
    {
      "epoch": 1.261455525606469,
      "grad_norm": 0.5310810804367065,
      "learning_rate": 0.000524891527253103,
      "loss": 3.7414,
      "step": 11700
    },
    {
      "epoch": 1.266846361185984,
      "grad_norm": 0.5255619287490845,
      "learning_rate": 0.0005245677280086346,
      "loss": 3.7409,
      "step": 11750
    },
    {
      "epoch": 1.2722371967654986,
      "grad_norm": 0.6119208931922913,
      "learning_rate": 0.0005242439287641661,
      "loss": 3.7545,
      "step": 11800
    },
    {
      "epoch": 1.2776280323450135,
      "grad_norm": 0.6038151979446411,
      "learning_rate": 0.0005239201295196978,
      "loss": 3.7357,
      "step": 11850
    },
    {
      "epoch": 1.2830188679245282,
      "grad_norm": 0.5870354175567627,
      "learning_rate": 0.0005235963302752293,
      "loss": 3.7318,
      "step": 11900
    },
    {
      "epoch": 1.2884097035040432,
      "grad_norm": 0.6072551012039185,
      "learning_rate": 0.0005232725310307609,
      "loss": 3.7151,
      "step": 11950
    },
    {
      "epoch": 1.2938005390835579,
      "grad_norm": 0.5396766662597656,
      "learning_rate": 0.0005229487317862924,
      "loss": 3.7287,
      "step": 12000
    },
    {
      "epoch": 1.2938005390835579,
      "eval_accuracy": 0.3499499816430946,
      "eval_loss": 3.6942965984344482,
      "eval_runtime": 183.9882,
      "eval_samples_per_second": 97.892,
      "eval_steps_per_second": 6.12,
      "step": 12000
    },
    {
      "epoch": 1.2991913746630728,
      "grad_norm": 0.5627809166908264,
      "learning_rate": 0.000522624932541824,
      "loss": 3.7433,
      "step": 12050
    },
    {
      "epoch": 1.3045822102425877,
      "grad_norm": 0.5790122151374817,
      "learning_rate": 0.0005223011332973557,
      "loss": 3.7371,
      "step": 12100
    },
    {
      "epoch": 1.3099730458221024,
      "grad_norm": 0.5617051720619202,
      "learning_rate": 0.0005219773340528872,
      "loss": 3.725,
      "step": 12150
    },
    {
      "epoch": 1.3153638814016173,
      "grad_norm": 0.5151467323303223,
      "learning_rate": 0.0005216535348084188,
      "loss": 3.7435,
      "step": 12200
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.571425199508667,
      "learning_rate": 0.0005213297355639503,
      "loss": 3.7168,
      "step": 12250
    },
    {
      "epoch": 1.326145552560647,
      "grad_norm": 0.554076611995697,
      "learning_rate": 0.0005210059363194819,
      "loss": 3.7248,
      "step": 12300
    },
    {
      "epoch": 1.3315363881401616,
      "grad_norm": 0.5426397919654846,
      "learning_rate": 0.0005206821370750134,
      "loss": 3.733,
      "step": 12350
    },
    {
      "epoch": 1.3369272237196765,
      "grad_norm": 0.5451375246047974,
      "learning_rate": 0.000520358337830545,
      "loss": 3.7304,
      "step": 12400
    },
    {
      "epoch": 1.3423180592991915,
      "grad_norm": 0.5756955146789551,
      "learning_rate": 0.0005200345385860766,
      "loss": 3.7188,
      "step": 12450
    },
    {
      "epoch": 1.3477088948787062,
      "grad_norm": 0.5725782513618469,
      "learning_rate": 0.0005197107393416082,
      "loss": 3.7372,
      "step": 12500
    },
    {
      "epoch": 1.353099730458221,
      "grad_norm": 0.5362164974212646,
      "learning_rate": 0.0005193869400971397,
      "loss": 3.7099,
      "step": 12550
    },
    {
      "epoch": 1.3584905660377358,
      "grad_norm": 0.5761519074440002,
      "learning_rate": 0.0005190631408526713,
      "loss": 3.7107,
      "step": 12600
    },
    {
      "epoch": 1.3638814016172507,
      "grad_norm": 0.6134148240089417,
      "learning_rate": 0.0005187393416082029,
      "loss": 3.7015,
      "step": 12650
    },
    {
      "epoch": 1.3692722371967654,
      "grad_norm": 0.556781530380249,
      "learning_rate": 0.0005184220183486238,
      "loss": 3.7163,
      "step": 12700
    },
    {
      "epoch": 1.3746630727762803,
      "grad_norm": 0.534551739692688,
      "learning_rate": 0.0005180982191041554,
      "loss": 3.7077,
      "step": 12750
    },
    {
      "epoch": 1.3800539083557952,
      "grad_norm": 0.5834203362464905,
      "learning_rate": 0.000517774419859687,
      "loss": 3.7058,
      "step": 12800
    },
    {
      "epoch": 1.38544474393531,
      "grad_norm": 0.5853343605995178,
      "learning_rate": 0.0005174506206152185,
      "loss": 3.7172,
      "step": 12850
    },
    {
      "epoch": 1.3908355795148248,
      "grad_norm": 0.5919023156166077,
      "learning_rate": 0.0005171268213707501,
      "loss": 3.7155,
      "step": 12900
    },
    {
      "epoch": 1.3962264150943398,
      "grad_norm": 0.5478957295417786,
      "learning_rate": 0.0005168030221262816,
      "loss": 3.724,
      "step": 12950
    },
    {
      "epoch": 1.4016172506738545,
      "grad_norm": 0.5379932522773743,
      "learning_rate": 0.0005164792228818132,
      "loss": 3.7051,
      "step": 13000
    },
    {
      "epoch": 1.4016172506738545,
      "eval_accuracy": 0.3524520403436865,
      "eval_loss": 3.672461986541748,
      "eval_runtime": 183.6622,
      "eval_samples_per_second": 98.066,
      "eval_steps_per_second": 6.131,
      "step": 13000
    },
    {
      "epoch": 1.4070080862533692,
      "grad_norm": 0.5632694363594055,
      "learning_rate": 0.0005161554236373448,
      "loss": 3.7037,
      "step": 13050
    },
    {
      "epoch": 1.412398921832884,
      "grad_norm": 0.5602201223373413,
      "learning_rate": 0.0005158316243928764,
      "loss": 3.698,
      "step": 13100
    },
    {
      "epoch": 1.417789757412399,
      "grad_norm": 0.6008055806159973,
      "learning_rate": 0.0005155078251484079,
      "loss": 3.7036,
      "step": 13150
    },
    {
      "epoch": 1.4231805929919137,
      "grad_norm": 0.5000778436660767,
      "learning_rate": 0.0005151840259039395,
      "loss": 3.7021,
      "step": 13200
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.5265121459960938,
      "learning_rate": 0.000514860226659471,
      "loss": 3.7143,
      "step": 13250
    },
    {
      "epoch": 1.4339622641509435,
      "grad_norm": 0.5470620393753052,
      "learning_rate": 0.0005145364274150027,
      "loss": 3.6872,
      "step": 13300
    },
    {
      "epoch": 1.4393530997304582,
      "grad_norm": 0.5715850591659546,
      "learning_rate": 0.0005142126281705343,
      "loss": 3.7053,
      "step": 13350
    },
    {
      "epoch": 1.444743935309973,
      "grad_norm": 0.5808437466621399,
      "learning_rate": 0.0005138888289260658,
      "loss": 3.7218,
      "step": 13400
    },
    {
      "epoch": 1.4501347708894878,
      "grad_norm": 0.5598849058151245,
      "learning_rate": 0.0005135650296815974,
      "loss": 3.7246,
      "step": 13450
    },
    {
      "epoch": 1.4555256064690028,
      "grad_norm": 0.6316524744033813,
      "learning_rate": 0.0005132412304371289,
      "loss": 3.7159,
      "step": 13500
    },
    {
      "epoch": 1.4609164420485174,
      "grad_norm": 0.6849909424781799,
      "learning_rate": 0.0005129174311926605,
      "loss": 3.7046,
      "step": 13550
    },
    {
      "epoch": 1.4663072776280324,
      "grad_norm": 0.6671061515808105,
      "learning_rate": 0.000512593631948192,
      "loss": 3.7215,
      "step": 13600
    },
    {
      "epoch": 1.4716981132075473,
      "grad_norm": 0.5725294947624207,
      "learning_rate": 0.0005122698327037237,
      "loss": 3.7105,
      "step": 13650
    },
    {
      "epoch": 1.477088948787062,
      "grad_norm": 0.6049585342407227,
      "learning_rate": 0.0005119460334592552,
      "loss": 3.6986,
      "step": 13700
    },
    {
      "epoch": 1.482479784366577,
      "grad_norm": 0.582939624786377,
      "learning_rate": 0.0005116222342147868,
      "loss": 3.699,
      "step": 13750
    },
    {
      "epoch": 1.4878706199460916,
      "grad_norm": 0.510508120059967,
      "learning_rate": 0.0005112984349703183,
      "loss": 3.7006,
      "step": 13800
    },
    {
      "epoch": 1.4932614555256065,
      "grad_norm": 0.5556900501251221,
      "learning_rate": 0.00051097463572585,
      "loss": 3.7061,
      "step": 13850
    },
    {
      "epoch": 1.4986522911051212,
      "grad_norm": 0.6585386395454407,
      "learning_rate": 0.0005106508364813815,
      "loss": 3.6874,
      "step": 13900
    },
    {
      "epoch": 1.5040431266846361,
      "grad_norm": 0.5924018025398254,
      "learning_rate": 0.0005103270372369131,
      "loss": 3.688,
      "step": 13950
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.5659307241439819,
      "learning_rate": 0.0005100032379924446,
      "loss": 3.7032,
      "step": 14000
    },
    {
      "epoch": 1.509433962264151,
      "eval_accuracy": 0.3539072284701858,
      "eval_loss": 3.6559207439422607,
      "eval_runtime": 183.7429,
      "eval_samples_per_second": 98.023,
      "eval_steps_per_second": 6.128,
      "step": 14000
    },
    {
      "epoch": 1.5148247978436657,
      "grad_norm": 0.5375577807426453,
      "learning_rate": 0.0005096794387479762,
      "loss": 3.7013,
      "step": 14050
    },
    {
      "epoch": 1.5202156334231804,
      "grad_norm": 0.631818950176239,
      "learning_rate": 0.0005093556395035078,
      "loss": 3.6999,
      "step": 14100
    },
    {
      "epoch": 1.5256064690026954,
      "grad_norm": 0.56558758020401,
      "learning_rate": 0.0005090318402590394,
      "loss": 3.6931,
      "step": 14150
    },
    {
      "epoch": 1.5309973045822103,
      "grad_norm": 0.5545461773872375,
      "learning_rate": 0.0005087080410145709,
      "loss": 3.6975,
      "step": 14200
    },
    {
      "epoch": 1.536388140161725,
      "grad_norm": 0.5795785784721375,
      "learning_rate": 0.0005083842417701025,
      "loss": 3.6976,
      "step": 14250
    },
    {
      "epoch": 1.54177897574124,
      "grad_norm": 0.5446183681488037,
      "learning_rate": 0.000508060442525634,
      "loss": 3.6944,
      "step": 14300
    },
    {
      "epoch": 1.5471698113207548,
      "grad_norm": 0.5753372311592102,
      "learning_rate": 0.0005077366432811656,
      "loss": 3.6866,
      "step": 14350
    },
    {
      "epoch": 1.5525606469002695,
      "grad_norm": 0.5649062991142273,
      "learning_rate": 0.0005074128440366971,
      "loss": 3.6891,
      "step": 14400
    },
    {
      "epoch": 1.5579514824797842,
      "grad_norm": 0.5094801783561707,
      "learning_rate": 0.0005070890447922288,
      "loss": 3.6906,
      "step": 14450
    },
    {
      "epoch": 1.5633423180592994,
      "grad_norm": 0.5297936797142029,
      "learning_rate": 0.0005067652455477604,
      "loss": 3.6888,
      "step": 14500
    },
    {
      "epoch": 1.568733153638814,
      "grad_norm": 0.537732720375061,
      "learning_rate": 0.0005064414463032919,
      "loss": 3.6874,
      "step": 14550
    },
    {
      "epoch": 1.5741239892183287,
      "grad_norm": 0.5864264369010925,
      "learning_rate": 0.0005061176470588235,
      "loss": 3.6742,
      "step": 14600
    },
    {
      "epoch": 1.5795148247978437,
      "grad_norm": 0.5446892380714417,
      "learning_rate": 0.0005057938478143551,
      "loss": 3.6769,
      "step": 14650
    },
    {
      "epoch": 1.5849056603773586,
      "grad_norm": 0.5335270762443542,
      "learning_rate": 0.0005054700485698867,
      "loss": 3.6846,
      "step": 14700
    },
    {
      "epoch": 1.5902964959568733,
      "grad_norm": 0.5628997087478638,
      "learning_rate": 0.0005051527253103076,
      "loss": 3.664,
      "step": 14750
    },
    {
      "epoch": 1.595687331536388,
      "grad_norm": 0.5562835931777954,
      "learning_rate": 0.0005048289260658392,
      "loss": 3.6578,
      "step": 14800
    },
    {
      "epoch": 1.6010781671159031,
      "grad_norm": 0.5934857726097107,
      "learning_rate": 0.0005045051268213707,
      "loss": 3.6825,
      "step": 14850
    },
    {
      "epoch": 1.6064690026954178,
      "grad_norm": 0.5580174922943115,
      "learning_rate": 0.0005041813275769023,
      "loss": 3.6859,
      "step": 14900
    },
    {
      "epoch": 1.6118598382749325,
      "grad_norm": 0.5808007121086121,
      "learning_rate": 0.0005038575283324338,
      "loss": 3.6708,
      "step": 14950
    },
    {
      "epoch": 1.6172506738544474,
      "grad_norm": 0.504628598690033,
      "learning_rate": 0.0005035337290879654,
      "loss": 3.6641,
      "step": 15000
    },
    {
      "epoch": 1.6172506738544474,
      "eval_accuracy": 0.3560253078652413,
      "eval_loss": 3.6341910362243652,
      "eval_runtime": 183.7265,
      "eval_samples_per_second": 98.032,
      "eval_steps_per_second": 6.129,
      "step": 15000
    },
    {
      "epoch": 1.6226415094339623,
      "grad_norm": 0.5475382208824158,
      "learning_rate": 0.000503209929843497,
      "loss": 3.6619,
      "step": 15050
    },
    {
      "epoch": 1.628032345013477,
      "grad_norm": 0.561851978302002,
      "learning_rate": 0.0005028861305990286,
      "loss": 3.6737,
      "step": 15100
    },
    {
      "epoch": 1.633423180592992,
      "grad_norm": 0.5445704460144043,
      "learning_rate": 0.0005025623313545601,
      "loss": 3.6936,
      "step": 15150
    },
    {
      "epoch": 1.6388140161725069,
      "grad_norm": 0.5345889329910278,
      "learning_rate": 0.0005022385321100917,
      "loss": 3.6804,
      "step": 15200
    },
    {
      "epoch": 1.6442048517520216,
      "grad_norm": 0.4933413863182068,
      "learning_rate": 0.0005019147328656232,
      "loss": 3.6721,
      "step": 15250
    },
    {
      "epoch": 1.6495956873315363,
      "grad_norm": 0.5393181443214417,
      "learning_rate": 0.0005015909336211549,
      "loss": 3.6716,
      "step": 15300
    },
    {
      "epoch": 1.6549865229110512,
      "grad_norm": 0.5043426752090454,
      "learning_rate": 0.0005012671343766864,
      "loss": 3.6568,
      "step": 15350
    },
    {
      "epoch": 1.6603773584905661,
      "grad_norm": 0.5442577004432678,
      "learning_rate": 0.000500943335132218,
      "loss": 3.675,
      "step": 15400
    },
    {
      "epoch": 1.6657681940700808,
      "grad_norm": 0.5932238101959229,
      "learning_rate": 0.0005006195358877495,
      "loss": 3.6634,
      "step": 15450
    },
    {
      "epoch": 1.6711590296495957,
      "grad_norm": 0.5122867226600647,
      "learning_rate": 0.0005002957366432812,
      "loss": 3.6645,
      "step": 15500
    },
    {
      "epoch": 1.6765498652291106,
      "grad_norm": 0.5891242027282715,
      "learning_rate": 0.0004999719373988127,
      "loss": 3.651,
      "step": 15550
    },
    {
      "epoch": 1.6819407008086253,
      "grad_norm": 0.570210874080658,
      "learning_rate": 0.0004996481381543442,
      "loss": 3.6675,
      "step": 15600
    },
    {
      "epoch": 1.68733153638814,
      "grad_norm": 0.5468628406524658,
      "learning_rate": 0.0004993243389098758,
      "loss": 3.6734,
      "step": 15650
    },
    {
      "epoch": 1.692722371967655,
      "grad_norm": 0.6194903254508972,
      "learning_rate": 0.0004990005396654074,
      "loss": 3.6724,
      "step": 15700
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.5616596937179565,
      "learning_rate": 0.000498676740420939,
      "loss": 3.6535,
      "step": 15750
    },
    {
      "epoch": 1.7035040431266846,
      "grad_norm": 0.5665090084075928,
      "learning_rate": 0.0004983529411764705,
      "loss": 3.6628,
      "step": 15800
    },
    {
      "epoch": 1.7088948787061995,
      "grad_norm": 0.5654871463775635,
      "learning_rate": 0.0004980291419320022,
      "loss": 3.6803,
      "step": 15850
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.5348817110061646,
      "learning_rate": 0.0004977053426875337,
      "loss": 3.6793,
      "step": 15900
    },
    {
      "epoch": 1.719676549865229,
      "grad_norm": 0.556675136089325,
      "learning_rate": 0.0004973815434430653,
      "loss": 3.6665,
      "step": 15950
    },
    {
      "epoch": 1.7250673854447438,
      "grad_norm": 0.5713870525360107,
      "learning_rate": 0.0004970577441985968,
      "loss": 3.6638,
      "step": 16000
    },
    {
      "epoch": 1.7250673854447438,
      "eval_accuracy": 0.35771866312183,
      "eval_loss": 3.618290662765503,
      "eval_runtime": 183.7921,
      "eval_samples_per_second": 97.997,
      "eval_steps_per_second": 6.126,
      "step": 16000
    },
    {
      "epoch": 1.7304582210242587,
      "grad_norm": 0.5948731899261475,
      "learning_rate": 0.0004967339449541284,
      "loss": 3.6498,
      "step": 16050
    },
    {
      "epoch": 1.7358490566037736,
      "grad_norm": 0.5521835684776306,
      "learning_rate": 0.00049641014570966,
      "loss": 3.6514,
      "step": 16100
    },
    {
      "epoch": 1.7412398921832883,
      "grad_norm": 0.6063022613525391,
      "learning_rate": 0.0004960863464651916,
      "loss": 3.6574,
      "step": 16150
    },
    {
      "epoch": 1.7466307277628033,
      "grad_norm": 0.6571112871170044,
      "learning_rate": 0.0004957625472207231,
      "loss": 3.6673,
      "step": 16200
    },
    {
      "epoch": 1.7520215633423182,
      "grad_norm": 0.5960469841957092,
      "learning_rate": 0.0004954387479762547,
      "loss": 3.65,
      "step": 16250
    },
    {
      "epoch": 1.7574123989218329,
      "grad_norm": 0.518771231174469,
      "learning_rate": 0.0004951149487317862,
      "loss": 3.6426,
      "step": 16300
    },
    {
      "epoch": 1.7628032345013476,
      "grad_norm": 0.5418721437454224,
      "learning_rate": 0.0004947911494873178,
      "loss": 3.6663,
      "step": 16350
    },
    {
      "epoch": 1.7681940700808625,
      "grad_norm": 0.4927101135253906,
      "learning_rate": 0.0004944673502428493,
      "loss": 3.646,
      "step": 16400
    },
    {
      "epoch": 1.7735849056603774,
      "grad_norm": 0.5151129364967346,
      "learning_rate": 0.000494143550998381,
      "loss": 3.6555,
      "step": 16450
    },
    {
      "epoch": 1.778975741239892,
      "grad_norm": 0.5475391149520874,
      "learning_rate": 0.0004938197517539125,
      "loss": 3.6402,
      "step": 16500
    },
    {
      "epoch": 1.784366576819407,
      "grad_norm": 0.5400313138961792,
      "learning_rate": 0.0004934959525094441,
      "loss": 3.6485,
      "step": 16550
    },
    {
      "epoch": 1.789757412398922,
      "grad_norm": 0.5347439646720886,
      "learning_rate": 0.0004931721532649756,
      "loss": 3.6607,
      "step": 16600
    },
    {
      "epoch": 1.7951482479784366,
      "grad_norm": 0.5058249235153198,
      "learning_rate": 0.0004928483540205073,
      "loss": 3.6536,
      "step": 16650
    },
    {
      "epoch": 1.8005390835579513,
      "grad_norm": 0.5410025119781494,
      "learning_rate": 0.0004925245547760388,
      "loss": 3.6574,
      "step": 16700
    },
    {
      "epoch": 1.8059299191374663,
      "grad_norm": 0.5334374904632568,
      "learning_rate": 0.0004922072315164598,
      "loss": 3.6566,
      "step": 16750
    },
    {
      "epoch": 1.8113207547169812,
      "grad_norm": 0.5640711784362793,
      "learning_rate": 0.0004918834322719913,
      "loss": 3.6645,
      "step": 16800
    },
    {
      "epoch": 1.8167115902964959,
      "grad_norm": 0.5393502116203308,
      "learning_rate": 0.0004915596330275229,
      "loss": 3.6718,
      "step": 16850
    },
    {
      "epoch": 1.8221024258760108,
      "grad_norm": 0.5351586937904358,
      "learning_rate": 0.0004912358337830544,
      "loss": 3.6445,
      "step": 16900
    },
    {
      "epoch": 1.8274932614555257,
      "grad_norm": 0.5266234874725342,
      "learning_rate": 0.000490912034538586,
      "loss": 3.6559,
      "step": 16950
    },
    {
      "epoch": 1.8328840970350404,
      "grad_norm": 0.5003353357315063,
      "learning_rate": 0.0004905882352941175,
      "loss": 3.6368,
      "step": 17000
    },
    {
      "epoch": 1.8328840970350404,
      "eval_accuracy": 0.3592136182052694,
      "eval_loss": 3.602701187133789,
      "eval_runtime": 183.6818,
      "eval_samples_per_second": 98.055,
      "eval_steps_per_second": 6.13,
      "step": 17000
    },
    {
      "epoch": 1.838274932614555,
      "grad_norm": 0.5567882657051086,
      "learning_rate": 0.0004902644360496492,
      "loss": 3.6407,
      "step": 17050
    },
    {
      "epoch": 1.8436657681940702,
      "grad_norm": 0.5742266178131104,
      "learning_rate": 0.0004899406368051808,
      "loss": 3.6327,
      "step": 17100
    },
    {
      "epoch": 1.849056603773585,
      "grad_norm": 0.5474602580070496,
      "learning_rate": 0.0004896168375607123,
      "loss": 3.6396,
      "step": 17150
    },
    {
      "epoch": 1.8544474393530996,
      "grad_norm": 0.5230289697647095,
      "learning_rate": 0.0004892930383162439,
      "loss": 3.6459,
      "step": 17200
    },
    {
      "epoch": 1.8598382749326146,
      "grad_norm": 0.4841998517513275,
      "learning_rate": 0.0004889692390717754,
      "loss": 3.6515,
      "step": 17250
    },
    {
      "epoch": 1.8652291105121295,
      "grad_norm": 0.5501573085784912,
      "learning_rate": 0.0004886454398273071,
      "loss": 3.6547,
      "step": 17300
    },
    {
      "epoch": 1.8706199460916442,
      "grad_norm": 0.543190598487854,
      "learning_rate": 0.0004883216405828386,
      "loss": 3.6554,
      "step": 17350
    },
    {
      "epoch": 1.8760107816711589,
      "grad_norm": 0.5470271706581116,
      "learning_rate": 0.00048799784133837017,
      "loss": 3.6487,
      "step": 17400
    },
    {
      "epoch": 1.881401617250674,
      "grad_norm": 0.5897737145423889,
      "learning_rate": 0.0004876740420939017,
      "loss": 3.6389,
      "step": 17450
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.5379341244697571,
      "learning_rate": 0.0004873502428494333,
      "loss": 3.6411,
      "step": 17500
    },
    {
      "epoch": 1.8921832884097034,
      "grad_norm": 0.5637924671173096,
      "learning_rate": 0.0004870264436049649,
      "loss": 3.6377,
      "step": 17550
    },
    {
      "epoch": 1.8975741239892183,
      "grad_norm": 0.5570976734161377,
      "learning_rate": 0.00048670264436049643,
      "loss": 3.6509,
      "step": 17600
    },
    {
      "epoch": 1.9029649595687332,
      "grad_norm": 0.6049391031265259,
      "learning_rate": 0.00048637884511602803,
      "loss": 3.6425,
      "step": 17650
    },
    {
      "epoch": 1.908355795148248,
      "grad_norm": 0.5467279553413391,
      "learning_rate": 0.0004860550458715596,
      "loss": 3.6374,
      "step": 17700
    },
    {
      "epoch": 1.9137466307277629,
      "grad_norm": 0.5799625515937805,
      "learning_rate": 0.0004857312466270912,
      "loss": 3.6579,
      "step": 17750
    },
    {
      "epoch": 1.9191374663072778,
      "grad_norm": 0.5544312000274658,
      "learning_rate": 0.00048540744738262274,
      "loss": 3.6415,
      "step": 17800
    },
    {
      "epoch": 1.9245283018867925,
      "grad_norm": 0.577680766582489,
      "learning_rate": 0.00048508364813815434,
      "loss": 3.6342,
      "step": 17850
    },
    {
      "epoch": 1.9299191374663072,
      "grad_norm": 0.5421233177185059,
      "learning_rate": 0.00048475984889368584,
      "loss": 3.6444,
      "step": 17900
    },
    {
      "epoch": 1.935309973045822,
      "grad_norm": 0.5379252433776855,
      "learning_rate": 0.0004844360496492175,
      "loss": 3.6261,
      "step": 17950
    },
    {
      "epoch": 1.940700808625337,
      "grad_norm": 0.5494712591171265,
      "learning_rate": 0.000484112250404749,
      "loss": 3.6291,
      "step": 18000
    },
    {
      "epoch": 1.940700808625337,
      "eval_accuracy": 0.3600362292189129,
      "eval_loss": 3.5892419815063477,
      "eval_runtime": 184.0541,
      "eval_samples_per_second": 97.857,
      "eval_steps_per_second": 6.118,
      "step": 18000
    },
    {
      "epoch": 1.9460916442048517,
      "grad_norm": 0.5733945369720459,
      "learning_rate": 0.00048378845116028055,
      "loss": 3.6336,
      "step": 18050
    },
    {
      "epoch": 1.9514824797843666,
      "grad_norm": 0.567290186882019,
      "learning_rate": 0.00048346465191581215,
      "loss": 3.633,
      "step": 18100
    },
    {
      "epoch": 1.9568733153638815,
      "grad_norm": 0.5201252698898315,
      "learning_rate": 0.0004831408526713437,
      "loss": 3.6308,
      "step": 18150
    },
    {
      "epoch": 1.9622641509433962,
      "grad_norm": 0.5754126310348511,
      "learning_rate": 0.0004828170534268753,
      "loss": 3.6186,
      "step": 18200
    },
    {
      "epoch": 1.967654986522911,
      "grad_norm": 0.5907006859779358,
      "learning_rate": 0.00048249325418240686,
      "loss": 3.6398,
      "step": 18250
    },
    {
      "epoch": 1.9730458221024259,
      "grad_norm": 0.6127999424934387,
      "learning_rate": 0.00048216945493793846,
      "loss": 3.6509,
      "step": 18300
    },
    {
      "epoch": 1.9784366576819408,
      "grad_norm": 0.5465409159660339,
      "learning_rate": 0.00048184565569347,
      "loss": 3.6337,
      "step": 18350
    },
    {
      "epoch": 1.9838274932614555,
      "grad_norm": 0.6550977230072021,
      "learning_rate": 0.0004815218564490016,
      "loss": 3.636,
      "step": 18400
    },
    {
      "epoch": 1.9892183288409704,
      "grad_norm": 0.5740667581558228,
      "learning_rate": 0.00048119805720453317,
      "loss": 3.6441,
      "step": 18450
    },
    {
      "epoch": 1.9946091644204853,
      "grad_norm": 0.5781065225601196,
      "learning_rate": 0.0004808742579600647,
      "loss": 3.6359,
      "step": 18500
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1652530431747437,
      "learning_rate": 0.0004805504587155963,
      "loss": 3.6409,
      "step": 18550
    },
    {
      "epoch": 2.0053908355795147,
      "grad_norm": 0.5303277373313904,
      "learning_rate": 0.0004802266594711278,
      "loss": 3.5688,
      "step": 18600
    },
    {
      "epoch": 2.01078167115903,
      "grad_norm": 0.5572181344032288,
      "learning_rate": 0.0004799028602266594,
      "loss": 3.5383,
      "step": 18650
    },
    {
      "epoch": 2.0161725067385445,
      "grad_norm": 0.577308177947998,
      "learning_rate": 0.000479579060982191,
      "loss": 3.5517,
      "step": 18700
    },
    {
      "epoch": 2.0215633423180592,
      "grad_norm": 0.5191822648048401,
      "learning_rate": 0.0004792552617377226,
      "loss": 3.5297,
      "step": 18750
    },
    {
      "epoch": 2.026954177897574,
      "grad_norm": 0.5460506081581116,
      "learning_rate": 0.00047893146249325413,
      "loss": 3.5495,
      "step": 18800
    },
    {
      "epoch": 2.032345013477089,
      "grad_norm": 0.5392601490020752,
      "learning_rate": 0.0004786141392336751,
      "loss": 3.5583,
      "step": 18850
    },
    {
      "epoch": 2.0377358490566038,
      "grad_norm": 0.5567241311073303,
      "learning_rate": 0.00047829033998920663,
      "loss": 3.5447,
      "step": 18900
    },
    {
      "epoch": 2.0431266846361185,
      "grad_norm": 0.5814675092697144,
      "learning_rate": 0.00047796654074473824,
      "loss": 3.539,
      "step": 18950
    },
    {
      "epoch": 2.0485175202156336,
      "grad_norm": 0.6219574213027954,
      "learning_rate": 0.0004776427415002698,
      "loss": 3.5668,
      "step": 19000
    },
    {
      "epoch": 2.0485175202156336,
      "eval_accuracy": 0.3623585760430596,
      "eval_loss": 3.5776376724243164,
      "eval_runtime": 183.6098,
      "eval_samples_per_second": 98.094,
      "eval_steps_per_second": 6.133,
      "step": 19000
    },
    {
      "epoch": 2.0539083557951483,
      "grad_norm": 0.5591341257095337,
      "learning_rate": 0.0004773189422558014,
      "loss": 3.5422,
      "step": 19050
    },
    {
      "epoch": 2.059299191374663,
      "grad_norm": 0.5593909025192261,
      "learning_rate": 0.00047699514301133294,
      "loss": 3.566,
      "step": 19100
    },
    {
      "epoch": 2.0646900269541777,
      "grad_norm": 0.5888485908508301,
      "learning_rate": 0.00047667134376686455,
      "loss": 3.5615,
      "step": 19150
    },
    {
      "epoch": 2.070080862533693,
      "grad_norm": 0.5862798094749451,
      "learning_rate": 0.0004763475445223961,
      "loss": 3.5691,
      "step": 19200
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.5456637740135193,
      "learning_rate": 0.0004760237452779276,
      "loss": 3.5466,
      "step": 19250
    },
    {
      "epoch": 2.0808625336927222,
      "grad_norm": 0.5669352412223816,
      "learning_rate": 0.0004756999460334592,
      "loss": 3.544,
      "step": 19300
    },
    {
      "epoch": 2.0862533692722374,
      "grad_norm": 0.6276587247848511,
      "learning_rate": 0.00047537614678899075,
      "loss": 3.5469,
      "step": 19350
    },
    {
      "epoch": 2.091644204851752,
      "grad_norm": 0.5461747646331787,
      "learning_rate": 0.00047505234754452235,
      "loss": 3.5435,
      "step": 19400
    },
    {
      "epoch": 2.0970350404312668,
      "grad_norm": 0.7402114868164062,
      "learning_rate": 0.0004747285483000539,
      "loss": 3.5519,
      "step": 19450
    },
    {
      "epoch": 2.1024258760107815,
      "grad_norm": 0.6080474853515625,
      "learning_rate": 0.0004744047490555855,
      "loss": 3.5453,
      "step": 19500
    },
    {
      "epoch": 2.1078167115902966,
      "grad_norm": 0.5156586170196533,
      "learning_rate": 0.00047408094981111706,
      "loss": 3.5511,
      "step": 19550
    },
    {
      "epoch": 2.1132075471698113,
      "grad_norm": 0.6247926354408264,
      "learning_rate": 0.00047375715056664866,
      "loss": 3.548,
      "step": 19600
    },
    {
      "epoch": 2.118598382749326,
      "grad_norm": 0.5468766093254089,
      "learning_rate": 0.0004734333513221802,
      "loss": 3.5704,
      "step": 19650
    },
    {
      "epoch": 2.123989218328841,
      "grad_norm": 0.5999529957771301,
      "learning_rate": 0.00047310955207771177,
      "loss": 3.5414,
      "step": 19700
    },
    {
      "epoch": 2.129380053908356,
      "grad_norm": 0.5577241778373718,
      "learning_rate": 0.00047278575283324337,
      "loss": 3.561,
      "step": 19750
    },
    {
      "epoch": 2.1347708894878705,
      "grad_norm": 0.5632361769676208,
      "learning_rate": 0.0004724619535887749,
      "loss": 3.5539,
      "step": 19800
    },
    {
      "epoch": 2.1401617250673857,
      "grad_norm": 0.5894713401794434,
      "learning_rate": 0.0004721381543443065,
      "loss": 3.5463,
      "step": 19850
    },
    {
      "epoch": 2.1455525606469004,
      "grad_norm": 0.5929577946662903,
      "learning_rate": 0.000471814355099838,
      "loss": 3.5588,
      "step": 19900
    },
    {
      "epoch": 2.150943396226415,
      "grad_norm": 0.5735053420066833,
      "learning_rate": 0.0004714905558553697,
      "loss": 3.5507,
      "step": 19950
    },
    {
      "epoch": 2.1563342318059298,
      "grad_norm": 0.572102427482605,
      "learning_rate": 0.0004711667566109012,
      "loss": 3.5744,
      "step": 20000
    },
    {
      "epoch": 2.1563342318059298,
      "eval_accuracy": 0.3623309782095547,
      "eval_loss": 3.5741710662841797,
      "eval_runtime": 184.0834,
      "eval_samples_per_second": 97.842,
      "eval_steps_per_second": 6.117,
      "step": 20000
    },
    {
      "epoch": 2.161725067385445,
      "grad_norm": 0.5751939415931702,
      "learning_rate": 0.00047084295736643273,
      "loss": 3.5742,
      "step": 20050
    },
    {
      "epoch": 2.1671159029649596,
      "grad_norm": 0.5660809874534607,
      "learning_rate": 0.00047051915812196433,
      "loss": 3.5769,
      "step": 20100
    },
    {
      "epoch": 2.1725067385444743,
      "grad_norm": 0.5410208702087402,
      "learning_rate": 0.0004701953588774959,
      "loss": 3.57,
      "step": 20150
    },
    {
      "epoch": 2.177897574123989,
      "grad_norm": 0.5733333826065063,
      "learning_rate": 0.0004698715596330275,
      "loss": 3.5658,
      "step": 20200
    },
    {
      "epoch": 2.183288409703504,
      "grad_norm": 0.6063035130500793,
      "learning_rate": 0.00046954776038855904,
      "loss": 3.5663,
      "step": 20250
    },
    {
      "epoch": 2.188679245283019,
      "grad_norm": 0.5596222281455994,
      "learning_rate": 0.00046922396114409064,
      "loss": 3.5569,
      "step": 20300
    },
    {
      "epoch": 2.1940700808625335,
      "grad_norm": 0.5413583517074585,
      "learning_rate": 0.0004689001618996222,
      "loss": 3.5729,
      "step": 20350
    },
    {
      "epoch": 2.1994609164420487,
      "grad_norm": 0.550186276435852,
      "learning_rate": 0.0004685763626551538,
      "loss": 3.5729,
      "step": 20400
    },
    {
      "epoch": 2.2048517520215634,
      "grad_norm": 0.6196731925010681,
      "learning_rate": 0.00046825256341068535,
      "loss": 3.5705,
      "step": 20450
    },
    {
      "epoch": 2.210242587601078,
      "grad_norm": 0.5624878406524658,
      "learning_rate": 0.0004679287641662169,
      "loss": 3.5411,
      "step": 20500
    },
    {
      "epoch": 2.215633423180593,
      "grad_norm": 0.55267733335495,
      "learning_rate": 0.0004676049649217485,
      "loss": 3.5695,
      "step": 20550
    },
    {
      "epoch": 2.221024258760108,
      "grad_norm": 0.6480850577354431,
      "learning_rate": 0.00046728116567728,
      "loss": 3.5455,
      "step": 20600
    },
    {
      "epoch": 2.2264150943396226,
      "grad_norm": 0.561083197593689,
      "learning_rate": 0.0004669573664328116,
      "loss": 3.5533,
      "step": 20650
    },
    {
      "epoch": 2.2318059299191373,
      "grad_norm": 0.5583511590957642,
      "learning_rate": 0.00046663356718834316,
      "loss": 3.5504,
      "step": 20700
    },
    {
      "epoch": 2.2371967654986524,
      "grad_norm": 0.6693071126937866,
      "learning_rate": 0.00046630976794387476,
      "loss": 3.5581,
      "step": 20750
    },
    {
      "epoch": 2.242587601078167,
      "grad_norm": 0.59278404712677,
      "learning_rate": 0.0004659859686994063,
      "loss": 3.5522,
      "step": 20800
    },
    {
      "epoch": 2.247978436657682,
      "grad_norm": 0.5818943977355957,
      "learning_rate": 0.0004656621694549379,
      "loss": 3.5674,
      "step": 20850
    },
    {
      "epoch": 2.2533692722371965,
      "grad_norm": 0.5376104116439819,
      "learning_rate": 0.00046533837021046947,
      "loss": 3.5601,
      "step": 20900
    },
    {
      "epoch": 2.2587601078167117,
      "grad_norm": 0.5341477990150452,
      "learning_rate": 0.000465014570966001,
      "loss": 3.5495,
      "step": 20950
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.5756168961524963,
      "learning_rate": 0.0004646907717215326,
      "loss": 3.5621,
      "step": 21000
    },
    {
      "epoch": 2.2641509433962264,
      "eval_accuracy": 0.3640750743647527,
      "eval_loss": 3.559407949447632,
      "eval_runtime": 183.7723,
      "eval_samples_per_second": 98.007,
      "eval_steps_per_second": 6.127,
      "step": 21000
    },
    {
      "epoch": 2.269541778975741,
      "grad_norm": 0.5583627223968506,
      "learning_rate": 0.0004643734484619536,
      "loss": 3.5553,
      "step": 21050
    },
    {
      "epoch": 2.274932614555256,
      "grad_norm": 0.6221109628677368,
      "learning_rate": 0.0004640496492174851,
      "loss": 3.5521,
      "step": 21100
    },
    {
      "epoch": 2.280323450134771,
      "grad_norm": 0.5557717680931091,
      "learning_rate": 0.00046372584997301673,
      "loss": 3.5767,
      "step": 21150
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.5979974269866943,
      "learning_rate": 0.0004634020507285483,
      "loss": 3.5602,
      "step": 21200
    },
    {
      "epoch": 2.2911051212938007,
      "grad_norm": 0.6088611483573914,
      "learning_rate": 0.0004630782514840798,
      "loss": 3.5701,
      "step": 21250
    },
    {
      "epoch": 2.2964959568733154,
      "grad_norm": 0.561540961265564,
      "learning_rate": 0.0004627544522396114,
      "loss": 3.5368,
      "step": 21300
    },
    {
      "epoch": 2.30188679245283,
      "grad_norm": 0.6222971677780151,
      "learning_rate": 0.00046243065299514293,
      "loss": 3.559,
      "step": 21350
    },
    {
      "epoch": 2.3072776280323453,
      "grad_norm": 0.5412083864212036,
      "learning_rate": 0.00046210685375067454,
      "loss": 3.5516,
      "step": 21400
    },
    {
      "epoch": 2.31266846361186,
      "grad_norm": 0.5675464868545532,
      "learning_rate": 0.0004617830545062061,
      "loss": 3.5741,
      "step": 21450
    },
    {
      "epoch": 2.3180592991913747,
      "grad_norm": 0.5713619589805603,
      "learning_rate": 0.0004614592552617377,
      "loss": 3.5468,
      "step": 21500
    },
    {
      "epoch": 2.3234501347708894,
      "grad_norm": 0.5480716824531555,
      "learning_rate": 0.00046113545601726924,
      "loss": 3.5626,
      "step": 21550
    },
    {
      "epoch": 2.3288409703504045,
      "grad_norm": 0.6332092881202698,
      "learning_rate": 0.00046081165677280085,
      "loss": 3.5526,
      "step": 21600
    },
    {
      "epoch": 2.334231805929919,
      "grad_norm": 0.6008698344230652,
      "learning_rate": 0.0004604878575283324,
      "loss": 3.5542,
      "step": 21650
    },
    {
      "epoch": 2.339622641509434,
      "grad_norm": 0.6072392463684082,
      "learning_rate": 0.00046016405828386395,
      "loss": 3.5722,
      "step": 21700
    },
    {
      "epoch": 2.3450134770889486,
      "grad_norm": 0.5895165801048279,
      "learning_rate": 0.00045984025903939555,
      "loss": 3.5609,
      "step": 21750
    },
    {
      "epoch": 2.3504043126684637,
      "grad_norm": 0.6073868274688721,
      "learning_rate": 0.0004595164597949271,
      "loss": 3.5427,
      "step": 21800
    },
    {
      "epoch": 2.3557951482479784,
      "grad_norm": 0.5244640111923218,
      "learning_rate": 0.0004591926605504587,
      "loss": 3.5541,
      "step": 21850
    },
    {
      "epoch": 2.361185983827493,
      "grad_norm": 0.5883101224899292,
      "learning_rate": 0.0004588688613059902,
      "loss": 3.5479,
      "step": 21900
    },
    {
      "epoch": 2.3665768194070083,
      "grad_norm": 0.5923961997032166,
      "learning_rate": 0.00045854506206152186,
      "loss": 3.5506,
      "step": 21950
    },
    {
      "epoch": 2.371967654986523,
      "grad_norm": 0.5549956560134888,
      "learning_rate": 0.00045822126281705336,
      "loss": 3.5697,
      "step": 22000
    },
    {
      "epoch": 2.371967654986523,
      "eval_accuracy": 0.3651554100282921,
      "eval_loss": 3.5489094257354736,
      "eval_runtime": 183.5303,
      "eval_samples_per_second": 98.136,
      "eval_steps_per_second": 6.135,
      "step": 22000
    },
    {
      "epoch": 2.3773584905660377,
      "grad_norm": 0.5528952479362488,
      "learning_rate": 0.00045789746357258497,
      "loss": 3.5474,
      "step": 22050
    },
    {
      "epoch": 2.382749326145553,
      "grad_norm": 0.6946583390235901,
      "learning_rate": 0.0004575736643281165,
      "loss": 3.5487,
      "step": 22100
    },
    {
      "epoch": 2.3881401617250675,
      "grad_norm": 0.5659127235412598,
      "learning_rate": 0.00045724986508364807,
      "loss": 3.5513,
      "step": 22150
    },
    {
      "epoch": 2.393530997304582,
      "grad_norm": 0.5336350798606873,
      "learning_rate": 0.00045692606583917967,
      "loss": 3.5623,
      "step": 22200
    },
    {
      "epoch": 2.398921832884097,
      "grad_norm": 0.5657646656036377,
      "learning_rate": 0.0004566022665947112,
      "loss": 3.5658,
      "step": 22250
    },
    {
      "epoch": 2.404312668463612,
      "grad_norm": 0.5422862768173218,
      "learning_rate": 0.0004562849433351322,
      "loss": 3.5705,
      "step": 22300
    },
    {
      "epoch": 2.4097035040431267,
      "grad_norm": 0.5550121068954468,
      "learning_rate": 0.0004559611440906638,
      "loss": 3.5507,
      "step": 22350
    },
    {
      "epoch": 2.4150943396226414,
      "grad_norm": 0.5688802003860474,
      "learning_rate": 0.00045563734484619533,
      "loss": 3.5711,
      "step": 22400
    },
    {
      "epoch": 2.420485175202156,
      "grad_norm": 0.5675523281097412,
      "learning_rate": 0.0004553135456017269,
      "loss": 3.5616,
      "step": 22450
    },
    {
      "epoch": 2.4258760107816713,
      "grad_norm": 0.5429262518882751,
      "learning_rate": 0.0004549897463572585,
      "loss": 3.5557,
      "step": 22500
    },
    {
      "epoch": 2.431266846361186,
      "grad_norm": 0.6132704615592957,
      "learning_rate": 0.00045466594711279,
      "loss": 3.5504,
      "step": 22550
    },
    {
      "epoch": 2.4366576819407006,
      "grad_norm": 0.5489976406097412,
      "learning_rate": 0.00045434214786832164,
      "loss": 3.5518,
      "step": 22600
    },
    {
      "epoch": 2.442048517520216,
      "grad_norm": 0.5603680610656738,
      "learning_rate": 0.00045401834862385314,
      "loss": 3.5373,
      "step": 22650
    },
    {
      "epoch": 2.4474393530997305,
      "grad_norm": 0.5937761068344116,
      "learning_rate": 0.00045369454937938474,
      "loss": 3.5433,
      "step": 22700
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.5349526405334473,
      "learning_rate": 0.0004533707501349163,
      "loss": 3.5468,
      "step": 22750
    },
    {
      "epoch": 2.4582210242587603,
      "grad_norm": 0.5487583875656128,
      "learning_rate": 0.0004530469508904479,
      "loss": 3.5434,
      "step": 22800
    },
    {
      "epoch": 2.463611859838275,
      "grad_norm": 0.6188079714775085,
      "learning_rate": 0.00045272315164597945,
      "loss": 3.5518,
      "step": 22850
    },
    {
      "epoch": 2.4690026954177897,
      "grad_norm": 0.6644785404205322,
      "learning_rate": 0.000452399352401511,
      "loss": 3.5607,
      "step": 22900
    },
    {
      "epoch": 2.4743935309973044,
      "grad_norm": 0.579646646976471,
      "learning_rate": 0.0004520755531570426,
      "loss": 3.5622,
      "step": 22950
    },
    {
      "epoch": 2.4797843665768196,
      "grad_norm": 0.5342908501625061,
      "learning_rate": 0.00045175175391257415,
      "loss": 3.537,
      "step": 23000
    },
    {
      "epoch": 2.4797843665768196,
      "eval_accuracy": 0.36602680618856426,
      "eval_loss": 3.5380289554595947,
      "eval_runtime": 183.7529,
      "eval_samples_per_second": 98.017,
      "eval_steps_per_second": 6.128,
      "step": 23000
    },
    {
      "epoch": 2.4851752021563343,
      "grad_norm": 0.570767343044281,
      "learning_rate": 0.00045142795466810576,
      "loss": 3.5417,
      "step": 23050
    },
    {
      "epoch": 2.490566037735849,
      "grad_norm": 0.5754830837249756,
      "learning_rate": 0.0004511041554236373,
      "loss": 3.5566,
      "step": 23100
    },
    {
      "epoch": 2.4959568733153636,
      "grad_norm": 0.5624446868896484,
      "learning_rate": 0.0004507803561791689,
      "loss": 3.5579,
      "step": 23150
    },
    {
      "epoch": 2.501347708894879,
      "grad_norm": 0.6118834614753723,
      "learning_rate": 0.00045045655693470046,
      "loss": 3.5519,
      "step": 23200
    },
    {
      "epoch": 2.5067385444743935,
      "grad_norm": 0.5770419836044312,
      "learning_rate": 0.00045013275769023207,
      "loss": 3.5446,
      "step": 23250
    },
    {
      "epoch": 2.512129380053908,
      "grad_norm": 0.5695980787277222,
      "learning_rate": 0.00044980895844576356,
      "loss": 3.544,
      "step": 23300
    },
    {
      "epoch": 2.5175202156334233,
      "grad_norm": 0.6304803490638733,
      "learning_rate": 0.0004494851592012951,
      "loss": 3.5394,
      "step": 23350
    },
    {
      "epoch": 2.522911051212938,
      "grad_norm": 0.5704509019851685,
      "learning_rate": 0.0004491613599568267,
      "loss": 3.5319,
      "step": 23400
    },
    {
      "epoch": 2.5283018867924527,
      "grad_norm": 0.5579874515533447,
      "learning_rate": 0.00044883756071235827,
      "loss": 3.5454,
      "step": 23450
    },
    {
      "epoch": 2.533692722371968,
      "grad_norm": 0.5578489899635315,
      "learning_rate": 0.0004485137614678899,
      "loss": 3.5406,
      "step": 23500
    },
    {
      "epoch": 2.5390835579514826,
      "grad_norm": 0.5366498231887817,
      "learning_rate": 0.0004481899622234214,
      "loss": 3.5523,
      "step": 23550
    },
    {
      "epoch": 2.5444743935309972,
      "grad_norm": 0.5812051892280579,
      "learning_rate": 0.00044786616297895303,
      "loss": 3.5521,
      "step": 23600
    },
    {
      "epoch": 2.5498652291105124,
      "grad_norm": 0.577279806137085,
      "learning_rate": 0.0004475423637344846,
      "loss": 3.5546,
      "step": 23650
    },
    {
      "epoch": 2.555256064690027,
      "grad_norm": 0.5395628213882446,
      "learning_rate": 0.00044721856449001613,
      "loss": 3.5627,
      "step": 23700
    },
    {
      "epoch": 2.560646900269542,
      "grad_norm": 0.5482959151268005,
      "learning_rate": 0.00044689476524554774,
      "loss": 3.5664,
      "step": 23750
    },
    {
      "epoch": 2.5660377358490565,
      "grad_norm": 0.6769027709960938,
      "learning_rate": 0.0004465709660010793,
      "loss": 3.5453,
      "step": 23800
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.5909700989723206,
      "learning_rate": 0.0004462471667566109,
      "loss": 3.5611,
      "step": 23850
    },
    {
      "epoch": 2.5768194070080863,
      "grad_norm": 0.6435213088989258,
      "learning_rate": 0.00044592336751214244,
      "loss": 3.5433,
      "step": 23900
    },
    {
      "epoch": 2.582210242587601,
      "grad_norm": 0.5850532650947571,
      "learning_rate": 0.00044559956826767405,
      "loss": 3.5489,
      "step": 23950
    },
    {
      "epoch": 2.5876010781671157,
      "grad_norm": 0.663982093334198,
      "learning_rate": 0.00044527576902320554,
      "loss": 3.5319,
      "step": 24000
    },
    {
      "epoch": 2.5876010781671157,
      "eval_accuracy": 0.36721959759099165,
      "eval_loss": 3.5314929485321045,
      "eval_runtime": 183.664,
      "eval_samples_per_second": 98.065,
      "eval_steps_per_second": 6.131,
      "step": 24000
    },
    {
      "epoch": 2.592991913746631,
      "grad_norm": 0.5680175423622131,
      "learning_rate": 0.00044495196977873715,
      "loss": 3.5434,
      "step": 24050
    },
    {
      "epoch": 2.5983827493261455,
      "grad_norm": 0.6317371129989624,
      "learning_rate": 0.0004446281705342687,
      "loss": 3.538,
      "step": 24100
    },
    {
      "epoch": 2.6037735849056602,
      "grad_norm": 0.5761317610740662,
      "learning_rate": 0.00044430437128980025,
      "loss": 3.5633,
      "step": 24150
    },
    {
      "epoch": 2.6091644204851754,
      "grad_norm": 0.5837670564651489,
      "learning_rate": 0.00044398057204533185,
      "loss": 3.5289,
      "step": 24200
    },
    {
      "epoch": 2.61455525606469,
      "grad_norm": 0.5544631481170654,
      "learning_rate": 0.0004436567728008634,
      "loss": 3.54,
      "step": 24250
    },
    {
      "epoch": 2.6199460916442048,
      "grad_norm": 0.5587742924690247,
      "learning_rate": 0.000443332973556395,
      "loss": 3.5479,
      "step": 24300
    },
    {
      "epoch": 2.62533692722372,
      "grad_norm": 0.5495783090591431,
      "learning_rate": 0.00044300917431192656,
      "loss": 3.5642,
      "step": 24350
    },
    {
      "epoch": 2.6307277628032346,
      "grad_norm": 0.5933278203010559,
      "learning_rate": 0.00044268537506745816,
      "loss": 3.526,
      "step": 24400
    },
    {
      "epoch": 2.6361185983827493,
      "grad_norm": 0.5730457305908203,
      "learning_rate": 0.0004423615758229897,
      "loss": 3.546,
      "step": 24450
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.5949668288230896,
      "learning_rate": 0.0004420377765785213,
      "loss": 3.5588,
      "step": 24500
    },
    {
      "epoch": 2.6469002695417787,
      "grad_norm": 0.5814875364303589,
      "learning_rate": 0.00044171397733405287,
      "loss": 3.5262,
      "step": 24550
    },
    {
      "epoch": 2.652291105121294,
      "grad_norm": 0.5717437267303467,
      "learning_rate": 0.00044139017808958437,
      "loss": 3.5318,
      "step": 24600
    },
    {
      "epoch": 2.6576819407008085,
      "grad_norm": 0.5098128914833069,
      "learning_rate": 0.00044106637884511597,
      "loss": 3.5355,
      "step": 24650
    },
    {
      "epoch": 2.6630727762803232,
      "grad_norm": 0.5785273909568787,
      "learning_rate": 0.0004407425796006475,
      "loss": 3.5384,
      "step": 24700
    },
    {
      "epoch": 2.6684636118598384,
      "grad_norm": 0.5735189318656921,
      "learning_rate": 0.0004404252563410685,
      "loss": 3.5379,
      "step": 24750
    },
    {
      "epoch": 2.673854447439353,
      "grad_norm": 0.5718660354614258,
      "learning_rate": 0.0004401014570966001,
      "loss": 3.5167,
      "step": 24800
    },
    {
      "epoch": 2.6792452830188678,
      "grad_norm": 0.5400899052619934,
      "learning_rate": 0.00043977765785213163,
      "loss": 3.5144,
      "step": 24850
    },
    {
      "epoch": 2.684636118598383,
      "grad_norm": 0.5528421998023987,
      "learning_rate": 0.0004394538586076632,
      "loss": 3.5536,
      "step": 24900
    },
    {
      "epoch": 2.6900269541778976,
      "grad_norm": 0.5778202414512634,
      "learning_rate": 0.0004391300593631948,
      "loss": 3.536,
      "step": 24950
    },
    {
      "epoch": 2.6954177897574123,
      "grad_norm": 0.5615628957748413,
      "learning_rate": 0.00043880626011872634,
      "loss": 3.5265,
      "step": 25000
    },
    {
      "epoch": 2.6954177897574123,
      "eval_accuracy": 0.36784598148924214,
      "eval_loss": 3.5187320709228516,
      "eval_runtime": 183.9267,
      "eval_samples_per_second": 97.925,
      "eval_steps_per_second": 6.122,
      "step": 25000
    },
    {
      "epoch": 2.7008086253369274,
      "grad_norm": 0.5722801685333252,
      "learning_rate": 0.00043848246087425794,
      "loss": 3.5369,
      "step": 25050
    },
    {
      "epoch": 2.706199460916442,
      "grad_norm": 0.5277543067932129,
      "learning_rate": 0.0004381586616297895,
      "loss": 3.5222,
      "step": 25100
    },
    {
      "epoch": 2.711590296495957,
      "grad_norm": 0.5880123972892761,
      "learning_rate": 0.0004378348623853211,
      "loss": 3.5126,
      "step": 25150
    },
    {
      "epoch": 2.7169811320754715,
      "grad_norm": 0.5826494097709656,
      "learning_rate": 0.00043751106314085265,
      "loss": 3.5171,
      "step": 25200
    },
    {
      "epoch": 2.7223719676549867,
      "grad_norm": 0.5825526118278503,
      "learning_rate": 0.00043718726389638425,
      "loss": 3.5533,
      "step": 25250
    },
    {
      "epoch": 2.7277628032345014,
      "grad_norm": 0.5669980049133301,
      "learning_rate": 0.00043686346465191575,
      "loss": 3.5104,
      "step": 25300
    },
    {
      "epoch": 2.733153638814016,
      "grad_norm": 0.6107063889503479,
      "learning_rate": 0.0004365396654074473,
      "loss": 3.5351,
      "step": 25350
    },
    {
      "epoch": 2.7385444743935308,
      "grad_norm": 0.6292389035224915,
      "learning_rate": 0.0004362158661629789,
      "loss": 3.5373,
      "step": 25400
    },
    {
      "epoch": 2.743935309973046,
      "grad_norm": 0.5547637343406677,
      "learning_rate": 0.00043589206691851045,
      "loss": 3.5098,
      "step": 25450
    },
    {
      "epoch": 2.7493261455525606,
      "grad_norm": 0.5594532489776611,
      "learning_rate": 0.00043556826767404206,
      "loss": 3.5693,
      "step": 25500
    },
    {
      "epoch": 2.7547169811320753,
      "grad_norm": 0.5568592548370361,
      "learning_rate": 0.0004352444684295736,
      "loss": 3.5577,
      "step": 25550
    },
    {
      "epoch": 2.7601078167115904,
      "grad_norm": 0.5955683588981628,
      "learning_rate": 0.0004349206691851052,
      "loss": 3.5345,
      "step": 25600
    },
    {
      "epoch": 2.765498652291105,
      "grad_norm": 0.6182993054389954,
      "learning_rate": 0.00043459686994063676,
      "loss": 3.5233,
      "step": 25650
    },
    {
      "epoch": 2.77088948787062,
      "grad_norm": 0.5633643865585327,
      "learning_rate": 0.00043427307069616837,
      "loss": 3.5408,
      "step": 25700
    },
    {
      "epoch": 2.776280323450135,
      "grad_norm": 0.5655704140663147,
      "learning_rate": 0.0004339492714516999,
      "loss": 3.5331,
      "step": 25750
    },
    {
      "epoch": 2.7816711590296497,
      "grad_norm": 0.560209333896637,
      "learning_rate": 0.00043362547220723147,
      "loss": 3.5319,
      "step": 25800
    },
    {
      "epoch": 2.7870619946091644,
      "grad_norm": 0.5956665873527527,
      "learning_rate": 0.0004333016729627631,
      "loss": 3.5201,
      "step": 25850
    },
    {
      "epoch": 2.7924528301886795,
      "grad_norm": 0.586514949798584,
      "learning_rate": 0.0004329778737182946,
      "loss": 3.5297,
      "step": 25900
    },
    {
      "epoch": 2.797843665768194,
      "grad_norm": 0.5690777897834778,
      "learning_rate": 0.00043265407447382623,
      "loss": 3.5196,
      "step": 25950
    },
    {
      "epoch": 2.803234501347709,
      "grad_norm": 0.5629533529281616,
      "learning_rate": 0.0004323302752293577,
      "loss": 3.5313,
      "step": 26000
    },
    {
      "epoch": 2.803234501347709,
      "eval_accuracy": 0.36869542976617575,
      "eval_loss": 3.510869026184082,
      "eval_runtime": 183.6283,
      "eval_samples_per_second": 98.084,
      "eval_steps_per_second": 6.132,
      "step": 26000
    },
    {
      "epoch": 2.8086253369272236,
      "grad_norm": 0.5801641345024109,
      "learning_rate": 0.00043200647598488933,
      "loss": 3.5313,
      "step": 26050
    },
    {
      "epoch": 2.8140161725067383,
      "grad_norm": 0.5631428956985474,
      "learning_rate": 0.0004316826767404209,
      "loss": 3.503,
      "step": 26100
    },
    {
      "epoch": 2.8194070080862534,
      "grad_norm": 0.5923662185668945,
      "learning_rate": 0.0004313588774959525,
      "loss": 3.5071,
      "step": 26150
    },
    {
      "epoch": 2.824797843665768,
      "grad_norm": 0.5873034000396729,
      "learning_rate": 0.00043103507825148404,
      "loss": 3.5307,
      "step": 26200
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.5601552724838257,
      "learning_rate": 0.0004307112790070156,
      "loss": 3.5429,
      "step": 26250
    },
    {
      "epoch": 2.835579514824798,
      "grad_norm": 0.5437904596328735,
      "learning_rate": 0.0004303874797625472,
      "loss": 3.5588,
      "step": 26300
    },
    {
      "epoch": 2.8409703504043127,
      "grad_norm": 0.7464920878410339,
      "learning_rate": 0.00043006368051807874,
      "loss": 3.529,
      "step": 26350
    },
    {
      "epoch": 2.8463611859838274,
      "grad_norm": 0.5931203365325928,
      "learning_rate": 0.00042973988127361035,
      "loss": 3.5378,
      "step": 26400
    },
    {
      "epoch": 2.8517520215633425,
      "grad_norm": 0.6171684265136719,
      "learning_rate": 0.0004294160820291419,
      "loss": 3.5438,
      "step": 26450
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.5586252808570862,
      "learning_rate": 0.0004290922827846735,
      "loss": 3.5151,
      "step": 26500
    },
    {
      "epoch": 2.862533692722372,
      "grad_norm": 0.5578598976135254,
      "learning_rate": 0.00042876848354020505,
      "loss": 3.5306,
      "step": 26550
    },
    {
      "epoch": 2.867924528301887,
      "grad_norm": 0.5919521450996399,
      "learning_rate": 0.00042844468429573655,
      "loss": 3.5387,
      "step": 26600
    },
    {
      "epoch": 2.8733153638814017,
      "grad_norm": 0.5993890762329102,
      "learning_rate": 0.00042812088505126815,
      "loss": 3.5445,
      "step": 26650
    },
    {
      "epoch": 2.8787061994609164,
      "grad_norm": 0.5829864144325256,
      "learning_rate": 0.0004277970858067997,
      "loss": 3.5301,
      "step": 26700
    },
    {
      "epoch": 2.884097035040431,
      "grad_norm": 0.6206557154655457,
      "learning_rate": 0.0004274732865623313,
      "loss": 3.5432,
      "step": 26750
    },
    {
      "epoch": 2.889487870619946,
      "grad_norm": 0.5877953171730042,
      "learning_rate": 0.00042714948731786286,
      "loss": 3.5212,
      "step": 26800
    },
    {
      "epoch": 2.894878706199461,
      "grad_norm": 0.6274372935295105,
      "learning_rate": 0.00042682568807339447,
      "loss": 3.5414,
      "step": 26850
    },
    {
      "epoch": 2.9002695417789757,
      "grad_norm": 0.5919215679168701,
      "learning_rate": 0.000426501888828926,
      "loss": 3.5075,
      "step": 26900
    },
    {
      "epoch": 2.9056603773584904,
      "grad_norm": 0.5824446678161621,
      "learning_rate": 0.0004261780895844576,
      "loss": 3.5275,
      "step": 26950
    },
    {
      "epoch": 2.9110512129380055,
      "grad_norm": 0.5630289316177368,
      "learning_rate": 0.00042585429033998917,
      "loss": 3.5433,
      "step": 27000
    },
    {
      "epoch": 2.9110512129380055,
      "eval_accuracy": 0.3697433868691464,
      "eval_loss": 3.5023281574249268,
      "eval_runtime": 183.854,
      "eval_samples_per_second": 97.964,
      "eval_steps_per_second": 6.124,
      "step": 27000
    },
    {
      "epoch": 2.91644204851752,
      "grad_norm": 0.5719215273857117,
      "learning_rate": 0.0004255304910955207,
      "loss": 3.5288,
      "step": 27050
    },
    {
      "epoch": 2.921832884097035,
      "grad_norm": 0.5788906216621399,
      "learning_rate": 0.0004252066918510523,
      "loss": 3.5314,
      "step": 27100
    },
    {
      "epoch": 2.92722371967655,
      "grad_norm": 0.6185005903244019,
      "learning_rate": 0.0004248828926065839,
      "loss": 3.5273,
      "step": 27150
    },
    {
      "epoch": 2.9326145552560647,
      "grad_norm": 0.6660767793655396,
      "learning_rate": 0.0004245590933621155,
      "loss": 3.516,
      "step": 27200
    },
    {
      "epoch": 2.9380053908355794,
      "grad_norm": 0.5727997422218323,
      "learning_rate": 0.00042423529411764703,
      "loss": 3.5295,
      "step": 27250
    },
    {
      "epoch": 2.9433962264150946,
      "grad_norm": 0.5961682200431824,
      "learning_rate": 0.00042391149487317864,
      "loss": 3.5323,
      "step": 27300
    },
    {
      "epoch": 2.9487870619946093,
      "grad_norm": 0.6300620436668396,
      "learning_rate": 0.00042358769562871013,
      "loss": 3.5177,
      "step": 27350
    },
    {
      "epoch": 2.954177897574124,
      "grad_norm": 0.6043277978897095,
      "learning_rate": 0.00042326389638424174,
      "loss": 3.5079,
      "step": 27400
    },
    {
      "epoch": 2.9595687331536387,
      "grad_norm": 0.5596343874931335,
      "learning_rate": 0.0004229400971397733,
      "loss": 3.5159,
      "step": 27450
    },
    {
      "epoch": 2.964959568733154,
      "grad_norm": 0.5775852203369141,
      "learning_rate": 0.00042261629789530484,
      "loss": 3.5263,
      "step": 27500
    },
    {
      "epoch": 2.9703504043126685,
      "grad_norm": 0.5495153069496155,
      "learning_rate": 0.00042229249865083644,
      "loss": 3.5161,
      "step": 27550
    },
    {
      "epoch": 2.975741239892183,
      "grad_norm": 0.5802893042564392,
      "learning_rate": 0.000421968699406368,
      "loss": 3.5265,
      "step": 27600
    },
    {
      "epoch": 2.981132075471698,
      "grad_norm": 0.6191282272338867,
      "learning_rate": 0.0004216449001618996,
      "loss": 3.5039,
      "step": 27650
    },
    {
      "epoch": 2.986522911051213,
      "grad_norm": 0.5658198595046997,
      "learning_rate": 0.00042132110091743115,
      "loss": 3.5147,
      "step": 27700
    },
    {
      "epoch": 2.9919137466307277,
      "grad_norm": 0.6845032572746277,
      "learning_rate": 0.00042099730167296275,
      "loss": 3.521,
      "step": 27750
    },
    {
      "epoch": 2.9973045822102424,
      "grad_norm": 0.5844449400901794,
      "learning_rate": 0.0004206735024284943,
      "loss": 3.5218,
      "step": 27800
    },
    {
      "epoch": 3.0026954177897576,
      "grad_norm": 0.5795214772224426,
      "learning_rate": 0.0004203497031840259,
      "loss": 3.4854,
      "step": 27850
    },
    {
      "epoch": 3.0080862533692723,
      "grad_norm": 0.7003495693206787,
      "learning_rate": 0.00042002590393955746,
      "loss": 3.4273,
      "step": 27900
    },
    {
      "epoch": 3.013477088948787,
      "grad_norm": 0.549725353717804,
      "learning_rate": 0.00041970210469508896,
      "loss": 3.431,
      "step": 27950
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.5706000328063965,
      "learning_rate": 0.00041937830545062056,
      "loss": 3.4434,
      "step": 28000
    },
    {
      "epoch": 3.018867924528302,
      "eval_accuracy": 0.37032381059585134,
      "eval_loss": 3.499523639678955,
      "eval_runtime": 183.7469,
      "eval_samples_per_second": 98.021,
      "eval_steps_per_second": 6.128,
      "step": 28000
    },
    {
      "epoch": 3.024258760107817,
      "grad_norm": 0.6284692883491516,
      "learning_rate": 0.0004190545062061521,
      "loss": 3.4299,
      "step": 28050
    },
    {
      "epoch": 3.0296495956873315,
      "grad_norm": 0.6581982374191284,
      "learning_rate": 0.0004187307069616837,
      "loss": 3.4238,
      "step": 28100
    },
    {
      "epoch": 3.035040431266846,
      "grad_norm": 0.6106514930725098,
      "learning_rate": 0.00041840690771721527,
      "loss": 3.4642,
      "step": 28150
    },
    {
      "epoch": 3.0404312668463613,
      "grad_norm": 0.5879032611846924,
      "learning_rate": 0.00041808310847274687,
      "loss": 3.4451,
      "step": 28200
    },
    {
      "epoch": 3.045822102425876,
      "grad_norm": 0.5639373660087585,
      "learning_rate": 0.0004177593092282784,
      "loss": 3.4552,
      "step": 28250
    },
    {
      "epoch": 3.0512129380053907,
      "grad_norm": 0.593923807144165,
      "learning_rate": 0.00041743550998381,
      "loss": 3.4516,
      "step": 28300
    },
    {
      "epoch": 3.056603773584906,
      "grad_norm": 0.6003124117851257,
      "learning_rate": 0.0004171117107393416,
      "loss": 3.4434,
      "step": 28350
    },
    {
      "epoch": 3.0619946091644206,
      "grad_norm": 0.5810577869415283,
      "learning_rate": 0.00041678791149487313,
      "loss": 3.42,
      "step": 28400
    },
    {
      "epoch": 3.0673854447439353,
      "grad_norm": 0.606003999710083,
      "learning_rate": 0.00041646411225040473,
      "loss": 3.4639,
      "step": 28450
    },
    {
      "epoch": 3.07277628032345,
      "grad_norm": 0.6627290844917297,
      "learning_rate": 0.0004161403130059363,
      "loss": 3.4576,
      "step": 28500
    },
    {
      "epoch": 3.078167115902965,
      "grad_norm": 0.615278959274292,
      "learning_rate": 0.0004158165137614679,
      "loss": 3.4439,
      "step": 28550
    },
    {
      "epoch": 3.08355795148248,
      "grad_norm": 0.5780900716781616,
      "learning_rate": 0.00041549271451699944,
      "loss": 3.4401,
      "step": 28600
    },
    {
      "epoch": 3.0889487870619945,
      "grad_norm": 0.6600579023361206,
      "learning_rate": 0.00041516891527253104,
      "loss": 3.4401,
      "step": 28650
    },
    {
      "epoch": 3.0943396226415096,
      "grad_norm": 0.5386844277381897,
      "learning_rate": 0.00041484511602806254,
      "loss": 3.45,
      "step": 28700
    },
    {
      "epoch": 3.0997304582210243,
      "grad_norm": 0.5708300471305847,
      "learning_rate": 0.0004145213167835941,
      "loss": 3.4373,
      "step": 28750
    },
    {
      "epoch": 3.105121293800539,
      "grad_norm": 0.5719059109687805,
      "learning_rate": 0.0004141975175391257,
      "loss": 3.4413,
      "step": 28800
    },
    {
      "epoch": 3.1105121293800537,
      "grad_norm": 0.5947633385658264,
      "learning_rate": 0.00041387371829465725,
      "loss": 3.4252,
      "step": 28850
    },
    {
      "epoch": 3.115902964959569,
      "grad_norm": 0.6137687563896179,
      "learning_rate": 0.00041354991905018885,
      "loss": 3.4438,
      "step": 28900
    },
    {
      "epoch": 3.1212938005390836,
      "grad_norm": 0.6112726330757141,
      "learning_rate": 0.0004132261198057204,
      "loss": 3.4593,
      "step": 28950
    },
    {
      "epoch": 3.1266846361185983,
      "grad_norm": 0.6146498918533325,
      "learning_rate": 0.000412902320561252,
      "loss": 3.4286,
      "step": 29000
    },
    {
      "epoch": 3.1266846361185983,
      "eval_accuracy": 0.371247034183611,
      "eval_loss": 3.4940013885498047,
      "eval_runtime": 183.9217,
      "eval_samples_per_second": 97.928,
      "eval_steps_per_second": 6.122,
      "step": 29000
    },
    {
      "epoch": 3.1320754716981134,
      "grad_norm": 0.6162520051002502,
      "learning_rate": 0.00041258499730167296,
      "loss": 3.4511,
      "step": 29050
    },
    {
      "epoch": 3.137466307277628,
      "grad_norm": 0.6008721590042114,
      "learning_rate": 0.0004122611980572045,
      "loss": 3.4405,
      "step": 29100
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.5806031227111816,
      "learning_rate": 0.00041193739881273606,
      "loss": 3.4575,
      "step": 29150
    },
    {
      "epoch": 3.1482479784366575,
      "grad_norm": 0.5788300037384033,
      "learning_rate": 0.00041161359956826766,
      "loss": 3.4583,
      "step": 29200
    },
    {
      "epoch": 3.1536388140161726,
      "grad_norm": 0.6230771541595459,
      "learning_rate": 0.0004112898003237992,
      "loss": 3.4449,
      "step": 29250
    },
    {
      "epoch": 3.1590296495956873,
      "grad_norm": 0.5821077227592468,
      "learning_rate": 0.0004109660010793308,
      "loss": 3.4518,
      "step": 29300
    },
    {
      "epoch": 3.164420485175202,
      "grad_norm": 0.6511071920394897,
      "learning_rate": 0.0004106422018348623,
      "loss": 3.4639,
      "step": 29350
    },
    {
      "epoch": 3.169811320754717,
      "grad_norm": 0.5678673982620239,
      "learning_rate": 0.0004103184025903939,
      "loss": 3.4581,
      "step": 29400
    },
    {
      "epoch": 3.175202156334232,
      "grad_norm": 0.593774139881134,
      "learning_rate": 0.00040999460334592547,
      "loss": 3.4671,
      "step": 29450
    },
    {
      "epoch": 3.1805929919137466,
      "grad_norm": 0.5744243860244751,
      "learning_rate": 0.000409670804101457,
      "loss": 3.4629,
      "step": 29500
    },
    {
      "epoch": 3.1859838274932613,
      "grad_norm": 0.8164382576942444,
      "learning_rate": 0.00040934700485698863,
      "loss": 3.4392,
      "step": 29550
    },
    {
      "epoch": 3.1913746630727764,
      "grad_norm": 0.6126202940940857,
      "learning_rate": 0.0004090232056125202,
      "loss": 3.4606,
      "step": 29600
    },
    {
      "epoch": 3.196765498652291,
      "grad_norm": 0.5741614103317261,
      "learning_rate": 0.0004086994063680518,
      "loss": 3.4472,
      "step": 29650
    },
    {
      "epoch": 3.202156334231806,
      "grad_norm": 0.5594901442527771,
      "learning_rate": 0.00040837560712358333,
      "loss": 3.4545,
      "step": 29700
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.5645290613174438,
      "learning_rate": 0.00040805180787911494,
      "loss": 3.4618,
      "step": 29750
    },
    {
      "epoch": 3.2129380053908356,
      "grad_norm": 0.5808700919151306,
      "learning_rate": 0.0004077280086346465,
      "loss": 3.4444,
      "step": 29800
    },
    {
      "epoch": 3.2183288409703503,
      "grad_norm": 0.6301125288009644,
      "learning_rate": 0.0004074042093901781,
      "loss": 3.4651,
      "step": 29850
    },
    {
      "epoch": 3.223719676549865,
      "grad_norm": 0.5941428542137146,
      "learning_rate": 0.00040708041014570964,
      "loss": 3.4735,
      "step": 29900
    },
    {
      "epoch": 3.22911051212938,
      "grad_norm": 0.5769610404968262,
      "learning_rate": 0.00040675661090124114,
      "loss": 3.4638,
      "step": 29950
    },
    {
      "epoch": 3.234501347708895,
      "grad_norm": 0.6222258806228638,
      "learning_rate": 0.0004064328116567728,
      "loss": 3.4506,
      "step": 30000
    },
    {
      "epoch": 3.234501347708895,
      "eval_accuracy": 0.3719082956588499,
      "eval_loss": 3.4874448776245117,
      "eval_runtime": 183.6565,
      "eval_samples_per_second": 98.069,
      "eval_steps_per_second": 6.131,
      "step": 30000
    },
    {
      "epoch": 3.2398921832884096,
      "grad_norm": 0.6240214109420776,
      "learning_rate": 0.0004061090124123043,
      "loss": 3.4467,
      "step": 30050
    },
    {
      "epoch": 3.2452830188679247,
      "grad_norm": 0.608329713344574,
      "learning_rate": 0.0004057852131678359,
      "loss": 3.4485,
      "step": 30100
    },
    {
      "epoch": 3.2506738544474394,
      "grad_norm": 0.5454065799713135,
      "learning_rate": 0.00040546141392336745,
      "loss": 3.4499,
      "step": 30150
    },
    {
      "epoch": 3.256064690026954,
      "grad_norm": 0.667070209980011,
      "learning_rate": 0.00040513761467889906,
      "loss": 3.4693,
      "step": 30200
    },
    {
      "epoch": 3.2614555256064692,
      "grad_norm": 0.6167458295822144,
      "learning_rate": 0.0004048138154344306,
      "loss": 3.4631,
      "step": 30250
    },
    {
      "epoch": 3.266846361185984,
      "grad_norm": 0.7352801561355591,
      "learning_rate": 0.0004044900161899622,
      "loss": 3.4654,
      "step": 30300
    },
    {
      "epoch": 3.2722371967654986,
      "grad_norm": 0.6545478701591492,
      "learning_rate": 0.00040416621694549376,
      "loss": 3.4401,
      "step": 30350
    },
    {
      "epoch": 3.2776280323450133,
      "grad_norm": 0.6410787105560303,
      "learning_rate": 0.0004038424177010253,
      "loss": 3.4697,
      "step": 30400
    },
    {
      "epoch": 3.2830188679245285,
      "grad_norm": 0.5979618430137634,
      "learning_rate": 0.0004035186184565569,
      "loss": 3.4568,
      "step": 30450
    },
    {
      "epoch": 3.288409703504043,
      "grad_norm": 0.6123748421669006,
      "learning_rate": 0.00040319481921208847,
      "loss": 3.4417,
      "step": 30500
    },
    {
      "epoch": 3.293800539083558,
      "grad_norm": 0.607407808303833,
      "learning_rate": 0.00040287101996762007,
      "loss": 3.4708,
      "step": 30550
    },
    {
      "epoch": 3.2991913746630726,
      "grad_norm": 0.7021310329437256,
      "learning_rate": 0.0004025472207231516,
      "loss": 3.4618,
      "step": 30600
    },
    {
      "epoch": 3.3045822102425877,
      "grad_norm": 0.6325151920318604,
      "learning_rate": 0.00040222342147868323,
      "loss": 3.4408,
      "step": 30650
    },
    {
      "epoch": 3.3099730458221024,
      "grad_norm": 0.6190793514251709,
      "learning_rate": 0.0004018996222342147,
      "loss": 3.4592,
      "step": 30700
    },
    {
      "epoch": 3.315363881401617,
      "grad_norm": 0.5823326110839844,
      "learning_rate": 0.00040157582298974633,
      "loss": 3.4269,
      "step": 30750
    },
    {
      "epoch": 3.3207547169811322,
      "grad_norm": 0.5816695690155029,
      "learning_rate": 0.0004012520237452779,
      "loss": 3.4634,
      "step": 30800
    },
    {
      "epoch": 3.326145552560647,
      "grad_norm": 0.6027852892875671,
      "learning_rate": 0.00040092822450080943,
      "loss": 3.4441,
      "step": 30850
    },
    {
      "epoch": 3.3315363881401616,
      "grad_norm": 0.6045222878456116,
      "learning_rate": 0.00040060442525634103,
      "loss": 3.4673,
      "step": 30900
    },
    {
      "epoch": 3.3369272237196768,
      "grad_norm": 0.597585141658783,
      "learning_rate": 0.0004002806260118726,
      "loss": 3.4441,
      "step": 30950
    },
    {
      "epoch": 3.3423180592991915,
      "grad_norm": 0.6446326375007629,
      "learning_rate": 0.0003999568267674042,
      "loss": 3.4563,
      "step": 31000
    },
    {
      "epoch": 3.3423180592991915,
      "eval_accuracy": 0.3723845212661408,
      "eval_loss": 3.4811089038848877,
      "eval_runtime": 183.6383,
      "eval_samples_per_second": 98.079,
      "eval_steps_per_second": 6.132,
      "step": 31000
    },
    {
      "epoch": 3.347708894878706,
      "grad_norm": 0.6195496916770935,
      "learning_rate": 0.00039963302752293574,
      "loss": 3.4571,
      "step": 31050
    },
    {
      "epoch": 3.353099730458221,
      "grad_norm": 0.6132037043571472,
      "learning_rate": 0.00039930922827846735,
      "loss": 3.4524,
      "step": 31100
    },
    {
      "epoch": 3.358490566037736,
      "grad_norm": 0.5746621489524841,
      "learning_rate": 0.0003989854290339989,
      "loss": 3.463,
      "step": 31150
    },
    {
      "epoch": 3.3638814016172507,
      "grad_norm": 0.596961498260498,
      "learning_rate": 0.00039866810577441985,
      "loss": 3.4495,
      "step": 31200
    },
    {
      "epoch": 3.3692722371967654,
      "grad_norm": 0.5609466433525085,
      "learning_rate": 0.0003983443065299514,
      "loss": 3.4515,
      "step": 31250
    },
    {
      "epoch": 3.37466307277628,
      "grad_norm": 0.5683972835540771,
      "learning_rate": 0.000398020507285483,
      "loss": 3.4771,
      "step": 31300
    },
    {
      "epoch": 3.3800539083557952,
      "grad_norm": 0.6360358595848083,
      "learning_rate": 0.0003976967080410145,
      "loss": 3.4486,
      "step": 31350
    },
    {
      "epoch": 3.38544474393531,
      "grad_norm": 0.6113670468330383,
      "learning_rate": 0.0003973729087965461,
      "loss": 3.448,
      "step": 31400
    },
    {
      "epoch": 3.3908355795148246,
      "grad_norm": 0.5796744227409363,
      "learning_rate": 0.00039704910955207765,
      "loss": 3.4534,
      "step": 31450
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.5753939151763916,
      "learning_rate": 0.00039672531030760926,
      "loss": 3.4462,
      "step": 31500
    },
    {
      "epoch": 3.4016172506738545,
      "grad_norm": 0.6340164542198181,
      "learning_rate": 0.0003964015110631408,
      "loss": 3.4674,
      "step": 31550
    },
    {
      "epoch": 3.407008086253369,
      "grad_norm": 0.5975732207298279,
      "learning_rate": 0.00039607771181867236,
      "loss": 3.445,
      "step": 31600
    },
    {
      "epoch": 3.4123989218328843,
      "grad_norm": 0.5981366634368896,
      "learning_rate": 0.00039575391257420397,
      "loss": 3.4462,
      "step": 31650
    },
    {
      "epoch": 3.417789757412399,
      "grad_norm": 0.6517634987831116,
      "learning_rate": 0.0003954301133297355,
      "loss": 3.4725,
      "step": 31700
    },
    {
      "epoch": 3.4231805929919137,
      "grad_norm": 0.6406164169311523,
      "learning_rate": 0.0003951063140852671,
      "loss": 3.4464,
      "step": 31750
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.6119227409362793,
      "learning_rate": 0.00039478251484079867,
      "loss": 3.4483,
      "step": 31800
    },
    {
      "epoch": 3.4339622641509435,
      "grad_norm": 0.6088457107543945,
      "learning_rate": 0.0003944587155963303,
      "loss": 3.4796,
      "step": 31850
    },
    {
      "epoch": 3.439353099730458,
      "grad_norm": 0.6567962169647217,
      "learning_rate": 0.0003941349163518618,
      "loss": 3.4615,
      "step": 31900
    },
    {
      "epoch": 3.444743935309973,
      "grad_norm": 0.6715599894523621,
      "learning_rate": 0.00039381111710739343,
      "loss": 3.4414,
      "step": 31950
    },
    {
      "epoch": 3.450134770889488,
      "grad_norm": 0.6421747803688049,
      "learning_rate": 0.000393487317862925,
      "loss": 3.4568,
      "step": 32000
    },
    {
      "epoch": 3.450134770889488,
      "eval_accuracy": 0.3734006430729818,
      "eval_loss": 3.473811388015747,
      "eval_runtime": 183.9,
      "eval_samples_per_second": 97.939,
      "eval_steps_per_second": 6.123,
      "step": 32000
    },
    {
      "epoch": 3.4555256064690028,
      "grad_norm": 0.593186616897583,
      "learning_rate": 0.0003931635186184565,
      "loss": 3.4593,
      "step": 32050
    },
    {
      "epoch": 3.4609164420485174,
      "grad_norm": 0.697148859500885,
      "learning_rate": 0.0003928397193739881,
      "loss": 3.453,
      "step": 32100
    },
    {
      "epoch": 3.466307277628032,
      "grad_norm": 0.6392279863357544,
      "learning_rate": 0.00039251592012951963,
      "loss": 3.4662,
      "step": 32150
    },
    {
      "epoch": 3.4716981132075473,
      "grad_norm": 0.5713638663291931,
      "learning_rate": 0.00039219212088505124,
      "loss": 3.4484,
      "step": 32200
    },
    {
      "epoch": 3.477088948787062,
      "grad_norm": 0.6339240074157715,
      "learning_rate": 0.0003918683216405828,
      "loss": 3.4653,
      "step": 32250
    },
    {
      "epoch": 3.4824797843665767,
      "grad_norm": 0.6048734784126282,
      "learning_rate": 0.0003915445223961144,
      "loss": 3.4757,
      "step": 32300
    },
    {
      "epoch": 3.487870619946092,
      "grad_norm": 0.616313099861145,
      "learning_rate": 0.00039122072315164594,
      "loss": 3.4397,
      "step": 32350
    },
    {
      "epoch": 3.4932614555256065,
      "grad_norm": 0.6406932473182678,
      "learning_rate": 0.0003908969239071775,
      "loss": 3.4706,
      "step": 32400
    },
    {
      "epoch": 3.498652291105121,
      "grad_norm": 0.6056227087974548,
      "learning_rate": 0.0003905731246627091,
      "loss": 3.4449,
      "step": 32450
    },
    {
      "epoch": 3.5040431266846364,
      "grad_norm": 0.5830230712890625,
      "learning_rate": 0.00039024932541824065,
      "loss": 3.4704,
      "step": 32500
    },
    {
      "epoch": 3.509433962264151,
      "grad_norm": 0.5504217147827148,
      "learning_rate": 0.00038992552617377225,
      "loss": 3.4591,
      "step": 32550
    },
    {
      "epoch": 3.5148247978436657,
      "grad_norm": 0.586064875125885,
      "learning_rate": 0.0003896017269293038,
      "loss": 3.4541,
      "step": 32600
    },
    {
      "epoch": 3.5202156334231804,
      "grad_norm": 0.6093249917030334,
      "learning_rate": 0.0003892779276848354,
      "loss": 3.4655,
      "step": 32650
    },
    {
      "epoch": 3.525606469002695,
      "grad_norm": 0.6133269667625427,
      "learning_rate": 0.0003889541284403669,
      "loss": 3.4392,
      "step": 32700
    },
    {
      "epoch": 3.5309973045822103,
      "grad_norm": 0.6184484958648682,
      "learning_rate": 0.0003886303291958985,
      "loss": 3.4513,
      "step": 32750
    },
    {
      "epoch": 3.536388140161725,
      "grad_norm": 0.6110262870788574,
      "learning_rate": 0.00038830652995143006,
      "loss": 3.4623,
      "step": 32800
    },
    {
      "epoch": 3.5417789757412397,
      "grad_norm": 0.6329076886177063,
      "learning_rate": 0.0003879827307069616,
      "loss": 3.4245,
      "step": 32850
    },
    {
      "epoch": 3.547169811320755,
      "grad_norm": 0.6302708387374878,
      "learning_rate": 0.0003876589314624932,
      "loss": 3.4471,
      "step": 32900
    },
    {
      "epoch": 3.5525606469002695,
      "grad_norm": 0.6204167008399963,
      "learning_rate": 0.00038733513221802477,
      "loss": 3.4626,
      "step": 32950
    },
    {
      "epoch": 3.557951482479784,
      "grad_norm": 0.635398805141449,
      "learning_rate": 0.00038701133297355637,
      "loss": 3.4606,
      "step": 33000
    },
    {
      "epoch": 3.557951482479784,
      "eval_accuracy": 0.37382395472390706,
      "eval_loss": 3.468949556350708,
      "eval_runtime": 183.5943,
      "eval_samples_per_second": 98.102,
      "eval_steps_per_second": 6.133,
      "step": 33000
    },
    {
      "epoch": 3.5633423180592994,
      "grad_norm": 0.5870746374130249,
      "learning_rate": 0.0003866875337290879,
      "loss": 3.4628,
      "step": 33050
    },
    {
      "epoch": 3.568733153638814,
      "grad_norm": 0.6441285014152527,
      "learning_rate": 0.00038636373448461953,
      "loss": 3.4496,
      "step": 33100
    },
    {
      "epoch": 3.5741239892183287,
      "grad_norm": 0.6544851660728455,
      "learning_rate": 0.0003860399352401511,
      "loss": 3.4598,
      "step": 33150
    },
    {
      "epoch": 3.579514824797844,
      "grad_norm": 0.601290225982666,
      "learning_rate": 0.0003857161359956827,
      "loss": 3.4587,
      "step": 33200
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.6271600127220154,
      "learning_rate": 0.0003853988127361036,
      "loss": 3.4506,
      "step": 33250
    },
    {
      "epoch": 3.5902964959568733,
      "grad_norm": 0.6283750534057617,
      "learning_rate": 0.0003850750134916352,
      "loss": 3.4413,
      "step": 33300
    },
    {
      "epoch": 3.595687331536388,
      "grad_norm": 0.5692456364631653,
      "learning_rate": 0.0003847512142471667,
      "loss": 3.4524,
      "step": 33350
    },
    {
      "epoch": 3.601078167115903,
      "grad_norm": 0.6084824204444885,
      "learning_rate": 0.0003844274150026983,
      "loss": 3.4519,
      "step": 33400
    },
    {
      "epoch": 3.606469002695418,
      "grad_norm": 0.6203867197036743,
      "learning_rate": 0.00038410361575822984,
      "loss": 3.4636,
      "step": 33450
    },
    {
      "epoch": 3.6118598382749325,
      "grad_norm": 0.5996057987213135,
      "learning_rate": 0.00038377981651376144,
      "loss": 3.4679,
      "step": 33500
    },
    {
      "epoch": 3.617250673854447,
      "grad_norm": 0.5934234857559204,
      "learning_rate": 0.000383456017269293,
      "loss": 3.4491,
      "step": 33550
    },
    {
      "epoch": 3.6226415094339623,
      "grad_norm": 0.6062605381011963,
      "learning_rate": 0.00038313221802482454,
      "loss": 3.451,
      "step": 33600
    },
    {
      "epoch": 3.628032345013477,
      "grad_norm": 0.6698931455612183,
      "learning_rate": 0.00038280841878035615,
      "loss": 3.4472,
      "step": 33650
    },
    {
      "epoch": 3.6334231805929917,
      "grad_norm": 0.6485372185707092,
      "learning_rate": 0.0003824846195358877,
      "loss": 3.4337,
      "step": 33700
    },
    {
      "epoch": 3.638814016172507,
      "grad_norm": 0.6247066259384155,
      "learning_rate": 0.0003821608202914193,
      "loss": 3.4591,
      "step": 33750
    },
    {
      "epoch": 3.6442048517520216,
      "grad_norm": 0.6120551228523254,
      "learning_rate": 0.00038183702104695085,
      "loss": 3.4388,
      "step": 33800
    },
    {
      "epoch": 3.6495956873315363,
      "grad_norm": 0.6113671660423279,
      "learning_rate": 0.00038151322180248246,
      "loss": 3.4713,
      "step": 33850
    },
    {
      "epoch": 3.6549865229110514,
      "grad_norm": 0.5842575430870056,
      "learning_rate": 0.000381189422558014,
      "loss": 3.4447,
      "step": 33900
    },
    {
      "epoch": 3.660377358490566,
      "grad_norm": 0.6196116209030151,
      "learning_rate": 0.0003808656233135456,
      "loss": 3.4459,
      "step": 33950
    },
    {
      "epoch": 3.665768194070081,
      "grad_norm": 0.6134396195411682,
      "learning_rate": 0.00038054182406907716,
      "loss": 3.4583,
      "step": 34000
    },
    {
      "epoch": 3.665768194070081,
      "eval_accuracy": 0.3740601661020157,
      "eval_loss": 3.4624133110046387,
      "eval_runtime": 184.0257,
      "eval_samples_per_second": 97.872,
      "eval_steps_per_second": 6.119,
      "step": 34000
    },
    {
      "epoch": 3.671159029649596,
      "grad_norm": 0.5997918248176575,
      "learning_rate": 0.00038021802482460866,
      "loss": 3.4519,
      "step": 34050
    },
    {
      "epoch": 3.6765498652291106,
      "grad_norm": 0.5810784101486206,
      "learning_rate": 0.00037989422558014027,
      "loss": 3.4694,
      "step": 34100
    },
    {
      "epoch": 3.6819407008086253,
      "grad_norm": 0.6275520324707031,
      "learning_rate": 0.0003795704263356718,
      "loss": 3.4657,
      "step": 34150
    },
    {
      "epoch": 3.68733153638814,
      "grad_norm": 0.6059765815734863,
      "learning_rate": 0.0003792466270912034,
      "loss": 3.4458,
      "step": 34200
    },
    {
      "epoch": 3.6927223719676547,
      "grad_norm": 0.6004732251167297,
      "learning_rate": 0.00037892282784673497,
      "loss": 3.4397,
      "step": 34250
    },
    {
      "epoch": 3.69811320754717,
      "grad_norm": 0.6046389937400818,
      "learning_rate": 0.0003785990286022666,
      "loss": 3.4392,
      "step": 34300
    },
    {
      "epoch": 3.7035040431266846,
      "grad_norm": 0.6276936531066895,
      "learning_rate": 0.00037827522935779813,
      "loss": 3.4455,
      "step": 34350
    },
    {
      "epoch": 3.7088948787061993,
      "grad_norm": 0.5531956553459167,
      "learning_rate": 0.00037795143011332973,
      "loss": 3.4406,
      "step": 34400
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.576456606388092,
      "learning_rate": 0.0003776276308688613,
      "loss": 3.4619,
      "step": 34450
    },
    {
      "epoch": 3.719676549865229,
      "grad_norm": 0.5860700011253357,
      "learning_rate": 0.00037730383162439283,
      "loss": 3.4433,
      "step": 34500
    },
    {
      "epoch": 3.725067385444744,
      "grad_norm": 0.6148476004600525,
      "learning_rate": 0.00037698003237992444,
      "loss": 3.4385,
      "step": 34550
    },
    {
      "epoch": 3.730458221024259,
      "grad_norm": 0.6846517324447632,
      "learning_rate": 0.000376656233135456,
      "loss": 3.4564,
      "step": 34600
    },
    {
      "epoch": 3.7358490566037736,
      "grad_norm": 0.6337952613830566,
      "learning_rate": 0.0003763324338909876,
      "loss": 3.4555,
      "step": 34650
    },
    {
      "epoch": 3.7412398921832883,
      "grad_norm": 0.6312965154647827,
      "learning_rate": 0.0003760086346465191,
      "loss": 3.4656,
      "step": 34700
    },
    {
      "epoch": 3.7466307277628035,
      "grad_norm": 0.5984307527542114,
      "learning_rate": 0.0003756848354020507,
      "loss": 3.4617,
      "step": 34750
    },
    {
      "epoch": 3.752021563342318,
      "grad_norm": 0.5848837494850159,
      "learning_rate": 0.00037536103615758224,
      "loss": 3.4411,
      "step": 34800
    },
    {
      "epoch": 3.757412398921833,
      "grad_norm": 0.6520708799362183,
      "learning_rate": 0.0003750372369131138,
      "loss": 3.4426,
      "step": 34850
    },
    {
      "epoch": 3.7628032345013476,
      "grad_norm": 0.5814857482910156,
      "learning_rate": 0.0003747134376686454,
      "loss": 3.4595,
      "step": 34900
    },
    {
      "epoch": 3.7681940700808623,
      "grad_norm": 0.6383914351463318,
      "learning_rate": 0.00037438963842417695,
      "loss": 3.4509,
      "step": 34950
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.6446661353111267,
      "learning_rate": 0.00037406583917970856,
      "loss": 3.459,
      "step": 35000
    },
    {
      "epoch": 3.7735849056603774,
      "eval_accuracy": 0.3753658478548823,
      "eval_loss": 3.4548258781433105,
      "eval_runtime": 183.6058,
      "eval_samples_per_second": 98.096,
      "eval_steps_per_second": 6.133,
      "step": 35000
    },
    {
      "epoch": 3.778975741239892,
      "grad_norm": 0.6061877012252808,
      "learning_rate": 0.0003737420399352401,
      "loss": 3.4709,
      "step": 35050
    },
    {
      "epoch": 3.784366576819407,
      "grad_norm": 0.6421599984169006,
      "learning_rate": 0.0003734182406907717,
      "loss": 3.4359,
      "step": 35100
    },
    {
      "epoch": 3.789757412398922,
      "grad_norm": 0.6605613231658936,
      "learning_rate": 0.00037309444144630326,
      "loss": 3.457,
      "step": 35150
    },
    {
      "epoch": 3.7951482479784366,
      "grad_norm": 0.6076908111572266,
      "learning_rate": 0.00037277064220183487,
      "loss": 3.4568,
      "step": 35200
    },
    {
      "epoch": 3.8005390835579513,
      "grad_norm": 0.611754298210144,
      "learning_rate": 0.0003724468429573664,
      "loss": 3.4381,
      "step": 35250
    },
    {
      "epoch": 3.8059299191374665,
      "grad_norm": 0.6356371641159058,
      "learning_rate": 0.00037212951969778737,
      "loss": 3.4514,
      "step": 35300
    },
    {
      "epoch": 3.811320754716981,
      "grad_norm": 0.671958863735199,
      "learning_rate": 0.00037180572045331887,
      "loss": 3.4529,
      "step": 35350
    },
    {
      "epoch": 3.816711590296496,
      "grad_norm": 0.6135779023170471,
      "learning_rate": 0.00037148192120885047,
      "loss": 3.4536,
      "step": 35400
    },
    {
      "epoch": 3.822102425876011,
      "grad_norm": 0.683573305606842,
      "learning_rate": 0.000371158121964382,
      "loss": 3.4466,
      "step": 35450
    },
    {
      "epoch": 3.8274932614555257,
      "grad_norm": 0.5822563171386719,
      "learning_rate": 0.0003708343227199136,
      "loss": 3.4516,
      "step": 35500
    },
    {
      "epoch": 3.8328840970350404,
      "grad_norm": 0.6443066596984863,
      "learning_rate": 0.0003705105234754452,
      "loss": 3.4611,
      "step": 35550
    },
    {
      "epoch": 3.838274932614555,
      "grad_norm": 0.642926037311554,
      "learning_rate": 0.0003701867242309768,
      "loss": 3.4675,
      "step": 35600
    },
    {
      "epoch": 3.8436657681940702,
      "grad_norm": 0.6716592311859131,
      "learning_rate": 0.00036986292498650833,
      "loss": 3.4654,
      "step": 35650
    },
    {
      "epoch": 3.849056603773585,
      "grad_norm": 0.5893973112106323,
      "learning_rate": 0.0003695391257420399,
      "loss": 3.4669,
      "step": 35700
    },
    {
      "epoch": 3.8544474393530996,
      "grad_norm": 0.6722517013549805,
      "learning_rate": 0.0003692153264975715,
      "loss": 3.4484,
      "step": 35750
    },
    {
      "epoch": 3.8598382749326143,
      "grad_norm": 0.6753908395767212,
      "learning_rate": 0.00036889152725310304,
      "loss": 3.4549,
      "step": 35800
    },
    {
      "epoch": 3.8652291105121295,
      "grad_norm": 0.6143805980682373,
      "learning_rate": 0.00036856772800863464,
      "loss": 3.4301,
      "step": 35850
    },
    {
      "epoch": 3.870619946091644,
      "grad_norm": 0.6016655564308167,
      "learning_rate": 0.0003682439287641662,
      "loss": 3.4456,
      "step": 35900
    },
    {
      "epoch": 3.876010781671159,
      "grad_norm": 0.6141828298568726,
      "learning_rate": 0.0003679201295196978,
      "loss": 3.4712,
      "step": 35950
    },
    {
      "epoch": 3.881401617250674,
      "grad_norm": 0.5878786444664001,
      "learning_rate": 0.00036759633027522935,
      "loss": 3.4485,
      "step": 36000
    },
    {
      "epoch": 3.881401617250674,
      "eval_accuracy": 0.3757107121208055,
      "eval_loss": 3.449120044708252,
      "eval_runtime": 184.1528,
      "eval_samples_per_second": 97.805,
      "eval_steps_per_second": 6.114,
      "step": 36000
    },
    {
      "epoch": 3.8867924528301887,
      "grad_norm": 0.6707215309143066,
      "learning_rate": 0.00036727253103076084,
      "loss": 3.4586,
      "step": 36050
    },
    {
      "epoch": 3.8921832884097034,
      "grad_norm": 0.6566523909568787,
      "learning_rate": 0.00036694873178629245,
      "loss": 3.4523,
      "step": 36100
    },
    {
      "epoch": 3.8975741239892185,
      "grad_norm": 0.5984740257263184,
      "learning_rate": 0.000366624932541824,
      "loss": 3.4565,
      "step": 36150
    },
    {
      "epoch": 3.9029649595687332,
      "grad_norm": 0.5919893383979797,
      "learning_rate": 0.0003663011332973556,
      "loss": 3.4475,
      "step": 36200
    },
    {
      "epoch": 3.908355795148248,
      "grad_norm": 0.5963663458824158,
      "learning_rate": 0.00036597733405288715,
      "loss": 3.4628,
      "step": 36250
    },
    {
      "epoch": 3.913746630727763,
      "grad_norm": 0.6338704824447632,
      "learning_rate": 0.00036565353480841876,
      "loss": 3.441,
      "step": 36300
    },
    {
      "epoch": 3.9191374663072778,
      "grad_norm": 0.6194099187850952,
      "learning_rate": 0.0003653297355639503,
      "loss": 3.4647,
      "step": 36350
    },
    {
      "epoch": 3.9245283018867925,
      "grad_norm": 0.6529048681259155,
      "learning_rate": 0.0003650059363194819,
      "loss": 3.465,
      "step": 36400
    },
    {
      "epoch": 3.929919137466307,
      "grad_norm": 0.632722020149231,
      "learning_rate": 0.00036468213707501347,
      "loss": 3.4408,
      "step": 36450
    },
    {
      "epoch": 3.935309973045822,
      "grad_norm": 0.6227572560310364,
      "learning_rate": 0.000364358337830545,
      "loss": 3.4502,
      "step": 36500
    },
    {
      "epoch": 3.940700808625337,
      "grad_norm": 0.6491719484329224,
      "learning_rate": 0.0003640345385860766,
      "loss": 3.4509,
      "step": 36550
    },
    {
      "epoch": 3.9460916442048517,
      "grad_norm": 0.5961635112762451,
      "learning_rate": 0.00036371073934160817,
      "loss": 3.4478,
      "step": 36600
    },
    {
      "epoch": 3.9514824797843664,
      "grad_norm": 0.6824967265129089,
      "learning_rate": 0.0003633869400971398,
      "loss": 3.4401,
      "step": 36650
    },
    {
      "epoch": 3.9568733153638815,
      "grad_norm": 0.6491476893424988,
      "learning_rate": 0.00036306314085267127,
      "loss": 3.4488,
      "step": 36700
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.6334487199783325,
      "learning_rate": 0.00036273934160820293,
      "loss": 3.4413,
      "step": 36750
    },
    {
      "epoch": 3.967654986522911,
      "grad_norm": 0.6231988668441772,
      "learning_rate": 0.00036241554236373443,
      "loss": 3.4347,
      "step": 36800
    },
    {
      "epoch": 3.973045822102426,
      "grad_norm": 0.6323678493499756,
      "learning_rate": 0.00036209174311926603,
      "loss": 3.4487,
      "step": 36850
    },
    {
      "epoch": 3.9784366576819408,
      "grad_norm": 0.6366806030273438,
      "learning_rate": 0.0003617679438747976,
      "loss": 3.4424,
      "step": 36900
    },
    {
      "epoch": 3.9838274932614555,
      "grad_norm": 0.6505522131919861,
      "learning_rate": 0.00036144414463032913,
      "loss": 3.4522,
      "step": 36950
    },
    {
      "epoch": 3.9892183288409706,
      "grad_norm": 0.6528891921043396,
      "learning_rate": 0.00036112034538586074,
      "loss": 3.4463,
      "step": 37000
    },
    {
      "epoch": 3.9892183288409706,
      "eval_accuracy": 0.3762563669234098,
      "eval_loss": 3.4432127475738525,
      "eval_runtime": 183.5509,
      "eval_samples_per_second": 98.125,
      "eval_steps_per_second": 6.135,
      "step": 37000
    },
    {
      "epoch": 3.9946091644204853,
      "grad_norm": 0.6302518248558044,
      "learning_rate": 0.0003607965461413923,
      "loss": 3.4597,
      "step": 37050
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.2225829362869263,
      "learning_rate": 0.0003604727468969239,
      "loss": 3.4339,
      "step": 37100
    },
    {
      "epoch": 4.005390835579515,
      "grad_norm": 0.6290598511695862,
      "learning_rate": 0.00036014894765245544,
      "loss": 3.3275,
      "step": 37150
    },
    {
      "epoch": 4.010781671159029,
      "grad_norm": 0.6681774258613586,
      "learning_rate": 0.00035982514840798705,
      "loss": 3.3563,
      "step": 37200
    },
    {
      "epoch": 4.0161725067385445,
      "grad_norm": 0.6569148898124695,
      "learning_rate": 0.0003595013491635186,
      "loss": 3.3489,
      "step": 37250
    },
    {
      "epoch": 4.02156334231806,
      "grad_norm": 0.5988460183143616,
      "learning_rate": 0.0003591775499190502,
      "loss": 3.3536,
      "step": 37300
    },
    {
      "epoch": 4.026954177897574,
      "grad_norm": 0.6734127998352051,
      "learning_rate": 0.00035886022665947105,
      "loss": 3.3418,
      "step": 37350
    },
    {
      "epoch": 4.032345013477089,
      "grad_norm": 0.6245633363723755,
      "learning_rate": 0.0003585364274150027,
      "loss": 3.361,
      "step": 37400
    },
    {
      "epoch": 4.037735849056604,
      "grad_norm": 0.6742115616798401,
      "learning_rate": 0.0003582126281705342,
      "loss": 3.3584,
      "step": 37450
    },
    {
      "epoch": 4.0431266846361185,
      "grad_norm": 0.5911452770233154,
      "learning_rate": 0.0003578888289260658,
      "loss": 3.3485,
      "step": 37500
    },
    {
      "epoch": 4.048517520215634,
      "grad_norm": 0.6306824088096619,
      "learning_rate": 0.00035756502968159736,
      "loss": 3.3633,
      "step": 37550
    },
    {
      "epoch": 4.053908355795148,
      "grad_norm": 0.6810359358787537,
      "learning_rate": 0.00035724123043712896,
      "loss": 3.3582,
      "step": 37600
    },
    {
      "epoch": 4.059299191374663,
      "grad_norm": 0.6800394654273987,
      "learning_rate": 0.0003569174311926605,
      "loss": 3.3723,
      "step": 37650
    },
    {
      "epoch": 4.064690026954178,
      "grad_norm": 0.6404582262039185,
      "learning_rate": 0.00035659363194819206,
      "loss": 3.3558,
      "step": 37700
    },
    {
      "epoch": 4.070080862533692,
      "grad_norm": 0.7083536386489868,
      "learning_rate": 0.00035626983270372367,
      "loss": 3.3775,
      "step": 37750
    },
    {
      "epoch": 4.0754716981132075,
      "grad_norm": 0.6967944502830505,
      "learning_rate": 0.0003559460334592552,
      "loss": 3.3606,
      "step": 37800
    },
    {
      "epoch": 4.080862533692723,
      "grad_norm": 0.6590530872344971,
      "learning_rate": 0.0003556222342147868,
      "loss": 3.3595,
      "step": 37850
    },
    {
      "epoch": 4.086253369272237,
      "grad_norm": 0.6163628101348877,
      "learning_rate": 0.0003552984349703184,
      "loss": 3.3587,
      "step": 37900
    },
    {
      "epoch": 4.091644204851752,
      "grad_norm": 0.6308746337890625,
      "learning_rate": 0.00035497463572585,
      "loss": 3.3714,
      "step": 37950
    },
    {
      "epoch": 4.097035040431267,
      "grad_norm": 0.6695423722267151,
      "learning_rate": 0.00035465083648138153,
      "loss": 3.3631,
      "step": 38000
    },
    {
      "epoch": 4.097035040431267,
      "eval_accuracy": 0.37669934474702943,
      "eval_loss": 3.446185827255249,
      "eval_runtime": 183.7978,
      "eval_samples_per_second": 97.994,
      "eval_steps_per_second": 6.126,
      "step": 38000
    },
    {
      "epoch": 4.1024258760107815,
      "grad_norm": 0.6689504384994507,
      "learning_rate": 0.00035432703723691314,
      "loss": 3.3665,
      "step": 38050
    },
    {
      "epoch": 4.107816711590297,
      "grad_norm": 0.6203464865684509,
      "learning_rate": 0.00035400323799244463,
      "loss": 3.3684,
      "step": 38100
    },
    {
      "epoch": 4.113207547169812,
      "grad_norm": 0.6366583704948425,
      "learning_rate": 0.0003536794387479762,
      "loss": 3.3796,
      "step": 38150
    },
    {
      "epoch": 4.118598382749326,
      "grad_norm": 0.6430602669715881,
      "learning_rate": 0.0003533556395035078,
      "loss": 3.3595,
      "step": 38200
    },
    {
      "epoch": 4.123989218328841,
      "grad_norm": 0.6412113904953003,
      "learning_rate": 0.00035303184025903934,
      "loss": 3.3632,
      "step": 38250
    },
    {
      "epoch": 4.129380053908355,
      "grad_norm": 0.654319167137146,
      "learning_rate": 0.00035270804101457094,
      "loss": 3.3759,
      "step": 38300
    },
    {
      "epoch": 4.1347708894878705,
      "grad_norm": 0.6605011224746704,
      "learning_rate": 0.0003523842417701025,
      "loss": 3.3537,
      "step": 38350
    },
    {
      "epoch": 4.140161725067386,
      "grad_norm": 0.601024866104126,
      "learning_rate": 0.0003520604425256341,
      "loss": 3.3774,
      "step": 38400
    },
    {
      "epoch": 4.1455525606469,
      "grad_norm": 0.6261166930198669,
      "learning_rate": 0.00035173664328116565,
      "loss": 3.3613,
      "step": 38450
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.6527787446975708,
      "learning_rate": 0.00035141284403669725,
      "loss": 3.3736,
      "step": 38500
    },
    {
      "epoch": 4.15633423180593,
      "grad_norm": 0.6282969117164612,
      "learning_rate": 0.0003510890447922288,
      "loss": 3.3998,
      "step": 38550
    },
    {
      "epoch": 4.1617250673854445,
      "grad_norm": 0.6190927028656006,
      "learning_rate": 0.00035077172153264976,
      "loss": 3.3699,
      "step": 38600
    },
    {
      "epoch": 4.16711590296496,
      "grad_norm": 0.6161026358604431,
      "learning_rate": 0.0003504479222881813,
      "loss": 3.3852,
      "step": 38650
    },
    {
      "epoch": 4.172506738544475,
      "grad_norm": 0.7209351658821106,
      "learning_rate": 0.0003501241230437129,
      "loss": 3.3512,
      "step": 38700
    },
    {
      "epoch": 4.177897574123989,
      "grad_norm": 0.625611424446106,
      "learning_rate": 0.0003498003237992444,
      "loss": 3.3763,
      "step": 38750
    },
    {
      "epoch": 4.183288409703504,
      "grad_norm": 0.6233698129653931,
      "learning_rate": 0.000349476524554776,
      "loss": 3.3798,
      "step": 38800
    },
    {
      "epoch": 4.188679245283019,
      "grad_norm": 0.6900334358215332,
      "learning_rate": 0.00034915272531030756,
      "loss": 3.3821,
      "step": 38850
    },
    {
      "epoch": 4.1940700808625335,
      "grad_norm": 0.6348626613616943,
      "learning_rate": 0.0003488289260658391,
      "loss": 3.3889,
      "step": 38900
    },
    {
      "epoch": 4.199460916442049,
      "grad_norm": 0.6387774348258972,
      "learning_rate": 0.0003485051268213707,
      "loss": 3.3821,
      "step": 38950
    },
    {
      "epoch": 4.204851752021563,
      "grad_norm": 0.6410725712776184,
      "learning_rate": 0.00034818132757690227,
      "loss": 3.3635,
      "step": 39000
    },
    {
      "epoch": 4.204851752021563,
      "eval_accuracy": 0.3774232989385373,
      "eval_loss": 3.443272113800049,
      "eval_runtime": 183.5709,
      "eval_samples_per_second": 98.115,
      "eval_steps_per_second": 6.134,
      "step": 39000
    },
    {
      "epoch": 4.210242587601078,
      "grad_norm": 0.6588302850723267,
      "learning_rate": 0.0003478575283324339,
      "loss": 3.3933,
      "step": 39050
    },
    {
      "epoch": 4.215633423180593,
      "grad_norm": 0.6347865462303162,
      "learning_rate": 0.0003475337290879654,
      "loss": 3.3851,
      "step": 39100
    },
    {
      "epoch": 4.2210242587601075,
      "grad_norm": 0.6326488852500916,
      "learning_rate": 0.00034720992984349703,
      "loss": 3.3934,
      "step": 39150
    },
    {
      "epoch": 4.226415094339623,
      "grad_norm": 0.6076704263687134,
      "learning_rate": 0.0003468861305990286,
      "loss": 3.3747,
      "step": 39200
    },
    {
      "epoch": 4.231805929919138,
      "grad_norm": 0.6273050308227539,
      "learning_rate": 0.0003465623313545602,
      "loss": 3.3727,
      "step": 39250
    },
    {
      "epoch": 4.237196765498652,
      "grad_norm": 0.6443808674812317,
      "learning_rate": 0.00034623853211009173,
      "loss": 3.384,
      "step": 39300
    },
    {
      "epoch": 4.242587601078167,
      "grad_norm": 0.6146422624588013,
      "learning_rate": 0.00034591473286562323,
      "loss": 3.3702,
      "step": 39350
    },
    {
      "epoch": 4.247978436657682,
      "grad_norm": 0.6165491938591003,
      "learning_rate": 0.0003455909336211549,
      "loss": 3.3923,
      "step": 39400
    },
    {
      "epoch": 4.2533692722371965,
      "grad_norm": 0.6656982898712158,
      "learning_rate": 0.0003452671343766864,
      "loss": 3.3772,
      "step": 39450
    },
    {
      "epoch": 4.258760107816712,
      "grad_norm": 0.6853116750717163,
      "learning_rate": 0.000344943335132218,
      "loss": 3.3865,
      "step": 39500
    },
    {
      "epoch": 4.264150943396227,
      "grad_norm": 0.6323865056037903,
      "learning_rate": 0.00034461953588774954,
      "loss": 3.3894,
      "step": 39550
    },
    {
      "epoch": 4.269541778975741,
      "grad_norm": 0.6815229654312134,
      "learning_rate": 0.00034429573664328115,
      "loss": 3.3976,
      "step": 39600
    },
    {
      "epoch": 4.274932614555256,
      "grad_norm": 0.6554942727088928,
      "learning_rate": 0.0003439719373988127,
      "loss": 3.3929,
      "step": 39650
    },
    {
      "epoch": 4.280323450134771,
      "grad_norm": 0.7081801891326904,
      "learning_rate": 0.00034364813815434425,
      "loss": 3.3724,
      "step": 39700
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.6656792163848877,
      "learning_rate": 0.00034332433890987585,
      "loss": 3.383,
      "step": 39750
    },
    {
      "epoch": 4.291105121293801,
      "grad_norm": 0.6686912775039673,
      "learning_rate": 0.0003430005396654074,
      "loss": 3.3819,
      "step": 39800
    },
    {
      "epoch": 4.296495956873315,
      "grad_norm": 0.6758848428726196,
      "learning_rate": 0.000342676740420939,
      "loss": 3.3823,
      "step": 39850
    },
    {
      "epoch": 4.30188679245283,
      "grad_norm": 0.691242516040802,
      "learning_rate": 0.00034235294117647056,
      "loss": 3.3879,
      "step": 39900
    },
    {
      "epoch": 4.307277628032345,
      "grad_norm": 0.6893483400344849,
      "learning_rate": 0.00034202914193200216,
      "loss": 3.3772,
      "step": 39950
    },
    {
      "epoch": 4.3126684636118595,
      "grad_norm": 0.5819874405860901,
      "learning_rate": 0.0003417053426875337,
      "loss": 3.3887,
      "step": 40000
    },
    {
      "epoch": 4.3126684636118595,
      "eval_accuracy": 0.3775175009922725,
      "eval_loss": 3.4382338523864746,
      "eval_runtime": 184.2231,
      "eval_samples_per_second": 97.767,
      "eval_steps_per_second": 6.112,
      "step": 40000
    },
    {
      "epoch": 4.318059299191375,
      "grad_norm": 0.6500782370567322,
      "learning_rate": 0.0003413815434430653,
      "loss": 3.3864,
      "step": 40050
    },
    {
      "epoch": 4.32345013477089,
      "grad_norm": 0.6664853692054749,
      "learning_rate": 0.0003410577441985968,
      "loss": 3.3832,
      "step": 40100
    },
    {
      "epoch": 4.328840970350404,
      "grad_norm": 0.6425719857215881,
      "learning_rate": 0.00034073394495412837,
      "loss": 3.3954,
      "step": 40150
    },
    {
      "epoch": 4.334231805929919,
      "grad_norm": 0.6592696905136108,
      "learning_rate": 0.00034041014570965997,
      "loss": 3.3912,
      "step": 40200
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 0.6312816739082336,
      "learning_rate": 0.0003400863464651915,
      "loss": 3.3784,
      "step": 40250
    },
    {
      "epoch": 4.345013477088949,
      "grad_norm": 0.637238085269928,
      "learning_rate": 0.0003397625472207231,
      "loss": 3.3924,
      "step": 40300
    },
    {
      "epoch": 4.350404312668464,
      "grad_norm": 0.5931002497673035,
      "learning_rate": 0.0003394387479762547,
      "loss": 3.3745,
      "step": 40350
    },
    {
      "epoch": 4.355795148247978,
      "grad_norm": 0.6598531603813171,
      "learning_rate": 0.0003391149487317863,
      "loss": 3.3874,
      "step": 40400
    },
    {
      "epoch": 4.361185983827493,
      "grad_norm": 0.6362660527229309,
      "learning_rate": 0.00033879114948731783,
      "loss": 3.3833,
      "step": 40450
    },
    {
      "epoch": 4.366576819407008,
      "grad_norm": 0.623309850692749,
      "learning_rate": 0.00033846735024284944,
      "loss": 3.3982,
      "step": 40500
    },
    {
      "epoch": 4.3719676549865225,
      "grad_norm": 0.6290128231048584,
      "learning_rate": 0.000338143550998381,
      "loss": 3.3897,
      "step": 40550
    },
    {
      "epoch": 4.377358490566038,
      "grad_norm": 0.6916012167930603,
      "learning_rate": 0.00033781975175391254,
      "loss": 3.3787,
      "step": 40600
    },
    {
      "epoch": 4.382749326145553,
      "grad_norm": 0.6463659405708313,
      "learning_rate": 0.00033749595250944414,
      "loss": 3.3871,
      "step": 40650
    },
    {
      "epoch": 4.388140161725067,
      "grad_norm": 0.675667941570282,
      "learning_rate": 0.00033717215326497564,
      "loss": 3.3882,
      "step": 40700
    },
    {
      "epoch": 4.393530997304582,
      "grad_norm": 0.6485020518302917,
      "learning_rate": 0.0003368483540205073,
      "loss": 3.3871,
      "step": 40750
    },
    {
      "epoch": 4.398921832884097,
      "grad_norm": 0.7107369303703308,
      "learning_rate": 0.0003365245547760388,
      "loss": 3.3755,
      "step": 40800
    },
    {
      "epoch": 4.404312668463612,
      "grad_norm": 0.6821770668029785,
      "learning_rate": 0.0003362007555315704,
      "loss": 3.3906,
      "step": 40850
    },
    {
      "epoch": 4.409703504043127,
      "grad_norm": 0.6535558104515076,
      "learning_rate": 0.00033587695628710195,
      "loss": 3.3996,
      "step": 40900
    },
    {
      "epoch": 4.415094339622642,
      "grad_norm": 0.6308371424674988,
      "learning_rate": 0.00033555315704263355,
      "loss": 3.3959,
      "step": 40950
    },
    {
      "epoch": 4.420485175202156,
      "grad_norm": 0.5909251570701599,
      "learning_rate": 0.0003352293577981651,
      "loss": 3.392,
      "step": 41000
    },
    {
      "epoch": 4.420485175202156,
      "eval_accuracy": 0.37843866017516364,
      "eval_loss": 3.4319839477539062,
      "eval_runtime": 183.6983,
      "eval_samples_per_second": 98.047,
      "eval_steps_per_second": 6.13,
      "step": 41000
    },
    {
      "epoch": 4.425876010781671,
      "grad_norm": 0.6647576689720154,
      "learning_rate": 0.00033490555855369665,
      "loss": 3.4006,
      "step": 41050
    },
    {
      "epoch": 4.431266846361186,
      "grad_norm": 0.6007199883460999,
      "learning_rate": 0.00033458175930922826,
      "loss": 3.3697,
      "step": 41100
    },
    {
      "epoch": 4.436657681940701,
      "grad_norm": 0.6572251319885254,
      "learning_rate": 0.0003342579600647598,
      "loss": 3.3942,
      "step": 41150
    },
    {
      "epoch": 4.442048517520216,
      "grad_norm": 0.6462733745574951,
      "learning_rate": 0.0003339341608202914,
      "loss": 3.406,
      "step": 41200
    },
    {
      "epoch": 4.44743935309973,
      "grad_norm": 0.7077494859695435,
      "learning_rate": 0.00033361036157582297,
      "loss": 3.4034,
      "step": 41250
    },
    {
      "epoch": 4.452830188679245,
      "grad_norm": 0.6289212703704834,
      "learning_rate": 0.00033328656233135457,
      "loss": 3.3715,
      "step": 41300
    },
    {
      "epoch": 4.45822102425876,
      "grad_norm": 0.6613530516624451,
      "learning_rate": 0.0003329627630868861,
      "loss": 3.3796,
      "step": 41350
    },
    {
      "epoch": 4.463611859838275,
      "grad_norm": 0.653528094291687,
      "learning_rate": 0.0003326389638424177,
      "loss": 3.397,
      "step": 41400
    },
    {
      "epoch": 4.46900269541779,
      "grad_norm": 0.6634072065353394,
      "learning_rate": 0.0003323151645979492,
      "loss": 3.4074,
      "step": 41450
    },
    {
      "epoch": 4.474393530997305,
      "grad_norm": 0.643779456615448,
      "learning_rate": 0.00033199136535348077,
      "loss": 3.4053,
      "step": 41500
    },
    {
      "epoch": 4.479784366576819,
      "grad_norm": 0.6200483441352844,
      "learning_rate": 0.0003316675661090124,
      "loss": 3.3927,
      "step": 41550
    },
    {
      "epoch": 4.485175202156334,
      "grad_norm": 0.7240083813667297,
      "learning_rate": 0.00033134376686454393,
      "loss": 3.3854,
      "step": 41600
    },
    {
      "epoch": 4.490566037735849,
      "grad_norm": 0.6716325283050537,
      "learning_rate": 0.00033101996762007553,
      "loss": 3.398,
      "step": 41650
    },
    {
      "epoch": 4.495956873315364,
      "grad_norm": 0.676361083984375,
      "learning_rate": 0.0003306961683756071,
      "loss": 3.3881,
      "step": 41700
    },
    {
      "epoch": 4.501347708894879,
      "grad_norm": 0.6593371629714966,
      "learning_rate": 0.0003303723691311387,
      "loss": 3.3887,
      "step": 41750
    },
    {
      "epoch": 4.506738544474393,
      "grad_norm": 0.7852563261985779,
      "learning_rate": 0.00033004856988667024,
      "loss": 3.4041,
      "step": 41800
    },
    {
      "epoch": 4.512129380053908,
      "grad_norm": 0.651554524898529,
      "learning_rate": 0.0003297247706422018,
      "loss": 3.3813,
      "step": 41850
    },
    {
      "epoch": 4.517520215633423,
      "grad_norm": 0.61871337890625,
      "learning_rate": 0.0003294009713977334,
      "loss": 3.3981,
      "step": 41900
    },
    {
      "epoch": 4.5229110512129385,
      "grad_norm": 0.7037910223007202,
      "learning_rate": 0.00032907717215326494,
      "loss": 3.3993,
      "step": 41950
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.6677039265632629,
      "learning_rate": 0.00032875337290879655,
      "loss": 3.3866,
      "step": 42000
    },
    {
      "epoch": 4.528301886792453,
      "eval_accuracy": 0.37861880666316006,
      "eval_loss": 3.4264838695526123,
      "eval_runtime": 183.841,
      "eval_samples_per_second": 97.971,
      "eval_steps_per_second": 6.125,
      "step": 42000
    },
    {
      "epoch": 4.533692722371968,
      "grad_norm": 0.6433460116386414,
      "learning_rate": 0.00032842957366432805,
      "loss": 3.385,
      "step": 42050
    },
    {
      "epoch": 4.539083557951482,
      "grad_norm": 0.6643568277359009,
      "learning_rate": 0.0003281057744198597,
      "loss": 3.3937,
      "step": 42100
    },
    {
      "epoch": 4.544474393530997,
      "grad_norm": 0.6829423308372498,
      "learning_rate": 0.0003277819751753912,
      "loss": 3.393,
      "step": 42150
    },
    {
      "epoch": 4.549865229110512,
      "grad_norm": 0.6842221021652222,
      "learning_rate": 0.0003274581759309228,
      "loss": 3.3902,
      "step": 42200
    },
    {
      "epoch": 4.555256064690027,
      "grad_norm": 0.6358245015144348,
      "learning_rate": 0.00032713437668645436,
      "loss": 3.4025,
      "step": 42250
    },
    {
      "epoch": 4.560646900269542,
      "grad_norm": 0.7178597450256348,
      "learning_rate": 0.0003268105774419859,
      "loss": 3.38,
      "step": 42300
    },
    {
      "epoch": 4.566037735849057,
      "grad_norm": 0.6716541051864624,
      "learning_rate": 0.0003264867781975175,
      "loss": 3.4133,
      "step": 42350
    },
    {
      "epoch": 4.571428571428571,
      "grad_norm": 0.6570898294448853,
      "learning_rate": 0.00032616297895304906,
      "loss": 3.3732,
      "step": 42400
    },
    {
      "epoch": 4.576819407008086,
      "grad_norm": 0.6482057571411133,
      "learning_rate": 0.00032583917970858067,
      "loss": 3.3954,
      "step": 42450
    },
    {
      "epoch": 4.5822102425876015,
      "grad_norm": 0.6238645315170288,
      "learning_rate": 0.0003255153804641122,
      "loss": 3.4011,
      "step": 42500
    },
    {
      "epoch": 4.587601078167116,
      "grad_norm": 0.6898269653320312,
      "learning_rate": 0.0003251915812196438,
      "loss": 3.3792,
      "step": 42550
    },
    {
      "epoch": 4.592991913746631,
      "grad_norm": 0.6453298926353455,
      "learning_rate": 0.00032486778197517537,
      "loss": 3.4067,
      "step": 42600
    },
    {
      "epoch": 4.598382749326145,
      "grad_norm": 0.6815885901451111,
      "learning_rate": 0.000324543982730707,
      "loss": 3.3896,
      "step": 42650
    },
    {
      "epoch": 4.60377358490566,
      "grad_norm": 0.624555766582489,
      "learning_rate": 0.0003242266594711278,
      "loss": 3.3894,
      "step": 42700
    },
    {
      "epoch": 4.609164420485175,
      "grad_norm": 0.6440547704696655,
      "learning_rate": 0.0003239028602266595,
      "loss": 3.3971,
      "step": 42750
    },
    {
      "epoch": 4.6145552560646905,
      "grad_norm": 0.6493710875511169,
      "learning_rate": 0.000323579060982191,
      "loss": 3.3733,
      "step": 42800
    },
    {
      "epoch": 4.619946091644205,
      "grad_norm": 0.6572422981262207,
      "learning_rate": 0.0003232552617377226,
      "loss": 3.3984,
      "step": 42850
    },
    {
      "epoch": 4.62533692722372,
      "grad_norm": 0.6159889698028564,
      "learning_rate": 0.00032293146249325413,
      "loss": 3.3971,
      "step": 42900
    },
    {
      "epoch": 4.630727762803234,
      "grad_norm": 0.7039486169815063,
      "learning_rate": 0.00032260766324878574,
      "loss": 3.3921,
      "step": 42950
    },
    {
      "epoch": 4.636118598382749,
      "grad_norm": 0.681053876876831,
      "learning_rate": 0.0003222838640043173,
      "loss": 3.3958,
      "step": 43000
    },
    {
      "epoch": 4.636118598382749,
      "eval_accuracy": 0.3791799988287219,
      "eval_loss": 3.422306776046753,
      "eval_runtime": 183.9806,
      "eval_samples_per_second": 97.896,
      "eval_steps_per_second": 6.12,
      "step": 43000
    },
    {
      "epoch": 4.6415094339622645,
      "grad_norm": 0.6886916160583496,
      "learning_rate": 0.00032196006475984884,
      "loss": 3.3878,
      "step": 43050
    },
    {
      "epoch": 4.646900269541779,
      "grad_norm": 0.6496508121490479,
      "learning_rate": 0.00032163626551538044,
      "loss": 3.3883,
      "step": 43100
    },
    {
      "epoch": 4.652291105121294,
      "grad_norm": 0.6063190698623657,
      "learning_rate": 0.000321312466270912,
      "loss": 3.406,
      "step": 43150
    },
    {
      "epoch": 4.657681940700809,
      "grad_norm": 0.6703405380249023,
      "learning_rate": 0.00032099514301133295,
      "loss": 3.3759,
      "step": 43200
    },
    {
      "epoch": 4.663072776280323,
      "grad_norm": 0.6515152454376221,
      "learning_rate": 0.00032067134376686455,
      "loss": 3.3939,
      "step": 43250
    },
    {
      "epoch": 4.668463611859838,
      "grad_norm": 0.6611012816429138,
      "learning_rate": 0.0003203475445223961,
      "loss": 3.3818,
      "step": 43300
    },
    {
      "epoch": 4.6738544474393535,
      "grad_norm": 0.7121708393096924,
      "learning_rate": 0.0003200237452779277,
      "loss": 3.3905,
      "step": 43350
    },
    {
      "epoch": 4.679245283018868,
      "grad_norm": 0.6578240394592285,
      "learning_rate": 0.00031969994603345926,
      "loss": 3.3888,
      "step": 43400
    },
    {
      "epoch": 4.684636118598383,
      "grad_norm": 0.646023154258728,
      "learning_rate": 0.00031937614678899075,
      "loss": 3.3989,
      "step": 43450
    },
    {
      "epoch": 4.690026954177897,
      "grad_norm": 0.6373788714408875,
      "learning_rate": 0.00031905234754452236,
      "loss": 3.3742,
      "step": 43500
    },
    {
      "epoch": 4.695417789757412,
      "grad_norm": 0.6402143239974976,
      "learning_rate": 0.0003187285483000539,
      "loss": 3.3918,
      "step": 43550
    },
    {
      "epoch": 4.7008086253369274,
      "grad_norm": 0.7260469794273376,
      "learning_rate": 0.0003184047490555855,
      "loss": 3.4,
      "step": 43600
    },
    {
      "epoch": 4.706199460916442,
      "grad_norm": 0.657831072807312,
      "learning_rate": 0.00031808094981111706,
      "loss": 3.3965,
      "step": 43650
    },
    {
      "epoch": 4.711590296495957,
      "grad_norm": 0.6721468567848206,
      "learning_rate": 0.00031775715056664867,
      "loss": 3.3898,
      "step": 43700
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 0.6404152512550354,
      "learning_rate": 0.0003174333513221802,
      "loss": 3.407,
      "step": 43750
    },
    {
      "epoch": 4.722371967654986,
      "grad_norm": 0.6717345118522644,
      "learning_rate": 0.00031710955207771177,
      "loss": 3.3918,
      "step": 43800
    },
    {
      "epoch": 4.727762803234501,
      "grad_norm": 0.6559435725212097,
      "learning_rate": 0.0003167857528332434,
      "loss": 3.3983,
      "step": 43850
    },
    {
      "epoch": 4.7331536388140165,
      "grad_norm": 0.6558781862258911,
      "learning_rate": 0.0003164619535887749,
      "loss": 3.3887,
      "step": 43900
    },
    {
      "epoch": 4.738544474393531,
      "grad_norm": 0.6264235377311707,
      "learning_rate": 0.00031613815434430653,
      "loss": 3.3874,
      "step": 43950
    },
    {
      "epoch": 4.743935309973046,
      "grad_norm": 0.6540461778640747,
      "learning_rate": 0.0003158143550998381,
      "loss": 3.3875,
      "step": 44000
    },
    {
      "epoch": 4.743935309973046,
      "eval_accuracy": 0.3794974825669158,
      "eval_loss": 3.418121099472046,
      "eval_runtime": 183.5372,
      "eval_samples_per_second": 98.133,
      "eval_steps_per_second": 6.135,
      "step": 44000
    },
    {
      "epoch": 4.74932614555256,
      "grad_norm": 0.6550005078315735,
      "learning_rate": 0.0003154905558553697,
      "loss": 3.3855,
      "step": 44050
    },
    {
      "epoch": 4.754716981132075,
      "grad_norm": 0.6489050984382629,
      "learning_rate": 0.0003151667566109012,
      "loss": 3.3913,
      "step": 44100
    },
    {
      "epoch": 4.7601078167115904,
      "grad_norm": 0.7171805500984192,
      "learning_rate": 0.00031484295736643284,
      "loss": 3.386,
      "step": 44150
    },
    {
      "epoch": 4.765498652291106,
      "grad_norm": 0.7200286388397217,
      "learning_rate": 0.00031451915812196434,
      "loss": 3.3915,
      "step": 44200
    },
    {
      "epoch": 4.77088948787062,
      "grad_norm": 0.6571013927459717,
      "learning_rate": 0.0003141953588774959,
      "loss": 3.3739,
      "step": 44250
    },
    {
      "epoch": 4.776280323450135,
      "grad_norm": 0.7125272750854492,
      "learning_rate": 0.0003138715596330275,
      "loss": 3.3972,
      "step": 44300
    },
    {
      "epoch": 4.781671159029649,
      "grad_norm": 0.6634894013404846,
      "learning_rate": 0.00031354776038855904,
      "loss": 3.3768,
      "step": 44350
    },
    {
      "epoch": 4.787061994609164,
      "grad_norm": 0.6476519703865051,
      "learning_rate": 0.00031322396114409065,
      "loss": 3.4184,
      "step": 44400
    },
    {
      "epoch": 4.7924528301886795,
      "grad_norm": 0.6488552689552307,
      "learning_rate": 0.0003129001618996222,
      "loss": 3.3751,
      "step": 44450
    },
    {
      "epoch": 4.797843665768194,
      "grad_norm": 0.6651248335838318,
      "learning_rate": 0.0003125763626551538,
      "loss": 3.3902,
      "step": 44500
    },
    {
      "epoch": 4.803234501347709,
      "grad_norm": 0.6538376212120056,
      "learning_rate": 0.00031225256341068535,
      "loss": 3.3826,
      "step": 44550
    },
    {
      "epoch": 4.808625336927224,
      "grad_norm": 0.6490738391876221,
      "learning_rate": 0.00031192876416621696,
      "loss": 3.3966,
      "step": 44600
    },
    {
      "epoch": 4.814016172506738,
      "grad_norm": 0.7015892267227173,
      "learning_rate": 0.0003116049649217485,
      "loss": 3.3805,
      "step": 44650
    },
    {
      "epoch": 4.819407008086253,
      "grad_norm": 0.6848888993263245,
      "learning_rate": 0.00031128116567728,
      "loss": 3.3929,
      "step": 44700
    },
    {
      "epoch": 4.824797843665769,
      "grad_norm": 0.722858726978302,
      "learning_rate": 0.00031095736643281166,
      "loss": 3.3867,
      "step": 44750
    },
    {
      "epoch": 4.830188679245283,
      "grad_norm": 0.6747796535491943,
      "learning_rate": 0.00031063356718834316,
      "loss": 3.386,
      "step": 44800
    },
    {
      "epoch": 4.835579514824798,
      "grad_norm": 0.6656998991966248,
      "learning_rate": 0.00031030976794387476,
      "loss": 3.3898,
      "step": 44850
    },
    {
      "epoch": 4.840970350404312,
      "grad_norm": 0.6500997543334961,
      "learning_rate": 0.0003099859686994063,
      "loss": 3.3885,
      "step": 44900
    },
    {
      "epoch": 4.846361185983827,
      "grad_norm": 0.6383869647979736,
      "learning_rate": 0.0003096621694549379,
      "loss": 3.3816,
      "step": 44950
    },
    {
      "epoch": 4.8517520215633425,
      "grad_norm": 0.6764130592346191,
      "learning_rate": 0.00030933837021046947,
      "loss": 3.4055,
      "step": 45000
    },
    {
      "epoch": 4.8517520215633425,
      "eval_accuracy": 0.3802766324254334,
      "eval_loss": 3.4115288257598877,
      "eval_runtime": 184.1198,
      "eval_samples_per_second": 97.822,
      "eval_steps_per_second": 6.116,
      "step": 45000
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.7333670854568481,
      "learning_rate": 0.0003090145709660011,
      "loss": 3.3761,
      "step": 45050
    },
    {
      "epoch": 4.862533692722372,
      "grad_norm": 0.666333794593811,
      "learning_rate": 0.0003086907717215326,
      "loss": 3.3802,
      "step": 45100
    },
    {
      "epoch": 4.867924528301887,
      "grad_norm": 0.6561823487281799,
      "learning_rate": 0.0003083669724770642,
      "loss": 3.3913,
      "step": 45150
    },
    {
      "epoch": 4.873315363881401,
      "grad_norm": 0.6639450192451477,
      "learning_rate": 0.0003080431732325958,
      "loss": 3.3775,
      "step": 45200
    },
    {
      "epoch": 4.878706199460916,
      "grad_norm": 0.6635633707046509,
      "learning_rate": 0.00030771937398812733,
      "loss": 3.394,
      "step": 45250
    },
    {
      "epoch": 4.884097035040432,
      "grad_norm": 0.6871430277824402,
      "learning_rate": 0.00030739557474365894,
      "loss": 3.3784,
      "step": 45300
    },
    {
      "epoch": 4.889487870619946,
      "grad_norm": 0.6663231253623962,
      "learning_rate": 0.0003070717754991905,
      "loss": 3.3862,
      "step": 45350
    },
    {
      "epoch": 4.894878706199461,
      "grad_norm": 0.6640026569366455,
      "learning_rate": 0.0003067479762547221,
      "loss": 3.3841,
      "step": 45400
    },
    {
      "epoch": 4.900269541778976,
      "grad_norm": 0.7088777422904968,
      "learning_rate": 0.0003064241770102536,
      "loss": 3.3933,
      "step": 45450
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 0.6727973222732544,
      "learning_rate": 0.00030610037776578514,
      "loss": 3.3949,
      "step": 45500
    },
    {
      "epoch": 4.9110512129380055,
      "grad_norm": 0.6286665201187134,
      "learning_rate": 0.00030577657852131674,
      "loss": 3.394,
      "step": 45550
    },
    {
      "epoch": 4.916442048517521,
      "grad_norm": 0.623975396156311,
      "learning_rate": 0.0003054527792768483,
      "loss": 3.3913,
      "step": 45600
    },
    {
      "epoch": 4.921832884097035,
      "grad_norm": 0.6779341101646423,
      "learning_rate": 0.0003051289800323799,
      "loss": 3.406,
      "step": 45650
    },
    {
      "epoch": 4.92722371967655,
      "grad_norm": 0.6502783894538879,
      "learning_rate": 0.00030480518078791145,
      "loss": 3.402,
      "step": 45700
    },
    {
      "epoch": 4.932614555256064,
      "grad_norm": 0.656953752040863,
      "learning_rate": 0.00030448138154344305,
      "loss": 3.3801,
      "step": 45750
    },
    {
      "epoch": 4.938005390835579,
      "grad_norm": 0.6836448311805725,
      "learning_rate": 0.0003041575822989746,
      "loss": 3.4076,
      "step": 45800
    },
    {
      "epoch": 4.943396226415095,
      "grad_norm": 0.669844925403595,
      "learning_rate": 0.0003038337830545062,
      "loss": 3.3775,
      "step": 45850
    },
    {
      "epoch": 4.948787061994609,
      "grad_norm": 0.6585819721221924,
      "learning_rate": 0.00030350998381003776,
      "loss": 3.3735,
      "step": 45900
    },
    {
      "epoch": 4.954177897574124,
      "grad_norm": 0.6468183398246765,
      "learning_rate": 0.0003031861845655693,
      "loss": 3.3727,
      "step": 45950
    },
    {
      "epoch": 4.959568733153639,
      "grad_norm": 0.6753085255622864,
      "learning_rate": 0.0003028623853211009,
      "loss": 3.3721,
      "step": 46000
    },
    {
      "epoch": 4.959568733153639,
      "eval_accuracy": 0.3807414494795038,
      "eval_loss": 3.407444953918457,
      "eval_runtime": 183.8672,
      "eval_samples_per_second": 97.957,
      "eval_steps_per_second": 6.124,
      "step": 46000
    },
    {
      "epoch": 4.964959568733153,
      "grad_norm": 0.6388835906982422,
      "learning_rate": 0.00030253858607663247,
      "loss": 3.3795,
      "step": 46050
    },
    {
      "epoch": 4.9703504043126685,
      "grad_norm": 0.6600316166877747,
      "learning_rate": 0.00030221478683216407,
      "loss": 3.3892,
      "step": 46100
    },
    {
      "epoch": 4.975741239892184,
      "grad_norm": 0.6317592263221741,
      "learning_rate": 0.00030189098758769557,
      "loss": 3.388,
      "step": 46150
    },
    {
      "epoch": 4.981132075471698,
      "grad_norm": 0.670013964176178,
      "learning_rate": 0.00030156718834322717,
      "loss": 3.3928,
      "step": 46200
    },
    {
      "epoch": 4.986522911051213,
      "grad_norm": 0.65596604347229,
      "learning_rate": 0.0003012433890987587,
      "loss": 3.3911,
      "step": 46250
    },
    {
      "epoch": 4.991913746630727,
      "grad_norm": 0.6848031878471375,
      "learning_rate": 0.0003009195898542903,
      "loss": 3.3845,
      "step": 46300
    },
    {
      "epoch": 4.997304582210242,
      "grad_norm": 0.6780948638916016,
      "learning_rate": 0.0003005957906098219,
      "loss": 3.3775,
      "step": 46350
    },
    {
      "epoch": 5.002695417789758,
      "grad_norm": 0.7113416790962219,
      "learning_rate": 0.00030027199136535343,
      "loss": 3.3608,
      "step": 46400
    },
    {
      "epoch": 5.008086253369272,
      "grad_norm": 0.641831636428833,
      "learning_rate": 0.00029994819212088503,
      "loss": 3.2879,
      "step": 46450
    },
    {
      "epoch": 5.013477088948787,
      "grad_norm": 0.6682975888252258,
      "learning_rate": 0.0002996243928764166,
      "loss": 3.295,
      "step": 46500
    },
    {
      "epoch": 5.018867924528302,
      "grad_norm": 0.6478630900382996,
      "learning_rate": 0.0002993005936319482,
      "loss": 3.3073,
      "step": 46550
    },
    {
      "epoch": 5.024258760107816,
      "grad_norm": 0.6761943697929382,
      "learning_rate": 0.00029897679438747974,
      "loss": 3.2986,
      "step": 46600
    },
    {
      "epoch": 5.0296495956873315,
      "grad_norm": 0.7165064215660095,
      "learning_rate": 0.0002986529951430113,
      "loss": 3.286,
      "step": 46650
    },
    {
      "epoch": 5.035040431266847,
      "grad_norm": 0.6782529354095459,
      "learning_rate": 0.0002983291958985429,
      "loss": 3.297,
      "step": 46700
    },
    {
      "epoch": 5.040431266846361,
      "grad_norm": 0.6282607316970825,
      "learning_rate": 0.00029800539665407444,
      "loss": 3.2874,
      "step": 46750
    },
    {
      "epoch": 5.045822102425876,
      "grad_norm": 0.7574204802513123,
      "learning_rate": 0.000297681597409606,
      "loss": 3.296,
      "step": 46800
    },
    {
      "epoch": 5.051212938005391,
      "grad_norm": 0.6350822448730469,
      "learning_rate": 0.0002973577981651376,
      "loss": 3.3101,
      "step": 46850
    },
    {
      "epoch": 5.056603773584905,
      "grad_norm": 0.648804783821106,
      "learning_rate": 0.00029703399892066915,
      "loss": 3.3177,
      "step": 46900
    },
    {
      "epoch": 5.061994609164421,
      "grad_norm": 0.6287740468978882,
      "learning_rate": 0.00029671019967620076,
      "loss": 3.3077,
      "step": 46950
    },
    {
      "epoch": 5.067385444743936,
      "grad_norm": 0.671330988407135,
      "learning_rate": 0.0002963864004317323,
      "loss": 3.3139,
      "step": 47000
    },
    {
      "epoch": 5.067385444743936,
      "eval_accuracy": 0.3809589725609084,
      "eval_loss": 3.4086132049560547,
      "eval_runtime": 183.6167,
      "eval_samples_per_second": 98.09,
      "eval_steps_per_second": 6.132,
      "step": 47000
    },
    {
      "epoch": 5.07277628032345,
      "grad_norm": 0.6910858750343323,
      "learning_rate": 0.00029606260118726386,
      "loss": 3.2859,
      "step": 47050
    },
    {
      "epoch": 5.078167115902965,
      "grad_norm": 0.643031120300293,
      "learning_rate": 0.0002957388019427954,
      "loss": 3.2986,
      "step": 47100
    },
    {
      "epoch": 5.083557951482479,
      "grad_norm": 0.7260094881057739,
      "learning_rate": 0.000295415002698327,
      "loss": 3.3228,
      "step": 47150
    },
    {
      "epoch": 5.0889487870619945,
      "grad_norm": 0.6842207908630371,
      "learning_rate": 0.00029509120345385856,
      "loss": 3.3173,
      "step": 47200
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 0.6923130750656128,
      "learning_rate": 0.0002947738801942795,
      "loss": 3.3,
      "step": 47250
    },
    {
      "epoch": 5.099730458221024,
      "grad_norm": 0.652860701084137,
      "learning_rate": 0.0002944500809498111,
      "loss": 3.3124,
      "step": 47300
    },
    {
      "epoch": 5.105121293800539,
      "grad_norm": 0.7349758744239807,
      "learning_rate": 0.00029412628170534267,
      "loss": 3.314,
      "step": 47350
    },
    {
      "epoch": 5.110512129380054,
      "grad_norm": 0.6797593235969543,
      "learning_rate": 0.0002938024824608742,
      "loss": 3.3092,
      "step": 47400
    },
    {
      "epoch": 5.115902964959568,
      "grad_norm": 0.7050797939300537,
      "learning_rate": 0.00029347868321640577,
      "loss": 3.316,
      "step": 47450
    },
    {
      "epoch": 5.121293800539084,
      "grad_norm": 0.7074447870254517,
      "learning_rate": 0.0002931548839719374,
      "loss": 3.3116,
      "step": 47500
    },
    {
      "epoch": 5.126684636118599,
      "grad_norm": 0.7114849090576172,
      "learning_rate": 0.0002928310847274689,
      "loss": 3.3062,
      "step": 47550
    },
    {
      "epoch": 5.132075471698113,
      "grad_norm": 0.7198069095611572,
      "learning_rate": 0.00029250728548300053,
      "loss": 3.3031,
      "step": 47600
    },
    {
      "epoch": 5.137466307277628,
      "grad_norm": 0.6557970643043518,
      "learning_rate": 0.0002921834862385321,
      "loss": 3.3167,
      "step": 47650
    },
    {
      "epoch": 5.142857142857143,
      "grad_norm": 0.6805382370948792,
      "learning_rate": 0.0002918596869940637,
      "loss": 3.3198,
      "step": 47700
    },
    {
      "epoch": 5.1482479784366575,
      "grad_norm": 0.6633416414260864,
      "learning_rate": 0.0002915358877495952,
      "loss": 3.3215,
      "step": 47750
    },
    {
      "epoch": 5.153638814016173,
      "grad_norm": 0.6749739646911621,
      "learning_rate": 0.0002912120885051268,
      "loss": 3.3283,
      "step": 47800
    },
    {
      "epoch": 5.159029649595688,
      "grad_norm": 0.7563609480857849,
      "learning_rate": 0.00029088828926065834,
      "loss": 3.3213,
      "step": 47850
    },
    {
      "epoch": 5.164420485175202,
      "grad_norm": 0.7092325687408447,
      "learning_rate": 0.00029056449001618994,
      "loss": 3.3062,
      "step": 47900
    },
    {
      "epoch": 5.169811320754717,
      "grad_norm": 0.7057313919067383,
      "learning_rate": 0.0002902406907717215,
      "loss": 3.3222,
      "step": 47950
    },
    {
      "epoch": 5.175202156334231,
      "grad_norm": 0.6930835247039795,
      "learning_rate": 0.0002899168915272531,
      "loss": 3.3123,
      "step": 48000
    },
    {
      "epoch": 5.175202156334231,
      "eval_accuracy": 0.3812406008461235,
      "eval_loss": 3.4109530448913574,
      "eval_runtime": 183.8217,
      "eval_samples_per_second": 97.981,
      "eval_steps_per_second": 6.126,
      "step": 48000
    },
    {
      "epoch": 5.180592991913747,
      "grad_norm": 0.6858886480331421,
      "learning_rate": 0.00028959309228278465,
      "loss": 3.3438,
      "step": 48050
    },
    {
      "epoch": 5.185983827493262,
      "grad_norm": 0.7007202506065369,
      "learning_rate": 0.00028926929303831625,
      "loss": 3.3228,
      "step": 48100
    },
    {
      "epoch": 5.191374663072776,
      "grad_norm": 0.6837854981422424,
      "learning_rate": 0.0002889454937938478,
      "loss": 3.3293,
      "step": 48150
    },
    {
      "epoch": 5.196765498652291,
      "grad_norm": 0.6684266924858093,
      "learning_rate": 0.00028862169454937935,
      "loss": 3.3229,
      "step": 48200
    },
    {
      "epoch": 5.202156334231806,
      "grad_norm": 0.6720783710479736,
      "learning_rate": 0.0002882978953049109,
      "loss": 3.3037,
      "step": 48250
    },
    {
      "epoch": 5.2075471698113205,
      "grad_norm": 0.652512788772583,
      "learning_rate": 0.0002879740960604425,
      "loss": 3.3299,
      "step": 48300
    },
    {
      "epoch": 5.212938005390836,
      "grad_norm": 0.6807644367218018,
      "learning_rate": 0.00028765029681597406,
      "loss": 3.317,
      "step": 48350
    },
    {
      "epoch": 5.218328840970351,
      "grad_norm": 0.6457434892654419,
      "learning_rate": 0.00028732649757150566,
      "loss": 3.3284,
      "step": 48400
    },
    {
      "epoch": 5.223719676549865,
      "grad_norm": 0.7819386720657349,
      "learning_rate": 0.0002870026983270372,
      "loss": 3.3175,
      "step": 48450
    },
    {
      "epoch": 5.22911051212938,
      "grad_norm": 0.6470807790756226,
      "learning_rate": 0.00028667889908256877,
      "loss": 3.3181,
      "step": 48500
    },
    {
      "epoch": 5.234501347708895,
      "grad_norm": 0.6543563604354858,
      "learning_rate": 0.00028635509983810037,
      "loss": 3.3024,
      "step": 48550
    },
    {
      "epoch": 5.2398921832884096,
      "grad_norm": 0.6741393208503723,
      "learning_rate": 0.0002860313005936319,
      "loss": 3.3287,
      "step": 48600
    },
    {
      "epoch": 5.245283018867925,
      "grad_norm": 0.6699310541152954,
      "learning_rate": 0.00028570750134916347,
      "loss": 3.3222,
      "step": 48650
    },
    {
      "epoch": 5.250673854447439,
      "grad_norm": 0.6818548440933228,
      "learning_rate": 0.0002853837021046951,
      "loss": 3.3207,
      "step": 48700
    },
    {
      "epoch": 5.256064690026954,
      "grad_norm": 0.6772786974906921,
      "learning_rate": 0.00028505990286022663,
      "loss": 3.3115,
      "step": 48750
    },
    {
      "epoch": 5.261455525606469,
      "grad_norm": 0.6387109756469727,
      "learning_rate": 0.0002847361036157582,
      "loss": 3.3251,
      "step": 48800
    },
    {
      "epoch": 5.2668463611859835,
      "grad_norm": 0.6841785907745361,
      "learning_rate": 0.0002844123043712898,
      "loss": 3.3394,
      "step": 48850
    },
    {
      "epoch": 5.272237196765499,
      "grad_norm": 0.6936202049255371,
      "learning_rate": 0.00028408850512682133,
      "loss": 3.3173,
      "step": 48900
    },
    {
      "epoch": 5.277628032345014,
      "grad_norm": 0.6657775044441223,
      "learning_rate": 0.00028376470588235294,
      "loss": 3.3317,
      "step": 48950
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 0.6841873526573181,
      "learning_rate": 0.0002834409066378845,
      "loss": 3.3424,
      "step": 49000
    },
    {
      "epoch": 5.283018867924528,
      "eval_accuracy": 0.38162577533342584,
      "eval_loss": 3.40484619140625,
      "eval_runtime": 183.6772,
      "eval_samples_per_second": 98.058,
      "eval_steps_per_second": 6.13,
      "step": 49000
    },
    {
      "epoch": 5.288409703504043,
      "grad_norm": 0.6730915307998657,
      "learning_rate": 0.00028311710739341604,
      "loss": 3.336,
      "step": 49050
    },
    {
      "epoch": 5.293800539083558,
      "grad_norm": 0.7052405476570129,
      "learning_rate": 0.00028279330814894764,
      "loss": 3.3385,
      "step": 49100
    },
    {
      "epoch": 5.2991913746630726,
      "grad_norm": 0.6705338954925537,
      "learning_rate": 0.0002824695089044792,
      "loss": 3.3141,
      "step": 49150
    },
    {
      "epoch": 5.304582210242588,
      "grad_norm": 0.6564133763313293,
      "learning_rate": 0.00028214570966001075,
      "loss": 3.3388,
      "step": 49200
    },
    {
      "epoch": 5.309973045822103,
      "grad_norm": 0.6649319529533386,
      "learning_rate": 0.0002818283864004317,
      "loss": 3.3264,
      "step": 49250
    },
    {
      "epoch": 5.315363881401617,
      "grad_norm": 0.7245952486991882,
      "learning_rate": 0.0002815045871559633,
      "loss": 3.3236,
      "step": 49300
    },
    {
      "epoch": 5.320754716981132,
      "grad_norm": 0.6721333861351013,
      "learning_rate": 0.00028118078791149485,
      "loss": 3.3403,
      "step": 49350
    },
    {
      "epoch": 5.3261455525606465,
      "grad_norm": 0.6484825015068054,
      "learning_rate": 0.0002808569886670264,
      "loss": 3.3165,
      "step": 49400
    },
    {
      "epoch": 5.331536388140162,
      "grad_norm": 0.6736767292022705,
      "learning_rate": 0.00028053318942255795,
      "loss": 3.316,
      "step": 49450
    },
    {
      "epoch": 5.336927223719677,
      "grad_norm": 0.6363009810447693,
      "learning_rate": 0.00028020939017808956,
      "loss": 3.343,
      "step": 49500
    },
    {
      "epoch": 5.342318059299191,
      "grad_norm": 0.698854923248291,
      "learning_rate": 0.0002798855909336211,
      "loss": 3.3392,
      "step": 49550
    },
    {
      "epoch": 5.347708894878706,
      "grad_norm": 0.6946191787719727,
      "learning_rate": 0.0002795617916891527,
      "loss": 3.3336,
      "step": 49600
    },
    {
      "epoch": 5.353099730458221,
      "grad_norm": 0.6818442940711975,
      "learning_rate": 0.00027923799244468426,
      "loss": 3.321,
      "step": 49650
    },
    {
      "epoch": 5.3584905660377355,
      "grad_norm": 0.6966042518615723,
      "learning_rate": 0.00027891419320021587,
      "loss": 3.3464,
      "step": 49700
    },
    {
      "epoch": 5.363881401617251,
      "grad_norm": 0.6865735054016113,
      "learning_rate": 0.0002785903939557474,
      "loss": 3.3428,
      "step": 49750
    },
    {
      "epoch": 5.369272237196766,
      "grad_norm": 0.6621612310409546,
      "learning_rate": 0.00027826659471127897,
      "loss": 3.3162,
      "step": 49800
    },
    {
      "epoch": 5.37466307277628,
      "grad_norm": 0.6775622963905334,
      "learning_rate": 0.0002779427954668105,
      "loss": 3.348,
      "step": 49850
    },
    {
      "epoch": 5.380053908355795,
      "grad_norm": 0.6511054039001465,
      "learning_rate": 0.0002776189962223421,
      "loss": 3.3274,
      "step": 49900
    },
    {
      "epoch": 5.38544474393531,
      "grad_norm": 0.7020261883735657,
      "learning_rate": 0.0002772951969778737,
      "loss": 3.3409,
      "step": 49950
    },
    {
      "epoch": 5.390835579514825,
      "grad_norm": 0.6863080263137817,
      "learning_rate": 0.0002769713977334053,
      "loss": 3.3402,
      "step": 50000
    },
    {
      "epoch": 5.390835579514825,
      "eval_accuracy": 0.3820112757793916,
      "eval_loss": 3.4008445739746094,
      "eval_runtime": 183.6696,
      "eval_samples_per_second": 98.062,
      "eval_steps_per_second": 6.131,
      "step": 50000
    },
    {
      "epoch": 5.39622641509434,
      "grad_norm": 0.6434275507926941,
      "learning_rate": 0.00027664759848893683,
      "loss": 3.3234,
      "step": 50050
    },
    {
      "epoch": 5.401617250673855,
      "grad_norm": 0.7132904529571533,
      "learning_rate": 0.00027632379924446844,
      "loss": 3.3187,
      "step": 50100
    },
    {
      "epoch": 5.407008086253369,
      "grad_norm": 0.6604653000831604,
      "learning_rate": 0.000276,
      "loss": 3.3325,
      "step": 50150
    },
    {
      "epoch": 5.412398921832884,
      "grad_norm": 0.702043890953064,
      "learning_rate": 0.00027567620075553154,
      "loss": 3.33,
      "step": 50200
    },
    {
      "epoch": 5.4177897574123985,
      "grad_norm": 0.6745196580886841,
      "learning_rate": 0.0002753524015110631,
      "loss": 3.3392,
      "step": 50250
    },
    {
      "epoch": 5.423180592991914,
      "grad_norm": 0.6715883612632751,
      "learning_rate": 0.0002750286022665947,
      "loss": 3.3213,
      "step": 50300
    },
    {
      "epoch": 5.428571428571429,
      "grad_norm": 0.6874634623527527,
      "learning_rate": 0.00027470480302212624,
      "loss": 3.3519,
      "step": 50350
    },
    {
      "epoch": 5.433962264150943,
      "grad_norm": 0.6819542050361633,
      "learning_rate": 0.00027438100377765785,
      "loss": 3.3322,
      "step": 50400
    },
    {
      "epoch": 5.439353099730458,
      "grad_norm": 0.7074691653251648,
      "learning_rate": 0.0002740572045331894,
      "loss": 3.3365,
      "step": 50450
    },
    {
      "epoch": 5.444743935309973,
      "grad_norm": 0.7091408967971802,
      "learning_rate": 0.00027373340528872095,
      "loss": 3.3268,
      "step": 50500
    },
    {
      "epoch": 5.450134770889488,
      "grad_norm": 0.6710211634635925,
      "learning_rate": 0.00027340960604425255,
      "loss": 3.3326,
      "step": 50550
    },
    {
      "epoch": 5.455525606469003,
      "grad_norm": 0.6949871182441711,
      "learning_rate": 0.0002730858067997841,
      "loss": 3.3433,
      "step": 50600
    },
    {
      "epoch": 5.460916442048518,
      "grad_norm": 0.7077759504318237,
      "learning_rate": 0.00027276200755531565,
      "loss": 3.3538,
      "step": 50650
    },
    {
      "epoch": 5.466307277628032,
      "grad_norm": 0.7271976470947266,
      "learning_rate": 0.00027243820831084726,
      "loss": 3.3146,
      "step": 50700
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 0.6720749735832214,
      "learning_rate": 0.0002721144090663788,
      "loss": 3.3346,
      "step": 50750
    },
    {
      "epoch": 5.4770889487870615,
      "grad_norm": 0.6647154092788696,
      "learning_rate": 0.00027179060982191036,
      "loss": 3.3343,
      "step": 50800
    },
    {
      "epoch": 5.482479784366577,
      "grad_norm": 0.633385419845581,
      "learning_rate": 0.00027146681057744197,
      "loss": 3.3266,
      "step": 50850
    },
    {
      "epoch": 5.487870619946092,
      "grad_norm": 0.7008586525917053,
      "learning_rate": 0.0002711430113329735,
      "loss": 3.3252,
      "step": 50900
    },
    {
      "epoch": 5.493261455525606,
      "grad_norm": 0.6773853898048401,
      "learning_rate": 0.0002708192120885051,
      "loss": 3.3414,
      "step": 50950
    },
    {
      "epoch": 5.498652291105121,
      "grad_norm": 0.6934987306594849,
      "learning_rate": 0.00027049541284403667,
      "loss": 3.3439,
      "step": 51000
    },
    {
      "epoch": 5.498652291105121,
      "eval_accuracy": 0.38229887997343653,
      "eval_loss": 3.3964803218841553,
      "eval_runtime": 184.2643,
      "eval_samples_per_second": 97.745,
      "eval_steps_per_second": 6.111,
      "step": 51000
    },
    {
      "epoch": 5.504043126684636,
      "grad_norm": 0.6530951857566833,
      "learning_rate": 0.0002701716135995683,
      "loss": 3.3507,
      "step": 51050
    },
    {
      "epoch": 5.509433962264151,
      "grad_norm": 0.6792340874671936,
      "learning_rate": 0.0002698478143550998,
      "loss": 3.3366,
      "step": 51100
    },
    {
      "epoch": 5.514824797843666,
      "grad_norm": 0.6732293367385864,
      "learning_rate": 0.0002695240151106314,
      "loss": 3.3349,
      "step": 51150
    },
    {
      "epoch": 5.520215633423181,
      "grad_norm": 0.6773748993873596,
      "learning_rate": 0.00026920021586616293,
      "loss": 3.3177,
      "step": 51200
    },
    {
      "epoch": 5.525606469002695,
      "grad_norm": 0.6929314136505127,
      "learning_rate": 0.00026887641662169453,
      "loss": 3.338,
      "step": 51250
    },
    {
      "epoch": 5.53099730458221,
      "grad_norm": 0.6713197231292725,
      "learning_rate": 0.0002685590933621155,
      "loss": 3.33,
      "step": 51300
    },
    {
      "epoch": 5.536388140161725,
      "grad_norm": 0.6788355708122253,
      "learning_rate": 0.00026823529411764704,
      "loss": 3.3213,
      "step": 51350
    },
    {
      "epoch": 5.54177897574124,
      "grad_norm": 0.7311290502548218,
      "learning_rate": 0.0002679114948731786,
      "loss": 3.3131,
      "step": 51400
    },
    {
      "epoch": 5.547169811320755,
      "grad_norm": 0.6731926202774048,
      "learning_rate": 0.0002675876956287102,
      "loss": 3.3122,
      "step": 51450
    },
    {
      "epoch": 5.55256064690027,
      "grad_norm": 0.7098152041435242,
      "learning_rate": 0.00026726389638424174,
      "loss": 3.3427,
      "step": 51500
    },
    {
      "epoch": 5.557951482479784,
      "grad_norm": 0.7369858026504517,
      "learning_rate": 0.0002669400971397733,
      "loss": 3.3459,
      "step": 51550
    },
    {
      "epoch": 5.563342318059299,
      "grad_norm": 0.6638474464416504,
      "learning_rate": 0.0002666162978953049,
      "loss": 3.3208,
      "step": 51600
    },
    {
      "epoch": 5.568733153638814,
      "grad_norm": 0.6835918426513672,
      "learning_rate": 0.00026629249865083645,
      "loss": 3.3362,
      "step": 51650
    },
    {
      "epoch": 5.574123989218329,
      "grad_norm": 0.7660952806472778,
      "learning_rate": 0.00026596869940636805,
      "loss": 3.3392,
      "step": 51700
    },
    {
      "epoch": 5.579514824797844,
      "grad_norm": 0.6250906586647034,
      "learning_rate": 0.0002656449001618996,
      "loss": 3.3312,
      "step": 51750
    },
    {
      "epoch": 5.584905660377358,
      "grad_norm": 0.6872050762176514,
      "learning_rate": 0.0002653211009174312,
      "loss": 3.3224,
      "step": 51800
    },
    {
      "epoch": 5.590296495956873,
      "grad_norm": 0.7721320986747742,
      "learning_rate": 0.0002649973016729627,
      "loss": 3.333,
      "step": 51850
    },
    {
      "epoch": 5.595687331536388,
      "grad_norm": 0.6676695346832275,
      "learning_rate": 0.0002646735024284943,
      "loss": 3.3208,
      "step": 51900
    },
    {
      "epoch": 5.601078167115903,
      "grad_norm": 0.7069310545921326,
      "learning_rate": 0.00026434970318402586,
      "loss": 3.3394,
      "step": 51950
    },
    {
      "epoch": 5.606469002695418,
      "grad_norm": 0.6760640144348145,
      "learning_rate": 0.00026402590393955746,
      "loss": 3.3189,
      "step": 52000
    },
    {
      "epoch": 5.606469002695418,
      "eval_accuracy": 0.3827234954590155,
      "eval_loss": 3.3922245502471924,
      "eval_runtime": 183.5334,
      "eval_samples_per_second": 98.135,
      "eval_steps_per_second": 6.135,
      "step": 52000
    },
    {
      "epoch": 5.611859838274933,
      "grad_norm": 0.7278321981430054,
      "learning_rate": 0.000263702104695089,
      "loss": 3.335,
      "step": 52050
    },
    {
      "epoch": 5.617250673854447,
      "grad_norm": 0.7073381543159485,
      "learning_rate": 0.0002633783054506206,
      "loss": 3.3279,
      "step": 52100
    },
    {
      "epoch": 5.622641509433962,
      "grad_norm": 0.7076802253723145,
      "learning_rate": 0.00026305450620615217,
      "loss": 3.3472,
      "step": 52150
    },
    {
      "epoch": 5.628032345013477,
      "grad_norm": 0.7682281732559204,
      "learning_rate": 0.0002627307069616837,
      "loss": 3.3518,
      "step": 52200
    },
    {
      "epoch": 5.633423180592992,
      "grad_norm": 0.6639510989189148,
      "learning_rate": 0.0002624069077172153,
      "loss": 3.3299,
      "step": 52250
    },
    {
      "epoch": 5.638814016172507,
      "grad_norm": 0.6877766847610474,
      "learning_rate": 0.0002620831084727469,
      "loss": 3.3161,
      "step": 52300
    },
    {
      "epoch": 5.644204851752022,
      "grad_norm": 0.6921620965003967,
      "learning_rate": 0.0002617593092282784,
      "loss": 3.3437,
      "step": 52350
    },
    {
      "epoch": 5.649595687331536,
      "grad_norm": 0.6771333813667297,
      "learning_rate": 0.00026143550998381003,
      "loss": 3.3196,
      "step": 52400
    },
    {
      "epoch": 5.654986522911051,
      "grad_norm": 0.7312127351760864,
      "learning_rate": 0.0002611117107393416,
      "loss": 3.3373,
      "step": 52450
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 0.7787958383560181,
      "learning_rate": 0.00026078791149487313,
      "loss": 3.3444,
      "step": 52500
    },
    {
      "epoch": 5.665768194070081,
      "grad_norm": 0.6783169507980347,
      "learning_rate": 0.00026046411225040474,
      "loss": 3.3308,
      "step": 52550
    },
    {
      "epoch": 5.671159029649596,
      "grad_norm": 0.6774412393569946,
      "learning_rate": 0.0002601403130059363,
      "loss": 3.3452,
      "step": 52600
    },
    {
      "epoch": 5.67654986522911,
      "grad_norm": 0.7152056097984314,
      "learning_rate": 0.0002598165137614679,
      "loss": 3.3428,
      "step": 52650
    },
    {
      "epoch": 5.681940700808625,
      "grad_norm": 0.7162232398986816,
      "learning_rate": 0.00025949271451699944,
      "loss": 3.3342,
      "step": 52700
    },
    {
      "epoch": 5.6873315363881405,
      "grad_norm": 0.7322828769683838,
      "learning_rate": 0.000259168915272531,
      "loss": 3.3471,
      "step": 52750
    },
    {
      "epoch": 5.692722371967655,
      "grad_norm": 0.7306454181671143,
      "learning_rate": 0.0002588451160280626,
      "loss": 3.3263,
      "step": 52800
    },
    {
      "epoch": 5.69811320754717,
      "grad_norm": 0.7096463441848755,
      "learning_rate": 0.00025852131678359415,
      "loss": 3.326,
      "step": 52850
    },
    {
      "epoch": 5.703504043126685,
      "grad_norm": 0.7072291374206543,
      "learning_rate": 0.0002581975175391257,
      "loss": 3.3271,
      "step": 52900
    },
    {
      "epoch": 5.708894878706199,
      "grad_norm": 0.7425927519798279,
      "learning_rate": 0.0002578737182946573,
      "loss": 3.343,
      "step": 52950
    },
    {
      "epoch": 5.714285714285714,
      "grad_norm": 0.7667298316955566,
      "learning_rate": 0.00025754991905018885,
      "loss": 3.3341,
      "step": 53000
    },
    {
      "epoch": 5.714285714285714,
      "eval_accuracy": 0.38343777954350794,
      "eval_loss": 3.385892391204834,
      "eval_runtime": 183.7702,
      "eval_samples_per_second": 98.008,
      "eval_steps_per_second": 6.127,
      "step": 53000
    },
    {
      "epoch": 5.719676549865229,
      "grad_norm": 0.666398286819458,
      "learning_rate": 0.00025722611980572046,
      "loss": 3.3137,
      "step": 53050
    },
    {
      "epoch": 5.725067385444744,
      "grad_norm": 0.7412814497947693,
      "learning_rate": 0.000256902320561252,
      "loss": 3.327,
      "step": 53100
    },
    {
      "epoch": 5.730458221024259,
      "grad_norm": 0.7019506692886353,
      "learning_rate": 0.00025657852131678356,
      "loss": 3.3345,
      "step": 53150
    },
    {
      "epoch": 5.735849056603773,
      "grad_norm": 0.7068491578102112,
      "learning_rate": 0.0002562547220723151,
      "loss": 3.3335,
      "step": 53200
    },
    {
      "epoch": 5.741239892183288,
      "grad_norm": 0.6603403091430664,
      "learning_rate": 0.0002559309228278467,
      "loss": 3.3237,
      "step": 53250
    },
    {
      "epoch": 5.7466307277628035,
      "grad_norm": 0.6823389530181885,
      "learning_rate": 0.00025561359956826767,
      "loss": 3.3499,
      "step": 53300
    },
    {
      "epoch": 5.752021563342318,
      "grad_norm": 0.6933403611183167,
      "learning_rate": 0.0002552898003237992,
      "loss": 3.327,
      "step": 53350
    },
    {
      "epoch": 5.757412398921833,
      "grad_norm": 0.6974020004272461,
      "learning_rate": 0.0002549660010793308,
      "loss": 3.3137,
      "step": 53400
    },
    {
      "epoch": 5.762803234501348,
      "grad_norm": 0.6720824241638184,
      "learning_rate": 0.0002546422018348624,
      "loss": 3.3461,
      "step": 53450
    },
    {
      "epoch": 5.768194070080862,
      "grad_norm": 0.7010537385940552,
      "learning_rate": 0.0002543184025903939,
      "loss": 3.3161,
      "step": 53500
    },
    {
      "epoch": 5.773584905660377,
      "grad_norm": 0.6914909482002258,
      "learning_rate": 0.0002539946033459255,
      "loss": 3.3437,
      "step": 53550
    },
    {
      "epoch": 5.7789757412398925,
      "grad_norm": 0.6911107301712036,
      "learning_rate": 0.0002536708041014571,
      "loss": 3.3169,
      "step": 53600
    },
    {
      "epoch": 5.784366576819407,
      "grad_norm": 0.6994060277938843,
      "learning_rate": 0.00025334700485698863,
      "loss": 3.3496,
      "step": 53650
    },
    {
      "epoch": 5.789757412398922,
      "grad_norm": 0.7192985415458679,
      "learning_rate": 0.00025302320561252023,
      "loss": 3.3223,
      "step": 53700
    },
    {
      "epoch": 5.795148247978437,
      "grad_norm": 0.7265368103981018,
      "learning_rate": 0.0002526994063680518,
      "loss": 3.3361,
      "step": 53750
    },
    {
      "epoch": 5.800539083557951,
      "grad_norm": 0.7237759232521057,
      "learning_rate": 0.0002523756071235834,
      "loss": 3.329,
      "step": 53800
    },
    {
      "epoch": 5.8059299191374665,
      "grad_norm": 0.698350191116333,
      "learning_rate": 0.00025205180787911494,
      "loss": 3.3318,
      "step": 53850
    },
    {
      "epoch": 5.811320754716981,
      "grad_norm": 0.6901752948760986,
      "learning_rate": 0.0002517280086346465,
      "loss": 3.3252,
      "step": 53900
    },
    {
      "epoch": 5.816711590296496,
      "grad_norm": 0.6650895476341248,
      "learning_rate": 0.00025140420939017804,
      "loss": 3.3167,
      "step": 53950
    },
    {
      "epoch": 5.822102425876011,
      "grad_norm": 0.6690970063209534,
      "learning_rate": 0.00025108041014570965,
      "loss": 3.3311,
      "step": 54000
    },
    {
      "epoch": 5.822102425876011,
      "eval_accuracy": 0.384017551352886,
      "eval_loss": 3.3810160160064697,
      "eval_runtime": 183.9693,
      "eval_samples_per_second": 97.902,
      "eval_steps_per_second": 6.121,
      "step": 54000
    },
    {
      "epoch": 5.827493261455525,
      "grad_norm": 0.6725430488586426,
      "learning_rate": 0.0002507566109012412,
      "loss": 3.3441,
      "step": 54050
    },
    {
      "epoch": 5.83288409703504,
      "grad_norm": 0.6652386784553528,
      "learning_rate": 0.0002504328116567728,
      "loss": 3.3496,
      "step": 54100
    },
    {
      "epoch": 5.8382749326145555,
      "grad_norm": 0.7626335620880127,
      "learning_rate": 0.00025010901241230435,
      "loss": 3.3252,
      "step": 54150
    },
    {
      "epoch": 5.84366576819407,
      "grad_norm": 0.6746981143951416,
      "learning_rate": 0.0002497852131678359,
      "loss": 3.3347,
      "step": 54200
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 0.7410542368888855,
      "learning_rate": 0.0002494614139233675,
      "loss": 3.3627,
      "step": 54250
    },
    {
      "epoch": 5.8544474393531,
      "grad_norm": 0.6656700372695923,
      "learning_rate": 0.00024913761467889906,
      "loss": 3.3415,
      "step": 54300
    },
    {
      "epoch": 5.859838274932614,
      "grad_norm": 0.6493524312973022,
      "learning_rate": 0.0002488138154344306,
      "loss": 3.3432,
      "step": 54350
    },
    {
      "epoch": 5.8652291105121295,
      "grad_norm": 0.7260926961898804,
      "learning_rate": 0.0002484900161899622,
      "loss": 3.3449,
      "step": 54400
    },
    {
      "epoch": 5.870619946091644,
      "grad_norm": 0.7246960401535034,
      "learning_rate": 0.00024816621694549376,
      "loss": 3.3164,
      "step": 54450
    },
    {
      "epoch": 5.876010781671159,
      "grad_norm": 0.7035079002380371,
      "learning_rate": 0.00024784241770102537,
      "loss": 3.3455,
      "step": 54500
    },
    {
      "epoch": 5.881401617250674,
      "grad_norm": 0.6615075469017029,
      "learning_rate": 0.0002475186184565569,
      "loss": 3.332,
      "step": 54550
    },
    {
      "epoch": 5.886792452830189,
      "grad_norm": 0.6515076160430908,
      "learning_rate": 0.00024719481921208847,
      "loss": 3.3244,
      "step": 54600
    },
    {
      "epoch": 5.892183288409703,
      "grad_norm": 0.6946157813072205,
      "learning_rate": 0.0002468710199676201,
      "loss": 3.3463,
      "step": 54650
    },
    {
      "epoch": 5.8975741239892185,
      "grad_norm": 0.6916877627372742,
      "learning_rate": 0.0002465472207231516,
      "loss": 3.3441,
      "step": 54700
    },
    {
      "epoch": 5.902964959568733,
      "grad_norm": 0.7089888453483582,
      "learning_rate": 0.0002462234214786832,
      "loss": 3.3255,
      "step": 54750
    },
    {
      "epoch": 5.908355795148248,
      "grad_norm": 0.6872110366821289,
      "learning_rate": 0.0002458996222342148,
      "loss": 3.3591,
      "step": 54800
    },
    {
      "epoch": 5.913746630727763,
      "grad_norm": 0.692903459072113,
      "learning_rate": 0.00024557582298974633,
      "loss": 3.3261,
      "step": 54850
    },
    {
      "epoch": 5.919137466307277,
      "grad_norm": 0.696246862411499,
      "learning_rate": 0.0002452520237452779,
      "loss": 3.3555,
      "step": 54900
    },
    {
      "epoch": 5.9245283018867925,
      "grad_norm": 0.6855026483535767,
      "learning_rate": 0.0002449282245008095,
      "loss": 3.3346,
      "step": 54950
    },
    {
      "epoch": 5.929919137466308,
      "grad_norm": 0.736362099647522,
      "learning_rate": 0.00024460442525634104,
      "loss": 3.3347,
      "step": 55000
    },
    {
      "epoch": 5.929919137466308,
      "eval_accuracy": 0.38405981732624583,
      "eval_loss": 3.3812179565429688,
      "eval_runtime": 183.7282,
      "eval_samples_per_second": 98.031,
      "eval_steps_per_second": 6.129,
      "step": 55000
    },
    {
      "epoch": 5.935309973045822,
      "grad_norm": 0.702866792678833,
      "learning_rate": 0.00024428062601187264,
      "loss": 3.3561,
      "step": 55050
    },
    {
      "epoch": 5.940700808625337,
      "grad_norm": 0.7051444053649902,
      "learning_rate": 0.0002439568267674042,
      "loss": 3.3295,
      "step": 55100
    },
    {
      "epoch": 5.946091644204852,
      "grad_norm": 0.7245047092437744,
      "learning_rate": 0.00024363302752293574,
      "loss": 3.3433,
      "step": 55150
    },
    {
      "epoch": 5.951482479784366,
      "grad_norm": 0.7037590742111206,
      "learning_rate": 0.00024330922827846732,
      "loss": 3.3271,
      "step": 55200
    },
    {
      "epoch": 5.9568733153638815,
      "grad_norm": 0.6878694295883179,
      "learning_rate": 0.0002429854290339989,
      "loss": 3.3147,
      "step": 55250
    },
    {
      "epoch": 5.962264150943396,
      "grad_norm": 0.6708479523658752,
      "learning_rate": 0.00024266810577441985,
      "loss": 3.3289,
      "step": 55300
    },
    {
      "epoch": 5.967654986522911,
      "grad_norm": 0.701370358467102,
      "learning_rate": 0.00024234430652995143,
      "loss": 3.3266,
      "step": 55350
    },
    {
      "epoch": 5.973045822102426,
      "grad_norm": 0.6717433929443359,
      "learning_rate": 0.00024202050728548298,
      "loss": 3.3393,
      "step": 55400
    },
    {
      "epoch": 5.97843665768194,
      "grad_norm": 0.6901810169219971,
      "learning_rate": 0.00024169670804101456,
      "loss": 3.3462,
      "step": 55450
    },
    {
      "epoch": 5.9838274932614555,
      "grad_norm": 0.6869121789932251,
      "learning_rate": 0.0002413729087965461,
      "loss": 3.3139,
      "step": 55500
    },
    {
      "epoch": 5.989218328840971,
      "grad_norm": 0.7031119465827942,
      "learning_rate": 0.00024104910955207768,
      "loss": 3.336,
      "step": 55550
    },
    {
      "epoch": 5.994609164420485,
      "grad_norm": 0.6958100199699402,
      "learning_rate": 0.00024072531030760926,
      "loss": 3.3504,
      "step": 55600
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.451558232307434,
      "learning_rate": 0.00024040151106314084,
      "loss": 3.3206,
      "step": 55650
    },
    {
      "epoch": 6.005390835579515,
      "grad_norm": 0.6931526064872742,
      "learning_rate": 0.0002400777118186724,
      "loss": 3.2332,
      "step": 55700
    },
    {
      "epoch": 6.010781671159029,
      "grad_norm": 0.7126358151435852,
      "learning_rate": 0.00023975391257420397,
      "loss": 3.2352,
      "step": 55750
    },
    {
      "epoch": 6.0161725067385445,
      "grad_norm": 0.7339010238647461,
      "learning_rate": 0.00023943011332973555,
      "loss": 3.2684,
      "step": 55800
    },
    {
      "epoch": 6.02156334231806,
      "grad_norm": 0.7038446068763733,
      "learning_rate": 0.00023910631408526712,
      "loss": 3.2338,
      "step": 55850
    },
    {
      "epoch": 6.026954177897574,
      "grad_norm": 0.7177115082740784,
      "learning_rate": 0.0002387825148407987,
      "loss": 3.2345,
      "step": 55900
    },
    {
      "epoch": 6.032345013477089,
      "grad_norm": 0.7070705890655518,
      "learning_rate": 0.00023845871559633025,
      "loss": 3.2623,
      "step": 55950
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 0.7214942574501038,
      "learning_rate": 0.0002381349163518618,
      "loss": 3.2357,
      "step": 56000
    },
    {
      "epoch": 6.037735849056604,
      "eval_accuracy": 0.38422996774856333,
      "eval_loss": 3.3826565742492676,
      "eval_runtime": 183.9778,
      "eval_samples_per_second": 97.898,
      "eval_steps_per_second": 6.12,
      "step": 56000
    },
    {
      "epoch": 6.0431266846361185,
      "grad_norm": 0.6994975209236145,
      "learning_rate": 0.00023781111710739338,
      "loss": 3.2527,
      "step": 56050
    },
    {
      "epoch": 6.048517520215634,
      "grad_norm": 0.6794683933258057,
      "learning_rate": 0.00023748731786292496,
      "loss": 3.2424,
      "step": 56100
    },
    {
      "epoch": 6.053908355795148,
      "grad_norm": 0.6963556408882141,
      "learning_rate": 0.00023716351861845654,
      "loss": 3.2607,
      "step": 56150
    },
    {
      "epoch": 6.059299191374663,
      "grad_norm": 0.673784613609314,
      "learning_rate": 0.0002368397193739881,
      "loss": 3.2584,
      "step": 56200
    },
    {
      "epoch": 6.064690026954178,
      "grad_norm": 0.7150440812110901,
      "learning_rate": 0.0002365159201295197,
      "loss": 3.2551,
      "step": 56250
    },
    {
      "epoch": 6.070080862533692,
      "grad_norm": 0.7197666764259338,
      "learning_rate": 0.00023619212088505127,
      "loss": 3.2584,
      "step": 56300
    },
    {
      "epoch": 6.0754716981132075,
      "grad_norm": 0.6947667002677917,
      "learning_rate": 0.0002358683216405828,
      "loss": 3.2609,
      "step": 56350
    },
    {
      "epoch": 6.080862533692723,
      "grad_norm": 0.7144685983657837,
      "learning_rate": 0.00023554452239611437,
      "loss": 3.2739,
      "step": 56400
    },
    {
      "epoch": 6.086253369272237,
      "grad_norm": 0.7323464155197144,
      "learning_rate": 0.00023522072315164595,
      "loss": 3.2539,
      "step": 56450
    },
    {
      "epoch": 6.091644204851752,
      "grad_norm": 0.7219662070274353,
      "learning_rate": 0.00023489692390717752,
      "loss": 3.2484,
      "step": 56500
    },
    {
      "epoch": 6.097035040431267,
      "grad_norm": 0.7247258424758911,
      "learning_rate": 0.0002345731246627091,
      "loss": 3.2491,
      "step": 56550
    },
    {
      "epoch": 6.1024258760107815,
      "grad_norm": 0.761326014995575,
      "learning_rate": 0.00023424932541824068,
      "loss": 3.2653,
      "step": 56600
    },
    {
      "epoch": 6.107816711590297,
      "grad_norm": 0.706102728843689,
      "learning_rate": 0.00023392552617377226,
      "loss": 3.2608,
      "step": 56650
    },
    {
      "epoch": 6.113207547169812,
      "grad_norm": 0.6789826154708862,
      "learning_rate": 0.00023360172692930384,
      "loss": 3.2632,
      "step": 56700
    },
    {
      "epoch": 6.118598382749326,
      "grad_norm": 0.7313393354415894,
      "learning_rate": 0.00023327792768483539,
      "loss": 3.2674,
      "step": 56750
    },
    {
      "epoch": 6.123989218328841,
      "grad_norm": 0.7755120992660522,
      "learning_rate": 0.0002329606044252563,
      "loss": 3.2678,
      "step": 56800
    },
    {
      "epoch": 6.129380053908355,
      "grad_norm": 0.6921339631080627,
      "learning_rate": 0.0002326368051807879,
      "loss": 3.2611,
      "step": 56850
    },
    {
      "epoch": 6.1347708894878705,
      "grad_norm": 0.6921306252479553,
      "learning_rate": 0.00023231300593631947,
      "loss": 3.2698,
      "step": 56900
    },
    {
      "epoch": 6.140161725067386,
      "grad_norm": 0.7285143733024597,
      "learning_rate": 0.00023198920669185104,
      "loss": 3.2675,
      "step": 56950
    },
    {
      "epoch": 6.1455525606469,
      "grad_norm": 0.7773783802986145,
      "learning_rate": 0.00023166540744738262,
      "loss": 3.2576,
      "step": 57000
    },
    {
      "epoch": 6.1455525606469,
      "eval_accuracy": 0.3845002961334457,
      "eval_loss": 3.383111000061035,
      "eval_runtime": 185.0654,
      "eval_samples_per_second": 97.322,
      "eval_steps_per_second": 6.084,
      "step": 57000
    },
    {
      "epoch": 6.150943396226415,
      "grad_norm": 0.724994957447052,
      "learning_rate": 0.0002313416082029142,
      "loss": 3.2746,
      "step": 57050
    },
    {
      "epoch": 6.15633423180593,
      "grad_norm": 0.69971764087677,
      "learning_rate": 0.00023101780895844572,
      "loss": 3.278,
      "step": 57100
    },
    {
      "epoch": 6.1617250673854445,
      "grad_norm": 0.7013365626335144,
      "learning_rate": 0.0002306940097139773,
      "loss": 3.2754,
      "step": 57150
    },
    {
      "epoch": 6.16711590296496,
      "grad_norm": 0.7439168095588684,
      "learning_rate": 0.00023037021046950888,
      "loss": 3.2606,
      "step": 57200
    },
    {
      "epoch": 6.172506738544475,
      "grad_norm": 0.6989354491233826,
      "learning_rate": 0.00023004641122504046,
      "loss": 3.2701,
      "step": 57250
    },
    {
      "epoch": 6.177897574123989,
      "grad_norm": 0.7361764311790466,
      "learning_rate": 0.00022972261198057203,
      "loss": 3.2753,
      "step": 57300
    },
    {
      "epoch": 6.183288409703504,
      "grad_norm": 0.7835276126861572,
      "learning_rate": 0.0002293988127361036,
      "loss": 3.281,
      "step": 57350
    },
    {
      "epoch": 6.188679245283019,
      "grad_norm": 0.7380451560020447,
      "learning_rate": 0.00022907501349163516,
      "loss": 3.2676,
      "step": 57400
    },
    {
      "epoch": 6.1940700808625335,
      "grad_norm": 0.7116109132766724,
      "learning_rate": 0.00022875121424716674,
      "loss": 3.2601,
      "step": 57450
    },
    {
      "epoch": 6.199460916442049,
      "grad_norm": 0.7472301125526428,
      "learning_rate": 0.00022842741500269832,
      "loss": 3.2762,
      "step": 57500
    },
    {
      "epoch": 6.204851752021563,
      "grad_norm": 0.7332467436790466,
      "learning_rate": 0.00022810361575822987,
      "loss": 3.2991,
      "step": 57550
    },
    {
      "epoch": 6.210242587601078,
      "grad_norm": 0.6753149032592773,
      "learning_rate": 0.00022777981651376145,
      "loss": 3.2702,
      "step": 57600
    },
    {
      "epoch": 6.215633423180593,
      "grad_norm": 0.7721616625785828,
      "learning_rate": 0.00022745601726929302,
      "loss": 3.2752,
      "step": 57650
    },
    {
      "epoch": 6.2210242587601075,
      "grad_norm": 0.7246688604354858,
      "learning_rate": 0.00022713221802482457,
      "loss": 3.2696,
      "step": 57700
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 0.6795578002929688,
      "learning_rate": 0.00022680841878035615,
      "loss": 3.2596,
      "step": 57750
    },
    {
      "epoch": 6.231805929919138,
      "grad_norm": 0.7370179891586304,
      "learning_rate": 0.00022648461953588773,
      "loss": 3.2688,
      "step": 57800
    },
    {
      "epoch": 6.237196765498652,
      "grad_norm": 0.6983503699302673,
      "learning_rate": 0.0002261608202914193,
      "loss": 3.2753,
      "step": 57850
    },
    {
      "epoch": 6.242587601078167,
      "grad_norm": 0.7345255613327026,
      "learning_rate": 0.00022583702104695088,
      "loss": 3.2752,
      "step": 57900
    },
    {
      "epoch": 6.247978436657682,
      "grad_norm": 0.7167491316795349,
      "learning_rate": 0.00022551322180248246,
      "loss": 3.2838,
      "step": 57950
    },
    {
      "epoch": 6.2533692722371965,
      "grad_norm": 0.7173773646354675,
      "learning_rate": 0.00022518942255801399,
      "loss": 3.2761,
      "step": 58000
    },
    {
      "epoch": 6.2533692722371965,
      "eval_accuracy": 0.38490709254542316,
      "eval_loss": 3.379185914993286,
      "eval_runtime": 184.7659,
      "eval_samples_per_second": 97.48,
      "eval_steps_per_second": 6.094,
      "step": 58000
    },
    {
      "epoch": 6.258760107816712,
      "grad_norm": 0.7806297540664673,
      "learning_rate": 0.00022486562331354556,
      "loss": 3.268,
      "step": 58050
    },
    {
      "epoch": 6.264150943396227,
      "grad_norm": 0.7050454616546631,
      "learning_rate": 0.00022454182406907714,
      "loss": 3.2624,
      "step": 58100
    },
    {
      "epoch": 6.269541778975741,
      "grad_norm": 0.6826767325401306,
      "learning_rate": 0.00022421802482460872,
      "loss": 3.2622,
      "step": 58150
    },
    {
      "epoch": 6.274932614555256,
      "grad_norm": 0.7257795333862305,
      "learning_rate": 0.0002238942255801403,
      "loss": 3.2999,
      "step": 58200
    },
    {
      "epoch": 6.280323450134771,
      "grad_norm": 0.7288961410522461,
      "learning_rate": 0.00022357042633567187,
      "loss": 3.2753,
      "step": 58250
    },
    {
      "epoch": 6.285714285714286,
      "grad_norm": 0.7374154925346375,
      "learning_rate": 0.00022324662709120345,
      "loss": 3.262,
      "step": 58300
    },
    {
      "epoch": 6.291105121293801,
      "grad_norm": 0.7150074243545532,
      "learning_rate": 0.00022292282784673503,
      "loss": 3.2877,
      "step": 58350
    },
    {
      "epoch": 6.296495956873315,
      "grad_norm": 0.7542361617088318,
      "learning_rate": 0.00022259902860226655,
      "loss": 3.2719,
      "step": 58400
    },
    {
      "epoch": 6.30188679245283,
      "grad_norm": 0.7025394439697266,
      "learning_rate": 0.00022227522935779813,
      "loss": 3.2783,
      "step": 58450
    },
    {
      "epoch": 6.307277628032345,
      "grad_norm": 0.7773345708847046,
      "learning_rate": 0.0002219514301133297,
      "loss": 3.2935,
      "step": 58500
    },
    {
      "epoch": 6.3126684636118595,
      "grad_norm": 0.7307718396186829,
      "learning_rate": 0.00022162763086886129,
      "loss": 3.2527,
      "step": 58550
    },
    {
      "epoch": 6.318059299191375,
      "grad_norm": 0.7027918696403503,
      "learning_rate": 0.00022130383162439286,
      "loss": 3.2868,
      "step": 58600
    },
    {
      "epoch": 6.32345013477089,
      "grad_norm": 0.7043139934539795,
      "learning_rate": 0.00022098003237992444,
      "loss": 3.2659,
      "step": 58650
    },
    {
      "epoch": 6.328840970350404,
      "grad_norm": 0.6800243854522705,
      "learning_rate": 0.00022065623313545602,
      "loss": 3.2702,
      "step": 58700
    },
    {
      "epoch": 6.334231805929919,
      "grad_norm": 0.7141492962837219,
      "learning_rate": 0.00022033243389098757,
      "loss": 3.2907,
      "step": 58750
    },
    {
      "epoch": 6.339622641509434,
      "grad_norm": 0.7111701369285583,
      "learning_rate": 0.00022000863464651915,
      "loss": 3.2788,
      "step": 58800
    },
    {
      "epoch": 6.345013477088949,
      "grad_norm": 0.7029032111167908,
      "learning_rate": 0.0002196848354020507,
      "loss": 3.2843,
      "step": 58850
    },
    {
      "epoch": 6.350404312668464,
      "grad_norm": 0.7716480493545532,
      "learning_rate": 0.00021936103615758227,
      "loss": 3.29,
      "step": 58900
    },
    {
      "epoch": 6.355795148247978,
      "grad_norm": 0.6854252815246582,
      "learning_rate": 0.00021903723691311385,
      "loss": 3.3005,
      "step": 58950
    },
    {
      "epoch": 6.361185983827493,
      "grad_norm": 0.6938695311546326,
      "learning_rate": 0.00021871343766864543,
      "loss": 3.289,
      "step": 59000
    },
    {
      "epoch": 6.361185983827493,
      "eval_accuracy": 0.3852709710667139,
      "eval_loss": 3.37485408782959,
      "eval_runtime": 185.1638,
      "eval_samples_per_second": 97.271,
      "eval_steps_per_second": 6.081,
      "step": 59000
    },
    {
      "epoch": 6.366576819407008,
      "grad_norm": 0.7637062668800354,
      "learning_rate": 0.00021838963842417698,
      "loss": 3.2879,
      "step": 59050
    },
    {
      "epoch": 6.3719676549865225,
      "grad_norm": 0.71692955493927,
      "learning_rate": 0.00021806583917970856,
      "loss": 3.2819,
      "step": 59100
    },
    {
      "epoch": 6.377358490566038,
      "grad_norm": 0.7055894732475281,
      "learning_rate": 0.00021774203993524014,
      "loss": 3.2831,
      "step": 59150
    },
    {
      "epoch": 6.382749326145553,
      "grad_norm": 0.6954150199890137,
      "learning_rate": 0.00021741824069077171,
      "loss": 3.2762,
      "step": 59200
    },
    {
      "epoch": 6.388140161725067,
      "grad_norm": 0.7507545948028564,
      "learning_rate": 0.00021709444144630326,
      "loss": 3.2795,
      "step": 59250
    },
    {
      "epoch": 6.393530997304582,
      "grad_norm": 0.7440924644470215,
      "learning_rate": 0.00021677064220183484,
      "loss": 3.2736,
      "step": 59300
    },
    {
      "epoch": 6.398921832884097,
      "grad_norm": 0.7490218281745911,
      "learning_rate": 0.00021644684295736642,
      "loss": 3.2633,
      "step": 59350
    },
    {
      "epoch": 6.404312668463612,
      "grad_norm": 0.7519710659980774,
      "learning_rate": 0.00021612304371289797,
      "loss": 3.2532,
      "step": 59400
    },
    {
      "epoch": 6.409703504043127,
      "grad_norm": 0.7268811464309692,
      "learning_rate": 0.00021579924446842955,
      "loss": 3.2831,
      "step": 59450
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 0.7365671396255493,
      "learning_rate": 0.00021547544522396113,
      "loss": 3.2734,
      "step": 59500
    },
    {
      "epoch": 6.420485175202156,
      "grad_norm": 0.784694492816925,
      "learning_rate": 0.0002151516459794927,
      "loss": 3.2898,
      "step": 59550
    },
    {
      "epoch": 6.425876010781671,
      "grad_norm": 0.7644874453544617,
      "learning_rate": 0.00021482784673502428,
      "loss": 3.2836,
      "step": 59600
    },
    {
      "epoch": 6.431266846361186,
      "grad_norm": 0.7010469436645508,
      "learning_rate": 0.00021450404749055586,
      "loss": 3.2646,
      "step": 59650
    },
    {
      "epoch": 6.436657681940701,
      "grad_norm": 0.7214942574501038,
      "learning_rate": 0.00021418024824608738,
      "loss": 3.2619,
      "step": 59700
    },
    {
      "epoch": 6.442048517520216,
      "grad_norm": 0.7231842875480652,
      "learning_rate": 0.00021385644900161896,
      "loss": 3.278,
      "step": 59750
    },
    {
      "epoch": 6.44743935309973,
      "grad_norm": 0.728193998336792,
      "learning_rate": 0.00021353264975715054,
      "loss": 3.2837,
      "step": 59800
    },
    {
      "epoch": 6.452830188679245,
      "grad_norm": 0.768265426158905,
      "learning_rate": 0.00021320885051268211,
      "loss": 3.2929,
      "step": 59850
    },
    {
      "epoch": 6.45822102425876,
      "grad_norm": 0.7742714285850525,
      "learning_rate": 0.0002128850512682137,
      "loss": 3.2794,
      "step": 59900
    },
    {
      "epoch": 6.463611859838275,
      "grad_norm": 0.6836204528808594,
      "learning_rate": 0.00021256125202374527,
      "loss": 3.2765,
      "step": 59950
    },
    {
      "epoch": 6.46900269541779,
      "grad_norm": 0.9617937207221985,
      "learning_rate": 0.00021223745277927685,
      "loss": 3.2893,
      "step": 60000
    },
    {
      "epoch": 6.46900269541779,
      "eval_accuracy": 0.38564234663726377,
      "eval_loss": 3.3716087341308594,
      "eval_runtime": 184.7027,
      "eval_samples_per_second": 97.513,
      "eval_steps_per_second": 6.096,
      "step": 60000
    },
    {
      "epoch": 6.474393530997305,
      "grad_norm": 0.7370874881744385,
      "learning_rate": 0.00021191365353480843,
      "loss": 3.2907,
      "step": 60050
    },
    {
      "epoch": 6.479784366576819,
      "grad_norm": 0.7189035415649414,
      "learning_rate": 0.00021158985429033995,
      "loss": 3.2801,
      "step": 60100
    },
    {
      "epoch": 6.485175202156334,
      "grad_norm": 0.7072881460189819,
      "learning_rate": 0.00021126605504587153,
      "loss": 3.2779,
      "step": 60150
    },
    {
      "epoch": 6.490566037735849,
      "grad_norm": 0.711500883102417,
      "learning_rate": 0.0002109422558014031,
      "loss": 3.2626,
      "step": 60200
    },
    {
      "epoch": 6.495956873315364,
      "grad_norm": 0.7228755950927734,
      "learning_rate": 0.00021061845655693468,
      "loss": 3.2895,
      "step": 60250
    },
    {
      "epoch": 6.501347708894879,
      "grad_norm": 0.7080203890800476,
      "learning_rate": 0.00021029465731246626,
      "loss": 3.2776,
      "step": 60300
    },
    {
      "epoch": 6.506738544474393,
      "grad_norm": 0.7516891360282898,
      "learning_rate": 0.00020997085806799784,
      "loss": 3.2753,
      "step": 60350
    },
    {
      "epoch": 6.512129380053908,
      "grad_norm": 0.6993798017501831,
      "learning_rate": 0.0002096470588235294,
      "loss": 3.2688,
      "step": 60400
    },
    {
      "epoch": 6.517520215633423,
      "grad_norm": 0.7494493126869202,
      "learning_rate": 0.00020932325957906097,
      "loss": 3.2716,
      "step": 60450
    },
    {
      "epoch": 6.5229110512129385,
      "grad_norm": 0.7326953411102295,
      "learning_rate": 0.00020899946033459254,
      "loss": 3.2797,
      "step": 60500
    },
    {
      "epoch": 6.528301886792453,
      "grad_norm": 0.7331379652023315,
      "learning_rate": 0.0002086756610901241,
      "loss": 3.272,
      "step": 60550
    },
    {
      "epoch": 6.533692722371968,
      "grad_norm": 0.7465304136276245,
      "learning_rate": 0.00020835186184565567,
      "loss": 3.2809,
      "step": 60600
    },
    {
      "epoch": 6.539083557951482,
      "grad_norm": 0.6983660459518433,
      "learning_rate": 0.00020802806260118725,
      "loss": 3.283,
      "step": 60650
    },
    {
      "epoch": 6.544474393530997,
      "grad_norm": 0.7236700654029846,
      "learning_rate": 0.00020770426335671883,
      "loss": 3.2601,
      "step": 60700
    },
    {
      "epoch": 6.549865229110512,
      "grad_norm": 0.794366717338562,
      "learning_rate": 0.00020738694009713975,
      "loss": 3.2861,
      "step": 60750
    },
    {
      "epoch": 6.555256064690027,
      "grad_norm": 0.7264977097511292,
      "learning_rate": 0.00020706314085267133,
      "loss": 3.3007,
      "step": 60800
    },
    {
      "epoch": 6.560646900269542,
      "grad_norm": 0.7221018075942993,
      "learning_rate": 0.00020673934160820288,
      "loss": 3.2797,
      "step": 60850
    },
    {
      "epoch": 6.566037735849057,
      "grad_norm": 0.7276595830917358,
      "learning_rate": 0.00020641554236373446,
      "loss": 3.2759,
      "step": 60900
    },
    {
      "epoch": 6.571428571428571,
      "grad_norm": 0.7512069940567017,
      "learning_rate": 0.00020609174311926604,
      "loss": 3.2959,
      "step": 60950
    },
    {
      "epoch": 6.576819407008086,
      "grad_norm": 0.7264222502708435,
      "learning_rate": 0.0002057679438747976,
      "loss": 3.2819,
      "step": 61000
    },
    {
      "epoch": 6.576819407008086,
      "eval_accuracy": 0.38633196651622226,
      "eval_loss": 3.366112470626831,
      "eval_runtime": 184.763,
      "eval_samples_per_second": 97.482,
      "eval_steps_per_second": 6.094,
      "step": 61000
    },
    {
      "epoch": 6.5822102425876015,
      "grad_norm": 0.7526605725288391,
      "learning_rate": 0.00020544414463032916,
      "loss": 3.2952,
      "step": 61050
    },
    {
      "epoch": 6.587601078167116,
      "grad_norm": 0.7065187692642212,
      "learning_rate": 0.00020512034538586074,
      "loss": 3.263,
      "step": 61100
    },
    {
      "epoch": 6.592991913746631,
      "grad_norm": 0.767744779586792,
      "learning_rate": 0.00020479654614139232,
      "loss": 3.3016,
      "step": 61150
    },
    {
      "epoch": 6.598382749326145,
      "grad_norm": 0.7380595803260803,
      "learning_rate": 0.0002044727468969239,
      "loss": 3.2851,
      "step": 61200
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 0.6881290674209595,
      "learning_rate": 0.00020414894765245547,
      "loss": 3.2706,
      "step": 61250
    },
    {
      "epoch": 6.609164420485175,
      "grad_norm": 0.6972199082374573,
      "learning_rate": 0.00020382514840798702,
      "loss": 3.2987,
      "step": 61300
    },
    {
      "epoch": 6.6145552560646905,
      "grad_norm": 0.7250159382820129,
      "learning_rate": 0.0002035013491635186,
      "loss": 3.2942,
      "step": 61350
    },
    {
      "epoch": 6.619946091644205,
      "grad_norm": 0.7611603736877441,
      "learning_rate": 0.00020317754991905015,
      "loss": 3.2798,
      "step": 61400
    },
    {
      "epoch": 6.62533692722372,
      "grad_norm": 0.7625829577445984,
      "learning_rate": 0.00020285375067458173,
      "loss": 3.2658,
      "step": 61450
    },
    {
      "epoch": 6.630727762803234,
      "grad_norm": 0.7896705865859985,
      "learning_rate": 0.0002025299514301133,
      "loss": 3.2919,
      "step": 61500
    },
    {
      "epoch": 6.636118598382749,
      "grad_norm": 0.75810706615448,
      "learning_rate": 0.00020220615218564489,
      "loss": 3.278,
      "step": 61550
    },
    {
      "epoch": 6.6415094339622645,
      "grad_norm": 0.7625492215156555,
      "learning_rate": 0.00020188235294117646,
      "loss": 3.2772,
      "step": 61600
    },
    {
      "epoch": 6.646900269541779,
      "grad_norm": 0.7140247821807861,
      "learning_rate": 0.00020155855369670804,
      "loss": 3.2875,
      "step": 61650
    },
    {
      "epoch": 6.652291105121294,
      "grad_norm": 0.7512028217315674,
      "learning_rate": 0.00020123475445223956,
      "loss": 3.2969,
      "step": 61700
    },
    {
      "epoch": 6.657681940700809,
      "grad_norm": 0.7267550826072693,
      "learning_rate": 0.00020091095520777114,
      "loss": 3.2861,
      "step": 61750
    },
    {
      "epoch": 6.663072776280323,
      "grad_norm": 0.7511851787567139,
      "learning_rate": 0.00020058715596330272,
      "loss": 3.2806,
      "step": 61800
    },
    {
      "epoch": 6.668463611859838,
      "grad_norm": 0.777556836605072,
      "learning_rate": 0.0002002633567188343,
      "loss": 3.2767,
      "step": 61850
    },
    {
      "epoch": 6.6738544474393535,
      "grad_norm": 0.7368999123573303,
      "learning_rate": 0.00019993955747436588,
      "loss": 3.274,
      "step": 61900
    },
    {
      "epoch": 6.679245283018868,
      "grad_norm": 0.7757512331008911,
      "learning_rate": 0.00019961575822989745,
      "loss": 3.2747,
      "step": 61950
    },
    {
      "epoch": 6.684636118598383,
      "grad_norm": 0.7147890329360962,
      "learning_rate": 0.00019929195898542903,
      "loss": 3.2998,
      "step": 62000
    },
    {
      "epoch": 6.684636118598383,
      "eval_accuracy": 0.38667759135236013,
      "eval_loss": 3.3609557151794434,
      "eval_runtime": 185.0014,
      "eval_samples_per_second": 97.356,
      "eval_steps_per_second": 6.086,
      "step": 62000
    },
    {
      "epoch": 6.690026954177897,
      "grad_norm": 0.7304677367210388,
      "learning_rate": 0.0001989681597409606,
      "loss": 3.2825,
      "step": 62050
    },
    {
      "epoch": 6.695417789757412,
      "grad_norm": 0.7841406464576721,
      "learning_rate": 0.00019864436049649216,
      "loss": 3.2635,
      "step": 62100
    },
    {
      "epoch": 6.7008086253369274,
      "grad_norm": 0.7494974732398987,
      "learning_rate": 0.0001983205612520237,
      "loss": 3.3044,
      "step": 62150
    },
    {
      "epoch": 6.706199460916442,
      "grad_norm": 0.7478502988815308,
      "learning_rate": 0.0001979967620075553,
      "loss": 3.2954,
      "step": 62200
    },
    {
      "epoch": 6.711590296495957,
      "grad_norm": 0.7201170325279236,
      "learning_rate": 0.00019767296276308686,
      "loss": 3.2798,
      "step": 62250
    },
    {
      "epoch": 6.716981132075472,
      "grad_norm": 0.7911655902862549,
      "learning_rate": 0.00019734916351861844,
      "loss": 3.2812,
      "step": 62300
    },
    {
      "epoch": 6.722371967654986,
      "grad_norm": 0.8429829478263855,
      "learning_rate": 0.00019702536427415002,
      "loss": 3.3016,
      "step": 62350
    },
    {
      "epoch": 6.727762803234501,
      "grad_norm": 0.7752325534820557,
      "learning_rate": 0.0001967015650296816,
      "loss": 3.2843,
      "step": 62400
    },
    {
      "epoch": 6.7331536388140165,
      "grad_norm": 0.7346392273902893,
      "learning_rate": 0.00019637776578521315,
      "loss": 3.2862,
      "step": 62450
    },
    {
      "epoch": 6.738544474393531,
      "grad_norm": 0.7346991300582886,
      "learning_rate": 0.00019605396654074473,
      "loss": 3.2908,
      "step": 62500
    },
    {
      "epoch": 6.743935309973046,
      "grad_norm": 0.7638818025588989,
      "learning_rate": 0.0001957301672962763,
      "loss": 3.2882,
      "step": 62550
    },
    {
      "epoch": 6.74932614555256,
      "grad_norm": 0.7539122104644775,
      "learning_rate": 0.00019540636805180785,
      "loss": 3.2801,
      "step": 62600
    },
    {
      "epoch": 6.754716981132075,
      "grad_norm": 0.7578554153442383,
      "learning_rate": 0.00019508256880733943,
      "loss": 3.2811,
      "step": 62650
    },
    {
      "epoch": 6.7601078167115904,
      "grad_norm": 0.7557489275932312,
      "learning_rate": 0.000194758769562871,
      "loss": 3.2902,
      "step": 62700
    },
    {
      "epoch": 6.765498652291106,
      "grad_norm": 0.7585029006004333,
      "learning_rate": 0.00019443497031840256,
      "loss": 3.2708,
      "step": 62750
    },
    {
      "epoch": 6.77088948787062,
      "grad_norm": 0.7600674629211426,
      "learning_rate": 0.00019411117107393414,
      "loss": 3.2731,
      "step": 62800
    },
    {
      "epoch": 6.776280323450135,
      "grad_norm": 0.7798412442207336,
      "learning_rate": 0.00019378737182946572,
      "loss": 3.2889,
      "step": 62850
    },
    {
      "epoch": 6.781671159029649,
      "grad_norm": 0.8409514427185059,
      "learning_rate": 0.0001934635725849973,
      "loss": 3.2864,
      "step": 62900
    },
    {
      "epoch": 6.787061994609164,
      "grad_norm": 0.7428311705589294,
      "learning_rate": 0.00019313977334052887,
      "loss": 3.3036,
      "step": 62950
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 0.7958593964576721,
      "learning_rate": 0.00019281597409606042,
      "loss": 3.2773,
      "step": 63000
    },
    {
      "epoch": 6.7924528301886795,
      "eval_accuracy": 0.38693759771290015,
      "eval_loss": 3.3576364517211914,
      "eval_runtime": 185.2969,
      "eval_samples_per_second": 97.201,
      "eval_steps_per_second": 6.077,
      "step": 63000
    },
    {
      "epoch": 6.797843665768194,
      "grad_norm": 0.757053017616272,
      "learning_rate": 0.00019249217485159197,
      "loss": 3.291,
      "step": 63050
    },
    {
      "epoch": 6.803234501347709,
      "grad_norm": 0.703620433807373,
      "learning_rate": 0.00019216837560712355,
      "loss": 3.2964,
      "step": 63100
    },
    {
      "epoch": 6.808625336927224,
      "grad_norm": 0.7836263179779053,
      "learning_rate": 0.00019184457636265513,
      "loss": 3.2749,
      "step": 63150
    },
    {
      "epoch": 6.814016172506738,
      "grad_norm": 0.7800487279891968,
      "learning_rate": 0.0001915207771181867,
      "loss": 3.2889,
      "step": 63200
    },
    {
      "epoch": 6.819407008086253,
      "grad_norm": 0.7118497490882874,
      "learning_rate": 0.00019119697787371828,
      "loss": 3.2898,
      "step": 63250
    },
    {
      "epoch": 6.824797843665769,
      "grad_norm": 0.7383244037628174,
      "learning_rate": 0.00019087317862924986,
      "loss": 3.2806,
      "step": 63300
    },
    {
      "epoch": 6.830188679245283,
      "grad_norm": 0.8963047862052917,
      "learning_rate": 0.00019054937938478144,
      "loss": 3.2644,
      "step": 63350
    },
    {
      "epoch": 6.835579514824798,
      "grad_norm": 0.7486133575439453,
      "learning_rate": 0.00019022558014031302,
      "loss": 3.295,
      "step": 63400
    },
    {
      "epoch": 6.840970350404312,
      "grad_norm": 0.7660999894142151,
      "learning_rate": 0.00018990178089584454,
      "loss": 3.2999,
      "step": 63450
    },
    {
      "epoch": 6.846361185983827,
      "grad_norm": 0.7520319223403931,
      "learning_rate": 0.00018957798165137612,
      "loss": 3.2928,
      "step": 63500
    },
    {
      "epoch": 6.8517520215633425,
      "grad_norm": 0.7498881220817566,
      "learning_rate": 0.0001892541824069077,
      "loss": 3.2875,
      "step": 63550
    },
    {
      "epoch": 6.857142857142857,
      "grad_norm": 0.7376906275749207,
      "learning_rate": 0.00018893038316243927,
      "loss": 3.2675,
      "step": 63600
    },
    {
      "epoch": 6.862533692722372,
      "grad_norm": 0.7770941853523254,
      "learning_rate": 0.00018860658391797085,
      "loss": 3.2695,
      "step": 63650
    },
    {
      "epoch": 6.867924528301887,
      "grad_norm": 0.7603805661201477,
      "learning_rate": 0.00018828278467350243,
      "loss": 3.278,
      "step": 63700
    },
    {
      "epoch": 6.873315363881401,
      "grad_norm": 0.7244340181350708,
      "learning_rate": 0.000187958985429034,
      "loss": 3.2774,
      "step": 63750
    },
    {
      "epoch": 6.878706199460916,
      "grad_norm": 0.8087451457977295,
      "learning_rate": 0.00018763518618456556,
      "loss": 3.2684,
      "step": 63800
    },
    {
      "epoch": 6.884097035040432,
      "grad_norm": 0.7538607120513916,
      "learning_rate": 0.0001873113869400971,
      "loss": 3.2879,
      "step": 63850
    },
    {
      "epoch": 6.889487870619946,
      "grad_norm": 0.7293424606323242,
      "learning_rate": 0.00018698758769562868,
      "loss": 3.2841,
      "step": 63900
    },
    {
      "epoch": 6.894878706199461,
      "grad_norm": 0.8110401034355164,
      "learning_rate": 0.00018666378845116026,
      "loss": 3.302,
      "step": 63950
    },
    {
      "epoch": 6.900269541778976,
      "grad_norm": 0.7437310814857483,
      "learning_rate": 0.00018633998920669184,
      "loss": 3.2882,
      "step": 64000
    },
    {
      "epoch": 6.900269541778976,
      "eval_accuracy": 0.38752508387731305,
      "eval_loss": 3.353229522705078,
      "eval_runtime": 184.6958,
      "eval_samples_per_second": 97.517,
      "eval_steps_per_second": 6.097,
      "step": 64000
    },
    {
      "epoch": 6.90566037735849,
      "grad_norm": 0.7894251346588135,
      "learning_rate": 0.00018601618996222342,
      "loss": 3.2894,
      "step": 64050
    },
    {
      "epoch": 6.9110512129380055,
      "grad_norm": 0.7877970933914185,
      "learning_rate": 0.00018569239071775497,
      "loss": 3.2938,
      "step": 64100
    },
    {
      "epoch": 6.916442048517521,
      "grad_norm": 0.7726397514343262,
      "learning_rate": 0.00018536859147328655,
      "loss": 3.2889,
      "step": 64150
    },
    {
      "epoch": 6.921832884097035,
      "grad_norm": 0.7730385065078735,
      "learning_rate": 0.00018504479222881812,
      "loss": 3.2796,
      "step": 64200
    },
    {
      "epoch": 6.92722371967655,
      "grad_norm": 0.7642452716827393,
      "learning_rate": 0.0001847209929843497,
      "loss": 3.264,
      "step": 64250
    },
    {
      "epoch": 6.932614555256064,
      "grad_norm": 0.7660939693450928,
      "learning_rate": 0.00018439719373988125,
      "loss": 3.306,
      "step": 64300
    },
    {
      "epoch": 6.938005390835579,
      "grad_norm": 0.7493078112602234,
      "learning_rate": 0.00018407339449541283,
      "loss": 3.2839,
      "step": 64350
    },
    {
      "epoch": 6.943396226415095,
      "grad_norm": 0.7475608587265015,
      "learning_rate": 0.00018374959525094438,
      "loss": 3.2844,
      "step": 64400
    },
    {
      "epoch": 6.948787061994609,
      "grad_norm": 0.7330486178398132,
      "learning_rate": 0.00018342579600647596,
      "loss": 3.2808,
      "step": 64450
    },
    {
      "epoch": 6.954177897574124,
      "grad_norm": 1.2176107168197632,
      "learning_rate": 0.00018310199676200753,
      "loss": 3.2776,
      "step": 64500
    },
    {
      "epoch": 6.959568733153639,
      "grad_norm": 0.710422933101654,
      "learning_rate": 0.0001827781975175391,
      "loss": 3.2817,
      "step": 64550
    },
    {
      "epoch": 6.964959568733153,
      "grad_norm": 0.7552564144134521,
      "learning_rate": 0.0001824543982730707,
      "loss": 3.2731,
      "step": 64600
    },
    {
      "epoch": 6.9703504043126685,
      "grad_norm": 0.7585906386375427,
      "learning_rate": 0.00018213059902860227,
      "loss": 3.2843,
      "step": 64650
    },
    {
      "epoch": 6.975741239892184,
      "grad_norm": 0.7237086892127991,
      "learning_rate": 0.00018180679978413382,
      "loss": 3.2802,
      "step": 64700
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 0.8057036995887756,
      "learning_rate": 0.00018148300053966537,
      "loss": 3.2895,
      "step": 64750
    },
    {
      "epoch": 6.986522911051213,
      "grad_norm": 0.7535995244979858,
      "learning_rate": 0.00018116567728008632,
      "loss": 3.277,
      "step": 64800
    },
    {
      "epoch": 6.991913746630727,
      "grad_norm": 0.7500599026679993,
      "learning_rate": 0.0001808418780356179,
      "loss": 3.2713,
      "step": 64850
    },
    {
      "epoch": 6.997304582210242,
      "grad_norm": 0.8092964291572571,
      "learning_rate": 0.00018051807879114948,
      "loss": 3.2817,
      "step": 64900
    },
    {
      "epoch": 7.002695417789758,
      "grad_norm": 0.7631925940513611,
      "learning_rate": 0.0001802007555315704,
      "loss": 3.2317,
      "step": 64950
    },
    {
      "epoch": 7.008086253369272,
      "grad_norm": 0.7095732092857361,
      "learning_rate": 0.00017987695628710198,
      "loss": 3.1839,
      "step": 65000
    },
    {
      "epoch": 7.008086253369272,
      "eval_accuracy": 0.3875124801423266,
      "eval_loss": 3.354387044906616,
      "eval_runtime": 184.7817,
      "eval_samples_per_second": 97.472,
      "eval_steps_per_second": 6.094,
      "step": 65000
    },
    {
      "epoch": 7.013477088948787,
      "grad_norm": 0.9158305525779724,
      "learning_rate": 0.00017955315704263356,
      "loss": 3.2062,
      "step": 65050
    },
    {
      "epoch": 7.018867924528302,
      "grad_norm": 0.7549683451652527,
      "learning_rate": 0.0001792293577981651,
      "loss": 3.2018,
      "step": 65100
    },
    {
      "epoch": 7.024258760107816,
      "grad_norm": 0.7751328349113464,
      "learning_rate": 0.00017890555855369668,
      "loss": 3.2119,
      "step": 65150
    },
    {
      "epoch": 7.0296495956873315,
      "grad_norm": 0.7429438829421997,
      "learning_rate": 0.00017858175930922826,
      "loss": 3.2053,
      "step": 65200
    },
    {
      "epoch": 7.035040431266847,
      "grad_norm": 0.7653865218162537,
      "learning_rate": 0.00017825796006475984,
      "loss": 3.2041,
      "step": 65250
    },
    {
      "epoch": 7.040431266846361,
      "grad_norm": 0.7375260591506958,
      "learning_rate": 0.00017793416082029142,
      "loss": 3.1978,
      "step": 65300
    },
    {
      "epoch": 7.045822102425876,
      "grad_norm": 0.7559249997138977,
      "learning_rate": 0.000177610361575823,
      "loss": 3.2147,
      "step": 65350
    },
    {
      "epoch": 7.051212938005391,
      "grad_norm": 0.7759532332420349,
      "learning_rate": 0.00017728656233135452,
      "loss": 3.1921,
      "step": 65400
    },
    {
      "epoch": 7.056603773584905,
      "grad_norm": 0.7644091844558716,
      "learning_rate": 0.0001769627630868861,
      "loss": 3.1814,
      "step": 65450
    },
    {
      "epoch": 7.061994609164421,
      "grad_norm": 0.7937024235725403,
      "learning_rate": 0.00017663896384241767,
      "loss": 3.1995,
      "step": 65500
    },
    {
      "epoch": 7.067385444743936,
      "grad_norm": 0.7552359104156494,
      "learning_rate": 0.00017631516459794925,
      "loss": 3.1924,
      "step": 65550
    },
    {
      "epoch": 7.07277628032345,
      "grad_norm": 0.7388563752174377,
      "learning_rate": 0.00017599136535348083,
      "loss": 3.2205,
      "step": 65600
    },
    {
      "epoch": 7.078167115902965,
      "grad_norm": 0.7597045302391052,
      "learning_rate": 0.0001756675661090124,
      "loss": 3.2121,
      "step": 65650
    },
    {
      "epoch": 7.083557951482479,
      "grad_norm": 0.8127196431159973,
      "learning_rate": 0.00017534376686454398,
      "loss": 3.2061,
      "step": 65700
    },
    {
      "epoch": 7.0889487870619945,
      "grad_norm": 0.7847875356674194,
      "learning_rate": 0.00017501996762007556,
      "loss": 3.1891,
      "step": 65750
    },
    {
      "epoch": 7.09433962264151,
      "grad_norm": 0.7454573512077332,
      "learning_rate": 0.00017469616837560709,
      "loss": 3.2224,
      "step": 65800
    },
    {
      "epoch": 7.099730458221024,
      "grad_norm": 0.739266574382782,
      "learning_rate": 0.00017437236913113866,
      "loss": 3.2168,
      "step": 65850
    },
    {
      "epoch": 7.105121293800539,
      "grad_norm": 0.7397430539131165,
      "learning_rate": 0.00017404856988667024,
      "loss": 3.2299,
      "step": 65900
    },
    {
      "epoch": 7.110512129380054,
      "grad_norm": 0.7771823406219482,
      "learning_rate": 0.00017372477064220182,
      "loss": 3.2127,
      "step": 65950
    },
    {
      "epoch": 7.115902964959568,
      "grad_norm": 0.8010558485984802,
      "learning_rate": 0.0001734009713977334,
      "loss": 3.2252,
      "step": 66000
    },
    {
      "epoch": 7.115902964959568,
      "eval_accuracy": 0.3877516251484063,
      "eval_loss": 3.3572838306427,
      "eval_runtime": 184.947,
      "eval_samples_per_second": 97.385,
      "eval_steps_per_second": 6.088,
      "step": 66000
    },
    {
      "epoch": 7.121293800539084,
      "grad_norm": 0.7524279952049255,
      "learning_rate": 0.00017307717215326497,
      "loss": 3.2191,
      "step": 66050
    },
    {
      "epoch": 7.126684636118599,
      "grad_norm": 0.7754007577896118,
      "learning_rate": 0.00017275337290879655,
      "loss": 3.2188,
      "step": 66100
    },
    {
      "epoch": 7.132075471698113,
      "grad_norm": 0.8016265630722046,
      "learning_rate": 0.0001724295736643281,
      "loss": 3.2193,
      "step": 66150
    },
    {
      "epoch": 7.137466307277628,
      "grad_norm": 0.7195385098457336,
      "learning_rate": 0.00017210577441985968,
      "loss": 3.2144,
      "step": 66200
    },
    {
      "epoch": 7.142857142857143,
      "grad_norm": 0.7985889911651611,
      "learning_rate": 0.00017178197517539123,
      "loss": 3.2224,
      "step": 66250
    },
    {
      "epoch": 7.1482479784366575,
      "grad_norm": 0.7573408484458923,
      "learning_rate": 0.0001714581759309228,
      "loss": 3.2186,
      "step": 66300
    },
    {
      "epoch": 7.153638814016173,
      "grad_norm": 0.8037809729576111,
      "learning_rate": 0.00017113437668645439,
      "loss": 3.2227,
      "step": 66350
    },
    {
      "epoch": 7.159029649595688,
      "grad_norm": 0.7795762419700623,
      "learning_rate": 0.00017081057744198596,
      "loss": 3.2077,
      "step": 66400
    },
    {
      "epoch": 7.164420485175202,
      "grad_norm": 0.7938854098320007,
      "learning_rate": 0.00017048677819751751,
      "loss": 3.2172,
      "step": 66450
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 0.7674042582511902,
      "learning_rate": 0.0001701629789530491,
      "loss": 3.2177,
      "step": 66500
    },
    {
      "epoch": 7.175202156334231,
      "grad_norm": 0.782508134841919,
      "learning_rate": 0.00016983917970858067,
      "loss": 3.2215,
      "step": 66550
    },
    {
      "epoch": 7.180592991913747,
      "grad_norm": 0.7263006567955017,
      "learning_rate": 0.00016951538046411225,
      "loss": 3.2079,
      "step": 66600
    },
    {
      "epoch": 7.185983827493262,
      "grad_norm": 0.7729055881500244,
      "learning_rate": 0.0001691915812196438,
      "loss": 3.215,
      "step": 66650
    },
    {
      "epoch": 7.191374663072776,
      "grad_norm": 0.7278570532798767,
      "learning_rate": 0.00016886778197517538,
      "loss": 3.2179,
      "step": 66700
    },
    {
      "epoch": 7.196765498652291,
      "grad_norm": 0.7499024271965027,
      "learning_rate": 0.00016854398273070693,
      "loss": 3.2216,
      "step": 66750
    },
    {
      "epoch": 7.202156334231806,
      "grad_norm": 0.8059422373771667,
      "learning_rate": 0.0001682201834862385,
      "loss": 3.211,
      "step": 66800
    },
    {
      "epoch": 7.2075471698113205,
      "grad_norm": 0.7714557647705078,
      "learning_rate": 0.00016789638424177008,
      "loss": 3.2301,
      "step": 66850
    },
    {
      "epoch": 7.212938005390836,
      "grad_norm": 0.7622218728065491,
      "learning_rate": 0.00016757258499730166,
      "loss": 3.2388,
      "step": 66900
    },
    {
      "epoch": 7.218328840970351,
      "grad_norm": 0.7789103984832764,
      "learning_rate": 0.00016724878575283324,
      "loss": 3.227,
      "step": 66950
    },
    {
      "epoch": 7.223719676549865,
      "grad_norm": 0.7892781496047974,
      "learning_rate": 0.00016692498650836481,
      "loss": 3.2308,
      "step": 67000
    },
    {
      "epoch": 7.223719676549865,
      "eval_accuracy": 0.388054006135194,
      "eval_loss": 3.3542799949645996,
      "eval_runtime": 184.5856,
      "eval_samples_per_second": 97.575,
      "eval_steps_per_second": 6.1,
      "step": 67000
    },
    {
      "epoch": 7.22911051212938,
      "grad_norm": 0.772681474685669,
      "learning_rate": 0.0001666011872638964,
      "loss": 3.1994,
      "step": 67050
    },
    {
      "epoch": 7.234501347708895,
      "grad_norm": 0.7405024766921997,
      "learning_rate": 0.00016627738801942792,
      "loss": 3.2434,
      "step": 67100
    },
    {
      "epoch": 7.2398921832884096,
      "grad_norm": 0.7997522354125977,
      "learning_rate": 0.00016596006475984887,
      "loss": 3.2065,
      "step": 67150
    },
    {
      "epoch": 7.245283018867925,
      "grad_norm": 0.8065983057022095,
      "learning_rate": 0.00016563626551538045,
      "loss": 3.2268,
      "step": 67200
    },
    {
      "epoch": 7.250673854447439,
      "grad_norm": 0.7845258116722107,
      "learning_rate": 0.00016531246627091202,
      "loss": 3.2351,
      "step": 67250
    },
    {
      "epoch": 7.256064690026954,
      "grad_norm": 0.7305331826210022,
      "learning_rate": 0.0001649886670264436,
      "loss": 3.237,
      "step": 67300
    },
    {
      "epoch": 7.261455525606469,
      "grad_norm": 0.7165262699127197,
      "learning_rate": 0.00016466486778197518,
      "loss": 3.2162,
      "step": 67350
    },
    {
      "epoch": 7.2668463611859835,
      "grad_norm": 0.7781831622123718,
      "learning_rate": 0.0001643410685375067,
      "loss": 3.2421,
      "step": 67400
    },
    {
      "epoch": 7.272237196765499,
      "grad_norm": 0.8223204016685486,
      "learning_rate": 0.00016401726929303828,
      "loss": 3.2393,
      "step": 67450
    },
    {
      "epoch": 7.277628032345014,
      "grad_norm": 0.7840092778205872,
      "learning_rate": 0.00016369347004856986,
      "loss": 3.229,
      "step": 67500
    },
    {
      "epoch": 7.283018867924528,
      "grad_norm": 0.7795689702033997,
      "learning_rate": 0.00016336967080410143,
      "loss": 3.2406,
      "step": 67550
    },
    {
      "epoch": 7.288409703504043,
      "grad_norm": 0.8028538823127747,
      "learning_rate": 0.000163045871559633,
      "loss": 3.2256,
      "step": 67600
    },
    {
      "epoch": 7.293800539083558,
      "grad_norm": 0.7559919953346252,
      "learning_rate": 0.0001627220723151646,
      "loss": 3.228,
      "step": 67650
    },
    {
      "epoch": 7.2991913746630726,
      "grad_norm": 0.7705023884773254,
      "learning_rate": 0.00016239827307069617,
      "loss": 3.2099,
      "step": 67700
    },
    {
      "epoch": 7.304582210242588,
      "grad_norm": 0.7943682670593262,
      "learning_rate": 0.00016207447382622775,
      "loss": 3.2247,
      "step": 67750
    },
    {
      "epoch": 7.309973045822103,
      "grad_norm": 0.7587892413139343,
      "learning_rate": 0.00016175067458175932,
      "loss": 3.2219,
      "step": 67800
    },
    {
      "epoch": 7.315363881401617,
      "grad_norm": 0.8514966368675232,
      "learning_rate": 0.00016142687533729085,
      "loss": 3.2224,
      "step": 67850
    },
    {
      "epoch": 7.320754716981132,
      "grad_norm": 0.7471888065338135,
      "learning_rate": 0.00016110307609282242,
      "loss": 3.237,
      "step": 67900
    },
    {
      "epoch": 7.3261455525606465,
      "grad_norm": 0.7487116456031799,
      "learning_rate": 0.000160779276848354,
      "loss": 3.2122,
      "step": 67950
    },
    {
      "epoch": 7.331536388140162,
      "grad_norm": 0.8473713397979736,
      "learning_rate": 0.00016045547760388558,
      "loss": 3.2282,
      "step": 68000
    },
    {
      "epoch": 7.331536388140162,
      "eval_accuracy": 0.3882555572420898,
      "eval_loss": 3.352534532546997,
      "eval_runtime": 184.9997,
      "eval_samples_per_second": 97.357,
      "eval_steps_per_second": 6.086,
      "step": 68000
    },
    {
      "epoch": 7.336927223719677,
      "grad_norm": 0.7282635569572449,
      "learning_rate": 0.00016013167835941716,
      "loss": 3.2332,
      "step": 68050
    },
    {
      "epoch": 7.342318059299191,
      "grad_norm": 0.7940888404846191,
      "learning_rate": 0.00015980787911494873,
      "loss": 3.2392,
      "step": 68100
    },
    {
      "epoch": 7.347708894878706,
      "grad_norm": 0.8099002242088318,
      "learning_rate": 0.00015948407987048029,
      "loss": 3.2271,
      "step": 68150
    },
    {
      "epoch": 7.353099730458221,
      "grad_norm": 0.7751094102859497,
      "learning_rate": 0.00015916028062601186,
      "loss": 3.2459,
      "step": 68200
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 0.7973043918609619,
      "learning_rate": 0.00015883648138154344,
      "loss": 3.2324,
      "step": 68250
    },
    {
      "epoch": 7.363881401617251,
      "grad_norm": 0.787239670753479,
      "learning_rate": 0.000158512682137075,
      "loss": 3.2199,
      "step": 68300
    },
    {
      "epoch": 7.369272237196766,
      "grad_norm": 0.7251448631286621,
      "learning_rate": 0.00015818888289260657,
      "loss": 3.2329,
      "step": 68350
    },
    {
      "epoch": 7.37466307277628,
      "grad_norm": 0.7860444188117981,
      "learning_rate": 0.00015786508364813815,
      "loss": 3.225,
      "step": 68400
    },
    {
      "epoch": 7.380053908355795,
      "grad_norm": 0.8149095773696899,
      "learning_rate": 0.0001575412844036697,
      "loss": 3.2095,
      "step": 68450
    },
    {
      "epoch": 7.38544474393531,
      "grad_norm": 0.8455824255943298,
      "learning_rate": 0.00015721748515920127,
      "loss": 3.2296,
      "step": 68500
    },
    {
      "epoch": 7.390835579514825,
      "grad_norm": 0.7530245780944824,
      "learning_rate": 0.00015689368591473285,
      "loss": 3.2297,
      "step": 68550
    },
    {
      "epoch": 7.39622641509434,
      "grad_norm": 0.7616867423057556,
      "learning_rate": 0.00015656988667026443,
      "loss": 3.2266,
      "step": 68600
    },
    {
      "epoch": 7.401617250673855,
      "grad_norm": 0.7562954425811768,
      "learning_rate": 0.000156246087425796,
      "loss": 3.2275,
      "step": 68650
    },
    {
      "epoch": 7.407008086253369,
      "grad_norm": 0.7328892350196838,
      "learning_rate": 0.00015592228818132756,
      "loss": 3.2381,
      "step": 68700
    },
    {
      "epoch": 7.412398921832884,
      "grad_norm": 0.8182516694068909,
      "learning_rate": 0.0001555984889368591,
      "loss": 3.2225,
      "step": 68750
    },
    {
      "epoch": 7.4177897574123985,
      "grad_norm": 0.8202837109565735,
      "learning_rate": 0.0001552746896923907,
      "loss": 3.2338,
      "step": 68800
    },
    {
      "epoch": 7.423180592991914,
      "grad_norm": 0.775603711605072,
      "learning_rate": 0.00015495089044792226,
      "loss": 3.2351,
      "step": 68850
    },
    {
      "epoch": 7.428571428571429,
      "grad_norm": 0.776694118976593,
      "learning_rate": 0.00015462709120345384,
      "loss": 3.23,
      "step": 68900
    },
    {
      "epoch": 7.433962264150943,
      "grad_norm": 0.786277711391449,
      "learning_rate": 0.00015430329195898542,
      "loss": 3.2334,
      "step": 68950
    },
    {
      "epoch": 7.439353099730458,
      "grad_norm": 0.7692110538482666,
      "learning_rate": 0.000153979492714517,
      "loss": 3.2182,
      "step": 69000
    },
    {
      "epoch": 7.439353099730458,
      "eval_accuracy": 0.388615632912307,
      "eval_loss": 3.346757173538208,
      "eval_runtime": 184.8049,
      "eval_samples_per_second": 97.46,
      "eval_steps_per_second": 6.093,
      "step": 69000
    },
    {
      "epoch": 7.444743935309973,
      "grad_norm": 0.8124752640724182,
      "learning_rate": 0.00015365569347004858,
      "loss": 3.2203,
      "step": 69050
    },
    {
      "epoch": 7.450134770889488,
      "grad_norm": 0.7867438793182373,
      "learning_rate": 0.00015333189422558015,
      "loss": 3.2298,
      "step": 69100
    },
    {
      "epoch": 7.455525606469003,
      "grad_norm": 0.7936617136001587,
      "learning_rate": 0.00015300809498111168,
      "loss": 3.2292,
      "step": 69150
    },
    {
      "epoch": 7.460916442048518,
      "grad_norm": 0.776745617389679,
      "learning_rate": 0.00015268429573664325,
      "loss": 3.2207,
      "step": 69200
    },
    {
      "epoch": 7.466307277628032,
      "grad_norm": 0.798880934715271,
      "learning_rate": 0.00015236049649217483,
      "loss": 3.2378,
      "step": 69250
    },
    {
      "epoch": 7.471698113207547,
      "grad_norm": 0.7821260690689087,
      "learning_rate": 0.0001520366972477064,
      "loss": 3.2253,
      "step": 69300
    },
    {
      "epoch": 7.4770889487870615,
      "grad_norm": 0.767823338508606,
      "learning_rate": 0.000151712898003238,
      "loss": 3.2376,
      "step": 69350
    },
    {
      "epoch": 7.482479784366577,
      "grad_norm": 0.7894257307052612,
      "learning_rate": 0.00015138909875876956,
      "loss": 3.2231,
      "step": 69400
    },
    {
      "epoch": 7.487870619946092,
      "grad_norm": 0.8050634860992432,
      "learning_rate": 0.00015106529951430114,
      "loss": 3.2228,
      "step": 69450
    },
    {
      "epoch": 7.493261455525606,
      "grad_norm": 0.7465404868125916,
      "learning_rate": 0.0001507415002698327,
      "loss": 3.2237,
      "step": 69500
    },
    {
      "epoch": 7.498652291105121,
      "grad_norm": 0.8405658006668091,
      "learning_rate": 0.00015041770102536424,
      "loss": 3.2376,
      "step": 69550
    },
    {
      "epoch": 7.504043126684636,
      "grad_norm": 0.7720454335212708,
      "learning_rate": 0.00015009390178089582,
      "loss": 3.2108,
      "step": 69600
    },
    {
      "epoch": 7.509433962264151,
      "grad_norm": 0.7862969636917114,
      "learning_rate": 0.0001497701025364274,
      "loss": 3.2289,
      "step": 69650
    },
    {
      "epoch": 7.514824797843666,
      "grad_norm": 0.761938214302063,
      "learning_rate": 0.00014944630329195898,
      "loss": 3.238,
      "step": 69700
    },
    {
      "epoch": 7.520215633423181,
      "grad_norm": 0.8123347163200378,
      "learning_rate": 0.00014912250404749055,
      "loss": 3.244,
      "step": 69750
    },
    {
      "epoch": 7.525606469002695,
      "grad_norm": 0.8718035817146301,
      "learning_rate": 0.0001487987048030221,
      "loss": 3.2249,
      "step": 69800
    },
    {
      "epoch": 7.53099730458221,
      "grad_norm": 0.755567193031311,
      "learning_rate": 0.00014847490555855368,
      "loss": 3.227,
      "step": 69850
    },
    {
      "epoch": 7.536388140161725,
      "grad_norm": 0.7703545093536377,
      "learning_rate": 0.00014815110631408526,
      "loss": 3.2319,
      "step": 69900
    },
    {
      "epoch": 7.54177897574124,
      "grad_norm": 0.7938928604125977,
      "learning_rate": 0.0001478273070696168,
      "loss": 3.2262,
      "step": 69950
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 0.7993608713150024,
      "learning_rate": 0.0001475035078251484,
      "loss": 3.2238,
      "step": 70000
    },
    {
      "epoch": 7.547169811320755,
      "eval_accuracy": 0.3888174013249785,
      "eval_loss": 3.3437108993530273,
      "eval_runtime": 184.8174,
      "eval_samples_per_second": 97.453,
      "eval_steps_per_second": 6.093,
      "step": 70000
    },
    {
      "epoch": 7.55256064690027,
      "grad_norm": 0.78631591796875,
      "learning_rate": 0.00014717970858067997,
      "loss": 3.2313,
      "step": 70050
    },
    {
      "epoch": 7.557951482479784,
      "grad_norm": 0.7872461080551147,
      "learning_rate": 0.00014685590933621154,
      "loss": 3.2356,
      "step": 70100
    },
    {
      "epoch": 7.563342318059299,
      "grad_norm": 0.7788323760032654,
      "learning_rate": 0.0001465321100917431,
      "loss": 3.2341,
      "step": 70150
    },
    {
      "epoch": 7.568733153638814,
      "grad_norm": 0.8200633525848389,
      "learning_rate": 0.00014620831084727467,
      "loss": 3.2179,
      "step": 70200
    },
    {
      "epoch": 7.574123989218329,
      "grad_norm": 0.806707501411438,
      "learning_rate": 0.00014588451160280625,
      "loss": 3.2329,
      "step": 70250
    },
    {
      "epoch": 7.579514824797844,
      "grad_norm": 0.7861495614051819,
      "learning_rate": 0.0001455607123583378,
      "loss": 3.2287,
      "step": 70300
    },
    {
      "epoch": 7.584905660377358,
      "grad_norm": 0.8309231996536255,
      "learning_rate": 0.00014523691311386938,
      "loss": 3.2446,
      "step": 70350
    },
    {
      "epoch": 7.590296495956873,
      "grad_norm": 0.7764883637428284,
      "learning_rate": 0.00014491311386940096,
      "loss": 3.2258,
      "step": 70400
    },
    {
      "epoch": 7.595687331536388,
      "grad_norm": 0.7940296530723572,
      "learning_rate": 0.00014458931462493253,
      "loss": 3.2383,
      "step": 70450
    },
    {
      "epoch": 7.601078167115903,
      "grad_norm": 0.8480539321899414,
      "learning_rate": 0.00014426551538046408,
      "loss": 3.2404,
      "step": 70500
    },
    {
      "epoch": 7.606469002695418,
      "grad_norm": 0.7697724103927612,
      "learning_rate": 0.00014394171613599566,
      "loss": 3.2288,
      "step": 70550
    },
    {
      "epoch": 7.611859838274933,
      "grad_norm": 0.7845086455345154,
      "learning_rate": 0.00014361791689152724,
      "loss": 3.2292,
      "step": 70600
    },
    {
      "epoch": 7.617250673854447,
      "grad_norm": 0.7915938496589661,
      "learning_rate": 0.00014329411764705882,
      "loss": 3.2291,
      "step": 70650
    },
    {
      "epoch": 7.622641509433962,
      "grad_norm": 0.7984021306037903,
      "learning_rate": 0.0001429703184025904,
      "loss": 3.2314,
      "step": 70700
    },
    {
      "epoch": 7.628032345013477,
      "grad_norm": 0.8050603866577148,
      "learning_rate": 0.00014264651915812194,
      "loss": 3.2322,
      "step": 70750
    },
    {
      "epoch": 7.633423180592992,
      "grad_norm": 0.8283957839012146,
      "learning_rate": 0.00014232271991365352,
      "loss": 3.2444,
      "step": 70800
    },
    {
      "epoch": 7.638814016172507,
      "grad_norm": 0.7875947952270508,
      "learning_rate": 0.0001419989206691851,
      "loss": 3.2222,
      "step": 70850
    },
    {
      "epoch": 7.644204851752022,
      "grad_norm": 0.8083146810531616,
      "learning_rate": 0.00014168159740960602,
      "loss": 3.2393,
      "step": 70900
    },
    {
      "epoch": 7.649595687331536,
      "grad_norm": 0.7859264612197876,
      "learning_rate": 0.0001413577981651376,
      "loss": 3.2201,
      "step": 70950
    },
    {
      "epoch": 7.654986522911051,
      "grad_norm": 0.7985530495643616,
      "learning_rate": 0.00014103399892066918,
      "loss": 3.2513,
      "step": 71000
    },
    {
      "epoch": 7.654986522911051,
      "eval_accuracy": 0.38947138305673384,
      "eval_loss": 3.3401072025299072,
      "eval_runtime": 184.8335,
      "eval_samples_per_second": 97.444,
      "eval_steps_per_second": 6.092,
      "step": 71000
    },
    {
      "epoch": 7.660377358490566,
      "grad_norm": 0.7884781956672668,
      "learning_rate": 0.00014071019967620073,
      "loss": 3.2415,
      "step": 71050
    },
    {
      "epoch": 7.665768194070081,
      "grad_norm": 0.7903490662574768,
      "learning_rate": 0.0001403864004317323,
      "loss": 3.2499,
      "step": 71100
    },
    {
      "epoch": 7.671159029649596,
      "grad_norm": 0.7995983958244324,
      "learning_rate": 0.00014006260118726389,
      "loss": 3.233,
      "step": 71150
    },
    {
      "epoch": 7.67654986522911,
      "grad_norm": 0.7693688273429871,
      "learning_rate": 0.00013973880194279546,
      "loss": 3.2488,
      "step": 71200
    },
    {
      "epoch": 7.681940700808625,
      "grad_norm": 0.7897132039070129,
      "learning_rate": 0.00013941500269832704,
      "loss": 3.2375,
      "step": 71250
    },
    {
      "epoch": 7.6873315363881405,
      "grad_norm": 0.8603083491325378,
      "learning_rate": 0.0001390912034538586,
      "loss": 3.2326,
      "step": 71300
    },
    {
      "epoch": 7.692722371967655,
      "grad_norm": 0.7997660636901855,
      "learning_rate": 0.00013876740420939017,
      "loss": 3.2474,
      "step": 71350
    },
    {
      "epoch": 7.69811320754717,
      "grad_norm": 0.8124864101409912,
      "learning_rate": 0.00013844360496492175,
      "loss": 3.2522,
      "step": 71400
    },
    {
      "epoch": 7.703504043126685,
      "grad_norm": 0.7789794206619263,
      "learning_rate": 0.00013811980572045333,
      "loss": 3.2292,
      "step": 71450
    },
    {
      "epoch": 7.708894878706199,
      "grad_norm": 0.8284264206886292,
      "learning_rate": 0.00013779600647598488,
      "loss": 3.2384,
      "step": 71500
    },
    {
      "epoch": 7.714285714285714,
      "grad_norm": 0.7541694045066833,
      "learning_rate": 0.00013747220723151645,
      "loss": 3.2273,
      "step": 71550
    },
    {
      "epoch": 7.719676549865229,
      "grad_norm": 0.8304938673973083,
      "learning_rate": 0.00013714840798704803,
      "loss": 3.2218,
      "step": 71600
    },
    {
      "epoch": 7.725067385444744,
      "grad_norm": 0.8099672198295593,
      "learning_rate": 0.00013682460874257958,
      "loss": 3.244,
      "step": 71650
    },
    {
      "epoch": 7.730458221024259,
      "grad_norm": 0.8018233180046082,
      "learning_rate": 0.00013650080949811116,
      "loss": 3.2449,
      "step": 71700
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 0.8082437515258789,
      "learning_rate": 0.00013617701025364274,
      "loss": 3.2408,
      "step": 71750
    },
    {
      "epoch": 7.741239892183288,
      "grad_norm": 0.7693942785263062,
      "learning_rate": 0.0001358532110091743,
      "loss": 3.2298,
      "step": 71800
    },
    {
      "epoch": 7.7466307277628035,
      "grad_norm": 0.7721502780914307,
      "learning_rate": 0.00013552941176470587,
      "loss": 3.2468,
      "step": 71850
    },
    {
      "epoch": 7.752021563342318,
      "grad_norm": 0.76530522108078,
      "learning_rate": 0.00013520561252023744,
      "loss": 3.2439,
      "step": 71900
    },
    {
      "epoch": 7.757412398921833,
      "grad_norm": 0.786440908908844,
      "learning_rate": 0.00013488181327576902,
      "loss": 3.244,
      "step": 71950
    },
    {
      "epoch": 7.762803234501348,
      "grad_norm": 0.7962386012077332,
      "learning_rate": 0.00013455801403130057,
      "loss": 3.2518,
      "step": 72000
    },
    {
      "epoch": 7.762803234501348,
      "eval_accuracy": 0.39001149656205963,
      "eval_loss": 3.3362839221954346,
      "eval_runtime": 184.9207,
      "eval_samples_per_second": 97.399,
      "eval_steps_per_second": 6.089,
      "step": 72000
    },
    {
      "epoch": 7.768194070080862,
      "grad_norm": 0.8243362903594971,
      "learning_rate": 0.00013423421478683215,
      "loss": 3.2376,
      "step": 72050
    },
    {
      "epoch": 7.773584905660377,
      "grad_norm": 0.7997215390205383,
      "learning_rate": 0.00013391041554236373,
      "loss": 3.2442,
      "step": 72100
    },
    {
      "epoch": 7.7789757412398925,
      "grad_norm": 0.8042483329772949,
      "learning_rate": 0.00013358661629789528,
      "loss": 3.2313,
      "step": 72150
    },
    {
      "epoch": 7.784366576819407,
      "grad_norm": 0.7405474781990051,
      "learning_rate": 0.00013326281705342685,
      "loss": 3.2229,
      "step": 72200
    },
    {
      "epoch": 7.789757412398922,
      "grad_norm": 0.784062385559082,
      "learning_rate": 0.00013293901780895843,
      "loss": 3.2333,
      "step": 72250
    },
    {
      "epoch": 7.795148247978437,
      "grad_norm": 0.8437650203704834,
      "learning_rate": 0.00013261521856449,
      "loss": 3.2217,
      "step": 72300
    },
    {
      "epoch": 7.800539083557951,
      "grad_norm": 0.7900862097740173,
      "learning_rate": 0.00013229141932002156,
      "loss": 3.2094,
      "step": 72350
    },
    {
      "epoch": 7.8059299191374665,
      "grad_norm": 0.80324786901474,
      "learning_rate": 0.00013196762007555314,
      "loss": 3.2356,
      "step": 72400
    },
    {
      "epoch": 7.811320754716981,
      "grad_norm": 0.8037505149841309,
      "learning_rate": 0.00013164382083108472,
      "loss": 3.2382,
      "step": 72450
    },
    {
      "epoch": 7.816711590296496,
      "grad_norm": 0.8118758797645569,
      "learning_rate": 0.0001313200215866163,
      "loss": 3.2327,
      "step": 72500
    },
    {
      "epoch": 7.822102425876011,
      "grad_norm": 0.8497474789619446,
      "learning_rate": 0.00013099622234214784,
      "loss": 3.2207,
      "step": 72550
    },
    {
      "epoch": 7.827493261455525,
      "grad_norm": 0.7857134938240051,
      "learning_rate": 0.00013067242309767942,
      "loss": 3.2506,
      "step": 72600
    },
    {
      "epoch": 7.83288409703504,
      "grad_norm": 0.8757586479187012,
      "learning_rate": 0.000130348623853211,
      "loss": 3.2352,
      "step": 72650
    },
    {
      "epoch": 7.8382749326145555,
      "grad_norm": 0.7962746024131775,
      "learning_rate": 0.00013002482460874258,
      "loss": 3.2401,
      "step": 72700
    },
    {
      "epoch": 7.84366576819407,
      "grad_norm": 0.8165475726127625,
      "learning_rate": 0.00012970102536427413,
      "loss": 3.2086,
      "step": 72750
    },
    {
      "epoch": 7.849056603773585,
      "grad_norm": 0.7961111068725586,
      "learning_rate": 0.0001293772261198057,
      "loss": 3.2404,
      "step": 72800
    },
    {
      "epoch": 7.8544474393531,
      "grad_norm": 0.8070231676101685,
      "learning_rate": 0.00012905342687533728,
      "loss": 3.2371,
      "step": 72850
    },
    {
      "epoch": 7.859838274932614,
      "grad_norm": 0.7886551022529602,
      "learning_rate": 0.00012872962763086886,
      "loss": 3.2402,
      "step": 72900
    },
    {
      "epoch": 7.8652291105121295,
      "grad_norm": 0.8020292520523071,
      "learning_rate": 0.00012840582838640044,
      "loss": 3.2398,
      "step": 72950
    },
    {
      "epoch": 7.870619946091644,
      "grad_norm": 0.8239029049873352,
      "learning_rate": 0.000128082029141932,
      "loss": 3.2309,
      "step": 73000
    },
    {
      "epoch": 7.870619946091644,
      "eval_accuracy": 0.39013916370524165,
      "eval_loss": 3.332185983657837,
      "eval_runtime": 184.799,
      "eval_samples_per_second": 97.463,
      "eval_steps_per_second": 6.093,
      "step": 73000
    },
    {
      "epoch": 7.876010781671159,
      "grad_norm": 0.850242018699646,
      "learning_rate": 0.00012775822989746357,
      "loss": 3.2415,
      "step": 73050
    },
    {
      "epoch": 7.881401617250674,
      "grad_norm": 0.7743697166442871,
      "learning_rate": 0.0001274409066378845,
      "loss": 3.2358,
      "step": 73100
    },
    {
      "epoch": 7.886792452830189,
      "grad_norm": 0.7658908367156982,
      "learning_rate": 0.00012711710739341607,
      "loss": 3.2371,
      "step": 73150
    },
    {
      "epoch": 7.892183288409703,
      "grad_norm": 0.8552940487861633,
      "learning_rate": 0.00012679330814894765,
      "loss": 3.2308,
      "step": 73200
    },
    {
      "epoch": 7.8975741239892185,
      "grad_norm": 0.791025698184967,
      "learning_rate": 0.00012646950890447922,
      "loss": 3.2368,
      "step": 73250
    },
    {
      "epoch": 7.902964959568733,
      "grad_norm": 0.8062611222267151,
      "learning_rate": 0.00012614570966001077,
      "loss": 3.2394,
      "step": 73300
    },
    {
      "epoch": 7.908355795148248,
      "grad_norm": 0.8666378259658813,
      "learning_rate": 0.00012582191041554235,
      "loss": 3.2175,
      "step": 73350
    },
    {
      "epoch": 7.913746630727763,
      "grad_norm": 0.8134450316429138,
      "learning_rate": 0.00012549811117107393,
      "loss": 3.2308,
      "step": 73400
    },
    {
      "epoch": 7.919137466307277,
      "grad_norm": 0.822311282157898,
      "learning_rate": 0.0001251743119266055,
      "loss": 3.2146,
      "step": 73450
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 0.8420822620391846,
      "learning_rate": 0.00012485051268213706,
      "loss": 3.2317,
      "step": 73500
    },
    {
      "epoch": 7.929919137466308,
      "grad_norm": 0.7920177578926086,
      "learning_rate": 0.00012452671343766864,
      "loss": 3.2235,
      "step": 73550
    },
    {
      "epoch": 7.935309973045822,
      "grad_norm": 0.8194290995597839,
      "learning_rate": 0.00012420291419320021,
      "loss": 3.2235,
      "step": 73600
    },
    {
      "epoch": 7.940700808625337,
      "grad_norm": 0.8023219704627991,
      "learning_rate": 0.0001238791149487318,
      "loss": 3.2397,
      "step": 73650
    },
    {
      "epoch": 7.946091644204852,
      "grad_norm": 1.2114663124084473,
      "learning_rate": 0.00012355531570426334,
      "loss": 3.2168,
      "step": 73700
    },
    {
      "epoch": 7.951482479784366,
      "grad_norm": 0.8042489290237427,
      "learning_rate": 0.00012323151645979492,
      "loss": 3.2347,
      "step": 73750
    },
    {
      "epoch": 7.9568733153638815,
      "grad_norm": 0.8316463232040405,
      "learning_rate": 0.0001229077172153265,
      "loss": 3.2324,
      "step": 73800
    },
    {
      "epoch": 7.962264150943396,
      "grad_norm": 0.8015411496162415,
      "learning_rate": 0.00012258391797085805,
      "loss": 3.221,
      "step": 73850
    },
    {
      "epoch": 7.967654986522911,
      "grad_norm": 0.8200330138206482,
      "learning_rate": 0.00012226011872638963,
      "loss": 3.2256,
      "step": 73900
    },
    {
      "epoch": 7.973045822102426,
      "grad_norm": 0.7811427712440491,
      "learning_rate": 0.00012193631948192119,
      "loss": 3.228,
      "step": 73950
    },
    {
      "epoch": 7.97843665768194,
      "grad_norm": 0.8133488297462463,
      "learning_rate": 0.00012161252023745277,
      "loss": 3.2519,
      "step": 74000
    },
    {
      "epoch": 7.97843665768194,
      "eval_accuracy": 0.39057507909115335,
      "eval_loss": 3.328582525253296,
      "eval_runtime": 185.1358,
      "eval_samples_per_second": 97.285,
      "eval_steps_per_second": 6.082,
      "step": 74000
    },
    {
      "epoch": 7.9838274932614555,
      "grad_norm": 0.8701046109199524,
      "learning_rate": 0.00012128872099298435,
      "loss": 3.2368,
      "step": 74050
    },
    {
      "epoch": 7.989218328840971,
      "grad_norm": 0.8377940058708191,
      "learning_rate": 0.00012096492174851591,
      "loss": 3.2394,
      "step": 74100
    },
    {
      "epoch": 7.994609164420485,
      "grad_norm": 0.8556270599365234,
      "learning_rate": 0.00012064112250404749,
      "loss": 3.2404,
      "step": 74150
    },
    {
      "epoch": 8.0,
      "grad_norm": 1.6344754695892334,
      "learning_rate": 0.00012031732325957905,
      "loss": 3.2383,
      "step": 74200
    },
    {
      "epoch": 8.005390835579515,
      "grad_norm": 0.801658034324646,
      "learning_rate": 0.00011999352401511062,
      "loss": 3.1543,
      "step": 74250
    },
    {
      "epoch": 8.01078167115903,
      "grad_norm": 0.7476630210876465,
      "learning_rate": 0.00011966972477064219,
      "loss": 3.1637,
      "step": 74300
    },
    {
      "epoch": 8.016172506738544,
      "grad_norm": 0.8232449293136597,
      "learning_rate": 0.00011934592552617377,
      "loss": 3.1622,
      "step": 74350
    },
    {
      "epoch": 8.021563342318059,
      "grad_norm": 0.8696622252464294,
      "learning_rate": 0.00011902212628170532,
      "loss": 3.1583,
      "step": 74400
    },
    {
      "epoch": 8.026954177897574,
      "grad_norm": 0.8294352293014526,
      "learning_rate": 0.0001186983270372369,
      "loss": 3.1607,
      "step": 74450
    },
    {
      "epoch": 8.032345013477089,
      "grad_norm": 0.7915950417518616,
      "learning_rate": 0.00011837452779276848,
      "loss": 3.1799,
      "step": 74500
    },
    {
      "epoch": 8.037735849056604,
      "grad_norm": 0.8357707858085632,
      "learning_rate": 0.00011805072854830005,
      "loss": 3.158,
      "step": 74550
    },
    {
      "epoch": 8.04312668463612,
      "grad_norm": 0.8107123970985413,
      "learning_rate": 0.0001177269293038316,
      "loss": 3.1785,
      "step": 74600
    },
    {
      "epoch": 8.048517520215633,
      "grad_norm": 0.8061670064926147,
      "learning_rate": 0.00011740313005936318,
      "loss": 3.175,
      "step": 74650
    },
    {
      "epoch": 8.053908355795148,
      "grad_norm": 0.7983211874961853,
      "learning_rate": 0.00011707933081489476,
      "loss": 3.1642,
      "step": 74700
    },
    {
      "epoch": 8.059299191374663,
      "grad_norm": 0.8365211486816406,
      "learning_rate": 0.00011675553157042632,
      "loss": 3.1511,
      "step": 74750
    },
    {
      "epoch": 8.064690026954178,
      "grad_norm": 0.8603684306144714,
      "learning_rate": 0.00011643173232595789,
      "loss": 3.1714,
      "step": 74800
    },
    {
      "epoch": 8.070080862533693,
      "grad_norm": 0.7773607969284058,
      "learning_rate": 0.00011610793308148947,
      "loss": 3.1512,
      "step": 74850
    },
    {
      "epoch": 8.075471698113208,
      "grad_norm": 0.7770172953605652,
      "learning_rate": 0.00011578413383702104,
      "loss": 3.152,
      "step": 74900
    },
    {
      "epoch": 8.080862533692722,
      "grad_norm": 0.78849858045578,
      "learning_rate": 0.00011546033459255261,
      "loss": 3.1676,
      "step": 74950
    },
    {
      "epoch": 8.086253369272237,
      "grad_norm": 0.814720630645752,
      "learning_rate": 0.00011513653534808419,
      "loss": 3.1583,
      "step": 75000
    },
    {
      "epoch": 8.086253369272237,
      "eval_accuracy": 0.3906881867473682,
      "eval_loss": 3.3347113132476807,
      "eval_runtime": 184.6754,
      "eval_samples_per_second": 97.528,
      "eval_steps_per_second": 6.097,
      "step": 75000
    },
    {
      "epoch": 8.091644204851752,
      "grad_norm": 0.8267271518707275,
      "learning_rate": 0.00011481273610361575,
      "loss": 3.165,
      "step": 75050
    },
    {
      "epoch": 8.097035040431267,
      "grad_norm": 0.8477714657783508,
      "learning_rate": 0.00011448893685914731,
      "loss": 3.1511,
      "step": 75100
    },
    {
      "epoch": 8.102425876010782,
      "grad_norm": 0.7959654331207275,
      "learning_rate": 0.00011416513761467889,
      "loss": 3.1745,
      "step": 75150
    },
    {
      "epoch": 8.107816711590296,
      "grad_norm": 0.8167853951454163,
      "learning_rate": 0.00011384133837021047,
      "loss": 3.1673,
      "step": 75200
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 0.7871206402778625,
      "learning_rate": 0.00011351753912574202,
      "loss": 3.1734,
      "step": 75250
    },
    {
      "epoch": 8.118598382749326,
      "grad_norm": 0.7872896194458008,
      "learning_rate": 0.0001131937398812736,
      "loss": 3.1801,
      "step": 75300
    },
    {
      "epoch": 8.123989218328841,
      "grad_norm": 0.7880840301513672,
      "learning_rate": 0.00011286994063680517,
      "loss": 3.1839,
      "step": 75350
    },
    {
      "epoch": 8.129380053908356,
      "grad_norm": 0.883547306060791,
      "learning_rate": 0.00011254614139233675,
      "loss": 3.1856,
      "step": 75400
    },
    {
      "epoch": 8.134770889487871,
      "grad_norm": 0.8173453211784363,
      "learning_rate": 0.0001122223421478683,
      "loss": 3.1721,
      "step": 75450
    },
    {
      "epoch": 8.140161725067385,
      "grad_norm": 0.7894411087036133,
      "learning_rate": 0.00011189854290339988,
      "loss": 3.1769,
      "step": 75500
    },
    {
      "epoch": 8.1455525606469,
      "grad_norm": 0.7918045520782471,
      "learning_rate": 0.00011157474365893146,
      "loss": 3.1614,
      "step": 75550
    },
    {
      "epoch": 8.150943396226415,
      "grad_norm": 0.825491726398468,
      "learning_rate": 0.00011125094441446302,
      "loss": 3.1587,
      "step": 75600
    },
    {
      "epoch": 8.15633423180593,
      "grad_norm": 0.8395944833755493,
      "learning_rate": 0.00011092714516999459,
      "loss": 3.1649,
      "step": 75650
    },
    {
      "epoch": 8.161725067385445,
      "grad_norm": 0.8326493501663208,
      "learning_rate": 0.00011060334592552616,
      "loss": 3.1851,
      "step": 75700
    },
    {
      "epoch": 8.167115902964959,
      "grad_norm": 0.795487642288208,
      "learning_rate": 0.00011027954668105773,
      "loss": 3.1673,
      "step": 75750
    },
    {
      "epoch": 8.172506738544474,
      "grad_norm": 0.8097192049026489,
      "learning_rate": 0.0001099557474365893,
      "loss": 3.1614,
      "step": 75800
    },
    {
      "epoch": 8.177897574123989,
      "grad_norm": 0.8157842755317688,
      "learning_rate": 0.00010963194819212088,
      "loss": 3.1657,
      "step": 75850
    },
    {
      "epoch": 8.183288409703504,
      "grad_norm": 0.7984054088592529,
      "learning_rate": 0.00010930814894765243,
      "loss": 3.1681,
      "step": 75900
    },
    {
      "epoch": 8.18867924528302,
      "grad_norm": 0.8114879727363586,
      "learning_rate": 0.00010898434970318401,
      "loss": 3.1797,
      "step": 75950
    },
    {
      "epoch": 8.194070080862534,
      "grad_norm": 0.7779016494750977,
      "learning_rate": 0.00010866055045871559,
      "loss": 3.1571,
      "step": 76000
    },
    {
      "epoch": 8.194070080862534,
      "eval_accuracy": 0.3906270151715287,
      "eval_loss": 3.3333170413970947,
      "eval_runtime": 185.1654,
      "eval_samples_per_second": 97.27,
      "eval_steps_per_second": 6.081,
      "step": 76000
    },
    {
      "epoch": 8.199460916442048,
      "grad_norm": 0.8344588279724121,
      "learning_rate": 0.00010833675121424717,
      "loss": 3.1813,
      "step": 76050
    },
    {
      "epoch": 8.204851752021563,
      "grad_norm": 0.8400987386703491,
      "learning_rate": 0.00010801295196977872,
      "loss": 3.1983,
      "step": 76100
    },
    {
      "epoch": 8.210242587601078,
      "grad_norm": 0.7751808762550354,
      "learning_rate": 0.0001076891527253103,
      "loss": 3.1688,
      "step": 76150
    },
    {
      "epoch": 8.215633423180593,
      "grad_norm": 0.8149611353874207,
      "learning_rate": 0.00010736535348084187,
      "loss": 3.1826,
      "step": 76200
    },
    {
      "epoch": 8.221024258760108,
      "grad_norm": 0.8001996874809265,
      "learning_rate": 0.00010704155423637345,
      "loss": 3.1785,
      "step": 76250
    },
    {
      "epoch": 8.226415094339623,
      "grad_norm": 0.8465657830238342,
      "learning_rate": 0.000106717754991905,
      "loss": 3.1606,
      "step": 76300
    },
    {
      "epoch": 8.231805929919137,
      "grad_norm": 0.8427553772926331,
      "learning_rate": 0.00010639395574743658,
      "loss": 3.1798,
      "step": 76350
    },
    {
      "epoch": 8.237196765498652,
      "grad_norm": 0.7900227308273315,
      "learning_rate": 0.00010607015650296816,
      "loss": 3.1745,
      "step": 76400
    },
    {
      "epoch": 8.242587601078167,
      "grad_norm": 0.7699751257896423,
      "learning_rate": 0.00010574635725849972,
      "loss": 3.175,
      "step": 76450
    },
    {
      "epoch": 8.247978436657682,
      "grad_norm": 0.8397601246833801,
      "learning_rate": 0.00010542255801403128,
      "loss": 3.164,
      "step": 76500
    },
    {
      "epoch": 8.253369272237197,
      "grad_norm": 0.874344527721405,
      "learning_rate": 0.00010509875876956286,
      "loss": 3.1902,
      "step": 76550
    },
    {
      "epoch": 8.25876010781671,
      "grad_norm": 0.7956122159957886,
      "learning_rate": 0.00010477495952509443,
      "loss": 3.1647,
      "step": 76600
    },
    {
      "epoch": 8.264150943396226,
      "grad_norm": 0.7708241939544678,
      "learning_rate": 0.000104451160280626,
      "loss": 3.1618,
      "step": 76650
    },
    {
      "epoch": 8.269541778975741,
      "grad_norm": 0.8957728743553162,
      "learning_rate": 0.00010412736103615758,
      "loss": 3.1832,
      "step": 76700
    },
    {
      "epoch": 8.274932614555256,
      "grad_norm": 0.8150855898857117,
      "learning_rate": 0.00010380356179168913,
      "loss": 3.1755,
      "step": 76750
    },
    {
      "epoch": 8.280323450134771,
      "grad_norm": 0.8756967186927795,
      "learning_rate": 0.00010347976254722071,
      "loss": 3.1797,
      "step": 76800
    },
    {
      "epoch": 8.285714285714286,
      "grad_norm": 0.7948899865150452,
      "learning_rate": 0.00010315596330275229,
      "loss": 3.1907,
      "step": 76850
    },
    {
      "epoch": 8.2911051212938,
      "grad_norm": 0.8002045154571533,
      "learning_rate": 0.00010283216405828387,
      "loss": 3.2053,
      "step": 76900
    },
    {
      "epoch": 8.296495956873315,
      "grad_norm": 0.8174263834953308,
      "learning_rate": 0.00010250836481381542,
      "loss": 3.1878,
      "step": 76950
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 0.7678391933441162,
      "learning_rate": 0.000102184565569347,
      "loss": 3.1771,
      "step": 77000
    },
    {
      "epoch": 8.30188679245283,
      "eval_accuracy": 0.3910068656673281,
      "eval_loss": 3.3296449184417725,
      "eval_runtime": 184.8006,
      "eval_samples_per_second": 97.462,
      "eval_steps_per_second": 6.093,
      "step": 77000
    },
    {
      "epoch": 8.307277628032345,
      "grad_norm": 0.8229707479476929,
      "learning_rate": 0.00010186724230976793,
      "loss": 3.1903,
      "step": 77050
    },
    {
      "epoch": 8.31266846361186,
      "grad_norm": 0.8444199562072754,
      "learning_rate": 0.0001015434430652995,
      "loss": 3.1739,
      "step": 77100
    },
    {
      "epoch": 8.318059299191376,
      "grad_norm": 0.8374699354171753,
      "learning_rate": 0.00010121964382083107,
      "loss": 3.1627,
      "step": 77150
    },
    {
      "epoch": 8.323450134770889,
      "grad_norm": 0.8763797283172607,
      "learning_rate": 0.00010089584457636265,
      "loss": 3.1634,
      "step": 77200
    },
    {
      "epoch": 8.328840970350404,
      "grad_norm": 0.844142496585846,
      "learning_rate": 0.00010057204533189423,
      "loss": 3.1628,
      "step": 77250
    },
    {
      "epoch": 8.33423180592992,
      "grad_norm": 0.8463607430458069,
      "learning_rate": 0.00010024824608742578,
      "loss": 3.1713,
      "step": 77300
    },
    {
      "epoch": 8.339622641509434,
      "grad_norm": 0.816167414188385,
      "learning_rate": 9.992444684295736e-05,
      "loss": 3.1854,
      "step": 77350
    },
    {
      "epoch": 8.34501347708895,
      "grad_norm": 0.868462860584259,
      "learning_rate": 9.960064759848894e-05,
      "loss": 3.1387,
      "step": 77400
    },
    {
      "epoch": 8.350404312668463,
      "grad_norm": 0.8351553082466125,
      "learning_rate": 9.92768483540205e-05,
      "loss": 3.1847,
      "step": 77450
    },
    {
      "epoch": 8.355795148247978,
      "grad_norm": 0.8340010046958923,
      "learning_rate": 9.895304910955206e-05,
      "loss": 3.1897,
      "step": 77500
    },
    {
      "epoch": 8.361185983827493,
      "grad_norm": 0.8664101362228394,
      "learning_rate": 9.862924986508364e-05,
      "loss": 3.1756,
      "step": 77550
    },
    {
      "epoch": 8.366576819407008,
      "grad_norm": 0.9245196580886841,
      "learning_rate": 9.83054506206152e-05,
      "loss": 3.1933,
      "step": 77600
    },
    {
      "epoch": 8.371967654986523,
      "grad_norm": 0.8184807300567627,
      "learning_rate": 9.798165137614678e-05,
      "loss": 3.1595,
      "step": 77650
    },
    {
      "epoch": 8.377358490566039,
      "grad_norm": 0.8540685176849365,
      "learning_rate": 9.765785213167835e-05,
      "loss": 3.1975,
      "step": 77700
    },
    {
      "epoch": 8.382749326145552,
      "grad_norm": 0.8300887942314148,
      "learning_rate": 9.733405288720992e-05,
      "loss": 3.1769,
      "step": 77750
    },
    {
      "epoch": 8.388140161725067,
      "grad_norm": 0.8534173369407654,
      "learning_rate": 9.701025364274149e-05,
      "loss": 3.1714,
      "step": 77800
    },
    {
      "epoch": 8.393530997304582,
      "grad_norm": 0.8436583280563354,
      "learning_rate": 9.668645439827307e-05,
      "loss": 3.1783,
      "step": 77850
    },
    {
      "epoch": 8.398921832884097,
      "grad_norm": 0.8904436230659485,
      "learning_rate": 9.636265515380464e-05,
      "loss": 3.176,
      "step": 77900
    },
    {
      "epoch": 8.404312668463612,
      "grad_norm": 0.9166279435157776,
      "learning_rate": 9.60388559093362e-05,
      "loss": 3.1848,
      "step": 77950
    },
    {
      "epoch": 8.409703504043126,
      "grad_norm": 0.8319751024246216,
      "learning_rate": 9.571505666486777e-05,
      "loss": 3.1673,
      "step": 78000
    },
    {
      "epoch": 8.409703504043126,
      "eval_accuracy": 0.3911980947498816,
      "eval_loss": 3.3280577659606934,
      "eval_runtime": 184.9642,
      "eval_samples_per_second": 97.376,
      "eval_steps_per_second": 6.088,
      "step": 78000
    },
    {
      "epoch": 8.415094339622641,
      "grad_norm": 0.8120921850204468,
      "learning_rate": 9.539125742039935e-05,
      "loss": 3.1822,
      "step": 78050
    },
    {
      "epoch": 8.420485175202156,
      "grad_norm": 0.8287458419799805,
      "learning_rate": 9.506745817593093e-05,
      "loss": 3.1704,
      "step": 78100
    },
    {
      "epoch": 8.425876010781671,
      "grad_norm": 0.8519405722618103,
      "learning_rate": 9.474365893146248e-05,
      "loss": 3.1791,
      "step": 78150
    },
    {
      "epoch": 8.431266846361186,
      "grad_norm": 0.9082578420639038,
      "learning_rate": 9.441985968699406e-05,
      "loss": 3.1903,
      "step": 78200
    },
    {
      "epoch": 8.436657681940702,
      "grad_norm": 0.8822596669197083,
      "learning_rate": 9.409606044252563e-05,
      "loss": 3.2137,
      "step": 78250
    },
    {
      "epoch": 8.442048517520215,
      "grad_norm": 0.8615216016769409,
      "learning_rate": 9.37722611980572e-05,
      "loss": 3.1818,
      "step": 78300
    },
    {
      "epoch": 8.44743935309973,
      "grad_norm": 0.8587093353271484,
      "learning_rate": 9.344846195358876e-05,
      "loss": 3.1776,
      "step": 78350
    },
    {
      "epoch": 8.452830188679245,
      "grad_norm": 0.8063041567802429,
      "learning_rate": 9.312466270912034e-05,
      "loss": 3.1846,
      "step": 78400
    },
    {
      "epoch": 8.45822102425876,
      "grad_norm": 0.8895376324653625,
      "learning_rate": 9.28008634646519e-05,
      "loss": 3.1933,
      "step": 78450
    },
    {
      "epoch": 8.463611859838275,
      "grad_norm": 0.8046665787696838,
      "learning_rate": 9.247706422018348e-05,
      "loss": 3.1872,
      "step": 78500
    },
    {
      "epoch": 8.46900269541779,
      "grad_norm": 0.8373346924781799,
      "learning_rate": 9.215326497571505e-05,
      "loss": 3.1793,
      "step": 78550
    },
    {
      "epoch": 8.474393530997304,
      "grad_norm": 0.8134214878082275,
      "learning_rate": 9.182946573124661e-05,
      "loss": 3.1889,
      "step": 78600
    },
    {
      "epoch": 8.479784366576819,
      "grad_norm": 0.8032200932502747,
      "learning_rate": 9.150566648677819e-05,
      "loss": 3.1765,
      "step": 78650
    },
    {
      "epoch": 8.485175202156334,
      "grad_norm": 0.8005744218826294,
      "learning_rate": 9.118186724230976e-05,
      "loss": 3.1795,
      "step": 78700
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 0.8798218965530396,
      "learning_rate": 9.085806799784134e-05,
      "loss": 3.189,
      "step": 78750
    },
    {
      "epoch": 8.495956873315365,
      "grad_norm": 0.8683854937553406,
      "learning_rate": 9.053426875337289e-05,
      "loss": 3.1934,
      "step": 78800
    },
    {
      "epoch": 8.501347708894878,
      "grad_norm": 0.813085675239563,
      "learning_rate": 9.021046950890447e-05,
      "loss": 3.1887,
      "step": 78850
    },
    {
      "epoch": 8.506738544474393,
      "grad_norm": 0.8143117427825928,
      "learning_rate": 8.988667026443605e-05,
      "loss": 3.1808,
      "step": 78900
    },
    {
      "epoch": 8.512129380053908,
      "grad_norm": 0.8620724678039551,
      "learning_rate": 8.956287101996761e-05,
      "loss": 3.1792,
      "step": 78950
    },
    {
      "epoch": 8.517520215633423,
      "grad_norm": 0.8598374724388123,
      "learning_rate": 8.923907177549918e-05,
      "loss": 3.1806,
      "step": 79000
    },
    {
      "epoch": 8.517520215633423,
      "eval_accuracy": 0.3915852249891646,
      "eval_loss": 3.322683095932007,
      "eval_runtime": 184.9558,
      "eval_samples_per_second": 97.38,
      "eval_steps_per_second": 6.088,
      "step": 79000
    },
    {
      "epoch": 8.522911051212938,
      "grad_norm": 0.8073253035545349,
      "learning_rate": 8.891527253103075e-05,
      "loss": 3.1988,
      "step": 79050
    },
    {
      "epoch": 8.528301886792454,
      "grad_norm": 0.8984984755516052,
      "learning_rate": 8.859147328656233e-05,
      "loss": 3.1804,
      "step": 79100
    },
    {
      "epoch": 8.533692722371967,
      "grad_norm": 0.843517541885376,
      "learning_rate": 8.82676740420939e-05,
      "loss": 3.1896,
      "step": 79150
    },
    {
      "epoch": 8.539083557951482,
      "grad_norm": 0.8529115319252014,
      "learning_rate": 8.794387479762546e-05,
      "loss": 3.2095,
      "step": 79200
    },
    {
      "epoch": 8.544474393530997,
      "grad_norm": 0.8642679452896118,
      "learning_rate": 8.762007555315704e-05,
      "loss": 3.1868,
      "step": 79250
    },
    {
      "epoch": 8.549865229110512,
      "grad_norm": 0.8819208145141602,
      "learning_rate": 8.72962763086886e-05,
      "loss": 3.1621,
      "step": 79300
    },
    {
      "epoch": 8.555256064690028,
      "grad_norm": 0.8184266686439514,
      "learning_rate": 8.697247706422018e-05,
      "loss": 3.1841,
      "step": 79350
    },
    {
      "epoch": 8.560646900269543,
      "grad_norm": 0.9021525979042053,
      "learning_rate": 8.664867781975174e-05,
      "loss": 3.1821,
      "step": 79400
    },
    {
      "epoch": 8.566037735849056,
      "grad_norm": 0.823969841003418,
      "learning_rate": 8.632487857528331e-05,
      "loss": 3.1666,
      "step": 79450
    },
    {
      "epoch": 8.571428571428571,
      "grad_norm": 0.8354356288909912,
      "learning_rate": 8.600107933081489e-05,
      "loss": 3.1763,
      "step": 79500
    },
    {
      "epoch": 8.576819407008086,
      "grad_norm": 0.9359567165374756,
      "learning_rate": 8.567728008634646e-05,
      "loss": 3.176,
      "step": 79550
    },
    {
      "epoch": 8.582210242587601,
      "grad_norm": 0.8543683290481567,
      "learning_rate": 8.535348084187804e-05,
      "loss": 3.1616,
      "step": 79600
    },
    {
      "epoch": 8.587601078167117,
      "grad_norm": 0.8099145293235779,
      "learning_rate": 8.502968159740959e-05,
      "loss": 3.1743,
      "step": 79650
    },
    {
      "epoch": 8.59299191374663,
      "grad_norm": 0.8237737417221069,
      "learning_rate": 8.470588235294117e-05,
      "loss": 3.173,
      "step": 79700
    },
    {
      "epoch": 8.598382749326145,
      "grad_norm": 0.8273354768753052,
      "learning_rate": 8.438208310847275e-05,
      "loss": 3.1819,
      "step": 79750
    },
    {
      "epoch": 8.60377358490566,
      "grad_norm": 0.8201135396957397,
      "learning_rate": 8.405828386400431e-05,
      "loss": 3.1865,
      "step": 79800
    },
    {
      "epoch": 8.609164420485175,
      "grad_norm": 0.835292637348175,
      "learning_rate": 8.373448461953587e-05,
      "loss": 3.165,
      "step": 79850
    },
    {
      "epoch": 8.61455525606469,
      "grad_norm": 0.8473345041275024,
      "learning_rate": 8.341068537506745e-05,
      "loss": 3.2078,
      "step": 79900
    },
    {
      "epoch": 8.619946091644206,
      "grad_norm": 0.8289116024971008,
      "learning_rate": 8.308688613059902e-05,
      "loss": 3.1844,
      "step": 79950
    },
    {
      "epoch": 8.625336927223719,
      "grad_norm": 0.8131082653999329,
      "learning_rate": 8.27630868861306e-05,
      "loss": 3.1973,
      "step": 80000
    },
    {
      "epoch": 8.625336927223719,
      "eval_accuracy": 0.39217618804598753,
      "eval_loss": 3.3202123641967773,
      "eval_runtime": 184.6219,
      "eval_samples_per_second": 97.556,
      "eval_steps_per_second": 6.099,
      "step": 80000
    },
    {
      "epoch": 8.630727762803234,
      "grad_norm": 0.8290890455245972,
      "learning_rate": 8.243928764166216e-05,
      "loss": 3.1902,
      "step": 80050
    },
    {
      "epoch": 8.63611859838275,
      "grad_norm": 0.8842519521713257,
      "learning_rate": 8.211548839719374e-05,
      "loss": 3.1727,
      "step": 80100
    },
    {
      "epoch": 8.641509433962264,
      "grad_norm": 0.8551476001739502,
      "learning_rate": 8.179816513761467e-05,
      "loss": 3.1815,
      "step": 80150
    },
    {
      "epoch": 8.64690026954178,
      "grad_norm": 0.8610032796859741,
      "learning_rate": 8.147436589314624e-05,
      "loss": 3.1982,
      "step": 80200
    },
    {
      "epoch": 8.652291105121293,
      "grad_norm": 0.8614827394485474,
      "learning_rate": 8.115056664867782e-05,
      "loss": 3.1849,
      "step": 80250
    },
    {
      "epoch": 8.657681940700808,
      "grad_norm": 0.8478200435638428,
      "learning_rate": 8.082676740420938e-05,
      "loss": 3.1728,
      "step": 80300
    },
    {
      "epoch": 8.663072776280323,
      "grad_norm": 0.8627998232841492,
      "learning_rate": 8.050296815974096e-05,
      "loss": 3.1857,
      "step": 80350
    },
    {
      "epoch": 8.668463611859838,
      "grad_norm": 0.8678106069564819,
      "learning_rate": 8.017916891527252e-05,
      "loss": 3.1834,
      "step": 80400
    },
    {
      "epoch": 8.673854447439354,
      "grad_norm": 0.8390142321586609,
      "learning_rate": 7.985536967080409e-05,
      "loss": 3.1827,
      "step": 80450
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 0.8010692000389099,
      "learning_rate": 7.953157042633566e-05,
      "loss": 3.1803,
      "step": 80500
    },
    {
      "epoch": 8.684636118598382,
      "grad_norm": 0.9025766253471375,
      "learning_rate": 7.920777118186724e-05,
      "loss": 3.1869,
      "step": 80550
    },
    {
      "epoch": 8.690026954177897,
      "grad_norm": 0.83486008644104,
      "learning_rate": 7.888397193739879e-05,
      "loss": 3.1919,
      "step": 80600
    },
    {
      "epoch": 8.695417789757412,
      "grad_norm": 0.8118559122085571,
      "learning_rate": 7.856017269293037e-05,
      "loss": 3.1851,
      "step": 80650
    },
    {
      "epoch": 8.700808625336927,
      "grad_norm": 0.8715183734893799,
      "learning_rate": 7.823637344846195e-05,
      "loss": 3.1723,
      "step": 80700
    },
    {
      "epoch": 8.706199460916443,
      "grad_norm": 0.867709219455719,
      "learning_rate": 7.791257420399353e-05,
      "loss": 3.1792,
      "step": 80750
    },
    {
      "epoch": 8.711590296495956,
      "grad_norm": 0.8244636654853821,
      "learning_rate": 7.758877495952508e-05,
      "loss": 3.1918,
      "step": 80800
    },
    {
      "epoch": 8.716981132075471,
      "grad_norm": 0.8116188049316406,
      "learning_rate": 7.726497571505665e-05,
      "loss": 3.1863,
      "step": 80850
    },
    {
      "epoch": 8.722371967654986,
      "grad_norm": 0.8385442495346069,
      "learning_rate": 7.694117647058823e-05,
      "loss": 3.1733,
      "step": 80900
    },
    {
      "epoch": 8.727762803234501,
      "grad_norm": 0.8450207710266113,
      "learning_rate": 7.661737722611981e-05,
      "loss": 3.1752,
      "step": 80950
    },
    {
      "epoch": 8.733153638814017,
      "grad_norm": 0.855089545249939,
      "learning_rate": 7.629357798165137e-05,
      "loss": 3.1893,
      "step": 81000
    },
    {
      "epoch": 8.733153638814017,
      "eval_accuracy": 0.3925210523119107,
      "eval_loss": 3.317495346069336,
      "eval_runtime": 184.9457,
      "eval_samples_per_second": 97.385,
      "eval_steps_per_second": 6.088,
      "step": 81000
    },
    {
      "epoch": 8.738544474393532,
      "grad_norm": 0.841830849647522,
      "learning_rate": 7.596977873718294e-05,
      "loss": 3.1854,
      "step": 81050
    },
    {
      "epoch": 8.743935309973045,
      "grad_norm": 0.8838796019554138,
      "learning_rate": 7.564597949271451e-05,
      "loss": 3.1856,
      "step": 81100
    },
    {
      "epoch": 8.74932614555256,
      "grad_norm": 0.8529191613197327,
      "learning_rate": 7.532218024824608e-05,
      "loss": 3.1826,
      "step": 81150
    },
    {
      "epoch": 8.754716981132075,
      "grad_norm": 0.8259932398796082,
      "learning_rate": 7.499838100377764e-05,
      "loss": 3.1631,
      "step": 81200
    },
    {
      "epoch": 8.76010781671159,
      "grad_norm": 0.9050982594490051,
      "learning_rate": 7.467458175930922e-05,
      "loss": 3.1785,
      "step": 81250
    },
    {
      "epoch": 8.765498652291106,
      "grad_norm": 0.8278824687004089,
      "learning_rate": 7.435078251484078e-05,
      "loss": 3.1802,
      "step": 81300
    },
    {
      "epoch": 8.77088948787062,
      "grad_norm": 0.8377477526664734,
      "learning_rate": 7.402698327037236e-05,
      "loss": 3.1874,
      "step": 81350
    },
    {
      "epoch": 8.776280323450134,
      "grad_norm": 0.8223797678947449,
      "learning_rate": 7.370318402590393e-05,
      "loss": 3.1866,
      "step": 81400
    },
    {
      "epoch": 8.78167115902965,
      "grad_norm": 0.8468875288963318,
      "learning_rate": 7.33793847814355e-05,
      "loss": 3.1894,
      "step": 81450
    },
    {
      "epoch": 8.787061994609164,
      "grad_norm": 0.8402231931686401,
      "learning_rate": 7.305558553696707e-05,
      "loss": 3.198,
      "step": 81500
    },
    {
      "epoch": 8.79245283018868,
      "grad_norm": 0.965043842792511,
      "learning_rate": 7.273178629249865e-05,
      "loss": 3.1848,
      "step": 81550
    },
    {
      "epoch": 8.797843665768195,
      "grad_norm": 0.8373719453811646,
      "learning_rate": 7.240798704803021e-05,
      "loss": 3.1881,
      "step": 81600
    },
    {
      "epoch": 8.80323450134771,
      "grad_norm": 0.859524130821228,
      "learning_rate": 7.208418780356179e-05,
      "loss": 3.1844,
      "step": 81650
    },
    {
      "epoch": 8.808625336927223,
      "grad_norm": 0.87333744764328,
      "learning_rate": 7.176038855909337e-05,
      "loss": 3.1909,
      "step": 81700
    },
    {
      "epoch": 8.814016172506738,
      "grad_norm": 0.8479647636413574,
      "learning_rate": 7.143658931462493e-05,
      "loss": 3.1834,
      "step": 81750
    },
    {
      "epoch": 8.819407008086253,
      "grad_norm": 0.8975802659988403,
      "learning_rate": 7.11127900701565e-05,
      "loss": 3.1805,
      "step": 81800
    },
    {
      "epoch": 8.824797843665769,
      "grad_norm": 0.8739588260650635,
      "learning_rate": 7.078899082568807e-05,
      "loss": 3.1803,
      "step": 81850
    },
    {
      "epoch": 8.830188679245284,
      "grad_norm": 0.8485525250434875,
      "learning_rate": 7.046519158121964e-05,
      "loss": 3.1731,
      "step": 81900
    },
    {
      "epoch": 8.835579514824797,
      "grad_norm": 0.8152867555618286,
      "learning_rate": 7.014139233675121e-05,
      "loss": 3.2022,
      "step": 81950
    },
    {
      "epoch": 8.840970350404312,
      "grad_norm": 0.8222323060035706,
      "learning_rate": 6.981759309228278e-05,
      "loss": 3.1843,
      "step": 82000
    },
    {
      "epoch": 8.840970350404312,
      "eval_accuracy": 0.3926712106028703,
      "eval_loss": 3.3136684894561768,
      "eval_runtime": 184.9068,
      "eval_samples_per_second": 97.406,
      "eval_steps_per_second": 6.09,
      "step": 82000
    },
    {
      "epoch": 8.846361185983827,
      "grad_norm": 0.8213887214660645,
      "learning_rate": 6.949379384781434e-05,
      "loss": 3.1873,
      "step": 82050
    },
    {
      "epoch": 8.851752021563343,
      "grad_norm": 0.8546909093856812,
      "learning_rate": 6.916999460334592e-05,
      "loss": 3.1781,
      "step": 82100
    },
    {
      "epoch": 8.857142857142858,
      "grad_norm": 0.8814563751220703,
      "learning_rate": 6.884619535887748e-05,
      "loss": 3.1857,
      "step": 82150
    },
    {
      "epoch": 8.862533692722373,
      "grad_norm": 0.8545170426368713,
      "learning_rate": 6.852239611440906e-05,
      "loss": 3.1847,
      "step": 82200
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 0.8242815732955933,
      "learning_rate": 6.819859686994062e-05,
      "loss": 3.1869,
      "step": 82250
    },
    {
      "epoch": 8.873315363881401,
      "grad_norm": 0.8673837184906006,
      "learning_rate": 6.78747976254722e-05,
      "loss": 3.1932,
      "step": 82300
    },
    {
      "epoch": 8.878706199460916,
      "grad_norm": 0.8401963114738464,
      "learning_rate": 6.755747436589314e-05,
      "loss": 3.1965,
      "step": 82350
    },
    {
      "epoch": 8.884097035040432,
      "grad_norm": 0.8353580236434937,
      "learning_rate": 6.72336751214247e-05,
      "loss": 3.172,
      "step": 82400
    },
    {
      "epoch": 8.889487870619947,
      "grad_norm": 0.8120338320732117,
      "learning_rate": 6.690987587695628e-05,
      "loss": 3.1932,
      "step": 82450
    },
    {
      "epoch": 8.89487870619946,
      "grad_norm": 0.8129428029060364,
      "learning_rate": 6.658607663248785e-05,
      "loss": 3.1704,
      "step": 82500
    },
    {
      "epoch": 8.900269541778975,
      "grad_norm": 0.8249984979629517,
      "learning_rate": 6.626227738801942e-05,
      "loss": 3.1927,
      "step": 82550
    },
    {
      "epoch": 8.90566037735849,
      "grad_norm": 0.8213885426521301,
      "learning_rate": 6.593847814355099e-05,
      "loss": 3.1605,
      "step": 82600
    },
    {
      "epoch": 8.911051212938006,
      "grad_norm": 0.8929885029792786,
      "learning_rate": 6.561467889908257e-05,
      "loss": 3.1752,
      "step": 82650
    },
    {
      "epoch": 8.91644204851752,
      "grad_norm": 0.8327133655548096,
      "learning_rate": 6.529087965461413e-05,
      "loss": 3.1778,
      "step": 82700
    },
    {
      "epoch": 8.921832884097036,
      "grad_norm": 0.8178892731666565,
      "learning_rate": 6.496708041014571e-05,
      "loss": 3.1733,
      "step": 82750
    },
    {
      "epoch": 8.92722371967655,
      "grad_norm": 0.8787632584571838,
      "learning_rate": 6.464328116567727e-05,
      "loss": 3.1881,
      "step": 82800
    },
    {
      "epoch": 8.932614555256064,
      "grad_norm": 0.8857139945030212,
      "learning_rate": 6.431948192120885e-05,
      "loss": 3.1853,
      "step": 82850
    },
    {
      "epoch": 8.93800539083558,
      "grad_norm": 0.8443358540534973,
      "learning_rate": 6.399568267674041e-05,
      "loss": 3.1897,
      "step": 82900
    },
    {
      "epoch": 8.943396226415095,
      "grad_norm": 0.7980090379714966,
      "learning_rate": 6.367188343227199e-05,
      "loss": 3.1898,
      "step": 82950
    },
    {
      "epoch": 8.94878706199461,
      "grad_norm": 0.9037806391716003,
      "learning_rate": 6.334808418780356e-05,
      "loss": 3.1679,
      "step": 83000
    },
    {
      "epoch": 8.94878706199461,
      "eval_accuracy": 0.3930123806706078,
      "eval_loss": 3.31054425239563,
      "eval_runtime": 184.7094,
      "eval_samples_per_second": 97.51,
      "eval_steps_per_second": 6.096,
      "step": 83000
    },
    {
      "epoch": 8.954177897574123,
      "grad_norm": 0.8617427945137024,
      "learning_rate": 6.302428494333512e-05,
      "loss": 3.1946,
      "step": 83050
    },
    {
      "epoch": 8.959568733153638,
      "grad_norm": 0.8506762385368347,
      "learning_rate": 6.27004856988667e-05,
      "loss": 3.181,
      "step": 83100
    },
    {
      "epoch": 8.964959568733153,
      "grad_norm": 0.8383913040161133,
      "learning_rate": 6.237668645439826e-05,
      "loss": 3.1833,
      "step": 83150
    },
    {
      "epoch": 8.970350404312669,
      "grad_norm": 0.8611315488815308,
      "learning_rate": 6.205288720992984e-05,
      "loss": 3.1827,
      "step": 83200
    },
    {
      "epoch": 8.975741239892184,
      "grad_norm": 0.8510012626647949,
      "learning_rate": 6.17290879654614e-05,
      "loss": 3.1884,
      "step": 83250
    },
    {
      "epoch": 8.981132075471699,
      "grad_norm": 0.826678991317749,
      "learning_rate": 6.140528872099298e-05,
      "loss": 3.1951,
      "step": 83300
    },
    {
      "epoch": 8.986522911051212,
      "grad_norm": 0.8059219121932983,
      "learning_rate": 6.108148947652455e-05,
      "loss": 3.1925,
      "step": 83350
    },
    {
      "epoch": 8.991913746630727,
      "grad_norm": 0.8477720618247986,
      "learning_rate": 6.075769023205612e-05,
      "loss": 3.1788,
      "step": 83400
    },
    {
      "epoch": 8.997304582210242,
      "grad_norm": 0.8244584798812866,
      "learning_rate": 6.043389098758769e-05,
      "loss": 3.1788,
      "step": 83450
    },
    {
      "epoch": 9.002695417789758,
      "grad_norm": 0.8307106494903564,
      "learning_rate": 6.0110091743119265e-05,
      "loss": 3.155,
      "step": 83500
    },
    {
      "epoch": 9.008086253369273,
      "grad_norm": 0.880568265914917,
      "learning_rate": 5.978629249865083e-05,
      "loss": 3.1189,
      "step": 83550
    },
    {
      "epoch": 9.013477088948788,
      "grad_norm": 0.8291214108467102,
      "learning_rate": 5.94624932541824e-05,
      "loss": 3.1419,
      "step": 83600
    },
    {
      "epoch": 9.018867924528301,
      "grad_norm": 0.814835250377655,
      "learning_rate": 5.913869400971397e-05,
      "loss": 3.1524,
      "step": 83650
    },
    {
      "epoch": 9.024258760107816,
      "grad_norm": 0.8376071453094482,
      "learning_rate": 5.881489476524554e-05,
      "loss": 3.1034,
      "step": 83700
    },
    {
      "epoch": 9.029649595687331,
      "grad_norm": 0.8205394744873047,
      "learning_rate": 5.849109552077711e-05,
      "loss": 3.1062,
      "step": 83750
    },
    {
      "epoch": 9.035040431266847,
      "grad_norm": 0.8897798657417297,
      "learning_rate": 5.8167296276308683e-05,
      "loss": 3.1204,
      "step": 83800
    },
    {
      "epoch": 9.040431266846362,
      "grad_norm": 0.8899309039115906,
      "learning_rate": 5.784349703184026e-05,
      "loss": 3.133,
      "step": 83850
    },
    {
      "epoch": 9.045822102425875,
      "grad_norm": 0.8886575698852539,
      "learning_rate": 5.7519697787371825e-05,
      "loss": 3.1176,
      "step": 83900
    },
    {
      "epoch": 9.05121293800539,
      "grad_norm": 0.8485926389694214,
      "learning_rate": 5.7195898542903396e-05,
      "loss": 3.102,
      "step": 83950
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 0.8616915941238403,
      "learning_rate": 5.687209929843497e-05,
      "loss": 3.1293,
      "step": 84000
    },
    {
      "epoch": 9.056603773584905,
      "eval_accuracy": 0.3930330247192925,
      "eval_loss": 3.3129842281341553,
      "eval_runtime": 184.7976,
      "eval_samples_per_second": 97.463,
      "eval_steps_per_second": 6.093,
      "step": 84000
    },
    {
      "epoch": 9.06199460916442,
      "grad_norm": 0.8321478962898254,
      "learning_rate": 5.654830005396654e-05,
      "loss": 3.127,
      "step": 84050
    },
    {
      "epoch": 9.067385444743936,
      "grad_norm": 0.883215069770813,
      "learning_rate": 5.62245008094981e-05,
      "loss": 3.1293,
      "step": 84100
    },
    {
      "epoch": 9.07277628032345,
      "grad_norm": 0.8449257612228394,
      "learning_rate": 5.590070156502968e-05,
      "loss": 3.1404,
      "step": 84150
    },
    {
      "epoch": 9.078167115902964,
      "grad_norm": 0.8433477878570557,
      "learning_rate": 5.5576902320561244e-05,
      "loss": 3.117,
      "step": 84200
    },
    {
      "epoch": 9.08355795148248,
      "grad_norm": 0.8466358780860901,
      "learning_rate": 5.525310307609282e-05,
      "loss": 3.1382,
      "step": 84250
    },
    {
      "epoch": 9.088948787061994,
      "grad_norm": 0.8381649255752563,
      "learning_rate": 5.4929303831624386e-05,
      "loss": 3.1199,
      "step": 84300
    },
    {
      "epoch": 9.09433962264151,
      "grad_norm": 0.840488076210022,
      "learning_rate": 5.460550458715596e-05,
      "loss": 3.1297,
      "step": 84350
    },
    {
      "epoch": 9.099730458221025,
      "grad_norm": 0.8325525522232056,
      "learning_rate": 5.428170534268753e-05,
      "loss": 3.0981,
      "step": 84400
    },
    {
      "epoch": 9.10512129380054,
      "grad_norm": 0.8756667971611023,
      "learning_rate": 5.39579060982191e-05,
      "loss": 3.111,
      "step": 84450
    },
    {
      "epoch": 9.110512129380053,
      "grad_norm": 0.8472950458526611,
      "learning_rate": 5.363410685375067e-05,
      "loss": 3.1322,
      "step": 84500
    },
    {
      "epoch": 9.115902964959568,
      "grad_norm": 0.818108081817627,
      "learning_rate": 5.331030760928224e-05,
      "loss": 3.1101,
      "step": 84550
    },
    {
      "epoch": 9.121293800539084,
      "grad_norm": 0.8659102916717529,
      "learning_rate": 5.2986508364813804e-05,
      "loss": 3.1253,
      "step": 84600
    },
    {
      "epoch": 9.126684636118599,
      "grad_norm": 0.8552064895629883,
      "learning_rate": 5.266270912034538e-05,
      "loss": 3.1252,
      "step": 84650
    },
    {
      "epoch": 9.132075471698114,
      "grad_norm": 0.8596740365028381,
      "learning_rate": 5.233890987587695e-05,
      "loss": 3.1311,
      "step": 84700
    },
    {
      "epoch": 9.137466307277627,
      "grad_norm": 0.8284486532211304,
      "learning_rate": 5.2015110631408524e-05,
      "loss": 3.1365,
      "step": 84750
    },
    {
      "epoch": 9.142857142857142,
      "grad_norm": 0.8481780886650085,
      "learning_rate": 5.1691311386940094e-05,
      "loss": 3.131,
      "step": 84800
    },
    {
      "epoch": 9.148247978436657,
      "grad_norm": 0.824164092540741,
      "learning_rate": 5.136751214247166e-05,
      "loss": 3.1193,
      "step": 84850
    },
    {
      "epoch": 9.153638814016173,
      "grad_norm": 0.851373016834259,
      "learning_rate": 5.1050188882892604e-05,
      "loss": 3.1248,
      "step": 84900
    },
    {
      "epoch": 9.159029649595688,
      "grad_norm": 0.8321095705032349,
      "learning_rate": 5.072638963842417e-05,
      "loss": 3.1121,
      "step": 84950
    },
    {
      "epoch": 9.164420485175203,
      "grad_norm": 0.7981374263763428,
      "learning_rate": 5.0402590393955746e-05,
      "loss": 3.1214,
      "step": 85000
    },
    {
      "epoch": 9.164420485175203,
      "eval_accuracy": 0.39314189491288265,
      "eval_loss": 3.31256365776062,
      "eval_runtime": 184.9292,
      "eval_samples_per_second": 97.394,
      "eval_steps_per_second": 6.089,
      "step": 85000
    },
    {
      "epoch": 9.169811320754716,
      "grad_norm": 0.8480749726295471,
      "learning_rate": 5.007879114948731e-05,
      "loss": 3.1318,
      "step": 85050
    },
    {
      "epoch": 9.175202156334231,
      "grad_norm": 0.8506357073783875,
      "learning_rate": 4.975499190501889e-05,
      "loss": 3.1311,
      "step": 85100
    },
    {
      "epoch": 9.180592991913747,
      "grad_norm": 0.864138662815094,
      "learning_rate": 4.943119266055045e-05,
      "loss": 3.1381,
      "step": 85150
    },
    {
      "epoch": 9.185983827493262,
      "grad_norm": 0.853118360042572,
      "learning_rate": 4.910739341608202e-05,
      "loss": 3.1358,
      "step": 85200
    },
    {
      "epoch": 9.191374663072777,
      "grad_norm": 0.8572016358375549,
      "learning_rate": 4.87835941716136e-05,
      "loss": 3.1298,
      "step": 85250
    },
    {
      "epoch": 9.19676549865229,
      "grad_norm": 0.8740936517715454,
      "learning_rate": 4.8459794927145164e-05,
      "loss": 3.1435,
      "step": 85300
    },
    {
      "epoch": 9.202156334231805,
      "grad_norm": 0.8369245529174805,
      "learning_rate": 4.813599568267674e-05,
      "loss": 3.1389,
      "step": 85350
    },
    {
      "epoch": 9.20754716981132,
      "grad_norm": 0.8236960768699646,
      "learning_rate": 4.7812196438208306e-05,
      "loss": 3.123,
      "step": 85400
    },
    {
      "epoch": 9.212938005390836,
      "grad_norm": 0.8511966466903687,
      "learning_rate": 4.748839719373988e-05,
      "loss": 3.124,
      "step": 85450
    },
    {
      "epoch": 9.21832884097035,
      "grad_norm": 0.8575359582901001,
      "learning_rate": 4.716459794927145e-05,
      "loss": 3.1427,
      "step": 85500
    },
    {
      "epoch": 9.223719676549866,
      "grad_norm": 0.8404532670974731,
      "learning_rate": 4.684079870480302e-05,
      "loss": 3.1437,
      "step": 85550
    },
    {
      "epoch": 9.22911051212938,
      "grad_norm": 0.9211980104446411,
      "learning_rate": 4.651699946033459e-05,
      "loss": 3.1337,
      "step": 85600
    },
    {
      "epoch": 9.234501347708894,
      "grad_norm": 0.820880115032196,
      "learning_rate": 4.619320021586616e-05,
      "loss": 3.1308,
      "step": 85650
    },
    {
      "epoch": 9.23989218328841,
      "grad_norm": 0.8520514369010925,
      "learning_rate": 4.5869400971397725e-05,
      "loss": 3.1308,
      "step": 85700
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 0.8518171906471252,
      "learning_rate": 4.55456017269293e-05,
      "loss": 3.1324,
      "step": 85750
    },
    {
      "epoch": 9.25067385444744,
      "grad_norm": 0.8360389471054077,
      "learning_rate": 4.5221802482460866e-05,
      "loss": 3.1236,
      "step": 85800
    },
    {
      "epoch": 9.256064690026955,
      "grad_norm": 0.8417567610740662,
      "learning_rate": 4.4898003237992444e-05,
      "loss": 3.1152,
      "step": 85850
    },
    {
      "epoch": 9.261455525606468,
      "grad_norm": 0.8565623164176941,
      "learning_rate": 4.457420399352401e-05,
      "loss": 3.1323,
      "step": 85900
    },
    {
      "epoch": 9.266846361185983,
      "grad_norm": 0.8710352182388306,
      "learning_rate": 4.425040474905558e-05,
      "loss": 3.1216,
      "step": 85950
    },
    {
      "epoch": 9.272237196765499,
      "grad_norm": 0.8659188151359558,
      "learning_rate": 4.392660550458716e-05,
      "loss": 3.1466,
      "step": 86000
    },
    {
      "epoch": 9.272237196765499,
      "eval_accuracy": 0.3936142090162122,
      "eval_loss": 3.3102121353149414,
      "eval_runtime": 184.7159,
      "eval_samples_per_second": 97.507,
      "eval_steps_per_second": 6.096,
      "step": 86000
    },
    {
      "epoch": 9.277628032345014,
      "grad_norm": 0.8267226219177246,
      "learning_rate": 4.360280626011872e-05,
      "loss": 3.1328,
      "step": 86050
    },
    {
      "epoch": 9.283018867924529,
      "grad_norm": 0.848037838935852,
      "learning_rate": 4.32790070156503e-05,
      "loss": 3.1186,
      "step": 86100
    },
    {
      "epoch": 9.288409703504042,
      "grad_norm": 0.841865062713623,
      "learning_rate": 4.295520777118186e-05,
      "loss": 3.1225,
      "step": 86150
    },
    {
      "epoch": 9.293800539083557,
      "grad_norm": 0.8285828232765198,
      "learning_rate": 4.263788451160281e-05,
      "loss": 3.1311,
      "step": 86200
    },
    {
      "epoch": 9.299191374663073,
      "grad_norm": 0.8383690714836121,
      "learning_rate": 4.231408526713437e-05,
      "loss": 3.1298,
      "step": 86250
    },
    {
      "epoch": 9.304582210242588,
      "grad_norm": 0.8436566591262817,
      "learning_rate": 4.199028602266594e-05,
      "loss": 3.129,
      "step": 86300
    },
    {
      "epoch": 9.309973045822103,
      "grad_norm": 0.8680644631385803,
      "learning_rate": 4.1666486778197514e-05,
      "loss": 3.1218,
      "step": 86350
    },
    {
      "epoch": 9.315363881401618,
      "grad_norm": 0.8822382092475891,
      "learning_rate": 4.1342687533729085e-05,
      "loss": 3.1494,
      "step": 86400
    },
    {
      "epoch": 9.320754716981131,
      "grad_norm": 0.8971915245056152,
      "learning_rate": 4.101888828926065e-05,
      "loss": 3.1238,
      "step": 86450
    },
    {
      "epoch": 9.326145552560646,
      "grad_norm": 0.8807439208030701,
      "learning_rate": 4.0695089044792226e-05,
      "loss": 3.1397,
      "step": 86500
    },
    {
      "epoch": 9.331536388140162,
      "grad_norm": 0.8657439351081848,
      "learning_rate": 4.037128980032379e-05,
      "loss": 3.1451,
      "step": 86550
    },
    {
      "epoch": 9.336927223719677,
      "grad_norm": 0.8694212436676025,
      "learning_rate": 4.004749055585537e-05,
      "loss": 3.1326,
      "step": 86600
    },
    {
      "epoch": 9.342318059299192,
      "grad_norm": 0.868144690990448,
      "learning_rate": 3.972369131138694e-05,
      "loss": 3.1345,
      "step": 86650
    },
    {
      "epoch": 9.347708894878707,
      "grad_norm": 0.9292327761650085,
      "learning_rate": 3.93998920669185e-05,
      "loss": 3.1448,
      "step": 86700
    },
    {
      "epoch": 9.35309973045822,
      "grad_norm": 0.8244055509567261,
      "learning_rate": 3.907609282245008e-05,
      "loss": 3.1568,
      "step": 86750
    },
    {
      "epoch": 9.358490566037736,
      "grad_norm": 0.8296806216239929,
      "learning_rate": 3.8752293577981645e-05,
      "loss": 3.1391,
      "step": 86800
    },
    {
      "epoch": 9.36388140161725,
      "grad_norm": 0.8584550023078918,
      "learning_rate": 3.842849433351322e-05,
      "loss": 3.1285,
      "step": 86850
    },
    {
      "epoch": 9.369272237196766,
      "grad_norm": 0.936661422252655,
      "learning_rate": 3.810469508904479e-05,
      "loss": 3.1412,
      "step": 86900
    },
    {
      "epoch": 9.374663072776281,
      "grad_norm": 0.8722965717315674,
      "learning_rate": 3.7780895844576364e-05,
      "loss": 3.1186,
      "step": 86950
    },
    {
      "epoch": 9.380053908355794,
      "grad_norm": 0.8549427390098572,
      "learning_rate": 3.745709660010793e-05,
      "loss": 3.1247,
      "step": 87000
    },
    {
      "epoch": 9.380053908355794,
      "eval_accuracy": 0.3936190983961639,
      "eval_loss": 3.3095529079437256,
      "eval_runtime": 184.7197,
      "eval_samples_per_second": 97.504,
      "eval_steps_per_second": 6.096,
      "step": 87000
    },
    {
      "epoch": 9.38544474393531,
      "grad_norm": 0.8127095699310303,
      "learning_rate": 3.71332973556395e-05,
      "loss": 3.1343,
      "step": 87050
    },
    {
      "epoch": 9.390835579514825,
      "grad_norm": 0.8294982314109802,
      "learning_rate": 3.680949811117107e-05,
      "loss": 3.1216,
      "step": 87100
    },
    {
      "epoch": 9.39622641509434,
      "grad_norm": 0.8491564393043518,
      "learning_rate": 3.648569886670264e-05,
      "loss": 3.1247,
      "step": 87150
    },
    {
      "epoch": 9.401617250673855,
      "grad_norm": 0.8749399781227112,
      "learning_rate": 3.616189962223421e-05,
      "loss": 3.1505,
      "step": 87200
    },
    {
      "epoch": 9.40700808625337,
      "grad_norm": 0.8547865152359009,
      "learning_rate": 3.583810037776578e-05,
      "loss": 3.1383,
      "step": 87250
    },
    {
      "epoch": 9.412398921832883,
      "grad_norm": 0.8318867683410645,
      "learning_rate": 3.5514301133297354e-05,
      "loss": 3.1398,
      "step": 87300
    },
    {
      "epoch": 9.417789757412399,
      "grad_norm": 0.8504889607429504,
      "learning_rate": 3.5190501888828925e-05,
      "loss": 3.1407,
      "step": 87350
    },
    {
      "epoch": 9.423180592991914,
      "grad_norm": 0.8165703415870667,
      "learning_rate": 3.4866702644360496e-05,
      "loss": 3.1343,
      "step": 87400
    },
    {
      "epoch": 9.428571428571429,
      "grad_norm": 0.8537405729293823,
      "learning_rate": 3.4542903399892067e-05,
      "loss": 3.142,
      "step": 87450
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 0.8769167065620422,
      "learning_rate": 3.421910415542363e-05,
      "loss": 3.1425,
      "step": 87500
    },
    {
      "epoch": 9.439353099730457,
      "grad_norm": 0.815278172492981,
      "learning_rate": 3.38953049109552e-05,
      "loss": 3.1374,
      "step": 87550
    },
    {
      "epoch": 9.444743935309972,
      "grad_norm": 0.8766458630561829,
      "learning_rate": 3.357150566648677e-05,
      "loss": 3.1123,
      "step": 87600
    },
    {
      "epoch": 9.450134770889488,
      "grad_norm": 0.8410953283309937,
      "learning_rate": 3.324770642201834e-05,
      "loss": 3.1393,
      "step": 87650
    },
    {
      "epoch": 9.455525606469003,
      "grad_norm": 0.8328889012336731,
      "learning_rate": 3.2923907177549914e-05,
      "loss": 3.134,
      "step": 87700
    },
    {
      "epoch": 9.460916442048518,
      "grad_norm": 0.9297603368759155,
      "learning_rate": 3.2600107933081485e-05,
      "loss": 3.1313,
      "step": 87750
    },
    {
      "epoch": 9.466307277628033,
      "grad_norm": 0.8465941548347473,
      "learning_rate": 3.2276308688613056e-05,
      "loss": 3.1189,
      "step": 87800
    },
    {
      "epoch": 9.471698113207546,
      "grad_norm": 0.8548228144645691,
      "learning_rate": 3.195250944414463e-05,
      "loss": 3.1317,
      "step": 87850
    },
    {
      "epoch": 9.477088948787062,
      "grad_norm": 0.8719295263290405,
      "learning_rate": 3.16287101996762e-05,
      "loss": 3.119,
      "step": 87900
    },
    {
      "epoch": 9.482479784366577,
      "grad_norm": 0.8304482102394104,
      "learning_rate": 3.130491095520777e-05,
      "loss": 3.1184,
      "step": 87950
    },
    {
      "epoch": 9.487870619946092,
      "grad_norm": 0.8402472734451294,
      "learning_rate": 3.098111171073934e-05,
      "loss": 3.132,
      "step": 88000
    },
    {
      "epoch": 9.487870619946092,
      "eval_accuracy": 0.3939127871519264,
      "eval_loss": 3.3066344261169434,
      "eval_runtime": 185.0557,
      "eval_samples_per_second": 97.327,
      "eval_steps_per_second": 6.085,
      "step": 88000
    },
    {
      "epoch": 9.493261455525607,
      "grad_norm": 0.8302611112594604,
      "learning_rate": 3.065731246627091e-05,
      "loss": 3.1315,
      "step": 88050
    },
    {
      "epoch": 9.498652291105122,
      "grad_norm": 0.8773887753486633,
      "learning_rate": 3.033351322180248e-05,
      "loss": 3.1278,
      "step": 88100
    },
    {
      "epoch": 9.504043126684635,
      "grad_norm": 0.8220431804656982,
      "learning_rate": 3.000971397733405e-05,
      "loss": 3.1538,
      "step": 88150
    },
    {
      "epoch": 9.50943396226415,
      "grad_norm": 0.8599470257759094,
      "learning_rate": 2.968591473286562e-05,
      "loss": 3.1354,
      "step": 88200
    },
    {
      "epoch": 9.514824797843666,
      "grad_norm": 0.8439090251922607,
      "learning_rate": 2.936211548839719e-05,
      "loss": 3.1212,
      "step": 88250
    },
    {
      "epoch": 9.520215633423181,
      "grad_norm": 0.9031540751457214,
      "learning_rate": 2.903831624392876e-05,
      "loss": 3.141,
      "step": 88300
    },
    {
      "epoch": 9.525606469002696,
      "grad_norm": 0.8797536492347717,
      "learning_rate": 2.8714516999460332e-05,
      "loss": 3.1318,
      "step": 88350
    },
    {
      "epoch": 9.530997304582211,
      "grad_norm": 0.8907245397567749,
      "learning_rate": 2.83907177549919e-05,
      "loss": 3.1219,
      "step": 88400
    },
    {
      "epoch": 9.536388140161725,
      "grad_norm": 0.8620273470878601,
      "learning_rate": 2.806691851052347e-05,
      "loss": 3.1271,
      "step": 88450
    },
    {
      "epoch": 9.54177897574124,
      "grad_norm": 0.8356373310089111,
      "learning_rate": 2.774311926605504e-05,
      "loss": 3.1419,
      "step": 88500
    },
    {
      "epoch": 9.547169811320755,
      "grad_norm": 0.8548893928527832,
      "learning_rate": 2.7419320021586613e-05,
      "loss": 3.1212,
      "step": 88550
    },
    {
      "epoch": 9.55256064690027,
      "grad_norm": 0.8545761704444885,
      "learning_rate": 2.7095520777118187e-05,
      "loss": 3.1353,
      "step": 88600
    },
    {
      "epoch": 9.557951482479785,
      "grad_norm": 0.8439458608627319,
      "learning_rate": 2.6771721532649758e-05,
      "loss": 3.1304,
      "step": 88650
    },
    {
      "epoch": 9.563342318059298,
      "grad_norm": 0.8117093443870544,
      "learning_rate": 2.6447922288181325e-05,
      "loss": 3.1321,
      "step": 88700
    },
    {
      "epoch": 9.568733153638814,
      "grad_norm": 0.8481546640396118,
      "learning_rate": 2.6124123043712896e-05,
      "loss": 3.1397,
      "step": 88750
    },
    {
      "epoch": 9.574123989218329,
      "grad_norm": 0.8625066876411438,
      "learning_rate": 2.5800323799244467e-05,
      "loss": 3.1234,
      "step": 88800
    },
    {
      "epoch": 9.579514824797844,
      "grad_norm": 0.8870933651924133,
      "learning_rate": 2.5476524554776038e-05,
      "loss": 3.132,
      "step": 88850
    },
    {
      "epoch": 9.584905660377359,
      "grad_norm": 0.8924064636230469,
      "learning_rate": 2.515272531030761e-05,
      "loss": 3.156,
      "step": 88900
    },
    {
      "epoch": 9.590296495956874,
      "grad_norm": 0.912037193775177,
      "learning_rate": 2.4828926065839176e-05,
      "loss": 3.1337,
      "step": 88950
    },
    {
      "epoch": 9.595687331536388,
      "grad_norm": 0.8235248923301697,
      "learning_rate": 2.4505126821370747e-05,
      "loss": 3.1121,
      "step": 89000
    },
    {
      "epoch": 9.595687331536388,
      "eval_accuracy": 0.3942704724586117,
      "eval_loss": 3.304278612136841,
      "eval_runtime": 184.915,
      "eval_samples_per_second": 97.402,
      "eval_steps_per_second": 6.089,
      "step": 89000
    },
    {
      "epoch": 9.601078167115903,
      "grad_norm": 0.8647744655609131,
      "learning_rate": 2.4181327576902318e-05,
      "loss": 3.1443,
      "step": 89050
    },
    {
      "epoch": 9.606469002695418,
      "grad_norm": 0.879808783531189,
      "learning_rate": 2.385752833243389e-05,
      "loss": 3.1359,
      "step": 89100
    },
    {
      "epoch": 9.611859838274933,
      "grad_norm": 0.891708254814148,
      "learning_rate": 2.353372908796546e-05,
      "loss": 3.1443,
      "step": 89150
    },
    {
      "epoch": 9.617250673854448,
      "grad_norm": 0.9006338119506836,
      "learning_rate": 2.3209929843497027e-05,
      "loss": 3.1291,
      "step": 89200
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 0.854828417301178,
      "learning_rate": 2.2886130599028598e-05,
      "loss": 3.135,
      "step": 89250
    },
    {
      "epoch": 9.628032345013477,
      "grad_norm": 0.8300639986991882,
      "learning_rate": 2.256233135456017e-05,
      "loss": 3.113,
      "step": 89300
    },
    {
      "epoch": 9.633423180592992,
      "grad_norm": 0.815534234046936,
      "learning_rate": 2.2238532110091743e-05,
      "loss": 3.1478,
      "step": 89350
    },
    {
      "epoch": 9.638814016172507,
      "grad_norm": 0.8253652453422546,
      "learning_rate": 2.1914732865623314e-05,
      "loss": 3.1443,
      "step": 89400
    },
    {
      "epoch": 9.644204851752022,
      "grad_norm": 0.8685674667358398,
      "learning_rate": 2.1590933621154885e-05,
      "loss": 3.1428,
      "step": 89450
    },
    {
      "epoch": 9.649595687331537,
      "grad_norm": 0.8638211488723755,
      "learning_rate": 2.1267134376686453e-05,
      "loss": 3.1376,
      "step": 89500
    },
    {
      "epoch": 9.65498652291105,
      "grad_norm": 0.8781023621559143,
      "learning_rate": 2.0943335132218024e-05,
      "loss": 3.1339,
      "step": 89550
    },
    {
      "epoch": 9.660377358490566,
      "grad_norm": 0.8183006644248962,
      "learning_rate": 2.0619535887749594e-05,
      "loss": 3.1411,
      "step": 89600
    },
    {
      "epoch": 9.66576819407008,
      "grad_norm": 0.8766468167304993,
      "learning_rate": 2.0295736643281165e-05,
      "loss": 3.1244,
      "step": 89650
    },
    {
      "epoch": 9.671159029649596,
      "grad_norm": 0.8593804240226746,
      "learning_rate": 1.9971937398812736e-05,
      "loss": 3.1434,
      "step": 89700
    },
    {
      "epoch": 9.676549865229111,
      "grad_norm": 0.8419660329818726,
      "learning_rate": 1.9648138154344304e-05,
      "loss": 3.1206,
      "step": 89750
    },
    {
      "epoch": 9.681940700808624,
      "grad_norm": 0.8620070219039917,
      "learning_rate": 1.9324338909875875e-05,
      "loss": 3.1244,
      "step": 89800
    },
    {
      "epoch": 9.68733153638814,
      "grad_norm": 0.8452453017234802,
      "learning_rate": 1.9000539665407446e-05,
      "loss": 3.1357,
      "step": 89850
    },
    {
      "epoch": 9.692722371967655,
      "grad_norm": 0.8461737036705017,
      "learning_rate": 1.8676740420939016e-05,
      "loss": 3.1241,
      "step": 89900
    },
    {
      "epoch": 9.69811320754717,
      "grad_norm": 0.9039965867996216,
      "learning_rate": 1.8352941176470587e-05,
      "loss": 3.1415,
      "step": 89950
    },
    {
      "epoch": 9.703504043126685,
      "grad_norm": 0.8599136471748352,
      "learning_rate": 1.8029141932002158e-05,
      "loss": 3.1384,
      "step": 90000
    },
    {
      "epoch": 9.703504043126685,
      "eval_accuracy": 0.3945016858038809,
      "eval_loss": 3.3018898963928223,
      "eval_runtime": 184.7082,
      "eval_samples_per_second": 97.511,
      "eval_steps_per_second": 6.096,
      "step": 90000
    },
    {
      "epoch": 9.7088948787062,
      "grad_norm": 0.8381307721138,
      "learning_rate": 1.770534268753373e-05,
      "loss": 3.1231,
      "step": 90050
    },
    {
      "epoch": 9.714285714285714,
      "grad_norm": 0.8417419791221619,
      "learning_rate": 1.7381543443065297e-05,
      "loss": 3.1217,
      "step": 90100
    },
    {
      "epoch": 9.719676549865229,
      "grad_norm": 0.8387212753295898,
      "learning_rate": 1.7057744198596867e-05,
      "loss": 3.1338,
      "step": 90150
    },
    {
      "epoch": 9.725067385444744,
      "grad_norm": 0.8630876541137695,
      "learning_rate": 1.673394495412844e-05,
      "loss": 3.1339,
      "step": 90200
    },
    {
      "epoch": 9.730458221024259,
      "grad_norm": 0.836463987827301,
      "learning_rate": 1.641014570966001e-05,
      "loss": 3.1234,
      "step": 90250
    },
    {
      "epoch": 9.735849056603774,
      "grad_norm": 0.872067928314209,
      "learning_rate": 1.608634646519158e-05,
      "loss": 3.1269,
      "step": 90300
    },
    {
      "epoch": 9.74123989218329,
      "grad_norm": 0.8544020056724548,
      "learning_rate": 1.576254722072315e-05,
      "loss": 3.1358,
      "step": 90350
    },
    {
      "epoch": 9.746630727762803,
      "grad_norm": 0.8575515747070312,
      "learning_rate": 1.5438747976254722e-05,
      "loss": 3.1237,
      "step": 90400
    },
    {
      "epoch": 9.752021563342318,
      "grad_norm": 0.849241316318512,
      "learning_rate": 1.5114948731786291e-05,
      "loss": 3.1376,
      "step": 90450
    },
    {
      "epoch": 9.757412398921833,
      "grad_norm": 0.8438499569892883,
      "learning_rate": 1.4791149487317862e-05,
      "loss": 3.1372,
      "step": 90500
    },
    {
      "epoch": 9.762803234501348,
      "grad_norm": 0.9010724425315857,
      "learning_rate": 1.4467350242849431e-05,
      "loss": 3.1429,
      "step": 90550
    },
    {
      "epoch": 9.768194070080863,
      "grad_norm": 0.8614817261695862,
      "learning_rate": 1.4143550998381002e-05,
      "loss": 3.1199,
      "step": 90600
    },
    {
      "epoch": 9.773584905660378,
      "grad_norm": 0.8398626446723938,
      "learning_rate": 1.3819751753912573e-05,
      "loss": 3.1367,
      "step": 90650
    },
    {
      "epoch": 9.778975741239892,
      "grad_norm": 0.8385025262832642,
      "learning_rate": 1.3495952509444142e-05,
      "loss": 3.1227,
      "step": 90700
    },
    {
      "epoch": 9.784366576819407,
      "grad_norm": 0.8450053930282593,
      "learning_rate": 1.3172153264975715e-05,
      "loss": 3.1295,
      "step": 90750
    },
    {
      "epoch": 9.789757412398922,
      "grad_norm": 0.852215588092804,
      "learning_rate": 1.2848354020507286e-05,
      "loss": 3.1309,
      "step": 90800
    },
    {
      "epoch": 9.795148247978437,
      "grad_norm": 0.8445461392402649,
      "learning_rate": 1.2524554776038855e-05,
      "loss": 3.1305,
      "step": 90850
    },
    {
      "epoch": 9.800539083557952,
      "grad_norm": 0.8316650986671448,
      "learning_rate": 1.2200755531570426e-05,
      "loss": 3.1438,
      "step": 90900
    },
    {
      "epoch": 9.805929919137466,
      "grad_norm": 0.8489454984664917,
      "learning_rate": 1.1876956287101995e-05,
      "loss": 3.1357,
      "step": 90950
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 0.8464455008506775,
      "learning_rate": 1.1553157042633566e-05,
      "loss": 3.1282,
      "step": 91000
    },
    {
      "epoch": 9.81132075471698,
      "eval_accuracy": 0.39473615873578455,
      "eval_loss": 3.2998392581939697,
      "eval_runtime": 184.8391,
      "eval_samples_per_second": 97.441,
      "eval_steps_per_second": 6.092,
      "step": 91000
    },
    {
      "epoch": 9.816711590296496,
      "grad_norm": 0.8439268469810486,
      "learning_rate": 1.1229357798165137e-05,
      "loss": 3.1357,
      "step": 91050
    },
    {
      "epoch": 9.822102425876011,
      "grad_norm": 0.8137937784194946,
      "learning_rate": 1.0905558553696706e-05,
      "loss": 3.119,
      "step": 91100
    },
    {
      "epoch": 9.827493261455526,
      "grad_norm": 0.8356099128723145,
      "learning_rate": 1.0581759309228278e-05,
      "loss": 3.1498,
      "step": 91150
    },
    {
      "epoch": 9.832884097035041,
      "grad_norm": 0.8601353168487549,
      "learning_rate": 1.025796006475985e-05,
      "loss": 3.1239,
      "step": 91200
    },
    {
      "epoch": 9.838274932614555,
      "grad_norm": 0.8289896249771118,
      "learning_rate": 9.934160820291419e-06,
      "loss": 3.1346,
      "step": 91250
    },
    {
      "epoch": 9.84366576819407,
      "grad_norm": 0.8336220383644104,
      "learning_rate": 9.61036157582299e-06,
      "loss": 3.134,
      "step": 91300
    },
    {
      "epoch": 9.849056603773585,
      "grad_norm": 0.8082159161567688,
      "learning_rate": 9.286562331354559e-06,
      "loss": 3.1424,
      "step": 91350
    },
    {
      "epoch": 9.8544474393531,
      "grad_norm": 0.8767507076263428,
      "learning_rate": 8.96276308688613e-06,
      "loss": 3.1319,
      "step": 91400
    },
    {
      "epoch": 9.859838274932615,
      "grad_norm": 0.8523329496383667,
      "learning_rate": 8.6389638424177e-06,
      "loss": 3.1263,
      "step": 91450
    },
    {
      "epoch": 9.865229110512129,
      "grad_norm": 0.8412721157073975,
      "learning_rate": 8.315164597949271e-06,
      "loss": 3.1274,
      "step": 91500
    },
    {
      "epoch": 9.870619946091644,
      "grad_norm": 0.8239290118217468,
      "learning_rate": 7.99136535348084e-06,
      "loss": 3.1269,
      "step": 91550
    },
    {
      "epoch": 9.876010781671159,
      "grad_norm": 0.8275209665298462,
      "learning_rate": 7.667566109012411e-06,
      "loss": 3.1202,
      "step": 91600
    },
    {
      "epoch": 9.881401617250674,
      "grad_norm": 0.8666613101959229,
      "learning_rate": 7.343766864543982e-06,
      "loss": 3.1307,
      "step": 91650
    },
    {
      "epoch": 9.88679245283019,
      "grad_norm": 0.8908191323280334,
      "learning_rate": 7.019967620075552e-06,
      "loss": 3.1524,
      "step": 91700
    },
    {
      "epoch": 9.892183288409704,
      "grad_norm": 0.8692740797996521,
      "learning_rate": 6.702644360496492e-06,
      "loss": 3.1283,
      "step": 91750
    },
    {
      "epoch": 9.897574123989218,
      "grad_norm": 0.8492558002471924,
      "learning_rate": 6.378845116028062e-06,
      "loss": 3.1457,
      "step": 91800
    },
    {
      "epoch": 9.902964959568733,
      "grad_norm": 0.8426559567451477,
      "learning_rate": 6.055045871559633e-06,
      "loss": 3.1189,
      "step": 91850
    },
    {
      "epoch": 9.908355795148248,
      "grad_norm": 0.8520486354827881,
      "learning_rate": 5.731246627091203e-06,
      "loss": 3.1194,
      "step": 91900
    },
    {
      "epoch": 9.913746630727763,
      "grad_norm": 0.8753141164779663,
      "learning_rate": 5.407447382622774e-06,
      "loss": 3.1268,
      "step": 91950
    },
    {
      "epoch": 9.919137466307278,
      "grad_norm": 0.8538247346878052,
      "learning_rate": 5.083648138154344e-06,
      "loss": 3.1198,
      "step": 92000
    },
    {
      "epoch": 9.919137466307278,
      "eval_accuracy": 0.3948759950024018,
      "eval_loss": 3.298466920852661,
      "eval_runtime": 185.0411,
      "eval_samples_per_second": 97.335,
      "eval_steps_per_second": 6.085,
      "step": 92000
    },
    {
      "epoch": 9.924528301886792,
      "grad_norm": 0.8613267540931702,
      "learning_rate": 4.7598488936859145e-06,
      "loss": 3.1107,
      "step": 92050
    },
    {
      "epoch": 9.929919137466307,
      "grad_norm": 0.8902148604393005,
      "learning_rate": 4.4360496492174846e-06,
      "loss": 3.1319,
      "step": 92100
    },
    {
      "epoch": 9.935309973045822,
      "grad_norm": 0.9252253770828247,
      "learning_rate": 4.112250404749055e-06,
      "loss": 3.1135,
      "step": 92150
    },
    {
      "epoch": 9.940700808625337,
      "grad_norm": 0.8792473673820496,
      "learning_rate": 3.788451160280626e-06,
      "loss": 3.1454,
      "step": 92200
    },
    {
      "epoch": 9.946091644204852,
      "grad_norm": 0.8873490691184998,
      "learning_rate": 3.464651915812196e-06,
      "loss": 3.1286,
      "step": 92250
    },
    {
      "epoch": 9.951482479784367,
      "grad_norm": 0.8667479753494263,
      "learning_rate": 3.1408526713437664e-06,
      "loss": 3.1376,
      "step": 92300
    },
    {
      "epoch": 9.95687331536388,
      "grad_norm": 0.8077408075332642,
      "learning_rate": 2.817053426875337e-06,
      "loss": 3.1289,
      "step": 92350
    },
    {
      "epoch": 9.962264150943396,
      "grad_norm": 0.8791921734809875,
      "learning_rate": 2.4932541824069074e-06,
      "loss": 3.1286,
      "step": 92400
    },
    {
      "epoch": 9.967654986522911,
      "grad_norm": 0.9087523818016052,
      "learning_rate": 2.169454937938478e-06,
      "loss": 3.1201,
      "step": 92450
    },
    {
      "epoch": 9.973045822102426,
      "grad_norm": 0.828881561756134,
      "learning_rate": 1.8456556934700485e-06,
      "loss": 3.1279,
      "step": 92500
    },
    {
      "epoch": 9.978436657681941,
      "grad_norm": 0.8658260703086853,
      "learning_rate": 1.521856449001619e-06,
      "loss": 3.1078,
      "step": 92550
    },
    {
      "epoch": 9.983827493261456,
      "grad_norm": 0.8751649260520935,
      "learning_rate": 1.1980572045331894e-06,
      "loss": 3.1215,
      "step": 92600
    },
    {
      "epoch": 9.98921832884097,
      "grad_norm": 0.8642011284828186,
      "learning_rate": 8.742579600647598e-07,
      "loss": 3.1244,
      "step": 92650
    },
    {
      "epoch": 9.994609164420485,
      "grad_norm": 0.8841090798377991,
      "learning_rate": 5.504587155963303e-07,
      "loss": 3.1176,
      "step": 92700
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.7963483333587646,
      "learning_rate": 2.2665947112790068e-07,
      "loss": 3.1182,
      "step": 92750
    },
    {
      "epoch": 10.0,
      "step": 92750,
      "total_flos": 7.75449427968e+17,
      "train_loss": 3.455072832380022,
      "train_runtime": 79721.2665,
      "train_samples_per_second": 37.227,
      "train_steps_per_second": 1.163
    }
  ],
  "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
}