{
  "best_metric": 3.4408767223358154,
  "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M_low_1000_6910/checkpoint-40000",
  "epoch": 4.3126684636118595,
  "eval_steps": 1000,
  "global_step": 40000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.005390835579514825,
      "grad_norm": 4.540576457977295,
      "learning_rate": 0.000276,
      "loss": 9.0234,
      "step": 50
    },
    {
      "epoch": 0.01078167115902965,
      "grad_norm": 1.3573050498962402,
      "learning_rate": 0.0005759999999999999,
      "loss": 6.9351,
      "step": 100
    },
    {
      "epoch": 0.016172506738544475,
      "grad_norm": 1.575430989265442,
      "learning_rate": 0.000599702104695089,
      "loss": 6.533,
      "step": 150
    },
    {
      "epoch": 0.0215633423180593,
      "grad_norm": 1.919303297996521,
      "learning_rate": 0.0005993783054506205,
      "loss": 6.2466,
      "step": 200
    },
    {
      "epoch": 0.026954177897574125,
      "grad_norm": 1.5167046785354614,
      "learning_rate": 0.0005990545062061521,
      "loss": 6.0902,
      "step": 250
    },
    {
      "epoch": 0.03234501347708895,
      "grad_norm": 2.198723793029785,
      "learning_rate": 0.0005987307069616836,
      "loss": 5.9584,
      "step": 300
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 1.1930028200149536,
      "learning_rate": 0.0005984069077172153,
      "loss": 5.8824,
      "step": 350
    },
    {
      "epoch": 0.0431266846361186,
      "grad_norm": 1.5988584756851196,
      "learning_rate": 0.0005980831084727469,
      "loss": 5.8242,
      "step": 400
    },
    {
      "epoch": 0.04851752021563342,
      "grad_norm": 1.7518678903579712,
      "learning_rate": 0.0005977593092282784,
      "loss": 5.7575,
      "step": 450
    },
    {
      "epoch": 0.05390835579514825,
      "grad_norm": 1.542738914489746,
      "learning_rate": 0.00059743550998381,
      "loss": 5.65,
      "step": 500
    },
    {
      "epoch": 0.05929919137466307,
      "grad_norm": 1.1888893842697144,
      "learning_rate": 0.0005971117107393416,
      "loss": 5.6048,
      "step": 550
    },
    {
      "epoch": 0.0646900269541779,
      "grad_norm": 1.1427416801452637,
      "learning_rate": 0.0005967879114948732,
      "loss": 5.5294,
      "step": 600
    },
    {
      "epoch": 0.07008086253369272,
      "grad_norm": 1.5147260427474976,
      "learning_rate": 0.0005964641122504047,
      "loss": 5.4641,
      "step": 650
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 1.2342205047607422,
      "learning_rate": 0.0005961403130059363,
      "loss": 5.3932,
      "step": 700
    },
    {
      "epoch": 0.08086253369272237,
      "grad_norm": 0.8510802388191223,
      "learning_rate": 0.0005958165137614678,
      "loss": 5.3301,
      "step": 750
    },
    {
      "epoch": 0.0862533692722372,
      "grad_norm": 1.226401686668396,
      "learning_rate": 0.0005954927145169995,
      "loss": 5.2543,
      "step": 800
    },
    {
      "epoch": 0.09164420485175202,
      "grad_norm": 1.0827720165252686,
      "learning_rate": 0.0005951689152725309,
      "loss": 5.2325,
      "step": 850
    },
    {
      "epoch": 0.09703504043126684,
      "grad_norm": 1.071048378944397,
      "learning_rate": 0.0005948451160280626,
      "loss": 5.1791,
      "step": 900
    },
    {
      "epoch": 0.10242587601078167,
      "grad_norm": 0.8528303503990173,
      "learning_rate": 0.0005945213167835941,
      "loss": 5.1133,
      "step": 950
    },
    {
      "epoch": 0.1078167115902965,
      "grad_norm": 1.1848195791244507,
      "learning_rate": 0.0005941975175391257,
      "loss": 5.1184,
      "step": 1000
    },
    {
      "epoch": 0.1078167115902965,
      "eval_accuracy": 0.22659494561977292,
      "eval_loss": 5.030709743499756,
      "eval_runtime": 186.5313,
      "eval_samples_per_second": 96.558,
      "eval_steps_per_second": 6.037,
      "step": 1000
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 1.076619029045105,
      "learning_rate": 0.0005938737182946572,
      "loss": 5.052,
      "step": 1050
    },
    {
      "epoch": 0.11859838274932614,
      "grad_norm": 1.5022995471954346,
      "learning_rate": 0.0005935499190501888,
      "loss": 5.0099,
      "step": 1100
    },
    {
      "epoch": 0.12398921832884097,
      "grad_norm": 0.9214362502098083,
      "learning_rate": 0.0005932261198057204,
      "loss": 5.017,
      "step": 1150
    },
    {
      "epoch": 0.1293800539083558,
      "grad_norm": 1.1205987930297852,
      "learning_rate": 0.000592902320561252,
      "loss": 4.9733,
      "step": 1200
    },
    {
      "epoch": 0.1347708894878706,
      "grad_norm": 1.3065948486328125,
      "learning_rate": 0.0005925785213167835,
      "loss": 4.9494,
      "step": 1250
    },
    {
      "epoch": 0.14016172506738545,
      "grad_norm": 1.0144490003585815,
      "learning_rate": 0.0005922547220723151,
      "loss": 4.9026,
      "step": 1300
    },
    {
      "epoch": 0.14555256064690028,
      "grad_norm": 0.8630505204200745,
      "learning_rate": 0.0005919309228278468,
      "loss": 4.8527,
      "step": 1350
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 0.9217053651809692,
      "learning_rate": 0.0005916071235833783,
      "loss": 4.8597,
      "step": 1400
    },
    {
      "epoch": 0.15633423180592992,
      "grad_norm": 1.303423523902893,
      "learning_rate": 0.0005912833243389097,
      "loss": 4.816,
      "step": 1450
    },
    {
      "epoch": 0.16172506738544473,
      "grad_norm": 0.9101192951202393,
      "learning_rate": 0.0005909595250944414,
      "loss": 4.8184,
      "step": 1500
    },
    {
      "epoch": 0.16711590296495957,
      "grad_norm": 1.0290004014968872,
      "learning_rate": 0.000590635725849973,
      "loss": 4.7953,
      "step": 1550
    },
    {
      "epoch": 0.1725067385444744,
      "grad_norm": 0.9769102334976196,
      "learning_rate": 0.0005903119266055045,
      "loss": 4.7756,
      "step": 1600
    },
    {
      "epoch": 0.1778975741239892,
      "grad_norm": 1.380980134010315,
      "learning_rate": 0.0005899881273610361,
      "loss": 4.7469,
      "step": 1650
    },
    {
      "epoch": 0.18328840970350405,
      "grad_norm": 1.0679956674575806,
      "learning_rate": 0.0005896643281165677,
      "loss": 4.728,
      "step": 1700
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 0.9642156362533569,
      "learning_rate": 0.0005893405288720993,
      "loss": 4.6778,
      "step": 1750
    },
    {
      "epoch": 0.1940700808625337,
      "grad_norm": 0.9266220927238464,
      "learning_rate": 0.0005890167296276308,
      "loss": 4.7051,
      "step": 1800
    },
    {
      "epoch": 0.19946091644204852,
      "grad_norm": 1.5473623275756836,
      "learning_rate": 0.0005886929303831624,
      "loss": 4.6514,
      "step": 1850
    },
    {
      "epoch": 0.20485175202156333,
      "grad_norm": 0.8920683860778809,
      "learning_rate": 0.0005883691311386939,
      "loss": 4.6281,
      "step": 1900
    },
    {
      "epoch": 0.21024258760107817,
      "grad_norm": 0.6964493989944458,
      "learning_rate": 0.0005880453318942256,
      "loss": 4.6063,
      "step": 1950
    },
    {
      "epoch": 0.215633423180593,
      "grad_norm": 0.9296350479125977,
      "learning_rate": 0.0005877215326497571,
      "loss": 4.5882,
      "step": 2000
    },
    {
      "epoch": 0.215633423180593,
      "eval_accuracy": 0.2696204026654292,
      "eval_loss": 4.511911392211914,
      "eval_runtime": 186.0484,
      "eval_samples_per_second": 96.808,
      "eval_steps_per_second": 6.052,
      "step": 2000
    },
    {
      "epoch": 0.2210242587601078,
      "grad_norm": 0.8285171389579773,
      "learning_rate": 0.0005873977334052887,
      "loss": 4.5739,
      "step": 2050
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 1.025179386138916,
      "learning_rate": 0.0005870739341608202,
      "loss": 4.5558,
      "step": 2100
    },
    {
      "epoch": 0.23180592991913745,
      "grad_norm": 0.9890318512916565,
      "learning_rate": 0.0005867501349163519,
      "loss": 4.53,
      "step": 2150
    },
    {
      "epoch": 0.2371967654986523,
      "grad_norm": 1.0426075458526611,
      "learning_rate": 0.0005864263356718833,
      "loss": 4.5316,
      "step": 2200
    },
    {
      "epoch": 0.24258760107816713,
      "grad_norm": 0.7025043964385986,
      "learning_rate": 0.000586102536427415,
      "loss": 4.5041,
      "step": 2250
    },
    {
      "epoch": 0.24797843665768193,
      "grad_norm": 0.8847193717956543,
      "learning_rate": 0.0005857787371829465,
      "loss": 4.4971,
      "step": 2300
    },
    {
      "epoch": 0.25336927223719674,
      "grad_norm": 0.9764680862426758,
      "learning_rate": 0.0005854549379384781,
      "loss": 4.4769,
      "step": 2350
    },
    {
      "epoch": 0.2587601078167116,
      "grad_norm": 0.9064006805419922,
      "learning_rate": 0.0005851311386940096,
      "loss": 4.4604,
      "step": 2400
    },
    {
      "epoch": 0.2641509433962264,
      "grad_norm": 0.8132376074790955,
      "learning_rate": 0.0005848073394495412,
      "loss": 4.4506,
      "step": 2450
    },
    {
      "epoch": 0.2695417789757412,
      "grad_norm": 0.7548365592956543,
      "learning_rate": 0.0005844835402050728,
      "loss": 4.4258,
      "step": 2500
    },
    {
      "epoch": 0.2749326145552561,
      "grad_norm": 0.824798047542572,
      "learning_rate": 0.0005841597409606044,
      "loss": 4.3996,
      "step": 2550
    },
    {
      "epoch": 0.2803234501347709,
      "grad_norm": 0.9818605184555054,
      "learning_rate": 0.000583835941716136,
      "loss": 4.4099,
      "step": 2600
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.8206386566162109,
      "learning_rate": 0.0005835121424716675,
      "loss": 4.3904,
      "step": 2650
    },
    {
      "epoch": 0.29110512129380056,
      "grad_norm": 0.8314753770828247,
      "learning_rate": 0.0005831883432271992,
      "loss": 4.3802,
      "step": 2700
    },
    {
      "epoch": 0.29649595687331537,
      "grad_norm": 0.8359651565551758,
      "learning_rate": 0.0005828645439827307,
      "loss": 4.3767,
      "step": 2750
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 0.673989474773407,
      "learning_rate": 0.0005825407447382622,
      "loss": 4.3286,
      "step": 2800
    },
    {
      "epoch": 0.30727762803234504,
      "grad_norm": 0.7831907868385315,
      "learning_rate": 0.0005822169454937938,
      "loss": 4.3547,
      "step": 2850
    },
    {
      "epoch": 0.31266846361185985,
      "grad_norm": 0.8069139719009399,
      "learning_rate": 0.0005818931462493254,
      "loss": 4.3245,
      "step": 2900
    },
    {
      "epoch": 0.31805929919137466,
      "grad_norm": 0.8398146033287048,
      "learning_rate": 0.0005815693470048569,
      "loss": 4.3271,
      "step": 2950
    },
    {
      "epoch": 0.32345013477088946,
      "grad_norm": 0.7504966855049133,
      "learning_rate": 0.0005812455477603885,
      "loss": 4.3179,
      "step": 3000
    },
    {
      "epoch": 0.32345013477088946,
      "eval_accuracy": 0.29801085898691393,
      "eval_loss": 4.238383769989014,
      "eval_runtime": 185.9371,
      "eval_samples_per_second": 96.866,
      "eval_steps_per_second": 6.056,
      "step": 3000
    },
    {
      "epoch": 0.3288409703504043,
      "grad_norm": 0.7269638180732727,
      "learning_rate": 0.0005809217485159201,
      "loss": 4.3182,
      "step": 3050
    },
    {
      "epoch": 0.33423180592991913,
      "grad_norm": 0.826326847076416,
      "learning_rate": 0.0005805979492714517,
      "loss": 4.2988,
      "step": 3100
    },
    {
      "epoch": 0.33962264150943394,
      "grad_norm": 0.9023220539093018,
      "learning_rate": 0.0005802741500269832,
      "loss": 4.2764,
      "step": 3150
    },
    {
      "epoch": 0.3450134770889488,
      "grad_norm": 0.7048584222793579,
      "learning_rate": 0.0005799503507825148,
      "loss": 4.2794,
      "step": 3200
    },
    {
      "epoch": 0.3504043126684636,
      "grad_norm": 0.8540084362030029,
      "learning_rate": 0.0005796265515380463,
      "loss": 4.2674,
      "step": 3250
    },
    {
      "epoch": 0.3557951482479784,
      "grad_norm": 0.9496080875396729,
      "learning_rate": 0.000579302752293578,
      "loss": 4.276,
      "step": 3300
    },
    {
      "epoch": 0.3611859838274933,
      "grad_norm": 0.8834256529808044,
      "learning_rate": 0.0005789789530491095,
      "loss": 4.251,
      "step": 3350
    },
    {
      "epoch": 0.3665768194070081,
      "grad_norm": 0.8843632340431213,
      "learning_rate": 0.0005786551538046411,
      "loss": 4.2396,
      "step": 3400
    },
    {
      "epoch": 0.3719676549865229,
      "grad_norm": 0.6297584176063538,
      "learning_rate": 0.0005783313545601726,
      "loss": 4.2318,
      "step": 3450
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 0.8427825570106506,
      "learning_rate": 0.0005780075553157043,
      "loss": 4.2304,
      "step": 3500
    },
    {
      "epoch": 0.38274932614555257,
      "grad_norm": 0.6668996810913086,
      "learning_rate": 0.0005776837560712357,
      "loss": 4.2258,
      "step": 3550
    },
    {
      "epoch": 0.3881401617250674,
      "grad_norm": 0.7428986430168152,
      "learning_rate": 0.0005773599568267673,
      "loss": 4.2192,
      "step": 3600
    },
    {
      "epoch": 0.3935309973045822,
      "grad_norm": 0.7153974175453186,
      "learning_rate": 0.0005770361575822989,
      "loss": 4.205,
      "step": 3650
    },
    {
      "epoch": 0.39892183288409705,
      "grad_norm": 0.7248367667198181,
      "learning_rate": 0.0005767123583378305,
      "loss": 4.1946,
      "step": 3700
    },
    {
      "epoch": 0.40431266846361186,
      "grad_norm": 0.6670223474502563,
      "learning_rate": 0.000576388559093362,
      "loss": 4.1958,
      "step": 3750
    },
    {
      "epoch": 0.40970350404312667,
      "grad_norm": 0.6554595232009888,
      "learning_rate": 0.0005760647598488936,
      "loss": 4.1611,
      "step": 3800
    },
    {
      "epoch": 0.41509433962264153,
      "grad_norm": 0.6678674221038818,
      "learning_rate": 0.0005757409606044253,
      "loss": 4.1724,
      "step": 3850
    },
    {
      "epoch": 0.42048517520215634,
      "grad_norm": 0.7696819305419922,
      "learning_rate": 0.0005754171613599568,
      "loss": 4.1913,
      "step": 3900
    },
    {
      "epoch": 0.42587601078167114,
      "grad_norm": 1.3464781045913696,
      "learning_rate": 0.0005750933621154884,
      "loss": 4.1607,
      "step": 3950
    },
    {
      "epoch": 0.431266846361186,
      "grad_norm": 0.7259769439697266,
      "learning_rate": 0.0005747695628710199,
      "loss": 4.1615,
      "step": 4000
    },
    {
      "epoch": 0.431266846361186,
      "eval_accuracy": 0.3115824738980451,
      "eval_loss": 4.0961432456970215,
      "eval_runtime": 185.8978,
      "eval_samples_per_second": 96.887,
      "eval_steps_per_second": 6.057,
      "step": 4000
    },
    {
      "epoch": 0.4366576819407008,
      "grad_norm": 0.7618716359138489,
      "learning_rate": 0.0005744457636265515,
      "loss": 4.1695,
      "step": 4050
    },
    {
      "epoch": 0.4420485175202156,
      "grad_norm": 0.6841301321983337,
      "learning_rate": 0.0005741219643820831,
      "loss": 4.1634,
      "step": 4100
    },
    {
      "epoch": 0.4474393530997305,
      "grad_norm": 0.591198205947876,
      "learning_rate": 0.0005737981651376146,
      "loss": 4.1427,
      "step": 4150
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 0.6859827637672424,
      "learning_rate": 0.0005734743658931462,
      "loss": 4.1505,
      "step": 4200
    },
    {
      "epoch": 0.4582210242587601,
      "grad_norm": 0.6032926440238953,
      "learning_rate": 0.0005731505666486778,
      "loss": 4.138,
      "step": 4250
    },
    {
      "epoch": 0.4636118598382749,
      "grad_norm": 0.6650081872940063,
      "learning_rate": 0.0005728267674042093,
      "loss": 4.1412,
      "step": 4300
    },
    {
      "epoch": 0.46900269541778977,
      "grad_norm": 0.8347264528274536,
      "learning_rate": 0.0005725029681597409,
      "loss": 4.1275,
      "step": 4350
    },
    {
      "epoch": 0.4743935309973046,
      "grad_norm": 0.6658074855804443,
      "learning_rate": 0.0005721791689152725,
      "loss": 4.1125,
      "step": 4400
    },
    {
      "epoch": 0.4797843665768194,
      "grad_norm": 0.6057379245758057,
      "learning_rate": 0.0005718553696708041,
      "loss": 4.1205,
      "step": 4450
    },
    {
      "epoch": 0.48517520215633425,
      "grad_norm": 0.6311624050140381,
      "learning_rate": 0.0005715315704263356,
      "loss": 4.1149,
      "step": 4500
    },
    {
      "epoch": 0.49056603773584906,
      "grad_norm": 0.7309545278549194,
      "learning_rate": 0.0005712077711818672,
      "loss": 4.125,
      "step": 4550
    },
    {
      "epoch": 0.49595687331536387,
      "grad_norm": 0.8806862235069275,
      "learning_rate": 0.0005708839719373987,
      "loss": 4.1193,
      "step": 4600
    },
    {
      "epoch": 0.5013477088948787,
      "grad_norm": 0.7140979170799255,
      "learning_rate": 0.0005705601726929304,
      "loss": 4.075,
      "step": 4650
    },
    {
      "epoch": 0.5067385444743935,
      "grad_norm": 0.738926112651825,
      "learning_rate": 0.0005702363734484619,
      "loss": 4.0752,
      "step": 4700
    },
    {
      "epoch": 0.5121293800539084,
      "grad_norm": 0.5916194319725037,
      "learning_rate": 0.0005699125742039935,
      "loss": 4.083,
      "step": 4750
    },
    {
      "epoch": 0.5175202156334232,
      "grad_norm": 0.7595686912536621,
      "learning_rate": 0.000569588774959525,
      "loss": 4.0794,
      "step": 4800
    },
    {
      "epoch": 0.522911051212938,
      "grad_norm": 0.6311149597167969,
      "learning_rate": 0.0005692649757150567,
      "loss": 4.0744,
      "step": 4850
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 0.7534050941467285,
      "learning_rate": 0.0005689411764705881,
      "loss": 4.0754,
      "step": 4900
    },
    {
      "epoch": 0.5336927223719676,
      "grad_norm": 0.7457841038703918,
      "learning_rate": 0.0005686173772261197,
      "loss": 4.0669,
      "step": 4950
    },
    {
      "epoch": 0.5390835579514824,
      "grad_norm": 0.7841017842292786,
      "learning_rate": 0.0005682935779816514,
      "loss": 4.0664,
      "step": 5000
    },
    {
      "epoch": 0.5390835579514824,
      "eval_accuracy": 0.32084295952647335,
      "eval_loss": 3.998095750808716,
      "eval_runtime": 185.8924,
      "eval_samples_per_second": 96.889,
      "eval_steps_per_second": 6.057,
      "step": 5000
    },
    {
      "epoch": 0.5444743935309974,
      "grad_norm": 0.6891273856163025,
      "learning_rate": 0.0005679697787371829,
      "loss": 4.0733,
      "step": 5050
    },
    {
      "epoch": 0.5498652291105122,
      "grad_norm": 0.5636075735092163,
      "learning_rate": 0.0005676459794927145,
      "loss": 4.0588,
      "step": 5100
    },
    {
      "epoch": 0.555256064690027,
      "grad_norm": 0.7175034880638123,
      "learning_rate": 0.000567322180248246,
      "loss": 4.0517,
      "step": 5150
    },
    {
      "epoch": 0.5606469002695418,
      "grad_norm": 0.7910956144332886,
      "learning_rate": 0.0005669983810037777,
      "loss": 4.0616,
      "step": 5200
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 0.6039184331893921,
      "learning_rate": 0.0005666745817593092,
      "loss": 4.0371,
      "step": 5250
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 0.5606033205986023,
      "learning_rate": 0.0005663507825148408,
      "loss": 4.039,
      "step": 5300
    },
    {
      "epoch": 0.5768194070080862,
      "grad_norm": 0.7192058563232422,
      "learning_rate": 0.0005660269832703723,
      "loss": 4.0264,
      "step": 5350
    },
    {
      "epoch": 0.5822102425876011,
      "grad_norm": 0.596314549446106,
      "learning_rate": 0.0005657031840259039,
      "loss": 4.0295,
      "step": 5400
    },
    {
      "epoch": 0.5876010781671159,
      "grad_norm": 0.7095101475715637,
      "learning_rate": 0.0005653793847814355,
      "loss": 4.0314,
      "step": 5450
    },
    {
      "epoch": 0.5929919137466307,
      "grad_norm": 0.688493013381958,
      "learning_rate": 0.000565055585536967,
      "loss": 4.0088,
      "step": 5500
    },
    {
      "epoch": 0.5983827493261455,
      "grad_norm": 0.6665045619010925,
      "learning_rate": 0.0005647317862924986,
      "loss": 4.0332,
      "step": 5550
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 0.6965456008911133,
      "learning_rate": 0.0005644079870480302,
      "loss": 3.9809,
      "step": 5600
    },
    {
      "epoch": 0.6091644204851752,
      "grad_norm": 0.625882625579834,
      "learning_rate": 0.0005640841878035617,
      "loss": 4.0153,
      "step": 5650
    },
    {
      "epoch": 0.6145552560646901,
      "grad_norm": 0.6585850715637207,
      "learning_rate": 0.0005637603885590933,
      "loss": 4.0181,
      "step": 5700
    },
    {
      "epoch": 0.6199460916442049,
      "grad_norm": 0.6237135529518127,
      "learning_rate": 0.0005634365893146248,
      "loss": 4.0169,
      "step": 5750
    },
    {
      "epoch": 0.6253369272237197,
      "grad_norm": 0.6779605150222778,
      "learning_rate": 0.0005631127900701565,
      "loss": 4.0086,
      "step": 5800
    },
    {
      "epoch": 0.6307277628032345,
      "grad_norm": 0.6086890697479248,
      "learning_rate": 0.000562788990825688,
      "loss": 3.9814,
      "step": 5850
    },
    {
      "epoch": 0.6361185983827493,
      "grad_norm": 0.6210436820983887,
      "learning_rate": 0.0005624651915812196,
      "loss": 3.9972,
      "step": 5900
    },
    {
      "epoch": 0.6415094339622641,
      "grad_norm": 0.7014210820198059,
      "learning_rate": 0.0005621413923367511,
      "loss": 3.9997,
      "step": 5950
    },
    {
      "epoch": 0.6469002695417789,
      "grad_norm": 0.6019284725189209,
      "learning_rate": 0.0005618175930922828,
      "loss": 4.0053,
      "step": 6000
    },
    {
      "epoch": 0.6469002695417789,
      "eval_accuracy": 0.32752848036658616,
      "eval_loss": 3.9227378368377686,
      "eval_runtime": 185.8855,
      "eval_samples_per_second": 96.893,
      "eval_steps_per_second": 6.057,
      "step": 6000
    },
    {
      "epoch": 0.6522911051212938,
      "grad_norm": 0.6057129502296448,
      "learning_rate": 0.0005614937938478143,
      "loss": 3.9961,
      "step": 6050
    },
    {
      "epoch": 0.6576819407008087,
      "grad_norm": 0.6797792911529541,
      "learning_rate": 0.0005611699946033459,
      "loss": 3.9911,
      "step": 6100
    },
    {
      "epoch": 0.6630727762803235,
      "grad_norm": 0.5967230200767517,
      "learning_rate": 0.0005608461953588774,
      "loss": 3.9835,
      "step": 6150
    },
    {
      "epoch": 0.6684636118598383,
      "grad_norm": 0.6430239677429199,
      "learning_rate": 0.000560522396114409,
      "loss": 3.9776,
      "step": 6200
    },
    {
      "epoch": 0.6738544474393531,
      "grad_norm": 0.6402201056480408,
      "learning_rate": 0.0005601985968699405,
      "loss": 3.9578,
      "step": 6250
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 0.6837700009346008,
      "learning_rate": 0.0005598747976254721,
      "loss": 3.9743,
      "step": 6300
    },
    {
      "epoch": 0.6846361185983828,
      "grad_norm": 0.5868032574653625,
      "learning_rate": 0.0005595509983810038,
      "loss": 3.9487,
      "step": 6350
    },
    {
      "epoch": 0.6900269541778976,
      "grad_norm": 0.6631322503089905,
      "learning_rate": 0.0005592271991365353,
      "loss": 3.9597,
      "step": 6400
    },
    {
      "epoch": 0.6954177897574124,
      "grad_norm": 0.6614171862602234,
      "learning_rate": 0.0005589033998920669,
      "loss": 3.976,
      "step": 6450
    },
    {
      "epoch": 0.7008086253369272,
      "grad_norm": 0.6480840444564819,
      "learning_rate": 0.0005585796006475984,
      "loss": 3.9484,
      "step": 6500
    },
    {
      "epoch": 0.706199460916442,
      "grad_norm": 0.6667515635490417,
      "learning_rate": 0.0005582558014031301,
      "loss": 3.9724,
      "step": 6550
    },
    {
      "epoch": 0.7115902964959568,
      "grad_norm": 0.5956810116767883,
      "learning_rate": 0.0005579320021586616,
      "loss": 3.9531,
      "step": 6600
    },
    {
      "epoch": 0.7169811320754716,
      "grad_norm": 0.7182953357696533,
      "learning_rate": 0.0005576082029141932,
      "loss": 3.9483,
      "step": 6650
    },
    {
      "epoch": 0.7223719676549866,
      "grad_norm": 0.540218710899353,
      "learning_rate": 0.0005572844036697247,
      "loss": 3.9354,
      "step": 6700
    },
    {
      "epoch": 0.7277628032345014,
      "grad_norm": 0.5549564957618713,
      "learning_rate": 0.0005569606044252563,
      "loss": 3.9416,
      "step": 6750
    },
    {
      "epoch": 0.7331536388140162,
      "grad_norm": 0.955643355846405,
      "learning_rate": 0.0005566368051807879,
      "loss": 3.9594,
      "step": 6800
    },
    {
      "epoch": 0.738544474393531,
      "grad_norm": 0.7693118453025818,
      "learning_rate": 0.0005563130059363194,
      "loss": 3.945,
      "step": 6850
    },
    {
      "epoch": 0.7439353099730458,
      "grad_norm": 0.587993860244751,
      "learning_rate": 0.000555989206691851,
      "loss": 3.9435,
      "step": 6900
    },
    {
      "epoch": 0.7493261455525606,
      "grad_norm": 0.5942090749740601,
      "learning_rate": 0.0005556654074473826,
      "loss": 3.9466,
      "step": 6950
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 0.6959335207939148,
      "learning_rate": 0.0005553416082029141,
      "loss": 3.9311,
      "step": 7000
    },
    {
      "epoch": 0.7547169811320755,
      "eval_accuracy": 0.3327442535932325,
      "eval_loss": 3.868434190750122,
      "eval_runtime": 185.8133,
      "eval_samples_per_second": 96.931,
      "eval_steps_per_second": 6.06,
      "step": 7000
    },
    {
      "epoch": 0.7601078167115903,
      "grad_norm": 0.6941900253295898,
      "learning_rate": 0.0005550178089584457,
      "loss": 3.9224,
      "step": 7050
    },
    {
      "epoch": 0.7654986522911051,
      "grad_norm": 0.5540989637374878,
      "learning_rate": 0.0005546940097139772,
      "loss": 3.9235,
      "step": 7100
    },
    {
      "epoch": 0.77088948787062,
      "grad_norm": 0.6097570061683655,
      "learning_rate": 0.0005543702104695089,
      "loss": 3.9264,
      "step": 7150
    },
    {
      "epoch": 0.7762803234501348,
      "grad_norm": 0.5658494830131531,
      "learning_rate": 0.0005540464112250404,
      "loss": 3.9229,
      "step": 7200
    },
    {
      "epoch": 0.7816711590296496,
      "grad_norm": 0.6700815558433533,
      "learning_rate": 0.000553722611980572,
      "loss": 3.9357,
      "step": 7250
    },
    {
      "epoch": 0.7870619946091644,
      "grad_norm": 0.5885155200958252,
      "learning_rate": 0.0005533988127361035,
      "loss": 3.9183,
      "step": 7300
    },
    {
      "epoch": 0.7924528301886793,
      "grad_norm": 0.5735599398612976,
      "learning_rate": 0.0005530750134916352,
      "loss": 3.9369,
      "step": 7350
    },
    {
      "epoch": 0.7978436657681941,
      "grad_norm": 0.6089330911636353,
      "learning_rate": 0.0005527512142471668,
      "loss": 3.9099,
      "step": 7400
    },
    {
      "epoch": 0.8032345013477089,
      "grad_norm": 0.6232162714004517,
      "learning_rate": 0.0005524274150026982,
      "loss": 3.9112,
      "step": 7450
    },
    {
      "epoch": 0.8086253369272237,
      "grad_norm": 0.606532096862793,
      "learning_rate": 0.0005521036157582299,
      "loss": 3.9004,
      "step": 7500
    },
    {
      "epoch": 0.8140161725067385,
      "grad_norm": 0.7327796816825867,
      "learning_rate": 0.0005517798165137614,
      "loss": 3.9057,
      "step": 7550
    },
    {
      "epoch": 0.8194070080862533,
      "grad_norm": 0.5931469202041626,
      "learning_rate": 0.000551456017269293,
      "loss": 3.881,
      "step": 7600
    },
    {
      "epoch": 0.8247978436657682,
      "grad_norm": 0.6132405400276184,
      "learning_rate": 0.0005511322180248245,
      "loss": 3.8988,
      "step": 7650
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 0.5921282172203064,
      "learning_rate": 0.0005508084187803562,
      "loss": 3.9097,
      "step": 7700
    },
    {
      "epoch": 0.8355795148247979,
      "grad_norm": 0.6329267024993896,
      "learning_rate": 0.0005504846195358877,
      "loss": 3.9026,
      "step": 7750
    },
    {
      "epoch": 0.8409703504043127,
      "grad_norm": 0.6082526445388794,
      "learning_rate": 0.0005501608202914193,
      "loss": 3.8919,
      "step": 7800
    },
    {
      "epoch": 0.8463611859838275,
      "grad_norm": 0.5989628434181213,
      "learning_rate": 0.0005498370210469508,
      "loss": 3.895,
      "step": 7850
    },
    {
      "epoch": 0.8517520215633423,
      "grad_norm": 0.592002809047699,
      "learning_rate": 0.0005495132218024824,
      "loss": 3.88,
      "step": 7900
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.5714187026023865,
      "learning_rate": 0.000549189422558014,
      "loss": 3.8903,
      "step": 7950
    },
    {
      "epoch": 0.862533692722372,
      "grad_norm": 0.6176041960716248,
      "learning_rate": 0.0005488656233135456,
      "loss": 3.8888,
      "step": 8000
    },
    {
      "epoch": 0.862533692722372,
      "eval_accuracy": 0.3374015509765124,
      "eval_loss": 3.820584535598755,
      "eval_runtime": 185.9631,
      "eval_samples_per_second": 96.853,
      "eval_steps_per_second": 6.055,
      "step": 8000
    },
    {
      "epoch": 0.8679245283018868,
      "grad_norm": 0.6291443109512329,
      "learning_rate": 0.0005485418240690771,
      "loss": 3.8699,
      "step": 8050
    },
    {
      "epoch": 0.8733153638814016,
      "grad_norm": 0.6424714922904968,
      "learning_rate": 0.0005482180248246087,
      "loss": 3.8858,
      "step": 8100
    },
    {
      "epoch": 0.8787061994609164,
      "grad_norm": 0.6014903783798218,
      "learning_rate": 0.0005478942255801403,
      "loss": 3.8827,
      "step": 8150
    },
    {
      "epoch": 0.8840970350404312,
      "grad_norm": 0.5504648089408875,
      "learning_rate": 0.0005475704263356718,
      "loss": 3.8694,
      "step": 8200
    },
    {
      "epoch": 0.889487870619946,
      "grad_norm": 0.5945785045623779,
      "learning_rate": 0.0005472466270912034,
      "loss": 3.8799,
      "step": 8250
    },
    {
      "epoch": 0.894878706199461,
      "grad_norm": 0.5891640186309814,
      "learning_rate": 0.000546922827846735,
      "loss": 3.8689,
      "step": 8300
    },
    {
      "epoch": 0.9002695417789758,
      "grad_norm": 0.5595161318778992,
      "learning_rate": 0.0005465990286022665,
      "loss": 3.8689,
      "step": 8350
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 0.5921164155006409,
      "learning_rate": 0.0005462752293577981,
      "loss": 3.8853,
      "step": 8400
    },
    {
      "epoch": 0.9110512129380054,
      "grad_norm": 0.5961281657218933,
      "learning_rate": 0.0005459514301133296,
      "loss": 3.8789,
      "step": 8450
    },
    {
      "epoch": 0.9164420485175202,
      "grad_norm": 0.5500056147575378,
      "learning_rate": 0.0005456276308688613,
      "loss": 3.875,
      "step": 8500
    },
    {
      "epoch": 0.921832884097035,
      "grad_norm": 0.5573350787162781,
      "learning_rate": 0.0005453038316243929,
      "loss": 3.868,
      "step": 8550
    },
    {
      "epoch": 0.9272237196765498,
      "grad_norm": 0.5513932108879089,
      "learning_rate": 0.0005449800323799244,
      "loss": 3.8786,
      "step": 8600
    },
    {
      "epoch": 0.9326145552560647,
      "grad_norm": 0.5722886323928833,
      "learning_rate": 0.000544656233135456,
      "loss": 3.8585,
      "step": 8650
    },
    {
      "epoch": 0.9380053908355795,
      "grad_norm": 0.6074110269546509,
      "learning_rate": 0.0005443324338909875,
      "loss": 3.8613,
      "step": 8700
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 0.5840967893600464,
      "learning_rate": 0.0005440086346465192,
      "loss": 3.8728,
      "step": 8750
    },
    {
      "epoch": 0.9487870619946092,
      "grad_norm": 0.580897867679596,
      "learning_rate": 0.0005436848354020506,
      "loss": 3.8535,
      "step": 8800
    },
    {
      "epoch": 0.954177897574124,
      "grad_norm": 0.5885988473892212,
      "learning_rate": 0.0005433610361575823,
      "loss": 3.8595,
      "step": 8850
    },
    {
      "epoch": 0.9595687331536388,
      "grad_norm": 0.568071186542511,
      "learning_rate": 0.0005430372369131138,
      "loss": 3.8457,
      "step": 8900
    },
    {
      "epoch": 0.9649595687331537,
      "grad_norm": 0.612762987613678,
      "learning_rate": 0.0005427134376686454,
      "loss": 3.857,
      "step": 8950
    },
    {
      "epoch": 0.9703504043126685,
      "grad_norm": 0.6389222145080566,
      "learning_rate": 0.0005423896384241769,
      "loss": 3.8417,
      "step": 9000
    },
    {
      "epoch": 0.9703504043126685,
      "eval_accuracy": 0.3407095967989121,
      "eval_loss": 3.7826406955718994,
      "eval_runtime": 185.8291,
      "eval_samples_per_second": 96.922,
      "eval_steps_per_second": 6.059,
      "step": 9000
    },
    {
      "epoch": 0.9757412398921833,
      "grad_norm": 0.6433724761009216,
      "learning_rate": 0.0005420658391797086,
      "loss": 3.8386,
      "step": 9050
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 0.616686999797821,
      "learning_rate": 0.0005417420399352401,
      "loss": 3.8486,
      "step": 9100
    },
    {
      "epoch": 0.9865229110512129,
      "grad_norm": 0.6407685875892639,
      "learning_rate": 0.0005414182406907717,
      "loss": 3.8421,
      "step": 9150
    },
    {
      "epoch": 0.9919137466307277,
      "grad_norm": 0.731094479560852,
      "learning_rate": 0.0005410944414463032,
      "loss": 3.8593,
      "step": 9200
    },
    {
      "epoch": 0.9973045822102425,
      "grad_norm": 0.617189884185791,
      "learning_rate": 0.0005407706422018348,
      "loss": 3.8551,
      "step": 9250
    },
    {
      "epoch": 1.0026954177897573,
      "grad_norm": 0.5850970149040222,
      "learning_rate": 0.0005404468429573664,
      "loss": 3.8018,
      "step": 9300
    },
    {
      "epoch": 1.0080862533692723,
      "grad_norm": 0.6619550585746765,
      "learning_rate": 0.000540123043712898,
      "loss": 3.7742,
      "step": 9350
    },
    {
      "epoch": 1.013477088948787,
      "grad_norm": 0.5740201473236084,
      "learning_rate": 0.0005397992444684295,
      "loss": 3.7547,
      "step": 9400
    },
    {
      "epoch": 1.0188679245283019,
      "grad_norm": 0.6828320622444153,
      "learning_rate": 0.0005394754452239611,
      "loss": 3.7691,
      "step": 9450
    },
    {
      "epoch": 1.0242587601078168,
      "grad_norm": 0.5637569427490234,
      "learning_rate": 0.0005391516459794927,
      "loss": 3.7799,
      "step": 9500
    },
    {
      "epoch": 1.0296495956873315,
      "grad_norm": 0.5780271887779236,
      "learning_rate": 0.0005388278467350242,
      "loss": 3.7877,
      "step": 9550
    },
    {
      "epoch": 1.0350404312668464,
      "grad_norm": 0.5565044283866882,
      "learning_rate": 0.0005385040474905557,
      "loss": 3.7821,
      "step": 9600
    },
    {
      "epoch": 1.0404312668463611,
      "grad_norm": 0.6038862466812134,
      "learning_rate": 0.0005381802482460874,
      "loss": 3.7916,
      "step": 9650
    },
    {
      "epoch": 1.045822102425876,
      "grad_norm": 0.5745878219604492,
      "learning_rate": 0.000537856449001619,
      "loss": 3.7901,
      "step": 9700
    },
    {
      "epoch": 1.0512129380053907,
      "grad_norm": 0.5736984014511108,
      "learning_rate": 0.0005375326497571505,
      "loss": 3.763,
      "step": 9750
    },
    {
      "epoch": 1.0566037735849056,
      "grad_norm": 0.6319218277931213,
      "learning_rate": 0.000537208850512682,
      "loss": 3.7697,
      "step": 9800
    },
    {
      "epoch": 1.0619946091644206,
      "grad_norm": 0.5360982418060303,
      "learning_rate": 0.0005368850512682137,
      "loss": 3.7678,
      "step": 9850
    },
    {
      "epoch": 1.0673854447439353,
      "grad_norm": 0.5352616906166077,
      "learning_rate": 0.0005365612520237453,
      "loss": 3.7801,
      "step": 9900
    },
    {
      "epoch": 1.0727762803234502,
      "grad_norm": 0.6189696788787842,
      "learning_rate": 0.0005362374527792768,
      "loss": 3.7794,
      "step": 9950
    },
    {
      "epoch": 1.0781671159029649,
      "grad_norm": 0.6562055945396423,
      "learning_rate": 0.0005359136535348084,
      "loss": 3.7812,
      "step": 10000
    },
    {
      "epoch": 1.0781671159029649,
      "eval_accuracy": 0.34422147543885173,
      "eval_loss": 3.759232521057129,
      "eval_runtime": 185.6272,
      "eval_samples_per_second": 97.028,
      "eval_steps_per_second": 6.066,
      "step": 10000
    },
    {
      "epoch": 1.0835579514824798,
      "grad_norm": 0.5553216338157654,
      "learning_rate": 0.0005355898542903399,
      "loss": 3.7574,
      "step": 10050
    },
    {
      "epoch": 1.0889487870619945,
      "grad_norm": 0.603709876537323,
      "learning_rate": 0.0005352660550458716,
      "loss": 3.7606,
      "step": 10100
    },
    {
      "epoch": 1.0943396226415094,
      "grad_norm": 0.6040270924568176,
      "learning_rate": 0.000534942255801403,
      "loss": 3.7491,
      "step": 10150
    },
    {
      "epoch": 1.0997304582210243,
      "grad_norm": 0.6243166327476501,
      "learning_rate": 0.0005346184565569347,
      "loss": 3.7686,
      "step": 10200
    },
    {
      "epoch": 1.105121293800539,
      "grad_norm": 0.5870617628097534,
      "learning_rate": 0.0005342946573124662,
      "loss": 3.7855,
      "step": 10250
    },
    {
      "epoch": 1.110512129380054,
      "grad_norm": 0.6204577088356018,
      "learning_rate": 0.0005339708580679978,
      "loss": 3.786,
      "step": 10300
    },
    {
      "epoch": 1.1159029649595686,
      "grad_norm": 0.5694855451583862,
      "learning_rate": 0.0005336470588235293,
      "loss": 3.7614,
      "step": 10350
    },
    {
      "epoch": 1.1212938005390836,
      "grad_norm": 0.5424656271934509,
      "learning_rate": 0.000533323259579061,
      "loss": 3.7652,
      "step": 10400
    },
    {
      "epoch": 1.1266846361185983,
      "grad_norm": 0.5861342549324036,
      "learning_rate": 0.0005329994603345925,
      "loss": 3.7618,
      "step": 10450
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 0.5731029510498047,
      "learning_rate": 0.0005326756610901241,
      "loss": 3.7604,
      "step": 10500
    },
    {
      "epoch": 1.137466307277628,
      "grad_norm": 0.5382975339889526,
      "learning_rate": 0.0005323518618456556,
      "loss": 3.7614,
      "step": 10550
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 0.6149365901947021,
      "learning_rate": 0.0005320280626011872,
      "loss": 3.7718,
      "step": 10600
    },
    {
      "epoch": 1.1482479784366577,
      "grad_norm": 0.5981982350349426,
      "learning_rate": 0.0005317042633567188,
      "loss": 3.7589,
      "step": 10650
    },
    {
      "epoch": 1.1536388140161726,
      "grad_norm": 0.644930362701416,
      "learning_rate": 0.0005313804641122504,
      "loss": 3.7657,
      "step": 10700
    },
    {
      "epoch": 1.1590296495956873,
      "grad_norm": 0.5735489130020142,
      "learning_rate": 0.0005310566648677819,
      "loss": 3.7608,
      "step": 10750
    },
    {
      "epoch": 1.1644204851752022,
      "grad_norm": 0.6112673282623291,
      "learning_rate": 0.0005307328656233135,
      "loss": 3.7487,
      "step": 10800
    },
    {
      "epoch": 1.169811320754717,
      "grad_norm": 0.5764835476875305,
      "learning_rate": 0.000530409066378845,
      "loss": 3.7451,
      "step": 10850
    },
    {
      "epoch": 1.1752021563342319,
      "grad_norm": 1.4614336490631104,
      "learning_rate": 0.0005300852671343766,
      "loss": 3.7583,
      "step": 10900
    },
    {
      "epoch": 1.1805929919137466,
      "grad_norm": 0.6848381161689758,
      "learning_rate": 0.0005297614678899081,
      "loss": 3.754,
      "step": 10950
    },
    {
      "epoch": 1.1859838274932615,
      "grad_norm": 0.5512279272079468,
      "learning_rate": 0.0005294376686454398,
      "loss": 3.7435,
      "step": 11000
    },
    {
      "epoch": 1.1859838274932615,
      "eval_accuracy": 0.3471423910219684,
      "eval_loss": 3.725170612335205,
      "eval_runtime": 185.7111,
      "eval_samples_per_second": 96.984,
      "eval_steps_per_second": 6.063,
      "step": 11000
    },
    {
      "epoch": 1.1913746630727764,
      "grad_norm": 0.5805233120918274,
      "learning_rate": 0.0005291138694009714,
      "loss": 3.7513,
      "step": 11050
    },
    {
      "epoch": 1.196765498652291,
      "grad_norm": 0.6513531804084778,
      "learning_rate": 0.0005287900701565029,
      "loss": 3.7442,
      "step": 11100
    },
    {
      "epoch": 1.202156334231806,
      "grad_norm": 0.6177436113357544,
      "learning_rate": 0.0005284662709120345,
      "loss": 3.7456,
      "step": 11150
    },
    {
      "epoch": 1.2075471698113207,
      "grad_norm": 0.6252652406692505,
      "learning_rate": 0.0005281424716675661,
      "loss": 3.7444,
      "step": 11200
    },
    {
      "epoch": 1.2129380053908356,
      "grad_norm": 0.6261422634124756,
      "learning_rate": 0.0005278186724230977,
      "loss": 3.7377,
      "step": 11250
    },
    {
      "epoch": 1.2183288409703503,
      "grad_norm": 0.5394766330718994,
      "learning_rate": 0.0005274948731786292,
      "loss": 3.7344,
      "step": 11300
    },
    {
      "epoch": 1.2237196765498652,
      "grad_norm": 0.6933181285858154,
      "learning_rate": 0.0005271710739341608,
      "loss": 3.7523,
      "step": 11350
    },
    {
      "epoch": 1.2291105121293802,
      "grad_norm": 0.6169976592063904,
      "learning_rate": 0.0005268472746896923,
      "loss": 3.7456,
      "step": 11400
    },
    {
      "epoch": 1.2345013477088949,
      "grad_norm": 0.6621821522712708,
      "learning_rate": 0.000526523475445224,
      "loss": 3.7365,
      "step": 11450
    },
    {
      "epoch": 1.2398921832884098,
      "grad_norm": 0.5416802763938904,
      "learning_rate": 0.0005261996762007554,
      "loss": 3.7232,
      "step": 11500
    },
    {
      "epoch": 1.2452830188679245,
      "grad_norm": 0.5890310406684875,
      "learning_rate": 0.0005258758769562871,
      "loss": 3.7459,
      "step": 11550
    },
    {
      "epoch": 1.2506738544474394,
      "grad_norm": 0.6583136916160583,
      "learning_rate": 0.0005255520777118186,
      "loss": 3.7463,
      "step": 11600
    },
    {
      "epoch": 1.256064690026954,
      "grad_norm": 0.531416118144989,
      "learning_rate": 0.0005252282784673502,
      "loss": 3.7356,
      "step": 11650
    },
    {
      "epoch": 1.261455525606469,
      "grad_norm": 0.562391996383667,
      "learning_rate": 0.0005249044792228817,
      "loss": 3.7385,
      "step": 11700
    },
    {
      "epoch": 1.266846361185984,
      "grad_norm": 0.5567585229873657,
      "learning_rate": 0.0005245806799784133,
      "loss": 3.7339,
      "step": 11750
    },
    {
      "epoch": 1.2722371967654986,
      "grad_norm": 0.6060622930526733,
      "learning_rate": 0.0005242568807339449,
      "loss": 3.73,
      "step": 11800
    },
    {
      "epoch": 1.2776280323450135,
      "grad_norm": 0.6515873670578003,
      "learning_rate": 0.0005239330814894765,
      "loss": 3.7276,
      "step": 11850
    },
    {
      "epoch": 1.2830188679245282,
      "grad_norm": 0.5297693610191345,
      "learning_rate": 0.000523609282245008,
      "loss": 3.7483,
      "step": 11900
    },
    {
      "epoch": 1.2884097035040432,
      "grad_norm": 0.6324127316474915,
      "learning_rate": 0.0005232854830005396,
      "loss": 3.7445,
      "step": 11950
    },
    {
      "epoch": 1.2938005390835579,
      "grad_norm": 0.5799328088760376,
      "learning_rate": 0.0005229616837560712,
      "loss": 3.7391,
      "step": 12000
    },
    {
      "epoch": 1.2938005390835579,
      "eval_accuracy": 0.3490961872506484,
      "eval_loss": 3.7054829597473145,
      "eval_runtime": 185.967,
      "eval_samples_per_second": 96.851,
      "eval_steps_per_second": 6.055,
      "step": 12000
    },
    {
      "epoch": 1.2991913746630728,
      "grad_norm": 0.5518076419830322,
      "learning_rate": 0.0005226378845116028,
      "loss": 3.7192,
      "step": 12050
    },
    {
      "epoch": 1.3045822102425877,
      "grad_norm": 0.6213856339454651,
      "learning_rate": 0.0005223140852671344,
      "loss": 3.7319,
      "step": 12100
    },
    {
      "epoch": 1.3099730458221024,
      "grad_norm": 0.7218828201293945,
      "learning_rate": 0.0005219902860226659,
      "loss": 3.7434,
      "step": 12150
    },
    {
      "epoch": 1.3153638814016173,
      "grad_norm": 0.563077986240387,
      "learning_rate": 0.0005216664867781975,
      "loss": 3.7417,
      "step": 12200
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.6648295521736145,
      "learning_rate": 0.000521342687533729,
      "loss": 3.7238,
      "step": 12250
    },
    {
      "epoch": 1.326145552560647,
      "grad_norm": 0.5917858481407166,
      "learning_rate": 0.0005210188882892606,
      "loss": 3.7323,
      "step": 12300
    },
    {
      "epoch": 1.3315363881401616,
      "grad_norm": 0.5561532378196716,
      "learning_rate": 0.0005206950890447922,
      "loss": 3.7224,
      "step": 12350
    },
    {
      "epoch": 1.3369272237196765,
      "grad_norm": 0.5533449053764343,
      "learning_rate": 0.0005203712898003238,
      "loss": 3.7372,
      "step": 12400
    },
    {
      "epoch": 1.3423180592991915,
      "grad_norm": 0.6743930578231812,
      "learning_rate": 0.0005200474905558553,
      "loss": 3.7268,
      "step": 12450
    },
    {
      "epoch": 1.3477088948787062,
      "grad_norm": 0.6915057897567749,
      "learning_rate": 0.0005197236913113869,
      "loss": 3.7267,
      "step": 12500
    },
    {
      "epoch": 1.353099730458221,
      "grad_norm": 0.5675690174102783,
      "learning_rate": 0.0005193998920669184,
      "loss": 3.7359,
      "step": 12550
    },
    {
      "epoch": 1.3584905660377358,
      "grad_norm": 0.5956441164016724,
      "learning_rate": 0.0005190760928224501,
      "loss": 3.718,
      "step": 12600
    },
    {
      "epoch": 1.3638814016172507,
      "grad_norm": 0.5554563403129578,
      "learning_rate": 0.0005187522935779816,
      "loss": 3.7277,
      "step": 12650
    },
    {
      "epoch": 1.3692722371967654,
      "grad_norm": 0.5844613909721375,
      "learning_rate": 0.0005184284943335132,
      "loss": 3.7423,
      "step": 12700
    },
    {
      "epoch": 1.3746630727762803,
      "grad_norm": 0.5583260655403137,
      "learning_rate": 0.0005181046950890447,
      "loss": 3.7263,
      "step": 12750
    },
    {
      "epoch": 1.3800539083557952,
      "grad_norm": 0.5783796310424805,
      "learning_rate": 0.0005177808958445764,
      "loss": 3.7188,
      "step": 12800
    },
    {
      "epoch": 1.38544474393531,
      "grad_norm": 0.5987197160720825,
      "learning_rate": 0.0005174570966001078,
      "loss": 3.7124,
      "step": 12850
    },
    {
      "epoch": 1.3908355795148248,
      "grad_norm": 0.5812613368034363,
      "learning_rate": 0.0005171332973556395,
      "loss": 3.7193,
      "step": 12900
    },
    {
      "epoch": 1.3962264150943398,
      "grad_norm": 0.5962573885917664,
      "learning_rate": 0.000516809498111171,
      "loss": 3.7173,
      "step": 12950
    },
    {
      "epoch": 1.4016172506738545,
      "grad_norm": 0.5796558260917664,
      "learning_rate": 0.0005164856988667026,
      "loss": 3.7123,
      "step": 13000
    },
    {
      "epoch": 1.4016172506738545,
      "eval_accuracy": 0.35139039297685115,
      "eval_loss": 3.679816722869873,
      "eval_runtime": 185.9335,
      "eval_samples_per_second": 96.868,
      "eval_steps_per_second": 6.056,
      "step": 13000
    },
    {
      "epoch": 1.4070080862533692,
      "grad_norm": 0.6270114779472351,
      "learning_rate": 0.0005161618996222341,
      "loss": 3.7239,
      "step": 13050
    },
    {
      "epoch": 1.412398921832884,
      "grad_norm": 0.5918281674385071,
      "learning_rate": 0.0005158381003777657,
      "loss": 3.7342,
      "step": 13100
    },
    {
      "epoch": 1.417789757412399,
      "grad_norm": 0.6168618202209473,
      "learning_rate": 0.0005155143011332973,
      "loss": 3.7145,
      "step": 13150
    },
    {
      "epoch": 1.4231805929919137,
      "grad_norm": 0.5616515278816223,
      "learning_rate": 0.0005151905018888289,
      "loss": 3.7179,
      "step": 13200
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.5521336793899536,
      "learning_rate": 0.0005148667026443604,
      "loss": 3.7139,
      "step": 13250
    },
    {
      "epoch": 1.4339622641509435,
      "grad_norm": 0.5330696105957031,
      "learning_rate": 0.000514542903399892,
      "loss": 3.7103,
      "step": 13300
    },
    {
      "epoch": 1.4393530997304582,
      "grad_norm": 0.6266703009605408,
      "learning_rate": 0.0005142191041554237,
      "loss": 3.6965,
      "step": 13350
    },
    {
      "epoch": 1.444743935309973,
      "grad_norm": 0.620352566242218,
      "learning_rate": 0.0005138953049109552,
      "loss": 3.7071,
      "step": 13400
    },
    {
      "epoch": 1.4501347708894878,
      "grad_norm": 0.6346424221992493,
      "learning_rate": 0.0005135715056664868,
      "loss": 3.7169,
      "step": 13450
    },
    {
      "epoch": 1.4555256064690028,
      "grad_norm": 0.5509406328201294,
      "learning_rate": 0.0005132477064220183,
      "loss": 3.7026,
      "step": 13500
    },
    {
      "epoch": 1.4609164420485174,
      "grad_norm": 0.6062074303627014,
      "learning_rate": 0.0005129239071775499,
      "loss": 3.7131,
      "step": 13550
    },
    {
      "epoch": 1.4663072776280324,
      "grad_norm": 0.617310643196106,
      "learning_rate": 0.0005126001079330814,
      "loss": 3.7147,
      "step": 13600
    },
    {
      "epoch": 1.4716981132075473,
      "grad_norm": 0.6272128224372864,
      "learning_rate": 0.000512276308688613,
      "loss": 3.6887,
      "step": 13650
    },
    {
      "epoch": 1.477088948787062,
      "grad_norm": 0.5724652409553528,
      "learning_rate": 0.0005119525094441446,
      "loss": 3.6875,
      "step": 13700
    },
    {
      "epoch": 1.482479784366577,
      "grad_norm": 0.5527543425559998,
      "learning_rate": 0.0005116287101996762,
      "loss": 3.7039,
      "step": 13750
    },
    {
      "epoch": 1.4878706199460916,
      "grad_norm": 0.6792011857032776,
      "learning_rate": 0.0005113049109552077,
      "loss": 3.7191,
      "step": 13800
    },
    {
      "epoch": 1.4932614555256065,
      "grad_norm": 0.6160116791725159,
      "learning_rate": 0.0005109811117107393,
      "loss": 3.7143,
      "step": 13850
    },
    {
      "epoch": 1.4986522911051212,
      "grad_norm": 0.5848609209060669,
      "learning_rate": 0.0005106573124662708,
      "loss": 3.728,
      "step": 13900
    },
    {
      "epoch": 1.5040431266846361,
      "grad_norm": 0.5350325107574463,
      "learning_rate": 0.0005103335132218025,
      "loss": 3.7018,
      "step": 13950
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 0.5794714689254761,
      "learning_rate": 0.000510009713977334,
      "loss": 3.7017,
      "step": 14000
    },
    {
      "epoch": 1.509433962264151,
      "eval_accuracy": 0.35348728505878285,
      "eval_loss": 3.6589605808258057,
      "eval_runtime": 186.0,
      "eval_samples_per_second": 96.833,
      "eval_steps_per_second": 6.054,
      "step": 14000
    },
    {
      "epoch": 1.5148247978436657,
      "grad_norm": 0.5505036115646362,
      "learning_rate": 0.0005096859147328656,
      "loss": 3.6933,
      "step": 14050
    },
    {
      "epoch": 1.5202156334231804,
      "grad_norm": 0.7258388996124268,
      "learning_rate": 0.0005093685914732865,
      "loss": 3.6907,
      "step": 14100
    },
    {
      "epoch": 1.5256064690026954,
      "grad_norm": 0.5728726387023926,
      "learning_rate": 0.0005090447922288181,
      "loss": 3.6884,
      "step": 14150
    },
    {
      "epoch": 1.5309973045822103,
      "grad_norm": 0.5695355534553528,
      "learning_rate": 0.0005087209929843496,
      "loss": 3.7208,
      "step": 14200
    },
    {
      "epoch": 1.536388140161725,
      "grad_norm": 0.5693349242210388,
      "learning_rate": 0.0005083971937398812,
      "loss": 3.6952,
      "step": 14250
    },
    {
      "epoch": 1.54177897574124,
      "grad_norm": 0.6146219968795776,
      "learning_rate": 0.0005080733944954127,
      "loss": 3.6983,
      "step": 14300
    },
    {
      "epoch": 1.5471698113207548,
      "grad_norm": 0.5958489775657654,
      "learning_rate": 0.0005077495952509444,
      "loss": 3.6792,
      "step": 14350
    },
    {
      "epoch": 1.5525606469002695,
      "grad_norm": 0.5772738456726074,
      "learning_rate": 0.0005074257960064759,
      "loss": 3.6917,
      "step": 14400
    },
    {
      "epoch": 1.5579514824797842,
      "grad_norm": 0.6064009070396423,
      "learning_rate": 0.0005071019967620075,
      "loss": 3.6984,
      "step": 14450
    },
    {
      "epoch": 1.5633423180592994,
      "grad_norm": 0.5335804224014282,
      "learning_rate": 0.000506778197517539,
      "loss": 3.6907,
      "step": 14500
    },
    {
      "epoch": 1.568733153638814,
      "grad_norm": 0.6177234053611755,
      "learning_rate": 0.0005064543982730707,
      "loss": 3.7017,
      "step": 14550
    },
    {
      "epoch": 1.5741239892183287,
      "grad_norm": 0.6319202780723572,
      "learning_rate": 0.0005061305990286023,
      "loss": 3.6945,
      "step": 14600
    },
    {
      "epoch": 1.5795148247978437,
      "grad_norm": 0.626189112663269,
      "learning_rate": 0.0005058067997841338,
      "loss": 3.7006,
      "step": 14650
    },
    {
      "epoch": 1.5849056603773586,
      "grad_norm": 0.5782404541969299,
      "learning_rate": 0.0005054830005396654,
      "loss": 3.6892,
      "step": 14700
    },
    {
      "epoch": 1.5902964959568733,
      "grad_norm": 0.6475991010665894,
      "learning_rate": 0.0005051592012951969,
      "loss": 3.6788,
      "step": 14750
    },
    {
      "epoch": 1.595687331536388,
      "grad_norm": 0.6086334586143494,
      "learning_rate": 0.0005048354020507286,
      "loss": 3.7018,
      "step": 14800
    },
    {
      "epoch": 1.6010781671159031,
      "grad_norm": 0.582402229309082,
      "learning_rate": 0.00050451160280626,
      "loss": 3.7014,
      "step": 14850
    },
    {
      "epoch": 1.6064690026954178,
      "grad_norm": 0.5852870941162109,
      "learning_rate": 0.0005041878035617917,
      "loss": 3.6901,
      "step": 14900
    },
    {
      "epoch": 1.6118598382749325,
      "grad_norm": 0.5692827701568604,
      "learning_rate": 0.0005038640043173232,
      "loss": 3.6855,
      "step": 14950
    },
    {
      "epoch": 1.6172506738544474,
      "grad_norm": 0.6348862051963806,
      "learning_rate": 0.0005035402050728548,
      "loss": 3.6819,
      "step": 15000
    },
    {
      "epoch": 1.6172506738544474,
      "eval_accuracy": 0.35549605964869696,
      "eval_loss": 3.6405038833618164,
      "eval_runtime": 185.8619,
      "eval_samples_per_second": 96.905,
      "eval_steps_per_second": 6.058,
      "step": 15000
    },
    {
      "epoch": 1.6226415094339623,
      "grad_norm": 0.520592987537384,
      "learning_rate": 0.0005032164058283863,
      "loss": 3.6832,
      "step": 15050
    },
    {
      "epoch": 1.628032345013477,
      "grad_norm": 0.6037412285804749,
      "learning_rate": 0.0005028926065839179,
      "loss": 3.6782,
      "step": 15100
    },
    {
      "epoch": 1.633423180592992,
      "grad_norm": 0.5562775135040283,
      "learning_rate": 0.0005025688073394495,
      "loss": 3.68,
      "step": 15150
    },
    {
      "epoch": 1.6388140161725069,
      "grad_norm": 0.5715218186378479,
      "learning_rate": 0.0005022450080949811,
      "loss": 3.6867,
      "step": 15200
    },
    {
      "epoch": 1.6442048517520216,
      "grad_norm": 0.5829628109931946,
      "learning_rate": 0.0005019212088505126,
      "loss": 3.6872,
      "step": 15250
    },
    {
      "epoch": 1.6495956873315363,
      "grad_norm": 0.5366255044937134,
      "learning_rate": 0.0005015974096060442,
      "loss": 3.6895,
      "step": 15300
    },
    {
      "epoch": 1.6549865229110512,
      "grad_norm": 0.5752381682395935,
      "learning_rate": 0.0005012736103615758,
      "loss": 3.6725,
      "step": 15350
    },
    {
      "epoch": 1.6603773584905661,
      "grad_norm": 0.6247972846031189,
      "learning_rate": 0.0005009498111171074,
      "loss": 3.6772,
      "step": 15400
    },
    {
      "epoch": 1.6657681940700808,
      "grad_norm": 0.5819849371910095,
      "learning_rate": 0.0005006260118726389,
      "loss": 3.6774,
      "step": 15450
    },
    {
      "epoch": 1.6711590296495957,
      "grad_norm": 0.5812469124794006,
      "learning_rate": 0.0005003022126281705,
      "loss": 3.6797,
      "step": 15500
    },
    {
      "epoch": 1.6765498652291106,
      "grad_norm": 0.6708167195320129,
      "learning_rate": 0.000499978413383702,
      "loss": 3.6718,
      "step": 15550
    },
    {
      "epoch": 1.6819407008086253,
      "grad_norm": 0.6221471428871155,
      "learning_rate": 0.0004996546141392336,
      "loss": 3.6679,
      "step": 15600
    },
    {
      "epoch": 1.68733153638814,
      "grad_norm": 0.5361509323120117,
      "learning_rate": 0.0004993308148947651,
      "loss": 3.677,
      "step": 15650
    },
    {
      "epoch": 1.692722371967655,
      "grad_norm": 0.5450946688652039,
      "learning_rate": 0.0004990070156502968,
      "loss": 3.6676,
      "step": 15700
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.5308447480201721,
      "learning_rate": 0.0004986832164058284,
      "loss": 3.6624,
      "step": 15750
    },
    {
      "epoch": 1.7035040431266846,
      "grad_norm": 0.6182087659835815,
      "learning_rate": 0.0004983594171613599,
      "loss": 3.6766,
      "step": 15800
    },
    {
      "epoch": 1.7088948787061995,
      "grad_norm": 0.6142793297767639,
      "learning_rate": 0.0004980356179168915,
      "loss": 3.6822,
      "step": 15850
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 0.6230039000511169,
      "learning_rate": 0.000497711818672423,
      "loss": 3.7059,
      "step": 15900
    },
    {
      "epoch": 1.719676549865229,
      "grad_norm": 0.5526073575019836,
      "learning_rate": 0.0004973880194279547,
      "loss": 3.6721,
      "step": 15950
    },
    {
      "epoch": 1.7250673854447438,
      "grad_norm": 0.6014766693115234,
      "learning_rate": 0.0004970642201834862,
      "loss": 3.6808,
      "step": 16000
    },
    {
      "epoch": 1.7250673854447438,
      "eval_accuracy": 0.35676077926285754,
      "eval_loss": 3.621253490447998,
      "eval_runtime": 185.5708,
      "eval_samples_per_second": 97.057,
      "eval_steps_per_second": 6.068,
      "step": 16000
    },
    {
      "epoch": 1.7304582210242587,
      "grad_norm": 0.5674132108688354,
      "learning_rate": 0.0004967404209390178,
      "loss": 3.6468,
      "step": 16050
    },
    {
      "epoch": 1.7358490566037736,
      "grad_norm": 0.5384870767593384,
      "learning_rate": 0.0004964230976794387,
      "loss": 3.6563,
      "step": 16100
    },
    {
      "epoch": 1.7412398921832883,
      "grad_norm": 0.6142412424087524,
      "learning_rate": 0.0004960992984349703,
      "loss": 3.6694,
      "step": 16150
    },
    {
      "epoch": 1.7466307277628033,
      "grad_norm": 0.5819224119186401,
      "learning_rate": 0.0004957754991905018,
      "loss": 3.6593,
      "step": 16200
    },
    {
      "epoch": 1.7520215633423182,
      "grad_norm": 0.5680378675460815,
      "learning_rate": 0.0004954516999460334,
      "loss": 3.6677,
      "step": 16250
    },
    {
      "epoch": 1.7574123989218329,
      "grad_norm": 0.5487083196640015,
      "learning_rate": 0.0004951279007015649,
      "loss": 3.659,
      "step": 16300
    },
    {
      "epoch": 1.7628032345013476,
      "grad_norm": 0.5697620511054993,
      "learning_rate": 0.0004948041014570966,
      "loss": 3.6536,
      "step": 16350
    },
    {
      "epoch": 1.7681940700808625,
      "grad_norm": 0.6107510328292847,
      "learning_rate": 0.0004944803022126281,
      "loss": 3.6661,
      "step": 16400
    },
    {
      "epoch": 1.7735849056603774,
      "grad_norm": 0.5611395835876465,
      "learning_rate": 0.0004941565029681597,
      "loss": 3.6655,
      "step": 16450
    },
    {
      "epoch": 1.778975741239892,
      "grad_norm": 0.5468912124633789,
      "learning_rate": 0.0004938327037236912,
      "loss": 3.6518,
      "step": 16500
    },
    {
      "epoch": 1.784366576819407,
      "grad_norm": 0.5951088666915894,
      "learning_rate": 0.0004935089044792229,
      "loss": 3.6503,
      "step": 16550
    },
    {
      "epoch": 1.789757412398922,
      "grad_norm": 0.5842005610466003,
      "learning_rate": 0.0004931851052347544,
      "loss": 3.6569,
      "step": 16600
    },
    {
      "epoch": 1.7951482479784366,
      "grad_norm": 0.5568435192108154,
      "learning_rate": 0.000492861305990286,
      "loss": 3.6647,
      "step": 16650
    },
    {
      "epoch": 1.8005390835579513,
      "grad_norm": 0.5812478065490723,
      "learning_rate": 0.0004925375067458175,
      "loss": 3.6551,
      "step": 16700
    },
    {
      "epoch": 1.8059299191374663,
      "grad_norm": 0.5109179615974426,
      "learning_rate": 0.0004922137075013491,
      "loss": 3.649,
      "step": 16750
    },
    {
      "epoch": 1.8113207547169812,
      "grad_norm": 0.5683979392051697,
      "learning_rate": 0.0004918899082568807,
      "loss": 3.6546,
      "step": 16800
    },
    {
      "epoch": 1.8167115902964959,
      "grad_norm": 0.5695392489433289,
      "learning_rate": 0.0004915661090124123,
      "loss": 3.6442,
      "step": 16850
    },
    {
      "epoch": 1.8221024258760108,
      "grad_norm": 0.5891880989074707,
      "learning_rate": 0.0004912423097679439,
      "loss": 3.6499,
      "step": 16900
    },
    {
      "epoch": 1.8274932614555257,
      "grad_norm": 0.5595072507858276,
      "learning_rate": 0.0004909185105234754,
      "loss": 3.6561,
      "step": 16950
    },
    {
      "epoch": 1.8328840970350404,
      "grad_norm": 0.5379965305328369,
      "learning_rate": 0.000490594711279007,
      "loss": 3.6642,
      "step": 17000
    },
    {
      "epoch": 1.8328840970350404,
      "eval_accuracy": 0.35826377465999526,
      "eval_loss": 3.608161211013794,
      "eval_runtime": 186.0988,
      "eval_samples_per_second": 96.782,
      "eval_steps_per_second": 6.051,
      "step": 17000
    },
    {
      "epoch": 1.838274932614555,
      "grad_norm": 0.6132093667984009,
      "learning_rate": 0.0004902709120345385,
      "loss": 3.6563,
      "step": 17050
    },
    {
      "epoch": 1.8436657681940702,
      "grad_norm": 0.5843748450279236,
      "learning_rate": 0.00048994711279007,
      "loss": 3.6496,
      "step": 17100
    },
    {
      "epoch": 1.849056603773585,
      "grad_norm": 0.581844687461853,
      "learning_rate": 0.0004896233135456017,
      "loss": 3.6445,
      "step": 17150
    },
    {
      "epoch": 1.8544474393530996,
      "grad_norm": 0.5657000541687012,
      "learning_rate": 0.0004892995143011333,
      "loss": 3.644,
      "step": 17200
    },
    {
      "epoch": 1.8598382749326146,
      "grad_norm": 0.5997892618179321,
      "learning_rate": 0.0004889757150566648,
      "loss": 3.6407,
      "step": 17250
    },
    {
      "epoch": 1.8652291105121295,
      "grad_norm": 0.5654188990592957,
      "learning_rate": 0.0004886519158121964,
      "loss": 3.6647,
      "step": 17300
    },
    {
      "epoch": 1.8706199460916442,
      "grad_norm": 0.6340252757072449,
      "learning_rate": 0.000488328116567728,
      "loss": 3.6677,
      "step": 17350
    },
    {
      "epoch": 1.8760107816711589,
      "grad_norm": 0.5622780323028564,
      "learning_rate": 0.0004880043173232595,
      "loss": 3.6476,
      "step": 17400
    },
    {
      "epoch": 1.881401617250674,
      "grad_norm": 0.5995896458625793,
      "learning_rate": 0.0004876805180787911,
      "loss": 3.657,
      "step": 17450
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.6025211811065674,
      "learning_rate": 0.000487363194819212,
      "loss": 3.642,
      "step": 17500
    },
    {
      "epoch": 1.8921832884097034,
      "grad_norm": 0.6041113138198853,
      "learning_rate": 0.0004870393955747436,
      "loss": 3.6554,
      "step": 17550
    },
    {
      "epoch": 1.8975741239892183,
      "grad_norm": 0.5311667919158936,
      "learning_rate": 0.0004867155963302752,
      "loss": 3.641,
      "step": 17600
    },
    {
      "epoch": 1.9029649595687332,
      "grad_norm": 0.5440093874931335,
      "learning_rate": 0.0004863917970858068,
      "loss": 3.64,
      "step": 17650
    },
    {
      "epoch": 1.908355795148248,
      "grad_norm": 0.5968238711357117,
      "learning_rate": 0.00048606799784133833,
      "loss": 3.6522,
      "step": 17700
    },
    {
      "epoch": 1.9137466307277629,
      "grad_norm": 0.604533851146698,
      "learning_rate": 0.00048574419859686994,
      "loss": 3.6414,
      "step": 17750
    },
    {
      "epoch": 1.9191374663072778,
      "grad_norm": 0.605301558971405,
      "learning_rate": 0.0004854203993524015,
      "loss": 3.6572,
      "step": 17800
    },
    {
      "epoch": 1.9245283018867925,
      "grad_norm": 0.5855521559715271,
      "learning_rate": 0.000485096600107933,
      "loss": 3.6647,
      "step": 17850
    },
    {
      "epoch": 1.9299191374663072,
      "grad_norm": 0.5575717091560364,
      "learning_rate": 0.00048477280086346464,
      "loss": 3.6396,
      "step": 17900
    },
    {
      "epoch": 1.935309973045822,
      "grad_norm": 0.5737566351890564,
      "learning_rate": 0.00048444900161899614,
      "loss": 3.6505,
      "step": 17950
    },
    {
      "epoch": 1.940700808625337,
      "grad_norm": 0.5678514838218689,
      "learning_rate": 0.00048412520237452774,
      "loss": 3.6455,
      "step": 18000
    },
    {
      "epoch": 1.940700808625337,
      "eval_accuracy": 0.3600217783848335,
      "eval_loss": 3.5910346508026123,
      "eval_runtime": 185.5537,
      "eval_samples_per_second": 97.066,
      "eval_steps_per_second": 6.068,
      "step": 18000
    },
    {
      "epoch": 1.9460916442048517,
      "grad_norm": 0.5508000254631042,
      "learning_rate": 0.0004838014031300593,
      "loss": 3.6286,
      "step": 18050
    },
    {
      "epoch": 1.9514824797843666,
      "grad_norm": 0.5589747428894043,
      "learning_rate": 0.0004834776038855909,
      "loss": 3.6585,
      "step": 18100
    },
    {
      "epoch": 1.9568733153638815,
      "grad_norm": 0.5658039450645447,
      "learning_rate": 0.00048315380464112245,
      "loss": 3.6593,
      "step": 18150
    },
    {
      "epoch": 1.9622641509433962,
      "grad_norm": 0.5849958658218384,
      "learning_rate": 0.00048283000539665405,
      "loss": 3.6465,
      "step": 18200
    },
    {
      "epoch": 1.967654986522911,
      "grad_norm": 0.6154955625534058,
      "learning_rate": 0.0004825062061521856,
      "loss": 3.6483,
      "step": 18250
    },
    {
      "epoch": 1.9730458221024259,
      "grad_norm": 0.556688666343689,
      "learning_rate": 0.00048218240690771716,
      "loss": 3.6273,
      "step": 18300
    },
    {
      "epoch": 1.9784366576819408,
      "grad_norm": 0.6788551211357117,
      "learning_rate": 0.00048185860766324876,
      "loss": 3.6541,
      "step": 18350
    },
    {
      "epoch": 1.9838274932614555,
      "grad_norm": 0.5126660466194153,
      "learning_rate": 0.0004815348084187803,
      "loss": 3.6298,
      "step": 18400
    },
    {
      "epoch": 1.9892183288409704,
      "grad_norm": 0.5663277506828308,
      "learning_rate": 0.0004812110091743119,
      "loss": 3.6338,
      "step": 18450
    },
    {
      "epoch": 1.9946091644204853,
      "grad_norm": 0.6142022013664246,
      "learning_rate": 0.00048088720992984347,
      "loss": 3.6218,
      "step": 18500
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.1090173721313477,
      "learning_rate": 0.00048056341068537507,
      "loss": 3.6398,
      "step": 18550
    },
    {
      "epoch": 2.0053908355795147,
      "grad_norm": 0.5922869443893433,
      "learning_rate": 0.00048023961144090657,
      "loss": 3.5547,
      "step": 18600
    },
    {
      "epoch": 2.01078167115903,
      "grad_norm": 0.5729225277900696,
      "learning_rate": 0.00047991581219643817,
      "loss": 3.5452,
      "step": 18650
    },
    {
      "epoch": 2.0161725067385445,
      "grad_norm": 0.6253648400306702,
      "learning_rate": 0.0004795920129519697,
      "loss": 3.5539,
      "step": 18700
    },
    {
      "epoch": 2.0215633423180592,
      "grad_norm": 0.5972828269004822,
      "learning_rate": 0.0004792682137075013,
      "loss": 3.5445,
      "step": 18750
    },
    {
      "epoch": 2.026954177897574,
      "grad_norm": 0.6110801100730896,
      "learning_rate": 0.0004789444144630329,
      "loss": 3.5542,
      "step": 18800
    },
    {
      "epoch": 2.032345013477089,
      "grad_norm": 0.5777032971382141,
      "learning_rate": 0.00047862061521856443,
      "loss": 3.5513,
      "step": 18850
    },
    {
      "epoch": 2.0377358490566038,
      "grad_norm": 0.5737757086753845,
      "learning_rate": 0.00047829681597409603,
      "loss": 3.5633,
      "step": 18900
    },
    {
      "epoch": 2.0431266846361185,
      "grad_norm": 0.5594393014907837,
      "learning_rate": 0.0004779730167296276,
      "loss": 3.5611,
      "step": 18950
    },
    {
      "epoch": 2.0485175202156336,
      "grad_norm": 0.6480301022529602,
      "learning_rate": 0.0004776492174851592,
      "loss": 3.5613,
      "step": 19000
    },
    {
      "epoch": 2.0485175202156336,
      "eval_accuracy": 0.36164266216525,
      "eval_loss": 3.581942558288574,
      "eval_runtime": 185.9777,
      "eval_samples_per_second": 96.845,
      "eval_steps_per_second": 6.054,
      "step": 19000
    },
    {
      "epoch": 2.0539083557951483,
      "grad_norm": 0.5580395460128784,
      "learning_rate": 0.00047732541824069074,
      "loss": 3.5699,
      "step": 19050
    },
    {
      "epoch": 2.059299191374663,
      "grad_norm": 0.6071756482124329,
      "learning_rate": 0.0004770016189962223,
      "loss": 3.5623,
      "step": 19100
    },
    {
      "epoch": 2.0646900269541777,
      "grad_norm": 0.5564524531364441,
      "learning_rate": 0.0004766778197517539,
      "loss": 3.5577,
      "step": 19150
    },
    {
      "epoch": 2.070080862533693,
      "grad_norm": 0.550567626953125,
      "learning_rate": 0.0004763540205072854,
      "loss": 3.5784,
      "step": 19200
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.5679172873497009,
      "learning_rate": 0.00047603022126281705,
      "loss": 3.5672,
      "step": 19250
    },
    {
      "epoch": 2.0808625336927222,
      "grad_norm": 0.583683967590332,
      "learning_rate": 0.00047570642201834855,
      "loss": 3.5566,
      "step": 19300
    },
    {
      "epoch": 2.0862533692722374,
      "grad_norm": 0.6317183971405029,
      "learning_rate": 0.00047538262277388015,
      "loss": 3.5737,
      "step": 19350
    },
    {
      "epoch": 2.091644204851752,
      "grad_norm": 0.5760337710380554,
      "learning_rate": 0.0004750588235294117,
      "loss": 3.566,
      "step": 19400
    },
    {
      "epoch": 2.0970350404312668,
      "grad_norm": 0.7079157829284668,
      "learning_rate": 0.0004747350242849433,
      "loss": 3.5478,
      "step": 19450
    },
    {
      "epoch": 2.1024258760107815,
      "grad_norm": 0.5852265954017639,
      "learning_rate": 0.00047441122504047486,
      "loss": 3.5721,
      "step": 19500
    },
    {
      "epoch": 2.1078167115902966,
      "grad_norm": 0.6242895126342773,
      "learning_rate": 0.0004740874257960064,
      "loss": 3.564,
      "step": 19550
    },
    {
      "epoch": 2.1132075471698113,
      "grad_norm": 0.5763869285583496,
      "learning_rate": 0.000473763626551538,
      "loss": 3.5606,
      "step": 19600
    },
    {
      "epoch": 2.118598382749326,
      "grad_norm": 0.568956196308136,
      "learning_rate": 0.00047343982730706956,
      "loss": 3.5628,
      "step": 19650
    },
    {
      "epoch": 2.123989218328841,
      "grad_norm": 0.5798338651657104,
      "learning_rate": 0.00047311602806260117,
      "loss": 3.5561,
      "step": 19700
    },
    {
      "epoch": 2.129380053908356,
      "grad_norm": 0.6521062254905701,
      "learning_rate": 0.0004727922288181327,
      "loss": 3.5471,
      "step": 19750
    },
    {
      "epoch": 2.1347708894878705,
      "grad_norm": 0.5969342589378357,
      "learning_rate": 0.0004724684295736643,
      "loss": 3.569,
      "step": 19800
    },
    {
      "epoch": 2.1401617250673857,
      "grad_norm": 0.585344672203064,
      "learning_rate": 0.0004721446303291959,
      "loss": 3.56,
      "step": 19850
    },
    {
      "epoch": 2.1455525606469004,
      "grad_norm": 0.5696685314178467,
      "learning_rate": 0.0004718208310847275,
      "loss": 3.5634,
      "step": 19900
    },
    {
      "epoch": 2.150943396226415,
      "grad_norm": 0.5736604928970337,
      "learning_rate": 0.000471497031840259,
      "loss": 3.5546,
      "step": 19950
    },
    {
      "epoch": 2.1563342318059298,
      "grad_norm": 0.5983167290687561,
      "learning_rate": 0.0004711732325957905,
      "loss": 3.5589,
      "step": 20000
    },
    {
      "epoch": 2.1563342318059298,
      "eval_accuracy": 0.36254372056389544,
      "eval_loss": 3.573596715927124,
      "eval_runtime": 185.5769,
      "eval_samples_per_second": 97.054,
      "eval_steps_per_second": 6.068,
      "step": 20000
    },
    {
      "epoch": 2.161725067385445,
      "grad_norm": 0.5717964768409729,
      "learning_rate": 0.00047084943335132213,
      "loss": 3.5619,
      "step": 20050
    },
    {
      "epoch": 2.1671159029649596,
      "grad_norm": 0.6264322996139526,
      "learning_rate": 0.0004705256341068537,
      "loss": 3.5579,
      "step": 20100
    },
    {
      "epoch": 2.1725067385444743,
      "grad_norm": 0.5710529088973999,
      "learning_rate": 0.0004702018348623853,
      "loss": 3.5685,
      "step": 20150
    },
    {
      "epoch": 2.177897574123989,
      "grad_norm": 0.5462448000907898,
      "learning_rate": 0.00046987803561791684,
      "loss": 3.5703,
      "step": 20200
    },
    {
      "epoch": 2.183288409703504,
      "grad_norm": 0.5277117490768433,
      "learning_rate": 0.00046955423637344844,
      "loss": 3.5762,
      "step": 20250
    },
    {
      "epoch": 2.188679245283019,
      "grad_norm": 0.6546245217323303,
      "learning_rate": 0.00046923043712898,
      "loss": 3.5575,
      "step": 20300
    },
    {
      "epoch": 2.1940700808625335,
      "grad_norm": 0.5462058782577515,
      "learning_rate": 0.0004689066378845116,
      "loss": 3.5511,
      "step": 20350
    },
    {
      "epoch": 2.1994609164420487,
      "grad_norm": 0.5982887744903564,
      "learning_rate": 0.00046858283864004315,
      "loss": 3.5645,
      "step": 20400
    },
    {
      "epoch": 2.2048517520215634,
      "grad_norm": 0.7412492036819458,
      "learning_rate": 0.0004682590393955747,
      "loss": 3.5573,
      "step": 20450
    },
    {
      "epoch": 2.210242587601078,
      "grad_norm": 0.5417401790618896,
      "learning_rate": 0.0004679352401511063,
      "loss": 3.5739,
      "step": 20500
    },
    {
      "epoch": 2.215633423180593,
      "grad_norm": 0.5726649761199951,
      "learning_rate": 0.0004676114409066378,
      "loss": 3.5713,
      "step": 20550
    },
    {
      "epoch": 2.221024258760108,
      "grad_norm": 0.6224658489227295,
      "learning_rate": 0.00046728764166216946,
      "loss": 3.5665,
      "step": 20600
    },
    {
      "epoch": 2.2264150943396226,
      "grad_norm": 0.6279109120368958,
      "learning_rate": 0.00046696384241770095,
      "loss": 3.576,
      "step": 20650
    },
    {
      "epoch": 2.2318059299191373,
      "grad_norm": 0.6470381617546082,
      "learning_rate": 0.00046664004317323256,
      "loss": 3.5569,
      "step": 20700
    },
    {
      "epoch": 2.2371967654986524,
      "grad_norm": 0.5726243853569031,
      "learning_rate": 0.0004663162439287641,
      "loss": 3.5754,
      "step": 20750
    },
    {
      "epoch": 2.242587601078167,
      "grad_norm": 0.5811974406242371,
      "learning_rate": 0.00046599244468429566,
      "loss": 3.5715,
      "step": 20800
    },
    {
      "epoch": 2.247978436657682,
      "grad_norm": 0.5972831845283508,
      "learning_rate": 0.00046566864543982726,
      "loss": 3.5719,
      "step": 20850
    },
    {
      "epoch": 2.2533692722371965,
      "grad_norm": 0.5632637739181519,
      "learning_rate": 0.0004653448461953588,
      "loss": 3.5633,
      "step": 20900
    },
    {
      "epoch": 2.2587601078167117,
      "grad_norm": 0.6027004718780518,
      "learning_rate": 0.0004650210469508904,
      "loss": 3.5614,
      "step": 20950
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 0.5686562061309814,
      "learning_rate": 0.00046469724770642197,
      "loss": 3.5597,
      "step": 21000
    },
    {
      "epoch": 2.2641509433962264,
      "eval_accuracy": 0.36353789448739793,
      "eval_loss": 3.5626556873321533,
      "eval_runtime": 185.9797,
      "eval_samples_per_second": 96.844,
      "eval_steps_per_second": 6.054,
      "step": 21000
    },
    {
      "epoch": 2.269541778975741,
      "grad_norm": 0.6364426016807556,
      "learning_rate": 0.0004643734484619536,
      "loss": 3.5565,
      "step": 21050
    },
    {
      "epoch": 2.274932614555256,
      "grad_norm": 0.5865249037742615,
      "learning_rate": 0.0004640496492174851,
      "loss": 3.5483,
      "step": 21100
    },
    {
      "epoch": 2.280323450134771,
      "grad_norm": 0.5965092778205872,
      "learning_rate": 0.00046372584997301673,
      "loss": 3.5742,
      "step": 21150
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 0.5928227305412292,
      "learning_rate": 0.0004634020507285483,
      "loss": 3.5535,
      "step": 21200
    },
    {
      "epoch": 2.2911051212938007,
      "grad_norm": 0.5737338662147522,
      "learning_rate": 0.0004630782514840798,
      "loss": 3.5516,
      "step": 21250
    },
    {
      "epoch": 2.2964959568733154,
      "grad_norm": 0.6345467567443848,
      "learning_rate": 0.0004627544522396114,
      "loss": 3.5598,
      "step": 21300
    },
    {
      "epoch": 2.30188679245283,
      "grad_norm": 0.5607115030288696,
      "learning_rate": 0.00046243065299514293,
      "loss": 3.5408,
      "step": 21350
    },
    {
      "epoch": 2.3072776280323453,
      "grad_norm": 0.5754068493843079,
      "learning_rate": 0.00046210685375067454,
      "loss": 3.5766,
      "step": 21400
    },
    {
      "epoch": 2.31266846361186,
      "grad_norm": 0.6035879254341125,
      "learning_rate": 0.0004617830545062061,
      "loss": 3.536,
      "step": 21450
    },
    {
      "epoch": 2.3180592991913747,
      "grad_norm": 0.6164117455482483,
      "learning_rate": 0.00046146573124662704,
      "loss": 3.5671,
      "step": 21500
    },
    {
      "epoch": 2.3234501347708894,
      "grad_norm": 0.5974962115287781,
      "learning_rate": 0.00046114193200215864,
      "loss": 3.5658,
      "step": 21550
    },
    {
      "epoch": 2.3288409703504045,
      "grad_norm": 0.6097601652145386,
      "learning_rate": 0.0004608181327576902,
      "loss": 3.5402,
      "step": 21600
    },
    {
      "epoch": 2.334231805929919,
      "grad_norm": 0.5905078053474426,
      "learning_rate": 0.00046049433351322175,
      "loss": 3.5607,
      "step": 21650
    },
    {
      "epoch": 2.339622641509434,
      "grad_norm": 0.557483971118927,
      "learning_rate": 0.00046017053426875335,
      "loss": 3.5734,
      "step": 21700
    },
    {
      "epoch": 2.3450134770889486,
      "grad_norm": 0.6139761805534363,
      "learning_rate": 0.0004598467350242849,
      "loss": 3.5557,
      "step": 21750
    },
    {
      "epoch": 2.3504043126684637,
      "grad_norm": 0.5611399412155151,
      "learning_rate": 0.0004595229357798165,
      "loss": 3.5681,
      "step": 21800
    },
    {
      "epoch": 2.3557951482479784,
      "grad_norm": 0.605839192867279,
      "learning_rate": 0.00045919913653534806,
      "loss": 3.5671,
      "step": 21850
    },
    {
      "epoch": 2.361185983827493,
      "grad_norm": 0.5682869553565979,
      "learning_rate": 0.00045887533729087966,
      "loss": 3.5595,
      "step": 21900
    },
    {
      "epoch": 2.3665768194070083,
      "grad_norm": 0.5689018368721008,
      "learning_rate": 0.00045855153804641116,
      "loss": 3.5521,
      "step": 21950
    },
    {
      "epoch": 2.371967654986523,
      "grad_norm": 0.5656780004501343,
      "learning_rate": 0.0004582277388019427,
      "loss": 3.5455,
      "step": 22000
    },
    {
      "epoch": 2.371967654986523,
      "eval_accuracy": 0.36478490368084476,
      "eval_loss": 3.5518763065338135,
      "eval_runtime": 185.6801,
      "eval_samples_per_second": 97.0,
      "eval_steps_per_second": 6.064,
      "step": 22000
    },
    {
      "epoch": 2.3773584905660377,
      "grad_norm": 0.5644449591636658,
      "learning_rate": 0.0004579039395574743,
      "loss": 3.547,
      "step": 22050
    },
    {
      "epoch": 2.382749326145553,
      "grad_norm": 0.5704251527786255,
      "learning_rate": 0.00045758014031300586,
      "loss": 3.5652,
      "step": 22100
    },
    {
      "epoch": 2.3881401617250675,
      "grad_norm": 0.5917501449584961,
      "learning_rate": 0.00045725634106853747,
      "loss": 3.5594,
      "step": 22150
    },
    {
      "epoch": 2.393530997304582,
      "grad_norm": 0.552393913269043,
      "learning_rate": 0.000456932541824069,
      "loss": 3.5463,
      "step": 22200
    },
    {
      "epoch": 2.398921832884097,
      "grad_norm": 0.6430604457855225,
      "learning_rate": 0.0004566087425796006,
      "loss": 3.5386,
      "step": 22250
    },
    {
      "epoch": 2.404312668463612,
      "grad_norm": 0.6215885281562805,
      "learning_rate": 0.0004562849433351322,
      "loss": 3.5672,
      "step": 22300
    },
    {
      "epoch": 2.4097035040431267,
      "grad_norm": 0.5455368161201477,
      "learning_rate": 0.0004559611440906638,
      "loss": 3.5753,
      "step": 22350
    },
    {
      "epoch": 2.4150943396226414,
      "grad_norm": 0.6595668792724609,
      "learning_rate": 0.00045563734484619533,
      "loss": 3.5592,
      "step": 22400
    },
    {
      "epoch": 2.420485175202156,
      "grad_norm": 0.575469970703125,
      "learning_rate": 0.0004553135456017269,
      "loss": 3.5671,
      "step": 22450
    },
    {
      "epoch": 2.4258760107816713,
      "grad_norm": 0.6219996213912964,
      "learning_rate": 0.0004549897463572585,
      "loss": 3.5634,
      "step": 22500
    },
    {
      "epoch": 2.431266846361186,
      "grad_norm": 0.564261257648468,
      "learning_rate": 0.00045466594711279,
      "loss": 3.5432,
      "step": 22550
    },
    {
      "epoch": 2.4366576819407006,
      "grad_norm": 0.649264395236969,
      "learning_rate": 0.00045434214786832164,
      "loss": 3.5624,
      "step": 22600
    },
    {
      "epoch": 2.442048517520216,
      "grad_norm": 0.5878944396972656,
      "learning_rate": 0.00045401834862385314,
      "loss": 3.5565,
      "step": 22650
    },
    {
      "epoch": 2.4474393530997305,
      "grad_norm": 0.6890021562576294,
      "learning_rate": 0.00045369454937938474,
      "loss": 3.5634,
      "step": 22700
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.6277121305465698,
      "learning_rate": 0.0004533707501349163,
      "loss": 3.5508,
      "step": 22750
    },
    {
      "epoch": 2.4582210242587603,
      "grad_norm": 0.5444703698158264,
      "learning_rate": 0.0004530469508904479,
      "loss": 3.5509,
      "step": 22800
    },
    {
      "epoch": 2.463611859838275,
      "grad_norm": 0.5967130661010742,
      "learning_rate": 0.00045272315164597945,
      "loss": 3.5403,
      "step": 22850
    },
    {
      "epoch": 2.4690026954177897,
      "grad_norm": 0.5967291593551636,
      "learning_rate": 0.000452399352401511,
      "loss": 3.5545,
      "step": 22900
    },
    {
      "epoch": 2.4743935309973044,
      "grad_norm": 0.5806335210800171,
      "learning_rate": 0.0004520755531570426,
      "loss": 3.5388,
      "step": 22950
    },
    {
      "epoch": 2.4797843665768196,
      "grad_norm": 0.610866904258728,
      "learning_rate": 0.00045175175391257415,
      "loss": 3.5547,
      "step": 23000
    },
    {
      "epoch": 2.4797843665768196,
      "eval_accuracy": 0.365600560909668,
      "eval_loss": 3.542806625366211,
      "eval_runtime": 186.022,
      "eval_samples_per_second": 96.822,
      "eval_steps_per_second": 6.053,
      "step": 23000
    },
    {
      "epoch": 2.4851752021563343,
      "grad_norm": 0.620668351650238,
      "learning_rate": 0.00045142795466810576,
      "loss": 3.5601,
      "step": 23050
    },
    {
      "epoch": 2.490566037735849,
      "grad_norm": 0.593329131603241,
      "learning_rate": 0.0004511041554236373,
      "loss": 3.5578,
      "step": 23100
    },
    {
      "epoch": 2.4959568733153636,
      "grad_norm": 0.6220084428787231,
      "learning_rate": 0.0004507803561791689,
      "loss": 3.5704,
      "step": 23150
    },
    {
      "epoch": 2.501347708894879,
      "grad_norm": 0.5983256101608276,
      "learning_rate": 0.00045045655693470046,
      "loss": 3.5619,
      "step": 23200
    },
    {
      "epoch": 2.5067385444743935,
      "grad_norm": 0.6088994741439819,
      "learning_rate": 0.00045013275769023207,
      "loss": 3.5623,
      "step": 23250
    },
    {
      "epoch": 2.512129380053908,
      "grad_norm": 0.5717225670814514,
      "learning_rate": 0.00044980895844576356,
      "loss": 3.5584,
      "step": 23300
    },
    {
      "epoch": 2.5175202156334233,
      "grad_norm": 0.5791255235671997,
      "learning_rate": 0.0004494851592012951,
      "loss": 3.5562,
      "step": 23350
    },
    {
      "epoch": 2.522911051212938,
      "grad_norm": 0.5934580564498901,
      "learning_rate": 0.00044916783594171607,
      "loss": 3.5359,
      "step": 23400
    },
    {
      "epoch": 2.5283018867924527,
      "grad_norm": 0.5999855995178223,
      "learning_rate": 0.00044884403669724767,
      "loss": 3.5516,
      "step": 23450
    },
    {
      "epoch": 2.533692722371968,
      "grad_norm": 0.5984418988227844,
      "learning_rate": 0.0004485202374527792,
      "loss": 3.5462,
      "step": 23500
    },
    {
      "epoch": 2.5390835579514826,
      "grad_norm": 0.558322548866272,
      "learning_rate": 0.00044819643820831083,
      "loss": 3.5491,
      "step": 23550
    },
    {
      "epoch": 2.5444743935309972,
      "grad_norm": 0.6213822960853577,
      "learning_rate": 0.0004478726389638424,
      "loss": 3.5522,
      "step": 23600
    },
    {
      "epoch": 2.5498652291105124,
      "grad_norm": 0.5522050261497498,
      "learning_rate": 0.00044754883971937393,
      "loss": 3.5409,
      "step": 23650
    },
    {
      "epoch": 2.555256064690027,
      "grad_norm": 0.5763940811157227,
      "learning_rate": 0.00044722504047490553,
      "loss": 3.5296,
      "step": 23700
    },
    {
      "epoch": 2.560646900269542,
      "grad_norm": 0.6493868231773376,
      "learning_rate": 0.0004469012412304371,
      "loss": 3.5384,
      "step": 23750
    },
    {
      "epoch": 2.5660377358490565,
      "grad_norm": 0.5779409408569336,
      "learning_rate": 0.0004465774419859687,
      "loss": 3.5574,
      "step": 23800
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 0.5564289689064026,
      "learning_rate": 0.00044625364274150024,
      "loss": 3.5447,
      "step": 23850
    },
    {
      "epoch": 2.5768194070080863,
      "grad_norm": 0.5602936148643494,
      "learning_rate": 0.00044592984349703184,
      "loss": 3.5445,
      "step": 23900
    },
    {
      "epoch": 2.582210242587601,
      "grad_norm": 0.635210394859314,
      "learning_rate": 0.00044560604425256334,
      "loss": 3.5371,
      "step": 23950
    },
    {
      "epoch": 2.5876010781671157,
      "grad_norm": 0.5645261406898499,
      "learning_rate": 0.000445282245008095,
      "loss": 3.5427,
      "step": 24000
    },
    {
      "epoch": 2.5876010781671157,
      "eval_accuracy": 0.366985124659088,
      "eval_loss": 3.5324132442474365,
      "eval_runtime": 185.6401,
      "eval_samples_per_second": 97.021,
      "eval_steps_per_second": 6.065,
      "step": 24000
    },
    {
      "epoch": 2.592991913746631,
      "grad_norm": 0.5728445053100586,
      "learning_rate": 0.0004449584457636265,
      "loss": 3.5657,
      "step": 24050
    },
    {
      "epoch": 2.5983827493261455,
      "grad_norm": 0.6249076724052429,
      "learning_rate": 0.00044463464651915805,
      "loss": 3.5435,
      "step": 24100
    },
    {
      "epoch": 2.6037735849056602,
      "grad_norm": 0.5713481903076172,
      "learning_rate": 0.00044431084727468965,
      "loss": 3.5558,
      "step": 24150
    },
    {
      "epoch": 2.6091644204851754,
      "grad_norm": 0.6012597680091858,
      "learning_rate": 0.0004439870480302212,
      "loss": 3.5396,
      "step": 24200
    },
    {
      "epoch": 2.61455525606469,
      "grad_norm": 0.6245263814926147,
      "learning_rate": 0.0004436632487857528,
      "loss": 3.5574,
      "step": 24250
    },
    {
      "epoch": 2.6199460916442048,
      "grad_norm": 0.6450279951095581,
      "learning_rate": 0.00044333944954128436,
      "loss": 3.5498,
      "step": 24300
    },
    {
      "epoch": 2.62533692722372,
      "grad_norm": 0.5852833986282349,
      "learning_rate": 0.00044301565029681596,
      "loss": 3.5417,
      "step": 24350
    },
    {
      "epoch": 2.6307277628032346,
      "grad_norm": 0.6017525792121887,
      "learning_rate": 0.0004426918510523475,
      "loss": 3.545,
      "step": 24400
    },
    {
      "epoch": 2.6361185983827493,
      "grad_norm": 0.5775270462036133,
      "learning_rate": 0.0004423680518078791,
      "loss": 3.5449,
      "step": 24450
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.5868250131607056,
      "learning_rate": 0.00044204425256341067,
      "loss": 3.5523,
      "step": 24500
    },
    {
      "epoch": 2.6469002695417787,
      "grad_norm": 0.6670169234275818,
      "learning_rate": 0.0004417204533189422,
      "loss": 3.5373,
      "step": 24550
    },
    {
      "epoch": 2.652291105121294,
      "grad_norm": 0.6082099676132202,
      "learning_rate": 0.0004413966540744738,
      "loss": 3.5459,
      "step": 24600
    },
    {
      "epoch": 2.6576819407008085,
      "grad_norm": 0.5733329653739929,
      "learning_rate": 0.0004410728548300053,
      "loss": 3.5421,
      "step": 24650
    },
    {
      "epoch": 2.6630727762803232,
      "grad_norm": 0.5629079937934875,
      "learning_rate": 0.0004407490555855369,
      "loss": 3.548,
      "step": 24700
    },
    {
      "epoch": 2.6684636118598384,
      "grad_norm": 0.6059092283248901,
      "learning_rate": 0.0004404252563410685,
      "loss": 3.5628,
      "step": 24750
    },
    {
      "epoch": 2.673854447439353,
      "grad_norm": 0.5731385946273804,
      "learning_rate": 0.0004401014570966001,
      "loss": 3.5559,
      "step": 24800
    },
    {
      "epoch": 2.6792452830188678,
      "grad_norm": 0.577727198600769,
      "learning_rate": 0.00043977765785213163,
      "loss": 3.5576,
      "step": 24850
    },
    {
      "epoch": 2.684636118598383,
      "grad_norm": 0.6294382214546204,
      "learning_rate": 0.0004394538586076632,
      "loss": 3.529,
      "step": 24900
    },
    {
      "epoch": 2.6900269541778976,
      "grad_norm": 0.5616303086280823,
      "learning_rate": 0.0004391300593631948,
      "loss": 3.5487,
      "step": 24950
    },
    {
      "epoch": 2.6954177897574123,
      "grad_norm": 0.5719131231307983,
      "learning_rate": 0.00043880626011872634,
      "loss": 3.5285,
      "step": 25000
    },
    {
      "epoch": 2.6954177897574123,
      "eval_accuracy": 0.3679841879625421,
      "eval_loss": 3.5216171741485596,
      "eval_runtime": 186.3995,
      "eval_samples_per_second": 96.626,
      "eval_steps_per_second": 6.041,
      "step": 25000
    },
    {
      "epoch": 2.7008086253369274,
      "grad_norm": 0.5909615159034729,
      "learning_rate": 0.00043848246087425794,
      "loss": 3.5423,
      "step": 25050
    },
    {
      "epoch": 2.706199460916442,
      "grad_norm": 0.5884896516799927,
      "learning_rate": 0.0004381586616297895,
      "loss": 3.5422,
      "step": 25100
    },
    {
      "epoch": 2.711590296495957,
      "grad_norm": 0.6020126938819885,
      "learning_rate": 0.0004378348623853211,
      "loss": 3.5438,
      "step": 25150
    },
    {
      "epoch": 2.7169811320754715,
      "grad_norm": 0.6269355416297913,
      "learning_rate": 0.00043751106314085265,
      "loss": 3.546,
      "step": 25200
    },
    {
      "epoch": 2.7223719676549867,
      "grad_norm": 0.6308701634407043,
      "learning_rate": 0.00043718726389638425,
      "loss": 3.5375,
      "step": 25250
    },
    {
      "epoch": 2.7277628032345014,
      "grad_norm": 0.6092113852500916,
      "learning_rate": 0.00043686346465191575,
      "loss": 3.5448,
      "step": 25300
    },
    {
      "epoch": 2.733153638814016,
      "grad_norm": 0.6175215244293213,
      "learning_rate": 0.0004365396654074473,
      "loss": 3.5399,
      "step": 25350
    },
    {
      "epoch": 2.7385444743935308,
      "grad_norm": 0.5836145877838135,
      "learning_rate": 0.0004362158661629789,
      "loss": 3.5344,
      "step": 25400
    },
    {
      "epoch": 2.743935309973046,
      "grad_norm": 0.6724771857261658,
      "learning_rate": 0.00043589206691851045,
      "loss": 3.5503,
      "step": 25450
    },
    {
      "epoch": 2.7493261455525606,
      "grad_norm": 0.5822784304618835,
      "learning_rate": 0.00043556826767404206,
      "loss": 3.5273,
      "step": 25500
    },
    {
      "epoch": 2.7547169811320753,
      "grad_norm": 0.6363187432289124,
      "learning_rate": 0.0004352444684295736,
      "loss": 3.5201,
      "step": 25550
    },
    {
      "epoch": 2.7601078167115904,
      "grad_norm": 0.5680950880050659,
      "learning_rate": 0.0004349206691851052,
      "loss": 3.5242,
      "step": 25600
    },
    {
      "epoch": 2.765498652291105,
      "grad_norm": 0.5622973442077637,
      "learning_rate": 0.00043459686994063676,
      "loss": 3.5318,
      "step": 25650
    },
    {
      "epoch": 2.77088948787062,
      "grad_norm": 0.5824505090713501,
      "learning_rate": 0.00043427307069616837,
      "loss": 3.5197,
      "step": 25700
    },
    {
      "epoch": 2.776280323450135,
      "grad_norm": 0.6137266159057617,
      "learning_rate": 0.0004339492714516999,
      "loss": 3.5221,
      "step": 25750
    },
    {
      "epoch": 2.7816711590296497,
      "grad_norm": 0.6163187026977539,
      "learning_rate": 0.00043362547220723147,
      "loss": 3.5376,
      "step": 25800
    },
    {
      "epoch": 2.7870619946091644,
      "grad_norm": 0.6002925038337708,
      "learning_rate": 0.0004333016729627631,
      "loss": 3.5233,
      "step": 25850
    },
    {
      "epoch": 2.7924528301886795,
      "grad_norm": 0.5682194828987122,
      "learning_rate": 0.0004329778737182946,
      "loss": 3.5258,
      "step": 25900
    },
    {
      "epoch": 2.797843665768194,
      "grad_norm": 0.5632787942886353,
      "learning_rate": 0.00043265407447382623,
      "loss": 3.5361,
      "step": 25950
    },
    {
      "epoch": 2.803234501347709,
      "grad_norm": 0.6110450029373169,
      "learning_rate": 0.0004323302752293577,
      "loss": 3.5395,
      "step": 26000
    },
    {
      "epoch": 2.803234501347709,
      "eval_accuracy": 0.3686300207277114,
      "eval_loss": 3.513617992401123,
      "eval_runtime": 185.827,
      "eval_samples_per_second": 96.924,
      "eval_steps_per_second": 6.059,
      "step": 26000
    },
    {
      "epoch": 2.8086253369272236,
      "grad_norm": 0.6710889339447021,
      "learning_rate": 0.00043200647598488933,
      "loss": 3.5389,
      "step": 26050
    },
    {
      "epoch": 2.8140161725067383,
      "grad_norm": 0.6085184812545776,
      "learning_rate": 0.0004316826767404209,
      "loss": 3.5215,
      "step": 26100
    },
    {
      "epoch": 2.8194070080862534,
      "grad_norm": 0.5876900553703308,
      "learning_rate": 0.0004313588774959525,
      "loss": 3.5307,
      "step": 26150
    },
    {
      "epoch": 2.824797843665768,
      "grad_norm": 0.6706977486610413,
      "learning_rate": 0.00043103507825148404,
      "loss": 3.5284,
      "step": 26200
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.6303845047950745,
      "learning_rate": 0.0004307112790070156,
      "loss": 3.5225,
      "step": 26250
    },
    {
      "epoch": 2.835579514824798,
      "grad_norm": 0.6239638328552246,
      "learning_rate": 0.0004303874797625472,
      "loss": 3.5361,
      "step": 26300
    },
    {
      "epoch": 2.8409703504043127,
      "grad_norm": 0.6235447525978088,
      "learning_rate": 0.00043006368051807874,
      "loss": 3.5407,
      "step": 26350
    },
    {
      "epoch": 2.8463611859838274,
      "grad_norm": 0.6071142554283142,
      "learning_rate": 0.00042973988127361035,
      "loss": 3.5293,
      "step": 26400
    },
    {
      "epoch": 2.8517520215633425,
      "grad_norm": 0.5975500345230103,
      "learning_rate": 0.0004294160820291419,
      "loss": 3.5295,
      "step": 26450
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.6414698362350464,
      "learning_rate": 0.0004290922827846735,
      "loss": 3.5255,
      "step": 26500
    },
    {
      "epoch": 2.862533692722372,
      "grad_norm": 0.6065353155136108,
      "learning_rate": 0.00042876848354020505,
      "loss": 3.5175,
      "step": 26550
    },
    {
      "epoch": 2.867924528301887,
      "grad_norm": 0.618472158908844,
      "learning_rate": 0.00042844468429573655,
      "loss": 3.5289,
      "step": 26600
    },
    {
      "epoch": 2.8733153638814017,
      "grad_norm": 0.6460848450660706,
      "learning_rate": 0.00042812088505126815,
      "loss": 3.5247,
      "step": 26650
    },
    {
      "epoch": 2.8787061994609164,
      "grad_norm": 0.5703945159912109,
      "learning_rate": 0.0004277970858067997,
      "loss": 3.5233,
      "step": 26700
    },
    {
      "epoch": 2.884097035040431,
      "grad_norm": 0.5636091828346252,
      "learning_rate": 0.0004274732865623313,
      "loss": 3.5415,
      "step": 26750
    },
    {
      "epoch": 2.889487870619946,
      "grad_norm": 0.6512492895126343,
      "learning_rate": 0.00042714948731786286,
      "loss": 3.5292,
      "step": 26800
    },
    {
      "epoch": 2.894878706199461,
      "grad_norm": 0.5482248067855835,
      "learning_rate": 0.00042682568807339447,
      "loss": 3.5223,
      "step": 26850
    },
    {
      "epoch": 2.9002695417789757,
      "grad_norm": 0.6073907017707825,
      "learning_rate": 0.000426501888828926,
      "loss": 3.5464,
      "step": 26900
    },
    {
      "epoch": 2.9056603773584904,
      "grad_norm": 0.5725634098052979,
      "learning_rate": 0.0004261780895844576,
      "loss": 3.5388,
      "step": 26950
    },
    {
      "epoch": 2.9110512129380055,
      "grad_norm": 0.5964205861091614,
      "learning_rate": 0.00042585429033998917,
      "loss": 3.523,
      "step": 27000
    },
    {
      "epoch": 2.9110512129380055,
      "eval_accuracy": 0.3692841111123546,
      "eval_loss": 3.5060484409332275,
      "eval_runtime": 186.1917,
      "eval_samples_per_second": 96.734,
      "eval_steps_per_second": 6.048,
      "step": 27000
    },
    {
      "epoch": 2.91644204851752,
      "grad_norm": 0.6369139552116394,
      "learning_rate": 0.0004255304910955207,
      "loss": 3.5417,
      "step": 27050
    },
    {
      "epoch": 2.921832884097035,
      "grad_norm": 0.5514616966247559,
      "learning_rate": 0.0004252066918510523,
      "loss": 3.5463,
      "step": 27100
    },
    {
      "epoch": 2.92722371967655,
      "grad_norm": 0.6309407353401184,
      "learning_rate": 0.0004248828926065839,
      "loss": 3.5301,
      "step": 27150
    },
    {
      "epoch": 2.9326145552560647,
      "grad_norm": 0.5894890427589417,
      "learning_rate": 0.0004245590933621155,
      "loss": 3.52,
      "step": 27200
    },
    {
      "epoch": 2.9380053908355794,
      "grad_norm": 0.5637946724891663,
      "learning_rate": 0.00042423529411764703,
      "loss": 3.5323,
      "step": 27250
    },
    {
      "epoch": 2.9433962264150946,
      "grad_norm": 0.6292248964309692,
      "learning_rate": 0.00042391149487317864,
      "loss": 3.5295,
      "step": 27300
    },
    {
      "epoch": 2.9487870619946093,
      "grad_norm": 0.6090145707130432,
      "learning_rate": 0.00042358769562871013,
      "loss": 3.5224,
      "step": 27350
    },
    {
      "epoch": 2.954177897574124,
      "grad_norm": 0.6164013147354126,
      "learning_rate": 0.0004232703723691311,
      "loss": 3.5252,
      "step": 27400
    },
    {
      "epoch": 2.9595687331536387,
      "grad_norm": 0.5353987812995911,
      "learning_rate": 0.00042295304910955204,
      "loss": 3.5225,
      "step": 27450
    },
    {
      "epoch": 2.964959568733154,
      "grad_norm": 0.5768761038780212,
      "learning_rate": 0.0004226292498650836,
      "loss": 3.5289,
      "step": 27500
    },
    {
      "epoch": 2.9703504043126685,
      "grad_norm": 0.5933774709701538,
      "learning_rate": 0.0004223054506206152,
      "loss": 3.5313,
      "step": 27550
    },
    {
      "epoch": 2.975741239892183,
      "grad_norm": 0.5720575451850891,
      "learning_rate": 0.00042198165137614674,
      "loss": 3.5331,
      "step": 27600
    },
    {
      "epoch": 2.981132075471698,
      "grad_norm": 0.5767709016799927,
      "learning_rate": 0.00042165785213167835,
      "loss": 3.5353,
      "step": 27650
    },
    {
      "epoch": 2.986522911051213,
      "grad_norm": 0.6661293506622314,
      "learning_rate": 0.0004213340528872099,
      "loss": 3.5243,
      "step": 27700
    },
    {
      "epoch": 2.9919137466307277,
      "grad_norm": 0.5878481864929199,
      "learning_rate": 0.00042101025364274145,
      "loss": 3.537,
      "step": 27750
    },
    {
      "epoch": 2.9973045822102424,
      "grad_norm": 0.641791582107544,
      "learning_rate": 0.00042068645439827305,
      "loss": 3.5329,
      "step": 27800
    },
    {
      "epoch": 3.0026954177897576,
      "grad_norm": 0.5974509119987488,
      "learning_rate": 0.0004203626551538046,
      "loss": 3.4783,
      "step": 27850
    },
    {
      "epoch": 3.0080862533692723,
      "grad_norm": 0.6066641807556152,
      "learning_rate": 0.0004200388559093362,
      "loss": 3.4297,
      "step": 27900
    },
    {
      "epoch": 3.013477088948787,
      "grad_norm": 0.6649867296218872,
      "learning_rate": 0.0004197150566648677,
      "loss": 3.4268,
      "step": 27950
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 0.5743685960769653,
      "learning_rate": 0.00041939125742039936,
      "loss": 3.4475,
      "step": 28000
    },
    {
      "epoch": 3.018867924528302,
      "eval_accuracy": 0.37044593644175483,
      "eval_loss": 3.5023696422576904,
      "eval_runtime": 185.967,
      "eval_samples_per_second": 96.851,
      "eval_steps_per_second": 6.055,
      "step": 28000
    },
    {
      "epoch": 3.024258760107817,
      "grad_norm": 0.6364364624023438,
      "learning_rate": 0.00041906745817593086,
      "loss": 3.4426,
      "step": 28050
    },
    {
      "epoch": 3.0296495956873315,
      "grad_norm": 0.6206678748130798,
      "learning_rate": 0.00041874365893146247,
      "loss": 3.449,
      "step": 28100
    },
    {
      "epoch": 3.035040431266846,
      "grad_norm": 0.6234203577041626,
      "learning_rate": 0.000418419859686994,
      "loss": 3.4393,
      "step": 28150
    },
    {
      "epoch": 3.0404312668463613,
      "grad_norm": 0.6210460066795349,
      "learning_rate": 0.00041809606044252557,
      "loss": 3.4433,
      "step": 28200
    },
    {
      "epoch": 3.045822102425876,
      "grad_norm": 0.5988819599151611,
      "learning_rate": 0.00041777226119805717,
      "loss": 3.426,
      "step": 28250
    },
    {
      "epoch": 3.0512129380053907,
      "grad_norm": 0.575577974319458,
      "learning_rate": 0.0004174484619535887,
      "loss": 3.4469,
      "step": 28300
    },
    {
      "epoch": 3.056603773584906,
      "grad_norm": 0.6115937829017639,
      "learning_rate": 0.00041712466270912033,
      "loss": 3.4201,
      "step": 28350
    },
    {
      "epoch": 3.0619946091644206,
      "grad_norm": 0.6175453066825867,
      "learning_rate": 0.0004168008634646519,
      "loss": 3.45,
      "step": 28400
    },
    {
      "epoch": 3.0673854447439353,
      "grad_norm": 0.6189521551132202,
      "learning_rate": 0.0004164770642201835,
      "loss": 3.4192,
      "step": 28450
    },
    {
      "epoch": 3.07277628032345,
      "grad_norm": 0.6038718223571777,
      "learning_rate": 0.00041615326497571503,
      "loss": 3.4449,
      "step": 28500
    },
    {
      "epoch": 3.078167115902965,
      "grad_norm": 0.6216437816619873,
      "learning_rate": 0.0004158294657312466,
      "loss": 3.4561,
      "step": 28550
    },
    {
      "epoch": 3.08355795148248,
      "grad_norm": 0.6171916723251343,
      "learning_rate": 0.0004155056664867782,
      "loss": 3.4401,
      "step": 28600
    },
    {
      "epoch": 3.0889487870619945,
      "grad_norm": 0.6580452919006348,
      "learning_rate": 0.0004151818672423097,
      "loss": 3.4528,
      "step": 28650
    },
    {
      "epoch": 3.0943396226415096,
      "grad_norm": 0.5904680490493774,
      "learning_rate": 0.0004148580679978413,
      "loss": 3.4483,
      "step": 28700
    },
    {
      "epoch": 3.0997304582210243,
      "grad_norm": 0.6261874437332153,
      "learning_rate": 0.00041453426875337284,
      "loss": 3.4755,
      "step": 28750
    },
    {
      "epoch": 3.105121293800539,
      "grad_norm": 0.618054986000061,
      "learning_rate": 0.00041421046950890445,
      "loss": 3.4339,
      "step": 28800
    },
    {
      "epoch": 3.1105121293800537,
      "grad_norm": 0.6256803274154663,
      "learning_rate": 0.000413886670264436,
      "loss": 3.4433,
      "step": 28850
    },
    {
      "epoch": 3.115902964959569,
      "grad_norm": 0.5824188590049744,
      "learning_rate": 0.0004135628710199676,
      "loss": 3.4396,
      "step": 28900
    },
    {
      "epoch": 3.1212938005390836,
      "grad_norm": 0.6012649536132812,
      "learning_rate": 0.00041323907177549915,
      "loss": 3.4695,
      "step": 28950
    },
    {
      "epoch": 3.1266846361185983,
      "grad_norm": 0.6452707648277283,
      "learning_rate": 0.0004129152725310307,
      "loss": 3.4418,
      "step": 29000
    },
    {
      "epoch": 3.1266846361185983,
      "eval_accuracy": 0.37053557507420176,
      "eval_loss": 3.497108221054077,
      "eval_runtime": 185.6539,
      "eval_samples_per_second": 97.014,
      "eval_steps_per_second": 6.065,
      "step": 29000
    },
    {
      "epoch": 3.1320754716981134,
      "grad_norm": 0.6206106543540955,
      "learning_rate": 0.0004125914732865623,
      "loss": 3.4605,
      "step": 29050
    },
    {
      "epoch": 3.137466307277628,
      "grad_norm": 0.5886445045471191,
      "learning_rate": 0.00041226767404209386,
      "loss": 3.4591,
      "step": 29100
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 0.6129039525985718,
      "learning_rate": 0.00041194387479762546,
      "loss": 3.4516,
      "step": 29150
    },
    {
      "epoch": 3.1482479784366575,
      "grad_norm": 0.6028342247009277,
      "learning_rate": 0.000411620075553157,
      "loss": 3.46,
      "step": 29200
    },
    {
      "epoch": 3.1536388140161726,
      "grad_norm": 0.5974977612495422,
      "learning_rate": 0.0004112962763086886,
      "loss": 3.4634,
      "step": 29250
    },
    {
      "epoch": 3.1590296495956873,
      "grad_norm": 0.6546943783760071,
      "learning_rate": 0.0004109724770642201,
      "loss": 3.4667,
      "step": 29300
    },
    {
      "epoch": 3.164420485175202,
      "grad_norm": 0.7015277743339539,
      "learning_rate": 0.00041064867781975177,
      "loss": 3.4572,
      "step": 29350
    },
    {
      "epoch": 3.169811320754717,
      "grad_norm": 0.6459450721740723,
      "learning_rate": 0.00041032487857528327,
      "loss": 3.4515,
      "step": 29400
    },
    {
      "epoch": 3.175202156334232,
      "grad_norm": 0.6202012300491333,
      "learning_rate": 0.0004100010793308148,
      "loss": 3.4346,
      "step": 29450
    },
    {
      "epoch": 3.1805929919137466,
      "grad_norm": 0.6145125031471252,
      "learning_rate": 0.0004096772800863464,
      "loss": 3.4599,
      "step": 29500
    },
    {
      "epoch": 3.1859838274932613,
      "grad_norm": 0.6208085417747498,
      "learning_rate": 0.0004093599568267674,
      "loss": 3.4478,
      "step": 29550
    },
    {
      "epoch": 3.1913746630727764,
      "grad_norm": 0.6099492907524109,
      "learning_rate": 0.0004090361575822989,
      "loss": 3.4527,
      "step": 29600
    },
    {
      "epoch": 3.196765498652291,
      "grad_norm": 0.6475285291671753,
      "learning_rate": 0.00040871235833783053,
      "loss": 3.4711,
      "step": 29650
    },
    {
      "epoch": 3.202156334231806,
      "grad_norm": 0.6295750737190247,
      "learning_rate": 0.0004083885590933621,
      "loss": 3.4526,
      "step": 29700
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 0.6257476806640625,
      "learning_rate": 0.00040806475984889363,
      "loss": 3.4639,
      "step": 29750
    },
    {
      "epoch": 3.2129380053908356,
      "grad_norm": 0.6265885829925537,
      "learning_rate": 0.00040774096060442524,
      "loss": 3.4713,
      "step": 29800
    },
    {
      "epoch": 3.2183288409703503,
      "grad_norm": 0.6368427872657776,
      "learning_rate": 0.0004074171613599568,
      "loss": 3.4778,
      "step": 29850
    },
    {
      "epoch": 3.223719676549865,
      "grad_norm": 0.6148289442062378,
      "learning_rate": 0.0004070933621154884,
      "loss": 3.4449,
      "step": 29900
    },
    {
      "epoch": 3.22911051212938,
      "grad_norm": 0.6030235290527344,
      "learning_rate": 0.00040676956287101994,
      "loss": 3.4455,
      "step": 29950
    },
    {
      "epoch": 3.234501347708895,
      "grad_norm": 0.6139891147613525,
      "learning_rate": 0.00040644576362655155,
      "loss": 3.47,
      "step": 30000
    },
    {
      "epoch": 3.234501347708895,
      "eval_accuracy": 0.37173064818727325,
      "eval_loss": 3.490469455718994,
      "eval_runtime": 186.1537,
      "eval_samples_per_second": 96.753,
      "eval_steps_per_second": 6.049,
      "step": 30000
    },
    {
      "epoch": 3.2398921832884096,
      "grad_norm": 0.6158934235572815,
      "learning_rate": 0.00040612196438208304,
      "loss": 3.4574,
      "step": 30050
    },
    {
      "epoch": 3.2452830188679247,
      "grad_norm": 0.6587799191474915,
      "learning_rate": 0.00040579816513761465,
      "loss": 3.4634,
      "step": 30100
    },
    {
      "epoch": 3.2506738544474394,
      "grad_norm": 0.5982083082199097,
      "learning_rate": 0.0004054743658931462,
      "loss": 3.4684,
      "step": 30150
    },
    {
      "epoch": 3.256064690026954,
      "grad_norm": 0.6240605115890503,
      "learning_rate": 0.00040515056664867775,
      "loss": 3.4522,
      "step": 30200
    },
    {
      "epoch": 3.2614555256064692,
      "grad_norm": 0.6786925792694092,
      "learning_rate": 0.00040482676740420935,
      "loss": 3.458,
      "step": 30250
    },
    {
      "epoch": 3.266846361185984,
      "grad_norm": 0.6130402088165283,
      "learning_rate": 0.0004045029681597409,
      "loss": 3.4428,
      "step": 30300
    },
    {
      "epoch": 3.2722371967654986,
      "grad_norm": 0.6382995843887329,
      "learning_rate": 0.0004041791689152725,
      "loss": 3.4597,
      "step": 30350
    },
    {
      "epoch": 3.2776280323450133,
      "grad_norm": 0.6383445262908936,
      "learning_rate": 0.00040385536967080406,
      "loss": 3.4696,
      "step": 30400
    },
    {
      "epoch": 3.2830188679245285,
      "grad_norm": 0.6179215908050537,
      "learning_rate": 0.00040353157042633567,
      "loss": 3.4591,
      "step": 30450
    },
    {
      "epoch": 3.288409703504043,
      "grad_norm": 0.6441600918769836,
      "learning_rate": 0.0004032077711818672,
      "loss": 3.4795,
      "step": 30500
    },
    {
      "epoch": 3.293800539083558,
      "grad_norm": 0.603977620601654,
      "learning_rate": 0.0004028839719373988,
      "loss": 3.47,
      "step": 30550
    },
    {
      "epoch": 3.2991913746630726,
      "grad_norm": 0.5777272582054138,
      "learning_rate": 0.00040256017269293037,
      "loss": 3.4623,
      "step": 30600
    },
    {
      "epoch": 3.3045822102425877,
      "grad_norm": 0.7024667859077454,
      "learning_rate": 0.00040223637344846187,
      "loss": 3.4629,
      "step": 30650
    },
    {
      "epoch": 3.3099730458221024,
      "grad_norm": 0.6607717871665955,
      "learning_rate": 0.00040191257420399347,
      "loss": 3.4535,
      "step": 30700
    },
    {
      "epoch": 3.315363881401617,
      "grad_norm": 0.6349992752075195,
      "learning_rate": 0.000401588774959525,
      "loss": 3.459,
      "step": 30750
    },
    {
      "epoch": 3.3207547169811322,
      "grad_norm": 0.6338915228843689,
      "learning_rate": 0.00040126497571505663,
      "loss": 3.4629,
      "step": 30800
    },
    {
      "epoch": 3.326145552560647,
      "grad_norm": 0.5931634902954102,
      "learning_rate": 0.0004009411764705882,
      "loss": 3.4538,
      "step": 30850
    },
    {
      "epoch": 3.3315363881401616,
      "grad_norm": 0.6514967083930969,
      "learning_rate": 0.0004006173772261198,
      "loss": 3.4761,
      "step": 30900
    },
    {
      "epoch": 3.3369272237196768,
      "grad_norm": 0.6070339679718018,
      "learning_rate": 0.00040029357798165133,
      "loss": 3.4592,
      "step": 30950
    },
    {
      "epoch": 3.3423180592991915,
      "grad_norm": 0.5832754969596863,
      "learning_rate": 0.00039996977873718294,
      "loss": 3.4761,
      "step": 31000
    },
    {
      "epoch": 3.3423180592991915,
      "eval_accuracy": 0.37260769429771173,
      "eval_loss": 3.483612298965454,
      "eval_runtime": 186.2865,
      "eval_samples_per_second": 96.684,
      "eval_steps_per_second": 6.044,
      "step": 31000
    },
    {
      "epoch": 3.347708894878706,
      "grad_norm": 0.6117115616798401,
      "learning_rate": 0.0003996459794927145,
      "loss": 3.4521,
      "step": 31050
    },
    {
      "epoch": 3.353099730458221,
      "grad_norm": 0.585542619228363,
      "learning_rate": 0.00039932218024824604,
      "loss": 3.4714,
      "step": 31100
    },
    {
      "epoch": 3.358490566037736,
      "grad_norm": 0.6312812566757202,
      "learning_rate": 0.00039899838100377764,
      "loss": 3.4496,
      "step": 31150
    },
    {
      "epoch": 3.3638814016172507,
      "grad_norm": 0.5648176670074463,
      "learning_rate": 0.0003986745817593092,
      "loss": 3.4531,
      "step": 31200
    },
    {
      "epoch": 3.3692722371967654,
      "grad_norm": 0.6473795771598816,
      "learning_rate": 0.0003983507825148408,
      "loss": 3.4693,
      "step": 31250
    },
    {
      "epoch": 3.37466307277628,
      "grad_norm": 0.606723964214325,
      "learning_rate": 0.00039802698327037235,
      "loss": 3.4553,
      "step": 31300
    },
    {
      "epoch": 3.3800539083557952,
      "grad_norm": 0.6226693391799927,
      "learning_rate": 0.00039770318402590396,
      "loss": 3.4419,
      "step": 31350
    },
    {
      "epoch": 3.38544474393531,
      "grad_norm": 0.6616231799125671,
      "learning_rate": 0.00039737938478143545,
      "loss": 3.4517,
      "step": 31400
    },
    {
      "epoch": 3.3908355795148246,
      "grad_norm": 0.5993637442588806,
      "learning_rate": 0.000397055585536967,
      "loss": 3.4709,
      "step": 31450
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.6602583527565002,
      "learning_rate": 0.0003967317862924986,
      "loss": 3.4721,
      "step": 31500
    },
    {
      "epoch": 3.4016172506738545,
      "grad_norm": 0.6435751914978027,
      "learning_rate": 0.00039640798704803016,
      "loss": 3.4573,
      "step": 31550
    },
    {
      "epoch": 3.407008086253369,
      "grad_norm": 0.5985782742500305,
      "learning_rate": 0.00039608418780356176,
      "loss": 3.4437,
      "step": 31600
    },
    {
      "epoch": 3.4123989218328843,
      "grad_norm": 0.6137441396713257,
      "learning_rate": 0.0003957603885590933,
      "loss": 3.4644,
      "step": 31650
    },
    {
      "epoch": 3.417789757412399,
      "grad_norm": 0.6449898481369019,
      "learning_rate": 0.0003954365893146249,
      "loss": 3.469,
      "step": 31700
    },
    {
      "epoch": 3.4231805929919137,
      "grad_norm": 0.6611258387565613,
      "learning_rate": 0.00039511279007015647,
      "loss": 3.4661,
      "step": 31750
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.726662278175354,
      "learning_rate": 0.0003947954668105774,
      "loss": 3.4528,
      "step": 31800
    },
    {
      "epoch": 3.4339622641509435,
      "grad_norm": 0.6143780946731567,
      "learning_rate": 0.00039447166756610897,
      "loss": 3.4565,
      "step": 31850
    },
    {
      "epoch": 3.439353099730458,
      "grad_norm": 0.622866153717041,
      "learning_rate": 0.0003941478683216406,
      "loss": 3.4671,
      "step": 31900
    },
    {
      "epoch": 3.444743935309973,
      "grad_norm": 0.6208969354629517,
      "learning_rate": 0.0003938240690771721,
      "loss": 3.4502,
      "step": 31950
    },
    {
      "epoch": 3.450134770889488,
      "grad_norm": 0.6072452664375305,
      "learning_rate": 0.00039350026983270373,
      "loss": 3.4649,
      "step": 32000
    },
    {
      "epoch": 3.450134770889488,
      "eval_accuracy": 0.3730710988642405,
      "eval_loss": 3.4763433933258057,
      "eval_runtime": 186.4885,
      "eval_samples_per_second": 96.58,
      "eval_steps_per_second": 6.038,
      "step": 32000
    },
    {
      "epoch": 3.4555256064690028,
      "grad_norm": 0.6111364960670471,
      "learning_rate": 0.00039317647058823523,
      "loss": 3.4865,
      "step": 32050
    },
    {
      "epoch": 3.4609164420485174,
      "grad_norm": 0.6374538540840149,
      "learning_rate": 0.00039285267134376683,
      "loss": 3.4763,
      "step": 32100
    },
    {
      "epoch": 3.466307277628032,
      "grad_norm": 0.661354660987854,
      "learning_rate": 0.0003925288720992984,
      "loss": 3.4483,
      "step": 32150
    },
    {
      "epoch": 3.4716981132075473,
      "grad_norm": 0.6084034442901611,
      "learning_rate": 0.00039220507285482993,
      "loss": 3.462,
      "step": 32200
    },
    {
      "epoch": 3.477088948787062,
      "grad_norm": 0.6169175505638123,
      "learning_rate": 0.00039188127361036154,
      "loss": 3.4514,
      "step": 32250
    },
    {
      "epoch": 3.4824797843665767,
      "grad_norm": 0.6797444224357605,
      "learning_rate": 0.0003915574743658931,
      "loss": 3.4557,
      "step": 32300
    },
    {
      "epoch": 3.487870619946092,
      "grad_norm": 0.6499859094619751,
      "learning_rate": 0.0003912336751214247,
      "loss": 3.4477,
      "step": 32350
    },
    {
      "epoch": 3.4932614555256065,
      "grad_norm": 0.6400001049041748,
      "learning_rate": 0.00039090987587695624,
      "loss": 3.453,
      "step": 32400
    },
    {
      "epoch": 3.498652291105121,
      "grad_norm": 0.5897339582443237,
      "learning_rate": 0.00039058607663248785,
      "loss": 3.4623,
      "step": 32450
    },
    {
      "epoch": 3.5040431266846364,
      "grad_norm": 0.6736600399017334,
      "learning_rate": 0.0003902622773880194,
      "loss": 3.4705,
      "step": 32500
    },
    {
      "epoch": 3.509433962264151,
      "grad_norm": 0.6231984496116638,
      "learning_rate": 0.000389938478143551,
      "loss": 3.4694,
      "step": 32550
    },
    {
      "epoch": 3.5148247978436657,
      "grad_norm": 0.630050539970398,
      "learning_rate": 0.00038961467889908255,
      "loss": 3.4581,
      "step": 32600
    },
    {
      "epoch": 3.5202156334231804,
      "grad_norm": 0.6342304348945618,
      "learning_rate": 0.00038929087965461405,
      "loss": 3.4617,
      "step": 32650
    },
    {
      "epoch": 3.525606469002695,
      "grad_norm": 0.6055838465690613,
      "learning_rate": 0.00038896708041014566,
      "loss": 3.4754,
      "step": 32700
    },
    {
      "epoch": 3.5309973045822103,
      "grad_norm": 0.6358813047409058,
      "learning_rate": 0.0003886432811656772,
      "loss": 3.4524,
      "step": 32750
    },
    {
      "epoch": 3.536388140161725,
      "grad_norm": 0.615064263343811,
      "learning_rate": 0.0003883194819212088,
      "loss": 3.4788,
      "step": 32800
    },
    {
      "epoch": 3.5417789757412397,
      "grad_norm": 0.6038545370101929,
      "learning_rate": 0.00038799568267674036,
      "loss": 3.4807,
      "step": 32850
    },
    {
      "epoch": 3.547169811320755,
      "grad_norm": 0.6157904267311096,
      "learning_rate": 0.00038767188343227197,
      "loss": 3.4717,
      "step": 32900
    },
    {
      "epoch": 3.5525606469002695,
      "grad_norm": 0.6286138296127319,
      "learning_rate": 0.0003873480841878035,
      "loss": 3.4596,
      "step": 32950
    },
    {
      "epoch": 3.557951482479784,
      "grad_norm": 0.621459424495697,
      "learning_rate": 0.0003870242849433351,
      "loss": 3.456,
      "step": 33000
    },
    {
      "epoch": 3.557951482479784,
      "eval_accuracy": 0.37320006984207627,
      "eval_loss": 3.4715840816497803,
      "eval_runtime": 186.603,
      "eval_samples_per_second": 96.52,
      "eval_steps_per_second": 6.034,
      "step": 33000
    },
    {
      "epoch": 3.5633423180592994,
      "grad_norm": 0.619020402431488,
      "learning_rate": 0.00038670048569886667,
      "loss": 3.4662,
      "step": 33050
    },
    {
      "epoch": 3.568733153638814,
      "grad_norm": 0.6311852335929871,
      "learning_rate": 0.0003863766864543982,
      "loss": 3.4676,
      "step": 33100
    },
    {
      "epoch": 3.5741239892183287,
      "grad_norm": 0.6569680571556091,
      "learning_rate": 0.00038605288720992983,
      "loss": 3.4707,
      "step": 33150
    },
    {
      "epoch": 3.579514824797844,
      "grad_norm": 0.652968168258667,
      "learning_rate": 0.0003857290879654614,
      "loss": 3.4672,
      "step": 33200
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 0.5941240191459656,
      "learning_rate": 0.000385405288720993,
      "loss": 3.4523,
      "step": 33250
    },
    {
      "epoch": 3.5902964959568733,
      "grad_norm": 0.6448426246643066,
      "learning_rate": 0.00038508148947652453,
      "loss": 3.4765,
      "step": 33300
    },
    {
      "epoch": 3.595687331536388,
      "grad_norm": 0.623894453048706,
      "learning_rate": 0.00038475769023205614,
      "loss": 3.4635,
      "step": 33350
    },
    {
      "epoch": 3.601078167115903,
      "grad_norm": 0.6404545307159424,
      "learning_rate": 0.00038443389098758763,
      "loss": 3.4621,
      "step": 33400
    },
    {
      "epoch": 3.606469002695418,
      "grad_norm": 0.6643368005752563,
      "learning_rate": 0.00038411009174311924,
      "loss": 3.4779,
      "step": 33450
    },
    {
      "epoch": 3.6118598382749325,
      "grad_norm": 0.5894840955734253,
      "learning_rate": 0.0003837862924986508,
      "loss": 3.4547,
      "step": 33500
    },
    {
      "epoch": 3.617250673854447,
      "grad_norm": 0.6586086750030518,
      "learning_rate": 0.00038346249325418234,
      "loss": 3.4542,
      "step": 33550
    },
    {
      "epoch": 3.6226415094339623,
      "grad_norm": 0.667512059211731,
      "learning_rate": 0.00038313869400971395,
      "loss": 3.4488,
      "step": 33600
    },
    {
      "epoch": 3.628032345013477,
      "grad_norm": 0.6525899171829224,
      "learning_rate": 0.0003828148947652455,
      "loss": 3.4643,
      "step": 33650
    },
    {
      "epoch": 3.6334231805929917,
      "grad_norm": 0.6395457983016968,
      "learning_rate": 0.0003824910955207771,
      "loss": 3.4696,
      "step": 33700
    },
    {
      "epoch": 3.638814016172507,
      "grad_norm": 0.6152315735816956,
      "learning_rate": 0.00038216729627630865,
      "loss": 3.4498,
      "step": 33750
    },
    {
      "epoch": 3.6442048517520216,
      "grad_norm": 0.6241964101791382,
      "learning_rate": 0.00038184349703184026,
      "loss": 3.4541,
      "step": 33800
    },
    {
      "epoch": 3.6495956873315363,
      "grad_norm": 0.6240250468254089,
      "learning_rate": 0.0003815196977873718,
      "loss": 3.4388,
      "step": 33850
    },
    {
      "epoch": 3.6549865229110514,
      "grad_norm": 0.6336662173271179,
      "learning_rate": 0.0003811958985429034,
      "loss": 3.4326,
      "step": 33900
    },
    {
      "epoch": 3.660377358490566,
      "grad_norm": 0.6259089112281799,
      "learning_rate": 0.00038087209929843496,
      "loss": 3.458,
      "step": 33950
    },
    {
      "epoch": 3.665768194070081,
      "grad_norm": 0.6685935854911804,
      "learning_rate": 0.00038054830005396646,
      "loss": 3.4552,
      "step": 34000
    },
    {
      "epoch": 3.665768194070081,
      "eval_accuracy": 0.3743405992054649,
      "eval_loss": 3.4655990600585938,
      "eval_runtime": 185.5692,
      "eval_samples_per_second": 97.058,
      "eval_steps_per_second": 6.068,
      "step": 34000
    },
    {
      "epoch": 3.671159029649596,
      "grad_norm": 0.7009271383285522,
      "learning_rate": 0.00038022450080949806,
      "loss": 3.456,
      "step": 34050
    },
    {
      "epoch": 3.6765498652291106,
      "grad_norm": 0.6516711711883545,
      "learning_rate": 0.0003799007015650296,
      "loss": 3.4562,
      "step": 34100
    },
    {
      "epoch": 3.6819407008086253,
      "grad_norm": 0.6520816087722778,
      "learning_rate": 0.0003795769023205612,
      "loss": 3.4495,
      "step": 34150
    },
    {
      "epoch": 3.68733153638814,
      "grad_norm": 0.6681199669837952,
      "learning_rate": 0.00037925310307609277,
      "loss": 3.4625,
      "step": 34200
    },
    {
      "epoch": 3.6927223719676547,
      "grad_norm": 0.6544855237007141,
      "learning_rate": 0.0003789293038316244,
      "loss": 3.4528,
      "step": 34250
    },
    {
      "epoch": 3.69811320754717,
      "grad_norm": 0.6212813258171082,
      "learning_rate": 0.0003786055045871559,
      "loss": 3.4579,
      "step": 34300
    },
    {
      "epoch": 3.7035040431266846,
      "grad_norm": 0.673931896686554,
      "learning_rate": 0.0003782817053426875,
      "loss": 3.4549,
      "step": 34350
    },
    {
      "epoch": 3.7088948787061993,
      "grad_norm": 0.6321856379508972,
      "learning_rate": 0.0003779579060982191,
      "loss": 3.4634,
      "step": 34400
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 0.635897696018219,
      "learning_rate": 0.00037764058283864003,
      "loss": 3.4341,
      "step": 34450
    },
    {
      "epoch": 3.719676549865229,
      "grad_norm": 0.6528010964393616,
      "learning_rate": 0.0003773167835941716,
      "loss": 3.4754,
      "step": 34500
    },
    {
      "epoch": 3.725067385444744,
      "grad_norm": 0.6384183168411255,
      "learning_rate": 0.0003769929843497032,
      "loss": 3.4525,
      "step": 34550
    },
    {
      "epoch": 3.730458221024259,
      "grad_norm": 0.6027995347976685,
      "learning_rate": 0.00037666918510523474,
      "loss": 3.4561,
      "step": 34600
    },
    {
      "epoch": 3.7358490566037736,
      "grad_norm": 0.6157944798469543,
      "learning_rate": 0.00037634538586076634,
      "loss": 3.4602,
      "step": 34650
    },
    {
      "epoch": 3.7412398921832883,
      "grad_norm": 0.6281530857086182,
      "learning_rate": 0.00037602158661629784,
      "loss": 3.4652,
      "step": 34700
    },
    {
      "epoch": 3.7466307277628035,
      "grad_norm": 0.6518787145614624,
      "learning_rate": 0.0003756977873718294,
      "loss": 3.4708,
      "step": 34750
    },
    {
      "epoch": 3.752021563342318,
      "grad_norm": 0.602342426776886,
      "learning_rate": 0.000375373988127361,
      "loss": 3.4545,
      "step": 34800
    },
    {
      "epoch": 3.757412398921833,
      "grad_norm": 0.6385865211486816,
      "learning_rate": 0.00037505018888289254,
      "loss": 3.4642,
      "step": 34850
    },
    {
      "epoch": 3.7628032345013476,
      "grad_norm": 0.6197490692138672,
      "learning_rate": 0.00037472638963842415,
      "loss": 3.4549,
      "step": 34900
    },
    {
      "epoch": 3.7681940700808623,
      "grad_norm": 0.6509697437286377,
      "learning_rate": 0.0003744025903939557,
      "loss": 3.4502,
      "step": 34950
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.6715763807296753,
      "learning_rate": 0.0003740787911494873,
      "loss": 3.4719,
      "step": 35000
    },
    {
      "epoch": 3.7735849056603774,
      "eval_accuracy": 0.3748824511569957,
      "eval_loss": 3.459099292755127,
      "eval_runtime": 186.0396,
      "eval_samples_per_second": 96.813,
      "eval_steps_per_second": 6.052,
      "step": 35000
    },
    {
      "epoch": 3.778975741239892,
      "grad_norm": 0.6326753497123718,
      "learning_rate": 0.00037375499190501885,
      "loss": 3.4577,
      "step": 35050
    },
    {
      "epoch": 3.784366576819407,
      "grad_norm": 0.6375369429588318,
      "learning_rate": 0.0003734311926605504,
      "loss": 3.4588,
      "step": 35100
    },
    {
      "epoch": 3.789757412398922,
      "grad_norm": 0.6651833653450012,
      "learning_rate": 0.000373107393416082,
      "loss": 3.4441,
      "step": 35150
    },
    {
      "epoch": 3.7951482479784366,
      "grad_norm": 0.6140313744544983,
      "learning_rate": 0.00037278359417161356,
      "loss": 3.4485,
      "step": 35200
    },
    {
      "epoch": 3.8005390835579513,
      "grad_norm": 0.6524780988693237,
      "learning_rate": 0.00037245979492714517,
      "loss": 3.4576,
      "step": 35250
    },
    {
      "epoch": 3.8059299191374665,
      "grad_norm": 0.642940104007721,
      "learning_rate": 0.0003721359956826767,
      "loss": 3.4505,
      "step": 35300
    },
    {
      "epoch": 3.811320754716981,
      "grad_norm": 0.6456888914108276,
      "learning_rate": 0.0003718121964382083,
      "loss": 3.4707,
      "step": 35350
    },
    {
      "epoch": 3.816711590296496,
      "grad_norm": 0.5969304442405701,
      "learning_rate": 0.0003714883971937398,
      "loss": 3.4509,
      "step": 35400
    },
    {
      "epoch": 3.822102425876011,
      "grad_norm": 0.6570595502853394,
      "learning_rate": 0.0003711645979492714,
      "loss": 3.4573,
      "step": 35450
    },
    {
      "epoch": 3.8274932614555257,
      "grad_norm": 0.6479651927947998,
      "learning_rate": 0.00037084079870480297,
      "loss": 3.4496,
      "step": 35500
    },
    {
      "epoch": 3.8328840970350404,
      "grad_norm": 0.6049849987030029,
      "learning_rate": 0.0003705169994603345,
      "loss": 3.4529,
      "step": 35550
    },
    {
      "epoch": 3.838274932614555,
      "grad_norm": 0.6642239093780518,
      "learning_rate": 0.00037019320021586613,
      "loss": 3.4626,
      "step": 35600
    },
    {
      "epoch": 3.8436657681940702,
      "grad_norm": 0.6449977159500122,
      "learning_rate": 0.0003698694009713977,
      "loss": 3.4487,
      "step": 35650
    },
    {
      "epoch": 3.849056603773585,
      "grad_norm": 0.6435646414756775,
      "learning_rate": 0.0003695456017269293,
      "loss": 3.4493,
      "step": 35700
    },
    {
      "epoch": 3.8544474393530996,
      "grad_norm": 0.5950155258178711,
      "learning_rate": 0.00036922180248246083,
      "loss": 3.4587,
      "step": 35750
    },
    {
      "epoch": 3.8598382749326143,
      "grad_norm": 0.6295285224914551,
      "learning_rate": 0.00036889800323799244,
      "loss": 3.4448,
      "step": 35800
    },
    {
      "epoch": 3.8652291105121295,
      "grad_norm": 0.6502915620803833,
      "learning_rate": 0.000368574203993524,
      "loss": 3.4552,
      "step": 35850
    },
    {
      "epoch": 3.870619946091644,
      "grad_norm": 0.6561721563339233,
      "learning_rate": 0.0003682504047490556,
      "loss": 3.4666,
      "step": 35900
    },
    {
      "epoch": 3.876010781671159,
      "grad_norm": 0.6062862277030945,
      "learning_rate": 0.00036792660550458714,
      "loss": 3.4686,
      "step": 35950
    },
    {
      "epoch": 3.881401617250674,
      "grad_norm": 0.6284040212631226,
      "learning_rate": 0.00036760280626011864,
      "loss": 3.4387,
      "step": 36000
    },
    {
      "epoch": 3.881401617250674,
      "eval_accuracy": 0.3754272367364975,
      "eval_loss": 3.4520866870880127,
      "eval_runtime": 185.7016,
      "eval_samples_per_second": 96.989,
      "eval_steps_per_second": 6.063,
      "step": 36000
    },
    {
      "epoch": 3.8867924528301887,
      "grad_norm": 0.6462699770927429,
      "learning_rate": 0.0003672790070156503,
      "loss": 3.4757,
      "step": 36050
    },
    {
      "epoch": 3.8921832884097034,
      "grad_norm": 0.6437435746192932,
      "learning_rate": 0.0003669552077711818,
      "loss": 3.4396,
      "step": 36100
    },
    {
      "epoch": 3.8975741239892185,
      "grad_norm": 0.6968255043029785,
      "learning_rate": 0.0003666314085267134,
      "loss": 3.4475,
      "step": 36150
    },
    {
      "epoch": 3.9029649595687332,
      "grad_norm": 0.6313275694847107,
      "learning_rate": 0.00036630760928224495,
      "loss": 3.4507,
      "step": 36200
    },
    {
      "epoch": 3.908355795148248,
      "grad_norm": 0.6161664724349976,
      "learning_rate": 0.00036598381003777656,
      "loss": 3.4589,
      "step": 36250
    },
    {
      "epoch": 3.913746630727763,
      "grad_norm": 0.6281296610832214,
      "learning_rate": 0.0003656600107933081,
      "loss": 3.4604,
      "step": 36300
    },
    {
      "epoch": 3.9191374663072778,
      "grad_norm": 0.6347154974937439,
      "learning_rate": 0.0003653362115488397,
      "loss": 3.4742,
      "step": 36350
    },
    {
      "epoch": 3.9245283018867925,
      "grad_norm": 0.6264120936393738,
      "learning_rate": 0.00036501241230437126,
      "loss": 3.4417,
      "step": 36400
    },
    {
      "epoch": 3.929919137466307,
      "grad_norm": 0.6858906745910645,
      "learning_rate": 0.0003646886130599028,
      "loss": 3.4477,
      "step": 36450
    },
    {
      "epoch": 3.935309973045822,
      "grad_norm": 0.6262073516845703,
      "learning_rate": 0.0003643648138154344,
      "loss": 3.4607,
      "step": 36500
    },
    {
      "epoch": 3.940700808625337,
      "grad_norm": 0.6087321043014526,
      "learning_rate": 0.00036404101457096597,
      "loss": 3.4752,
      "step": 36550
    },
    {
      "epoch": 3.9460916442048517,
      "grad_norm": 0.6700966954231262,
      "learning_rate": 0.00036371721532649757,
      "loss": 3.4274,
      "step": 36600
    },
    {
      "epoch": 3.9514824797843664,
      "grad_norm": 0.6734328866004944,
      "learning_rate": 0.0003633934160820291,
      "loss": 3.4459,
      "step": 36650
    },
    {
      "epoch": 3.9568733153638815,
      "grad_norm": 0.7123139500617981,
      "learning_rate": 0.00036306961683756073,
      "loss": 3.4517,
      "step": 36700
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 0.6552482843399048,
      "learning_rate": 0.0003627458175930922,
      "loss": 3.4328,
      "step": 36750
    },
    {
      "epoch": 3.967654986522911,
      "grad_norm": 0.5720398426055908,
      "learning_rate": 0.0003624220183486238,
      "loss": 3.4444,
      "step": 36800
    },
    {
      "epoch": 3.973045822102426,
      "grad_norm": 0.621894359588623,
      "learning_rate": 0.0003620982191041554,
      "loss": 3.4538,
      "step": 36850
    },
    {
      "epoch": 3.9784366576819408,
      "grad_norm": 0.6294187307357788,
      "learning_rate": 0.00036177441985968693,
      "loss": 3.4514,
      "step": 36900
    },
    {
      "epoch": 3.9838274932614555,
      "grad_norm": 0.598534882068634,
      "learning_rate": 0.00036145062061521854,
      "loss": 3.432,
      "step": 36950
    },
    {
      "epoch": 3.9892183288409706,
      "grad_norm": 0.6319286227226257,
      "learning_rate": 0.0003611268213707501,
      "loss": 3.4191,
      "step": 37000
    },
    {
      "epoch": 3.9892183288409706,
      "eval_accuracy": 0.3759199775827362,
      "eval_loss": 3.4474434852600098,
      "eval_runtime": 185.7735,
      "eval_samples_per_second": 96.951,
      "eval_steps_per_second": 6.061,
      "step": 37000
    },
    {
      "epoch": 3.9946091644204853,
      "grad_norm": 0.6325300335884094,
      "learning_rate": 0.0003608030221262817,
      "loss": 3.4729,
      "step": 37050
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.3337475061416626,
      "learning_rate": 0.00036047922288181324,
      "loss": 3.4358,
      "step": 37100
    },
    {
      "epoch": 4.005390835579515,
      "grad_norm": 0.6611919403076172,
      "learning_rate": 0.00036015542363734485,
      "loss": 3.3648,
      "step": 37150
    },
    {
      "epoch": 4.010781671159029,
      "grad_norm": 0.6550231575965881,
      "learning_rate": 0.0003598316243928764,
      "loss": 3.3604,
      "step": 37200
    },
    {
      "epoch": 4.0161725067385445,
      "grad_norm": 0.6367225646972656,
      "learning_rate": 0.00035950782514840795,
      "loss": 3.3585,
      "step": 37250
    },
    {
      "epoch": 4.02156334231806,
      "grad_norm": 0.6150023341178894,
      "learning_rate": 0.00035918402590393955,
      "loss": 3.3617,
      "step": 37300
    },
    {
      "epoch": 4.026954177897574,
      "grad_norm": 0.6218328475952148,
      "learning_rate": 0.00035886022665947105,
      "loss": 3.3433,
      "step": 37350
    },
    {
      "epoch": 4.032345013477089,
      "grad_norm": 0.5891315937042236,
      "learning_rate": 0.0003585364274150027,
      "loss": 3.3488,
      "step": 37400
    },
    {
      "epoch": 4.037735849056604,
      "grad_norm": 0.6458793878555298,
      "learning_rate": 0.0003582126281705342,
      "loss": 3.3586,
      "step": 37450
    },
    {
      "epoch": 4.0431266846361185,
      "grad_norm": 0.6779893040657043,
      "learning_rate": 0.0003578888289260658,
      "loss": 3.3562,
      "step": 37500
    },
    {
      "epoch": 4.048517520215634,
      "grad_norm": 1.0542553663253784,
      "learning_rate": 0.00035756502968159736,
      "loss": 3.3662,
      "step": 37550
    },
    {
      "epoch": 4.053908355795148,
      "grad_norm": 0.7193336486816406,
      "learning_rate": 0.00035724123043712896,
      "loss": 3.3626,
      "step": 37600
    },
    {
      "epoch": 4.059299191374663,
      "grad_norm": 0.6875600218772888,
      "learning_rate": 0.0003569174311926605,
      "loss": 3.3683,
      "step": 37650
    },
    {
      "epoch": 4.064690026954178,
      "grad_norm": 0.6467178463935852,
      "learning_rate": 0.00035659363194819206,
      "loss": 3.3686,
      "step": 37700
    },
    {
      "epoch": 4.070080862533692,
      "grad_norm": 0.5828800797462463,
      "learning_rate": 0.00035626983270372367,
      "loss": 3.3742,
      "step": 37750
    },
    {
      "epoch": 4.0754716981132075,
      "grad_norm": 0.6709010601043701,
      "learning_rate": 0.0003559460334592552,
      "loss": 3.3674,
      "step": 37800
    },
    {
      "epoch": 4.080862533692723,
      "grad_norm": 0.6646392941474915,
      "learning_rate": 0.0003556222342147868,
      "loss": 3.3727,
      "step": 37850
    },
    {
      "epoch": 4.086253369272237,
      "grad_norm": 0.6608548760414124,
      "learning_rate": 0.0003552984349703184,
      "loss": 3.3827,
      "step": 37900
    },
    {
      "epoch": 4.091644204851752,
      "grad_norm": 0.6172977089881897,
      "learning_rate": 0.00035497463572585,
      "loss": 3.3811,
      "step": 37950
    },
    {
      "epoch": 4.097035040431267,
      "grad_norm": 0.6528482437133789,
      "learning_rate": 0.00035465083648138153,
      "loss": 3.3624,
      "step": 38000
    },
    {
      "epoch": 4.097035040431267,
      "eval_accuracy": 0.3762181211068991,
      "eval_loss": 3.4528567790985107,
      "eval_runtime": 185.7274,
      "eval_samples_per_second": 96.975,
      "eval_steps_per_second": 6.063,
      "step": 38000
    },
    {
      "epoch": 4.1024258760107815,
      "grad_norm": 0.6787949204444885,
      "learning_rate": 0.00035432703723691314,
      "loss": 3.3806,
      "step": 38050
    },
    {
      "epoch": 4.107816711590297,
      "grad_norm": 0.6272634863853455,
      "learning_rate": 0.00035400323799244463,
      "loss": 3.3655,
      "step": 38100
    },
    {
      "epoch": 4.113207547169812,
      "grad_norm": 0.6547327637672424,
      "learning_rate": 0.0003536794387479762,
      "loss": 3.3788,
      "step": 38150
    },
    {
      "epoch": 4.118598382749326,
      "grad_norm": 0.6879836320877075,
      "learning_rate": 0.0003533556395035078,
      "loss": 3.3821,
      "step": 38200
    },
    {
      "epoch": 4.123989218328841,
      "grad_norm": 0.6509316563606262,
      "learning_rate": 0.00035303184025903934,
      "loss": 3.3707,
      "step": 38250
    },
    {
      "epoch": 4.129380053908355,
      "grad_norm": 0.6268203258514404,
      "learning_rate": 0.00035270804101457094,
      "loss": 3.3752,
      "step": 38300
    },
    {
      "epoch": 4.1347708894878705,
      "grad_norm": 0.6729905605316162,
      "learning_rate": 0.0003523842417701025,
      "loss": 3.3579,
      "step": 38350
    },
    {
      "epoch": 4.140161725067386,
      "grad_norm": 0.6678585410118103,
      "learning_rate": 0.0003520604425256341,
      "loss": 3.3793,
      "step": 38400
    },
    {
      "epoch": 4.1455525606469,
      "grad_norm": 0.630298912525177,
      "learning_rate": 0.00035173664328116565,
      "loss": 3.378,
      "step": 38450
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.6240480542182922,
      "learning_rate": 0.00035141284403669725,
      "loss": 3.3717,
      "step": 38500
    },
    {
      "epoch": 4.15633423180593,
      "grad_norm": 0.6326809525489807,
      "learning_rate": 0.00035109552077711815,
      "loss": 3.3701,
      "step": 38550
    },
    {
      "epoch": 4.1617250673854445,
      "grad_norm": 0.6675199866294861,
      "learning_rate": 0.00035077172153264976,
      "loss": 3.3688,
      "step": 38600
    },
    {
      "epoch": 4.16711590296496,
      "grad_norm": 0.6258735656738281,
      "learning_rate": 0.0003504479222881813,
      "loss": 3.3733,
      "step": 38650
    },
    {
      "epoch": 4.172506738544475,
      "grad_norm": 0.6700466871261597,
      "learning_rate": 0.0003501241230437129,
      "loss": 3.3901,
      "step": 38700
    },
    {
      "epoch": 4.177897574123989,
      "grad_norm": 0.6353086829185486,
      "learning_rate": 0.0003498003237992444,
      "loss": 3.3659,
      "step": 38750
    },
    {
      "epoch": 4.183288409703504,
      "grad_norm": 0.6454593539237976,
      "learning_rate": 0.000349476524554776,
      "loss": 3.3805,
      "step": 38800
    },
    {
      "epoch": 4.188679245283019,
      "grad_norm": 0.6273381114006042,
      "learning_rate": 0.00034915272531030756,
      "loss": 3.3904,
      "step": 38850
    },
    {
      "epoch": 4.1940700808625335,
      "grad_norm": 0.628691554069519,
      "learning_rate": 0.0003488289260658391,
      "loss": 3.4075,
      "step": 38900
    },
    {
      "epoch": 4.199460916442049,
      "grad_norm": 0.689763069152832,
      "learning_rate": 0.0003485051268213707,
      "loss": 3.3711,
      "step": 38950
    },
    {
      "epoch": 4.204851752021563,
      "grad_norm": 1.091294527053833,
      "learning_rate": 0.00034818132757690227,
      "loss": 3.3861,
      "step": 39000
    },
    {
      "epoch": 4.204851752021563,
      "eval_accuracy": 0.3769242562248054,
      "eval_loss": 3.447742462158203,
      "eval_runtime": 186.0302,
      "eval_samples_per_second": 96.818,
      "eval_steps_per_second": 6.053,
      "step": 39000
    },
    {
      "epoch": 4.210242587601078,
      "grad_norm": 0.6612212657928467,
      "learning_rate": 0.0003478575283324339,
      "loss": 3.3949,
      "step": 39050
    },
    {
      "epoch": 4.215633423180593,
      "grad_norm": 0.7378313541412354,
      "learning_rate": 0.0003475337290879654,
      "loss": 3.3936,
      "step": 39100
    },
    {
      "epoch": 4.2210242587601075,
      "grad_norm": 0.6191502809524536,
      "learning_rate": 0.00034720992984349703,
      "loss": 3.3829,
      "step": 39150
    },
    {
      "epoch": 4.226415094339623,
      "grad_norm": 0.7121553421020508,
      "learning_rate": 0.0003468861305990286,
      "loss": 3.3724,
      "step": 39200
    },
    {
      "epoch": 4.231805929919138,
      "grad_norm": 0.6000710129737854,
      "learning_rate": 0.0003465623313545602,
      "loss": 3.3759,
      "step": 39250
    },
    {
      "epoch": 4.237196765498652,
      "grad_norm": 0.6634258031845093,
      "learning_rate": 0.00034623853211009173,
      "loss": 3.3785,
      "step": 39300
    },
    {
      "epoch": 4.242587601078167,
      "grad_norm": 0.6099295020103455,
      "learning_rate": 0.0003459212088505127,
      "loss": 3.3904,
      "step": 39350
    },
    {
      "epoch": 4.247978436657682,
      "grad_norm": 0.7189794182777405,
      "learning_rate": 0.0003455974096060442,
      "loss": 3.3855,
      "step": 39400
    },
    {
      "epoch": 4.2533692722371965,
      "grad_norm": 0.6642246842384338,
      "learning_rate": 0.0003452736103615758,
      "loss": 3.3947,
      "step": 39450
    },
    {
      "epoch": 4.258760107816712,
      "grad_norm": 0.6335538625717163,
      "learning_rate": 0.00034494981111710734,
      "loss": 3.382,
      "step": 39500
    },
    {
      "epoch": 4.264150943396227,
      "grad_norm": 0.6120173335075378,
      "learning_rate": 0.00034462601187263894,
      "loss": 3.3705,
      "step": 39550
    },
    {
      "epoch": 4.269541778975741,
      "grad_norm": 0.6543063521385193,
      "learning_rate": 0.0003443022126281705,
      "loss": 3.3853,
      "step": 39600
    },
    {
      "epoch": 4.274932614555256,
      "grad_norm": 0.6222617030143738,
      "learning_rate": 0.00034397841338370204,
      "loss": 3.3943,
      "step": 39650
    },
    {
      "epoch": 4.280323450134771,
      "grad_norm": 0.6919296979904175,
      "learning_rate": 0.00034365461413923365,
      "loss": 3.3755,
      "step": 39700
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 0.6370746493339539,
      "learning_rate": 0.0003433308148947652,
      "loss": 3.3996,
      "step": 39750
    },
    {
      "epoch": 4.291105121293801,
      "grad_norm": 0.6165392398834229,
      "learning_rate": 0.0003430070156502968,
      "loss": 3.3853,
      "step": 39800
    },
    {
      "epoch": 4.296495956873315,
      "grad_norm": 0.6337375044822693,
      "learning_rate": 0.00034268321640582835,
      "loss": 3.3893,
      "step": 39850
    },
    {
      "epoch": 4.30188679245283,
      "grad_norm": 0.6721922159194946,
      "learning_rate": 0.00034235941716135996,
      "loss": 3.3897,
      "step": 39900
    },
    {
      "epoch": 4.307277628032345,
      "grad_norm": 0.7578305006027222,
      "learning_rate": 0.0003420356179168915,
      "loss": 3.3968,
      "step": 39950
    },
    {
      "epoch": 4.3126684636118595,
      "grad_norm": 0.6459212899208069,
      "learning_rate": 0.0003417118186724231,
      "loss": 3.4024,
      "step": 40000
    },
    {
      "epoch": 4.3126684636118595,
      "eval_accuracy": 0.37721848824500703,
      "eval_loss": 3.4408767223358154,
      "eval_runtime": 185.5969,
      "eval_samples_per_second": 97.044,
      "eval_steps_per_second": 6.067,
      "step": 40000
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.344276717568e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}