{
  "best_metric": 3.3902640342712402,
  "best_model_checkpoint": "bert_tiny_lda_book/checkpoint-350000",
  "epoch": 25.0,
  "eval_steps": 10000,
  "global_step": 355850,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0351271603203597,
      "grad_norm": 2.4042296409606934,
      "learning_rate": 5e-06,
      "loss": 11.8352,
      "step": 500
    },
    {
      "epoch": 0.0702543206407194,
      "grad_norm": 0.9798582196235657,
      "learning_rate": 1e-05,
      "loss": 9.7702,
      "step": 1000
    },
    {
      "epoch": 0.1053814809610791,
      "grad_norm": 0.7924838066101074,
      "learning_rate": 1.5e-05,
      "loss": 8.5368,
      "step": 1500
    },
    {
      "epoch": 0.1405086412814388,
      "grad_norm": 0.6194170117378235,
      "learning_rate": 2e-05,
      "loss": 8.3348,
      "step": 2000
    },
    {
      "epoch": 0.1756358016017985,
      "grad_norm": 0.8334989547729492,
      "learning_rate": 2.5e-05,
      "loss": 8.2443,
      "step": 2500
    },
    {
      "epoch": 0.2107629619221582,
      "grad_norm": 0.7570953369140625,
      "learning_rate": 3e-05,
      "loss": 8.1834,
      "step": 3000
    },
    {
      "epoch": 0.2458901222425179,
      "grad_norm": 0.8500428795814514,
      "learning_rate": 3.5e-05,
      "loss": 8.1446,
      "step": 3500
    },
    {
      "epoch": 0.2810172825628776,
      "grad_norm": 0.6219971776008606,
      "learning_rate": 4e-05,
      "loss": 8.1138,
      "step": 4000
    },
    {
      "epoch": 0.31614444288323734,
      "grad_norm": 0.6914583444595337,
      "learning_rate": 4.5e-05,
      "loss": 8.0879,
      "step": 4500
    },
    {
      "epoch": 0.351271603203597,
      "grad_norm": 0.6896765828132629,
      "learning_rate": 5e-05,
      "loss": 8.0668,
      "step": 5000
    },
    {
      "epoch": 0.38639876352395675,
      "grad_norm": 0.7136889100074768,
      "learning_rate": 5.500000000000001e-05,
      "loss": 8.0429,
      "step": 5500
    },
    {
      "epoch": 0.4215259238443164,
      "grad_norm": 0.805734395980835,
      "learning_rate": 6e-05,
      "loss": 8.0306,
      "step": 6000
    },
    {
      "epoch": 0.45665308416467615,
      "grad_norm": 0.8246288299560547,
      "learning_rate": 6.500000000000001e-05,
      "loss": 8.0146,
      "step": 6500
    },
    {
      "epoch": 0.4917802444850358,
      "grad_norm": 0.7797628045082092,
      "learning_rate": 7e-05,
      "loss": 8.0038,
      "step": 7000
    },
    {
      "epoch": 0.5269074048053956,
      "grad_norm": 0.7447306513786316,
      "learning_rate": 7.500000000000001e-05,
      "loss": 7.9902,
      "step": 7500
    },
    {
      "epoch": 0.5620345651257552,
      "grad_norm": 0.7058584690093994,
      "learning_rate": 8e-05,
      "loss": 7.9779,
      "step": 8000
    },
    {
      "epoch": 0.5971617254461149,
      "grad_norm": 0.7725171446800232,
      "learning_rate": 8.5e-05,
      "loss": 7.9681,
      "step": 8500
    },
    {
      "epoch": 0.6322888857664747,
      "grad_norm": 1.0565847158432007,
      "learning_rate": 9e-05,
      "loss": 7.9563,
      "step": 9000
    },
    {
      "epoch": 0.6674160460868344,
      "grad_norm": 0.5960007309913635,
      "learning_rate": 9.5e-05,
      "loss": 7.9522,
      "step": 9500
    },
    {
      "epoch": 0.702543206407194,
      "grad_norm": 0.6335282921791077,
      "learning_rate": 0.0001,
      "loss": 7.9444,
      "step": 10000
    },
    {
      "epoch": 0.702543206407194,
      "eval_accuracy": 0.16441983733189408,
      "eval_loss": 7.782281875610352,
      "eval_runtime": 132.1614,
      "eval_samples_per_second": 913.322,
      "eval_steps_per_second": 5.713,
      "step": 10000
    },
    {
      "epoch": 0.7376703667275537,
      "grad_norm": 0.7276875376701355,
      "learning_rate": 9.985542865404077e-05,
      "loss": 7.9223,
      "step": 10500
    },
    {
      "epoch": 0.7727975270479135,
      "grad_norm": 0.7996670007705688,
      "learning_rate": 9.971085730808155e-05,
      "loss": 7.8921,
      "step": 11000
    },
    {
      "epoch": 0.8079246873682732,
      "grad_norm": 0.8018574714660645,
      "learning_rate": 9.956628596212231e-05,
      "loss": 7.8549,
      "step": 11500
    },
    {
      "epoch": 0.8430518476886328,
      "grad_norm": 1.1303762197494507,
      "learning_rate": 9.942171461616309e-05,
      "loss": 7.7991,
      "step": 12000
    },
    {
      "epoch": 0.8781790080089925,
      "grad_norm": 1.0644803047180176,
      "learning_rate": 9.927714327020385e-05,
      "loss": 7.7166,
      "step": 12500
    },
    {
      "epoch": 0.9133061683293523,
      "grad_norm": 0.9369208812713623,
      "learning_rate": 9.913257192424462e-05,
      "loss": 7.6594,
      "step": 13000
    },
    {
      "epoch": 0.948433328649712,
      "grad_norm": 1.0729360580444336,
      "learning_rate": 9.89880005782854e-05,
      "loss": 7.6285,
      "step": 13500
    },
    {
      "epoch": 0.9835604889700716,
      "grad_norm": 0.9288055300712585,
      "learning_rate": 9.884342923232616e-05,
      "loss": 7.6188,
      "step": 14000
    },
    {
      "epoch": 1.0186876492904313,
      "grad_norm": 0.8967397809028625,
      "learning_rate": 9.869885788636692e-05,
      "loss": 7.6067,
      "step": 14500
    },
    {
      "epoch": 1.053814809610791,
      "grad_norm": 0.6617307066917419,
      "learning_rate": 9.855428654040769e-05,
      "loss": 7.602,
      "step": 15000
    },
    {
      "epoch": 1.0889419699311507,
      "grad_norm": 0.8365678191184998,
      "learning_rate": 9.840971519444846e-05,
      "loss": 7.5909,
      "step": 15500
    },
    {
      "epoch": 1.1240691302515105,
      "grad_norm": 0.741174042224884,
      "learning_rate": 9.826514384848924e-05,
      "loss": 7.564,
      "step": 16000
    },
    {
      "epoch": 1.1591962905718702,
      "grad_norm": 1.1366857290267944,
      "learning_rate": 9.812057250253e-05,
      "loss": 7.514,
      "step": 16500
    },
    {
      "epoch": 1.1943234508922298,
      "grad_norm": 1.1493654251098633,
      "learning_rate": 9.797600115657077e-05,
      "loss": 7.4461,
      "step": 17000
    },
    {
      "epoch": 1.2294506112125896,
      "grad_norm": 1.6501221656799316,
      "learning_rate": 9.783142981061155e-05,
      "loss": 7.0418,
      "step": 17500
    },
    {
      "epoch": 1.2645777715329491,
      "grad_norm": 1.5519462823867798,
      "learning_rate": 9.768685846465231e-05,
      "loss": 6.4769,
      "step": 18000
    },
    {
      "epoch": 1.299704931853309,
      "grad_norm": 1.3463587760925293,
      "learning_rate": 9.754228711869307e-05,
      "loss": 6.1009,
      "step": 18500
    },
    {
      "epoch": 1.3348320921736687,
      "grad_norm": 1.3636152744293213,
      "learning_rate": 9.739771577273384e-05,
      "loss": 5.7742,
      "step": 19000
    },
    {
      "epoch": 1.3699592524940285,
      "grad_norm": 1.180375099182129,
      "learning_rate": 9.725314442677463e-05,
      "loss": 5.5886,
      "step": 19500
    },
    {
      "epoch": 1.405086412814388,
      "grad_norm": 1.1276618242263794,
      "learning_rate": 9.71085730808154e-05,
      "loss": 5.4537,
      "step": 20000
    },
    {
      "epoch": 1.405086412814388,
      "eval_accuracy": 0.46576995310731434,
      "eval_loss": 5.008655548095703,
      "eval_runtime": 135.4608,
      "eval_samples_per_second": 891.077,
      "eval_steps_per_second": 5.574,
      "step": 20000
    },
    {
      "epoch": 1.4402135731347478,
      "grad_norm": 1.3548939228057861,
      "learning_rate": 9.696400173485616e-05,
      "loss": 5.3526,
      "step": 20500
    },
    {
      "epoch": 1.4753407334551074,
      "grad_norm": 1.2437584400177002,
      "learning_rate": 9.681943038889692e-05,
      "loss": 5.2688,
      "step": 21000
    },
    {
      "epoch": 1.5104678937754672,
      "grad_norm": 1.1224002838134766,
      "learning_rate": 9.66748590429377e-05,
      "loss": 5.1964,
      "step": 21500
    },
    {
      "epoch": 1.545595054095827,
      "grad_norm": 1.071878433227539,
      "learning_rate": 9.653028769697846e-05,
      "loss": 5.1344,
      "step": 22000
    },
    {
      "epoch": 1.5807222144161865,
      "grad_norm": 1.0790518522262573,
      "learning_rate": 9.638571635101923e-05,
      "loss": 5.0876,
      "step": 22500
    },
    {
      "epoch": 1.6158493747365463,
      "grad_norm": 1.0892592668533325,
      "learning_rate": 9.624114500506e-05,
      "loss": 5.0403,
      "step": 23000
    },
    {
      "epoch": 1.650976535056906,
      "grad_norm": 1.170798659324646,
      "learning_rate": 9.609657365910077e-05,
      "loss": 4.9966,
      "step": 23500
    },
    {
      "epoch": 1.6861036953772657,
      "grad_norm": 1.1245207786560059,
      "learning_rate": 9.595200231314155e-05,
      "loss": 4.9514,
      "step": 24000
    },
    {
      "epoch": 1.7212308556976255,
      "grad_norm": 1.046912431716919,
      "learning_rate": 9.580743096718231e-05,
      "loss": 4.9167,
      "step": 24500
    },
    {
      "epoch": 1.7563580160179852,
      "grad_norm": 1.04922616481781,
      "learning_rate": 9.566285962122307e-05,
      "loss": 4.8802,
      "step": 25000
    },
    {
      "epoch": 1.7914851763383448,
      "grad_norm": 1.0708765983581543,
      "learning_rate": 9.551828827526385e-05,
      "loss": 4.8482,
      "step": 25500
    },
    {
      "epoch": 1.8266123366587044,
      "grad_norm": 1.0325334072113037,
      "learning_rate": 9.537371692930462e-05,
      "loss": 4.8153,
      "step": 26000
    },
    {
      "epoch": 1.8617394969790642,
      "grad_norm": 1.1176384687423706,
      "learning_rate": 9.522914558334538e-05,
      "loss": 4.7857,
      "step": 26500
    },
    {
      "epoch": 1.896866657299424,
      "grad_norm": 1.0515133142471313,
      "learning_rate": 9.508457423738616e-05,
      "loss": 4.7617,
      "step": 27000
    },
    {
      "epoch": 1.9319938176197837,
      "grad_norm": 0.9994978308677673,
      "learning_rate": 9.494000289142692e-05,
      "loss": 4.737,
      "step": 27500
    },
    {
      "epoch": 1.9671209779401433,
      "grad_norm": 1.0131473541259766,
      "learning_rate": 9.47954315454677e-05,
      "loss": 4.7113,
      "step": 28000
    },
    {
      "epoch": 2.002248138260503,
      "grad_norm": 1.0716643333435059,
      "learning_rate": 9.465086019950846e-05,
      "loss": 4.6987,
      "step": 28500
    },
    {
      "epoch": 2.0373752985808626,
      "grad_norm": 1.129737138748169,
      "learning_rate": 9.450628885354923e-05,
      "loss": 4.6703,
      "step": 29000
    },
    {
      "epoch": 2.0725024589012224,
      "grad_norm": 1.1458584070205688,
      "learning_rate": 9.436171750758999e-05,
      "loss": 4.6505,
      "step": 29500
    },
    {
      "epoch": 2.107629619221582,
      "grad_norm": 0.9604290723800659,
      "learning_rate": 9.421714616163077e-05,
      "loss": 4.6397,
      "step": 30000
    },
    {
      "epoch": 2.107629619221582,
      "eval_accuracy": 0.5606812094914088,
      "eval_loss": 4.264980792999268,
      "eval_runtime": 133.6788,
      "eval_samples_per_second": 902.955,
      "eval_steps_per_second": 5.648,
      "step": 30000
    },
    {
      "epoch": 2.142756779541942,
      "grad_norm": 1.01162850856781,
      "learning_rate": 9.407257481567155e-05,
      "loss": 4.6205,
      "step": 30500
    },
    {
      "epoch": 2.1778839398623013,
      "grad_norm": 1.0250729322433472,
      "learning_rate": 9.392800346971231e-05,
      "loss": 4.6006,
      "step": 31000
    },
    {
      "epoch": 2.213011100182661,
      "grad_norm": 1.037964940071106,
      "learning_rate": 9.378343212375307e-05,
      "loss": 4.5895,
      "step": 31500
    },
    {
      "epoch": 2.248138260503021,
      "grad_norm": 1.0791294574737549,
      "learning_rate": 9.363886077779385e-05,
      "loss": 4.5752,
      "step": 32000
    },
    {
      "epoch": 2.2832654208233807,
      "grad_norm": 0.9663081169128418,
      "learning_rate": 9.349428943183462e-05,
      "loss": 4.5613,
      "step": 32500
    },
    {
      "epoch": 2.3183925811437405,
      "grad_norm": 1.1115280389785767,
      "learning_rate": 9.334971808587538e-05,
      "loss": 4.5437,
      "step": 33000
    },
    {
      "epoch": 2.3535197414641003,
      "grad_norm": 1.0173557996749878,
      "learning_rate": 9.320514673991614e-05,
      "loss": 4.5312,
      "step": 33500
    },
    {
      "epoch": 2.3886469017844596,
      "grad_norm": 0.9513857364654541,
      "learning_rate": 9.306057539395692e-05,
      "loss": 4.5205,
      "step": 34000
    },
    {
      "epoch": 2.4237740621048194,
      "grad_norm": 0.9750761389732361,
      "learning_rate": 9.29160040479977e-05,
      "loss": 4.5038,
      "step": 34500
    },
    {
      "epoch": 2.458901222425179,
      "grad_norm": 1.0778580904006958,
      "learning_rate": 9.277143270203846e-05,
      "loss": 4.4922,
      "step": 35000
    },
    {
      "epoch": 2.494028382745539,
      "grad_norm": 1.0201412439346313,
      "learning_rate": 9.262686135607923e-05,
      "loss": 4.4831,
      "step": 35500
    },
    {
      "epoch": 2.5291555430658983,
      "grad_norm": 1.0821698904037476,
      "learning_rate": 9.248229001012e-05,
      "loss": 4.4723,
      "step": 36000
    },
    {
      "epoch": 2.564282703386258,
      "grad_norm": 0.990523636341095,
      "learning_rate": 9.233771866416077e-05,
      "loss": 4.4593,
      "step": 36500
    },
    {
      "epoch": 2.599409863706618,
      "grad_norm": 0.9600923657417297,
      "learning_rate": 9.219314731820153e-05,
      "loss": 4.4483,
      "step": 37000
    },
    {
      "epoch": 2.6345370240269776,
      "grad_norm": 1.0416806936264038,
      "learning_rate": 9.20485759722423e-05,
      "loss": 4.4404,
      "step": 37500
    },
    {
      "epoch": 2.6696641843473374,
      "grad_norm": 1.0397828817367554,
      "learning_rate": 9.190400462628307e-05,
      "loss": 4.4294,
      "step": 38000
    },
    {
      "epoch": 2.704791344667697,
      "grad_norm": 0.94615238904953,
      "learning_rate": 9.175943328032385e-05,
      "loss": 4.4186,
      "step": 38500
    },
    {
      "epoch": 2.739918504988057,
      "grad_norm": 1.0596972703933716,
      "learning_rate": 9.161486193436461e-05,
      "loss": 4.4075,
      "step": 39000
    },
    {
      "epoch": 2.7750456653084163,
      "grad_norm": 1.002113938331604,
      "learning_rate": 9.147029058840538e-05,
      "loss": 4.4017,
      "step": 39500
    },
    {
      "epoch": 2.810172825628776,
      "grad_norm": 1.009598970413208,
      "learning_rate": 9.132571924244616e-05,
      "loss": 4.3898,
      "step": 40000
    },
    {
      "epoch": 2.810172825628776,
      "eval_accuracy": 0.5915942292045733,
      "eval_loss": 4.037876129150391,
      "eval_runtime": 136.3883,
      "eval_samples_per_second": 885.017,
      "eval_steps_per_second": 5.536,
      "step": 40000
    },
    {
      "epoch": 2.845299985949136,
      "grad_norm": 1.0175772905349731,
      "learning_rate": 9.118114789648692e-05,
      "loss": 4.3774,
      "step": 40500
    },
    {
      "epoch": 2.8804271462694957,
      "grad_norm": 0.9904564023017883,
      "learning_rate": 9.103657655052768e-05,
      "loss": 4.3698,
      "step": 41000
    },
    {
      "epoch": 2.915554306589855,
      "grad_norm": 0.9531894326210022,
      "learning_rate": 9.089200520456845e-05,
      "loss": 4.3605,
      "step": 41500
    },
    {
      "epoch": 2.950681466910215,
      "grad_norm": 1.0265958309173584,
      "learning_rate": 9.074743385860923e-05,
      "loss": 4.3564,
      "step": 42000
    },
    {
      "epoch": 2.9858086272305746,
      "grad_norm": 1.0228315591812134,
      "learning_rate": 9.060286251265e-05,
      "loss": 4.3459,
      "step": 42500
    },
    {
      "epoch": 3.0209357875509344,
      "grad_norm": 1.1415255069732666,
      "learning_rate": 9.045829116669077e-05,
      "loss": 4.3307,
      "step": 43000
    },
    {
      "epoch": 3.056062947871294,
      "grad_norm": 1.043325662612915,
      "learning_rate": 9.031371982073153e-05,
      "loss": 4.3277,
      "step": 43500
    },
    {
      "epoch": 3.091190108191654,
      "grad_norm": 1.03691565990448,
      "learning_rate": 9.016914847477231e-05,
      "loss": 4.3218,
      "step": 44000
    },
    {
      "epoch": 3.1263172685120133,
      "grad_norm": 1.024833083152771,
      "learning_rate": 9.002457712881307e-05,
      "loss": 4.3112,
      "step": 44500
    },
    {
      "epoch": 3.161444428832373,
      "grad_norm": 1.0092805624008179,
      "learning_rate": 8.988000578285384e-05,
      "loss": 4.3064,
      "step": 45000
    },
    {
      "epoch": 3.196571589152733,
      "grad_norm": 1.0393027067184448,
      "learning_rate": 8.973543443689461e-05,
      "loss": 4.296,
      "step": 45500
    },
    {
      "epoch": 3.2316987494730927,
      "grad_norm": 1.0920835733413696,
      "learning_rate": 8.959086309093538e-05,
      "loss": 4.2906,
      "step": 46000
    },
    {
      "epoch": 3.2668259097934524,
      "grad_norm": 1.0164620876312256,
      "learning_rate": 8.944629174497616e-05,
      "loss": 4.283,
      "step": 46500
    },
    {
      "epoch": 3.301953070113812,
      "grad_norm": 0.9901977181434631,
      "learning_rate": 8.930172039901692e-05,
      "loss": 4.2753,
      "step": 47000
    },
    {
      "epoch": 3.3370802304341716,
      "grad_norm": 1.0199934244155884,
      "learning_rate": 8.915714905305768e-05,
      "loss": 4.2695,
      "step": 47500
    },
    {
      "epoch": 3.3722073907545314,
      "grad_norm": 1.081707239151001,
      "learning_rate": 8.901257770709845e-05,
      "loss": 4.2655,
      "step": 48000
    },
    {
      "epoch": 3.407334551074891,
      "grad_norm": 0.9450724720954895,
      "learning_rate": 8.886800636113923e-05,
      "loss": 4.2551,
      "step": 48500
    },
    {
      "epoch": 3.442461711395251,
      "grad_norm": 1.0958073139190674,
      "learning_rate": 8.872343501518e-05,
      "loss": 4.2467,
      "step": 49000
    },
    {
      "epoch": 3.4775888717156107,
      "grad_norm": 0.9844788908958435,
      "learning_rate": 8.857886366922077e-05,
      "loss": 4.2399,
      "step": 49500
    },
    {
      "epoch": 3.51271603203597,
      "grad_norm": 0.9961697459220886,
      "learning_rate": 8.843429232326153e-05,
      "loss": 4.2383,
      "step": 50000
    },
    {
      "epoch": 3.51271603203597,
      "eval_accuracy": 0.6112833055627742,
      "eval_loss": 3.8977701663970947,
      "eval_runtime": 135.3224,
      "eval_samples_per_second": 891.988,
      "eval_steps_per_second": 5.579,
      "step": 50000
    },
    {
      "epoch": 3.54784319235633,
      "grad_norm": 1.0457099676132202,
      "learning_rate": 8.828972097730231e-05,
      "loss": 4.2353,
      "step": 50500
    },
    {
      "epoch": 3.5829703526766896,
      "grad_norm": 1.0435949563980103,
      "learning_rate": 8.814514963134307e-05,
      "loss": 4.2229,
      "step": 51000
    },
    {
      "epoch": 3.6180975129970494,
      "grad_norm": 0.966555118560791,
      "learning_rate": 8.800057828538384e-05,
      "loss": 4.2241,
      "step": 51500
    },
    {
      "epoch": 3.6532246733174087,
      "grad_norm": 1.0896607637405396,
      "learning_rate": 8.78560069394246e-05,
      "loss": 4.216,
      "step": 52000
    },
    {
      "epoch": 3.6883518336377685,
      "grad_norm": 0.986983060836792,
      "learning_rate": 8.771143559346538e-05,
      "loss": 4.2088,
      "step": 52500
    },
    {
      "epoch": 3.7234789939581283,
      "grad_norm": 1.003517746925354,
      "learning_rate": 8.756686424750616e-05,
      "loss": 4.207,
      "step": 53000
    },
    {
      "epoch": 3.758606154278488,
      "grad_norm": 0.9948218464851379,
      "learning_rate": 8.742229290154692e-05,
      "loss": 4.2021,
      "step": 53500
    },
    {
      "epoch": 3.793733314598848,
      "grad_norm": 1.0764601230621338,
      "learning_rate": 8.727772155558768e-05,
      "loss": 4.1926,
      "step": 54000
    },
    {
      "epoch": 3.8288604749192077,
      "grad_norm": 0.9770961999893188,
      "learning_rate": 8.713315020962846e-05,
      "loss": 4.1896,
      "step": 54500
    },
    {
      "epoch": 3.8639876352395675,
      "grad_norm": 0.9867109060287476,
      "learning_rate": 8.698857886366923e-05,
      "loss": 4.1839,
      "step": 55000
    },
    {
      "epoch": 3.899114795559927,
      "grad_norm": 1.0001426935195923,
      "learning_rate": 8.684400751770999e-05,
      "loss": 4.1775,
      "step": 55500
    },
    {
      "epoch": 3.9342419558802866,
      "grad_norm": 1.045431137084961,
      "learning_rate": 8.669943617175075e-05,
      "loss": 4.1746,
      "step": 56000
    },
    {
      "epoch": 3.9693691162006464,
      "grad_norm": 0.982204258441925,
      "learning_rate": 8.655486482579153e-05,
      "loss": 4.172,
      "step": 56500
    },
    {
      "epoch": 4.004496276521006,
      "grad_norm": 1.1098988056182861,
      "learning_rate": 8.641029347983231e-05,
      "loss": 4.1631,
      "step": 57000
    },
    {
      "epoch": 4.0396234368413655,
      "grad_norm": 1.0858879089355469,
      "learning_rate": 8.626572213387307e-05,
      "loss": 4.1566,
      "step": 57500
    },
    {
      "epoch": 4.074750597161725,
      "grad_norm": 1.000654935836792,
      "learning_rate": 8.612115078791384e-05,
      "loss": 4.153,
      "step": 58000
    },
    {
      "epoch": 4.109877757482085,
      "grad_norm": 1.0371450185775757,
      "learning_rate": 8.597657944195461e-05,
      "loss": 4.1496,
      "step": 58500
    },
    {
      "epoch": 4.145004917802445,
      "grad_norm": 0.9899897575378418,
      "learning_rate": 8.583200809599538e-05,
      "loss": 4.1445,
      "step": 59000
    },
    {
      "epoch": 4.180132078122805,
      "grad_norm": 0.9598783254623413,
      "learning_rate": 8.568743675003614e-05,
      "loss": 4.1447,
      "step": 59500
    },
    {
      "epoch": 4.215259238443164,
      "grad_norm": 1.0509892702102661,
      "learning_rate": 8.55428654040769e-05,
      "loss": 4.1379,
      "step": 60000
    },
    {
      "epoch": 4.215259238443164,
      "eval_accuracy": 0.6234003989470186,
      "eval_loss": 3.8116791248321533,
      "eval_runtime": 137.8325,
      "eval_samples_per_second": 875.744,
      "eval_steps_per_second": 5.478,
      "step": 60000
    },
    {
      "epoch": 4.250386398763524,
      "grad_norm": 0.9857525825500488,
      "learning_rate": 8.539829405811768e-05,
      "loss": 4.1377,
      "step": 60500
    },
    {
      "epoch": 4.285513559083884,
      "grad_norm": 1.021316409111023,
      "learning_rate": 8.525372271215846e-05,
      "loss": 4.1296,
      "step": 61000
    },
    {
      "epoch": 4.320640719404244,
      "grad_norm": 0.9933499097824097,
      "learning_rate": 8.510915136619923e-05,
      "loss": 4.1251,
      "step": 61500
    },
    {
      "epoch": 4.355767879724603,
      "grad_norm": 1.018663763999939,
      "learning_rate": 8.496458002023999e-05,
      "loss": 4.1228,
      "step": 62000
    },
    {
      "epoch": 4.3908950400449624,
      "grad_norm": 1.0634220838546753,
      "learning_rate": 8.482000867428075e-05,
      "loss": 4.1238,
      "step": 62500
    },
    {
      "epoch": 4.426022200365322,
      "grad_norm": 1.0175825357437134,
      "learning_rate": 8.467543732832153e-05,
      "loss": 4.1183,
      "step": 63000
    },
    {
      "epoch": 4.461149360685682,
      "grad_norm": 1.0975725650787354,
      "learning_rate": 8.45308659823623e-05,
      "loss": 4.1127,
      "step": 63500
    },
    {
      "epoch": 4.496276521006042,
      "grad_norm": 1.1212193965911865,
      "learning_rate": 8.438629463640307e-05,
      "loss": 4.1103,
      "step": 64000
    },
    {
      "epoch": 4.531403681326402,
      "grad_norm": 1.1338615417480469,
      "learning_rate": 8.424172329044384e-05,
      "loss": 4.1077,
      "step": 64500
    },
    {
      "epoch": 4.566530841646761,
      "grad_norm": 1.099275827407837,
      "learning_rate": 8.409715194448461e-05,
      "loss": 4.1028,
      "step": 65000
    },
    {
      "epoch": 4.601658001967121,
      "grad_norm": 0.9647557139396667,
      "learning_rate": 8.395258059852538e-05,
      "loss": 4.1004,
      "step": 65500
    },
    {
      "epoch": 4.636785162287481,
      "grad_norm": 1.000170111656189,
      "learning_rate": 8.380800925256614e-05,
      "loss": 4.0951,
      "step": 66000
    },
    {
      "epoch": 4.671912322607841,
      "grad_norm": 0.978367805480957,
      "learning_rate": 8.36634379066069e-05,
      "loss": 4.0958,
      "step": 66500
    },
    {
      "epoch": 4.7070394829282005,
      "grad_norm": 1.0026097297668457,
      "learning_rate": 8.351886656064768e-05,
      "loss": 4.0884,
      "step": 67000
    },
    {
      "epoch": 4.742166643248559,
      "grad_norm": 1.0012880563735962,
      "learning_rate": 8.337429521468846e-05,
      "loss": 4.0881,
      "step": 67500
    },
    {
      "epoch": 4.777293803568919,
      "grad_norm": 0.9903104901313782,
      "learning_rate": 8.322972386872923e-05,
      "loss": 4.0849,
      "step": 68000
    },
    {
      "epoch": 4.812420963889279,
      "grad_norm": 1.0347023010253906,
      "learning_rate": 8.308515252276999e-05,
      "loss": 4.0828,
      "step": 68500
    },
    {
      "epoch": 4.847548124209639,
      "grad_norm": 1.0888615846633911,
      "learning_rate": 8.294058117681077e-05,
      "loss": 4.0782,
      "step": 69000
    },
    {
      "epoch": 4.8826752845299985,
      "grad_norm": 1.0173075199127197,
      "learning_rate": 8.279600983085153e-05,
      "loss": 4.0765,
      "step": 69500
    },
    {
      "epoch": 4.917802444850358,
      "grad_norm": 1.0210330486297607,
      "learning_rate": 8.26514384848923e-05,
      "loss": 4.0736,
      "step": 70000
    },
    {
      "epoch": 4.917802444850358,
      "eval_accuracy": 0.6324019458058838,
      "eval_loss": 3.7461562156677246,
      "eval_runtime": 135.0029,
      "eval_samples_per_second": 894.099,
      "eval_steps_per_second": 5.592,
      "step": 70000
    },
    {
      "epoch": 4.952929605170718,
      "grad_norm": 0.9816343784332275,
      "learning_rate": 8.250686713893306e-05,
      "loss": 4.0713,
      "step": 70500
    },
    {
      "epoch": 4.988056765491078,
      "grad_norm": 1.0771396160125732,
      "learning_rate": 8.236229579297384e-05,
      "loss": 4.0671,
      "step": 71000
    },
    {
      "epoch": 5.023183925811438,
      "grad_norm": 1.0552209615707397,
      "learning_rate": 8.221772444701461e-05,
      "loss": 4.0653,
      "step": 71500
    },
    {
      "epoch": 5.0583110861317975,
      "grad_norm": 1.0150943994522095,
      "learning_rate": 8.207315310105538e-05,
      "loss": 4.0587,
      "step": 72000
    },
    {
      "epoch": 5.093438246452157,
      "grad_norm": 0.9984225034713745,
      "learning_rate": 8.192858175509614e-05,
      "loss": 4.0547,
      "step": 72500
    },
    {
      "epoch": 5.128565406772516,
      "grad_norm": 1.0152759552001953,
      "learning_rate": 8.178401040913692e-05,
      "loss": 4.0512,
      "step": 73000
    },
    {
      "epoch": 5.163692567092876,
      "grad_norm": 1.0121378898620605,
      "learning_rate": 8.163943906317768e-05,
      "loss": 4.05,
      "step": 73500
    },
    {
      "epoch": 5.198819727413236,
      "grad_norm": 1.1056668758392334,
      "learning_rate": 8.149486771721845e-05,
      "loss": 4.0468,
      "step": 74000
    },
    {
      "epoch": 5.2339468877335955,
      "grad_norm": 1.088919997215271,
      "learning_rate": 8.135029637125921e-05,
      "loss": 4.0482,
      "step": 74500
    },
    {
      "epoch": 5.269074048053955,
      "grad_norm": 1.0417137145996094,
      "learning_rate": 8.120572502529999e-05,
      "loss": 4.0441,
      "step": 75000
    },
    {
      "epoch": 5.304201208374315,
      "grad_norm": 0.9883053302764893,
      "learning_rate": 8.106115367934077e-05,
      "loss": 4.0409,
      "step": 75500
    },
    {
      "epoch": 5.339328368694675,
      "grad_norm": 1.0172115564346313,
      "learning_rate": 8.091658233338153e-05,
      "loss": 4.0376,
      "step": 76000
    },
    {
      "epoch": 5.374455529015035,
      "grad_norm": 1.0904879570007324,
      "learning_rate": 8.07720109874223e-05,
      "loss": 4.0364,
      "step": 76500
    },
    {
      "epoch": 5.409582689335394,
      "grad_norm": 1.1110398769378662,
      "learning_rate": 8.062743964146306e-05,
      "loss": 4.0314,
      "step": 77000
    },
    {
      "epoch": 5.444709849655754,
      "grad_norm": 1.1156582832336426,
      "learning_rate": 8.048286829550384e-05,
      "loss": 4.0271,
      "step": 77500
    },
    {
      "epoch": 5.479837009976113,
      "grad_norm": 1.023111343383789,
      "learning_rate": 8.03382969495446e-05,
      "loss": 4.0284,
      "step": 78000
    },
    {
      "epoch": 5.514964170296473,
      "grad_norm": 1.020240068435669,
      "learning_rate": 8.019372560358536e-05,
      "loss": 4.0266,
      "step": 78500
    },
    {
      "epoch": 5.550091330616833,
      "grad_norm": 1.0575897693634033,
      "learning_rate": 8.004915425762614e-05,
      "loss": 4.0235,
      "step": 79000
    },
    {
      "epoch": 5.5852184909371925,
      "grad_norm": 1.0241798162460327,
      "learning_rate": 7.990458291166692e-05,
      "loss": 4.0231,
      "step": 79500
    },
    {
      "epoch": 5.620345651257552,
      "grad_norm": 0.9985337257385254,
      "learning_rate": 7.976001156570768e-05,
      "loss": 4.0187,
      "step": 80000
    },
    {
      "epoch": 5.620345651257552,
      "eval_accuracy": 0.639119809184786,
      "eval_loss": 3.698474645614624,
      "eval_runtime": 143.3347,
      "eval_samples_per_second": 842.127,
      "eval_steps_per_second": 5.267,
      "step": 80000
    },
    {
      "epoch": 5.655472811577912,
      "grad_norm": 1.0854889154434204,
      "learning_rate": 7.961544021974845e-05,
      "loss": 4.0141,
      "step": 80500
    },
    {
      "epoch": 5.690599971898272,
      "grad_norm": 1.0414183139801025,
      "learning_rate": 7.947086887378921e-05,
      "loss": 4.0224,
      "step": 81000
    },
    {
      "epoch": 5.725727132218632,
      "grad_norm": 0.9985676407814026,
      "learning_rate": 7.932629752782999e-05,
      "loss": 4.0142,
      "step": 81500
    },
    {
      "epoch": 5.760854292538991,
      "grad_norm": 1.0887261629104614,
      "learning_rate": 7.918172618187075e-05,
      "loss": 4.009,
      "step": 82000
    },
    {
      "epoch": 5.795981452859351,
      "grad_norm": 1.014569878578186,
      "learning_rate": 7.903715483591153e-05,
      "loss": 4.0107,
      "step": 82500
    },
    {
      "epoch": 5.83110861317971,
      "grad_norm": 0.9678444266319275,
      "learning_rate": 7.88925834899523e-05,
      "loss": 4.0085,
      "step": 83000
    },
    {
      "epoch": 5.866235773500071,
      "grad_norm": 1.025954246520996,
      "learning_rate": 7.874801214399307e-05,
      "loss": 4.0054,
      "step": 83500
    },
    {
      "epoch": 5.90136293382043,
      "grad_norm": 1.0009151697158813,
      "learning_rate": 7.860344079803384e-05,
      "loss": 4.0035,
      "step": 84000
    },
    {
      "epoch": 5.936490094140789,
      "grad_norm": 0.9895545840263367,
      "learning_rate": 7.84588694520746e-05,
      "loss": 4.0013,
      "step": 84500
    },
    {
      "epoch": 5.971617254461149,
      "grad_norm": 0.9882386326789856,
      "learning_rate": 7.831429810611536e-05,
      "loss": 3.9992,
      "step": 85000
    },
    {
      "epoch": 6.006744414781509,
      "grad_norm": 1.0791679620742798,
      "learning_rate": 7.816972676015614e-05,
      "loss": 3.9977,
      "step": 85500
    },
    {
      "epoch": 6.041871575101869,
      "grad_norm": 1.081634759902954,
      "learning_rate": 7.802515541419692e-05,
      "loss": 3.9924,
      "step": 86000
    },
    {
      "epoch": 6.076998735422229,
      "grad_norm": 0.9757462739944458,
      "learning_rate": 7.788058406823768e-05,
      "loss": 3.9942,
      "step": 86500
    },
    {
      "epoch": 6.112125895742588,
      "grad_norm": 1.098532795906067,
      "learning_rate": 7.773601272227845e-05,
      "loss": 3.9876,
      "step": 87000
    },
    {
      "epoch": 6.147253056062948,
      "grad_norm": 1.0299136638641357,
      "learning_rate": 7.759144137631922e-05,
      "loss": 3.9859,
      "step": 87500
    },
    {
      "epoch": 6.182380216383308,
      "grad_norm": 1.0599994659423828,
      "learning_rate": 7.744687003035999e-05,
      "loss": 3.9855,
      "step": 88000
    },
    {
      "epoch": 6.217507376703667,
      "grad_norm": 1.0483022928237915,
      "learning_rate": 7.730229868440075e-05,
      "loss": 3.9841,
      "step": 88500
    },
    {
      "epoch": 6.252634537024027,
      "grad_norm": 1.0010693073272705,
      "learning_rate": 7.715772733844152e-05,
      "loss": 3.9807,
      "step": 89000
    },
    {
      "epoch": 6.287761697344386,
      "grad_norm": 1.0573371648788452,
      "learning_rate": 7.70131559924823e-05,
      "loss": 3.9805,
      "step": 89500
    },
    {
      "epoch": 6.322888857664746,
      "grad_norm": 1.0268819332122803,
      "learning_rate": 7.686858464652307e-05,
      "loss": 3.9803,
      "step": 90000
    },
    {
      "epoch": 6.322888857664746,
      "eval_accuracy": 0.644372543386169,
      "eval_loss": 3.6643807888031006,
      "eval_runtime": 135.5289,
      "eval_samples_per_second": 890.629,
      "eval_steps_per_second": 5.571,
      "step": 90000
    },
    {
      "epoch": 6.358016017985106,
      "grad_norm": 1.0589687824249268,
      "learning_rate": 7.672401330056384e-05,
      "loss": 3.9789,
      "step": 90500
    },
    {
      "epoch": 6.393143178305466,
      "grad_norm": 1.0754573345184326,
      "learning_rate": 7.65794419546046e-05,
      "loss": 3.9763,
      "step": 91000
    },
    {
      "epoch": 6.4282703386258255,
      "grad_norm": 1.0275828838348389,
      "learning_rate": 7.643487060864536e-05,
      "loss": 3.9729,
      "step": 91500
    },
    {
      "epoch": 6.463397498946185,
      "grad_norm": 1.0694172382354736,
      "learning_rate": 7.629029926268614e-05,
      "loss": 3.9724,
      "step": 92000
    },
    {
      "epoch": 6.498524659266545,
      "grad_norm": 1.030853271484375,
      "learning_rate": 7.61457279167269e-05,
      "loss": 3.9721,
      "step": 92500
    },
    {
      "epoch": 6.533651819586905,
      "grad_norm": 0.9506504535675049,
      "learning_rate": 7.600115657076767e-05,
      "loss": 3.971,
      "step": 93000
    },
    {
      "epoch": 6.568778979907265,
      "grad_norm": 1.0954880714416504,
      "learning_rate": 7.585658522480845e-05,
      "loss": 3.9693,
      "step": 93500
    },
    {
      "epoch": 6.603906140227624,
      "grad_norm": 0.9772918224334717,
      "learning_rate": 7.571201387884922e-05,
      "loss": 3.9644,
      "step": 94000
    },
    {
      "epoch": 6.639033300547983,
      "grad_norm": 1.0400298833847046,
      "learning_rate": 7.556744253288999e-05,
      "loss": 3.9623,
      "step": 94500
    },
    {
      "epoch": 6.674160460868343,
      "grad_norm": 1.0729682445526123,
      "learning_rate": 7.542287118693075e-05,
      "loss": 3.9631,
      "step": 95000
    },
    {
      "epoch": 6.709287621188703,
      "grad_norm": 1.0170788764953613,
      "learning_rate": 7.527829984097152e-05,
      "loss": 3.9611,
      "step": 95500
    },
    {
      "epoch": 6.744414781509063,
      "grad_norm": 1.0005239248275757,
      "learning_rate": 7.51337284950123e-05,
      "loss": 3.9666,
      "step": 96000
    },
    {
      "epoch": 6.7795419418294225,
      "grad_norm": 1.033534049987793,
      "learning_rate": 7.498915714905306e-05,
      "loss": 3.9567,
      "step": 96500
    },
    {
      "epoch": 6.814669102149782,
      "grad_norm": 1.0282751321792603,
      "learning_rate": 7.484458580309382e-05,
      "loss": 3.9585,
      "step": 97000
    },
    {
      "epoch": 6.849796262470142,
      "grad_norm": 1.1123915910720825,
      "learning_rate": 7.47000144571346e-05,
      "loss": 3.9509,
      "step": 97500
    },
    {
      "epoch": 6.884923422790502,
      "grad_norm": 1.0020571947097778,
      "learning_rate": 7.455544311117538e-05,
      "loss": 3.9531,
      "step": 98000
    },
    {
      "epoch": 6.920050583110862,
      "grad_norm": 1.0285818576812744,
      "learning_rate": 7.441087176521614e-05,
      "loss": 3.9507,
      "step": 98500
    },
    {
      "epoch": 6.955177743431221,
      "grad_norm": 1.0496854782104492,
      "learning_rate": 7.42663004192569e-05,
      "loss": 3.9489,
      "step": 99000
    },
    {
      "epoch": 6.99030490375158,
      "grad_norm": 1.0300986766815186,
      "learning_rate": 7.412172907329767e-05,
      "loss": 3.9465,
      "step": 99500
    },
    {
      "epoch": 7.02543206407194,
      "grad_norm": 1.0604512691497803,
      "learning_rate": 7.397715772733845e-05,
      "loss": 3.9462,
      "step": 100000
    },
    {
      "epoch": 7.02543206407194,
      "eval_accuracy": 0.648548853567911,
      "eval_loss": 3.633261203765869,
      "eval_runtime": 144.5331,
      "eval_samples_per_second": 835.144,
      "eval_steps_per_second": 5.224,
      "step": 100000
    },
    {
      "epoch": 7.0605592243923,
      "grad_norm": 1.0465333461761475,
      "learning_rate": 7.383258638137921e-05,
      "loss": 3.9439,
      "step": 100500
    },
    {
      "epoch": 7.09568638471266,
      "grad_norm": 0.977771520614624,
      "learning_rate": 7.368801503541999e-05,
      "loss": 3.9459,
      "step": 101000
    },
    {
      "epoch": 7.1308135450330195,
      "grad_norm": 1.09309983253479,
      "learning_rate": 7.354344368946075e-05,
      "loss": 3.9432,
      "step": 101500
    },
    {
      "epoch": 7.165940705353379,
      "grad_norm": 1.0659376382827759,
      "learning_rate": 7.339887234350153e-05,
      "loss": 3.9407,
      "step": 102000
    },
    {
      "epoch": 7.201067865673739,
      "grad_norm": 0.9782362580299377,
      "learning_rate": 7.32543009975423e-05,
      "loss": 3.9393,
      "step": 102500
    },
    {
      "epoch": 7.236195025994099,
      "grad_norm": 1.0496647357940674,
      "learning_rate": 7.310972965158306e-05,
      "loss": 3.9384,
      "step": 103000
    },
    {
      "epoch": 7.271322186314459,
      "grad_norm": 1.0655579566955566,
      "learning_rate": 7.296515830562382e-05,
      "loss": 3.9384,
      "step": 103500
    },
    {
      "epoch": 7.306449346634818,
      "grad_norm": 1.0276716947555542,
      "learning_rate": 7.28205869596646e-05,
      "loss": 3.933,
      "step": 104000
    },
    {
      "epoch": 7.341576506955178,
      "grad_norm": 1.0266293287277222,
      "learning_rate": 7.267601561370536e-05,
      "loss": 3.9291,
      "step": 104500
    },
    {
      "epoch": 7.376703667275537,
      "grad_norm": 0.9779165983200073,
      "learning_rate": 7.253144426774614e-05,
      "loss": 3.9322,
      "step": 105000
    },
    {
      "epoch": 7.411830827595897,
      "grad_norm": 1.0079984664916992,
      "learning_rate": 7.23868729217869e-05,
      "loss": 3.9352,
      "step": 105500
    },
    {
      "epoch": 7.446957987916257,
      "grad_norm": 0.9881858229637146,
      "learning_rate": 7.224230157582768e-05,
      "loss": 3.9338,
      "step": 106000
    },
    {
      "epoch": 7.482085148236616,
      "grad_norm": 0.9892914891242981,
      "learning_rate": 7.209773022986845e-05,
      "loss": 3.9285,
      "step": 106500
    },
    {
      "epoch": 7.517212308556976,
      "grad_norm": 1.154355764389038,
      "learning_rate": 7.195315888390921e-05,
      "loss": 3.9259,
      "step": 107000
    },
    {
      "epoch": 7.552339468877336,
      "grad_norm": 1.0398566722869873,
      "learning_rate": 7.180858753794997e-05,
      "loss": 3.9247,
      "step": 107500
    },
    {
      "epoch": 7.587466629197696,
      "grad_norm": 0.9936870336532593,
      "learning_rate": 7.166401619199075e-05,
      "loss": 3.9255,
      "step": 108000
    },
    {
      "epoch": 7.6225937895180556,
      "grad_norm": 1.001603603363037,
      "learning_rate": 7.151944484603153e-05,
      "loss": 3.9269,
      "step": 108500
    },
    {
      "epoch": 7.657720949838415,
      "grad_norm": 1.019192099571228,
      "learning_rate": 7.13748735000723e-05,
      "loss": 3.9223,
      "step": 109000
    },
    {
      "epoch": 7.692848110158775,
      "grad_norm": 1.051631212234497,
      "learning_rate": 7.123030215411306e-05,
      "loss": 3.9212,
      "step": 109500
    },
    {
      "epoch": 7.727975270479135,
      "grad_norm": 1.01295804977417,
      "learning_rate": 7.108573080815382e-05,
      "loss": 3.9217,
      "step": 110000
    },
    {
      "epoch": 7.727975270479135,
      "eval_accuracy": 0.6526292954175039,
      "eval_loss": 3.606437921524048,
      "eval_runtime": 137.8832,
      "eval_samples_per_second": 875.422,
      "eval_steps_per_second": 5.476,
      "step": 110000
    },
    {
      "epoch": 7.763102430799494,
      "grad_norm": 0.9921694397926331,
      "learning_rate": 7.09411594621946e-05,
      "loss": 3.9205,
      "step": 110500
    },
    {
      "epoch": 7.798229591119854,
      "grad_norm": 0.9742565155029297,
      "learning_rate": 7.079658811623536e-05,
      "loss": 3.9163,
      "step": 111000
    },
    {
      "epoch": 7.833356751440213,
      "grad_norm": 1.0286446809768677,
      "learning_rate": 7.065201677027613e-05,
      "loss": 3.9164,
      "step": 111500
    },
    {
      "epoch": 7.868483911760573,
      "grad_norm": 1.0960822105407715,
      "learning_rate": 7.05074454243169e-05,
      "loss": 3.9167,
      "step": 112000
    },
    {
      "epoch": 7.903611072080933,
      "grad_norm": 1.0249719619750977,
      "learning_rate": 7.036287407835768e-05,
      "loss": 3.9157,
      "step": 112500
    },
    {
      "epoch": 7.938738232401293,
      "grad_norm": 1.02247154712677,
      "learning_rate": 7.021830273239845e-05,
      "loss": 3.908,
      "step": 113000
    },
    {
      "epoch": 7.9738653927216525,
      "grad_norm": 0.9692552089691162,
      "learning_rate": 7.007373138643921e-05,
      "loss": 3.9104,
      "step": 113500
    },
    {
      "epoch": 8.008992553042011,
      "grad_norm": 1.059390664100647,
      "learning_rate": 6.992916004047997e-05,
      "loss": 3.9115,
      "step": 114000
    },
    {
      "epoch": 8.044119713362372,
      "grad_norm": 1.0319334268569946,
      "learning_rate": 6.978458869452075e-05,
      "loss": 3.9056,
      "step": 114500
    },
    {
      "epoch": 8.079246873682731,
      "grad_norm": 0.9874253273010254,
      "learning_rate": 6.964001734856152e-05,
      "loss": 3.9054,
      "step": 115000
    },
    {
      "epoch": 8.114374034003092,
      "grad_norm": 1.0460796356201172,
      "learning_rate": 6.949544600260228e-05,
      "loss": 3.9045,
      "step": 115500
    },
    {
      "epoch": 8.14950119432345,
      "grad_norm": 1.0085667371749878,
      "learning_rate": 6.935087465664306e-05,
      "loss": 3.9038,
      "step": 116000
    },
    {
      "epoch": 8.184628354643811,
      "grad_norm": 1.0626521110534668,
      "learning_rate": 6.920630331068383e-05,
      "loss": 3.905,
      "step": 116500
    },
    {
      "epoch": 8.21975551496417,
      "grad_norm": 0.977799117565155,
      "learning_rate": 6.90617319647246e-05,
      "loss": 3.9006,
      "step": 117000
    },
    {
      "epoch": 8.25488267528453,
      "grad_norm": 1.0686109066009521,
      "learning_rate": 6.891716061876536e-05,
      "loss": 3.9011,
      "step": 117500
    },
    {
      "epoch": 8.29000983560489,
      "grad_norm": 1.0719889402389526,
      "learning_rate": 6.877258927280613e-05,
      "loss": 3.8962,
      "step": 118000
    },
    {
      "epoch": 8.325136995925249,
      "grad_norm": 1.0647310018539429,
      "learning_rate": 6.86280179268469e-05,
      "loss": 3.8959,
      "step": 118500
    },
    {
      "epoch": 8.36026415624561,
      "grad_norm": 1.0350008010864258,
      "learning_rate": 6.848344658088767e-05,
      "loss": 3.8951,
      "step": 119000
    },
    {
      "epoch": 8.395391316565968,
      "grad_norm": 0.988900899887085,
      "learning_rate": 6.833887523492845e-05,
      "loss": 3.8961,
      "step": 119500
    },
    {
      "epoch": 8.430518476886329,
      "grad_norm": 1.0585073232650757,
      "learning_rate": 6.819430388896921e-05,
      "loss": 3.8974,
      "step": 120000
    },
    {
      "epoch": 8.430518476886329,
      "eval_accuracy": 0.6558340280780799,
      "eval_loss": 3.5809683799743652,
      "eval_runtime": 136.7242,
      "eval_samples_per_second": 882.843,
      "eval_steps_per_second": 5.522,
      "step": 120000
    },
    {
      "epoch": 8.465645637206688,
      "grad_norm": 0.9938855171203613,
      "learning_rate": 6.804973254300999e-05,
      "loss": 3.8963,
      "step": 120500
    },
    {
      "epoch": 8.500772797527048,
      "grad_norm": 1.0587878227233887,
      "learning_rate": 6.790516119705075e-05,
      "loss": 3.8954,
      "step": 121000
    },
    {
      "epoch": 8.535899957847407,
      "grad_norm": 1.003372073173523,
      "learning_rate": 6.776058985109152e-05,
      "loss": 3.8969,
      "step": 121500
    },
    {
      "epoch": 8.571027118167768,
      "grad_norm": 1.0167999267578125,
      "learning_rate": 6.761601850513228e-05,
      "loss": 3.8929,
      "step": 122000
    },
    {
      "epoch": 8.606154278488127,
      "grad_norm": 1.0059306621551514,
      "learning_rate": 6.747144715917306e-05,
      "loss": 3.8903,
      "step": 122500
    },
    {
      "epoch": 8.641281438808488,
      "grad_norm": 1.0564857721328735,
      "learning_rate": 6.732687581321382e-05,
      "loss": 3.8892,
      "step": 123000
    },
    {
      "epoch": 8.676408599128846,
      "grad_norm": 1.0630180835723877,
      "learning_rate": 6.71823044672546e-05,
      "loss": 3.8853,
      "step": 123500
    },
    {
      "epoch": 8.711535759449205,
      "grad_norm": 1.0301406383514404,
      "learning_rate": 6.703773312129536e-05,
      "loss": 3.8835,
      "step": 124000
    },
    {
      "epoch": 8.746662919769566,
      "grad_norm": 1.0411887168884277,
      "learning_rate": 6.689316177533613e-05,
      "loss": 3.8786,
      "step": 124500
    },
    {
      "epoch": 8.781790080089925,
      "grad_norm": 1.0367854833602905,
      "learning_rate": 6.67485904293769e-05,
      "loss": 3.8835,
      "step": 125000
    },
    {
      "epoch": 8.816917240410286,
      "grad_norm": 1.0663400888442993,
      "learning_rate": 6.660401908341767e-05,
      "loss": 3.8817,
      "step": 125500
    },
    {
      "epoch": 8.852044400730644,
      "grad_norm": 1.0770373344421387,
      "learning_rate": 6.645944773745843e-05,
      "loss": 3.8802,
      "step": 126000
    },
    {
      "epoch": 8.887171561051005,
      "grad_norm": 1.044163465499878,
      "learning_rate": 6.631487639149921e-05,
      "loss": 3.8823,
      "step": 126500
    },
    {
      "epoch": 8.922298721371364,
      "grad_norm": 1.0890542268753052,
      "learning_rate": 6.617030504553999e-05,
      "loss": 3.8802,
      "step": 127000
    },
    {
      "epoch": 8.957425881691725,
      "grad_norm": 1.0345556735992432,
      "learning_rate": 6.602573369958075e-05,
      "loss": 3.8837,
      "step": 127500
    },
    {
      "epoch": 8.992553042012084,
      "grad_norm": 1.0826619863510132,
      "learning_rate": 6.588116235362152e-05,
      "loss": 3.8832,
      "step": 128000
    },
    {
      "epoch": 9.027680202332444,
      "grad_norm": 1.0863028764724731,
      "learning_rate": 6.573659100766228e-05,
      "loss": 3.8743,
      "step": 128500
    },
    {
      "epoch": 9.062807362652803,
      "grad_norm": 1.0319151878356934,
      "learning_rate": 6.559201966170306e-05,
      "loss": 3.8748,
      "step": 129000
    },
    {
      "epoch": 9.097934522973162,
      "grad_norm": 1.0857160091400146,
      "learning_rate": 6.544744831574382e-05,
      "loss": 3.8745,
      "step": 129500
    },
    {
      "epoch": 9.133061683293523,
      "grad_norm": 1.1056162118911743,
      "learning_rate": 6.530287696978458e-05,
      "loss": 3.8714,
      "step": 130000
    },
    {
      "epoch": 9.133061683293523,
      "eval_accuracy": 0.6580812197085962,
      "eval_loss": 3.569554328918457,
      "eval_runtime": 136.7612,
      "eval_samples_per_second": 882.604,
      "eval_steps_per_second": 5.521,
      "step": 130000
    },
    {
      "epoch": 9.168188843613882,
      "grad_norm": 1.0687092542648315,
      "learning_rate": 6.515830562382536e-05,
      "loss": 3.8724,
      "step": 130500
    },
    {
      "epoch": 9.203316003934242,
      "grad_norm": 1.0319805145263672,
      "learning_rate": 6.501373427786614e-05,
      "loss": 3.8747,
      "step": 131000
    },
    {
      "epoch": 9.238443164254601,
      "grad_norm": 1.0324089527130127,
      "learning_rate": 6.48691629319069e-05,
      "loss": 3.8724,
      "step": 131500
    },
    {
      "epoch": 9.273570324574962,
      "grad_norm": 1.0121489763259888,
      "learning_rate": 6.472459158594767e-05,
      "loss": 3.8745,
      "step": 132000
    },
    {
      "epoch": 9.30869748489532,
      "grad_norm": 1.1079241037368774,
      "learning_rate": 6.458002023998843e-05,
      "loss": 3.8721,
      "step": 132500
    },
    {
      "epoch": 9.343824645215681,
      "grad_norm": 1.0359601974487305,
      "learning_rate": 6.443544889402921e-05,
      "loss": 3.8686,
      "step": 133000
    },
    {
      "epoch": 9.37895180553604,
      "grad_norm": 1.1285004615783691,
      "learning_rate": 6.429087754806997e-05,
      "loss": 3.8674,
      "step": 133500
    },
    {
      "epoch": 9.414078965856401,
      "grad_norm": 0.971286952495575,
      "learning_rate": 6.414630620211074e-05,
      "loss": 3.8672,
      "step": 134000
    },
    {
      "epoch": 9.44920612617676,
      "grad_norm": 1.089293122291565,
      "learning_rate": 6.400173485615151e-05,
      "loss": 3.8639,
      "step": 134500
    },
    {
      "epoch": 9.484333286497119,
      "grad_norm": 1.1344711780548096,
      "learning_rate": 6.385716351019229e-05,
      "loss": 3.8638,
      "step": 135000
    },
    {
      "epoch": 9.51946044681748,
      "grad_norm": 1.0056766271591187,
      "learning_rate": 6.371259216423306e-05,
      "loss": 3.8648,
      "step": 135500
    },
    {
      "epoch": 9.554587607137838,
      "grad_norm": 1.0078498125076294,
      "learning_rate": 6.356802081827382e-05,
      "loss": 3.8634,
      "step": 136000
    },
    {
      "epoch": 9.589714767458199,
      "grad_norm": 0.9975442886352539,
      "learning_rate": 6.342344947231458e-05,
      "loss": 3.8648,
      "step": 136500
    },
    {
      "epoch": 9.624841927778558,
      "grad_norm": 1.0246551036834717,
      "learning_rate": 6.327887812635536e-05,
      "loss": 3.861,
      "step": 137000
    },
    {
      "epoch": 9.659969088098919,
      "grad_norm": 1.0477834939956665,
      "learning_rate": 6.313430678039613e-05,
      "loss": 3.8612,
      "step": 137500
    },
    {
      "epoch": 9.695096248419278,
      "grad_norm": 1.0582610368728638,
      "learning_rate": 6.29897354344369e-05,
      "loss": 3.8566,
      "step": 138000
    },
    {
      "epoch": 9.730223408739638,
      "grad_norm": 1.062447190284729,
      "learning_rate": 6.284516408847767e-05,
      "loss": 3.8633,
      "step": 138500
    },
    {
      "epoch": 9.765350569059997,
      "grad_norm": 1.0355119705200195,
      "learning_rate": 6.270059274251843e-05,
      "loss": 3.8561,
      "step": 139000
    },
    {
      "epoch": 9.800477729380358,
      "grad_norm": 1.036462426185608,
      "learning_rate": 6.255602139655921e-05,
      "loss": 3.8573,
      "step": 139500
    },
    {
      "epoch": 9.835604889700717,
      "grad_norm": 0.9981861114501953,
      "learning_rate": 6.241145005059997e-05,
      "loss": 3.8565,
      "step": 140000
    },
    {
      "epoch": 9.835604889700717,
      "eval_accuracy": 0.6612540934142709,
      "eval_loss": 3.5454251766204834,
      "eval_runtime": 137.6124,
      "eval_samples_per_second": 877.145,
      "eval_steps_per_second": 5.486,
      "step": 140000
    },
    {
      "epoch": 9.870732050021076,
      "grad_norm": 1.2044545412063599,
      "learning_rate": 6.226687870464074e-05,
      "loss": 3.8562,
      "step": 140500
    },
    {
      "epoch": 9.905859210341436,
      "grad_norm": 1.0026911497116089,
      "learning_rate": 6.212230735868151e-05,
      "loss": 3.8556,
      "step": 141000
    },
    {
      "epoch": 9.940986370661795,
      "grad_norm": 1.0099139213562012,
      "learning_rate": 6.197773601272228e-05,
      "loss": 3.8558,
      "step": 141500
    },
    {
      "epoch": 9.976113530982156,
      "grad_norm": 1.0034656524658203,
      "learning_rate": 6.183316466676306e-05,
      "loss": 3.8505,
      "step": 142000
    },
    {
      "epoch": 10.011240691302515,
      "grad_norm": 1.0413298606872559,
      "learning_rate": 6.168859332080382e-05,
      "loss": 3.8533,
      "step": 142500
    },
    {
      "epoch": 10.046367851622875,
      "grad_norm": 1.030550479888916,
      "learning_rate": 6.154402197484458e-05,
      "loss": 3.8523,
      "step": 143000
    },
    {
      "epoch": 10.081495011943234,
      "grad_norm": 1.0325989723205566,
      "learning_rate": 6.139945062888536e-05,
      "loss": 3.8486,
      "step": 143500
    },
    {
      "epoch": 10.116622172263595,
      "grad_norm": 1.1221369504928589,
      "learning_rate": 6.125487928292613e-05,
      "loss": 3.8473,
      "step": 144000
    },
    {
      "epoch": 10.151749332583954,
      "grad_norm": 1.0764824151992798,
      "learning_rate": 6.111030793696689e-05,
      "loss": 3.8478,
      "step": 144500
    },
    {
      "epoch": 10.186876492904315,
      "grad_norm": 1.030349850654602,
      "learning_rate": 6.096573659100766e-05,
      "loss": 3.8474,
      "step": 145000
    },
    {
      "epoch": 10.222003653224673,
      "grad_norm": 0.9821157455444336,
      "learning_rate": 6.082116524504844e-05,
      "loss": 3.8496,
      "step": 145500
    },
    {
      "epoch": 10.257130813545032,
      "grad_norm": 0.9990749955177307,
      "learning_rate": 6.067659389908921e-05,
      "loss": 3.8484,
      "step": 146000
    },
    {
      "epoch": 10.292257973865393,
      "grad_norm": 1.0158008337020874,
      "learning_rate": 6.053202255312997e-05,
      "loss": 3.8415,
      "step": 146500
    },
    {
      "epoch": 10.327385134185752,
      "grad_norm": 1.0147724151611328,
      "learning_rate": 6.0387451207170744e-05,
      "loss": 3.8438,
      "step": 147000
    },
    {
      "epoch": 10.362512294506113,
      "grad_norm": 1.0989351272583008,
      "learning_rate": 6.024287986121151e-05,
      "loss": 3.844,
      "step": 147500
    },
    {
      "epoch": 10.397639454826471,
      "grad_norm": 0.9766375422477722,
      "learning_rate": 6.009830851525228e-05,
      "loss": 3.8406,
      "step": 148000
    },
    {
      "epoch": 10.432766615146832,
      "grad_norm": 1.0558122396469116,
      "learning_rate": 5.995373716929304e-05,
      "loss": 3.8444,
      "step": 148500
    },
    {
      "epoch": 10.467893775467191,
      "grad_norm": 1.0255546569824219,
      "learning_rate": 5.980916582333381e-05,
      "loss": 3.8405,
      "step": 149000
    },
    {
      "epoch": 10.503020935787552,
      "grad_norm": 0.9918079376220703,
      "learning_rate": 5.966459447737459e-05,
      "loss": 3.8371,
      "step": 149500
    },
    {
      "epoch": 10.53814809610791,
      "grad_norm": 1.0484111309051514,
      "learning_rate": 5.952002313141536e-05,
      "loss": 3.8382,
      "step": 150000
    },
    {
      "epoch": 10.53814809610791,
      "eval_accuracy": 0.6631513108850643,
      "eval_loss": 3.5309653282165527,
      "eval_runtime": 137.5219,
      "eval_samples_per_second": 877.722,
      "eval_steps_per_second": 5.49,
      "step": 150000
    },
    {
      "epoch": 10.573275256428271,
      "grad_norm": 1.0241247415542603,
      "learning_rate": 5.9375451785456126e-05,
      "loss": 3.8433,
      "step": 150500
    },
    {
      "epoch": 10.60840241674863,
      "grad_norm": 1.0452362298965454,
      "learning_rate": 5.9230880439496896e-05,
      "loss": 3.8382,
      "step": 151000
    },
    {
      "epoch": 10.643529577068989,
      "grad_norm": 1.0399458408355713,
      "learning_rate": 5.908630909353766e-05,
      "loss": 3.838,
      "step": 151500
    },
    {
      "epoch": 10.67865673738935,
      "grad_norm": 0.9991360306739807,
      "learning_rate": 5.894173774757843e-05,
      "loss": 3.837,
      "step": 152000
    },
    {
      "epoch": 10.713783897709709,
      "grad_norm": 1.0450655221939087,
      "learning_rate": 5.8797166401619195e-05,
      "loss": 3.834,
      "step": 152500
    },
    {
      "epoch": 10.74891105803007,
      "grad_norm": 1.0295804738998413,
      "learning_rate": 5.865259505565997e-05,
      "loss": 3.8362,
      "step": 153000
    },
    {
      "epoch": 10.784038218350428,
      "grad_norm": 1.1099650859832764,
      "learning_rate": 5.8508023709700743e-05,
      "loss": 3.8372,
      "step": 153500
    },
    {
      "epoch": 10.819165378670789,
      "grad_norm": 1.0414799451828003,
      "learning_rate": 5.8363452363741514e-05,
      "loss": 3.8355,
      "step": 154000
    },
    {
      "epoch": 10.854292538991148,
      "grad_norm": 0.9899589419364929,
      "learning_rate": 5.821888101778228e-05,
      "loss": 3.8346,
      "step": 154500
    },
    {
      "epoch": 10.889419699311508,
      "grad_norm": 1.027968168258667,
      "learning_rate": 5.807430967182305e-05,
      "loss": 3.8331,
      "step": 155000
    },
    {
      "epoch": 10.924546859631867,
      "grad_norm": 1.020301103591919,
      "learning_rate": 5.792973832586381e-05,
      "loss": 3.8343,
      "step": 155500
    },
    {
      "epoch": 10.959674019952226,
      "grad_norm": 1.1077617406845093,
      "learning_rate": 5.7785166979904584e-05,
      "loss": 3.8324,
      "step": 156000
    },
    {
      "epoch": 10.994801180272587,
      "grad_norm": 1.005629301071167,
      "learning_rate": 5.764059563394536e-05,
      "loss": 3.8301,
      "step": 156500
    },
    {
      "epoch": 11.029928340592946,
      "grad_norm": 1.0447512865066528,
      "learning_rate": 5.7496024287986125e-05,
      "loss": 3.8237,
      "step": 157000
    },
    {
      "epoch": 11.065055500913306,
      "grad_norm": 1.0114192962646484,
      "learning_rate": 5.7351452942026896e-05,
      "loss": 3.828,
      "step": 157500
    },
    {
      "epoch": 11.100182661233665,
      "grad_norm": 1.0450975894927979,
      "learning_rate": 5.720688159606766e-05,
      "loss": 3.828,
      "step": 158000
    },
    {
      "epoch": 11.135309821554026,
      "grad_norm": 1.0884100198745728,
      "learning_rate": 5.706231025010843e-05,
      "loss": 3.8235,
      "step": 158500
    },
    {
      "epoch": 11.170436981874385,
      "grad_norm": 1.06899094581604,
      "learning_rate": 5.6917738904149195e-05,
      "loss": 3.8251,
      "step": 159000
    },
    {
      "epoch": 11.205564142194746,
      "grad_norm": 1.0569485425949097,
      "learning_rate": 5.6773167558189966e-05,
      "loss": 3.8258,
      "step": 159500
    },
    {
      "epoch": 11.240691302515105,
      "grad_norm": 1.1100273132324219,
      "learning_rate": 5.6628596212230736e-05,
      "loss": 3.8272,
      "step": 160000
    },
    {
      "epoch": 11.240691302515105,
      "eval_accuracy": 0.6646739994277903,
      "eval_loss": 3.5181100368499756,
      "eval_runtime": 141.0802,
      "eval_samples_per_second": 855.584,
      "eval_steps_per_second": 5.352,
      "step": 160000
    },
    {
      "epoch": 11.275818462835465,
      "grad_norm": 1.0628185272216797,
      "learning_rate": 5.6484024866271514e-05,
      "loss": 3.8221,
      "step": 160500
    },
    {
      "epoch": 11.310945623155824,
      "grad_norm": 1.1270934343338013,
      "learning_rate": 5.633945352031228e-05,
      "loss": 3.8258,
      "step": 161000
    },
    {
      "epoch": 11.346072783476183,
      "grad_norm": 1.0492292642593384,
      "learning_rate": 5.619488217435305e-05,
      "loss": 3.8212,
      "step": 161500
    },
    {
      "epoch": 11.381199943796544,
      "grad_norm": 1.1114745140075684,
      "learning_rate": 5.605031082839381e-05,
      "loss": 3.8213,
      "step": 162000
    },
    {
      "epoch": 11.416327104116903,
      "grad_norm": 1.0274219512939453,
      "learning_rate": 5.5905739482434584e-05,
      "loss": 3.8176,
      "step": 162500
    },
    {
      "epoch": 11.451454264437263,
      "grad_norm": 1.0416280031204224,
      "learning_rate": 5.576116813647535e-05,
      "loss": 3.8137,
      "step": 163000
    },
    {
      "epoch": 11.486581424757622,
      "grad_norm": 1.0962536334991455,
      "learning_rate": 5.561659679051612e-05,
      "loss": 3.8165,
      "step": 163500
    },
    {
      "epoch": 11.521708585077983,
      "grad_norm": 1.0611205101013184,
      "learning_rate": 5.5472025444556896e-05,
      "loss": 3.8171,
      "step": 164000
    },
    {
      "epoch": 11.556835745398342,
      "grad_norm": 1.0393544435501099,
      "learning_rate": 5.532745409859767e-05,
      "loss": 3.8154,
      "step": 164500
    },
    {
      "epoch": 11.591962905718702,
      "grad_norm": 1.0397778749465942,
      "learning_rate": 5.518288275263843e-05,
      "loss": 3.8151,
      "step": 165000
    },
    {
      "epoch": 11.627090066039061,
      "grad_norm": 1.0978599786758423,
      "learning_rate": 5.50383114066792e-05,
      "loss": 3.815,
      "step": 165500
    },
    {
      "epoch": 11.662217226359422,
      "grad_norm": 1.0579622983932495,
      "learning_rate": 5.4893740060719965e-05,
      "loss": 3.8131,
      "step": 166000
    },
    {
      "epoch": 11.69734438667978,
      "grad_norm": 1.090484380722046,
      "learning_rate": 5.4749168714760736e-05,
      "loss": 3.8132,
      "step": 166500
    },
    {
      "epoch": 11.73247154700014,
      "grad_norm": 1.0129539966583252,
      "learning_rate": 5.46045973688015e-05,
      "loss": 3.8079,
      "step": 167000
    },
    {
      "epoch": 11.7675987073205,
      "grad_norm": 1.121238350868225,
      "learning_rate": 5.446002602284227e-05,
      "loss": 3.8077,
      "step": 167500
    },
    {
      "epoch": 11.80272586764086,
      "grad_norm": 1.1320801973342896,
      "learning_rate": 5.431545467688305e-05,
      "loss": 3.8114,
      "step": 168000
    },
    {
      "epoch": 11.83785302796122,
      "grad_norm": 1.0639028549194336,
      "learning_rate": 5.417088333092382e-05,
      "loss": 3.8069,
      "step": 168500
    },
    {
      "epoch": 11.872980188281579,
      "grad_norm": 1.129622459411621,
      "learning_rate": 5.402631198496458e-05,
      "loss": 3.8055,
      "step": 169000
    },
    {
      "epoch": 11.90810734860194,
      "grad_norm": 1.1511698961257935,
      "learning_rate": 5.3881740639005354e-05,
      "loss": 3.8066,
      "step": 169500
    },
    {
      "epoch": 11.943234508922298,
      "grad_norm": 1.1098284721374512,
      "learning_rate": 5.373716929304612e-05,
      "loss": 3.8059,
      "step": 170000
    },
    {
      "epoch": 11.943234508922298,
      "eval_accuracy": 0.6666074675524029,
      "eval_loss": 3.5011649131774902,
      "eval_runtime": 136.426,
      "eval_samples_per_second": 884.773,
      "eval_steps_per_second": 5.534,
      "step": 170000
    },
    {
      "epoch": 11.978361669242659,
      "grad_norm": 1.084730625152588,
      "learning_rate": 5.359259794708689e-05,
      "loss": 3.8061,
      "step": 170500
    },
    {
      "epoch": 12.013488829563018,
      "grad_norm": 1.167523980140686,
      "learning_rate": 5.344802660112765e-05,
      "loss": 3.8064,
      "step": 171000
    },
    {
      "epoch": 12.048615989883379,
      "grad_norm": 1.1096934080123901,
      "learning_rate": 5.330345525516843e-05,
      "loss": 3.8035,
      "step": 171500
    },
    {
      "epoch": 12.083743150203738,
      "grad_norm": 1.0423945188522339,
      "learning_rate": 5.31588839092092e-05,
      "loss": 3.8004,
      "step": 172000
    },
    {
      "epoch": 12.118870310524096,
      "grad_norm": 1.0941615104675293,
      "learning_rate": 5.3014312563249965e-05,
      "loss": 3.8028,
      "step": 172500
    },
    {
      "epoch": 12.153997470844457,
      "grad_norm": 1.0533970594406128,
      "learning_rate": 5.2869741217290736e-05,
      "loss": 3.8052,
      "step": 173000
    },
    {
      "epoch": 12.189124631164816,
      "grad_norm": 1.0236855745315552,
      "learning_rate": 5.272516987133151e-05,
      "loss": 3.8012,
      "step": 173500
    },
    {
      "epoch": 12.224251791485177,
      "grad_norm": 1.1091432571411133,
      "learning_rate": 5.258059852537227e-05,
      "loss": 3.796,
      "step": 174000
    },
    {
      "epoch": 12.259378951805536,
      "grad_norm": 1.1524672508239746,
      "learning_rate": 5.243602717941304e-05,
      "loss": 3.8003,
      "step": 174500
    },
    {
      "epoch": 12.294506112125896,
      "grad_norm": 1.0929014682769775,
      "learning_rate": 5.229145583345382e-05,
      "loss": 3.7954,
      "step": 175000
    },
    {
      "epoch": 12.329633272446255,
      "grad_norm": 1.0436333417892456,
      "learning_rate": 5.214688448749458e-05,
      "loss": 3.802,
      "step": 175500
    },
    {
      "epoch": 12.364760432766616,
      "grad_norm": 1.1149826049804688,
      "learning_rate": 5.2002313141535354e-05,
      "loss": 3.7975,
      "step": 176000
    },
    {
      "epoch": 12.399887593086975,
      "grad_norm": 1.1571273803710938,
      "learning_rate": 5.185774179557612e-05,
      "loss": 3.7953,
      "step": 176500
    },
    {
      "epoch": 12.435014753407334,
      "grad_norm": 1.053972840309143,
      "learning_rate": 5.171317044961689e-05,
      "loss": 3.7971,
      "step": 177000
    },
    {
      "epoch": 12.470141913727694,
      "grad_norm": 1.0854816436767578,
      "learning_rate": 5.156859910365765e-05,
      "loss": 3.7949,
      "step": 177500
    },
    {
      "epoch": 12.505269074048053,
      "grad_norm": 1.2815136909484863,
      "learning_rate": 5.1424027757698423e-05,
      "loss": 3.7962,
      "step": 178000
    },
    {
      "epoch": 12.540396234368414,
      "grad_norm": 1.2136986255645752,
      "learning_rate": 5.127945641173919e-05,
      "loss": 3.798,
      "step": 178500
    },
    {
      "epoch": 12.575523394688773,
      "grad_norm": 1.0453957319259644,
      "learning_rate": 5.113488506577997e-05,
      "loss": 3.7924,
      "step": 179000
    },
    {
      "epoch": 12.610650555009133,
      "grad_norm": 1.0897852182388306,
      "learning_rate": 5.0990313719820736e-05,
      "loss": 3.7927,
      "step": 179500
    },
    {
      "epoch": 12.645777715329492,
      "grad_norm": 1.0974160432815552,
      "learning_rate": 5.0845742373861507e-05,
      "loss": 3.7935,
      "step": 180000
    },
    {
      "epoch": 12.645777715329492,
      "eval_accuracy": 0.6682870526036502,
      "eval_loss": 3.4848945140838623,
      "eval_runtime": 139.026,
      "eval_samples_per_second": 868.226,
      "eval_steps_per_second": 5.431,
      "step": 180000
    },
    {
      "epoch": 12.680904875649853,
      "grad_norm": 1.1538357734680176,
      "learning_rate": 5.070117102790227e-05,
      "loss": 3.791,
      "step": 180500
    },
    {
      "epoch": 12.716032035970212,
      "grad_norm": 1.1396054029464722,
      "learning_rate": 5.055659968194304e-05,
      "loss": 3.7921,
      "step": 181000
    },
    {
      "epoch": 12.751159196290573,
      "grad_norm": 1.0357263088226318,
      "learning_rate": 5.0412028335983805e-05,
      "loss": 3.791,
      "step": 181500
    },
    {
      "epoch": 12.786286356610931,
      "grad_norm": 1.1179182529449463,
      "learning_rate": 5.0267456990024576e-05,
      "loss": 3.7919,
      "step": 182000
    },
    {
      "epoch": 12.82141351693129,
      "grad_norm": 1.0229140520095825,
      "learning_rate": 5.0122885644065354e-05,
      "loss": 3.7939,
      "step": 182500
    },
    {
      "epoch": 12.856540677251651,
      "grad_norm": 1.1127890348434448,
      "learning_rate": 4.997831429810612e-05,
      "loss": 3.7887,
      "step": 183000
    },
    {
      "epoch": 12.89166783757201,
      "grad_norm": 1.084653377532959,
      "learning_rate": 4.983374295214689e-05,
      "loss": 3.7897,
      "step": 183500
    },
    {
      "epoch": 12.92679499789237,
      "grad_norm": 1.1412302255630493,
      "learning_rate": 4.968917160618766e-05,
      "loss": 3.7859,
      "step": 184000
    },
    {
      "epoch": 12.96192215821273,
      "grad_norm": 1.173107624053955,
      "learning_rate": 4.954460026022842e-05,
      "loss": 3.7911,
      "step": 184500
    },
    {
      "epoch": 12.99704931853309,
      "grad_norm": 1.0742864608764648,
      "learning_rate": 4.9400028914269194e-05,
      "loss": 3.7894,
      "step": 185000
    },
    {
      "epoch": 13.032176478853449,
      "grad_norm": 1.1034468412399292,
      "learning_rate": 4.9255457568309965e-05,
      "loss": 3.7836,
      "step": 185500
    },
    {
      "epoch": 13.06730363917381,
      "grad_norm": 1.0950945615768433,
      "learning_rate": 4.9110886222350736e-05,
      "loss": 3.7844,
      "step": 186000
    },
    {
      "epoch": 13.102430799494169,
      "grad_norm": 1.0757458209991455,
      "learning_rate": 4.89663148763915e-05,
      "loss": 3.7851,
      "step": 186500
    },
    {
      "epoch": 13.13755795981453,
      "grad_norm": 1.0367438793182373,
      "learning_rate": 4.882174353043227e-05,
      "loss": 3.7813,
      "step": 187000
    },
    {
      "epoch": 13.172685120134888,
      "grad_norm": 1.0611573457717896,
      "learning_rate": 4.867717218447304e-05,
      "loss": 3.7803,
      "step": 187500
    },
    {
      "epoch": 13.207812280455247,
      "grad_norm": 1.0296107530593872,
      "learning_rate": 4.853260083851381e-05,
      "loss": 3.7827,
      "step": 188000
    },
    {
      "epoch": 13.242939440775608,
      "grad_norm": 1.0659501552581787,
      "learning_rate": 4.8388029492554576e-05,
      "loss": 3.7825,
      "step": 188500
    },
    {
      "epoch": 13.278066601095967,
      "grad_norm": 1.042000412940979,
      "learning_rate": 4.824345814659535e-05,
      "loss": 3.7803,
      "step": 189000
    },
    {
      "epoch": 13.313193761416327,
      "grad_norm": 1.1316543817520142,
      "learning_rate": 4.809888680063612e-05,
      "loss": 3.7835,
      "step": 189500
    },
    {
      "epoch": 13.348320921736686,
      "grad_norm": 1.1366993188858032,
      "learning_rate": 4.795431545467688e-05,
      "loss": 3.7815,
      "step": 190000
    },
    {
      "epoch": 13.348320921736686,
      "eval_accuracy": 0.6694518509673556,
      "eval_loss": 3.478436231613159,
      "eval_runtime": 139.0366,
      "eval_samples_per_second": 868.16,
      "eval_steps_per_second": 5.43,
      "step": 190000
    },
    {
      "epoch": 13.383448082057047,
      "grad_norm": 1.1076951026916504,
      "learning_rate": 4.780974410871765e-05,
      "loss": 3.7764,
      "step": 190500
    },
    {
      "epoch": 13.418575242377406,
      "grad_norm": 1.1507115364074707,
      "learning_rate": 4.766517276275842e-05,
      "loss": 3.7813,
      "step": 191000
    },
    {
      "epoch": 13.453702402697767,
      "grad_norm": 1.0347645282745361,
      "learning_rate": 4.7520601416799194e-05,
      "loss": 3.7782,
      "step": 191500
    },
    {
      "epoch": 13.488829563018125,
      "grad_norm": 1.1612026691436768,
      "learning_rate": 4.737603007083996e-05,
      "loss": 3.7791,
      "step": 192000
    },
    {
      "epoch": 13.523956723338486,
      "grad_norm": 1.076613426208496,
      "learning_rate": 4.723145872488073e-05,
      "loss": 3.7797,
      "step": 192500
    },
    {
      "epoch": 13.559083883658845,
      "grad_norm": 1.1691707372665405,
      "learning_rate": 4.70868873789215e-05,
      "loss": 3.7779,
      "step": 193000
    },
    {
      "epoch": 13.594211043979204,
      "grad_norm": 1.1474121809005737,
      "learning_rate": 4.694231603296227e-05,
      "loss": 3.7792,
      "step": 193500
    },
    {
      "epoch": 13.629338204299565,
      "grad_norm": 1.121609091758728,
      "learning_rate": 4.6797744687003034e-05,
      "loss": 3.7739,
      "step": 194000
    },
    {
      "epoch": 13.664465364619923,
      "grad_norm": 1.0955735445022583,
      "learning_rate": 4.665317334104381e-05,
      "loss": 3.7746,
      "step": 194500
    },
    {
      "epoch": 13.699592524940284,
      "grad_norm": 1.0900901556015015,
      "learning_rate": 4.6508601995084576e-05,
      "loss": 3.7761,
      "step": 195000
    },
    {
      "epoch": 13.734719685260643,
      "grad_norm": 1.1831464767456055,
      "learning_rate": 4.6364030649125346e-05,
      "loss": 3.7751,
      "step": 195500
    },
    {
      "epoch": 13.769846845581004,
      "grad_norm": 1.0997892618179321,
      "learning_rate": 4.621945930316611e-05,
      "loss": 3.7761,
      "step": 196000
    },
    {
      "epoch": 13.804974005901363,
      "grad_norm": 1.0963380336761475,
      "learning_rate": 4.607488795720689e-05,
      "loss": 3.7724,
      "step": 196500
    },
    {
      "epoch": 13.840101166221723,
      "grad_norm": 1.162185549736023,
      "learning_rate": 4.593031661124765e-05,
      "loss": 3.7751,
      "step": 197000
    },
    {
      "epoch": 13.875228326542082,
      "grad_norm": 1.071937084197998,
      "learning_rate": 4.578574526528842e-05,
      "loss": 3.7708,
      "step": 197500
    },
    {
      "epoch": 13.910355486862443,
      "grad_norm": 1.1475390195846558,
      "learning_rate": 4.564117391932919e-05,
      "loss": 3.7717,
      "step": 198000
    },
    {
      "epoch": 13.945482647182802,
      "grad_norm": 1.047614336013794,
      "learning_rate": 4.5496602573369964e-05,
      "loss": 3.771,
      "step": 198500
    },
    {
      "epoch": 13.98060980750316,
      "grad_norm": 1.1338860988616943,
      "learning_rate": 4.535203122741073e-05,
      "loss": 3.7759,
      "step": 199000
    },
    {
      "epoch": 14.015736967823521,
      "grad_norm": 1.0949747562408447,
      "learning_rate": 4.52074598814515e-05,
      "loss": 3.7713,
      "step": 199500
    },
    {
      "epoch": 14.05086412814388,
      "grad_norm": 1.144314169883728,
      "learning_rate": 4.506288853549227e-05,
      "loss": 3.7719,
      "step": 200000
    },
    {
      "epoch": 14.05086412814388,
      "eval_accuracy": 0.6709508502461347,
      "eval_loss": 3.467146635055542,
      "eval_runtime": 136.415,
      "eval_samples_per_second": 884.844,
      "eval_steps_per_second": 5.535,
      "step": 200000
    },
    {
      "epoch": 14.08599128846424,
      "grad_norm": 1.12237548828125,
      "learning_rate": 4.491831718953304e-05,
      "loss": 3.7681,
      "step": 200500
    },
    {
      "epoch": 14.1211184487846,
      "grad_norm": 1.114652395248413,
      "learning_rate": 4.4773745843573805e-05,
      "loss": 3.7697,
      "step": 201000
    },
    {
      "epoch": 14.15624560910496,
      "grad_norm": 1.165330410003662,
      "learning_rate": 4.4629174497614575e-05,
      "loss": 3.7669,
      "step": 201500
    },
    {
      "epoch": 14.19137276942532,
      "grad_norm": 1.0363140106201172,
      "learning_rate": 4.4484603151655346e-05,
      "loss": 3.7652,
      "step": 202000
    },
    {
      "epoch": 14.22649992974568,
      "grad_norm": 1.0935860872268677,
      "learning_rate": 4.434003180569612e-05,
      "loss": 3.7666,
      "step": 202500
    },
    {
      "epoch": 14.261627090066039,
      "grad_norm": 1.089687705039978,
      "learning_rate": 4.419546045973688e-05,
      "loss": 3.7684,
      "step": 203000
    },
    {
      "epoch": 14.2967542503864,
      "grad_norm": 1.118787169456482,
      "learning_rate": 4.405088911377765e-05,
      "loss": 3.7666,
      "step": 203500
    },
    {
      "epoch": 14.331881410706758,
      "grad_norm": 1.1370214223861694,
      "learning_rate": 4.390631776781842e-05,
      "loss": 3.7693,
      "step": 204000
    },
    {
      "epoch": 14.367008571027117,
      "grad_norm": 1.1776578426361084,
      "learning_rate": 4.3761746421859187e-05,
      "loss": 3.7663,
      "step": 204500
    },
    {
      "epoch": 14.402135731347478,
      "grad_norm": 1.0615099668502808,
      "learning_rate": 4.361717507589996e-05,
      "loss": 3.7644,
      "step": 205000
    },
    {
      "epoch": 14.437262891667837,
      "grad_norm": 1.094874382019043,
      "learning_rate": 4.347260372994073e-05,
      "loss": 3.7671,
      "step": 205500
    },
    {
      "epoch": 14.472390051988198,
      "grad_norm": 1.0545638799667358,
      "learning_rate": 4.33280323839815e-05,
      "loss": 3.7642,
      "step": 206000
    },
    {
      "epoch": 14.507517212308557,
      "grad_norm": 1.0935437679290771,
      "learning_rate": 4.318346103802226e-05,
      "loss": 3.7627,
      "step": 206500
    },
    {
      "epoch": 14.542644372628917,
      "grad_norm": 1.0788770914077759,
      "learning_rate": 4.3038889692063034e-05,
      "loss": 3.7619,
      "step": 207000
    },
    {
      "epoch": 14.577771532949276,
      "grad_norm": 1.0779285430908203,
      "learning_rate": 4.2894318346103804e-05,
      "loss": 3.7616,
      "step": 207500
    },
    {
      "epoch": 14.612898693269637,
      "grad_norm": 1.1573363542556763,
      "learning_rate": 4.2749747000144575e-05,
      "loss": 3.7632,
      "step": 208000
    },
    {
      "epoch": 14.648025853589996,
      "grad_norm": 1.1256577968597412,
      "learning_rate": 4.260517565418534e-05,
      "loss": 3.762,
      "step": 208500
    },
    {
      "epoch": 14.683153013910356,
      "grad_norm": 1.1223344802856445,
      "learning_rate": 4.246060430822611e-05,
      "loss": 3.762,
      "step": 209000
    },
    {
      "epoch": 14.718280174230715,
      "grad_norm": 1.0763652324676514,
      "learning_rate": 4.231603296226688e-05,
      "loss": 3.7624,
      "step": 209500
    },
    {
      "epoch": 14.753407334551074,
      "grad_norm": 1.0835922956466675,
      "learning_rate": 4.217146161630765e-05,
      "loss": 3.7614,
      "step": 210000
    },
    {
      "epoch": 14.753407334551074,
      "eval_accuracy": 0.6724020681050233,
      "eval_loss": 3.4574368000030518,
      "eval_runtime": 134.2712,
      "eval_samples_per_second": 898.972,
      "eval_steps_per_second": 5.623,
      "step": 210000
    },
    {
      "epoch": 14.788534494871435,
      "grad_norm": 1.123087763786316,
      "learning_rate": 4.2026890270348416e-05,
      "loss": 3.7604,
      "step": 210500
    },
    {
      "epoch": 14.823661655191794,
      "grad_norm": 1.1097919940948486,
      "learning_rate": 4.1882318924389186e-05,
      "loss": 3.7583,
      "step": 211000
    },
    {
      "epoch": 14.858788815512154,
      "grad_norm": 1.0269588232040405,
      "learning_rate": 4.173774757842996e-05,
      "loss": 3.7607,
      "step": 211500
    },
    {
      "epoch": 14.893915975832513,
      "grad_norm": 1.1044747829437256,
      "learning_rate": 4.159317623247073e-05,
      "loss": 3.758,
      "step": 212000
    },
    {
      "epoch": 14.929043136152874,
      "grad_norm": 1.1890785694122314,
      "learning_rate": 4.144860488651149e-05,
      "loss": 3.7611,
      "step": 212500
    },
    {
      "epoch": 14.964170296473233,
      "grad_norm": 1.1136363744735718,
      "learning_rate": 4.130403354055227e-05,
      "loss": 3.7614,
      "step": 213000
    },
    {
      "epoch": 14.999297456793593,
      "grad_norm": 1.0864790678024292,
      "learning_rate": 4.1159462194593033e-05,
      "loss": 3.7589,
      "step": 213500
    },
    {
      "epoch": 15.034424617113952,
      "grad_norm": 1.122802495956421,
      "learning_rate": 4.1014890848633804e-05,
      "loss": 3.7549,
      "step": 214000
    },
    {
      "epoch": 15.069551777434313,
      "grad_norm": 1.045833945274353,
      "learning_rate": 4.087031950267457e-05,
      "loss": 3.7564,
      "step": 214500
    },
    {
      "epoch": 15.104678937754672,
      "grad_norm": 1.1709121465682983,
      "learning_rate": 4.0725748156715346e-05,
      "loss": 3.757,
      "step": 215000
    },
    {
      "epoch": 15.13980609807503,
      "grad_norm": 1.0737279653549194,
      "learning_rate": 4.058117681075611e-05,
      "loss": 3.7524,
      "step": 215500
    },
    {
      "epoch": 15.174933258395392,
      "grad_norm": 1.0796692371368408,
      "learning_rate": 4.043660546479688e-05,
      "loss": 3.7561,
      "step": 216000
    },
    {
      "epoch": 15.21006041871575,
      "grad_norm": 1.0271170139312744,
      "learning_rate": 4.0292034118837645e-05,
      "loss": 3.7574,
      "step": 216500
    },
    {
      "epoch": 15.245187579036111,
      "grad_norm": 1.0409847497940063,
      "learning_rate": 4.014746277287842e-05,
      "loss": 3.7543,
      "step": 217000
    },
    {
      "epoch": 15.28031473935647,
      "grad_norm": 1.113217830657959,
      "learning_rate": 4.0002891426919186e-05,
      "loss": 3.7537,
      "step": 217500
    },
    {
      "epoch": 15.31544189967683,
      "grad_norm": 1.226541519165039,
      "learning_rate": 3.985832008095996e-05,
      "loss": 3.7532,
      "step": 218000
    },
    {
      "epoch": 15.35056905999719,
      "grad_norm": 1.1041687726974487,
      "learning_rate": 3.971374873500072e-05,
      "loss": 3.7541,
      "step": 218500
    },
    {
      "epoch": 15.38569622031755,
      "grad_norm": 1.0691336393356323,
      "learning_rate": 3.95691773890415e-05,
      "loss": 3.7529,
      "step": 219000
    },
    {
      "epoch": 15.42082338063791,
      "grad_norm": 1.135039210319519,
      "learning_rate": 3.942460604308226e-05,
      "loss": 3.7538,
      "step": 219500
    },
    {
      "epoch": 15.45595054095827,
      "grad_norm": 1.1539758443832397,
      "learning_rate": 3.928003469712303e-05,
      "loss": 3.7509,
      "step": 220000
    },
    {
      "epoch": 15.45595054095827,
      "eval_accuracy": 0.6740203555555773,
      "eval_loss": 3.4487600326538086,
      "eval_runtime": 136.4267,
      "eval_samples_per_second": 884.768,
      "eval_steps_per_second": 5.534,
      "step": 220000
    },
    {
      "epoch": 15.491077701278629,
      "grad_norm": 1.036586880683899,
      "learning_rate": 3.9135463351163804e-05,
      "loss": 3.7533,
      "step": 220500
    },
    {
      "epoch": 15.526204861598988,
      "grad_norm": 1.0971709489822388,
      "learning_rate": 3.899089200520457e-05,
      "loss": 3.754,
      "step": 221000
    },
    {
      "epoch": 15.561332021919348,
      "grad_norm": 1.0935513973236084,
      "learning_rate": 3.884632065924534e-05,
      "loss": 3.751,
      "step": 221500
    },
    {
      "epoch": 15.596459182239707,
      "grad_norm": 1.0288581848144531,
      "learning_rate": 3.870174931328611e-05,
      "loss": 3.7519,
      "step": 222000
    },
    {
      "epoch": 15.631586342560068,
      "grad_norm": 1.0612101554870605,
      "learning_rate": 3.855717796732688e-05,
      "loss": 3.7492,
      "step": 222500
    },
    {
      "epoch": 15.666713502880427,
      "grad_norm": 1.1137163639068604,
      "learning_rate": 3.8412606621367644e-05,
      "loss": 3.7508,
      "step": 223000
    },
    {
      "epoch": 15.701840663200787,
      "grad_norm": 1.1050989627838135,
      "learning_rate": 3.8268035275408415e-05,
      "loss": 3.7503,
      "step": 223500
    },
    {
      "epoch": 15.736967823521146,
      "grad_norm": 1.0861667394638062,
      "learning_rate": 3.812346392944918e-05,
      "loss": 3.7487,
      "step": 224000
    },
    {
      "epoch": 15.772094983841507,
      "grad_norm": 1.169948697090149,
      "learning_rate": 3.797889258348996e-05,
      "loss": 3.7458,
      "step": 224500
    },
    {
      "epoch": 15.807222144161866,
      "grad_norm": 1.1400046348571777,
      "learning_rate": 3.783432123753072e-05,
      "loss": 3.7459,
      "step": 225000
    },
    {
      "epoch": 15.842349304482227,
      "grad_norm": 1.094469666481018,
      "learning_rate": 3.768974989157149e-05,
      "loss": 3.7459,
      "step": 225500
    },
    {
      "epoch": 15.877476464802585,
      "grad_norm": 1.0615177154541016,
      "learning_rate": 3.754517854561226e-05,
      "loss": 3.7502,
      "step": 226000
    },
    {
      "epoch": 15.912603625122944,
      "grad_norm": 1.1277542114257812,
      "learning_rate": 3.740060719965303e-05,
      "loss": 3.7495,
      "step": 226500
    },
    {
      "epoch": 15.947730785443305,
      "grad_norm": 1.0778846740722656,
      "learning_rate": 3.72560358536938e-05,
      "loss": 3.7461,
      "step": 227000
    },
    {
      "epoch": 15.982857945763664,
      "grad_norm": 1.1428011655807495,
      "learning_rate": 3.711146450773457e-05,
      "loss": 3.7475,
      "step": 227500
    },
    {
      "epoch": 16.017985106084023,
      "grad_norm": 1.1642730236053467,
      "learning_rate": 3.696689316177534e-05,
      "loss": 3.7459,
      "step": 228000
    },
    {
      "epoch": 16.053112266404383,
      "grad_norm": 1.0904488563537598,
      "learning_rate": 3.682232181581611e-05,
      "loss": 3.7432,
      "step": 228500
    },
    {
      "epoch": 16.088239426724744,
      "grad_norm": 1.189934253692627,
      "learning_rate": 3.667775046985687e-05,
      "loss": 3.7424,
      "step": 229000
    },
    {
      "epoch": 16.123366587045105,
      "grad_norm": 1.0590540170669556,
      "learning_rate": 3.6533179123897644e-05,
      "loss": 3.7443,
      "step": 229500
    },
    {
      "epoch": 16.158493747365462,
      "grad_norm": 1.1176244020462036,
      "learning_rate": 3.6388607777938415e-05,
      "loss": 3.7456,
      "step": 230000
    },
    {
      "epoch": 16.158493747365462,
      "eval_accuracy": 0.6744550930748897,
      "eval_loss": 3.4445083141326904,
      "eval_runtime": 135.5044,
      "eval_samples_per_second": 890.79,
      "eval_steps_per_second": 5.572,
      "step": 230000
    },
    {
      "epoch": 16.193620907685823,
      "grad_norm": 1.0993369817733765,
      "learning_rate": 3.6244036431979186e-05,
      "loss": 3.7406,
      "step": 230500
    },
    {
      "epoch": 16.228748068006183,
      "grad_norm": 1.0925554037094116,
      "learning_rate": 3.609946508601995e-05,
      "loss": 3.7432,
      "step": 231000
    },
    {
      "epoch": 16.26387522832654,
      "grad_norm": 1.2375068664550781,
      "learning_rate": 3.595489374006073e-05,
      "loss": 3.7397,
      "step": 231500
    },
    {
      "epoch": 16.2990023886469,
      "grad_norm": 1.1000934839248657,
      "learning_rate": 3.581032239410149e-05,
      "loss": 3.7432,
      "step": 232000
    },
    {
      "epoch": 16.33412954896726,
      "grad_norm": 1.1983340978622437,
      "learning_rate": 3.566575104814226e-05,
      "loss": 3.7408,
      "step": 232500
    },
    {
      "epoch": 16.369256709287622,
      "grad_norm": 1.1128774881362915,
      "learning_rate": 3.5521179702183026e-05,
      "loss": 3.7412,
      "step": 233000
    },
    {
      "epoch": 16.40438386960798,
      "grad_norm": 1.1636124849319458,
      "learning_rate": 3.5376608356223804e-05,
      "loss": 3.74,
      "step": 233500
    },
    {
      "epoch": 16.43951102992834,
      "grad_norm": 1.082008719444275,
      "learning_rate": 3.523203701026457e-05,
      "loss": 3.7394,
      "step": 234000
    },
    {
      "epoch": 16.4746381902487,
      "grad_norm": 1.1031757593154907,
      "learning_rate": 3.508746566430534e-05,
      "loss": 3.7424,
      "step": 234500
    },
    {
      "epoch": 16.50976535056906,
      "grad_norm": 1.0792458057403564,
      "learning_rate": 3.49428943183461e-05,
      "loss": 3.7391,
      "step": 235000
    },
    {
      "epoch": 16.54489251088942,
      "grad_norm": 1.128117561340332,
      "learning_rate": 3.479832297238687e-05,
      "loss": 3.7379,
      "step": 235500
    },
    {
      "epoch": 16.58001967120978,
      "grad_norm": 1.1126629114151,
      "learning_rate": 3.4653751626427644e-05,
      "loss": 3.7415,
      "step": 236000
    },
    {
      "epoch": 16.61514683153014,
      "grad_norm": 1.1816858053207397,
      "learning_rate": 3.4509180280468415e-05,
      "loss": 3.7388,
      "step": 236500
    },
    {
      "epoch": 16.650273991850497,
      "grad_norm": 1.2019652128219604,
      "learning_rate": 3.436460893450918e-05,
      "loss": 3.7436,
      "step": 237000
    },
    {
      "epoch": 16.685401152170858,
      "grad_norm": 1.11158287525177,
      "learning_rate": 3.422003758854995e-05,
      "loss": 3.7393,
      "step": 237500
    },
    {
      "epoch": 16.72052831249122,
      "grad_norm": 1.1717816591262817,
      "learning_rate": 3.407546624259072e-05,
      "loss": 3.7387,
      "step": 238000
    },
    {
      "epoch": 16.75565547281158,
      "grad_norm": 1.1399754285812378,
      "learning_rate": 3.393089489663149e-05,
      "loss": 3.7405,
      "step": 238500
    },
    {
      "epoch": 16.790782633131936,
      "grad_norm": 1.1949219703674316,
      "learning_rate": 3.378632355067226e-05,
      "loss": 3.7365,
      "step": 239000
    },
    {
      "epoch": 16.825909793452297,
      "grad_norm": 1.0693680047988892,
      "learning_rate": 3.3641752204713026e-05,
      "loss": 3.7377,
      "step": 239500
    },
    {
      "epoch": 16.861036953772658,
      "grad_norm": 1.1649973392486572,
      "learning_rate": 3.3497180858753797e-05,
      "loss": 3.736,
      "step": 240000
    },
    {
      "epoch": 16.861036953772658,
      "eval_accuracy": 0.6753275933579761,
      "eval_loss": 3.4378044605255127,
      "eval_runtime": 147.3318,
      "eval_samples_per_second": 819.28,
      "eval_steps_per_second": 5.124,
      "step": 240000
    },
    {
      "epoch": 16.89616411409302,
      "grad_norm": 1.2256054878234863,
      "learning_rate": 3.335260951279456e-05,
      "loss": 3.7354,
      "step": 240500
    },
    {
      "epoch": 16.931291274413375,
      "grad_norm": 1.1687901020050049,
      "learning_rate": 3.320803816683534e-05,
      "loss": 3.7399,
      "step": 241000
    },
    {
      "epoch": 16.966418434733736,
      "grad_norm": 1.1551425457000732,
      "learning_rate": 3.30634668208761e-05,
      "loss": 3.7345,
      "step": 241500
    },
    {
      "epoch": 17.001545595054097,
      "grad_norm": 1.160975456237793,
      "learning_rate": 3.291889547491687e-05,
      "loss": 3.7395,
      "step": 242000
    },
    {
      "epoch": 17.036672755374454,
      "grad_norm": 1.1405760049819946,
      "learning_rate": 3.277432412895764e-05,
      "loss": 3.7382,
      "step": 242500
    },
    {
      "epoch": 17.071799915694815,
      "grad_norm": 1.0844274759292603,
      "learning_rate": 3.2629752782998414e-05,
      "loss": 3.7338,
      "step": 243000
    },
    {
      "epoch": 17.106927076015175,
      "grad_norm": 1.2275018692016602,
      "learning_rate": 3.248518143703918e-05,
      "loss": 3.7349,
      "step": 243500
    },
    {
      "epoch": 17.142054236335536,
      "grad_norm": 1.1637654304504395,
      "learning_rate": 3.234061009107995e-05,
      "loss": 3.7318,
      "step": 244000
    },
    {
      "epoch": 17.177181396655893,
      "grad_norm": 1.1421692371368408,
      "learning_rate": 3.219603874512072e-05,
      "loss": 3.7326,
      "step": 244500
    },
    {
      "epoch": 17.212308556976254,
      "grad_norm": 1.1641732454299927,
      "learning_rate": 3.205146739916149e-05,
      "loss": 3.7338,
      "step": 245000
    },
    {
      "epoch": 17.247435717296614,
      "grad_norm": 1.0615341663360596,
      "learning_rate": 3.1906896053202255e-05,
      "loss": 3.7315,
      "step": 245500
    },
    {
      "epoch": 17.282562877616975,
      "grad_norm": 1.1757549047470093,
      "learning_rate": 3.1762324707243026e-05,
      "loss": 3.7352,
      "step": 246000
    },
    {
      "epoch": 17.317690037937332,
      "grad_norm": 1.0860717296600342,
      "learning_rate": 3.1617753361283796e-05,
      "loss": 3.7314,
      "step": 246500
    },
    {
      "epoch": 17.352817198257693,
      "grad_norm": 1.085434913635254,
      "learning_rate": 3.147318201532457e-05,
      "loss": 3.7307,
      "step": 247000
    },
    {
      "epoch": 17.387944358578054,
      "grad_norm": 1.1140954494476318,
      "learning_rate": 3.132861066936533e-05,
      "loss": 3.7336,
      "step": 247500
    },
    {
      "epoch": 17.42307151889841,
      "grad_norm": 1.0659854412078857,
      "learning_rate": 3.11840393234061e-05,
      "loss": 3.7261,
      "step": 248000
    },
    {
      "epoch": 17.45819867921877,
      "grad_norm": 1.1735544204711914,
      "learning_rate": 3.103946797744687e-05,
      "loss": 3.7299,
      "step": 248500
    },
    {
      "epoch": 17.493325839539132,
      "grad_norm": 1.0841625928878784,
      "learning_rate": 3.0894896631487643e-05,
      "loss": 3.7279,
      "step": 249000
    },
    {
      "epoch": 17.528452999859493,
      "grad_norm": 1.1285041570663452,
      "learning_rate": 3.075032528552841e-05,
      "loss": 3.7275,
      "step": 249500
    },
    {
      "epoch": 17.56358016017985,
      "grad_norm": 1.1707388162612915,
      "learning_rate": 3.060575393956918e-05,
      "loss": 3.728,
      "step": 250000
    },
    {
      "epoch": 17.56358016017985,
      "eval_accuracy": 0.6762921466145205,
      "eval_loss": 3.4329540729522705,
      "eval_runtime": 139.2714,
      "eval_samples_per_second": 866.696,
      "eval_steps_per_second": 5.421,
      "step": 250000
    },
    {
      "epoch": 17.59870732050021,
      "grad_norm": 1.1041512489318848,
      "learning_rate": 3.046118259360995e-05,
      "loss": 3.7328,
      "step": 250500
    },
    {
      "epoch": 17.63383448082057,
      "grad_norm": 1.1091268062591553,
      "learning_rate": 3.0316611247650716e-05,
      "loss": 3.7256,
      "step": 251000
    },
    {
      "epoch": 17.668961641140932,
      "grad_norm": 1.0993403196334839,
      "learning_rate": 3.0172039901691484e-05,
      "loss": 3.7298,
      "step": 251500
    },
    {
      "epoch": 17.70408880146129,
      "grad_norm": 1.0888632535934448,
      "learning_rate": 3.0027468555732258e-05,
      "loss": 3.7292,
      "step": 252000
    },
    {
      "epoch": 17.73921596178165,
      "grad_norm": 1.1624661684036255,
      "learning_rate": 2.9882897209773025e-05,
      "loss": 3.7275,
      "step": 252500
    },
    {
      "epoch": 17.77434312210201,
      "grad_norm": 1.1163383722305298,
      "learning_rate": 2.9738325863813793e-05,
      "loss": 3.7274,
      "step": 253000
    },
    {
      "epoch": 17.809470282422367,
      "grad_norm": 1.1039903163909912,
      "learning_rate": 2.959375451785456e-05,
      "loss": 3.7292,
      "step": 253500
    },
    {
      "epoch": 17.844597442742728,
      "grad_norm": 1.1737174987792969,
      "learning_rate": 2.9449183171895334e-05,
      "loss": 3.726,
      "step": 254000
    },
    {
      "epoch": 17.87972460306309,
      "grad_norm": 1.0892473459243774,
      "learning_rate": 2.93046118259361e-05,
      "loss": 3.7291,
      "step": 254500
    },
    {
      "epoch": 17.91485176338345,
      "grad_norm": 1.0831233263015747,
      "learning_rate": 2.916004047997687e-05,
      "loss": 3.7283,
      "step": 255000
    },
    {
      "epoch": 17.949978923703807,
      "grad_norm": 1.2125272750854492,
      "learning_rate": 2.9015469134017636e-05,
      "loss": 3.7286,
      "step": 255500
    },
    {
      "epoch": 17.985106084024167,
      "grad_norm": 1.1826454401016235,
      "learning_rate": 2.887089778805841e-05,
      "loss": 3.7257,
      "step": 256000
    },
    {
      "epoch": 18.020233244344528,
      "grad_norm": 1.0991463661193848,
      "learning_rate": 2.8726326442099178e-05,
      "loss": 3.7279,
      "step": 256500
    },
    {
      "epoch": 18.05536040466489,
      "grad_norm": 1.1094375848770142,
      "learning_rate": 2.8581755096139945e-05,
      "loss": 3.7249,
      "step": 257000
    },
    {
      "epoch": 18.090487564985246,
      "grad_norm": 1.1255062818527222,
      "learning_rate": 2.843718375018072e-05,
      "loss": 3.7224,
      "step": 257500
    },
    {
      "epoch": 18.125614725305606,
      "grad_norm": 1.1805857419967651,
      "learning_rate": 2.8292612404221487e-05,
      "loss": 3.7252,
      "step": 258000
    },
    {
      "epoch": 18.160741885625967,
      "grad_norm": 1.1080857515335083,
      "learning_rate": 2.8148041058262254e-05,
      "loss": 3.7242,
      "step": 258500
    },
    {
      "epoch": 18.195869045946324,
      "grad_norm": 1.0906721353530884,
      "learning_rate": 2.800346971230302e-05,
      "loss": 3.7257,
      "step": 259000
    },
    {
      "epoch": 18.230996206266685,
      "grad_norm": 1.1306054592132568,
      "learning_rate": 2.7858898366343792e-05,
      "loss": 3.7258,
      "step": 259500
    },
    {
      "epoch": 18.266123366587045,
      "grad_norm": 1.2493983507156372,
      "learning_rate": 2.7714327020384563e-05,
      "loss": 3.7223,
      "step": 260000
    },
    {
      "epoch": 18.266123366587045,
      "eval_accuracy": 0.6771566661947445,
      "eval_loss": 3.4269628524780273,
      "eval_runtime": 138.9975,
      "eval_samples_per_second": 868.404,
      "eval_steps_per_second": 5.432,
      "step": 260000
    },
    {
      "epoch": 18.301250526907406,
      "grad_norm": 1.1990478038787842,
      "learning_rate": 2.756975567442533e-05,
      "loss": 3.7225,
      "step": 260500
    },
    {
      "epoch": 18.336377687227763,
      "grad_norm": 1.1356704235076904,
      "learning_rate": 2.7425184328466098e-05,
      "loss": 3.7205,
      "step": 261000
    },
    {
      "epoch": 18.371504847548124,
      "grad_norm": 1.0688560009002686,
      "learning_rate": 2.728061298250687e-05,
      "loss": 3.7249,
      "step": 261500
    },
    {
      "epoch": 18.406632007868485,
      "grad_norm": 1.089322566986084,
      "learning_rate": 2.7136041636547636e-05,
      "loss": 3.723,
      "step": 262000
    },
    {
      "epoch": 18.441759168188845,
      "grad_norm": 1.2481943368911743,
      "learning_rate": 2.6991470290588407e-05,
      "loss": 3.719,
      "step": 262500
    },
    {
      "epoch": 18.476886328509202,
      "grad_norm": 1.0918136835098267,
      "learning_rate": 2.6846898944629178e-05,
      "loss": 3.7179,
      "step": 263000
    },
    {
      "epoch": 18.512013488829563,
      "grad_norm": 1.190004825592041,
      "learning_rate": 2.6702327598669945e-05,
      "loss": 3.7188,
      "step": 263500
    },
    {
      "epoch": 18.547140649149924,
      "grad_norm": 1.1181012392044067,
      "learning_rate": 2.6557756252710713e-05,
      "loss": 3.7188,
      "step": 264000
    },
    {
      "epoch": 18.58226780947028,
      "grad_norm": 1.0874921083450317,
      "learning_rate": 2.641318490675148e-05,
      "loss": 3.7183,
      "step": 264500
    },
    {
      "epoch": 18.61739496979064,
      "grad_norm": 1.209376335144043,
      "learning_rate": 2.6268613560792254e-05,
      "loss": 3.7206,
      "step": 265000
    },
    {
      "epoch": 18.652522130111002,
      "grad_norm": 1.1031169891357422,
      "learning_rate": 2.612404221483302e-05,
      "loss": 3.7215,
      "step": 265500
    },
    {
      "epoch": 18.687649290431363,
      "grad_norm": 1.1632440090179443,
      "learning_rate": 2.597947086887379e-05,
      "loss": 3.7189,
      "step": 266000
    },
    {
      "epoch": 18.72277645075172,
      "grad_norm": 1.1632835865020752,
      "learning_rate": 2.5834899522914556e-05,
      "loss": 3.7221,
      "step": 266500
    },
    {
      "epoch": 18.75790361107208,
      "grad_norm": 1.1051679849624634,
      "learning_rate": 2.569032817695533e-05,
      "loss": 3.7195,
      "step": 267000
    },
    {
      "epoch": 18.79303077139244,
      "grad_norm": 1.134438157081604,
      "learning_rate": 2.5545756830996098e-05,
      "loss": 3.7214,
      "step": 267500
    },
    {
      "epoch": 18.828157931712802,
      "grad_norm": 1.124955654144287,
      "learning_rate": 2.5401185485036865e-05,
      "loss": 3.7189,
      "step": 268000
    },
    {
      "epoch": 18.86328509203316,
      "grad_norm": 1.1189135313034058,
      "learning_rate": 2.5256614139077633e-05,
      "loss": 3.7182,
      "step": 268500
    },
    {
      "epoch": 18.89841225235352,
      "grad_norm": 1.2001628875732422,
      "learning_rate": 2.5112042793118407e-05,
      "loss": 3.7186,
      "step": 269000
    },
    {
      "epoch": 18.93353941267388,
      "grad_norm": 1.1196178197860718,
      "learning_rate": 2.4967471447159174e-05,
      "loss": 3.7174,
      "step": 269500
    },
    {
      "epoch": 18.968666572994238,
      "grad_norm": 1.0874722003936768,
      "learning_rate": 2.4822900101199945e-05,
      "loss": 3.7195,
      "step": 270000
    },
    {
      "epoch": 18.968666572994238,
      "eval_accuracy": 0.6780094265912249,
      "eval_loss": 3.421011447906494,
      "eval_runtime": 135.7318,
      "eval_samples_per_second": 889.298,
      "eval_steps_per_second": 5.562,
      "step": 270000
    },
    {
      "epoch": 19.0037937333146,
      "grad_norm": 1.1497821807861328,
      "learning_rate": 2.4678328755240712e-05,
      "loss": 3.719,
      "step": 270500
    },
    {
      "epoch": 19.03892089363496,
      "grad_norm": 1.166760802268982,
      "learning_rate": 2.4533757409281483e-05,
      "loss": 3.7147,
      "step": 271000
    },
    {
      "epoch": 19.07404805395532,
      "grad_norm": 1.112505316734314,
      "learning_rate": 2.438918606332225e-05,
      "loss": 3.7169,
      "step": 271500
    },
    {
      "epoch": 19.109175214275677,
      "grad_norm": 1.0959382057189941,
      "learning_rate": 2.424461471736302e-05,
      "loss": 3.7152,
      "step": 272000
    },
    {
      "epoch": 19.144302374596037,
      "grad_norm": 1.1270334720611572,
      "learning_rate": 2.410004337140379e-05,
      "loss": 3.7155,
      "step": 272500
    },
    {
      "epoch": 19.179429534916398,
      "grad_norm": 1.2032307386398315,
      "learning_rate": 2.395547202544456e-05,
      "loss": 3.7152,
      "step": 273000
    },
    {
      "epoch": 19.21455669523676,
      "grad_norm": 1.2107353210449219,
      "learning_rate": 2.3810900679485327e-05,
      "loss": 3.7132,
      "step": 273500
    },
    {
      "epoch": 19.249683855557116,
      "grad_norm": 1.0875844955444336,
      "learning_rate": 2.3666329333526098e-05,
      "loss": 3.7146,
      "step": 274000
    },
    {
      "epoch": 19.284811015877477,
      "grad_norm": 1.1689175367355347,
      "learning_rate": 2.3521757987566865e-05,
      "loss": 3.7141,
      "step": 274500
    },
    {
      "epoch": 19.319938176197837,
      "grad_norm": 1.15033757686615,
      "learning_rate": 2.3377186641607636e-05,
      "loss": 3.7155,
      "step": 275000
    },
    {
      "epoch": 19.355065336518194,
      "grad_norm": 1.1002916097640991,
      "learning_rate": 2.3232615295648403e-05,
      "loss": 3.7166,
      "step": 275500
    },
    {
      "epoch": 19.390192496838555,
      "grad_norm": 1.0819830894470215,
      "learning_rate": 2.3088043949689174e-05,
      "loss": 3.7114,
      "step": 276000
    },
    {
      "epoch": 19.425319657158916,
      "grad_norm": 1.110970377922058,
      "learning_rate": 2.2943472603729945e-05,
      "loss": 3.7161,
      "step": 276500
    },
    {
      "epoch": 19.460446817479276,
      "grad_norm": 1.1547013521194458,
      "learning_rate": 2.2798901257770712e-05,
      "loss": 3.7109,
      "step": 277000
    },
    {
      "epoch": 19.495573977799634,
      "grad_norm": 1.1552664041519165,
      "learning_rate": 2.265432991181148e-05,
      "loss": 3.7151,
      "step": 277500
    },
    {
      "epoch": 19.530701138119994,
      "grad_norm": 1.1739439964294434,
      "learning_rate": 2.250975856585225e-05,
      "loss": 3.7121,
      "step": 278000
    },
    {
      "epoch": 19.565828298440355,
      "grad_norm": 1.1144962310791016,
      "learning_rate": 2.2365187219893018e-05,
      "loss": 3.7142,
      "step": 278500
    },
    {
      "epoch": 19.600955458760716,
      "grad_norm": 1.122421145439148,
      "learning_rate": 2.2220615873933785e-05,
      "loss": 3.7112,
      "step": 279000
    },
    {
      "epoch": 19.636082619081073,
      "grad_norm": 1.150740146636963,
      "learning_rate": 2.2076044527974556e-05,
      "loss": 3.7137,
      "step": 279500
    },
    {
      "epoch": 19.671209779401433,
      "grad_norm": 1.1842594146728516,
      "learning_rate": 2.1931473182015323e-05,
      "loss": 3.7104,
      "step": 280000
    },
    {
      "epoch": 19.671209779401433,
      "eval_accuracy": 0.6789739412605793,
      "eval_loss": 3.4155983924865723,
      "eval_runtime": 139.0053,
      "eval_samples_per_second": 868.356,
      "eval_steps_per_second": 5.431,
      "step": 280000
    },
    {
      "epoch": 19.706336939721794,
      "grad_norm": 1.103110671043396,
      "learning_rate": 2.1786901836056094e-05,
      "loss": 3.7144,
      "step": 280500
    },
    {
      "epoch": 19.74146410004215,
      "grad_norm": 1.0898276567459106,
      "learning_rate": 2.164233049009686e-05,
      "loss": 3.7129,
      "step": 281000
    },
    {
      "epoch": 19.776591260362512,
      "grad_norm": 1.1253899335861206,
      "learning_rate": 2.1497759144137632e-05,
      "loss": 3.7117,
      "step": 281500
    },
    {
      "epoch": 19.811718420682872,
      "grad_norm": 1.1312872171401978,
      "learning_rate": 2.1353187798178403e-05,
      "loss": 3.7105,
      "step": 282000
    },
    {
      "epoch": 19.846845581003233,
      "grad_norm": 1.2385282516479492,
      "learning_rate": 2.120861645221917e-05,
      "loss": 3.7113,
      "step": 282500
    },
    {
      "epoch": 19.88197274132359,
      "grad_norm": 1.1235606670379639,
      "learning_rate": 2.106404510625994e-05,
      "loss": 3.7094,
      "step": 283000
    },
    {
      "epoch": 19.91709990164395,
      "grad_norm": 1.2216262817382812,
      "learning_rate": 2.091947376030071e-05,
      "loss": 3.7147,
      "step": 283500
    },
    {
      "epoch": 19.95222706196431,
      "grad_norm": 1.1158368587493896,
      "learning_rate": 2.077490241434148e-05,
      "loss": 3.7087,
      "step": 284000
    },
    {
      "epoch": 19.987354222284672,
      "grad_norm": 1.1191985607147217,
      "learning_rate": 2.0630331068382247e-05,
      "loss": 3.7104,
      "step": 284500
    },
    {
      "epoch": 20.02248138260503,
      "grad_norm": 1.1592704057693481,
      "learning_rate": 2.0485759722423017e-05,
      "loss": 3.7128,
      "step": 285000
    },
    {
      "epoch": 20.05760854292539,
      "grad_norm": 1.1146634817123413,
      "learning_rate": 2.0341188376463785e-05,
      "loss": 3.7073,
      "step": 285500
    },
    {
      "epoch": 20.09273570324575,
      "grad_norm": 1.1642574071884155,
      "learning_rate": 2.0196617030504556e-05,
      "loss": 3.7073,
      "step": 286000
    },
    {
      "epoch": 20.127862863566108,
      "grad_norm": 1.0674467086791992,
      "learning_rate": 2.0052045684545323e-05,
      "loss": 3.707,
      "step": 286500
    },
    {
      "epoch": 20.16299002388647,
      "grad_norm": 1.1526765823364258,
      "learning_rate": 1.9907474338586094e-05,
      "loss": 3.7079,
      "step": 287000
    },
    {
      "epoch": 20.19811718420683,
      "grad_norm": 1.11869215965271,
      "learning_rate": 1.976290299262686e-05,
      "loss": 3.7074,
      "step": 287500
    },
    {
      "epoch": 20.23324434452719,
      "grad_norm": 1.1538423299789429,
      "learning_rate": 1.9618331646667632e-05,
      "loss": 3.708,
      "step": 288000
    },
    {
      "epoch": 20.268371504847547,
      "grad_norm": 1.2199963331222534,
      "learning_rate": 1.9473760300708403e-05,
      "loss": 3.7048,
      "step": 288500
    },
    {
      "epoch": 20.303498665167908,
      "grad_norm": 1.1896425485610962,
      "learning_rate": 1.932918895474917e-05,
      "loss": 3.7088,
      "step": 289000
    },
    {
      "epoch": 20.33862582548827,
      "grad_norm": 1.1344529390335083,
      "learning_rate": 1.918461760878994e-05,
      "loss": 3.7076,
      "step": 289500
    },
    {
      "epoch": 20.37375298580863,
      "grad_norm": 1.134081482887268,
      "learning_rate": 1.9040046262830708e-05,
      "loss": 3.7086,
      "step": 290000
    },
    {
      "epoch": 20.37375298580863,
      "eval_accuracy": 0.6796801377476835,
      "eval_loss": 3.4104647636413574,
      "eval_runtime": 135.3877,
      "eval_samples_per_second": 891.558,
      "eval_steps_per_second": 5.577,
      "step": 290000
    },
    {
      "epoch": 20.408880146128986,
      "grad_norm": 1.1230337619781494,
      "learning_rate": 1.889547491687148e-05,
      "loss": 3.7059,
      "step": 290500
    },
    {
      "epoch": 20.444007306449347,
      "grad_norm": 1.1675646305084229,
      "learning_rate": 1.8750903570912246e-05,
      "loss": 3.7082,
      "step": 291000
    },
    {
      "epoch": 20.479134466769708,
      "grad_norm": 1.174786925315857,
      "learning_rate": 1.8606332224953017e-05,
      "loss": 3.7032,
      "step": 291500
    },
    {
      "epoch": 20.514261627090065,
      "grad_norm": 1.1421759128570557,
      "learning_rate": 1.8461760878993785e-05,
      "loss": 3.7056,
      "step": 292000
    },
    {
      "epoch": 20.549388787410425,
      "grad_norm": 1.1594599485397339,
      "learning_rate": 1.8317189533034555e-05,
      "loss": 3.7045,
      "step": 292500
    },
    {
      "epoch": 20.584515947730786,
      "grad_norm": 1.1131938695907593,
      "learning_rate": 1.8172618187075323e-05,
      "loss": 3.7061,
      "step": 293000
    },
    {
      "epoch": 20.619643108051147,
      "grad_norm": 1.1961661577224731,
      "learning_rate": 1.8028046841116093e-05,
      "loss": 3.7078,
      "step": 293500
    },
    {
      "epoch": 20.654770268371504,
      "grad_norm": 1.0960444211959839,
      "learning_rate": 1.788347549515686e-05,
      "loss": 3.7034,
      "step": 294000
    },
    {
      "epoch": 20.689897428691864,
      "grad_norm": 1.1319866180419922,
      "learning_rate": 1.7738904149197628e-05,
      "loss": 3.7053,
      "step": 294500
    },
    {
      "epoch": 20.725024589012225,
      "grad_norm": 1.16361403465271,
      "learning_rate": 1.75943328032384e-05,
      "loss": 3.7053,
      "step": 295000
    },
    {
      "epoch": 20.760151749332586,
      "grad_norm": 1.1448746919631958,
      "learning_rate": 1.7449761457279166e-05,
      "loss": 3.7062,
      "step": 295500
    },
    {
      "epoch": 20.795278909652943,
      "grad_norm": 1.0977457761764526,
      "learning_rate": 1.7305190111319937e-05,
      "loss": 3.7028,
      "step": 296000
    },
    {
      "epoch": 20.830406069973304,
      "grad_norm": 1.1030011177062988,
      "learning_rate": 1.7160618765360705e-05,
      "loss": 3.7036,
      "step": 296500
    },
    {
      "epoch": 20.865533230293664,
      "grad_norm": 1.0698529481887817,
      "learning_rate": 1.7016047419401475e-05,
      "loss": 3.7017,
      "step": 297000
    },
    {
      "epoch": 20.90066039061402,
      "grad_norm": 1.0933773517608643,
      "learning_rate": 1.6871476073442243e-05,
      "loss": 3.7061,
      "step": 297500
    },
    {
      "epoch": 20.935787550934382,
      "grad_norm": 1.1930452585220337,
      "learning_rate": 1.6726904727483014e-05,
      "loss": 3.7048,
      "step": 298000
    },
    {
      "epoch": 20.970914711254743,
      "grad_norm": 1.1529662609100342,
      "learning_rate": 1.658233338152378e-05,
      "loss": 3.7051,
      "step": 298500
    },
    {
      "epoch": 21.006041871575103,
      "grad_norm": 1.1338509321212769,
      "learning_rate": 1.6437762035564552e-05,
      "loss": 3.7037,
      "step": 299000
    },
    {
      "epoch": 21.04116903189546,
      "grad_norm": 1.1432526111602783,
      "learning_rate": 1.629319068960532e-05,
      "loss": 3.7043,
      "step": 299500
    },
    {
      "epoch": 21.07629619221582,
      "grad_norm": 1.0891770124435425,
      "learning_rate": 1.614861934364609e-05,
      "loss": 3.7002,
      "step": 300000
    },
    {
      "epoch": 21.07629619221582,
      "eval_accuracy": 0.6803094698179593,
      "eval_loss": 3.4070346355438232,
      "eval_runtime": 135.9572,
      "eval_samples_per_second": 887.824,
      "eval_steps_per_second": 5.553,
      "step": 300000
    },
    {
      "epoch": 21.111423352536182,
      "grad_norm": 1.1525535583496094,
      "learning_rate": 1.6004047997686857e-05,
      "loss": 3.7015,
      "step": 300500
    },
    {
      "epoch": 21.14655051285654,
      "grad_norm": 1.1834564208984375,
      "learning_rate": 1.5859476651727628e-05,
      "loss": 3.7037,
      "step": 301000
    },
    {
      "epoch": 21.1816776731769,
      "grad_norm": 1.1772528886795044,
      "learning_rate": 1.57149053057684e-05,
      "loss": 3.7023,
      "step": 301500
    },
    {
      "epoch": 21.21680483349726,
      "grad_norm": 1.163122296333313,
      "learning_rate": 1.5570333959809166e-05,
      "loss": 3.7057,
      "step": 302000
    },
    {
      "epoch": 21.25193199381762,
      "grad_norm": 1.0979679822921753,
      "learning_rate": 1.5425762613849937e-05,
      "loss": 3.6989,
      "step": 302500
    },
    {
      "epoch": 21.287059154137978,
      "grad_norm": 1.1861684322357178,
      "learning_rate": 1.5281191267890704e-05,
      "loss": 3.7015,
      "step": 303000
    },
    {
      "epoch": 21.32218631445834,
      "grad_norm": 1.112101674079895,
      "learning_rate": 1.5136619921931475e-05,
      "loss": 3.7004,
      "step": 303500
    },
    {
      "epoch": 21.3573134747787,
      "grad_norm": 1.1137944459915161,
      "learning_rate": 1.4992048575972243e-05,
      "loss": 3.6977,
      "step": 304000
    },
    {
      "epoch": 21.39244063509906,
      "grad_norm": 1.1729108095169067,
      "learning_rate": 1.4847477230013013e-05,
      "loss": 3.7014,
      "step": 304500
    },
    {
      "epoch": 21.427567795419417,
      "grad_norm": 1.1489734649658203,
      "learning_rate": 1.470290588405378e-05,
      "loss": 3.698,
      "step": 305000
    },
    {
      "epoch": 21.462694955739778,
      "grad_norm": 1.221509575843811,
      "learning_rate": 1.4558334538094551e-05,
      "loss": 3.6999,
      "step": 305500
    },
    {
      "epoch": 21.49782211606014,
      "grad_norm": 1.23967444896698,
      "learning_rate": 1.4413763192135319e-05,
      "loss": 3.6997,
      "step": 306000
    },
    {
      "epoch": 21.532949276380496,
      "grad_norm": 1.2451962232589722,
      "learning_rate": 1.4269191846176088e-05,
      "loss": 3.6991,
      "step": 306500
    },
    {
      "epoch": 21.568076436700856,
      "grad_norm": 1.192785620689392,
      "learning_rate": 1.4124620500216857e-05,
      "loss": 3.6975,
      "step": 307000
    },
    {
      "epoch": 21.603203597021217,
      "grad_norm": 1.1372467279434204,
      "learning_rate": 1.3980049154257626e-05,
      "loss": 3.6983,
      "step": 307500
    },
    {
      "epoch": 21.638330757341578,
      "grad_norm": 1.20252525806427,
      "learning_rate": 1.3835477808298397e-05,
      "loss": 3.7064,
      "step": 308000
    },
    {
      "epoch": 21.673457917661935,
      "grad_norm": 1.1753140687942505,
      "learning_rate": 1.3690906462339164e-05,
      "loss": 3.7002,
      "step": 308500
    },
    {
      "epoch": 21.708585077982296,
      "grad_norm": 1.142754316329956,
      "learning_rate": 1.3546335116379935e-05,
      "loss": 3.6979,
      "step": 309000
    },
    {
      "epoch": 21.743712238302656,
      "grad_norm": 1.1272801160812378,
      "learning_rate": 1.3401763770420702e-05,
      "loss": 3.6971,
      "step": 309500
    },
    {
      "epoch": 21.778839398623017,
      "grad_norm": 1.1040281057357788,
      "learning_rate": 1.3257192424461473e-05,
      "loss": 3.698,
      "step": 310000
    },
    {
      "epoch": 21.778839398623017,
      "eval_accuracy": 0.6811742246203961,
      "eval_loss": 3.401301145553589,
      "eval_runtime": 136.6607,
      "eval_samples_per_second": 883.253,
      "eval_steps_per_second": 5.525,
      "step": 310000
    },
    {
      "epoch": 21.813966558943374,
      "grad_norm": 1.1779478788375854,
      "learning_rate": 1.311262107850224e-05,
      "loss": 3.6953,
      "step": 310500
    },
    {
      "epoch": 21.849093719263735,
      "grad_norm": 1.2097868919372559,
      "learning_rate": 1.2968049732543011e-05,
      "loss": 3.6949,
      "step": 311000
    },
    {
      "epoch": 21.884220879584095,
      "grad_norm": 1.1302045583724976,
      "learning_rate": 1.2823478386583779e-05,
      "loss": 3.6998,
      "step": 311500
    },
    {
      "epoch": 21.919348039904452,
      "grad_norm": 1.0829895734786987,
      "learning_rate": 1.267890704062455e-05,
      "loss": 3.6946,
      "step": 312000
    },
    {
      "epoch": 21.954475200224813,
      "grad_norm": 1.1204338073730469,
      "learning_rate": 1.2534335694665317e-05,
      "loss": 3.698,
      "step": 312500
    },
    {
      "epoch": 21.989602360545174,
      "grad_norm": 1.1794413328170776,
      "learning_rate": 1.2389764348706088e-05,
      "loss": 3.6991,
      "step": 313000
    },
    {
      "epoch": 22.024729520865534,
      "grad_norm": 1.151856780052185,
      "learning_rate": 1.2245193002746857e-05,
      "loss": 3.6959,
      "step": 313500
    },
    {
      "epoch": 22.05985668118589,
      "grad_norm": 1.172200322151184,
      "learning_rate": 1.2100621656787626e-05,
      "loss": 3.6929,
      "step": 314000
    },
    {
      "epoch": 22.094983841506252,
      "grad_norm": 1.1896703243255615,
      "learning_rate": 1.1956050310828395e-05,
      "loss": 3.6986,
      "step": 314500
    },
    {
      "epoch": 22.130111001826613,
      "grad_norm": 1.1245017051696777,
      "learning_rate": 1.1811478964869162e-05,
      "loss": 3.6945,
      "step": 315000
    },
    {
      "epoch": 22.165238162146974,
      "grad_norm": 1.1526340246200562,
      "learning_rate": 1.1666907618909931e-05,
      "loss": 3.6914,
      "step": 315500
    },
    {
      "epoch": 22.20036532246733,
      "grad_norm": 1.0916378498077393,
      "learning_rate": 1.15223362729507e-05,
      "loss": 3.6925,
      "step": 316000
    },
    {
      "epoch": 22.23549248278769,
      "grad_norm": 1.1332939863204956,
      "learning_rate": 1.137776492699147e-05,
      "loss": 3.6958,
      "step": 316500
    },
    {
      "epoch": 22.270619643108052,
      "grad_norm": 1.162461280822754,
      "learning_rate": 1.123319358103224e-05,
      "loss": 3.6953,
      "step": 317000
    },
    {
      "epoch": 22.30574680342841,
      "grad_norm": 1.13168466091156,
      "learning_rate": 1.108862223507301e-05,
      "loss": 3.6952,
      "step": 317500
    },
    {
      "epoch": 22.34087396374877,
      "grad_norm": 1.085618495941162,
      "learning_rate": 1.0944050889113779e-05,
      "loss": 3.6935,
      "step": 318000
    },
    {
      "epoch": 22.37600112406913,
      "grad_norm": 1.1397286653518677,
      "learning_rate": 1.0799479543154548e-05,
      "loss": 3.6963,
      "step": 318500
    },
    {
      "epoch": 22.41112828438949,
      "grad_norm": 1.1780164241790771,
      "learning_rate": 1.0654908197195317e-05,
      "loss": 3.6935,
      "step": 319000
    },
    {
      "epoch": 22.44625544470985,
      "grad_norm": 1.1332197189331055,
      "learning_rate": 1.0510336851236086e-05,
      "loss": 3.6918,
      "step": 319500
    },
    {
      "epoch": 22.48138260503021,
      "grad_norm": 1.2165111303329468,
      "learning_rate": 1.0365765505276855e-05,
      "loss": 3.6915,
      "step": 320000
    },
    {
      "epoch": 22.48138260503021,
      "eval_accuracy": 0.6813662909472582,
      "eval_loss": 3.398728370666504,
      "eval_runtime": 136.4626,
      "eval_samples_per_second": 884.535,
      "eval_steps_per_second": 5.533,
      "step": 320000
    },
    {
      "epoch": 22.51650976535057,
      "grad_norm": 1.1656314134597778,
      "learning_rate": 1.0221194159317624e-05,
      "loss": 3.6948,
      "step": 320500
    },
    {
      "epoch": 22.55163692567093,
      "grad_norm": 1.1767264604568481,
      "learning_rate": 1.0076622813358393e-05,
      "loss": 3.6989,
      "step": 321000
    },
    {
      "epoch": 22.586764085991287,
      "grad_norm": 1.1991188526153564,
      "learning_rate": 9.932051467399162e-06,
      "loss": 3.6952,
      "step": 321500
    },
    {
      "epoch": 22.621891246311648,
      "grad_norm": 1.197310447692871,
      "learning_rate": 9.787480121439931e-06,
      "loss": 3.6924,
      "step": 322000
    },
    {
      "epoch": 22.65701840663201,
      "grad_norm": 1.1502188444137573,
      "learning_rate": 9.6429087754807e-06,
      "loss": 3.6942,
      "step": 322500
    },
    {
      "epoch": 22.692145566952366,
      "grad_norm": 1.1106452941894531,
      "learning_rate": 9.49833742952147e-06,
      "loss": 3.6942,
      "step": 323000
    },
    {
      "epoch": 22.727272727272727,
      "grad_norm": 1.1478626728057861,
      "learning_rate": 9.353766083562238e-06,
      "loss": 3.6924,
      "step": 323500
    },
    {
      "epoch": 22.762399887593087,
      "grad_norm": 1.1238915920257568,
      "learning_rate": 9.209194737603008e-06,
      "loss": 3.6901,
      "step": 324000
    },
    {
      "epoch": 22.797527047913448,
      "grad_norm": 1.1469053030014038,
      "learning_rate": 9.064623391643777e-06,
      "loss": 3.6945,
      "step": 324500
    },
    {
      "epoch": 22.832654208233805,
      "grad_norm": 1.1498441696166992,
      "learning_rate": 8.920052045684546e-06,
      "loss": 3.6937,
      "step": 325000
    },
    {
      "epoch": 22.867781368554166,
      "grad_norm": 1.1367475986480713,
      "learning_rate": 8.775480699725315e-06,
      "loss": 3.6929,
      "step": 325500
    },
    {
      "epoch": 22.902908528874526,
      "grad_norm": 1.1764289140701294,
      "learning_rate": 8.630909353766084e-06,
      "loss": 3.6948,
      "step": 326000
    },
    {
      "epoch": 22.938035689194887,
      "grad_norm": 1.205005407333374,
      "learning_rate": 8.486338007806853e-06,
      "loss": 3.6905,
      "step": 326500
    },
    {
      "epoch": 22.973162849515244,
      "grad_norm": 1.176540493965149,
      "learning_rate": 8.341766661847622e-06,
      "loss": 3.6866,
      "step": 327000
    },
    {
      "epoch": 23.008290009835605,
      "grad_norm": 1.1330305337905884,
      "learning_rate": 8.197195315888391e-06,
      "loss": 3.6926,
      "step": 327500
    },
    {
      "epoch": 23.043417170155966,
      "grad_norm": 1.1623578071594238,
      "learning_rate": 8.05262396992916e-06,
      "loss": 3.6917,
      "step": 328000
    },
    {
      "epoch": 23.078544330476323,
      "grad_norm": 1.1747080087661743,
      "learning_rate": 7.90805262396993e-06,
      "loss": 3.6868,
      "step": 328500
    },
    {
      "epoch": 23.113671490796683,
      "grad_norm": 1.2508447170257568,
      "learning_rate": 7.763481278010698e-06,
      "loss": 3.6907,
      "step": 329000
    },
    {
      "epoch": 23.148798651117044,
      "grad_norm": 1.1551564931869507,
      "learning_rate": 7.618909932051468e-06,
      "loss": 3.6928,
      "step": 329500
    },
    {
      "epoch": 23.183925811437405,
      "grad_norm": 1.1579604148864746,
      "learning_rate": 7.474338586092237e-06,
      "loss": 3.6909,
      "step": 330000
    },
    {
      "epoch": 23.183925811437405,
      "eval_accuracy": 0.6818435829692937,
      "eval_loss": 3.396228551864624,
      "eval_runtime": 136.1565,
      "eval_samples_per_second": 886.524,
      "eval_steps_per_second": 5.545,
      "step": 330000
    },
    {
      "epoch": 23.219052971757762,
      "grad_norm": 1.1311612129211426,
      "learning_rate": 7.3297672401330065e-06,
      "loss": 3.6861,
      "step": 330500
    },
    {
      "epoch": 23.254180132078123,
      "grad_norm": 1.278993010520935,
      "learning_rate": 7.1851958941737756e-06,
      "loss": 3.6901,
      "step": 331000
    },
    {
      "epoch": 23.289307292398483,
      "grad_norm": 1.1452503204345703,
      "learning_rate": 7.040624548214545e-06,
      "loss": 3.6892,
      "step": 331500
    },
    {
      "epoch": 23.324434452718844,
      "grad_norm": 1.1354631185531616,
      "learning_rate": 6.896053202255314e-06,
      "loss": 3.6949,
      "step": 332000
    },
    {
      "epoch": 23.3595616130392,
      "grad_norm": 1.1522068977355957,
      "learning_rate": 6.751481856296083e-06,
      "loss": 3.6907,
      "step": 332500
    },
    {
      "epoch": 23.39468877335956,
      "grad_norm": 1.1194485425949097,
      "learning_rate": 6.606910510336851e-06,
      "loss": 3.6859,
      "step": 333000
    },
    {
      "epoch": 23.429815933679922,
      "grad_norm": 1.160535216331482,
      "learning_rate": 6.46233916437762e-06,
      "loss": 3.6874,
      "step": 333500
    },
    {
      "epoch": 23.46494309400028,
      "grad_norm": 1.1213781833648682,
      "learning_rate": 6.317767818418389e-06,
      "loss": 3.6884,
      "step": 334000
    },
    {
      "epoch": 23.50007025432064,
      "grad_norm": 1.1670405864715576,
      "learning_rate": 6.173196472459159e-06,
      "loss": 3.6906,
      "step": 334500
    },
    {
      "epoch": 23.535197414641,
      "grad_norm": 1.0877423286437988,
      "learning_rate": 6.028625126499928e-06,
      "loss": 3.6919,
      "step": 335000
    },
    {
      "epoch": 23.57032457496136,
      "grad_norm": 1.1785836219787598,
      "learning_rate": 5.884053780540697e-06,
      "loss": 3.6887,
      "step": 335500
    },
    {
      "epoch": 23.60545173528172,
      "grad_norm": 1.1489542722702026,
      "learning_rate": 5.739482434581466e-06,
      "loss": 3.6874,
      "step": 336000
    },
    {
      "epoch": 23.64057889560208,
      "grad_norm": 1.1612714529037476,
      "learning_rate": 5.5949110886222355e-06,
      "loss": 3.6848,
      "step": 336500
    },
    {
      "epoch": 23.67570605592244,
      "grad_norm": 1.1745002269744873,
      "learning_rate": 5.4503397426630046e-06,
      "loss": 3.6903,
      "step": 337000
    },
    {
      "epoch": 23.7108332162428,
      "grad_norm": 1.1236752271652222,
      "learning_rate": 5.305768396703774e-06,
      "loss": 3.6833,
      "step": 337500
    },
    {
      "epoch": 23.745960376563158,
      "grad_norm": 1.1431001424789429,
      "learning_rate": 5.161197050744543e-06,
      "loss": 3.6867,
      "step": 338000
    },
    {
      "epoch": 23.78108753688352,
      "grad_norm": 1.1044416427612305,
      "learning_rate": 5.016625704785312e-06,
      "loss": 3.6854,
      "step": 338500
    },
    {
      "epoch": 23.81621469720388,
      "grad_norm": 1.1204737424850464,
      "learning_rate": 4.872054358826081e-06,
      "loss": 3.6883,
      "step": 339000
    },
    {
      "epoch": 23.851341857524236,
      "grad_norm": 1.150956392288208,
      "learning_rate": 4.72748301286685e-06,
      "loss": 3.6874,
      "step": 339500
    },
    {
      "epoch": 23.886469017844597,
      "grad_norm": 1.191972017288208,
      "learning_rate": 4.582911666907619e-06,
      "loss": 3.6883,
      "step": 340000
    },
    {
      "epoch": 23.886469017844597,
      "eval_accuracy": 0.682490029681647,
      "eval_loss": 3.3932995796203613,
      "eval_runtime": 137.5476,
      "eval_samples_per_second": 877.558,
      "eval_steps_per_second": 5.489,
      "step": 340000
    },
    {
      "epoch": 23.921596178164958,
      "grad_norm": 1.17784583568573,
      "learning_rate": 4.438340320948388e-06,
      "loss": 3.6861,
      "step": 340500
    },
    {
      "epoch": 23.956723338485318,
      "grad_norm": 1.1280088424682617,
      "learning_rate": 4.293768974989157e-06,
      "loss": 3.6877,
      "step": 341000
    },
    {
      "epoch": 23.991850498805675,
      "grad_norm": 1.0852442979812622,
      "learning_rate": 4.149197629029926e-06,
      "loss": 3.6878,
      "step": 341500
    },
    {
      "epoch": 24.026977659126036,
      "grad_norm": 1.108507752418518,
      "learning_rate": 4.004626283070695e-06,
      "loss": 3.6895,
      "step": 342000
    },
    {
      "epoch": 24.062104819446397,
      "grad_norm": 1.2238554954528809,
      "learning_rate": 3.8600549371114645e-06,
      "loss": 3.6869,
      "step": 342500
    },
    {
      "epoch": 24.097231979766757,
      "grad_norm": 1.1142975091934204,
      "learning_rate": 3.7154835911522336e-06,
      "loss": 3.6853,
      "step": 343000
    },
    {
      "epoch": 24.132359140087114,
      "grad_norm": 1.1270474195480347,
      "learning_rate": 3.5709122451930026e-06,
      "loss": 3.6875,
      "step": 343500
    },
    {
      "epoch": 24.167486300407475,
      "grad_norm": 1.140702486038208,
      "learning_rate": 3.426340899233772e-06,
      "loss": 3.6897,
      "step": 344000
    },
    {
      "epoch": 24.202613460727836,
      "grad_norm": 1.1378307342529297,
      "learning_rate": 3.2817695532745412e-06,
      "loss": 3.6873,
      "step": 344500
    },
    {
      "epoch": 24.237740621048193,
      "grad_norm": 1.1261718273162842,
      "learning_rate": 3.1371982073153103e-06,
      "loss": 3.6895,
      "step": 345000
    },
    {
      "epoch": 24.272867781368554,
      "grad_norm": 1.2305586338043213,
      "learning_rate": 2.9926268613560794e-06,
      "loss": 3.6867,
      "step": 345500
    },
    {
      "epoch": 24.307994941688914,
      "grad_norm": 1.1221165657043457,
      "learning_rate": 2.8480555153968485e-06,
      "loss": 3.6842,
      "step": 346000
    },
    {
      "epoch": 24.343122102009275,
      "grad_norm": 1.1741433143615723,
      "learning_rate": 2.7034841694376176e-06,
      "loss": 3.683,
      "step": 346500
    },
    {
      "epoch": 24.378249262329632,
      "grad_norm": 1.1471890211105347,
      "learning_rate": 2.5589128234783866e-06,
      "loss": 3.6866,
      "step": 347000
    },
    {
      "epoch": 24.413376422649993,
      "grad_norm": 1.2041131258010864,
      "learning_rate": 2.414341477519156e-06,
      "loss": 3.6845,
      "step": 347500
    },
    {
      "epoch": 24.448503582970353,
      "grad_norm": 1.133492350578308,
      "learning_rate": 2.269770131559925e-06,
      "loss": 3.6856,
      "step": 348000
    },
    {
      "epoch": 24.483630743290714,
      "grad_norm": 1.1298537254333496,
      "learning_rate": 2.125198785600694e-06,
      "loss": 3.684,
      "step": 348500
    },
    {
      "epoch": 24.51875790361107,
      "grad_norm": 1.0692825317382812,
      "learning_rate": 1.980627439641463e-06,
      "loss": 3.6832,
      "step": 349000
    },
    {
      "epoch": 24.553885063931432,
      "grad_norm": 1.1416022777557373,
      "learning_rate": 1.8360560936822323e-06,
      "loss": 3.6816,
      "step": 349500
    },
    {
      "epoch": 24.589012224251793,
      "grad_norm": 1.1415365934371948,
      "learning_rate": 1.6914847477230013e-06,
      "loss": 3.6867,
      "step": 350000
    },
    {
      "epoch": 24.589012224251793,
      "eval_accuracy": 0.6828799528534023,
      "eval_loss": 3.3902640342712402,
      "eval_runtime": 136.9373,
      "eval_samples_per_second": 881.469,
      "eval_steps_per_second": 5.513,
      "step": 350000
    },
    {
      "epoch": 24.62413938457215,
      "grad_norm": 1.1606513261795044,
      "learning_rate": 1.5469134017637704e-06,
      "loss": 3.6842,
      "step": 350500
    },
    {
      "epoch": 24.65926654489251,
      "grad_norm": 1.1781175136566162,
      "learning_rate": 1.4023420558045395e-06,
      "loss": 3.6819,
      "step": 351000
    },
    {
      "epoch": 24.69439370521287,
      "grad_norm": 1.1214185953140259,
      "learning_rate": 1.2577707098453088e-06,
      "loss": 3.6832,
      "step": 351500
    },
    {
      "epoch": 24.72952086553323,
      "grad_norm": 1.1757885217666626,
      "learning_rate": 1.1131993638860779e-06,
      "loss": 3.6851,
      "step": 352000
    },
    {
      "epoch": 24.76464802585359,
      "grad_norm": 1.1694167852401733,
      "learning_rate": 9.68628017926847e-07,
      "loss": 3.6842,
      "step": 352500
    },
    {
      "epoch": 24.79977518617395,
      "grad_norm": 1.1446681022644043,
      "learning_rate": 8.240566719676161e-07,
      "loss": 3.6846,
      "step": 353000
    },
    {
      "epoch": 24.83490234649431,
      "grad_norm": 1.1173033714294434,
      "learning_rate": 6.794853260083851e-07,
      "loss": 3.6828,
      "step": 353500
    },
    {
      "epoch": 24.870029506814667,
      "grad_norm": 1.1385812759399414,
      "learning_rate": 5.349139800491543e-07,
      "loss": 3.6858,
      "step": 354000
    },
    {
      "epoch": 24.905156667135028,
      "grad_norm": 1.1264246702194214,
      "learning_rate": 3.903426340899234e-07,
      "loss": 3.6856,
      "step": 354500
    },
    {
      "epoch": 24.94028382745539,
      "grad_norm": 1.1489578485488892,
      "learning_rate": 2.457712881306925e-07,
      "loss": 3.6796,
      "step": 355000
    },
    {
      "epoch": 24.97541098777575,
      "grad_norm": 1.1137019395828247,
      "learning_rate": 1.0119994217146162e-07,
      "loss": 3.6846,
      "step": 355500
    },
    {
      "epoch": 25.0,
      "step": 355850,
      "total_flos": 3.038176470544589e+18,
      "train_loss": 4.100663753765192,
      "train_runtime": 96704.9936,
      "train_samples_per_second": 588.734,
      "train_steps_per_second": 3.68
    }
  ],
  "logging_steps": 500,
  "max_steps": 355850,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 25,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.038176470544589e+18,
  "train_batch_size": 160,
  "trial_name": null,
  "trial_params": null
}