{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 1024,
  "global_step": 23204,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011032818324821687,
      "grad_norm": 0.11718960851430893,
      "learning_rate": 0.000498046875,
      "loss": 8.126412391662598,
      "step": 256
    },
    {
      "epoch": 0.022065636649643373,
      "grad_norm": 0.5918834209442139,
      "learning_rate": 0.000998046875,
      "loss": 7.521966934204102,
      "step": 512
    },
    {
      "epoch": 0.03309845497446506,
      "grad_norm": 0.6679062247276306,
      "learning_rate": 0.000999688448778502,
      "loss": 7.341925144195557,
      "step": 768
    },
    {
      "epoch": 0.04413127329928675,
      "grad_norm": 1.1320008039474487,
      "learning_rate": 0.0009987492950653055,
      "loss": 7.3012261390686035,
      "step": 1024
    },
    {
      "epoch": 0.04413127329928675,
      "eval_bleu": 0.4137140409513573,
      "eval_loss": 7.362073738691904,
      "eval_mse_loss": 1.3527455637449903,
      "eval_recon_loss": 6.009328172150959,
      "step": 1024
    },
    {
      "epoch": 0.04413127329928675,
      "eval_bleu": 0.4137140409513573,
      "eval_loss": 7.362073738691904,
      "eval_mse_loss": 1.3527455637449903,
      "eval_recon_loss": 6.009328172150959,
      "eval_runtime": 122.6264,
      "eval_samples_per_second": 244.645,
      "eval_steps_per_second": 3.825,
      "step": 1024
    },
    {
      "epoch": 0.05516409162410843,
      "grad_norm": 0.583996593952179,
      "learning_rate": 0.0009971837136430763,
      "loss": 7.304327011108398,
      "step": 1280
    },
    {
      "epoch": 0.06619690994893011,
      "grad_norm": 0.4174424707889557,
      "learning_rate": 0.0009949936708776692,
      "loss": 7.320664882659912,
      "step": 1536
    },
    {
      "epoch": 0.07722972827375181,
      "grad_norm": 0.6613262295722961,
      "learning_rate": 0.0009921819174566252,
      "loss": 7.289535999298096,
      "step": 1792
    },
    {
      "epoch": 0.0882625465985735,
      "grad_norm": 0.8233232498168945,
      "learning_rate": 0.000988751984934317,
      "loss": 7.30844259262085,
      "step": 2048
    },
    {
      "epoch": 0.0882625465985735,
      "eval_bleu": 0.4137402227214601,
      "eval_loss": 7.285478673255774,
      "eval_mse_loss": 1.312072853543865,
      "eval_recon_loss": 5.973405825812171,
      "step": 2048
    },
    {
      "epoch": 0.0882625465985735,
      "eval_bleu": 0.4137402227214601,
      "eval_loss": 7.285478673255774,
      "eval_mse_loss": 1.312072853543865,
      "eval_recon_loss": 5.973405825812171,
      "eval_runtime": 120.626,
      "eval_samples_per_second": 248.703,
      "eval_steps_per_second": 3.888,
      "step": 2048
    },
    {
      "epoch": 0.09929536492339518,
      "grad_norm": 0.6929961442947388,
      "learning_rate": 0.0009847081812963268,
      "loss": 7.224529266357422,
      "step": 2304
    },
    {
      "epoch": 0.11032818324821686,
      "grad_norm": 0.4629116356372833,
      "learning_rate": 0.0009800555855486275,
      "loss": 7.270728588104248,
      "step": 2560
    },
    {
      "epoch": 0.12136100157303854,
      "grad_norm": 0.8039212226867676,
      "learning_rate": 0.0009748000413383664,
      "loss": 7.244544982910156,
      "step": 2816
    },
    {
      "epoch": 0.13239381989786023,
      "grad_norm": 0.7267772555351257,
      "learning_rate": 0.0009689481496142604,
      "loss": 7.242873191833496,
      "step": 3072
    },
    {
      "epoch": 0.13239381989786023,
      "eval_bleu": 0.41751728447473074,
      "eval_loss": 7.207233533676245,
      "eval_mse_loss": 1.2975257886752392,
      "eval_recon_loss": 5.909707753643044,
      "step": 3072
    },
    {
      "epoch": 0.13239381989786023,
      "eval_bleu": 0.41751728447473074,
      "eval_loss": 7.207233533676245,
      "eval_mse_loss": 1.2975257886752392,
      "eval_recon_loss": 5.909707753643044,
      "eval_runtime": 120.6478,
      "eval_samples_per_second": 248.658,
      "eval_steps_per_second": 3.887,
      "step": 3072
    },
    {
      "epoch": 0.14342663822268192,
      "grad_norm": 0.6957461833953857,
      "learning_rate": 0.0009625072603358231,
      "loss": 7.226602077484131,
      "step": 3328
    },
    {
      "epoch": 0.15445945654750362,
      "grad_norm": 0.7835198044776917,
      "learning_rate": 0.0009554854632418371,
      "loss": 7.225058555603027,
      "step": 3584
    },
    {
      "epoch": 0.1654922748723253,
      "grad_norm": 0.698896586894989,
      "learning_rate": 0.000947891577689663,
      "loss": 7.229842662811279,
      "step": 3840
    },
    {
      "epoch": 0.176525093197147,
      "grad_norm": 0.5856732130050659,
      "learning_rate": 0.0009397351415781539,
      "loss": 7.255417346954346,
      "step": 4096
    },
    {
      "epoch": 0.176525093197147,
      "eval_bleu": 0.4222208504795976,
      "eval_loss": 7.213959849465376,
      "eval_mse_loss": 1.284232896782442,
      "eval_recon_loss": 5.929726949378626,
      "step": 4096
    },
    {
      "epoch": 0.176525093197147,
      "eval_bleu": 0.4222208504795976,
      "eval_loss": 7.213959849465376,
      "eval_mse_loss": 1.284232896782442,
      "eval_recon_loss": 5.929726949378626,
      "eval_runtime": 120.9546,
      "eval_samples_per_second": 248.027,
      "eval_steps_per_second": 3.877,
      "step": 4096
    },
    {
      "epoch": 0.18755791152196866,
      "grad_norm": 0.8607048392295837,
      "learning_rate": 0.000931026399368079,
      "loss": 7.224533557891846,
      "step": 4352
    },
    {
      "epoch": 0.19859072984679035,
      "grad_norm": 0.6015220284461975,
      "learning_rate": 0.0009217762892151117,
      "loss": 7.171858310699463,
      "step": 4608
    },
    {
      "epoch": 0.20962354817161205,
      "grad_norm": 0.5955944061279297,
      "learning_rate": 0.0009119964292315354,
      "loss": 7.180020809173584,
      "step": 4864
    },
    {
      "epoch": 0.22065636649643372,
      "grad_norm": 0.9265109896659851,
      "learning_rate": 0.0009016991028939279,
      "loss": 7.237715244293213,
      "step": 5120
    },
    {
      "epoch": 0.22065636649643372,
      "eval_bleu": 0.42211295717074476,
      "eval_loss": 7.192519418720497,
      "eval_mse_loss": 1.2717160906618847,
      "eval_recon_loss": 5.920803332379632,
      "step": 5120
    },
    {
      "epoch": 0.22065636649643372,
      "eval_bleu": 0.42211295717074476,
      "eval_loss": 7.192519418720497,
      "eval_mse_loss": 1.2717160906618847,
      "eval_recon_loss": 5.920803332379632,
      "eval_runtime": 123.6227,
      "eval_samples_per_second": 242.674,
      "eval_steps_per_second": 3.794,
      "step": 5120
    },
    {
      "epoch": 0.23168918482125542,
      "grad_norm": 1.0009490251541138,
      "learning_rate": 0.0008908972436151494,
      "loss": 7.181396484375,
      "step": 5376
    },
    {
      "epoch": 0.2427220031460771,
      "grad_norm": 0.5662420988082886,
      "learning_rate": 0.0008796044185000127,
      "loss": 7.214040756225586,
      "step": 5632
    },
    {
      "epoch": 0.2537548214708988,
      "grad_norm": 0.5989573001861572,
      "learning_rate": 0.0008678348113050368,
      "loss": 7.233224391937256,
      "step": 5888
    },
    {
      "epoch": 0.26478763979572045,
      "grad_norm": 0.6971510648727417,
      "learning_rate": 0.0008556032046236897,
      "loss": 7.238555431365967,
      "step": 6144
    },
    {
      "epoch": 0.26478763979572045,
      "eval_bleu": 0.42175834100767645,
      "eval_loss": 7.158817366496332,
      "eval_mse_loss": 1.2668565204148607,
      "eval_recon_loss": 5.891960855231865,
      "step": 6144
    },
    {
      "epoch": 0.26478763979572045,
      "eval_bleu": 0.42175834100767645,
      "eval_loss": 7.158817366496332,
      "eval_mse_loss": 1.2668565204148607,
      "eval_recon_loss": 5.891960855231865,
      "eval_runtime": 122.7695,
      "eval_samples_per_second": 244.36,
      "eval_steps_per_second": 3.82,
      "step": 6144
    },
    {
      "epoch": 0.2758204581205422,
      "grad_norm": 0.5552608966827393,
      "learning_rate": 0.000842924961319492,
      "loss": 7.156804084777832,
      "step": 6400
    },
    {
      "epoch": 0.28685327644536385,
      "grad_norm": 0.5489348769187927,
      "learning_rate": 0.0008298160052303045,
      "loss": 7.185822486877441,
      "step": 6656
    },
    {
      "epoch": 0.2978860947701855,
      "grad_norm": 0.6533068418502808,
      "learning_rate": 0.0008162928011680314,
      "loss": 7.207901477813721,
      "step": 6912
    },
    {
      "epoch": 0.30891891309500724,
      "grad_norm": 0.6510070562362671,
      "learning_rate": 0.000802372334238864,
      "loss": 7.252415180206299,
      "step": 7168
    },
    {
      "epoch": 0.30891891309500724,
      "eval_bleu": 0.42033993073908893,
      "eval_loss": 7.211322486527693,
      "eval_mse_loss": 1.262001540106751,
      "eval_recon_loss": 5.949320952267026,
      "step": 7168
    },
    {
      "epoch": 0.30891891309500724,
      "eval_bleu": 0.42033993073908893,
      "eval_loss": 7.211322486527693,
      "eval_mse_loss": 1.262001540106751,
      "eval_recon_loss": 5.949320952267026,
      "eval_runtime": 122.2468,
      "eval_samples_per_second": 245.405,
      "eval_steps_per_second": 3.837,
      "step": 7168
    },
    {
      "epoch": 0.3199517314198289,
      "grad_norm": 0.629555881023407,
      "learning_rate": 0.0007880720885100349,
      "loss": 7.190337181091309,
      "step": 7424
    },
    {
      "epoch": 0.3309845497446506,
      "grad_norm": 0.7832825183868408,
      "learning_rate": 0.0007734100250498788,
      "loss": 7.169421672821045,
      "step": 7680
    },
    {
      "epoch": 0.3420173680694723,
      "grad_norm": 1.125791072845459,
      "learning_rate": 0.000758404559368781,
      "loss": 7.217397212982178,
      "step": 7936
    },
    {
      "epoch": 0.353050186394294,
      "grad_norm": 1.4169881343841553,
      "learning_rate": 0.0007430745382893488,
      "loss": 7.163691520690918,
      "step": 8192
    },
    {
      "epoch": 0.353050186394294,
      "eval_bleu": 0.4273075759012065,
      "eval_loss": 7.0710403741295655,
      "eval_mse_loss": 1.2539655610696594,
      "eval_recon_loss": 5.817074826022964,
      "step": 8192
    },
    {
      "epoch": 0.353050186394294,
      "eval_bleu": 0.4273075759012065,
      "eval_loss": 7.0710403741295655,
      "eval_mse_loss": 1.2539655610696594,
      "eval_recon_loss": 5.817074826022964,
      "eval_runtime": 119.9191,
      "eval_samples_per_second": 250.169,
      "eval_steps_per_second": 3.911,
      "step": 8192
    },
    {
      "epoch": 0.36408300471911564,
      "grad_norm": 0.6753491759300232,
      "learning_rate": 0.0007274392162748551,
      "loss": 7.200345516204834,
      "step": 8448
    },
    {
      "epoch": 0.3751158230439373,
      "grad_norm": 0.5481099486351013,
      "learning_rate": 0.000711518231245687,
      "loss": 7.261969566345215,
      "step": 8704
    },
    {
      "epoch": 0.38614864136875904,
      "grad_norm": 0.9790687561035156,
      "learning_rate": 0.0006953315799141723,
      "loss": 7.222372055053711,
      "step": 8960
    },
    {
      "epoch": 0.3971814596935807,
      "grad_norm": 0.6765880584716797,
      "learning_rate": 0.0006788995926687669,
      "loss": 7.16965913772583,
      "step": 9216
    },
    {
      "epoch": 0.3971814596935807,
      "eval_bleu": 0.4247464090003691,
      "eval_loss": 7.120412382235659,
      "eval_mse_loss": 1.2528218912925801,
      "eval_recon_loss": 5.867590499076762,
      "step": 9216
    },
    {
      "epoch": 0.3971814596935807,
      "eval_bleu": 0.4247464090003691,
      "eval_loss": 7.120412382235659,
      "eval_mse_loss": 1.2528218912925801,
      "eval_recon_loss": 5.867590499076762,
      "eval_runtime": 122.0364,
      "eval_samples_per_second": 245.828,
      "eval_steps_per_second": 3.843,
      "step": 9216
    },
    {
      "epoch": 0.4082142780184024,
      "grad_norm": 0.65935879945755,
      "learning_rate": 0.0006622429080391422,
      "loss": 7.197686195373535,
      "step": 9472
    },
    {
      "epoch": 0.4192470963432241,
      "grad_norm": 0.6421726942062378,
      "learning_rate": 0.0006453824467742515,
      "loss": 7.234926700592041,
      "step": 9728
    },
    {
      "epoch": 0.43027991466804577,
      "grad_norm": 0.5790821313858032,
      "learning_rate": 0.0006283393855659275,
      "loss": 7.189432621002197,
      "step": 9984
    },
    {
      "epoch": 0.44131273299286744,
      "grad_norm": 1.3850535154342651,
      "learning_rate": 0.0006111351304510173,
      "loss": 7.234661102294922,
      "step": 10240
    },
    {
      "epoch": 0.44131273299286744,
      "eval_bleu": 0.418702569877124,
      "eval_loss": 7.200763400429602,
      "eval_mse_loss": 1.2514713542547815,
      "eval_recon_loss": 5.949292033465941,
      "step": 10240
    },
    {
      "epoch": 0.44131273299286744,
      "eval_bleu": 0.418702569877124,
      "eval_loss": 7.200763400429602,
      "eval_mse_loss": 1.2514713542547815,
      "eval_recon_loss": 5.949292033465941,
      "eval_runtime": 120.1297,
      "eval_samples_per_second": 249.73,
      "eval_steps_per_second": 3.904,
      "step": 10240
    },
    {
      "epoch": 0.45234555131768917,
      "grad_norm": 0.7837287187576294,
      "learning_rate": 0.0005937912899254605,
      "loss": 7.249295711517334,
      "step": 10496
    },
    {
      "epoch": 0.46337836964251083,
      "grad_norm": 0.6780616641044617,
      "learning_rate": 0.0005763296478040787,
      "loss": 7.172219276428223,
      "step": 10752
    },
    {
      "epoch": 0.4744111879673325,
      "grad_norm": 0.7398590445518494,
      "learning_rate": 0.0005587721358601663,
      "loss": 7.186757564544678,
      "step": 11008
    },
    {
      "epoch": 0.4854440062921542,
      "grad_norm": 0.9151716232299805,
      "learning_rate": 0.0005411408062792448,
      "loss": 7.159968852996826,
      "step": 11264
    },
    {
      "epoch": 0.4854440062921542,
      "eval_bleu": 0.4226532958201888,
      "eval_loss": 7.1653254210059325,
      "eval_mse_loss": 1.2460996743712598,
      "eval_recon_loss": 5.9192257395177,
      "step": 11264
    },
    {
      "epoch": 0.4854440062921542,
      "eval_bleu": 0.4226532958201888,
      "eval_loss": 7.1653254210059325,
      "eval_mse_loss": 1.2460996743712598,
      "eval_recon_loss": 5.9192257395177,
      "eval_runtime": 121.8437,
      "eval_samples_per_second": 246.217,
      "eval_steps_per_second": 3.849,
      "step": 11264
    },
    {
      "epoch": 0.4964768246169759,
      "grad_norm": 0.8292572498321533,
      "learning_rate": 0.0005234578039615789,
      "loss": 7.178791046142578,
      "step": 11520
    },
    {
      "epoch": 0.5075096429417976,
      "grad_norm": 0.6656509637832642,
      "learning_rate": 0.0005057453387082458,
      "loss": 7.218838691711426,
      "step": 11776
    },
    {
      "epoch": 0.5185424612666193,
      "grad_norm": 0.5986668467521667,
      "learning_rate": 0.0004880256573256866,
      "loss": 7.1541748046875,
      "step": 12032
    },
    {
      "epoch": 0.5295752795914409,
      "grad_norm": 0.9848335385322571,
      "learning_rate": 0.0004703210156837805,
      "loss": 7.192195415496826,
      "step": 12288
    },
    {
      "epoch": 0.5295752795914409,
      "eval_bleu": 0.4265346559703667,
      "eval_loss": 7.1317149693015285,
      "eval_mse_loss": 1.2411813456366565,
      "eval_recon_loss": 5.890533620106386,
      "step": 12288
    },
    {
      "epoch": 0.5295752795914409,
      "eval_bleu": 0.4265346559703667,
      "eval_loss": 7.1317149693015285,
      "eval_mse_loss": 1.2411813456366565,
      "eval_recon_loss": 5.890533620106386,
      "eval_runtime": 121.6345,
      "eval_samples_per_second": 246.641,
      "eval_steps_per_second": 3.856,
      "step": 12288
    },
    {
      "epoch": 0.5406080979162626,
      "grad_norm": 0.4458930194377899,
      "learning_rate": 0.0004526536507625343,
      "loss": 7.163888931274414,
      "step": 12544
    },
    {
      "epoch": 0.5516409162410844,
      "grad_norm": 0.6304380893707275,
      "learning_rate": 0.00043504575272249973,
      "loss": 7.172398567199707,
      "step": 12800
    },
    {
      "epoch": 0.562673734565906,
      "grad_norm": 0.4299321174621582,
      "learning_rate": 0.0004175194370339921,
      "loss": 7.1996941566467285,
      "step": 13056
    },
    {
      "epoch": 0.5737065528907277,
      "grad_norm": 0.7110609412193298,
      "learning_rate": 0.0004000967167001243,
      "loss": 7.176135540008545,
      "step": 13312
    },
    {
      "epoch": 0.5737065528907277,
      "eval_bleu": 0.4210106981552223,
      "eval_loss": 7.17737479606417,
      "eval_mse_loss": 1.24177454910807,
      "eval_recon_loss": 5.9356002522938285,
      "step": 13312
    },
    {
      "epoch": 0.5737065528907277,
      "eval_bleu": 0.4210106981552223,
      "eval_loss": 7.17737479606417,
      "eval_mse_loss": 1.24177454910807,
      "eval_recon_loss": 5.9356002522938285,
      "eval_runtime": 121.837,
      "eval_samples_per_second": 246.231,
      "eval_steps_per_second": 3.849,
      "step": 13312
    },
    {
      "epoch": 0.5847393712155494,
      "grad_norm": 0.6122337579727173,
      "learning_rate": 0.00038279947460853446,
      "loss": 7.161331653594971,
      "step": 13568
    },
    {
      "epoch": 0.595772189540371,
      "grad_norm": 0.3825603723526001,
      "learning_rate": 0.00036564943604654345,
      "loss": 7.182938575744629,
      "step": 13824
    },
    {
      "epoch": 0.6068050078651928,
      "grad_norm": 0.789337158203125,
      "learning_rate": 0.00034866814141425254,
      "loss": 7.209775447845459,
      "step": 14080
    },
    {
      "epoch": 0.6178378261900145,
      "grad_norm": 0.5527840256690979,
      "learning_rate": 0.0003318769191698637,
      "loss": 7.182364463806152,
      "step": 14336
    },
    {
      "epoch": 0.6178378261900145,
      "eval_bleu": 0.4225600590459926,
      "eval_loss": 7.143745667390478,
      "eval_mse_loss": 1.2410022706619457,
      "eval_recon_loss": 5.902743396474354,
      "step": 14336
    },
    {
      "epoch": 0.6178378261900145,
      "eval_bleu": 0.4225600590459926,
      "eval_loss": 7.143745667390478,
      "eval_mse_loss": 1.2410022706619457,
      "eval_recon_loss": 5.902743396474354,
      "eval_runtime": 120.9077,
      "eval_samples_per_second": 248.123,
      "eval_steps_per_second": 3.879,
      "step": 14336
    },
    {
      "epoch": 0.6288706445148361,
      "grad_norm": 1.0367250442504883,
      "learning_rate": 0.00031529685904119485,
      "loss": 7.168930530548096,
      "step": 14592
    },
    {
      "epoch": 0.6399034628396578,
      "grad_norm": 0.6979912519454956,
      "learning_rate": 0.0002989487855370421,
      "loss": 7.185887813568115,
      "step": 14848
    },
    {
      "epoch": 0.6509362811644795,
      "grad_norm": 0.9760459661483765,
      "learning_rate": 0.00028285323179165424,
      "loss": 7.213273048400879,
      "step": 15104
    },
    {
      "epoch": 0.6619690994893012,
      "grad_norm": 0.5782520771026611,
      "learning_rate": 0.0002670304137751759,
      "loss": 7.2389631271362305,
      "step": 15360
    },
    {
      "epoch": 0.6619690994893012,
      "eval_bleu": 0.4214572010245335,
      "eval_loss": 7.171980977820944,
      "eval_mse_loss": 1.238288759422709,
      "eval_recon_loss": 5.933692220431655,
      "step": 15360
    },
    {
      "epoch": 0.6619690994893012,
      "eval_bleu": 0.4214572010245335,
      "eval_loss": 7.171980977820944,
      "eval_mse_loss": 1.238288759422709,
      "eval_recon_loss": 5.933692220431655,
      "eval_runtime": 122.1262,
      "eval_samples_per_second": 245.648,
      "eval_steps_per_second": 3.84,
      "step": 15360
    },
    {
      "epoch": 0.6730019178141229,
      "grad_norm": 0.458218514919281,
      "learning_rate": 0.0002515002049024435,
      "loss": 7.234853744506836,
      "step": 15616
    },
    {
      "epoch": 0.6840347361389446,
      "grad_norm": 0.6497246026992798,
      "learning_rate": 0.00023628211107203429,
      "loss": 7.127511978149414,
      "step": 15872
    },
    {
      "epoch": 0.6950675544637662,
      "grad_norm": 0.47331640124320984,
      "learning_rate": 0.00022139524616691188,
      "loss": 7.157367706298828,
      "step": 16128
    },
    {
      "epoch": 0.706100372788588,
      "grad_norm": 0.5251054167747498,
      "learning_rate": 0.000206858308047443,
      "loss": 7.214804649353027,
      "step": 16384
    },
    {
      "epoch": 0.706100372788588,
      "eval_bleu": 0.4238607122458957,
      "eval_loss": 7.129481719246805,
      "eval_mse_loss": 1.2356282551405526,
      "eval_recon_loss": 5.893853451397373,
      "step": 16384
    },
    {
      "epoch": 0.706100372788588,
      "eval_bleu": 0.4238607122458957,
      "eval_loss": 7.129481719246805,
      "eval_mse_loss": 1.2356282551405526,
      "eval_recon_loss": 5.893853451397373,
      "eval_runtime": 122.2268,
      "eval_samples_per_second": 245.445,
      "eval_steps_per_second": 3.837,
      "step": 16384
    },
    {
      "epoch": 0.7171331911134096,
      "grad_norm": 1.383571982383728,
      "learning_rate": 0.00019268955506693798,
      "loss": 7.159656524658203,
      "step": 16640
    },
    {
      "epoch": 0.7281660094382313,
      "grad_norm": 0.8240185976028442,
      "learning_rate": 0.00017890678313921,
      "loss": 7.187053203582764,
      "step": 16896
    },
    {
      "epoch": 0.739198827763053,
      "grad_norm": 0.9478136301040649,
      "learning_rate": 0.00016552730338695792,
      "loss": 7.169220447540283,
      "step": 17152
    },
    {
      "epoch": 0.7502316460878746,
      "grad_norm": 1.9408297538757324,
      "learning_rate": 0.00015256792039904465,
      "loss": 7.1799798011779785,
      "step": 17408
    },
    {
      "epoch": 0.7502316460878746,
      "eval_bleu": 0.4214146319298905,
      "eval_loss": 7.153863178895735,
      "eval_mse_loss": 1.2358840449786643,
      "eval_recon_loss": 5.917979137221379,
      "step": 17408
    },
    {
      "epoch": 0.7502316460878746,
      "eval_bleu": 0.4214146319298905,
      "eval_loss": 7.153863178895735,
      "eval_mse_loss": 1.2358840449786643,
      "eval_recon_loss": 5.917979137221379,
      "eval_runtime": 121.0512,
      "eval_samples_per_second": 247.829,
      "eval_steps_per_second": 3.874,
      "step": 17408
    },
    {
      "epoch": 0.7612644644126964,
      "grad_norm": 0.42684435844421387,
      "learning_rate": 0.00014004491112398103,
      "loss": 7.144793510437012,
      "step": 17664
    },
    {
      "epoch": 0.7722972827375181,
      "grad_norm": 1.1356747150421143,
      "learning_rate": 0.00012797400442612433,
      "loss": 7.137683868408203,
      "step": 17920
    },
    {
      "epoch": 0.7833301010623397,
      "grad_norm": 0.804119884967804,
      "learning_rate": 0.00011637036133026895,
      "loss": 7.187657833099365,
      "step": 18176
    },
    {
      "epoch": 0.7943629193871614,
      "grad_norm": 0.5904266834259033,
      "learning_rate": 0.00010524855597944216,
      "loss": 7.176294326782227,
      "step": 18432
    },
    {
      "epoch": 0.7943629193871614,
      "eval_bleu": 0.42259771720418604,
      "eval_loss": 7.143732512175148,
      "eval_mse_loss": 1.2345045111072597,
      "eval_recon_loss": 5.909227997509401,
      "step": 18432
    },
    {
      "epoch": 0.7943629193871614,
      "eval_bleu": 0.42259771720418604,
      "eval_loss": 7.143732512175148,
      "eval_mse_loss": 1.2345045111072597,
      "eval_recon_loss": 5.909227997509401,
      "eval_runtime": 120.9093,
      "eval_samples_per_second": 248.12,
      "eval_steps_per_second": 3.879,
      "step": 18432
    },
    {
      "epoch": 0.8053957377119831,
      "grad_norm": 0.7857964634895325,
      "learning_rate": 9.462255732982089e-05,
      "loss": 7.177124977111816,
      "step": 18688
    },
    {
      "epoch": 0.8164285560368048,
      "grad_norm": 0.5492353439331055,
      "learning_rate": 8.450571160576348e-05,
      "loss": 7.213504791259766,
      "step": 18944
    },
    {
      "epoch": 0.8274613743616265,
      "grad_norm": 1.4939922094345093,
      "learning_rate": 7.491072553698764e-05,
      "loss": 7.130130767822266,
      "step": 19200
    },
    {
      "epoch": 0.8384941926864482,
      "grad_norm": 0.6733608841896057,
      "learning_rate": 6.584965039895586e-05,
      "loss": 7.213354587554932,
      "step": 19456
    },
    {
      "epoch": 0.8384941926864482,
      "eval_bleu": 0.42425975662031007,
      "eval_loss": 7.170489285800503,
      "eval_mse_loss": 1.2338256403835597,
      "eval_recon_loss": 5.936663639825036,
      "step": 19456
    },
    {
      "epoch": 0.8384941926864482,
      "eval_bleu": 0.42425975662031007,
      "eval_loss": 7.170489285800503,
      "eval_mse_loss": 1.2338256403835597,
      "eval_recon_loss": 5.936663639825036,
      "eval_runtime": 120.4436,
      "eval_samples_per_second": 249.079,
      "eval_steps_per_second": 3.894,
      "step": 19456
    },
    {
      "epoch": 0.8495270110112698,
      "grad_norm": 0.6659030318260193,
      "learning_rate": 5.73338668765051e-05,
      "loss": 7.2163591384887695,
      "step": 19712
    },
    {
      "epoch": 0.8605598293360915,
      "grad_norm": 0.5635488033294678,
      "learning_rate": 4.9374070769740984e-05,
      "loss": 7.225249290466309,
      "step": 19968
    },
    {
      "epoch": 0.8715926476609133,
      "grad_norm": 1.0609259605407715,
      "learning_rate": 4.198025956014095e-05,
      "loss": 7.194465637207031,
      "step": 20224
    },
    {
      "epoch": 0.8826254659857349,
      "grad_norm": 1.3461576700210571,
      "learning_rate": 3.516171985374755e-05,
      "loss": 7.195147514343262,
      "step": 20480
    },
    {
      "epoch": 0.8826254659857349,
      "eval_bleu": 0.42110197936728205,
      "eval_loss": 7.151369848231009,
      "eval_mse_loss": 1.233785225892626,
      "eval_recon_loss": 5.917584622084205,
      "step": 20480
    },
    {
      "epoch": 0.8826254659857349,
      "eval_bleu": 0.42110197936728205,
      "eval_loss": 7.151369848231009,
      "eval_mse_loss": 1.233785225892626,
      "eval_recon_loss": 5.917584622084205,
      "eval_runtime": 120.8036,
      "eval_samples_per_second": 248.337,
      "eval_steps_per_second": 3.882,
      "step": 20480
    },
    {
      "epoch": 0.8936582843105566,
      "grad_norm": 0.49557796120643616,
      "learning_rate": 2.8927015717215733e-05,
      "loss": 7.159825801849365,
      "step": 20736
    },
    {
      "epoch": 0.9046911026353783,
      "grad_norm": 0.4444299638271332,
      "learning_rate": 2.3283977921370547e-05,
      "loss": 7.15433931350708,
      "step": 20992
    },
    {
      "epoch": 0.9157239209601999,
      "grad_norm": 0.49471017718315125,
      "learning_rate": 1.8239694105780413e-05,
      "loss": 7.151764392852783,
      "step": 21248
    },
    {
      "epoch": 0.9267567392850217,
      "grad_norm": 0.27193328738212585,
      "learning_rate": 1.3800499876701955e-05,
      "loss": 7.129524230957031,
      "step": 21504
    },
    {
      "epoch": 0.9267567392850217,
      "eval_bleu": 0.421951847647465,
      "eval_loss": 7.1731349867798375,
      "eval_mse_loss": 1.2322176178889488,
      "eval_recon_loss": 5.940917385158254,
      "step": 21504
    },
    {
      "epoch": 0.9267567392850217,
      "eval_bleu": 0.421951847647465,
      "eval_loss": 7.1731349867798375,
      "eval_mse_loss": 1.2322176178889488,
      "eval_recon_loss": 5.940917385158254,
      "eval_runtime": 119.5557,
      "eval_samples_per_second": 250.929,
      "eval_steps_per_second": 3.923,
      "step": 21504
    },
    {
      "epoch": 0.9377895576098434,
      "grad_norm": 0.38439345359802246,
      "learning_rate": 9.971970849576406e-06,
      "loss": 7.190091609954834,
      "step": 21760
    },
    {
      "epoch": 0.948822375934665,
      "grad_norm": 0.5374737977981567,
      "learning_rate": 6.758915646072339e-06,
      "loss": 7.214999675750732,
      "step": 22016
    },
    {
      "epoch": 0.9598551942594867,
      "grad_norm": 0.55973219871521,
      "learning_rate": 4.1653698544703575e-06,
      "loss": 7.165055274963379,
      "step": 22272
    },
    {
      "epoch": 0.9708880125843083,
      "grad_norm": 0.2763802111148834,
      "learning_rate": 2.1945909609756286e-06,
      "loss": 7.138909816741943,
      "step": 22528
    },
    {
      "epoch": 0.9708880125843083,
      "eval_bleu": 0.42247179102439836,
      "eval_loss": 7.163274863635554,
      "eval_mse_loss": 1.2323216156664687,
      "eval_recon_loss": 5.930953238310336,
      "step": 22528
    },
    {
      "epoch": 0.9708880125843083,
      "eval_bleu": 0.42247179102439836,
      "eval_loss": 7.163274863635554,
      "eval_mse_loss": 1.2323216156664687,
      "eval_recon_loss": 5.930953238310336,
      "eval_runtime": 119.657,
      "eval_samples_per_second": 250.717,
      "eval_steps_per_second": 3.92,
      "step": 22528
    },
    {
      "epoch": 0.9819208309091301,
      "grad_norm": 0.19876758754253387,
      "learning_rate": 8.490542583243222e-07,
      "loss": 7.1233296394348145,
      "step": 22784
    },
    {
      "epoch": 0.9929536492339518,
      "grad_norm": 0.17880426347255707,
      "learning_rate": 1.3044973682302396e-07,
      "loss": 7.214967250823975,
      "step": 23040
    }
  ],
  "logging_steps": 256,
  "max_steps": 23204,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1024,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}