{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9932104752667313,
  "eval_steps": 1024,
  "global_step": 21504,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011823934229365849,
      "grad_norm": 0.24549634754657745,
      "learning_rate": 0.000498046875,
      "loss": 4.18386697769165,
      "step": 256
    },
    {
      "epoch": 0.023647868458731697,
      "grad_norm": 0.2271815687417984,
      "learning_rate": 0.000998046875,
      "loss": 2.770956039428711,
      "step": 512
    },
    {
      "epoch": 0.03547180268809755,
      "grad_norm": 0.10401463508605957,
      "learning_rate": 0.000999640996023194,
      "loss": 1.3252710103988647,
      "step": 768
    },
    {
      "epoch": 0.047295736917463395,
      "grad_norm": 0.14152967929840088,
      "learning_rate": 0.0009985588674043958,
      "loss": 0.9731757044792175,
      "step": 1024
    },
    {
      "epoch": 0.047295736917463395,
      "eval_bleu": 0.3892877345908514,
      "eval_ce_loss": 0.9097963889984235,
      "eval_loss": 0.9097963889984235,
      "step": 1024
    },
    {
      "epoch": 0.047295736917463395,
      "eval_bleu": 0.3892877345908514,
      "eval_ce_loss": 0.9097963889984235,
      "eval_loss": 0.9097963889984235,
      "eval_runtime": 121.8207,
      "eval_samples_per_second": 229.789,
      "eval_steps_per_second": 3.595,
      "step": 1024
    },
    {
      "epoch": 0.05911967114682925,
      "grad_norm": 0.15999071300029755,
      "learning_rate": 0.0009967551747861387,
      "loss": 0.8742998242378235,
      "step": 1280
    },
    {
      "epoch": 0.0709436053761951,
      "grad_norm": 0.16329661011695862,
      "learning_rate": 0.000994232528651847,
      "loss": 0.8178759217262268,
      "step": 1536
    },
    {
      "epoch": 0.08276753960556095,
      "grad_norm": 0.15794983506202698,
      "learning_rate": 0.0009909945800260092,
      "loss": 0.7751419544219971,
      "step": 1792
    },
    {
      "epoch": 0.09459147383492679,
      "grad_norm": 0.16746914386749268,
      "learning_rate": 0.0009870460151900522,
      "loss": 0.7435040473937988,
      "step": 2048
    },
    {
      "epoch": 0.09459147383492679,
      "eval_bleu": 0.44248013946091835,
      "eval_ce_loss": 0.7321214277178185,
      "eval_loss": 0.7321214277178185,
      "step": 2048
    },
    {
      "epoch": 0.09459147383492679,
      "eval_bleu": 0.44248013946091835,
      "eval_ce_loss": 0.7321214277178185,
      "eval_loss": 0.7321214277178185,
      "eval_runtime": 118.3848,
      "eval_samples_per_second": 236.458,
      "eval_steps_per_second": 3.7,
      "step": 2048
    },
    {
      "epoch": 0.10641540806429264,
      "grad_norm": 0.1850295215845108,
      "learning_rate": 0.0009823925488998885,
      "loss": 0.7177194952964783,
      "step": 2304
    },
    {
      "epoch": 0.1182393422936585,
      "grad_norm": 0.22075414657592773,
      "learning_rate": 0.0009770409161149525,
      "loss": 0.7007547616958618,
      "step": 2560
    },
    {
      "epoch": 0.13006327652302435,
      "grad_norm": 0.18754425644874573,
      "learning_rate": 0.0009709988622506973,
      "loss": 0.6803107857704163,
      "step": 2816
    },
    {
      "epoch": 0.1418872107523902,
      "grad_norm": 0.27784082293510437,
      "learning_rate": 0.000964275131968659,
      "loss": 0.667098343372345,
      "step": 3072
    },
    {
      "epoch": 0.1418872107523902,
      "eval_bleu": 0.45531419834435466,
      "eval_ce_loss": 0.6618382356482554,
      "eval_loss": 0.6618382356482554,
      "step": 3072
    },
    {
      "epoch": 0.1418872107523902,
      "eval_bleu": 0.45531419834435466,
      "eval_ce_loss": 0.6618382356482554,
      "eval_loss": 0.6618382356482554,
      "eval_runtime": 118.8245,
      "eval_samples_per_second": 235.583,
      "eval_steps_per_second": 3.686,
      "step": 3072
    },
    {
      "epoch": 0.15371114498175603,
      "grad_norm": 0.19028860330581665,
      "learning_rate": 0.0009568794565203123,
      "loss": 0.6527599692344666,
      "step": 3328
    },
    {
      "epoch": 0.1655350792111219,
      "grad_norm": 0.20103782415390015,
      "learning_rate": 0.0009488225396630347,
      "loss": 0.6440002918243408,
      "step": 3584
    },
    {
      "epoch": 0.17735901344048774,
      "grad_norm": 0.20544974505901337,
      "learning_rate": 0.0009401160421685646,
      "loss": 0.6339991688728333,
      "step": 3840
    },
    {
      "epoch": 0.18918294766985358,
      "grad_norm": 0.20827855169773102,
      "learning_rate": 0.0009307725649463714,
      "loss": 0.6260778903961182,
      "step": 4096
    },
    {
      "epoch": 0.18918294766985358,
      "eval_bleu": 0.46845902018751817,
      "eval_ce_loss": 0.6253500647468654,
      "eval_loss": 0.6253500647468654,
      "step": 4096
    },
    {
      "epoch": 0.18918294766985358,
      "eval_bleu": 0.46845902018751817,
      "eval_ce_loss": 0.6253500647468654,
      "eval_loss": 0.6253500647468654,
      "eval_runtime": 119.8358,
      "eval_samples_per_second": 233.595,
      "eval_steps_per_second": 3.655,
      "step": 4096
    },
    {
      "epoch": 0.20100688189921945,
      "grad_norm": 0.21191295981407166,
      "learning_rate": 0.0009208056308063659,
      "loss": 0.6236065030097961,
      "step": 4352
    },
    {
      "epoch": 0.2128308161285853,
      "grad_norm": 0.24691034853458405,
      "learning_rate": 0.0009102296648873445,
      "loss": 0.612799346446991,
      "step": 4608
    },
    {
      "epoch": 0.22465475035795113,
      "grad_norm": 0.20382587611675262,
      "learning_rate": 0.0008990599737794927,
      "loss": 0.6066073179244995,
      "step": 4864
    },
    {
      "epoch": 0.236478684587317,
      "grad_norm": 0.20030871033668518,
      "learning_rate": 0.0008873127233711644,
      "loss": 0.599780797958374,
      "step": 5120
    },
    {
      "epoch": 0.236478684587317,
      "eval_bleu": 0.47697047709820556,
      "eval_ce_loss": 0.5962924270733306,
      "eval_loss": 0.5962924270733306,
      "step": 5120
    },
    {
      "epoch": 0.236478684587317,
      "eval_bleu": 0.47697047709820556,
      "eval_ce_loss": 0.5962924270733306,
      "eval_loss": 0.5962924270733306,
      "eval_runtime": 120.1746,
      "eval_samples_per_second": 232.936,
      "eval_steps_per_second": 3.645,
      "step": 5120
    },
    {
      "epoch": 0.24830261881668284,
      "grad_norm": 0.2779148519039154,
      "learning_rate": 0.0008750049154520011,
      "loss": 0.5962232351303101,
      "step": 5376
    },
    {
      "epoch": 0.2601265530460487,
      "grad_norm": 0.25544700026512146,
      "learning_rate": 0.0008621543631062487,
      "loss": 0.588756263256073,
      "step": 5632
    },
    {
      "epoch": 0.27195048727541454,
      "grad_norm": 0.2710427939891815,
      "learning_rate": 0.0008487796649318904,
      "loss": 0.5883460640907288,
      "step": 5888
    },
    {
      "epoch": 0.2837744215047804,
      "grad_norm": 0.2350832223892212,
      "learning_rate": 0.0008349001781229053,
      "loss": 0.5858848094940186,
      "step": 6144
    },
    {
      "epoch": 0.2837744215047804,
      "eval_bleu": 0.48177328442502626,
      "eval_ce_loss": 0.5815416241753592,
      "eval_loss": 0.5815416241753592,
      "step": 6144
    },
    {
      "epoch": 0.2837744215047804,
      "eval_bleu": 0.48177328442502626,
      "eval_ce_loss": 0.5815416241753592,
      "eval_loss": 0.5815416241753592,
      "eval_runtime": 120.4092,
      "eval_samples_per_second": 232.482,
      "eval_steps_per_second": 3.638,
      "step": 6144
    },
    {
      "epoch": 0.2955983557341462,
      "grad_norm": 0.2077747881412506,
      "learning_rate": 0.0008205359904536107,
      "loss": 0.579054594039917,
      "step": 6400
    },
    {
      "epoch": 0.30742228996351206,
      "grad_norm": 0.23695871233940125,
      "learning_rate": 0.0008057078912056363,
      "loss": 0.5759257674217224,
      "step": 6656
    },
    {
      "epoch": 0.3192462241928779,
      "grad_norm": 0.20875827968120575,
      "learning_rate": 0.0007904373410796086,
      "loss": 0.5730624794960022,
      "step": 6912
    },
    {
      "epoch": 0.3310701584222438,
      "grad_norm": 0.32154712080955505,
      "learning_rate": 0.0007747464411350876,
      "loss": 0.569648802280426,
      "step": 7168
    },
    {
      "epoch": 0.3310701584222438,
      "eval_bleu": 0.48784394354563465,
      "eval_ce_loss": 0.571360412185595,
      "eval_loss": 0.571360412185595,
      "step": 7168
    },
    {
      "epoch": 0.3310701584222438,
      "eval_bleu": 0.48784394354563465,
      "eval_ce_loss": 0.571360412185595,
      "eval_loss": 0.571360412185595,
      "eval_runtime": 120.5359,
      "eval_samples_per_second": 232.238,
      "eval_steps_per_second": 3.634,
      "step": 7168
    },
    {
      "epoch": 0.34289409265160964,
      "grad_norm": 0.22952647507190704,
      "learning_rate": 0.000758657900803716,
      "loss": 0.5667076110839844,
      "step": 7424
    },
    {
      "epoch": 0.3547180268809755,
      "grad_norm": 0.23854298889636993,
      "learning_rate": 0.000742195005021869,
      "loss": 0.5655595064163208,
      "step": 7680
    },
    {
      "epoch": 0.3665419611103413,
      "grad_norm": 0.2523816227912903,
      "learning_rate": 0.0007253815805303786,
      "loss": 0.5611915588378906,
      "step": 7936
    },
    {
      "epoch": 0.37836589533970716,
      "grad_norm": 0.21064619719982147,
      "learning_rate": 0.0007082419613901028,
      "loss": 0.5565797686576843,
      "step": 8192
    },
    {
      "epoch": 0.37836589533970716,
      "eval_bleu": 0.4903212059069678,
      "eval_ce_loss": 0.5588067457692264,
      "eval_loss": 0.5588067457692264,
      "step": 8192
    },
    {
      "epoch": 0.37836589533970716,
      "eval_bleu": 0.4903212059069678,
      "eval_ce_loss": 0.5588067457692264,
      "eval_loss": 0.5588067457692264,
      "eval_runtime": 119.8012,
      "eval_samples_per_second": 233.662,
      "eval_steps_per_second": 3.656,
      "step": 8192
    },
    {
      "epoch": 0.390189829569073,
      "grad_norm": 0.2359577864408493,
      "learning_rate": 0.0006908009537632514,
      "loss": 0.557487428188324,
      "step": 8448
    },
    {
      "epoch": 0.4020137637984389,
      "grad_norm": 0.28248342871665955,
      "learning_rate": 0.0006730838000114403,
      "loss": 0.557571530342102,
      "step": 8704
    },
    {
      "epoch": 0.41383769802780473,
      "grad_norm": 0.22688360512256622,
      "learning_rate": 0.0006551161421624341,
      "loss": 0.5515068173408508,
      "step": 8960
    },
    {
      "epoch": 0.4256616322571706,
      "grad_norm": 0.22482667863368988,
      "learning_rate": 0.0006369239847984517,
      "loss": 0.5492411851882935,
      "step": 9216
    },
    {
      "epoch": 0.4256616322571706,
      "eval_bleu": 0.49488542715926903,
      "eval_ce_loss": 0.5491624213244817,
      "eval_loss": 0.5491624213244817,
      "step": 9216
    },
    {
      "epoch": 0.4256616322571706,
      "eval_bleu": 0.49488542715926903,
      "eval_ce_loss": 0.5491624213244817,
      "eval_loss": 0.5491624213244817,
      "eval_runtime": 119.2059,
      "eval_samples_per_second": 234.829,
      "eval_steps_per_second": 3.674,
      "step": 9216
    },
    {
      "epoch": 0.4374855664865364,
      "grad_norm": 0.23678575456142426,
      "learning_rate": 0.0006185336574197479,
      "loss": 0.5475496649742126,
      "step": 9472
    },
    {
      "epoch": 0.44930950071590225,
      "grad_norm": 0.23306351900100708,
      "learning_rate": 0.0005999717763379407,
      "loss": 0.5489075779914856,
      "step": 9728
    },
    {
      "epoch": 0.4611334349452681,
      "grad_norm": 0.24446141719818115,
      "learning_rate": 0.0005812652061542363,
      "loss": 0.5464962720870972,
      "step": 9984
    },
    {
      "epoch": 0.472957369174634,
      "grad_norm": 0.26656046509742737,
      "learning_rate": 0.0005624410208783071,
      "loss": 0.5421015024185181,
      "step": 10240
    },
    {
      "epoch": 0.472957369174634,
      "eval_bleu": 0.49757119992436116,
      "eval_ce_loss": 0.5445564628055651,
      "eval_loss": 0.5445564628055651,
      "step": 10240
    },
    {
      "epoch": 0.472957369174634,
      "eval_bleu": 0.49757119992436116,
      "eval_ce_loss": 0.5445564628055651,
      "eval_loss": 0.5445564628055651,
      "eval_runtime": 119.9032,
      "eval_samples_per_second": 233.463,
      "eval_steps_per_second": 3.653,
      "step": 10240
    },
    {
      "epoch": 0.48478130340399983,
      "grad_norm": 0.25631916522979736,
      "learning_rate": 0.0005435264647440881,
      "loss": 0.5437694787979126,
      "step": 10496
    },
    {
      "epoch": 0.49660523763336567,
      "grad_norm": 0.24505920708179474,
      "learning_rate": 0.000524548912779213,
      "loss": 0.5406980514526367,
      "step": 10752
    },
    {
      "epoch": 0.5084291718627315,
      "grad_norm": 0.2363251894712448,
      "learning_rate": 0.0005055358311851499,
      "loss": 0.5423585772514343,
      "step": 11008
    },
    {
      "epoch": 0.5202531060920974,
      "grad_norm": 0.2904762923717499,
      "learning_rate": 0.0004865147375853812,
      "loss": 0.5392704010009766,
      "step": 11264
    },
    {
      "epoch": 0.5202531060920974,
      "eval_bleu": 0.49735706994160356,
      "eval_ce_loss": 0.5361337422916334,
      "eval_loss": 0.5361337422916334,
      "step": 11264
    },
    {
      "epoch": 0.5202531060920974,
      "eval_bleu": 0.49735706994160356,
      "eval_ce_loss": 0.5361337422916334,
      "eval_loss": 0.5361337422916334,
      "eval_runtime": 118.0528,
      "eval_samples_per_second": 237.123,
      "eval_steps_per_second": 3.71,
      "step": 11264
    },
    {
      "epoch": 0.5320770403214632,
      "grad_norm": 0.2444256991147995,
      "learning_rate": 0.0004675131611991607,
      "loss": 0.536561906337738,
      "step": 11520
    },
    {
      "epoch": 0.5439009745508291,
      "grad_norm": 0.225338414311409,
      "learning_rate": 0.0004485586029984899,
      "loss": 0.5376089811325073,
      "step": 11776
    },
    {
      "epoch": 0.5557249087801949,
      "grad_norm": 0.24263811111450195,
      "learning_rate": 0.00042967849590597266,
      "loss": 0.5349382162094116,
      "step": 12032
    },
    {
      "epoch": 0.5675488430095608,
      "grad_norm": 0.24349793791770935,
      "learning_rate": 0.0004109001650911621,
      "loss": 0.5335611701011658,
      "step": 12288
    },
    {
      "epoch": 0.5675488430095608,
      "eval_bleu": 0.5003156304933045,
      "eval_ce_loss": 0.5322896588474648,
      "eval_loss": 0.5322896588474648,
      "step": 12288
    },
    {
      "epoch": 0.5675488430095608,
      "eval_bleu": 0.5003156304933045,
      "eval_ce_loss": 0.5322896588474648,
      "eval_loss": 0.5322896588474648,
      "eval_runtime": 123.196,
      "eval_samples_per_second": 227.223,
      "eval_steps_per_second": 3.555,
      "step": 12288
    },
    {
      "epoch": 0.5793727772389267,
      "grad_norm": 0.22607196867465973,
      "learning_rate": 0.0003922507884228551,
      "loss": 0.5312694907188416,
      "step": 12544
    },
    {
      "epoch": 0.5911967114682924,
      "grad_norm": 0.2581818997859955,
      "learning_rate": 0.00037375735713457723,
      "loss": 0.5293662548065186,
      "step": 12800
    },
    {
      "epoch": 0.6030206456976583,
      "grad_norm": 0.2328752726316452,
      "learning_rate": 0.00035544663676018276,
      "loss": 0.5298049449920654,
      "step": 13056
    },
    {
      "epoch": 0.6148445799270241,
      "grad_norm": 0.22804544866085052,
      "learning_rate": 0.00033734512839611255,
      "loss": 0.5289351940155029,
      "step": 13312
    },
    {
      "epoch": 0.6148445799270241,
      "eval_bleu": 0.505403307988388,
      "eval_ce_loss": 0.527684580585728,
      "eval_loss": 0.527684580585728,
      "step": 13312
    },
    {
      "epoch": 0.6148445799270241,
      "eval_bleu": 0.505403307988388,
      "eval_ce_loss": 0.527684580585728,
      "eval_loss": 0.527684580585728,
      "eval_runtime": 119.1767,
      "eval_samples_per_second": 234.887,
      "eval_steps_per_second": 3.675,
      "step": 13312
    },
    {
      "epoch": 0.62666851415639,
      "grad_norm": 0.25363898277282715,
      "learning_rate": 0.0003194790303463687,
      "loss": 0.5256697535514832,
      "step": 13568
    },
    {
      "epoch": 0.6384924483857558,
      "grad_norm": 0.2509095370769501,
      "learning_rate": 0.00030187420020572406,
      "loss": 0.5276732444763184,
      "step": 13824
    },
    {
      "epoch": 0.6503163826151217,
      "grad_norm": 0.2644820809364319,
      "learning_rate": 0.00028455611743603626,
      "loss": 0.5258468985557556,
      "step": 14080
    },
    {
      "epoch": 0.6621403168444876,
      "grad_norm": 0.2666020691394806,
      "learning_rate": 0.0002675498464898373,
      "loss": 0.5252395272254944,
      "step": 14336
    },
    {
      "epoch": 0.6621403168444876,
      "eval_bleu": 0.5072326089834556,
      "eval_ce_loss": 0.5239842616395863,
      "eval_loss": 0.5239842616395863,
      "step": 14336
    },
    {
      "epoch": 0.6621403168444876,
      "eval_bleu": 0.5072326089834556,
      "eval_ce_loss": 0.5239842616395863,
      "eval_loss": 0.5239842616395863,
      "eval_runtime": 121.1147,
      "eval_samples_per_second": 231.128,
      "eval_steps_per_second": 3.616,
      "step": 14336
    },
    {
      "epoch": 0.6739642510738534,
      "grad_norm": 0.305530309677124,
      "learning_rate": 0.0002508800005345623,
      "loss": 0.5255744457244873,
      "step": 14592
    },
    {
      "epoch": 0.6857881853032193,
      "grad_norm": 0.22708480060100555,
      "learning_rate": 0.00023457070582992562,
      "loss": 0.5249497294425964,
      "step": 14848
    },
    {
      "epoch": 0.6976121195325851,
      "grad_norm": 0.22963008284568787,
      "learning_rate": 0.00021864556680999692,
      "loss": 0.5229964852333069,
      "step": 15104
    },
    {
      "epoch": 0.709436053761951,
      "grad_norm": 0.22102124989032745,
      "learning_rate": 0.0002031276319205152,
      "loss": 0.5240696668624878,
      "step": 15360
    },
    {
      "epoch": 0.709436053761951,
      "eval_bleu": 0.5042547631198441,
      "eval_ce_loss": 0.5232533255940703,
      "eval_loss": 0.5232533255940703,
      "step": 15360
    },
    {
      "epoch": 0.709436053761951,
      "eval_bleu": 0.5042547631198441,
      "eval_ce_loss": 0.5232533255940703,
      "eval_loss": 0.5232533255940703,
      "eval_runtime": 121.5857,
      "eval_samples_per_second": 230.233,
      "eval_steps_per_second": 3.602,
      "step": 15360
    },
    {
      "epoch": 0.7212599879913169,
      "grad_norm": 0.23232056200504303,
      "learning_rate": 0.00018803936026088542,
      "loss": 0.5210784077644348,
      "step": 15616
    },
    {
      "epoch": 0.7330839222206826,
      "grad_norm": 0.2541489899158478,
      "learning_rate": 0.00017340258907913464,
      "loss": 0.5206863284111023,
      "step": 15872
    },
    {
      "epoch": 0.7449078564500485,
      "grad_norm": 0.2417968064546585,
      "learning_rate": 0.0001592385021668743,
      "loss": 0.5179308652877808,
      "step": 16128
    },
    {
      "epoch": 0.7567317906794143,
      "grad_norm": 0.24219931662082672,
      "learning_rate": 0.0001455675992000087,
      "loss": 0.5221379995346069,
      "step": 16384
    },
    {
      "epoch": 0.7567317906794143,
      "eval_bleu": 0.5066068090628344,
      "eval_ce_loss": 0.5208139295039111,
      "eval_loss": 0.5208139295039111,
      "step": 16384
    },
    {
      "epoch": 0.7567317906794143,
      "eval_bleu": 0.5066068090628344,
      "eval_ce_loss": 0.5208139295039111,
      "eval_loss": 0.5208139295039111,
      "eval_runtime": 120.8269,
      "eval_samples_per_second": 231.679,
      "eval_steps_per_second": 3.625,
      "step": 16384
    },
    {
      "epoch": 0.7685557249087802,
      "grad_norm": 0.2539346218109131,
      "learning_rate": 0.000132409666069565,
      "loss": 0.5175387859344482,
      "step": 16640
    },
    {
      "epoch": 0.780379659138146,
      "grad_norm": 0.22402986884117126,
      "learning_rate": 0.0001197837462455823,
      "loss": 0.5182628035545349,
      "step": 16896
    },
    {
      "epoch": 0.7922035933675119,
      "grad_norm": 0.23848603665828705,
      "learning_rate": 0.00010770811321550749,
      "loss": 0.516541063785553,
      "step": 17152
    },
    {
      "epoch": 0.8040275275968778,
      "grad_norm": 0.22720517218112946,
      "learning_rate": 9.620024403698591e-05,
      "loss": 0.5187172889709473,
      "step": 17408
    },
    {
      "epoch": 0.8040275275968778,
      "eval_bleu": 0.5108638112211565,
      "eval_ce_loss": 0.5187523302422267,
      "eval_loss": 0.5187523302422267,
      "step": 17408
    },
    {
      "epoch": 0.8040275275968778,
      "eval_bleu": 0.5108638112211565,
      "eval_ce_loss": 0.5187523302422267,
      "eval_loss": 0.5187523302422267,
      "eval_runtime": 120.6635,
      "eval_samples_per_second": 231.992,
      "eval_steps_per_second": 3.63,
      "step": 17408
    },
    {
      "epoch": 0.8158514618262436,
      "grad_norm": 0.23969151079654694,
      "learning_rate": 8.527679404332429e-05,
      "loss": 0.516950249671936,
      "step": 17664
    },
    {
      "epoch": 0.8276753960556095,
      "grad_norm": 0.2770597040653229,
      "learning_rate": 7.495357273823544e-05,
      "loss": 0.5192644000053406,
      "step": 17920
    },
    {
      "epoch": 0.8394993302849753,
      "grad_norm": 0.22556257247924805,
      "learning_rate": 6.524552091475183e-05,
      "loss": 0.5170964598655701,
      "step": 18176
    },
    {
      "epoch": 0.8513232645143411,
      "grad_norm": 0.23745474219322205,
      "learning_rate": 5.6166689031422024e-05,
      "loss": 0.5167249441146851,
      "step": 18432
    },
    {
      "epoch": 0.8513232645143411,
      "eval_bleu": 0.5090472136817807,
      "eval_ce_loss": 0.5178870771709643,
      "eval_loss": 0.5178870771709643,
      "step": 18432
    },
    {
      "epoch": 0.8513232645143411,
      "eval_bleu": 0.5090472136817807,
      "eval_ce_loss": 0.5178870771709643,
      "eval_loss": 0.5178870771709643,
      "eval_runtime": 118.7626,
      "eval_samples_per_second": 235.705,
      "eval_steps_per_second": 3.688,
      "step": 18432
    },
    {
      "epoch": 0.8631471987437069,
      "grad_norm": 0.2293296754360199,
      "learning_rate": 4.773021687709067e-05,
      "loss": 0.5173235535621643,
      "step": 18688
    },
    {
      "epoch": 0.8749711329730728,
      "grad_norm": 0.24479453265666962,
      "learning_rate": 3.994831455368719e-05,
      "loss": 0.5143179297447205,
      "step": 18944
    },
    {
      "epoch": 0.8867950672024387,
      "grad_norm": 0.2324228435754776,
      "learning_rate": 3.283224480455282e-05,
      "loss": 0.5171309113502502,
      "step": 19200
    },
    {
      "epoch": 0.8986190014318045,
      "grad_norm": 0.21480675041675568,
      "learning_rate": 2.639230671387627e-05,
      "loss": 0.5147438645362854,
      "step": 19456
    },
    {
      "epoch": 0.8986190014318045,
      "eval_bleu": 0.5083390638340656,
      "eval_ce_loss": 0.5156200128739283,
      "eval_loss": 0.5156200128739283,
      "step": 19456
    },
    {
      "epoch": 0.8986190014318045,
      "eval_bleu": 0.5083390638340656,
      "eval_ce_loss": 0.5156200128739283,
      "eval_loss": 0.5156200128739283,
      "eval_runtime": 121.597,
      "eval_samples_per_second": 230.211,
      "eval_steps_per_second": 3.602,
      "step": 19456
    },
    {
      "epoch": 0.9104429356611704,
      "grad_norm": 0.23823200166225433,
      "learning_rate": 2.063782080083576e-05,
      "loss": 0.5157837867736816,
      "step": 19712
    },
    {
      "epoch": 0.9222668698905362,
      "grad_norm": 0.22363300621509552,
      "learning_rate": 1.557711553001523e-05,
      "loss": 0.5182561874389648,
      "step": 19968
    },
    {
      "epoch": 0.9340908041199021,
      "grad_norm": 0.23723983764648438,
      "learning_rate": 1.1217515257622269e-05,
      "loss": 0.5126692652702332,
      "step": 20224
    },
    {
      "epoch": 0.945914738349268,
      "grad_norm": 0.24191035330295563,
      "learning_rate": 7.565329630950746e-06,
      "loss": 0.5159070491790771,
      "step": 20480
    },
    {
      "epoch": 0.945914738349268,
      "eval_bleu": 0.5118187869284645,
      "eval_ce_loss": 0.5148512013426655,
      "eval_loss": 0.5148512013426655,
      "step": 20480
    },
    {
      "epoch": 0.945914738349268,
      "eval_bleu": 0.5118187869284645,
      "eval_ce_loss": 0.5148512013426655,
      "eval_loss": 0.5148512013426655,
      "eval_runtime": 120.5786,
      "eval_samples_per_second": 232.156,
      "eval_steps_per_second": 3.632,
      "step": 20480
    },
    {
      "epoch": 0.9577386725786338,
      "grad_norm": 0.22184860706329346,
      "learning_rate": 4.62584445643166e-06,
      "loss": 0.5139285326004028,
      "step": 20736
    },
    {
      "epoch": 0.9695626068079997,
      "grad_norm": 0.2565344572067261,
      "learning_rate": 2.40331404948807e-06,
      "loss": 0.5169987082481384,
      "step": 20992
    },
    {
      "epoch": 0.9813865410373654,
      "grad_norm": 0.24548502266407013,
      "learning_rate": 9.009550772663965e-07,
      "loss": 0.5152971744537354,
      "step": 21248
    },
    {
      "epoch": 0.9932104752667313,
      "grad_norm": 0.24441944062709808,
      "learning_rate": 1.2094190315575791e-07,
      "loss": 0.5162725448608398,
      "step": 21504
    },
    {
      "epoch": 0.9932104752667313,
      "eval_bleu": 0.5100991731531398,
      "eval_ce_loss": 0.5133357846165356,
      "eval_loss": 0.5133357846165356,
      "step": 21504
    },
    {
      "epoch": 0.9932104752667313,
      "eval_bleu": 0.5100991731531398,
      "eval_ce_loss": 0.5133357846165356,
      "eval_loss": 0.5133357846165356,
      "eval_runtime": 120.5305,
      "eval_samples_per_second": 232.248,
      "eval_steps_per_second": 3.634,
      "step": 21504
    }
  ],
  "logging_steps": 256,
  "max_steps": 21651,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}