{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7567317906794143,
  "eval_steps": 1024,
  "global_step": 16384,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011823934229365849,
      "grad_norm": 0.0007770129013806581,
      "learning_rate": 0.000498046875,
      "loss": 0.638597846031189,
      "step": 256
    },
    {
      "epoch": 0.023647868458731697,
      "grad_norm": 0.0036502168513834476,
      "learning_rate": 0.000998046875,
      "loss": 0.6255820393562317,
      "step": 512
    },
    {
      "epoch": 0.03547180268809755,
      "grad_norm": 0.00819552130997181,
      "learning_rate": 0.000999640996023194,
      "loss": 0.5880061388015747,
      "step": 768
    },
    {
      "epoch": 0.047295736917463395,
      "grad_norm": 0.014988946728408337,
      "learning_rate": 0.0009985588674043958,
      "loss": 0.5405965447425842,
      "step": 1024
    },
    {
      "epoch": 0.047295736917463395,
      "eval_bleu": 0.020278462916643203,
      "eval_ce_loss": 0.4954925418988755,
      "eval_loss": 0.4954925418988755,
      "step": 1024
    },
    {
      "epoch": 0.047295736917463395,
      "eval_bleu": 0.020278462916643203,
      "eval_ce_loss": 0.4954925418988755,
      "eval_loss": 0.4954925418988755,
      "eval_runtime": 131.1195,
      "eval_samples_per_second": 213.492,
      "eval_steps_per_second": 3.34,
      "step": 1024
    },
    {
      "epoch": 0.05911967114682925,
      "grad_norm": 0.018301010131835938,
      "learning_rate": 0.0009967551747861387,
      "loss": 0.4193720519542694,
      "step": 1280
    },
    {
      "epoch": 0.0709436053761951,
      "grad_norm": 0.02350299246609211,
      "learning_rate": 0.000994232528651847,
      "loss": 0.32118111848831177,
      "step": 1536
    },
    {
      "epoch": 0.08276753960556095,
      "grad_norm": 0.024243628606200218,
      "learning_rate": 0.0009909945800260092,
      "loss": 0.27625754475593567,
      "step": 1792
    },
    {
      "epoch": 0.09459147383492679,
      "grad_norm": 0.022370202466845512,
      "learning_rate": 0.0009870460151900522,
      "loss": 0.25225532054901123,
      "step": 2048
    },
    {
      "epoch": 0.09459147383492679,
      "eval_bleu": 0.16173067052738044,
      "eval_ce_loss": 0.24119392792655997,
      "eval_loss": 0.24119392792655997,
      "step": 2048
    },
    {
      "epoch": 0.09459147383492679,
      "eval_bleu": 0.16173067052738044,
      "eval_ce_loss": 0.24119392792655997,
      "eval_loss": 0.24119392792655997,
      "eval_runtime": 124.4827,
      "eval_samples_per_second": 224.875,
      "eval_steps_per_second": 3.519,
      "step": 2048
    },
    {
      "epoch": 0.10641540806429264,
      "grad_norm": 0.025183985009789467,
      "learning_rate": 0.0009823925488998885,
      "loss": 0.23278099298477173,
      "step": 2304
    },
    {
      "epoch": 0.1182393422936585,
      "grad_norm": 0.025138691067695618,
      "learning_rate": 0.0009770409161149525,
      "loss": 0.21929927170276642,
      "step": 2560
    },
    {
      "epoch": 0.13006327652302435,
      "grad_norm": 0.02561408467590809,
      "learning_rate": 0.0009709988622506973,
      "loss": 0.20501913130283356,
      "step": 2816
    },
    {
      "epoch": 0.1418872107523902,
      "grad_norm": 0.028821036219596863,
      "learning_rate": 0.000964275131968659,
      "loss": 0.19426275789737701,
      "step": 3072
    },
    {
      "epoch": 0.1418872107523902,
      "eval_bleu": 0.19048343372542764,
      "eval_ce_loss": 0.18918462373214226,
      "eval_loss": 0.18918462373214226,
      "step": 3072
    },
    {
      "epoch": 0.1418872107523902,
      "eval_bleu": 0.19048343372542764,
      "eval_ce_loss": 0.18918462373214226,
      "eval_loss": 0.18918462373214226,
      "eval_runtime": 125.1618,
      "eval_samples_per_second": 223.654,
      "eval_steps_per_second": 3.499,
      "step": 3072
    },
    {
      "epoch": 0.15371114498175603,
      "grad_norm": 0.030421536415815353,
      "learning_rate": 0.0009568794565203123,
      "loss": 0.1839032918214798,
      "step": 3328
    },
    {
      "epoch": 0.1655350792111219,
      "grad_norm": 0.032538577914237976,
      "learning_rate": 0.0009488225396630347,
      "loss": 0.17660538852214813,
      "step": 3584
    },
    {
      "epoch": 0.17735901344048774,
      "grad_norm": 0.030282698571681976,
      "learning_rate": 0.0009401160421685646,
      "loss": 0.1695183515548706,
      "step": 3840
    },
    {
      "epoch": 0.18918294766985358,
      "grad_norm": 0.029157549142837524,
      "learning_rate": 0.0009307725649463714,
      "loss": 0.16315339505672455,
      "step": 4096
    },
    {
      "epoch": 0.18918294766985358,
      "eval_bleu": 0.20316795180629035,
      "eval_ce_loss": 0.16114722188885353,
      "eval_loss": 0.16114722188885353,
      "step": 4096
    },
    {
      "epoch": 0.18918294766985358,
      "eval_bleu": 0.20316795180629035,
      "eval_ce_loss": 0.16114722188885353,
      "eval_loss": 0.16114722188885353,
      "eval_runtime": 124.8908,
      "eval_samples_per_second": 224.14,
      "eval_steps_per_second": 3.507,
      "step": 4096
    },
    {
      "epoch": 0.20100688189921945,
      "grad_norm": 0.033927738666534424,
      "learning_rate": 0.0009208056308063659,
      "loss": 0.15908151865005493,
      "step": 4352
    },
    {
      "epoch": 0.2128308161285853,
      "grad_norm": 0.037180621176958084,
      "learning_rate": 0.0009102296648873445,
      "loss": 0.15225762128829956,
      "step": 4608
    },
    {
      "epoch": 0.22465475035795113,
      "grad_norm": 0.037015024572610855,
      "learning_rate": 0.0008990599737794927,
      "loss": 0.1476851999759674,
      "step": 4864
    },
    {
      "epoch": 0.236478684587317,
      "grad_norm": 0.03508929908275604,
      "learning_rate": 0.0008873127233711644,
      "loss": 0.14329124987125397,
      "step": 5120
    },
    {
      "epoch": 0.236478684587317,
      "eval_bleu": 0.21566794230701403,
      "eval_ce_loss": 0.14114715776418987,
      "eval_loss": 0.14114715776418987,
      "step": 5120
    },
    {
      "epoch": 0.236478684587317,
      "eval_bleu": 0.21566794230701403,
      "eval_ce_loss": 0.14114715776418987,
      "eval_loss": 0.14114715776418987,
      "eval_runtime": 125.5711,
      "eval_samples_per_second": 222.926,
      "eval_steps_per_second": 3.488,
      "step": 5120
    },
    {
      "epoch": 0.24830261881668284,
      "grad_norm": 0.03854870796203613,
      "learning_rate": 0.0008750049154520011,
      "loss": 0.14005360007286072,
      "step": 5376
    },
    {
      "epoch": 0.2601265530460487,
      "grad_norm": 0.03188326954841614,
      "learning_rate": 0.0008621543631062487,
      "loss": 0.13561338186264038,
      "step": 5632
    },
    {
      "epoch": 0.27195048727541454,
      "grad_norm": 0.03463684767484665,
      "learning_rate": 0.0008487796649318904,
      "loss": 0.1329711377620697,
      "step": 5888
    },
    {
      "epoch": 0.2837744215047804,
      "grad_norm": 0.0358298160135746,
      "learning_rate": 0.0008349001781229053,
      "loss": 0.1301661729812622,
      "step": 6144
    },
    {
      "epoch": 0.2837744215047804,
      "eval_bleu": 0.22173654323693984,
      "eval_ce_loss": 0.12855397619000852,
      "eval_loss": 0.12855397619000852,
      "step": 6144
    },
    {
      "epoch": 0.2837744215047804,
      "eval_bleu": 0.22173654323693984,
      "eval_ce_loss": 0.12855397619000852,
      "eval_loss": 0.12855397619000852,
      "eval_runtime": 126.9306,
      "eval_samples_per_second": 220.538,
      "eval_steps_per_second": 3.451,
      "step": 6144
    },
    {
      "epoch": 0.2955983557341462,
      "grad_norm": 0.037201233208179474,
      "learning_rate": 0.0008205359904536107,
      "loss": 0.1271919459104538,
      "step": 6400
    },
    {
      "epoch": 0.30742228996351206,
      "grad_norm": 0.0340813584625721,
      "learning_rate": 0.0008057078912056363,
      "loss": 0.12491822242736816,
      "step": 6656
    },
    {
      "epoch": 0.3192462241928779,
      "grad_norm": 0.031119871884584427,
      "learning_rate": 0.0007904373410796086,
      "loss": 0.12252337485551834,
      "step": 6912
    },
    {
      "epoch": 0.3310701584222438,
      "grad_norm": 0.036261022090911865,
      "learning_rate": 0.0007747464411350876,
      "loss": 0.12051380425691605,
      "step": 7168
    },
    {
      "epoch": 0.3310701584222438,
      "eval_bleu": 0.228738826346753,
      "eval_ce_loss": 0.12006643695169933,
      "eval_loss": 0.12006643695169933,
      "step": 7168
    },
    {
      "epoch": 0.3310701584222438,
      "eval_bleu": 0.228738826346753,
      "eval_ce_loss": 0.12006643695169933,
      "eval_loss": 0.12006643695169933,
      "eval_runtime": 128.0461,
      "eval_samples_per_second": 218.617,
      "eval_steps_per_second": 3.421,
      "step": 7168
    },
    {
      "epoch": 0.34289409265160964,
      "grad_norm": 0.03925757110118866,
      "learning_rate": 0.000758657900803716,
      "loss": 0.11830627918243408,
      "step": 7424
    },
    {
      "epoch": 0.3547180268809755,
      "grad_norm": 0.03783626854419708,
      "learning_rate": 0.000742195005021869,
      "loss": 0.11684451997280121,
      "step": 7680
    },
    {
      "epoch": 0.3665419611103413,
      "grad_norm": 0.04456119239330292,
      "learning_rate": 0.0007253815805303786,
      "loss": 0.11479435861110687,
      "step": 7936
    },
    {
      "epoch": 0.37836589533970716,
      "grad_norm": 0.0406075082719326,
      "learning_rate": 0.0007082419613901028,
      "loss": 0.11298805475234985,
      "step": 8192
    },
    {
      "epoch": 0.37836589533970716,
      "eval_bleu": 0.23316787828393487,
      "eval_ce_loss": 0.11301100160501319,
      "eval_loss": 0.11301100160501319,
      "step": 8192
    },
    {
      "epoch": 0.37836589533970716,
      "eval_bleu": 0.23316787828393487,
      "eval_ce_loss": 0.11301100160501319,
      "eval_loss": 0.11301100160501319,
      "eval_runtime": 129.7081,
      "eval_samples_per_second": 215.815,
      "eval_steps_per_second": 3.377,
      "step": 8192
    },
    {
      "epoch": 0.390189829569073,
      "grad_norm": 0.04124804586172104,
      "learning_rate": 0.0006908009537632514,
      "loss": 0.11198760569095612,
      "step": 8448
    },
    {
      "epoch": 0.4020137637984389,
      "grad_norm": 0.036579761654138565,
      "learning_rate": 0.0006730838000114403,
      "loss": 0.11078736931085587,
      "step": 8704
    },
    {
      "epoch": 0.41383769802780473,
      "grad_norm": 0.042201053351163864,
      "learning_rate": 0.0006551161421624341,
      "loss": 0.10881925374269485,
      "step": 8960
    },
    {
      "epoch": 0.4256616322571706,
      "grad_norm": 0.032566700130701065,
      "learning_rate": 0.0006369239847984517,
      "loss": 0.10761021077632904,
      "step": 9216
    },
    {
      "epoch": 0.4256616322571706,
      "eval_bleu": 0.23706037291352922,
      "eval_ce_loss": 0.10713473501610973,
      "eval_loss": 0.10713473501610973,
      "step": 9216
    },
    {
      "epoch": 0.4256616322571706,
      "eval_bleu": 0.23706037291352922,
      "eval_ce_loss": 0.10713473501610973,
      "eval_loss": 0.10713473501610973,
      "eval_runtime": 136.316,
      "eval_samples_per_second": 205.354,
      "eval_steps_per_second": 3.213,
      "step": 9216
    },
    {
      "epoch": 0.4374855664865364,
      "grad_norm": 0.03181539848446846,
      "learning_rate": 0.0006185336574197479,
      "loss": 0.10633238404989243,
      "step": 9472
    },
    {
      "epoch": 0.44930950071590225,
      "grad_norm": 0.039569102227687836,
      "learning_rate": 0.0005999717763379407,
      "loss": 0.10582899302244186,
      "step": 9728
    },
    {
      "epoch": 0.4611334349452681,
      "grad_norm": 0.03662619739770889,
      "learning_rate": 0.0005812652061542363,
      "loss": 0.10484181344509125,
      "step": 9984
    },
    {
      "epoch": 0.472957369174634,
      "grad_norm": 0.037710174918174744,
      "learning_rate": 0.0005624410208783071,
      "loss": 0.10320662707090378,
      "step": 10240
    },
    {
      "epoch": 0.472957369174634,
      "eval_bleu": 0.2394408170470903,
      "eval_ce_loss": 0.10344514005804716,
      "eval_loss": 0.10344514005804716,
      "step": 10240
    },
    {
      "epoch": 0.472957369174634,
      "eval_bleu": 0.2394408170470903,
      "eval_ce_loss": 0.10344514005804716,
      "eval_loss": 0.10344514005804716,
      "eval_runtime": 130.5108,
      "eval_samples_per_second": 214.488,
      "eval_steps_per_second": 3.356,
      "step": 10240
    },
    {
      "epoch": 0.48478130340399983,
      "grad_norm": 0.035840343683958054,
      "learning_rate": 0.0005435264647440881,
      "loss": 0.10283803939819336,
      "step": 10496
    },
    {
      "epoch": 0.49660523763336567,
      "grad_norm": 0.04071688652038574,
      "learning_rate": 0.000524548912779213,
      "loss": 0.1021987721323967,
      "step": 10752
    },
    {
      "epoch": 0.5084291718627315,
      "grad_norm": 0.036369238048791885,
      "learning_rate": 0.0005055358311851499,
      "loss": 0.10196679085493088,
      "step": 11008
    },
    {
      "epoch": 0.5202531060920974,
      "grad_norm": 0.035920336842536926,
      "learning_rate": 0.0004865147375853812,
      "loss": 0.10061318427324295,
      "step": 11264
    },
    {
      "epoch": 0.5202531060920974,
      "eval_bleu": 0.2421649824093764,
      "eval_ce_loss": 0.09971236017400815,
      "eval_loss": 0.09971236017400815,
      "step": 11264
    },
    {
      "epoch": 0.5202531060920974,
      "eval_bleu": 0.2421649824093764,
      "eval_ce_loss": 0.09971236017400815,
      "eval_loss": 0.09971236017400815,
      "eval_runtime": 124.8326,
      "eval_samples_per_second": 224.244,
      "eval_steps_per_second": 3.509,
      "step": 11264
    },
    {
      "epoch": 0.5320770403214632,
      "grad_norm": 0.035242412239313126,
      "learning_rate": 0.0004675131611991607,
      "loss": 0.10020274668931961,
      "step": 11520
    },
    {
      "epoch": 0.5439009745508291,
      "grad_norm": 0.03995301201939583,
      "learning_rate": 0.0004485586029984899,
      "loss": 0.09980272501707077,
      "step": 11776
    },
    {
      "epoch": 0.5557249087801949,
      "grad_norm": 0.0381169356405735,
      "learning_rate": 0.00042967849590597266,
      "loss": 0.09913930296897888,
      "step": 12032
    },
    {
      "epoch": 0.5675488430095608,
      "grad_norm": 0.03928132355213165,
      "learning_rate": 0.0004109001650911621,
      "loss": 0.09818586707115173,
      "step": 12288
    },
    {
      "epoch": 0.5675488430095608,
      "eval_bleu": 0.24165027099086678,
      "eval_ce_loss": 0.09799085275087183,
      "eval_loss": 0.09799085275087183,
      "step": 12288
    },
    {
      "epoch": 0.5675488430095608,
      "eval_bleu": 0.24165027099086678,
      "eval_ce_loss": 0.09799085275087183,
      "eval_loss": 0.09799085275087183,
      "eval_runtime": 128.0046,
      "eval_samples_per_second": 218.687,
      "eval_steps_per_second": 3.422,
      "step": 12288
    },
    {
      "epoch": 0.5793727772389267,
      "grad_norm": 0.038329754024744034,
      "learning_rate": 0.0003922507884228551,
      "loss": 0.09753931313753128,
      "step": 12544
    },
    {
      "epoch": 0.5911967114682924,
      "grad_norm": 0.03905988112092018,
      "learning_rate": 0.00037375735713457723,
      "loss": 0.09739061444997787,
      "step": 12800
    },
    {
      "epoch": 0.6030206456976583,
      "grad_norm": 0.03997718170285225,
      "learning_rate": 0.00035544663676018276,
      "loss": 0.09687047451734543,
      "step": 13056
    },
    {
      "epoch": 0.6148445799270241,
      "grad_norm": 0.037294477224349976,
      "learning_rate": 0.00033734512839611255,
      "loss": 0.09683524817228317,
      "step": 13312
    },
    {
      "epoch": 0.6148445799270241,
      "eval_bleu": 0.244063034514669,
      "eval_ce_loss": 0.0961443424803209,
      "eval_loss": 0.0961443424803209,
      "step": 13312
    },
    {
      "epoch": 0.6148445799270241,
      "eval_bleu": 0.244063034514669,
      "eval_ce_loss": 0.0961443424803209,
      "eval_loss": 0.0961443424803209,
      "eval_runtime": 127.2652,
      "eval_samples_per_second": 219.958,
      "eval_steps_per_second": 3.442,
      "step": 13312
    },
    {
      "epoch": 0.62666851415639,
      "grad_norm": 0.04453560709953308,
      "learning_rate": 0.0003194790303463687,
      "loss": 0.09584437310695648,
      "step": 13568
    },
    {
      "epoch": 0.6384924483857558,
      "grad_norm": 0.03323407843708992,
      "learning_rate": 0.00030187420020572406,
      "loss": 0.09573940187692642,
      "step": 13824
    },
    {
      "epoch": 0.6503163826151217,
      "grad_norm": 0.03651966527104378,
      "learning_rate": 0.00028455611743603626,
      "loss": 0.09577071666717529,
      "step": 14080
    },
    {
      "epoch": 0.6621403168444876,
      "grad_norm": 0.03129328042268753,
      "learning_rate": 0.0002675498464898373,
      "loss": 0.09538843482732773,
      "step": 14336
    },
    {
      "epoch": 0.6621403168444876,
      "eval_bleu": 0.24573789493234413,
      "eval_ce_loss": 0.09470595116620739,
      "eval_loss": 0.09470595116620739,
      "step": 14336
    },
    {
      "epoch": 0.6621403168444876,
      "eval_bleu": 0.24573789493234413,
      "eval_ce_loss": 0.09470595116620739,
      "eval_loss": 0.09470595116620739,
      "eval_runtime": 128.6793,
      "eval_samples_per_second": 217.541,
      "eval_steps_per_second": 3.404,
      "step": 14336
    },
    {
      "epoch": 0.6739642510738534,
      "grad_norm": 0.03330269455909729,
      "learning_rate": 0.0002508800005345623,
      "loss": 0.09491758048534393,
      "step": 14592
    },
    {
      "epoch": 0.6857881853032193,
      "grad_norm": 0.03013775125145912,
      "learning_rate": 0.00023457070582992562,
      "loss": 0.09489757567644119,
      "step": 14848
    },
    {
      "epoch": 0.6976121195325851,
      "grad_norm": 0.03159448131918907,
      "learning_rate": 0.00021864556680999692,
      "loss": 0.09417803585529327,
      "step": 15104
    },
    {
      "epoch": 0.709436053761951,
      "grad_norm": 0.028872501105070114,
      "learning_rate": 0.0002031276319205152,
      "loss": 0.094399593770504,
      "step": 15360
    },
    {
      "epoch": 0.709436053761951,
      "eval_bleu": 0.242515567114501,
      "eval_ce_loss": 0.09441615134205449,
      "eval_loss": 0.09441615134205449,
      "step": 15360
    },
    {
      "epoch": 0.709436053761951,
      "eval_bleu": 0.242515567114501,
      "eval_ce_loss": 0.09441615134205449,
      "eval_loss": 0.09441615134205449,
      "eval_runtime": 128.6332,
      "eval_samples_per_second": 217.619,
      "eval_steps_per_second": 3.405,
      "step": 15360
    },
    {
      "epoch": 0.7212599879913169,
      "grad_norm": 0.03194403648376465,
      "learning_rate": 0.00018803936026088542,
      "loss": 0.09390134364366531,
      "step": 15616
    },
    {
      "epoch": 0.7330839222206826,
      "grad_norm": 0.03124140575528145,
      "learning_rate": 0.00017340258907913464,
      "loss": 0.09355755150318146,
      "step": 15872
    },
    {
      "epoch": 0.7449078564500485,
      "grad_norm": 0.027917077764868736,
      "learning_rate": 0.0001592385021668743,
      "loss": 0.09316008538007736,
      "step": 16128
    },
    {
      "epoch": 0.7567317906794143,
      "grad_norm": 0.02921847254037857,
      "learning_rate": 0.0001455675992000087,
      "loss": 0.09386321902275085,
      "step": 16384
    },
    {
      "epoch": 0.7567317906794143,
      "eval_bleu": 0.24441726952160492,
      "eval_ce_loss": 0.0937359631061554,
      "eval_loss": 0.0937359631061554,
      "step": 16384
    },
    {
      "epoch": 0.7567317906794143,
      "eval_bleu": 0.24441726952160492,
      "eval_ce_loss": 0.0937359631061554,
      "eval_loss": 0.0937359631061554,
      "eval_runtime": 130.1985,
      "eval_samples_per_second": 215.002,
      "eval_steps_per_second": 3.364,
      "step": 16384
    }
  ],
  "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
}