{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9974326059050065,
  "eval_steps": 500,
  "global_step": 1038,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0019255455712451862,
      "grad_norm": 4.675525031091207,
      "learning_rate": 9.615384615384617e-08,
      "loss": 2.296,
      "step": 1
    },
    {
      "epoch": 0.009627727856225931,
      "grad_norm": 4.383312697034829,
      "learning_rate": 4.807692307692308e-07,
      "loss": 2.1961,
      "step": 5
    },
    {
      "epoch": 0.019255455712451863,
      "grad_norm": 3.163432373683522,
      "learning_rate": 9.615384615384617e-07,
      "loss": 2.1263,
      "step": 10
    },
    {
      "epoch": 0.028883183568677792,
      "grad_norm": 2.265167060038816,
      "learning_rate": 1.4423076923076922e-06,
      "loss": 1.7774,
      "step": 15
    },
    {
      "epoch": 0.038510911424903725,
      "grad_norm": 1.5336348895834757,
      "learning_rate": 1.9230769230769234e-06,
      "loss": 1.351,
      "step": 20
    },
    {
      "epoch": 0.04813863928112965,
      "grad_norm": 0.5295614063340365,
      "learning_rate": 2.403846153846154e-06,
      "loss": 0.939,
      "step": 25
    },
    {
      "epoch": 0.057766367137355584,
      "grad_norm": 0.4334072719900553,
      "learning_rate": 2.8846153846153845e-06,
      "loss": 0.8218,
      "step": 30
    },
    {
      "epoch": 0.06739409499358151,
      "grad_norm": 0.3248649181280309,
      "learning_rate": 3.365384615384616e-06,
      "loss": 0.6933,
      "step": 35
    },
    {
      "epoch": 0.07702182284980745,
      "grad_norm": 0.2543051022910018,
      "learning_rate": 3.846153846153847e-06,
      "loss": 0.6536,
      "step": 40
    },
    {
      "epoch": 0.08664955070603338,
      "grad_norm": 0.24243639068039996,
      "learning_rate": 4.326923076923077e-06,
      "loss": 0.5727,
      "step": 45
    },
    {
      "epoch": 0.0962772785622593,
      "grad_norm": 0.21784634136238626,
      "learning_rate": 4.807692307692308e-06,
      "loss": 0.5824,
      "step": 50
    },
    {
      "epoch": 0.10590500641848524,
      "grad_norm": 0.24558020757705826,
      "learning_rate": 5.288461538461539e-06,
      "loss": 0.5447,
      "step": 55
    },
    {
      "epoch": 0.11553273427471117,
      "grad_norm": 0.27633240690641786,
      "learning_rate": 5.769230769230769e-06,
      "loss": 0.5196,
      "step": 60
    },
    {
      "epoch": 0.1251604621309371,
      "grad_norm": 0.1831824766725952,
      "learning_rate": 6.25e-06,
      "loss": 0.5183,
      "step": 65
    },
    {
      "epoch": 0.13478818998716302,
      "grad_norm": 0.25923345195442493,
      "learning_rate": 6.730769230769232e-06,
      "loss": 0.4999,
      "step": 70
    },
    {
      "epoch": 0.14441591784338895,
      "grad_norm": 0.17651629460554236,
      "learning_rate": 7.211538461538462e-06,
      "loss": 0.4656,
      "step": 75
    },
    {
      "epoch": 0.1540436456996149,
      "grad_norm": 0.23934885323800617,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.4705,
      "step": 80
    },
    {
      "epoch": 0.16367137355584083,
      "grad_norm": 0.17293134941389127,
      "learning_rate": 8.173076923076923e-06,
      "loss": 0.4743,
      "step": 85
    },
    {
      "epoch": 0.17329910141206675,
      "grad_norm": 0.16896953176677149,
      "learning_rate": 8.653846153846155e-06,
      "loss": 0.4696,
      "step": 90
    },
    {
      "epoch": 0.18292682926829268,
      "grad_norm": 0.14455896542895968,
      "learning_rate": 9.134615384615384e-06,
      "loss": 0.4671,
      "step": 95
    },
    {
      "epoch": 0.1925545571245186,
      "grad_norm": 0.14120629212505248,
      "learning_rate": 9.615384615384616e-06,
      "loss": 0.4368,
      "step": 100
    },
    {
      "epoch": 0.20218228498074453,
      "grad_norm": 0.13472865892209046,
      "learning_rate": 9.999971715689765e-06,
      "loss": 0.4397,
      "step": 105
    },
    {
      "epoch": 0.21181001283697048,
      "grad_norm": 0.15889811986141222,
      "learning_rate": 9.99898179843121e-06,
      "loss": 0.432,
      "step": 110
    },
    {
      "epoch": 0.2214377406931964,
      "grad_norm": 0.15298006269669823,
      "learning_rate": 9.996577985645488e-06,
      "loss": 0.4484,
      "step": 115
    },
    {
      "epoch": 0.23106546854942234,
      "grad_norm": 0.14657423882488327,
      "learning_rate": 9.992760957219083e-06,
      "loss": 0.4411,
      "step": 120
    },
    {
      "epoch": 0.24069319640564826,
      "grad_norm": 0.13147480990020327,
      "learning_rate": 9.987531792747726e-06,
      "loss": 0.4518,
      "step": 125
    },
    {
      "epoch": 0.2503209242618742,
      "grad_norm": 0.14062203555872538,
      "learning_rate": 9.980891971231052e-06,
      "loss": 0.4325,
      "step": 130
    },
    {
      "epoch": 0.25994865211810014,
      "grad_norm": 0.11622965182668187,
      "learning_rate": 9.972843370654286e-06,
      "loss": 0.4153,
      "step": 135
    },
    {
      "epoch": 0.26957637997432604,
      "grad_norm": 0.11345089120636126,
      "learning_rate": 9.963388267457071e-06,
      "loss": 0.4244,
      "step": 140
    },
    {
      "epoch": 0.279204107830552,
      "grad_norm": 0.12785723323300235,
      "learning_rate": 9.952529335889615e-06,
      "loss": 0.4483,
      "step": 145
    },
    {
      "epoch": 0.2888318356867779,
      "grad_norm": 0.1129536919207956,
      "learning_rate": 9.940269647256319e-06,
      "loss": 0.4269,
      "step": 150
    },
    {
      "epoch": 0.29845956354300385,
      "grad_norm": 0.11813817138680209,
      "learning_rate": 9.926612669047085e-06,
      "loss": 0.4292,
      "step": 155
    },
    {
      "epoch": 0.3080872913992298,
      "grad_norm": 0.12171485620288995,
      "learning_rate": 9.911562263956593e-06,
      "loss": 0.4101,
      "step": 160
    },
    {
      "epoch": 0.3177150192554557,
      "grad_norm": 0.11639352426762696,
      "learning_rate": 9.895122688791792e-06,
      "loss": 0.4239,
      "step": 165
    },
    {
      "epoch": 0.32734274711168165,
      "grad_norm": 0.12486408455362112,
      "learning_rate": 9.877298593267906e-06,
      "loss": 0.4269,
      "step": 170
    },
    {
      "epoch": 0.33697047496790755,
      "grad_norm": 0.10765095070330241,
      "learning_rate": 9.85809501869334e-06,
      "loss": 0.4249,
      "step": 175
    },
    {
      "epoch": 0.3465982028241335,
      "grad_norm": 0.10473147389747606,
      "learning_rate": 9.837517396543799e-06,
      "loss": 0.428,
      "step": 180
    },
    {
      "epoch": 0.35622593068035946,
      "grad_norm": 0.11066926415967249,
      "learning_rate": 9.815571546926074e-06,
      "loss": 0.4067,
      "step": 185
    },
    {
      "epoch": 0.36585365853658536,
      "grad_norm": 0.11382788234866194,
      "learning_rate": 9.792263676931906e-06,
      "loss": 0.418,
      "step": 190
    },
    {
      "epoch": 0.3754813863928113,
      "grad_norm": 0.11003473176727298,
      "learning_rate": 9.767600378882379e-06,
      "loss": 0.4333,
      "step": 195
    },
    {
      "epoch": 0.3851091142490372,
      "grad_norm": 0.10956539331792438,
      "learning_rate": 9.741588628463384e-06,
      "loss": 0.4308,
      "step": 200
    },
    {
      "epoch": 0.39473684210526316,
      "grad_norm": 0.10379012730393708,
      "learning_rate": 9.714235782752633e-06,
      "loss": 0.4234,
      "step": 205
    },
    {
      "epoch": 0.40436456996148906,
      "grad_norm": 0.10979551297884416,
      "learning_rate": 9.68554957813881e-06,
      "loss": 0.3863,
      "step": 210
    },
    {
      "epoch": 0.413992297817715,
      "grad_norm": 0.10147077391922903,
      "learning_rate": 9.655538128133432e-06,
      "loss": 0.4286,
      "step": 215
    },
    {
      "epoch": 0.42362002567394097,
      "grad_norm": 0.10441472907725119,
      "learning_rate": 9.62420992107605e-06,
      "loss": 0.4122,
      "step": 220
    },
    {
      "epoch": 0.43324775353016687,
      "grad_norm": 0.11600725245110252,
      "learning_rate": 9.59157381773344e-06,
      "loss": 0.4001,
      "step": 225
    },
    {
      "epoch": 0.4428754813863928,
      "grad_norm": 0.10739569680474984,
      "learning_rate": 9.557639048793453e-06,
      "loss": 0.4231,
      "step": 230
    },
    {
      "epoch": 0.4525032092426187,
      "grad_norm": 0.11403638891480855,
      "learning_rate": 9.522415212254229e-06,
      "loss": 0.4093,
      "step": 235
    },
    {
      "epoch": 0.4621309370988447,
      "grad_norm": 0.09497909577287572,
      "learning_rate": 9.485912270709542e-06,
      "loss": 0.4113,
      "step": 240
    },
    {
      "epoch": 0.47175866495507063,
      "grad_norm": 0.09985569093273962,
      "learning_rate": 9.448140548531004e-06,
      "loss": 0.3785,
      "step": 245
    },
    {
      "epoch": 0.4813863928112965,
      "grad_norm": 0.11631044044684512,
      "learning_rate": 9.409110728947964e-06,
      "loss": 0.3933,
      "step": 250
    },
    {
      "epoch": 0.4910141206675225,
      "grad_norm": 0.0997854084788479,
      "learning_rate": 9.368833851025882e-06,
      "loss": 0.395,
      "step": 255
    },
    {
      "epoch": 0.5006418485237484,
      "grad_norm": 0.10619811125132311,
      "learning_rate": 9.327321306544097e-06,
      "loss": 0.4137,
      "step": 260
    },
    {
      "epoch": 0.5102695763799743,
      "grad_norm": 0.10149827880932093,
      "learning_rate": 9.284584836773796e-06,
      "loss": 0.4077,
      "step": 265
    },
    {
      "epoch": 0.5198973042362003,
      "grad_norm": 0.10925773008402428,
      "learning_rate": 9.240636529157158e-06,
      "loss": 0.4092,
      "step": 270
    },
    {
      "epoch": 0.5295250320924262,
      "grad_norm": 0.10181499630559736,
      "learning_rate": 9.195488813888588e-06,
      "loss": 0.3975,
      "step": 275
    },
    {
      "epoch": 0.5391527599486521,
      "grad_norm": 0.08480738919090135,
      "learning_rate": 9.149154460398993e-06,
      "loss": 0.4149,
      "step": 280
    },
    {
      "epoch": 0.5487804878048781,
      "grad_norm": 0.11272795463115028,
      "learning_rate": 9.101646573744118e-06,
      "loss": 0.4149,
      "step": 285
    },
    {
      "epoch": 0.558408215661104,
      "grad_norm": 0.11249381051877479,
      "learning_rate": 9.052978590897964e-06,
      "loss": 0.4081,
      "step": 290
    },
    {
      "epoch": 0.5680359435173299,
      "grad_norm": 0.09372931069614396,
      "learning_rate": 9.003164276952315e-06,
      "loss": 0.391,
      "step": 295
    },
    {
      "epoch": 0.5776636713735558,
      "grad_norm": 0.10766510897950467,
      "learning_rate": 8.95221772122345e-06,
      "loss": 0.3918,
      "step": 300
    },
    {
      "epoch": 0.5872913992297818,
      "grad_norm": 0.09999194407255885,
      "learning_rate": 8.900153333267198e-06,
      "loss": 0.3843,
      "step": 305
    },
    {
      "epoch": 0.5969191270860077,
      "grad_norm": 0.09518035736085621,
      "learning_rate": 8.846985838803357e-06,
      "loss": 0.4125,
      "step": 310
    },
    {
      "epoch": 0.6065468549422336,
      "grad_norm": 0.08726122835506574,
      "learning_rate": 8.792730275550736e-06,
      "loss": 0.4079,
      "step": 315
    },
    {
      "epoch": 0.6161745827984596,
      "grad_norm": 0.09353747112555597,
      "learning_rate": 8.73740198897393e-06,
      "loss": 0.4142,
      "step": 320
    },
    {
      "epoch": 0.6258023106546855,
      "grad_norm": 0.09406069607300888,
      "learning_rate": 8.681016627943063e-06,
      "loss": 0.4104,
      "step": 325
    },
    {
      "epoch": 0.6354300385109114,
      "grad_norm": 0.09374910799397031,
      "learning_rate": 8.623590140307715e-06,
      "loss": 0.395,
      "step": 330
    },
    {
      "epoch": 0.6450577663671374,
      "grad_norm": 0.10471444555170399,
      "learning_rate": 8.565138768386276e-06,
      "loss": 0.421,
      "step": 335
    },
    {
      "epoch": 0.6546854942233633,
      "grad_norm": 0.0994921653563304,
      "learning_rate": 8.50567904437203e-06,
      "loss": 0.4156,
      "step": 340
    },
    {
      "epoch": 0.6643132220795892,
      "grad_norm": 0.08678612456218597,
      "learning_rate": 8.445227785657245e-06,
      "loss": 0.4055,
      "step": 345
    },
    {
      "epoch": 0.6739409499358151,
      "grad_norm": 0.09635413003916711,
      "learning_rate": 8.383802090076589e-06,
      "loss": 0.3856,
      "step": 350
    },
    {
      "epoch": 0.6835686777920411,
      "grad_norm": 0.09507719640106775,
      "learning_rate": 8.321419331071242e-06,
      "loss": 0.4057,
      "step": 355
    },
    {
      "epoch": 0.693196405648267,
      "grad_norm": 0.1015426141791736,
      "learning_rate": 8.258097152775045e-06,
      "loss": 0.4212,
      "step": 360
    },
    {
      "epoch": 0.7028241335044929,
      "grad_norm": 0.10014829884316678,
      "learning_rate": 8.1938534650241e-06,
      "loss": 0.3968,
      "step": 365
    },
    {
      "epoch": 0.7124518613607189,
      "grad_norm": 0.10046140588576453,
      "learning_rate": 8.128706438291193e-06,
      "loss": 0.3972,
      "step": 370
    },
    {
      "epoch": 0.7220795892169448,
      "grad_norm": 0.08519030816328944,
      "learning_rate": 8.062674498546542e-06,
      "loss": 0.4011,
      "step": 375
    },
    {
      "epoch": 0.7317073170731707,
      "grad_norm": 0.10853398520099701,
      "learning_rate": 7.995776322046236e-06,
      "loss": 0.3968,
      "step": 380
    },
    {
      "epoch": 0.7413350449293966,
      "grad_norm": 0.08705459909141744,
      "learning_rate": 7.92803083004991e-06,
      "loss": 0.3795,
      "step": 385
    },
    {
      "epoch": 0.7509627727856226,
      "grad_norm": 0.09871394890968048,
      "learning_rate": 7.859457183469119e-06,
      "loss": 0.3924,
      "step": 390
    },
    {
      "epoch": 0.7605905006418485,
      "grad_norm": 0.08824376485660647,
      "learning_rate": 7.790074777447938e-06,
      "loss": 0.351,
      "step": 395
    },
    {
      "epoch": 0.7702182284980744,
      "grad_norm": 0.09406342927041313,
      "learning_rate": 7.719903235877289e-06,
      "loss": 0.4028,
      "step": 400
    },
    {
      "epoch": 0.7798459563543004,
      "grad_norm": 0.0910039354481899,
      "learning_rate": 7.648962405844587e-06,
      "loss": 0.3965,
      "step": 405
    },
    {
      "epoch": 0.7894736842105263,
      "grad_norm": 0.10182384336867278,
      "learning_rate": 7.577272352020269e-06,
      "loss": 0.3767,
      "step": 410
    },
    {
      "epoch": 0.7991014120667522,
      "grad_norm": 0.09839014599757313,
      "learning_rate": 7.5048533509827474e-06,
      "loss": 0.3864,
      "step": 415
    },
    {
      "epoch": 0.8087291399229781,
      "grad_norm": 0.08845839135994245,
      "learning_rate": 7.43172588548347e-06,
      "loss": 0.3961,
      "step": 420
    },
    {
      "epoch": 0.8183568677792041,
      "grad_norm": 0.08848040955413092,
      "learning_rate": 7.357910638653629e-06,
      "loss": 0.3688,
      "step": 425
    },
    {
      "epoch": 0.82798459563543,
      "grad_norm": 0.09423759349695249,
      "learning_rate": 7.283428488154227e-06,
      "loss": 0.3879,
      "step": 430
    },
    {
      "epoch": 0.8376123234916559,
      "grad_norm": 0.08488822732748169,
      "learning_rate": 7.208300500271097e-06,
      "loss": 0.4036,
      "step": 435
    },
    {
      "epoch": 0.8472400513478819,
      "grad_norm": 0.07899711812098008,
      "learning_rate": 7.1325479239565875e-06,
      "loss": 0.4006,
      "step": 440
    },
    {
      "epoch": 0.8568677792041078,
      "grad_norm": 0.08340972320218538,
      "learning_rate": 7.056192184819582e-06,
      "loss": 0.3922,
      "step": 445
    },
    {
      "epoch": 0.8664955070603337,
      "grad_norm": 0.08930316294422015,
      "learning_rate": 6.9792548790655465e-06,
      "loss": 0.3619,
      "step": 450
    },
    {
      "epoch": 0.8761232349165597,
      "grad_norm": 0.07843462308791004,
      "learning_rate": 6.901757767388331e-06,
      "loss": 0.3683,
      "step": 455
    },
    {
      "epoch": 0.8857509627727856,
      "grad_norm": 0.0888434612182392,
      "learning_rate": 6.823722768815446e-06,
      "loss": 0.4155,
      "step": 460
    },
    {
      "epoch": 0.8953786906290115,
      "grad_norm": 0.0815067949974782,
      "learning_rate": 6.7451719545085585e-06,
      "loss": 0.3608,
      "step": 465
    },
    {
      "epoch": 0.9050064184852374,
      "grad_norm": 0.09498095494824059,
      "learning_rate": 6.666127541520958e-06,
      "loss": 0.394,
      "step": 470
    },
    {
      "epoch": 0.9146341463414634,
      "grad_norm": 0.09361211338328486,
      "learning_rate": 6.586611886513756e-06,
      "loss": 0.3995,
      "step": 475
    },
    {
      "epoch": 0.9242618741976893,
      "grad_norm": 0.0904791289087828,
      "learning_rate": 6.506647479432604e-06,
      "loss": 0.3887,
      "step": 480
    },
    {
      "epoch": 0.9338896020539152,
      "grad_norm": 0.08454411427597165,
      "learning_rate": 6.426256937146707e-06,
      "loss": 0.378,
      "step": 485
    },
    {
      "epoch": 0.9435173299101413,
      "grad_norm": 0.09668080121210802,
      "learning_rate": 6.34546299705195e-06,
      "loss": 0.3813,
      "step": 490
    },
    {
      "epoch": 0.9531450577663672,
      "grad_norm": 0.10076895625991092,
      "learning_rate": 6.264288510639922e-06,
      "loss": 0.3629,
      "step": 495
    },
    {
      "epoch": 0.962772785622593,
      "grad_norm": 0.08864741394595178,
      "learning_rate": 6.182756437034677e-06,
      "loss": 0.3924,
      "step": 500
    },
    {
      "epoch": 0.972400513478819,
      "grad_norm": 0.09234930786675101,
      "learning_rate": 6.100889836499052e-06,
      "loss": 0.4034,
      "step": 505
    },
    {
      "epoch": 0.982028241335045,
      "grad_norm": 0.10439073600863705,
      "learning_rate": 6.018711863912381e-06,
      "loss": 0.3867,
      "step": 510
    },
    {
      "epoch": 0.9916559691912709,
      "grad_norm": 0.0832820482402092,
      "learning_rate": 5.936245762221436e-06,
      "loss": 0.3866,
      "step": 515
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.08006018844085963,
      "learning_rate": 5.853514855866481e-06,
      "loss": 0.3859,
      "step": 520
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3565569818019867,
      "eval_runtime": 14.4227,
      "eval_samples_per_second": 17.472,
      "eval_steps_per_second": 4.368,
      "step": 520
    },
    {
      "epoch": 1.009627727856226,
      "grad_norm": 0.09401110743601077,
      "learning_rate": 5.770542544184252e-06,
      "loss": 0.3628,
      "step": 525
    },
    {
      "epoch": 1.0192554557124518,
      "grad_norm": 0.0946480182617177,
      "learning_rate": 5.687352294789776e-06,
      "loss": 0.3708,
      "step": 530
    },
    {
      "epoch": 1.0288831835686778,
      "grad_norm": 0.10846300152568576,
      "learning_rate": 5.603967636938864e-06,
      "loss": 0.3625,
      "step": 535
    },
    {
      "epoch": 1.0385109114249038,
      "grad_norm": 0.09681851328737975,
      "learning_rate": 5.52041215487317e-06,
      "loss": 0.3666,
      "step": 540
    },
    {
      "epoch": 1.0481386392811296,
      "grad_norm": 0.12033097836718823,
      "learning_rate": 5.4367094811497105e-06,
      "loss": 0.3617,
      "step": 545
    },
    {
      "epoch": 1.0577663671373556,
      "grad_norm": 0.09615564194204358,
      "learning_rate": 5.352883289956701e-06,
      "loss": 0.37,
      "step": 550
    },
    {
      "epoch": 1.0673940949935816,
      "grad_norm": 0.09285850703825273,
      "learning_rate": 5.268957290417632e-06,
      "loss": 0.3856,
      "step": 555
    },
    {
      "epoch": 1.0770218228498074,
      "grad_norm": 0.08976885470379573,
      "learning_rate": 5.184955219885457e-06,
      "loss": 0.3747,
      "step": 560
    },
    {
      "epoch": 1.0866495507060334,
      "grad_norm": 0.1022360329654874,
      "learning_rate": 5.1009008372288015e-06,
      "loss": 0.365,
      "step": 565
    },
    {
      "epoch": 1.0962772785622592,
      "grad_norm": 0.1035446669504653,
      "learning_rate": 5.016817916112075e-06,
      "loss": 0.3624,
      "step": 570
    },
    {
      "epoch": 1.1059050064184852,
      "grad_norm": 0.09569503568277107,
      "learning_rate": 4.932730238271414e-06,
      "loss": 0.3559,
      "step": 575
    },
    {
      "epoch": 1.1155327342747112,
      "grad_norm": 0.08861451080285555,
      "learning_rate": 4.848661586788334e-06,
      "loss": 0.3682,
      "step": 580
    },
    {
      "epoch": 1.125160462130937,
      "grad_norm": 0.09786515991475879,
      "learning_rate": 4.764635739362988e-06,
      "loss": 0.3715,
      "step": 585
    },
    {
      "epoch": 1.134788189987163,
      "grad_norm": 0.12183470701722215,
      "learning_rate": 4.68067646158898e-06,
      "loss": 0.3773,
      "step": 590
    },
    {
      "epoch": 1.144415917843389,
      "grad_norm": 0.0848742632850072,
      "learning_rate": 4.596807500231573e-06,
      "loss": 0.3601,
      "step": 595
    },
    {
      "epoch": 1.1540436456996148,
      "grad_norm": 0.10570291615511458,
      "learning_rate": 4.513052576511227e-06,
      "loss": 0.3635,
      "step": 600
    },
    {
      "epoch": 1.1636713735558408,
      "grad_norm": 0.10634802539511004,
      "learning_rate": 4.429435379394377e-06,
      "loss": 0.3568,
      "step": 605
    },
    {
      "epoch": 1.1732991014120668,
      "grad_norm": 0.13446103864517522,
      "learning_rate": 4.34597955889332e-06,
      "loss": 0.3586,
      "step": 610
    },
    {
      "epoch": 1.1829268292682926,
      "grad_norm": 0.0941264293467416,
      "learning_rate": 4.262708719377118e-06,
      "loss": 0.3481,
      "step": 615
    },
    {
      "epoch": 1.1925545571245186,
      "grad_norm": 0.1410741383886165,
      "learning_rate": 4.17964641289543e-06,
      "loss": 0.385,
      "step": 620
    },
    {
      "epoch": 1.2021822849807444,
      "grad_norm": 0.09241491761949897,
      "learning_rate": 4.096816132517108e-06,
      "loss": 0.3785,
      "step": 625
    },
    {
      "epoch": 1.2118100128369704,
      "grad_norm": 0.11124645421714398,
      "learning_rate": 4.014241305685514e-06,
      "loss": 0.3681,
      "step": 630
    },
    {
      "epoch": 1.2214377406931964,
      "grad_norm": 0.11023070742649112,
      "learning_rate": 3.9319452875923725e-06,
      "loss": 0.3411,
      "step": 635
    },
    {
      "epoch": 1.2310654685494224,
      "grad_norm": 0.09552140786807434,
      "learning_rate": 3.849951354572057e-06,
      "loss": 0.3459,
      "step": 640
    },
    {
      "epoch": 1.2406931964056482,
      "grad_norm": 0.12231047507514385,
      "learning_rate": 3.7682826975182e-06,
      "loss": 0.3535,
      "step": 645
    },
    {
      "epoch": 1.2503209242618742,
      "grad_norm": 0.10830754166725957,
      "learning_rate": 3.686962415324452e-06,
      "loss": 0.3572,
      "step": 650
    },
    {
      "epoch": 1.2599486521181,
      "grad_norm": 0.14342247712840153,
      "learning_rate": 3.6060135083512656e-06,
      "loss": 0.366,
      "step": 655
    },
    {
      "epoch": 1.269576379974326,
      "grad_norm": 0.11323866001401849,
      "learning_rate": 3.5254588719205494e-06,
      "loss": 0.3539,
      "step": 660
    },
    {
      "epoch": 1.279204107830552,
      "grad_norm": 0.09929728772523248,
      "learning_rate": 3.4453212898400355e-06,
      "loss": 0.348,
      "step": 665
    },
    {
      "epoch": 1.2888318356867778,
      "grad_norm": 0.17492229609956267,
      "learning_rate": 3.365623427959175e-06,
      "loss": 0.3389,
      "step": 670
    },
    {
      "epoch": 1.2984595635430038,
      "grad_norm": 0.09292411937812878,
      "learning_rate": 3.2863878277584125e-06,
      "loss": 0.3523,
      "step": 675
    },
    {
      "epoch": 1.3080872913992299,
      "grad_norm": 0.10385494142168447,
      "learning_rate": 3.2076368999736175e-06,
      "loss": 0.3546,
      "step": 680
    },
    {
      "epoch": 1.3177150192554556,
      "grad_norm": 0.15116008215242238,
      "learning_rate": 3.1293929182575154e-06,
      "loss": 0.3595,
      "step": 685
    },
    {
      "epoch": 1.3273427471116817,
      "grad_norm": 0.13134864128367904,
      "learning_rate": 3.0516780128798794e-06,
      "loss": 0.3604,
      "step": 690
    },
    {
      "epoch": 1.3369704749679077,
      "grad_norm": 0.1385810651745671,
      "learning_rate": 2.9745141644682684e-06,
      "loss": 0.3817,
      "step": 695
    },
    {
      "epoch": 1.3465982028241335,
      "grad_norm": 0.11308291681302851,
      "learning_rate": 2.8979231977911127e-06,
      "loss": 0.3436,
      "step": 700
    },
    {
      "epoch": 1.3562259306803595,
      "grad_norm": 0.1311446028434404,
      "learning_rate": 2.8219267755848613e-06,
      "loss": 0.3804,
      "step": 705
    },
    {
      "epoch": 1.3658536585365852,
      "grad_norm": 0.10654779695663232,
      "learning_rate": 2.746546392426963e-06,
      "loss": 0.3426,
      "step": 710
    },
    {
      "epoch": 1.3754813863928113,
      "grad_norm": 0.14977867299262188,
      "learning_rate": 2.671803368656413e-06,
      "loss": 0.373,
      "step": 715
    },
    {
      "epoch": 1.3851091142490373,
      "grad_norm": 0.10008323376413787,
      "learning_rate": 2.5977188443435874e-06,
      "loss": 0.3707,
      "step": 720
    },
    {
      "epoch": 1.3947368421052633,
      "grad_norm": 0.12287994467841869,
      "learning_rate": 2.524313773311032e-06,
      "loss": 0.3419,
      "step": 725
    },
    {
      "epoch": 1.404364569961489,
      "grad_norm": 0.13242489561288942,
      "learning_rate": 2.4516089172069852e-06,
      "loss": 0.3505,
      "step": 730
    },
    {
      "epoch": 1.413992297817715,
      "grad_norm": 0.15290371838598157,
      "learning_rate": 2.379624839633187e-06,
      "loss": 0.3433,
      "step": 735
    },
    {
      "epoch": 1.4236200256739409,
      "grad_norm": 0.15079312975317574,
      "learning_rate": 2.308381900328767e-06,
      "loss": 0.3521,
      "step": 740
    },
    {
      "epoch": 1.4332477535301669,
      "grad_norm": 0.13283749811762047,
      "learning_rate": 2.2379002494117467e-06,
      "loss": 0.3694,
      "step": 745
    },
    {
      "epoch": 1.4428754813863929,
      "grad_norm": 0.13332742256673105,
      "learning_rate": 2.1681998216798476e-06,
      "loss": 0.3425,
      "step": 750
    },
    {
      "epoch": 1.4525032092426187,
      "grad_norm": 0.14406157700699168,
      "learning_rate": 2.099300330972209e-06,
      "loss": 0.3358,
      "step": 755
    },
    {
      "epoch": 1.4621309370988447,
      "grad_norm": 0.14024887794079086,
      "learning_rate": 2.0312212645935755e-06,
      "loss": 0.3461,
      "step": 760
    },
    {
      "epoch": 1.4717586649550707,
      "grad_norm": 0.15414062139512394,
      "learning_rate": 1.9639818778025836e-06,
      "loss": 0.3477,
      "step": 765
    },
    {
      "epoch": 1.4813863928112965,
      "grad_norm": 0.1369666143345331,
      "learning_rate": 1.8976011883656632e-06,
      "loss": 0.3611,
      "step": 770
    },
    {
      "epoch": 1.4910141206675225,
      "grad_norm": 0.15223059852909734,
      "learning_rate": 1.8320979711781212e-06,
      "loss": 0.3476,
      "step": 775
    },
    {
      "epoch": 1.5006418485237485,
      "grad_norm": 0.1714137202575772,
      "learning_rate": 1.767490752953896e-06,
      "loss": 0.3129,
      "step": 780
    },
    {
      "epoch": 1.5102695763799743,
      "grad_norm": 0.14319611843869498,
      "learning_rate": 1.7037978069855472e-06,
      "loss": 0.3407,
      "step": 785
    },
    {
      "epoch": 1.5198973042362003,
      "grad_norm": 0.16393330833863626,
      "learning_rate": 1.641037147975872e-06,
      "loss": 0.3526,
      "step": 790
    },
    {
      "epoch": 1.529525032092426,
      "grad_norm": 0.16364031655094152,
      "learning_rate": 1.5792265269427082e-06,
      "loss": 0.3456,
      "step": 795
    },
    {
      "epoch": 1.539152759948652,
      "grad_norm": 0.16348327211177935,
      "learning_rate": 1.5183834261982804e-06,
      "loss": 0.3398,
      "step": 800
    },
    {
      "epoch": 1.548780487804878,
      "grad_norm": 0.15134397957607432,
      "learning_rate": 1.4585250544045666e-06,
      "loss": 0.3479,
      "step": 805
    },
    {
      "epoch": 1.558408215661104,
      "grad_norm": 0.17575731164707303,
      "learning_rate": 1.399668341706053e-06,
      "loss": 0.322,
      "step": 810
    },
    {
      "epoch": 1.5680359435173299,
      "grad_norm": 0.17640101887621162,
      "learning_rate": 1.3418299349412761e-06,
      "loss": 0.3439,
      "step": 815
    },
    {
      "epoch": 1.5776636713735557,
      "grad_norm": 0.15625925540478658,
      "learning_rate": 1.2850261929344748e-06,
      "loss": 0.338,
      "step": 820
    },
    {
      "epoch": 1.5872913992297817,
      "grad_norm": 0.1475923645387349,
      "learning_rate": 1.2292731818687204e-06,
      "loss": 0.344,
      "step": 825
    },
    {
      "epoch": 1.5969191270860077,
      "grad_norm": 0.1598002892634283,
      "learning_rate": 1.1745866707418146e-06,
      "loss": 0.3524,
      "step": 830
    },
    {
      "epoch": 1.6065468549422337,
      "grad_norm": 0.16235140780790008,
      "learning_rate": 1.1209821269062254e-06,
      "loss": 0.3314,
      "step": 835
    },
    {
      "epoch": 1.6161745827984597,
      "grad_norm": 0.1817156515805022,
      "learning_rate": 1.0684747116943683e-06,
      "loss": 0.3279,
      "step": 840
    },
    {
      "epoch": 1.6258023106546855,
      "grad_norm": 0.162000455486421,
      "learning_rate": 1.0170792761304083e-06,
      "loss": 0.3562,
      "step": 845
    },
    {
      "epoch": 1.6354300385109113,
      "grad_norm": 0.22932643721397972,
      "learning_rate": 9.668103567298615e-07,
      "loss": 0.3374,
      "step": 850
    },
    {
      "epoch": 1.6450577663671373,
      "grad_norm": 0.18052968600391292,
      "learning_rate": 9.176821713881229e-07,
      "loss": 0.3409,
      "step": 855
    },
    {
      "epoch": 1.6546854942233633,
      "grad_norm": 0.19204286956569327,
      "learning_rate": 8.697086153591289e-07,
      "loss": 0.3411,
      "step": 860
    },
    {
      "epoch": 1.6643132220795893,
      "grad_norm": 0.15338857676660417,
      "learning_rate": 8.229032573252671e-07,
      "loss": 0.3509,
      "step": 865
    },
    {
      "epoch": 1.673940949935815,
      "grad_norm": 0.2064824927113351,
      "learning_rate": 7.772793355596597e-07,
      "loss": 0.3535,
      "step": 870
    },
    {
      "epoch": 1.6835686777920411,
      "grad_norm": 0.1901669621594226,
      "learning_rate": 7.328497541818891e-07,
      "loss": 0.35,
      "step": 875
    },
    {
      "epoch": 1.693196405648267,
      "grad_norm": 0.20680883575996672,
      "learning_rate": 6.896270795082394e-07,
      "loss": 0.3284,
      "step": 880
    },
    {
      "epoch": 1.702824133504493,
      "grad_norm": 0.208535850753021,
      "learning_rate": 6.476235364974848e-07,
      "loss": 0.3388,
      "step": 885
    },
    {
      "epoch": 1.712451861360719,
      "grad_norm": 0.20588654883512666,
      "learning_rate": 6.06851005293217e-07,
      "loss": 0.3361,
      "step": 890
    },
    {
      "epoch": 1.722079589216945,
      "grad_norm": 0.20944379197373206,
      "learning_rate": 5.673210178637129e-07,
      "loss": 0.3285,
      "step": 895
    },
    {
      "epoch": 1.7317073170731707,
      "grad_norm": 0.192174693171731,
      "learning_rate": 5.290447547402594e-07,
      "loss": 0.3357,
      "step": 900
    },
    {
      "epoch": 1.7413350449293965,
      "grad_norm": 0.2791441208177208,
      "learning_rate": 4.920330418549007e-07,
      "loss": 0.3306,
      "step": 905
    },
    {
      "epoch": 1.7509627727856225,
      "grad_norm": 0.2268656547573065,
      "learning_rate": 4.5629634747845764e-07,
      "loss": 0.3012,
      "step": 910
    },
    {
      "epoch": 1.7605905006418485,
      "grad_norm": 0.20835714245970288,
      "learning_rate": 4.218447792597219e-07,
      "loss": 0.3425,
      "step": 915
    },
    {
      "epoch": 1.7702182284980745,
      "grad_norm": 0.23100114988163314,
      "learning_rate": 3.8868808136663995e-07,
      "loss": 0.3227,
      "step": 920
    },
    {
      "epoch": 1.7798459563543005,
      "grad_norm": 0.21698851745344763,
      "learning_rate": 3.56835631730309e-07,
      "loss": 0.3357,
      "step": 925
    },
    {
      "epoch": 1.7894736842105263,
      "grad_norm": 0.19672243352771557,
      "learning_rate": 3.262964393925433e-07,
      "loss": 0.3239,
      "step": 930
    },
    {
      "epoch": 1.7991014120667521,
      "grad_norm": 0.23389991198328033,
      "learning_rate": 2.970791419577995e-07,
      "loss": 0.3047,
      "step": 935
    },
    {
      "epoch": 1.8087291399229781,
      "grad_norm": 0.23366499263913967,
      "learning_rate": 2.6919200315013606e-07,
      "loss": 0.287,
      "step": 940
    },
    {
      "epoch": 1.8183568677792041,
      "grad_norm": 0.1966679887252355,
      "learning_rate": 2.42642910475932e-07,
      "loss": 0.3255,
      "step": 945
    },
    {
      "epoch": 1.8279845956354301,
      "grad_norm": 0.2366162252630318,
      "learning_rate": 2.1743937299301242e-07,
      "loss": 0.3249,
      "step": 950
    },
    {
      "epoch": 1.837612323491656,
      "grad_norm": 0.17846484227641432,
      "learning_rate": 1.9358851918680478e-07,
      "loss": 0.3319,
      "step": 955
    },
    {
      "epoch": 1.847240051347882,
      "grad_norm": 0.22183877870838897,
      "learning_rate": 1.7109709495415073e-07,
      "loss": 0.2987,
      "step": 960
    },
    {
      "epoch": 1.8568677792041077,
      "grad_norm": 0.1922746561314516,
      "learning_rate": 1.499714616953124e-07,
      "loss": 0.3273,
      "step": 965
    },
    {
      "epoch": 1.8664955070603337,
      "grad_norm": 0.2129526026233563,
      "learning_rate": 1.3021759451473548e-07,
      "loss": 0.3165,
      "step": 970
    },
    {
      "epoch": 1.8761232349165597,
      "grad_norm": 0.22092907981647764,
      "learning_rate": 1.1184108053107268e-07,
      "loss": 0.3201,
      "step": 975
    },
    {
      "epoch": 1.8857509627727858,
      "grad_norm": 0.21605685840962405,
      "learning_rate": 9.484711729694229e-08,
      "loss": 0.3334,
      "step": 980
    },
    {
      "epoch": 1.8953786906290115,
      "grad_norm": 0.23787534108452493,
      "learning_rate": 7.924051132886357e-08,
      "loss": 0.3173,
      "step": 985
    },
    {
      "epoch": 1.9050064184852373,
      "grad_norm": 0.17867415785242957,
      "learning_rate": 6.502567674780524e-08,
      "loss": 0.3351,
      "step": 990
    },
    {
      "epoch": 1.9146341463414633,
      "grad_norm": 0.2159405119405128,
      "learning_rate": 5.220663403070592e-08,
      "loss": 0.3144,
      "step": 995
    },
    {
      "epoch": 1.9242618741976893,
      "grad_norm": 0.1818905292820818,
      "learning_rate": 4.078700887333365e-08,
      "loss": 0.3399,
      "step": 1000
    },
    {
      "epoch": 1.9338896020539154,
      "grad_norm": 0.2160735551084158,
      "learning_rate": 3.0770031164812366e-08,
      "loss": 0.3399,
      "step": 1005
    },
    {
      "epoch": 1.9435173299101414,
      "grad_norm": 0.19974665047237547,
      "learning_rate": 2.2158534074083193e-08,
      "loss": 0.3046,
      "step": 1010
    },
    {
      "epoch": 1.9531450577663672,
      "grad_norm": 0.23888039492400656,
      "learning_rate": 1.495495324858709e-08,
      "loss": 0.329,
      "step": 1015
    },
    {
      "epoch": 1.962772785622593,
      "grad_norm": 0.23943732877322751,
      "learning_rate": 9.16132612537035e-09,
      "loss": 0.3211,
      "step": 1020
    },
    {
      "epoch": 1.972400513478819,
      "grad_norm": 0.23280339101940556,
      "learning_rate": 4.779291354822802e-09,
      "loss": 0.3058,
      "step": 1025
    },
    {
      "epoch": 1.982028241335045,
      "grad_norm": 0.2491328973768298,
      "learning_rate": 1.8100883372085266e-09,
      "loss": 0.3212,
      "step": 1030
    },
    {
      "epoch": 1.991655969191271,
      "grad_norm": 0.22142463430680218,
      "learning_rate": 2.545568721179503e-10,
      "loss": 0.3358,
      "step": 1035
    },
    {
      "epoch": 1.9974326059050065,
      "eval_loss": 0.3029652535915375,
      "eval_runtime": 12.342,
      "eval_samples_per_second": 20.418,
      "eval_steps_per_second": 5.105,
      "step": 1038
    },
    {
      "epoch": 1.9974326059050065,
      "step": 1038,
      "total_flos": 2.6515469798631014e+18,
      "train_loss": 0.4156393744697461,
      "train_runtime": 8276.9589,
      "train_samples_per_second": 6.023,
      "train_steps_per_second": 0.125
    }
  ],
  "logging_steps": 5,
  "max_steps": 1038,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.6515469798631014e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}