{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0458919092563981,
  "eval_steps": 500,
  "global_step": 1500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00021416224319652444,
      "grad_norm": 0.7752708792686462,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.9009,
      "step": 7
    },
    {
      "epoch": 0.0004283244863930489,
      "grad_norm": 0.6635493636131287,
      "learning_rate": 8.666666666666668e-06,
      "loss": 1.8402,
      "step": 14
    },
    {
      "epoch": 0.0006424867295895734,
      "grad_norm": 0.5666959285736084,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.6104,
      "step": 21
    },
    {
      "epoch": 0.0008566489727860978,
      "grad_norm": 0.4543763995170593,
      "learning_rate": 1.8e-05,
      "loss": 1.5991,
      "step": 28
    },
    {
      "epoch": 0.0010708112159826223,
      "grad_norm": 0.37796229124069214,
      "learning_rate": 2.2666666666666668e-05,
      "loss": 1.5406,
      "step": 35
    },
    {
      "epoch": 0.0012849734591791468,
      "grad_norm": 0.553122341632843,
      "learning_rate": 2.733333333333333e-05,
      "loss": 1.705,
      "step": 42
    },
    {
      "epoch": 0.0014991357023756713,
      "grad_norm": 0.5612793564796448,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.556,
      "step": 49
    },
    {
      "epoch": 0.0017132979455721955,
      "grad_norm": 0.495226114988327,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 1.6175,
      "step": 56
    },
    {
      "epoch": 0.00192746018876872,
      "grad_norm": 0.622969925403595,
      "learning_rate": 4.133333333333333e-05,
      "loss": 1.5846,
      "step": 63
    },
    {
      "epoch": 0.0021416224319652445,
      "grad_norm": 0.6336926221847534,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.4601,
      "step": 70
    },
    {
      "epoch": 0.002355784675161769,
      "grad_norm": 0.4700409173965454,
      "learning_rate": 5.0666666666666674e-05,
      "loss": 1.6889,
      "step": 77
    },
    {
      "epoch": 0.0025699469183582935,
      "grad_norm": 0.5999062061309814,
      "learning_rate": 5.5333333333333334e-05,
      "loss": 1.6322,
      "step": 84
    },
    {
      "epoch": 0.002784109161554818,
      "grad_norm": 0.7579219341278076,
      "learning_rate": 6e-05,
      "loss": 1.4542,
      "step": 91
    },
    {
      "epoch": 0.0029982714047513425,
      "grad_norm": 0.6173705458641052,
      "learning_rate": 6.466666666666666e-05,
      "loss": 1.4948,
      "step": 98
    },
    {
      "epoch": 0.0032124336479478666,
      "grad_norm": 0.4720444679260254,
      "learning_rate": 6.933333333333334e-05,
      "loss": 1.5726,
      "step": 105
    },
    {
      "epoch": 0.003426595891144391,
      "grad_norm": 1.1178276538848877,
      "learning_rate": 7.4e-05,
      "loss": 1.5583,
      "step": 112
    },
    {
      "epoch": 0.0036407581343409156,
      "grad_norm": 0.660765528678894,
      "learning_rate": 7.866666666666666e-05,
      "loss": 1.3969,
      "step": 119
    },
    {
      "epoch": 0.00385492037753744,
      "grad_norm": 0.7253082394599915,
      "learning_rate": 8.333333333333334e-05,
      "loss": 1.5692,
      "step": 126
    },
    {
      "epoch": 0.004069082620733965,
      "grad_norm": 0.8589861989021301,
      "learning_rate": 8.800000000000001e-05,
      "loss": 1.3925,
      "step": 133
    },
    {
      "epoch": 0.004283244863930489,
      "grad_norm": 0.6899182796478271,
      "learning_rate": 9.266666666666666e-05,
      "loss": 1.4918,
      "step": 140
    },
    {
      "epoch": 0.004497407107127013,
      "grad_norm": 0.8465347290039062,
      "learning_rate": 9.733333333333335e-05,
      "loss": 1.4168,
      "step": 147
    },
    {
      "epoch": 0.004711569350323538,
      "grad_norm": 0.8710561990737915,
      "learning_rate": 9.999878153526974e-05,
      "loss": 1.4505,
      "step": 154
    },
    {
      "epoch": 0.004925731593520062,
      "grad_norm": 1.1197599172592163,
      "learning_rate": 9.998646205897309e-05,
      "loss": 1.6068,
      "step": 161
    },
    {
      "epoch": 0.005139893836716587,
      "grad_norm": 0.7392743229866028,
      "learning_rate": 9.996087868740243e-05,
      "loss": 1.626,
      "step": 168
    },
    {
      "epoch": 0.005354056079913111,
      "grad_norm": 0.8371026515960693,
      "learning_rate": 9.992203820909906e-05,
      "loss": 1.5677,
      "step": 175
    },
    {
      "epoch": 0.005568218323109636,
      "grad_norm": 0.6864525675773621,
      "learning_rate": 9.986995093037421e-05,
      "loss": 1.4513,
      "step": 182
    },
    {
      "epoch": 0.00578238056630616,
      "grad_norm": 0.777845025062561,
      "learning_rate": 9.980463067257437e-05,
      "loss": 1.3807,
      "step": 189
    },
    {
      "epoch": 0.005996542809502685,
      "grad_norm": 0.9488137364387512,
      "learning_rate": 9.972609476841367e-05,
      "loss": 1.5597,
      "step": 196
    },
    {
      "epoch": 0.006210705052699209,
      "grad_norm": 0.779089093208313,
      "learning_rate": 9.963436405737476e-05,
      "loss": 1.6287,
      "step": 203
    },
    {
      "epoch": 0.006424867295895733,
      "grad_norm": 0.885947585105896,
      "learning_rate": 9.952946288017899e-05,
      "loss": 1.7233,
      "step": 210
    },
    {
      "epoch": 0.006639029539092258,
      "grad_norm": 0.9541194438934326,
      "learning_rate": 9.941141907232765e-05,
      "loss": 1.747,
      "step": 217
    },
    {
      "epoch": 0.006853191782288782,
      "grad_norm": 0.9196546673774719,
      "learning_rate": 9.928026395671576e-05,
      "loss": 1.5667,
      "step": 224
    },
    {
      "epoch": 0.007067354025485307,
      "grad_norm": 1.2334916591644287,
      "learning_rate": 9.913603233532067e-05,
      "loss": 1.6394,
      "step": 231
    },
    {
      "epoch": 0.007281516268681831,
      "grad_norm": 0.9803845286369324,
      "learning_rate": 9.89787624799672e-05,
      "loss": 1.4262,
      "step": 238
    },
    {
      "epoch": 0.007495678511878356,
      "grad_norm": 1.4271188974380493,
      "learning_rate": 9.880849612217238e-05,
      "loss": 1.5741,
      "step": 245
    },
    {
      "epoch": 0.00770984075507488,
      "grad_norm": 0.9486837983131409,
      "learning_rate": 9.862527844207189e-05,
      "loss": 1.5092,
      "step": 252
    },
    {
      "epoch": 0.007924002998271404,
      "grad_norm": 1.0058749914169312,
      "learning_rate": 9.842915805643155e-05,
      "loss": 1.5684,
      "step": 259
    },
    {
      "epoch": 0.00813816524146793,
      "grad_norm": 1.3838095664978027,
      "learning_rate": 9.822018700574695e-05,
      "loss": 1.5235,
      "step": 266
    },
    {
      "epoch": 0.008352327484664454,
      "grad_norm": 0.9538054466247559,
      "learning_rate": 9.799842074043439e-05,
      "loss": 1.4871,
      "step": 273
    },
    {
      "epoch": 0.008566489727860978,
      "grad_norm": 1.0226269960403442,
      "learning_rate": 9.776391810611718e-05,
      "loss": 1.5704,
      "step": 280
    },
    {
      "epoch": 0.008780651971057502,
      "grad_norm": 1.5275094509124756,
      "learning_rate": 9.751674132801107e-05,
      "loss": 1.6827,
      "step": 287
    },
    {
      "epoch": 0.008994814214254026,
      "grad_norm": 1.4683727025985718,
      "learning_rate": 9.725695599441258e-05,
      "loss": 1.493,
      "step": 294
    },
    {
      "epoch": 0.009208976457450552,
      "grad_norm": 1.014497995376587,
      "learning_rate": 9.698463103929542e-05,
      "loss": 1.5605,
      "step": 301
    },
    {
      "epoch": 0.009423138700647076,
      "grad_norm": 1.4134587049484253,
      "learning_rate": 9.669983872401867e-05,
      "loss": 1.7188,
      "step": 308
    },
    {
      "epoch": 0.0096373009438436,
      "grad_norm": 1.3481359481811523,
      "learning_rate": 9.640265461815234e-05,
      "loss": 1.4565,
      "step": 315
    },
    {
      "epoch": 0.009851463187040124,
      "grad_norm": 0.8728109002113342,
      "learning_rate": 9.609315757942503e-05,
      "loss": 1.5941,
      "step": 322
    },
    {
      "epoch": 0.01006562543023665,
      "grad_norm": 0.9764572381973267,
      "learning_rate": 9.577142973279896e-05,
      "loss": 1.548,
      "step": 329
    },
    {
      "epoch": 0.010279787673433174,
      "grad_norm": 1.630246877670288,
      "learning_rate": 9.543755644867822e-05,
      "loss": 1.5262,
      "step": 336
    },
    {
      "epoch": 0.010493949916629698,
      "grad_norm": 1.1554211378097534,
      "learning_rate": 9.50916263202557e-05,
      "loss": 1.4859,
      "step": 343
    },
    {
      "epoch": 0.010708112159826222,
      "grad_norm": 1.3077260255813599,
      "learning_rate": 9.473373114000492e-05,
      "loss": 1.4612,
      "step": 350
    },
    {
      "epoch": 0.010922274403022746,
      "grad_norm": 1.9986414909362793,
      "learning_rate": 9.436396587532296e-05,
      "loss": 1.627,
      "step": 357
    },
    {
      "epoch": 0.011136436646219272,
      "grad_norm": 1.4336645603179932,
      "learning_rate": 9.398242864333083e-05,
      "loss": 1.5606,
      "step": 364
    },
    {
      "epoch": 0.011350598889415796,
      "grad_norm": 1.0279778242111206,
      "learning_rate": 9.358922068483812e-05,
      "loss": 1.532,
      "step": 371
    },
    {
      "epoch": 0.01156476113261232,
      "grad_norm": 1.2006853818893433,
      "learning_rate": 9.318444633747883e-05,
      "loss": 1.4958,
      "step": 378
    },
    {
      "epoch": 0.011778923375808844,
      "grad_norm": 1.528941035270691,
      "learning_rate": 9.276821300802534e-05,
      "loss": 1.6738,
      "step": 385
    },
    {
      "epoch": 0.01199308561900537,
      "grad_norm": 1.008375883102417,
      "learning_rate": 9.234063114388809e-05,
      "loss": 1.5629,
      "step": 392
    },
    {
      "epoch": 0.012207247862201894,
      "grad_norm": 1.8260549306869507,
      "learning_rate": 9.190181420380836e-05,
      "loss": 1.5956,
      "step": 399
    },
    {
      "epoch": 0.012421410105398418,
      "grad_norm": 1.231460452079773,
      "learning_rate": 9.145187862775209e-05,
      "loss": 1.5006,
      "step": 406
    },
    {
      "epoch": 0.012635572348594942,
      "grad_norm": 1.2324597835540771,
      "learning_rate": 9.099094380601244e-05,
      "loss": 1.5578,
      "step": 413
    },
    {
      "epoch": 0.012849734591791466,
      "grad_norm": 1.2664161920547485,
      "learning_rate": 9.051913204752972e-05,
      "loss": 1.3945,
      "step": 420
    },
    {
      "epoch": 0.013063896834987992,
      "grad_norm": 1.8512499332427979,
      "learning_rate": 9.003656854743667e-05,
      "loss": 1.5641,
      "step": 427
    },
    {
      "epoch": 0.013278059078184516,
      "grad_norm": 1.3970890045166016,
      "learning_rate": 8.954338135383804e-05,
      "loss": 1.5878,
      "step": 434
    },
    {
      "epoch": 0.01349222132138104,
      "grad_norm": 1.42117440700531,
      "learning_rate": 8.903970133383297e-05,
      "loss": 1.535,
      "step": 441
    },
    {
      "epoch": 0.013706383564577564,
      "grad_norm": 1.23691725730896,
      "learning_rate": 8.852566213878947e-05,
      "loss": 1.582,
      "step": 448
    },
    {
      "epoch": 0.01392054580777409,
      "grad_norm": 1.0575220584869385,
      "learning_rate": 8.800140016888009e-05,
      "loss": 1.5825,
      "step": 455
    },
    {
      "epoch": 0.014134708050970614,
      "grad_norm": 1.1797815561294556,
      "learning_rate": 8.746705453688814e-05,
      "loss": 1.6639,
      "step": 462
    },
    {
      "epoch": 0.014348870294167138,
      "grad_norm": 1.1307278871536255,
      "learning_rate": 8.692276703129421e-05,
      "loss": 1.6416,
      "step": 469
    },
    {
      "epoch": 0.014563032537363662,
      "grad_norm": 1.3289051055908203,
      "learning_rate": 8.636868207865244e-05,
      "loss": 1.504,
      "step": 476
    },
    {
      "epoch": 0.014777194780560188,
      "grad_norm": 1.5391404628753662,
      "learning_rate": 8.580494670526725e-05,
      "loss": 1.3852,
      "step": 483
    },
    {
      "epoch": 0.014991357023756712,
      "grad_norm": 0.9621486067771912,
      "learning_rate": 8.523171049817974e-05,
      "loss": 1.4558,
      "step": 490
    },
    {
      "epoch": 0.015205519266953236,
      "grad_norm": 1.2261931896209717,
      "learning_rate": 8.464912556547486e-05,
      "loss": 1.5033,
      "step": 497
    },
    {
      "epoch": 0.01541968151014976,
      "grad_norm": 1.3401846885681152,
      "learning_rate": 8.405734649591963e-05,
      "loss": 1.7003,
      "step": 504
    },
    {
      "epoch": 0.015633843753346286,
      "grad_norm": 1.2691761255264282,
      "learning_rate": 8.345653031794292e-05,
      "loss": 1.5769,
      "step": 511
    },
    {
      "epoch": 0.01584800599654281,
      "grad_norm": 1.6369774341583252,
      "learning_rate": 8.284683645796813e-05,
      "loss": 1.6432,
      "step": 518
    },
    {
      "epoch": 0.016062168239739334,
      "grad_norm": 1.509771704673767,
      "learning_rate": 8.222842669810935e-05,
      "loss": 1.6568,
      "step": 525
    },
    {
      "epoch": 0.01627633048293586,
      "grad_norm": 1.4711761474609375,
      "learning_rate": 8.160146513324254e-05,
      "loss": 1.5285,
      "step": 532
    },
    {
      "epoch": 0.016490492726132382,
      "grad_norm": 1.41879403591156,
      "learning_rate": 8.096611812746301e-05,
      "loss": 1.4978,
      "step": 539
    },
    {
      "epoch": 0.016704654969328908,
      "grad_norm": 1.2458244562149048,
      "learning_rate": 8.032255426994069e-05,
      "loss": 1.5741,
      "step": 546
    },
    {
      "epoch": 0.01691881721252543,
      "grad_norm": 1.1006284952163696,
      "learning_rate": 7.967094433018508e-05,
      "loss": 1.6395,
      "step": 553
    },
    {
      "epoch": 0.017132979455721956,
      "grad_norm": 1.5933326482772827,
      "learning_rate": 7.901146121273164e-05,
      "loss": 1.5952,
      "step": 560
    },
    {
      "epoch": 0.017347141698918482,
      "grad_norm": 1.4252536296844482,
      "learning_rate": 7.834427991126155e-05,
      "loss": 1.5419,
      "step": 567
    },
    {
      "epoch": 0.017561303942115004,
      "grad_norm": 1.452164888381958,
      "learning_rate": 7.766957746216721e-05,
      "loss": 1.4545,
      "step": 574
    },
    {
      "epoch": 0.01777546618531153,
      "grad_norm": 1.2681009769439697,
      "learning_rate": 7.698753289757565e-05,
      "loss": 1.5318,
      "step": 581
    },
    {
      "epoch": 0.017989628428508053,
      "grad_norm": 1.7576844692230225,
      "learning_rate": 7.629832719784245e-05,
      "loss": 1.4869,
      "step": 588
    },
    {
      "epoch": 0.01820379067170458,
      "grad_norm": 1.075482726097107,
      "learning_rate": 7.560214324352858e-05,
      "loss": 1.4676,
      "step": 595
    },
    {
      "epoch": 0.018417952914901104,
      "grad_norm": 1.2565999031066895,
      "learning_rate": 7.489916576687319e-05,
      "loss": 1.6566,
      "step": 602
    },
    {
      "epoch": 0.018632115158097626,
      "grad_norm": 1.6566146612167358,
      "learning_rate": 7.418958130277483e-05,
      "loss": 1.5047,
      "step": 609
    },
    {
      "epoch": 0.018846277401294152,
      "grad_norm": 1.0712546110153198,
      "learning_rate": 7.347357813929454e-05,
      "loss": 1.5777,
      "step": 616
    },
    {
      "epoch": 0.019060439644490675,
      "grad_norm": 1.2899247407913208,
      "learning_rate": 7.275134626769369e-05,
      "loss": 1.5059,
      "step": 623
    },
    {
      "epoch": 0.0192746018876872,
      "grad_norm": 1.3111728429794312,
      "learning_rate": 7.202307733201975e-05,
      "loss": 1.6085,
      "step": 630
    },
    {
      "epoch": 0.019488764130883726,
      "grad_norm": 1.4129036664962769,
      "learning_rate": 7.128896457825364e-05,
      "loss": 1.5525,
      "step": 637
    },
    {
      "epoch": 0.01970292637408025,
      "grad_norm": 1.9603643417358398,
      "learning_rate": 7.054920280303198e-05,
      "loss": 1.5722,
      "step": 644
    },
    {
      "epoch": 0.019917088617276774,
      "grad_norm": 1.5986700057983398,
      "learning_rate": 6.980398830195785e-05,
      "loss": 1.5231,
      "step": 651
    },
    {
      "epoch": 0.0201312508604733,
      "grad_norm": 1.2295145988464355,
      "learning_rate": 6.905351881751372e-05,
      "loss": 1.6469,
      "step": 658
    },
    {
      "epoch": 0.020345413103669822,
      "grad_norm": 1.1692911386489868,
      "learning_rate": 6.82979934865906e-05,
      "loss": 1.528,
      "step": 665
    },
    {
      "epoch": 0.020559575346866348,
      "grad_norm": 1.3855468034744263,
      "learning_rate": 6.753761278764719e-05,
      "loss": 1.4652,
      "step": 672
    },
    {
      "epoch": 0.02077373759006287,
      "grad_norm": 1.2203513383865356,
      "learning_rate": 6.677257848751277e-05,
      "loss": 1.6676,
      "step": 679
    },
    {
      "epoch": 0.020987899833259396,
      "grad_norm": 1.0921286344528198,
      "learning_rate": 6.600309358784857e-05,
      "loss": 1.6842,
      "step": 686
    },
    {
      "epoch": 0.021202062076455922,
      "grad_norm": 1.1615856885910034,
      "learning_rate": 6.52293622712814e-05,
      "loss": 1.478,
      "step": 693
    },
    {
      "epoch": 0.021416224319652444,
      "grad_norm": 1.7459659576416016,
      "learning_rate": 6.445158984722358e-05,
      "loss": 1.5996,
      "step": 700
    },
    {
      "epoch": 0.02163038656284897,
      "grad_norm": 1.4326015710830688,
      "learning_rate": 6.366998269739441e-05,
      "loss": 1.5972,
      "step": 707
    },
    {
      "epoch": 0.021844548806045493,
      "grad_norm": 1.4477587938308716,
      "learning_rate": 6.28847482210567e-05,
      "loss": 1.4253,
      "step": 714
    },
    {
      "epoch": 0.02205871104924202,
      "grad_norm": 1.6338754892349243,
      "learning_rate": 6.209609477998338e-05,
      "loss": 1.4285,
      "step": 721
    },
    {
      "epoch": 0.022272873292438544,
      "grad_norm": 1.1946927309036255,
      "learning_rate": 6.130423164316893e-05,
      "loss": 1.5696,
      "step": 728
    },
    {
      "epoch": 0.022487035535635067,
      "grad_norm": 1.4626315832138062,
      "learning_rate": 6.05093689312997e-05,
      "loss": 1.5052,
      "step": 735
    },
    {
      "epoch": 0.022701197778831592,
      "grad_norm": 1.8981741666793823,
      "learning_rate": 5.97117175609986e-05,
      "loss": 1.6392,
      "step": 742
    },
    {
      "epoch": 0.022915360022028118,
      "grad_norm": 1.125319242477417,
      "learning_rate": 5.891148918885834e-05,
      "loss": 1.6573,
      "step": 749
    },
    {
      "epoch": 0.02312952226522464,
      "grad_norm": 2.1538913249969482,
      "learning_rate": 5.810889615527838e-05,
      "loss": 1.5932,
      "step": 756
    },
    {
      "epoch": 0.023343684508421166,
      "grad_norm": 1.2083415985107422,
      "learning_rate": 5.730415142812059e-05,
      "loss": 1.4371,
      "step": 763
    },
    {
      "epoch": 0.02355784675161769,
      "grad_norm": 1.177195429801941,
      "learning_rate": 5.649746854619814e-05,
      "loss": 1.4984,
      "step": 770
    },
    {
      "epoch": 0.023772008994814214,
      "grad_norm": 2.0190370082855225,
      "learning_rate": 5.568906156261309e-05,
      "loss": 1.7023,
      "step": 777
    },
    {
      "epoch": 0.02398617123801074,
      "grad_norm": 1.478585124015808,
      "learning_rate": 5.487914498795747e-05,
      "loss": 1.5875,
      "step": 784
    },
    {
      "epoch": 0.024200333481207262,
      "grad_norm": 1.2122184038162231,
      "learning_rate": 5.4067933733392915e-05,
      "loss": 1.5887,
      "step": 791
    },
    {
      "epoch": 0.02441449572440379,
      "grad_norm": 1.549975872039795,
      "learning_rate": 5.325564305362404e-05,
      "loss": 1.457,
      "step": 798
    },
    {
      "epoch": 0.02462865796760031,
      "grad_norm": 1.2659565210342407,
      "learning_rate": 5.244248848978067e-05,
      "loss": 1.5007,
      "step": 805
    },
    {
      "epoch": 0.024842820210796836,
      "grad_norm": 1.7076735496520996,
      "learning_rate": 5.162868581222406e-05,
      "loss": 1.593,
      "step": 812
    },
    {
      "epoch": 0.025056982453993362,
      "grad_norm": 1.2901355028152466,
      "learning_rate": 5.0814450963292295e-05,
      "loss": 1.4844,
      "step": 819
    },
    {
      "epoch": 0.025271144697189885,
      "grad_norm": 1.728499412536621,
      "learning_rate": 5e-05,
      "loss": 1.5555,
      "step": 826
    },
    {
      "epoch": 0.02548530694038641,
      "grad_norm": 2.449018716812134,
      "learning_rate": 4.918554903670771e-05,
      "loss": 1.6124,
      "step": 833
    },
    {
      "epoch": 0.025699469183582933,
      "grad_norm": 1.1849424839019775,
      "learning_rate": 4.837131418777595e-05,
      "loss": 1.3471,
      "step": 840
    },
    {
      "epoch": 0.02591363142677946,
      "grad_norm": 1.4348622560501099,
      "learning_rate": 4.755751151021934e-05,
      "loss": 1.6534,
      "step": 847
    },
    {
      "epoch": 0.026127793669975984,
      "grad_norm": 1.300132393836975,
      "learning_rate": 4.674435694637597e-05,
      "loss": 1.5182,
      "step": 854
    },
    {
      "epoch": 0.026341955913172507,
      "grad_norm": 1.4465512037277222,
      "learning_rate": 4.593206626660709e-05,
      "loss": 1.604,
      "step": 861
    },
    {
      "epoch": 0.026556118156369032,
      "grad_norm": 1.1964848041534424,
      "learning_rate": 4.512085501204253e-05,
      "loss": 1.3865,
      "step": 868
    },
    {
      "epoch": 0.026770280399565558,
      "grad_norm": 1.3393502235412598,
      "learning_rate": 4.431093843738692e-05,
      "loss": 1.5336,
      "step": 875
    },
    {
      "epoch": 0.02698444264276208,
      "grad_norm": 1.347054362297058,
      "learning_rate": 4.350253145380189e-05,
      "loss": 1.5255,
      "step": 882
    },
    {
      "epoch": 0.027198604885958606,
      "grad_norm": 2.0334134101867676,
      "learning_rate": 4.269584857187943e-05,
      "loss": 1.4529,
      "step": 889
    },
    {
      "epoch": 0.02741276712915513,
      "grad_norm": 1.097373127937317,
      "learning_rate": 4.1891103844721636e-05,
      "loss": 1.3301,
      "step": 896
    },
    {
      "epoch": 0.027626929372351654,
      "grad_norm": 1.9113101959228516,
      "learning_rate": 4.108851081114169e-05,
      "loss": 1.4297,
      "step": 903
    },
    {
      "epoch": 0.02784109161554818,
      "grad_norm": 1.67792809009552,
      "learning_rate": 4.028828243900141e-05,
      "loss": 1.4526,
      "step": 910
    },
    {
      "epoch": 0.028055253858744703,
      "grad_norm": 1.7338939905166626,
      "learning_rate": 3.949063106870031e-05,
      "loss": 1.483,
      "step": 917
    },
    {
      "epoch": 0.02826941610194123,
      "grad_norm": 1.3107962608337402,
      "learning_rate": 3.869576835683109e-05,
      "loss": 1.4511,
      "step": 924
    },
    {
      "epoch": 0.02848357834513775,
      "grad_norm": 1.466441035270691,
      "learning_rate": 3.790390522001662e-05,
      "loss": 1.5808,
      "step": 931
    },
    {
      "epoch": 0.028697740588334276,
      "grad_norm": 1.8877427577972412,
      "learning_rate": 3.711525177894331e-05,
      "loss": 1.5342,
      "step": 938
    },
    {
      "epoch": 0.028911902831530802,
      "grad_norm": 1.3630350828170776,
      "learning_rate": 3.6330017302605576e-05,
      "loss": 1.5243,
      "step": 945
    },
    {
      "epoch": 0.029126065074727325,
      "grad_norm": 1.4589847326278687,
      "learning_rate": 3.554841015277641e-05,
      "loss": 1.6611,
      "step": 952
    },
    {
      "epoch": 0.02934022731792385,
      "grad_norm": 1.4928864240646362,
      "learning_rate": 3.477063772871861e-05,
      "loss": 1.3335,
      "step": 959
    },
    {
      "epoch": 0.029554389561120376,
      "grad_norm": 1.400469183921814,
      "learning_rate": 3.399690641215142e-05,
      "loss": 1.3706,
      "step": 966
    },
    {
      "epoch": 0.0297685518043169,
      "grad_norm": 1.7362520694732666,
      "learning_rate": 3.322742151248725e-05,
      "loss": 1.5866,
      "step": 973
    },
    {
      "epoch": 0.029982714047513424,
      "grad_norm": 2.1456429958343506,
      "learning_rate": 3.246238721235283e-05,
      "loss": 1.4713,
      "step": 980
    },
    {
      "epoch": 0.030196876290709947,
      "grad_norm": 1.4593017101287842,
      "learning_rate": 3.1702006513409396e-05,
      "loss": 1.5291,
      "step": 987
    },
    {
      "epoch": 0.030411038533906472,
      "grad_norm": 1.3535205125808716,
      "learning_rate": 3.09464811824863e-05,
      "loss": 1.4651,
      "step": 994
    },
    {
      "epoch": 0.030625200777102998,
      "grad_norm": 1.5841355323791504,
      "learning_rate": 3.019601169804216e-05,
      "loss": 1.5082,
      "step": 1001
    },
    {
      "epoch": 0.03083936302029952,
      "grad_norm": 1.669594168663025,
      "learning_rate": 2.9450797196968023e-05,
      "loss": 1.4725,
      "step": 1008
    },
    {
      "epoch": 0.031053525263496046,
      "grad_norm": 0.9601817727088928,
      "learning_rate": 2.8711035421746367e-05,
      "loss": 1.4469,
      "step": 1015
    },
    {
      "epoch": 0.03126768750669257,
      "grad_norm": 1.3222395181655884,
      "learning_rate": 2.7976922667980272e-05,
      "loss": 1.1912,
      "step": 1022
    },
    {
      "epoch": 0.031481849749889095,
      "grad_norm": 1.4175786972045898,
      "learning_rate": 2.7248653732306316e-05,
      "loss": 1.5412,
      "step": 1029
    },
    {
      "epoch": 0.03169601199308562,
      "grad_norm": 1.1599923372268677,
      "learning_rate": 2.6526421860705473e-05,
      "loss": 1.6639,
      "step": 1036
    },
    {
      "epoch": 0.031910174236282146,
      "grad_norm": 1.7723807096481323,
      "learning_rate": 2.581041869722519e-05,
      "loss": 1.5578,
      "step": 1043
    },
    {
      "epoch": 0.03212433647947867,
      "grad_norm": 1.012082815170288,
      "learning_rate": 2.5100834233126823e-05,
      "loss": 1.4582,
      "step": 1050
    },
    {
      "epoch": 0.03233849872267519,
      "grad_norm": 1.5531741380691528,
      "learning_rate": 2.4397856756471432e-05,
      "loss": 1.5486,
      "step": 1057
    },
    {
      "epoch": 0.03255266096587172,
      "grad_norm": 1.1736657619476318,
      "learning_rate": 2.3701672802157566e-05,
      "loss": 1.561,
      "step": 1064
    },
    {
      "epoch": 0.03276682320906824,
      "grad_norm": 1.1771007776260376,
      "learning_rate": 2.3012467102424373e-05,
      "loss": 1.442,
      "step": 1071
    },
    {
      "epoch": 0.032980985452264765,
      "grad_norm": 1.25017511844635,
      "learning_rate": 2.23304225378328e-05,
      "loss": 1.4735,
      "step": 1078
    },
    {
      "epoch": 0.03319514769546129,
      "grad_norm": 1.349295735359192,
      "learning_rate": 2.1655720088738453e-05,
      "loss": 1.4819,
      "step": 1085
    },
    {
      "epoch": 0.033409309938657816,
      "grad_norm": 1.1505281925201416,
      "learning_rate": 2.0988538787268374e-05,
      "loss": 1.3994,
      "step": 1092
    },
    {
      "epoch": 0.03362347218185434,
      "grad_norm": 1.6927034854888916,
      "learning_rate": 2.0329055669814934e-05,
      "loss": 1.5193,
      "step": 1099
    },
    {
      "epoch": 0.03383763442505086,
      "grad_norm": 1.2424277067184448,
      "learning_rate": 1.9677445730059346e-05,
      "loss": 1.3923,
      "step": 1106
    },
    {
      "epoch": 0.03405179666824739,
      "grad_norm": 1.8569272756576538,
      "learning_rate": 1.9033881872537006e-05,
      "loss": 1.4679,
      "step": 1113
    },
    {
      "epoch": 0.03426595891144391,
      "grad_norm": 1.278309941291809,
      "learning_rate": 1.8398534866757454e-05,
      "loss": 1.3875,
      "step": 1120
    },
    {
      "epoch": 0.034480121154640435,
      "grad_norm": 1.5193406343460083,
      "learning_rate": 1.7771573301890664e-05,
      "loss": 1.6248,
      "step": 1127
    },
    {
      "epoch": 0.034694283397836964,
      "grad_norm": 1.487230658531189,
      "learning_rate": 1.715316354203188e-05,
      "loss": 1.5312,
      "step": 1134
    },
    {
      "epoch": 0.034908445641033486,
      "grad_norm": 1.7162104845046997,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 1.6004,
      "step": 1141
    },
    {
      "epoch": 0.03512260788423001,
      "grad_norm": 1.3494603633880615,
      "learning_rate": 1.594265350408039e-05,
      "loss": 1.447,
      "step": 1148
    },
    {
      "epoch": 0.03533677012742654,
      "grad_norm": 1.5886056423187256,
      "learning_rate": 1.5350874434525142e-05,
      "loss": 1.444,
      "step": 1155
    },
    {
      "epoch": 0.03555093237062306,
      "grad_norm": 1.3253085613250732,
      "learning_rate": 1.4768289501820265e-05,
      "loss": 1.3363,
      "step": 1162
    },
    {
      "epoch": 0.03576509461381958,
      "grad_norm": 1.9405076503753662,
      "learning_rate": 1.4195053294732758e-05,
      "loss": 1.362,
      "step": 1169
    },
    {
      "epoch": 0.035979256857016105,
      "grad_norm": 1.1877350807189941,
      "learning_rate": 1.3631317921347563e-05,
      "loss": 1.5041,
      "step": 1176
    },
    {
      "epoch": 0.036193419100212634,
      "grad_norm": 1.3068554401397705,
      "learning_rate": 1.3077232968705805e-05,
      "loss": 1.4786,
      "step": 1183
    },
    {
      "epoch": 0.03640758134340916,
      "grad_norm": 1.0625447034835815,
      "learning_rate": 1.2532945463111855e-05,
      "loss": 1.2748,
      "step": 1190
    },
    {
      "epoch": 0.03662174358660568,
      "grad_norm": 1.4709055423736572,
      "learning_rate": 1.1998599831119912e-05,
      "loss": 1.491,
      "step": 1197
    },
    {
      "epoch": 0.03683590582980221,
      "grad_norm": 1.5545629262924194,
      "learning_rate": 1.1474337861210543e-05,
      "loss": 1.4981,
      "step": 1204
    },
    {
      "epoch": 0.03705006807299873,
      "grad_norm": 2.0076568126678467,
      "learning_rate": 1.096029866616704e-05,
      "loss": 1.5171,
      "step": 1211
    },
    {
      "epoch": 0.03726423031619525,
      "grad_norm": 1.0814955234527588,
      "learning_rate": 1.0456618646161954e-05,
      "loss": 1.3583,
      "step": 1218
    },
    {
      "epoch": 0.03747839255939178,
      "grad_norm": 1.3351236581802368,
      "learning_rate": 9.963431452563332e-06,
      "loss": 1.4998,
      "step": 1225
    },
    {
      "epoch": 0.037692554802588304,
      "grad_norm": 1.345676064491272,
      "learning_rate": 9.480867952470284e-06,
      "loss": 1.4631,
      "step": 1232
    },
    {
      "epoch": 0.03790671704578483,
      "grad_norm": 1.1110200881958008,
      "learning_rate": 9.00905619398757e-06,
      "loss": 1.4986,
      "step": 1239
    },
    {
      "epoch": 0.03812087928898135,
      "grad_norm": 1.1065698862075806,
      "learning_rate": 8.548121372247918e-06,
      "loss": 1.4631,
      "step": 1246
    },
    {
      "epoch": 0.03833504153217788,
      "grad_norm": 1.2057925462722778,
      "learning_rate": 8.098185796191631e-06,
      "loss": 1.439,
      "step": 1253
    },
    {
      "epoch": 0.0385492037753744,
      "grad_norm": 1.4041844606399536,
      "learning_rate": 7.659368856111926e-06,
      "loss": 1.4684,
      "step": 1260
    },
    {
      "epoch": 0.03876336601857092,
      "grad_norm": 1.3971853256225586,
      "learning_rate": 7.2317869919746705e-06,
      "loss": 1.439,
      "step": 1267
    },
    {
      "epoch": 0.03897752826176745,
      "grad_norm": 1.2180246114730835,
      "learning_rate": 6.815553662521185e-06,
      "loss": 1.5764,
      "step": 1274
    },
    {
      "epoch": 0.039191690504963975,
      "grad_norm": 1.822922706604004,
      "learning_rate": 6.410779315161886e-06,
      "loss": 1.4794,
      "step": 1281
    },
    {
      "epoch": 0.0394058527481605,
      "grad_norm": 1.2133909463882446,
      "learning_rate": 6.017571356669183e-06,
      "loss": 1.4784,
      "step": 1288
    },
    {
      "epoch": 0.039620014991357026,
      "grad_norm": 1.4232499599456787,
      "learning_rate": 5.636034124677042e-06,
      "loss": 1.405,
      "step": 1295
    },
    {
      "epoch": 0.03983417723455355,
      "grad_norm": 1.8573203086853027,
      "learning_rate": 5.266268859995083e-06,
      "loss": 1.4153,
      "step": 1302
    },
    {
      "epoch": 0.04004833947775007,
      "grad_norm": 1.6481467485427856,
      "learning_rate": 4.908373679744316e-06,
      "loss": 1.3081,
      "step": 1309
    },
    {
      "epoch": 0.0402625017209466,
      "grad_norm": 2.072063446044922,
      "learning_rate": 4.562443551321788e-06,
      "loss": 1.2842,
      "step": 1316
    },
    {
      "epoch": 0.04047666396414312,
      "grad_norm": 1.2372018098831177,
      "learning_rate": 4.228570267201049e-06,
      "loss": 1.5924,
      "step": 1323
    },
    {
      "epoch": 0.040690826207339645,
      "grad_norm": 1.206811785697937,
      "learning_rate": 3.90684242057498e-06,
      "loss": 1.4247,
      "step": 1330
    },
    {
      "epoch": 0.04090498845053617,
      "grad_norm": 1.6587467193603516,
      "learning_rate": 3.5973453818476556e-06,
      "loss": 1.4644,
      "step": 1337
    },
    {
      "epoch": 0.041119150693732696,
      "grad_norm": 1.1888831853866577,
      "learning_rate": 3.3001612759813393e-06,
      "loss": 1.5327,
      "step": 1344
    },
    {
      "epoch": 0.04133331293692922,
      "grad_norm": 1.0991014242172241,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 1.4681,
      "step": 1351
    },
    {
      "epoch": 0.04154747518012574,
      "grad_norm": 1.5934925079345703,
      "learning_rate": 2.743044005587425e-06,
      "loss": 1.3098,
      "step": 1358
    },
    {
      "epoch": 0.04176163742332227,
      "grad_norm": 1.4214099645614624,
      "learning_rate": 2.4832586719889416e-06,
      "loss": 1.4065,
      "step": 1365
    },
    {
      "epoch": 0.04197579966651879,
      "grad_norm": 1.2822719812393188,
      "learning_rate": 2.2360818938828187e-06,
      "loss": 1.4728,
      "step": 1372
    },
    {
      "epoch": 0.042189961909715315,
      "grad_norm": 1.119520664215088,
      "learning_rate": 2.0015792595656226e-06,
      "loss": 1.4284,
      "step": 1379
    },
    {
      "epoch": 0.042404124152911844,
      "grad_norm": 1.3280854225158691,
      "learning_rate": 1.7798129942530551e-06,
      "loss": 1.4559,
      "step": 1386
    },
    {
      "epoch": 0.04261828639610837,
      "grad_norm": 1.1883761882781982,
      "learning_rate": 1.5708419435684462e-06,
      "loss": 1.3757,
      "step": 1393
    },
    {
      "epoch": 0.04283244863930489,
      "grad_norm": 1.3203167915344238,
      "learning_rate": 1.374721557928116e-06,
      "loss": 1.6273,
      "step": 1400
    },
    {
      "epoch": 0.04304661088250142,
      "grad_norm": 1.2010129690170288,
      "learning_rate": 1.191503877827621e-06,
      "loss": 1.4944,
      "step": 1407
    },
    {
      "epoch": 0.04326077312569794,
      "grad_norm": 1.6128370761871338,
      "learning_rate": 1.0212375200327973e-06,
      "loss": 1.6092,
      "step": 1414
    },
    {
      "epoch": 0.04347493536889446,
      "grad_norm": 1.0786170959472656,
      "learning_rate": 8.639676646793382e-07,
      "loss": 1.4227,
      "step": 1421
    },
    {
      "epoch": 0.043689097612090985,
      "grad_norm": 1.0348191261291504,
      "learning_rate": 7.197360432842359e-07,
      "loss": 1.5162,
      "step": 1428
    },
    {
      "epoch": 0.043903259855287514,
      "grad_norm": 1.0884389877319336,
      "learning_rate": 5.885809276723608e-07,
      "loss": 1.5015,
      "step": 1435
    },
    {
      "epoch": 0.04411742209848404,
      "grad_norm": 1.5776199102401733,
      "learning_rate": 4.705371198210129e-07,
      "loss": 1.5789,
      "step": 1442
    },
    {
      "epoch": 0.04433158434168056,
      "grad_norm": 1.4270485639572144,
      "learning_rate": 3.65635942625242e-07,
      "loss": 1.5011,
      "step": 1449
    },
    {
      "epoch": 0.04454574658487709,
      "grad_norm": 1.2836881875991821,
      "learning_rate": 2.7390523158633554e-07,
      "loss": 1.4921,
      "step": 1456
    },
    {
      "epoch": 0.04475990882807361,
      "grad_norm": 1.3851501941680908,
      "learning_rate": 1.953693274256374e-07,
      "loss": 1.5536,
      "step": 1463
    },
    {
      "epoch": 0.04497407107127013,
      "grad_norm": 1.9082521200180054,
      "learning_rate": 1.3004906962578721e-07,
      "loss": 1.3908,
      "step": 1470
    },
    {
      "epoch": 0.04518823331446666,
      "grad_norm": 1.3173164129257202,
      "learning_rate": 7.796179090094891e-08,
      "loss": 1.4139,
      "step": 1477
    },
    {
      "epoch": 0.045402395557663185,
      "grad_norm": 1.4804590940475464,
      "learning_rate": 3.9121312597573125e-08,
      "loss": 1.4134,
      "step": 1484
    },
    {
      "epoch": 0.04561655780085971,
      "grad_norm": 1.5944395065307617,
      "learning_rate": 1.3537941026914303e-08,
      "loss": 1.301,
      "step": 1491
    },
    {
      "epoch": 0.045830720044056236,
      "grad_norm": 1.4432575702667236,
      "learning_rate": 1.2184647302626583e-09,
      "loss": 1.5252,
      "step": 1498
    }
  ],
  "logging_steps": 7,
  "max_steps": 1500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.119015678246912e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}